The vaccine industry 3 43
however, with the growth of new markets in emerging econo-
mies and with the pressing need for new vaccines for the devel-
oping world. The current efforts of PDPs and public creation of
markets in response to this need will be successful if lessons
learned from the industrial vaccine effort are incorporated into
these government and philanthropically driven experiments.
Access the complete reference list online at http://www.expertconsult.com
1. Warren KS. New scientific opportunities and old obstacles in vaccine
development. Proc Natl Acad Sci U S A 1986;83:9275–7.
5. Halsted SB, Gellin BG. Immunizing children: can one shot do it all? In:
Medical and Health Annual 1994. Chicago, IL: Encyclopedia Britannica;
1994.
6. Cohen J. Public health: U.S. vaccine Supply falls seriously short. Science
2002;295:1998–2001.
7. Peter G, des Vignes-Kendrick M, Eickhoff TC, et al. Lessons learned from a
review of the development of selected vaccines. National Vaccine Advisory
Committee. Pediatrics 1999;104(4pt 1):942–50.
8. Marcuse EK, Braiman J, Douglas RG, et al., For the National Vaccine
Advisory Committee United States vaccine research: a delicate fabric of
political and private collaboration. Pediatrics 1997;100:1015–20.
Acknowledgments
Special thanks to Centerview Partners for updated industry data
reflected in Tables 3-1, 3-5, and 3-6, and in Figure 3-2, and to
Alan Engbring for compositional support.
10. Gregerson J. Vaccine development: the long road from initial idea to product
licensure. In: Levin MM, Woodrow GC, Kaspe JB, et al, editors. New
Generation Vaccines. New York: Marcel Dekker; 1987. p. 1165–83.
11. DiMasi J, Hansen R, Grabowski H. Cost of new drug development. J Health
Econ 2003;22:151.
16. Pasternak A, Sabow A, Chadwick-Jones A. Structural shift: promising yet
challenging new markets for vaccines. Mercer Management Consulting;
2006.
17. Berndt ER, Hurvitz JA. Vaccine advance-purchase agreements for low income
countries: practical issues. Health Aff 2005;24:653–65.
18. Pauley MV. Improving vaccine supply and development: who needs what?
Health Aff 2005;24:680–9.
 SECTION ONE: General aspects of vaccination
Vaccine manufacturing
4 Phillip L. Gomez
James M. Robinson
Joseph A. Rogalewicz
The vast majority of the more than 1 billion doses of vaccines
manufactured worldwide each year are given to perfectly
healthy people.1–4 It is this fact that drives the requirements
for vaccines to be among the most rigorously designed, moni-
tored, and compliant products manufactured today. The ability
to manufacture these vaccines safely and consistently is built
on four competencies:
1. the manufacturing process that defines how the product
is made;
2. the compliance of the organization to successfully complete
that process;
3. the testing of the product and supporting operations; and
4. the regulatory authorization to release and distribute the
product.
This chapter examines how each of these components is estab-
lished during the development of a new vaccine and how the
field of vaccine manufacturing is responding to emerging chal-
lenges for increased capacity (eg, pandemic influenza), increased
safety assurance (eg, barrier isolator filling), and increasing
complexities of manufacture (eg, conjugate vaccines). All of
this must be accomplished while consistently delivering more
than 1 billion doses annually at the relatively low cost of similar
therapeutic products.
In the United States, vaccines are regulated as biologic prod-
ucts. The Food and Drug Administration's (FDA) Center for
Biologics Evaluation and Research (CBER) is responsible for reg-
ulating vaccines in the United States. Current authority for the
regulation of vaccines resides primarily in Section 351 of the
Public Health Service Act and specific sections of the Federal
Food, Drug and Cosmetic Act.5,6 Section 351 of the Public
Health Service Act gives the federal government the author-
ity to license biologic products and the establishments where
they are produced.7 Vaccines undergo a rigorous review of labo-
ratory, nonclinical, and clinical data to ensure safety, efficacy,
purity, and potency. Vaccines approved for marketing may also
be required to undergo additional studies to further evaluate the
vaccine and often to address specific questions about the vac-
cine's safety, effectiveness, or possible side effects.8
In the European Union, animal and human vaccines are reg-
ulated by the European Medicines Agency (EMA), whose main
responsibility is the promotion of public and animal health.
The EMA's Committee on Medicinal Products for Human Use
through its Vaccine Working Party has oversight for human vac-
cines. Vaccines are licensed through a centralized procedure that
allows for simultaneous licensure within all countries within
the European Union. Human vaccines manufacturing is regu-
lated under a Good Manufacturing Practices (GMP) Directive
200/94/EEC, Annex 16, and Annex 2.
Harmonization of licensing and regulating procedures for vac-
cines worldwide has obvious benefits in rapidly delivering safe and
effective vaccines to the market. Impediments to harmonization
include lack of standardized regulatory procedures and mutual rec-
ognition of licenses and inspections between countries and world-
wide regulatory agencies. Harmonization of regulation continues
to progress as joint FDA-EMA establishment inspections programs
have become reality and adherence to harmonized International
Conference on Harmonisation (ICH) guidance expected.
New vaccines are subjected to a well-defined regulatory pro-
cess for approval. The approval process consists of four princi-
pal elements:
– Preparation of preclinical materials for proof-of-concept
testing in animal models, manufacture of clinical
materials according to current GMP (cGMP), and
toxicology analysis in an appropriate animal system
– Submission of an investigational new drug application
(IND) for submission to FDA for review
– Testing for safety and effectiveness through clinical and
further nonclinical studies (phase 1 to 3 clinical studies).
– Submission of all clinical, nonclinical, and
manufacturing data to the FDA and EMA in the form
of a Biologics License Application (BLA) for final review
and licensure.
This chapter outlines the basics of manufacturing a vaccine and
a description of some examples of currently licensed products.
It then moves to the regulatory requirements for vaccine manu-
facturing including cGMP compliance and then discusses the
development of new vaccines. The final section examines the
great challenges in the field to deliver a product held to an ever-
increasing standard of safety while providing sufficient doses
at reasonable costs for an ever-increasing number of diseases.
Manufacturing basics
The manufacture of vaccines is composed of several basic steps
that result in the finished product. A summary of these steps
with examples for pathogens that have a licensed vaccine is
given in Table 4-1. The first step is the generation of the anti-
gen used to induce an immune response. This step includes the
generation of the pathogen itself (for subsequent inactivation
or isolation of a subunit) or generation of a recombinant pro-
tein derived from the pathogen. Vaccines under development
use additional methods that will be discussed later. Viruses are
grown on cells, primary cells such as chicken fibroblasts (yellow
fever), or they are grown on continuous cell lines such as MRC-5
Examples of Licensed Vaccine Manufacturing Processes

Vaccine manufacturing
4
45
 46
Examples of Licensed Vaccine Manufacturing Processes—cont'd

SECTION ONE General aspects of vaccination

ND, not disclosed.
Source: package inserts.
48 SECTION ONE General aspects of vaccination
strains by the Therapeutic Goods Administration of Australia.
These viral strains are used to prepare the inoculums for vac-
cine production.
The substrate most commonly used by producers of influ-
enza vaccine is the 11-day-old embryonated chicken egg.
A monovalent virus (suspension) is received from the CBER
or the CDC. The monovalent virus suspension is passed in
eggs. The inoculated eggs are incubated for a specific time
and temperature regimen under controlled relative humidity
and then harvested. The harvested allantoic fluids, which
contain the live virus, are tested for infectivity, titer, speci-
ficity, and sterility. These fluids are then stored wet frozen
at extremely low temperatures to maintain the stability of
the monovalent seed virus (MSV).15 This MSV is also certi-
fied by the CBER.
Once the MSV is introduced to the egg by automated inocula-
tors, the virus is grown at incubated temperatures, and then the
allantoic fluid is harvested and purified by high-speed centrifuga-
tion on a sucrose gradient or by chromatography. The purified
virus is often split using a detergent before final filtration. The
virus is inactivated using formaldehyde before or after the primary
purification step, depending on manufacturer. This is repeated for
three strains of virus, and the individually tested and released
inactivated viral concentrates are combined and diluted to final
vaccine strength. The overall process is outlined in Figure 4-2.
Recombinant protein (hepatitis B)
In July 1986, a recombinant hepatitis B vaccine was licensed
in the United States. This vaccine built on the knowledge that
heat-inactivated serum containing hepatitis B virus (HBV) and
hepatitis B surface antigen (HBsAg) was not infectious, but was
immunogenic and partially protective against subsequent expo-
sure to HBV.16 It was determined that the HBsAg was the com-
ponent that conferred protection to HBV on immunization.17 To
produce this vaccine, the HBsAg or "S" gene was inserted into
an expression vector that was capable of directing the synthesis
of large quantities of HbsAg in . The
HbsAg particles expressed by and purified from the yeast cells
have been demonstrated to be equivalent to the HBsAg derived
from the plasma of the blood of hepatitis B chronic carriers.16,18,19
The recombinant cells expressing HBsAg are
grown in stirred tank fermenters. The medium used in this
process is a complex fermentation medium that consists of an
extract of yeast, soy peptone, dextrose, amino acids, and mineral
salts. In-process testing is conducted on the fermentation prod-
uct to determine the percentage of host cells with the expression
construct.7 At the end of the fermentation process, the HBsAg
is harvested by lysing the yeast cells. It is separated by hydro-
phobic interaction and size-exclusion chromatography. The
resulting HBsAg is assembled into 22-nm-diameter lipoprotein
particles. The HBsAg is purified to greater than 99% for pro-
tein by a series of physical and chemical methods. The purified
protein is treated in phosphate buffer with formaldehyde, sterile
filtered, and then coprecipitated with alum (potassium alumi-
num sulfate) to form bulk vaccine adjuvanted with amorphous
aluminum hydroxyphosphate sulfate. The vaccine contains no
detectable yeast DNA but may contain not more than 1% yeast
protein.7,18,20 In a second recombinant hepatitis B vaccine, the
surface antigen expressed in cells is purified by sev-
eral physiochemical steps and formulated as a suspension of the
antigen absorbed on aluminum hydroxide. The procedures used
in its manufacturing result in a product that contains no more
than 5% yeast protein. No substances of human origin are used
in its manufacture.19 Vaccines against hepatitis B prepared from
recombinant yeast cultures are noninfectious19 and are free of
association with human blood and blood products.18
Each lot of hepatitis B vaccine is tested for safety, in mice
and guinea pigs, and for sterility.18 QC product testing for purity
and identity includes numerous chemical, biochemical, and
physical assays on the final product to assure thorough char-
acterization and lot-to-lot consistency. Quantitative immuno-
assays using monoclonal antibodies can be used to measure
the presence of high levels of key epitopes on the yeast-derived
HbsAg. A mouse potency assay is also used to measure the
immunogenicity of hepatitis B vaccines. The effective dose
capable of seroconverting 50% of the mice (ED50) is calculated.20
Hepatitis B vaccines are sterile suspensions for intramuscu-
lar injection. The vaccine is supplied in four formulations: pedi-
atric, adolescent/high-risk infant, adult, and dialysis.
All formulations contain approximately 0.5 mg of aluminum
(provided as amorphous aluminum hydroxyphosphate sulfate)
per milliliter of vaccine.18
The QC testing requirements for the release of recombinant
hepatitis B vaccine are summarized in Table 4-2.
Most vaccines are still released by the CBER on a lot-by-lot
basis, but for several extensively characterized vaccines, this
requirement has been eliminated. They include hepatitis B vac-
cines and human papillomavirus (HPV) vaccines, which are
manufactured using recombinant DNA processes. Their manu-
facturing process includes significant purification, and they are
extensively characterized by their analytical methods. In addition,
hepatitis B vaccine had to demonstrate a "track record" of contin-
ued safety, purity, and potency to qualify for this exemption.7,21
Conjugate vaccine (Haemophilus influenzae type b)
The production of type b conjugate includes
the separate production of capsular polysaccharide from
type b and a carrier protein such as
tetanus protein from (ie, purified tetanus
toxoid), CRM protein from or
outer membrane protein complex of .
The capsular polysaccharide is produced in industrial bio-
reactors using approved seeds of type b. A crude
intermediate is recovered from fermentation supernatant, using
a cationic detergent. The resulting material is harvested by
continuous-flow centrifugation. The paste is then resuspended
in buffer, and the polysaccharide is selectively dissociated from
disrupted paste by increasing the ionic strength.
The polysaccharide is then further purified by phenol extrac-
tion, ultrafiltration, and ethanol precipitation. The final mate-
rial is precipitated with alcohol, dried under vacuum, and stored
at 35°C for further processing.
Tetanus protein is prepared in bioreactors using approved
seeds of . The crude toxin is recovered from the cul-
ture supernatant by continuous-flow centrifugation and diafil-
tration. Crude toxin is then purified by a combination of
fractional ammonium sulfate precipitation and ultrafiltration.
Testing Requirements for the Release of Recombinant
Hepatitis B Vaccine
 Vaccine manufacturing 4 49

The egg-based influenza vaccine manufacturing process flow. CBER, Center for Biologics Evaluation and Research (of the US Food
and Drug Administration); QA, quality assurance; QC, quality control.

The resulting purified toxin is detoxified using formaldehyde,

concentrated by ultrafiltration, and stored at more than 2°C up
to 8°C for further processing.
The industrial conjugation process was initially developed using
tetanus toxoid by the J.B. Robbins team at the National Institute
of Allergy and Infectious Diseases (NIAID), Bethesda, Maryland.22
Conjugate preparation is a two-step process that involves:
– activation of the Hib capsular polysaccharide and
– conjugation of activated polysaccharide to tetanus
protein through a spacer.
Activation includes chemical fragmentation of the native polysac-
charide to a specified molecular weight target and covalent linkage
of adipic acid dihydrazide. The activated polysaccharide is then
covalently linked to the purified tetanus protein by carbodiimide-
mediated condensation using 1-ethyl-3(3-dimethylaminopropyl)
carbodiimide. Purification of the conjugated material is per-
formed to obtain high-molecular-weight conjugate molecules
devoid of chemical residues and free protein and polysaccharide.
Conjugate bulk is then diluted in an appropriate buffer, filled
into unit-dose and/or multidose vials, and lyophilized.
 SECTION ONE
Live attenuated vaccine (measles)
The measles virus, isolated in 1954, is part of the genus
Morbillivirus in the family Paramyxoviridae. Current vaccines
are derived from Edmonston, Moraten, or Schwarz strains.
Such vaccines have been on the market since the 1960s and
in combination (MMR) since the 1970s. The final vaccine is a
live attenuated viral vaccine inducing immunity in more than
90% of recipients.
The manufacture of measles starts with specific pathogen-
free (SPF) embryonated chicken eggs that are incubated several
days. The embryos are collected and treated with trypsin to pre-
pare the chick embryo fibroblast for cell culture. All of the oper-
ations are done under strict aseptic conditions, performed by
well-trained operators.
The cell culture is grown in roller bottles using fetal calf sera
and M199 Hanks media for optimal cell growth. The chick
embryo fibroblast cells are further infected by the viral working
seed and incubated several days for viral culture. At the end of
the viral culture, the cells are disrupted by mechanical lysis to
release the virus. The virus is purified by centrifugation and fil-
tration and stored frozen.
After release of all QC tests, the vaccine is formulated
alone or with mumps and rubella vaccines and lyophilized to
obtain the stable product. The vaccine is reconstituted just
before use.
Virus-like particle–based vaccines
Traditional viral vaccines rely on attenuated virus strains or
inactivation of infectious virus. Subunit vaccines based on viral
proteins expressed in heterologous systems have been effective
for some pathogens but have often had poor immunogenicity
due to incorrect folding or modification.23 Virus-like particles
(VLPs) are designed to mimic the overall structure of virus par-
ticles and, thus, preserve the native antigenic conformation
of the immunogenic proteins. VLPs have been produced for a
wide range of taxonomically and structurally distinct viruses
and have unique advantages in terms of safety and
immunogenicity over previous approaches.1 Attenuation or
inactivation of the VLP is not required; this is particularly
important as epitopes are commonly modified by inactivation
treatments.24 However, if a viral vector (eg, baculovirus) is used
as the expression system, inactivation may be required if the
purification process cannot eliminate residual viral activity.
For a VLP to be a realistic vaccine candidate, it needs to
be produced in a safe expression system that is easy to scale
up to large-scale production1 and by an accompanying puri-
fication (and inactivation) process that will maintain native
structure and immunogenicity and that will meet the require-
ments of today's global regulatory authorities. A number of
expression systems have been demonstrated to manufacture
multimeric VLPs, including the baculovirus expression system
(BVES) in Sf9 and High Five cells,
, Chinese hamster ovary, human function liver cells 4,
baby hamster kidney, transgenic plants (potato, tobacco, soy-
bean), , , human embryonic kid-
ney 293 (HEK293), and lupin callus with yields ranging from
0.3 to 10 g/mL or as high as 300 to 500 g/mL with
and HEK293 (purified).2 The BVES has proven quite versatile,
demonstrating the capability of preparing vaccine candidates
for papillomavirus, feline calicivirus, hepatitis E virus, porcine
parvovirus, chicken anemia virus, porcine circovirus, SV40,
poliovirus, bluetongue virus, rotavirus, hepatitis C virus,
human immunodeficiency virus (HIV), simian immunodefi-
ciency virus, feline immunodeficiency virus, Newcastle disease
virus, SARS coronavirus, Hantaan virus, influenza A virus, and
infectious bursal disease virus.1
Many pathogenic viruses such as influenza, HIV, and hepa-
titis C are surrounded by an envelope, a membrane that con-
sists of a lipid bilayer derived from the host cell, inserted with
virus glycoprotein spikes. These proteins are targets of neu-
tralizing antibodies and are essential components of vaccine.
Owing to inherent properties of the lipid envelope, assembly
of VLPs in insect cells for these viruses is a different type of
technical challenge to those produced viruses with multiple
capsids.1 For these targets, production of VLPs is a challenging
task because the synthesis and assembly of one or more recom-
binant proteins may be required. This is the case for VLPs of
rotavirus (RLP), which is an RNA virus with capsids formed
by 1860 monomers of four different proteins. In addition, the
production of most VLPs requires the simultaneous expression
and assembly of several recombinant proteins, which, for the
case of RLP, needs to occur in a single host cell.25 Purification of
VLPs also constitutes a particularly challenging task. VLPs are
structures of several nanometers in diameter and of molecular
weights in the range of 106 Da. Also, for guaranteeing the qual-
ity of the product, it is not sufficient to demonstrate the absence
of contaminant proteins; it is also necessary to show that pro-
teins are correctly assembled into VLPs.
The HPV type 16 major 55-kDa capsids protein, L1, when
produced in certain recombinant expression systems such as
can form irregularly shaped VLPs with a broad size
distribution. These HPV VLPs are inherently unstable and tend
to aggregate in solution. The primary challenge of HPV vaccine
formulation development was the preparation of aqueous HPV
VLP solutions that are stable under a variety of purification,
processing, and storage conditions. By treating the HPV VLPs
through a process of disassembly and reassembly, the stability
and in vitro potency of the vaccine are enhanced significantly.
In addition, the in vivo immunogenicity of the vaccine was also
improved by as much as approximately 10-fold, as shown in
mouse potency studies.26 The disassembly and reassembly of
particles may also be important to remove residual proteins from
the expression system/host cell used in the production and is a
serious processing challenge, particularly for enveloped VLPs.
Product development
Vaccine development involves the process of taking a new anti-
gen or immunogen identified in the research process and devel-
oping this substance into a final vaccine that can be evaluated
through preclinical and clinical studies to determine the safety
and efficacy of the resultant vaccine. During this process, the
product's components, in-process materials, final product
specifications, and manufacturing process are defined. The
manufacturing scale used during development is usually sig-
nificantly smaller than that used in the final manufacturing
process. Phase 1 and, sometimes, phase 2 clinical trial vaccines
are typically produced in product development, but it is usually
anticipated that at least one of the three or more consistency
lots used for phase 3 clinical trials will be manufactured at full-
scale production volume. The product manufactured during
the development phase is manufactured according to cGMP.27
Current GMP considerations
Historically, US manufacturers were bound to the cGMP as
detailed in Sections 210 through 226 and Section 600 of the US
Code of Federal Regulations (CFR),28,29 which apply specifically
to approved drug and biological products. Federal regulations
set forth detailed cGMP that provide principles and methods
to ensure that the product meets safety requirements and the
manufacturing process will consistently produce a product

that meets the specified identity, strength, quality, and purity
specifications set forth for licensed vaccine products.
During the 1960s and 1970s, domestic and international
drug and biologic manufacturers saw a rapid increase in laws,
regulations, and guidelines for reporting and evaluating the data
on the safety, quality, and efficacy of new products. The indus-
try, at the time, was becoming more international and seeking
new global markets, but the registration of vaccines remained
a national responsibility. Although different regulatory systems
were based on the same fundamental obligations to evaluate
quality, safety, and efficacy, the detailed technical requirements
had diverged to an extent that industry found it necessary to
duplicate many time-consuming and expensive test procedures
to market new products internationally.
The need to harmonize regulation was also fueled by con-
cerns about rising costs of health care, escalation of the cost of
research and development, and a public expectation that there
be a minimum of delay in making safe and efficacious new
treatments available to patients. To this end, the International
Conference on Harmonization (ICH) was born in 1990.30 This
committee includes regulatory and scientific representation
from the United States, Europe, and Asia and has been provid-
ing regulatory and scientific guidance and requirements for the
manufacture of pharmaceuticals and biologics. These guide-
lines and requirements do not take the place of the codified US
cGMP; however, manufacturers applying the ICH guidance will
by default be in compliance with the US cGMP and interna-
tional cGMP requirements. The harmonization of currently
licensed and new vaccine specifications and testing requirements
remains a challenge to industry and regulatory authorities.
The application of cGMP to materials produced for human
clinical trials is not codified or well defined in domestic and inter-
national regulations. Although not specifically noted in the current
CFR, the introduction of the cGMP in 1978 included a preamble
stating that cGMP requirements are applicable to investigational
materials. It is intended that manufacturers apply the cGMP where
applicable and practical, taking into account the intent of the clini-
cal trial phase and manufacturing process development. The appli-
cation of the cGMP is summarized in initial IND and subsequent
supplements that further build on the manufacturer's control and
assurance of the safety and efficacy of the product. The European
Union has issued a directive (EMEA 2003) specific to the manu-
facture of investigational materials, which though lacking specific
detail on application of the cGMP, provides further focus on the
quality of investigation supplies.31 The directive has gone as far as
to require inspection and certification of manufacturers preparing
investigational materials intended for use in European clinical trials.
Clinical materials produced for phase 1 clinical trials are
used to demonstrate candidate product safety in a relatively
small number of healthy human patients (tens of volunteers)
and to verify the ability to manufacture the product duplicating
the theoretical process used to manufacture preclinical materi-
als used in animal toxicology studies. The application of appro-
priate written controls, accurate and consistent data recording,
and controlled equipment in the preparation and testing of even
early-stage candidate vaccine materials is critical to ensuring
the desired outcome and to set the foundation for subsequent
development of the potential candidate vaccine.
The expectations for phase 1 cGMP applications include the
following32:
– Personnel with adequate education, experience, and
training to perform function, including cGMP training.
– Written quality unit responsibilities with appropriate
training to disposition components, procedures, assays,
batches, and investigation of deviations.
– Adequately designed facilities and utilities, including
HVAC design to minimize contamination; appropriate
Vaccine manufacturing 4 51
water source; and quality and basic procedural controls
to prevent contamination and mix-ups.
– Appropriate equipment for intended function that is
properly designed, maintained, calibrated, and operated
per written instructions.
– Appropriate qualification and controls to assure safety of
product (eg, viral clearance, toxin inactivation).
– Written component and raw material controls including
established acceptance criteria, appropriate disposition,
and traceability.
– Written production records and procedures that include
records of components and equipment, changes to
procedures and processes, and microbial control records
where appropriate.
– Aseptic processing performed under adequate conditions
by trained personnel.
– Written test procedures performed under controlled
conditions using scientifically sound analytical
methodology with specified acceptance criteria.
Laboratory instrumentation should be operated per
written procedure, maintained, and calibrated.
– Qualified safety testing (eg, sterility, endotoxin).
– Packaging to protect product from contamination and
controlled labeling operations designed to prevent mix-ups.
– Adequate and documented storage and shipping controls
to ensure the integrity of the product.
Phase 2 clinical trials are intended to demonstrate safety and dose
response in a larger target population than would be expected to
receive the vaccine (hundreds of volunteers). Manufacture of phase
2 clinical trial materials will be used to develop initial consistency
of product manufacture, incorporating modifications and improve-
ments based on the phase 1 production and testing experience.
Identification of key candidate process control points for monitoring
and trending and evaluation of equipment and materials to assure
applicability of GMP conformance is considered at this stage.
Phase 3 clinical trials are intended to demonstrate safety and
efficacy in a statistically significant target population (up to tens
of thousands of volunteers) and to demonstrate the ability to
consistently manufacture materials meeting predescribed qual-
ity attributes. Modifications and improvements based on phase
1 and 2 production and testing experience are incorporated into
the manufacturing process, and specifications for process and
control points are defined. All processes and systems necessary
for the manufacture and testing of late-stage clinical materials
should be validated according to cGMP, essentially mimicking
the requirements applied to approved products.
A risk-based approach to cGMP application consistent with
the phase of development will provide for assurance and evi-
dence of product safety and assurance that the desired target
molecule has been derived from the process. These controls at
minimum should include the following:
– removing potential variability surrounding the
manufacturing process through validation and/
or monitoring potential contributors such as the
environment, utilities, and equipment;
– validating finished product safety testing, including
general safety, sterility, and endotoxin;
– ensuring that process design is capable of removing
undesirable contaminants from process streams through
validation or testing per process step; and
– comprehensive documentation of manufacturing and
testing experience.
The cGMP are detailed in Sections 210 through 226 and
Section 600 of the CFR.28,29
52 SECTION ONE General aspects of vaccination

Identity Test Methods

Analytical testing
The analytical testing of vaccines provides evidence that the
vaccine and any of its intermediates meet the specifications
defined within the BLA. Safety, efficacy, and potency tests
associated with a licensed vaccine are maintained within the
approved filing and published in 21 CFR Part 610. In addition,
the USP has prepared monographs for all approved vaccines to
provide standardized requirements and continues to add mono-
graphs as new vaccines proceed to commercialization.33 Most
bulk vaccines must be tested for safety and efficacy by the man-
ufacturer and CBER before release for final formulation and
packaging. In the mid-1990s, CBER developed the concept of
a which was defined as a chemi-
cal entity whose identity, purity, impurities, potency, and quan-
tity can be determined and controlled through analytical testing
and control of the manufacturing process.34 The advantage of
well-characterized products is that the quantifiable analytical
measurements can relate molecule structure to function. The
application of this definition allowed manufacturers of biologics
to eliminate the need for CBER release of biologics before distri-
bution into the market and also allowed for process changes to
take place after the product had been licensed. Current analyti-
cal and process technologies allow the application of this def-
inition for most recombinant DNA proteins and monoclonal developed based on process capability to control contaminating
antibody products; however, with the exception of one currently microorganisms and inherent process impurities. Most vaccines
licensed vaccine, vaccines in general do not meet the criteria are large molecules that depend on their biochemical makeup
for “well-characterized” owing to their complexity, size, and and physical configuration to provide the desired immunologic
structure and the inability to fully characterize and quantify response. Unlike small molecules, which can be steam steril-
all analytical and biological parameters.35 Recent vaccine devel- ized, most vaccines are prepared aseptically. Sterility testing is
opment has focused on molecules that can meet criteria for used for bulk vaccines and final dosage forms but still provides
well-characterized products. Chemical, microbial, and physical limited assurance of sterility because only a sample of the bulk
assays for vaccines are developed in concurrence with product and of a finished product lot can be tested. Current advances
and process development. Process step outputs are tested using in sterility and bioburden testing include the use of process
a variety of analytical techniques designed to understand and analysis technology and rapid micro testing sensitive down to
characterize the structure of the target molecule and any asso- one bacterial cell per sample. These are currently being vali-
ciated impurities. Many of these tests are used during process dated and implemented to reduce current sterility testing time
validation activities to demonstrate the capability and robust- from 14 days down to hours. Traditionally, endotoxin or pyro-
ness of the process. gen testing has been performed using animal response to assess
Identity testing of biologicals includes a wide variety of pyrogenicity of compounds. Reliance on the endotoxin limulus
tests designed to be specific to the moiety based on unique amebocyte lysate (LAL) test to assess pyrogenicity has become
characteristics of its molecular structure or other properties. standard for testing new vaccines. Current testing has tradition-
Identity testing may include one or many complementary ally been performed according to the 1993 Points to Consider
tests including physicochemical, biological, or immunolog- and ICH guidance regarding adventitious agent testing, includ-
ical assays. Identity tests traditionally used were relatively ing viral testing and mycoplasma testing. Advances in technol-
simple tests, but advances in analytical technologies are ogy also involve PCR to detect the presence of contaminating
providing for better specificity, which, in multivalent and adventitious agents in the process and in final formulation.
combination vaccines, becomes critical to ensuring that all In addition to pyrogen testing, general safety testing (abnor-
components are adequately distinguished. One example of mal toxicity testing) is performed. This testing is required only
an advanced technology is the replacement of multiple colo- for vaccines with product-specific safety tests and is performed
rimetric assays by single NMR spectral analysis for multi- in laboratory animals by injecting mice and guinea pigs and
valent polysaccharide identification.36 Current methods are assessing the animals for distress. There is currently no
listed in Table 4-3. alternative for abnormal toxicity testing. For example, specific
Purity tests are developed to identify potential process-
or product-related impurities that may be traced to the pro-
cess, equipment, or inherent product impurities. These tests Product- and Process-Related Impurity Testing
may include chromatographic methods that allow for quan-
titation of trace amounts of impurities from raw materials
or process-related impurities such as chromatographic resin
leakage. Inherent or copurified protein contaminants can be
detected using electrophoresis techniques that can quantify
trace amounts of proteinaceous contaminants. Detection of
residual DNA traditionally used spectral analysis for quanti-
tation, although advances in polymerase chain reaction (PCR)
assays are allowing even greater sensitivity.37 Examples of cur-
rent methods are listed in Table 4-4.
Microbial tests, including bioburden, sterility, endotoxin,
and adventitious viral and mycoplasma agent testing, are also
 Vaccine manufacturing 4 53

safety testing in animals is used to ensure effective inactiva- of conjugate vaccines introduces greater complexity into the
tion of exotoxins produced by host bacterial cells. Advances supply chain as many components must be manufactured sepa-
in replacing animal testing include use of Vero cell assays, rately, conjugated, purified, and then formulated into a single
which are extremely sensitive to toxins, and assays that mea- vaccine. Current trends in the field are aimed at enhancing the
sure enzymatic activity of toxins. Also, biochemical assays supply chain robustness (cell culture–derived flu vaccine) and
using fluorescent peptides mimicking natural substrate have developing the ability to manufacture new types of vaccine
been developed for tetanus and pertussis toxins38 Live virus (plasmid DNA, viral vectors, peptides, live bacteria, irradiated
vaccines also use safety tests to assure that viruses have not sporozoites).
reverted to wild virulent types. Many animal models are used
to assess viral vaccine safety. Advances in PCR technologies
provide for a sensitive alternative by assessing any mutations
in the genome of the virus. Quantitation tests are performed The currently licensed flu vaccine is made in embryonated
to determine the content or mass of the target active moiety chicken eggs. This process requires extremely large quanti-
in vaccine formulations through physiochemical procedures ties of pathogen-free eggs for the manufacturing process. New
and to detect changes in the molecule over time. Traditional technologies are under development to use continuous cell
tests include colorimetric and spectrophotometric assays lines for the production of influenza virus vaccine. A vari-
that, in general, do not have the precision and sensitivity to ety of cell lines are currently under development and prom-
detect changes in the target molecule. Separation assays, such ise to provide a more robust system for the manufacture of
as high-performance liquid chromatography, capillary elec- bulk influenza vaccine. Cell lines under development include
trophoresis, and sodium dodecyl sulfate–polyacrylamide gel MRC-5, Vero, and PER.C6.39 These processes will eliminate
electrophoresis, can indicate stability and lack of product deg- the risk that an avian influenza virus could infect the flocks
radation. Protein-based vaccines will routinely include a total that produce the eggs for the current vaccine manufacturing.
protein test. Antigenicity provides a quantitative measure of Such an infection would potentially eliminate the supply of
antigen present by measuring antigen-antibody interaction. the vaccine entirely. A number of cell culture–derived influ-
Antigen assays fall into four groups: enza vaccines have been approved in Europe, although not at
a scale capable of providing sufficient vaccine to replace egg-
– radioimmunoassay,
derived vaccine or to respond to a pandemic event. Strong
– enzyme immunoassays, industry, academic, and government focus in recent years
– nonlabeled immunoassays, and promises to advance the development of this technology.
– biosensor analysis.38 Manufacturers are developing processes to produce influenza
vaccine using cell culture, which uses bioreactors similar to
Potency tests are used to assure that the vaccine produces the
the 2,000 L model (Figure 4-3) at the Vaccine Pilot Plant of
desired immunologic effect per specification. Most potency
the Vaccine Research Center, NIAID, National Institutes of
tests fall into one of four categories:
Health. Contrary to popular belief, there is no reduction in
– traditional animal-based assays that measure biological production cycle time for a batch of vaccine made by cell cul-
response in an animal model; ture compared with the use of embryonated eggs, although
– cell culture–based assays that measure biochemical and availability of the latter can be a problem during a pandemic
physiological response at the cellular level; event. Furthermore, the limiting factor in timely supply of
– biological activity or response induced by immunologic vaccine can be the capacity to perform the release testing,
interactions, and sterile filtration, and sterile filling of the vaccine.
– ligand or receptor binding assays, which are based on
in vivo attributes of the active molecule.37
Cell culture–based enzyme-linked immunosorbent assay
testing using known antibody-antigen interaction contin-
ues to replace or accentuate classical animal-based models.
It is desirable that the potency assay provide for the ability to
detect changes in the activity of the molecule over time. The
challenge for vaccine manufacturers is to develop the appropri-
ate correlates between chemical and biological approaches to
replace current animal models using a corresponding assay.38
Physical and chemical tests are performed on final formu-
lations and dosage forms to characterize the material. These
tests include pH, quantitation of preservatives, quantitation of
adjuvants, uniformity, particulate matter, loss on drying, and
residual moisture and dissolution for lyophilized final dosage
forms. A subset of the release testing provides for assessment of
stability to ensure that the vaccines continue to meet product
specifications over time.

Industry's response to new challenges

New technologies for manufacturing and testing
As described earlier, currently licensed vaccines are manufac-
tured as live attenuated, inactivated, purified subunit, conju-
gate, or recombinant protein antigens. The recent introduction Typical bioreactor used in vaccine manufacturing.
54 SECTION ONE
Several vaccines are under development that use plasmid DNA
as the delivery vehicle. Known as "genetic" vaccination, the pro-
tein antigen is delivered as DNA sequence, which is taken up
by the host and expressed in vivo. This particularly provides an
advantage for viral vaccines in which the host expression of the
protein may provide protein in a more native conformation and,
thus, a better immune response.
Manufacture of plasmid DNA vaccines is done in
grown in large bioreactors and a variety of downstream pro-
cessing unit operations.40 The isolation of the plasmid DNA
from the after lysis (chemical or heat) is followed by sol-
ids removal and chromatographic or selective precipitation.
The plasmid must be separated from genomic DNA, host cell
proteins, host cell RNA, and endotoxins. Although plasmid
productivity can be relatively high, human clinical studies for
prophylactic vaccines have used doses up to 8 mg of plasmid
DNA, making their potential manufacture a significant eco-
nomic and engineering challenge.41
Another genetic immunization strategy uses viruses to
deliver genetic sequences for pathogen protein. Viral vectors
include recombinant adenoviruses, poxviruses, and alphavi-
ruses.42 Some of the vectors are developed to be replication-
incompetent, like many of the adenoviral vectors. These
vectors contain deletions in the native viruses to render them
replication-incompetent for safety reasons. The deletion also
allows for the insertion of the gene of interest into the vector.
As the vectors cannot replicate on their own, they require
complementary cell lines called "packaging cell lines" that
provide the missing genetic information for replication. After
the vectors are expanded on the packaging cell line, the vec-
tors must be purified to separate them from host cell pro-
tein, DNA, and RNA. Other vectors like poxvirus vectors are
replication-competent and can be grown on permissive cell
lines without requiring genetic modifications. Production
techniques usually include cell lysis, chromatography, and
ultrafiltration to isolate the purified vectors.43
One of the most novel vaccine production technologies recently
proposed is the production of irradiated
sporozoites in mosquitoes for malaria. is
the pathogen responsible for malaria, and it has been reported
that a vaccine containing irradiated sporozoites conveys more
than 90% efficacy.44 Current estimates are that enough irradi-
ated sporozoites could be harvested from a single mosquito to
immunize a single human.45 One company, Sanaria, is explor-
ing whether sufficient quantities could be grown in mosquitoes
and harvested in a consistent and quality manner that would
allow for licensure of this type of vaccine. Initial clinical test-
ing showed suboptimal immunogenicity and protection when
given intravenously.46 Although certainly unconventional, the
fact that malaria currently causes approximately 300 million
clinical cases and 1 million deaths each year warrants further
development.
Increased capacity and responsiveness (avian
influenza pandemic preparedness)
One of the most recent major shifts in the industry has been
around the concern of an avian influenza pandemic and pre-
paring a strong response to reduce the impact of such a pan-
demic. The pandemic would be triggered by the combination of
an avian influenza strain with a human strain or mutation of
an avian strain allowing it to infect and spread in the human
population. As humans do not have natural immunity to avian
strains, even healthy adults are not expected to have the back-
ground antibody levels to fight infection. Such an event could
trigger a global epidemic taking millions of lives and also taking
a major toll on global economics.
As a first line of defense against the possibility of a pandemic,
the government and vaccine manufacturers have teamed up to
secure the supply chain for influenza vaccine, prepare and clini-
cally test avian influenza vaccine from circulating avian strains,
stockpile vaccine from circulating avian strains, and expand
manufacturing and distribution infrastructure for preparation
and delivery of vaccine. In addition to the fear the thought of a
pandemic brings to everyone who tries to understand it, a pan-
demic triggers many new issues for vaccine manufacturing.
During previous threats of pandemic influenza (eg, swine flu in
1976), the borders of countries were closed to transport of vac-
cines, leaving countries without manufacturing infrastructure
also without vaccine, at least for an initial response. The man-
ufacturers are not able to produce global supplies from a sin-
gle location if borders are blocked, and a more capital-intensive
regional approach (ie, multiple small plants) would be required.
The production of influenza vaccine requires many components
outside the manufacturing plant itself. All components of this
supply chain—eggs, vials, stoppers, reagents, and the labor to
prepare and deliver these components to the manufacturer—are
at risk during a pandemic. The vendors supporting the vaccine
manufacturers need to be protected and prioritized. One might
expect major disruptions in supply of goods during a pandemic
event; some firms could close or restrict operations owing to
the risk of illness or to absenteeism. Noncritical workers may
be restricted from the operating site or the critical workforce
sequestered to protect them from the spread of disease while
they support the production of vaccine to fight the disease.
This applies not only to vaccine manufacturers, but also to the
industries and agencies that support them.
Today's vaccines are still made in embryonated chicken eggs.
Traditionally, influenza vaccine was made January through
July, and chickens were replaced each year to maintain produc-
tivity and egg quality. If a pandemic were to start during the
time when chickens were not producing eggs for vaccines, the
response would be rather slow. Manufacturers have now estab-
lished year-round egg supplies allowing a strong and instant
response to a pandemic at any point in the year. (In the United
States, the Department of Health and Human Services supported
and funded this contingency supply.) Also, avian strains are not
abnormal in the bird market in the United States. (It is the threat
of human infection that is new.) Protecting the flocks from dis-
ease is always a concern and focus of manufacturers and the ven-
dors that support them. Biosecurity measures have been in place
since 1983, when a major avian outbreak destroyed the majority
of egg-producing flocks established by US manufacturers. These
measures have secured the supply of eggs ever since. Regardless,
contingency supplies have been added to secure production quan-
tity of eggs even if some chicken flocks are lost to avian influenza.
An outbreak of avian influenza would require termination of
the interpandemic production (annual flu vaccine) for up to
2 years, as excess capacity does not exist, even in support of
today's growing influenza vaccine needs. Even with routine

annual vaccination, more than 200,000 people are hospitalized
annually due to influenza. This number is expected to
increase significantly without annual vaccination. However,
the lack of inherent protection against avian strains makes
this vaccine still more important than the routine influenza
vaccination.
The current targeted vaccination response in the United
States is to prepare up to 600 million doses of monovalent
avian flu vaccine within 6 months. Approximately 166–173
million trivalent doses of influenza vaccine were available in
the US during the 2011–2012 flu season.46a To react in a more
urgent manner, facilities are being built and expanded globally
to support a larger, faster responsiveness with vaccine to con-
trol a pandemic outbreak. The ideal situation is one in which
the interpandemic production of influenza vaccine does not
have to be interrupted for a pandemic event.
The bottleneck for production in a pandemic event is expected to
be the supply of vaccine concentrate itself, but the need to fill and
finish large quantities in a short time is also not to be overlooked.
To meet the 600 million dose target in 6 months, 4 million doses
will need to be filled each day in addition to other products that
will continue to be supplied to maintain vaccination against all
disease. Supplies (and suppliers) of filling components could be
impacted by a pandemic event because absenteeism is expected
to be high and critical supplies could be at risk.
As transportation is an industry expecting significant impact in
case of a pandemic, vaccine distribution is not an aspect of sup-
ply to be taken for granted. An added consideration is the scar-
city of vaccine in the early response period, increasing the need
for security of shipments from loss due to temperature excur-
sion or interruption of shipments because of theft.
Overall, the planning and preparation for a pandemic event
is an all-encompassing exercise of every part of the vaccine busi-
ness and supporting agencies. The cooperation and assistance
of the governments globally has benefited all parties and will
benefit consumers in the end.
Growth in emerging market manufacturing
Many countries have long had domestic vaccine production
capability. Countries like China, India, Brazil, and South Korea
have more recently been developing new state-of-the art man-
ufacturing capability for the introduction of novel vaccines to
domestic populations and for exporting vaccines to the inter-
national market.
The expansion in emerging market manufacturing has been
driven by several key enablers. The first is the establishment of
central funds for the procurement and distribution of vaccines
in emerging markets. These organizations allow manufactur-
ers to work with single-point contact to procure and coordinate
distribution of vaccines to the markets they serve. Examples of
such organizations are the Pan American Health Organization,
which procures vaccines for much of Central and South
America, and the GAVI Alliance, formerly called the Global
Alliance for Vaccines and Immunization. This eliminates the
need to establish separate supply chain and sales channels into
these diverse markets.
Most of the vaccines exported into emerging markets use
the World Health Organization (WHO) qualification pro-
cess. The process established by WHO examines the quality
of the vaccine and manufacturing process and, in a separate
Vaccine manufacturing 4 55
evaluation, the technical and commercial ability of the
organization to manufacture the vaccine.47 This process pro-
vides a standard regulatory framework to enable manufactur-
ers to gain approval that is recognized in many of the emerging
markets and replaces the need to gain regulatory approval in
individual countries.
The global transition of manufacturing of many industrial
products into lower cost geographic regions is also evident in
vaccines. Most global manufacturers like GlaxoSmithKline,
Sanofi Pasteur, and Merck have all established joint ventures
or made acquisitions in India, China, or Brazil. These ventures
have in some cases, however, run into challenges for the multi-
national firms. Sanofi, for example acquired Shanta Biotechnics
in 2009.48 Shanta was an India-based vaccine manufacturer
supplying vaccines under the WHO prequalification program.
Soon after the acquisition, however, Shanta was cited for qual-
ity problems by WHO and lost its prequalification status for
one of its vaccines.49
Therapeutic vaccines
There has been substantial interest and development during the
past decade to use vaccines to elicit the body's immune system
to treat diseases after onset. These target diseases can be caused
by infectious agents, cancer, or autoimmunity. Current efforts
are being devoted to developing therapeutic vaccines against
tumors, AIDS, hepatitis B, tuberculosis, malaria, and autoim-
mune diseases such as myasthenia gravis, systemic lupus ery-
thematosus, and rheumatoid arthritis.50
In 2010, the FDA approved the first therapeutic cancer vac-
cine, sipuleucel-T (Provenge), which is developed and manufac-
tured by Dendreon. The vaccine was approved for use in men
with metastatic prostate cancer whose tumors are no longer
responding to hormonal therapy and has been demonstrated to
provide for more than a 4-month median improvement in over-
all survival compared with a placebo vaccine.51
The approval validates the concept of an active treatment
approach such as immunotherapy, which is intended to train
the immune system to attack cancer cells and potentially get a
response with long-lasting effect. Though not a cure, the vac-
cine provides significant clinical benefit.51
Sipuleucel-T is customized to each patient. Before treat-
ment, patients undergo a procedure called leukapheresis to iso-
late antigen-presenting cells (APCs) from their blood. These
APCs include dendritic cells and macrophages, among other
cells, that can “present” markers, or antigens, on their surfaces
that are recognized by other immune cells, thereby sparking an
immune response.51
The APCs are cultured with a proprietary manufactured pro-
tein. The end result is a vaccine with hundreds of millions of
“activated” APCs loaded with an antigen commonly found on
most prostate cancer cells, called prostatic acid phosphatase
(PAP). The vaccine is returned to the patient's treating physi-
cian and infused into the patient, with the intent of spurring
immune system cells, T cells, to neutralize tumor cells that
express PAP.51
Single-use manufacturing technologies
Historically, vaccines were each manufactured in a dedicated
facility to ensure segregation of each product from other products.
The development of more sophisticated engineering controls
for air handling, automated cleaning, and analytical testing of
residual product resulted in an FDA guideline in 1994 on the
use of pilot manufacturing facilities (which are multiproduct)
for the launch of biologics.52 This has resulted in a tremendous
amount of innovation in technologies that can enable rapid con-
version of facilities from one product to another. Typically, cells
would be grown in bioreactors made of stainless steel, as shown in
Figure 4-3. After use, these reactors have to be extensively cleaned
56 SECTION ONE General aspects of vaccination
Single-use bioreactor for cell culture (courtesy GE
Healthcare).
and tested to ensure they are cleaned adequately. Newer technolo-
gies use plastic bags, which are used once and then discarded.
An example of this system is shown in Figure 4-4. The use of
single-use technologies allows faster changeover from one batch
to another and from one product to another and reduced cleaning
and sterilization validation. Recent reports have shown substan-
tial savings in the cost and lead time of new facilities and a reduc-
tion in overall operating expenses using the new technologies.53–55
Removal of preservatives
The elimination of preservatives is an extension of increased
vaccine purity and a specific issue related to public perception
of vaccine safety. Examples of preservatives used in vaccine
formulation include thimerosal (a derivative of mercury), phe-
nol, benzethonium and formaldehyde, and 2-phenoxyethanol.
Benzethonium or 2-phenoxyethonol is often mixed with form-
aldehyde to improve effectiveness against a broader spectrum
of potential contaminants. These compounds have bactericidal
and/or bacteriostatic properties. Preservatives have been neces-
sary to improve the safety of vaccines historically, as the ability
to make a sterile product through the manufacturing, formula-
tion, and filling process with legacy manufacturing processes and
facilities is challenging, if not impossible, without preservatives.
Furthermore, a preservative is required in a multidose container
to prevent contamination of future doses during the extraction
of the first doses from the vial. The risk of an infection or sep-
sis due to a dose from a contaminated, unpreserved vial is con-
sidered far greater than the risk of an adverse event from the
preservative itself. To eliminate preservatives, the facilities and
processes used to manufacture vaccines had to be overhauled.
In some cases, the preservatives are used as inactivation
agents integrated in the manufacturing process and cannot be
fully removed. In these cases, the levels have been significantly
reduced by diafiltration but not fully eliminated.
Conversion to unit-dose presentations
A significant impact of the elimination of the preservatives was
the consequential switch from multidose vials to single-dose
presentations. The impact is three-fold: decreased production
capacity, increased product consumption per unit, and higher
consumption of storage space at the manufacturer, distributor,
and physician's office.
The largest initial impact on the manufacturer of the
removal of preservatives is the greater number of units that now
need to be filled. As an example, Sanofi Pasteur has typically
filled about 50 million doses of flu vaccine in 12 to 14 weeks
in 10-dose vials on a single filling line. In all, 5 million units
(10 doses each) needed to be filled, inspected, and packaged in
this period. With the elimination of preservatives in influenza
vaccine, 50 million units would now need to be filled in the
same time, and all existing filling lines at the manufacturing
site combined are not capable of filling that number of vials
and/or syringes in the time required to meet the need for the
influenza vaccine demand. A large investment in new lines is
necessary before making the switch to unit-dose presentations.
Because the same change to single dose is already underway for
other products, the capacity is already consumed.
The effort necessary to produce the high volume of vaccines
to satisfy the health need is increased 10-fold. Likewise, fill-
ing is not the only challenge. Every vial and syringe also needs
to be inspected for product and container/closure defects, pack-
aged with inserts into cartons, stored while awaiting regulatory
release, and shipped to customers. The entire supply chain for
this product is expanded 10-fold with this pending change.
A great benefit to multidose product presentations for man-
ufacturers and consumers is the savings in product for filling.
It is impossible to get the contents of a single dose vial into a
syringe for administration. Therefore, manufacturers need to
fill extra product that will never be administered to aid practi-
tioners in delivering the intended dose. This is called "overfill".
Overfill for a 10-dose vial is about 16% to 24% (or 0.58-0.62 mL
fill volume for a 0.50-mL dose). (Some practitioners claim the
ability to get 11 doses from a 10-dose vial.) When a single dose
is put into a vial, a higher overfill is needed to ensure a full
withdrawable dose. For unit-dose vials, overfill is 28% to 44%
(or 0.64-0.72 mL per vial for a 0.5-mL dose). The conversion of
a product from multidose vials to unit-dose vials results in an
additional loss of product owing to overfill of 20%. (Therefore,
Sanofi Pasteur's stated capacity of 50 million doses of flu vac-
cine is reduced to 40 million doses by this change alone.) This
change was not recognized in the rush to remove preservatives
from acellular pertussis vaccines and resulted in a nationwide
shortage of DTaP vaccine, as manufacturers did not have time
to increase capacity before the switch.
An alternative presentation, with less overfill for unit-dose
unpreserved product, is the prefilled syringe. The fill volume
for a prefilled syringe varies with design, but the fill volume for
some commercially available designs is as low as 0.53 mL, equal
to the lower milliliter per dose consumption range of 10-dose
vials. Syringes are more complicated to handle on a filling line
and also carry a higher per-unit cost of materials, but the sav-
ings of bulk make the change positive for manufacturers who
are limited for total bulk capacity. The US vaccine market has
been a largely vial market. Some products have been successfully
moved to the syringe in the United States as unpreserved pre-
sentations have been developed, including tetanus-diphtheria
vaccine (DECAVAC). However, total syringe filling capacity has
been limited, and major expansions are underway globally for
added syringe filling capacity to support the industry.
A negative impact from the conversion of products to unit-
dose presentations has been one of space. The storage capacity
of manufacturers, distributors, and practitioners will need to
be expanded because the package for 10 unit-dose vials is more
than two times larger than that for a 10-dose vial, and likewise,
the space required for 10 syringes can be more than 5 times
more than for a 10-dose vial. The industry is developing more
compact package designs to minimize the impact of the change
on consumers. The investment in new packaging equipment is
large and urgent.
The issue of preservatives, and their removal has had a great
cascading impact on the manufacturing and supply of vaccines
and has consumed a great deal of technical and engineering
effort and capital investment to resolve and to deliver existing
products to consumers in new forms. The true added benefit to
consumers may be difficult to measure, but the added costs and
complexity to the consumer are obvious.

Access the complete reference list online at http://www.expertconsult.com
2. Mathieu M, editor. Biologics Development: A Regulatory Overview. 2nd ed.
Waltham, MA: Parexel lnternational; 1997. p. 123–4.
6. Code of Federal Regulations, Title 21, Sec. 601.2(a). Washington, DC: Office
of the Federal Register, National Archives & Records Administration; April 1,
2001. http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/cfrsearch.cfm
[accessed 07.03.12].
9. Code of Federal Regulations, Title 21, Sec. 615. Washington, DC: Office of the
Federal Register, National Archives & Records Administration; April 1, 2001.
11. Stability testing drug substances and drug products. Rockville, MD: Center
for Drug Evaluation and Research and Center for Biologics Evaluation and
Research; June 1998. http://www.fda.gov/RegulatoryInformation/Guidances/
ucm128179.htm [accessed 07.03.12].
15. Code of Federal Regulations, Title 21, Sec. 610.18. Washington, DC: Office
of the Federal Register, National Archives & Records Administration; April 1,
2001.
Vaccine manufacturing 4 57
25. Palomares LA, Ramirez OT. Challenges for the production of virus-like
particles in insect cells: the case of rotavirus-like particles. Biochem Eng J
2009;45:158–67.
28. Code of Federal Regulations, Title 21, Sections 211–226. Washington, DC:
Office of the Federal Register, National Archives & Records Administration;
April 1, 2001.
29. Code of Federal Regulations, Title 21, Sec. 600. Washington, DC: Office of
the Federal Register, National Archives & Records Administration; April 1,
2001.
52. FDA Guidance Document Concerning Use of Pilot Manufacturing Facilities
for the Development and Manufacture of Biological Products. Fed Regist
1995;60:35750.
55. Robinson JM. Disposable process equipment: developing a disposable
process. Vaccines Europe. December 4, 2008.
 SECTION ONE: General aspects of vaccination
Evolution of adjuvants across
5 the centuries
Nathalie Garçon
Stanley Hem
Martin Friede
Adjuvants are substances that are added to vaccine antigens
to enhance and modulate the immunogenicity of the antigen.
The first adjuvants developed focused on increasing antibody
responses, and this has often been sufficient for the vaccines con-
sidered. During the last two decades, however, it has been realized
that simply increasing antibody responses is not always sufficient
for candidate vaccines to be effective. It has been observed that
adjuvants can be used very effectively to do the following:
– Provide a strong priming response in naïve populations,
effectively reducing the number of doses required to
induce protection
– Increase the duration of the immune response
– Enhance specific arms of the immune response such as
cell-mediated immunity (CMI), a critical target for many
of the remaining infectious diseases to which we do not
have vaccines
– Increase the breadth of the immune response to variable
antigens, enabling broader cross-protection
– Enhance the immune response in poorly responsive
populations, such as elderly and immunosuppressed
populations
– Allow for dose sparing of antigens where antigen supply
is limited
Generally speaking, adjuvants are useful for antigens such as
inactivated, subunit, and recombinant proteins, which can lose
some of the immunological information needed to trigger an
immune response. They are not required for live attenuated
vaccines, which carry the necessary immune-stimulating sig-
nals themselves. As discussed later, however, some preliminary
research suggests that adjuvants can have an effect on live vac-
cines as well.
Chance and necessity: The discovery
of adjuvants
The use of adjuvants has been known for more than a century, and
it is only recently that their mechanism of action has been eluci-
dated, in part owing to the progress in microbiology and immu-
nology. The first recorded observation of immune potentiation by
“adjuvants” is probably that of Coley, who in 1893 observed that
administration of killed bacteria (Coley's toxins) could in some
cases cure certain forms of cancer. It was only in the 1990s that it
was determined that this effect was due to immune stimulation
mediated via bacterial DNA. From there on, the specific oligonu-
cleotide sequences that could stimulate the immune response and
enhance it to a coadministered antigen were discovered.
It took another two decades to recognize the usefulness of
adjuvants to enhance humoral immunity: In 1925 Ramon1
observed that administering diphtheria toxoid to horses with
a variety of substances, including starch, plant extracts, or fish
oils, substantially enhanced the antibody response to the tox-
oid. A year later, Glenny2 observed a similar effect with alu-
minum potassium sulfate, or alum. Alum was used thereafter
as an adjuvant for numerous human vaccines, and to this day,
other aluminum salts, in the form of aluminum oxyhydroxide
or hydroxyphosphate, are the most widely used adjuvants in
human vaccines. The starch and fish-oils shown by Ramon to
act as adjuvants have, in the last decades, been tested in vac-
cines in the form of inulin and squalene, respectively.
During the following 80 years, a wide variety of substances
were tested as adjuvants, but many of them failed to be accepted
for human use. In the 1940s Jules Freund developed a water-in-oil
emulsion, the Freund adjuvant, in which the vaccine antigen is
emulsified as water droplets in a continuous mineral oil phase,
containing killed mycobacterium (Freund complete adjuvant) or
not (Freund incomplete adjuvant). The latter was briefly used
for a commercial influenza vaccine in the United Kingdom in
the 1960s, but was soon withdrawn owing to unacceptable reac-
togenicity. This, however, led to the development of oil-in-water
emulsions, in which oil droplets are present in a continuous
aqueous phase.
The first oil-in-water emulsions were based on a nonmetab-
olizable oil (squalane) and replaced later with metabolizable oils
(squalene), as opposed to mineral oils as in the original Freund
adjuvant. However, more recently, a water-in-oil emulsion sim-
ilar in structure to the Freund adjuvant has been introduced,
using mineral oil with a higher degree of purity that allows for
use in human vaccine candidates.
In the 1970s liposomes and virosomes were developed to
adsorb or encapsulate antigen. Liposomes consist of lipid layers
that form nanospheres and can encapsulate or integrate anti-
gens into their membranes. Several licensed vaccines contain
virosomes, which are reconstituted empty envelopes of influ-
enza viruses and similar to liposomes in structure.
A turning point: Better understanding of
immunology and its impact on development
of adjuvants
For most of the 20th century, adjuvant discovery and develop-
ment was based on observations and experimentation with no
clear immunological knowledge of the mechanism behind the
adjuvant effect. This, however, dramatically changed in 1996
64 SECTION ONE
reactogenic for use, partially due to MDP. However, because it
was less effective without the immunostimulant, the emulsion
was abandoned. Later, Chiron Corporation developed a range
of oil-in-water emulsions by replacing squalane with squalene
as another vehicle for muramyl derivatives. One of these emul-
sions (MF59) demonstrated some adjuvant properties as such
and was, therefore, further evaluated. MF59 and the majority of
the later developed oil-in-water emulsions used squalene, a nat-
ural, metabolizable product found in all plant and animal cells
where it is a precursor of cholesterol. The commercial source
is generally from shark liver, where it is abundant; alternative
sources such as phytosqualene are being explored.38 However, to
date, only squalene from shark origin allows for a product with
a purity level acceptable for human use.
Despite extensive clinical studies with a wide range of anti-
gens, MF59 was approved only in one vaccine, Fluad, an influ-
enza vaccine for older adults, and licensed in several primarily
European countries from 1997 onward. While there was benefit
of the adjuvanted vaccine in terms of antibody response to the
influenza hemagglutinin in the target population,39 the really
significant benefit of MF59 and other oil-in-water emulsions
became clear during investigations on pandemic influenza vac-
cines. The emergence of avian H5N1 influenza with occasional
human-to-human transmission and the fear that this could
become a pandemic led to intensive research in academic and
pharmaceutical environments for ways to immunize a largely
immunologically naïve population in the context of limited
antigen supplies. This was especially critical when it was shown
that for an H5N1 pandemic strain, sixfold more antigen was
required to induce an immune response to a level equivalent to
the seasonal influenza vaccine (90 g compared with 15 g). 40 It
was shown that MF59 enabled immunization with very signifi-
cantly reduced doses of antigen, down to 7.5 g, nearly a 12-fold
dose reduction.41
In parallel to the development of MF59, several other
oil-in-water emulsions were developed. For example an oil-
in-water emulsion containing -tocopherol as the immuno-
stimulating compound was formulated by GlaxoSmithKline
(GSK). This emulsion was tested earlier as part of the initial
development of a malaria vaccine, alone (AS03) or in combi-
nation with the immunostimulants MPL and QS21 (described
later).42 AS03 demonstrated very potent dose-sparing poten-
tial for pandemic influenza antigens, allowing for dose sparing
down to 3.75 g.43
In response to the need for dose sparing and the demon-
strated potential of oil-in-water emulsions, Sanofi Pasteur
developed AF03. This adjuvant is also squalene-based; however,
unlike the other emulsions, which are made by microfluidiza-
tion of the components, the emulsification of AF03 is achieved
without mechanical energy and uses a temperature-induced
self-emulsification process (PCT application WO2007080308).
Composition of various oil-in-water emulsions
EMA, European Medicines Agency.
With the outbreak of the H1N1 influenza pandemic,
European regulatory authorities approved three oil-in-water
emulsions containing pandemic influenza vaccines, with
MF59, AS03, and AF03 as adjuvants.
Other emulsions are under development, such as SE (stable
emulsion),44 a squalene-based emulsion, originally developed
by researchers at Corixa as a vehicle for MPL and synthetic
TLR4 agonists. This emulsion differs from the others in that
the emulsifier is a natural phospholipid rather than a sur-
factant such as Tween-80. SE has been tested in clinical tri-
als in combination with MPL in the context of a
vaccine.45
CoVaccine is an experimental adjuvant comprising sucrose
fatty acid sulfate ester, combined with squalane, in the form
of an oil-in-water emulsion. This adjuvant has been reported
to allow for dose sparing in the context of influenza vac-
cines.46 A single immunization with CoVaccine HT-adjuvanted
H5N1 influenza virus vaccine induces protective cellular and
humoral immune responses in ferrets and is undergoing clini-
cal evaluation.
Table 5-4 gives an overview of oil-in-water emulsions used as
adjuvants in licensed and investigational vaccines.
Adjuvant effect of oil-in-water emulsions on naïve vs
primed persons
Oil-in-water emulsion adjuvants are very effective at enhanc-
ing the immunogenicity and allowing for dose reduction in
pandemic influenza vaccines when the vaccinees are naïve. In
contrast, the adjuvant effect for seasonal vaccines in healthy
adults is quite poor.47 This suggests that these adjuvants are
excellent for priming but do not boost preexisting immune
responses very well. The benefit for priming is also very evi-
dent in the studies on the use of MF59-adjuvanted seasonal
vaccine in infants, who usually respond poorly to a single
administration of seasonal vaccine. These studies show a
strong effect of MF59 on immunogenicity48 and on efficacy
of seasonal influenza vaccines with the effect being strongest
in the youngest ages.48a The adjuvanted vaccine demonstrated
89% efficacy against vaccine-matched strains during two influ-
enza seasons compared with 45% for the nonadjuvanted sea-
sonal influenza vaccines group.
The situation is slightly different in the elderly population.
Older adults have, in general, been primed to seasonal influ-
enza; however, immunological senescence results in a decreased
ability to induce sufficient antibody responses to conventional
influenza vaccines. Fluad, the MF59-containing seasonal influ-
enza vaccine that has been licensed in numerous countries
for the last decade, has been shown to enhance the immune
response in this nonresponder population.49 There are no
efficacy data, however, that correlate this increase in immuno-
genicity with increased efficacy.

Enhancing the breadth of the immune response
Possibly one of the most important breakthroughs in adjuvant
research during the last 2 years was the observation during the
development of H5N1 pandemic influenza vaccines that oil-in-
water adjuvants not only enhance the immune response and
allow for dose reduction, but also enhance diversity and affinity
of the antibodies induced.50 This qualitative and quantitative
expansion of the antibody repertoire has tremendous relevance
for vaccination against pathogens, which undergo frequent anti-
genic drifts, such as influenza, as this would reduce the need for
a perfect match between the antigen and the circulating patho-
gen strain.
This observation was first demonstrated in the ferret chal-
lenge model with the AS03-adjuvanted H5N1 vaccine through
cross-neutralizing antibodies and lethal challenge.51 The cross-
neutralizing antibodies were confirmed later in clinical settings
using the same AS03-adjuvanted H5N1 vaccine.52 Data pub-
lished with MF59-adjuvanted H5N1 vaccine showed the induc-
tion of epitope spreading from HA2 to HA1 hemagglutinin and
to neuraminidase, suggesting that this is a common feature of
this family of adjuvants. MF59 adjuvant enhances diversity and
affinity of antibody-mediated immune response to pandemic
influenza vaccines.53
Mode of action
For many years oil-in-water emulsions were classified as vehi-
cles, and it was assumed that the mode of action was primarily
through enhanced delivery of the antigen to APCs or to the lymph
nodes, even though most antigens do not associate physically
to the oil droplets. It has, however, recently been shown that
these emulsions stimulate the immune response indirectly.
Using gene microarray analysis, Mosca and colleagues54 demon-
strated that skeletal muscle fibers are the target of MF59, where
the adjuvant induces production of PTX3 and JunB, which in
turn stimulate production of TNF- , IL-1B, and CCLs, result-
ing in activation of resident APCs and recruitment and activa-
tion of circulating APCs. A similar mechanism has also recently
been demonstrated for the -tocopherol–containing oil-in-water
emulsion AS03. Morel and colleagues55 showed that similar
to MF59, AS03 induces expression of a range of cytokines,
granulocyte recruitment at the injection site and increased
antigen uptake by monocytes and migration to the draining
lymph node. In this adjuvant, however, the expression of some
of the cytokines such as IL-6 was modulated by the presence of
-tocopherol, which also enhanced the magnitude of the
immune response, suggesting an immunomodulatory action
of -tocopherol independent of the oil emulsion. The same
authors also demonstrated that the adjuvant effect is local and
that temporal and spatial colocalization of antigen and adju-
vant were required, ie, injecting the adjuvant in a site distant
to the antigen or at a later time resulted in no adjuvant effect,
consistent with a local and short-lived direct impact of the
immune system.
Safety of squalene-containing oil-in-water emulsions
During the H1N1 pandemic, two influenza vaccines contain-
ing an oil-in-water emulsion as adjuvant were approved and
distributed in Europe (Pandemrix and Focetria); however, the
uptake of these vaccines was hampered by broad media claims
of the dangers of squalene. Many of these concerns seem to
have arisen from a 2002 report that soldiers returning from the
Gulf War with the so-called Gulf War syndrome had antisqua-
lene antibodies, which had been induced through receiving vac-
cines allegedly containing squalene.56 Although the vaccines in
question did not apparently contain squalene, the WHO Global
Advisory Committee on Vaccine Safety reviewed all available
data, including data from clinical trials with the squalene-
containing Fluad vaccine. Animal studies have also suggested
that squalene can induce arthritis.57 However, in the animal
Evolution of adjuvants across the centuries 5 65
model used, the requirement for a specific breed and the use of
a complex protocol irrelevant to the vaccination practice make
it difficult to assess the relevance of these data for safety evalu-
ation in humans. The committee concluded that there was no
evidence that squalene could induce pathological antisqualene
antibodies. Before recommending the use of squalene-containing
oil-in-water emulsion for the H1N1 pandemic, the WHO also
reviewed all clinical data from more than 35,000 volunteers
of all ages and concluded that there were no significant safety
concerns.
Thorough safety surveillance by authorities during the 2009-
2010 pandemic season showed a positive benefit-risk profile for
the vaccines. Since August 2010, an increased number of cases
of narcolepsy was reported in children and adolescents vacci-
nated with an AS03-adjuvanted H1N1 pandemic vaccine in
three northern European countries. Interim reports of epidemi-
ologic studies conducted in those countries have suggested an
increased risk of narcolepsy in some vaccinated persons.58
The European Medicines Agency Committee for Medicinal
Products for Human Use (CHMP) undertook a thorough
review of all available data as of July 2011, and concluded that
the overall benefit-risk profile of the AS03-adjuvanted H1N1
vaccine remains positive. The CHMP acknowledged that fur-
ther research in the area of genetic and environmental factors
in particular need to be explored before definitive conclusions
can be drawn.
Other oil alternatives for squalene
Little investigation into alternative metabolizable oils has been
reported. As mentioned earlier, squalane was used in the SAF
emulsion and is included in the CoVaccine adjuvant; however,
it is not clear whether this oil can be metabolized or whether
it is eliminated through the skin.59 Miglyol, a metabolizable
semisynthetic mixed triglyceride has been evaluated as an oil-
in-water emulsion adjuvant60 and was shown to enhance immu-
nity but at lower titers than a nonmetabolizable mineral oil. In
the future, synthetic oils may overcome some of the challenges
associated with the use of animal-derived squalene.
Although the TLRs and their role in triggering the innate
immune response were understood only after their ligands were
identified as adjuvants, it is useful to classify these adjuvants
through their specific receptor.
Although it has long been known that LPS, a major component
of the outer membrane of gram-negative bacteria, is a potent
stimulator of the immune system, its use in adjuvants has been
curtailed by its toxic effects. Early studies demonstrated that
removing the core carbohydrate, generating lipid A, reduced the
pyrogenicity without reducing the immune-stimulating activity.
Lipid A was still too pyrogenic for use; however, it was shown
that formulating lipid A in liposomes reduced the pyrogenicity
a further 100- to 1,000-fold, and liposomal lipid A formulations
were used in clinical trials for candidate malaria vaccines with-
out severe adverse events.61 Ribi61a demonstrated in 1984 that
a significantly less toxic molecule could be obtained from LPS
through sequential acidic and basic hydrolysis steps. The phos-
phate from the reducing-end glucosamine of lipid A derived from
RC595 is removed by mild acid hydroly-
sis, resulting in a molecule, referred to as monophosphoryl lipid
A (MPLA, referring to all other TLR4 agonists), that was sig-
nificantly less toxic than the lipid A and that still stimulated
the immune response (Ribi et al).5a Ribi and colleagues then
observed that if this MPLA was subjected to a further mild alkali
hydrolysis, which selectively removed the acyl chain from the 3
position of the disaccharide backbone, the resulting molecule,
66 SECTION ONE
3-deacylated MPLA (3d-MPL, or MPL), was even less pyrogenic
but still exhibited adjuvant activity. GSK developed a combina-
tion of adjuvants, AS04, in which MPL is formulated with alu-
minum salt. The adjuvant system AS04 is used in two approved
vaccines (Cervarix and Fendrix). Cervarix is aimed at preventing
HPV infection and subsequent development of cervical cancer.
Extensive clinical trials demonstrated an acceptable safety pro-
file and efficacy of the vaccine against the vaccine strains during
a period of more than 8 years in clinical trial follow-up62 and also
against divergent HPV strains not present in the vaccine.63
Mechanism of action of TLR4 agonists
Immune recognition of TLR4 agonists such as LPS or MPL
is initiated by extraction of monomers from the aggregates
by LPS-binding protein in the serum.64 Monomeric LPS is
transferred from LPS-binding protein to another accessory
protein, CD14, which in turn transfers LPS to MD2, a secreted
glycoprotein that associates with the extracellular domain of
TLR4 to form the heterodimeric receptor that is responsible
for the physiological recognition of LPS.65 Lipid A bound to the
TLR4/MD2 complex activates two distinct intracellular signal-
ing pathways, which have come to be known by the names of
the TLR4-proximal adaptor proteins, MyD88 and TRIF.66 The
Myd88 pathway leads to the activation of mitogen-activated
protein kinase– and nuclear factor- B–dependent proinflam-
matory responses, while the TRIF pathway activates kinases
responsible for type I interferon responses.67 This is shown
schematically in Figure 5-1. To this extent, the TLR4 receptor
is unique among the TLRs by its ability to induce two distinct
signaling pathways. The activation of these pathways is depen-
dent on the structure of the agonist, where minor changes in
the number and length of the acyl chains can have a major
effect on pathway activation. MPL has been reported to activate
the TLR4-Tram-TRIF-based signaling pathway and the TLR4-
Mal-Myd88 pathway.68
Species specificity of TLR4 receptor
One challenge exists, however, in the development of TLR4
agonists for use in humans—the variability in the receptor
specificity across species. Human, but not murine, TLR4/MD2
transmits proinflammatory signals in response to hexa-acylated
but not penta-acylated LPS.69 Moreover, in humans, but less so
in rodents, tetra-acylated and penta-acylated forms can inhibit
the adjuvant activity. As a result, molecules shown to work in
rodents will not necessarily function the same way in humans.
One example of this is the OM-174 molecule, a tri-acylated
molecule that functioned in mice70 but apparently did not in
humans.
Other TLR4 agonists
As MPL is isolated from a gram-negative bacterium, ,
and such extraction can present manufacturing hurdles, many
groups started to develop MPL analogs by chemical synthesis.
The structure of MPL and these analogs is shown in Figure 5-3.
RC529, a member of the aminoalkyl glucopyranosides devel-
oped by Corixa scientists to provide a synthetic alternative to
MPL, is structurally similar to a hexa-acyl MPL, but the reduc-
ing terminal glucosamine has been replaced by a nonsaccha-
ride backbone.71,72 This molecule, formulated with alum, is in a
hepatitis B vaccine approved in Argentina (SUPERVAX) that is
reported to increase immune response compared with nonadju-
vanted vaccine.73
Glucopyranosyl lipid adjuvant (GLA) is a synthetic hexa-acyl
form of MPL, but it is designed on the form of
LPS rather than the LPS. The molecule has a sin-
gle acyl chain on the 2 amine and has an acyl chain on the 3
hydroxyl.74 Clinical studies are in progress.
E6020 is a synthetic TLR4 agonist that is not based on a
saccharide backbone, but it is able to stimulate the TLR4/MD2
pathway.75 E6020 is in phase 1 trials.
The Brazilian vaccine manufacturer, Butantan, has initi-
ated a program to produce MPL from the LPS extracted from
, a side product of the whole-cell pertussis
vaccine process. This has been reported to enhance immunity
to influenza vaccines.76 Hepta-acyl sulphonyl sucrose, a compo-
nent of the CoVaccine HT adjuvant, currently in clinical devel-
opment, bears strong resemblance to the backbone structure of
TLR4 agonists and is thought to also act through TLR4.46 One
question remains, however, concerning the ability of synthetic
TLR4 agonists to overcome polymorphism associated with the
TLR4 receptor.77 Indeed, it is not clear whether all recipients
will respond equally to a given TLR4 agonist molecule and
whether a mix of various molecules, as present in the differ-
ent naturally produced TLR4 agonists, will not be necessary to
compensate for human diversity.
Formulation challenges
Ensuring an optimal and reproducible adjuvant effect requires
some formulation know-how. In the currently approved vac-
cines with TLR4 agonists (Cervarix and Fendrix), the MPL is
combined with an aluminum salt, and the combination of the
two adjuvants is referred to as AS04. The combination of MPL
with an aluminum salt adjuvant demonstrates some of the
challenges of using MPL and analogous multiacylated disaccha-
ride TLR4 agonists. Only a well-defined and controlled process
can ensure that those molecules, which are insoluble in water,
will not clump into aggregates that would lead to difficulty dur-
ing sterilization by filtration or variability in the adjuvant activ-
ity. Various other approaches have been adopted to formulate
TLR4 agonists, such as their incorporation into small unila-
mellar liposomes permitting a stable formulation. However, the
activity is dependent on the ratio of lipid to agonist and how the
agonist is combined with the liposomes.78 Other alternatives
include the incorporation of the agonists into oil-in-water emul-
sions or application of thermal or sonic energy to form colloids,
which can be further stabilized by the addition of small quan-
tities of lipids. The immune enhancement induced by TLR4
agonists is highly dependent on the physical structures, and
determination of the consistency and stability of these requires
detailed physicochemical characterization.79 Successful use of
these adjuvants, therefore, requires careful consideration of
how to formulate them and how these amphipathic molecules
will interact with other components of the vaccine, such as sur-
factants, aluminum salts, and antigens.
Safety of TLR4 agonists
MPL, by far the most widely used of the TLR4 agonists, has
demonstrated an acceptable safety profile. A meta-analysis of
data from 11 clinical trials and more than 74,000 volunteers (2/3
receiving the vaccine that contains AS04 and 1/3 the control)
has shown no increase in severe adverse events over aluminum
alone or hepatitis A vaccine controls.62 It has been administered
to millions of young women worldwide with few reported severe
adverse events. This was further supported by the mechanism of
action of AS04, which demonstrated that the adjuvant effect is
local and that temporal and spatial colocalization of antigen and
adjuvant were required; injecting the adjuvant in a site distant
to the antigen or long after injection of the antigen resulted in
no adjuvant effect.80
RC529 (Figure 5-3), a synthetic TLR4 agonist in an approved
hepatitis B vaccine, has a limited history of clinical use, but as
for MPL, no severe adverse events have been associated with the
vaccine containing it. It is, however, not certain that other TLR4
agonists will have the same safety profile. Minor changes in the
structure of TLR4 agonists can modify the way they activate
the MyD88 or TRAM-TRIF pathway, promoting type 1 inter-
feron or proinflammatory responses, and, as described earlier,
the difference in receptor specificity between rodent and human
TLR4 receptors may prevent these being detected in preclinical
models. The acceptability and extensive history of whole-cell
 Evolution of adjuvants across the centuries 5 67

Structure Name Structure Name

Structures of TLR4 agonists in development or in licensed vaccines. GLA, glucopyranosyl lipid adjuvant; MPL, monophosphoryl lipid A.
68 SECTION ONE
pertussis vaccines, which contain residual bacterial LPS, sug-
gests that this is a pathway that can be stimulated without long-
term adverse effects.
Bacterial DNA is recognized by mammalian species as a PAMP
and serves as a potent activator in the innate immune system
through interaction with the intracellular TLR9. Fragments of
bacterial DNA and synthetic single-stranded oligodeoxynucleo-
tides containing unmethlyated CpG motifs (CpG ODNs) found
in bacterial DNA have also been demonstrated to be powerful
adjuvants.81,82 In vivo, however, the lability of phosphodiester
bonds results in rapid degradation of DNA oligonucleotides, and
it has been found that replacing these labile ester bonds in the
oligonucleotide sequences with phosphorothioate bonds stabi-
lizes the oligonucleotides and significantly enhances the activ-
ity.83 Most preclinical and clinical studies with CpG-containing
oligonucleotides contained this modification.
While obtaining pure CpG-containing oligonucleotides
is easy, the development of this class of molecules as adju-
vants for human vaccines has been severely hampered by the
fact that the specific hexameric CpG motifs inducing optimal
immune enhancement differ between species.81 Hence, eval-
uation of efficacy and safety in rodents is not readily trans-
lated to humans. Furthermore, the distribution of the TLR9
receptor and the pathways that its activation promotes also
differ across species: In humans, CpG motifs are recognized by
TLR9 found on NK cells,84 B cells, and plasmacytoid dendritic
cells81 but are not recognized by myeloid dendritic cells and
monocytes. In contrast, in mice, these latter express TLR9,
which recognizes CpG motifs. In both species, CpG motifs
trigger B-cell activation and induce directly or indirectly the
production of Th1 and proinflammatory cytokines, includ-
ing IL-1, IL-6, IL-18, TNF- , and interferon- ; in some cases,
CpG can redirect preestablished Th2 responses toward Th1.85
CpG ODNs have also been shown to act as adjuvants for anti-
gens delivered by the mucosal route of administration.86
Numerous clinical evaluations of CpG as an adjuvant
have been performed, most commonly with a sequence of
CpG referred to as 7909, a sequence found to be optimal for
human use. CpG has been evaluated with and without alum
for malaria vaccines,87,88 conjugate pneumonia vaccines,89
and hepatitis B vaccines90,91 and in HIV-infected patients in
whom it has been shown that CpG permits rapid and long-
term seroprotection.90 Furthermore, it has been recently dem-
onstrated in an elderly target population that when chemically
conjugated CpG (1080 sequence) to the HBV antigen, the
seroprotection induced was faster, superior, and more dura-
ble than three doses of a licensed comparator HBV vaccine
(Engerix-B(®).90a
Other TLR9 agonists
It was thought for a long time that TLR9 receptors were spe-
cific for unmethylated CpG sequences; however, it has been
shown that poly(dI:dC), a DNA version of the well-known
TLR3 agonist poly(I:C), is also a potent adjuvant, acting
through the TLR9/MyD88-dependent pathway.92 This oligo-
nucleotide, in conjunction with a cationic peptide, is being
developed as an adjuvant termed IC31 in clinical trials with
recombinant tuberculosis antigens and has been reported to
induce long-lasting T-cell responses to the antigens.93 It is
thought that the mode of action of IC31 is dependent not only
on the poly(dI:dC) TLR9 agonist, but also on the interaction
of the oligonucleotide with the coformulated cationic peptide.
The complex formed entraps the antigen and acts as a physi-
cal depot at the site of injection, resulting in sustained antigen
and adjuvant release.94 As for the examples given with TLR4,
this demonstrates how critical formulation parameters are to
successful adjuvant development.
The molecules described in the preceding sections account for
the vast majority of late-stage clinical trials and approved vac-
cines containing adjuvants. However, numerous other TLR ago-
nists are entering or are in early phase clinical trials. These are
briefly described in the following paragraphs.
Flagellin is a protein that polymerizes to form the flagella of
flagellated bacteria and is recognized by TLR5. This protein has
been expressed as a fusion partner to influenza hemagglutinin
or influenza M2e, in which it serves as an adjuvant, enhanc-
ing the immune response to the influenza antigen as shown in
clinical trials.95–97
Numerous small molecules have been found to act as ago-
nists of TLR7 and TLR8. The best known of these are the
small molecule nucleoside analogs, imiquimod and resiquimod
(R-848),98,99 that are TLR7 and TLR7/TLR8 ligands, respectively
(both produced by 3 M). Other well-known agonists include
loxoribine100 and bropirimine.101 The use of these molecules
as adjuvants is complicated by their low molecular weight and
rapid removal from the site of injection. Several formulation
methods have been shown to enable these molecules to be used
as adjuvants. Aldara cream, a 5% topical preparation of imiqui-
mod that is licensed for treatment of genital warts and basal cell
carcinoma, has been tested as an adjuvant for topical admin-
istration at the site of subcutaneous or intradermal injection
of antigens.102,103 This has been shown to have some adjuvant
activity in human clinical trials for cancer.104
There are several other adjuvants in advanced clinical develop-
ment that do not fall into any of the aforementioned catego-
ries. The best known of these are the saponins, for which the
mechanism of action is not fully understood. In this category,
however, one could also include adjuvants such as virosomes,
cationic liposomes, and polyelectrolytes, a class of polymers,
which includes polyoxidonium (a component of the Grippol
influenza vaccine produced in Russia) and polyphosphazenes,
which are being evaluated in several clinical trials.
Saponins
Saponins are triterpenoid molecules with a complex sugar back-
bone extracted from a variety of plants. The most widely used of
these extracts is the saponin extract from the South American
tree Molina, referred to as Quil-A. It has been
used as an adjuvant in veterinary vaccines since the 1970s. This
mixture of saponins, however, with varying adjuvant activity and
toxicity was found to be too reactogenic for use in humans. The
Quil-A saponins have an exquisite affinity for cholesterol, and
when in contact with membranes containing cholesterol, they
form a complex creating pores in the membrane. At the site of
injection, this results in considerable reactogenicity. However,
this affinity for cholesterol has also enabled the development
of two adjuvants for human use based on the Quil-A saponins:
ISCOMS and AS01. ISCOMS are complexes of partially puri-
fied saponins from Quil-A, combined with cholesterol, to form
small porous particles 50 to 60 nm in diameter with a char-
acteristic buckminsterfullerene shape and a high density.105,106
Originally the antigen was associated with these structures
through entrapment, conjugation, or hydrophobic interaction,
processes that were tedious and not readily applicable to large-
scale manufacturing. More recently, it has been shown that
association of the antigen with the particle is not required,
and simple coadministration with antigen is adequate. As a
result, simpler formulations (Iscomatrix and Iscoprep), which
can be simply mixed with the antigen, have been developed
by CSL (Australia).107,108 These cholesterol-saponin complexes
are less reactogenic than the parent saponin, yet maintain a
strong adjuvant effect. In clinical trials, they have been shown

to be slightly more reactogenic than placebo or active control;
however, the reactogenicity is generally mild and is accept-
able. No other vaccine-related severe adverse events have been
reported.109
An alternative approach to reduce the toxicity of Quil-A sapo-
nins was taken by Kensil and colleagues,110 who isolated a pure
component, termed QS21, from the mixture of saponins that
was shown to be less toxic yet to retain adjuvant activity. QS21,
however, had two drawbacks: It is chemically unstable at even
mildly alkaline conditions, and, while less toxic than the parent
mixture, it was still reactogenic. It was found that by combining
QS21 with liposomes that contained cholesterol, the stability of
the molecule was significantly enhanced and the reactogenicity
abrogated.111 The adjuvant activity of this combination could be
further enhanced by the addition of the TLR4 agonist MPL, and
the resulting combination of adjuvants is termed AS01 by its
developers (GSK). This formulation has been tested in the clini-
cal setting and shown to induce higher CD4 T-cell responses
to a plasmodium antigen than the same immunostimulants
combined with an oil-in-water emulsion.112 Further challenge
studies demonstrated the higher efficacy induced by this for-
mulation, which led to the selection of AS01 for the phase 3
efficacy study of the candidate malaria vaccine.113,114 This once
again demonstrates the critical aspect of formulation and the
challenges to identifying appropriate formulations for optimal
activity of adjuvants. QS21 in its pure form is also being used
as an adjuvant for cancer vaccines.115
The exact mechanism of action of QS21 is not fully eluci-
dated. The loss of adjuvant and lytic activity when the mol-
ecule is hydrolyzed suggests that membrane lysis has a role.110
However, the adjuvant activity is also lost when the aldehyde
function on the triterpenoid backbone is removed.116 Several
synthetic analogs have been developed that may permit a more
detailed analysis of critical components of the molecule and a
better understanding of its mechanism of action.117,118
There has been recently a renewed interest in saponin-based
adjuvants from other plant sources, particularly from research-
ers in China and India.
Virosomes
Virosomes are reconstituted liposomes containing viral (typi-
cally influenza virus) proteins in the liposomal membrane and,
optionally, with additional antigens incorporated in the lipo-
somal membrane or attached to the membrane.119 Virosomes
have been used for influenza vaccines (licensed in Europe as
Inflexal) and as adjuvants for hepatitis A vaccine (licensed in
Europe as Epaxal). They are under investigation as adjuvants for
numerous other targets.
Polyelectrolytes and polycations
The influenza vaccine Grippol licensed in Russia contains polyox-
idonium,120 a polyelectrolyte copolymer of N-oxide 1,4-ethylene
piperazine and (N-carboxyethyl)-1,4-ethylene piperidium bro-
mide, that has been shown to have immunostimulatory
properties.121 While little has been published about this spe-
cific polymer and its use as an adjuvant, an analogous polymer,
poly(carboxylatophenoxy)phosphazene (or PCPP) has been
widely investigated as an adjuvant for vaccines and has been
shown to exert adjuvant activity.122,123
Polycations such as polyarginine,124 chitosan,125,126 and cat-
ionic lipids127 have also been shown to exert adjuvant activity.
As for polyelectrolytes, the mode of action is not fully under-
stood but is likely to involve interaction with cell membranes
similar to that described for alum.25 Another cationic adjuvant,
CAF01, is a liposome-based adjuvant composed of the cationic
quaternary ammonium lipid dimethyldioctadecylammonium
(DDA) and the synthetic analog of mycobacterial cord factor;
trehalose dibehenate (TDB). CAF01 was originally developed
as a CMI-promoting adjuvant for a subunit vaccine against
Evolution of adjuvants across the centuries 5 69
tuberculosis, but the adjuvant has since been demonstrated
to promote a diverse immune response resulting in CMI and
humoral responses128,129 and is in clinical evaluation.
Mucosal adjuvants
There is a strong interest in the vaccine community in admin-
istering vaccines via the mucosal route. This route offers the
advantages of being easier to administer and not requiring
trained health care workers, but also has the potential of induc-
ing mucosal immunity at the portal of entry of many pathogens.
While mucosal delivery and induction of mucosal immunity
can easily be achieved with live attenuated vaccines (for exam-
ple, oral poliovirus vaccines, oral rotavirus vaccines, nasally
administered live attenuated influenza vaccines), inducing sys-
temic or mucosal immunity with nonlive antigens is much
more difficult. These tend to be nonimmunogenic or poorly
immunogenic via the mucosal routes, and, hence, numerous
adjuvants have been evaluated to enhance the immunogenicity.
The most widely investigated are the bacterial ADP-
ribosylating exotoxins such as cholera holotoxin and
heat-labile enterotoxin.130 These are potent mucosal adju-
vants; however, their toxicity has, for the most part, precluded
their use as adjuvants, and, hence, mutations have been intro-
duced in an attempt to reduce their toxicity while aiming to
retain some adjuvant activity:131 these include adjuvants such
as mLT(R192G), dmLT,132 and LTK63.133,134 For nasal delivery,
these adjuvants have, however, presented unacceptable safety
concerns: The native toxin was used as an adjuvant in an intra-
nasal influenza vaccine that was licensed in Switzerland but
was associated with occasional severe adverse events in the
form of Bell palsy (partial facial analysis) ascribed to the adju-
vant,135 and similar effects were seen in an experimental vac-
cine using the LTH63 detoxified mutant.136 These adjuvants all
contain the pentamer B subunit of the toxin, which binds gan-
glioside gM1 on nerves and can be transported through retroax-
onal transport to the root of the nerve, and it is thought that
the neurological adverse events may have arisen through such
an interaction.136 An alternative adjuvant, which does not con-
tain the B subunit and, hence, does not bind gM1 gangliosides,
is CTA1-DD, currently in development.137,138 These exotoxins
have also been evaluated for oral delivery in several clinical tri-
als; however, their susceptibility to acidic environments requires
enteric formulations, and none are in advanced development.
Future directions
The last decade has seen an unprecedented surge in understand-
ing how adjuvants work. While the history of adjuvant discov-
ery and development has evolved through serendipity, we are
now in a position in which identification, selection, and use of
adjuvants can be rationalized. This has opened the door to vac-
cine candidates against as yet unmet medical needs, vaccines
that will not only prevent infectious diseases, but also could
treat life-threatening diseases such as cancers or allow for the
management of chronic disorders such as Alzheimer disease.
There are, however, a number of factors and unknowns that
needs to be considered. The most important of these is how
to demonstrate the safety of adjuvants. As pointed out several
times, there is an underlying concern that is not necessarily
based on sound science that adjuvants can induce immune-
mediated disorders. The absence of suitable animal models and
the differences in receptors and receptor distribution between
animals and humans can make it difficult in some cases to
demonstrate that this is not the case.
Despite the occurrence of autoimmune reactions in the gen-
eral population, the fear that adjuvants can induce or exacerbate
autoimmune disorders is one that is at the forefront of follow-
up and surveillance. Holding clinical development of a vaccine
70 SECTION ONE
based on a single adverse event demonstrates the challenges of
developing vaccines with novel adjuvants. The risk-benefit bal-
ance associated with the vaccine being developed will always be
the guiding principle to ascertain the value of a given approach.
A large safety database to demonstrate rare events in such stud-
ies is not always possible, which underlines the difficulty intro-
ducing new approaches into the field of vaccines. The paucity of
epidemiologic data on such immune-mediated disorders, which
are key to establishing the background rate of disease, needs to
be addressed so that factual analysis can be undertaken when
such events occur.
The safety evaluation of a vaccine encompasses all constitu-
ents of the product. It cannot be assumed that an adjuvant that
is safe in one vaccine with a given antigen will be safe when
added to another vaccine, even if the latter vaccine has been
shown to be safe without adjuvant. A rational approach requires
nonclinical toxicology, determination of the mode of action of
the adjuvant, an evaluation of differences in receptors and activ-
ity in animal and human cells, controlled clinical trials, and
postmarketing surveillance.139
Also, as pointed out, adequate formulation is critical for the
activity of many adjuvants. Yet the know-how for adjuvant for-
mulation is not widely available, while predicting how the phys-
icochemical parameters of an adjuvant and its interaction with
the antigen affect immune responses is key to its selection. All
these points emphasize the criticality of process development,
robustness, and reproducibility and the ability to characterize
adjuvants in a relevant and efficient way.
Access the complete reference list online at http://www.expertconsult.com
5 Janeway Jr CA, Medzhitov R. Innate immune recognition. Annu Rev
Immunol 2002;20:197–216.
21 Marrack P, McKee AS, Munks MW. Towards an understanding of the
adjuvant action of aluminium. Nat Rev Immunol 2009;9:287–93.
39 Minutello M, Senatore F, Cecchinelli G, et al. Safety and immunogenicity
of an inactivated subunit influenza virus vaccine combined with MF59
adjuvant emulsion in elderly subjects, immunized for three consecutive
influenza seasons. Vaccine 1999;17:99–104.
55 Morel S, Didierlaurent A, Bourguignon P, et al. Adjuvant system AS03
containing alpha-tocopherol modulates innate immune response and leads
to improved adaptive immunity. Vaccine 2011;29:2461–73.
There is a host of adjuvants in clinical development; however,
none tested in humans so far has the ability to induce
functional CD8+ T cells to a level that can be seen with live
viral vaccines. Live viral vaccines have their limitations, in
particular with respect to repeated boosting and their use in
immune-suppressed persons. There is, therefore, a need to pur-
sue research into adjuvants capable of inducing CD8+ T-cell
responses.
Finally, because vaccines are built on the combination of
antigen(s) and adjuvant(s), the need for relevant and immu-
nogenic antigens should not be overlooked. Because more
and more vaccines will require the induction of humoral
immunity and CMI, there is a need for researchers to seek to
improve the intrinsic immunogenicity of the antigen and to
ensure optimum immune responses if the addition of an adju-
vant is required.
It is only through the appropriate combination of antigen
and adjuvant, selected on the basis of the targeted disease, rel-
evant protective immune response, and target population, that
adjuvants will fulfill their promises and find their place as a rel-
evant and effective tool for improving human health.
Acknowledgment
This chapter includes a significant contribution written for the
previous edition by our late colleague, Dr. Stanley Hem.
63 Paavonen J, Naud P, Salmeron J, et al. Efficacy of human papillomavirus
(HPV)-16/18 AS04-adjuvanted vaccine against cervical infection and
precancer caused by oncogenic HPV types (PATRICIA): final analysis of a
double-blind, randomised study in young women. Lancet 2009;374:301–14.
82 Klinman DM, Currie D, Gursel I, et al. Use of CpG oligodeoxynucleotides
as immune adjuvants. Immunol Rev 2004;199:201–16.
114 Cohen J, Benns S, Vekemans J, et al. malaria vaccine candidate RTS,S/AS is
in phase III clinical trials [in French]. Ann Pharm Fr 2010;68:370–9.
139 Garçon N, Segal L, Tavares F, et al. The safety evaluation of adjuvants during
vaccine development: the AS04 experience. Vaccine 2011;29:4453–9.
 SECTION ONE: General aspects of vaccination
Vaccine additives and manufacturing
residuals in the United States: licensed
vaccines
In addition to one or more immunogens,* a vaccine may con-
tain any of several added substances—for example, an adjuvant
or a preservative. Residual components from the manufactur-
ing process, in varying amounts, are also present in the vaccine.
This chapter addresses the types and amounts of additives that
are present in vaccines, the rationale for their inclusion, and the
applicable federal regulations. Additionally, residual materials
from the manufacturing process that are present in the final for-
mulation of the vaccine, as well as relevant federal regulations
regarding these residuals, are discussed. Finally, albeit to a lim-
ited extent, several issues and concerns that currently pertain to
the use of, or presence of, some of these additives and residuals
are examined. This chapter focuses on vaccines licensed in the
United States; vaccines not licensed in the United States may
contain the same types of additives and residuals, although the
amounts that are present in any given vaccine may differ.
For the purposes of this chapter, the term additives refers to
materials that are added to the immunogen by the manufacturer
for a specific purpose. Additives include adjuvants, preservatives
(ie, antimicrobial agents), and stabilizers, as well as materials
that are added to affect pH and isotonicity. In addition to addi-
tives, vaccines contain residuals that remain from the licensed
manufacturing process. The final formulation—immunogen plus
additives and residuals—defines the specific vaccine; although
not all manufacturing residuals can be identified and quantified,
their presence and quantity is assumed to be constant because of
the constancy of the manufacturing process. Some information
regarding additives and residuals is considered to be a trade secret
and thus confidential, and cannot be discussed in this chapter.
Vaccine manufacturing includes in-process and release
tests, along with their respective specifications, for the allow-
able quantity of additives and certain residuals that may be
present in the vaccine. These tests and their accompanying
specifications are detailed in the product license; some of the
specifications may be provided in the vaccine's package insert.
A manufacturer may not remove, change, or adjust the quan-
tity of an additive, or change the manufacturing process, with-
out submitting a license supplement to the US Food and Drug
Administration (FDA) describing that change (along with the
data supporting the change), and without obtaining approval of
that change in the licensed vaccine. FDA regulations (in Title
21 of the US Code of Federal Regulations [21 CFR]) define the use
and labeling of preservatives and adjuvants, and these regulations
*An immunogen is a preparation consisting of all or a portion of a
disease-causing organism, or the nucleic acid that encodes one or more
of the proteins from that organism, or all or a portion of a human tissue,
and it is administered to an individual to induce an immune response to
the immunogen for the treatment or prevention of a disease or condition.
Theresa M. Finn
William Egan
6
are described, as appropriate, in this chapter. FDA biologics
regulations (21 CFR 610.61) address whether the use of, and
quantity of, additives and residuals should be disclosed on the
vaccine label. The regulations that concern us in this chapter
are stated as follows:
“The following shall appear on the label affixed to each pack-
age containing a product . . .
(e) The preservative used and its concentration . . .
(l) Known sensitizing substances, or reference to an enclosed
circular containing appropriate information;
(m) The type and calculated amount of antibiotics added
during manufacture;
(n) The inactive ingredients when a safety factor, or reference
to an enclosed circular containing appropriate information;
(o) The adjuvant, if present;
(p) The source of the product when a factor in safe
administration;
(q) The identity of each microorganism used in manufacture,
and, where applicable, the production medium and
method of inactivation . . .".
Vaccine additives
Preservatives
Preservatives are added to vaccine formulations to prevent the
growth of bacteria or fungi that may inadvertently be introduced
into the vaccine during use. In some cases, preservatives are
used during the manufacturing process (eg, in buffers and col-
umn washes) to prevent microbial growth. Improvements in
manufacturing technology, however, have decreased this need
for the addition of preservatives to control bioburden during the
manufacturing process. The CFR requires that, with certain
defined exceptions, or with the approval of the Center Director
(discussed later), preservatives must be added to multidose vials
of vaccines. In the past, tragic consequences followed the use
of multidose vials that did not contain a preservative, which
served, in part, as the impetus for this requirement (see Wilson1
for a discussion of incidents related to the lack of preservatives
in vaccines). Specifically, 21 CFR 610.15(a) states that “prod-
ucts in multiple-dose containers shall contain a preservative,
except that a preservative need not be added to Yellow Fever
Vaccine; Polio-virus Vaccine Live Oral; viral vaccines labeled for
use with the jet injector; dried vaccines when the accompanying
diluent contains a preservative; or to an Allergenic Product in
50 percent or more volume in volume (v/v) glycerin”.
72 SECTION ONE General aspects of vaccination
Although the regulation does not specify a quantity, it does
require that the preservative used “shall be sufficiently nontoxic
so that the amount present in the recommended dose of the
product will not be toxic to the recipient, and in the combina-
tion used it shall not denature the specific substances in the
product to result in a decrease below the minimum acceptable
potency within the dating period when stored at the recom-
mended temperature”.
The CFR does not, however, provide a definition of a preser-
vative. The definition (ie, antimicrobial effectiveness) that has
been used by the FDA for vaccines and other biologicals is found
in the US Pharmacopoeia (USP).2 This is a functional definition,
wherein the final formulation of the vaccine, including the pre-
servative, is challenged with specified quantities of the follow-
ing organisms:
and . The
test sample (preservative-containing vaccine plus the microor-
ganism) is incubated at 20° to 25° C, and the number of viable
microorganisms is determined on days 7, 14, and 28. A preser-
vative is deemed acceptable if the following are achieved:
t Bacteria: a reduction of not less than 1.0 log10 from
the initial count at 7 days, and not less than a 3.0 log10
reduction from the initial count after 14 days, and no
increase in the 14-day count at 28 days
t Yeasts and molds: remain at or below the level of the initial
inoculum on days 7, 14 and 28.
Note that the antimicrobial agent is not tested by itself; it is
only the final vaccine formulation that is tested.
Preservatives cannot completely eliminate the risk of bacte-
rial or fungal contamination of vaccines; moreover, they do not
address any potential viral contamination. Although it occurs
rarely, and not in the recent past, the scientific literature does
contain reports 3,3a (see also Wilson1) of bacterial contamination
of vaccines despite the presence of a preservative, emphasizing
the need for meticulous attention to technique when withdraw-
ing vaccines from multiuse vials. At present, only four preserva-
tives are used in US-licensed vaccines: phenol, benzethonium
chloride, 2-phenoxyethanol, and thimerosal (spelled
in some other countries). Recently, the FDA amended the bio-
logics regulations to permit exceptions or alternatives to the
regulation for constituent materials (21 CFR 610.15), which
includes preservatives and adjuvants. The following section
was added to the regulation: “(d) The Director of the Center for
Biologics Evaluation and Research or the Director of the Center
for Drug Evaluation and Research may approve an exception
or alternative to any requirement in this Section. Requests for
such exceptions must be made in writing”.
The amended regulation could, as an example, allow the use
of particular vial adaptors to prevent contamination of prod-
ucts in multidose vials without the use of preservatives. As
noted in the final rule,4 the Director of the Center for Biologics
Evaluation and Research or the Director of the Center for Drug
Evaluation and Research “would not approve an exception or
alternative when the data or conditions of use, including the
indication and patient population, do not provide a sufficient
scientific and regulatory basis for such an approval”. The
amended regulation took effect in May 2011.
As noted, a preservative “shall not denature the specific sub-
stances in the product to result in a decrease below the mini-
mum acceptable potency within the dating period when stored
at the recommended temperature”. Certain preservatives are not
compatible with certain antigens; compatibility must be estab-
lished. For example, it has been known for a number of years
that thimerosal has a deleterious effect on the potency of inac-
tivated poliovirus vaccine (IPV).5,6 An alternative preservative is
necessary for IPV. A preservative that is used in other products,
2-phenoxyethanol,7 has been found to be compatible with IPV
vaccine formulations; it is used as a preservative in both of
the currently US-licensed IPV vaccines (IPOL [Sanofi Pasteur]
and Poliovax [Sanofi Pasteur; not currently marketed in the
United States]).
Phenol is currently used in three US-licensed vaccines: the
polysaccharide vaccines Pneumovax 23 (a 23-valent pneumo-
coccal polysaccharide vaccine from Merck & Co) and Typhim Vi
( capsular polysaccharide vaccine from Sanofi
Pasteur), and ACAM2000 (the smallpox vaccine from Sanofi
Pasteur); each of these vaccines contains 0.25% phenol as a pre-
servative (phenol is contained in the diluent for ACAM2000).
According to the Minimum Requirements of the National
Institutes of Health (NIH),8,9 phenolic compounds (such as phe-
nol or the various creosols) are not permitted as preservatives
in diphtheria- and tetanus toxoid–containing products. This
requirement is also reflected in other regulations or require-
ments, such as those of the World Health Organization (WHO).10
It has been reported that phenol affects diphtheria toxoid, “so
that its immunizing power falls rapidly”.11 Benzethonium chlo-
ride is currently used in only one US-licensed vaccine, Anthrax
Vaccine Adsorbed (BioThrax; the preservative is 25 g/mL
benzethonium chloride and 100 g/mL formaldehyde), manu-
factured by EmergentBioDefense Operations Lansing, Inc.
In recent years, considerable controversy has surrounded the
use of thimerosal, an organomercurial, in vaccines. Although
allergic responses to thimerosal have been described,12 a more
recent controversy, arising in the late 1990s, centered on the
hypothesis that exposure to thimerosal, a derivative of ethyl
mercury, may be causally linked to autism and other neuro-
developmental disorders in children. Although there were no
clear or definitive data to support a link between thimero-
sal and neurodevelopmental disorders, the US Public Health
Service (PHS), first in July 199913 and again in June 2000,14
in an effort to reduce mercury exposure in children from all
sources, recommended that thimerosal be removed from pedi-
atric vaccines as expeditiously as possible. The July 1999 PHS
statement was issued jointly with the American Academy of
Pediatrics; the June 2000 PHS statement was issued jointly
with the American Academy of Pediatricians, the American
Academy of Family Physicians, and the Advisory Committee
on Immunization Practices (ACIP). Letters from the Center
for Biologics Evaluation and Research (CBER) of the FDA, in
199915 and again in 2000,16 to the various vaccine manufac-
turers noted that the removal of thimerosal from vaccines was
merited and requested manufacturers' timelines for thimerosal
removal or submission of an explanation as to why thimerosal
removal was not currently feasible.
In 2001, the Immunization Safety Review Committee of
the National Academy of Science's Institute of Medicine (IOM)
reviewed the issues surrounding thimerosal and vaccines and
concluded that the evidence was inadequate to accept or reject
a causal relationship between thimerosal exposure from child-
hood vaccines and the neurodevelopmental disorders of autism,
attention deficit hyperactivity disorder, and speech or language
delay.17 At that time, the committee's conclusion was based on
there having been no published epidemiologic studies examin-
ing the potential association between thimerosal-containing
vaccines and neurodevelopmental disorders. Nevertheless, the
Committee believed that the effort to remove thimerosal from
vaccines was “a prudent measure in support of the public health
goal to reduce mercury exposure of infants and children as
much as possible”. Furthermore, in this regard, the Committee
urged that “full consideration be given to removing thimerosal
from any biological product to which infants, children and preg-
nant women are exposed”.
In 2004, the IOM's Immunization Safety Review Committee
issued its final report, examining the hypothesis that vac-
cines, specifically measles-mumps-rubella (MMR) vaccines
and thimerosal-containing vaccines, are causally associated
with autism. In this report, the committee incorporated new
epidemiologic evidence from the United States, Denmark,
Sweden, and the United Kingdom, and studies of biologic
 Vaccine additives and manufacturing residuals in the United States: licensed vaccines 6 73

mechanisms related to vaccines and autism since its report
in 2001. The committee concluded that this body of evidence
favors rejection of a causal relationship between thimerosal-
containing vaccines and autism, and that the hypotheses that
were generated concerning a biological mechanism for such
causality were theoretical only.18 The European Agency for the
Evaluation of Medicinal Products (EMEA, now called European
Medicines Agency [EMA]) also noted, as a precautionary mea-
sure, “that, although there is no evidence of harm caused by
the level of exposure from vaccines, it would be prudent to pro-
mote the general use of vaccines without thiomersal and other
mercury-containing preservatives”.19 Of note, the WHO contin-
ues to recommend the use of vaccines containing thimerosal
because the need for multidose preservative-containing vac-
cines and, thus, the benefits of using such vaccines outweighs
the theoretical risk of toxicity,20 and the Global Advisory
Committee on Vaccine Safety has stated that they remain of the
view that there is no evidence supporting a causal association
between neurobehavioral disorders and thimerosal-containing
vaccines.21 A more comprehensive update on the thimerosal-
autism hypothesis for vaccines may be found in Chapter 76.
At present, with the exception of the inactivated influenza
vaccine (more on this later), all of the US-licensed, routinely rec-
ommended pediatric vaccines (hepatitis B, diphtheria–tetanus
toxoid–acellular pertussis [DTaP], type
b, IPV, pneumococcal conjugate, human papilloma virus [HPV],
rotavirus, MMR, and varicella) that are being manufactured are
either thimerosal free or contain only trace amounts ( 1 g of
mercury per dose) as a residual from the manufacturing pro-
cess (Table 6-1). The currently used varicella, MMR, IPV, and

Preservatives, Adjuvants, and Inactivation Residues Noted in Labels of Selected US-Licensed Vaccines

*Dose is 0.5 mL except where noted.

DTaP, diphtheria–tetanus toxoid–acellular pertussis; MPL, 3- -desacyl-4’-monophosphoryl lipid A; %NN, amount in final container not noted on label; NN, not noted
on label; Td, tetanus-diphtheria vaccine for adolescents and adults; Tdap, tetanus, diphtheria, and pertussis for adolescents and adults.
74 SECTION ONE
pneumococcal conjugate vaccines have always been thimerosal
free, as are the recently licensed MMR plus varicella (MMRV)
and rotavirus vaccines (ProQuad, and Rotateq, Merck; Rotarix,
GlaxoSmithKline Biologicals). Of the six US-licensed trivalent
inactivated influenza vaccines, four are approved for pediatric
use: Fluzone (Sanofi Pasteur), for use in infants 6 months of age
and older; Fluvirin (Novartis Vaccines and Diagnostics), for use
in children 4 years of age and older; Afluria (CSL, Ltd.), for use in
children 5 years of age and older; and Fluarix (GlaxoSmithKline
Biologicals), for use in children 3 years of age and older. These
four inactivated influenza vaccines are available in either
thimerosal-free presentations (Fluzone, Fluarix, and Afluria) or
with trace thimerosal ( 1 g of mercury per 0.5 mL dose) (Fluvirin).
FluMist (MedImmune Vaccines, Inc.) is a live influenza vaccine
and does not contain thimerosal.
Two Tdap (tetanus, diphtheria, and pertussis for adolescents
and adults) vaccines (tetanus, diphtheria, and acellular pertus-
sis vaccines, with the lower case letters indicating a reduced
antigen content for the diphtheria toxoid and one or more of
the pertussis antigens) for use in adolescents and adults, Adacel
(Sanofi Pasteur) for use in persons 11 through 64 years of age,
and Boostrix (GlaxoSmithKline) for use in persons 10 years of
age and older, are licensed in the United States. Neither product
contains thimerosal. The meningococcal conjugate vaccines,
Menactra (Sanofi Pasteur) and Menveo (Novartis Vaccines and
Diagnostics, Inc.) do not contain any preservative.
A pediatric presentation of the diphtheria and tetanus tox-
oid vaccine (DT) is available from Sanofi Pasteur in a presenta-
tion containing only a trace of thimerosal as a manufacturing
residual.
Adjuvants
Adjuvants are materials that enhance and direct the immune
response (see Chapter 5). Vaccine adjuvants are not licensed
separately; rather, the adjuvant is a constituent of the licensed
vaccine, and it is the vaccine formulation, in toto, that is tested
in clinical trials and is licensed. As a consequence, an adju-
vant cannot be added or removed, or its amount in a licensed
vaccine changed, without submitting a supplement to the vac-
cine license and obtaining approval from the FDA. The various
aluminum salts (aluminum hydroxide, aluminum phosphate,
alum [potassium aluminum sulfate], or mixed aluminum salts)
are the most commonly used adjuvants in US-licensed vaccines.
Recently, one vaccine, Cervarix (GlaxoSmithKline), which con-
tains AS04, an adjuvant system composed of an aluminum salt
and monophosphoryl lipid A, a detoxified lipopolysaccharide
(LPS), has been licensed.
Despite worldwide use of aluminum salts for more than
50 years, surprisingly little has been known about their
mechanism of action as adjuvants (see, eg, Chapter 5 and
HogenEsch22). For many years, the prevailing thought was that
the aluminum salts functioned as depots for the vaccine immu-
nogens. More recently, it was shown that the aluminum salts
also activate inflammasomes, clusters of proteins found inside
certain cells. Inflammasomes respond to stresses such as infec-
tion or injury by releasing cytokines, which, in turn, stimulate
an immune response.23
The specific aluminum salt (hydroxide, phosphate, sul-
fate, or mixed) and the quantity of aluminum that is con-
tained in a number of commonly used vaccines are presented
in Table 6-1. (The aluminum content that is listed for some
vaccines noted in Table 6-1 represents the upper limit of the
specification; the vaccine may routinely contain less alumi-
num.) Currently, live vaccines do not contain an adjuvant.
By regulation (21 CFR 610.15[a]), the aluminum content of
a vaccine cannot exceed 0.85 mg of aluminum per dose if the
amount is assayed, or 1.14 mg/dose if determined by calcula-
tion based on the amount of the aluminum compound that is
added. To harmonize with WHO recommendations, this regu-
lation was amended in 1981 to permit up to 1.25 mg of alu-
minum per dose. However, the higher amount was permitted
only “provided that data demonstrating that the amount of
aluminum used is safe and necessary to produce the intended
effect are submitted to and approved by the Director, Center
for Biologics Evaluation and Research” (21 CFR 610.15[a]).
Recently, the FDA amended the CFR requirements4 for alumi-
num salts, to permit, when justified, the use of a greater alu-
minum content in a vaccine; this change may have a greater
impact on certain therapeutic vaccines than on the preventive
vaccines.
Concerns have been raised in recent years about the use of alu-
minum in vaccines and potential adverse outcomes that may be
associated with its use at the levels that exist in individual vac-
cines and through the additive effects of multiple vaccinations.
These concerns about the use of aluminum in vaccines prompted
a workshop that was sponsored by the National Vaccine Program
Office in May 2000. The general use of aluminum salts in vac-
cines24 and aluminum toxicokinetics25 were reviewed during the
workshop. In their overall summary of the workshop, Eickhoff and
Meyers26 noted that “based on 70 years of experience, the use of
salts of aluminum as adjuvants in vaccines has proven safe and
effective”. A recent study by FDA scientists, using updated param-
eters, including current recommended vaccines and aluminum
excretion data, concluded that the risk from aluminum exposure
from vaccines and the environment to infants was extremely low.26a
Stabilizers
Various stabilizers are added to vaccines to help protect the vac-
cine from adverse conditions such as the freeze-drying process
(for those vaccines that are freeze-dried) or heat. For freeze-dried
(lyophilized) preparations of vaccines, it is also necessary to add
materials that provide a bulk matrix for the vaccine. The amount
of an immunogen that is contained in a vaccine can be extremely
small, on the order of tens of micrograms or less. If sufficient
amounts of various materials were not added to the vaccine before
lyophilization, the vaccine would not be readily observable and
would undoubtedly adhere to the vial wall. As an illustration of
this latter point, ActHIB (Sanofi Pasteur), a polysaccharide con-
jugate vaccine, contains approximately 10 g of purified polysac-
charide conjugated to 24 g of tetanus toxoid. The viral mass for
the live viral vaccines is even smaller, on the order of nanograms
(about 103 to 105 viral particles per dose). Thus there is a need to
provide a matrix to contain these vaccines during freeze-drying.
The types of material that are added to vaccines as stabilizers
include sugars (such as sucrose and lactose), amino acids (such
as glycine or the monosodium salt of glutamic acid), and pro-
teins (such as gelatin). The stabilizers that are used for a num-
ber of common vaccines are listed in Table 6-2.
Added proteins are of concern for two main reasons. The first
concern arises from the potential for animal- and human-derived
protein to contain one or more adventitious agents. The second
concern arises from the potential for animal- or human-derived
protein to elicit an allergic reaction in susceptible individuals. The
two animal- or human-derived proteins that are used as stabilizers
in US-licensed vaccines are human serum albumin (HSA) and gela-
tin. The FDA has, to date, required that, if blood-derived HSA is
used in vaccine manufacture, only US-licensed HSA may be used.
Additionally, an FDA guidance recommends that the following state-
ment appear in the Warnings section of the package insert for blood-
derived HSA-containing products27: “This product contains albumin,
a derivative of human blood. Based on effective donor screening and
product manufacturing processes, it carries an extremely remote risk
for transmission of viral diseases. Although there is a theoretical
risk for transmission of Creutzfeldt-Jakob disease (CJD), no cases of
transmission of CJD or viral disease have ever been identified that
were associated with the use of albumin”.
 Vaccine additives and manufacturing residuals in the United States: licensed vaccines 6 75

Vaccine Stabilizers, Manufacturing Residuals, and Cell Lines Noted in Labels of Selected US-Licensed Vaccines

* Dose is 0.5 mL except where noted.

BSA, bovine serum albumin; FBS, fetal bovine serum; MSG, monosodium glutamate; %NN, amount in final container not noted on label; NN, not noted on label;
rHSA, recombinant human serum albumin.
76 SECTION ONE General aspects of vaccination
For HSA produced as a recombinant DNA protein, this warn-
ing would not be required (because it would not be derived from
blood), nor would it be necessary for the albumin to be sepa-
rately licensed by the FDA. In August 2005, the FDA approved
a supplement for use of recombinant HSA in MMR-II (Merck).
Gelatin or processed gelatin is also used as a stabilizer.
Gelatin may be bovine or porcine derived. Despite the use of
a harsh manufacturing procedure (extremes of heat and pH)
in the production of gelatin, there is concern about the pres-
ence of the bovine spongiform encephalopathy (BSE) agent in
bovine-derived material. Thus, any bovine-derived gelatin that
is added to vaccines, or used in the vaccine manufacturing pro-
cess, must not be sourced from a country in which BSE exists or
that presents an undue risk for BSE (see “Transmissible spongi-
form encephalopathy agents”, later).28 A second concern for gel-
atin relates to allergic responses. Allergic responses to gelatin,
although rare, have been described in the literature.29–31 It has
been hypothesized that in Japan, use of partially hydrolyzed gel-
atin, which contained a small amount of high-molecular-weight
gelatin, contributed to an increase in the incidence of allergic
reactions.31,32 Nakayama and Aizawa noted that a change to
hydrolyzed modified porcine gelatin, together with discontinu-
ation of the use of gelatin-containing DTaP vaccines, may have
contributed to a decrease in the incidence of allergic reactions
after administration of monovalent measles and mumps vac-
cine in Japan.32 An allergic response to gelatin is a contraindica-
tion to receiving gelatin-containing vaccines.
Various buffers (eg, phosphate buffer) are also used in vac-
cines to maintain a particular pH range, and salts (eg, NaCl)
may be added to achieve isotonicity.
Manufacturing residuals
In principle, any or all of the materials that are used in the
manufacturing process may be present in the final vaccine for-
mulation. For the purposes of this chapter, materials that are
present in the final vaccine formulation that derive from the
manufacturing process are termed residuals. Various steps in
the manufacturing process may remove or reduce the amounts
of many of these residuals. However, for various vaccines, an
acceptable technology to remove these manufacturing residu-
als may not exist, or there may be no perceived need (eg, with
regard to safety or a potential for an adverse effect on efficacy)
for their removal. As a general principle, it is not possible to
remove a particular substance completely, nor is it possible to
demonstrate that a particular substance has been completely
removed. For many substances, the residual amount may be
below the limits of detection by current analytic technologies,
and may, for practical reasons, be considered absent.
Bacterial and viral inactivation substances must be noted
in the package label (21 CFR 610.61[q]). Residual bacterial or
cellular culture components, such as antibiotics that are used
during manufacture, as well as sensitizing substances (gener-
ally proteins), and other inactive ingredients when considered a
safety factor, also must be noted in the label (21 CFR 610.61 [l]
[m][n]). There may be some overlap between these categories;
however, they are grouped in this manner for convenience and
to aid a discussion of these materials as they are affected by cur-
rent regulations. Residual bacterial or cell culture components
may be included in these categories, but other residuals, such as
DNA and endotoxin, are not generally noted in labeling.
Inactivation residuals
Various agents may be used to inactivate bacteria and viruses
or to detoxify bacterial toxins. The goal of these chemical treat-
ments is to inactivate the bacterium or virus or to remove toxic
activity while still preserving the antigenicity of the product
against the homologous organism or toxin. After inactivation,
a virus or bacterium may be processed further to furnish par-
ticular antigens. For example, after inactivation, the influenza
viruses may be split by various chemical treatments (eg, deter-
gent) so that more purified vaccines can be produced.
Formaldehyde has a long and extensive history of use in the
preparation of bacterial and viral vaccines. It was used by Ramon
in 1923 to detoxify diphtheria, yielding a diphtheria toxoid vac-
cine termed an anatoxine.33 Requirements for the use of formalde-
hyde and the permitted residual amount of formaldehyde that is
allowed for diphtheria toxoid are provided in the NIH Minimum
Requirements.8 Similar NIH Minimum Requirements exist for
tetanus toxoid.9 These documents note that residual, free form-
aldehyde in the finished product should not be in excess of 0.02%
(ie, 0.1 mg for a 0.5-mL vaccine dose). Formaldehyde is also used
to inactivate viruses (eg, the polio and influenza viruses) when
preparing vaccines. The amount of residual formaldehyde that
is present or allowed in these and various other US-licensed vac-
cines is provided in Table 6-1; in the diphtheria and tetanus tox-
oids, the amount does not exceed 0.02%.
Concern about the presence of residual formaldehyde in
vaccines stems from the known toxic effects of formaldehyde,
and from its carcinogenicity potential. The US Environmental
Protection Agency (EPA) has established a reference dose (RfD)
for formaldehyde through the oral route.34 The RfD is defined
by the EPA as “an estimate (with uncertainty spanning perhaps
an order of magnitude) of a continuous inhalation exposure or
a daily exposure to the human population (including sensitive
subgroups) that is likely to be without an appreciable risk of dele-
terious non-cancer effects during a lifetime”.35 The RfD for form-
aldehyde (oral administration) is 0.2 mg/kg of bodyweight per
day.34 The amount of residual formaldehyde in vaccines (which
are administered infrequently, not daily) is below this level.
Formaldehyde has been further classified by the EPA as a
“probable human carcinogen” (the EPA's B1 classification).34
The bulk of the carcinogenicity studies on formaldehyde have
focused on chronic respiratory exposure because this is the
primary route of industrial and routine household exposure.
There are fewer data regarding ingested or parenteral exposure
to formaldehyde, and the EPA has not developed a risk estimate
for either oral or parenteral exposure.34 Data regarding carcino-
genicity studies may be found in documents from the EPA34 and
the International Agency for Research on Cancer.36
One point regarding formaldehyde should be made.
Formaldehyde is naturally present in the human body as the
result of various biochemical processes.36,37 The steady-state
level of formaldehyde in the bloodstream of humans is approxi-
mately 2.6 mg/L.38 The amount of formaldehyde that is natu-
rally, continuously present in the blood of humans, or turned
over in a particular day, is in excess of the amount that is pres-
ent as a vaccine residual.
Inactivating agents other than formaldehyde are used in var-
ious US-licensed vaccines and include glutaraldehyde, which
is used to inactivate pertussis toxin (PT) in seven acellular
pertussis-containing vaccines, Adacel, Boostrix, Daptacel,
Infanrix, Kinrix, Pediarix, and Pentacel, and -propiolac-
tone, which is used in the inactivation of two influenza virus
vaccines, Afluria and Fluvirin, and two rabies vaccines, RabAvert
(Novartis) and Imovax (Sanofi Pasteur). Hydrogen peroxide was
used to inactivate PT in a previously licensed DTaP vaccine,
Certiva (North American Vaccines).
Residual cell culture materials
The CFR permits the addition of antibiotics (with the excep-
tion of penicillin) in viral vaccine manufacture (21 CFR
610.15[c]). Antibiotics that have been used include streptomycin,
 Vaccine additives and manufacturing residuals in the United States: licensed vaccines
polymyxin B, neomycin, and gentamicin. Although a manufac-
turer need not specifically test for these antibiotics in the final
container, the calculated amount of residual antibiotics (based on
dilutions of the amount that was added) must be noted on the
package label (21 CFR 610.61[m]). The amount of residual anti-
biotics in several US-licensed vaccines can be found in Table 6-2.
Although the amended regulation for Constituent Materials4 (21
CFR 610.15) applies equally to antibiotics, no examples of poten-
tial changes for the antibiotic regulation, unlike for preservatives
and adjuvants, were presented in the Federal Register notice.
Sensitizing substances
The CFR's biologics labeling regulations (21 CFR 610.61[l])
provide that "known sensitizing substances" should be listed on
the product label. Furthermore, 21 CFR 610.15(b) states that
“extraneous protein known to be capable of producing allergenic
effects in human subjects shall not be added to a final virus
medium of cell culture produced vaccines intended for injec-
tion”. These regulations address the possibility that animal-
derived proteins present in the final formulation of a vaccine
can cause allergic reactions in some vaccine recipients. (Other
sensitizing substances, such as preservatives and stabilizers,
are addressed in other sections of these regulations.) Animal-
derived materials are used extensively in vaccine manufac-
turing, particularly in viral cultures. When viral vaccines are
grown in embryonated chicken eggs (influenza and yellow fever
vaccines) or chick embryo fibroblast cell culture (measles or
mumps virus vaccines), the label will state that residual chicken
proteins may be present in the final formulation (see Table 6-2).
Although hypersensitivity to any component of the vaccine is a
contraindication, the MMR-II and yellow fever vaccine (YF-Vax)
package inserts address the vaccination of persons with hyper-
sensitivity to eggs or egg protein.
The two US-licensed hepatitis B vaccines, Engerix-B
(GlaxoSmithKline) and Recombivax HB (Merck & Co.), are
recombinant DNA–derived proteins and are produced in yeast;
their package inserts note that residual yeast protein may be
present (Engerix-B contains not more than 5% and Recombivax
HB not more than 1% yeast protein). Hepatitis B vaccine is
contraindicated in persons with a history of hypersensitivity to
yeast; however, ACIP has noted that “no evidence exists that
documents adverse reactions after vaccination of persons with
a history of yeast allergy”.39 The human papillomavirus vac-
cine (HPV), Gardasil (Merck), is also manufactured in yeast
cells; according to the package insert, the vaccine contains less
than 7 g of yeast proteins per dose; hypersensitivity to yeast is
given as a contraindication to receiving the vaccine. The other
HPV, Cervarix (GlaxoSmithKline Biologicals), is manufactured
in insect cells; levels of insect cell proteins are reduced to less
than 40 ng/dose.
It has recently been reported40 that anaphylactic reactions
to DTaP and Tdap vaccines may be caused by the presence of
residual amounts of casein, a milk protein that is used in the
culture media for these vaccines; the reactions occurred in chil-
dren who already had high levels of milk allergy. The authors40
note, however, that these anaphylactic reactions are exceedingly
rare, and that many highly sensitive children with milk allergy
tolerate the vaccines. Further studies will be necessary to define
whether there is a subpopulation of children with milk allergies
who might be at risk for an anaphylactic reaction.
The Food, Drug and Cosmetic Act (Section 502[e][1][A][iii])
states that all inactive ingredients should be noted in labeling; it
also states that this requirement is not necessary if trade secret
information would be disclosed. The CFR additionally notes
that an inactive ingredient should be listed in the labeling if
the ingredient's presence is considered a safety factor (21 CFR
610.61[n]). In some cases, even in the absence of any evidence
6 77
that a particular material might pose a safety factor, manufac-
turers have elected to disclose the presence of residual mate-
rials such as detergents, solvents, and chelating agents (see
Table 6-2 for examples of manufacturing residuals). In addition,
many manufacturers provide a brief summary of manufactur-
ing methods, including the reagents used in various steps, such
as precipitation (ammonium sulfate) or bacterial culture (eg, an
antifoam agent such as polydimethylsiloxane). Many of these
substances are removed or markedly reduced in subsequent
manufacturing steps.
Bacterial and cellular residuals
For bacterial vaccines, manufacturing residuals may include vari-
ous bacterial cell constituents. Naturally, whole-cell vaccines—
such as the previously used whole-cell pertussis vaccine—contain
high levels of these components. At present, no parenteral whole-
cell bacterial vaccine is in use in the United States; an oral, live
attenuated bacterial vaccine, Typhoid Vaccine Live Oral Ty21a
(Vivotif, Berna Biotech Ltd.), is in use.
Vaccines derived from gram-negative bacteria may contain
lipopolysaccharide, commonly termed endotoxin, a component
of the bacterium's outer membrane. Stimulation of the innate
immune system by LPS can produce an inflammatory response
that can result in fever, shock, and death.41 Different organ-
isms have different LPS structures, so their respective response
potentials differ. Two tests are currently used to detect LPS in
vaccines, the limulus amebocyte lysate (LAL) test and the rab-
bit pyrogen test. The LAL test is more commonly used. The
lysate from the amebocytes of the horseshoe crab,
clots in the presence of LPS and forms the basis for
this test.42,43 The limulus lysate that is used to test for bacterial
endotoxin in US-licensed vaccines (and other FDA-regulated
biological products) is itself a US-licensed product. Before the
development and acceptance of the LAL test, manufacturers
performed the rabbit pyrogenicity test, which is still an accept-
able test for endotoxin. The amount of endotoxin remaining
in a final vaccine formulation depends on a number of factors,
including the purification steps that are used in vaccine pro-
duction. Although endotoxin testing of antigens derived from
gram-negative bacteria is performed during the manufacturing
process, and there may be a release specification for this test,
the labeling may not contain this information. One of the DTaP
vaccine labels (Tripedia [Sanofi Pasteur]) includes the amount of
endotoxin contributed by the inactivated pertussis components
(< 50 endotoxin units/mL).
Residual bacterial protein may be present in the final vac-
cine formulation of bacterial vaccines. The consequences of
this residual protein can vary, and its presence may be neu-
tral or harmful. For example, it has been recognized for many
years that the presence of residual protein may contribute to
increased reactogenicity of diphtheria toxoid.44,45 However, it
has also been held that any such protein contributed to the
immunogenicity of the vaccine.46
Polysaccharide, conjugated polysaccharide, and purified pro-
tein vaccines undergo a number of purification steps that reduce
the amount of residual bacterial protein. However, these puri-
fication steps may not totally eliminate cellular or media pro-
tein components. During development of the product, a number
of assessments of purity are performed, such as silver staining
or polyacrylamide gel electrophoresis (PAGE) gel immunoblot-
ting. After licensure, purity and quality of the vaccine antigen is
often assessed by sodium dodecyl sulfate–PAGE as a release test.
However, there is a limit to the sensitivity of these methods, as
is illustrated in publications of the National Institute of Allergy
and Infectious Disease–sponsored multicenter acellular pertussis
trial,47,48 which show that some children vaccinated with a fourth
and fifth dose of Tripedia (DTaP; a two-component pertussis vac-
cine containing PT and filamentous hemagglutinin [FHA]), after
78 SECTION ONE
previous doses of Tripedia, or a fourth dose of Tripedia after a
primary series with whole-cell pertussis vaccine, had a booster
response to pertactin and fimbriae, suggesting that there was suf-
ficient antigen in Tripedia to stimulate an immune response.
Cell substrates used in viral vaccine manufacture include
two diploid cell strains of human origin (MRC-5 and WI-38);
a simian-derived, continuous cell line (Vero cells); a simian-
derived diploid cell strain (FRhL cells); and chick embryo and
chick embryo fibroblasts. Residual protein from these cell lines
is present to varying degrees in the vaccines produced from
them. There is a particular concern, as noted previously, for
residual egg proteins in sensitive individuals. The US-licensed
Japanese encephalitis vaccine, JE-Vax (distributed by Sanofi
Pasteur and no longer available), is produced in mouse brains;
the package insert notes that there is less than 2 ng/mL (the
limit of detection of the assay) of myelin basic protein.
Residual cellular DNA from primary and diploid cells, as well
as from bacterial cells, may be present in the final vaccine formu-
lation, and this DNA is not considered to pose any risk. Residual
DNA from continuous cell lines, such as Vero cells, has been con-
sidered by a WHO study group.49 The WHO Expert Committee on
Biological Standardization assessed the risk of a transformational
event as negligible and concluded that levels of up to 10 ng/dose
of injected product are acceptable.48 This limit was a revision of
an earlier, more conservative limit ( 100 pg per parenteral dose)
proposed by the 1986 WHO Expert Committee.50 In revising this
limit, the 1997 Expert Committee considered data from human
and animal experience. This included data from nonhuman pri-
mates showing that milligram amounts of DNA containing an
activated oncogene from human tumor cells had not caused a
tumor during 10 years of evaluation; consideration that human
blood contains substantial amounts of DNA in plasma; and
consideration that contaminating DNA in a biological product
would probably be in small fragments that are unlikely to encode
a functional gene.49 The committee concluded that continuous-
cell-line DNA could be considered a contaminant rather than a
significant risk factor requiring removal to extremely low levels,
hence the revised limit of 10 ng/dose.
Manufacturers do not necessarily need to demonstrate that
each lot meets this specification through specific testing on
each lot; they may be able to validate that the purification pro-
cess can remove DNA to this level or below. This limit of 10 ng
of residual DNA per dose does not apply to products derived
from microorganisms, diploid cell strains, or primary animal
cells, or to oral vaccines made with continuous cell lines.49 The
US-licensed Japanese encephalitis vaccine Ixiaro (Intercell), IPV
vaccine (Ipol, Sanofi Pasteur), and the IPV component of the
DTaP-HepB-IPV vaccine, Pediarix (GlaxoSmithKline) are pro-
duced in the continuous Vero cell line. Ipol contains less than
10 pg of DNA per dose.51
Adventitious agents
Use of animal-derived materials, such as gelatin, fetal bovine
serum (commonly referred to as fetal calf serum), or primary
animal-derived cells, in vaccine manufacture raises concerns
about the potential presence of adventitious contaminants.
Regulations (21 CFR 610.18) require that cultures used in the
manufacture of products be free from extraneous organisms,
and that cell lines should be tested for the presence of detect-
able microbial agents. A 1993 FDA guidance document52 stated
that master cell banks should be tested for adventitious agents
and that animal-derived materials should be free from contami-
nants and adventitious agents, including viruses and the agent
of BSE. A final guidance was published in February 2010.53
Manufacturers are required to perform such testing as nec-
essary, and to ensure that the certification provided with any
raw material is adequate (eg, documentation that bovine-derived
gelatin is not sourced from a country where BSE exists or which
presents an undue risk for BSE28). Adventitious agent testing, per-
formed to ensure that cell substrates used in vaccine manufacture
are free from bacteria, fungi, mycoplasma, and mycobacteria, is
described in detail in US52,53 and international54,55 guidance docu-
ments. The use of polymerase chain reaction–based reverse tran-
scriptase assays to test for adventitious retroviruses is addressed
in a CBER letter to manufacturers.56 The testing that has been
carried out for adventitious agents is not described in the product
labeling. However, the manufacturing method, including the cell
lines and culture media used, are described.
Transmissible spongiform encephalopathy agents
The lack of sensitive, specific, or readily accessible pre-mortem
diagnostic tests for transmissible spongiform encephalopathies
(TSEs), such as BSE or CJD, limits surveillance and the ability to
test for contamination of products for these agents. Diagnosis of
a TSE disease is definitively confirmed by postmortem examina-
tion of brain tissue. However, even this examination is limited
in sensitivity because, at earlier stages of disease, the TSE agent
may be present in undetectably low concentrations. Because
of these limits on testing, potential contamination of animal-
derived materials with TSE agents is controlled through restricted
sourcing. Manufacturers of US-licensed vaccines are required to
ensure that bovine-derived materials are sourced from a country
that is not on the US Department of Agriculture (USDA) list of
countries where BSE exists or that present an undue risk for BSE.
The list of such countries is maintained by the USDA (see 9 CFR
94.18) and is updated as necessary.57 A proposed rule, published
in January 2007, harmonizes requirements for bovine materials
for pharmaceutical products with those of USDA-regulated meat
and FDA-regulated foods and animal feeds.58
In 2000, it was discovered that some manufacturers of
US-licensed vaccines had not followed FDA recommenda-
tions with regard to the sourcing of bovine materials from
BSE-free countries. A joint session of the FDA's Transmissible
Spongiform Encephalopathy Advisory Committee and the
Vaccines and Related Biological Products Advisory Committee
recommended that, in the future, bovine-derived materials
used to make working cell and seed banks and in routine pro-
duction be replaced with material from the countries not on
the USDA list. The committees determined that the risk of
contamination of the existing master cell or seed banks was so
remote that they did not warrant re-derivation. The advisory
committees also recommended that the public be informed of
affected vaccines. All manufacturers have complied with the
committees' recommendations and no longer are any affected
vaccines identified on the CBER website.59 Since the joint
Advisory Committees meeting in 2000, several cases of BSE
have been found in Canada and the United States; the FDA has
not required that manufacturers find a new source of bovine-
derived materials obtained from these countries for use in
manufacture of vaccines.
Summary
A final vaccine contains materials in addition to the active
immunogen. Some of these materials are added by the man-
ufacturer to effect a specific purpose—for example, stabilizers
and adjuvants. Others are residual materials from the manu-
facturing process. Although not all of the results of all final
release and in-process testing are contained in the labeling that
accompanies a product, the CFR does specify which informa-
tion should be included. Package inserts also contain informa-
tion on manufacturing methods and growth conditions.
All material in this chapter is in the public domain.
 Vaccine additives and manufacturing residuals in the United States: licensed vaccines
Access the complete reference list online at http://www.expertconsult.com
1. Wilson GS. The Hazards of Immunization. New York: Athlone Press; 1967.
4. Food and Drug Administration. Revision of the requirements for constituent
materials: final rule. Fed Regist 2011;76:20513–8.
6. Sawyer LA, McInnis J, Patel A, et al. Deleterious effect of thimerosal on the
potency of inactivated poliovirus vaccine. Vaccine 1994;12:851–6.
8. NIH Minimum Requirements: diphtheria toxoid. 4th rev. Bethesda, MD:
National Institutes of Health; March 1, 1947.
9. NIH Minimum Requirements: tetanus toxoid. 4th rev. Bethesda, MD:
National Institutes of Health; December 15, 1952.
24. Baylor NW, Egan W, Richman P. Aluminum salts in vaccines: US
perspective. Vaccine 2002;20:S18–23 (corrigendum: Vaccine 2002;20:3428).
25. Keith LS, Jones DE, Chou CHJ. Aluminum toxicokinetics regarding infant
diet and vaccinations. Vaccine 2002;20(Suppl.):S13–7.
6 79
43. TenCate JW, Bueller HR, Sturk A, et al., editors. Bacterial Endotoxins:
Structure, Biomedical Significance, and Detection with the Limulus
Amebocyte Lysate Test. New York: Alan R. Liss; 1985.
53. US Food and Drug Administration. Guidance for industry: characterization and
qualification of cell substrates and other biological starting materials used in
the production of viral vaccines for infectious disease indications, www.fda.gov/
downloads/BiologicsBloodVaccines/GuidanceComplianceRegulatoryInformation/
Guidances/Vaccines/UCM202439.pdf; 2010.
57. US Department of Agriculture, Animal and Plant Health Inspection Service.
Animal and animal product import, www.aphis.usda.gov/import_export/
animals/animal_import/animal_imports_bse.shtml.
59. Center for Biologics Evaluation and Research. Bovine spongiform
encephalopathy (BSE), www.fda.gov/cber/bse/bse.htm.
 SECTION ONE: General aspects of vaccination
Passive immunization
7 E. Richard Stiehm
Margaret A. Keller
Passive immunity (e.g., antibody therapy) has been used for a
century for the prevention and treatment of infectious diseases.
In bacterial disease, antibodies neutralize toxins, facilitate opso-
nization, and, with complement, promote bacteriolysis; in viral
disease, antibodies block viral entry into uninfected cells, promote
antibody-dependent cell-mediated cytotoxicity by Fc-bearing cells
(e.g., natural killer cells, macrophages, others), and neutralize
virus alone or with the participation of complement.
Before the era of antibiotics, antibodies were the only specific
agents for the treatment of certain infections. Although this
role has been largely supplanted by antimicrobials, antibody
still has a crucial role in the treatment of certain infectious
diseases (Table 7-1). This chapter discusses the role of antibody
in prevention and treatment of bacterial and viral diseases.
Other reviews are available.1–4
Passive immunity—general considerations
Four types of preparations are used in passive immunity: (1)
human immune serum globulin (IG) for general use, available
in three forms, intramuscular (IGIM), subcutaneous (SCIG),
and intravenous (IGIV); (2) high-titered IGIMs and IGIVs with
a known antibody content for specific illnesses; (3) animal sera
and antitoxins; and (4) monoclonal antibodies. Those used in
infectious diseases are listed in Table 7-1. Only one therapeutic
monoclonal antibody, an antibody to respiratory syncytial virus
(palivizumab), is licensed for use in infectious disease; others are
under study. Whole blood, plasma, serum, breast milk, and even
animal colostrum have also been used in passive immunization.
Antibody treatment is not always effective, its duration is
short and variable (1-6 weeks), and undesirable reactions may
occur, especially if the product is not of human origin. The
special high-titered IGIMs and IGIVs are identical to regular
IGIM and IGIV except that the former are derived from patients
immunized or convalescing from a specific infection and the
antibody content to the specific antigen is assayed; thus, they
are useful in several disorders in which IGIM or IGIV is of little
or no value.
Side effects are not uncommon with antibody therapy use.
Reactions to equine antisera, such as diphtheria or botulism
antitoxins, are of special risk. Patients should be asked about
a history of allergic symptoms following exposure to horses or
injections of antisera. All patients require skin testing before
their use, using a scratch test of dilute antitoxin before admin-
istration. Positive (histamine) and negative (saline) controls
should be done and recent administration of antihistamines
noted since they may negate the skin test result. A negative
scratch test should be followed by an intradermal test.
If both skin tests are negative, the antitoxin can be adminis-
tered intramuscularly with careful observation for 30 minutes
at a locale where trained personnel and drugs are available to
treat an adverse effect. If either test is positive or if the patient
has had a prior reaction to animal exposure or antitoxin, desen-
sitization must be done by giving a series of gradually increas-
ing doses at short intervals under careful observation. Details
are given in the package inserts and in the American Academy
of Pediatrics Red Book on infectious diseases.5
Side effects of animal sera occur in 20% to 40% of patients,
including febrile reactions, urticaria, headaches, and muscle
pain. Serious reactions include anaphylaxis, serum sickness,
and, very rarely, peripheral neuritis. Equine antiserum adminis-
tration should always be done under close medical supervision.1,5
Human IGs, high-titered and regular immunoglobulins,
have replaced most animal antitoxins since they are much
less likely to cause adverse reactions and have a longer half-life
in the circulation. Human IGs are made from large pools of
plasma from healthy donors, while high-titered IGs (e.g., hepa-
titis B IG, tetanus IG) are derived from immunized donors or
donors with high titers to the desired antibody. Human IGs for
general use were initially 16% solutions for IM use, but now are
also available for IV or subcutaneous use. All of these products
have a wide range of antibody specificities at different titers, but
must contain minimal protective titers to polioviruses, hepati-
tis B surface antigen, diphtheria, and measles.
Adverse effects to human IGs may occur in up to 10% of
patients and include mild reactions such as headache, chills,
fever, and malaise; serious but rare side effects include throm-
bosis, renal insufficiency, aseptic meningitis, and anaphylaxis.
The use of subcutaneous immunoglobulin lessens the likeli-
hood of these reactions.6,7 Side reactions are more common in
patients receiving their first IG dose, patients receiving large
doses, and patients given the product over a short period.
The remainder of this chapter will discuss the role of passive
immunity in bacterial and viral disease, particularly disorders for
which active immunization is available or under development
(Table 7-2).
Bacterial diseases
Anthrax
Anthrax is a rare but serious infectious disease, predominantly
of ruminant animals, caused by an aerobic gram-positive rod
8
Humans can be infected through the skin
(cutaneous anthrax), by ingestion (gastrointestinal anthrax),

Antibody Preparations Available in the United States for Passive Immunity to Infectious Agents
*Many others are available for treatment of cancer, immune-related diseases, and transplant rejection.
or by inhaling the spores (inhalational anthrax). Inhalational
anthrax, the most serious form, results from exposure to ani-
mal hides or carcasses, infected soil, or, rarely, by deliberate
spore exposure in the bioterrorism setting, as occurred among
US postal workers in 2001.9
A vaccine is available for people at high risk for exposure and
for military personnel. It is effective against cutaneous anthrax
but of unproven efficacy against inhalational anthrax.10
Before the antibiotic era and as early as 1903, anthrax anti-
toxin (usually equine) was used in the therapy for anthrax.11
Thus, a human anthrax antitoxin would be of value in a bio-
terrorism attack, before and after exposure. A contract was
awarded to Cangene (Winnipeg, Canada) in 2006 to manufac-
ture human anthrax IG, derived from immunized donors (mili-
tary recruits and volunteers).12
Several monoclonal antibodies against anthrax antigens
have been produced. The best studied is a human monoclonal
antibody, raxibacumab (ABthrax).13 It is safe and well tolerated,
either IV or IM, in human volunteers; provides good antibody
titers; and has a half-life of 15 to 19 days.14 The US govern-
ment has contracted with Human Genome Sciences (Rockville,
MD) to provide 20,000 treatment doses of this monoclonal
antibody.12
No commercial antitoxin is available in the United States.
The only source of antitoxin is the limited supply of immune
Passive immunization 7 81
plasma stockpiled by the US Army. This should be used in
conjunction with antibiotics and with vaccine, the latter to
provide immunity after metabolism of the antibody.
Bacterial respiratory infections
Respiratory tract infections secondary to group A
, ,
, and mul-
tiple viruses are more frequent and severe in patients with pri-
mary and secondary antibody immunodeficiencies.4,6,15–17 These
infections can be markedly reduced by regular administration
of immunoglobulin. Furthermore, specific antisera to some of
these organisms were used in the early 1930s for the treatment
of severe infections (e.g., meningitis, epiglottitis) even after the
introduction of sulfonamides.18
Vaccines against and have dra-
matically decreased these infections in children. Vaccines given
to expectant mothers also provide transplacental antibody to
protect newborn children.19–21
Low-dose standard IGIM (e.g., 100-200 mg/kg/month) pro-
vides some protection to major infection (sepsis, meningitis,
pneumonia) but has little effect on decreasing the incidence
of otitis, sinusitis, or bronchitis, possibly because insufficient
antibody reaches mucosal surfaces.22,23
82 SECTION ONE General aspects of vaccination

Summary of the Efficacy of Antibody in the Prevention and Treatment of Infectious Diseases*

CMV: cytomegalovirus
EBV: Epstein-Barr virus
HIV: human immunodeficiency virus
HSV: herpes simplex virus
NR: not recommended
RSV: respiratory syncytial virus
VZV: varicella-zoster virus
*Modified from Stiehm and Keller.1
†
Recommended for immunodeficient patients.

Santosham et al24 prepared an experimental high-titered
human IG from donors immunized with pneumococcal, menin-
gococcal and b polysaccharide vaccines (termed
bacterial polysaccharide IG). This IG significantly reduced
invasive disease due to and when
given to Apache Indian infants. It also reduced the incidence of
pneumococcal otitis media but not the total number of otitis
episodes.25
In larger doses (400 mg/kg), IGIV reduced the frequency of otitis
due to pneumococcus in HIV-infected infants26,27 and otitis-prone
infants.28,29 Larger monthly doses (750 mg/kg) reduced the number
of viral respiratory infections in young infants.28 Thus, IGIV may
be beneficial in recurrent respiratory infections and chronic sinus-
itis not responding to prophylactic antibiotics, even in patients
with no apparent immunodeficient illness.29–31
In addition, IGIV has been used in the management of
refractory measles and adenoviral pneumonia.32,33
Botulism (Clostridium botulinum)
Botulism is a severe paralytic poisoning resulting from the
ingestion or absorption of neurotoxin or spores of .
Several variants are recognized: food poisoning from ingestion
of contaminated canned food, wound botulism from a contami-
nated soft tissue infection, inhalational botulism among people
working with the toxin or in a bioterrorist event, infantile botu-
lism (see next section), and adult-type infant botulism in adults
with preexisting gastrointestinal disease.34–36 In the latter two
types, ingested spores multiply in the gastrointestinal tract
to elaborate toxin; the absorbed toxin results in a severe and
prolonged paralytic disorder.
A few cases of botulism have been associated with use of
botulinum toxin for cosmetic use.37,38
A heptavalent equine antitoxin to types A, B, C, D, E, F, and
G toxins is available in the United States through the Centers
for Disease Control and Prevention (CDC).39 This antitoxin is
composed of more than 90% Fab and F(ab) 2 equine immuno-
globulin fragments. These fragments are cleared more rapidly
from the circulation than is intact IgG. Sensitivity testing must
be conducted before IV use. Additional doses may be needed
in severe wound or intestinal colonization in which toxin con-
tinues to be produced. Antitoxin can also be given prophylac-
tically to people known to have ingested contaminated food.
Early antitoxin therapy for wound botulism associated with
skin infections from subcutaneous injection of heroin (“skin
popping”) has been advocated.40 Equine antitoxin is not used
for infantile botulism.
Botulism, infantile (C. botulinum)
This severe paralytic disorder of infants results from the inges-
tion of spores in infant formulae or food (e.g.,
honey) resulting in the slow onset of constipation, abdomi-
nal bloating, poor feeding, and respiratory paralysis.41 Affected
infants must be hospitalized for tube feeding and respira-
tory support, often for 6 to 9 months. Human 5% botulinum
immune globulin (BIG, Baby BIG) for IV use is available for
treatment of infantile botulism.42 Despite its expense ($50,000
per vial), it is cost-effective because of the shortened hospital
stay needed.43
Clostridium difficile gastroenteritis
infection of the gastrointestinal tract usually results
in antibiotic-associated diarrhea, often with pseudomembra-
nous colitis and sometimes toxic megacolon.44 Toxic strains of
release two distinct toxins, both with potent cyto-
toxic and inflammatory properties.44 Infection generally leads
to an antibody response to the toxin, and high levels of these
Passive immunization 7 83
antibodies acquired after colonization may result in the asymp-
tomatic carrier state.45
In some patients with recurrent symptomatic infection,
many of whom are immunodeficient or immunosuppressed,
antibiotic-resistant diarrhea develops; many have low or absent
IgG antibodies to toxin A or B.45,46 The patients may respond
to IGIV, 300 to 500 mg/kg every 1 to 3 weeks.47–49 Such ther-
apy has been reported to increase antitoxin levels, control diar-
rhea, and prevent relapse in some patients.47–50 However Juang
et al51 reviewed IGIV use in 18 patients and 18 patients with
similarly severe disease not given IGIV and could find no differ-
ences in clinical outcome. Thus the routine use of IGIV is not
recommended.
Monoclonal antibodies to toxins A and B have
been studied.52,53 Lowy et al53 assessed the effectiveness of one
infusion of two human monoclonal antibodies (to toxins A and
B). They studied recurrences in 202 symptomatic patients with
diarrhea, following discontinuation of antibiotics; all
received antimicrobial treatment and half received the mono-
clonal antibody. Recurrence (defined by stool cultures) occurred
in 7% of the antibody-treated group and 25% of the placebo
group. The diarrhea was lessened in 28% in the antibody-
treated group and 50% in the control group. Serum antitoxin
levels were considerably higher in the antibody-treated group.
Oral passive antibody therapy is also under study using
immunoglobulin or milk whey protein from toxin–
immunized cows.54–56 Much of the latter retains its antibody
activity throughout the small intestine.56
Diphtheria
Diphtheria is caused by toxigenic strains of
and occasionally .
Diphtheria is exceedingly uncommon because of widespread
immunization, but an epidemic in the former USSR occurred
in the 1990s.57,58
Many of the adverse consequences of diphtheria result
from the toxin elaborated and absorbed from the diphtheritic
membrane. This toxin not only has a local effect of causing
membrane formation, but also allows its distribution via the
bloodstream to the heart, nervous system, kidney, and other
organs. Antitoxin neutralizes circulating toxin and competes
with and partially neutralizes loosely bound toxin; however,
it has no effect on tissue-bound toxin. Thus, optimal passive
immunity must be initiated at an early stage of the disease via
the IV route so that toxin can be neutralized before it becomes
tissue-bound.
Antitoxin of animal origin remains the mainstay of treat-
ment, as it was in the preantibiotic era. Diphtheria was the first
illness for which antiserum was used as standard therapy. Emil
von Behring was awarded the first Nobel Prize in Medicine (in
1901) for this achievement. Subsequent studies have demon-
strated antitoxin efficacy,59 the need for early administration,60
and the value of the IV route rather than the IM route.61,62
Antitoxin was used in the 1890 s at Boston Children's Hospital
to prevent new patients and nurses from becoming infected dur-
ing diphtheria epidemics.63
Human IG preparations have insufficient titers of diphtheria
antitoxin so they cannot be used instead of antitoxin.64 A high-
titered human IG is not available in the United States but has
been produced in the Ukraine.65
Equine diphtheria antitoxin is indicated for all suspected
or proven cases of diphtheria. It is available in vials contain-
ing 20,000 U from the CDC. Before administration, skin tests
must be performed to determine sensitivity. If the patient has a
history of serum reactions or these skin test results are positive,
desensitization must be done.5
The amount of antitoxin given depends on the location and
the extensiveness of the membrane, the degree of systemic
84 SECTION ONE General aspects of vaccination
toxicity, and the duration of illness. The preferred route is IV,
although the IM route has historically been used in milder cases.
In all cases, diphtheria antitoxin should be given promptly,
rather than awaiting bacterial confirmation of the diagnosis.
In pharyngeal or laryngeal disease of 48 hours' duration or
less, 20,000 to 40,000 U is given; in nasopharyngeal disease,
40,000 to 60,000 U is given; and in extensive disease with
neck edema or disease of more than 3 days' duration, 80,000
to 120,000 U is given.66 In cutaneous diphtheria, antitoxin is
of uncertain value. When used, the dose is 20,000 to 40,000 U.
Although antimicrobial therapy is a valuable aid in the treat-
ment of diphtheria, it is not a substitute for antitoxin therapy.
Routine use of antitoxin in an asymptomatic, exposed, sus-
ceptible patient is not routinely recommended. With heavy
exposure or an extremely susceptible host, antitoxin, 5000
to 10,000 U IM, can be given in addition to antibiotics and
diphtheria immunization; proof of efficacy is lacking.1
The world supply of diphtheria antitoxin is limited. There
are no US manufacturers.
Pertussis
Pertussis antiserum was used in the 1930s for the treatment
of pertussis.67 Human pertussis IG was developed in the 1960s
but was shown to have no additive benefit to antibiotics in the
treatment of pertussis.68–70
No agent is available for passive immunity to pertussis.
Serious bacterial infections
Some success has been achieved with IGIV in serious bacterial
infections associated with sepsis, postoperative care, burns, and
trauma.71 Controlled studies suggest but have not proven effi-
cacy.71 Pseudomonal and antisera and monoclo-
nal antibodies have yet to show efficacy.
Staphylococcal infection
Staphylococcal infections are ubiquitous and of varying sever-
ity, ranging from superficial skin infection to deep-seated cel-
lulitis, osteomyelitis, severe pneumonia, and overwhelming
toxic shock syndrome.72,73 Passive immunity with IGIV may be
beneficial in severe illness, especially toxic shock syndrome or
infection with antibiotic resistant organisms.
Staphylococcal toxic shock syndrome, often associated with
tampon use by menstruating women, has a rapid course lead-
ing to multisystem organ failure.74 This is generally associated
with strains that release toxic shock syndrome toxin (TSST-1),
a powerful superantigen causing a cytokine storm as in strepto-
coccal toxic shock syndrome. IGIV contains neutralizing anti-
bodies to this toxin75 and, in addition, downregulates cytokine
release and action and, thus, can be used for adjunctive therapy.
IGIV may also be valuable in the prevention of nosocomial
infection, particularly coagulase-negative staphylococcal infec-
tion in premature infants. An IGIV containing opsonizing anti-
body may be effective in prevention due to decreased transfer of
maternal antibody to very premature infants. One study showed
that IGIV could decrease the incidence of bacterial infection in
premature infants76; others have not.77,78 Two double-blind mul-
ticenter studies examined a hyperimmune human IGIV derived
from donors who received a vaccine or
had high-titer antibodies to staphylococcus; neither showed
clinical efficacy.79,80 Monoclonal antibodies are under study in
newborns.
A final use of IGIV is in the treatment of antibiotic-resistant
chronic staphylococcal infection along with antibiotics.
Waisbren81 treated patients with an antibiotic-IG combination,
with recovery in 13 of them. Plasma and immunoglobulin from
subjects immunized with staphylococcal toxoid are widely used
in Russia with apparent success.82
A major problem is the multiple virulence factors present
in the staphylococcus, thus limiting the effectiveness of any
hyperimmune or monoclonal antibodies.
Streptococcal infections
Antibody has a role in the prevention and treatment of group A
streptococcal infection as indicated by its rarity in newborns as
a result of transplacental antibody, the decrease of streptococ-
cal pharyngitis with increasing age, and the success of equine
antiserum in the treatment of erysipelas and scarlet fever in the
1920s and 1930s.83,84
Invasive group A streptococcal infections, including septice-
mia, necrotizing fasciitis/myositis, and toxic shock syndrome,
are of increasing severity and frequency.85,86 Streptococcal pyro-
genic exotoxins, including types A, B, and C, and mitogenic fac-
tor associated with certain strains are potent superantigens that
activate T cells to release multiple cytokines, with resultant
shock, fever, and organ failure. IGIV contains neutralizing anti-
body of varying titers to these superantigens.87,88 Thus, IGIV is
recommended, in addition to antibiotics in severe illness, not
only to neutralize exotoxins, but also to dampen cytokine pro-
duction and action. Controlled studies are unavailable, but case
reports and large series compared with historical controls sug-
gest its value.89 A high dose of IGIV is recommended (e.g., 1-2 g
as one dose or divided over several days).90
Tetanus
Antitoxin for the treatment of tetanus was introduced into
medicine by Behring and Kitasato in 1890; large doses (50-
100 mL) of serum from horses immunized with tetanus toxin
were used.91 Extensive studies have been done to determine the
optimal dose of antitoxin92 and the possible benefit of intrathe-
cal antitoxin, particularly in tetanus neonatorum, a common
problem in developing countries.93
The mechanism of action of tetanus antitoxin is to neutral-
ize toxin before its transport to the nervous system via the cir-
culation. Antitoxin also can neutralize toxin locally and prevent
its systemic absorption. Thus, antitoxin can be given locally, at
the site of toxin production, intravenously (in severe cases), and
intramuscularly (in less severe cases).
Since the 1960s, a human tetanus IG (TIG) has been avail-
able, Standards for TIG have been established.94 Treatment rec-
ommendations for dosing vary from 500 U to 3000 to 6000
U with use of TIG intramuscularly and for infiltration of the
wound, although efficacy has not been demonstrated for the
latter.95
TIG is also given to unimmunized or incompletely immu-
nized patients who sustain contaminated or deep puncture
wounds.95 The recommended dose of TIG is 250 IU intramus-
cularly, along with initiation of active immunization at a site
different from the TIG site.95 A dose of 250 U of TIG given
intramuscularly at a site different from that of the toxoid vac-
cine does not interfere with antibody response.95
Antitoxin and TIG have been used intrathecally in severe
cases and in tetanus neonatorum. A recent meta-analysis sug-
gests that intrathecal antibody is more beneficial than paren-
teral antibody. Kabura et al93 studied 842 patients in 12 trials;
8 trials showed a benefit of intrathecal therapy as defined by
mortality with an overall relative risk of 0.71 (95% confidence
interval, 0.62-0.81) compared with IM use. The benefit held
for all ages and doses. in the United States, TIG is not licensed
for this use.
When TIG is unavailable, human IGIV can be used; it con-
tains variable titers of tetanus antitoxin; a minimal dose of
200 to 400 mg/kg is suggested for tetanus prophylaxis96 and

treatment.95 Equine antitoxin is still used in some developing
countries without access to TIG. Skin testing for sensitivity
must be done before its use. Equine tetanus antitoxin is also
available for veterinary use.
A further discussion of passive immunity is given in the
chapter on tetanus toxoid vaccine (Chapter 33).
Viral infections
Argentine hemorrhagic fever
Argentine hemorrhagic fever (AHF) is caused by the Junin
virus, an arenavirus endemic in rodents of the pampas areas of
Argentina. Maiztegui et al97 showed that early treatment (before
the ninth day of disease) with immune plasma reduced mortal-
ity to 1.1%, compared with 16.5% for patients given normal
plasma. Enria et al98 reported that the live virus AHF vaccine
reduces the incidence of AHF, which has resulted in fewer early
diagnoses in nonvaccinated subjects when immune plasma
is maximally effective. The result is increased mortality in
patients who acquire AHF.
Ribavirin has been used to treat a small number of AHF cases
with late disease, but the mortality remains high, 28.5%99; its
effect in early disease is not known. A protocol combining rib-
avirin with immune plasma was not completed because of a
ribavirin shortage.98 IGIV or a monoclonal antibody with high
neutralizing titers is not available, and supplies of high-titered
immune plasma are limited.98
Cytomegalovirus
Cytomegalovirus antibody-enriched human IG (CMVIG) has
been used for prophylaxis in solid organ transplantation, treat-
ment of CMV pneumonia in immunocompromised patients,
and prevention of symptomatic CMV of infants born to preg-
nant women who contracted CMV infection in pregnancy.
Standard IGIV preparations also have CMV antibodies, usually
of lower titers than CMVIG. The use of CMVIG or IGIV is no
longer recommended for CMV prophylaxis in human stem cell
transplantation.100,101
For solid organ transplantation in high-risk situations, such as
a CMV-seropositive donor for a CMV-negative recipient (D+/R ),
CMVIG is used to prevent CMV infection, in addition to anti-
viral treatment. In a meta-analysis of 11 randomized trials of
CMVIG to prevent CMV infection and disease in solid organ
transplants, a benefit was shown in survival but not CMV
infection or rejection.102 These studies were done from 1987
to 2003, and included different immunoglobulin preparations,
different immunosuppression treatment, different antiviral
prophylaxis regimens, and different diagnostic criteria.
International consensus, recognizing the increased risk of
CMV infection in the D+/R situation, allows for the optional
use in heart transplantation, lung transplantation, and intesti-
nal transplantation. However, only limited data support its use
when appropriate antiviral therapy is given.103 In a recent sur-
vey of management practices in lung transplantation, adjunc-
tive CMVIG was used by 32% of centers for D+/R patients,
14% for D+/R+ patients, and 7% for D /R patients.104
Because of the high mortality despite antiviral therapy,
CMVIG is often used in severe CMV disease, particularly in
CMV pneumonitis in transplant recipients. Consensus guide-
lines103 recommend CMVIG for children who received solid
organ transplants and have CMV pneumonitis, CMV enteritis,
or hypogammaglobulinemia. Boeckh and Ljungman100 also use
IGIV or CMVIG in hematopoietic cell transplant recipients with
CMV pneumonitis. The American Society of Transplantation
Infectious Diseases Community of Practice states that IGIV or
Passive immunization 7 85
CMVIG may be considered in severe CMV disease and CMV
pneumonitis.105
Nigro et al106 studied the value of CMVIG in preventing
symptomatic CMV disease in infants of women who had pri-
mary CMV infection in pregnancy. CMVIG was given during
pregnancy to 31 women with primary CMV acquired more than
6 weeks before enrollment and CMV-positive amniotic fluid,
compared with 14 women who did not receive CMVIG. Of the
treated women, 9 also received CMVIG into the amniotic sac or
umbilical cord because of ultrasound evidence of fetal infection.
Of the infants of the 14 untreated women, 7 had CMV disease,
whereas only 1 of the infants of 31 mothers in the treatment
group had symptomatic CMV. A second group of 37 pregnant
women with primary CMV received monthly CMVIG, and 6 of
their infants had congenital infection compared with 19 symp-
tomatic infants of 47 mothers who did not receive CMVIG.
Subsequently Nigro et al107 reported the long-term follow-up
of three fetuses with CMV-associated cerebral and other ultra-
sound abnormalities in which the mothers were treated with
CMVIG during pregnancy. Ventriculomegaly regressed, and
ascites, hepatic echodensities, periventricular echodensities,
and intestinal echodensities resolved. Their mental and motor
developments were normal at 4, 4.7, and 7 years of age.
A randomized trial will be needed to establish the value of
CMVIG in preventing CMV infection or disease. Until then,
CMVIG should be considered for pregnancies complicated
by CMV infection if abortion is not an option.
Ebola hemorrhagic fever
Ebola, caused by a filovirus, is a severe hemorrhagic fever with
high mortality and without specific therapy. Goat, equine, and
murine hyperimmune sera and monoclonal antibodies were
protective in preexposure and postexposure trials in guinea pigs
and mice.108–112 Goat hyperimmune sera was given to four adults
following Ebola laboratory accidents113; in the definitely exposed
patient, a mild case of Ebola developed. The other three with
questionable exposure did not become ill. All received human
interferon-2.
Equine sera did not protect cynomolgus monkeys against
Ebola if given on the day of and 5 days after infection.114 Jahrling
et al115 gave rhesus macaque immune whole blood to other rhe-
sus macaques 3 or 4 days after viral challenge: all monkeys died.
Rhesus macaques could not be protected by a monoclonal anti-
body protective in guinea pigs. Oswald et al116 gave KZ52, a
neutralizing human monoclonal antibody protective in guinea
pigs, to 4 macaques 1 day before and 4 days after challenge; all
animals died.
Thus, antibody to Ebola is successful in guinea pigs and
mice but not in nonhuman primates and, thus, probably not
in humans.
Enteroviral infections in newborns
IGIV has been used for disseminated neonatal enteroviral infec-
tion in neonates and to prevent its spread in nursery outbreaks.
Several case reports have described the successful use of IGIV
in the treatment of disseminated enteroviral infection,117–119
particularly if treatment is started early.120 By contrast, Abzug
et al121 reported no clinical benefit of IGIV, despite a reduced
period of viremia and viruria if the IGIV had a high neutral-
izing titer.
Similarly, the value of IGIV in neonatal enteroviral nursery
outbreaks is conflicting. Pasic et al,122 using IGIV in a nurs-
ery outbreak of echovirus 6 and echovirus 4, found that trans-
mission continued although clinical disease was attenuated.
Nagington et al123 used IGIM to control the spread of a nursery
outbreak of echovirus 11. By contrast, Kinney et al124 found no
such benefit in their echovirus 11 nursery outbreak. In sum,
 SECTION ONE
if the strain of the enterovirus is identified and the IGIV used
has titers to the strain, IGIV use should be considered; proof of
efficacy is lacking.
Enteroviral meningoencephalitis
In patients with profound antibody immunodeficiencies, IGIV
is used to prevent central nervous system enteroviral infec-
tion and to treat chronic enteroviral meningoencephalitis.
Multiple authors125–129 have summarized the value of IGIV and/
or intraventricular IGIV in these cases, primarily in patients
with X-linked agammaglobulinemia. Maintenance of a high
serum trough level is recommended, e.g., 900 to 1000 mg/
dL by McKinney et al125 or more than 800 mg/dL by Quartier
et al,128 as determined by the clinical, virologic, and inflam-
matory response of the cerebrospinal fluid (CSF). The IGIV
should contain antibodies to the enteroviral serotype causing
the infection.
Wang and Liu130 reviewed the use of IGIV in enterovirus 71
brainstem encephalitis in Taiwan: they recommend IGIV for
encephalitic patients with autonomic nervous system dysfunc-
tion and patients with acute flaccid paralysis and CSF pleo-
cytosis. A randomized trial of IGIV and milrinone for severe
enterovirus 71 is in progress. IGIV preparations from China
have high titers to enterovirus 71, but IGIV from Canada
lacks such antibodies.131 Monoclonal antibodies are under
development.132,133
Herpes simplex virus
Although maternal antibody to herpes simplex virus (HSV)
can dramatically lower perinatal transmission of herpes infec-
tion to the infant, a hyperimmune immunoglobulin prepara-
tion or monoclonal antibody is not available to prevent vertical
HSV transmission. Current IGIV has low and variable titers
of herpes simplex antibodies.134 As reviewed by Kimberlin and
Whitley,134 human studies have not been performed, although
animal studies in mice135 and guinea pigs136 support prophylac-
tic and treatment roles for antibody in HSV infection.
In adults with recurrent genital HSV infection, monthly
IGIV resulted in fewer recurrences than intermittent acyclovir
treatment.137 However, use of antibody to prevent and treat her-
pes simplex infections cannot be recommended.
Parvovirus
Parvovirus B19, the cause of a minor childhood illness with
rash (erythema infectiosum, fifth disease), can also cause
serious infections in immunocompetent and immunocom-
promised patients and fetuses.138 IGIV, an excellent source of
neutralizing antibody for parvovirus,139 is used for parvovirus-
induced pure red cell aplasia (PRCA) in immunocompromised
patients. Before highly active antiretroviral treatment (HAART)
for HIV, chronic parvovirus B19-PRCA was treated effectively
with IVIG.140 However, HAART alone with immune reconstitu-
tion may effectively treat this condition.141
Parvovirus B19-PRCA and other cytopenias in transplant
recipients142–145 and oncology patients146–148 have been treated
successfully with IGIV. Side effects have included acute renal
failure149,150 and pancreatic allograft thrombosis.151
In addition, IGIV has been used in other parvovirus B19–
induced diseases, including severe chronic arthritis,152 polyarter-
itis nodosa,153 acute fulminant hepatitis,154 and chronic fatigue
syndrome.155 In 2010, Dennert et al156 treated 17 patients with
chronic dilated cardiomyopathy who had a high parvovirus viral
load in endocardial biopsies with IGIV; their parvovirus viral
load decreased and the cardiac ejection fraction improved.
Parvovirus infection during pregnancy can result in severe
hydrops fetalis and fetal death; intrauterine transfusions may
be needed.157 Immunoglobulin therapy can also be given during
pregnancy to the mother and indirectly to the fetus. Matsuda
et al158 injected immunoglobulin into the peritoneal cavity of
a fetus with hydrops, thus avoiding intrauterine transfusion.
Selbing et al159 treated a pregnant woman with a severely affected
infant with high-dose IGIV at 24 weeks' gestation. Severe con-
genital anemia due to parvovirus has also been treated with
IGIV and transfusion.160–162
Poliovirus
In the early 1950s, before the availability of polio vaccine,
Hammon et al163,164 conducted a series of controlled clinical tri-
als of the value of IGIM in the prevention of polio. Their stud-
ies, the largest passive immunity controlled trial ever conducted
(55,000 children), showed that IGIM provided significant but
short-lived protection against poliovirus infection, and if pro-
phylaxis failed, there was attenuation of disease severity. With
widespread immunization now available, IGIM is indicated
only for unimmunized or immunodeficient subjects traveling
to an area with known poliovirus infections.
Postpolio syndrome
Gonzalez et al,165 in 2004, reported that patients with postpolio
syndrome had elevated CSF messenger RNA levels of TNF-
and interferon- that could be reduced by IGIV treatment. This
group performed a subsequent randomized placebo-controlled
trial in 73 postpolio patients who received one 90-g dose of
IGIV and 69 untreated control subjects; muscle strength was
increased by 8.3% in the IGIV-treated patients compared with
the control subjects.166 There also was improvement in fitness,
pain, and physical activity but no change in quality of life, sleep
quality, or balance.
Farbu et al167 treated 10 patients with postpolio syndrome
with 2 g/kg of IGIV and compared them with 10 untreated con-
trol subjects. After 3 months, the IGIV-treated patients had
decreased pain but no change in muscle strength or fatigue.
IGIV is unproven in postpolio syndrome.
Rabies
Rabies is the ideal disease for passive immunization because
the exact moment, the exact source, and the exact location of
exposure usually are known. Furthermore, the long incubation
period and the fact that virus remains localized in the wound for
several days enhance the effectiveness of passive immunization.
Optimal prevention of rabies following exposure requires
vaccine and antibody administration in addition to wound
cleansing.168 Several vaccines are available, and antibody is
available as human rabies immune globulin (RIG) or, in some
countries, equine rabies antiserum.
The first dose of vaccine and RIG should be given as soon
as possible after exposure (at different sites and with different
syringes).169,170 As much as possible of the RIG should be given
at the wound site—dilution of the RIG may be necessary to
ensure that all contaminated areas are injected.169,170 The dose
of RIG of 20 IU/kg should not be exceeded since it may interfere
with the antibody response to the vaccine.
Failure of combined vaccine-antibody prophylaxis has
been reported, possibly associated with inadequate RIG dose,
failure to infiltrate all wounds with RIG, immune compro-
mise with failure of the vaccine to stimulate an antibody
response, and direct inoculation of the virus into a nerve of
the face.169,171
A further description of preexposure and postexposure
prophylaxis is given in the chapter on rabies vaccine (Chapter 29).
Monoclonal antibodies against rabies have been prepared
and tested for safety in humans.172–174

Respiratory syncytial virus
Respiratory syncytial virus (RSV) is the leading cause of serious
respiratory illness in young children. RSV is a major cause of
hospitalizations and morbidity for infants worldwide, and there
is no effective vaccine.
Passive immunity for RSV first used a human hyperim-
mune immunoglobulin that is no longer available.175 This was
replaced by the monoclonal antibody palivizumab (Synagis),176 a
humanized IgG1 monoclonal antibody directed at the F (fusion)
glycoprotein of RSV.177,178 A new more-potent monoclonal anti-
body (motavizumab) has recently been studied and shown to be
efficacious for prophylaxis; it is not yet licensed worldwide.179
Palivizumab is given intramuscularly (15 mg/kg every 30 days)
during the RSV season. Its use significantly decreases hospital-
izations due to RSV in premature infants and premature infants
with chronic lung disease. Although expensive, palivizumab is
now the US standard of care to prevent RSV hospitalization in
targeted groups. The timing and duration of use is dependent on
the RSV season in different geographic regions.180
Palivizumab decreased hospitalizations due to RSV by 55%
for premature infants and infants with chronic lung disease in
the RSV impact study.176 Palivizumab decreased RSV by 45%
in infants with hemodynamically significant heart disease.181
Both of these studies were randomized, double-blind, and
placebo-controlled.
The Committee on Infectious Diseases for the American
Academy of Pediatrics182 recommends palivizumab for the fol-
lowing cases: (1) infants with chronic lung disease younger than
24 months requiring medical therapy within 6 months before
the start of the RSV season (second season of treatment can be
considered); (2) infants born before 32 weeks' gestation; (3) pre-
mature infants born at 32 to less than 35 weeks' gestation with
increased risk of exposure because of attendance at child care or
one or more siblings or household contacts younger than 5 years;
(4) infants with congenital abnormalities of the airway or neu-
romuscular disease (may consider) and (5) children younger
than 24 months with hemodynamically significant congenital
heart disease, particularly children undergoing treatment for
congestive heart failure, infants with moderate to severe pulmo-
nary hypertension, and infants with cyanotic heart disease.181
Although palivizumab has been used as prophylaxis in neo-
natal intensive care unit (NICU) outbreaks, the Committee
does not recommend its use in this situation, but rather rec-
ommends strict infection control practices. Although there are
several reports of its use to control NICU outbreaks, none are
controlled studies.183 Katz and Sullivan184 found no change in
nosocomial RSV rates after initiation of monthly RSV prophy-
laxis use in all susceptible infants in the NICU.
In studying the pharmacokinetics of palivizumab in very
premature infants, Wu et al185 found that optimal levels were
attained only after the second dose; thus, additional studies are
needed to determine optimum dosing intervals. Abadesso et al186
administered palivizumab after a fifth case in an NICU out-
break; retrospective analysis suggested some benefit. Cox et al187
Access the complete reference list online at http://www.expertconsult.com
1. Stiehm ER, Keller MA. Passive immunization. In: Feigin RD, Cherry JD,
Demmler-Harrison GJ, et al, (eds). Feigin and Cherry's Textbook of Pediatric
Infectious Diseases. 6th ed. Philadelphia, PA: Saunders Elsevier; 2009, 3401–3446.
3. Keller MA, Stiehm ER. Passive immunity in prevention and treatment of
infectious diseases. Clin Microbiol Rev 13:602–614, 2000.
42. Arnon SS, Schechter R, Maslanka SE, et al. Human botulism immune globulin
for the treatment of infant botulism. N Engl J Med 345:462–471, 2006.
103. Kotton CN, Kumar D, Caliendo AM, et al. International consensus
guidelines on the management of cytomegalovirus in solid organ
transplantation. Transplantation 89:779–795, 2010.
106. Nigro G, Adler SP, La Torre R, et al., Congenital Cytomegalovirus
Collaborating Group. Passive immunization during pregnancy for congenital
cytomegalovirus infection. for the N Engl J Med 353:1350–1362, 2005.
Passive immunization 7 87
also used palivizumab after 7 infants in the NICU had been
infected. These data do not yet support its use in NICU
nosocomial outbreaks.
Since RSV can cause devastating illness in a transplantation
unit, palivizumab has been used for prophylaxis188 and treatment189
in this setting. Several studies190–192 indicated its safety but are incon-
clusive about its efficacy. Palivizumab was used without ribavirin to
treat 19 patients with RSV infection in a stem cell transplantation
unit.193 However, palivizumab did not prevent progression to lower
respiratory infection or alter the survival rate. Thus, the role for
palivizumab in RSV treatment remains undefined.
A Cochrane meta-analysis of the value of palivizumab in
cystic fibrosis was inconclusive.194 Its use in infants with neuro-
muscular diseases is supported by their increased risk for severe
RSV.195
Smallpox (variola) and smallpox vaccine (vaccinia)
Although smallpox (variola) has been eradicated since 1977, it
has tremendous biologic warfare potential since routine vacci-
nation has been discontinued. The virus is present in many
laboratories throughout the world. Furthermore, smallpox vac-
cine is still used in military populations and laboratory person-
nel working with vaccinia and related viruses.
Kempe et al196,197 showed that human vaccinia immune
globulin (VIG) from the plasma of recently vaccinated subjects
markedly reduced transmission of smallpox among close con-
tacts of index patients. Its main use has been to treat patients
with complications of smallpox vaccination, including gener-
alized vaccinia, eczema vaccinatum, vaccinia necrosum, and
autoinoculation.198,199
VIG is available for IM and IV use. Further details are given
in the chapter on smallpox and vaccinia (Chapter 32).
Varicella-zoster infections
Varicella-zoster immune globulin (VZIG) has been used since
1978 for the prevention or modification of varicella in exposed,
susceptible, high-risk subjects.1 VZIG is prepared from human
plasma with high titers to VZV. It is now available as VariZIG
for IM injection through FFF Enterprises (Temecula, CA) as
outlined in the Red Book.200 Another VZIG preparation has also
been used intravenously.201
VZIG is usually given to susceptible subjects as soon as pos-
sible within 96 hours of exposure to prevent or modify vari-
cella severity.200 Exposed subjects who are candidates for VZIG
include immunocompromised patients, newborn infants whose
mothers have chickenpox in the perinatal period, premature
infants, and pregnant women. The latter are at risk for devel-
oping varicella pneumonia and giving birth to a child with the
congenital varicella syndrome. Prompt use of VZIG in exposed
pregnant women diminishes the likelihood of delivering an
infant with congenital varicella.202
Further details are presented in the chapter on varicella vac-
cine (Chapter 37).
125. McKinney RE, Jr, Katz SL, Wilfert CM. Chronic enteroviral meningoencephalitis
in agammaglobulinemic patients. Rev Infect Dis 9:334–356, 1987.
150. Bassols AC. Parvovirus B19 and the new century. Clin Infect Dis 46:537–539,
2008.
156. Dennert R, Velthuis S, Schalla S, et al. Intravenous immunoglobulin therapy
for patients with idiopathic cardiomyopathy and endomyocardial biopsy-
proven high PVB19 viral load. Antivir Ther 15:193–201, 2010.
176. Impact-RSV Study Group. Palivizumab, a humanized respiratory syncytial
virus monoclonal antibody, reduces hospitalization from respiratory
syncytial virus infection in high-risk infants. Pediatrics 102:531–537, 1998.
182. Committee on Infectious Diseases. Modified recommendations for use
of palivizumab for prevention of respiratory syncytial virus infections.
Pediatrics 12:1694–1701, 2009.
 SECTION ONE: General aspects of vaccination
General immunization practices
8 Andrew T. Kroger
William L. Atkinson
Larry K. Pickering
Recommendations for immunization practices are based
on scientific knowledge of vaccine characteristics, biology of
immunization, epidemiology of specific diseases, and host
characteristics. In addition, experience and judgment of public
health officials and specialists in clinical and preventive medi-
cine have a key role in developing recommendations that max-
imize benefits and minimize risks and costs associated with
immunization. General guidelines for immunization practices
are based on evidence and expert opinion of benefits, costs, and
risks of vaccinations as they apply to the current epidemiology
of disease and use of vaccines in the United States. However,
many of the principles are universal and are applicable to
other countries where different public health infrastructures
may exist.
Vaccine storage and handling
Vaccines must be properly shipped, stored, and handled to
avoid loss of their biologic activities. Recommended stor-
age and handling requirements for each vaccine are given in
each manufacturer's product label.1 Correct shipping, storage,
and handling practices also are published in recommendations
of the major vaccine policy-making committees, such as the
Advisory Committee on Immunization Practices (ACIP) of the
Centers for Disease Control and Prevention, the Committee
on Infectious Diseases of the American Academy of Pediatrics
(AAP), and the World Health Organization (see Chapter 70).2-5
Failure to adhere to these requirements can result in loss of vac-
cine potency, leading to an inadequate immune response in the
vaccinee. Visible evidence of altered vaccine integrity may not
be present. The manufacturer should be contacted when ques-
tions arise about the correct handling of a vaccine. New vaccines
or new formulations of an existing vaccine may have different
shipping, storage, and handling requirements. Table 8-1 gives
recommended storage practices for the most commonly used
vaccines in the United States.
Refrigerators without freezers and stand-alone freezers are
usually the most effective at maintaining the precise temper-
atures required for vaccine storage and are preferred to combi-
nation refrigerator/freezer units. A combination refrigerator/
freezer unit sold for home use is acceptable for storage of vac-
cines only if the refrigerator and freezer compartments each
have a separate external door. Freezer storage units may be
manual defrost or automatic defrost (“frost-free”). Automatic
defrost freezers periodically and transiently increase the
freezer temperature to reduce the formation of ice. This type
of freezer unit is acceptable for the storage of vaccines that
must be stored in a frozen state (varicella, measles-mumps-
rubella-varicella [MMRV], oral polio, zoster).
Exposure to higher or lower temperatures than recom-
mended can damage a vaccine (Table 8-1). For example, live
virus vaccines such as oral poliovirus vaccine (OPV), varicella,
combination MMRV, and zoster vaccine are sensitive to tem-
peratures above freezing and should be kept frozen until just
before administration. Measles-mumps-rubella (MMR, without
varicella) vaccine, rotavirus vaccine, and yellow fever vaccine
should be stored at refrigerator temperature (2-8°C [35-46°F]).3,6
However, vaccines composed of purified antigens or inactivated
microorganisms, such as hepatitis A, hepatitis B,
type b (Hib), human papillomavirus, and inactivated
influenza, can lose their potency if frozen and should be kept
at refrigerator temperature and never frozen.3,4 Diluents should
not be frozen and may be kept at room or refrigerator tempera-
ture. Maintenance of a “cold chain” from vaccine production to
use helps ensure vaccine potency at the time of administration.
Temperature monitoring and control are important for storage
and handling of all vaccines, particularly during transport and
field use. Temperatures should be monitored at least twice a
day, preferably using a thermometer that records current, maxi-
mum, and minimum temperatures. Whereas maintenance of
cold and freezing temperatures may be a problem in tropical
climates, data suggest that inappropriate freezing of inactivated
vaccines is a problem in maintaining vaccine stability in cold
and temperate climates. Shipping containers should be sturdy,
the correct size for the amount of vaccine to be shipped, and
contain a temperature monitor. Appropriate insulation (eg, pan-
els and boxes of polystyrene, isocyanurate, or polyurethane) and
cold source (eg, dry ice, gel packs, or bottles with frozen liq-
uid) should be used to maintain the recommended temperature.
Loose fillers do not provide reliable temperature insulation.3
Vaccines should not be reconstituted until immediately
before use. If not administered within the interval recom-
mended by the manufacturer, reconstituted vaccine should
be discarded.5 Only the diluent provided by the manufacturer
should be used to reconstitute a lyophilized vaccine. With the
exception of OPV, live virus vaccines should not be refrozen
after thawing (Table 8-1). Certain vaccines (eg, MMR, varicella,
and MMRV) also must be protected from light to prevent inac-
tivation of the vaccine virus.3
Certain vaccines are distributed in multidose vials. When
opened, the remaining doses from partially used multidose vials
can be administered until the expiration date printed on the vial
or vaccine packaging unless otherwise specified by the manu-
facturer, provided that the vial has been stored correctly and
that the vaccine is not visibly contaminated.5
 General immunization practices 8 89

Vaccine Storage Temperature Recommendations

90 SECTION ONE General aspects of vaccination
Vaccine Storage Temperature Recommendations—cont'd
DT: diphtheria and tetanus toxoids
DTaP: DT and acellular pertussis
Td: tetanus and diphtheria toxoids
Tdap: tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis
*DTaP-Tripedia is sometimes used as a diluent for ActHib.
†
Protect from light.
‡
There are two meningococcal conjugate vaccines; Menactra is nonlyophilized, and Menveo is lyophilized. Both powder and diluent should be stored at 35-46°F.
§
The lyophilized pellet may be stored at freezer temperature; the reconstituted vaccine should be stored at refrigerator temperature.
Adapted from Centers for Disease Control and Prevention. General Recommendations on Immunization: Recommendations of the Advisory Committee on
Immunization Practices (ACIP). MMWR 2011;60(No. RR-2):1-61.
Vaccine administration
Complete and accurate records documenting administration
of all vaccines should be maintained by healthcare provid-
ers who administer vaccines and vaccine recipients (or their
parents). For each immunization, the following information
should be recorded: (1) date of vaccination; (2) product admin-
istered, manufacturer, lot number, and expiration date; (3) site
and route of administration; and (4) name, address, and title of
healthcare provider administering the vaccine.
Infection control and sterile injection technique
Infection resulting from administration of vaccines is unlikely
if appropriate precautions are taken. Hands should be washed
with soap and water or cleansed with an alcohol-based water-
less antiseptic hand rub before each patient contact to reduce
the risk of bacterial contamination and transmission of micro-
organisms between recipients and healthcare personnel. In
general, use of protective gloves is not necessary when adminis-
tering vaccines unless the healthcare provider will have contact
with potentially infectious body fluids or has open lesions on
the hands.2,5
Failure to follow relevant infection control guidelines can
result in transmission of bloodborne pathogens or bacterial
infection and abscess formation. Contamination of an injection
site can occur from bacteria on the skin at the site of injection.
To prevent such contamination, skin at the injection site should
be prepared with isopropyl alcohol (70%) or another disinfect-
ing agent and allowed to dry before injection. Transmission of
pathogens also can occur if needles, syringes, vaccines, or other
equipment used to administer vaccines becomes contaminated.
To prevent such contamination, syringes and needles must be
sterile. A separate needle and syringe should be used for each
injection. Disposable needles and syringes should be discarded
after a single use in a labeled, puncture-proof container to pre-
vent inadvertent needle-stick injury or reuse. Because recapping
and removing a used needle from a syringe can result in injury
to the user, needles should not be recapped after use.5 The nee-
dle and syringe should be discarded as a single unit without
removing the needle from the syringe. Single-use disposable
needles and syringes should not be sterilized and reused.
If only reusable (ie, nondisposable) needles and syringes are
available, they must be thoroughly cleaned and sterilized after
each injection to prevent transmission of bloodborne or other
pathogens between patients. Reusable syringes are usually glass
rather than plastic. Because of its inert characteristics, glass
can be cleaned and sterilized more easily than plastic. Because
hypodermic needles enter deep tissues, great care must be taken
to ensure that all contaminants are removed from the needle
and syringe.7 Liquid germicides alone are insufficient for nee-
dle sterilization because of the restricted access of the chemical
agent to the narrow lumen of the needle. Strict adherence to the
recommended time and temperature for the sterilization proce-
dure used must be observed.
The majority of vaccines have a similar appearance after
they are drawn into a syringe. Cases in which the wrong vac-
cine was administered often are attributable to the practice of
prefilling syringes or drawing doses of a vaccine into multiple
syringes before their immediate need.5 The routine practice of
prefilling syringes should be discouraged because of the poten-
tial for such administration errors. To prevent errors, vaccine
doses should not be drawn into a syringe until immediately
before administration. In certain circumstances in which a sin-
gle vaccine type is used (eg, in a community influenza vaccina-
tion campaign), filling multiple syringes before their immediate
use may be considered. Care should be taken to ensure that
the cold chain is maintained until the vaccine is administered.
When the syringes are filled, the type of vaccine, lot number,
and date of filling must be labeled carefully on each syringe,
and the doses should be administered as soon as possible after
filling. Vaccine drawn into syringes by the user (that is, not by
the manufacturer) generally should be discarded at the end of
the clinic day.
Route of administration
One or more routes of administration (eg, intramuscular, subcu-
taneous, intradermal, intranasal, and oral) are recommended for
each vaccine and are listed in the manufacturer's product label
and in published recommendations of immunization advisory
committees (Table 8-2).2,5 These routes usually are determined
during prelicensure vaccine studies and are based on vaccine com-
position and immunogenicity. Vaccines should be administered
in sites where they elicit the desired immune response and where
the likelihood of local tissue, neural, or vascular injury is mini-
mal.2 To avoid unnecessary local and systemic adverse events and
to ensure the appropriate immune response, persons administer-
ing vaccines should not deviate from the recommended route of
 General immunization practices 8 99

†
Recommended and Minimum Ages and Intervals Between Vaccine Doses* —cont'd

*
Combination vaccines are available. Use of licensed combination vaccines is generally preferred to separate injections of their equivalent component vaccines.
When administering combination vaccines, the minimum age for administration is the oldest age for any of the individual components; the minimum interval between
doses is equal to the greatest interval of any of the individual components.
†
Information on travel vaccines, including typhoid, Japanese encephalitis, and yellow fever, is available at http://www.cdc.gov/travel. Information on other vaccines that
are licensed in the United States but not distributed, including anthrax and smallpox, is available at http://www.bt.cdc.gov.
§
Combination vaccines containing the HepB component are available (see Table 8-2). These vaccines should not be administered to infants younger than 6 wk
because of the other components (ie, Hib, DTaP, HepA, and IPV).
¶
HepB-3 should be administered at least 8 wk after HepB-2 and at least 16 wk after HepB-1 and should not be administered before age 24 wk.
**
Calendar months.
††
The minimum recommended interval between DTaP-3 and DTaP-4 is 6 mo. However, DTaP-4 need not be repeated if administered at least 4 mo after DTaP-3.
§§
For Hib and PCV, children receiving the first dose of vaccine at age 7 mo require fewer doses to complete the series.
¶¶
If polyribosylribitol phosphate-meningococcal outer membrane protein conjugate (PRP-OMP; Pedvax-Hib, Merck Vaccine Division) was administered at ages 2 and
4 mo, a dose at age 6 mo is not necessary.
***
A fourth dose is not needed if the third dose was administered at 4 y and at least 6 mo after the previous dose.
†††
Combination MMRV vaccine can be used for children aged 12 mo to 12 y.
§§§
The minimum interval from varicella-1 to varicella-2 for persons beginning the series at age 13 y is 4 wk.
¶¶¶
One dose of influenza vaccine per season is recommended for most persons. Children younger than 9 y who are receiving influenza vaccine for the first time or
who received only one dose the previous season (if it was their first vaccination season) should receive two doses this season.
****
The minimum age for inactivated influenza vaccine varies by vaccine manufacturer. See package insert for vaccine-specific minimum ages.
††††
Revaccination with meningococcal vaccine is recommended for previously vaccinated persons who remain at high risk for meningococcal disease. (Source:
CDC. Updated recommendations from the Advisory Committee on Immunization Practices (ACIP) for revaccination of persons at prolonged increased risk for
meningococcal disease. MMWR 2009;58:1042-1043).
§§§§
Only one dose of Tdap is recommended. Subsequent doses should be given as Td. For one brand of Tdap, the minimum age is 11 y. For management of a
tetanus-prone wound in persons who have received a primary series of tetanus-toxoid–containing vaccine, the minimum interval after a previous dose of any tetanus-
containing vaccine is 5 y.
¶¶¶¶
A second dose of PPSV 5 y after the first dose is recommended for persons aged 65 y at highest risk for serious pneumococcal infection and persons likely to
have a rapid decline in pneumococcal antibody concentration. (Source: CDC. Prevention of pneumococcal disease: recommendations of the Advisory Committee on
Immunization Practices [ACIP]. MMWR 1997;46[RR-8]).
*****
Bivalent HPV vaccine is approved for females aged 10-25 y. Quadrivalent HPV vaccine is approved for males and females aged 9-26 y.
†††††
The minimum age for HPV-3 is based on the baseline minimum age for the first dose (ie, 108 mo) and the minimum interval of 24 wk between the first and third
dose. Dose 3 need not be repeated if it is administered at least 16 wk after the first dose.
§§§§§
The first dose of rotavirus must be administered at age 6 wk through 14 wk and 6 days. The vaccine series should not be started for infants aged 15 wk, 0 days.
Rotavirus should not be administered to children older than 8 mo, 0 days of age regardless of the number of doses received between 6 wk and 8 mo, 0 days of age.
¶¶¶¶¶
If 2 doses of Rotarix (GlaxoSmithKline) are administered as age-appropriate, a third dose is not necessary.
******
Herpes zoster vaccine is recommended as a single dose for persons aged 60 y.

Guidelines for Spacing of Live and Inactivated Antigens

*The American Academy of Pediatrics suggests a 1-mo interval between
tetanus toxoid, reduced diphtheria toxoid, and reduced acellular pertussis
vaccine and tetravalent meningococcal conjugate vaccine if these vaccines
are not administered on the same day (Pediatrics 117:965-978, 2006185).
†
Live oral vaccines (eg, Ty21a typhoid vaccine and rotavirus vaccine) can be
administered on the same day or at any interval before or after inactivated or
live injectable vaccines.
Simultaneous administration of
different vaccines
Simultaneous administration of all indicated vaccines is an
essential component of childhood vaccination programs.2,5
Simultaneous administration of different vaccines is particularly
important when return of the recipient for further vaccination
is uncertain, imminent exposure to several vaccine-preventable
diseases is expected, or a vaccinee is preparing for international
travel on short notice.
Unless specifically licensed for injection in the same syringe,
different vaccines administered simultaneously should be
injected separately and at different anatomic sites. If both upper
and lower limbs must be used for simultaneous administration
of different vaccines, the anterolateral thigh is often chosen for
intramuscular injections and the triceps region for subcutane-
ous injections. If more than one injection must be administered
in a single limb of an infant or young child, the thigh usually is
preferred because of its large muscle mass. The distance sepa-
rating two injections in the same limb should be sufficient (eg,
1 to 2 inches) to minimize the chance of overlapping local reac-
tions.5,12,13 In general, different vaccines, including live virus
products, can be administered simultaneously without reducing
their safety and effectiveness62 (Table 8-4). Studies of cortisol
concentration and behavioral responses to vaccination indicate
that responses are similar in infants who receive two injections
during one visit and infants who receive a single injection, sug-
gesting that a second injection does not increase stress.63,64
Increased severity or incidence of adverse reactions has not
been observed after simultaneous administration of the most
widely used vaccines.5 Similarly, simultaneous administration
of vaccines generally does not cause immunologic interference
except possibly between pneumococcal conjugate vaccine and
Menactra brand meningococcal conjugate vaccine.62,64
100 SECTION ONE General aspects of vaccination
Interference by immune globulins
Passively acquired antibodies can interfere with the immune
response to certain vaccines, both live and inactivated, and to
toxoids. The result can be the absence of seroconversion or a
blunting of the immune response with lower final antibody con-
centrations in the vaccinee. Passively acquired antibody does
not affect the immune response to all vaccines.
Interference with live virus vaccines
To elicit an adequate immune response, live vaccine virus
must replicate in the recipient. The probable mechanism by
which passively acquired immune globulin blunts the immune
response is neutralization of vaccine virus, resulting in inhibi-
tion of viral replication and insufficient antigenic mass.65 For
example, persisting transplacentally acquired maternal measles
antibodies inhibit the response to live measles vaccine in
infants for as long as 12 months and perhaps longer.66,67 The age
to which inhibition persists has been correlated with concen-
trations of maternal or cord blood antibodies.68-70 Rubella vac-
cine virus may be less susceptible than measles vaccine virus
to these transplacentally acquired maternal antibodies.69,71
The effect of blood and immune globulin preparations on the
response to mumps and varicella vaccines is unknown, but
commercial immune globulin preparations contain antibodies
to these viruses. The effect of blood and immune globulin prep-
arations on the response to live rotavirus and influenza vaccines
is unknown.
Intramuscular or intravenous administration of immune
globulin–containing preparations (eg, immune globulin, hyper-
immune globulins, intravenous immune globulin, and blood)
before or simultaneously with certain vaccines also can affect
the immune response to live virus vaccines. When partially
attenuated Edmonston B measles vaccine, which is no longer
available in the United States, was administered concurrently
with measles immune globulin in an effort to reduce the inci-
dence of adverse events associated with this vaccine, the rate
of seroconversion was not affected but the geometric mean
titer of serum measles antibody was diminished.72 In a study
of an investigational bacterial polysaccharide immune globu-
lin (BPIG), children had a reduced immune response to live
measles vaccine for as long as 5 months after receipt of BPIG.73
The measles antibody seroconversion rate and geometric mean
titer were lower among children who received BPIG compared
with children who received placebo. Blunting of the immune
response to live rubella vaccine also occurred after receipt of
BPIG but was less marked and of shorter duration.
Although passively acquired antibodies can interfere with
the response to rubella vaccine, the low dose of anti-Rh(D) glob-
ulin administered to postpartum women has not been demon-
strated to inhibit the immune response to RA27/3 strain rubella
vaccine.74 Parenterally administered immune globulin prepara-
tions also do not seem to adversely affect the immune response
to yellow fever vaccine.75 Although high concentrations of pas-
sively acquired antibodies may reduce the serum antibody
response to live poliovirus vaccine, they have little effect on
replication of vaccine virus and development of gastrointesti-
nal tract immunity.47,75-77 Data are insufficient to determine the
extent to which passively acquired antibodies interfere with the
immune response to other live viral or bacterial vaccines, such
as varicella, mumps, and typhoid (Ty21a strain). A humanized
mouse monoclonal antibody product (palivizumab) is available
for prevention of respiratory syncytial virus infection among
infants and young children. This product contains only anti-
body to respiratory syncytial virus; hence, it will not interfere
with immune response to live vaccines.78
Interference with inactivated and
component vaccines
Interference with current inactivated and component vaccines
is less marked than with live vaccines and requires exposure to
large doses of passively acquired antibodies.79 The mechanism
by which passively acquired antibodies interfere with the immu-
nologic response to inactivated and toxoid vaccines is not clear.
Moderate doses of parenterally administered immune globu-
lins have not inhibited development of a protective immune
response to DTP, tetanus toxoid, hepatitis B vaccines, and
Hib conjugate vaccines.80,81 Although the concurrent adminis-
tration of inactivated hepatitis A vaccine and immune globu-
lin can result in lower serum antibody concentrations than if
vaccine alone is administered, seroconversion rates have not
been diminished.82,83 Infants with high concentrations of pas-
sively acquired maternal antibody to hepatitis A virus had lower
serum antibody concentrations after receipt of hepatitis A vac-
cine but had seroconversion rates similar to those of vaccinated
infants without maternal antibodies.84
Recommendations for spacing administration
of vaccines and immune globulins
Interference of immune globulins with the immune response
to vaccines is dose-related and more likely to occur and to per-
sist for a longer period after receipt of larger doses of immune
globulins.73,85, The recommended interval between administra-
tion of immune globulin preparations and vaccines is based
on whether evidence suggests interference between immune
globulin and the vaccine, the dose of the immune globulin
administered, and the expected half-life of immunoglobulin G.
Recommended intervals between administration of immune
globulin preparations and various live and killed vaccines are
listed in Tables 8-5 and 8-6.
In the United States, inactivated and component (subunit)
vaccines may be administered simultaneously with or at any
time before or after receipt of an immune globulin prepara-
tion.2,5 The vaccine and immune globulin preparation should
be administered at different sites, and the standard recom-
mended doses of the corresponding vaccines should be given.
Supplemental doses are not indicated.
Recommendations for administration of live virus vaccines
vary on the basis of the aforementioned considerations. After
receipt of an immune globulin preparation or other blood prod-
uct, measles vaccine should be deferred during the intervals
listed in Table 8-6.5,77,86 Human blood and immune globulin
preparations also contain rubella, mumps, and varicella anti-
bodies. High doses of passively acquired antibodies can inhibit
the immune response to live rubella vaccine for as long as
3 months.73,87 The effect of immune globulin preparations on
the response to live mumps and live varicella vaccines has not
been defined. To reduce the possibility of interference, post-
ponement of administration of rubella, mumps, and varicella
vaccines for the intervals indicated in Table 8-6 is prudent.5
Immune globulin preparations administered too soon after
vaccination with MMR or varicella vaccines can interfere with
the immune response. If administration of an immune globulin
preparation becomes necessary less than 2 weeks after receipt of
MMR, its component vaccines, or varicella vaccine, readmin-
istration of the vaccine is recommended after the appropriate
interval listed in Tables 8-5 and 8-6, unless serologic testing
indicates an antibody response.5,45 For example, if whole blood
is administered less than 14 days after receipt of varicella vac-
cine, the vaccine should be readministered at least 6 months
after the whole blood unless serologic testing indicates an ade-
quate immune response to the initial dose of varicella vaccine.
 General immunization practices 8 101

Guidelines for Administering Antibody-Containing Products* and Vaccines

*Blood products containing substantial amounts of immune globulin include intramuscular and intravenous immune globulin, specific hyperimmune globulin (eg, hepatitis B
immune globulin, tetanus immune globulin, varicella zoster immune globulin, and rabies immune globulin), whole blood, packed red blood cells, plasma, and platelet
products.
†
Yellow fever vaccine; rotavirus vaccine; oral Ty21a typhoid vaccine; live, attenuated influenza vaccine; and zoster vaccine are exceptions to these recommendations.
These live, attenuated vaccines can be administered at any time before or after or simultaneously with an antibody-containing product.
§
The duration of interference of antibody-containing products with the immune response to the measles component of measles-containing vaccine, and possibly
varicella vaccine, is dose-related (see Table 8-6).

Recommended Intervals Between Administration of Antibody-Containing Products and Measles- or Varicella-Containing Vaccine, by
Product and Indication for Vaccination
102 SECTION ONE

Recommended Intervals Between Administration of Antibody-Containing Products and Measles- or Varicella-Containing Vaccine, by
Product and Indication for Vaccination—cont'd

*
This table is not intended for determining the correct indications and dosages for using antibody-containing products. Unvaccinated persons might not be
protected fully against measles during the entire recommended interval, and additional doses of IG or measles vaccine might be indicated after measles exposure.
Concentrations of measles antibody in an IG preparation can vary by manufacturer's lot. Rates of antibody clearance after receipt of an IG preparation also might
vary. Recommended intervals are extrapolated from an estimated half-life of 30 days for passively acquired antibody and an observed interference with the immune
response to measles vaccine for 5 mo after a dose of 80 mg immune globulin G (IgG)/kg.
†
Does not include zoster vaccine. Zoster vaccine may be given with antibody-containing blood products.
§
Assumes a serum IgG concentration of 16 mg/mL.
¶
Measles and varicella vaccinations are recommended for children with asymptomatic or mildly symptomatic human immunodeficiency virus (HIV) infection but are
contraindicated for persons with severe immunosuppression from HIV or any other immunosuppressive disorder.
**
The investigational VariZIG, similar to licensed varicella-zoster IG (VZIG), is a purified human IG preparation made from plasma containing high levels of antivaricella
antibodies (IgG). The interval between VariZIG and varicella vaccine (alone or as measles, mumps, rubella, and varicella [MMRV]) is 5 mo.
††
Contains antibody only to respiratory syncytial virus.

Although data are not available on the effect of passive
antibody on the response to rotavirus vaccine, the ACIP rec-
ommends that live rotavirus vaccine may be administered at
any time before, concurrent with, or after administration of
any blood product, including antibody-containing products.24
Because the immune responses to OPV, zoster, and yellow
fever vaccines have not been demonstrated to be adversely
affected by immune globulin preparations, these vaccines can
be administered at any time in relation to receipt of immune
globulin preparations.75 Live oral typhoid (Ty21a) vaccine also
is recommended for administration irrespective of the receipt
of immune globulin preparations.5,88 Live attenuated influenza
vaccine can be administered at any time before or after receipt
of an antibody-containing blood product.5
Interchangeability of vaccines from different
manufacturers
Combination and monovalent vaccines against the same dis-
eases with similar antigens and produced by the same manu-
facturer are considered interchangeable in most situations.2,5
However, supporting data on the safety, immunogenicity, and
efficacy of using comparable vaccines from different manufac-
turers for different doses of a vaccination series frequently are
limited or unavailable. When the same vaccine cannot be used
to complete an immunization series, similar vaccines produced
by different manufacturers or produced by the same manu-
facturer in different countries generally have been considered
acceptable to complete the immunization series provided each
vaccine is given according to licensed recommendations.
Some diseases have serologic correlates of immunity that
can be used to evaluate vaccine interchangeability. For exam-
ple, in studies in which one or more doses of hepatitis B vac-
cine produced by one manufacturer were followed by doses from
another manufacturer, the immune response was comparable to
that resulting from use of a single vaccine type.89,90 Whereas Hib
conjugate vaccines differ in antigen composition, interchange-
ability of different products has been validated on the basis of
the accepted serologic correlate of immunity against Hib inva-
sive disease.91-93
Determination of vaccine interchangeability is more difficult
for diseases without serologic correlates of immunity. For exam-
ple, in the absence of such a correlate for
infection, interchangeability of acellular pertussis vaccines is
difficult to assess. Available data from one study indicate that,
for the first three doses of the DTaP series, one or two doses of
Tripedia (manufactured by Aventis Pasteur) followed by Infanrix
(manufactured by GlaxoSmithKline) for the remaining doses(s)
is comparable to three doses of Tripedia with regard to immuno-
genicity, as measured by antibodies to diphtheria, tetanus, and
pertussis toxoids and filamentous hemagglutinin.94 However, in
the absence of a clear serologic correlate of protection for pertus-
sis, the relevance of these immunogenicity data for protection
against pertussis is unknown. When feasible, acellular pertussis
vaccine from the same manufacturer is preferred for the entire
primary vaccination series.50 However, any DTaP vaccine can be
used to complete the DTaP series if the product(s) administered
for earlier doses are unknown or unavailable.5,50,95
Hypersensitivity to vaccine components
Types of reactions
Hypersensitivity reactions after vaccination can be local or sys-
temic and can vary in severity from mild discomfort at the site
of vaccination to severe anaphylaxis. Onset can be immediate
or delayed. Serious allergic reactions are rare. Whether a specific
hypersensitivity reaction is caused by a vaccine component or
an unrelated environmental allergen can be difficult to deter-
mine. However, symptoms occurring immediately after vacci-
nation that are suggestive of an anaphylactic reaction generally
contraindicate further administration of that vaccine to the
recipient.2,5
Urticaria and anaphylactic reactions have been reported after
administration of DTP, diphtheria and tetanus toxoids (DT,
Td), and tetanus toxoid.96-98 Although immunoglobulin E–type
antibodies to tetanus and diphtheria antigens have been identi-
fied in some patients with these symptoms, transient urticaria-
like rashes are not a contraindication to subsequent vaccination
because they are unlikely to be anaphylactic unless they appear

within minutes after vaccination.2,98-100 A serum sickness–type
reaction caused by circulating complexes of vaccine antigen
and previously acquired antibody is the probable cause of these
reactions, and subsequent vaccination at a 10-year interval is
unlikely to result in the necessary ratio of antigen-to-antibody
concentration to form immune complexes.98,101
Tetanus toxoid is contraindicated in persons who experi-
enced an immediate anaphylactic reaction to tetanus toxoid–
containing vaccine, unless the person can be desensitized to the
toxoid.2 Because of the importance of tetanus immunization and
the uncertainty about which vaccine component might be the
cause of the reaction, the patient may be referred to an allergist
for evaluation and possible desensitization.98,102,103 On occasion,
a history of an allergic reaction to tetanus vaccine may refer to
a reaction to tetanus antitoxin of equine origin given for teta-
nus prophylaxis before human-derived tetanus immune globulin
became available in the 1960s. Before use of tetanus toxoid is
discontinued because of an alleged episode of anaphylaxis, skin
testing and possible desensitization should be considered.102,104
Urticaria also has been reported following pneumococcal
conjugate, MMR, varicella, and smallpox vaccines.105-108
Immediate or delayed onset of generalized urticaria and angio-
edema that can progress to respiratory distress and hypotension
has been reported after receipt of inactivated mouse brain-
derived Japanese encephalitis vaccine.109-111 The pathogenesis
of these reactions is not known. Mouse brain derived Japanese
encephalitis vaccine is no longer available in the United States.
A vero cell derived Japanese encephalitis vaccine is available in
the United States for people 17 years of age and older.112
Vaccine components causing hypersensitivity
Proteins
Egg protein is a constituent of vaccines prepared with use of
embryonated chicken eggs, such as influenza and yellow fever
vaccines. On rare occasions, these vaccines can induce anaphy-
laxis or other immediate hypersensitivity reactions, and these
reactions are sometimes attributed to egg protein antigen.2,5,113-115
As a result, these vaccines are generally contraindicated in per-
sons with a history of anaphylactic reactions to egg ingestion
unless desensitization has been successfully completed. For
example, persons needing yellow fever vaccine who have a his-
tory of systemic anaphylaxis-like symptoms after egg ingestion
can be skin tested with yellow fever vaccine before vaccination
and desensitized if necessary.113 Although possible, skin testing
and desensitization with influenza vaccine often are precluded
by the risk of reactions, need for yearly vaccination, and avail-
ability of chemoprophylaxis with antiviral agents active against
influenza virus.114,116,117 The ACIP has recently issued recom-
mendations regarding influenza vaccination of persons with a
history of egg allergy.118
Measles and mumps vaccines are produced in chick embryo
fibroblast cell culture. Persons with hypersensitivity to eggs are
at low risk for anaphylactic reactions to these vaccines, and skin
testing with vaccine is not predictive of allergic reaction after
immunization.2,119-121 Neither skin testing nor administration of
gradually increasing doses of vaccine is required when these vac-
cines are administered to persons who are allergic to eggs.2,5,45,86,122
Live virus vaccines, such as measles, mumps, rubella, yellow
fever, and varicella, contain gelatin as a stabilizer. Persons with
a history of allergy to gelatin have experienced, on rare occa-
sions, an anaphylactic reaction after vaccination with such a
vaccine.105,108,123-125 Skin testing of persons with a history of sys-
temic anaphylaxis-like symptoms after gelatin ingestion may
be useful to identify persons at risk for severe hypersensitivity
reactions to vaccination. A regimen for administering particular
General immunization practices 8 103
vaccines to persons with anaphylaxis to components contained
in those vaccines has been published.126 Because gelatin used as
a vaccine stabilizer may be of porcine origin, whereas ingested
food gelatin may be of bovine origin, absence of a history of
allergy to gelatin-containing foods does not eliminate the pos-
sibility of a gelatin-mediated reaction to vaccine.
Approximately 6% of persons who receive a booster dose of
human diploid rabies vaccine have a serum sickness–type ill-
ness.127,128 This reaction is thought to be caused by sensitization
to human albumin that has been altered chemically by a virus-
inactivating agent used in the production of the vaccine.2,129
Anaphylaxis following recombinant hepatitis B vaccines
rarely has been reported and usually is attributed to hypersensi-
tivity to residual yeast protein in the vaccine.130
Latex
Latex is liquid sap from the commercial rubber tree. Latex con-
tains naturally occurring impurities (eg, plant proteins and pep-
tides), which are believed to be responsible for allergic reactions.
Latex is processed to form natural rubber latex and dry natural
rubber. Dry natural rubber and natural rubber latex might con-
tain the same plant impurities as latex but in lesser amounts.
Natural rubber latex is used to produce medical gloves, cath-
eters, and other products, whereas dry natural rubber is used in
syringe plungers, vial stoppers, and injection ports on intravas-
cular tubing. Synthetic rubber and synthetic latex also are used
in medical gloves, syringe plungers, and vial stoppers, but they
do not contain natural rubber or natural latex or the impurities
linked to allergic reactions.
The most common type of latex sensitivity is contact-type
(type 4) allergy, usually as a result of prolonged contact with
natural rubber latex–containing gloves.131 Although injection
procedure–associated latex allergies among patients with diabe-
tes mellitus have been described,132-134 allergic reactions, includ-
ing anaphylaxis, after vaccination procedures are rare. Only one
report of an allergic reaction after administering hepatitis B
vaccine in a patient with known severe allergy (anaphylaxis) to
latex has been published.135
If a person reports a severe (anaphylactic) allergy to latex,
vaccines supplied in vials or syringes that contain natural rub-
ber should not be administered, unless the benefit of vaccina-
tion outweighs the risk of an allergic reaction to the vaccine.
For latex allergies other than anaphylactic allergies, such as a
history of contact allergy to latex gloves, vaccines supplied in
vials or syringes that contain dry natural rubber or natural rub-
ber latex can be administered.5
Antimicrobial agents
Live virus vaccines may contain trace amounts of one or more
antimicrobial agents, such as neomycin, streptomycin, and
polymyxin B. Vaccine contents are listed in each manufactur-
er's product label for each vaccine. The most common aller-
gic response to neomycin is a delayed-type (cell-mediated) local
contact dermatitis consisting of an erythematous, pruritic pap-
ule that occurs 48 to 96 hours after vaccine administration.2,5,136
Such delayed-type reactions are not contraindications for vac-
cination.2,5,136,137 However, persons who have experienced an
anaphylactic reaction to neomycin or to another vaccine con-
stituent should not receive vaccines containing that antimi-
crobial agent.2,5,138,139 No vaccines licensed in the United States
contain penicillin or penicillin derivatives.
Thimerosal
Thimerosal is an organic mercurial compound in use since the
1930s and added to certain immunobiologic products as a pre-
servative. A joint statement issued by the US Public Health
104 SECTION ONE General aspects of vaccination
Service and the AAP in 1999140 established the goal of removing
thimerosal as soon as possible from vaccines routinely recom-
mended for infants. Although no evidence exists of any harm
caused by low concentrations of thimerosal in vaccines and the
risk was only theoretical,141 this goal was established as a pre-
cautionary measure.
Since mid-2001, vaccines produced in the United States that
are recommended routinely for young infants have been manu-
factured without thimerosal as a preservative and contain no
thimerosal or only trace amounts. Thimerosal as a preserva-
tive is present in certain other vaccines. Examples are tetanus
toxoid, Td, DT, certain formulations of influenza vaccine, and
meningococcal polysaccharide vaccine in multidose vials.142
Formulations of influenza vaccine with a reduced concentra-
tion of thimerosal or no thimerosal as a preservative are avail-
able in the United States.142
Receiving thimerosal-containing vaccines has been pos-
tulated to lead to induction of allergy in some persons.143,144
However, there is limited scientific evidence for this assertion.
Hypersensitivity to thimerosal usually consists of local delayed-
type hypersensitivity reactions.145-147 Thimerosal elicits posi-
tive delayed-type hypersensitivity patch tests in 1% to 18% of
persons tested, but these tests have limited or no clinical rel-
evance.148,149 The majority of patients do not experience reac-
tions to thimerosal administered as a component of vaccines,
even when patch or intradermal tests for thimerosal indicate
hypersensitivity.145,149 A localized or delayed-type hypersensitiv-
ity reaction to thimerosal is not a contraindication to receipt of
a vaccine that contains thimerosal.2,5
Management of acute vaccine
adverse reactions
Although rare after vaccination, the immediate onset and life-
threatening nature of an anaphylactic reaction require that
personnel and facilities providing vaccination be capable of pro-
viding initial care for suspected anaphylaxis. Epinephrine and
equipment for maintaining an airway should be available for
immediate use.
Anaphylaxis usually begins within several minutes of
administration of vaccine. Rapid recognition and initiation of
treatment are required to prevent possible progression to car-
diovascular collapse. If flushing, facial edema, urticaria, itching,
swelling of the mouth or throat, wheezing, difficulty breathing,
or other signs of anaphylaxis occur, the patient should be placed
in a recumbent position with the legs elevated. Aqueous epi-
nephrine (1:1,000) should be administered intramuscularly or
subcutaneously and can be repeated within 10 to 20 minutes.150
A dose of diphenhydramine hydrochloride may shorten the
reaction, but it will have little immediate effect. Maintenance
of an airway and oxygen administration may be necessary.
Arrangements should be made for immediate transfer to an
emergency facility for further evaluation and treatment. All
patients should be observed for at least 12 hours after onset of
symptoms.150
Syncope (vasovagal or vasodepressor reaction) can occur
after vaccination and is most common among adolescents and
young adults. In 2005, the Vaccine Adverse Event Reporting
System (VAERS) detected a trend of increasing syncope reports
that coincided with licensure of three vaccines for adolescents:
human papillomavirus, MCV4, and Tdap.151 Of particular
concern among adolescents has been the risk for serious sec-
ondary injuries, including skull fracture and cerebral hemor-
rhage after a fall and subsequent head injury. Of 463 VAERS
reports of syncope during January 1, 2005, to July 31, 2007,
a total of 41 listed syncope with secondary injury with informa-
tion on the timing after vaccination, and the majority of these
syncope reports (76%) occurred among adolescents. Among
all age groups, 80% of reported syncope episodes occur within
15 minutes of vaccine administration (additional information
available at http://www.cdc.gov/vaccinesafety/concern/syncope.
htm). Providers should take appropriate measures to prevent
injuries if a patient becomes weak or dizzy or loses conscious-
ness. Adolescents and adults should be seated or lying down
during vaccination. Vaccine providers, particularly when vac-
cinating adolescents, should consider observing patients (with
patients seated or lying down) for 15 minutes after vaccina-
tion to decrease the risk for injury should they faint.151 If syn-
cope develops, patients should be observed until the symptoms
resolve.
Special considerations
Vaccination of preterm infants
The immune response to vaccination is a function of postna-
tal rather than gestational age.152-154 Transplacentally acquired
maternal antibody is present in lower concentrations and, thus,
persists for a shorter interval in preterm infants than in gesta-
tionally mature infants.153,155-157 Because preterm infants have
less transplacentally acquired maternal antibody, inhibition of
the immune response in preterm infants may be less than that
in full-term infants.153,158
In the majority of cases, infants born prematurely, regardless
of birth weight, should be vaccinated at the same chronological
age and according to the same schedule and precautions as full-
term infants and children.5 Birth weight and size are not factors
in deciding whether to postpone routine vaccination of a clini-
cally stable preterm infant,159-163 except for hepatitis B vaccine.
The full recommended dose of each vaccine should be used.
Divided or reduced doses are not recommended.164
Decreased seroconversion rates might occur among certain
preterm infants with birth weights of less than 2,000 g after
administration of hepatitis B vaccine at birth.165 However, by
chronological age 1 month, all preterm infants, regardless of ini-
tial birth weight or gestational age, are likely to respond as ade-
quately as older and larger infants.166-168 Preterm infants born
to HBsAg-positive mothers and mothers with unknown HBsAg
status must receive immunoprophylaxis with hepatitis B vac-
cine and hepatitis B immunoglobulin (HBIG) within 12 hours
after birth. HBIG must be given within 12 hours of birth if the
infant weighs less than 2,000 g; if the infant weighs 2,000 g or
more, HBIG must be given within 7 days of birth. Note that
regardless of weight, hepatitis B vaccine must be given within
12 hours of birth. For infants who weigh less than 2,000 g at
birth, the initial vaccine dose should not be counted toward
completion of the hepatitis B vaccine series, and 3 additional
doses of hepatitis B vaccine should be administered, begin-
ning when the infant is 1 month old. Preterm infants weighing
less than 2,000 g and born to HBsAg-negative mothers should
receive the first dose of the hepatitis B vaccine series at chrono-
logical age 1 month or at hospital discharge.169
Several studies suggest that the incidence of adverse events
after vaccination of preterm infants is the same as or lower than
that of full-term infants vaccinated at the same chronological
age.157,170,171 A temporal association between receipt of DTP and
Hib vaccine and a transient increase or recurrence of apnea in
premature infants has been reported, although the significance
of this finding is unclear.172
A preterm infant who is still hospitalized at age 2 months
can receive the vaccines routinely scheduled at that age.
However, in countries in which OPV is used, IPV may be con-
sidered for hospitalized infants. Because poliovirus vaccine
strains are excreted after receipt of OPV, IPV will decrease the

risk of transmission of vaccine viruses in the hospital.23,173
ACIP supports rotavirus vaccination of preterm infants accord-
ing to the same schedule and precautions as full-term infants
and under the following conditions: the infant's chronological
age meets the age requirements for rotavirus vaccine (eg, age
6-14 weeks and 6 days for dose 1), the infant is clinically sta-
ble, and the vaccine is administered at the time of discharge
from the neonatal intensive care unit [NICU] or nursery or
after discharge from the NICU or nursery. Although the lower
level of maternal antibody to rotavirus in very preterm infants
theoretically could increase the risk for adverse reactions from
rotavirus vaccine, ACIP believes the benefits of vaccinating an
infant when age-eligible, clinically stable, and no longer in the
hospital outweigh the theoretic risks.24
Breastfeeding and immunization
Neither inactivated nor live virus vaccines administered to a
lactating mother or infant who is breastfeeding have adverse
consequences.2,5 Because inactivated and component vaccines
do not multiply in the body, they pose no special risk for lac-
tating women or their infants. Lactating women also may
safely receive live virus vaccines, such as MMR, LAIV, OPV,
and varicella without interruption of their breastfeeding sched-
ule.5,45,173,174 Although vaccines that contain attenuated live
viruses or bacteria replicate in the vaccine recipient, most live
vaccine strains are not known to be secreted in human milk. An
exception is attenuated rubella vaccine virus, which has been
detected in human milk and recovered from the nasopharynx
and throat of some breastfed infants after maternal immuni-
zation.175,176 In one study, transient seroconversion to rubella
virus without evidence of clinical disease was noted in 25% of
the breastfed infants.175 Breastfed infants who acquired rubella
vaccine virus and rubella-specific antibodies from human milk
have a normal immune response to rubella vaccine adminis-
tered at 15 to 18 months of age.177
Breastfeeding of infants does not adversely affect their devel-
opment of a protective immune response and is not a contrain-
dication for any routinely administered vaccine.2,5 Yellow fever
vaccine should be avoided in breastfeeding women. However,
when nursing mothers cannot avoid or postpone travel to areas
endemic for yellow fever in which risk of acquisition is high,
they should be vaccinated.178 Compared with infants who are
formula fed, breastfed infants may have an enhanced immune
response to certain oral and parenteral vaccines, such as conju-
gate Hib vaccine, OPV, and DT.179-181 However, the significance
of such an effect is unclear.
Vaccination during pregnancy
Risk for a developing fetus from vaccination of the mother
during pregnancy primarily is theoretical. No evidence exists
of risk from vaccinating pregnant women with inactivated
virus or bacterial vaccines or toxoids.182,183 Live vaccines pose
a theoretical risk to the fetus. Benefits of vaccinating pregnant
women usually outweigh potential risks when the likelihood
of disease exposure is high, when infection would pose a risk
to the mother or fetus, and when the vaccine is unlikely to
cause harm.
Recommendations for vaccination during pregnancy can
be found in the annual US adult immunization schedule.55
Pregnant women should receive Td vaccine if indicated.
Previously vaccinated pregnant women who have not received a
Td vaccination within the last 10 years should receive a booster
dose. Pregnant women who are not immunized or only par-
tially immunized against tetanus should complete the primary
General immunization practices 8 105
series.184 Women for whom the vaccine is indicated but who
have not completed the recommended three-dose series dur-
ing pregnancy should receive follow-up after delivery to ensure
the series is completed. If a tetanus and diphtheria booster vac-
cination is indicated during pregnancy for a woman who has
previously not received Tdap (ie, 10 years since previous Td),
health-care providers should administer Tdap, preferably dur-
ing the third or late second trimester (ie, after 20 weeks’ ges-
tation).184 Use of pertussis vaccine in pregnant women may
theoretically lead to passage of higher levels of pertussis anti-
body across the placenta to the fetus and result in protection
from pertussis in the first few months of life when the disease
is most severe.185
Women in the second and third trimesters of pregnancy are
at increased risk for hospitalization from influenza. Therefore,
routine influenza vaccination is recommended for all women
who will be pregnant (in any trimester) during influenza sea-
son, usually November through March in the United States.25
IPV can be administered to pregnant women who are at risk for
exposure to wild-type poliovirus infection.186,187 Hepatitis B vac-
cine is recommended for pregnant women at risk for hepatitis
B virus infection.169 Hepatitis A, pneumococcal polysaccharide,
meningococcal conjugate, and meningococcal polysaccharide
vaccines should be considered for women at increased risk for
those infections.188-192 Pregnant women who must travel to
areas where the risk for yellow fever is high should receive yel-
low fever vaccine because the limited theoretical risk from vac-
cination is substantially outweighed by the risk for yellow fever
infection.178,193
Pregnancy is a contraindication for smallpox (vaccinia),
measles, mumps, rubella, and varicella-containing vac-
cines. Smallpox (vaccinia) vaccine is the only vaccine known
to cause harm to a fetus when administered to a pregnant
woman. In addition to the vaccinee herself, smallpox (vac-
cinia) vaccine should not be administered to a household con-
tact of a pregnant woman. Although of theoretical concern,
no cases of congenital rubella or varicella syndrome or abnor-
malities attributable to fetal infection have been observed
among infants born to susceptible women who received
rubella or varicella vaccines during pregnancy.45,194 Because
of the importance of protecting women of childbearing age
against rubella and varicella, reasonable practices in any vac-
cination program include asking women if they are pregnant
or might become pregnant in the next 4 weeks, not vaccinat-
ing women who state that they are pregnant, explaining the
theoretical risk for the fetus if MMR or varicella vaccine were
administered to a women who is pregnant, and counseling
women who are vaccinated not to become pregnant during the
4 weeks after MMR or varicella vaccination.45,194,195 Routine
pregnancy testing of women of childbearing age before admin-
istering a live virus vaccine is not recommended.45 If vacci-
nation of an unknowingly pregnant woman occurs or if she
becomes pregnant within 4 weeks after MMR or varicella vac-
cination, she should be counseled about the theoretical basis
of concern for the fetus; however, MMR or varicella vaccina-
tion during pregnancy should not be regarded as a reason to
terminate pregnancy.45,174
Persons who receive MMR vaccine do not transmit the vac-
cine viruses to contacts.45 Transmission of varicella vaccine
virus to contacts is rare.174 MMR and varicella vaccines should
be administered when indicated to the children and other
household contacts of pregnant women.45,174
All pregnant women should be evaluated for immunity to
rubella and varicella and be tested for the presence of HBsAg
in every pregnancy.45,169,174,195 Women susceptible to rubella
and varicella should be vaccinated immediately after delivery.
A woman found to be HBsAg-positive should be followed up
carefully to ensure that the infant receives HBIG and begins
the hepatitis B vaccine series no later than 12 hours after birth

ulcer. Examination by Gram stain or culture of the vesicular
fluid should confirm the diagnosis, but prior antibiotic therapy
quickly renders the infected site culture-negative. Biopsy at the
lesion edge, examined by Gram stain, immunohistology, and
PCR, may be useful in people who have been treated with anti-
biotics. In addition, there should be a history of exposure to
materials contaminated with .
The diagnosis of inhalational anthrax is difficult, but it
should be suspected in cases with a history of exposure to an
aerosol that contains , followed by an initial phase
during which the symptoms of inhalational anthrax are non-
specific. Once the acute stage develops, a widened mediastinum
seen on a chest radiograph, often with pleural effusions, should
suggest the diagnosis. In untreated cases, culture of blood and
pleural effusions will readily establish the diagnosis, and more
rapid detection methods, including PCR for the PA gene, immu-
noassay for PA and capsule antigens, and mass spectrometry for
LF, may all be useful.95 In cases previously treated with antibiot-
ics, these assays for toxin and capsule in blood and pleural fluid,
as well as immunohistologic examination of pleural fluid or
transbronchial biopsy specimens, are of particular value14,15,17,97
and may be detected as long as 6 days after therapy was begun.98
Because primary pneumonia is not a feature of inhalational
anthrax, sputum examinations do not aid diagnosis. The radio-
graphic differential diagnosis should include histoplasmosis,
sarcoidosis, tuberculosis, and lymphoma. A computed tomog-
raphy scan of the chest may be helpful to detect mediasti-
nal hemorrhagic lymphadenopathy and edema, peribronchial
thickening, and pleural effusions.
Gastrointestinal anthrax is difficult to diagnose because of
its rarity and similarity to other more common severe gastro-
intestinal diseases. An epidemiologic history of ingesting con-
taminated meat, particularly in association with other similar
cases, suggests the diagnosis. Microbiologic cultures are not
helpful in confirming the diagnosis unless bacteremia is pres-
ent. The diagnosis of oral-oropharyngeal anthrax can be made
from the clinical and physical findings. Adequate data are not
available to assess the value of bacteriologic cultures in con-
firming this diagnosis.
Treatment and prevention with antibiotics
Mild cases of cutaneous anthrax may be treated effectively
orally with a penicillin, a tetracycline, or another antibiotic,
depending on antimicrobial resistance. If spreading infection
or prominent systemic symptoms are present, high-dose paren-
teral therapy should be given until there is a clinical response.
Effective therapy reduces the edema and systemic symptoms
but does not change the evolution of the skin lesion.
Treatment of inhalational anthrax requires high-dose intra-
venous therapy (ciprofloxacin or another fluoroquinolone) with
at least one or more antibiotics with good central nervous sys-
tem penetration because of the high incidence of meningi-
tis.14,15,22,99,100 This approach applies to gastrointestinal anthrax
as well. Regimens should be altered based on susceptibility test-
ing and clinical status. The successful treatment of 6 of the
11 inhalational cases in the 2001 bioweapon attacks suggests
that, with rapid treatment with effective antibiotics and mod-
ern supportive care, including aggressive management of respi-
ratory distress and pleural effusions, mortality is similar to that
of other causes of sepsis.
After exposure to an aerosol, postexposure prophylaxis or pre-
symptomatic (ie, empiric) treatment should include oral antibi-
otics for 60 days or more, depending on individual circumstances
(eg, extent of exposure, vaccination status; see later text).99,100
The US Food and Drug Administration (FDA) confirms the
evidence for safety and efficacy of ciprofloxacin, doxycycline,
and penicillin G procaine for this indication,101 and levofloxa-
cin is indicated as a second-line drug because of fewer safety
Anthrax vaccines 10 131
data.100 Amoxicillin is recommended for children and pregnant
or lactating women, depending on microbial resistance.15,99,100
Preexposure or postexposure vaccination may enable shorter
courses of antibiotics (see later text).99,102 Postexposure vaccina-
tion alone would not be expected to protect quickly enough to
be effective.22
Epidemiology
Several theories explain the ecology of soil contaminated with
spores. One theory suggests that
spores can persist for many years in some types of soil under
certain conditions. These conditions are a soil rich in nitrogen
and organic material and with adequate calcium, a pH greater
than 6.0, and an ambient temperature greater than 15.5°C. It
remains unclear whether there are cycles of germination and
replication within the soil or if amplification within mamma-
lian hosts serves to maintain the spore population in the soil
between cases in animals.
Animal anthrax results from animals ingesting
spores by eating contaminated feed or while grazing on pas-
tures. Soil becomes contaminated from contaminated fertilizer
or contaminated feed spread on the ground or from diseased
animals that contaminate the soil with their secretions before
or after death.103
The number of reported human anthrax cases in the United
States has declined steadily since adequate surveillance data
have been available. Between 1916 and 1925, the annual aver-
age number of cases was 127; between 1948 and 1957, 44
cases; between 1978 and 1987, 0.9 case; and between 1988
and 2000, 0.25 case. Of the 235 human cases reported from
1955 to 2000, 20 were fatal (Figure 10-2).104 Among these cases,
224 had cutaneous lesions (118 on an arm, 65 on the head or
neck, 11 on the trunk, 8 on a leg, and 22 at an unknown site)
and 11 were inhalational cases. These represent the naturally
occurring cases and do not include cases related to the 2001
letter mailings. As discussed, in 2001, there were 22 cases, 11
inhalational including 5 deaths, and 11 cutaneous cases associ-
ated with contaminated mail. In the years from 2002 to 2010,
there have been five cases of anthrax reported in the United
States. Three of those cases were related to exposure to drums
with animal skin drum heads; two of these were inhalational
anthrax, and one was the first confirmed case of gastrointesti-
nal anthrax reported in the United States. The other two cases
were cutaneous.
A new manifestation of anthrax has been reported from
the United Kingdom in 2010, injectional anthrax. More than
50 cases with varying clinical signs and symptoms and high
mortality have been reported from Scotland and England, as
described. These cases have been associated with injecting
heroin contaminated with .19 It has not been
determined whether the heroin itself was contaminated or
was in the cutting material.
The traditional classification of cases is related to the source
of infection, that is, whether it is acquired in an industrial, an
agricultural, or a laboratory setting. The basic epidemiologic
principles are the same in developing and developed countries.
Agricultural anthrax is a more significant problem in devel-
oping countries,104a and industrial anthrax occurs more com-
monly in developed countries. Industrial anthrax results from
the exposure of susceptible persons to contaminated animal
products that include wool, goat hair, hides, or bones. These
materials come from animals that were infected with
before death or are contaminated after death (eg, from con-
taminated soil with which the carcass or animal products came
in contact). The wool and hair from infected animals may be
clipped from live animals or pulled from carcasses. A hide may
be obtained from an animal that has died of anthrax. Bones can
132 SECTION TWO Licensed vaccines

Number of human
anthrax cases and deaths in the
United States, 1916-2010. From
National Office of Vital Statistics and
Centers for Disease Control and
Prevention.104

be collected from grazing areas on which animals die or from Sources of Infection in 263 Cases of Human Anthrax in the
rendering plants that may handle carcasses of animals that United States, 1955 to 2010
have died of anthrax.
Wool and goat hair are processed into yarn that is used in
the textile and carpet industries and in the preparation of other
cloth-like materials. Hides are processed into leather goods.
Bones are used in preparing bone meal, gelatin, and fertilizer.
In industrial cases, cutaneous anthrax results from spores that
gain entrance through the skin by entering preexisting wounds
or by being rubbed through the skin or on a hair fiber that may
penetrate the skin. At times, the processing of goat hair and
wool creates infectious aerosols that may result in inhalational
anthrax when inhaled. A rendering plant is another source of
potential infection.
A case of inhalational anthrax was reported in early 2006
in a 44-year-old male drum-head maker who brought some
goat skins into the United States from Africa. After cleaning
the skins, he developed inhalational anthrax.27,97 With vigor-
ous treatment, he survived. was recovered from
his work environment. Several additional cases of anthrax asso-
ciated with drums from Africa have been reported, including
one case of disseminated anthrax and another of inhalational
t
8PSLJOH
JO
NBJMQSPDFTTJOH

anthrax from Europe32 and two cutaneous105 and one presumed
gastrointestinal case106 from the United States.
t
3FDFJWJOH
NBJM
DPOGJSNFE
PS

Cases associated with agricultural settings result from contact
QSFTVNQUJWF
with diseased animals or with the products of animals that have
died of anthrax. Affected persons are primarily agricultural work- t
8PSLJOH
JO
NBJMQSPDFTTJOH

ers, veterinarians, or people who kill and butcher infected animals
or butcher the carcasses of animals that have died of anthrax. t
3FDFJWJOH
NBJM
DPOGJSNFE
PS

This contact results in cutaneous anthrax or, if the infected meat QSFTVNQUJWF
is ingested, gastrointestinal or oral-oropharyngeal anthrax. 

DPOGJSNFE
BOE

Laboratory-associated cases of anthrax are rare. These are 
TVTQFDUFE
essentially all cutaneous, although a few inhalational cases
have occurred. Rarely, cases have been reported after contact
with contaminated clothing, such as woolen coats or pilots' courses of prophylactic antibiotics while potential exposures were
leather helmets. Table 10-2 presents the sources of infection evaluated,108 and among the treated people, more than 10,000
of the 263 cases reported in the United States from 1955 to continued to receive antibiotics for 60 or more days with or with-
2010. The two vaccine-associated cases of cutaneous agricul- out postexposure vaccination as prophylaxis.109,110 Exposures may
tural anthrax resulted from the inadvertent injection of animal have resulted from opening contaminated letters, from working
vaccine into the hand of the vaccinator. Two additional vaccine- in buildings with high-speed automated mail-sorting machines,
associated cases of anthrax occurred in persons producing vac- or through contact with cross-contaminated pieces of mail or
cine and were thought to result from contamination of skin environments contaminated with spores.
lesions with the vaccine strain.107
Exposures related to bioterrorist events represent a new cat-
egory. The anthrax spore attacks in the fall of 2001 resulted in 11 Passive immunization
confirmed inhalational cases and 7 confirmed and 4 suspected
cutaneous cases reported from Florida, New York, New Jersey, In the era before antibiotics, animal antisera were common
the District of Columbia, and Connecticut.13–15 Exposure to con- therapeutic products.111 One of the first was anthrax antise-
taminated mail was the confirmed or apparent source of infec- rum, developed in France by Marchoux and in Italy by Sclavo
tion in all patients.15,16 More than 32,000 people received short in 1895.112,113 Although it was used initially for prophylaxis and

treatment of anthrax among livestock, Sclavo later used his
product to treat human disease—cutaneous or septicemic. He
reported 10 deaths among 164 treated patients (6% mortality,
compared with the Italian case-fatality rate of 24%), although
these were not controlled clinical trials. Sclavo injected 30 to
40 mL of antiserum subcutaneously, repeated 24 hours later. In
severe cases, he also injected 10 mL or more intravenously.
Between the 1910s and 1940s, clinicians in Europe and the
Americas treated patients with anthrax antiserum using 25 to
300 mL daily for 5 days, sometimes in combination with arseni-
cals.111–122 One patient with severe cutaneous anthrax recovered
after receiving 2,265 mL of antiserum.123 No controlled studies
were performed to demonstrate efficacy. Anthrax antiserum for
therapy of cutaneous anthrax was superseded by therapy with
sulfanilamide, followed by penicillin and other antibiotics.121,124
Equine anthrax antiserum produced by live-spore vaccination
has been licensed in China,125 the Soviet Union, and later Russia
for decades, and its use continues, although the magnitude or
frequency of use is unclear. The Lanzhou Institute of Biological
Products in China developed a lyophilized antianthrax F(ab)2 for-
mulation of equine IgG fragments for human use by intracutane-
ous, intramuscular, or intravenous administration, but it is little
used (Dong Shulin, personal communication, 2002).
Experimental evidence indicates that passive immunization
with equine antibody produced against attenuated Sterne veter-
inary vaccine strains or against crude toxins prevents disease in
animals when given before or shortly after spore challenge.23,126
Rhesus monkeys could be protected with one or two doses of
equine antianthrax spore hyperimmune serum when begun
1 day after low-dose aerosol challenge. Of immune serum–
treated animals 45% survived, compared with 10% of control
animals.
More recent studies by Little et al127 showed efficacy of anti-
PA antiserum prophylaxis against an intramuscular challenge
in animals. The anti-PA polyclonal antibody protected against
death, and anti-PA monoclonal antibody significantly delayed
mortality. Reuveny and colleagues128 similarly found in passive
immunization studies of guinea pigs that polyclonal anti-PA
antisera conferred protection against an intradermal challenge
dose of 40 median lethal doses (LD50).
Kobiler and colleagues129 challenged guinea pigs intranasally
with a 25 LD50 dose of spores. The animals were then treated with
anti-PA, anti-LF, or anti-Sterne vaccine antibodies. Intraperitoneal
administration of rabbit anti-PA serum 24 hours after infection
protected 90% of infected animals, with lesser efficacy seen with
anti-Sterne and anti-LF antibodies. Beedham and colleagues130
demonstrated that mice could be protected against challenge
with a vaccine strain using serum, but not spleen lymphocytes,
from PA-vaccinated animals, supporting the long-standing evi-
dence that antibody is the major mechanism of vaccine-induced
immunity.
Although the importance of anthrax toxins in pathogenesis
suggests that antiserum may have a role in treatment, modern
Western interest in such products for human use was not rekin-
dled until the anthrax bioweapon attacks in the fall of 2001.131
Modern experimental evidence shows that passive immuniza-
tion with antiserum prevents anthrax in animals when given
before or shortly after spore challenge.132 This includes pro-
tecting guinea pigs from intradermal challenge, Rhesus mon-
keys from low-dose aerosol challenge, and pretreated rats from
parenteral challenge. While additional passive immunization
data in animal models are collected, the US Strategic National
Stockpile is storing 10,000 therapeutic courses of human poly-
clonal anthrax immune globulin. This product was manu-
factured by fractionating the plasma of volunteers previously
given at least four doses of anthrax vaccine adsorbed (AVA
[BioThrax]). The anthrax immune globulin formed part of the
successful treatment of a 2006 case of inhalational anthrax in a
44-year-old man exposed via African hides97 and has been used
Anthrax vaccines 10 133
in treatment of anthrax in drug users in Scotland,133 but defini-
tive evidence of efficacy in humans is unavailable.
Several human monoclonal anti-PA antibodies have since
been produced and are undergoing testing. The need for thera-
peutic tools other than antibiotics may be especially great in
the case of antibiotic-resistant strains of , although
there remains no definitive evidence to date of efficacy in
humans. The US government has contracted for 20,000 treat-
ment courses of raxibacumab (ABthrax, Human Genome
Sciences, Rockville, MD), a monoclonal anti-PA antibody. This
antibody has been shown to protect rabbits and nonhuman pri-
mates from death when given prophylactically against an aero-
sol spore challenge and to give a significant increase in survival
when given therapeutically.134 Other monoclonal antibodies
are also under development, such as MDX-1303 developed by
Medarex.135 But it is unclear whether, to counter bioweapons,
polyclonal antibodies that neutralize several epitopes might be
preferable to monoclonal antibodies that target only one epit-
ope.136 If bacteria could be reengineered to modify that single
epitope, a monoclonal antibody might be rendered ineffective.
Indeed, it has been reported that modifications to PA can result
in lack of effectiveness of a neutralizing monoclonal antibody,
whereas a polyclonal antibody was still active.137
Mouse and humanized chimpanzee monoclonal antibod-
ies to the other major virulence factor, the antiphagocytic poly-
-glutamic acid capsule, have also been shown to be protective
in mouse models.138,139
Active immunization
History of vaccine development
Although there is great historical interest in Pasteur's develop-
ment of the first effective live bacterial vaccine, and live attenu-
ated veterinary vaccines are still used, human vaccines against
anthrax consist of proteins purified from anthrax cultures,
except as indicated in the following discussion. Early human
anthrax vaccines (presumably live) were used in the 1910s but
found little favor.113 Sterne developed live attenuated strains in
the 1930s, which led to worldwide use for domesticated ani-
mals.43 Russian investigators developed similar vaccines for
animal and human use. In 1946, Gladstone140 identified the
PA component of cultures of as being an effective
vaccine, and only later did subsequent studies identify PA as
the cell-binding component of the anthrax toxins. Belton and
Strange126 increased the yields of PA to allow large-scale pro-
duction, leading to the current British vaccine. Wright and col-
leagues141–143 used similar techniques to develop the precursors
to the American vaccine.
Properly, protective antigen is the term applied to one of
the toxin proteins, which is the plasmid-encoded binding com-
ponent of the anthrax toxins described previously. The major
effective immunogen in culture supernatants is the PA com-
ponent of the toxins, although smaller amounts of LF and EF
may be present; their contribution to protective immunity has
remained controversial.52 In older studies, EF enhanced the
protective efficacy of PA in some experimental animals.144,145
The results of these studies are difficult to interpret because
the preparations used may not have been pure and free of cross-
contamination. Studies using the PA gene cloned into
demonstrated conclusively that PA alone, in the absence
of EF, LF, or other proteins, protects animals
against experimental infection.146 Although other experiments
showed that purer preparations of PA, free of immunologically
detectable LF or EF,147 or recombinant PA,148 can protect experi-
mental animals, it remains unknown whether adding EF or LF
enhances the vaccine efficacy of PA. One study reported that a
DNA vaccine with the N-terminal domain of LF alone delayed
134 SECTION TWO
time to death but did not increase overall survival in rabbits
against a virulent strain.149 Similar results were observed with
anti-LF serum in guinea pigs,129 although antibody to LF150 and
vaccination with inactivated LF or the N-terminal domain151
can protect mice from infection with an attenuated strain.
Some protection in a similar mouse model was observed with
EF expressed in an adenovirus vector.152
Description of vaccines
The human anthrax vaccine licensed in the United States, AVA
marketed as BioThrax, is produced by Emergent BioSolutions
(Rockville, MD) from sterile filtrates of microaerophilic cultures
of an attenuated, unencapsulated, nonproteolytic strain (V770-
NP1-R) of . The cell-free culture filtrate, thought to
contain predominantly PA, is adsorbed to aluminum hydroxide,
and the final product contains no more than 2.4 mg of alumi-
num hydroxide per 0.5-mL dose. Formaldehyde, in a final con-
centration of no more than 0.02%, and 0.0025% benzethonium
chloride are present as preservatives. Current product-content
standards require 5 to 20 g/mL of total protein, of which at
least 35% is the 83-kDa PA protein, measured by densitometric
analysis on sodium dodecyl sulfate–polyacrylamide gel electro-
phoresis after pooling 12 sublots.153
Some lots produced in the 1980s seemed to contain small
amounts of LF and lesser amounts of EF, as determined by
induction of antibody responses in animal recipients,46,147,154,155
although this has not been reported in the limited observa-
tions in human vaccinees.155 Analysis found no detectable EF
by Western blotting. Enzyme-linked immunosorbent assay
(ELISA) studies found LF to be present in the range of 10 to
30 ng/mL of fermentation filtrate before adsorption.153 Analysis
by mouse macrophage cytotoxicity assay suggested that LF is
present in a biologically inactive form.153 Although it is clear
that PA by itself is an effective immunogen, it remains unre-
solved whether the small amounts of LF or EF that may be
present in some lots of the vaccine contribute to the vaccine's
protective efficacy. A more recent study of serum samples of
AVA vaccinees showed that low levels of antibody to LF and EF
were present by Western blot, but these did not contribute to
toxin neutralization.156
Potency testing of the AVA is performed by assessing biologic
activity after intradermally challenging vaccinated guinea pigs
with a lethal dose of spores. The vaccine is stored at 2°C to 8°C.
In December 2008, the FDA approved a revised recommended
schedule for vaccination consisting of 0.5 mL given intramuscu-
larly at 0 and 4 weeks and 6, 12, and 18 months instead of the
previous schedule with doses given subcutaneously at 0, 2, and
4 weeks followed by doses at 6, 12, and 18 months.102 Studies of
immunogenicity with even fewer doses are underway. With con-
tinued exposure, additional yearly boosters are recommended.
The vaccine is stable for 3 years after a successful potency test.
Anthrax Vaccine Precipitated, a similar vaccine produced
by the Health Protection Agency (Porton Down, Salisbury,
Wiltshire), was developed in the United Kingdom, first admin-
istered to humans in the early 1950s, and licensed there in
1979.157–162 This vaccine is made by precipitating the sterile
cell-free culture filtrate of a derivative of the attenuated, unen-
capsulated Sterne strain 34 F2 with aluminum potassium sul-
fate.161 LF and EF are present in this vaccine at levels higher
than believed to be found in lots of the US vaccine from the
1980s.157,162 The vaccine contains thimerosal as a preservative.
The British vaccine is administered intramuscularly in a regi-
men of three 0.5-mL doses at 0, 3, and 6 weeks, with a booster
dose 6 months after the third dose. Subsequent booster doses
are given annually.155 Other schedules are being studied.
A vaccine consisting of a suspension of live spores, named STI-1
for the Sanitary-Technical Institute, has been used for humans
in the Soviet Union and its subsequent independent republics
since 1953.103,163 This strain, similar to the Sterne strain used in
veterinary vaccines, is unencapsulated.163 Although this vaccine
has a reputation for causing substantial side effects, its develop-
ers assert that it is reasonably well tolerated and shows some
degree of protective efficacy.163,164 This vaccine, manufactured by
the Tbilisi Scientific Research Institute of Vaccines & Serums
(Tbilisi, Georgia), the Institute of Microbiology (Kirov [Viatka],
Russian Federation), and perhaps at other locations, is given by
scarification through a 10- to 20- L drop of vaccine containing 1.3
to 4 108 spores or subcutaneously.2,103,161,163,165–167 The ini-
tial dose is followed by a second dose 21 days later, with yearly
boosters.
Another live spore human vaccine given by scarification
has been manufactured by the Lanzhou Institute of Biological
Products (Lanzhou, Gansu, People's Republic of China) since
the 1960s, based on the avirulent strain A16R.85,104 A single
dose contains 1.6 to 2.4 10 8 colony-forming units. A single
booster dose is given 6 to 12 months after the first vaccination
(Dong Shulin, personal communication, 2002).
Immunogenicity of vaccine
The results of two studies indicated that immunization with the
licensed US vaccine induced an immune response (as measured
by indirect hemagglutination) to PA. In the first study, 83% of
vaccinees responded 2 weeks after the first three doses.168 In the
other, 91% responded after receiving two or more doses.169 The
titers fell over time, but 100% of vaccinees responded with an
anamnestic response to the annual booster dose. This hem-
agglutination assay correlated with results obtained by using
an ELISA against PA,170 which is the current test of choice.
Analysis using a more sensitive ELISA against PA demonstrated
that seroconversion occurs in 96% to 100% of vaccinees after
the second dose.171
By using a more sensitive validated ELISA assay, Pittman
and colleagues172 found that one dose of AVA evoked detectable
anti-PA IgG antibodies in 60% to 84% of vaccinees. After two
doses, 95% to 100% of vaccinees developed anti-PA antibodies.
Prolonging the interval between the first two doses did not impair
booster responses among Persian Gulf War troops given anthrax
vaccine after gaps of 18 to 24 months.173 A pilot study comparing
subcutaneous and intramuscular administration of AVA revealed
higher titers to PA when a 4-week interval between the first
two doses was compared with a 2-week interval, with fewer
injection-site or systemic events with the intramuscular route.171
As described earlier, based on a larger randomized clinical
trial,174 the FDA changed the licensed schedule to 0.5 mL given
at 0 and 4 weeks and 6, 12, and 18 months by the intramuscu-
lar route from the old schedule of 0, 2, and 4 weeks and 6, 12,
and 18 months given subcutaneously. While the change in route
from subcutaneous to intramuscular resulted in fewer solicited
injection-site adverse events, it had no effect on the occurrence
of systemic adverse events. Most important, in contrast to the
pilot study, the change in route and elimination of the week 2
dose resulted in statistically significantly lower concentrations
and titers of anti-PA antibody from week 8 through and includ-
ing month 6. The antibody concentration and titers using the
new schedule were noninferior to the old schedule at month 7,
only after the 6-month dose. However, the percentage of respond-
ers with a fourfold rise in titer was similar at week 8 and month
7 after the 6-month booster with the two schedules. The sig-
nificance of the lower immune response up to month 7 remains
unclear as the serologic correlation with immunity, while under
study, remains to be determined and will require extrapolation
from studies in animal models.
There also seems to be some variability in the response
of humans to vaccination with AVA, and it has recently been
reported that persons with DRB1-DQA1-DQB1 HLA class II hap-
lotypes produce significantly lower levels of anti-PA antibody.175

Soviet scientists developed a skin-test antigenic reagent known
as anthraxin in 1957, derived from the edematous fluid of
infected animals given an unencapsulated strain.167
Licensed there in 1962, the skin test product is an autoclaved
liquid composed of an undefined heat-stable polysaccharide-
protein-nucleic acid complex, without anthrax capsular or
toxigenic material.167,176 A positive skin test after a 0.1-mL
intradermal injection is defined as erythema of 8 mm or more
with local induration persisting for 48 hours.167 Anthraxin dem-
onstrated usefulness in identifying cases of anthrax163,177 and
identifying STI-1 vaccine-induced immunity in guinea pigs,
sheep, and humans.163 Experimental data show that guinea pigs
vaccinated against anthrax that developed a positive anthraxin
skin test were immune to a subsequent parenteral challenge.167
Positive and negative predictive values of individual test results
have not been published. There is limited experience with the
skin-test antigen in humans in Western countries, and its use-
fulness in predicting immunity in humans remains unknown.
After a naturally acquired infection, antibody to PA develops
in 68% to 93% of cases as reported in different series, depend-
ing on the time when samples are drawn.162,169,170,178 Antibody
to LF occurs in 42% to 55% of cases, whereas antibody to EF is
less frequently observed.162,170 Antibody to the anthrax capsule
occurs in 67% to 94% of cases.170,178 This reaction contrasts with
that of vaccinees, in which no response to capsule is expected
because the strain used to produce the vaccine is nonencapsu-
lated. In the 2001 epidemic of inhalational anthrax, antibody to
PA was detected in all survivors of confirmed cases.
In experimental animals, there is generally a correlation
between immunity and antibody titer to PA after immunization
with AVA.179 However, the live veterinary vaccine provides signifi-
cantly greater protection against similar challenge doses of anthrax
in experimentally infected animals than does the human vaccine,
even though it often induces lower levels of antibody to PA,146,154,155
suggesting that other antigens may be involved in protection.
More recent studies using live and protein-based vaccines
have demonstrated a strong correlation between antibodies
to PA and immunity. Barnard and Friedlander180 showed, for
the first time using live vaccines producing varying amounts
of PA, that protection was strongly correlated with antibody
titers to PA, a finding subsequently confirmed by Cohen and
colleagues.181 Pitt and colleagues,179 using AVA, found a simi-
lar in vitro correlation of immunity with antibody to PA, mea-
sured by ELISA and toxin neutralization, in a rabbit model of
inhalational anthrax. Reuveny and colleagues,128 using a PA vac-
cine to protect guinea pigs against an intradermal challenge,
found that toxin-neutralizing antibodies correlated better with
survival than did antibodies measured by ELISA, and similar
results were observed in rabbits challenged by the intranasal
route.182 Further analysis of the antibody response to different
epitopes on PA will increase our knowledge of the nature of the
protective antibodies as monoclonal antibodies to PA with syn-
ergistic, additive or antagonistic effects on neutralization have
been described.182a The overall evidence based on passive immu-
nity with the aforementioned antibodies and protection with PA
vaccines confirms that humoral immunity is the predominant
mechanism of protection. Vaccine-induced protection seems in
some animal models to be a function of the rapidity of the infec-
tious process, the level of antibody present at challenge, and the
speed of development of an anamnestic response.183
More recent studies sponsored by the National Institute of
Allergy and Infectious Diseases (NIAID) and the Centers for
Disease Control and Prevention (CDC) to further develop the
relationship between antibody response and protection in rabbits
(NIAID) and in nonhuman primates (NIAID and CDC) have
Anthrax vaccines 10 135
been conducted. Although the NIAID used second-generation
anthrax PA vaccines (see later text) and the CDC used the
licensed AVA vaccine, studies using both types of vaccines have
shown that antibody levels in both species correlate with pro-
tection against aerosol challenge. The approach under develop-
ment is to extrapolate the results of animal efficacy studies to
determine the levels of antibody that are predictive of efficacy
in humans. It is hoped this approach will be suitable for licen-
sure of new anthrax vaccines and/or indications using the FDA
Animal Rule developed for licensure of vaccines that cannot be
tested for efficacy in humans.184
The protective efficacy of different experimental PA-based vac-
cines derived from culture filtrates of has been
clearly demonstrated with the use of various animal models and
routes of challenge.52,160 A comprehensive, peer-reviewed evalu-
ation by the National Academy of Sciences reported that “The
committee finds that the available evidence from studies with
humans and animals, coupled with reasonable assumptions
of analogy, shows that AVA as licensed is an effective vaccine
for the protection of humans against anthrax, including inha-
lational anthrax, caused by all known or plausible engineered
153
strains of ”. The FDA independently affirmed
that AVA prevents anthrax regardless of route of exposure.185
A controlled field trial was conducted in the late 1950s with
a less potent vaccine similar to the currently licensed AVA.186
This vaccine was composed of an alum-precipitated, cell-free
culture supernatant from an attenuated, unencapsulated, non-
proteolytic strain of . This strain differed slightly
from that used to produce the licensed vaccine and was grown
under aerobic rather than microaerophilic conditions.142 The
study was conducted in a susceptible population working in four
mills in the northeastern United States, where raw imported
goat hair contaminated with was used. The results
indicated that vaccination, compared with inoculation with a
placebo, provided 92.5% protection against anthrax, combining
the cutaneous and inhalational cases (95% confidence interval
[CI], 65%-100%). No isolated assessment of the effectiveness of
the vaccine against inhalational anthrax could be made because
there were too few cases, although the only inhalational cases
observed occurred in nonvaccinated persons.
There have been no controlled clinical trials in humans of the
efficacy of the currently licensed US vaccine, although the differ-
ences between the AVA vaccine and the PA-based vaccine used in
the study by Brachman et al186 are minor from a regulatory per-
spective.153 The AVA vaccine has been tested extensively in ani-
mals and has protected guinea pigs against intramuscular154,155 and
aerosol150 challenge. More recent experiments show that this vac-
cine also protected rhesus monkeys against a lethal aerosol chal-
lenge with anthrax spores.148,187–190 Inhalational challenge studies
in nonhuman primates vaccinated with the licensed human vac-
cine or a recombinant PA vaccine are summarized in Table 10-3.
Additional studies with various formulations of recombinant PA
by different manufacturers have confirmed its efficacy in nonhu-
man primates (E. Nuzum, personal communication, 03.18.11).
The duration of immunity induced by vaccination has not been
clearly established. In the field trials that evaluated a vaccine simi-
lar to the currently licensed AVA, one case of cutaneous anthrax
occurred 5 months after receipt of three doses within 4 weeks and
just before the scheduled 6-month boosting dose.186 Although data
are insufficient to support any firm conclusions, this observation
suggests that the immunity induced by the initial dosing series
of the current vaccine may not be long lasting. Ongoing stud-
ies of clinical correlates of protection that involve reducing the
total number of vaccine doses by using longer intervals between

Diagnosis of meningitis requires analysis of a CSF sample,
which is usually obtained by lumbar puncture. With appropri-
ate transport and processing of specimens, analysis of CSF is
a highly sensitive method for the diagnosis of Hib meningitis.
However, in areas where appropriate laboratory equipment is
not readily available, diagnosis of Hib meningitis can be diffi-
cult. A recent review and meta-analysis estimated a global Hib
meningitis incidence rate of 31 per 100,000 (uncertainty range,
16-39 per 100,000) in children younger than 5 years in 2000.1
A study in Indonesia showed a vaccine-preventable, laboratory-
confirmed Hib meningitis incidence rate of 16 (95% CI, 1.4-31)
per 100,000 children younger than 2 years, but the rate of men-
ingitis with CSF findings consistent with a bacterial cause (not
necessarily culture-confirmed) prevented by Hib vaccine was 67
(95% CI, 22-112) per 100,000 children younger than 2 years.98
Thus, laboratory methods identified less than 30% of probable
Hib meningitis cases in this study.
Acute epiglottitis is the swelling and inflammation of the epiglot-
tis and surrounding structures. Classic symptoms include sore
throat, dysphagia, stridor, and high fever. If appropriate manage-
ment, including airway establishment and antibiotic therapy, is
not instituted, the disease can progress rapidly to airway obstruc-
tion and death. In the prevaccine era, Hib was responsible for 75%
to 90% of epiglottitis cases in US children.111 Epiglottitis also was
a common manifestation of Hib disease in other developed coun-
tries.112,113 Rates in Scandinavia were extremely high; annual rates
in Sweden ranged from 4.5 to 32 per 100,000, with the highest
rates in children 3 to 4 years old.113 However, epiglottitis was rare
among indigenous populations in developed countries, including
Australian Aboriginals and Alaskan/Canadian Inuit,114,115 and in
most developing countries.116 The reason for this geographic dif-
ference is not clear but may relate to age at exposure. In popula-
tions with low epiglottitis incidence, the majority of serious Hib
disease cases occurred in children younger than 1 year, suggesting
an early age of exposure to Hib.
Septic arthritis is a result of an organism invading the syno-
vial fluid and most commonly occurs in the lower extremities.
Common clinical manifestations include joint pain, local swell-
ing and erythema, limited range of motion, and fever. Sequelae
are fairly common, including limitation of motion, limping
gait, limb-length discrepancy, and abnormal bone growth.117,118
In the prevaccine era, Hib was the most common cause of sep-
tic arthritis in children younger than 2 years in the United
States.119 According to one study in Canada conducted before
the availability of Hib vaccines, Hib was identified in 41% of
cases of culture-positive septic arthritis.120 Hib was a less com-
mon cause of osteomyelitis in the prevaccine era. An estimated
5% of culture-positive osteomyelitis was caused by Hib.120
Before conjugate vaccine introduction, Hib was a common
cause of orbital, periorbital, and facial cellulitis.121 Periorbital
cellulitis due to Hib was associated with bacteremia in up to
80% of cases, which was often associated with meningeal seed-
ing and subsequent intracranial infection.121-123 Hib was also a
cause of facial cellulitis, including a syndrome of buccal cellu-
litis accompanied with bacteremia. In one series of facial cellu-
litis cases in the United States before the use of Hib vaccines,
82% were attributed to Hib.124
While non–type b encapsulated can cause menin-
gitis and sepsis, the overall disease invasive disease burden due
to non–type b encapsulated is low in relation to
Haemophilus influenzae vaccines 13 171
that caused by the type b strains. In the United States between
1998 and 2000, the rate of non–type b inva-
sive disease was 0.8 cases per 100,000 children younger than
5 years.125
Nontypeable strains are ubiquitous colonizers
of the respiratory tracts of children and adults.126,127 Studies
have documented pharyngeal carriage rates of around 50% in
children younger than 2 years.128 Nontypeable
is a well-known cause of otitis media, sinusitis, and pneumo-
nia.129,130 Rarely, the organisms can cause invasive disease as
well, particularly in children younger than 4 years and people
with underlying serious medical conditions, especially immu-
nocompromised people.127,131,132
Treatment and antibiotic resistance
Treatment of Hib disease requires early assessment and iden-
tification of the illness, administration of appropriate antibi-
otic therapy, and supportive management of sequelae. Therapy
should be based on patterns of antibiotic resistance in the
community. In the United States, the American Academy of
Pediatrics (AAP) recommends that initial empiric therapy for
suspected Hib meningitis should be cefotaxime or ceftriaxone.
Meropenem or the combination of ampicillin and chloram-
phenicol are alternative regimens.133 The WHO recommends
initial therapy with chloramphenicol in combination with
ampicillin or benzylpenicillin for suspected Hib meningitis.134 If
the cause is unknown, some experts recommend the addition of
vancomycin to the empiric regimen owing to widespread resis-
tant pneumococci.133 Subsequent therapy should be guided by
the antibiotic susceptibility profile of the organism and national
guidelines.
Chemoprophylaxis of household contacts is recommended
in certain cases because of the occurrence of secondary infec-
tions. In the prevaccine era, there were several documented
cases of Hib in close contacts of index cases, such as co–day-
care attendees and household contacts.57,135,136 Rates of second-
ary disease were documented to be around 0.6% in one study,
with contacts younger than 2 years at highest risk.137,138 One
national study showed that for Hib meningitis in the 30 days
following initial illness, the risk for household contacts was
585 times the risk in the age-adjusted general US population.139
Currently, chemoprophylaxis with rifampin is recommended
for all household contacts of a case of Hib if the household has a
child younger than 12 months who has not received the primary
series, a child younger than 4 years who is incompletely immu-
nized, or an immunocompromised child. Day-care contacts
should receive prophylaxis only if two or more cases of invasive
Hib disease have occurred within 60 days.133 Antibiotics that are
usually used for the treatment of meningitis often do not elimi-
nate Hib from the upper airway; therefore, the index case may
require a different antibiotic to eliminate carriage.140
Between 1972 and 1974, the first antibiotic-resistant Hib
isolates were reported in Europe and the United States. Since
then, the proportion of resistant strains has increased across
the world.141 Studies have shown that 20% to 60% of isolates
produce -lactamase and are resistant to ampicillin.55,142 Other
antibiotics to which resistance has been shown include chlor-
amphenicol, tetracycline, and trimethoprim-sulfamethoxazole
(TMP-SMX). Emerging resistance to the cephalosporins has
been shown; in Mali, of 207 Hib isolates from blood, 0.5%
were resistant to ceftriaxone.143,144 A study in Korea showed
that of 55 Hib strains, 15% had only intermediate suscepti-
bility to cefprozil and cefaclor.144 Multidrug-resistant Hib has
also been demonstrated. In Bangladesh, 31% of Hib isolates
were multidrug-resistant.145 In Spain, 11% of isolates in one
study were resistant to multiple antibiotics, with 3% of strains
172 SECTION TWO Licensed vaccines
resistant to ampicillin, tetracycline, and chloramphenicol.146 In
a study from Kenya looking at Hib susceptibility to amoxicil-
lin, chloramphenicol, and TMP-SMX, of a total of 236 blood
or CSF isolates from children admitted with meningitis or
sepsis, 40% were resistant to at least two antibiotics and 28%
were resistant to all three antibiotics. In addition, resistance
to these three antibiotics increased significantly during the
9 years of the study.147 Surveillance studies in Latin America
demonstrated that 21% of invasive (type b and
non–type b) isolates were -lactamase producers, 21% to 32%
were ampicillin-resistant, and 26% to 49% were TMP-SMX–
resistant.148,149 Significant levels of resistance to ampicillin,
chloramphenicol, and TMP-SMX have also been seen in Asian
and African countries.145,150 The problem of antibiotic resistance
is particularly challenging in the developing world because Hib
is often resistant to the antibiotics commonly used for empiric
therapy and antibiotics that remain effective against Hib are
expensive or unavailable. Although judicious use of antibiotics
and continued development of new antimicrobial drugs should
remain priorities, vaccination remains the most powerful tool
for reducing the burden of antibiotic-resistant Hib disease.
Epidemiology in the prevaccine era
(Box 13-2)
Statistics
Before the routine use of Hib conjugate vaccines, more than
95% of all invasive disease was due to sero-
type b.85 The annual incidence of invasive disease due to Hib
in the general US population was estimated at 20 to 88 per
100,000 children younger than 5 years,151-153 with approxi-
mately 20,000 recognized cases each year, more than 50% of
which were cases of meningitis.154 Reported rates of invasive
disease in Europe varied. Studies in Spain and France showed
invasive disease rates of 12 and 21 cases per 100,000, respec-
tively.155,156 Reported rates in Scandinavia were higher; the inci-
dences reported from Finland and Sweden were 41 and 54 cases
per 100,000 children younger than 5 years, respectively.155,157 It
is not clear whether these differences in reported incidence rates
reflect differences in surveillance methods or true differences in
disease incidence.1
Certain US populations were at higher risk of Hib disease.
For example, the incidence of invasive Hib disease in Navajo
and Apache children was found to be 152 and 250 per 100,000
children younger than 5 years, respectively.50,158 Alaskan
Box 13-2
,FZ
QPJOUT
FQJEFNJPMPHZ
JO
UIF
QSFWBDDJOF
FSB
t
#FGPSF
SPVUJOF
VTF
PG
)JC
DPOKVHBUF
WBDDJOFT
NPSF
UIBO


PG
JOWBTJWF

EJTFBTF
XBT
EVF
UP

serotype b.
t
5IF
BOOVBM
JODJEFODF
PG
)JC
JOWBTJWF
EJTFBTF
JO
UIF
HFOFSBM

64
QPQVMBUJPO
XBT
FTUJNBUFE
BU

UP

DBTFT
QFS
 

DIJMESFO
ZPVOHFS
UIBO

ZFBST
CFGPSF
)JC
DPOKVHBUF
WBDDJOF

introduction.
t
:PVOH
BHF

ZFBST
JT
UIF
QSFEPNJOBOU
SJTL
GBDUPS
GPS

JOWBTJWF
)JC
EJTFBTF
t
0UIFS
SJTL
GBDUPST
GPS
JOWBTJWF
)JC
EJTFBTF
JODMVEF
CFMPOHJOH

UP
DFSUBJO
SBDJBM
PS
FUIOJD
HSPVQT
MPXFS
TPDJPFDPOPNJD
TUBUVT

MJWJOH
JO
DSPXEFE
DPOEJUJPOT
BOE
IBWJOH
DPODPNJUBOU
TFSJPVT

NFEJDBM
DPOEJUJPOT
QBSUJDVMBSMZ
)*7
JOGFDUJPO
t
5IF
8)0
FTUJNBUFT
UIBU
JO

CFGPSF
XJEFTQSFBE
WBDDJOF

VTF
)JC
XBT
BTTPDJBUFE
XJUI
NPSF
UIBO

NJMMJPO
TFSJPVT

JMMOFTTFT
BOE
 
EFBUIT
QFS
ZFBS
5IF
NBKPSJUZ
PG
UIFTF

PDDVSSFE
JO
SFTPVSDFMJNJUFE
TFUUJOHT
XIFSF
UIF
)JC
DPOKVHBUF

WBDDJOFT
XFSF
OPU
JO
SPVUJOF
VTF
Eskimos were documented to have rates of invasive Hib disease
of 491 per 100,000 children younger than 5 years.159 Australia
reported similar population differences in rates of Hib inva-
sive disease. Nonindigenous children had an incidence rate for
Hib invasive disease ranging from 33 to 60 cases per 100,000
children younger than 5 years, while indigenous children were
found to have rates as high as 500 per 100,000 children younger
than 5 years.160
Studies in Africa have consistently found high rates of Hib
disease. For example, in Uganda and The Gambia, the incidence
rates of Hib meningitis in children younger than 5 years were 88
and 60 per 100,000 before vaccine introduction, respectively.161
Significant rates of Hib disease have also been documented in
the Middle East and the Pacific island countries.162-164
In Asia, data on Hib disease burden have emerged more
recently.165-167 A recent review showed that Hib was identified
in 60% of Asian studies that reported the etiology of bacterial
meningitis and that the incidence rate of Hib meningitis ranged
from 0.98 to 28 per 100,000 children younger than 5 years.166
A surveillance study of children younger than 5 years in Sri
Lanka showed Hib was responsible for 50% of the 108 men-
ingitis cases in which an etiology was identified.168 A prospec-
tive population-based surveillance study in India showed an
Hib meningitis incidence of 7.1 per 100,000 (95% CI, 3.1-14.0)
children younger than 5 years.169 An expert panel that reviewed
the literature on Hib disease in Asia concluded that many stud-
ies underestimated the true incidence of Hib because of antibi-
otic use before diagnostic testing, delayed contact with health
providers, low rates of lumbar puncture, and inadequate spec-
imen processing.165 Another possible explanation for the low
observed incidence of Hib meningitis in some countries is that
widespread antibiotic use may actually change the epidemiol-
ogy of Hib disease by altering transmission patterns or the pro-
gression of disease. In an effort to address these issues, vaccine
probe studies were carried out in Indonesia and Bangladesh.98,99
In Indonesia, Hib conjugate vaccine prevented 67 (95% CI,
22-112) cases of probable bacterial meningitis and 158 (95%
CI, 42-273) cases of clinically diagnosed meningitis or seizures
per 100,000 child years in children younger than 2 years.98 Of
note, the incidence of laboratory-confirmed Hib meningitis
was only 16 per 100,000 child years in children younger than
2 years, suggesting that even under clinical trial conditions, lab-
oratory confirmation of Hib meningitis is difficult. As noted,
the burden of Hib pneumonia is difficult to measure, but sev-
eral studies have shown that Hib conjugate vaccine prevents a
significant proportion of clinical pneumonias and pneumonias
with radiographic consolidation in children.1,98,99 The WHO
estimated that in 2000, before widespread global vaccine intro-
duction, Hib caused approximately 8 million cases of serious
Hib disease and 371,000 deaths annually.1
Risk factors
Age is a major risk factor for invasive Hib disease, with children
younger than 2 years at highest risk. Of note, the incidence of
invasive Hib disease in children younger than 2 months is low,
presumably owing to protection from transplacentally derived
maternal antibodies. The age at which peak attack rates occur
varies by geographic location. Figure 13-2 shows the distribu-
tion of Hib meningitis cases by age in different geographic loca-
tions. In Western Europe, the majority of cases occurred in the
second year of life, and fewer than 20% of cases had occurred by
6 months of age,170 while in indigenous and developing country
populations, the majority of cases occurred in the first year of
life, with 30% to 50% of cases having occurred by 6 months of
age.50,115,150,171 The US general population age distribution was
intermediate.116,170

meningitis cases Hib meningitis cases
$VNVMBUJWF
QSPQPSUJPO
PG
)JC
NFOJOHJUJT
DBTFT
CZ
BHF

BOE
HFPHSBQIJD
MPDBUJPO

4PVSDFT

$VNVMBUJWF
QSPQPSUJPO
PG
DBTFT
PG


NFOJOHJUJT
BU
TVDDFTTJWF
BHFT
MFTT
UIBO


NPOUIT
GPS
DIJMESFO

JO
4BOUJBHP
$IJMF
'SPN
'FSSFDDJP
$
0SUJ[
&
"TUSP[B
-
FU
BM
"
QPQVMBUJPO
CBTFE
SFUSPTQFDUJWF
BTTFTTNFOU
PG
UIF
EJTFBTF
CVSEFO
SFTVMUJOH
GSPN
JOWBTJWF


JO
JOGBOUT
BOE
ZPVOH
DIJMESFO
JO
4BOUJBHP
$IJMF

1FEJBUS
*OGFDU
%JT
+


1BHF


$VNVMBUJWF
EJTUSJCVUJPO
PG

)JC
NFOJOHJUJT
DBTFT
CZ
BHF
JO
5IF
(BNCJB
BOE
'JOMBOE
'SPN
8BUU
+1
-FWJOF

04
4BOUPTIBN
.
(MPCBM
SFEVDUJPO
PG
)JC
EJTFBTF
XIBU
BSF
UIF
OFYU
TUFQT

1SPDFFEJOHT
PG
UIF
NFFUJOH
4DPUUTEBMF
"SJ[POB
4FQUFNCFS


+
1FEJBUS

 
TVQQM 44

1BHF
4
The risk of invasive Hib disease has been reported to be ele-
vated in several ethnic groups. In the prevaccine era, African
American children in the United States had rates of Hib menin-
gitis that were up to fourfold higher than in white children.172-174
Native American populations in Alaska and the continental
United States also have been shown to have increased inci-
dence of invasive Hib disease.50,51,115,158 Whether differences in
rates between various ethnic groups are due to biological differ-
ences in the host or other factors remains unclear.175 One study
found that after controlling for various sociodemographic fac-
tors, African Americans did not have a higher rate of Hib dis-
ease than other races.175 Some studies have looked for genetic
markers that may be associated with increased disease risk in
certain communities.176,177 Genes that increased decreased
susceptibility to Hib disease have been identified. For example,
in Alaskan Eskimos, children with a Gm allotype were found
to have increased susceptibility to invasive disease if associated
with the HLA-DR8 allele but a decreased susceptibility if asso-
ciated with the HLA-DR5 allele.178
Several factors that are often used as surrogate markers for low
socioeconomic status have been associated with increased risk
of Hib disease, including low household income,172 low paren-
tal education,179 and large household size.180,181 The indepen-
dent effect of each of these and other related risk factors has
been difficult to isolate. In a multivariate analysis, Cochi et al175
showed that household crowding was associated with a 2.7-
times increased odds of Hib disease (95% CI, 1.3-5.6). Other
investigators have also found that household crowding resulted
in increased risk of Hib disease.182-184 An increased number of
Haemophilus influenzae vaccines 13 173
siblings, particularly of preschool and elementary school age,
and the presence of extended family groups has been shown to
be associated with increased risk of Hib disease.173,180,181,185 Day-
care attendance has also been associated with increased risk
of invasive Hib disease, particularly in children younger than
2 years.88,137,175,186 Hib has been shown to survive for up to 48
hours on toys that were placed in the mouths of children who
were asymptomatic Hib carriers; therefore, communal toys may
serve as fomites for the transmission of Hib.187
Serious medical conditions, particularly conditions resulting
in immunosuppression, are also recognized risk factors for
Hib disease. Hemoglobinopathies,188 complement deficiency,189
antibody deficiency,190 and asplenia191 have all been associated
with increased risk of Hib disease. The incidence and severity
of Hib disease is also higher in adults and children with human
immunodeficiency virus (HIV) infection.192,193 Antecedent viral
respiratory infections have been associated with increased Hib
carriage47 and increased susceptibility to Hib meningitis.59 This
could be due to the enhanced transmission of virus through
the increased production of respiratory secretions or damage to
the respiratory epithelium by the virus, resulting in increased
attachment or invasion by Hib into the bloodstream.194
Several studies have shown that in children younger than
6 months, breastfeeding is protective against invasive Hib
disease.88,175,181,186 Although the mechanism for protection is
unknown, it is thought to be due immunologic or nutritional
cofactors that are transmitted from the mother. Studies have
shown that human milk contains secretory antibody to the Hib
PRP capsule.195
Passive immunization
Bacterial polysaccharide immune globulin (BPIg)
Passive immunization with BPIg hyperimmune globulin was
tested in high-risk children in the late 1980s. BPIg was pre-
pared from the plasma of adult volunteers immunized with
Hib polysaccharide vaccine, a 4-valent meningococcal vaccine,
and pneumococcal polysaccharide (14- or 23-valent) vaccines.
Compared with conventional immune globulin, BPIg contained
10 to 60 times higher concentrations of antibody to Hib poly-
saccharide.196 Infants were randomized to receive BPIg or saline
placebo at 2, 6, and 10 months of age in a double-blind trial
among White Mountain Apache infants in Arizona. The point
estimate of efficacy for 4 months following BPIg administra-
tion was 86% (95% CI, 11%-100%). 197 However, because of
the necessity of frequent large-volume injections and the sub-
sequent availability of effective conjugate vaccines for this age
group, BPIg was not licensed for use in the United States.
Active immunization (Box 13-3)
Polysaccharide vaccine
The first Hib vaccine to be field tested was the pure capsular poly-
saccharide (PRP) vaccine. The level of antibody response to PRP
vaccine is age-dependent. After one dose, 8% to 20% of children
younger than 12 months had an antibody concentration of more
than 0.15 g/mL, and only 2% had a concentration of more than
174 SECTION TWO
Box 13-3
,FZ
QPJOUT
BDUJWF
JNNVOJ[BUJPO
t
5IF
QVSF
QPMZTBDDIBSJEF
131
WBDDJOF
JT
FGGFDUJWF
JO
DIJMESFO

PMEFS
UIBO

NPOUIT
CVU
OPU
JO
DIJMESFO
ZPVOHFS
UIBO

NPOUIT
t
5IF
)JC
DPOKVHBUF
WBDDJOFT
BSF
TBGF
BOE
FGGFDUJWF
JO
DIJMESFO

BT
ZPVOH
BT

XFFLT
t
*OWBTJWF
EJTFBTF
SBUFT
IBWF
EFDSFBTFE
ESBNBUJDBMMZ
XJUI

SPVUJOF
VTF
PG
)JC
DPOKVHBUF
WBDDJOF
JO
B
WBSJFUZ
PG
TFUUJOHT
1.0 g/mL.157,198 By contrast, up to 45% of children aged 12 to
17 months had an antibody concentration of more than 1.0 g/
mL, and 50% to 75% of children aged 18 to 23 months mounted
such an antibody response.41,157,199 More than 90% of 4 to 5 year-
olds responded with a strong antibody response.200 While a range of
doses elicited a significant immune response in older children, no
difference in response was seen in children younger than 12 months
who were given doses ranging from 0.2 to 50 g of PRP.13,201
There has been much interest in the correlation between the anti-
PRP antibody titer and protection against Hib disease. Whether
protection is correlated with the peak of the induced antibody
response, the circulating antibody response at the time of expo-
sure to the organism, or the level of antibody response elicited
by exposure remains unclear. An analysis of serum samples
from unimmunized persons in Finland showed that a minimum
serum antibody concentration of more than 0.15 g/mL corre-
lated with decreased incidence of Hib meningitis.199 Population-
level studies suggest that an antibody concentration of more than
0.15 g/mL is adequate to provide short-term protection against
invasive Hib disease, but that a concentration of 1.0 g/mL or
more is necessary for long-term protection. However, certain
individuals have been shown to have breakthrough illness events
even with apparently adequate antibody levels.203
Persistence of antibody response following immunization is
also age-dependent. In children younger than 2 years, antibody
concentrations rapidly declined to levels similar to unimmu-
nized children in the first few months after vaccination.41,204 In a
follow-up study of children vaccinated with 1 dose of PRP at ages
18 to 35 months, children who were 1.5 years postvaccination
had geometric mean antibody titers of more than 1.0 g/mL,
but children who were 3.5 years postvaccination had geometric
mean titers of only around 0.5 g/mL.41
A small trial in North Carolina failed to show efficacy of Hib
polysaccharide vaccine.205 However, the failure is likely due to
the small control group and the small number of invasive Hib
5IF

5ZQF
C
$POKVHBUF
7 BDDJOFT
"WBJMBCMF
JO
UIF
6OJUFE
4UBUFT
disease episodes in the study population.13 In a trial designed to
evaluate the efficacy of a meningococcal vaccine during an epi-
demic, the Hib polysaccharide vaccine was used as the control.
Finnish children aged 3 months to 5 years were randomized to
receive meningococcal vaccine or the Hib polysaccharide vac-
cine. Children aged 18 to 71 months who received Hib poly-
saccharide vaccine had a statistically significant reduction in
bacteremic Hib disease incidence, and vaccine efficacy was
estimated to be 90% (95% CI, 55%-98%). Of note, in children
younger than 18 months, there was no statistically significant
difference in rates of bacteremic Hib disease between the two
groups.157,206 Based on these results, the vaccine was licensed
in the United States in 1985 for use in children older than
23 months and for children 18 to 23 months of age considered
at high risk for invasive Hib disease.207 Subsequent case-control
studies in the United States demonstrated a range of efficacy of
55% to 92% in children 24 to 72 months old.208,209
In general, PRP vaccines were found to be extremely safe
in field trials and postmarketing studies, with no association
between PRP vaccine and serious reactions found in a 1-year
postmarketing surveillance study.210 There was some concern
about an increased risk for Hib infection in the 1 week follow-
ing vaccination with PRP.211-214 This corresponded to the obser-
vation that the anti-PRP antibody concentrations fell in the
first few days after vaccination.215 Black et al213 found a 6.4-fold
increased risk in the week after vaccination (95% CI, 2.1-19.2)
vs a 1.8-fold increase (95% CI, 0.3-10.2) shown by a Centers for
Disease Control and Prevention (CDC) study conducted across
the United States.211 The affected children in both studies were
thought to be high risk; they were African American, attended
day care, and had underlying medical conditions.212 Owing to
the subsequent availability of Hib conjugate vaccines, inter-
est in further defining the safety and efficacy of PRP vaccines
waned.
Conjugate vaccines
Four conjugate Hib vaccines have been developed. The manu-
facturers and formulations available for use in the United States
are shown in Table 13-2. The protein carrier, PRP component
size, and chemical linkages are different for all four conjugate
vaccines, and there are some differences in the elicited immune
response. The vaccines should all be stored between 2°C and 8°C.
PRP-diphtheria toxoid conjugate (PRP-D) was produced by
Connaught Laboratories. It contained medium-sized lengths of
the polysaccharide linked to the diphtheria toxoid carrier. There
were no added adjuvants. Thimerosal was added as a preservative.

PRP-D was the first Hib conjugate vaccine to be licensed in the
United States. It was licensed for children older than 15 to
59 months in 1987. This product is no longer available.
The b oligosaccharide conjugate (HbOC) vac-
cine contains oligosaccharides derived from purified PRP from
the Hib Eagan strain. These are coupled by reductive amination
to purified CRM197, a nontoxic variant of diphtheria toxin iso-
lated from C7 ( 197). The HbOC
vaccine was licensed in the United States in October 1990.
The PRP-OMP vaccine is based on a purified PRP from
the Hib Ross strain. This is covalently bonded to an outer
membrane protein complex of the B11 strain of
type B. The mixture is formulated with alumi-
num hydroxyphosphate sulfate and 0.9% sodium chloride. The
vaccine was licensed for use in all infants in the United States
in December 1990.
PRP-T is made by covalently binding PRP to tetanus toxoid.
The PRP-T currently available in the United States is produced
as a lyophilized powder that is reconstituted at the time of use
with 0.4% sodium chloride and contains no preservative. This
vaccine was licensed in the United States in March 1993. None
of the vaccine manufacturers report the presence of antibiotics
in any of the monovalent or combination Hib vaccines.
PRP-D elicited a geometric mean antibody response of more
than 1.0 g/mL in a group of 141 children 15 months or older
after just one dose.216 In children 9 to 15 months of age, all sub-
jects achieved antibody titers of more than 1.0 g/mL after two
doses; although titers waned in the following year, a booster
dose induced a strong antibody response.217 However, in chil-
dren younger than 6 months, the immune response was lim-
ited.218 In addition, the antibody response was suboptimal in
some high-risk children, including children with cancer.219
Clinical trials of the protective efficacy of PRP-D were conducted
in Finland and Alaska. Between 1985 and 1987, an open-label
non–placebo-controlled trial was conducted in Finland in which
all children born on the odd days of the month received 25-g
doses of PRP-D at ages 3, 4, 6, and 14 to 18 months. The pro-
tective efficacy of PRP-D against invasive Hib disease after three
doses was 90% (95% CI, 79%-96%). Following the booster dose,
the protective efficacy was 100%.220 The Alaskan trial recruited
Native Alaskan infants between 1984 and 1988. Infants were
randomized to receive PRP-D or placebo at 2, 4, and 6 months
of age. The protective efficacy after three doses was 35% (95%
CI, 57%-73%). 221 Because of the availability of other conjugate
vaccines with superior immunogenicity and efficacy, particu-
larly in young and high-risk children, PRP-D was not used in
most countries and is no longer manufactured.
The three other currently available conjugate vaccines are
highly immunogenic when given to children 18 to 24 months or
older.222,223 The response in younger infants varies between vac-
cines (Figure 13-3). HbOC, while relatively poorly immunogenic
after one or two doses at 2 and 4 months of age, elicits antibody
titers more than 1.0 g/mL in the majority of recipients after the
third dose at 6 months of age.224,225 The antibody levels induced
after the third dose at 6 months of age are slightly higher than with
PRP-OMP and are similar to those from PRP-T.226 The induced
antibody is bactericidal and has a statistically significantly greater
avidity than the antibody induced by PRP-OMP and PRP-T.227 The
clinical significance of this increased antibody avidity is unknown.
PRP-T has an immunogenicity pattern similar to that of
HbOC.224,226 Some studies have shown that this vaccine is the
most immunogenic of the three available Hib conjugate vac-
cines after completion of a three-dose primary series;224,225
whether that translates to improved efficacy is unclear. The
elicited antibodies are bactericidal, with a similar isotype dis-
tribution as HbOC.203
Haemophilus influenzae vaccines 13 175
*NNVOPHFOJDJUZ
PG
UIF

type b
DPOKVHBUF
WBDDJOFT
JO
JOGBOUT
WBDDJOBUFE
BU


BOE

NPOUIT
PG

age. 4PVSDF
*NNVOPHFOJDJUZ
PG
GPVS

type b conjugate
WBDDJOFT
JO
JOGBOUT
WBDDJOBUFE
BU


BOE


NPOUIT
PG
BHF
'SPN



UZQF
C
DPOKVHBUF
WBDDJOFT
SFDPNNFOEBUJPOT
GPS
JNNVOJ[BUJPO
XJUI

SFDFOUMZ
BOE
QSFWJPVTMZ
MJDFOTFE
WBDDJOFT
1FEJBUSJDT


1BHF


%BUB
GSPN
(SBOPGG
%.
"OEFSTPO
&-
0TUFSIPMN
.5
FU
BM
%JGGFSFODFT
JO

JNNVOPHFOJDJUZ
PG
UISFF


UZQF
C
DPOKVHBUF
WBDDJOFT
JO

JOGBOUT
+
1FEJBUS


Unlike the other Hib conjugate vaccines, PRP-OMP elicits
antibody titers of more than 1.0 g/mL in 70% to 80% of chil-
dren after just one dose given at 2 months of age.228-232 After
the second dose at 4 months of age, some studies showed that
well over 90% of children achieved antibody titers more than
1.0 g/mL. 233,234 A study in high-risk Native American children
showed that 1 month following the second dose at 4 months of
age, 67% of Apaches and 75% of Navajos had antibody titers
more than 1.0 g/mL. 228 A third dose did not result in sig-
nificantly increased antibody concentrations in most studies;
therefore, the primary series for this vaccine consists of only
two doses.224,225,229 In children given a booster dose of PRP-OMP
at 12 to 15 months of age, a clear booster response is seen.235
However, the peak antibody titer achieved after the booster dose
is lower than with the other two vaccines. Because of the ability
to induce antibody responses after the first dose in young infants,
PRP-OMP was evaluated in newborns as a possible means of
providing earlier protection against invasive Hib disease; unfor-
tunately, antibody levels were substantially diminished in chil-
dren through the first year of life, suggesting immune tolerance
if the vaccine is administered soon after birth.236 In contrast, a
study in Finland found that children receiving PRP-T at 2 days of
age had a geometric mean antibody concentration of 0.12 g/mL
at 4 months of age, which was significantly higher than in
unimmunized children. Subsequently vaccinating the children
again at 4 months of age elicited a strong antibody response,
suggesting that children vaccinated as neonates are protected
against invasive Hib disease.237 The antibody induced by PRP-
OMP has been shown to have lower avidity and bactericidal
activity compared with the antibody elicited by the other con-
jugate vaccines.227 The differences in antibody titers and avidity
induced after the primary series and booster doses of the differ-
ent conjugate vaccines have not been shown to have any clini-
cal significance. PRP-OMP may have a theoretical advantage
in populations in which high rates of Hib disease are seen in
children younger than 6 months or in which incomplete immu-
nization is more likely. In populations such as the Navajo and
Apache in the United States where high rates of Hib disease
176 SECTION TWO Licensed vaccines
were observed in children younger than 6 months, routine use
of PRP-OMP has virtually eliminated invasive Hib disease.238
Some studies showed differences in immunogenic response
to the Hib polysaccharide vaccine in certain ethnic groups; for
example, Siber et al239 showed that Apache children had 10-fold
lower geometric mean Hib antibody concentration after immu-
nization than white children (0.34 vs 3.6 g/mL; .01) at the
age of 24 months, approximately 5 to 6 months after immuniza-
tion. Similarly, PRP-D had very low immunogenicity in Alaskan
Native infants.221 However, ethnic differences in immune
response have not been shown with the other Hib conjugate
vaccines when tested in numerous developed and developing
country settings. Combination vaccines containing Hib conju-
gate vaccine have also been shown to be highly immunogenic
in a variety of settings.240-242 For example, the DTaP-HBV-IPV-
PRP-T hexavalent vaccine when studied in Western Europe and
the Philippines showed that 93% to 100% of children had an
anti-PRP titer of 0.15 g/mL 1 month after the primary vaccina-
tion series.241
Interchanging HbOC and PRP-OMP,243 as well as HbOC and
PRP-T,244 for the primary series or the booster dose has been
shown to be equally or more immunogenic than using a single
vaccine for the entire series. Use of HbOC for the booster dose
following a two-dose primary series with PRP-OMP has been
shown to induce a higher antibody response compared with giv-
ing PRP-OMP as the booster dose (geometric mean antibody
titer 7.46 vs 29.5 g/mL; 0.05). 245 Use of a single vaccine has
been highly effective in many different settings; whether certain
combinations of vaccines in the Hib series can improve effec-
tiveness has not been established.203
Administration of a vaccine's carrier protein at a time before
the administration of a conjugate vaccine that contains that
same protein is termed “carrier priming”. Carrier priming
may result in enhanced immune responses to Hib conjugate
vaccine, but the immunologic consequences of carrier prim-
ing are unclear.246-248 Because most children have early expo-
sure to vaccines containing diphtheria and tetanus toxoids, it
has been suggested that use of diphtheria or tetanus toxoid as
the carrier for Hib conjugate vaccine may confer an immuno-
logic advantage.249 Animal studies have shown immunologic
enhancement with carrier priming.250,251 Several clinical trials
have demonstrated an enhanced antibody response following
PRP-T administration in children who had previously received
diphtheria-tetanus–acellular pertussis (DTaP) vaccine and sug-
gested that the effect may be at least partially due to carrier
priming.252-254 In a clinical trial in Denmark, infants were given
PRP-T at 5 to 6 months of age and randomized to receive teta-
nus toxoid as part of the diphtheria-tetanus–inactivated polio
(DT/IPV) vaccine before, during, or after this dose. The primed
group that received prior priming with DT/IPV had threefold
higher antibody levels than the unprimed group. The group that
received only concomitant priming had no difference in anti-
body levels compared with the unprimed group.255 The clinical
significance in terms of vaccine efficacy of carrier priming for
the Hib conjugate vaccines remains unclear.
Two large prospective trials showed the clinical efficacy of
HbOC. In the late 1980s, an unblinded, nonrandomized study
was conducted in California, in which all children between
6 weeks and 6 months of age were given a three-dose series with
HbOC vaccine; findings were compared with those for unvacci-
nated children between 7 and 18 months of age. Vaccine efficacy
after two doses was 100% (95% CI, 47%-100%) and after three
doses was 100% (95% CI, 68%-100%); however, efficacy after
just the first dose was lower (26%; 95% CI, 166%-80%), but
the data are limited to the small number of cases occurring after
a single dose.256 A second trial in Finland compared the efficacy
of HbOC with PRP-D when given at 4, 6, and 14 to 18 months
of age. After two doses, the vaccine efficacy was 87% (95% CI,
69%-96%) for PRP-D and 95% (95% CI, 76%-99%) for HbOC.
There were no episodes of invasive Hib disease following the
booster dose in either group.257
PRP-OMP was evaluated in a randomized, double-blind,
placebo-controlled trial on the Navajo Indian reservation, a pop-
ulation previously established to be at high risk for invasive Hib
disease. Infants were randomized to receive the vaccine or placebo
at 2 and 4 months of age. Overall efficacy was 95% (95% CI, 72%-
99%). Of note, protection began after the first dose. Between the
first and second doses of vaccine or placebo, there were 8 cases of
Hib disease in the placebo group and none in the vaccine group
(vaccine efficacy, 100%; 95% CI, 15%-100%).258 The PRP-OMP
vaccine was subsequently used in postlicensure trials in several
populations, including Los Angeles County, California,259 Navajo
and Apache Indians,260 Australia,261 and Israel.262 Dramatic
reductions in Hib disease incidence have been demonstrated in
all populations in which the vaccine has been used.
Two large PRP-T trials were started in the United States but
were discontinued in October 1990 when the recommendations
for use of HbOC in infants were instituted. In the mid 1990s,
a large vaccine trial was conducted in The Gambia, compar-
ing PRP-T mixed with diphtheria-tetanus-pertussis (DTP) with
DTP alone. This trial showed a protective efficacy of 95% (95%
CI, 67%-100%) against all invasive Hib disease after three doses.
Efficacy after a single dose for all invasive Hib disease was 44%
(95% CI, 85%-85%). 96 PRP-T has been shown to reduce clini-
cal pneumonia and radiologically confirmed pneumonia in sev-
eral clinical trials. The efficacy for the prevention of clinical
pneumonia as defined by the WHO was 4.4% (95% CI, –5%-
12.9%), and for radiologically defined pneumonia, efficacy was
21.1% (95% CI, 4.6%-34.9%). The efficacy for the prevention
of blood culture–positive pneumonia after three doses of vac-
cine was 100% (95% CI, 55%-100%).96 PRP-T was also shown
to be highly efficacious against invasive Hib disease in trials in
Britain, Chile, and Finland.263-265 There have been no compara-
tive efficacy trials of the three commonly used Hib conjugate
vaccines. All three seem highly and equally efficacious.
Hib vaccine has been highly effective in developed and develop-
ing countries. In the United States, the introduction of the Hib
conjugate vaccine has resulted in a decrease in the incidence
of Hib disease of more than 98%.125,152 Similar success has
been documented in Canada and several countries in Western
Europe.266-269 Several developing countries have also noted dra-
matic declines in Hib disease with use of the Hib vaccines.161,270
In Chile, a 90% decline in the rate of invasive Hib disease was
demonstrated.264 After Hib conjugate vaccine introduction,
the annual incidence rates of Hib meningitis in The Gambia
decreased from more than 200 per 100,000 children younger
than 1 year in the early 1990s to none per 100,000 in 2002
and from 60 to no cases per 100,000 in children younger than
5 years.48 In Uganda, vaccine effectiveness with 2 or more doses
against Hib meningitis was 99% (95% CI, 92%-100%) using
neighborhood control subjects and 96% (95% CI, 80%-100%)
using hospitalized control subjects.271 In Senegal, the annual
rate of Hib meningitis in children younger than 1 year went
from 33 to 1.4 per 100,000; similar success was documented
in Mali.272,273 Some studies have shown decreases in rates of
clinical pneumonia, as well as pneumonia with radiographic
consolidation and clinical pneumonia, following Hib vaccine
introduction.100,101 Following routine introduction in Chile, Hib
vaccine was shown to have 80% effectiveness ( = .039) against
bacteremic pneumonia and empyema.274
 Haemophilus influenzae vaccines 13 181

Hib vaccination, unless there is strong epidemiologic evidence

to suggest low disease burden, lack of benefit, or overwhelm-
ing impediments to implementation.360 To sustain vaccination
programs, continued monitoring will be necessary to document
vaccine impact and the cost-effectiveness of vaccine use.

Future vaccines
As the pediatric immunization schedule becomes more and
more complex, developing more combination vaccines is an
area of active interest. There is also a great deal of investigation
182 SECTION TWO
into vaccine candidates for nontypeable disease
and type a. Several candidate vaccines have been
shown to reduce carriage in the ear and nasal mucosa of ani-
mal models.361-363 An oral vaccine against nontypeable
has been shown to reduce the incidence and severity of
exacerbations of chronic bronchitis in small clinical trials.364
Access the complete reference list online at http://www.expertconsult.com
1. Watt JP, Wolfson LJ, O'Brien KL, et al. Burden of disease caused by
type b in children younger than 5 years: global
estimates. Lancet 374:903–911, 2009.
3. Fitzwater SP, Watt JP, Levine OS, et al. type
b conjugate vaccines: Considerations for vaccination schedules and
implications for developing countries. Hum Vaccin 6:810–818, 2010.
15. Levine OS, Knoll MD, Jones A, et al. Global status of
type b and pneumococcal conjugate vaccines: evidence, policies,
and introductions. Curr Opin Infect Dis 23:236–241, 2010.
96. Mulholland K, Hilton S, Adegbola R, et al. Randomised trial of
type-b tetanus protein conjugate vaccine [corrected] for
prevention of pneumonia and meningitis in Gambian infants. Lancet
349:1191–1197, 1997.
97. Levine OS, Lagos R, Muñoz A, et al. Defining the burden of pneumonia in
children preventable by vaccination against type b.
Pediatr Infect Dis J 18:1060–1064, 1999.
A lipopolysaccharide-based conjugate vaccine showed good
safety and immunogenicity in a phase 1 study in adults and
may soon be evaluated in children.365 Future emphasis will
likely be on elucidating correlates of protection against nontype-
able disease that will facilitate the development
and testing of more vaccine candidates.9
98. Gessner BD, Sutanto A, Linehan M, et al. Incidences of vaccine-preventable
type b pneumonia and meningitis in Indonesian
children: hamlet-randomised vaccine-probe trial. Lancet 365:43–52, 2005.
99. Baqui AH, El Arifeen S, Saha SK, et al. Effectiveness of
type B conjugate vaccine on prevention of pneumonia and
meningitis in Bangladeshi children: a case-control study. Pediatr Infect Dis J
26:565–571, 2007.
171. Adegbola RA, Mulholland EK, Secka O, et al. Vaccination with a
type b conjugate vaccine reduces oropharyngeal
carriage of type b among Gambian children. J Infect Dis
177:1758–1761, 1998.
270. Gessner BD. type b vaccine impact in resource-poor
settings in Asia and Africa. Expert Rev Vaccines 8:91–102, 2009.
356. Hajjeh RA, Privor-Dumm L, Edmond K, et al. Supporting new vaccine
introduction decisions: lessons learned from the Hib Initiative experience.
Vaccine 28:7123–7129, 2010.
 SECTION TWO: Licensed vaccines
Hepatitis A vaccines
Although episodes of jaundice have been recognized since the
time of Hippocrates, the earliest outbreaks that on epidemio-
logic grounds seem likely to have been hepatitis A occurred in
Europe in the 17th and 18th centuries.1 Early in the 20th cen-
tury, Cockayne2 concluded that sporadic and epidemic forms of
jaundice were probably manifestations of the same disease, and
McDonald3 postulated that a virus might be involved. Hepatitis
has been a military problem for centuries, and major outbreaks
occurred among British, French, German, and Romanian troops
in World War I and among French, American, Commonwealth,
and Axis troops in World War II.
The first scientific evidence that hepatitis A was an enteri-
cally transmitted viral infection was obtained during World War
II from studies among experimentally infected volunteers. In
a classical series of experiments, Havens4 and Neefe and col-
leagues5 were able to show that volunteers in whom infectious
hepatitis had developed were protected from subsequent chal-
lenge with the same virus and with infectious material obtained
from a separate outbreak. Havens4 and Havens and Paul6 also
demonstrated that intramuscular injection of pooled nor-
mal human immune globulin could prevent or attenuate the
disease, and the practice was rapidly adopted. During an epi-
demic in 1945 in the Mediterranean arena, more than 2,700
American soldiers were immunized, with an 86% reduction in
the incidence of disease among immunized troops.7
After it became apparent that infectious hepatitis was clearly
distinct from serum hepatitis in mode of transmission and eti-
ology, MacCallum8 proposed that the diseases be known as hep-
atitis A and B, replacing the terms infectious hepatitis (hepatitis
A) and homologous serum hepatitis (hepatitis B). This sugges-
tion was adopted in 1952 by the World Health Organization's
First Expert Committee on Viral Hepatitis9 but not widely
accepted by physicians and virologists until the early 1970s.
By the end of World War II, volunteer studies had clearly
established that infectious hepatitis was enterically transmitted
and was caused by a filterable agent—presumably a virus—that
was relatively heat-stable but could be inactivated by chlorine.5
The disease seemed to be caused by a single agent, seemed to
be associated with lifelong immunity, and was preventable by
administration of normal immune globulin. In the 1950s, these
data were expanded and refined by a further series of studies
conducted by Ward et al10 and Krugman and colleagues11 at the
Willowbrook State School in New York and by Melnick and
Boggs12 and their colleagues at the Joliet prison in Illinois. Fecal
samples collected from the latter studies were critical in the
subsequent identification of the etiologic agent of the disease by
electron microscopy (Figure 14-1).13
Trudy V. Murphy
Stephen M. Feinstone
Beth P. Bell 14
Why the disease is important
Hepatitis A virus (HAV) infects more than 80% of the popula-
tion of many developing countries by late adolescence and also
is common in developed countries.14–16 In the United States,
several thousand hepatitis A cases are reported annually.17,18
Background
Clinical description
Numerous studies have been conducted to define the incuba-
tion period of hepatitis A after natural or experimental infection
in children or adults. Although disease has been seen as early as
15 days and as late as 50 days after exposure, the mean incuba-
tion period is about 28 days.5,19,20 The average incubation period
has been reported to be shorter among patients who acquired
HAV infection by parenteral transmission from contaminated
blood products and among nonhuman primates infected par-
enterally compared with those infected orally. The incubation
period also depends on the infectious dose.21–23
Infection with HAV, as evidenced by the detection of HAV-
specific IgM antibody in serum, may produce a wide spectrum
of outcomes ranging from inapparent (asymptomatic, with-
out elevation of serum aminotransferase levels), to subclinical
(asymptomatic, with elevation of serum aminotransferase lev-
els), to clinically evident (with symptoms). Symptoms typical of
acute hepatitis include jaundice and dark urine, but symptom-
atic hepatitis A without jaundice also occurs.
The frequency of symptoms with HAV infection is strongly influ-
enced by age. Children are less likely to have symptomatic infection
compared with adults; 50% to 90% of infections acquired before the
age of 5 years are asymptomatic, but 70% to 95% of infected adults
will have symptoms.24–26 Jaundice is rare among young children but
will occur in the majority of adults with hepatitis A.26,27
The clinical symptoms of acute hepatitis A in an individual
patient are indistinguishable from those caused by other forms of
viral hepatitis. The onset of the prodromal period, particularly in
older children and adults, can be abrupt and is characterized by
increasing fatigue, malaise, anorexia, fever, myalgias, dull abdom-
inal pain, nausea, and vomiting. Pediatric patients may have
diarrhea or, less commonly, upper respiratory symptoms.28,29 If
present, typical symptoms of hepatitis, beginning with darken-
ing of the urine and followed by jaundice and pale-colored stools,
188 SECTION TWO Licensed vaccines
Treatment
No specific therapy is available for hepatitis A, and management
is supportive. Most physicians do not recommend restrictions
in activity because studies have failed to demonstrate an impact
on the course of illness. Similarly, it has been customary to rec-
ommend against vigorous exercise and to encourage abstinence
from alcohol, although there are few objective data that demon-
strate a benefit. Medications, particularly those metabolized by
the liver or that are potentially hepatotoxic, should be used with
caution because their half-life may be prolonged. Hospitalization
may be necessary for patients who become dehydrated from
vomiting or in whom fulminant hepatitis develops.
Hepatitis A is occasionally complicated by cholestasis; approxi-
mately 7% of cases were affected in one series of 59 hospitalized
patients.30 A brief course of corticosteroids has been reported to
shorten the course and reduce symptoms, primarily itching, but
recovery is universal, even without treatment.34 Liver transplan-
tation may be indicated in some cases of fulminant hepatitis, but
because survival without transplantation is relatively high and no
single factor is predictive of poor outcome, it has been difficult to
establish criteria for choosing transplantation candidates.35,69,182–184
Persistent HAV infection has been demonstrated in some trans-
plant recipients, but whether this affects survival is unknown.185
Epidemiology
Worldwide disease patterns
Hepatitis A occurs worldwide, but major geographic differ-
ences exist in endemicity and resulting epidemiologic features
(Figure 14-4). The degree of endemicity is closely related to
hygienic and sanitary conditions and other indicators of the
level of development. Under conditions of overcrowding, espe-
cially when there is limited access to clean water and inadequate
disposal of human feces, HAV infects most people early in life,
when infection is rarely clinically apparent (Figure 14-5). Where
high standards of hygiene and sanitation apply, most children
reach adult life without encountering the virus. Distinct pat-
terns of HAV infection can be described, each with characteris-
tic age-specific profiles of anti-HAV prevalence and hepatitis A
incidence and prevailing environmental (hygienic and sanitary)
and socioeconomic conditions. These patterns present a chal-
lenge for conveying the risk of hepatitis A infection in maps of
the world (Figure 14-5).577
In areas of high endemicity, represented by the least devel-
oped countries (ie, parts of Africa, Asia, and Central and South
America), poor socioeconomic conditions allow HAV to spread
readily (Figure 14-4). Infection is nearly universal in early child-
hood, when asymptomatic infection predominates, and essen-
tially the entire population is infected before reaching adolescence,
as demonstrated by the age-specific prevalence of anti-HAV
(Figure 14-5).186–188 Susceptible adults in these areas are at high risk
of infection and disease, but reported disease rates are generally
low and outbreaks rare because of the high prevalence of immunity
in the population. High endemicity patterns also may be seen in
some ethnic or geographic groups within highly developed coun-
tries, such as the aboriginal children in northern Australia.189
In areas of moderate endemicity, HAV is not transmitted as
readily because of better sanitary and living conditions, and the
predominant age of infection is older than in areas of high ende-
micity (Figures 14-4 and 14-5).190 Paradoxically, the overall inci-
dence and average age in reported cases are often higher than in
highly endemic areas because high levels of virus circulate in a
population that includes many susceptible older children, ado-
lescents, and young adults, who are likely to have symptoms
with HAV infection.191 Large common-source food- and water-
associated outbreaks occur because of the relatively high rate
of virus transmission and large number of susceptible persons,
especially in the higher socioeconomic level. Such an outbreak

Patterns of hepatitis A virus infection worldwide.
occurred in Shanghai in 1988, with more than 300,000 cases
associated with consumption of clams harvested from water
contaminated with human sewage.19 Nevertheless, person-to-
person transmission in community-wide epidemics continues
to account for much of the disease in these countries.
Shifts in age-specific prevalence patterns that reflect a tran-
sition from high to intermediate endemicity are occurring in
many parts of the world (Figure 14-4). A feature of this tran-
sitional pattern is striking variations in hepatitis A epidemi-
ology between countries and within countries and cities, with
some areas displaying a pattern typical of high endemicity and
others a pattern of intermediate endemicity.192–204 Considerable
hepatitis A–related morbidity and associated costs occur with
this transition, even in developing countries.80,205 For exam-
ple, hepatitis A was the cause of the fulminant hepatitis of two
thirds of children admitted to two hospitals in Argentina dur-
ing a 15-year period, and, in one of these hospitals performing
liver transplantations, one third of liver transplantations among
children were for fulminant hepatitis A.80 Researchers in India
and Egypt have documented an increasing proportion of hospi-
talizations for acute viral hepatitis attributable to hepatitis A
in recent years in hospitals serving primarily patients of higher
socioeconomic status.206,207 For example, hepatitis A was the
etiologic agent of acute hepatitis in 25% of adults hospitalized
in one tertiary care hospital in northern India.207
In the United States, Canada, Western Europe, and other
developed countries, the endemicity of HAV infection is low
(Figures 14-4 and 14-5). Relatively fewer children are infected,
the incidence of disease is generally low, and disease often
occurs in the context of community-wide and child-care center
Hepatitis A incidence, United States, 1952 through 2009. (Data from the Centers for Disease Control and Prevention, National Notifiable Diseases
Surveillance System, Atlanta, GA.)
Hepatitis A vaccines 14 189
outbreaks.208–212 A cyclic pattern of disease incidence with peaks
every 5 to 10 years has been noted in some developed countries
with temperate climates. Population-based seroprevalence sur-
veys show a gradual increase in the prevalence of anti-HAV with
increasing age, primarily reflecting declining incidence, chang-
ing endemicity, and resultant lower childhood infection rates
over time (Figure 14-5).16 In countries that have very low ende-
micity (eg, Scandinavian countries), most cases occur in defined
risk groups such as travelers returning from areas of high or
intermediate endemicity and users of injection drugs.213
Epidemiology in the United States
Before hepatitis A vaccine was introduced in the United States
in the mid-1990s, hepatitis A incidence had been cyclic, with
peaks every 10 to 15 years (Figure 14-6). During the 1980s and
1990s, approximately 25,000 to 35,000 hepatitis A cases were
nationally reported each year.214,215 Modeling suggested that the
majority of infections were asymptomatic. One such analysis
estimated an average of 270,000 infections per year during the
1980s and 1990s, approximately 10 times the reported num-
ber of symptomatic cases.216 The highest hepatitis A rates were
reported among children 5 to 14 years old, with approximately
one third of cases occurring among children younger than
15 years.217 The same model estimated that more than half of
HAV infections were among children younger than 10 years,
and the majority of these infections were among asymptomatic
children from birth to 4 years.216 Thus, young children with
unrecognized asymptomatic infection constituted a major res-
ervoir for HAV transmission.
Hepatitis A incidence showed striking variation by region
and by racial/ethnic group during the 1990s. The highest rates
and the majority of cases consistently occurred in a limited
number of states and counties in the western and southwestern
United States. Among residents of 11 states comprising 22%
of the US population, the average annual hepatitis A incidence
was 20 cases per 100,000 or greater during 1987 to 1997 (twice
the national average of about 10 per 100,000). These states
accounted for an average of 50% of reported cases in the United
States. An additional 18% of cases were among residents of 6
states in the region with average annual rates above the national
average.217 Rates of hepatitis A among Native Americans and
Alaska Natives were more than 5 times higher than among
other racial groups, and rates among Hispanics were approxi-
mately three times higher than among non-Hispanics.218
190 SECTION TWO
Most cases of hepatitis A in the United States occurred in
the context of community-wide epidemics during which infec-
tion was transmitted from person to person in households and
extended family settings.208 Once initiated, these outbreaks
often persisted for several years and proved difficult to control219
and even failed to be controlled with early attempts to rapidly
vaccinate some portion of the population.77,220 Often no single
risk group accounted for the majority of cases and infections.
Children, most of whom were asymptomatic, had an important
role in sustaining HAV transmission. Serologic studies of mem-
bers of households with an adult case without an identified
source found that 25% to 40% of contacts younger than 6 years
had serologic evidence of recent HAV infection221,222 In one of
these studies, 52% of households of adults without an identified
source of infection included a child younger than 6 years, and
the presence of a young child was associated with household
transmission of HAV.221
Outbreaks in child-care centers also were common, particu-
larly in large centers that cared for children in diapers.26,223,224
In the 1970s, it was recognized that asymptomatic infections
among children receiving care maintained the outbreaks and
occasionally were the source of more extensive transmission
in the community.223,225,226 Outbreaks were often recognized
when one or more adult contacts (usually parents) became
ill.26,224 Despite the occurrence of outbreaks in child-care cen-
ters, studies of center employees did not show significantly
increased prevalence of HAV infection compared with control
populations.227
Hepatitis A cases among children in schools usually reflected
disease that had been acquired in the community. However,
multiple cases among children within a school could indicate
a common-source outbreak.180 Historically, HAV infection was
endemic in institutions for developmentally disabled people,
but with smaller facilities, improved conditions, and vaccina-
tion, the incidence and prevalence of infection have decreased
and outbreaks rarely are reported in the United States.17,214,228
Cyclic outbreaks occurred among men who have sex with
men and users of injecting and noninjecting drugs; dur-
ing outbreak years, this exposure accounted for up to 15% of
nationally reported cases.78,210,229–232 Other potential sources
of infection such as international travel and food-borne out-
breaks accounted for a small proportion of cases. For nearly
50% of patients with hepatitis A, no source of infection could
be identified.79
Incidence of acute hepatitis A by age group, United States, 1990 through 2009.
National Notifiable Diseases Surveillance System, Atlanta, GA).
Results of the Third National Health and Nutrition Examination
Survey, conducted during 1988 to 1994, indicated that about
30% of the US population had serologic evidence of prior HAV
infection.16 Anti-HAV prevalence was directly related to age,
increasing from 9% among children 6 to 11 years old to 75%
among persons older than 70 years, and was inversely related to
income. High seroprevalence of anti-HAV among older adults
probably was the result of widespread exposure to HAV in the
prevaccine era, when widespread cyclic increases in HAV infec-
tion occurred in the United States. Age-adjusted anti-HAV
prevalence was higher among foreign-born (68%) than among
US-born (25%) subjects and among Mexican Americans (70%)
than among non-Hispanic blacks (39%) and whites (23%).16
In 1995, when hepatitis A vaccines became available in the
United States, more than 31,000 cases of hepatitis A had been
reported (incidence, 12 cases per 100,000 persons) making hep-
atitis A one of the most frequently reported vaccine-preventable
diseases.214 Between 1996 and 1999, the Advisory Committee
on Immunization Practices (ACIP) made incremental recom-
mendations for hepatitis A vaccination217,218 and, in 2006, rec-
ommended routine hepatitis A vaccination for all US infants
starting at 1 year of age.17 The US hepatitis A rates have shown
dramatic declines. In 2009, 1,987 cases were reported to the
Centers for Disease Control and Prevention (CDC), for a his-
torically low incidence of 0.6 per 100,000 (Figure 14-6).18
Hepatitis A rates among children have declined more
sharply than among adults, and since 2002, rates among
adults and children have been similar (1.0 cases per 100,000)
(Figure 14-7).18,214 Previous disparities in rates across racial/
ethnic groups have largely disappeared (Figure 14-8). In 2009,
rates among Alaska Natives/Native Americans, which had
been more than 50 cases per 100,000, were lower than those
for other racial/ethnic populations (0.3 per 100,000). Rates
among Hispanics declined 94% from 1997 to 2007 (1.4 cases
per 100,000 population); in 2009, the rate was the lowest ever
recorded (0.8 per 100,000).18,214 The geographic variations that
characterized hepatitis A incidence in the past215 have disap-
peared. Significantly greater reductions in incidence occurred in
areas where vaccination was recommended first (Figure 14-9).233
After implementation of routine vaccination of children in
these states, rates equalized across the country.214,217
(Data from the Centers for Disease Control and Prevention,

Incidence of acute hepatitis A by race/ethnicity, United States, 1990 through 2009. (Data from the Centers for Disease Control and Prevention,
National Notifiable Diseases Surveillance System, Atlanta, GA.)
With the advent of routine vaccination of children, community-
wide outbreaks have largely ceased. Community outbreaks
that were sustained by transmission among adults in high-risk
groups are rare; when they occur, strategies such as providing
vaccination in settings such as jails have had some success.234
Results from National Health and Nutrition Examination
surveys conducted for the periods 1988-1994, 1999-2002,
and 2003-2006 indicate an overall decline in the age-
adjusted seroprevalence of anti-HAV in the US population
from 32.5% in 1988-1994 to 26.7% in 2003-2006. During
this period, significant declines were observed for non-
Hispanic whites (decreased to 23.8%), non-Hispanic blacks
(decreased to 37.3%), Mexican Americans (decreased to
58.4%), and for all adults 40 years or older (data not pro-
vided). Among persons 6 to 19 years old, the age-adjusted
increase in anti-HAV seroprevalence was 27% (8% in 1988-
1994 and 35% in 2003-2006), probably reflecting increasing
vaccination coverage.235
As the number and incidence of reported hepatitis A cases
has declined, the two most commonly reported potential
sources of infection have become international travel and
household or sexual contact with a person who has hepatitis
A.18,214 During the 2002-2004 period, these exposures each
accounted for about 13% of reported cases with information
available78,79; in 2009, the proportions had changed to 5.6%
and 15% of reported cases with information.18 The propor-
tion of cases among men who have sex with men and users
of injecting and noninjecting drugs declined from as high as
15% to 8.7% and 1.1% of cases with information, respectively,
in 2009. The number of reported cases occurring among chil-
dren and employees of child-care centers and members of
their households also declined, but the proportion remained
approximately 8%.18,78,79 Recognized food-borne or waterborne
outbreaks accounted for a small proportion of cases in most
years (3%-8%), but large outbreaks, such as those associated
with contaminated green onions in 2003, increased the pro-
portion of cases associated with food in some years to as high
as 16%.18,78,79,214 In 2009, 44% of reported cases of hepati-
tis A did not have a recognized source of infection; 41% of
cases did not have information on potential exposures or risk
behaviors.18
Hepatitis A vaccines 14 191
International travel
Hepatitis A is common among unvaccinated persons from devel-
oped countries who travel to regions with high, transitional, or
intermediate endemicity (Figure 14-4).236–238 In prospective stud-
ies during the prevaccine era of American and European travel-
ers, the risk of infection for persons who did not receive immune
globulin was 3 to 5 per 1,000 per month of stay, of the same
order of magnitude as that for malaria, 10 to 100 times greater
than for typhoid, and 1,000 times greater than for cholera.239,240
A more recent study of Swiss travelers estimated the risk to be 6
to 30 cases per 100,000 months of stay in developing countries
among persons who did not receive immune globulin or vaccine
before departure.241 The risk may be higher among travelers stay-
ing in areas with poor hygienic conditions242 and varies accord-
ing to the region and the length of stay. In the United States,
the number of cases among international travelers declined, but
the proportion of cases attributed to recent international travel
increased from 5% to 7% in the 1990s to 15% in 200918; declines
in the number of hepatitis A cases attributed to improved infra-
structure in endemic countries are now reported for travelers
from some European countries.243,244
In the United States and Europe, hepatitis A in travelers,
especially children, traveling to endemic countries to visit rel-
atives and friends accounts for an increasing proportion of
reported cases. Compared with persons traveling for work or
recreation, these persons generally have trips of longer duration
and, by staying in family settings rather than in hotels, may
have greater exposure to HAV circulating in the community.
These persons also are less likely to seek or to adhere to pretravel
advice.245–252 Travelers who acquire hepatitis A during their trip
may also transmit the virus to others on their return.253
Contacts of international adoptees
International adoptees with asymptomatic hepatitis A may
transmit the virus to family members and close contacts. Most
adoptee-associated cases of hepatitis A are among unvaccinated
nontraveling contacts of the adoptee exposed during the first
60 days after arrival in the United States; secondary and tertiary
transmission has been described.254–257 Serologic results from
270 adoptees who received testing within 4 months of arrival in
the United States showed 1% with evidence of acute hepatitis
A; rates varied by country of origin. All acutely infected adop-
tees were asymptomatic.257
222 SECTION TWO Licensed vaccines
Recommended Options for Adding Hepatitis B Vaccine to Childhood Immunization
Schedules, World Health Organization
*Option 1 is recommended in countries that are not yet targeting prevention of perinatal hepatitis B virus
(HBV) transmission; options II and III are recommended in countries that are targeting prevention of perinatal
HBV transmission.
†
Only given in countries where polio is highly endemic.
‡
To prevent perinatal HBV transmission, the birth dose should be given as close as possible to the time of
delivery, preferably within 12 hr.
§
Monovalent vaccine.
¶
Monovalent or combination vaccine.
BCG, bacille Calmette-Guérin; DPT, diphtheria-pertussis-tetanus; DTP, diphtheria-tetanus-pertussis; OPV, oral
poliovirus.
immunogenicity or final antibody concentration.449,458 Longer
intervals between the last two doses result in higher final antibody
concentrations but not seroconversion rates.459 Postvaccination
testing is recommended for certain persons whose subsequent
clinical management depends on knowledge of their immune
status, including some health care and public safety workers;
chronic hemodialysis patients, HIV-infected persons, and other
immunocompromised persons; and sex or needle-sharing part-
ners of HBsAg-positive persons. An anti-HBs concentration of
10 mIU/mL or more measured 1 to 3 months after administra-
tion of the last dose of the primary vaccination series is consid-
ered a reliable marker of protection against infection.
Of persons who did not respond to a primary three-dose vaccine
series with anti-HBs concentrations of 10 mIU/mL, 25% to
50% responded to an additional vaccine dose, and 44% to 100%
responded to a three-dose revaccination series.460–465 Better
response to revaccination occurs in persons who have measur-
able but low ( 10 mIU/mL) levels of antibody after the initial
series.460,461,464 One study identified no difference in response to
revaccination in persons receiving double the standard vaccine
dose.462 Persons who do not have protective levels of anti-HBs
1 to 2 months after revaccination either are primary nonre-
sponders or are infected with HBV. Genetic factors might con-
tribute to nonresponse to hepatitis B vaccination.461,466
A variety of hepatitis B vaccine schedules have been shown to
induce levels of seroprotection of greater than 95% in infants,
including doses administered at birth and 1 and 6 months of
age; at 2, 4, and 6 months of age; and 6, 10, and 14 weeks of
age.451–454,456 Programmatically, there is an advantage to admin-
istering the three doses of hepatitis B vaccine at the same time
as the three doses of other childhood vaccines (eg, DTP, Hib),
and these schedules will accommodate use of DTP- and Hib-
containing combination vaccines. A hepatitis B vaccine schedule
administered with DTP or Hib vaccine will prevent infections
acquired during childhood as well as infections acquired later in
life. However, these schedules will not prevent perinatal HBV
infections because they do not include a birth dose. Therefore,
concerns have been expressed that increased use of combina-
tion vaccines might jeopardize perinatal hepatitis B prevention
programs.467 To prevent perinatal HBV transmission in settings
where combination vaccines are used, a four-dose hepatitis B
vaccination schedule is needed, with the first dose administered
at birth. High titers of maternal antibodies do not interfere with
neonatal immune response to vaccination.467a Use of four-dose
hepatitis B vaccine schedules, including schedules with a birth
dose, has not increased vaccine reactogenicity.468,469
Certain premature infants with low birthweights (ie, 2,000 g)
might have decreased seroconversion rates after administration
of hepatitis B vaccine at birth.470 However, by age 1 month, all
premature infants, regardless of initial birthweight or gesta-
tional age, have a response to vaccination that is comparable to
that of full-term infants.471–474
WHO recommends multiple options for adding hepatitis
B vaccine to existing infant immunization schedules without
requiring additional visits for immunization (see Table 15–4).410
Schedules should optimize the percentage of children completing
the hepatitis B vaccine series, which can be achieved with earlier
administration of vaccine.457 The WHO-recommended minimal
interval between doses one and two is 4 weeks, and the minimum
interval between doses two and three is 4 weeks. Schedules with
these minimal intervals (eg, 6, 10, and 14 weeks) have been dem-
onstrated to have seroconversion rates similar to schedules with
longer intervals, albeit with lower final anti-HBs concentrations.
The WHO minimal recommended intervals for infant hepatitis
B vaccination are shorter than those recommended in the United
States. Although data are limited regarding long-term protection
for schedules with shorter intervals, alternative schedules are
often not feasible. In addition, concerns about long-term protec-
tion are of less practical significance in countries of high ende-
micity where most HBV infections are acquired in childhood.
Hepatitis B vaccine schedules that have been demonstrated to
induce seroprotection rates of greater than 95% in adolescents
include doses administered at 0, 1, and 6 months; 0, 2, and
4 months; and 0, 12, and 24 months.446,458,475–477 In addition, for
adolescents aged 11 to 15 years, the adult dose of hepatitis B vac-
cine can be used for administration at 0 and at 4 to 6 months.478,479
This two-dose schedule produces anti-HBs concentrations equiv-
alent to those obtained with the pediatric dose administered on
a three-dose schedule. After 10 years of follow-up in the Czech
Republic, 85.9% of the adolescents (12 to 15 years old) vacci-
nated with the two-dose schedule had anti-HBs levels of 10 mIU/
mL or higher, as did 85.1% of those who received three pediatric
260 SECTION TWO
are serologically distinct.86,106 Although some serologic cross-
reactivity may exist among HA or NA proteins of the same
influenza subtype but isolated from different species, the extent
of cross-protection afforded by such responses is not well under-
stood. During the 2009 H1N1 pandemic, most studies showed
that persons who had previously been vaccinated against or are
likely to have been infected with recently circulating seasonal
influenza A(H1N1) viruses appeared to have reduced or no sig-
nificant protection against the novel H1N1 virus;36,107–110 some
studies performed in Canada suggested that previous vaccina-
tion with seasonal influenza vaccines might have increased risk
for symptomatic infection with the 2009 H1N1 virus.111
The influenza A virus envelope also contains matrix (M1)
and transmembrane (M2) proteins. M1 protein is located inside
the viral membrane and is thought to add rigidity to the lipid
bilayer, whereas the M2 protein functions as a pH-activated
ion channel.86 A nonstructural gene encodes two proteins: the
multifunctional NS1, which is found only in the virus-infected
cells, and NEP (also known as NS2), which is a minor virion
component involved in nuclear export of the nucleocapsids con-
taining viral RNA. Also contained within the envelope are the
eight segmented nucleocapsids.86,106
Pathogenesis as it relates to prevention
Influenza viruses are spread predominantly by virus-laden
aerosols produced when infected people cough or sneeze. The
degree to which transmission occurs by airborne versus large
droplets is uncertain, but some studies have suggested the poten-
tial for transmission via aerosols.112,113 Transmission of virus
by direct contact with infected respiratory secretions can also
occur.114 A person's susceptibility to infection will depend on his
or her preexisting humoral and cellular immunity to influenza.
Immunity is not fully protective over time and across strains, as
the variability of HA and NA allows viruses that have mutations
in HA, and to a lesser extent in NA, to infect persons previously
infected with earlier viruses of the same subtype. Natural infec-
tion appears to provide broader-based immunity against con-
served epitopes than vaccination, but the extent of protection
afforded by such cross-reactive responses remains unclear.115
This topic will be discussed in more detail (see “Immune
responses to vaccination and correlates of protection”, later).
In temperate climates, influenza activity typically occurs
during the late autumn and winter months. However, in tropi-
cal climates, influenza can occur year round.116 Reasons for the
seasonality of influenza are not understood. In contrast to typi-
cal seasonal influenza, the 2009 influenza A(H1N1) pandemic
began in the United States in April 2009, and it included unusu-
ally high influenza activity during the late spring, as well as peak
transmission during late summer and early autumn 2009.36
The incubation period for influenza is commonly 2 days but
ranges from 1 to 4 days.117 Virus can be isolated from the naso-
pharynx of adults for up to 5 days after illness and longer from
more severely ill persons.118 Children may shed virus for up to
2 weeks, and severely immunocompromised persons can shed
virus for months.119–121
The transmissibility or the reproductive number (R0) for
influenza has been estimated to be approximately 1.5 to 2, and
the serial interval (the interval between the onset of symptoms
in a patient and the onset of symptoms in the contacts who
were infected by that patient) is usually estimated to be 2 to
3 days.122–126 These factors may lead to explosive outbreaks,
especially in highly susceptible populations, as can occur in a
pandemic.122,127–129
The pathogenesis of influenza virus replication and its rela-
tionship to the clinical manifestations and complications of
influenza are not well elucidated. Infection and viral replication
occurs primarily in the columnar epithelial cells of the respira-
tory tract,130,131 but viral replication can occur throughout the
respiratory tract. After infection, epithelial cells become vac-
uolated, lose cilia, and become necrotic. Regeneration of epi-
thelium takes approximately 3 to 4 weeks, during which time
pulmonary abnormalities can persist. Although constitutional
manifestations are pronounced in influenza, viremia is not
usually present in seasonal or pandemic 2009 H1N1 influ-
enza.132–136 However, viremia has been more frequently reported
among human H5N1 infections.137
In contrast to human influenza viruses, avian influenza
viruses were shown to predominantly infect ciliated respiratory
airway epithelial cells133 and to exhibit a binding preference for
lower respiratory tract epithelium because of a predominance of
sialic acid alpha-2,3 receptors preferentially recognized by avian
viruses.138 Further studies are needed to better understand the
distribution of sialic acid alpha-2,3 receptors, for which avian
viruses have a preference, and sialic acid alpha-2,6 receptors,
which are preferred by human influenza viruses, in upper respi-
ratory tract epithelium.
The incubation period of avian H5N1 influenza may be
longer than for classical human influenza; fever, respiratory
symptoms, diarrhea, lymphopenia, and thrombocytopenia
are common symptoms in hospitalized H5N1 patients, but
patients with fever and gastrointestinal symptoms only have
also been rarely reported.101,137 Pneumonia and multiorgan fail-
ure is common, and more than 50% of patients have progressed
to death. Detection of H5N1 virus in extrapulmonary tissues
and in feces, urine, cerebrospinal fluid, and serum has been
reported.101,137 In avian species, a polybasic amino acid sequence
at the cleavage site in the HA molecule is associated with the
systemic spread of highly pathogenic avian viruses.86

Diagnosis
Infections by other pathogens, including respiratory syncytial
virus, adenovirus, parainfluenza virus, rhinovirus, numerous other
respiratory viruses, , 146
and , can result in individual ill-
nesses similar to influenza. Nonetheless, certain epidemiologic
features of influenza outbreaks and epidemics can greatly aid in
the clinical diagnosis of influenza. For example, community-wide
or institutional outbreaks of febrile respiratory illness cases during
winter and spring months are characteristic of influenza, although
respiratory syncytial virus can exhibit a similar epidemiologic
pattern. In most instances, individual cases of influenza are diffi-
cult to identify reliably by clinical examination and routine labora-
tory findings alone, and the appropriate use of influenza diagnostic
tests is helpful. Such tests include virus isolation (culture) and
identification, direct detection of influenza virus in clinical speci-
mens, rapid point-of-care tests, molecular methods, and serologic
tests. Nasopharyngeal and nasal specimens generally have higher
yields than throat swab specimens for detecting seasonal influenza
viruses, and tests perform best when collected as close to illness
onset as possible (eg, 72 hours after onset). Posterior-pharyngeal
(throat) swabs are recommended by the World Health Organization
(WHO) for suspected influenza A(H5N1).137 Endotracheal aspirate
or bronchoalveolar lavage specimens are preferred in patients with
lower respiratory tract illness, and for patients with lower respi-
ratory tract disease, a negative upper respiratory tract specimen
should not be used to rule out influenza when clinical suspicion
for influenza is high.54,139
Although isolation of influenza viruses in cell culture or eggs
followed by hemagglutination-inhibition (HI) testing to identify
the virus was the gold standard for influenza diagnosis for many
years, molecular tests such as real-time reverse-transcription
polymerase chain reaction (RT-PCR) are now more widely
used, are more sensitive than virus culture,126 and have essen-
tially become the new gold standard. A new, less-used molec-
ular approach uses DNA microarrays containing nucleic acid
probes to type and subtype influenza viruses.140,141 However, iso-
lation of influenza viruses is critical for the antigenic analysis
used to identify viruses for use in influenza vaccines. The sensi-
tivity of viral culture depends on when in the course of illness the
specimen is collected and on its quality.142 Results are usually not
available for at least 3 days, although some rapid culture meth-
ods allow virus to be detected within 18 to 24 hours.143 Although
state health department laboratories and some hospital laborato-
ries can type and subtype viral isolates, further antigenic charac-
terization is generally conducted in specialized laboratories.
A variety of sensitive and specific radioimmunoassays, fluo-
roimmunoassays, and enzyme immunoassays can detect viral
antigens in clinical samples. Although these assays can produce
a result within a few hours, they are often less sensitive than
standard virus isolation, and they require specialized laboratory
equipment and reagents.142
Many commercial point-of-care tests are available for rapid
diagnostic testing for influenza. Rapid influenza diagnostic tests
(RIDTs) typically use immunoassays that detect influenza viral
proteins in specimens. Some RIDTs detect influenza A and
B viruses but do not distinguish between them, whereas oth-
ers detect only influenza A viruses, or detect and distinguish
between influenza A and B. In general, RIDTs are useful for rap-
idly determining whether influenza is the cause of outbreaks in
institutions or other settings and for documenting circulation
of influenza viruses in populations of patients. RIDTs appear
to vary considerably by manufacturer and test setting, but
they are generally considered to have high specificity (>90%)
and low to moderate sensitivity (20% to 70%) compared with
other influenza tests (eg, RT-PCR or virus culture). RIDTs
have higher sensitivity when used in young children than when
used in adults, possibly because young children with influenza
Inactivated influenza vaccines 17 261
may shed higher concentrations of influenza viruses than
adults.45,117 The specificity of available RIDTs for detection of
2009 H1N1 virus, like that for detection of other influenza
virus strains, is high ( 95%), but sensitivity is 11% to 70%.144–
During the 2009 H1N1 pandemic, sensitivity of point-of-
care tests was 11% to 70%, with most tests having a sensitivity
of less than 50%; negative test results were viewed as provid-
ing insufficient information to help in making decisions about
treatment or infection control. Specificity remained greater
than 90% in most evaluations.54,147,148
RIDTs are not sensitive for detecting avian influenza virus
infections in humans.149 Collection of additional specimens for
viral isolation or for RT-PCR testing is strongly recommended
to confirm rapid antigen test results for human influenza infec-
tions. However, for persons suspected of having H5N1 avian
influenza infection, viral isolation should not be done except
in laboratory facilities under Biosafety Level 3 enhancements.
Influenza H5N1-specific real-time RT-PCR testing conducted
under Biosafety Level 2 conditions is the preferred method
for diagnosis of human infection with H5N1. In the United
States, all state public health laboratories, several local public
health laboratories, and the Centers for Disease Control and
Prevention (CDC) are able to perform influenza H5N1 real-
time RT-PCR testing, and they are the recommended sites for
initial diagnosis.150
Influenza virus infections can also be detected by measuring
increases in influenza-specific antibody between acute and con-
valescent serum samples. Techniques for measuring antibody
against influenza in sera include HI, virus neutralization,
enzyme immunoassay, and complement fixation. In general, the
HI and virus neutralization tests provide the most strain-specific
results, and a fourfold or greater rise in titer in convalescent-
phase serum compared with acute-phase serum is the gold stan-
dard for serodiagnosis of influenza infection. When appropriate
respiratory specimens are not available, serologic methods may
provide the only means for documenting influenza infection.
Treatment and prevention with antiviral agents
Two classes of prescription medications, adamantanes and NA
inhibitors, have been approved in the United States for use
against influenza virus infections (Table 17-1). The adaman-
tane derivatives amantadine hydrochloride and rimantadine
hydrochloride have specific activity against influenza A viruses
but not B viruses. However, widespread emergence and now
predominance of adamantine resistance among human influ-
enza A(H3N2) and A(H1N1) viruses, and among both human
and animal influenza A(H5N1) viruses, has greatly limited
the usefulness of these medications for influenza treatment
or prophylaxis.151–154 The NA inhibitors zanamivir and osel-
tamivir have activity against both influenza A and B viruses.
In 2008, widespread and unexpected circulation of seasonal
influenza A(H1N1) viruses that were resistant to oseltamivir
further complicated antiviral treatment and chemoprophy-
laxis options.139,155–157 Interestingly, these oseltamivir-resistant
strains retained sensitivity to adamantanes. However, few 2009
pandemic influenza A(H1N1) viruses have exhibited resistance
to NA inhibitors, and as seasonal influenza A(H1N1) virus
strains became rare during and after the pandemic, oseltami-
vir and zanamivir again became the recommended medications
for treatment or chemoprophylaxis.139,158 The four drugs dif-
fer by routes of administration, approved usage for treatment
or chemoprophylaxis for different age groups, adverse events,
costs, and availability.158
Antiviral resistance to the adamantanes among circulating
influenza A viruses is now widespread, and they are not cur-
rently recommended for use. However, these drugs might again
 Inactivated influenza vaccines 17 267

Estimated Rates* of Influenza-Associated Hospitalization by Age Group and Risk Group (Selected Studies)

*
268 SECTION TWO
influenza matrix protein and nucleoprotein), resulting in vac-
cines referred to as subunit or purified surface antigen prepa-
rations. It is also possible to produce purified influenza virus
HA and NA vaccines using recombinant DNA technology,244–247
and clinical studies have been done using vaccines with this
technology.248
The original inactivated whole-virus influenza vaccines were
crude preparations, but the purity of influenza virus vaccines
has steadily increased,235 aided by the introduction of centrifu-
gation and chromatographic steps to reduce residual egg mate-
rials. Since dissolution of the lipid envelope allows retention
of immunogenicity with reduction in reactogenicity, splitting
influenza viruses to produce subvirion preparations has become
routine. An intact viral membrane is essential for infectivity of
enveloped viruses. Therefore, disruption of the viral envelope
adds assurance of viral inactivation. Subvirion vaccines were
first prepared using ethyl ether and polysorbate 80, but a variety
of detergents, including deoxycholate, tri--butyl phosphate,
Triton X-100, Triton N101, and cetyltrimethylammonium bro-
mide, are now used for commercial vaccine preparation.
The development of high-growth influenza A viruses suited to
maximal replication in eggs has helped increase vaccine produc-
tion. Since the early 1970s, influenza virus A/Puerto Rico/8/34
(PR8), a strain very well adapted to replication in eggs, has been
used to develop influenza A virus reassortants that combine the
HA and NA from wild-type viruses with the high-growth prop-
erties of the PR8 donor virus.249,250 Influenza A virus reassor-
tants derived from PR8 are often more uniformly spherical than
wild-type viruses,250 which may facilitate recovery of virus dur-
ing the various processing steps. The growth characteristics of
reassortants vary because the HA and NA also affect the adap-
tation and replication capabilities of the viruses in growth sub-
strates. The overall advantage provided by strains that replicate
well in eggs has meant that most, if not all, influenza vaccines
prepared in eggs over the past 35 years have been prepared using
a high-growth influenza A virus reassortant. An area of continu-
ing investigation is development of reliable high-growth donor
strains of influenza B viruses suited to vaccine production.
Vaccine constituents, including antibiotics and
preservatives
Hemagglutinin is the main immunogen in inactivated influ-
enza vaccines, and its level (ie, potency) is standardized by
single radial immunodiffusion (SRID).251–253 Although inacti-
vated influenza vaccines contain NA, M, and NP in varying
amounts depending on the process methods, their levels are
not specifically quantified. In persons who received inactivated
vaccine, serum antibody to NA was correlated with resistance
to experimental infection with a wild-type influenza A virus.254
The immune responses to other viral proteins (eg, M2)255 are
being investigated, but their relative contributions to protection
appear much less than the effects of antibodies directed against
HAs. Therefore, HA content continues to be the primary con-
cern in preparing inactivated vaccines.
The amount of HA, as measured by SRID, has been corre-
lated with the immunogenicity of subvirion and whole-virion
inactivated influenza virus vaccines.241,242,256 To test com-
mercially prepared influenza vaccines, most authorities favor
SRID performed with influenza strain–specific antigens and
antisera because other methods have correlated less well with
immunogenicity.
Antibiotics are not added to inactivated influenza virus vac-
cines as active ingredients. Aminoglycoside antibiotics are used
in some production schemes to reduce bacterial growth in eggs
during processing steps. Antimicrobial agents associated with
anaphylactic-type hypersensitivity responses are strongly dis-
couraged for use at any phase of production. Antibiotics that are
used in the production process are reduced to trace or undetect-
able amounts during purification of the viral proteins and dilution
of the concentrated vaccine components during formulation of the
final multivalent vaccine. Minimal amounts of the detergents or
solvents used for virus disruption may remain in the final vac-
cine preparation, but purification and dilution steps often reduce
the amount to the limits of detection.
Thimerosal, a mercury-containing compound with broad,
highly effective antimicrobial properties, is present in inacti-
vated influenza virus vaccines produced for multidose contain-
ers. Thimerosal is used to reduce bioburden (the total amount
of bacteria and fungi) during production of influenza vaccines
in eggs, or as a preservative to prevent growth of bacteria and
fungi in multidose containers. Although vaccines in single-dose
containers are formulated to contain no thimerosal, or thimero-
sal in amounts too low to have a preservative effect (such vac-
cines may be called preservative free), multidose containers are
designed to be entered several times, which raises the possibil-
ity of entrainment and growth of bacteria or fungi. Alternative
preservatives such as phenoxyethanol are being evaluated as
part of a continuing trend to limit the amount of mercury pres-
ent in vaccines of all kinds.257
Either formalin or -propiolactone is used in influenza vaccine
to inactivate the viruses. If processed properly, propiolactone
is chemically degraded to levels below the limits of detection.258
Although detectable quantities of formaldehyde persist in inac-
tivated vaccines when formalin is used, the steps used for puri-
fication also reduce the amount of free formaldehyde, which,
if it remains in high concentration, can reduce the potency of
vaccines or interfere with SRID measurements of potency.259–262
Adjuvants, which are substances that are intended to aug-
ment immune responses to vaccine antigens, have not been used
with influenza vaccines currently licensed in the United States.
However, data from studies conducted using candidate influenza
A(H5N1) vaccines, as well as practical experience gained outside
the United States using adjuvanted 2009 pandemic H1N1 vac-
cines (including vaccines with proprietary adjuvants such as
AS03 and MF59, or aluminum salts), have greatly increased
interest in adjuvanted influenza vaccines. Large clinical stud-
ies are underway to evaluate whether risk groups such as the
elderly, who respond less well to standard inactivated influenza
vaccines, might acquire greater protection from adjuvanted vac-
cines that exhibit improved immunogenicity (see “Preparations
available” and “Pandemic vaccines”, later).
Manufacture of vaccine
The requirements of national authorities for influenza vac-
cines generally reflect the guidelines published by WHO,263
but they may include items that are specific to individual
control authorities. To expedite and standardize clinical stud-
ies that are intended to support licensure of new influenza
vaccines, both the FDA and the European Medicines Agency
(EMA) have developed guidelines for industry.264,265 Licensure
of influenza vaccines during pandemics is evolving, as the
need for rapid licensure of a safe and effective vaccine is critical.
The EMA introduced a “mock-up” dossier procedure that pro-
vided a fast track for licensure of pandemic vaccines in the
EU, and this licensing mechanism was used during the 2009
pandemic.266,267 During the 2009 H1N1 pandemic, the FDA
licensed the monovalent 2009 H1N1 pandemic vaccine by rely-
ing on the safety and immunogenicity knowledge base devel-
oped during licensure of previous seasonal influenza vaccines
prepared in the same way. Licensure of vaccines prepared in
ways involves new product licensure pathways that were
not used during the 2009 H1N1 pandemic because licensure
of unadjuvanted vaccines allowed quicker implementation of
the urgently required immunization program. The processes by
which vaccines against the 2009 H1N1 virus were developed
were recently summarized.268
Recommendations for the antigenic composition of influenza
vaccines are made annually to ensure that current influenza vaccines

are effective against recently circulating strains in both the
northern and southern hemispheres. Laboratories from many
countries participate and help determine whether significant
changes have occurred in the antigenicity of the HAs of cir-
culating influenza viruses. Surveillance has made it clear that
antigenic changes occur not only by point mutations (antigenic
drift), but also by way of antigenic shift (at irregular and unpre-
dictable intervals). The WHO global influenza surveillance sys-
tem has been expanded to improve the timely identification of
antigenic changes necessary for updating vaccines.
Influenza vaccine viruses are usually isolates obtained
through the WHO surveillance network. The WHO and vari-
ous national authorities recommend use of certain strains based
primarily on the antigenic characteristics of their HAs and NAs.
Original isolates are passaged in eggs or primary chick kidney
cultures to develop reference strains that are distributed to man-
ufacturers to develop seed viruses. Often the original wild-type
strains grow relatively poorly in eggs, so considerable effort is
expended examining several antigenically similar strains for
several qualities needed to maximize virus yield during large-
scale production, including their potential for development of
high-growth reassortants, their growth characteristics, and their
optimal incubation conditions. Because the available production
time for trivalent influenza vaccine is limited by the need to dis-
tribute vaccine each autumn, the total amount of vaccine that
can be produced is limited by the least productive strain.
Because most influenza surveillance laboratories are not pre-
pared to handle tissue cultures in a way necessary to prevent
introduction of extraneous agents, a concern is that the possi-
ble introduction and amplification of extraneous agents during
subsequent passages could compromise the safety of vaccines.269
However, replication of influenza viruses in eggs might provide
a partial barrier to extraneous agents that might originate in the
clinical source material. The interest in mammalian cell sys-
tems stems from concern that the availability of eggs could be
reduced by an event requiring destruction of the egg-producing
hens (eg, an outbreak of avian influenza or Newcastle disease
virus), and also from concern that the HAs of viruses grown in
eggs may exhibit antigenic alterations.270–273 Therefore, strategies
are being developed to directly isolate influenza viruses in mam-
malian tissue culture to provide suitable starting seed viruses.
Regardless of the growth substrate used, the prevention of
the introduction of extraneous agents is paramount. When
using hen eggs, the main concern is the introduction of extra-
neous agents, such as avian leucosis virus, that originate in
the flocks of chickens providing the eggs. However, extrane-
ous agents could be introduced from the original human host
to mammalian tissue cultures, from the tissue cultures used
by the laboratory recovering the influenza virus, from the cell
substrate used for manufacturing, or from materials used to
support the growth of the cell substrate.274–277 For inactivated
vaccines produced in either eggs or mammalian tissue cultures,
these concerns are reduced when the process used to inactivate
influenza viruses inactivates other microorganisms effectively.
After replication of influenza viruses, a number of manufac-
turing steps are taken to increase the concentration of the active
immunizing ingredients of the vaccines (mainly the viral HAs)
and to reduce other (mainly nonviral) materials. For inactivated
influenza virus vaccines produced using eggs or mammalian
tissue cultures, removal and reduction of egg or tissue culture
proteins occurs throughout the manufacturing process, begin-
ning with concentration of virus by means of centrifugation
through a sucrose gradient or passage of the virus-containing
allantoic fluid over a chromatographic column. The resulting
fluids may be additionally purified by dialysis or diafiltration,
and the concentration of residual egg or cell proteins is reduced
further during the final formulation when the HA concentra-
tion is adjusted to achieve final target levels.
Disruption of the lipid envelope permits further purification
of the viral proteins and, in particular, the HAs and NAs. These
Inactivated influenza vaccines 17 269
proteins form rosette-like structures in which the hydrophilic
heads of the proteins are on the exterior, and the hydrophobic
tail portions of the molecules are buried internally. The dis-
ruption process varies in efficiency, so that HA and NA still
can be attached to lipid but in pieces smaller than the orig-
inal virion. Dissolution of the viral envelope and removal of
additional viral components results in vaccine products with
reduced reactogenicity.240–243,278–280
Chemical inactivation steps are used to minimize the micro-
bial load in the raw viral harvest from the growth substrate, and
careful handling and use of component reagents and buffers also
facilitate inactivated vaccine sterility. However, filtration steps
ensure elimination of undesirable microbes. Although the sterile
filtration step does not eliminate endotoxins, which may have
been formed previously by bacteria in the product, some of the
purification steps can help to reduce endotoxin levels.279 Because
endotoxins contribute to febrile responses to injected vaccines,
the limits for endotoxin levels in the finished product are set
well below clinically established thresholds for reactions.280
Producers
Each manufacturer has a somewhat different, proprietary pro-
cess, but all of the processes result in vaccines standardized to
ensure the immunogenicity of the HA. Worldwide interest in the
availability of inactivated influenza virus vaccine has increased
as demand for vaccines has grown. As an example, annual inac-
tivated influenza vaccine production for use in the United States
increased from approximately 20 million doses in 1989 to more
than 120 million doses in 2009. In addition, more than 100
million doses of an inactivated monovalent 2009 H1N1 vaccine
were also produced in 2009.199 Significant commercial influenza
vaccine production capability exists in many parts of the world,
and the recognition of health benefits combined with the need
for additional capacity to meet global demand for influenza vac-
cine is stimulating further expansion. By 2010, WHO estimated
that the world's manufacturing capacity was approximately
800 million doses of trivalent influenza vaccine, an increase
of 450 million additional doses since 2006. However, despite
this rapid increase in global capacity, more than 80% of sea-
sonal influenza vaccine doses produced in 2009-10 will have
originated from seven large manufacturers that are located
in the United States, Canada, Australia, western Europe,
Russia, China, and Japan. As a result, vaccine availability was
delayed and insufficient for much of the world during the peak
of the 2009 H1N1 pandemic, particularly in areas that had no
in-country manufacturing capacity.281 Information on influenza
vaccine manufacturers can be obtained from the WHO Web site
(www.who.int/csr/disease/influenza/manulist/en).
Preparations available
Monovalent, bivalent, trivalent, quadrivalent, and even pen-
tavalent influenza vaccines have been produced. Usually, the
vaccines have included both influenza A and influenza B virus
components. In recent years, monovalent vaccines have been
used only in unusual circumstances, such as in 1986, when
a supplemental A/Taiwan/21/86 vaccine was produced, or dur-
ing the 2009 H1N1 pandemic, when nearly simultaneous mon-
ovalent 2009 pandemic H1N1 and 2009-10 seasonal trivalent
vaccine programs were both implemented. Since 1978, most
vaccines have been trivalent and have incorporated influenza
A(H1N1) and A(H3N2) subtype viruses and an influenza B
virus. During the 1990s, influenza B viruses diverged into two
antigenically distinct lineages based on the HA.282 This led to
the use of a quadrivalent vaccine in at least one country (The
Netherlands),283 and renewed interest, including plans to apply
for licensure, in a quadrivalent seasonal vaccine in the United
States.192,192a However, the extent to which a quadrivalent prep-
aration could substantially reduce morbidity is dependent on
270 SECTION TWO Licensed vaccines
both vaccine match and whether there are substantial numbers
of influenza B viruses in circulation that are of a different lin-
eage compared to the B component in trivalent vaccine.192b
A number of adjuvants have been investigated for their poten-
tial to increase the immunogenicity of vaccines.284–299 Apart from
aluminum-containing compounds used in some commercial vac-
cines, three adjuvants have so far been licensed in Europe or used
in 2009 H1N1 pandemic vaccines. Two of these adjuvants are pro-
prietary oil-in-water emulsion preparations (MF59 and AS03).235A
third vaccine uses immunopotentiating reconstituted influenza
virosomes.286,289,290 Most adjuvants have combined a lipid or fatty
acid with a bacterial cell wall or bacterial protein, but other mate-
rials such as chitosan, polyinositol, aluminum, and cytokines also
have been evaluated. Generally, adjuvants have modestly improved
seasonal vaccine responses but often at the expense of increased
reactogenicity.293 For example, elderly recipients of an MF59 adju-
vanted influenza vaccine developed geometric mean HI antibody
titers that were 1.5 to 2 times higher against each of the three
influenza vaccine components than those of recipients of stan-
dard subunit vaccine without adjuvant. In addition, fourfold titer
rises were also increased, and there was evidence of a broader sero-
logic protection against drifted strains that circulated 1 and 2 years
after vaccination.289,300–303 Among children, vaccines adjuvanted
with oil-in-water preparations appear to provide equivalent or bet-
ter immunogenicity than unadjuvanted vaccines.304,305 In a trial
using seasonal trivalent vaccines among children 6 to 36 months
old, postvaccination HI antibody titers to all three vaccine strains
were significantly higher with MF59-adjuvanted vaccine than
with unadjuvanted vaccine after one and two doses, and also
among children given a dose 1 year after a two-dose series. MF59-
adjuvanted vaccine also elicited significantly greater serologic cross-
reactivity against A/H3N2 and A/H1N1 strains that were less well
matched with vaccine strains. Similar findings have been reported
for AS03-adjuvanted 2009 H1N1 vaccines.306,307 One randomized,
placebo-controlled trial among children 6 months through 5 years
old demonstrated that efficacy of an MF59-adjuvanted vaccine was
86%, compared with 43% efficacy using an unadjuvanted triva-
lent vaccine.308 The extensive experience gained with adjuvanted
influenza vaccines in Europe, Canada and other parts of the world
during the 2009 pandemic is expected to accelerate plans for more
widespread use in seasonal influenza vaccines.
However, among recipients of vaccine containing MF59,
there were more local reactions (soreness was six times more
frequent and erythema two times more frequent after the first
dose of vaccine containing MF59 compared with vaccine with-
out MF59 in one study of elderly persons193) and more systemic
reactions such as malaise or myalgia.301,303 Among children, a
higher frequency of injection site reactions, most prominently
injection site pain, has also been found to be more common
with MF59-adjuvanted vaccines. However, reactions are gener-
ally mild and self-limited, and experience in clinical studies and
during the 2009 H1N1 pandemic has not indicated that adverse
events other than injection site reactions are common, or that
new serious adverse events will result from use.304,305,309
For the virosome vaccine, influenza virus is inactivated, fol-
lowed by extraction of the HA and NA proteins, which are then
associated with lecithin to produce a conformation in which the
influenza glycoproteins are exposed in a lipid particle.310 When
compared with subvirion vaccine in geriatric patients, injection
of the virosome vaccine elicited higher geometric mean anti-
body titers for the two influenza A components in one study,286
and for the influenza A(H1N1) and B components in another.311
A small study in children with cystic fibrosis suggested that the
virosome vaccine might be more immunogenic than subunit
vaccine after one dose, but the differences were not striking.299
Few clinical efficacy studies comparing standard inactivated
influenza vaccines with adjuvanted vaccines have been pub-
lished.308 However, large clinical comparative trials are under-
way and results are expected as early as 2012.
The components of influenza viruses themselves (including
possibly the lipid envelope) play a role in producing reactions.280
Because whole-virus preparations already include cellular lipids
in the viral envelope, their increased immunogenicity in some
studies may relate to an adjuvant effect of the lipid envelope.
In parallel with the general adjuvant experience, the increased
reactogenicity of whole-virus vaccines (eg, febrile seizures in
children) has restricted their use in some groups because of an
unfavorable risk-benefit ratio.243,278
Adjuvanted vaccines were used in many non-US countries dur-
ing the 2009 H1N1 pandemic,31,312 and immunogenicity data indi-
cated that adjuvanted 2009 H1N1 vaccines produced high antibody
titers after a single dose.30,32,33,306,312–314 Studies using experimental
vaccines made with influenza A virus subtypes including H5 and
H9 have also demonstrated that higher and more broadly reactive
antibody responses were obtained with vaccines containing MF59
or AS03 as an adjuvant.296,315–317 However, some experiences sug-
gest that responses of the immunologically naïve may not always be
improved by incorporation of an adjuvant in influenza vaccines,318
particularly an aluminum-based adjuvant system.319 Systems other
than the classic adjuvant method are being pursued as means to
increase the immunogenicity and possibly effectiveness of inacti-
vated influenza vaccines. Several systems being explored combine a
number of influenza viral proteins (virus-like particles, which gen-
erally contain HA in combination with NA and M1) or incorporate
viral proteins in presentation systems with other agents or com-
pounds to evoke strong cellular and humoral immune responses
(such as inducers of innate immune responses).320–323
Timeline for influenza vaccine production
Production of influenza virus vaccines follows a similar sched-
ule each year, reflecting the need to produce and administer
vaccine before each influenza season (Figure 17-4). The founda-
tion for influenza vaccine production is the global influenza
virus surveillance that continues throughout the year in both
hemispheres. Information from global surveillance is used to
inform manufacturers about trends in antigenic changes in
influenza viruses, particularly trends that might signify the
need to make changes in vaccine composition. Strains show-
ing antigenic changes are examined and, when warranted, high-
growth reassortant viruses are produced.
Because several months are required for production of annual
influenza virus vaccines, the time allotted for collecting and
assessing surveillance data to ensure the best possible recom-
mendations for vaccine composition must be balanced against
the time needed by manufacturers to produce all components of
the vaccine. If recommendations are made too early, significant
antigenic changes may be missed. However, if recommenda-
tions are made too late, the total vaccine output can be ham-
pered. In addition to collecting and analyzing viral surveillance
data, time is needed to allow experience to be gained in han-
dling the vaccine candidate strains and to produce the reagents
needed to standardize the new vaccine component viruses.
Manufacturing of monovalent vaccine components begins
as early as possible and continues until sufficient material is
available to produce the manufacturer's final targeted num-
ber of doses. The potency of monovalent vaccine components
must be known before formulation of the trivalent vaccines can
begin. Currently, the potency of monovalent components and
trivalent vaccine is established using SRID.252,253 This technique
requires both an HA-specific antigen and an HA-specific antise-
rum. The antiserum is produced by immunizing animals with
preparations of purified HA, and it takes at least 3 to 6 weeks.
Once the potency of all three components can be determined,
formulation of the trivalent vaccine begins, followed by filling
the containers, labeling, packaging, and distribution. Despite
increases in vaccine production and distribution of multimil-
lions of doses, all manufacturing steps must be completed

Approximate production timetable for influenza vaccines manufactured for the Northern Hemisphere. The Southern Hemisphere
vaccine production timetable would start in approximately October of each year.
within 6 to 8 months because administration of vaccine typically
begins and is largely completed between October and December
in the northern hemisphere and from April to July in the south-
ern hemisphere. The time needed for production of a mon-
ovalent vaccine against a new pandemic strain would require
approximately 4 to 6 months from identification of a candidate
vaccine virus to the time when the first vaccine doses would be
available. During the 2009 H1N1 pandemic, these timelines
were shown to be approximately correct. However, widespread
availability of the 2009 pandemic H1N1 monovalent vaccine
lagged well behind the peak pandemic wave, again illustrating
the urgent need to more rapidly identify candidate viruses for
vaccine production, develop vaccine virus candidates, and scale
up production of a safe and immunogenic vaccine.281
Dosage and route
Inactivated influenza virus vaccines have been given by
intramuscular, subcutaneous, intradermal, intranasal, and
oral routes. The routes associated with the most reproducible
immunogenicity and lowest reactogenicity have been the intra-
muscular and subcutaneous routes.
Current dosages (based on the SRID content) recommended
for inactivated influenza vaccines that do not contain an adju-
vant are 15 g of each HA component per intramuscular vac-
cine dose for persons 3 years of age and older, and 7.5 g of each
HA component per intramuscular vaccine dose for children less
than 3 years of age. The inactivated influenza vaccine dosages
are based on extensive clinical studies241,242,278 undertaken when
influenza A(H1N1) viruses were reestablished in human popu-
lations during the late 1970s. These studies demonstrated that
children and adults who were unexposed to influenza through
vaccine or natural infection required two doses of inactivated
vaccine to achieve maximal antibody titers, whereas persons
with some degree of preexisting immunity needed only a sin-
gle dose. More recent studies have confirmed the need for two
doses in children less than 9 years of age to optimize the immune
response and vaccine effectiveness.324–327 The dosage chosen bal-
anced the desirability of inducing a maximal immune response
in an individual (a higher dose increases the probability of achi-
eving a maximal antibody titer) against the desirability of max-
imizing population coverage (a smaller dose will allow more
persons to be immunized, which might induce some degree of
herd immunity) and the desirability of minimizing adverse reac-
tions (a larger dose has a greater probability of causing immedi-
ate local and systemic reactions). More recent studies of dosage
responses in elderly persons suggest that increased dosages of
Inactivated influenza vaccines 17 271
modern subvirion vaccines can increase antibody responses
without a substantial penalty of increased reactogenicity.328–330
For example, one study demonstrated that the high-dosage vac-
cine increased the percentage of persons who seroconverted after
vaccination by 12%, 18%, and 25% for influenza B, influenza
A(H3N2), and influenza A(H1N1) virus components, respec-
tively. For the influenza A components, these increases met
FDA immunologic criteria for demonstrating superiority.329 In
2009, the FDA approved a trivalent vaccine containing 60 g
per vaccine strain (ie, a fourfold increased dose) for use in per-
sons 65 or older. Clinical trials underway in 2010-12 among
persons 65 years or older are comparing efficacy of a standard-
dosage trivalent vaccine with high-dosage trivalent vaccines.331
As vaccine availability increases globally, further studies may
be warranted to determine whether vaccine effectiveness can be
improved simply by dosage selection in other specific popula-
tions that respond poorly to standard-dosage vaccine.
The intradermal route requires less antigen than intramus-
cular injection to produce a similar immunologic response,
but it results in more local erythema at the injection site than
other routes.332–337,337a A prefilled microinjection device that
delivers vaccine intradermally more reliably than conventional
needles has been shown to provide equivalent or better immu-
nogenicity at comparable dosages and sometimes lower dosages,
but it also increased injection site–related adverse events.338,339
An influenza vaccine using a microinjection device and 9 g of
each HA per dose is now approved for use in 18- to 59-year-olds
by the EMA, in the United States for persons 18 to 64 years old,
and in Canada for adults (where a 15 g per HA formulation is
recommended for those 60 years old).340–342
The intranasal administration of inactivated influenza vac-
cine has not been studied as extensively as other routes of
delivery.343 A commercially available nasally administered for-
mulation adjuvanted with an heat-labile toxin
was withdrawn because of its association with cranial nerve pal-
sies.344 The ability of oral or intranasal administration to induce
protective antibody responses has been studied, but generally
very large dosages or an adjuvant is required to promote muco-
sal and systemic immune responses.345–347
Vaccine stability
Vaccine manufacturers in the United States and elsewhere have ongo-
ing programs to assess the stability of all vaccine products. Stability
of a product is influenced by the specific formulation of the product,
the addition of stabilizing compounds (such as gelatin or polysor-
bate), the compatibility of the product with the intended container
272 SECTION TWO
and closure and the preparative treatments needed to reduce adsorp-
tion or chemical interaction of the vaccine components with the
container, and the vaccine's specific temperature limits. Stability
assessment programs usually also examine sterility, pH, and the
measurable content of preservatives and other chemical ingredients.
Experience indicates that inactivated influenza vaccines are
sufficiently robust to maintain full potency above the minimum
release specifications for more than 1 year when stored at 4º to
8 º C. However, the potency of individual influenza virus vac-
cine components can decline at varying rates. For example, in
1996, the influenza A(H3N2) component, but not the influenza
A(H1N1) or influenza B components, in the vaccine from one
manufacturer lost potency faster than expected,348 and acceler-
ated potency loss was noted in several vaccine products during
the 2009 H1N1 pandemic.349
Immune responses to vaccination
and correlates of protection
Measurement of antibodies
Influenza vaccination induces primarily antibodies against the
major surface glycoproteins, HA and NA, although in some cases,
antibody to the NP and M1 proteins may also be induced. Although
the HI and neutralization tests both measure levels of strain-spe-
cific antibodies, the HI test142 is most commonly used for measur-
ing the antibody response to inactivated human influenza virus
vaccines. In the HI test, antibodies present in an immune serum
compete with red blood cells (RBCs) to bind the viral HA. RBCs
from chickens or turkeys are used most often; however, guinea pig
or human erythrocytes can be substituted in circumstances where
avian RBCs do not bind well to the influenza viruses in question.
The HI assay can be complicated by nonspecific inhibitors of hem-
agglutination in human and animal serum. The inhibitors can be
removed by pretreating serum with a receptor-destroying enzyme
from or with periodate.
The virus neutralization test is an alternative method for the
detection of strain-specific antibodies to influenza viruses and in
some cases may be more sensitive than the HI test, and it has
the added advantage of measuring functional antibody capable
of preventing virus infection. The microneutralization assay is a
higher throughput test typically using MDCK cells in a 96-well
plate format and an enzyme immunoassay endpoint to measure
neutralizing antibodies in small quantities of serum.350,351 When
detecting serum antibody responses to human influenza strains,
titers of neutralizing antibody measured using this method cor-
relate well with titers obtained by HI.351
However, the neutralization or microneutralization assay
is preferred when detecting antibody responses to avian influ-
enza viruses, as the traditional HI test using avian RBC is not
sufficiently sensitive.352 A modified HI assay that uses horse
RBCs, which express a predominance of sialic acid alpha-2,3
receptors to which most avian influenza viruses preferentially
bind, may also be used as an alternative or confirmatory assay
for the detection of antibodies against avian influenza virus HA.
Finally, enzyme immunoassays are typically performed to mea-
sure virus-specific IgG, IgM, or IgA antibodies in paired prevacci-
nation and postvaccination serum or respiratory samples. Such
immunoassays perform optimally when concentrated or puri-
fied preparations of viral antigen are used to detect antibodies.
Antibody responses
The robustness of the serum antibody response to inactivated
vaccines depends on age and preexisting antibody levels.
Recently, a systems biology approach has been used to identify
expression of kinase CaMKIV as a possible regulator of antibody
responses.353 In previously primed persons, the response is pre-
dominantly anti-HA IgG, but in young, unprimed children,
systemic IgM antibody may be more evident.354,355 In primed
persons, antibodies that are specific for the vaccine strain or
cross-reactive with earlier, antigenically related strains can be
detected using the HI assay. Numbers of influenza virus–spe-
cific antibody-forming cells in peripheral blood peak about
1 week after vaccination, whereas serum antibody levels peak
2 to 4 weeks after vaccination in healthy, previously primed
persons, but they may peak 4 weeks or later in unprimed per-
sons and the elderly.354,356–358 Elderly persons have lower HI
titers as a result of reduced quantities of vaccine-specific
antibodies, rather than because of a lack of antibody avidity or
affinity, according to one study that suggested this was because
they had fewer vaccine-specific plasmablasts and lower concen-
trations of plasmablast-derived polyclonal antibodies after vacci-
nation than younger persons.359 In children 9 years and younger,
and in unprimed persons, two doses of inactivated vaccine are
required to induce an optimal serum antibody response.324–327,360
Inactivated influenza vaccines delivered intramuscularly may
also induce local influenza virus–specific IgA in oral and respi-
ratory fluids, with responses being more substantial in persons
primed by natural infection.357 Multiple immunogenicity stud-
ies conducted with both unadjuvanted and adjuvanted 2009
H1N1 monovalent vaccines demonstrated excellent immuno-
genicity in healthy persons in all age groups.32,33,306,361–364 For
children younger than 9 years, immunogenicity of the H1N1
vaccines matched or exceeded what is typically seen with sea-
sonal influenza vaccines. However, children less than 9 years
old had higher rates of seroconversion and higher postvaccina-
tion antibody levels when two doses were administered.33,364
Increasing the amount of influenza HA per vaccine dose
above the currently recommended level of 15 g generally results
in a dosage-related increase in serum antibody response, with
only modest increases in the occurrence of injection site reac-
tions.360 In the elderly, increasing dosages of inactivated trivalent
vaccine resulted in significantly higher and more frequent serum
antibody responses and may offer one approach to enhancing
the protective effect of inactivated vaccines in this high-risk pop-
ulation.328–330 Multiple studies have shown that increasing the
influenza HA fourfold (60 g per strain) results in significantly
higher anti-HA IgG among older persons, and one high-dosage
vaccine was recently licensed in the United States for use in
persons 65 years or older.36 Studies are underway to determine
whether higher titers predict better vaccine effectiveness.
The monovalent 2009 H1N1 vaccines adjuvanted with
oil-in-water emulsions or aluminum preparations, and the
MF59 adjuvanted seasonal influenza vaccine now available in
several European countries elicit high HA titers using dosages
similar to, or twofold to fourfold less, than dosages used in
unadjuvanted vaccines.30,306,364,365,365a In addition, whole-virion
monovalent 2009 H1N1 vaccines were used in some countries
with immunogenicity data demonstrating satisfactory antibody
responses after a single dose in older children and adults.32 In a
randomized controlled trial comparing the immunogenicity of a
whole virion 2009 H1N1 vaccine (7.5 g HA) versus an AS03-
adjuvanted vaccine (3.8 g HA), adjuvanted vaccine was more
immunogenic based on antibody responses. On day 21 after one
dose of adjuvanted AS03 or whole-virion vaccine, the percent-
age of persons who developed titers of 40 or greater decreased
with age for both vaccines, ranging from 94% and 71% of 18-
to 44-year-olds to 51% and 32% of those aged 65 years or older,
respectively. On day 42 (21 days after the second dose), 100%
and 73% of participants aged 18 to 44 years who received adju-
vanted or whole-virion vaccine, but 76% and 36% of those aged
65 years or older, had titers of 40 or greater, respectively.365
Among persons with chronic medical or immunocompromis-
ing conditions, serologic response rates to vaccination are typi-
cally lower. Trivalent inactivated influenza vaccine (TIV) produces

adequate antibody responses against influenza among vaccinated
persons infected with human immunodeficiency virus (HIV)
who have no or minimal acquired immunodeficiency syndrome
(AIDS)-related symptoms, including those who have successfully
responded to highly active antiretroviral therapy.366–369 Among per-
sons who have advanced HIV disease and low CD4+ T-lymphocyte
cell counts, TIV might not induce protective antibody titers;369,370 a
second dose of vaccine does not improve the immune response in
these persons.370,371 A randomized, placebo-controlled trial deter-
mined that TIV was highly effective in preventing symptomatic,
laboratory-confirmed influenza virus infection among HIV-infected
persons with a mean of 400 CD4+ T-lymphocyte cells/mm3; how-
ever, a limited number of persons with CD4+ T-lymphocyte cell
counts of less than 200 were included in that study.203,371 A non-
randomized study of HIV-infected persons determined that influ-
enza vaccination was most effective among persons with greater
than 100 CD4+ T-cell counts and among those with less than
30,000 viral copies of HIV type-1 per milliliter.372 Initial studies
indicate that the response to 2009 H1N1 antigen is also reduced.
For example, a study among HIV-infected persons on highly active
antiretroviral therapy who received the 2009 H1N1 monovalent
vaccine have shown similar immunogenicity with an unadju-
vanted vaccine,373 improved immunogenicity with an adjuvanted
vaccine (68% with postvaccination HI titers of 1:40), and excel-
lent immunogenicity after two doses of an adjuvanted vaccine (92%
with postvaccination HI titers of 1:40).374 Among HIV-infected
children, use of an MF-59 adjuvanted vaccine elicited significantly
lower seroconversion responses (60%) compared to those among
uninfected children (82%) after 1 dose; after 2 doses seroconversion
responses were 73% and 89%, respectively (p>0.05).374a
Immunogenicity among persons with solid-organ trans-
plants, as measured by the proportion who developed seroprotec-
tive antibody levels, varies according to transplant type and time
since transplant. Those with kidney or lung transplants have
responses that are similar or moderately reduced compared with
healthy persons; however, those with heart or liver transplants
respond poorly, especially if vaccination occurs within 4 months
after the transplant procedure.375–379,379a While antibodies against
the globular head of HA detected by HI assay are generally asso-
ciated with protection, neutralizing antibodies against the con-
served stem region of HA have been detected in adults following
seasonal or 2009 pandemic H1N1 vaccination, although the fre-
quency of these antibodies is much lower than those targeting
the head of the molecule.379a,379b,379c Inactivated influenza vac-
cines also elicit antibodies that inhibit NA activity.379d However,
the NA content in licensed inactivated vaccines is not standard-
ized and varies by manufacturing process and vaccine virus.379e
Inactivated high HA-dose vaccines which also have higher NA
content, induced more frequent and higher titered anti-NA anti-
body responses in older adults than standard dose vaccines.379f
Cellular immune responses
CD4+ and CD8+ T cells also play an important role in immunity
to influenza and, in contrast to the strain-specific response of
antibodies, tend to be more cross-reactive among subtypes, recog-
nizing more conserved epitopes on the surface proteins or inter-
nal viral proteins. The ability to recognize a given T-cell epitope
depends on the human leukocyte antigen (HLA) phenotype of an
individual. CD4+ T cells provide help for the antibody response
and the induction of CD8+ T cells, whereas CD8+ cytotoxic T
lymphocytes (CTLs) have been associated with accelerated clear-
ance of virus and recovery from infection.380 The CTL responses
in humans who have received inactivated influenza vaccines
have not been studied as extensively as the humoral immune
response. Age, type of vaccine, and prevaccination immune sta-
tus influence the quantitative and phenotypic changes in T cells
after vaccination. Immunization of healthy adults with inacti-
vated whole-virus vaccine resulted in enhanced CTL responses
Inactivated influenza vaccines 17 273
in peripheral blood, whereas immunization with subunit
vaccine resulted in a poor CTL response.256,380,381 The duration
of the response varied greatly among individuals, persisting for
several months to 1 year. Children aged 6 months to 9 years who
received inactivated vaccine, including previously unvaccinated
children younger than 5 years who received two doses of vaccine,
demonstrated an increase in interferon- –producing CD8+ T
cells and an accompanying increase in expression of the cytotox-
icity molecule perforin, but similar responses were not detected
in adults.382 In another study, an oil-in-water emulsion adjuvant
(AS03) appeared to induce increased numbers of antigen-specific
memory B cells compared with unadjuvanted vaccine. Cross-
reactive and polyfunctional H5N1-specific CD4 T cells were
identified after a single dose of adjuvanted vaccine, and ampli-
fied by a second vaccination.317 AS03-adjuvanted (3.8 g hemag-
glutinin) 2009 H1N1 vaccine strongly enhanced both antibody
and CD4 T-cell responses as reported in a commentary.34
The number of influenza virus–specific CD8+ T cells
decreases in elderly persons compared with younger adults.383–385
Nevertheless, influenza vaccination may transiently enhance CTL
responses381 in this population. There is growing evidence that
cell-mediated immune responses may be a component in protect-
ing the elderly from influenza disease.386,387 In adults 60 years and
older with congestive heart failure, postvaccination production of
Granzyme B, a key mediator of cytolysis of virus-infected cells,
was significantly lower in subjects with influenza than in those
without laboratory-confirmed influenza.386 These results suggest
that the evaluation of cell-mediated immune responses to vacci-
nation, particularly in older adults, warrants further investigation.
Correlates of protection
Strain-specific virus-neutralizing antibody directed against
the HA is the primary immune mediator of protection against
infection and clinical illness, whereas antibody directed against
the NA restricts virus release from infected cells and may
reduce the severity of disease through enhancing viral clear-
ance.388 Based on serologic studies of influenza using human
serum, an HI titer of 32 to 40 is often referred to as the pro-
tective titer. However, this range of titers represents the level
at which approximately 50% of persons will be protected, and
there is no titer value that can guarantee protection from infec-
tion.205,389–391 Among participants in one randomized, placebo-
controlled vaccine effectiveness trial that demonstrated 68%
vaccine efficacy, lower prevaccination and postvaccination HI
titers were seen in the 22 vaccinated persons who developed
laboratory-confirmed influenza. However, all vaccinated cases
had postvaccination HI titers greater than 32, and HI titer sero-
conversion (fourfold or more increase in titer after vaccina-
tion) did not predict protection, suggesting that patients could
remain susceptible because of other unmeasured factors such
as lack of adequate cell-mediated immunity or protective anti-
bodies directed at other antigens.392 Nevertheless, elevated lev-
els of serum anti-HA antibody are generally correlated with
resistance to influenza infection, whereas lower antibody lev-
els are associated with increased risk of illness among persons
exposed to influenza viruses,390,393–395 although elderly persons
with HI titers of greater than or equal to 40 may still be sus-
ceptible to influenza disease.396 Currently, no similar immune
correlate exists for antibodies detected by neutralization assays.
Challenge studies using live virus have shown that protec-
tion against influenza is also associated with local neutralizing
antibody and IgA found at mucosal surfaces.254,355 In human
experimental challenge studies in healthy young adults, serum
antibodies to NA were associated with protection from infec-
tion after experimental challenge with wild-type influenza A
viruses.254 In older adults, measurements of cellular responses,
in combination with serum antibody responses, may provide
better correlation with protective efficacy.397 In summary, while

and circulating strains, the vaccine was 25% against influenza-
like illness and 49% against P&I for two doses, but for those
who needed two doses but received only one, the vaccine was
not effective against influenza-like illness and was 22% effec-
tive for P&I.326 Another two studies from the same year con-
firmed the need for two doses of vaccine in children less than
5 and less than 2 years, respectively, to optimize vaccine effec-
tiveness.324,327 Two studies on the immunogenicity of one ver-
sus two doses of influenza vaccine in 6- to 23-month-old and in
5- to 8-year-old children also confirmed the need for two doses
of influenza vaccine to provide optimal protection against influ-
enza in children younger than 9 years being vaccinated for the
first time.325,522
Three studies reported that vaccination decreased influenza-
related otitis media by approximately 30% to 50%,510,512,523 and
another study found that a virosomal vaccine protected against
otitis media among children with a history of recurrent otitis
media.524 However, a 2-year randomized study reported that
vaccination did not reduce cases of otitis media, even when the
vaccine reduced respiratory illness caused by culture-confirmed
influenza by 66%; the efficacy against laboratory-confirmed
influenza-related otitis media was not reported.508 Overall, vac-
cine efficacy estimates in school-age children are similar to
those found in healthy adults. Data for younger children are
limited but suggest that efficacy could be somewhat lower in
the younger age groups.
Limited studies have been conducted among children with
chronic medical conditions. In a nonrandomized study, vaccine
reduced culture-confirmed or serologically confirmed influenza
(against a drifted strain) by 22% to 54% in asthmatic children
2 to 6 years old and by 60% to 78% in 7- to 14-year-olds.511
A retrospective analysis using a computerized primary-care
database estimated that vaccine reduced medically attended
visits for respiratory illness or otitis media among asthmatic
children by 27% (95% CI, 7 to + 51%) for those 0 to 12 years
old, by 55% (20%-75%) for those less than 6 years old, and by
5% ( 81% to + 39%) for those 6 to 12 years old.514 An analy-
sis of military beneficiaries concluded that vaccination reduced
asthma exacerbations,525 and recent structured reviews have
found that asthma exacerbations are not increased among those
vaccinated.526,527 Overall, vaccine efficacy may be lower among
high-risk children, particularly immunosuppressed children,
than among healthy adults or older healthy children.
Although the risk for complications caused by influenza is
much higher for children less than 6 months old than for older
children, influenza vaccines are not approved for use in this age
group. Limited data indicate that inactivated influenza vaccines
are immunogenic and safe in infants as young as 6 weeks.493,494
However, infants with preexisting maternally derived antibody
have significantly lower seroresponse rates to vaccination,493
and vaccine effectiveness studies are needed in this age group.
Effectiveness of 2009 H1N1 monovalent vaccines
Preliminary estimates for an AS03-adjuvanted, monovalent
2009 H1N1 vaccine indicated very high vaccine effective-
ness. In a study conducted in Germany, vaccine effectiveness
was estimated to be 97% among persons aged 14 to 59 years,
and 83% among those 60 years or older, using a screening
method that compared vaccination rates in the general popu-
lation with rates for those who were ill (and probably overesti-
mated effectiveness to an uncertain degree).528,529 In the United
Kingdom, a case-control study estimated vaccine effectiveness
for all ages, in a sample drawn from among persons who sought
medical care for influenza-like illness, to be 72% in preventing
laboratory-confirmed infection.530 In Canada, a study that used
the test-negative case-control approach estimated effectiveness
of an AS03-aduvanted monovalent vaccine to be 93%, with
most persons in the sample either children or young adults.531
Inactivated influenza vaccines 17 283
One case-control study conducted in the United Kingdom
among children during the 2009 H1N1 pandemic indicated
that the AS03-adjuvanted vaccine was 96% to 100% effective
in preventing laboratory-confirmed influenza after one dose.532
Among children in Canada, a matched case-control study esti-
mated effectiveness of an AS03-adjuvanted pandemic vaccine
in preventing hospitalization to be 85% after a single dose.533
A study among sentinel medical practices in seven European
countries estimated effectiveness to be 72% overall, includ-
ing 78% in patients less than 65 years old, and 73% among
persons without chronic disease.534 In Stockholm, Sweden,
effectiveness of the AS03-adjuvanted monovalent 2009 H1N1
vaccine was estimated be 87% to 93% against medical visits
or hospitalization resulting from influenza.535 The effective-
ness of an AS03-adjuvanted monovalent 2009 H1N1 vaccine
in preventing laboratory-confirmed influenza among persons
with chronic medical conditions varied by age in a case-con-
trol study from the United Kingdom: among children younger
than 10, effectiveness was 77%, and among 10- to 24-year-
olds it was 100%, with lower confidence limits excluding 0%.
However, among adults 25 to 49, and 50 years or older, effec-
tiveness was 22% and 41%, respectively, with lower confidence
bounds including 0%, indicating that no significant effective-
ness was demonstrated.536 Estimates of the effectiveness of
the 2009 H1N1 monovalent vaccines that did not contain
adjuvant are limited, because the fall pandemic wave pre-
ceded large-scale vaccine use. One study using a test-negative
control design estimated the effectiveness of the unadjuvanted,
monovalent, inactivated vaccine to be 62% overall in prevent-
ing influenza-associated medical care visits. However, effec-
tiveness was convincingly demonstrated only in persons 10 to
49 years old; wide confidence limits in older and younger age
groups included 0%.537
Duration of immunity and protection
Inactivated seasonal vaccine induces a rapid systemic and local
immune response in healthy young adults.216 It has been shown
that up to 90% of normal subjects develop serum HI titers of 40
or greater within 2 weeks of vaccination, and that second doses
provide little or no further increase in titers.
Elderly subjects generally respond less well to influenza vac-
cines than young healthy adults and those with chronic debil-
itating medical conditions generally respond less well than
healthy subjects of similar age. Up to 50% of elderly vaccin-
ees may fail to respond to standard doses of inactivated influ-
enza vaccine with a fourfold increase in HI antibodies.310 In
addition, antibody responses in the elderly may be somewhat
delayed compared with those in younger persons.538 Although
some studies have noted that antibody responses return toward
the baseline more rapidly than in young adults,539,540 the rel-
evance of these observations to prevention of influenza illness
is uncertain.457
The duration of protection from illness after influenza vac-
cination has been studied in several clinical trials. In 1968,
a group of schoolchildren was vaccinated with the A/Hong
Kong/68 vaccine and was observed over three successive influ-
enza epidemics caused by the A/Hong Kong/68 virus. Three
years after vaccination, the vaccine was still 67% effective in
preventing influenza.541 In randomized trials conducted among
healthy college students, immunization with trivalent inacti-
vated vaccine before the 1982-83 epidemic provided 92% and
100% efficacy against influenza H3N2 and H1N1 infection–
related illnesses, respectively, during the first year, and a 68%
reduction against H1N1 infections during the second year with-
out revaccination.542 Because elderly persons who have received
repeated influenza vaccinations develop lower peak HI titers,
and these antibody levels return to baseline faster than in young
healthy adults receiving influenza vaccine for the first time,
284 SECTION TWO
immunity may be of shorter duration in this target group than
in the young adults just mentioned.
Although protection may persist for longer than a year
after vaccination in young healthy adults, annual immuniza-
tion with inactivated vaccine is recommended because one
or more vaccine antigens are usually updated each year, and
because reductions in serum antibody levels have been well
documented during the year after immunization. In partic-
ular, annual immunization of those 65 years of age and older
close to the influenza season will help maximize antibody levels
and protection in this important target group.
In contrast to the relatively short-lived antibody response
to vaccination, immunity to HA can persist for decades after
natural infection. During 1977 and 1978, influenza A(H1N1)
viruses similar to those circulating in 1950 reappeared and
spread throughout the world. Persons born before 1950 were
not affected, indicating that substantial immunity remained
after more than 20 years. Furthermore, studies of the immu-
nogenicity of vaccines against the H1N1 virus found high rates
of preexisting immunity among those born before 1957.543 In
contrast, disease occurred in persons less than 20 years of age,
regardless of previous infection by influenza A(H3N2) viruses,
and antibody studies confirmed that these persons were much
more likely to require two doses of H1N1-containing vaccine
to demonstrate an immune response than would be expected
for an immunologically naïve population, showing that inter-
subtypic immunity in humans is weak. The ability of natural
infection to engender long-lasting immunity was dramatically
demonstrated during the 2009 influenza A(H1N1) pandemic,
when approximately 20% to 30% of adults older than 60 had
cross-reacting antibodies to the pandemic strain, presumably as
a result of exposures to antigenically similar strains 50 or more
years previously and possibly subsequent boosting through
exposure to related strains.186,188 In addition, the surprisingly
low burden of severe illness in older populations worldwide,
particularly in contrast to seasonal influenza, suggested that
unmeasured partial or full immunity after natural infection was
retained in an even larger proportion of older persons.544,545
Effectiveness of vaccination for decreasing
transmission to contacts
Influenza vaccination has been shown in certain settings,
such as nursing homes and schools, to provide indirect ben-
efits to contacts who remain unvaccinated;395,505,507,546,547 this
finding is consistent with models of vaccine-induced herd immu-
nity. Furthermore, mass immunization among those who are
most likely to acquire and transmit influenza (usually school-
age children) might reduce the overall occurrence of influenza
in the community, including in unvaccinated persons.395,548,549
The community-level protective effect of vaccinating chil-
dren was shown during the 1968 pandemic, when vaccination
of more than 85% of schoolchildren in a community resulted
in lower illness rates among all age groups than in an unvac-
cinated community.395,550 Several recent studies have evaluated
the benefits of vaccinating school-age children with influenza
vaccine. Overall, studies reported decreases in influenza-related
illness in children, and, usually, modest benefit to adults, either
in the household or in the community.509,548,551,552 One cluster-
randomized trial conducted among Hutterite communities in
Canada randomized colonies to vaccination or no vaccination of
school-age children, and then assessed PCR-confirmed influenza
in children and among adults in each community. This study
found a protective effect of 61% among nonvaccine recipients
in colonies randomized to influenza vaccination of children.553
Although studies have not consistently demonstrated com-
munity benefits, the variability by season, vaccine coverage, and
circulating strains, as well as difficulty in monitoring outpa-
tient illness among adult contacts, have presented formidable
challenges to the conducting of this type of study. For exam-
ple, one community study in which an estimated 45% of
school children were vaccinated showed a 35% reduction in
laboratory-confirmed influenza-related emergency department
visits among children, compared with a control community,
but found no reduction among adults.553 Community impact
of pediatric vaccination on morbidity and mortality beyond
the household effect is difficult to quantify, and more studies
are needed to assess the costs and benefits to children and the
larger communities of school-based or universal vaccination
programs.554 (For further discussion, see Chapter 71).
Community level, multiyear studies have also shown popu-
lation-level impact. A retrospective study suggested that higher
rates of vaccination of schoolchildren in Japan had reduced excess
mortality rates among older adults, and that discontinuation of
this vaccination program led to increases in excess mortality
rates among elderly Japanese.555 This study remains controver-
sial because aging of the population and other variables might
not have been adequately controlled.556,557 The largest study to
examine the community effects of increasing overall vaccine cov-
erage described the experience in Ontario, Canada, which is the
only province to implement a universal influenza vaccination
program, beginning in 2000. On the basis of models developed
from administrative and viral surveillance data, influenza-related
mortality, hospitalizations, emergency department use, and phy-
sicians' office visits decreased substantially more in Ontario
after program introduction than in other provinces, with the
largest reductions observed in younger age groups.558 In addition,
prescribing of influenza-associated antibiotics was substantially
reduced compared with other provinces.559
Decreasing the transmission of influenza from caregivers
and household contacts might reduce influenza-like illness
and complications among persons at high risk. Observational
studies have demonstrated that vaccination of health care per-
sonnel in a nursing home facility is associated with decreased
deaths among patients.210,461 More recently, controlled trials
have shown decreases in mortality and influenza-like illness
in facilities where higher vaccine coverage levels among staff
were achieved.459,460 A comprehensive review concluded that
vaccination of health care personnel in settings where patients
were also vaccinated provided significant reductions, among
elderly patients, in deaths from all causes and deaths from
pneumonia.560
Safety
Common adverse events
In clinical trials, the most frequent adverse events associ-
ated with inactivated vaccines, both adjuvanted and unadju-
vanted, are local acute inflammatory reactions.241,242,278 Pain,
erythema, and induration, which are generally mild and rarely
interfere with daily activities, occur at the site of vaccine admin-
istration in up to 65% of recipients, more commonly with adju-
vanted than with unadjuvanted vaccines.315 Local reactions
rarely persist for longer than 24 to 48 hours. Although infection
and bruising are possible, they occur no more often than with
other injections.
The most common systemic reactions include fever, myal-
gia, arthralgia, and headache. These reactions have occurred
much less frequently (generally 15%) than local reactions and
have been seen most often in very young children and in oth-
ers exposed to influenza virus vaccines or to one of the antigens
in the vaccine for the first time.241,243,256,279,561 However, such
systemic side effects have not been observed in more recent
randomized trials of inactivated vaccines, where the only dif-
ference between placebo and vaccine groups have been sore
arm and redness at the injection site,561 and a recent 15-year

review of adverse-event reports did not identify any new safety
concerns.562 In young children, whole-virus influenza vaccines
and larger vaccine dosages appear to produce more systemic
reactions, including fever, than subvirion vaccines, so subvi-
rion inactivated vaccines are preferred for children.243,278 Recent
summaries of vaccine safety data for pregnant women con-
cluded that inactivated vaccine has an excellent safety record
in this population, with no identified adverse events related to
pregnancy outcomes.501,563
Large postlicensure studies in the United States among chil-
dren enrolled in health maintenance organizations strongly
support the safety of inactivated influenza vaccines. Among
approximately 250,000 children included in one study, there
was no increase in medically attended events during the 2 weeks
after vaccination.564 A study among approximately 45,000 chil-
dren 6 to 23 months old found no statistically significant asso-
ciation with any medical events other than a small increase in
gastritis and duodenitis (typically, self-limited nausea and vom-
iting). However, statistically significant decreases in upper respi-
ratory illness, asthma, and otitis media were also observed.565
Clinical immunogenicity and safety studies of the 2009
H1N1 monovalent unadjuvanted vaccines indicated that the
reactogenicity profile in children and adults is similar to sea-
sonal influenza vaccines.361,362 Ongoing comprehensive safety
monitoring of the pandemic 2009 H1N1 vaccine was imple-
mented as part of the pandemic immunization program, and
the recent estimation of background rates of potential adverse
events (in absence of vaccines) provides useful context for safety
data.566,567 Data from the Vaccine Adverse Event Reporting
System (VAERS) and the Vaccine Safety Datalink indicated that
the 2009 H1N1 monovalent vaccine had a safety profile similar
to that observed for seasonal influenza vaccines.568
During the 2009 H1N1 pandemic, most other countries
used vaccines that included oil-in-water emulsions as adju-
vants. Safety monitoring for these vaccines demonstrated that
adverse events were rare and typically not serious, consisting
mostly of local reactions, fever, or malaise. In France, a sponta-
neous adverse-event reporting system similar to VAERS found
an overall rate of adverse events of 35 to 102 per 100,000 vac-
cine doses (varying by vaccine manufacturer). Serious events
were uncommon, and the study's authors concluded that the
data did not reveal evidence of any unexpected safety issues.569
Similar findings were reported from the Netherlands, where
fever was common but transient.570
Rare adverse events
The main concern identified in large-scale use of influenza virus
vaccines has been a rare neurologic syndrome characterized by
ascending paralysis, paresthesia, and dysesthesia, and oligoclo-
nal antibodies in the cerebrospinal fluid in the absence of cere-
brospinal fluid pleocytosis. Guillain-Barré syndrome (GBS) is a
rare condition with an annual incidence of 10 to 20 cases per
1 million adult population,571 and it has been associated with
many respiratory and gastrointestinal illnesses and, in particular,
infection with species.572 Although the etiology
is not entirely understood, there is evidence that the induction of
antibodies to glycolipids is central to the neuropathic changes.573
Most patients make a complete or near-complete recovery, and
plasmapheresis and administration of immune globulin appear
to speed recovery.318 However, paralysis of respiratory muscula-
ture requiring assisted ventilation can occur, and the case-fatal-
ity rate, approximately 6%, increases with age.571,574
During the 1976 swine influenza virus immunization cam-
paign in the United States, an increased incidence of GBS
occurred in vaccine recipients. The increase in GBS cases above
the background rate was approximately one case for every
100,000 persons vaccinated.575 Subsequent observational stud-
ies in the United States did not identify a similar increase in
Inactivated influenza vaccines 17 285
GBS cases among recipients of inactivated vaccine between
1977 and 1991.576,577 However, a study of the influenza seasons
from 1992 to 1994 estimated an increased incidence of approxi-
mately one GBS case per million recipients of inactivated influ-
enza vaccine during the study years,578 which is substantially
less than the risk of developing severe complications from influ-
enza infection. A more recent study from Canada also found
an increased risk of GBS among adult vaccinees, with a simi-
lar incidence of approximately 1 per million persons vaccinated
above expected GBS incidence.579
In the years since the 1976 swine influenza vaccine cam-
paign, the question of whether GBS is associated with inac-
tivated influenza vaccine has proved difficult to resolve by
epidemiologic studies, given the low incidence of GBS overall
and the even lower possible risk of influenza vaccine–associated
GBS. Data from VAERS have found decreased reporting of GBS
occurring after vaccination over time, despite overall increased
reporting of non-GBS conditions occurring after admin-
istration of influenza vaccine.580,581 Data from the United
Kingdom's General Practice Research Database (GPRD) indi-
cated that influenza vaccine was associated with a decreased
risk for GBS, although whether this was associated with pro-
tection against influenza or confounding because of a “healthy
vaccinee” effect (eg, healthier persons might be more likely to
be vaccinated and also be at lower risk for GBS) is unclear.582
Another GPRD analysis identified no association between
vaccination and GBS for a 9-year period; only three cases of
GBS occurred within 6 weeks after administration of influ-
enza vaccine.583 A third GPRD analysis indicated that GBS
was associated with recently preceding influenza-like illness,
but not with influenza vaccination.584 Another study has also
indicated that GBS risk was associated with influenza illness,
rather than influenza vaccine, in recent years.584a
Data from the US safety systems monitoring influenza
A(H1N1) 2009 monovalent vaccines initially suggested a weak
signal between receipt of H1N1 vaccine and GBS in some safety
monitoring systems; other systems found weak signals associated
with vaccination and Bell's palsy or thrombocytopenia. However,
the National Vaccine Advisory Committee noted that additional
data needed to be analyzed before concluding whether the sig-
nals were spurious or if they represented a true association.585 No
patterns of adverse events related to maternal or fetal outcomes
were observed in VAERS.585a A subsequent multivariate analysis
within the Vaccine Safety Datalink did not confirm the signal for
an association with Bell's palsy.585b In the United States, the asso-
ciation between GBS and unadjuvanted H1N1 influenza vaccine
was estimated to be similar to previous estimates of approxi-
mately 1 GBS case per 1 million doses.586 Similar findings were
reported with the adjuvanted vaccine in Europe. 586a,586b
No experience similar to the 1976 vaccine campaign has
occurred during seasonal or pandemic influenza vaccination
campaigns,568,581 and the Advisory Committee on Immunization
Practices (ACIP) has stated that the potential benefits of influ-
enza vaccination in preventing serious illness, hospitalization,
and death greatly outweigh the possible risks for developing
vaccine-associated GBS.36 The CDC and the FDA jointly main-
tain the Vaccine Adverse Event Reporting System, a vaccine
safety surveillance system for continuously monitoring reports
of GBS and other adverse events potentially linked to influenza
and other vaccines.568,580,586
Febrile seizures after vaccination have been reported. Seizures
and fever were the leading serious adverse events reported to
VAERS.587,588 However, an analysis of data obtained from the
Vaccine Safety Datalink, which, unlike VAERS, provides data
that can be used to estimate adverse event incidence rates, did
not confirm an association between febrile seizures and influ-
enza vaccination.565 In April 2010, Australia's Therapeutic
Goods Administration reported preliminary data indicating
an elevated risk for febrile reactions, including febrile seizures,
 Inactivated influenza vaccines 17 289

avian influenza H9 and H7 subtypes, which have also been

transmitted from avian species to humans, are also considered
a pandemic threat for which vaccines are needed. In addition,
the 2009 H1N1 pandemic demonstrated that the appearance
of antigenic variants of HA subtypes that currently circulate in
humans, or subtypes that circulated before a substantial frac-
tion of the current population was born, can cause a pandemic.
For example, since H2N2 viruses circulated in humans from
1957 until 1968, persons born after this time lack immunity to
the H2 subtype and are therefore a susceptible population for
reintroduction of the H2 subtype. In addition, sporadic human
cases of Eurasian-lineage swine H1 infection have been docu-
mented. This H1 HA is most closely related to the avian H1;
thus humans have little to no immunity against this swine
virus. Swine H3 viruses are also drifting farther from human
H3 viruses, potentially increasing the vulnerability of humans
to swine-origin H3 viruses.622 Active investigation of all human
infections with novel influenza viruses is essential to identify
potential pandemic threats and to develop candidate vaccine
seed strains when indicated.
An emerging pandemic virus will create a surge in
global vaccine demand, and new approaches for immuni-
zation strategies may be needed to optimize protection of
unprimed persons when vaccine antigen may be limited.
The manufacture of vaccines from pathogenic avian influenza
viruses by traditional methods is not feasible for safety rea-
sons as well as because of technical issues. Strategies adopted
to overcome these include the use of reverse genetic systems
to generate modified recombinant strains,623 the use of baculo-
virus-expressed hemagglutinin or related nonpathogenic avian
influenza strains, production of vaccines using mammalian cell
lines (Madin-Darby canine kidney [MDCK] or Vero),235,237,238 and
the use of adjuvants to enhance immunogenicity.235 After the
emergence of HP H5N1 viruses in humans in 1997, two vac-
cine strategies were developed to overcome these limitations.
One approach was to use a surrogate apathogenic H5 vaccine
strain that was antigenically related but not identical to the HP
H5N1 strain, overcoming the need to grow and purify a vac-
cine under high-containment conditions. A surface-antigen
vaccine based on the apathogenic A/duck/Singapore/97 (H5N3)
virus, which possessed an HA that was antigenically similar to
A/Hong Kong/156/97 (H5N1), was administered with or with-
out MF59 adjuvant in two doses of 7.5, 15, or 30 g of H5
HA given 3 weeks apart.296 Although both vaccines were well
tolerated, the nonadjuvanted vaccine was poorly immunogenic,
with only a 36% response rate after two 30- g doses of vac-
cine. Persons who received the H5N3 vaccine formulated with
MF59 achieved significantly higher antibody responses, with a
majority of them showing seroconversion to the vaccine strain.
Follow-up revaccination 16 months later, using the same vac-
cine formulation, substantially boosted antibody titers in those
receiving the adjuvanted, but not the nonadjuvanted, vaccine.624
Boosting with the adjuvanted 1997 H5N3 vaccine, but not the
nonadjuvanted vaccine, also induced cross-reactive antibodies
against H5N1 strains isolated from humans in 2004.625 These
results suggested that the use of adjuvants may be beneficial
not only for dose-sparing of vaccine but also for enhancing the
cross-reactivity of the antibody response.
290 SECTION TWO Licensed vaccines
Traditionally prepared inactivated vaccines against avian
H9N2 or an early human H2N2 and one MF59 adjuvanted vac-
cine have also been evaluated in human studies.315,638,639 In one
study, a subunit vaccine was compared with a whole-virus A/
Hong Kong/1073/99 (H9N2) vaccines in subjects aged 18 to
60 years. Surprisingly, persons older than 32 years were found to
have reactivity to prevaccination sera to the H9N2 virus. This
age-related antibody response was attributed to cross-reactivity
with human influenza viruses that had been encountered early
in life, and this priming was sufficient to elicit a level of anti-H9
HA antibody response that was associated with protection after
a single dose of the H9N2 vaccine. However, in persons younger
than 32 years, who were essentially unprimed, even two doses
of vaccine was suboptimal, because a significant number of vol-
unteers failed to achieve antibody titers associated with pro-
tection. In this naïve population, the whole-virus vaccine was
more immunogenic than the subunit vaccine.639
In another study, alum adjuvant was shown to enhance
responses to whole-virus H9N2 and H2N2 vaccines, although,
once again, unprimed persons required two doses of vaccine
to achieve maximal mean antibody titers.638 Vaccine doses of
less than 15 g formulated with adjuvant induced antibody
titers similar to those induced by unadjuvanted full-dose vac-
cine. In the study of MF59-adjuvanted vaccine, seroconversion
was substantially higher among those who received the MF59-
containing vaccine, with similar responses for the dosage range
of 3.75 to 30 g of antigen. 315 In another study that examined
immune responses among subjects who received either unad-
juvanted whole-virus, alum-adjuvanted whole-virus, virosomal,
or intradermal whole-virus vaccines, immunologically naïve
subjects less than 40 years old required two doses of vaccine to
develop an immune response thought to correlate with protec-
tion, and alum-adjuvanted vaccines were most immunogenic.
Among immunologically primed subjects older than 40 years,
one dose of whole-virus or alum-adjuvanted vaccine induced
adequate immune responses.640
A recent study evaluated an inactivated split-product H7N1
vaccine produced in PER.C6 human cells administered to
adults, either alone or with aluminum hydroxide adjuvant.
Although alum adjuvant augmented responses, overall antibody
titers elicited by two doses of the H7N1 vaccine were modest,
suggesting that, as for H5N1 vaccines, more effective adjuvants
may be needed to improve the immunogenicity of candidate
pre-pandemic H7 vaccines.641
Taken together, these studies suggest that in unprimed pop-
ulations, two doses of inactivated avian influenza vaccines
are necessary to elicit a protective antibody response to avian
influenza viruses. Furthermore, the studies demonstrated the
feasibility of using adjuvants to either increase vaccine immu-
nogenicity or to reduce the dosage of vaccine antigen, which
may be important to extend a limited vaccine supply in a pan-
demic situation.
Public health considerations
Vaccination coverage levels
WHO encourages influenza vaccine use in persons at increased
risk for complications of influenza in all countries where epi-
demic surveillance is well established and where reduction of
influenza and its complications are public health priorities.
As the substantial annual burden of influenza related morbidity
and mortality has been appreciated, interest in expanding rec-
ommendations for annual vaccination to broader population
groups such as healthy children, and the number of countries that
have vaccination programs, has increased. The World Health
Assembly has stated that “…influenza may be a larger public
health problem in poor societies than realized…” and “strongly
encourages the implementation of epidemiologic surveillance,
disease burden assessments and, where appropriate infrastruc-
ture is available, demonstration projects to estimate the impact
of vaccination on disease in poor countries”. It also has urged
Member States with influenza vaccination policies to increase
vaccination coverage of all people at high risk, setting goals in
2003 that were not reached, including vaccination coverage
of elderly people of at least 50% by 2006 and 75% by 2010.642
Additional recommendations included targeted influenza vac-
cination when feasible for groups of people at increased risk for
severe illness and premature death due to influenza. High risk
groups in order of WHO prioritization include: elderly and dis-
abled persons, who live in institutionalized settings, elderly per-
sons with underlying chronic medical conditions, persons older
than six months with underlying chronic medical conditions,
elderly persons above a nationally defined age limit regardless of
underlying health status, and other groups defined by national
data and vaccine capacity including pregnant women, children
6 to 23 months of age, health care workers, and contacts of per-
sons in high risk groups. In addition, the WHO specifically rec-
ommends influenza vaccination for all pregnant women during
influenza season to protect both pregnant mothers and their
newborn infants. Furthermore, WHO has indicated that “explo-
ration of the safety and cost-effectiveness of introducing influ-
enza vaccination into national immunization programmes is
clearly warranted”.642 Vaccination programs capable of deliver-
ing annual influenza vaccination to a broad range of the pop-
ulation could potentially serve as a resilient and sustainable
platform for delivering vaccines or other medical assistance,
and monitoring outcomes for other urgently required public
health interventions.
Recommendations for influenza vaccination vary among
countries, with most recommending influenza vaccination for
patients with cardiopulmonary disorders and for older adults,
somewhat fewer countries recommending vaccine for those
with other medical conditions and few with recommendations
for vaccination of young children.595,643 Vaccine coverage varies
considerably among European countries that recommend rou-
tine vaccination. For example, in 2009, coverage among older
adults (>65 years) was over 80% in The Netherlands, but <50%
in many other countries.595a Although influenza vaccine use
increased dramatically during the 1990s in a number of devel-
oped countries in Europe and the Americas, vaccine distribution
levels (ie, the number of doses distributed per 1,000 total resi-
dent population) among these countries have continued to vary
5 to 10 fold or more during the past two decades. In May 2006,
a global action plan was developed that sought to increase influ-
enza vaccine production capacity to the point that two billion
people could be immunized within 6 months after a pandemic
virus vaccine candidate became available.644 In 2009, annual
global production capacity was estimated at 5 billion doses.645
However, only 534 million doses were produced during the
first 6 months of production during the 2009 H1N1 pandemic,
and 1.37 billion were projected to be available within 1 year of
monovalent vaccine production. This less than anticipated pro-
duction in the first year was attributed to a variety of reasons,
including the greatly reduced production yields, use of vac-
cines with HA content that was not as dose-sparing as could
have been achieved, use of unadjuvanted vaccine formulations,
shrinking vaccine demand, and the switch to Southern and
Northern Hemisphere trivalent seasonal vaccine production.281
In addition, nearly all of the world's vaccine produc-
tion capacity is contained in the Western Europe, Western
Pacific and North American WHO regions, and countries
in these regions also utilize much of the vaccine produced
(Figure 17-6).281 Thus, comparatively little vaccine is admin-
istered outside of these countries and in the developing world.

LAIV appeared to reduce more-severe disease in older healthy
adults, more data are needed before the live vaccine can be indi-
cated for those older than 50 years.
Special Considerations and Contraindications
The following precautions should be considered
before vaccinating with LAIV:

t
LAIV is not indicated for persons younger than 2 years.

t
LAIV is not indicated for persons 50 years and older.

t
LAIV is contraindicated for persons with a history of
hypersensitivity, especially anaphylaxis, to any component
in LAIV, including eggs.

t
LAIV is contraindicated for children or adolescents receiving
aspirin or other salicylates (because of the association of
Reye syndrome with wild-type influenza infection).

t
Persons with a history of Guillain-Barré syndrome should
not receive LAIV.

t
Pregnant women should not receive LAIV unless clearly
needed. This recommendation is based on a theoretical
concern: there has been no indication of fetal harm from
LAIV in animals studies, and it is not known whether LAIV
can cause fetal harm when administered to a pregnant
woman, or can affect reproductive capacity.

t
Persons with asthma, reactive airways disease, or other
chronic disorders of the pulmonary or cardiovascular
systems; persons with other underlying medical conditions,
including metabolic diseases such as diabetes, renal
dysfunction, and hemoglobinopathies; and persons with
known or suspected immune deficiency diseases or receiving
immunosuppressive therapies should not receive LAIV.
Close contacts of severely immunosuppressed persons (eg, bone
marrow transplant patients while they are in a protected envi-
ronment), if given LAIV, should avoid close contact with the
immunosuppressed patient for 7 days.
For vaccination of healthy persons aged 2 to 49 years in
close contact with members of all other immunocompetent
high-risk groups (eg, persons with heart disease, lung disease,
or diabetes), either TIV or LAIV is an acceptable vaccination
option.
Timing of LAIV administration
Vaccination with LAIV can begin as soon as vaccine is available
for that season. Children less than 9 years old receiving LAIV for
the first time should begin in October or earlier, because they
Access the complete reference list online at http://www.expertconsult.com
63. Maassab HF. Adaptation and growth characteristics of influenza virus at 25
degrees C. Nature 1967;213:612–4.
71. Jin H, Lu B, Zhou H, et al. Multiple amino acid residues confer temperature
sensitivity to human influenza virus vaccine strains (FluMist) derived from
cold-adapted A/Ann Arbor/6/60. Virology 2003;306:18–24.
73. Chen Z, Aspelund A, Kemble G, et al. Genetic mapping of the cold-adapted
phenotype of B/Ann Arbor/1/66, the master donor virus for live attenuated
influenza vaccines (FluMist). Virology 2006;345:416–23.
85c. Block SL, Yi T, Sheldon E, et al. A randomized double-blind noninferiority
study of quadrivalent live attenuated influenza vaccine in adults. Vaccine
2011;29:9391–97.
85d. Block SL, Falloon J, Hirschfield JA, et al. The immunogenicity and
safety of a quadrivalent live attenuated influenza vaccine in children.
Ped Inf Dis J. 2012; March 29. E pub ahead of print. DOI:10.1097/
INF.0b013e31825687b0.
87. Belshe RB, Edwards KM, Vesikari T, et al. Live attenuated versus
inactivated influenza vaccine in infants and young children. N Engl J Med
2007;356:685–96.
Influenza vaccine—live 18 311
will need a second dose a minimum of 4 weeks after the initial
dose. A recent study determined that different two-dose regi-
mens of TIV alone, LAIV alone, LAIV plus TIV, or TIV plus
LAIV administered to children aged 6 to 35 months at an inter-
val of 1 month were well tolerated and induced similar levels of
HAI antibody.233 This study was conducted with a small number
of children, but the findings suggest that flexibility in immuni-
zation regimens may be an effective approach to vaccination of
children against influenza. The safety and immunogenicity of
concurrent administration of LAIV with measles-mumps-rubella
II and Varivax vaccines have been evaluated in children 12 to
15 months of age;234 they were safe and well tolerated, and the
immune responses to the relevant viral antigens were similar
whether the vaccines were given concurrently or separately.
In the absence of specific data indicating lack of interference
with other vaccines, it is prudent to follow the ACIP General
Recommendations on Immunization. Inactivated vaccines do not
interfere with the immune response to other inactivated vaccines
or to live vaccines. An inactivated vaccine can be administered
either simultaneously or at any time before or after LAIV. A live
vaccine not administered on the same day should be adminis-
tered after a 4-week (or longer) interval when possible.
Administration of LAIV and influenza antiviral use
It is not known whether administering influenza antiviral med-
ications affects the safety or efficacy of LAIV; LAIV should not
be administered until 48 hours after cessation of influenza anti-
viral therapy, and influenza antiviral medications should not be
administered for 2 weeks after receipt of LAIV.
Conclusions
LAIV has the potential to significantly contribute to the con-
trol of influenza and influenza-associated illnesses. LAIV has
significant advantages in convenience of administration. The
high efficacy of LAIV compared with TIV against matched
strains and drifted strains are compelling reasons to use the
vaccine in children. Effectiveness of LAIV in adults has also
been demonstrated.
Acknowledgment
Contributors to this chapter in the prior edition were Robert
Belshe, Robert Walker, Jeffrey Stoddard, George Kemble, Hunein
Maassab, and Paul Mendelman. This research was supported
in part by the Intramural Research Program of the National
Institute of Allergy and Infectious Diseases, National Institutes
of Health.
132. Monto AS, Ohmit SE, Petrie JG, et al. Comparative efficacy of inactivated
and live attenuated influenza vaccines. N Engl J Med 2009;361:1260–7.
133. Ohmit SE, Victor JC, Rotthoff JR, et al. Prevention of antigenically drifted
influenza by inactivated and live attenuated vaccines. N Engl J Med
2006;355:2513–22.
136. Block SL, Toback SL, Yi T, et al. Efficacy of a single dose of live attenuated
influenza vaccine in previously unvaccinated children: a post hoc analysis of
three studies of children aged 2 to 6 years. Clin Ther 2009;31:2140–7.
148. Ambrose CS, Luke C, Coelingh K. Current status of live attenuated
influenza vaccine in the United States for seasonal and pandemic influenza.
Influenza Other Respi Viruses 2008;2:193–202.
155. Rhorer J, Ambrose CS, Dickinson S, et al. Efficacy of live attenuated
influenza vaccine in children: a meta-analysis of nine randomized clinical
trials. Vaccine 2009;27:1101–10.
206. Murphy BR, Coelingh K. Principles underlying the development and use
of live attenuated cold-adapted influenza A and B virus vaccines. Viral
Immunol 2002;15:295–323.
 SECTION TWO: Licensed vaccines

Japanese encephalitis vaccines

19 Scott B. Halstead
Julie Jacobson
Katrin Dubischar-Kastner

Japanese encephalitis (JE), a mosquito-borne flaviviral infec-

tion, is the leading recognized cause of childhood encephali-
tis in Asia. After, yellow fever (YF), JE is the second Flavivirus
that is vaccine-preventable. Despite this, thousands of cases
and deaths are reported annually. However, in many loca-
tions, the disease is not under systematic surveillance, and
official reports undoubtedly underestimate the true number
of cases ( Figure 19-1).1–4 Japanese encephalitis is transmit-
ted throughout Asia, in a region supporting 3.4 billion peo-
ple—50% of the world's population. Consequently, regional
JE-associated morbidity probably exceeds worldwide mor-
bidity from herpes encephalitis. 3,5–7 In many Asian coun-
tries, such as Japan, Korea, Taiwan, Singapore, Sri Lanka,
and Nepal, the integration of JE vaccine into routine immu-
nization programs has led to the near elimination of JE. 8
The major disease burden is now in developing countries
in Asia. With the near-eradication of poliomyelitis, JE is
the continent's leading cause of childhood viral neurologic
infection and disability.8 By any standard, JE is a major pub-
lic health problem that is controllable by proven effective
vaccines.
History of disease
Summer-fall encephalitis outbreaks consistent with JE were
recorded in Japan as early as 1871, the largest of which, in
1924, led to more than 6,000 cases, 60% of them fatal.9 A
filterable agent from human brain tissue was isolated in rab-
bits that year, and in 1934, Hayashi transmitted the disease
experimentally to monkeys.10 Soon after, the isolation of JE
and related St. Louis encephalitis (SLE) viruses made pos-
sible serologic confirmation of encephalitis cases occurring
elsewhere in the region, including a cluster of cases occur-
ring in 1934 through 1935 in Beijing.11 The virus initially
was called Japanese B encephalitis (the modifying “B” has
since fallen into disuse) to distinguish the disease from von
Economo type A encephalitis, which had different clini-
cal and epidemiologic characteristics. The mosquito-borne
mode of JE transmission was evidenced with the isolation
of JE virus (JEV) from mosquitoes
in 1938. Subsequent field studies established the role of
aquatic birds and pigs in the viral enzootic cycle. Viruses
isolated from human cases in Japan in 1935 and in Beijing
in 1949 provided the prototype Nakayama and Beijing and
P3 strains, respectively. These strains were most widely used
in vaccine production for many years.
Why the disease is important
Regional distribution
During the first half of the 20th century, JE was recognized
principally in temperate areas of Asia in the form of peren-
nial outbreaks in Japan, Korea, and China.2 Annual outbreaks
of several thousand cases recurred in Japan until as recently
as 1966, with a public health impact magnified further by
the concentration of outbreaks during the summer season.
In Korea, after 5,616 cases and 2,729 deaths were recorded in
1949, epidemics continued every 2 or 3 years, culminating in
an unprecedented 6,897 cases in 1958 (Figure 19-2).12 China
accounted for the majority of cases in the region; between 1965
and 1975, more than 1 million cases were reported, 175,000
in 1971 alone (Figure 19-3).13 Chinese public health efforts
that placed great emphasis on vaccination produced a dra-
matic decline in cases. For many years, vaccination coverage
remained low in rural provinces and incidence in the rural pop-
ulation remained stable. From 2000 to 2005, China has con-
tinued to report annual cases from 5,000 to 9,000. In 2005, in
an effort to improve coverage, China integrated JE into the rou-
tine Expanded Programme on Immunization (EPI) system in
all endemic provinces. In Japan, Korea, and Taiwan, the intro-
duction of routine immunization programs after 1965 led to
the near-elimination of the disease. Still, enzootic transmis-
sion of the virus continues in these locations (Figure 19-4),
and periodic outbreaks occur in unimmunized populations, as
in Korea in 1982 (Figure 19-2).14 In Japan, viral surveillance
using sentinel pigs and horses continues to document yearly
high levels of transmission of JEV, signifying continued risk of
human infection and disease should vaccination programs be
discontinued15–17 (I. Kurane, personal communication, 2010).
Today, JE vaccination has gained wide support from public
health authorities as there is a growing recognition of the scope
of the problem.
Although sporadic viral encephalitis cases were noted in north-
ern Thailand from early in the 20th century, JE was not rec-
ognized as a major public health problem in Southeast Asia
until 1969, when an epidemic of 685 cases was reported from
the Chiang Mai Valley.18 Yearly outbreaks producing thou-
sands of cases and hundreds of deaths followed in the north-
ern region, and JE became recognized as a leading cause of
314 SECTION TWO Licensed vaccines

A, Reported cases of Japanese encephalitis (JE) for Asian countries and for China only, 1970 to 2009. (Data from World Health Organization
and ministry of health reports.) B, Incidence of JE per 100,000 population by province, China between 1996 and 2005. (Wang H, Li Y, Liang X, et al. Japanese
encephalitis in mainland China. Jpn J Infect Dis 62:331-336, 2009, with permission.) Endemicity is as follows: high, >1/100,000; moderate, 0.5-1.0/100,000; low,
0.1-0.5/100,000.
330 SECTION TWO
Protective Efficacy of SA 14-14-2 Attenuated Japanese Encephalitis Vaccine, China
CI, confidence interval.
*Children 1 to 10 years old immunized with single primary dose.
†
Combination of 1- to 10-year-old children immunized in previous year(s), 1-year-old children given primary dose and 2-year-old children given booster dose.
‡
Children 1 to 7 years old immunized with single primary dose only.
Data from Hueyang Antiepidemic Station,290 Chendu Biologics Institute,352 and Wang et al.353
vaccine with intermittent use of an inactivated Vero cell–based
vaccine. Three provinces in northern and far western China
with higher altitudes are considered nonendemic and are not
using JE vaccines.292 Attenuation of the SA 14-14-2 virus was
produced empirically by serial passage in nonneural tissues,
and the underlying molecular basis of its neuroattenuation
is partially known. The nucleotide sequence of the neuro-
virulent parent SA 14 virus differs from that of SA 14-14-2
and two other attenuated SA 14-2–derived vaccine viruses in
seven amino acid substitutions found in all three attenuated
strains. Four were in the envelope protein (E138, E176, E315,
and E439), one was in NS protein 2B (NS2B63), one was in
NS3 (NS3105), and one was in NS4B (NS4B106).291,293–295
These mutations have been shown to be very stable. An
amino acid change at E138 in SA 14-14-2 virus was shown
to be sufficient for mouse neuroattenuation when introduced
into a wild-type JE complementary DNA infectious clone.
Nucleotide sequences of structural protein genes revealed
that the attenuated and parental viruses differed by eight
amino acid mutations in the E protein. Attenuated JEVs also
have been obtained by selecting neutralization escape vari-
ants. Attenuation was associated with single base changes
resulting in single E protein amino acid changes and was
linked with altered early virus-cell interactions but not with
replication.296,297
The reduced neurovirulence of the SA 14-14-2 strain was
confirmed in 3-week-old mice and monkeys (Table 19-5).268
Compared with the parent SA 14 strain, which killed weanling
mice by subcutaneous or intracerebral inoculation with median
lethal doses (LD50) of 105.5 and 108.3 LD50/mL, respectively, the
SA 14-14-2 virus produced no mortality and only minor clini-
cal signs in a few intracerebrally inoculated mice. Combined
intrathalamic and intraspinal inoculation of rhesus monkeys
produced no clinical illness and only minor inflammatory reac-
tions in the substantia nigra and cervical spinal cord. Mice were
more sensitive than monkeys to intracerebral infection, with
some animals showing mild neuronal lesions in the cerebral
cortex, hippocampus, or basal ganglia.298 Compared with his-
topathologic lesions produced by the parent SA 14 virus, the
inflammatory reaction and neuronal necrosis to SA 14-14-2
virus were significantly less. In 5-week-old mice inoculated
intracerebrally with the virus pair, ultrastructural studies
showed that the parent virus produced cytopathologic changes
in the majority of neurons, particularly in the rough ER and
Golgi apparatus of the neuronal secretory system, while it could
not be confirmed that the vaccine strain replicated at all and
neurons appeared normal.287
Further evidence of the strain's reduced neurotropism comes
from experimental studies in athymic nude mice. No deaths or
histopathologic abnormalities were observed after intraperito-
neal or subcutaneous inoculation of a viral dose greater than
107 median tissue culture infective doses (TCID50), and virus
could not be recovered from brain tissue.145 Although cyclo-
phosphamide increases susceptibility of mice (and also of
monkeys) to virulent JEV, immunosuppression with cyclophos-
phamide did not lead to encephalitis in mice inoculated periph-
erally with the SA 14-14-2 virus.146,299 The strain also did not
kill intracerebrally inoculated weanling hamsters. Phenotypic
characteristics of the vaccine strain (PHK8), such as small
plaque, reduced mouse neurovirulence, and genetic proper-
ties, were stable through at least 10 additional PHK cell culture
passages.288,291
SA 14-14-2 viremia has not been studied in human vaccin-
ees. SA 14-14-2 propagated in PDK cells was able to be iso-
lated from blood obtained from vaccinees, but at infectious
titers below the usual oral infection threshold of mosquitoes.
Inferences from several studies indicate a negligible potential
for mosquito transmission of attenuated JEVs from a vaccinated
pig or human; however, more definitive experiments on trans-
mission potential of the vaccine in JE vector mosquitoes are
needed.300 The SA 14-2-8 strain, closely related to SA 14-14-2,
propagated well in mosquitoes after intra-
thoracic inoculation but could not be transmitted. In oral
infection trials, only 11% of mosquitoes
ingesting vaccine became infected compared with 100% with
wild-type JE.300 The virus did not revert to a neurovirulent phe-
notype after mosquito passage.300 Intrathoracic inoculation of
the SA 14-14-2 strain into vector mosquitoes replicated to titers
similar to parental SA 14, but the authors did not test transmis-
sibility of vaccine virus.301

Several hundred ampules of seed virus, prepared from the
sixth passage level of SA 14-14-2 virus, are maintained at
the NICPBP, Beijing. Lyophilized seed virus (PHK5) is pro-
vided to the production institute, where it is passaged once
for the production seed (PHK6). The PHK cells are obtained
from 10- to 12-day-old golden Syrian hamsters maintained
in colonies at the Chengdu and Wuhan Production institutes.
Current Chinese pharmacopeia and WHO guidelines specify
the use of specific pathogen-free hamsters for the primary cell
cultures. A specific pathogen-free hamster colony has been
established at the Chengdu facility and is used for all vaccine
production.302 Monolayers are inoculated with diluted virus,
and cells are fed with minimal essential medium contain-
ing human albumin, gentamicin, and kanamycin. Infected
cell culture fluid with an infectious titer of approximately
107.2 plaque-forming units (PFU)/mL is harvested at 78 to 96
hours and coarsely filtered, and the resulting liquid vaccine
is lyophilized. Gelatin (1%) and sucrose (5%) are added as
stabilizers. Lyophilized vaccine is reconstituted and diluted
with sterile water for injection.303 The reconstituted vaccine
should be a transparent orange-red liquid. The PFU of vac-
cine is not to be lower than 105.7 PFU/mL in the reconstituted
vaccine.
Vaccine must meet standards according to the WHO guide-
lines and Chinese pharmacopeia for absence of neuroviru-
lence in adult mice, stability for reversion to neurovirulence
after intracerebral passage in suckling mice, and freedom from
adventitious agents, including retroviruses. A highly sensitive
product-enhanced reverse transcriptase (PERT) assay could
not detect any reverse transcriptase activity in finished or
bulk vaccine. A cocultivation assay on production cell culture
using human cell lines followed by PERT assay that allowed an
expanded opportunity of detecting cross-species infection of ret-
roviruses into human cells did not show any reverse transcrip-
tase activity in PHK cells. The finished vaccine must have an
infectious titer exceeding 105.7 PFU/mL.
Vaccine is produced by three manufacturers in China: Chengdu
Institute of Biological Products (CDIBP), Lanzhou Institute
of Biological Products, and Wuhan Institute of Biological
Products (Table 19-4). The Chinese pharmacopeia has been
revised to correspond with the WHO guidelines for the pro-
duction of live attenuated JE vaccines.155,304 In 1998, 200,000
doses of vaccine from CDIBP were donated to Nepal. In 2001,
CDIBP became the first and only manufacturer to officially
export vaccine outside of China when it successfully licensed
and began exporting vaccine to Nepal and to South Korea in
2002. Since that time, CDIBP has additionally licensed vac-
cine in Sri Lanka in 2002, India in 2006, Thailand in 2007,
Laos in 2008, and the Democratic People's Republic of Korea
and Cambodia in 2009. As of 2006, CDIBP is the only facil-
ity to fully comply with all WHO production standards for live
attenuated vaccine.
In one study, concurrent administration of killed JE vac-
cine with bacille Calmette-Guérin (BCG), measles, or DTP
vaccine at 6 to 10 months of age was not associated with
increased adverse events and led to higher JE ELISA antibod-
ies in the first two groups and to no change in the last.305
In the Philippines, live attenuated JE vaccine administered
together or sequentially with measles vaccines to 8, 9, and
Japanese encephalitis vaccines 19 331
10-month-old infants demonstrated minimal reactogenic-
ity and excellent rates of seroconversion to all vaccines.306 In
China, there were no differences in reactogenicity of vaccines
in children receiving measles, JE, or JE and measles vaccines
combined.307
In China, the vaccine is licensed for a 0.5-mL dose to be admin-
istered subcutaneously to children at 8 months of age and a sec-
ond opportunity again at 2 years. In some areas, a booster dose
is given at 7 years. Measles vaccine has been given concurrently.
Like the inactivated PHK cell–derived vaccine, SA 14-14-2 vac-
cine was distributed in annual spring campaigns rather than
according to an age-based schedule. In China, this vaccine is
now integrated into the routine immunization schedule in
most endemic provinces.
The infectious titer of lyophilized vaccine is not appreciably
changed after storage at 37°C for 7 to 10 days, at room tem-
perature for 4 months, or at 2°C to 8°C for at least 1.5 years.
After reconstitution with sterile saline or distilled water and
storage at 23°C, the vaccine's infectious titer is stable for 2 to 4
hours or 2 hours, respectively.303 The vaccine should be stored
and shipped at 8°C, protected from sunlight, and used within
18 months after the titration test is completed. Phenotypic and
genotypic attributes of the SA 14-14-2 vaccine have remained
stable over several decades.268
Immunogenicity of JE vaccines
The JE viral strains isolated in different geographic locales
or at different times exhibit minor antigenic differences
and biologic characteristics such as growth in cell culture
and neuroinvasiveness in experimentally infected mice.
However, there is no evidence of corresponding differences
in human pathogenicity and no evidence that immune
responses to one strain, whether as a live viral infection
or as a vaccine, does not protect against disease caused by
another. 107,308–311 In the modern era, antibodies raised by
administration of JE vaccines have been measured by IgM
and IgG ELISA, hemagglutination inhibition, complement-
fixation, and tissue culture–based neutralization tests of
varying formats. A neutralizing antibody titer of more than
1:10 generally is accepted as evidence of protection and
postvaccination seroconversion.312 Passively immunized
mice that acquire this level of neutralizing antibody are
protected against challenge from 10 5 LD 50 of JEV, a typical
dose transmitted by an infectious mosquito bite. Indirect
observations from human trials have associated efficacy
with this criterion. 175 Plaque reduction neutralization tests
are used most frequently, and procedural differences, such
as choice of challenge virus strain, cell systems, addition of
exogenous complement, and choice of end points (ranging
from 50% to 90% plaque reduction in serum dilution tests),
affect test sensitivity. Some laboratories still use log10 neu-
tralization indices (LNIs) in tests using a single serum dilu-
tion. However, despite procedural differences, neutralizing
antibody titers in three laboratories (the US CDC, Japan's
National Institutes of Health, and the Yale Arbovirus
Research Unit) were shown to be highly correlated (R.
DeFraites, unpublished data, 1998). This same result was
not replicated when several laboratories evaluated standard
serum samples for a test serum bank being developed by the
WHO. 313 An international standard of protective antibody
units is in the process of being established.312
332 SECTION TWO Licensed vaccines
Inactivated mouse brain vaccine
Among Asian children immunized with two doses of
Nakayama strain– or Beijing-1 strain–derived vaccines, neu-
tralizing antibody responses to the respective homologous
vaccine strains are in the range of 94% to 99%; responses to
strains representing a heterologous antigenic group are lower
(results of selected studies are shown in Table 19-7).314–318
The proportion of vaccinees retaining detectable neutralizing
antibodies and their geometric mean antibody titers (GMTs)
declined rapidly in the year after the primary two-dose series,
so that only 78% to 89% of Nakayama vaccine recipients and
88% to 100% of Beijing-1 vaccine recipients still had detect-
able levels before the scheduled 1-year booster. Antibody
persistence was greater among Beijing-1 vaccine recipients.
After booster immunization (or third vaccine dose), antibody
response rates were uniformly high (100%). During the his-
tory of the administration of JE vaccines in Asia, additional
booster doses have been given empirically, usually with the
goal of restoring neutralizing antibody titers to the detectable
range and according to the philosophy that more is better, as
in Figure 19-19. There are no studies of the efficacy of such
booster doses. In fact, the long-term efficacy of two or three
doses of killed vaccine has not been measured. There is evi-
dence of sustained protection of adults who do not receive
booster doses of JE vaccines in Korea, Japan, and Taiwan. This
evidence is compatible with solid protection being provided to
immunized persons who circulate very low titers of neutral-
izing antibodies.
Immunogenicity studies in Asian subjects should be inter-
preted in the light of the immunologic background of vaccin-
ees. Although some studies have been done in nonendemic
areas and in subjects without JE viral antibodies, in other
studies, undetected exposures to JE, dengue, and other flavi-
viruses prevalent in Asia may have resulted in an augmented
antibody response after immunization and apparently better
immune responses. Where the influence of previous flavivi-
ral infections was unlikely, vaccinees receiving two doses had
Homologous and Heterologous Neutralizing Antibody Responses in Children Immunized With Inactivated Nakayama- or Beijing-Strain
Japanese Encephalitis Vaccines
GMT, geometric mean antibody titer.
*Data from Konishi E, Kurane I, Mason PW, et al. Induction of Japanese encephalitis virus–specific cytotoxic T lymphocytes in humans by poxvirus-based JE vaccine
candidates. Vaccine 16:842-849, 1998.
†
.001.
‡
Data from Fu DW, Zhand PF. Establishment and characterization of Japanese B encephalitis virus persistent infection in the Sf9 cell line. Biologicals 24:225-233,
1996.
§
Data from Chambers TJ, Tsai TF, Pervikov Y, Monath TP. Vaccine development against dengue and Japanese encephalitis: report of a World Health Organization
meeting. Vaccine 15:1494-1502, 1997.
¶
.03.
lower seroconversion rates and lower GMTs than vaccinees
receiving three doses (Figure 19-20A).28,192,319–323 Such differ-
ences may be important to scientific advisory groups who must
establish immunization doses and schedules for nonendemic
populations. After administration of two doses of inactivated
vaccine, neutralizing antibody titers declined to less than 1:8
in 90% of vaccinees in as little as 6 to 12 months following
vaccination (Figure 19-20B).192 A three-dose primary sched-
ule was more immunogenic, resulting in seroconversion rates
exceeding 90% and significantly higher neutralizing antibody
titers.192,218,319–323 Administration of vaccines at long (days 0, 7,
and 30) vs short intervals (days 0, 7, and 14) using three doses
resulted in uniform seroconversions in all subjects but signifi-
cantly higher neutralizing antibody titers in vaccinees immu-
nized over the longer period (30 days). Vaccine prepared from
the Beijing-1 strain seems to be more immunogenic, despite its
smaller delivered volume, yielding higher seroconversion rates
and higher antibody titers to heterologous Nakayama virus
(Table 19-7).316–318,324–326 Similar but more marked differences
were seen in comparative neutralization of field viral strains
from Taiwan, paralleling those in experimentally immunized
mice (Figure 19-19).327 The clinical importance of these differ-
ences in strain reactivity in the neutralization test is uncertain.
Results of the efficacy trial comparing a monovalent Nakayama
strain vaccine with a bivalent vaccine also containing Beijing-1
antigen showed that the two were equally efficacious.328 Because
JE vaccines manufactured in Thailand, India, Vietnam, and
Taiwan all use the Nakayama strain, this strain seems to be
protective as field observations do not suggest a pattern of vac-
cine failure. Neutralizing activity present below the threshold
of detection in in vitro assays may be protective. B-cell mem-
ory, anamnestic antibody responses, and T-cell memory may
contribute to protect vaccinees whose serum samples measure
as seronegative, but nevertheless provide immunologic help to
clear infections on reexposure.
Although previous exposures to dengue and certain other fla-
viviruses probably enhance the immune response to JE vaccine,
antibody responses did not differ in persons with a history of YF
vaccination, unlike the accelerated response to inactivated TBE
vaccine seen among YF-vaccinated persons.
 Japanese encephalitis vaccines 19 333

B
A, Antibody response to inactivated mouse brain–derived Japanese encephalitis vaccine in a trial among US citizens. Only 77% of
vaccinees who received two doses seroconverted, compared with 99% of vaccinees who received three doses. Geometric mean antibody titers
also were higher in the latter group (28 vs 141). B, At 6 to 12 months after primary immunization, only 10% of vaccinees who were given two doses
retained protective levels of neutralizing antibody. Booster immunization led to a greater than 90% response. (Adapted from Poland JD, Cropp CB, Craven
RB, et al. Evaluation of the potency and safety of inactivated Japanese encephalitis vaccine in US inhabitants. J Infect Dis 161:878-882, 1990.)

adequate antibody response (see preceding discussion).192,319–323

Persons given inactivated vaccine are exposed only to viral
structural proteins, and, in contrast with convalescent patients,
they do not produce radioprecipitating antibodies to viral NS
proteins.329 Their memory T-cell proliferative responses to a
viral-like particle containing only structural proteins also dif-
fer from those of recovered patients, whose CD4 and CD8 cell
responses also include viral NS proteins.139 The implications of
these immune response differences are uncertain. While cellu-
lar responses have been studied following infection with wild-
type JE or immunization with JE vaccines, these have not been
correlated with protection or disease outcome.
Neutralizing antibodies measured in a mouse or tissue cul-
ture neutralization test have been generally accepted as a cor-
relate or surrogate for protection.330 Immunogenicity studies
in subjects from areas without endemic transmission (west-
ern countries and areas of India) indicate that three doses of
inactivated mouse brain–derived vaccines are necessary for an
The US CDC ACIP recommends three doses, on days 0, 7,
and 30 (Table 19-8).205 An abbreviated schedule in which doses
are administered on days 0, 7, and 14 also results in uniform
seroconversion; however, neutralizing antibody titers are sig-
nificantly lower. Although approximately 80% of vaccinees
respond after two doses, this schedule is not recommended.192
Recommendations for booster doses are based on limited data.
Neutralizing antibody titers were maintained for 3 years in 37
of 39 US Army vaccinees given the Biken vaccine; however, field
studies indicate greater variability in antibody persistence.319
Until further data become available, a booster dose is recom-
mended 2 years after the primary series and thereafter as deter-
mined by serologic monitoring.205 In endemic populations, two
doses are given within 1 to 4 weeks, with an additional dose
given at 1 year. Boosting schedules vary nationally.205,302
A field study in Thailand showed that seroconversion rates
were higher after immunization with lyophilized vaccine than
after immunization with liquid vaccine. A moderate loss of
potency was demonstrated after liquid vaccine was exposed to
simulated field conditions.331
338 SECTION TWO
vaccinees receiving an Indian-manufactured JE vaccine
found that 34 (97%) of 35 retained neutralizing antibodies
3 years after a primary series of three doses and 31 (91%) of
34 retained antibodies at 4.5 years, with GMTs of 71 and
32, respectively. However, the boosting effect of naturally
acquired flaviviral infections in these subjects cannot be ruled
out.363 During the history of the administration of JE vaccines
in Asia, additional booster doses have been given empirically,
usually with the goal of restoring neutralizing antibody titers
to the detectable range and according to the philosophy that
more is better as in Figure 19-19.
Flavivirus-naïve US Army soldiers who received a three-dose
primary immunization series retained protective neutralizing
antibody titers for at least 1 year (GMT, 76). Antibody titers at
12 months were unchanged from those observed 3 months after
immunization (GMT, 78). A booster dose given at 12 months
was followed by a significant anamnestic response (GMT,
1,117). In a limited number of subjects studied 3 years after the
primary series, 16 (94%) of 17 who had neither traveled to Asia
nor received a booster retained neutralizing antibody titers of
greater than 1:10 and their GMTs at 6 months and at 3 years
after primary immunization were unchanged.319 A large study
of 293 laboratory workers revealed that more than 50% lost
neutralizing antibody titers to below 1:10 just over 2 years after
immunization. The authors also noted significant lot-to-lot
variation in antibody response to vaccination.364 Although these
observations suggest that the first booster immunization is
needed no sooner than 2 to 3 years after primary immunization,
the interval for subsequent boosters has not been established.
Studies cited suggest that antibodies quickly wane no mat-
ter how many doses of inactivated vaccine are given. It seems
likely that recall B-cell memory responses to infected mosquito
bites are capable of raising protective levels of antibodies before
JEV invades the CNS. However, data from humans and animal
models on this point are missing.
Inactivated Vero cell vaccine
The persistence of antibodies following immunization with
inactivated Vero cell–derived SA 14-14-2 vaccine has been
investigated in several studies that have shown slightly vary-
ing results, with between 58% and 83% of subjects still having
neutralizing antibody titers of 1:10 or more at 12 to 15 months
after the primary immunization (Table 19-11).272,274,336 Reasons
for the differing results are not completely understood, but
the studies were carried out in different geographic areas, and
a high prevalence of prior TBE vaccination in the trial show-
ing higher antibody persistence might have contributed to this
difference.274
Booster doses of inactivated Vero cell–derived SA 14-14-2
vaccine are needed to sustain long-lasting titers of neutraliz-
ing antibodies at or above a dilution of 1:10. Based on data on
the persistence of antibodies and immunogenicity of booster
doses, the ACIP recommends that a booster dose should be
given at 12 months after the primary series. In the main booster
trial in 198 subjects, inactivated Vero cell–derived SA 14-14-2
vaccine was administered as a booster 15 months after pri-
mary immunization. The booster dose was highly immuno-
genic and yielded a GMT of 900 (a 40-fold increase relative to
prebooster titers), with 100% of subjects achieving a PRNT50
titer of 1:10 or more by 28 days after the booster dose. This
response lasted for at least 1 year with a GMT of 361 at 1 year
after the booster and 98.5% of subjects still having a PRNT50 of
1:10 or more (Figure 19-21B).274 Modeling of antibody decline
 Japanese encephalitis vaccines 19 339

Persistence of Neutralizing Antibodies Against Inactivated Vero Cell–Derived SA 14-14-2 Vaccine Following Two-Dose Primary
Immunization Schedule

GMT, geometric mean titer; SCR, seroconversion rate and SPR, seroprotection rate (both defined as the rate of subjects with PRNT50 1:10; PRNT, plaque reduction
neutralization test).

following the booster dose suggests that the majority of vaccine

recipients will continue to have protective antibody levels for
at least 4 years.274 In another study, booster doses were given
11 or 23 months after primary immunization to vaccine recipi-
ents whose titers had dropped to less than 1:10. All subjects
with complete primary immunization seroconverted following
the booster dose, regardless of the time point for booster dose
administration.274 No data are available on immunogenicity of
booster doses administered longer than 23 months after pri-
mary immunization.
Live attenuated vaccine
Persons given a single dose of SA 14-14-2 vaccine in Nepal in
1999 were followed up through the 2004 transmission sea-
son. A case-control study was performed to measure vaccine
efficacy. Of 35 children resident in the Bardiya and Banke dis-
tricts admitted to the Bheri Zonal Hospital with serologically
confirmed JE, only one had been vaccinated in 1999. Among
430 age and sex-matched village control subjects, 234 (54.4%)
had been vaccinated. The protective efficacy of JE vaccine 12
to 15 months after administration was 98.5% (CI, 90.1%-
99.2%).356 A 5-year follow-up found the single-dose efficacy was
maintained at 96.2%.357
As suggested by a long-term experience of up to 11 years of
sustained reduction in JE cases in areas where many children
received only a single dose of SA 14-14-2 vaccine, efficacy of
this and possibly other JE vaccines greatly exceeds the ability to
detect a circulating neutralizing antibody response to a single
administered dose.
Safety (adverse events)
Inactivated mouse brain–derived JE vaccine
Localized reaction
Local tenderness, redness, or swelling at the injection site
occurs in approximately 20% of persons immunized with inac-
tivated mouse brain–derived vaccines. Mild systemic symp-
toms, chiefly headache, low-grade fever, myalgias, malaise, and
gastrointestinal symptoms, are reported by 10% to 30% of vac-
cinees (Table 19-12).192,316,322,323 ,365–367
Systemic hypersensitivity reactions
Vaccine-related allergic adverse events not reported previously
from Asia were recognized after 1989 in Australia and sev-
eral European and North American countries as the vaccine
became used widely for travelers.192,196,365–374 Hypersensitivity
reactions have consisted principally of generalized urticaria,
angioedema, or both, which in a few patients were potentially
life threatening. These reactions generally have responded to
oral antihistamines or corticosteroids, but recalcitrant cases
have required hospitalization and parenteral steroid therapy.
A temporally related death was reported in a man with mul-
tiple hypersensitivities who also had received plague vac-
cine.196 Numerous lots and different manufacturers have been
implicated.370 In retrospect, allergic side effects, including urti-
caria, angioedema, and moderate dyspnea, were also observed

Reported Side Effects of Inactivated Mouse Brain–Derived Japanese Encephalitis Vaccine

*Local tenderness, redness, swelling, itching and numbness.

†
Chiefly fever, headache, malaise, rash; also chills, dizziness, myalgia, nausea, vomiting, abdominal pain, diarrhea, sore throat, blurred vision, increased salivation and
taste, difficulty concentrating, and emotional instability.
340 SECTION TWO Licensed vaccines
in recipients of the crude mouse brain vaccine adminis-
tered on Okinawa in 1945.247 The US Vaccine Adverse Event
Reporting System reported that between 1999 and 2009, there
were 300 adverse events reported, 106 (35.3%) were classified
as hypersensitivity reactions (8.4 per 100,000 doses) and 4 as
neurologic events (0.3 per 100,000 doses). Of the events, 23
were described as serious (1.8 per 100,000 doses). No cases
of encephalitis, meningitis, or Guillain-Barré syndrome were
reported.367
An important feature of the reactions is the potential for
delayed onset, particularly after a second dose. In a prospec-
tive study of 14,249 US Marines, the median interval between
immunization and onset was 18 to 24 hours after the first
dose, with 74% of reactions occurring within 48 hours.192,196,369
Among reactors to a second dose, there was a greater delay, with
a median interval of 96 hours and a range of 20 to 336 hours.
Reactions have developed after a second or third dose when
previous doses were given uneventfully. A nested case-control
study found an elevated risk with history of various allergic
disorders (eg, urticaria, OR, 11.4 [95% CI, 2.4-62.1]; allergic
rhinitis, OR, 9.2 [95% CI, 2.8-23.1]; asthma, rhinitis, or both,
OR, 6.5 [95% CI, 2.1-20.8); and any allergy, OR, 5.7 [95% CI,
1.8-18.1]).196 Another small study also implicated alcohol con-
sumption and receipt of another vaccine 1 to 9 days previously,
as opposed to simultaneously, as risk factors.375
Reported rates have varied according to the approach to
ascertainment (Table 19-13). Prospective and retrospective
studies have found risk of an allergic adverse event, usually
defined as objective urticaria or angioedema, in the range of 18
to 64 per 10,000 vaccinees.368,373–376 A large postmarketing study
in the United States and Japan determined the adverse event
rates to be 2.8 and 15.0 per 100,000 doses in Japan and the
United States, respectively.377 Hypersensitivity rates were 0.8
and 6.3 per 100,000 doses, respectively. A cluster of two deaths
resulting from anaphylactic shock in children receiving JE vac-
cine was reported in Korea in 1994. In a follow-up study to mea-
sure the incidence of JE vaccine–related adverse events, one case
of anaphylactic shock with syncope and collapse, three cases of
generalized urticaria, and three cases of severe erythema were
found in 15,487 Korean children immunized between May
15 and June 30, 1995. The rate of 0.03% was lower than that
observed in adult travelers, which could reflect biologic differ-
ences in reactivity or the sensitivity of surveillance (Y.M. Sohn,
unpublished data, 1996).
Although the pathogenesis of the hypersensitivity reactions
is not proven, in three Japanese children experiencing systemic
reactions, IgE antibodies to gelatin were demonstrated,
suggesting that gelatin, which is added as a vaccine stabilizer,
may be a provoking antigen.378 Further analysis of reactions
showed two patterns: One was a combination of urticarial rash
and wheezing, which was associated with the presence of anti–
gelatin IgE in the serum, and the second was a cardiovascular
Hypersensitivity Reactions* After Immunization With Inactivated Mouse Brain–Derived Japanese Encephalitis Virus Vaccine
CI, confidence interval.
*Generalized urticaria or angioedema, excluding pruritus only.
†
Rates varied by vaccine lot.
collapse syndrome apparently caused by another mechanism.378
A similar syndrome has been described in recipients of dip-
loid cell–derived rabies vaccine in whom symptoms developed
after a delay of as long as 1 week after booster immunization.379
Immunologic studies demonstrated IgE antibodies to human
albumin, which is added to the vaccine as a stabilizer and
chemically altered by the inactivating agent -propiolactone. 380
Allergic reactions in recipients of crude mouse brain vaccine
in Okinawa were attributed to formalin-altered proteins. In a
Danish case-control study, about one third of allergic reactions
could be attributed to an allergic predisposition in the vaccin-
ees. The main risk factors were young age, female sex, previous
allergic skin reactions or hay fever, skin reactivity to nickel, and
hyperresponsiveness to mosquito bites.381
From the beginning, the vaccine's neural tissue substrate raised
concern about the possibility of postvaccination neurologic
side effects.382 The manufacturing process purifies the infected
mouse brain suspension extensively, and MBP content is con-
trolled below 2 ng/mL, well below the dose considered to have
an encephalitogenic effect in a guinea pig test system. However,
measurements of other acute disseminated encephalomyeli-
tis (ADEM)-associated neural proteins (eg, proteolipid protein,
myelin-oligodendrocyte glycoprotein) have not been reported.
Experimental immunization of guinea pigs and cynomolgus
monkeys with adjuvant and 50 times the normal dose of vac-
cine did not result in clinical or histopathologic evidence of
encephalomyelitis.383,384
In 1945, in one of the first mass uses of mouse brain–derived
JE vaccine, 53,000 American soldiers on Okinawa were immu-
nized with a crude inactivated mouse brain suspension after a
JE outbreak occurred on the island.247 Acute vaccine-associated
side effects, including the occurrence of acute neurologic events,
were monitored. Eight neurologic reactions, principally polyneu-
ritis, were observed. However, similar cases were reported con-
currently in nonvaccinated soldiers, and it is unclear whether
the illnesses were vaccine-related. One case of Guillain-Barré
syndrome, temporally related to JE immunization, was reported
among approximately 20,000 American soldiers immunized
with the vaccine before US licensure.
An early prospective study in Japan to detect vaccine-associ-
ated adverse events found no neurologic complications occur-
ring within 1 month after vaccination in 38,384 subjects
receiving crude or purified vaccine.382 A country-wide study to
detect neurologic complications found 26 temporally related
cases (meningitis, seizures, demyelinating disease, polyneuri-
tis) between 1957 and 1966, but rates and comparisons with
nonimmunized control subjects were not available. Passive
surveillance of vaccine-related adverse events in Japan is con-
ducted through sentinel hospitals, clinics, pharmacies, and
manufacturers. Surveillance data on JE vaccine adverse events

come principally from the manufacturers (Biken and others).
Few neurologic complications temporally related to JE vaccina-
tion have been reported, but denominators of vaccinees were
not available in all years and the sensitivity of this passive sur-
veillance system is unknown (Table 19-14).259,383–385
In 1992, two anecdotally reported cases of temporally related,
vaccine-associated ADEM in Japan prompted a survey of 162
Japanese medical institutions to solicit additional cases.386 Five
more cases spanning 22 years were reported, including two with
elevated CSF MBP levels.387 Neither the numerator of cases nor
the denominator of vaccinees was defined rigorously, but the
authors estimated that ADEM occurred in fewer than 1 in 1
million vaccinees. An unrelated report described two deaths due
to anaphylactic shock, and four ADEM cases (two fatal) tempo-
rally related to vaccination were reported in Korea in 1994; one,
also fatal, was reported in 1996. An additional fatal case of acute
encephalopathy occurred in a 15-year-old girl who received her
ninth dose of JE vaccine and her third dose of hantaviral vac-
cine (also made in mouse brain) at 4 and 2 weeks, respectively,
before onset of stupor and seizures (Y.M. Sohn, unpublished
data, 2002).
An additional report of vaccination-associated ADEM cases
in Danish travelers, unprompted by previous reports from
Japan and Korea, suggests that the issue of neurologic compli-
cations should be reinvestigated.388 After a vaccinee developed
ADEM in 1995, a review of the national database disclosed two
similar temporally related cases in 1983 and 1989, all in adults.
Because JE vaccine distribution in Denmark is controlled, the
denominator of vaccinees and a rate for adverse events could be
estimated. The rate of temporally related ADEM, 1 in 50,000
to 75,000 vaccinees, is far above previous estimates of all neu-
rologic complications and in the same range as JE incidence in
countries where the disease is endemic. In the United States, a
postmarketing study revealed that no serious neurologic adverse
events were temporally associated with JE vaccine between
January 1993 and June 1999, while in Japan, 17 vaccine-related
neurologic disorders were reported from April 1996 to October
1998.377 Finally, in 2005, the government of Japan withdrew its
recommendations for two and three-dose immunizations using
inactivated mouse brain–derived JE vaccines.242 Biken ceased
manufacture of this vaccine. Remaining supplies on the world
market are dwindling.
There are few data on the safety and efficacy of inactivated JE
vaccines in immunocompromised persons. A small study of
children with various chronic diseases, including some oncol-
ogy patients, disclosed no difference in immunogenicity or
reactogenicity in recipients of mouse brain–derived vaccine.316
Reported Neurologic Manifestations Temporally Associated
With Japanese Encephalitis Vaccination,* Japan
*Inactivated mouse brain–derived vaccine.
†
1971 to 1973 data from Tokyo only: 2 cases per 883,373 vaccinees.
Data from Japanese encephalitis vaccine lyophilized, Oya, 239 Egashira
et al, 383 and Okinaka et al. 385
Japanese encephalitis vaccines 19 341
Infants vertically infected with HIV responded less well to the
vaccine (see earlier discussion), but no unusual adverse events
were recorded.332
No adverse outcomes of pregnancy have been associated directly
with JE vaccine. Currently, the US ACIP notes that the safety
of JE vaccine in pregnancy is unknown, and, since vaccination
poses a hypothetical risk to the fetus, routine vaccination dur-
ing pregnancy is not recommended. However, the ACIP also
advises that pregnant women traveling to a JE-endemic area
should be vaccinated if the risk of infection prevails over the
theoretical risks of immunization.
Spread to contacts is not a risk because the vaccine is inactivated.
Inactivated Vero cell vaccine
The adverse event profile of inactivated Vero cell–derived SA
14-14-2 vaccine has been thoroughly characterized in a series
of licensure-relevant clinical trials, comparing the local and sys-
temic safety profiles with alum-containing PBS and with inacti-
vated mouse brain–derived JE vaccine.334,389
Localized reactions
The localized reactions seen after immunization with inac-
tivated Vero cell–derived SA 14-14-2 vaccine are mainly
consistent with local reactions to other intramuscularly
administered, alum-adjuvanted vaccines. Throughout sev-
eral trials, the major local reactions have been injection site
pain and tenderness, which have been reported in 20% to 30%
of recipients of vaccine and placebo (Table 19-15). Local red-
ness, swelling, itching, and hardening were less commonly
observed. Most local reactions after immunization with inac-
tivated Vero cell–derived SA 14-14-2 vaccine are transient
and mild. The local reactogenicity profile seemed to be more
favorable compared with mouse brain–derived JE vaccine in
terms frequency of severe reactions, with any severe symptom
reported in 9 (2.1%) of 421 subjects vs 59 (13.8%) of 427 sub-
jects ( .0001).334
Common systemic reactions
Headache and myalgia are the most common systemic reac-
tions, reported in placebo and vaccinated subjects in clini-
cal trials of inactivated Vero cell–derived SA 14-14-2 vaccine
(Table 19-16). Other mostly mild systemic reactions observed
in vaccinees were fever, rash, nausea, vomiting, diarrhea, and
excessive fatigue. All reactions were seen at approximately
the same frequency in recipients of control vaccine (PBS +
alum).
Systemic hypersensitivity reactions
The serious systemic hypersensitivity reactions observed after
administration of inactivated mouse brain–derived JE vaccines
have not been observed with inactivated Vero cell–derived SA
14-14-2 vaccine in clinical trials. Adverse events that generally
might be associated with systemic hypersensitivity reactions
were analyzed in a meta-analysis of clinical trials includ-
ing 3,558 subjects receiving inactivated Vero cell–derived SA
14-14-2 vaccine; 657 subjects, a control vaccine (PBS + alum);
and 435 subjects, a mouse brain–derived inactivated JE vac-
cine.391 Potentially hypersensitivity-associated adverse events
(observed in inactivated Vero cell–derived SA 14-14-2 vac-
cine and control groups) are shown in Table 19-17.391 These
adverse events were reported at similar frequencies in recipi-
ents of inactivated Vero cell–derived SA 14-14-2 vaccine (3.5%)
or a PBS + alum control vaccine (3.7%) but the incidence was
342 SECTION TWO

Injection Site Solicited Adverse Reactions* After Inactivated Vero Cell–Derived SA 14-14-2 Vaccine or Control Vaccine† in Subjects in the
Safety Population With Evaluable Diary Cards

JE, Japanese encephalitis.

*Injection site reactions were assessed for 7 days after each dose.
†
The control vaccine was phosphate-buffered saline and aluminum hydroxide.
‡
Denominators used to calculate percentages are based on the number of evaluable diary card entries (defined as documented presence on any day [ie, entry of
“yes”] or absence on all days [ie, entry of “no”]) for each individual symptom and observation period.
§
Number of subjects who returned diary cards after each dose.
Sources: US Food and Drug Administration–approved prescribing information for IXIARO 390 and Tauber et al.389

Common Systemic Adverse Events* After Inactivated Vero Cell–Derived SA 14-14-2 Vaccine or Control Vaccine† in the Safety Population

JE, Japanese encephalitis.

*The adverse events are those observed at an incidence of 1% in the IXIARO or control group.
†
The control vaccine was phosphate-buffered saline and aluminum hydroxide.
‡
Number of subjects in the safety population (subjects treated with at least one dose) who received the respective dose.
§
These symptoms were solicited in a subject diary card. Percentages include unsolicited events that occurred after the 7-day period covered by the diary card.
Sources: US Food and Drug Administration–approved prescribing information for IXIARO 390 and Tauber et al. 389
 Japanese encephalitis vaccines 19 343

Potential Hypersensitivity- or Allergy-Associated Adverse Events*391

CI, confidence interval; JE, Japanese encephalitis; PBS, phosphate-buffered saline.

*Multiple occurrences of the same adverse event in one subject were counted only once.

significantly higher in the group receiving the mouse brain– other drugs or agents, and 3 subjects with preexisting allergies
derived JE vaccine (5.5%; for difference between inactivated experienced worsening of allergy symptoms or allergy flares. All
Vero cell–derived SA 14-14-2 vaccine and mouse brain–derived reactions were of mild to moderate severity; vaccination was
inactivated JE vaccine, =.044). Rash was the only potentially stopped after one dose in one case (worsening of allergy symp-
hypersensitivity-related adverse event reported in subjects, toms on the day of vaccination).391
with comparable incidence in all 3 vaccine groups (1.7%, inac-
tivated Vero cell–derived SA 14-14-2 vaccine; 2.5%, mouse
brain–derived inactivated JE vaccine; and 2.0%, PBS + alum).
To date, inactivated Vero cell–derived SA 14-14-2 vaccine has
Urticaria was reported in 4 of 3,558 subjects receiving inac-
been administered to approximately 5,800 subjects in phase 3
tivated Vero cell–derived SA 14-14-2 vaccine (2 generalized and
clinical trials, including approximately 1,400 children. In post-
2 localized reactions of mild to moderate severity) and 1 of 657
marketing use to date, the relatively high rate of hypersensi-
subjects receiving PBS + alum vaccination. All cases occurred
tivity reactions seen with mouse brain–derived JE vaccine has
between 6 and 28 days after administration of the vaccine. All
not been observed for inactivated Vero cell–derived SA 14-14-2
recovered successfully, 3 subjects without treatment, 1 with
vaccine.
antihistamines plus a corticosteroid, and 1 with antihistamines
During the first year after licensure, 25 adverse drug reac-
alone (both in the inactivated Vero cell–derived SA 14-14-2 vac-
tions (ADRs) were reported, translating into a reporting rate of
cine group). Only the reaction following the PBS + alum vacci-
10.1 per 100,000 doses distributed. Four of these ADRs (16%)
nation was judged as possibly related to the vaccination by the
were serious (reporting rate for serious ADRs, 1.6 per 100.000
investigator.
doses distributed). The serious ADRs were neuritis, meningism
Of the four inactivated Vero cell–derived SA 14-14-2 vac-
(a case of headache and neck pain, self-limited, initially reported
cine cases, in one case the urticaria was reported after the first
as mild encephalitis but later downgraded by the local regula-
vaccination and was localized to the abdomen.391 The subject
tory authority to be a nonserious case), pharyngeal spasm, and
had no reaction following the second dose. In another case,
iritis (initially reported as flu-like illness associated with eye
urticaria was reported 28 days after the second injection had
pain). No cases of anaphylaxis or angioedema were reported in
been given in the left arm and was localized to the right arm.
the first 12 months of postmarketing use of inactivated Vero
One subject reported generalized urticaria after the first vac-
cell–derived SA 14-14-2 vaccine (E. Schuller, personal commu-
cination and received antihistamines plus a corticosteroid but
nication, 2011).
had no reaction to the second dose. The subject was using lan-
soprazole concomitantly. The last subject reported generalized
urticaria after the second vaccination and complained of other
symptoms (fever, headache, nausea, vomiting, flu-like symp- Safety of inactivated Vero cell–derived SA 14-14-2 vac-
toms). The investigator noted that the urticaria was proba- cine has not been studied in immunocompromised persons.
bly cold-induced. Angioedema was not reported in any of the Postmarketing data through the first year of marketing did not
cases.391 contain reports on adverse events in immunocompromised per-
Eight subjects experienced adverse events reported as hyper- sons. Nevertheless, Chinese authorities recommend that SA
sensitivity, 6 (0.2%) in the inactivated Vero cell–derived SA 14-14-2 vaccine not be given to immunosuppressed persons.
14-14-2 vaccine group and 2 (0.5%) in the mouse brain–derived Naturally, concerns that the vaccine could cause JE in immu-
inactivated JE vaccine group. In the inactivated Vero cell– nocompromised persons are not applicable to an inactivated
derived SA 14-14-2 vaccine group, 3 events were reactions to vaccine.
344 SECTION TWO Licensed vaccines
Inactivated Vero cell–derived SA 14-14-2 vaccine is pregnancy
category B (United States). There are limited data from the use
of the vaccine in pregnant or breastfeeding women. In animal
studies, findings of unclear clinical relevance have been iden-
tified. In a reproductive and prenatal/postnatal toxicity study,
no vaccine-related effects were observed on reproduction, fetal
weight, survival, and development of the offspring. However,
incomplete ossification of parts of the skeleton was observed
in the group receiving two doses, but not in the group receiving
three doses. As a precautionary measure, use of the vaccine dur-
ing pregnancy or lactation should be avoided as per European
labeling.392
After reviewing data from 28 women who became preg-
nant during clinical trials with inactivated Vero cell–derived SA
14-14-2 vaccine, the European regulatory authority concluded
that there was currently no evidence for a causal relationship of
abnormal pregnancy outcomes with the vaccine (Intercell, on
file). Cases in which women are inadvertently vaccinated dur-
ing pregnancy should be reported to Intercell AG for follow-up
of the pregnancy and infant health status. Further data will be
generated as part of a vaccine safety surveillance program.
Spread to contacts is not a risk because the vaccine is inactivated.
Inactivated PHK cell–derived JE vaccine
Few adverse reactions have been reported in connection with
the P3 inactivated vaccine. Local reactions, including swelling
at the injection site, are observed in about 4% of vaccinees, and
mild systemic symptoms, such as headache and dizziness, are
reported by fewer than 1% of vaccinees. A temperature higher
than 38°C previously was a complication in 12% of vaccinees,
but, after a reduction in bovine serum in the currently formu-
lated vaccine, febrile reactions have been halved. An urticar-
ial allergic reaction was observed in only 1 of nearly 15,000
vaccinees surveyed.328 Recent clusters of reactions temporally
related to vaccination and consisting of acute asthenia, syn-
cope, and disorientation have been reported from disparate
areas of China. Some features of the reactions suggest that they
may be outbreaks of hysteria, but their consistency and occur-
rence in a widespread geographic distribution are difficult to
explain.
Live attenuated JE vaccine
An estimated 350 million Indian, Sri Lankan, Nepalese, and
Chinese adults and children have been immunized with the
live attenuated vaccine without apparent complication. Clinical
monitoring of experimentally immunized subjects has docu-
mented the absence of local or systemic symptoms after immu-
nization; specifically, headache and symptoms that might be
associated with neuroinvasive infection and fever and signs and
symptoms of systemic infection have not been observed after
immunization. In a study of 867 children in whom fever was
monitored during a 21-day period after immunization, temper-
atures above 37.6°C were recorded in fewer than 0.5% of vac-
cinees, and fever-onset days were distributed throughout the
observation interval, mitigating against a vaccine-related febrile
illness after a specific incubation period. In the same study,
symptoms were recorded from 588,512 other vaccinees; fever
was reported in 0.046% of subjects, rash in 0.01%, dizziness in
0.0003%, and nausea in 0.0003%, but these rates are difficult
to interpret in the absence of similar observations in control
subjects.284,393
A block-randomized cohort study of 13,266 vaccinated
and 12,951 nonvaccinated 1- to 2-year-old children followed
up prospectively for 30 days confirmed the vaccine's safety.
No cases of encephalitis or meningitis were detected in either
group, and rates of hospitalization, new onset of seizures, fever
lasting more than 3 days, and allergic, respiratory, and gas-
trointestinal symptoms were similar in the two groups. The
observations excluded a vaccination-related encephalitis risk
above 1 in 3,400.394
The rates of clinical encephalitis among populations vacci-
nated in field trials (Table 19-6) provide additional reassurance
that the SA 14-14-2 virus does not itself cause encephalitis at a
detectable rate. Rates of clinical encephalitis in children receiv-
ing SA 14-14-2 vaccine—1.16 to 6.75 per 100,000 population—
are lower than reported population-based incidence rates of
childhood encephalitis due to causes other than JE (56-639 per
100,000 population).
No rare adverse events have been reported.
Although experimental data suggest that JE SA 14-14-2 virus
may not be neurotropic in immunosuppressed animals, there
are no data on the vaccine's safety in immunocompromised per-
sons, specifically HIV-infected patients.
No observations on the vaccine's safety in pregnant women
have been reported. Live attenuated JE vaccine has a theoreti-
cal risk in pregnant women. Usage advisories are that pregnant
women should not be vaccinated. When JE vaccine must be
given to pregnant women or to immunocompromised persons,
available inactivated JE vaccine should be used rather than live
vaccine.
No spread to contacts has been reported. Transmissibility of
vaccine virus by vector mosquitoes is not well studied.
Indications for vaccine (who and why)
Endemic areas
In rural areas of Asia, intense JEV transmission in the enzo-
otic cycle leads to a high risk of exposure at an early age.
The risk of infection increases sharply with the waning of
the protective effects of maternal immunity and as increasing
outdoor activity places young children at risk of bites from
infected mosquitoes. In hyperenzootic settings, there may be
almost universal seroconversion by the age of 15 years, indi-
cating that by that age everyone has been exposed and had
an asymptomatic, mild illness, or clinical disease. Universal
primary immunization is indicated for children younger than
15 years, which ideally would be integrated into the routine
immunization schedule. For the most part, stepwise imple-
mentation of national JE vaccination programs, initially in
epidemic foci and in areas with hyperenzootic transmis-
sion, has been necessary because of economic considerations
(Figure 19-24).28,395 The current recommendation for an effec-
tive immunization strategy is to mount a one time catch-up
campaign to cover all at-risk age groups followed by the intro-
duction of routine immunization through the EPI. Experience
at the country level has shown that although most JE immu-
nization programs were initiated in limited areas, as sur-
veillance systems improved, immunization programs have

Mother with a child with Japanese encephalitis (JE)
in Gorakhpur, India. During the 2006 outbreak, two or three patients
occupied one bed owing to lack of space, a common situation during
JE outbreaks.
expanded to the entire country.161,396–398 Recommendations for
the scale and scope of immunization programs may change as
less expensive vaccines and simpler immunization schedules
become available. It is important to note that immunization
schedules and boosting may be different in endemic than in
nonendemic settings.
Although incidence may vary from year to year and within
countries, universal childhood immunization is desirable
because, even in economically advanced countries, viral
transmission cannot be eliminated and the cumulative risk
of acquiring the illness over a lifetime of exposure justifies
universal protection. This ongoing risk has been recently
well established in a cross-sectional study in Japan as recent
as 2001 that demonstrated annual infection rates of 0.2%
to 3.4% in populations in eight different prefectures across
Japan by measuring NS1 antibodies. These antibody levels
were 4.4% overall with no difference in males and females
and represented all age groups in the sample from 6 months to
69 years of age (Figure 19-21)16,399 (I. Kurane, personal commu-
nication, 2006). In South Korea, where an immunization pro-
gram successfully controlled JE, a fivefold to sixfold increase
in JE cases was observed in 2010.173 Furthermore, conditions
leading to epidemic transmission are unpredictable, and,
at intervals, outbreaks may lead to large numbers of cases,
even in urban areas. Despite this, Hong Kong and Singapore
have reported few JE cases despite the absence of mandatory
JE immunization. This may be the result of extensive urban-
ization and lack of vector and reservoir host habitat making
it likely that viral transmission is limited in these predomi-
nantly urban environments.174 Hong Kong has reported a few
cases, while on Singapore island, there is evidence of limited
Japanese encephalitis vaccines 19 345
JE transmission on the outskirts of Singapore city and sur-
rounding islands.400,401 The transmission and epidemiology of
JE continue to evolve. In Japan, there has been a shift from
normal to unusual species of mosquitoes dominantly involved
in transmission of JE and a decreased link with swine as the
amplifying host for transmission of viruses to humans.401
Unfortunately, there are virtually no data on the efficacy of
boosting immunity to killed JE vaccines; however, because of
waning neutralizing antibodies, it is prudent to give one or
more booster vaccinations. Waning immunity in adults and
elderly people in countries with a long history of immunizing
with inactivated vaccine such as Japan and South Korea has
brought up the question of the need for additional boosters
later in life.
Expatriates
With safe and effective new JE vaccines available, vaccination
is recommended for all expatriates whose principal residence
is in an area where JE is enzootic.180,203,205,367,402 Risk of acquir-
ing JE among expatriates is variable and depends principally
on the specific location of intended residence, housing condi-
tions, nature of activities, and the possibility of unanticipated
exposure to high-risk areas (see the next section, “Travelers”).
Risk varies regionally and within specific countries. Viral
transmission is seasonal in most areas and can fluctuate from
year to year in a given location. Figure 19-1, Figure 19-13, and
Table 19-18 summarize and extrapolate available data on loca-
tions and seasonality of risk by country and region. Patterns
of viral transmission may change, and physicians and travel-
ers are cautioned to consult public health officials for current
data and trends. The risk may be difficult to assess because
JE is poorly reported, accurate data may not be available, and
local immunization prevents disease, giving a false sense of
security.
Travelers
Administration of a licensed JE vaccine is recommended for
selected travelers to Asia and is not yet recommended as a
routine immunization. In the United States, the recommen-
dation is for persons 1 year or older spending 1 month or more
in enzootic settings during the season especially if travel-
ing, living, or working in rural areas.203,205,365 Currently, it is
recommended that persons 17 years old or older receive the
licensed inactivated Vero cell vaccine (IXIARO). Younger per-
sons should be immunized using remaining stocks of mouse
brain vaccine. Although the risk of acquiring JE during travel
is extremely low, estimated at less than 1 per million travel-
ers to Asia, the availability of a new safe and effective vaccine
and an accelerated immunization schedule changes the cost-
effectiveness calculations. The vast majority of visitors to Asia
on business or in tours are at low risk and need not be immu-
nized. Because JE viral transmission is confined to certain sea-
sons and occurs principally in rural areas, visitors with such
a travel itinerary have a higher risk of acquiring the disease.
For many Asian countries, the use of vaccines in endemic
areas masks viral activity, making it difficult to determine
risk accurately. Travelers and their physicians should weigh
individual risk factors and disease risk in the area and season
of anticipated travel (Figures 19-1 and 19-13 and Tables 19-3
and 19-18).202,203,205 For travelers, the recommended primary
vaccination series for JE-MB is three doses administered sub-
cutaneously on days 0, 7, and 30 (Table 19-8). An abbrevi-
ated schedule (days 0, 7, and 14) can be used when the longer
schedule is impractical.
Advanced age and pregnancy may affect risk and outcome
of JE. Repellents and other protective measures are recom-
mended regardless, because other vector-borne diseases may be
346 SECTION TWO

Risk of Japanese Encephalitis by Country, Region, and Season*


Risk of Japanese Encephalitis by Country, Region, and Season—Cont'd
EPI, Expanded Programme on Immunization.
*Assessments based on published reports, reports to the World Health Organization, and personal correspondence. Because virus is transmitted in a zoonotic cycle,
reports of human cases in countries with vigorous vaccination programs are not a reliable indicator of risk. For an update, refer to the US Centers for Disease Control
and Prevention. Travelers' Health (http://wwwnc.cdc.gov/travel).
transmitted in the same areas. General precautions are espe-
cially important to travelers in whom vaccine is contraindi-
cated, who are unable to complete immunization because of
departures on short notice, or who do not choose to be immu-
nized because their visits to high-risk areas are brief or carry
an equivocal risk. An expert advisory group has recommended
vaccinating all repeat travelers or other persons with prolonged
duration of stay and any traveler with a rural itinerary. This
group also recommends vaccinations for persons with chronic
conditions such as solid organ transplant or conditions lead-
ing to CNS leakage, hypertension, diabetes mellitus, or chronic
renal disease, and persons receiving anti-TNF therapy.180 The
recommendations for vaccination may change as data on cost
and risk profiles become available for vaccines licensed for use
in travelers.
Research laboratory workers
There have been 22 cases of laboratory-acquired JEV, princi-
pally in research settings where infectious JEV was used.403
Infection can be transmitted by percutaneous or mucous
membrane exposures and potentially by aerosols, especially
from preparations containing high viral concentrations,
which occur during viral purification. Immunization pre-
sumably protects against percutaneous exposures; however,
it is unknown whether vaccine-derived immunity, especially
from inactivated vaccine, protects against aerosol infection.
Immunization is advised for all research laboratory person-
nel who potentially may be exposed to field or virulent strains
of the virus. Although no formal biosafety recommendations
have been issued for work with the attenuated vaccine SA
14-14-2 strain, sufficient data are available on its attenuation
such that immunized workers should be permitted to handle
that virus under biosafety level 2 conditions, paralleling rec-
ommendations for the attenuated vaccine strains of YF, Junin,
Rift Valley fever, chikungunya, and Venezuelan equine enceph-
alitis viruses.370
Japanese encephalitis vaccines 19 347
Contraindications and precautions to
immunization
Mouse brain–derived JE vaccine is contraindicated in people
who have had an allergic reaction to the vaccine, to gelatin,
or to other rodent-derived products, including previous doses
of JE vaccine. A WHO Advisory Committee voted against
recommending for human use rabies vaccines derived from
neural tissue, and this may have implications for inactivated
mouse brain–derived vaccine.404 As previously noted, hypersen-
sitivity reactions to mouse brain–derived JE vaccine are more
common in persons with allergic conditions (eg, asthma; aller-
gic rhinitis; drug or Hymenoptera venom sensitivity; and food
allergy, especially to gelatin-containing foods). If these persons
are offered JE vaccine, they should be advised of the poten-
tial for vaccine-related angioedema and generalized urticaria.
Hypersensitivity to a protein found in mouse urine is com-
mon in animal caretakers and certain laboratory workers. It
is unknown whether this sensitivity carries a specific risk in
recipients of JE vaccine. Anecdotal reports of ADEM occurring
in temporal relationship to vaccination suggest that the mouse
brain–derived vaccine should not be used in persons who have
recovered from ADEM or Guillain-Barré syndrome or who have
multiple sclerosis or other demyelinating disorders. As a result
of perceived safety concerns, manufacturers in many countries
have announced plans to replace mouse brain–derived vaccines
with tissue culture–based inactivated vaccines or live attenu-
ated vaccine.366 There are no specific contraindications to the
use of PHK-derived inactivated JE vaccine except a history of an
allergic reaction to a previous dose. Contraindications for inac-
tivated Vero cell–derived vaccine are hypersensitivity to one of
the components, eg, protamine sulfate, or allergic reaction to a
previous dose of the vaccine. Live JE vaccines pose a theoreti-
cal risk to the developing fetus. No adverse outcomes of preg-
nancy have been associated directly with JE vaccine. Travelers
348 SECTION TWO Licensed vaccines

and their physicians must balance the theoretical risks of JE

vaccine in pregnancy against the potential risks of acquiring JE
and the adverse outcome of the disease. At present, live attenu-
ated SA 14-14-2 vaccine is not recommended for administra-
tion to pregnant women.304
A theoretical risk, explored by passive transfer of anti-
body in animal models, is the antibody-mediated enhance-
ment of JE infection and disease.233 A similar phenomenon
has occurred in humans whose antibody levels have waned
below protective thresholds after having been immunized
with inactivated respiratory syncytial virus or measles
viruses. A similar phenomenon has been demonstrated in
experimental animals using inactivated vs live attenuated
vaccines for JE or Murray Valley encephalitis.405,406 Antibody-
dependent enhancement is also postulated as a pathogenic
mechanism in severe dengue infections. Fundamental dif-
ferences in cellular pathogenesis may preclude this phe-
nomenon from occurring in JE where the brain and not
In Thailand, Japanese encephalitis (JE) cases failed
macrophages or pulmonary epithelial cells are major targets to respond to control measures directed against vector mosquitoes;
of viral infection.407 attack rates declined only after introduction of JE vaccine.

Public health considerations

Epidemiologic effects of vaccination
Japanese encephalitis has a devastating effect on health
resources and the affected communities. Epidemics frequently
result in pressure from the community to take action to control
the disease and provide a safe living environment. Vaccination
for JE has been the single most effective and reliable means
to control JE in endemic settings. Although a secular trend
toward declining JE incidence has been observed with wide-
spread use of JE vaccine, coincident socioeconomic changes
also may have contributed to falling disease incidence. It was
this same change in socioeconomic status that allowed devel-
oped Asian countries to afford vaccine and introduce it. In gen-
eral, the higher the socioeconomic status of each country, the
earlier that universal JE vaccine coverage was introduced. Less
affluent countries are catching up. In Nepal, successful clini-
cal trials of SA 14-14-2 vaccine led to the early introduction
of this vaccine and a sharp drop in JE morbidity and mortality
(J. Tandan, personal communication). Owing to interventions
managed by PATH, a not-for-profit organization in Seattle,
Washington, with funding from the Bill and Melinda Gates
Foundation, live attenuated JE vaccine is now routinely admin-
istered in Sri Lanka and large quantities have been imported to
combat epidemics in India.355–357 In Thailand, a vertical control
program based on vector control was implemented from 1970,
which demonstrated no impact on disease. The program was
then integrated with an increased focus on outbreak response,
again without impact on disease. In 1986, vaccine began to be
integrated into the routine immunization system in the high-
est risk areas of the country, which had a dramatic impact on
disease as immunization was expanded to additional provinces
(Figure 19-25).19,408,409
Observations from China, where development has been less
extensive, also reinforce the impact of immunization. The JE
incidence rates in Beijing and other areas of China where high
immunization rates are maintained have declined dramatically
and remained low (Figure 19-26 and Table 19-18).1,13,239 Until
recently, vaccine coverage, while high in prosperous regions in
China, remained low in many high-risk rural locations. The
principal barrier to immunization has been the cost of vac-
cine to the family because JE vaccine was not subsidized by
the government as a routine childhood vaccine. However, in
2005, JE vaccine was expanded into the EPI system in endemic
provinces.
Declining incidence of Japanese encephalitis (JE)
in Beijing and association with mass immunization, 1950 to 1985.
(Adapted from Gu PW, Ding ZF. Inactivated Japanese encephalitis (JE) vaccine
made from hamster cell culture (a review). JE HFRS Bull 2:15-26, 1987.)
Disease control strategies
As a zoonotic disease with natural viral reservoirs, JE never can
be eradicated. Although its transmission can be modulated by
factors mentioned earlier, these approaches alone or in combi-
nation cannot be relied on to reduce disease incidence. The only
effective strategy is vaccination. Successful control of JE by uni-
versal immunization in many Asian countries provides ample
proof that expansion of vaccine coverage throughout the conti-
nent will lead to the near-elimination of the disease.410
Cost-benefit information
The inactivated mouse brain–derived vaccine is troubled by
safety and other issues. Moreover, the vaccine's 91% efficacy,
when extrapolated to the entire cohort of children younger than
15 years in Asia—approximately 1 billion children—yields an
absolute number of primary vaccine failures of questionable
acceptability. Assuming a JE incidence rate of 1 per 10,000
population in children younger than 15 years, approximately
100,000 cases would occur in the absence of any immuniza-
tion. If every child was immunized but only 91% were pro-
tected, 9,000 cases might be expected annually as a result of
primary vaccine failure. Although additional booster doses pre-
sumably would improve efficacy, the strategy also would lead to
increased costs for a vaccine that already is considered costly
and of marginal benefit for the cost.395–398

In Thailand, it was estimated that incorporating the inac-
tivated vaccine into routine immunization at 18 months
(at a cost of $2.28 per person [monetary values, US dol-
lars throughout this section]) would prevent 124 cases (per
100,000 people), with a cost-effectiveness of $15,715 and
would save $72,922 (in treatment costs, disability care, and
loss of future earnings) for each prevented JE case. Vaccination
for JE was thought to be worth implementing unless the inci-
dence was below 3 per 100,000 population.395 In Shanghai,
China, a cost-effectiveness analysis estimated that immuni-
zation with inactivated P3 vaccine would prevent 420 JE cases
and 105 deaths, saving 6,456 disability-adjusted life-years
per 100,000 people. The live attenuated SA 14-14-2 vaccine
would prevent a similar number of cases and deaths. Both
vaccines resulted in cost savings compared with no vaccina-
tion, but the live vaccine would result in a greater cost sav-
ing ($512,456 per 100,000 people vs $348,246) because it
requires fewer doses.411
A partnership was developed between the main SA 14-14-2
vaccine producer in Chengdu, China, and PATH to increase the
supply of live vaccine with pricing well below a dollar per dose
for the poorest countries of Asia (GNP $1,000). This partner-
ship plans to make a WHO prequalified SA 14-14-2 vaccine
available throughout the region. In 2005, the Chinese pharma-
copeia was updated to match the WHO productions guidelines
for live attenuated JE vaccines. With the emphasis on quality
and the increased data available on the SA 14-14-2, the vac-
cine has been licensed in South Korea (since 2001), Nepal
(1999), India (2006), and Sri Lanka (2007) ). With the availabil-
ity of this vaccine at lower prices, cost-effectiveness has been
reevaluated and should result in increased access to JE vaccines
throughout Asia.411–413
Eradication or elimination, if feasible
Eradication of JEV from nature is not feasible. Elimination of
clinical JE in humans is possible through universal vaccine cov-
erage of the at-risk population. Disease control relies on indi-
vidual protection because there is no herd immunity with JE as
there is no human-to-human transmission.
New and future vaccines
A number of JE vaccines are in development, some have
been licensed, and others are approaching licensure. Vero-
based inactivated JE vaccines have been licensed by Biken and
Kaketsuken for the Japanese domestic market. Sanofi Pasteur
has licensed a live attenuated YF 17D/JE chimeric vaccine in
Australia and Thailand and is actively pursuing licensure in
other countries. As the postlicensure experiences with these
vaccines are somewhat limited, they are described in sum-
mary form.
Inactivated vaccines
Several groups have produced experimental inactivated whole-
virion vaccines from infected Vero cell cultures.302,414–416 Cell
culture medium with high viral infectious titers, harvested con-
tinuously from microcarrier cultures, inactivated with forma-
lin, and further concentrated has yielded candidate vaccines
meeting mouse protection potency standards established for
the inactivated mouse brain vaccine. The vaccine produced by
Intercell has been licensed widely in the world. Its properties
were described earlier in the chapter.
The Research Foundation for Microbiological Disease of
Osaka University, Biken, has produced Biken BK-VJE, a Vero-
based vaccine formulated in alum and using the Beijing-1 JE
Japanese encephalitis vaccines 19 349
strain. By using a bellows system for feeding Vero cells and
batch harvesting, high yields of virus have been obtained.417
Virus has been grown in serum-free medium using glycine and
sorbital stabilizers in the inactivation process.418 A 2001 phase
1 study of a two-dose regimen showed the vaccine to be well tol-
erated, with only mild elevations of liver function test results,
and immunogenic, with JE neutralizing antibody seroconver-
sions observed in all seronegative subjects. A two-dose phase
3 trial in 110 Flavivirus-susceptible children produced neutral-
izing antibody in 100% of subjects with no adverse events. A
second phase 3 open-label trial resulted in high levels of neu-
tralizing antibodies in all children previously immunized with
mouse brain vaccine and given one dose of BK-VJE. The vaccine
was licensed in Japan in 2009.415 Administration of this vac-
cine is as follows: For persons 3 years or older, 0.5 mL is given
intramuscularly with 0.25 mL given to children younger than
3 years. Two doses are given, 1 to 4 weeks apart. A booster dose
is given approximately 1 year after the two-dose series. A fourth
dose is given to 9 to 13 year olds (I. Kurane, personal commu-
nication, 2010).
Kaketsuken, one of the Japanese producers of mouse brain–
derived JE vaccines has produced a purified, unadjuvanted, Vero
cell–derived vaccine formulated without gelatin or thimerosal.
Vero DNA and protein concentrations are several picograms
per dose in the final product. In phase 1, 30 susceptible adults
were given three doses of vaccine at 0 and 4 weeks and 6 months
with no serious adverse events observed and all subjects devel-
oping JE neutralizing antibodies. Phase 2 studies involved 200
children aged 6 to 90 months who were given three doses with
100% seroconversions. Mild adverse events were noted in as
many as 45% of subjects following administration of the first
dose. The company has obtained a license to market this vac-
cine in Japan.415 Similarly, Vero cell inactivated vaccines are now
being produced in China by Beijing Tiantan Biologial Products
Co, Ltd, and Liaoning Chengda Biotechnology Co, Ltd.
An Australian-developed, formalin-inactivated JE vaccine
produced in Vero cells and formulated with a polysaccharide
adjuvant (Advax) produced neutralizing antibody responses in
mice and horses and resistance to JE challenge as good as or
better than licensed IXIARO or Chimeravax JE.416 This vac-
cine also provided protection against Murray Valley encepha-
litis virus.
Chimeric vaccines
By far the most promising genetic approach has been the con-
struction of flavivirus chimeras in which the YF 17D genome
contributes NS genes and SA 14-14-2 contributes prM and
E genes (ChimeriVax-JE; developed by Acambis, Cambridge,
Massachusetts, and licensed for US distribution to Sanofi
Pasteur, Lyon, France).419 The chimera is constructed from
DNA, transcribed to RNA and electroporated into Vero cells
(Figure 19-27). The resultant infectious clone has the neuro-
virulence properties of SA 14-14-2 but Vero cell growth char-
acteristics of 17D.420,421 Attenuation of the chimeric virus was
shown to depend on clusters of at least three of the six amino
acid changes in the JE E protein.421 The chimera JE-CV has
proved highly immunogenic in rhesus monkeys and protects
against intracerebral and intranasal challenge using a wild
virus strain.422,423 The chimera cannot be transmitted by mos-
quito vectors of JE or YF viruses.424 Clinical development of
ChimeriVax-JE to date has included 557 healthy adult subjects
between 18 and 59 years old who participated in seven clinical
trials. A total of 404 subjects received a single subcutaneous
vaccination with ChimeriVax-JE, 55 subjects received two sub-
cutaneous vaccinations separated by 30 days, and 98 subjects
received two subcutaneous vaccinations separated by 6 months.
The first clinical study, a phase 1 proof-of-principle trial, was
conducted in 36 healthy adult subjects, half of whom had
350 SECTION TWO
The ChimeriVax concept. ChimeriVax-JE results from the insertion of the prM and E structural genes from the SA 14-14-2 virus into
the nonstructural genes of yellow fever vaccine 17D (YF 17D). cDNA, complementary DNA. (Data from Acambis, Inc.)
previous immunity to YF.425 Assessments were made of viremia,
adverse events, laboratory values, and neutralizing antibodies
to JE and YF. Mild adverse events following vaccination with
ChimeriVax-JE were similar to those seen following vaccination
with YF 17D vaccine and consisted mainly of reactions at the
injection site, headache, fatigue, and fever. A low-level, transient
viremia was present in the majority of subjects (10/12 naïve and
11/12 YF-immune ChimeriVax-JE recipients) and was similar
in magnitude and duration to the viremia induced by YF 17D.
In 100% of ChimeriVax-JE subjects, including all YF-immune
subjects, neutralizing antibodies developed against JE (serocon-
version) by 30 days following vaccination; YF-immune subjects
responded with higher JE neutralizing antibody titers than did
naïve subjects.426
A double-blind phase 2 trial involving 99 adults doc-
umented short-duration viremia in most subjects given
between 102 and 105 PFU. Viremia titers were inverse to
inoculum dose. Irrespective of dose, nearly 100% serocon-
version in Flavivirus-naïve volunteers was observed to the
parental virus and a somewhat lower rate to three wild-type
JE strains. Although neutralizing antibody titers waned to
low levels at 6 months, immunologic memory could be elic-
ited uniformly by challenge with inactivated vaccine.427 In
humans and monkeys, prior YF immunity did not reduce the
response to the chimeric vaccine.426,428 Two pivotal phase 3
immunogenicity, safety, and tolerability trials of one dose of
JE-CV have been conducted.429,430 In one, more than 2,000
susceptible adults were studied looking for low-frequency
adverse events. None were found.429 Seroconversion after a
single JE-CV vaccination (99.1%) was statistically noninferior
to that after three doses of JE-VAX (95.1%). JE-CV elicited
a rapid immune response, with 93.6% of participants sero-
converting within 14 days. Local reaction rates at the site of
inoculation were significantly lower with JE-CV (67.6%) than
with JE-VAX (82.2%), and the reactogenicity profile of JE-CV
was comparable with that of placebo. Long-term immunity
was studied in 202 healthy adults who received JE-CV and
placebo 28 days apart in a cross-over design. A subgroup of
98 volunteers received a JE-CV booster at month 6. Nearly
all vaccine recipients achieved a seroprotective antibody titer
of 10 or more to JE-CV 28 days following a single dose of
JE-CV, and 97% were seroprotected at month 6, while 5 years
later, 65% circulated neutralizing antibodies against three or
four wild-type JE strains.430 Serologic responses of seronega-
tive 12- to 24-month-old Thai children were studied follow-
ing the administration of a single dose of JE-CV. Among 200
children, 96% raised neutralizing antibody titers to 1:10 or
greater at 28 days, but 15% of primary responders had titers
less than 1:10 at 1 year after single-dose immunization.431
Recently, Acambis has developed a manufacturing and mar-
keting agreement with Bharat Biotech, Hyderabad, India,
which will complete good manufacturing practices–compliant
fill/finish for the vaccine. Studies on the coadministration of
measles vaccine with ChimeriVax-JE are planned.
Vaccinia-vectored vaccines
Replication-deficient canary pox (ALVAC) and highly atten-
uated vaccinia viruses (NYVAC) have been used as vectors
for delivering PrM-E or PrM-E-NS1 genes.432 Several can-
didate JE vaccines are in various stages of clinical and pre-
clinical development or research.314,315 Recombinant vaccine
candidates engineered by inserting four JE viral genes—
prM, E, NS1, and NS2a—into vaccinia (NYVAC) or canary-
pox (ALVAC) viruses have been extensively evaluated. Both
recombinants expressed the encoded JE structural and NS
gene products and stimulated JE protective antibodies in
mice; two doses of the former also protected rhesus monkeys
against lethal viral challenge. In a phase 1 human trial, the
vaccines were somewhat more reactogenic than mouse brain
vaccine but were otherwise safe. Two doses of the NYVAC-JE
recombinant were nearly as immunogenic as three doses of

inactivated mouse brain JE vaccine, but only in vaccinia-naïve
volunteers. Japanese encephalitis neutralizing antibodies
were elicited in all five non–vaccinia-immune recipients but
in none of the five vaccinia-immune volunteers. Antibody
responses were observed in only 1 of 10 ALVAC-JE vaccin-
ees.433 Although the NYVAC-JE recombinant virus proved
safe and potent in vaccinia-nonimmune subjects, the appar-
ent failure to replicate in vaccinia-immune subjects places
some limits on utility.
Other genetically engineered constructs
Neuroattenuated JEV was produced by introducing defined
nucleotide substitutions into the E gene. Virions and virion sub-
units derived from complementary DNA protected mice from
death following challenge.434,435 The JE viral subunit proteins
produced in various expression systems, including
, , , and
Schneider 2 cells, have had variable success in elic-
iting mouse immunogenicity and protective potency, depend-
ing on the expressed epitopes and their conformation.435–438 The
identification of peptides mimicking the conformation of viral
epitopes may be expedited by screening pentapeptide libraries
against monoclonal antibodies with previously defined speci-
ficity.439 A vaccinia-JEV recombinant that releases extracellular
particles composed of JE, prM, and E proteins resulted in an
apparently more authentic configuration than when presented
as simple peptides.440 When given without adjuvant, extracel-
lular particles induced long-lasting antibody and memory T
cells in immunized mice. Immunogenic subviral particles con-
taining JE E protein also have been produced in other systems,
including an alphaviral recombinant virus.441,442 Novel deliv-
ery systems and adjuvants have been explored as a means to
improve immunogenicity, to direct the Th1 or Th2 response,
or to improve the convenience of immunization (eg, the inacti-
vated mouse brain vaccine has been microencapsulated in gly-
colide and lactide polymer microspheres designed to degrade at
specific intervals).443
A pseudoinfectious JEV (named RepliVAX) has been made
by deleting the capsid gene (C). The virus can be propagated in
C protein–expressing cells, permitting them to form infectious
particles structurally identical to live virus and secreted forms
of NS1, but, in normal cells, RepliVAX JE cannot produce infec-
tious particles. This vaccine was shown to protect mice and
hamsters from lethal challenge.444
Access the complete reference list online at http://www.expertconsult.com
2. Erlanger TE, Weiss S, Keiser J, et al. Past, present, and future of Japanese
encephalitis. Emerg Infect Dis 15:1–7, 2009.
3. Jmor F, Emsley HC, Fischer M, et al. The incidence of acute encephalitis
syndrome in Western industrialised and tropical countries. Virol J 5:134, 2008.
7. Solomon T. Flavivirus encephalitis. N Engl J Med 351:370–378, 2004.
53. Solomon T, Kneen R, Dung NM, et al. Poliomyelitis-like illness due to
Japanese encephalitis virus. Lancet 351:1094–1097, 1998.
74. Burke DS, Nisalak A, Ussery MA, et al. Kinetics of IgM and IgG responses
to Japanese encephalitis virus in human serum and cerebrospinal fluid. J
Infect Dis 151:1093–1099, 1985.
155. Hills S, Dabbagh A, Jacobson J, et al. Evidence and rationale for the World
Health Organization recommended standards for Japanese encephalitis
surveillance. BMC Infect Dis 9:214, 2009. doi:10.1186/1471-2334-9-214.
Japanese encephalitis vaccines 19 351
Subunit vaccines
Vaccines expressing JE envelope protein or domain III have been
presented to mice in an adenovirus construct or after expres-
sion in a baculovirus, J12#26 cells, or cells,
respectively.445–449
DNA vaccines
Another approach being explored is the intramuscular or intra-
cutaneous injection of naked DNA plasmids encoding viral
prM and E genes under the control of a cytomegalovirus imme-
diate early promoter. Using JE plasmids, immunized mice were
protected against challenge with wild-type viruses.450–453 In
swine, two intramuscular doses produced high antibody titers
and high anamnestic responses to challenge with live atten-
uated virus.454 Another vaccine in development includes the
secretory signal sequence derived from tissue plasminogen
activator fused to the full-length or partial JEV envelope pro-
tein gene. Cells transfected with the latter construct secreted
E protein and produced better protection against intracerebral
challenge in mice.455
Live attenuated viral vaccines
The potential for adventitious agents in the PHK cell sub-
strate of attenuated SA 14-14-2 vaccine requires extensive lot
release testing. As a result, this has stimulated attempts to
adapt the virus to more conventional cell systems. The strain
was adapted to PDK cells and, after passing monkey neuro-
virulence and other safety tests, was produced under good
manufacturing practices in its ninth PDK cell passage by the
WRAIR as a candidate vaccine.289 The IND vaccine, contain-
ing an infectious titer of more than 105.5 PFU/mL, was given
safely to adults and children in phase 1 human trials, but
neutralizing antibody responses to a single dose were detected
in only 2 of 4 and 14 (31%) of 45 vaccinees, respectively, with
geometric mean antibody titers (GMTs) ranging from 7 to
40 (Y.X. Yu, personal communication, 1997). In view of the
apparently low antibody responses, the PDK-passaged strain
was considered overattenuated, and development was discon-
tinued. Subsequently, the virus was adapted to Vero cells and
inactivated and is now widely licensed and marketed.266 The
YF 17D/JE chimeric vaccine described above is a live attenu-
ated single-dose vaccine.
192. Hills SL, Griggs AC, Fischer M. Japanese encephalitis in travelers from
non-endemic countries, 1973–2008. Am J Trop Med Hyg 82:930–936,
2010.
205. Centers for Disease Control and Prevention. Japanese encephalitis
vaccines: recommendations of the Advisory Committee on
Immunization Practices (ACIP). MMWR Recomm Rep 59(RR-1):1–27,
2010.
330. Hombach J, Cardosa MJ, Sabchareon A, et al. Scientific consultation on
immunological correlates of protection induced by dengue vaccines: report
from a meeting held at the World Health Organization 17–18 November
2005. Vaccine 25:4130–4139, 2007.
367. Halstead SB, Thomas SJ. New Japanese encephalitis vaccines: alternatives to
mouse brain. Exp Rev Vaccines 10:355–364, 2011.
354 SECTION TWO Licensed vaccines
16-year-old girl who reportedly had measles 8 years earlier. She
had a hemagglutination-inhibition (HI) antibody titer of 1:200
on the second day of rash and titers of 1:1,600 and 1:320 at 23
days and 6 months, respectively, after rash onset. The rapidity of
antibody appearance, the high titer achieved, and the absence of
immunoglobulin M (IgM) antibody in all of the specimens sug-
gested a secondary immune response. Although reports exam-
ining immunity after infection rely on antibody determination,
immunity relies heavily on T-lymphocyte memory and function.
Measles infection in the immunocompromised host (eg,
persons with malignancies or human immunodeficiency
virus [HIV] infection) can be prolonged, severe, and frequently
fatal.175–180 Infection in these persons may occur in the absence
of rash.175,178–181 The severity of illness is believed to be due pri-
marily to impaired cell-mediated immunity.182–185 Two espe-
cially severe complications are an acute progressive encephalitis
(measles inclusion body encephalitis)98,186,187 and a characteristic
giant cell pneumonia (Hecht pneumonia).175–177,188 Measles has
been found to be more severe in persons with HIV infection. In
the United States, the case-fatality rate has been reported to be
as high as 50% in HIV-infected children.189
An atypical variant of measles occurred in some recipi-
ents of killed measles vaccine who were subsequently exposed
to wild-type virus.49,51,163,165,190–202 Early studies found that
patients with atypical measles lacked antibody to the mea-
sles virus fusion (F) protein127,202,203 and had exaggerated cel-
lular responses to measles antigen.75,204 Affected patients had
extremely high levels of measles-specific circulating anti-
body.51,75,76,191,202 Studies in monkeys have shown that this ill-
ness is caused by antigen-antibody immune complexes.205 The
formalin-inactivated vaccine induced complement-fixing anti-
bodies that failed to undergo affinity maturation; after expo-
sure to measles, an anamnestic production of nonprotective,
complement-fixing antibodies resulted in immune complex
deposition and atypical measles.206 After an incubation period
of 1 to 2 weeks, a prodrome consisting of high fever, head-
ache, abdominal pain, myalgia and cough ensued. In the next
2 to 3 days, an unusual rash erupted on the extremities and
spread centripetally. Whereas the exanthem could be erythem-
atous and maculopapular, it was frequently petechial or vesicu-
lar and accompanied by edema, and was occasionally pruritic.
Hepatocellular enzymes were sometimes strikingly elevated.
The illness was frequently mistaken for Rocky Mountain spot-
ted fever and had to be differentiated from meningococcemia,
Henoch-Schönlein purpura, and drug eruptions.49,51 A nodular
pneumonitis with pleural effusion was common.199,207 Despite
the potential for serious illness, there was only one report of a
possible atypical measles-related fatality.208 This syndrome also
has been reported rarely after receipt of live measles vaccine
exclusively.209,210 Persons with atypical measles are believed to
be noncontagious.51,191,192 On the basis of humoral and cellular
immunity studies, they also are thought to be protected from
subsequent illness after exposure to measles.75,76,202 The syn-
drome probably can be prevented by appropriate immunization
with live vaccine.75,76,211,212
Virology
The measles virus is a pleomorphic, nonsegmented, single-
stranded, negative-sense RNA virus with a diameter of 120 to
250 nm.127,203,213 It is a member of the genus Morbillivirus in
the family Paramyxoviridae and is closely related to the canine
distemper, rinderpest, peste-des-petits-ruminants viruses, and
a phocine distemper virus of seals.214,215 Recent phylogenetic
analysis suggests divergence of measles virus from its closest
relative, rinderpest virus, occurred around the 11th to 12th cen-
turies.216 The only natural host for the measles virus is humans.
In susceptible populations, measles is one of the most trans-
missible viruses known. The viral structure is composed of eight
proteins encoded within a 16-kb genome, including replication
factors (polymerase [L] and phosphoprotein [P]), structural pro-
teins (hemagglutinin [H], fusion [F], nucleoprotein [N] and
matrix [M]) and two accessory proteins of unknown function (C
and V).203 The F and H proteins are essential in viral pathogen-
esis. The F protein is involved in the viral/host cell membrane
fusion and viral entry into the host cell. The measles H protein
is involved in the attachment and entry of measles virus into
host cells via interaction with cell surface receptors. Signaling
lymphocyte activation molecule (SLAM, also called CD150) is
a membrane glycoprotein expressed on activated T and B lym-
phocytes and antigen-presenting cells that acts as the principal
cellular receptor for measles virus, accounting for its lymphot-
ropism and immunosuppressive nature.217 Measles virus also
infects respiratory epithelial cells via a recently identified recep-
tor molecule (CD147/EMMPRIN), thereby facilitating trans-
mission via aerosol droplets.218 Vaccine and laboratory-adapted
strains of measles virus use ubiquitously expressed CD46 as an
alternate receptor. Other molecules have also been implicated in
measles virus infection, although their relevance remains to be
determined.
The entire 15,894-nucleotide–long genome of the prototype
Edmonston strain of measles and the genomes of the predomi-
nant measles vaccine strains219–221 have been fully sequenced,
as well as an increasing number of wild-type strains, including
some identified from SSPE cases.222–224 Although measles virus
is considered a monotypic virus, sequence analysis of the
and genes has shown that there are multiple, distinct lin-
eages of wild-type viruses.225–232 This sequence variability has
made it possible to use molecular techniques to help monitor
the transmission pathways of measles virus.233 Molecular epi-
demiologic data, when analyzed in conjunction with standard
epidemiologic information, can confirm or suggest the source
of outbreaks and, over time, can provide a measurement of
the effectiveness of vaccination control programs and moni-
tor elimination.231 For example, genetic analysis of wild-type
viruses isolated in the United States from 1988 to 2000 helped
to document the interruption of endemic transmission in the
United States.234 Comparison of genetic sequences from wild-
type strains isolated in the United States with those isolated
elsewhere in the world has suggested international importa-
tions of measles virus as sources of outbreaks.233 To facilitate
the expansion of virologic surveillance activities, the World
Health Organization (WHO) has recommended a standard
nomenclature and analysis protocol.235 The WHO currently
recognizes 24 genotypes (Figure 20-1) of measles virus based on
phylogenetic analysis of the gene and has established a set of
reference sequences.236 In 2000, the WHO established a global
measles and rubella laboratory network to provide standard-
ized procedures for case confirmation and virus characteriza-
tion.237 Although virologic surveillance is incomplete, increased
laboratory capability and improved surveillance in the past 2
years has seen a pattern for the global distribution of genotypes
(Figure 20-2).238 The biologic significance of differences in the
genetic sequence of wild-type strains is not known; the immune
response generated through vaccination protects against all
strains.
Measles virus can be cultured from clinical specimens and is
best isolated in the cell line, Vero/hSLAM.239 These are Vero cells
transfected with a plasmid encoding the gene for the human
SLAM molecule. The sensitivity of Vero/hSLAM cells for iso-
lation of measles virus is equivalent to that of the previously
used B95a cells, which were coinfected with Epstein-Barr (EB)
virus. In addition, Vero/hSLAM cells are sensitive to laboratory-
adapted measles strains, including vaccine viruses.240 Measles
virus can also be detected directly from clinical specimens, such
as oral fluid or throat swabs, through molecular techniques that
are being used to monitor measles transmission patterns with-
out the need for virus culture.
 Measles vaccine 20 357

reference laboratories and research work.277 Fluorescence tests are

available,278 as are enzyme immunoassays (EIAs), also referred
to as enzyme-linked immunosorbent assays (ELISAs).279–285
Currently EIA tests for measles IgG are the most widely used
because they are generally sensitive and convenient. Good cor-
relation has been shown between HI and Nt antibody in many
studies,286 as well as between EIA and other serologic methods
for the diagnosis of acute measles.280,281,287 The EIA for measles
IgG is based on significant changes in optical density values and
cannot be translated directly to antibody concentrations or titers.
Currently, the recommended laboratory method for the confir-
mation of clinically diagnosed measles is a serum-based IgM EIA
collected at the time the patient is first seen for medical care.288
A single specimen is adequate to detect the presence of IgM anti-
body.274,284,285 A number of commercial kits using an indirect EIA
method (with removal of the patient's IgG antibody before testing
for IgM) or an antibody capture EIA technique (without removal
of IgG) are now available and have been shown to have similar
sensitivity (83%-92%) and specificity (87%-100%).289 However, if
measles prevalence is low (eg, 1%), a modest reduction in the
specificity of the assay from 99% to 95% will decrease the posi-
tive predictive value of the assay from 48% to 15% (ie, only 15%
of IgM-positive clinical cases will be true measles). Patients with
parvovirus B19 or rubella infections may have false-positive reac-
tions (rate of about 4%) when tested with measles IgM EIAs.252,290
Correct interpretation of serologic data depends on proper
timing of specimens relative to rash onset. This is especially
important in interpreting negative IgM results. For example, in
one study, the sensitivity of an antibody capture IgM assay was
approximately 80% within the first 72 hours after rash onset and
rose to 100% between 3 and 14 days after rash onset.291 Sensitivity
of the assay within the first 3 days of rash onset is thought to be
similar for other commercially available kits. If the validity of the
initial measles IgM test is in doubt, a second convalescent speci-
men should be taken and tested for IgM and for a rise in IgG titer
compared with the initial specimen. In some cases, interpretation
may be difficult and requires precise information regarding dates
of rash onset, prior measles vaccination, and specimen collection.
Two approaches provide alternatives to venipuncture for col-
lection of diagnostic specimens: oral fluid and blood spots on fil-
ter paper. In the United Kingdom, a commercial IgM EIA-based
assay has been developed and is routinely used to test oral fluid
specimens for measles, with a reported sensitivity of 94%, speci-
ficity of 91%, and positive predictive value of 99% compared with
serum.292,293 Use of oral fluid samples has appeal because the tech-
nique is noninvasive, can be used for rubella and mumps testing,
does not require processing in the field, can be shipped without
cold chain, and can be used to detect not only measles IgM and
IgG but also the measles virus genome for molecular characteriza-
tion. Although still considered an invasive technique, fingerstick
as a method for sample collection may be more acceptable to par-
ents than phlebotomy.265of Bas
rash
been
onset,
devdprior
requires
measles
precise
vaccination,
information
andregarding
specimendates
collection.
Two approaches provide alternatives to venipuncture for col-
lection of diagnostic specimens: oral fluid and blood spots on fil-
ter paper. In the United Kingdom, a commercial IgM EIA-based
assay has been developed and is routinely used to test oral fluid
specimens for measles, with a reported sensitivity of 94%, speci-
ficity of 91%, and positive predictive value of 99% compared with
serum.

by measles directly among vaccinated persons and indirectly
among unvaccinated persons as a result of decreased trans-
mission.796 In the absence of vaccination, measles occurs in
epidemic cycles. The magnitude and frequency of the epidem-
ics depends on the population size, contact rates between per-
sons, and the rate at which new susceptible persons are added
to the population through births or migration.797 In England
and Wales in the 1940s, epidemics occurred every second year,
starting in the large cities of London, Manchester, and Liverpool
and spreading outward to towns and rural villages.39,798 In the
large population centers, chains of transmission were sus-
tained during the periods between epidemics, and these cities
were the reservoirs of measles. In towns and villages, transmis-
sion died out after an epidemic and had to be reintroduced for
each subsequent epidemic. Epidemic cycles of measles also can
be described in terms of the effective reproduction number, R,
defined as the average number of secondary cases produced by
a typical case in a population (see Chapter 71).532,799,800 When
R is less than 1, the average case gives rise to less than one
case, and the number of cases occurring subsequently begins
to decrease.
Introduction of measles vaccination leads to a reduction in
the size of epidemics and an increase in the interval between
measles epidemics.796 The widened interepidemic interval has
been referred to as the “honeymoon period”.801,802 Vaccination
of successive cohorts of young children results in a decrease in
incidence among vaccinated cohorts, an overall decline in mea-
sles incidence in all age groups (because of dampened transmis-
sion), and, when outbreaks occur, an increase in the proportion
of cases among older children.340,802,803 Susceptibility to measles
among older cohorts occurs most often because these persons
missed natural disease as young children and either missed get-
ting vaccinated (program failure) or failed to respond to vac-
cination (primary vaccine failure). Waning of vaccine-induced
immunity (secondary vaccine failure) has been found to occur
in 0% to 5% of vaccinees but does not appear to play a major role
in reducing overall population immunity to measles.170,629,637
As vaccination coverage increases among successive birth
cohorts, measles transmission decreases, reducing the risk
of measles even among unvaccinated persons. At some vac-
cine-induced immunity level lower than 100%, measles virus
transmission is interrupted.804–806 Mathematical models have
estimated the herd immunity threshold for measles in the
United States at 92% to 95%.807 If this level of population
immunity is maintained by a vaccination program, endemic
transmission of measles will die out and measles elimination
is achieved. Elimination does not result in zero measles cases
because importations of measles from endemic areas continue
to occur. However, spread from these importations is short lived
and will end without intervention. Documenting the elimina-
tion of measles is challenging. De Serres and associates808 have
proposed using the proportion of cases imported and the dis-
tribution of outbreak sizes to monitor elimination. These sur-
veillance parameters can be used to estimate R, which must be
maintained at less than 1 to achieve elimination. Global eradi-
cation of measles will occur when the last chain of transmission
of measles virus is interrupted and can be defined as the simul-
taneous achievement of measles elimination in every country.
In industrialized countries, a single dose of measles vaccine
administered in the second year of life induces immunity in
about 95% of vaccinees.530 With a primary vaccine failure rate
of 5%, 100% of the population would have to be vaccinated
to reach a 95% immunity level with a single-dose strategy.
Approximately 95% of persons who fail to respond to the first
dose respond to a second dose,517 and with high vaccine cover-
age, the herd immunity target can be reached if two doses are
administered.
In developing countries, high morbidity and mortality result-
ing from measles among infants led to a recommendation to
Measles vaccine 20 381
vaccinate infants at 9 months of age, a time when maternal
antibody may interfere with seroconversion.808 Studies have
found the median seroconversion rate with vaccination at age
9 months to be 85% (range, 70% to 98%).809 This leaves three
times more infants susceptible (15% of vaccinees) than does
a rate of 95%. After a single dose with 90% coverage and 85%
seroconversion, 77% of the population would be immune. To
improve immunity, some have considered providing a second
dose, usually through routine health services at an older age,
when seroconversion is 95%. However, if only first-dose recipi-
ents were revaccinated through routine health services, given
90% coverage with both doses, immunity levels could rise to
only 90%, leaving 10% susceptible. Alternatively, if vaccination
activities were intensified both within and outside the routine
health services, such that many children never vaccinated could
receive a first dose while previously vaccinated children receive
a second dose, the resulting population immunity level would
be much higher. This is known as a second opportunity strat-
egy. For example, if the second opportunity reached 90% cover-
age of prior first-dose recipients and previously unvaccinated
children, population immunity would increase to greater than
95%. In countries with poor access to preventive services, the
second opportunity for measles vaccination is most often pro-
vided through nationwide supplementary immunization activ-
ities or mass campaigns, because these are more effective at
reaching children who have never been vaccinated.
The impact of measles vaccination on overall childhood
mortality in developing countries was questioned by one study
that reported replacement mortality (ie, later death from other
causes).115 However, several subsequent studies have docu-
mented that measles vaccination increases overall child sur-
vival.810–813 In Bangladesh, investigators found that the overall
mortality rate in measles-vaccinated children was 45% lower
than in unvaccinated children and that this difference persisted
for several years after vaccination.810 A similary large impact
of measles vaccination on overall child mortality was found in
rural India, where children who received measles vaccination in
infancy had 27% lower mortality.813 A comparison of the propor-
tion of all-cause mortality resulting from measles, among chil-
dren ages less than 5 years, found measles accounted for about
7% of all child deaths in 1990 and 1% in 2008, contributing
23% to the overall reduction in mortality in this age group.814
Experiences with measles control and elimination
in various countries
Experience in industrialized countries has shown that a sin-
gle dose of measles vaccine, widely administered, can reduce
measles transmission, but a two-dose strategy is necessary
for elimination of indigenous transmission.815–818 Many coun-
tries introduced measles vaccine as a single-dose schedule
and then added a second opportunity for vaccination because
of the persistence of outbreaks despite high single-dose cover-
age (Table 20-10). Industrialized countries have found that the
costs of patient care and outbreak control outweigh the cost
of a second opportunity for immunization.819 Some countries
(eg, Canada, United Kingdom, Australia, and New Zealand)
have adopted a two-dose schedule and have conducted a one-
time, nationwide vaccination campaign to reduce susceptibil-
ity among school-age children. Other countries (eg, the United
States, Finland, and Sweden) have relied on a two-dose sched-
ule alone.
In the United States, the distribution and use of more than
350 million doses of live measles vaccine between the year of
licensure (1963) and the end of 2010 (see Figure 20-6) has been
associated with a marked reduction in measles (Figure 20-9) and
its associated complications and estimated savings of billions
382 SECTION TWO
Evolution of Measles Vaccination Policy in Selected Countries and the Expanded Programme on Immunization (EPI), 1963 to 2010
WHO, World Health Organization.
of dollars.26,356 Whereas approximately 4 million cases occurred
annually in prevaccine years, on average 400,000 to 500,000
cases were reported. In 2010, only 64 confirmed cases were
reported, with 87.5% of these cases linked to importations.
This represents a reduction of more than 99.9% compared
with the years preceding vaccine licensure. The reported occur-
rence of SSPE also has declined greatly (see Figure 20-8) and
was virtually eliminated in the late 1980s and early 1990s.
A small number of cases had been reported each year during
the late 1990s as a result of the resurgence of measles during
1989 to 1991.
The United States has embarked on three elimination
efforts (see Figure 20-9).22,23,805,820 The first, in 1966, was based
on a single-dose strategy with vaccination at 12 months of
age (see Table 20-10). The second, announced in 1978, had
three components: (1) a high level of population immunity
through vaccination with a single dose of measles vaccine, (2)
disease surveillance, and (3) prompt response to outbreaks.22,25
To reach high coverage, vaccination requirements for school
entry were enacted and enforced in every state. These require-
ments, which mandated not only measles vaccination but also
other childhood vaccinations, have become the major legacy of
this elimination initiative. Although sustained interruption of
transmission of measles was not achieved, measles was elimi-
nated from most of the country; 54% of counties in the United
States were measles free for the entire decade from 1980 to
1989, and only 17 (0.5%) counties reported measles every year
during that period.821 During this interval, outbreaks contin-
ued to occur in schools among school-age persons with his-
tories of vaccination and among unvaccinated preschool-age
children.338,499,556,560,633,822 In 1989, the United States introduced
a routine second dose of measles vaccine at 4 to 6 years or 11
to 12 years to address the problem of school outbreaks (see
Table 20-10).
After relatively low incidence during the 1980s, a 3-year epi-
demic of measles began in 1989 and resulted in over 55,000
cases and 123 deaths.823,824 The average annual incidence dur-
ing 1989 to 1991 was 7.4 per 100,000 population compared
with an average incidence for 1981 to 1988 of 1.8 per 100,000.
Incidence was increased in all age groups; however, the greatest
increases were in children younger than 1 year and 1 to 4
years, resulting in almost half of all cases occurring in children

Case confirmation
Traditionally, confirmation of meningococcal disease is achieved
through conventional cultures of blood, cerebrospinal fluid (CSF),
hemorrhagic skin lesions, or other infected sites. Gram staining
to demonstrate the characteristic gram-negative diplococcus can
also be used as a confirmatory tool. Latex agglutination test-
ing for detection of meningococcal capsular polysaccharides in
serum or CSF also may be used, but false-negative results are
common. For rapid capsular grouping, a gold particle immuno-
chromatography method for the detection of soluble polysaccha-
ride has been developed.126 This assay has the potential to be of
great value for rapid identification of disease and capsular groups
in settings such as the African meningitis belt and for selection
of meningococcal vaccines for mass vaccination campaigns.
The potential for meningococcal disease to progress rapidly
with fatal outcomes has led to recommendations for initiating
early antibiotic treatment, often before complete collection of diag-
nostic specimens.127 As a result, an apparent decrease has been
observed in the number of culture-confirmed cases among patients
with clinical syndromes suggestive of meningococcal infection.127
Because asymptomatic meningococcal carriage is common, a
positive pharyngeal swab cannot be used to establish a defini-
tive diagnosis. To enhance diagnostic accuracy in culture-negative
meningococcal disease, a polymerase chain reaction (PCR) assay
that detects meningococcal DNA in normally sterile body fluids
can be used to establish the diagnosis. PCR assays rely on con-
served regions of the gene (a capsular transport gene) or the
gene (a transcriptional regulator belonging to the LysR family)
for DNA detection.128 Capsular group-specific PCR assays using
sequences within the sialyltransferase ( ) gene are also available
to discriminate among capsular groups B, C, Y, and W135 menin-
gococci.128,129 For confirmation of group A disease, PCR amplifi-
cation targets two open reading frame sequences within the gene
cassette required for the synthesis of the capsule for group A.106
PCR assays for capsular groups 29E, X, and Z also have been developed
using group-specific regions in the gene.130
In an expanding number of European countries, a substan-
tial proportion of meningococcal cases are detected using PCR,
as opposed to the United States where this diagnostic test is
used primarily as a research tool. PCR has greatly enhanced
case detection in the United Kingdom: in 2008, 55% of all cases
were diagnosed based on positive PCR in settings of clinically
compatible illnesses.131,132 In Greece, 57% of cases of meningo-
coccal disease reported during the 2001-2004 period were diag-
nosed using PCR. In São Paulo, Brazil, the incorporation of PCR
into routine public health surveillance for bacterial meningitis
increased the number of diagnosed cases of meningococcal dis-
ease by 85%.133 The use of PCR for diagnosis of meningococcal
disease has dual benefits: providing a more accurate picture of
disease burden and permitting the diagnosis to be established in
a greater proportion of clinically compatible cases.
Treatment and prevention
Antibiotics
Antimicrobial therapy remains the cornerstone of management
of meningococcal disease. Although many antimicrobial agents
are effective, intravenous aqueous penicillin is generally consid-
ered the therapy of choice once meningococcal disease is con-
firmed. A small proportion of meningococcal isolates from the
United States have intermediate susceptibility to penicillin (min-
imum inhibitory concentration, 0.1-1.0 g/mL), but the clinical
importance of this finding is unclear.10,134 In Belgium, the preva-
lence of clinical isolates with decreased susceptibility to penicillin
Meningococcal vaccines 21 391
has increased; however, there has been no increase in resistance
to expanded-spectrum cephalosporins, such as ceftriaxone.134a
Early initiation of antibiotic therapy and attention to circulatory
status in patients suspected of having meningococcal infection
may help decrease mortality and morbidity.135,136 Because the clini-
cal features of meningococcal disease can overlap those of bacterial
meningitis or sepsis caused by other bacterial pathogens, depending
on the age of the patient, clinical manifestations, exposure history,
the results of Gram stains of CSF or petechial hemorrhages,
and the presence or absence of other epidemiologic risk factors,
broader initial empirical therapy should be considered before the
diagnosis is confirmed.10 Because of concerns about penicillin- or
cephalosporin-resistant , cefotaxime or ceftriax-
one, given by the intravenous or intramuscular route in combi-
nation with vancomycin, should be added empirically in patients
with meningitis of unknown etiology. In patients at risk for
meningitis caused by , coverage for this
organism is also included. For treatment of meningococcal disease
in patients with type I hypersensitivity to -lactam antimicrobials,
chloramphenicol also is effective, although resistance to this drug
has been reported in some countries.137 Antibiotic treatment for 5
to 7 days is adequate therapy for most systemic meningococcal ill-
nesses. In resource-poor settings, single-dose intramuscular chlor-
amphenicol in oil or ceftriaxone can be used.138
Other therapies
Patients with purpura fulminans continue to have the highest
mortality rates (up to 50%). This syndrome is a result of a mas-
sive release of proinflammatory mediators139 and development
of shock and disseminated intravascular coagulation from a
procoagulant state with local accumulation of thrombi in small
and large arterial vessels.
Activated protein C is an endogenous anticoagulant that sup-
presses excessive thrombosis and fibrin formation.140,141 The mol-
ecule also has a dual anti-inflammatory role,142 downregulating
TNF- and IL-1 production in animal models of sepsis and
decreasing levels of IL-6 in patients with sepsis. Decreased levels
of protein C are present in most patients with meningococcal
sepsis,140,143 and patients with the lowest levels have the worse
prognosis. The results of two open-label studies in patients with
meningococcemia suggested that treatment with protein C con-
centrate prepared from human plasma decreased mortality and
morbidity compared with those of historical control subjects.144,145
The results of a large, prospective, randomized trial in patients with
severe sepsis caused by a variety of gram-positive or gram-negative
bacteria or fungi indicated that treatment with recombinant human
activated protein C [drotrecogin alfa (activated)] decreased the 28-day
mortality rate by 19.4% compared with that of patients treated with
placebo (95% confidence interval [CI], 6.6%-30.5%; = .005).141
The main adverse event associated with drotrecogin alfa (activated)
therapy was an excess of serious bleeding (3.5% vs 2.0%; = .06).
Based on these data, the US Food and Drug Administration (FDA)
approved the use of drotrecogin alfa (activated) (Xigris, Eli Lilly) for
adjunctive therapy in adults with septic shock.
Since licensure, additional data about this agent have become
available. An increased risk of serious bleeding events and of
death was observed in patients with sepsis and risk factors for
bleeding at baseline who received drotrecogin alfa.146 In a ran-
domized, double-blind, placebo-controlled trial with activated
protein C in pediatric patients, enrollment was terminated
earlier than planned because of concerns about lack of efficacy
and an increased rate of central nervous system bleeding.147
The European Medicines Agency (EMEA) recommendations
supported using drotrecogin alfa only in high-risk patients, in
whom therapy could be started early. Furthermore, the EMEA
recommend that drotrecogin alfa not be used in patients with
single organ dysfunction, especially if they have had surgery
within 30 days (http://guidance.nice.org.uk/TA84/Guidance/
392 SECTION TWO Licensed vaccines
pdf/English). Further studies have failed to show a survival
benefit, and the FDA and EMEA announced on October 25, 2011,
that Eli Lilly was withdrawing drotrecogin alfa from the market
(http://www.fda.gov/Safety/MedWatch/SafetyInformation/
SafetyAlertsforHumanMedicalProducts/ucm277143.htm and
http://www.ema.europa.eu/ema/index.jsp?curl=pages/news_
and_events/news/2011/10/news_detail_001373.jsp&mid=
WC0b01ac058004d5c1&jsenabled=true).
Results from a phase 3 clinical trial, in children with menin-
gococcal disease, of bactericidal permeability-increasing pro-
tein, which neutralizes endotoxin and has bactericidal activity
against reported lower morbidity than patients
given placebo as adjunctive therapy, but there was no statisti-
cally significant difference in mortality, the primary end point
of the trial.148 Because of a low mortality rate in the control sub-
jects, this trial was underpowered to detect a significant differ-
ence for this end point.149
Dexamethasone therapy for 4 days, with the first dose given
before the initiation of antibiotic therapy, is reported to decrease
the rate of neurosensory deafness in patients with meningitis
caused by type b. In a randomized controlled trial
in adults with bacterial meningitis, dexamethasone use also
was associated with decreased mortality. A statistically sig-
nificant benefit was seen only in patients with pneumococcal
meningitis, but the study was not sufficiently powered to dem-
onstrate efficacy for other pathogens.150 High doses of cortico-
steroids have been demonstrated not to be beneficial in septic
shock. Although low-dose corticosteroids are often adminis-
tered,151,152 a recent randomized, controlled study in patients
with septic shock (CORTICUS trial) failed to demonstrate a
benefit in overall 28-day mortality rate (33% for hydrocorti-
sone recipients vs 31% for placebo recipients). Although the
duration of shock was shorter in patients who received cortico-
steroids, adverse events, including new episodes of sepsis and
septic shock, were more frequent in the hydrocortisone group.
Sprung et al153 concluded that for the majority of patients with
septic shock, the potential advantages of corticosteroids do not
outweigh the adverse risks.
Chemoprophylaxis
Close contacts of a patient with meningococcal disease are at
greatly increased risk of acquiring disease. The risk of acquir-
ing disease is greatest in household contacts and other close
contacts exposed to oral secretions through kissing or sharing
of eating utensils or glasses and in coattendees at day care. The
risk of transmission to health care workers is low, but persons
performing mouth-to-mouth resuscitation and unprotected
health care workers exposed during management of endotra-
cheal tubes are at increased risk.154 Numerous fatal cases of
meningococcal disease among microbiology laboratory workers,
primarily workers who processed cultures on an open bench,
also have been reported.155,156 Rifampin, ceftriaxone, and cipro-
floxacin have been demonstrated to eradicate nasopharyngeal
colonization, and these agents are recommended
for chemoprophylaxis of close contacts.157 Fluoroquinolone-
resistant isolates were reported in 2008 in North Dakota and
Minnesota, which prompted the Centers for Disease Control
and Prevention (CDC) to recommend ceftriaxone, rifampin, or
azithromycin for eradication of meningococcal carriage in areas
where resistance has been identified.158,159
Chemoprophylaxis should be offered to household members
and to other persons with a history of prolonged close con-
tact. In the United States157 but not in the United Kingdom,135
prophylaxis of day-care contacts and staff is recommended.
Because the risk of disease in contacts is highest in the first
week after onset of disease in the index patient, prophylaxis
should be administered to contacts as soon as feasible, pref-
erably within 24 hours of identification of the index patient.
Penicillin therapy does not reliably eradicate nasopharyngeal
colonization. Therefore, antimicrobial treatment is also rec-
ommended at the time of hospital discharge for eradication of
colonization of patients treated with penicillin. Contacts who
have been previously immunized with meningococcal vaccines
should still be considered for chemoprophylaxis because pri-
mary vaccine failure and/or lack of serum antibody persistence
may render vaccinated persons susceptible to disease. In addi-
tion, a substantial proportion of cases are caused by group B
which is not covered by current vaccines.
Chemoprophylaxis given to large populations in response
to outbreaks of meningococcal disease has not been demon-
strated to be efficacious in lowering the risk of disease and
increases the likelihood of emergence of antimicrobial resis-
tance. In small school-based outbreaks, chemoprophylaxis to
all people within the population may be considered. If under-
taken, the medication should be administered to all members
at the same time.
Epidemiology
Incidence and public health burden
Although meningococcal disease is a global problem that occurs
in all countries,26 the true burden of disease is unknown for
many parts of the world because of inadequate surveillance.160
The epidemiology is highly variable, dynamic, and fluid and
is influenced by natural variation and immunization policy.
In addition, the epidemiology varies substantially by capsular
group. Group A meningococcus is remarkable for its ability to
cause large epidemics, which represent the most serious pub-
lic health issue caused by . 4
Since World War
II, epidemic meningococcal disease caused by group A organ-
isms largely has been confined to developing countries, most
often reported from the meningitis belt of sub-Saharan Africa.
In this region, disease can spread rapidly and cause extremely
high rates of disease. Capsular group X and W135 strains also
cause substantial rates of disease in Africa.
Only incomplete information is available on the epidemiol-
ogy of meningococcal disease in Asia, in part because of absence
of surveillance in many countries and inadequate bacterial
detection methods.161 In the 1960s and 1970s, high rates of
capsular group A disease were reported in China, Mongolia, and
Nepal. As recently as 2005, group A epidemics were reported in
India, and a meningococcal capsular group A outbreak occurred
in the Philippines in 2004 and 2005, with 98 cases (33% mor-
tality) reported to the World Health Organization (WHO)
(http://www.who.int/csr/don/2005_01_28a/en/index.html).
In Europe, most cases of meningococcal disease are sporadic
and currently caused by capsular group B and C strains. The
burden of endemic meningococcal disease varies substantially
among countries and has declined dramatically in countries
that introduced group C meningococcal conjugate vaccines.
Using culture-confirmed cases, countries with the highest
reported incidence rates in 2008 were the United Kingdom (2.1
cases/100,000 population), Spain (1.3), and Slovenia (1.2).
The annual incidence of meningococcal disease in the
United States between 1978 and 2008 is shown in Figure 21-1.
The annual incidence of meningococcal disease since 2000 has
been remarkably low, with an incidence in 2009 of 0.28 per
100,000 (http://www.cdc.gov/abcs/reports-findings/survreports/
mening09.html).162 Although the reasons for the substantial
fluctuations in meningococcal disease incidence are unknown,
a study from Maryland found that increases in capsular group
C and Y disease incidence during the 1990s were associated
with antigenic shifts involving key meningococcal outer mem-
brane proteins.35

Meningococcal disease incidence per 100,000
population, by year United States, 1978-2008 (Centers for Disease
Control and Prevention. Summary of notifiable diseases: United States,
2008. MMWR Morb Mortal Wkly Rep 57:1-100, 2010).
The age-specific incidence of meningococcal disease
results from an interplay between the degree of immu-
nity in the population and virulence of circulating strains.
Disease incidence is highest among infants who have not
yet acquired natural immunity. In the European Union,
the reported rate of meningococcal disease (primarily sero-
group B) during 2008 among children younger than 5 years
was 8.51 per 100,000 (www.ecdc.europa.eu/en/publications/
Publications/1011_SUR_Annual_Epidemiological_Report_
on_Communicable_Diseases_in_Europe.pdf). During the
same period, the incidence in the United States was 3.64,
0.76, and 0.26 among children younger than 1, 1, and 2 to
4 years (http://www.cdc.gov/abcs/reports-findings/survreports/
mening08.html).
In the United States and Europe, there are peaks in inci-
dence among adolescents and young adults (representative data
for Europe, Figure 21-2).163 In the United States and United
Kingdom, college student groups have been at increased risk,
which may be related to increased transmission in dormito-
ries and residence halls and other social behaviors in this age
group.77,164–166 In the United States, the case fatality rate among
adolescents and young adults was disproportionately higher
than that in children younger than 15 years, particularly for dis-
ease caused by capsular group C strains.167
Age incidence of meningococcal disease, European
Union and European Economic Area, 2007. Data from European Centre for
Disease Prevention and Control (ECDC), Stockholm, Sweden, 2007. http://www.
ecdc.europa.eu/en/publications/publications/0910_sur_annual_epidemiological_
report_on_communicable_diseases_in_europe.pdf.
Meningococcal vaccines 21 393
In temperate climates, rates of meningococcal infections are
highest during the winter months. This association may be a
result of closer personal contact during the winter months, lack
of ventilation, or an increase in upper respiratory infections,
all factors that facilitate transmission of and/or invasion by the
organism.
Meningococcal capsular group A disease
Group A strains of are the major cause of epi-
demic and endemic meningococcal disease in the meningitis
belt of sub-Saharan Africa, which extends from Senegal on the
west coast to Eritrea on the east coast (Figure 21-3).26,168 These
strains rarely cause disease in the United States162 or in Europe
( www.ecdc.europa.eu/en/publications/Publications/1011_
SUR_Annual_Epidemiological_Report_on_Communicable_
Diseases_in_Europe.pdf). Notable exceptions during the past
several decades were an outbreak in the United States in the
Pacific Northwest in the 1970s among people living on skid
row,169 an epidemic in Finland in the 1970s,170,171 and, more
recently, in Russia, where group A isolates accounted for more
than 10% of cases.172,173
During epidemics, the annual incidence of disease during
group A epidemics in sub-Saharan Africa can exceed 1,000 per
100,000 population, or 3,000-fold higher than the current inci-
dence of meningococcal disease in the United States. During the
2009 epidemic season, there were more than 80,000 reported
cases, with 5,352 deaths, which was the largest number of cases
since an epidemic that occurred in 1996 (http://www.who.int/
mediacentre/factsheets/fs141/en/). In the beginning of 2011, a
group A epidemic involving 923 suspected cases occurred in five
districts in Chad, which resulted in the distribution of 752,000
The sub-Saharan African meningitis belt. Countries
shown in yellow experience explosive epidemics of meningococcal
disease every 5 to 10 years that are usually caused by capsular group
A strains (Courtesy of the World Health Organization).
 SECTION TWO
doses of bivalent A/C polysaccharide meningococcal vaccine for
a mass immunization campaign.174 An estimated 80% to 85% of
meningococcal disease cases in this region are currently caused
by group A strains, with most of the remainder being caused
by groups W135 and X. Why disease caused by strains with
these capsular groups is concentrated specifically in the sub-
Saharan region and why epidemics begin with the dry season
(generally November or December) and terminate abruptly with
the onset of increased humidity and rain (generally June and
July) is unknown. Most likely, environmental, host, and micro-
bial factors have a role.175 The case fatality rate during group A
epidemics in Africa varies between 10% and 15%.4 Genotypic
analyses of epidemiologically defined group A isolates indi-
cate that pandemic outbreaks of group A disease are caused by
the global spread of specific clonal lineages. Work by Olyhoek
and colleagues176 demonstrated how the lineage known as sub-
group III/ST-5 complex was responsible for two global pandem-
ics during the latter half of the 20th century. The first started
with large epidemics in China in the mid-1960s and subse-
quently spread, causing epidemics in Moscow (1969-1971),177
Norway (1973), Finland (1975), and, finally, Brazil in the mid-
1970s. After an interval of about 10 years, ST-5 strains of
reemerged, causing new outbreaks in China,
Nepal, and, in 1987, among pilgrims to the Hajj. Epidemics
from strains in this clonal complex followed in Sudan, Chad,
Ethiopia, Kenya, and Tanzania178 and spread into Zambia and
the Central African Republic by the mid-1990s.179 Mass immu-
nization with a new capsular group A conjugate vaccine recently
was introduced as part of demonstration projects in three coun-
tries in sub-Saharan Africa and is expected to have a major role
in decreasing or even eliminating group A epidemics (see sub-
sequent discussion).
Meningococcal capsular group B disease
Group B disease is one of the most important causes of endemic
disease in industrialized countries, accounting for up to 90%
of sporadic meningococcal cases in certain European countries
with successful group C vaccine programs and about a third
of cases in the United States in 2009 (http://www.cdc.gov/abcs/
reports-findings/survreports/mening09.pdf). As compared with
group A epidemics, group B epidemics begin more slowly, are
associated with lower rates of disease, and are more prolonged,
at times lasting a decade or longer. Group B epidemic disease
also is associated with emergence of hypervirulent strains. For
example, in the 1970s, an increase in the incidence of dis-
ease caused by ST-32 group B strains was noted in Norway
(B:15:P1.7,16) and in Spain (B:4:P1.19,15).180 A severe epi-
demic in Cuba in the 1980s was also caused by ST-32 group B
strains (B:4:P1.19,15), which spread to São Paulo, Brazil, in the
late 1980s.180 Group B meningococci strains from this genetic
complex became established in Canada in the 1990s, and sub-
stantial increases in incidence also were noted in the United
States in Oregon and Washington State (B:15:P1.7,16).181 The
incidence of disease in Oregon caused by this outbreak clone
peaked in the 1994-1996 period, but the clone has remained an
important cause of disease (Paul Cieslak, personal communi-
cation, May 2011) which underscores the tendency of group B
epidemics to be of long duration.
New Zealand experienced an epidemic of group B menin-
gococcal disease that started in 1991.182,183 The epidemic was
caused by a strain identified as B:4:P1.7-2,4, ST-41/44 com-
plex. A tailor-made outer membrane vesicle vaccine against
this strain was developed and was introduced in July 2004 for
mass immunization in New Zealand (see discussion on pros-
pects for control of disease caused by group B strains). Because
of decreased disease incidence, in April 2008, the decision was
made to discontinue the immunization program.184
Capsular group distribution of meningococcal disease
cases, by age group in the United States. (Data from Cohn AC, MacNeil JR,
Harrison LH, et al. Changes in disease epidemiology in the
United States, 1998-2007: implications for prevention of meningococcal disease.
Clin Infect Dis 50:184-191, 2010.)
One of the most important epidemiologic features of group B
meningococcal disease is that it is the predominant cause of dis-
ease in infants (Figure 21-4). This fact has two implications for
vaccine prevention: (1) There are currently no licensed vaccines
for prevention of group B disease. (2) The predominance of group
B disease in infants is one of the justifications made by Advisory
Committee on Immunization Practices (ACIP) for not recom-
mending universal quadrivalent (A, C, Y, and W135) meningococ-
cal conjugate vaccination in this age group (see later discussion).
Meningococcal capsular group C disease
The percentage of group C isolates responsible for endemic menin-
gococcal disease varies by country. In the United States, group C
isolates accounted for approximately 25% to 40% of disease from
1990 to 2009. Many countries in Europe have low rates of group C
disease since the introduction of group C conjugate vaccines. For
example, group C strains caused 30% to 39% of all meningococ-
cal disease in Ireland and the United Kingdom in 1999. Following
introductions of these vaccines in the United Kingdom in 1999
and in Ireland in 2000, by 2004 less than 10% of meningococcal
disease in these countries was caused by group C strains.
ST-11 group C outbreaks occurred in the 1960s in US Army
recruits185 and in the 1970s in an urban setting in São Paulo,
Brazil.186 Brazil has experienced three waves of group C menin-
gococcal disease since the 1970s, the first, second, and third
being caused by strains belonging to ST-11, ST-8, and ST-103
clonal complexes, respectively.187 In the 1990s, Spain had a
large outbreak of group C disease caused by strains with the
ST-8 clonal complex.188 In the 1990s, another clone emerged
belonging to the ST-11 clonal complex and caused a large group
C outbreak in Canada189,190 before spreading worldwide.191 The
increase of group C disease in the United Kingdom in the 1990s
was caused by ST-11 isolates, which caused outbreaks in uni-
versities and was associated with high fatality rates.192
Meningococcal capsular group Y disease
Capsular group Y strains cause all of the major meningococcal
clinical syndromes, but patients with group Y disease are more
likely to have pneumonia than patients with disease caused by
strains with other capsular groups.193,194 In the late 1980s, group
Y was a relatively uncommon cause of invasive
disease. However, in the United States, the incidence of serogroup
Y disease increased and the proportion of disease caused by group

Y strains rose from 2% of disease in 1989 through 1991195 to
11% in 1992 and to 33% in 1996.196,197 In 2009, 37% of IMD in
the United States was caused by group Y strains (http://www.cdc.
gov/abcs/reports-findings/survreports/mening09.html). Group
Y disease affects all age groups in the United States, including
infants, although it is the predominant capsular group among
strains causing disease in the 65-year-old and older group. Group
Y strains also caused a substantial proportion of recent cases
of disease in Colombia, parts of Canada,198,199 South Africa,200
Sweden,201 and other European countries.201a For reasons that
remain unexplained, disease caused by group Y strains remains
relatively rare in most other parts of the world, despite observa-
tions that group Y strains are some of the most common strains
found among encapsulated carrier isolates.61,64,202–204
Meningococcal capsular group W135 disease
Until 2000, capsular group W135 was a relatively uncommon
cause of invasive disease worldwide and had not been known to
cause outbreaks. Large numbers of pilgrims from the meningi-
tis belt of Africa congregate at the Hajj, and in 2000 and 2001,
outbreaks of group W135 disease during the Hajj pilgrimage
were reported in Saudi Arabia.205–208 A group W135 clone that
belonged to the ST-11 complex caused the outbreak. This evi-
dence suggested that a capsular group C to group W135 switch
had occurred at some point. Group W135 strains belonging to
the ST-11 complex were also isolated from patients with menin-
gococcal disease in other African countries and globally.209–212 In
2002, more than 10,000 cases caused by group W135 isolates
were reported from Burkina Faso.213,214
Control of emerging epidemics of W135 disease in sub-
Saharan Africa poses a challenge because the monovalent
group A meningococcal conjugate vaccine that is currently
being introduced into Africa does not contain group W135
polysaccharide. In response to a request from the WHO,
GlaxoSmithKline formulated a trivalent polysaccharide vac-
cine containing capsular groups A, C, and W135, which is
kept in reserve by WHO for use in group W135 outbreaks in
sub-Saharan Africa. Despite the recent emergence of W135
disease in Africa, serogroup A accounted for 80% to 85% of
cases in the meningitis belt in 2009.
Although there was an outbreak of 14 cases of group W135
meningococcal disease in Florida in the 2008-2009 period,215
group W135 disease remains an infrequent cause of disease in
the United States.162 For the 1998-2007 period, 2.5% of menin-
gococcal infections in the United States were caused by group
W135 strains.162 An increase in disease caused by group W135
strains has been reported from Argentina216 and southern
Brazil.217 For example, the proportion of meningococcal cases
caused by group W135 strains among children in Argentina
increased from 7% in 2006 to 28% in 2008.218
Meningococcal immunity
Natural acquisition of humoral immunity and
correlates of protection
Serum antibodies confer protection against meningococcal dis-
ease by activating complement-mediated bacteriolysis and, pos-
sibly, by enhancing phagocytosis (opsonic activity).219 Naturally
acquired serum antibodies are elicited by asymptomatic naso-
pharyngeal carriage of pathogenic and nonpathogenic menin-
gococci,220–222 or 62,223–225
Anticapsular antibodies
stimulated by gastrointestinal colonization with
or strains with cross-reacting polysaccha-
rides also may contribute.226–228
Meningococcal vaccines 21 395
In the 1960s, Goldscheider et al89 investigated the role of
serum complement-mediated bactericidal antibodies in pro-
tection against meningococcal disease in military recruits.
Recruits with serum bactericidal titers of 1:4 or greater who
were exposed to a group C meningococcal epidemic did not
develop disease. Virtually all cases occurred in persons whose
baseline serum titers were less than 1:4 as measured with
human complement. The importance of serum bactericidal
antibody in protection also is underscored by the greatly
increased risk of acquiring meningococcal disease in persons
with inherited complement deficiencies who lack SBA.93,229–233
Indirect evidence frequently cited as supporting protection
against meningococcal disease by SBA is an inverse relationship
between the average age of acquisition of naturally acquired
serum bactericidal antibody and the incidence of meningococ-
cal disease, which was first postulated by Goldschneider et al89
in the 1960s (Figure 21-5A). Serum bactericidal antibodies
were rarely detected in children 2 months to 2 years of age, the
age group with the highest incidence of disease. The majority
of adults, in whom disease was rare, had serum bactericidal
titers of 1:4 or greater measured against group A, B, and C
strains (data for group A not shown; bactericidal activity to Y
and W135 was not evaluated).
In more recent studies, however, the relationship between the
decline in incidence of meningococcal disease in the first 4 years
of life and age-related acquisition of serum bactericidal antibody
is less clear. For example, data from the United Kingdom for
the years 2000 to 2004 show that the incidence of group B dis-
ease declined rapidly beginning at age 7 months when the preva-
lence of serum bactericidal antibody was low. The prevalence of
group B serum bactericidal titers of 1:4 or more did not increase
until after age 12 to 14 years (Figure 21-5B),222 which coincided
with an increase in asymptomatic nasopharygneal meningococ-
cal carriage (Figure 21-5C).65 In the 10-year period from 1998 to
2007, the incidence of group B disease in the United States was
much lower than in the United Kingdom. The age incidence of
group B disease in the United States, however, also declined rap-
idly beginning at age 8 months162 when the prevalence of SBA
was low. Thus, natural acquisition of serum bactericidal anti-
bodies does not explain the rapid fall in the incidence of group B
disease beginning at 7 to 8 months of age in both countries. More
likely explanations include acquisition of protective antibodies
that lack complement-mediated bactericidal activity, enhanced
innate immunity from maturation of the alternative comple-
ment pathway, and changes in exposure to .234
In a recent study in the United States, fewer than 25% of healthy
adults had serum bactericidal titers of 1:4 or greater against a panel
of genetically diverse group B strains (Figure 21-6, upper panel). In
the San Francisco Bay area of California, the prevalence of SBA to
capsular group A, C and W135 strains also was low (Figure 21-6,
lower panel).235–237 Other studies confirm that the prevalence of
naturally acquired SBA in adults is less common than in the
1960s. In a study performed in British Columbia in 1991-1993,
which used group C strain C11 (the same strain as used in the
study by Goldschneider et al89), only 10% of adolescents, ages 13 to
19 years, had bactericidal titers of 1:4 or greater,238 compared with
almost 80% of adults in the 1960s (Figure 21-5A). In the United
Kingdom, only 10% to 30% of recently obtained serum samples
from unimmunized adults were positive for group C serum bacte-
ricidal antibody.239 Taken together, the data suggest that the preva-
lence of SBA in adults has decreased since the 1960s without an
apparent increase in incidence of disease. Indeed, in recent years,
the annual incidence of meningococcal disease in adults in the
United States is at an all time low ( 1/100,000; Figure 21-1).26,162
A historically low rate of asymptomatic meningococcal carriage in
US high school students was also recently found.60
In the aforementioned military study by Goldschneider et al,89
more than half of the recruits who lacked SBA and who became
396 SECTION TWO Licensed vaccines
Relationship between incidence of meningococcal
disease and age acquisition of serum bactericidal activity (SBA).
A, Historical data on age incidence of disease caused by capsular
group B and C strains and prevalence of serum bactericidal titers of
1:4 or greater (SBA, human complement) in the United States, 1965-
1966. (Adapted from Goldschneider et al.89) B, Incidence of group B
disease in the United Kingdom, 2000-2004, in relation to prevalence
of SBA, group B, (Adapted from published data;219,222 reprinted with
permission). C, Meta-analysis of meningococcal carriage by age from
published studies when swabs were plated immediately.65 Solid line
represents the fitted regression line. The greatest variation was seen
for 20- to 29-year-olds, with reported prevalence ranging from 2.6%
to 60.7%. (Figure modified from previously published figure.65) CI,
confidence interval.
Prevalence of serum bactericidal activity (human
complement) in serum samples from adults, San Francisco Bay Area,
California. Serum bactericidal titers 1:4 (human complement) in
serum samples from unvaccinated adults. Each strain tested against
30 to 100 serum samples. Upper panel, Representative group B strains
with different genetic lineages. (Courtesy of George Santos, Novartis
Vaccines, Emeryville, CA). Lower panel, Group A, C, Y, and W135
strains. (Figure prepared from published data.235,258)
colonized with the epidemic strain did not develop disease.89
Conceivably these recruits may have been protected by nonbacte-
ricidal serum antibody.219 C3b and the Fc portion of IgG antibody
are known ligands for interacting with receptors on phagocytic
cells and, thus, can serve as opsonins to enhance uptake of bac-
teria. Platonov and colleagues240 observed antibody-dependent
killing of meningococci by neutrophils with serum from late
complement component–deficient patients as a complement
source. In another study, there was a decreased risk of meningo-
coccal disease in late complement component–deficient patients
who were immunized with meningococcal polysaccharide vac-
cine compared with unimmunized patients.241 While these obser-
vations do not address the protective role of naturally acquired
opsonic antibodies, it seems that vaccine-induced opsonic anti-
body can confer protection against meningococcal disease.
Passive immunization
Antibodies administered to infant rats passively confer protec-
tion against meningococcal bacteremia.235,242–251 The antigenic
targets of the protective antibodies in these studies included the
capsular polysaccharide and protein antigens, such as PorA, fH
binding protein, neisserial heparin binding antigen, and NspA
(for discussion of proteins antigens, see section on group B
 Meningococcal vaccines 21 397

vaccines). In some studies of passive protection in the infant rat

bacteremia model, antibodies that lacked bactericidal activity
were protective.235,248,251,252 The mechanism was thought to be
opsonophagocytosis. Studies of the ability of passive immuniza-
tion to confer protection against bacteremia in nonhuman mod-
els of meningococcal bacteremia likely overestimate protective
activity because binding of certain complement downregulating
proteins, such as fH, to is specific for human fH
(see “Pathogenesis”). Nevertheless, passively administered IgG-
containing bactericidal antibody would be expected to confer
protection against meningococcal disease in humans.

Immunity following active immunization

Serum antibody
A number of antigen-binding assays have been used to assess
meningococcal vaccine immunogenicity.253,254 Some investigations
have suggested that an anticapsular antibody concentration of
2 g/mL or greater is sufficient to confer protection against menin-
gococcal disease.170,255 The results, however, of antigen binding
assays such as enzyme-linked immunosorbent assay do not con-
sistently distinguish between bactericidal and nonbactericidal anti-
capsular antibodies.242,243,254,256–258 Therefore, serologic assessments
of meningococcal vaccine immunogenicity rely on measurements
of complement-mediated SBA. In this assay, the test serum is
heated to inactivate endogenous complement. Dilutions of the test
serum are incubated for 60 minutes with the bacteria and a source
of exogenous complement. The results are expressed as the dilution
of serum giving 50% killing of the bacteria. The seminal study by
Goldschneider et al,89 which demonstrated a correlation between
naturally acquired SBA and protection against developing meningo-
coccal disease, used endogenous complement (a fixed dilution of the
human test serum that had been obtained to preserve endogenous
complement and that had not been heated).89 For many years, infant
rabbit serum was used as an exogenous source of complement for
testing vaccine-induced bactericidal activity because of the difficulty
of finding normal human serum without antibody that is suitable as
a complement source for the assay.257,259 Titers measured with rab-
bit complement are much higher than with human complement.260
One reason is the presence of downregulatory complement proteins
in human serum, such as fH, which demonstrate species specificity
in binding to the bacteria.100,261 In the absence of fH binding, the Effect of the addition of human factor H (fH) on survival of
. A and B, Survival of group C strain 4242 when
meningococcus is more susceptible to complement-mediated bac-
incubated with rabbit complement and dilutions of serum samples
tericidal activity. from children immunized with meningococcal polysaccharide vaccine.
Figures 21-7A and 21-7B show the percent survival of a group White circles, no added human fH; gray triangles, addition of a negative
C strain incubated with rabbit complement and different dilu- control complement protein, C1 esterase inhibitor, 25 g/mL; blue
tions of test serum obtained 1 month after immunization of circles, addition of human fH (25 g/mL).261 C, Bacteremia 18 hours after
two children with meningococcal polysaccharide vaccine. In the IP challenge of infant rats with group B strain H44/76 ( 1,000 colony-
forming units [CFU]/rat). Blue circles, human fH transgenic (Hu fH Tg) rats
absence of added human fH, the bacteria were killed by rabbit with serum human fH concentrations of 100-500 g/mL; gray triangles,
complement and antibody at a serum dilution of about 1:800 wild-type (WT) control littermates (human fH 12 g/mL). Adapted from Vu
( 50% survival after incubation for 1 hour compared with colony- D, Shaughnessy J, Lewis LA et al. Infect Immun 2012:80:643–650.
forming units at time 0). The addition of 25 g/mL of human fH
eliminated the killing by binding to the bacteria and downregu-
lating complement activation.261 The addition of a negative con- For meningococcal group B strains, there is consensus that only
trol complement protein (C1 esterase inhibitor) had no effect on human complement bactericidal assays are reliable for predicting pro-
bacterial survival. Wild-type infant rats whose natural fH does tection against disease.262 As of 2012, rabbit complement assays were
not bind to are naturally relatively resistant to still used in Europe for licensure of new polysaccharide-protein con-
disease caused by many pathogenic group B jugate vaccines against capsular group A, C, Y, or W135 strains
strains. After treatment with human fH, the rats became sus- (http://whqlibdoc.who.int/trs/WHO_TRS_926.pdf and http://
ceptible.261 Human fH transgenic infant rats also are highly sus- www.ema.europa.eu/docs/en_GB/document_library/Application_
ceptible to meningococcal bacteremia, whereas control wild-type withdrawal_assessment_report/2010/02/WC500074060.pdf )
littermates cleared the bacteremia (Figure 21-7C). Collectively, and were the basis of inferring efficacy for licensure in India of a
the data illustrate one mechanism by which the meningococcus new group A conjugate vaccine intended for Africa (see subse-
has uniquely adapted to cause disease in humans. quent discussion). For group C strains, bactericidal titers as high
398 SECTION TWO
as 1:128 using rabbit complement are needed to ensure that a
titer of 1:4 is present if measured with human complement.260
Using estimates of age-specific vaccine effectiveness from the
United Kingdom, a titer of 1:8 or greater measured with rabbit com-
plement has been suggested as the putative protective threshold fol-
lowing capsular group C conjugate vaccination.263,264 A titer of 1:128
measured with rabbit complement was the basis of inferring effi-
cacy for US licensure of a quadrivalent meningococcal polysaccha-
ride conjugate vaccine for adolescents (MCV4-DT; see subsequent
discussion). Because meningococci cause disease only in humans
and because of the exquisite human species specificity of binding
complement-downregulating molecules by and the
corresponding uncertainty in inferring vaccine efficacy from bacte-
ricidal titers measured with rabbit complement, the FDA requires
bactericidal data from assays using human complement to support
licensure of new meningococcal conjugate vaccines intended for
infants and children (http://www.fda.gov/downloads/advisorycom-
mittees/…/ucm248586.pdf).
Immunologic B-cell memory
Unconjugated polysaccharides elicit serum antibody responses
largely without T-cell help (so-called, T cell–independent anti-
gens).265,266 Conjugation of a polysaccharide to a carrier protein
confers T cell–dependent properties to the polysaccharide. The
immunologic hallmark of polysaccharide-protein conjugate
vaccines is their ability to induce memory B cells capable of
responding with a booster antibody response to a subsequent
exposure to the polysaccharide. In practice, measuring serum
antibody responses to an unconjugated polysaccharide vaccine
“probe” assesses induction of immunologic memory by the con-
jugate vaccine. Typically a reduced dose of the polysaccharide
vaccine is used (as low as 1 g), 267 based on the theory that a
suboptimal immunogenic dose may be more useful for distin-
guishing memory antibody responses from antibody responses
in subjects not primed by the conjugate vaccine.
The ability of a conjugate vaccine to prime for memory anti-
body responses on subsequently encountering meningococci was
thought to be an important second mechanism of protection from
developing disease in persons whose serum antibody concentra-
tions had declined to subprotective levels.264 Analysis of UK data,
however, indicates that persistence of functional antibody is the
determinant of MenC conjugate vaccine effectiveness and casts
doubt on the ability of memory antibody responses alone to pro-
tect against meningococcal group C disease.268 The inability of
memory antibody responses alone to confer protection is not sur-
prising since earlier studies had shown that a minimum of 5 to
7 days was required to elicit a serum group C memory antibody
response267 and most meningococcal disease occurs within a few
days after exposure to the organism.89,269,270
The need to assess induction of B-cell memory, therefore,
has limited value from the perspective of a potential mecha-
nism of protection. As well, the administration of “test” doses
of unconjugated polysaccharide vaccine to assess memory
antibody responses may be harmful by depleting antigen-
specific memory B cells and impairing subsequent antibody
responses.271 The impaired serum antibody responses taken
together with immunologic data showing loss of meningo-
coccal group C immunologic memory raise important ques-
tions about the safety of using unconjugated polysaccharide
vaccines for primary immunization of children younger than
2 years and as probes for assessing priming by conjugate vac-
cines during clinical trials.266 For these reasons, national
regulatory authorities are reassessing requirements for dem-
onstration of memory antibody responses to a polysaccharide
challenge dose for licensure of new conjugate vaccines.
Induction of memory B cells is accompanied by avidity mat-
uration, which reflects the effects of T cells on B cell somatic
mutation of antibody variable region genes. Memory B-cell
clones with mutated antibody combining site result in an
increase in antibody avidity compared with that of antibodies
elicited by the primary immunization (so-called avidity matura-
tion). An example of avidity maturation of group C and W135
anticapsular antibodies in vaccinated children is shown in
Table 21-1. During the 6 months of follow-up, there was a sig-
nificant increase in avidity of the anticapsular antibodies elicited
by a dose of the meningococcal conjugate vaccine but not by a
dose of the unconjugated polysaccharide vaccine. Thus, instead
of administering a polysaccharide probe to assess induction of
antigen-specific B-cell memory, it may be possible to use avidity
as a surrogate.272–274 Other surrogates include demonstration of
booster antibody responses to a second injection of the conjugate
vaccine (as demonstrated for groups C, Y, and W135 after quad-
rivalent meningococcal conjugate vaccination of adolescents; see
later discussion)275 and detection of antigen-specific peripheral
blood memory B-cell responses after vaccination.276,277 None of
these approaches risks induction of antibody hyporesponsive-
ness or loss of immunologic memory that can accompany group
C and, possibly, Y and W135 polysaccharide challenges
History of meningococcal vaccine development
Vaccines offering protection against meningococcal disease
were first introduced into routine use in the military more than
40 years ago; even today, however, no formulation offers com-
prehensive protection against strains from all pathogenic cap-
sular groups. The development of meningococcal vaccines has
been hampered by the biology of and its unique
relationship with the human host. The organism has evolved
an array of sophisticated mechanisms to evade host defenses,
which confound attempts to produce a truly comprehensive vac-
cine. These mechanisms include poorly immunogenic polysac-
charide capsules and lipopolysaccharides, some of which mimic
glycosylated structures on human tissue; lack of conservation
of surface-exposed proteins, which prevent use of such antigens
to elicit broadly cross-reactive antibodies; genetic exchange of
DNA permitting almost endless possibilities for reassortment
of nucleic acid sequences, including those encoding its prin-
cipal antigens; and the ability to delete genes encoding key
Antibody Avidity in Relation to Age and Vaccine*
±
ND, not done.
*Vaccines: MPSV4, a quadrivalent meningococcal polysaccharide vaccine;
and MCV4-DT, a quadrivalent meningococcal conjugate vaccine with each
of the four polysaccharides independently conjugated with diphtheria toxoid
(See Table 21-3). At 1 and 6 months post vaccination, the mean antibody
avidity is higher in the MCV4-DT group than the MPSV4 group ( = .002 for
capsular group C; .02 for capsular group W135). The increase in mean
avidity between 1 and 6 months is statistically significant for the MCV4-DT
group ( .001 for group C, and .05 for W135) but not for the MPSV4
group. (Data from Granoff DM, Morgan A, Welsch JA. Immunogenicity of
an investigational quadrivalent –diphtheria toxoid
conjugate vaccine in 2-year-old children. Vaccine 23:4307-4314, 2005.)

antigens.35,278,279 Only relatively recently, with the detailed anal-
yses of the meningococcal genome, has the real extent of these
mechanisms become fully appreciated.50,52,280,281
Early attempts to develop meningococcal vaccines used killed
whole bacterial cells.282–284 Between 1900 and 1940, several tri-
als were conducted, but the studies were poorly controlled, and,
in most cases, it was impossible to tell whether the vaccines
conferred protective immunity.285 Pursuit of a whole-cell vac-
cine was curtailed by the excessive reactogenicity of such prepa-
rations. Following the successful development of tetanus and
diphtheria toxoid vaccines in the 1930s, the protective poten-
tial of crude meningococcal culture filtrates containing inac-
tivated exotoxin was explored.286,287 Perhaps not surprisingly,
these preparations proved to be immunogenic, but they would
have certainly been contaminated with capsular polysaccharide
antigens, outer membranes, and endotoxin. Enthusiasm for
development of meningococcal vaccines subsequently waned
in the face of overwhelming optimism surrounding early suc-
cesses with antibiotic treatment and prevention of secondary
cases, most notably with sulfonamides. By the early 1960s,
sulfonamide-resistant isolates of had become
widespread and represented an important problem among mil-
itary recruits during the Vietnam War era, which promoted
renewed interest in meningococcal vaccine development.185,288
During the early 1940s, Scherp and Rake289 demonstrated
that serum from horses immunized with group-specific capsu-
lar polysaccharides protected mice against lethal challenge with
but purified preparations of capsular polysac-
charide had failed to elicit antibody responses in humans.290,291
In retrospect, the poor immunogenicity can be attributed to
the relatively low molecular size of the polysaccharide formula-
tion tested, as subsequent studies showed that polysaccharide
antigens of high molecular weight induced sufficient antibody
responses in humans.292,293 By the end of the 1960s, Gotschlich
and colleagues,294,295 working at the Walter Reed Army Institute,
had developed an alternative approach for the purification of
high-molecular-weight meningococcal polysaccharides that was
safe and immunogenic. As described subsequently, the principal
limitation of polysaccharide vaccines is that they do not induce
T cell–dependent immunity. The lack of T-cell immunity has
profound implications for the poor effectiveness of these vac-
cines in young children and the inability of polysaccharide vac-
cines to elicit long-term immunologic memory.267,296,297 The
chemical conjugation of polysaccharides to protein carrier mol-
ecules confers a T cell–dependent immune response.298 The
success of this approach was first demonstrated in humans with
type b vaccines in the 1980s.299–301 Subsequently,
similar conjugate vaccines were investigated for the meningo-
coccus for group A (N-acetyl mannosamine-1-phosphate) and
group C (alpha 2-9 N-acetyl neuraminic acid [NANA]) polysac-
charides,302 group C polysaccharide,303–306 and, more recently,
groups A, C, W135 (copolymer of NANA with galactose), and
Y (copolymer of NANA with glucose).236,245,258,307–309 These vac-
cines elicit high serum bactericidal antibody titers and boosta-
ble immune responses. As described subsequently, monovalent
group C conjugate vaccines were introduced with great success
in the United Kingdom in late 1999 and are now widely used in
other European countries, Canada, and Australia.
As of May 2012, two quadrivalent A, C, W135, and Y menin-
gococcal polysaccharide-protein conjugate vaccines were
licensed in the United States and recommended for routine
use in adolescents and young adults.163 Vaccine manufacturers
are developing other multivalent meningococcal conjugate
vaccines that contain capsular groups C and Y or C, Y, and
type b, (Hib) with a variety of carrier proteins
and that are intended for use in infants and toddlers. This
strategy, polysaccharide conjugated to protein carrier, is not
applicable for capsular group B (see later discussion), but a
number of promising group B recombinant protein and outer
Meningococcal vaccines 21 399
membrane vesicle vaccines are being investigated (reviewed
by Granoff310 and Sadarangani et al;311 see also later section
on group B vaccines), which makes the prospect for a compre-
hensive vaccine more hopeful.
The UK experience during the last decade revealed principles
that will need to be considered to achieve effective strategies for
conferring protection against meningococcal disease across the
entire lifespan. Currently available conjugate vaccines induce
immunologic memory and protective serum antibody titers in
all age groups. The presence of memory by itself, however, is
insufficient to provide protection based on evidence from UK
experiences with Hib and meningococcal group C conjugate
vaccine schedules that elicited long-term immunologic memory
but short-term protective levels of serum antibody.265,312 Thus,
protective concentrations of serum antibody seem to be neces-
sary at the time of exposure to the organism to prevent invasive
disease. In the absence of booster doses, the durability of protec-
tion is limited in young children and adolescents by relatively
rapid decline of serum titers of protective antibody.265
Polysaccharide meningococcal vaccines:
composition, immunogenicity, limitations,
efficacy, and safety
Meningococcal polysaccharide vaccines (Table 21-2) currently
licensed include bivalent (groups A and C) and quadrivalent
(groups A, C, W135, and Y) formulations, although only the
quadrivalent vaccines are readily available. In partnership with
the WHO, a trivalent vaccine (group A, C, and W135) manu-
factured by GlaxoSmithKline Biologicals is used in response to
W135 epidemics in sub-Saharan Africa. These vaccines will be
referred to as MPSV2, MPSV4, and MPSV3, respectively.
Immunogenicity of meningococcal polysaccharide
vaccine
Unconjugated capsular polysaccharide vaccines elicit serum
antibody responses largely in the absence of T-cell help (so-
called, T cell–independent antigens). These vaccines are immu-
nogenic in older children and adults but typically are poorly
immunogenic in infants, and repeated injections do not elicit
antibody boosting at any age. In adults, the serum antibody
responses reflect natural priming and existence of anticapsu-
lar B-cell populations in the memory state as evidenced by IgG
responses and hypermutation of the variable region genes, which
are features indicative of a secondary antibody response.313–315
Most adults develop protective serum bactericidal antibody
levels against strains from all four capsular groups contained
in the vaccine within 7 to 10 days after vaccination. Antibody
titers peak at approximately 2 to 4 weeks and then decline over
2 years to approximately 5% of their respective peak values and
remain above baseline preimmunization levels for 10 years.316
Immunized infants and young children frequently have low or
undetectable SBA despite relatively high concentrations of anti-
capsular antibody.237,243,317–319 These age-related differences in
antibody functional activity likely reflect differences in antibody
avidity and repertoire. The disparities underscore the difficul-
ties in establishing a serum anticapsular antibody concentra-
tion that correlates with protection from disease.320 Therefore,
the results of meningococcal bactericidal assays are specific
for predicting protection after vaccination263,321 but likely lack
sensitivity.219
In the 1970s, Gold and colleagues322 first noted that an
injection of group A meningococcal polysaccharide vaccine
given at 3 months of age primed infants for booster antibody
responses to a second injection given at 12 months. In contrast,
an injection of group C polysaccharide vaccine resulted in lower

495–497 495,497 485 498
The Hoshino, Torrii, Miyahara, and NK-M46 strains
used in Japan and South Korea and the Leningrad-3499,500 and Sofia-
6501 strains used in Russia and Eastern Europe500,501 have also been
linked to aseptic meningitis. For the Torii and Hoshino strains, the
frequency of aseptic meningitis ranges from 1 in 500 to 1 in 1,000
doses.495,497 There are insufficient data to identify the frequency of
aseptic meningitis attributable to the other vaccine strains.
In conclusion, it seems that many, though not all, attenuated
mumps vaccine strains cause aseptic meningitis, but the rates of
this complication vary according to the vaccine strain, the manu-
facturer, the index of clinical suspicion, and the intensity of surveil-
lance. In evaluating the importance of this complication, one has
to consider the incidence of clinical meningitis that would occur in
natural mumps, conservatively estimated between 1% and 10%.502
Assuming a rate of vaccine-associated meningitis of 1 in 10,000
and an incidence of mumps greater than 1% per year, vaccination
is clearly better than not vaccinating, a conclusion also reached
after evaluation of safety and effectiveness data from 64 differ-
ent MMR vaccine studies involving over 14 million children.410b
Different countries have made different judgments regarding the
use of Jeryl Lynn or other vaccine strains for the prevention of
mumps,503,504 and the WHO considers most to be acceptable.505 In
any case, the risk-benefit ratio is in favor of vaccination despite the
occasional case of aseptic meningitis.506 Notably, sequelae to post-
vaccine meningitis have been rare or absent.
Indications for vaccination
In the US, the ACIP recommends two doses of M-M-R II vac-
cine for all children and for certain high-risk groups of ado-
lescents and adults, including international travelers, people
attending universities and other higher educational institutions,
and people who work at health care facilities.441 ProQuad may be
given instead of M-M-R II, with some limitations (see
and ).
In general, people can be considered immune to mumps who
have documentation of vaccination with live mumps virus vac-
cine on or after their first birthday, have laboratory evidence of
mumps immunity, have documentation of physician-diagnosed
mumps, or were born before 1957 (although during mumps out-
breaks, vaccination with mumps-containing vaccine should be
considered for people born before 1957).
On June 24, 2009, the ACIP revised its recommendations
for evidence of mumps immunity for health care personnel to
include documented administration of two doses of vaccine con-
taining live mumps vaccine or laboratory evidence of immunity
or laboratory confirmation of disease. For unvaccinated person-
nel born before 1957 who lack laboratory evidence of mumps
immunity or laboratory confirmation of mumps, health care
facilities should consider vaccinating with two doses of mumps-
containing vaccine at the appropriate interval.507
Laboratory testing for mumps susceptibility before vaccina-
tion is unnecessary.441 There is no increased risk of adverse reac-
tions to vaccination of people who are already immune.
Precautions and contraindications
Most mumps vaccines are produced in chick embryo cell culture
and contain small quantities of ovalbumin; however, the risk for
serious allergic reactions in people allergic to eggs is extremely
low,395,508–510 and, therefore, egg allergy is not a contraindication
for vaccination. Although some physicians recommend cuta-
neous testing as a precaution in egg-allergic children,511,512 sev-
eral studies have found skin testing not helpful in predicting an
adverse reaction,508–510 and it is not required.441
Mumps vaccine 22 441
Caution should be exercised when administering M-M-R II
or ProQuad to people who have a history of an anaphylactic reac-
tion to gelatin or gelatin-containing products. Skin testing for
sensitivity to gelatin could be considered, but no specific pro-
tocols for this purpose have been published.441 Likewise, peo-
ple who have a history of anaphylactic reactions to neomycin
should not receive the vaccine; a history of contact dermatitis
to neomycin is not a contraindication to vaccination. Mumps-
containing vaccines used in the US and most other countries do
not contain penicillin.
Mumps-containing vaccine should be given at least 2 weeks
before the administration of immune globulin or deferred until
3 months after such administration because passively acquired
antibody can interfere with response to the vaccine. Interestingly,
coadministration of immune globulin with vaccine does not
seem to interfere with vaccine responses.321
Mumps-containing vaccine should not be given to pregnant
women because of the theoretical risk of fetal damage. Likewise,
vaccinated women should avoid pregnancy for 1 month after
vaccination.395
People with severe febrile illnesses should, in general, not be
vaccinated until they have recovered. Vaccination should not be
postponed because of minor illness.
The measles virus component in trivalent (MMR) and quad-
rivalent (MMRV) mumps-containing vaccines may temporarily
suppress tuberculin reactivity,395 and, therefore, tuberculin test-
ing should be done before vaccination or 4 to 6 weeks afterward.
Tuberculin testing may be done on the same day as vaccination,
but in such cases, the Mantoux test is recommended over the
multiple puncture test because the latter, if results are positive,
will require confirmation, which will have to be postponed 4 to
6 weeks.
Untreated active tuberculosis and a history of thrombocyto-
penia/thrombocytopenic purpura are contraindications for MMR
vaccination. The benefits of immunity to measles, mumps, and
rubella are thought to outweigh the risk of recurrence or exac-
erbation of thrombocytopenia after vaccination. However, if a
prior episode of thrombocytopenia occurred shortly after vac-
cination, avoidance of a subsequent dose of MMR or MMRV
should be considered.395
MMR or MMRV vaccine should not be given to people with
acquired immunodeficiency or suppressed immunity (eg, leu-
kemia, lymphoma, generalized malignancy, or therapy with
corticosteroids, alkylating drugs, antimetabolites, or radia-
tion). An exception is children infected with HIV. MMR vac-
cination is recommended for all HIV-infected people who do
not have evidence of severe immunosuppression and for whom
measles vaccination would otherwise be indicated. Because the
immunologic response to vaccines may decrease as HIV disease
progresses, HIV-infected infants without severe immunosup-
pression should be given MMR at age 12 months and a sec-
ond dose as soon as 1 month later.441 Data are not available on
safety, immunogenicity, or efficacy of ProQuad vaccine in HIV-
infected children, and, thus, this vaccine should not be used in
this population.395
Patients with leukemia in remission who have not received
chemotherapy in at least 3 months may receive mumps-
containing vaccine, as may people who have received short-term
(ie, 2 weeks’ duration) steroid therapy.441 There are limited
data on the safety and efficacy of M-M-R II or ProQuad vac-
cine in transplant recipients receiving immunosuppressive ther-
apy. Because the diseases prevented by M-M-R II or ProQuad
do not result in significant morbidity (mumps, varicella) or
are of a low contraction risk owing to population immunity
(measles, rubella), vaccination should be avoided. However, dur-
ing an outbreak, immunization may be advisable.513 M-M-R
II or ProQuad vaccination is recommended for persons com-
pleting immunosuppressive therapy who are determined to be
immunocompetent.395
 Pertussis vaccines 23 461

Additional Characteristics of Selected Acellular Pertussis Vaccines*

*Compositions may differ in various markets; local suppliers should be consulted as necessary. Aluminum content is per dose. Some products are no longer
available, but are included because knowledge of their composition is important to understanding the results of the efficacy trials.
†
Trade names (most common trade name, if multiple names exist) except as follows: HCPDT is the ‘hybrid’ formulation of evaluated in the 1993 Stockholm
trial (otherwise, used only in combinations); JNIH-6 and JNIH-7 were the acellular vaccines used in the 1986 Swedish trial; SKB-2 was an experimental 2-component
DTaP evaluated in the 1992 Stockholm trial.
‡
As aluminum phosphate.
§
As aluminum hydroxide.
As aluminum potassium sulfate.
¶
Pertussis components are formaldehyde-stabilized. Aluminum content was 0.35 mg in MAPT.
**PT component detoxified with both formaldehyde and glutaraldehyde. In MAPT, aluminum content was 0.50 mg.
††
A mixture of aluminum hydroxide and aluminum phosphate.
n/a, information not available; HCPDT, hybrid component pertussis-diphtheria-tetanus vaccine; MAPT, Multicenter Acellular Pertussis Trial; PBS, phosphate-buffered
saline; PS, polysorbate-80.

Pertussis Immunization Schedules Recommended by Selected National Authorities, as of September 2006*

Pertussis Immunization Schedules Recommended by Selected National Authorities, as of September 2006—cont'd

*Compiled from multiple sources. In many countries, products based on whole-cell pertussis vaccine are used in the national or public programs (shown in this
table), whereas products based on acellular vaccines are often used in private practice. In such countries, recommended schedules for private practice typically are
based on US or Western European schedules.
As of September 2006, the World Health organization (WHO) maintained an interactive resource that displayed recent (but not necessarily current) immunization
schedules by selected country at www.who.int/immunization_monitoring/en/globalsummary/scheduleselect.cfm. The same information, plus demographic data and
the incidence of vaccine-preventable diseases, was available at www-nt.who.int/vaccines/globalsummary/Immunization/CountryProfileSelect.cfm.
DTaP, diphtheria and tetanus toxoids and acellular pertussis; DTwP, diphtheria and tetanus toxoids and whole-cell pertussis; EPI, WHO's Expanded Programme on
Immunization; HB, hepatitis B virus; Hib, type b; IPV, inactivated poliovirus; Tdap, tetanus, diphtheria, and pertussis for adolescents and adults.

Current Major Vaccine Manufacturers and Their Merged,
Acquired, or Component Companies
*Acquired by Johnson & Johnson in 2011
Daptacel, Sanofi Pasteur
Daptacel is the US trade name for a five-component acellular per-
tussis vaccine manufactured by Sanofi Pasteur in Canada. The
acellular pertussis vaccine component contains approximately
10 g PT, 5 g FHA, 3 g PRN, and 5 g of combined FIM-2 and
FIM-3. Daptacel is supplied in thimerosal-free, single-dose vials.
Daptacel is marketed under various names (Tripacel, Triacel,
Pertacel, or Monocel) in many different countries, often in com-
bination with other components (see Chapter 40).
Infanrix, GlaxoSmithKline
Infanrix consists of acellular pertussis components manufac-
tured by GlaxoSmithKline Biologicals (Rixensart, Belgium) plus
diphtheria and tetanus toxoids manufactured by Chiron Behring
GmbH & Co (Marburg, Germany). The acellular pertussis vac-
cine component contains not less than 25 g PT, 25 g FHA, and
8 g PRN. The vaccine is supplied in single-dose, thimerosal-
free vials or prefilled syringes. The vaccine is currently licensed
in the United States as a five-dose series for children 6 weeks
to 7 years of age. Infanrix is marketed in many countries, often
in combination with other components (see Chapter 40). In the
United States, the combination of Infanrix, hepatitis B vaccine,
and inactivated poliovirus vaccine is marketed as Pediarix (see
Chapter 40). Pediarix is supplied in single-dose, thimerosal-free
vials or prefilled syringes and is licensed for a three-dose primary
series in infants born to hepatitis B surface antigen–negative
mothers.
Triavax, Sanofi Pasteur
This two-component DTaP vaccine is manufactured by Sanofi
Pasteur in France and contains approximately 25 g PT and 25
g FHA. The vaccine is not presently available as a stand-alone
Pertussis vaccines 23 463
DTaP but is marketed in many countries in combination with
other components such as Hib and IPV (see Chapter 40).
Tripedia, Sanofi Pasteur
Tripedia combines acellular pertussis vaccine manufactured
by Biken and Tanabe Corporation (Osaka, Japan) with diph-
theria and tetanus toxoids manufactured by Sanofi Pasteur.
The acellular pertussis vaccine component contains approxi-
mately 23.4 g PT and 23.4 g FHA. The vaccine is supplied
in preservative-free, single-dose vials. Under the trade name
TriHIBit, Tripedia is marketed bundled with the conjugate Hib
vaccine ActHIB. When used to reconstitute the ActHIB, the
resulting combination may be administered in a single injec-
tion as the fourth dose of DTaP and the booster dose of conju-
gate Hib vaccine (see Chapters 13 and 40). The manufacturer
has announced the discontinuation of this product once exist-
ing stocks are exhausted, likely in mid 2013.
The vaccines discussed next are intended primarily for use in
adolescents and adults, but in some jurisdictions they are also
licensed for use at 3 or 4 years of age for preschool boosting of
previously primed children.
Adacel, Sanofi Pasteur
Adacel (marketed in Europe as Covaxis) is an adolescent-adult
formulation of Sanofi Pasteur's five-component acellular per-
tussis vaccine. It is identical to Daptacel (Tripacel) except that
the quantities of PT (2.5 g) and diphtheria toxoids (2 Lf [limit
of flocculation]) are reduced. Adacel is licensed for use in the
United States in persons 11 to 64 years of age and is supplied in
thimerosal-free, single-dose vials. Combined with IPV, the prod-
uct is marketed as Repevax.
Boostrix, GlaxoSmithKline
Boostrix is an adolescent-adult formulation of GlaxoSmithKline's
three-component acellular pertussis vaccine. It differs from
Infanrix by containing reduced quantities of PT (8 g), FHA (8 g),
PRN (2.5 g), diphtheria (2.5 Lf), and tetanus (5 Lf) toxoids, and
adjuvant ( 0.39 mg aluminum). This Tdap vaccine is licensed
in Australia, Europe, and elsewhere. In the United States, it is
licensed for use in persons 10 years of age and older. Outside the
United States, it is also available combined with IPV.
Other acellular pertussis preparations
Several additional acellular pertussis vaccines were developed
during the 1980s and 1990s but were not licensed, and a num-
ber of vaccines that once were licensed have been removed from
the market. Interested readers are referred to the third edition
of this text313 for further details. Local manufacturers in India
and China have begun to develop acellular pertussis vaccines
for their local markets, and it is likely that these vaccines (prob-
ably as components of larger combinations) will be offered for
export (see Chapter 40).
A failure to adhere to the recommended schedule, causing a
delay between doses, should not interfere with the final immu-
nity achieved by any of the pertussis vaccines. There is no need
to start the series over again, regardless of the time between
doses. Partial doses should not be given; there is no evidence
that doing so reduces the frequency of serious adverse events,
and there is the risk that efficacy might be impaired.
Few data are available demonstrating the effects of a change dur-
ing the scheduled immunization series from one brand of DTaP
to another. Recommendations from ACIP and the American
Academy of Pediatrics Committee on Infectious Diseases (Red
Book Committee) therefore have indicated a preference to use
464 SECTION TWO
the same brand of DTaP vaccine for all doses of the vaccina-
tion series for a given child. However, the recommendations
also state that, if the vaccine provider does not know, or does
not have available, the brand of DTaP previously administered
to a child, any of the licensed DTaP vaccines may be used to
complete the vaccination series. Moreover, the one published
study evaluating a change from one DTaP (Tripedia) to another
(Infanrix) during the primary series found no adverse effect on
safety or immunogenicity.314 As a practical matter, free inter-
change of the various acellular pertussis vaccines has become
commonplace, fueled by periodic changes in public distribution
systems, the vagaries of patient migration and shifting provider
purchases, and intermittent shortages of one vaccine or another.
Simultaneous administration with other vaccines
DTP and DTaP are routinely administered concomitantly with
polio vaccine, conjugate Hib vaccine, conjugate pneumococcal
vaccine, HBV vaccine, measles-mumps-rubella vaccine, and
varicella vaccine, as appropriate to the age and previous vac-
cination status of the child, using separate syringes for injec-
tions at separate sites. Most studies have found that the adverse
reactions after multiple simultaneous vaccinations are only
slightly greater than would be expected from the most reacto-
genic vaccine alone. Further reductions in adverse reactions can
often be obtained by using vaccines that combine multiple anti-
gens (see Chapter 40), thereby reducing the number of injec-
tions given. Although reduced immunogenicity to the pertussis
antigens has occasionally been seen when given in combina-
tion or association with one or more of the other vaccines, no
data exist to suggest that concomitant administration of any
of these other vaccines decreases the efficacy of pertussis vac-
cine. Because whole-cell pertussis vaccine has adjuvant proper-
ties, replacement of the whole-cell component with acellular
vaccine has resulted in reports of diminution of the immune
responses to some of the simultaneously administered vaccines
(see Chapter 40), without evidence of impairment of efficacy.
Results of vaccination: whole-cell
pertussis vaccines
Although some observers have challenged whole-cell pertus-
sis vaccine as being ineffective,315,316 most authorities agree
that the widespread use of the whole-cell vaccine has had enor-
mous benefits.213,317,318 These benefits include the rapid decline
in morbidity and mortality from pertussis after implementa-
tion of whole-cell vaccine programs; the recurrence of disease
in countries where whole-cell immunization was discontinued,
acceptance rates declined, or vaccines became ineffective; the
inverse correlation of the pertussis attack rates and the number
of immunized children; and the lower attack rates in previously
immunized children than in unimmunized children under both
endemic and epidemic conditions.
Immune responses to whole-cell vaccines
Although many of the world's children are immunized with
locally produced whole-cell pertussis vaccines, few data exist to
compare their safety, immunogenicity, and efficacy. Most author-
ities had presumed that all whole-cell vaccines were very simi-
lar and that one was as immunogenic and effective as another.
However, when significant differences in immune responses
were noted with whole-cell pertussis products produced by differ-
ent manufacturers in the United States and Canada, it became
apparent that all whole-cell vaccines were not the same.256,319,320
The European acellular pertussis vaccine efficacy trials that also
used whole-cell vaccines produced in several different countries
demonstrated substantial differences in efficacy among the vari-
ous whole-cell products. Thus it appears possible that some of
the whole-cell pertussis vaccines produced in various countries
in the world might be less effective than others, and this would
have an impact on the global burden of pertussis disease.
Another concern with whole-cell pertussis vaccines is
the negative impact of high maternal antibody levels on
infant immune responses. It has been demonstrated that the
magnitude of the primary antibody response to whole-cell
vaccine in infants depends on the preimmunization (trans-
placental) levels of antibody to PT, with higher circulating
levels of maternally derived antibody being associated with
significantly lower levels of postimmunization antibody.283 In
contrast, PT antibody responses to an acellular vaccine con-
taining 12.5 g each of PT and FHA were superior to those of
the whole-cell vaccine and were not affected by prevaccina-
tion antibody levels.283 Studies in developing countries have
not evaluated the impact of maternal antibody on immune
responses to locally produced whole-cell vaccines in young
children.
Controlled clinical trials of whole-cell vaccines
It is well documented by controlled clinical trials that whole-
cell pertussis vaccine provides protection against clinical
whooping cough in the majority of immunized people. The
first convincing evidence was provided by studies in the Faroe
Islands during two epidemics.299 These studies showed that
pertussis vaccine not only protected against disease but also
ameliorated the severity of disease in immunized persons who
contracted the illness. Although studies with early vaccines
produced inconsistent results,14,213,318,321 clinical trials after
standardization of the vaccines by the mouse protection test317
demonstrated consistent efficacy.317,322,323
In more recent years, a number of field trials of acellular
pertussis vaccines (see “Efficacy trials, 1992 to 1997”, later)
have incorporated whole-cell pertussis vaccines as controls and
have provided some of the best data ever obtained about the
efficacy of the conventional whole-cell vaccines. As mentioned
previously, these studies showed that whole-cell vaccines vary
substantially in efficacy. The following estimates reflect effi-
cacy after three doses and, to the extent possible, consistent
case definitions; however, none of these studies was fully
blinded and randomized, and the estimates may be generous.
The Mainz324 and Munich325 studies produced efficacy esti-
mates of 98% and 96%, respectively, for the German-produced
Behringwerke vaccine; the US-made Wyeth-Lederle whole-cell
vaccine was reported to be 83% efficacious in the Erlangen
trial;258,326 and the Senegal trial reported the French-made
Sanofi Pasteur whole-cell vaccine to be 96% efficacious.327 In
each of these trials, the whole-cell vaccine was more efficacious
than the acellular product.
In marked contrast, the US-made Connaught whole-cell
vaccine had very low rates of efficacy after three doses: 48% in
Sweden and 36% in Italy.259,260 In the Swedish study, efficacy
was nearly 74% for the first 6 months after the third dose of vac-
cine but declined rapidly after that time.259 A British national
survey of reported whooping cough from 1989 to 1990 deter-
mined that the efficacy rates of the Wellcome whole-cell vac-
cine, administered at 3, 5, and 10 months of age, were 87% and
93% during epidemic and nonepidemic periods, respectively.328
Efficacy declined with age but remained high until the age of
8 years. A repeat survey was conducted in 1994 to determine
whether efficacy had been altered by the change to an acceler-
ated schedule of immunization at 2, 3, and 4 months of age
with no subsequent booster. Efficacy was not altered and was
94% overall for those subjects between 6 months and 5 years
of age.329 This accelerated schedule was also associated with a
reduced rate of adverse reactions.330

Other evidence of effectiveness of whole-cell
vaccines
There is no question that the widespread use of whole-cell per-
tussis vaccine in developed countries has resulted in a remark-
able decline in reported pertussis.213,240,241,331 However, it is also
clear that mortality rates from pertussis were declining in some
countries even before the advent of the vaccine.213,317,332,333 The
latter reductions probably reflect improved social and economic
conditions, better nutrition, and declines in concomitant infec-
tions that may have enhanced pertussis mortality.
Strong evidence of the benefits of pertussis vaccine was provided
by unintended experiments that occurred in three developed
countries when vaccine use was curtailed or abandoned. Japan
initiated widespread immunization against pertussis in 1950,
and over the ensuing years the numbers of reported cases and the
numbers of deaths declined remarkably.241,334 However, begin-
ning in 1975, adverse events temporally associated with admin-
istration of whole-cell pertussis vaccine to young children led to
a near-boycott of the vaccine, and epidemic pertussis recurred.
Hundreds of children died of pertussis during this period.241 A
similar experience occurred in England and Wales, in the context
of negative publicity surrounding adverse events associated with
vaccine. Rates of vaccine acceptance fell from approximately
75% to nearly 25% during the mid 1970s, major epidemics of
pertussis ensued, and numerous children died.240,335 In Sweden,
the administration of pertussis vaccine was suspended in 1979
when pertussis outbreaks occurred despite high vaccination cov-
erage, suggesting poor efficacy of the existing vaccine. The inci-
dence of whooping cough then increased more than fourfold
from 1980 to 1985, with several major outbreaks in subsequent
years.336,337 With the advent of routine pertussis immunization
using acellular vaccines after completion of the vaccine efficacy
trials in Sweden, pertussis rates again declined.338 However,
recent reports suggest that pertussis rates have again increased
and booster doses are needed.339,340 A review of the French expe-
rience with whole-cell pertussis vaccine over a period of 30 years
showed persistent high efficacy of the product.341
Further evidence of the efficacy of whole-cell pertussis vaccine
is provided by the observation that the reported incidence of
pertussis disease varies inversely with vaccine acceptance rates.
A study in England and Wales found that communities with
low pertussis vaccine acceptance rates (30%) had a 59% higher
reported incidence of pertussis among children than did areas
with high (50%) acceptance rates; areas with intermediate
acceptance rates had intermediate pertussis rates.342 These find-
ings were not explained by differences among the communities
such as crowding and social class; indeed, after adjustment for
these two social indicators, the inverse correlation with immu-
nization status was even stronger. In the United States, where
infant immunization for pertussis has been routine since the
late 1940s, national surveillance demonstrated a greater than
95% reduction in pertussis; surveillance data from 1992 to
1994 found an overall whole-cell pertussis vaccine efficacy of
64% after three doses and of 82% after four or more doses.261
Additional evidence for the efficacy of whole-cell pertussis vac-
cine is provided by community outbreaks of pertussis in which
Pertussis vaccines 23 465
the attack rates of the disease in immunized children were com-
pared with the rates in those who were incompletely or never
immunized.240,343 Although the methods of ascertainment and
analysis vary, most studies indicate that the efficacy of three or
more doses of pertussis vaccine in protecting children against
clinical disease during outbreaks is 80% to 90%, with incom-
plete immunization offering partial protection.343 In children
who contracted pertussis in spite of immunization, the disease
was milder and complications were far less frequent, despite
the fact that younger infants and children are more likely to
have received inadequate or no immunization.344,345 Studies
conducted in Senegal also documented that breakthrough cases
of pertussis among children vaccinated with whole-cell vaccine
were of lesser severity and reduced infectivity than those seen in
unimmunized children.346,347
Herd immunity after immunization
Because whole-cell pertussis vaccine is not 100% effective, it
might be considered curious that morbidity and mortality
from clinical pertussis have been negligible in countries with
widespread immunization programs. This is a particularly
interesting finding because surveillance statistics and studies
demonstrating pertussis to be a common cause of protracted
cough illness in adolescents and adults suggest that
remains ubiquitous in these countries.188,190,204,272–274 Attempts
to model mathematically the decline in pertussis incidence
attributable to widespread immunization with vaccines of 85%
efficacy have underestimated the rates of decline of the dis-
ease.348 Although other factors, such as social and economic
changes, very likely play a role, the most probable explanation is
herd immunity, a complex phenomenon that varies among dif-
ferent infectious diseases and is difficult to measure with preci-
sion349 (see Chapter 71). Herd immunity undoubtedly explains
the cycles of outbreaks of pertussis every 3 or 4 years; after an
outbreak, several years are required for the proportion of suscep-
tible persons to increase to a level that facilitates a new wave of
spread in a population. Finally, studies describing higher rates
of pertussis disease in immunized children living in communi-
ties with large numbers of unimmunized children provide addi-
tional evidence for the role of herd immunity.350
Duration of immunity after whole-cell vaccines
A number of studies have evaluated the duration of protection
after immunization with whole-cell vaccine. Those that provide
the longest period of evaluation indicate that protection declines
by 50% over a period of 6 to 12 years.351–353 These data are consis-
tent with the incidence and serosurvey data cited previously that
suggest an increase in rates of pertussis among 13- to 17-year-
olds, representing an interval of 7 to 12 years since last vaccina-
tion.190 It is likely that the duration of protection is influenced
by the vaccine used, the number of doses given, the vaccination
schedule, and the level of circulating capable of stimu-
lating an anamnestic response in previously immunized persons.
Adverse events with whole-cell vaccines
Despite their clear benefits in reducing the substantial mortality
and morbidity of pertussis, the whole-cell vaccines have long
been recognized as one of our most reactogenic vaccines. They
commonly cause reactions that are minor but burdensome, occa-
sionally cause reactions that are transient but frightening, and
uncommonly cause more serious, generally self-limited, adverse
effects. For some time, there was substantial suspicion that
whole-cell vaccines might be causally related to devastating out-
comes such as encephalopathy or sudden infant death syndrome
(SIDS), but several careful epidemiologic studies have largely dis-
pelled these concerns.354–356
466 SECTION TWO
Untoward events after pertussis immunization began to be
of increasing concern to the public and to physicians in the early
1970s, particularly in countries where widespread vaccination
had eliminated most disease. Vaccine-associated adverse events
loomed large in the eyes of young parents and physicians who
had never witnessed the morbidity and mortality of whooping
cough. Widespread publicity about the alleged dangers of per-
tussis vaccine, coupled with declining disease rates in some
countries and doubts about vaccine efficacy in others, resulted
in near-abandonment of pertussis vaccine in several countries.
Consequently, pertussis disease recurred.240,241,336 In the United
States, strong school-entry immunization laws enabled vac-
cination rates to be maintained despite widespread publicity
about these concerns. However, extensive litigation over alleged
personal injuries caused by the vaccine cost millions of dollars
and contributed to the cessation of pertussis vaccine production
by several manufacturers.
Establishing or disproving cause and effect, particularly for
events of major consequence, proved difficult. Although the
original allegations of causation were largely anecdotal and
based on the fallacious assumption that subsequences and con-
sequences were synonymous, they raised great concern and
stimulated the search for an improved vaccine. The relation-
ships between whole-cell pertussis vaccine and fatal or disabling
events were difficult to evaluate because of the rarity of such
events; because vaccine was administered to infants at an age
when disorders such as encephalopathy, infantile spasms, neu-
rologic conditions, and SIDS were most likely to occur; because
these disorders can arise from other causes; and because an
absolute negative can never be proven. Earlier estimates of the
rates of adverse events were not optimal because of lack of con-
sideration of background rates, ill-defined criteria, and uncer-
tainty of denominators. However, in the 1980s, more rigorous
epidemiologic or interventional studies greatly improved the
understanding of the incidence and spectrum of adverse events
after whole-cell pertussis vaccine. Many of these studies, partic-
ularly those related to serious untoward events, were evaluated
by a special committee of the Institute of Medicine (IOM) of the
US National Academy of Sciences.354–356 Although these evalua-
tions were reassuring to health care providers and to parents, it
was the development of less reactogenic acellular pertussis vac-
cines and their replacement of the whole-cell products that put
an end to the long debate in developed countries over adverse
events associated with whole-cell vaccines. Whole-cell pertus-
sis vaccines remain widely used in developing countries, where
concerns regarding adverse events do not seem to be issues.
In light of the continuing move toward acellular vaccines, we
restrict the discussion of adverse events associated with whole-
cell vaccines to a summary of the key studies and conclusions.
Readers interested in a more thorough review of these data are
referred to the second edition of this text.357
Common reactions
Minor local reactions, consisting of redness, swelling, and pain
at the site of injection, occur in about half of DTP recipients.
Reactions occur five times more frequently after DTP than after
DT.358–360 Similarly, minor systemic reactions such as fever, irri-
tability, and drowsiness are significantly more common after
DTP than after DT.358–360 About half of the children who receive
whole-cell pertussis vaccine experience some minor fever, with
less than 1% having an elevation in temperature to 40.5° C
(105° F).358–361
Participants in the Multicenter Acellular Pertussis Trial
(MAPT; more about this later) experienced somnolence at rates
of 62% for whole-cell recipients and 43% for acellular vaccine
recipients,361 suggesting that the somnolence was, at least in
part, an effect of the DTP vaccine. Some children were reported
to have an unusual, high-pitched cry. Somewhat more remark-
able was a period of excessive crying, which may last several
hours or longer after an injection. This incessant, inconsol-
able crying usually begins within 12 hours. Persistent crying
of 1 hour or more occurred in both the DTP and DT groups in
the study by Cody and coworkers360 but was at least four times
more common after DTP. Among those with persistent crying,
the cry was described as high-pitched or unusual in 3.5% of the
children.359–361
These common reactions vary somewhat in frequency and
severity among lots362 and manufacturers. The vaccine schedule
followed may also affect the incidence and severity of adverse
reactions. In 1990, the schedule for DTP vaccination in the
United Kingdom was changed from 3, 5, and 10 months of age
to 2, 3, and 4 months of age. The new schedule was associated
with a substantial reduction in postvaccination fever and red-
ness at the injection site.330 With the accelerated vaccination
schedule in the United Kingdom, reaction rates for the whole-
cell vaccine did not differ significantly from those of several
acellular vaccines.330
Reaction rates also vary with the number of prior DTP injec-
tions. In the Cody and colleagues study, local reactions increased
in frequency with successive doses, including the preschool
booster.359,360,362 The incidence of fever also increased with suc-
cessive doses through the 18-month booster, but was lower with
the preschool booster. On the other hand, persistent, inconsol-
able crying occurred most frequently with the initial dose and
less often thereafter. In the MAPT, the incidence and severity of
fever increased substantially with successive primary doses of
the reference whole-cell vaccine.361,363 Redness increased mod-
estly; frequency of pain, fussiness, anorexia, vomiting, and use
of antipyretics did not materially increase or decrease with suc-
cessive doses; and drowsiness decreased.361 In general, children
who experienced local or systemic reactions after pertussis vac-
cine have an enhanced likelihood of experiencing the same reac-
tion with a subsequent dose.364
Uncommon reactions
DTP is associated with febrile seizures (0.06% in the Cody and
coworkers trial),360 and seizures occur at increased rates after
DTP in children with personal or family histories of convul-
sions.360,364–371 However, simple febrile convulsions, although
distressing, are considered to be benign,372 with no evidence
that seizures after DTP induce epilepsy.373
Another worrisome but uncommon reaction to DTP is that of
a strange, shock-like state, termed a hypotonic-hyporesponsive
episode (HHE), that usually has its onset within 12 hours of
inoculation and may last for several hours but always resolves.374
Neither death nor adverse sequelae have been observed after
these episodes. The Cody and colleagues study360,362 detected an
incidence of HHE of 0.06%. The mechanism of this phenom-
enon is unknown, and HHE has been seen with other vaccines,
including acellular pertussis products, as described later.
Encephalopathy
The most serious reaction that has been attributed to whole-
cell pertussis vaccine is acute encephalopathy. The National
Childhood Encephalopathy Study, conducted in Great Britain
from 1976 to 1979, examined whether the frequency of DTP
vaccination in children with encephalopathy was greater than
expected. Based on 11 subjects who appeared to have residua
18 months after vaccination, it was estimated that acute enceph-
alopathy with permanent brain damage occurred at the widely
quoted rate of 1 per 310,000 doses, with a 95% CI of 1 in 54,000
to 5,310,000 doses. However, subsequent investigations cast
doubt on most of these 11 diagnoses.375–377 At 10-year follow-up,

the rates of death or other sequelae in these subjects were simi-
lar regardless of whether the onset of acute neurologic illness
was temporally associated with DTP vaccination.378
The results of the National Childhood Encephalopathy Study
have been subjected to extensive analysis, reanalysis, challenge,
and debate.377–383 None of this scrutiny has overturned the ini-
tial cautious interpretation that the data suggested but did not
prove a causal relationship between pertussis vaccine and per-
manent neurologic damage. Several US studies also failed to
show a relationship between vaccine and acute encephalopa-
thy leading to brain damage.367,371 In 1994, the IOM concluded
that the “balance of evidence is consistent with a causal rela-
tion between DTP and chronic nervous system dysfunction in
children whose serious acute neurologic illness occurred within
7 days of DTP vaccination”.355 However, the IOM was not able
to determine whether the pertussis vaccine increased the num-
ber of children with chronic neurologic illness or was simply
a precipitating event in children who would have nonetheless
developed chronic neurologic dysfunction as a result of underly-
ing brain or metabolic abnormalities.
Infantile spasms
Infantile spasms, which occur in about 40 per 100,000
infants,384 have been reported in temporal association with per-
tussis vaccination.385 Because infantile spasms typically occur
between 2 and 8 months of age, an association is occasionally
seen by chance alone. Four studies have demonstrated no foun-
dation for concern that DTP causes infantile spasms.238,376,386–388
Sudden infant death syndrome
Because SIDS occurs most often in the first 6 months of life,389
it is to be expected by chance alone that some instances would
be observed within a day or two of receipt of DTP. Several early
reports suggested clustering of SIDS cases within a few days after
the administration of DTP,390–394 but subsequent studies found
no evidence of a causal relationship between SIDS and receipt
of DTP.16,395–402 The IOM panel, after a careful review of all stud-
ies, also concluded that no causal relationship existed.354,403
Other serious conditions
The IOM panel examined the evidence concerning an associa-
tion between DTP and a variety of syndromes. Their conclu-
sions are summarized in Table 23-7.354,403
Increasing litigation over alleged vaccine injuries and with-
drawal from the marketplace of vaccines by several DTP man-
ufacturers prompted the US Congress in 1986 to pass the
National Childhood Vaccine Injury Act, which provides com-
pensation for certain untoward events that occur within speci-
fied time periods after vaccination. In 1995, program rules were
revised in light of the report of the IOM committee.354,403 The
replacement of whole-cell with acellular pertussis vaccines has
markedly reduced the rates of adverse reactions temporally
associated with pertussis vaccine,404 resulting in a correspond-
ing decline in vaccine injury claims.405
Results of vaccination: acellular pertussis
vaccines
Immune responses to acellular vaccines
Numerous immunogenicity and reactogenicity studies
have been published evaluating acellular pertussis vaccines.
However, making comparison among them is difficult because
of variations in study design, study populations, and serologic
Pertussis vaccines 23 467
Institute of Medicine Conclusions Regarding Causation of
Serious Adverse Events by Diphtheria and Tetanus Toxoids Plus Whole-
cell Pertussis (DTP) Vaccines
Data from Howson CP, Howe CJ, Fineberg HV, eds. Adverse effects of
pertussis and rubella vaccines: a report of the Committee to Review the
Adverse Consequences of Pertussis and Rubella Vaccines. Washington, DC:
National Academy Press, National Academy of Sciences; 1991; Howson
CP, Fineberg HV. Adverse events following pertussis and rubella vaccines:
summary of a report of the Institute of Medicine. JAMA 267:392-396, 1992;
and Howson CP, Fineberg HV. The ricochet of magic bullets: summary of
the Institute of Medicine report, adverse effects of pertussis and rubella
vaccines. Pediatrics 89:318-324, 1992.
assays. To provide such comparisons and facilitate selection of
candidate vaccines for anticipated efficacy trials, the National
Institute of Allergy and Infectious Diseases (NIAID) spon-
sored the MAPT in six of its Vaccine Treatment and Evaluation
Units from 1991 to 1992. Thirteen acellular and two whole-
cell vaccines were evaluated in the MAPT, including all but one
of the acellular vaccines subsequently evaluated in efficacy tri-
als. Although that vaccine was not available for the MAPT, it
was evaluated thereafter at one of the Vaccine Treatment and
Evaluation Units using the MAPT protocol, procedures, and
data forms; sera were evaluated in one of the MAPT reference
laboratories. Immunogenicity and reactogenicity results from
that study are presented here, along with the MAPT results,
to provide the most complete available comparison of these
vaccines.
Healthy infants enrolled in the MAPT were randomized to
receive one of the study vaccines (Table 23-8) at 2, 4, and 6 months
of age. Whole-cell vaccines made by Lederle Laboratories (the
reference or control vaccine) and the Massachusetts Public
Health Biologic Laboratories also were evaluated. Sera were
obtained before the first immunization and 1 month after the
third immunization. Serologic assays included ELISA antibody
to PT, FHA, PRN, and fimbrial proteins; Chinese hamster ovary
cell toxin neutralization and agglutination assays; and assays of
diphtheria and tetanus antitoxin.311
Each vaccine produced significant increases in antibod-
ies directed against its included antigens, which most often
equaled or exceeded those produced by the reference whole-
cell vaccine (see Table 23-8).311 Nonetheless, postimmuniza-
tion antibody levels differed substantially among the acellular
592 SECTION TWO: Licensed vaccines

If adults need primary polio vaccination, they should always Countries where IPV is Recommended by Health Authorities
receive IPV, because VAPP after OPV appears to be more common or By Medical associations for Routine Pediatric Immunization (2010).*
after 18 years of age. IPV-containing combination vaccines made
of low-dose diphtheria and tetanus toxoids plus IPV (some includ-
ing also low-dose acellular pertussis antigens) have been developed
and licensed in many countries.121–125 In principle, vaccination of
previously unvaccinated adults, to protect them from VAPP, is rec-
ommended when they are in contact with children excreting OPV.
Adults traveling to polio-epidemic or polio-endemic areas should
receive IPV as a booster before their first trip.267 Laboratory per-
sonnel working with wild polioviruses should have previously
completed vaccination. Health care workers should also be vacci-
nated because they may come into contact with wild poliovirus or
reverted attenuated viruses excreted by vaccinees.
IPV is universally recommended for subjects with known
congenital or acquired immunodeficiency, including HIV infec-
tion, in view of the VAPP risk in those patients after use of
OPV.268 Those receiving systemic steroid therapy or chemother-
apy are included in this indication. In developing countries OPV
is recommended for asymptomatic HIV-seropositive people,
because the risk of polio from wild viruses is considered larger
than the risk from VAPP. In industrialized countries where OPV
is still routinely used but IPV is available for some indications,
family contacts of immunocompromised people should receive
IPV instead of OPV to avoid transmitting the vaccine viruses to
the immunocompromised host.

Contraindications to IPV
Formal contraindications to IPV consist of previous severe reac-
tion to IPV vaccine or known or documented allergy to strep-
tomycin, neomycin, or polymyxin B. Neither pregnancy nor
breastfeeding is a contraindication.

Simultaneous use with other vaccines

No clinically relevant interference effects have been described
when IPV is used in association or in combination with
licensed DTwP/DTaP, Hib, or hepatitis B vaccines. The largest
experiences with IPV combinations have been between the mid-
1980s and the mid-1990s with DTwP-IPV and DTwP-IPV/Hib
in France and Canada, and since the mid-1990s with DTaP-
IPV-backboned combinations including HepB and/or Hib in the
United States, Canada, Europe, and elsewhere (see Chapter 40).
In combination vaccines, IPV is compatible with DTaP, Hib,
and hepatitis B, although generalization is difficult owing to
the pharmaceutical specificities of all of these drug substances,
which are the key driver of the mixability of these antigens.
Extemporaneous mixing of two distinct liquid finished products
should not be made by vaccinators.

Public health considerations

Results of vaccination programs with IPV
Experience with IPV in national programs has been longest in
Europe and in Canada. Some countries have used IPV exclu-
sively since the mid-1950s, and some as part of mixed or
sequential schedules with OPV, as reviewed by Murdin,203,
Plotkin,264 and Bonnet and Dutta.94
Table 27-14 lists the nearly 60 countries (2010) where IPV-
containing vaccines are recommended for routine pediatric vac-
cination as IPV-only schedules or as part of sequential schemes
with OPV or as IPV-only schedules supplemented by SIAs with *Either as IPV-only schedules or as part of sequential IPV/OPV schedules
or as part of IPV/OPV combined schedules.
OPV.263 Some of these experiences are reviewed in the next sec-
IPV, inactivated polio vaccine; OPV, oral poliovirus vaccine.263
tion for several WHO regions.
594 SECTION TWO: Licensed vaccines
Steps in US Decisions Regarding Polio Vaccination
ACIP, Advisory Committee on Immunization Practices; IOM, Institute of Medicine; IPV, inactivated poliovirus vaccine; OPV, oral poliovirus vaccine; RIVM, Rijksinstituut
Voor Volksgezondheid en Milien; VAPP, vaccine-associated paralytic poliomyelitis.
Israel has used both polio vaccines in an attempt to solve their
particular epidemiologic situation in which two communi-
ties that live close together have different hygienic conditions
and levels of vaccination coverage. After a brief experience
with IPV, Israel started routine OPV vaccination in 1960.
Vaccination coverage reached high levels among both Jewish
and Arab children. Nevertheless, sporadic poliomyelitis con-
tinued among Jews, and small epidemics continued to occur
in the West Bank and Gaza.290–294 In view of the failure of OPV
to control polio, in 1978 the Israelis introduced a combined
mixed schedule: OPV was administered at 1, 2.5, 4, 5.5, and
12 months of age, and IPV (as a DTwP-IPV vaccine) was given
at 2.5 and 4 months of age. For a time in the 1980s, there
were no cases in Israel proper and only sporadic cases in the
Palestinian areas.290–292 All was well until 1988, when an epi-
demic of 15 cases of type 1 polio occurred in Israel294 local-
ized in one of two districts that had adopted IPV vaccination
of infants. Although the analysis of this epidemic is contro-
versial, it is clear that antibody responses to OPV were sub-
optimal among Israeli young adults, resulting in a low level
of resistance. Conversely, the wild virus may have circulated
among infants immunized with IPV only, allowing spread to
their parents. The response to the epidemic included mass
vaccination with OPV and the institution of three doses of
DTP-IPV in the routine vaccination scheme together with four
simultaneous doses of OPV.294 Since 1988, no cases of polio
have been reported in Israel or its territories, despite an out-
break in neighboring Jordan from 1991 to 1992 that caused
wild virus circulation in Gaza.293
Since 2007, an IPV introduction program has been put in
place in the entire Yogyakarta province of Indonesia, and IPV
is the exclusive EPI vaccine used. Polioviruses were only iso-
lated a few times in sewage samples after the switch to IPV, but
some challenges in the environmental sampling occurred.295
Seroprevalence/seroconversion surveys and vaccination cover-
age evaluations have not yet been reported.
Australia and New Zealand are now using IPV-containing
acellular pertussis combinations based on a cost analysis show-
ing that the introduction of IPV in a combination vaccine cost-
ing $10 or less is cost-effective.296
Rationale for the use of IPV
Since the early 1960s, the well-recorded vaccinology contro-
versy is the choice of IPV or OPV for routine vaccination in
infancy. Table 27-16 summarizes the advantages and disadvan-
tages of IPV-only, OPV-only, or mixed and sequential IPV/OPV
schedules.264 In essence, the key arguments for IPV usage are
safety (no VAPP and no induction of VDPVs), predictable and
consistent immunogenicity, and the possibility of its inclusion
in combination vaccines. The key arguments for OPV usage
are induction of better mucosal intestinal immunity, ease of
administration to large populations, and low cost. The argu-
ment for a mixed schedule is to fuse the immunogenicity
advantages of each vaccine with less or null risk of VAPP when
IPV is started first.
VAPP, which is discussed in detail in Chapter 28, is an ines-
capable phenomenon that has been consistently observed with
OPV.307,308 In our view, the following circumstances should lead
to the choice of IPV for routine vaccination of infants in a par-
ticular country:
t Absence of paralytic polio and likelihood that wild
polioviruses are no longer circulating. This criterion applies
to countries where eradication of polio has been certified,
even if importation is possible by migrants.
t High vaccination coverage with routine DTP, equivalent to
90% or better in infants and children, so that introduction
of wild virus is unlikely to result in spread.
t Ability of the medical systems or of vaccinees to afford the
costs of IPV-containing vaccines, although the cost issue
is not straightforward, particularly when a full pharmaco-
economical cost-effectiveness analysis is made, including
the cost of National Immunization Days with OPV.
A dose of OPV costs $0.13 to $0.14 for UNICEF or for other types
of public markets, whereas a dose of IPV stand-alone costs 2.5,
but this price may be unduly high because volume orders are low.
In private markets, the price of stand-alone IPV vaccines and IPV-
containing combination vaccines vary widely, depending on the
type of markets and the type of vaccine used (the IPV-containing
combinations representing the vast majority of volumes distrib-
uted), and prices start at several US dollars. Public sector prices
also vary considerably depending on the type of market, volume
purchased, and contract conditions. However, given the right con-
ditions, the prices of IPV-containing combination vaccines (today

Advantages and Disadvantages of All-OPV, All-IPV, or IPV/OPV Mixed Vaccination Schedules
*WHO estimate is between 1/250,000 and 1/800,000 first OPV doses. From WHO. Wkly Epidemiol Rec 86:205-220, 2011.
IPV, inactivated polio vaccine; OPV, oral poliovirus vaccine; VAPP, vaccine-associated paralytic polio; VDPV, vaccine-derived polioviruses.
all of them are acellular pertussis combination vaccines) could be
in the $5 to $8 per dose range. If the IPV-containing combinations
were based on whole-cell pertussis rather than acellular pertussis
vaccines and ideally manufactured from non-Western manufac-
turing units with lower manufacturing cost structures, the prices
would be lower. Although these prices per dose are greater than
the price of OPV, the human and financial costs of VAPP and
VDPV, the high amount of wastage of OPV, and the cost of keep-
ing the oral vaccine frozen must be taken into account. An analy-
sis commissioned by the Bill and Melinda Gates Foundation297
concluded that if requested, current manufacturers could supply
IPV for all the world's children, although several years would be
required for ramping up. The report also estimated that this level
of production could lower the price per dose to between $0.50 and
$2.00. However, the price of IPV in combination vaccines was not
calculated. These prices are still substantially higher than the cost
of OPV and WHO is undertaking efforts to bring them down to
those of a course of OPV (see the following section). Khan298 con-
cluded that if the risk of recrudescence of wild or vaccine strain
polio is taken into account, the cost of IPV replacement of OPV
would be less than continued use of OPV.
In the scope of its Global Polio Eradication Initiative Strategic
Plan 2010-2012, WHO has actively engaged in promoting
research and development activities toward the emergence of
affordable IPV solutions. The major focus for IPV research has
been on its use post wild poliovirus circulation eradication when
OPV use must stop because of the potential for cVDPVs, although
research on possible pre-eradication IPV vaccination has also been
conducted. The hope for post-eradication IPV use is to minimize
the risks of reintroduction and spread of polioviruses as cVDPVs,
or through laboratory containment breaks, spread from chronic
immune-deficient poliovirus shedders, and even the potential for
reintroduction of polioviruses as part of bioterrorism. By assuring
continuing population immunity through IPV, those risks could
be markedly reduced. The Polio Research Committee,301 the polio
Poliovirus vaccine—inactivated 27 595
working group of the WHO Strategic Advisory Group of Experts,299
and the Independent Monitoring Board302 are regularly reviewing
progress made from a multipronged approach. This approach
includes developing a better understanding of the role that can be
played by IPV in boosting immune responses, particularly intes-
tinal immunity in populations shown to be poorly responsive to
OPV, to the development of new IPV vaccines, or to new ways of
using existing vaccines. The “low-dose” option consisting of the
use of the ID intradermal route (described above), or of one-dose
IPV regimen followed by one or more OPV (“light” mixed sequen-
tial schedule), or of two-dose IPV-only regimen is being explored
(see “Immune responses”). The use of classical (aluminum salts)
or new (squalene-based emulsions or other) adjuvants aimed at
reducing the amount of antigen is also under active investigation.
The most advanced option is the use of the Sabin strains allow-
ing potential new manufacturers, particularly those in the devel-
oping world who may be able to produce vaccine more cheaply,
to play a role in IPV supply (see “IPV manufactured from Sabin
strains”). Finally, the development of new and alternate poliovi-
rus seed strains usable for IPV manufacturing based on stabilized
Sabin-like strains or on more innovative options based on the use
of noninfectious genetic material transfected in expression sys-
tems is under investigation.
The most recent position of WHO300 is that “The national
choice of vaccines and vaccination schedules during the pre-
eradication period must include OPV or IPV, or a combination
of both, and should be based on assessments of the probabilities
and consequences of wild poliovirus importation”. However,
WHO notes that after eradication, use of OPV will have to stop
and the vaccine may become unavailable.
The criteria for adopting mixed sequential schedules begin-
ning with IPV and ending with OPV are the same, but with the
addition of a public health policy factor: the desire to prevent
polio by all possible means, taking advantage of both vaccines
but also eliminating VAPP.309,310
 SECTION TWO: Licensed vaccines
Poliovirus vaccine—live
28 Roland W. Sutter
Olen M. Kew
Stephen L. Cochi
R. Bruce Aylward
The written history of poliomyelitis can be traced back more
than 200 years to the first description of the disease as a
separate clinical entity by Michael Underwood in 1789.1
Since then, many important scientific discoveries and pub-
lic health milestones have been associated with poliomyelitis.
Undoubtedly, the most consequential of these was the devel-
opment of effective poliovirus vaccines,2 which paved the way
for the implementation of control programs, and a resolution
by the World Health Assembly that established the goal of
global eradication of poliomyelitis by the year 2000.3
Implementation of the eradication strategies led to elimina-
tion of poliomyelitis in the World Health Organization (WHO)
Regions of the Americas, the Western Pacific and Europe and
to substantial decreases in polio incidence in the rest of the
world. By 2012, only three countries (Afghanistan, Nigeria,
and Pakistan) have never interrupted indigenous wild poliovi-
rus transmission, and the introduction of innovative new tools,
such as monovalent oral poliovirus vaccines (mOPVs) in 2005
and bivalent OPV (bOPV) in 2009, may facilitate eradication.
Although the written history of poliomyelitis is relatively
succinct, an Egyptian stele from the 18th dynasty (dated
between 1403 and 1365 BCE) depicts a “crippled young man,
apparently a priest, with a withered and shortened right leg, and
with his foot held in a typical equinus position characteristic
of flaccid paralysis”.4 This inscription demonstrates that polio-
myelitis probably has affected humankind since ancient times.
Underwood1 introduced the term debility of the lower extremi-
ties in 1789; other researchers suggested a series of alternative
terms, including Lähmungszustände der unteren Extremitäten
in 1840,5 morning paralysis in 1843,6 paralysie essentielle
chez les enfants in 1851,7 paralysie atrophiques graisseues de
l'enfance in 1855,8 spinale Kinderlähmung in 1860,9 tephro-
myelitis anterior acuta parenchymatose in 1872,10 and polio-
myelitis anterior acuta in 1874.11 The last term is based on
anatomic location of lesions within the spinal cord—which was
discovered in the early 1870 s—and constructed from the Greek
words polios (ie, gray) and myelos (ie, marrow, the gray matter
of the spinal cord) with the ending “-itis” to imply inflamma-
tion. Although the terms Heine-Medin disease in 190712 and
infantile paralysis were proposed subsequently, poliomyelitis
(and the shortened term, polio) prevailed and became the stan-
dard designation for the disease.
In the late 19th and early 20th centuries, a change in the
epidemiology of poliomyelitis from a predominantly endemic
to an epidemic form was observed in Sweden and Norway,12–14
heralding similar changes in other industrialized countries.
Our understanding of these changes in the epidemiology was
greatly aided by groundbreaking investigations of the three larg-
est poliomyelitis outbreaks of the time: (1) 132 cases in Rutland
County, Vermont, in 1894 by Charles Caverly;15 (2) 1,031 cases
in Sweden in 1905 by Ivar Wickman;12,14 and (3) more than
9,000 cases in New York in 1916.16 Wickman14 was the first to
recognize that abortive cases might equal or outnumber cases
with paralytic manifestations and that these cases may be sig-
nificant in the propagation of the infection. These outbreaks
were due to an accumulation of a sufficient number of suscep-
tible children to sustain epidemic transmission of poliovirus,
presumably because improvements in hygiene and sanitation
had delayed poliovirus exposure from infancy to later in life.
Landsteiner and Popper17 reported in 1908 that a “filtrable
agent” (ie, virus) was the cause of poliomyelitis on the basis
of microscopic examination of spinal cords from two mon-
keys that had been injected intraperitoneally with a suspension
of ground-up cord from a fatal human case. Burnet and
Macnamara18 determined in 1931 that more than one serotype
of virus could cause poliomyelitis and that immunity to one
serotype did not confer immunity to another serotype. These
investigators based their findings on cross-immunity and
serologic tests and, most importantly, showed that three mon-
keys that had recovered from one serotype (and might have been
expected to be immune) developed paralytic disease after injec-
tion of another serotype. This report had profound implications,
although not immediately appreciated at the time, in terms of
redefining the epidemiology of the disease and with respect to
directing the subsequent development of vaccines. In 1948, an
effort was launched to determine the number of distinct polio-
virus strains. This effort was coordinated by the Committee
on Typing of the National Foundation for Infantile Paralysis,
which reported in 1951 that three and only three serotypes of
poliovirus, designated types 1, 2, and 3, were the cause of polio-
myelitis.19 Enders et al20 demonstrated in 1949 that poliovirus
could be grown in nonnervous, human embryonic tissue, work
that was later honored with the Nobel Prize. Thus, the deter-
mination of the number of poliovirus serotypes, the ability for
large-scale growth of the virus, and the finding that circulat-
ing antibody had a protective effect against poliomyelitis21–26 all
were essential preconditions for the development of effective
poliovirus vaccines.
Two approaches for vaccine development pursued at the time
were successful: inactivation of poliovirus by formalin,27 fully
developed by Jonas Salk and colleagues, licensed as inactivated
poliovirus vaccine (IPV) in 1955 after one of the largest con-
trolled field trials ever conducted;2 and the attenuation of the
three serotypes of poliovirus by Albert Sabin, licensed in 1961
as monovalent OPV and in 1963 as trivalent OPV (tOPV).28 The
widespread use of IPV and OPV rapidly controlled poliomyelitis.
The invention of the Drinker respirator (ie, “iron lung”)
beginning in 1928 and its widespread use in the 1930s and

1940s rapidly decreased the case-fatality rate of bulbar forms
of poliomyelitis.29 Epidemic poliomyelitis in the early part of
the 20th century was associated with a high case-fatality rate
(27.1% during the New York epidemic of 1916).16,30,31 Further
improvements in hygiene and sanitation delayed the median
age of poliovirus infection from younger than 5 years in the
1910s to 5 to 9 years in the 1940s32,33 and allowed the accu-
mulation of large numbers of people susceptible to poliomy-
elitis. Epidemics of ever-increasing magnitude began to occur
in the United States and Europe until the mid to late 1950s,
when vaccines became available. Because increasing age seemed
to be the primary risk factor for bulbar paralysis—the basis for
the “central dogma” of the epidemiology of poliomyelitis34—an
increasing proportion of cases required respiratory support, and
whole wards of iron lungs were devoted to caring for poliomy-
elitis victims in the 1940s and 1950s.
Poliomyelitis is intimately linked with some of the great-
est triumphs in medicine, including scientific breakthroughs,
public health achievements, and advancements in social jus-
tice. The concept of social justice, the indiscriminate benefit of
scientific discoveries or access to care and rehabilitation, was
pioneered by the National Foundation for Infantile Paralysis,
which raised funds through annual March of Dimes campaigns
that covered treatment and rehabilitation costs of poliomyelitis
victims.35 The Vaccines for Children Act (1993) ensures that
poliovirus vaccines are available for poor children in the United
States. Ultimately, the successful conclusion of the Global Polio
Eradication Initiative (GPEI) will benefit all children equally,
whether rich or poor; whether brown, white, or black; and
whether living in industrialized or developing countries.
The history of poliomyelitis has been reviewed in detail
by Paul4 and Eggers36 and in Chapter 25 of the fifth edition of
Vaccines.
Why is the disease important?
Poliomyelitis was the leading cause of permanent disability
in children in the prevaccine era.32 Besides the considerable
disease burden, poliomyelitis was much feared in the prevac-
cine era because it could strike anybody, no means existed of
protecting oneself or one's children, and, unlike the situation
with other diseases such as measles, from which most children
recover or die rapidly, society was reminded every day of the
devastating effects of this crippling disease.
Disease control programs using poliovirus vaccines have
prevented and continue to prevent millions of children from
becoming paralyzed. In 1988, when the global eradication target
was adopted, the WHO estimated that approximately 350,000
cases of paralytic poliomyelitis were occurring annually,37
despite availability of effective vaccines during the prior three
decades. Poliomyelitis has gained renewed attention in recent
years because the feasibility of its eradication has been dem-
onstrated37,38 and because of the visibility of the ongoing global
effort to eradicate poliovirus.39–41
Background
Clinical description
Poliomyelitis is an acute infection caused by any of three sero-
types of poliovirus that replicate initially in the gastrointestinal
(pharyngeal and intestinal) tract and rarely in the motor neu-
rons of the anterior horn cells in the spinal cord, where repli-
cation of virus results in cell destruction and flaccid paralysis
of the muscles the cells innervate (ie, spinal poliomyelitis).
On occasion, brainstem cells innervating respiratory muscles
Poliovirus vaccine—live 28 599
can be affected, resulting in difficulties in breathing (ie, bul-
bar paralysis). In addition to the acute paralysis, late mani-
festations with exacerbation of weakness or new paralysis (ie,
postpolio syndrome) can be observed in a significant proportion
of patients decades after the acute paralytic episode.
Poliovirus exposure in a person susceptible to poliomyeli-
tis results in one of the following consequences: (1) inapparent
infection without symptoms, (2) minor illness, (3) nonparalytic
poliomyelitis (aseptic meningitis), or (4) paralytic poliomyeli-
tis.32,42–44 Inapparent infection without symptoms is the most
frequent outcome (72%) after poliovirus exposure in suscepti-
ble people.45 Minor illness is the most frequent form (24%) of
the disease, characterized by transient illness associated with a
few days of fever, malaise, drowsiness, headache, nausea, vom-
iting, constipation, or sore throat, in various combinations.45
Nonparalytic poliomyelitis (aseptic meningitis) is a relatively
rare outcome (4%) of poliovirus infection. It begins usually as
a minor illness characterized by fever, sore throat, vomiting,
and malaise, and 1 to 2 days later, signs of meningeal irrita-
tion become apparent, including stiffness of the neck or back;
vomiting; severe headache; and pain in the limbs, back, and
neck.42 This form of the disease lasts 2 to 10 days, and recovery
is usually rapid and complete. In a small proportion of these
cases, the disease advances to transient mild muscle weakness
or paralysis.
Paralytic poliomyelitis is a rare outcome (usually 1%) of
poliovirus infections among susceptible people. Its clinical
course is characterized by a minor illness of several days and
a symptom-free period of 1 to 3 days, followed by rapid onset
of flaccid paralysis with fever and progression to the maxi-
mum extent of paralysis within a few days. This characteristic
clinical course of a minor illness followed by the major illness
with paralysis has been likened to the two humps of the drom-
edary.46,47 (In actuality, this is a misnomer, because the drom-
edary is a one-humped camel.) Among adolescent and adult
cases of poliomyelitis, the minor illness is often absent, and
these groups also seem to experience more severe pain in the
affected extremities. After temperature returns to normal, there
is usually no further progression of paralysis. If paralysis of an
extremity is not complete, it is more pronounced proximally.
Paralysis is usually asymmetric, associated with diminished
or complete loss of deep tendon reflexes and an intact sensory
system. Paralytic manifestations in extremities begin proxi-
mally and progress to involve distal muscle groups (ie, descend-
ing paralysis). Depending on the anatomic location of motor
neuron damage in the spinal cord or in the brainstem, spinal,
mixed spinal-bulbar, or bulbar paralysis involving primarily
respiratory muscles may be observed. The anterior horn cells
(and brainstem cells), just like other nerve cells of the central
nervous system (CNS), cannot be regenerated or replaced and
paralysis is permanent. Nevertheless, because of compensation
of other, still functioning muscles, partial or total recovery can
be achieved, usually within the first 6 months after onset of dis-
ease. Detailed clinical descriptions may be found in a number
of reviews and books.48–51
Postpolio syndrome, a term invented in the early 1980s,
refers to a disease entity that encompasses the late manifesta-
tions of acute paralytic poliomyelitis.52 Aside from previously
published case reports and case series, the first systematic
investigation of postpolio syndrome was published in 1984.53
After an interval of 15 to 40 years, many people (25%-40%)
who contracted paralytic poliomyelitis in their childhood may
experience muscle pain and exacerbation of existing weak-
ness or may develop new weakness or paralysis. Factors that
enhance the risk of postpolio syndrome include the following:
(1) increasing time since acute poliovirus infection, (2) pres-
ence of permanent residual impairment after recovery from
the acute illness, and (3) female sex. The exact cause of these
late effects is unknown, although it is not a consequence of
600 SECTION TWO Licensed vaccines
persistent infection. The pathogenesis of postpolio syndrome
is thought to involve late attrition of oversized motor units
that developed during the recovery process of paralytic polio-
myelitis.52 Postpolio syndrome has been described in people
infected during the era of wild poliovirus circulation. An excel-
lent summary of the current scientific knowledge of postpolio
syndrome has been published.54
Virology
Polioviruses are part of the Enterovirus genus and belong to
the family Picornaviridae (Italian, pico, implying small, and
RNA, the nucleic acid component).55–57 Polioviruses are small
(27-30 nm in diameter), nonenveloped viruses with capsids of
icosahedral symmetry enclosing a single-stranded, positive-
sense RNA genome about 7,500 nucleotides long. Polioviruses
share most of their biochemical and biophysical properties with
the other enteroviruses. They are stable at acid pH (3.0-5.0)
for 1 to 3 hours and resistant to inactivation by many com-
mon detergents and disinfectants, including soaps, nonionic
detergents, and lipid solvents such as ethanol, ether, and chlo-
roform. Polioviruses are rapidly inactivated by treatment with
0.3% formaldehyde or 0.5 ppm free residual chlorine, by desic-
cation, low humidity, or exposure to ultraviolet light. Infectivity
is stable for weeks at 4 °C and for days at 30 °C. Polioviruses are
readily inactivated at 55 °C, but magnesium chloride stabilizes
infectivity.58
There are three antigenic types (serotypes 1, 2, and 3)19,59–62
(Table 28-1), which, apart from their antigenic differences,
are otherwise similar. The complete genomic sequences have
been determined for representatives of each serotype,63 includ-
ing those of the three Sabin OPV strains. Only the sequences
encoding the capsid proteins are unique to polioviruses because
the flanking sequences are frequently exchanged by recombi-
nation with closely related enteroviruses (species C enterovi-
ruses) during circulation in nature.64–66 The poliovirion consists
of 60 copies each of four capsid proteins (VP1 through VP4)
that form a highly structured capsid shell.67 The three major
proteins (VP1, VP2, VP3) share a similar basic architecture
and were probably derived from a common ancestral protein.
The smallest protein, VP4, internalized in the native virion, is
formed by the cleavage of the precursor VP0 (VP4 VP2) during
the final maturation of the virion. The external surface of the
poliovirion is decorated by peptide loops extending from VP1,
VP2, and VP3, which form the neutralizing antigenic sites.67–69
Four neutralizing antigenic sites have been identified by pat-
terns of reactivity with neutralizing monoclonal antibodies,
and the assignments have been confirmed by high-resolution
X-ray crystallography.69–72 Sites 2, 3, and 4 are discontinuous
and formed from loops contributed by different capsid proteins.
The major type-specific differences in the sequences of the cap-
sid polypeptides primarily reside on the most surface-accessi-
ble peptide loops, which represent less than 4% of the total
capsid protein. Amino acid sequences of the underlying frame-
work of the capsid are highly conserved across poliovirus sero-
types.63 Although the neutralizing antigenic sites vary within
Polioviruses
*Virus recovered from feces.
†
Virus recovered from spinal cord.
each serotype, the range of variability is constrained such that
all polioviruses within a serotype can be neutralized by type-
specific antisera.
Polioviruses attach to and enter cells via a specific polio-
virus receptor (PVR or CD155) on the cytoplasmic mem-
brane, a glycoprotein of the immunoglobulin superfamily
whose host function is related to cellular adhesion and acti-
vation.73,74 The terminal residues of the receptor specifically
interact with sequences forming a canyon-like channel on
the virion surface. Each poliovirion has 60 receptor-binding
sites. 75–77 The PVR is expressed on the surface of human and
simian cells supporting poliovirus infection, but absent on
cells from nonprimates, rendering them resistant to infec-
tion. The human PVR gene has been cloned.73 On intro-
duction into resistant cells, the human gene converts them
into susceptible cells.78 The gene for the human PVR has
been introduced into a germ line of mice.79,80 The result-
ing transgenic animals become susceptible to polioviruses,
demonstrating that the primary block to infection of normal
mice by such strains is at the level of cell entry.79 Transgenic
mice expressing the human PVR gene become paralyzed on
intraspinal, intracerebral, intranasal, or intramuscular inoc-
ulation with wild poliovirus but do not develop disease on
inoculation with attenuated Sabin strains.79 Such transgenic
mice have proven useful for neurovirulence testing of OPV
lots, study of field isolates, and detailed studies of poliovi-
rus pathogenesis.81 Because of concerns that transgenic mice
infected with poliovirus may escape the laboratories, breed
with nontransgenic mice, and potentially establish a nonhu-
man reservoir of poliovirus transmission, specific guidelines
have been issued to address this theoretical risk.82
Polioviruses are among the simplest viruses in terms of
genetic organization and replication cycle.57 However, they
grow rapidly, yielding up to 10,000 infectious particles per
cell in a growth cycle of 4 to 5 hours. After attachment and
entry, the genomic RNA is uncoated and translated under the
control of the “internal ribosomal entry site” (IRES) to pro-
duce a single polypeptide, the polyprotein.83 The polyprotein
is cleaved after translation into the virus-specific proteins,
which include the capsid proteins, two virus-specific proteases
controlling proteolytic processing, and an RNA-dependent
RNA polymerase catalyzing the synthesis of RNA molecules
of negative and positive polarity. The negative-polarity RNA
serves as a template for synthesis of positive-polarity mes-
senger and genomic RNA. Host protein synthesis is rapidly
inhibited by the cleavage of a cellular protein required for ini-
tiation of translation of host messenger RNA but not for the
initiation of translation from the poliovirus IRES. Infected
cells are rapidly converted into factories for poliovirus syn-
thesis, show cytopathic effects after several hours, and release
virus through cell lysis and death. Once virion assembly
has started, production of capsid protein and replication of
RNA are closely linked, and integration of viral RNA into the
virion follows within several minutes. Morphogenesis seems
to involve the combination of viral RNA with a shell of viral
proteins (VP0, VP1, VP3), during which the VP0 procapsid
protein is cleaved to yield VP2 and VP4.

Pathogenesis as it relates to prevention
The pathogenesis of poliovirus infection indicates that preven-
tion through immunization can be accomplished by inhibiting
replication at and dissemination from the gastrointestinal tract,
by inhibiting the viremia that follows, or by doing both. After
a person is exposed to poliovirus by way of the oral cavity, the
virus attaches and enters specific cells that express the PVR.73
The virus replicates locally at the sites of virus implantation
(eg, tonsils, intestinal M cells, and Peyer patches of the ileum) or
at the lymph nodes that drain these tissues. The first approach
requires the presence of local secretory IgA antibody. The second
approach, because spread occurs primarily by way of the blood-
stream to other susceptible tissues (ie, other lymph nodes, brown
fat, and the CNS) or by way of retrograde axonal transport to the
CNS, requires the presence of circulating neutralizing antibody.
The host range of poliovirus and tissue tropism is determined by
the expression of the PVR.73 Tissue tropism refers to the ability of
poliovirus to replicate in specific cells.84 Early studies on poliovirus
pathogenesis were conducted in primates, but the development of
PVR transgenic mice has led to a renaissance in pathogenesis stud-
ies.85,86 Most of the earlier observations with primates48,49,51 have
been confirmed in the PVR transgenic mouse model,86 and new
tools have been used to investigate important unresolved ques-
tions about poliovirus pathogenesis. For example, in situ hybrid-
ization with nucleic acid probes of the PVR in transgenic mice
suggested PVR expression was limited to the CNS, thymus, lung,
kidney, and adrenal glands and, more recently, monocytes (mono-
nuclear phagocytes).87 Replication of poliovirus in motor neurons
results in cell destruction and paralysis. The chief limitation of the
mouse model is the low level of expression of the PVR in mouse
intestinal tissue, resulting in inefficient replication of virus at the
primary site of natural infection.86 The transgenic mouse model
has facilitated neurovirulence studies but has limited usefulness in
assessing the infectivity or transmissibility of the different poliovi-
rus strains because of the difficulties in orally infecting the mice.
Thus, the lack of an animal model to study infectivity and trans-
missibility of poliovirus remains a major obstacle in learning more
about these aspects of the poliovirus.
Poliovirus may be found in the blood of patients with the
abortive form (“minor illness”), and it can be detected sev-
eral days before onset of clinical signs of CNS involvement in
patients in whom nonparalytic or paralytic poliomyelitis devel-
ops.24,88 The virus is regularly present in the throat and in the
stools before the onset of illness. In persons who have clinical
or subclinical infection, virus is excreted in the feces for several
weeks45 and in saliva for 1 to 2 weeks. The mean duration of
wild poliovirus type 1 excretion in fecal specimens is 24 days
(median, 20-29 days), with an observed range of 1 to 114 days.45
For further details of pathogenesis and pathology, see
Bodian,48 Bodian and Horstmann,49 Sabin,51 Racaniello and
Ren,86 Mueller and Wimmer,89 Nathanson,90 and Koike and
Nomoto.81
Diagnosis
Paralytic poliomyelitis caused by imported wild poliovirus or
by vaccine-related poliovirus has become a rare disease in the
United States and other industrialized countries. Therefore,
physicians may not be familiar with the disease or consider
the diagnosis of poliomyelitis until other more frequent causes
of acute flaccid paralysis (AFP) have been ruled out. The diag-
nosis of paralytic poliomyelitis is dependent on the following:
(1) clinical course, (2) virologic testing, (3) special studies, and
(4) residual neurologic deficit 60 days after onset of symptoms.
For surveillance purposes, any case with physician-diagnosed
suspected poliomyelitis is investigated in the United States,
Poliovirus vaccine—live 28 601
and a case is confirmed if a panel of independent experts deter-
mines that the case definition* for paralytic poliomyelitis has
been met. The WHO is using a sensitive screening case defini-
tion: any case of AFP in a person younger than 15 years or a case
in a person of any age in whom poliomyelitis is suspected. This
definition attempts to capture all cases of paralysis that might
be caused by poliovirus, including cases erroneously diagnosed
as other clinical syndromes such as Guillain-Barré syndrome.
This sensitive screening definition is balanced by a specific case
classification system (ie, virologic case classification scheme)
that relies on isolation of poliovirus from stool specimens to
confirm cases of AFP as poliomyelitis. As countries improve
the quality and timeliness of AFP surveillance, the virologic
case classification has replaced the clinical case classification
scheme in all endemic and recently endemic countries (see
“Disease control strategies”).
Clinical course
The clinical course (see “Clinical description”) is helpful in rul-
ing in or ruling out paralytic poliomyelitis. Several studies in the
developing world have attempted to assess the sensitivity and
specificity of different clinical case definitions for paralytic polio-
myelitis and compared these with the “gold standard” of virologi-
cally confirmed poliomyelitis based on poliovirus isolation from
stool specimens. These studies reported similar findings.91–93 The
largest study reported a sensitivity of 64% and a specificity of
82% for a case definition that included age younger than 6 years,
fever at onset, and rapid progress to maximum extent of paral-
ysis ( 4 days).91 The addition of a specific pattern of paralysis
(proximal, unilateral, or absence of paralysis in all four extremi-
ties) increased the specificity with varying degrees of loss in sen-
sitivity. The case definitions and case classification schemes have
been reviewed.94 Data from India suggest that residual paralysis
60 days after onset of paralysis is the strongest predictor for con-
firmed poliomyelitis (ie, isolating wild poliovirus in stool sam-
ples during the initial examination). These findings highlight the
importance of a 60-day follow-up examination to assess whether
areas with wild poliovirus circulation may have been missed.
Virologic testing
Because AFP has many causes, including Guillain-Barré syn-
drome, transverse myelitis, and infection with nonpolio entero-
viruses (see “Differential diagnosis”), laboratory confirmation is
critical to establishing the diagnosis of poliomyelitis. The basic
approach is to attempt to isolate poliovirus from the stools of
patients with AFP and to characterize any poliovirus isolates to
determine whether they are vaccine-related or wild. Detailed
descriptions of standard laboratory principles and procedures
for investigation of enterovirus infections are available,95 and
standard typing antisera for identifying enteroviral isolates are
available through the WHO.96 The WHO has published a man-
ual for the virologic investigation of poliomyelitis cases that
includes protocols for the isolation of poliovirus.97 This man-
ual has become the standard guide for the isolation and charac-
terization of polioviruses for laboratories in the WHO's Global
Laboratory Network (GPLN) for Poliomyelitis Eradication98 and
is widely used in other diagnostic laboratories.99
Two cell lines are recommended for the isolation of polio-
viruses in stool specimens: (1) RD cells, derived from human
rhabdomyosarcoma and (2) L20B cells, derived from the
mouse L cell line and genetically altered to express the human
PVR.78 RD cells have the advantage of being very sensitive
*“A patient must have had paralysis clinically and epidemiologically
compatible with poliomyelitis and, at 60 days after onset of symptoms,
had residual neurologic deficit, had died, or had no information
available on neurologic residua.” This case definition was formerly
known as best available paralytic poliomyelitis case count.
602 SECTION TWO Licensed vaccines
to poliovirus infection and yielding high poliovirus titers in
culture. RD cells will support the replication of other human
enteroviruses but not Coxsackie B viruses. L20B cells will
support the replication of polioviruses but, like their paren-
teral mouse L cells, are resistant to infection by most nonpo-
lio enteroviruses. Thus, L20B cells are used for the selective
cultivation of polioviruses.
Poliovirus may be recovered from stool, throat swabs, or cere-
brospinal fluid obtained soon after the onset of illness and from
stool specimens obtained over longer periods (usually up to 8 weeks
after the start of infection).100 Poliovirus isolation rates from cere-
brospinal fluid are generally low; however, when virus is found, a
causal relationship between a poliovirus serotype and paralytic dis-
ease is strongly suggested. The WHO recommends that two stool
samples be obtained at least 24 hours apart to confirm the diagno-
sis because excretion of virus can be intermittent and the sensitiv-
ity of isolation is less than 100%. Wild poliovirus has been found
in stool samples of 63% to 93% of patients during the first 2 weeks
of illness, in 35% to 75% during the third and fourth weeks, and in
fewer than 50% during the fifth and sixth weeks.100 The duration
of viral shedding is reduced among children who were previously
vaccinated, had preexisting homologous antibody, or had a previ-
ous intestinal infection with homologous poliovirus.100
Clinical specimens are processed to produce a virus suspen-
sion largely free of bacteria and other debris, to which antibiot-
ics are added to inhibit the growth of residual bacteria, and the
suspension is inoculated onto cell cultures. Cells are monitored
daily for cytopathic effects, which appear typically within 3 to
6 days of incubation. Poliovirus isolates are identified by sero-
type in neutralization tests using pools of specific antiserum.
Polioviruses also can be identified by the reverse transcriptase–
polymerase chain reaction (RT-PCR) using group-specific,101
serotype-specific,102 and strain-specific103,104 primer sets.
Intratypic differentiation (ITD) of poliovirus isolates (testing
whether they are vaccine-related or wild) is performed through-
out the GPLN using one antigenic and one molecular method.98
The standard antigenic ITD method uses an enzyme-linked
immunosorbent assay system with preparations of highly spe-
cific cross-adsorbed antisera.105,106 The molecular ITD methods
use genotype-specific nucleic acid probes,107,108 genotype-specific
PCR primers,109–111 or PCR coupled to analyses of restriction
fragment length polymorphism.112 More recently, poliovirus
diagnostic RT-PCR has been adapted to the real-time format,113
which has become the standard for the GPLN.114 Some GPLN
laboratories use panels of neutralizing monoclonal antibod-
ies for preliminary ITD testing,105 but antigenic methods are
increasingly being replaced by molecular methods for routine
poliovirus identification, and antigenic characterization is used
primarily for research studies. New laboratory procedures have
significantly decreased the time required to detect and confirm
new polio cases from 42 to 21 days on average.
The purpose of ITD is to screen out polioviruses that are
closely related (99% VP1 sequence identity) to the Sabin OPV
strains and are unlikely to be of current epidemiologic impor-
tance. The remaining poliovirus isolates are wild polioviruses
or vaccine-derived polioviruses (VDPVs) (see “Vaccine-derived
polioviruses”). Since the beginning of 2001, GPLN laboratories
routinely sequence the complete VP1 region of any wild polio-
virus isolated from an AFP case. Analysis of the full approxi-
mately 900-nucleotide VP1 region, performed to obtain the
degree of phylogenetic resolution necessary to distinguish
among wild poliovirus isolates, has permitted reconstruction
of individual chains of transmission from sequence data.115 The
use of genomic sequencing has given rise to a new discipline
that combines the tools and concepts of classical epidemiology
with those of microbiology, biochemistry, genetics, and evolu-
tionary biology (see “Molecular epidemiology of poliovirus”).116
Serologic testing may be helpful in establishing the diagnosis
but often does not contribute and sometimes may cause confusion
for the following reasons: (1) antibody increases have already
occurred by the time the first specimen is collected, (2) anti-
body may be present to one or more serotypes because of pre-
vious or recent vaccination, and (3) heterotypic responses may
be observed to one serotype after exposure to another serotype.
There are currently no reliable means of distinguishing anti-
body induced by vaccine-related vs wild poliovirus. Standard
protocols for neutralization assays to determine levels of anti-
body to poliovirus are available, using Sabin poliovirus that
may be neutralized with increasing dilutions of serum sam-
ples.97,117,118 Paired serum specimens are required to demon-
strate a fourfold or greater rise in antibody titer between acute
and convalescent serum samples. The first serum specimen
should be obtained as soon as possible after onset of paralytic
manifestations, and the second specimen should be obtained
2 to 3 weeks later. Neutralizing antibodies appear early and
are usually already detectable at the time of onset of paralysis.
However, if the first specimen is obtained early enough, a rise
in titer may be demonstrated during the course of the disease.
In a study from Louisiana, specimens were obtained as soon as
possible after hospital admission from patients with poliomyeli-
tis and about 6 weeks later; in 36% of patients, a fourfold rise in
poliovirus antibody titer could be demonstrated; in 61%, recip-
rocal titers were more than 320 and did not change; and in 3%,
reciprocal titers were less than 320 and remained unchanged.119
Neutralizing assays continue to be the gold standard method
for the detection of type-specific antibody in serum samples.22
Neutralization antibody induced by a single serotype may not
be completely serotype-specific. In practical terms, this seldom
constitutes a problem because the heterotypic response results
in low levels of neutralizing antibody. Because of the limitations
described, serologic testing may be more important in excluding
(eg, no detectable antibody) the diagnosis of poliomyelitis than
in confirming it. However, intrathecal immune responses can be
measured and offer the advantage of attributing a causal relation-
ship between a poliovirus serotype and paralytic disease.120 Other
assays have been proposed but are used infrequently, includ-
ing indirect immunofluorescence,121 paper-radioactive virus
method,122 enzyme-linked immunosorbent assay,123 and micro-
indirect hemagglutination and hemagglutination-inhibition.124
The GPLN has recently developed standardized tests for polio-
virus IgA and IgM antibody in serum.
Special studies
Nerve conduction and electromyography studies can point to
the anatomic location of the paralysis125—destruction of ante-
rior horn cells in the spinal cord vs a demyelinating or axo-
nal degenerative process in the peripheral nerves—helping to
exclude the most frequent cause of AFP, Guillain-Barré syn-
drome. Magnetic resonance imaging has been used infrequently,
but, in at least one patient with poliomyelitis, it highlighted the
anterior column of the spinal cord.126 Analysis of spinal fluid
may be helpful in ruling out other causes. In paralytic polio-
myelitis, the cerebrospinal fluid contains an increased number
of leukocytes, usually 10 to 200/mL, and seldom more than
500/mL.127,128 At the onset of signs of CNS involvement, the
ratio of polymorphonuclear cells to lymphocytes is high, but
within a few days, the ratio is reversed. The total white blood
cell count slowly subsides to normal levels. The protein con-
tent of the cerebrospinal fluid initially is elevated only slightly
(average, in nonparalytic cases, about 46 mg/100 mL [range,
15-165 mg/100 mL]; in paralytic cases, 68 mg/100 mL [range,
25-250 mg/100 mL]), but it rises gradually in paralytic cases
until the third week, generally returning to normal by the sixth
week.127 Glucose levels are usually within the normal range. In
fatal cases, spinal cord and brainstem tissue samples should be
examined for the typical lesions caused by viral replication and
destruction of the motor neuron cells.

Residual neurologic deficit
The clinical case definition for paralytic poliomyelitis requires
a residual neurologic deficit at 60 days after onset of paralysis.
Such a neurologic deficit may be apparent as complete flac-
cid paralysis of one or more extremities or partial paralysis or
weakness of muscles or muscle groups. In the latter instance,
because of functional recovery (intact muscles may compen-
sate for muscles that are not innervated), it may be more dif-
ficult to establish a neurologic deficit. The most severe cases
of poliomyelitis in terms of complications and fatal outcomes
occur in persons with underlying immunodeficiency disorders.
Causes and Differential Diagnosis of Acute Flaccid Paralysis
EPN, ethyl -nitrophenyl thionobenzene phosphonate.
Poliovirus vaccine—live 28 603
Differential diagnosis
The differential diagnoses of AFP have been reviewed.129 The
list of underlying causes of AFP is extensive (Table 28-2). The
causes can be classified according to the pathophysiologic
mechanisms and anatomic sites of the etiologic factors. For
example, poliovirus damages primarily the anterior horn cells
of the spinal cord. This damage leads secondarily to paralysis
of extremity muscles (Figure 28-1). The distinguishing features
of poliomyelitis, Guillain-Barré syndrome, transverse myeli-
tis, and traumatic neuritis (neuritis secondary to the trauma of
injections) are given in Table 28-3.
 Poliovirus vaccine—live 28 605

In general, in the absence of wild virus-induced poliomyeli- of the lower extremities (many cases caused by intramuscular
tis, Guillain-Barré syndrome accounts for 50% or more of the injections, which increase the risk of paralytic manifestations
cases of AFP in industrialized countries such as the United in the extremity that, in the absence of intramuscular injec-
Kingdom and Australia and in developing countries in Latin tions, would not have become paralyzed)148 or mild paralysis or
America.130–132 At times, nonpolio enteroviruses have been weakness in partially immune children, may be seen.
associated with cases of polio-like paralytic disease, but this
has been uncommon. Coxsackie virus A7 has been associated
with outbreaks of paralytic disease,133,134 and enterovirus 71 has
been involved in several outbreaks of CNS disease, including
Epidemiology
polio-like paralysis, with some fatal cases,135 with recent large
outbreaks in East Asia.136,137 Two motor neuron diseases in General epidemiology
childhood are Werdnig-Hoffmann disease, a rapidly progressing,
often fatal disorder of early childhood, and Wohlfart-Kugelberg- Poliomyelitis was once a ubiquitous, highly contagious, sea-
Welander disease, a more benign disorder with a generally later sonal viral disease (more pronounced in moderate-climate
onset.138 Electromyographic findings are useful in establishing countries) caused by three serotypes of poliovirus (types 1, 2,
the diagnosis of these disorders.139 China paralytic syndrome, and 3) that infected nearly every person in a given population
a distinct disease entity that appears different from Guillain- in the absence of vaccination.32 Paralytic manifestations are a
Barré syndrome and poliomyelitis, has been described among rare outcome ( 1%) of poliovirus infections. Important excep-
children and adults in northern China.140 Early symptoms of tions are island or isolated populations (eg, Eskimo), which can
this disease include leg weakness and resistance to neck flex- remain unaffected by the virus for varying periods and, after
ion. The weakness ascends rapidly, affects symmetrically the reintroduction, can experience outbreaks of poliomyelitis that
arms and respiratory muscles, and progresses to a maximum affect all age groups that were not affected by the previous wave
extent of weakness within 6 days on average. Electromyography of infection.149 Poliovirus type 1 seems to be the most neuro-
indicates denervation potentials in weak muscles and suggests virulent of the three serotypes.150 Most epidemic and endemic
that this entity may be a reversible distal motor nerve termi- cases of poliomyelitis are caused by poliovirus type 1, followed
nal or anterior horn lesion. Tick-bite paralysis occurs infre- by type 3. The last naturally acquired cases of wild virus type
quently and is manifested by flaccid ascending paralysis that 2 were detected in 1999.46,151 Peak transmission occurs among
usually resolves rapidly after tick removal. Botulism toxins can infants and young children (tropical areas) and school-aged chil-
also cause descending paralysis—characterized by symmetric
impairment of cranial nerves, followed by a descending pattern
of weakness or paralysis of the extremities and trunk.141 A rela-
tively frequent complication among approximately 10% to 15%
of patients with diphtheria is paralysis of the soft palate and
peripheral nerves resulting from diphtheria toxin;142 tetanus
toxin can cause flaccid paralysis of the muscles innervated by
the affected cranial nerves (ie, cephalic tetanus).143 Case reports
have linked AFP and West Nile virus (a flavivirus) infection
among elderly adults in Mississippi and Louisiana, suggesting
that West Nile virus may damage the anterior horn cells of the
spinal cord to cause paralytic disease.144–146 Japanese encepha-
litis infection seems to be an important contributing cause of
AFP in areas where this virus is endemic.147
The following signs and symptoms help in distinguishing
poliomyelitis from other causes of AFP: (1) fever present at
onset of paralysis, (2) rapid progression to maximum paraly-
sis, (3) usually asymmetric paralysis, and (4) more pronounced
paralysis proximally than distally (ie, descending paralysis).
However, as poliomyelitis becomes an increasingly rare disease,
unusual clinical manifestations, such as symmetric paralysis
606 SECTION TWO Licensed vaccines
Until recently, excretion of poliovirus was thought to be lim-
ited to 4 to 6 weeks among immune-competent persons and less
than 3 years among immunodeficient persons. In 1997, a case
report suggested that an immunodeficient patient with com-
mon variable immunodeficiency disorder (CVID) who acquired
vaccine-associated paralytic poliomyelitis (VAPP) in 1981 may
have excreted VDPV for approximately 7 years before onset of
paralysis.158,159 Subsequent investigations have revealed that
as many as 45 persons with CVID, agammaglobulinemia,
or severe combined immunodeficiency disorder (SCID) have
excreted poliovirus for prolonged periods, all for 6 months or
longer (prolonged excretion), including 7 persons who excreted
poliovirus for 5 years or longer (chronic excreters), based on epi-
demiologic data and molecular sequencing information160 (see
“Molecular epidemiology of poliovirus”) and WHO unpub-
lished data (Table 28-4).158,159,161–169 Many of the long-term car-
riers stopped excreting poliovirus spontaneously. As of the end
of 2011, one of the seven chronic excreters is known to con-
tinue to excrete polioviruses, one has stopped excretion spon-
taneously, four have died, and for one, the excretion status is
unknown. None of the viruses examined from these carriers
showed recombination with other nonpolio enteroviruses. The
absence of recombination may suggest that the poliovirus had
replicated in a single person rather than through person-to-
person transmission in the community (see “Vaccine-derived
poliovirus”).158–167
Between 1976 and 1995, 48 outbreaks involving approxi-
mately 17,000 cases of paralytic poliomyelitis were reported in
the literature.170 These outbreaks involved primarily unvaccinated
or inadequately vaccinated subgroups and were caused predomi-
nantly by poliovirus type 1 (74%). On the basis of this review,
cases in developing countries occurred mostly among children
younger than 2 years, whereas cases in industrialized countries
tended to occur in older people who had remained susceptible
to poliomyelitis. More recently, as an expression of the chang-
ing epidemiology due to progress of polio eradication efforts,
outbreaks in Albania (1996),171 Namibia (2006),172 Cape Verde
(2000),173 Tajikistan (2010),174 and Congo-Brazzaville (2010)174,175
have involved a substantial proportion of cases in adolescents and
young adults associated with a high case-fatality rate.
Besides age and being unvaccinated or inadequately vacci-
nated, several factors have been shown to increase the risk of
acquiring paralytic manifestations, including intramuscular
injections with diphtheria and tetanus toxoids and pertussis
vaccine (DTP)176,177 or antibiotics,178,179 strenuous exercise,180–182
injury such as fractures, and pregnancy.183 Provocation polio-
myelitis describes the enhanced risk of paralytic manifesta-
tions that follows injection in the 30 days preceding paralysis
onset. Nathanson and Bodian184,185 demonstrated more than
40 years ago that retrograde axonal transport is responsible for
the poliovirus invasion of the CNS in provocation poliomy-
elitis. More recently, Gromeier and Wimmer85 and Gromeier
et al186 suggested that the temporary expression of the human
PVR on peripheral neurons during the repair process of injured
nerves may enhance poliovirus access into peripheral neu-
rons. Retrograde transport in the axon via the fast system
seems to shorten further the period from initial access of the
poliovirus to the peripheral neuron to the virus reaching the
motor neuron cells of the CNS.187 This limits the time during
which the immune system could develop an effective response.
Aggravation poliomyelitis describes the elevated risk of paralytic
disease that follows strenuous exercise shortly before paralysis
onset (in the preceding 24-48 hours).
Removal of tonsils and adenoids predisposes to bulbar polio-
myelitis.188 Clinical observations on this fact were reported
in the early 20th century.188,189 Rhesus monkeys, when inoc-
ulated with poliovirus in the tonsillopharyngeal region, devel-
oped poliomyelitis with greater frequency than when they were
inoculated by other routes.4 Later, von Magnus and Melnick190
demonstrated that, if cynomolgus monkeys were given
poliovirus by the oral route, their susceptibility was greatly
enhanced in animals that had recently removed tonsils. Ogra191
and Ogra and Karzon192 studied 40 children before and after
removal of tonsils and adenoids. The children ranged from 3
to 11 years old and had been immunized with live attenuated
poliovirus vaccine 6 months to 6 years previously. Before ton-
sillectomy, IgA poliovirus antibody was present in appreciable
titers in the nasopharynx of all children, but no IgM or IgG
antibody was detectable. Significantly, however, after tonsillec-
tomy, the preexisting IgA poliovirus antibody level in the naso-
pharynx sharply declined in all children studied. Mean antibody
titers decreased threefold to fourfold. Thus, removal of tonsils
may eliminate a valuable source of immunocompetent tissue
particularly important in conferring resistance to poliovirus.
Lower socioeconomic status has been shown to be a risk
for paralytic poliomyelitis in developing countries,193 probably
because children belonging to the lower socioeconomic group
experience more intense exposure to poliovirus (ie, a higher
virus inoculum, which has been shown in experimental stud-
ies to be a risk factor for paralytic disease30). In addition, the
children are also at higher risk for primary vaccine failure
after OPV because of more frequent concurrent enterovirus
infections.193–196
In a study of twins, concordance with regard to paralytic
poliomyelitis was found in 36% of monozygous pairs compared
with 6% of dizygous pairs.197 The authors concluded that the
data were consistent with “the theory that susceptibility may be
conditioned by the homozygous state of a recessive gene.”197 An
HLA complex study suggested that HLA-encoded genetic fac-
tors control resistance to the paralytic form of poliomyelitis.198
Data on genetic susceptibility to poliomyelitis were reviewed by
Wyatt,199 who proposed that multiple-linked genes determine
whether an infection with poliovirus results in paralytic disease.
The case-fatality rate is variable and depends primarily on
the age groups affected. The highest case-fatality rates have been
reported from epidemic cases in the early 20th century16,30,32
and among adolescents and young adults, but are commonly
between 5% and 10%.30,32 Even in the 1990s, the case-fatality
rate could be high, as occurred in a large outbreak of poliomy-
elitis in Albania in 1996, which reported a case-fatality rate of
10%.171 More recently, an outbreak of poliomyelitis occurred on
the Cape Verde Islands in 2000. Of the 33 reported cases, 7 peo-
ple died, for a case-fatality rate of 21%. The case-fatality rates
were 0%, 20%, and 57% among cases ages in person younger
than 5 years, 5 to 14 years, and 15 years or older, respectively.173
In 2011, a large outbreak of poliomyelitis due to poliovirus type
1 centered in Pointe Noir, western part of Congo-Brazzaville,
involved a substantial proportion of young adult men. As might
be expected with this age distribution, the corresponding case-
fatality rate was high ( 30%). 175,200
Apes, such as chimpanzees, gorillas, and orangutans, are
susceptible to poliovirus and can experience paralytic disease
after poliovirus infection; outbreaks of poliomyelitis have been
reported in captivity and in the wild.201–203 It is unlikely that
they have any role in the sustained transmission of this virus204
because of the limited size of the ape populations. Most mon-
keys cannot be infected by oral administration of poliovirus and
would not be expected to participate in the chain of transmission.
In short, there is no significant animal reservoir for poliovirus.204
Results from mathematical modeling suggest that the force
of poliovirus infection, measured primarily by the average age at
infection among populations in the prevaccine era, is substan-
tially higher in developing countries compared with industrial-
ized countries. For example, the basic reproductive number (a
measure of infectivity) of wild poliovirus is between 3 and 5 in
the United States, which means that, on average, an infected
person introduced into a fully susceptible population would
transmit the poliovirus to three to five contacts. In contrast,
Summary of Immunodeficient Persons Excreting Poliovirus for 6 Months or Longer, 1962-2009

Poliovirus vaccine—live
28
607
 608
Summary of Immunodeficient Persons Excreting Poliovirus for 6 Months or Longer, 1962-2009—cont'd

SECTION TWO
AGG, agammaglobulinemia; CDC, Centers for Disease Control and Prevention; CVID, common variable immunodeficiency disorder; HGG, hypogammaglobulinemia; ICF, immunodeficiency, centromeric region instability, facial
anomalies; MHC class II, major histocompatibility complex class II immunodeficiency disorder; NA, not available; ND, not done; SCID, severe combined immunodeficiency disorder; UK, United Kingdom; US, United States; WHO,
World Health Organization; WPRO, Western Pacific Regional Office; XLA, X-linked agammaglobulinemia.
*Year of diagnosis.
†
Age at disease onset.

the average infected person in a developing tropical setting
would be expected to have transmitted the infection to 10 to 12
contacts.205,206 As population immunity increases and many of
the contacts of an infected person are no longer susceptible, the
number of transmissions decreases. When the reproductive rate
is less than 1 because of high population immunity, transmis-
sion will eventually cease. The herd immunity threshold is the
level of immunity in the population (eg, in the “herd”) at which
an infected person, on average, would transmit the infection to
less than one susceptible contact.
Whereas poliomyelitis outbreaks in industrialized countries
can be prevented with overall population immunity levels of
approximately 66% to 80%, outbreaks in developing countries
with inadequate sanitation and hygiene could still occur with
immunity levels as high as 94% to 97% (Figure 28-2). These
findings may be helpful in explaining why there was no spread
to the general population after the outbreaks in the Netherlands,
Canada, and the United States207,208 and why widespread trans-
mission among well-vaccinated populations occurred in many
Herd immunity threshold levels for selected industrialized
and developing countries, based on basic reproductive rate, or R0.
Threshold values for herd immunity were calculated using 1 (1/R 0),
where R0 is 1 (life expectancy/average age at infection with poliovirus).
Herd immunity threshold values are shown by the dashed line. The
solid bars are the basic reproductive rate in a given population. (From
Patriarca PA, Sutter RW, Oostvogel PM. Outbreaks of paralytic poliomyelitis, 1976-
1995. J Infect Dis 175[suppl 1]:S165-S172, 1997, with permission.)
Reported cases of poliomyelitis, United States, 1920 to 1962. (Centers for Disease Control and Prevention.)
Poliovirus vaccine—live 28 609
outbreaks in developing countries.209,210 A seroprevalence survey
of poliovirus antibodies during the final stage of polio eradica-
tion in Egypt suggested that between 97% and 100% immunity
against poliovirus type 1 may have been needed to interrupt
type 1 transmission in a country with almost ideal conditions
for poliovirus circulation.211 Recent seroprevalence surveys
in Northern India, conducted in the 2009-2011 period, after
introduction and widespread use of bOPV (types 1 and 3), have
demonstrated high seroprevalence levels of antibodies against
poliovirus type 1 (98%-99%) among infants 6 to 7 months old
but considerably lower levels against poliovirus types 2 and 3.
Epidemiologic patterns and incidence of poliomyelitis
The epidemiology of poliomyelitis changed substantially dur-
ing the last century. Three epidemiologic patterns have been
observed: (1) endemic, (2) epidemic (prevaccine), and (3) vac-
cine era. Polioviruses probably circulated in an uninterrupted
endemic manner for many centuries, infecting new cohorts of
susceptible infants continuously, almost all early in life, when
maternally derived antibody transferred from mother to new-
born still provided some protection.
A change from endemic transmission to periodic epidem-
ics was first observed in some temperate-climate countries
(eg, Norway, Sweden, the United States) late in the 19th cen-
tury and at the beginning of the 20th century.12–15 The delay
in median age of poliovirus exposure permitted the accumula-
tion of sufficient children susceptible to poliomyelitis to per-
mit periodic outbreaks. In the United States, the median age
of poliovirus infection increased from younger than 5 years at
the beginning of the century to 5 to 9 years in the 1940s, before
poliovirus vaccine licensure.32 In contrast, approximately 80%
of the cases were in children younger than 5 years during the
large epidemic in New York in 1916.16 The generally accepted
explanation, supported by numerous studies, is that, in a
temperate-zone climate with increased economic development
and correspondingly improved resources for community san-
itation and household hygiene, exposure to polioviruses was
postponed to later in life. Epidemic transmission became the
primary epidemiologic pattern in temperate-climate coun-
tries, such as the industrialized countries in Europe and North
America, until poliomyelitis was brought under control after
introduction of effective vaccines (Figure 28-3).
In developing countries, particularly tropical areas, an
endemic epidemiologic pattern predominated until recently.
610 SECTION TWO Licensed vaccines
Poliovirus exposure occurred early in life. Although earlier theo-
ries suggested that paralytic poliomyelitis was not a health bur-
den in tropical countries because of early exposure of infants
to virus at a time when levels of maternally derived antibodies
protected them from paralytic disease, more recent studies have
disproved these theories. The history and scientific evidence for
this misconception—that poliomyelitis was not a significant
public health problem in developing countries—was reviewed
in detail in the corresponding chapter of the second edition of
Vaccines.212 In the last three decades, a series of lameness sur-
veys were conducted in many developing countries that reported
between 5 and 10 lameness cases per 1,000 children in the age
group studied,157 suggesting that approximately 1 in 100 to 1 in
200 children acquire paralytic disease attributable to poliovirus.
The WHO estimates that, in the absence of vaccination, at least
1 of every 200 children would become paralyzed by poliovirus
and there would be approximately 650,000 cases of paralytic
disease annually, the great majority of which would occur in
children from developing countries. With improving vaccina-
tion coverage, a shift from an endemic to an epidemic pattern of
poliovirus transmission has been observed in some developing
countries that experienced large epidemics.170,210,213–215
The vaccine era began in the United States and in many
European countries, Canada, Australia, New Zealand, and
Japan after introduction of IPV in 1955.2 The incidence of para-
lytic poliomyelitis decreased rapidly from 18,308 reported cases
in the United States in 1954, the year immediately preceding
IPV licensure, to 2,499 cases in 1957, a decline of 86% only
3 years after the availability and widespread use of IPV. The rela-
tive upswing in reported cases in 1959 (6,289 cases), with many
cases having a history of receiving several prior doses of IPV,
raised concerns regarding the clinical efficacy of IPV in prevent-
ing paralytic disease.216 Nevertheless, continued and acceler-
ated IPV use decreased the incidence of paralytic poliomyelitis
to nearly record low levels (2,525 cases) in the United States by
1960. Widespread use of IPV in other countries was followed
by substantial decreases in the incidence of poliomyelitis and,
in some European countries, including Finland, Iceland, the
Netherlands, and Sweden, resulted in the apparent elimination
of indigenous wild poliovirus transmission.217,218
The OPV era started in the United States with licensure
of mOPV in 1961, followed by licensure of tOPV in 1963.28
Although live attenuated oral poliovirus vaccine was developed
in the United States, the first large-scale production and the
large field trials that proved the safety and efficacy of the vaccine
took place in the former Soviet Union. A mass immunization
program was initiated in the Soviet Union in 1959 and com-
pleted in 1960, covering 77.5 million people, or 36.7% of the
entire population. The immunization campaign was followed by
a sharp decrease in the incidence of poliomyelitis, from 10.6 per
100,000 population in 1958 to 0.43 per 100,000 population in
1963. Between 1964 and 1979, the incidence remained at a level
of 0.01 to 0.1 per 100,000 population.219 Similar declines in the
incidence of poliomyelitis were observed in other European coun-
tries, Australia, New Zealand, Canada, and the United States
after the introduction of OPV. In the United States, mOPV was
administered initially in mass vaccination campaigns in 1962,
called Sabin Oral Saturdays/Sundays (SOS), followed by a rou-
tine vaccination program that administered vaccine to infants
year-round.28,220,221 The impact of administering OPV to a popu-
lation that already had high immunity levels generated by pre-
vious natural infection or vaccination with IPV was impressive.
Substantial reductions in the reported number of poliomyeli-
tis cases were observed from 988 cases in 1961 to 61 cases in
1965. In 1973, only seven cases of poliomyelitis were reported.
Epidemic poliomyelitis also was brought under control, with the
last outbreak in the general population occurring in Texas along
the US-Mexico border in 1970, followed by small outbreaks in
1972 and 1979 among religious groups whose members object
to vaccination.208 The last indigenously acquired case of polio-
myelitis due to wild poliovirus was detected in 1979. Between
1985 and 2000, aside from three imported cases of poliomyeli-
tis (the most recent was reported in 1993), all cases have been
vaccine-associated.222 Rarely has a serious disease been con-
trolled as rapidly and dramatically as has poliomyelitis in the
United States and other industrialized countries.
The history of controlling poliomyelitis in many developing,
particularly tropical, countries has been more recent. There is a
notable exception: Cuba seems to have interrupted wild poliovi-
rus after two rounds of mass vaccination campaigns in 1962.223
In many other developing countries, however, national vaccina-
tion programs were not operational until the late 1970s and early
1980s, and global OPV coverage with three doses among chil-
dren age 1 year only reached 80% by 1990.224 Wherever moder-
ately high levels of OPV coverage were achieved, the incidence
of poliomyelitis decreased by more than would be expected,225
but endemic transmission of polioviruses continued, and cases of
poliomyelitis continued to be reported. In addition to achieving
high routine coverage with three doses of OPV, control of polio-
myelitis required additional supplemental doses of OPV that were
incorporated into the routine vaccination schedule in some coun-
tries; other countries needed to administer supplemental doses of
OPV in mass campaigns. For example, in Brazil, control of polio-
myelitis could not be accomplished until mass vaccination cam-
paigns were initiated in 1980 (Figure 28-4). The impact of these
mass campaigns on poliomyelitis incidence was dramatic; the
number of reported cases decreased from 1,290 in 1980 to 122
in 1981, a decrease of more than 90%.226 Based on the experience
of the GPEI, once a poliovirus serotype has not been detected
for 6 months or longer, it seems that circulation has been inter-
rupted in that geographic location. However, while this may be
valid in general, the length of “silent circulation” (without detec-
tion of overt cases of paralytic disease) is very much dependent
on surveillance sensitivity and serotype. For example, for type 3,
because of the lower case-to-infection ratio, a minimum period of
at least 12 months may be needed, even under conditions of ade-
quate surveillance. For certification purposes, a minimum period
of 3 years or longer is anticipated to ensure confidence.227
Significance as a public health problem
In the absence of effective control programs with poliovirus vac-
cine, paralysis develops in approximately 1 of every 200 children
(see “General epidemiology”) after exposure to polioviruses,157
Cases of poliomyelitis by 4-week periods in Brazil, 1975 to
1982. Arrows indicate national immunization days. (From Risi JB. The control
of poliomyelitis in Brazil. Rev Infect Dis 6[suppl 2]:S400-S403, 1984, with permission.)

followed in most cases by permanent disability; 5% to 10% of
patients with paralytic disease have a fatal outcome.30,31 Thus,
with a global birth cohort of approximately 130 million surviv-
ing infants in 2009, approximately 650,000 children would be
expected to acquire paralytic poliomyelitis resulting in permanent
disability each year, and between 32,500 and 65,000 of the cases
would result in poliomyelitis-associated deaths. In the United
States, a report estimated that, in the absence of a control pro-
gram, more than $3 billion ($926 million in direct costs and $2.1
billion in indirect costs) would be required each year to cover the
treatment and other related costs of patients with poliomyelitis.228
A recent study modeling different control scenarios suggested
these scenarios will be more costly in the long run than finishing
the eradication effort.229 In addition to the acute manifestations
of poliomyelitis, patients may experience postpolio syndrome
decades after the acute episode; postpolio syndrome is associated
with new muscle pain, exacerbation of existing muscle weak-
ness, or the development of new weakness or paralysis52 that may
require additional therapy, rehabilitation, and respiratory support.
Despite the availability of two highly effective vaccines, polio-
myelitis still exerts a significant public health impact in the world.
In the United States in the prevaccine era, the peak incidence year
of poliomyelitis was 1952, when 57,879 cases of poliomyelitis
were reported (including 21,269 cases of paralytic disease).230 After
the availability and widespread use of poliovirus vaccines begin-
ning in 1955, poliomyelitis was rapidly controlled in industrialized
countries and in other areas where vaccines were used effectively.
Globally, the Expanded Programme on Immunization, a program of
the WHO established in 1974, provided leadership to national pro-
grams in virtually all developing countries to improve vaccination
coverage. Coverage levels apparently reached 80% with three doses
of OPV among children 1 year of age for the first time in 1990,224
resulting in substantial decreases in the global morbidity and mor-
tality burden of poliomyelitis. Despite this success, the WHO esti-
mated that approximately 350,000 cases of paralytic poliomyelitis
associated with permanent disability occurred in 1988, the year the
global polio eradication target was adopted.37 Because of rapid prog-
ress toward polio eradication, the worldwide reported incidence of
poliomyelitis was only 494 cases in 200139 (and the number of cases
associated with wild poliovirus isolation was 483). Between 2005
and 2010, the number of poliomyelitis cases globally has fluctuated
between 1,293 and 1,997, and the number of countries reporting
any wild poliovirus varied between 20 and 23. During this same
period, the reporting completeness has further improved39,231; thus,
it is unlikely that many cases of poliomyelitis remained undetected
in 2010 (and even more unlikely that significant areas of poliovirus
circulation were missed, but there are areas not accessible owing
to conflict and populations that are difficult to survey, such as
nomads). As of 13 March 2012, a total of 650 virologically and epi-
demiologically confirmed poliomyelitis cases had been reported to
the WHO for 2011.39 In 2010, 1413 virologically and epidemiologi-
cally confirmed poliomyelitis were reported to WHO. Of these, 266
(19%) were reported from polio-endemic countries (Afghanistan,
India, Nigeria, Pakistan); the rest, 1,147 (81%) were from reestab-
lished poliovirus transmission countries (Angola, Chad, Democratic
Republic of the Congo, Sudan) and other countries with imported
poliovirus, particularly from the large outbreaks in Tajikistan174 and
Congo-Brazzaville.175,200 A total of 20 countries reported virologi-
cally confirmed poliomyelitis cases in 2010.39,232
Paralytic poliomyelitis continues to be a serious, albeit declin-
ing, threat to children in polio-endemic countries and occasionally
to people residing in industrialized countries that have primar-
ily achieved good control of poliomyelitis for many years.233,234
Even in countries with well-vaccinated populations that have
eliminated indigenous wild poliovirus circulation for decades,
gaps in population immunity may persist, particularly in groups
objecting to vaccination (eg, religious groups such as the Amish
in the United States, the Netherlands Reformatory Church in
the Netherlands, and related groups in Canada) or groups that
Poliovirus vaccine—live 28 611
are not reached effectively by national vaccination programs (eg,
Roma, previously referred to as “gypsies”).207,235–237 In the United
States, the last two outbreaks of poliomyelitis occurred in 1972
and 1979 among members of religious groups objecting to vac-
cination.208 The 1979 outbreak was an extension of an outbreak
affecting first the Netherlands in 1978 and then Canada.207,236
An outbreak of poliomyelitis affecting the same religious group
in the Netherlands also occurred in 1992 to 1993.234 On the basis
of genomic sequence,111,112 the poliovirus type 1 strain causing
the epidemic in the Netherlands in 1978 had its origin in Turkey.
The recent outbreak was due to poliovirus type 3, which was
most likely imported from the Indian subcontinent.238 In Spain,
the last cases of poliomyelitis in 1980 to 1981 were detected
among Roma children.235 In both outbreaks in Bulgaria in 1990
to 1991 and in 2000,239,240 as well as in the Romanian outbreak in
1991 to 1992,241 Roma children were exclusively affected or con-
stituted a substantial proportion of poliomyelitis cases. Although
no cases of paralytic poliomyelitis were detected in the outbreaks
in the United States (1979), Canada (1978), and the Netherlands
(1978 and 1992-1993) beyond the affected unvaccinated or inad-
equately vaccinated subpopulations, wild poliovirus exposure of
people in the general population and the establishment of sub-
sequent endemic and epidemic transmission remain a concern.
Molecular epidemiology of poliovirus
The application of molecular tools, such as genomic sequencing of
poliovirus, has added a new dimension and resolving power to our
understanding of the epidemiology of poliomyelitis.116,242 Because
the wild poliovirus genome evolves rapidly (typically a little over
1% nucleotide substitutions per site per year)243–245 (Table 28-5),
links between poliomyelitis cases can be determined and importa-
tions from the remaining poliovirus reservoirs can be established.
These molecular methods offer an additional tool to monitor
the progress of the GPEI and suggest that lineages of poliovirus
genotypes (differing by 15% in their nucleotide sequences) dis-
appear sequentially through intensive immunization efforts.116,246
The experience in the Americas suggests that, if a genotype is
not detected for a year or more despite adequate surveillance, it
probably has become extinct.247
These methods have established the existence of numerous
poliovirus genotypes endemic to different regions of the world,116
the former Soviet Union,109 Europe, the Middle East, and the
Indian subcontinent,248 and demonstrated that poliovirus type
2 is usually the first serotype to be eliminated,249 that poliovi-
rus type 3 seems to circulate more locally than other serotypes,
and that poliovirus type 1 seems to be most commonly asso-
ciated with importations from neighboring countries and with
intercontinental or global spread of the virus.246,247,250 Genomic
sequencing of polioviruses suggested that the viruses respon-
sible for the epidemic in the Netherlands in 1992 to 1993 (type
3) and Albania in 1996 (type 1), as well as that in Bulgaria and
Georgia in 2001 (type 1),250 were probably imported from reser-
voirs in the Indian subcontinent.248
The molecular methods also have shown that, in some
instances, different genotypes of poliovirus can circulate con-
comitantly and cause poliomyelitis cases in a geographically lim-
ited area.214,247 More recently, molecular epidemiologic tools have
documented the elimination of local lineages in Uttar Pradesh
and Bihar, India, followed by spread of wild poliovirus from the
remaining domestic reservoirs to polio-free Indian states and
Nepal,251 exportation of wild poliovirus types 1 and 3 from India
to Angola with its subsequent spread to other countries in South
Central Africa,252,253 and exportation of wild poliovirus type 1
from Uttar Pradesh into Central Asia and Russia.253 Similarly, the
repeated spread of wild poliovirus from insecure reservoir areas in
northwestern and southwestern Pakistan–southern Afghanistan
to other provinces of Pakistan and Afghanistan has been continu-
ously monitored.254 In 2011, China experienced an importation of

Poliovirus Type 2, Sabin Strain, Passage History*
MKTC, SO + 1 = SOM
MKTC, monkey kidney tissue culture; MSD, Merck Sharp & Dohme.
*Data from WHO Consultative Group on Poliomyelitis Vaccines, 1985.
Poliovirus Type 3, Sabin Strain, Passage History*
SO + 1 = SOM
MKTC, monkey kidney tissue culture; MSD, Merck Sharp & Dohme.
*Data from WHO Consultative Group on Poliomyelitis Vaccines, 1985.
Poliovirus Type 3, Sabin Strain, RNA-Derived,
Passage History
MKTC, monkey kidney tissue culture. SO, Sabin Original.
*Data from WHO Consultative Group on Poliomyelitis Vaccines, 1985.
Poliovirus vaccine—live 28 615
through terminal dilutions and single-plaque passages, carefully
selecting by neurovirulence testing, and finally obtaining strain
LS-c, 2ab. Two further passes, in cynomolgus kidney, yielded
LS-c, 2ab/KP2, designated SO (Sabin original). Bettylee Hampil,
at Merck Sharp & Dohme, made one additional passage in
rhesus monkey kidney tissue culture to derive LS-c, 2ab/KP3,
designated SO + 1, or SOM. The current vaccine is SO + 4,
four tissue culture passages beyond the SO. A maximum of five
passages is permitted, after which earlier (grandmother) seeds
must be thawed and used to prepare new mother seeds. Because
of inherent difficulties in maintaining the genetic stability of
the Sabin type 3 seed stock, manufacturers have turned to an
RNA-derived passage and clone of the strain, labeled SOR. This
seed has yielded a vaccine of greater consistency and stability
than the original Sabin seed. Grandmother seeds are potentially
available through the WHO.
The genetic basis for the attenuation of the Sabin OPV strains
has been intensively investigated. Two basic approaches have
been taken: (1) sequence comparisons between the Sabin strains
and their neurovirulent wild parents (types 1 and 3) or neuro-
virulent revertants (types 2 and 3) obtained from patients with
VAPP; and (2) investigation of the contribution of specific nucle-
otide substitutions to attenuation using infectious complemen-
tary DNA (cDNA) clones (see also “Genetic stability of vaccine
seed strains”). The 57 nucleotide substitutions distinguishing the
Sabin 1 strain from its neurovirulent parent, Mahoney/USA41,
are scattered throughout the genome.303 Of the 57 nucleotide
substitutions, 6 mapped to the 5 -untranslated region (UTR), 49
to the coding region (21 of which encoded amino acid substitu-
tions), and 2 to the 3 -UTR. Infectious cDNA constructs con-
taining different combinations of Sabin 1 and Mahoney/USA41
sequences were tested for neurovirulence in monkeys or trans-
genic mice for temperature sensitivity and for other phenotypic
properties distinguishing Sabin 1 from Mahoney/USA41.303–305
The single most important determinant for attenuation in
Sabin 1 was the A G substitution at position 480 (abbrevi-
ated A480G) in the 5 -UTR.306 Four other determinants mapped
to the capsid region (one in VP4, one in VP3, and two in VP1),
and a weak determinant of attenuation and temperature sensi-
tivity mapped to the RNA-dependent RNA polymerase encoded
by the 3D gene (Figure 28-6).304,305,307 Only two nucleotide sub-
stitutions (G481A in the 5 -UTR and C2909U encoding a threo-
nine isoleucine substitution at position 143 of VP1) seem to
be responsible for the attenuated phenotype of Sabin 2.308,309 The
total number of sequence differences between P712/USA54 and
Sabin 2 is uncertain. However, because P712/USA54 has inher-
ently low neurovirulence,310 identification of critical attenuating
sites in Sabin 2 involved determination of the effects of intro-
duction of sequences derived from a neurovirulent revertant of
Sabin 2 (obtained from a case of VAPP) into infectious cDNA
constructs derived from Sabin 2.308,309 Of the 10 nucleotide sub-
stitutions distinguishing Leon/USA37 and Sabin 3, only three
(C472U in the 5 -UTR, C2034U encoding a serine phenyl-
alanine substitution at position 91 of VP3, and U2493C encod-
ing an isoleucine threonine substitution at position 6 of VP1)
seem to be the main determinants of attenuation.302,311,312
As with Sabin 1,304–306 the most important determinants of
attenuation in Sabin 2 and 3 map to the 5 -UTR.302,306,308,309 The
other substitutions are thought to contribute to the stability
of the attenuated phenotype. The high stability of the attenu-
ated phenotype of Sabin 1 is attributed to the larger number of
attenuating substitutions and their relative contributions to the
phenotype. Quantitative determination of the contributions of
each substitution is complicated by several factors: (1) The role
of minor determinants of attenuation is difficult to measure. (2)
Some substitutions have pleiotropic effects on phenotype. (3)
Some Sabin strain phenotypes require a combination of substi-
tutions. (4) Second-site mutations can suppress the attenuated
phenotype in various ways. (5) The outcome of experimental
616 SECTION TWO

Location of principal attenuating nucleotide (lower bars) and amino acid (upper bars) substitutions in each of the three Sabin OPV
strains. Abbreviations of nucleotide residues: A, adenine; C, cytosine; G, guanine; U, uracil. Abbreviations for amino acid residues: A, alanine;
C, cysteine; F, phenylalanine; H, histidine; I, isoleucine; L, leucine; M, methionine; S, serine; T, threonine; Y, tyrosine. Substitutions are shown as
nonattenuated parent-position-Sabin strain; nucleotide positions are numbered consecutively from residue 1 of the RNA genome; and amino acid
positions are indicated by the abbreviated name for viral protein (4, VP4; 2, VP2; 3, VP3; 1, VP1; 3D, 3D-polymerase) and numbered consecutively
from residue 1 of each protein. For example, a guanine (Mahoney) uracil (Sabin 1) substitution at RNA position 935 (G935U) encodes an alanine
(Mahoney) serine (Sabin 1) replacement at residue 65 of VP4 (A4065S). The Y3D073H substitution in Sabin 1 and the S3091F substitution in
Sabin 3 are important determinants of temperature sensitivity. (Constructed from findings reported by Bouchard et al,304 Ren et al,309 Macadam et al,311 Tatem
et al,312 and Westrop et al.302)

neurovirulence tests may vary with the choice of experimental
animals (monkeys vs transgenic mice) or the route of injection
(intraspinal vs intracerebral). Examples of substitutions with
pleiotropic effects are the serine phenyalanine substitution at
position 91 of VP3, which confers attenuation and temperature
sensitivity to Sabin 3,313 and the tyrosine histidine substitu-
tion at position 73 of the 3D polymerase of Sabin 1, which is
an important determinant of temperature sensitivity304,307 and a
minor contributor to attenuation.307 The relationship between
poliovirus neurovirulence in experimental animals (where
virus is introduced directly into the CNS) and pathogenicity for
humans (where virus is introduced by ingestion) cannot be mea-
sured in controlled experiments and remains ambiguous. Clearly,
the Sabin OPV strains are several orders of magnitude less neuro-
virulent than wild polioviruses, as indicated by the very low inci-
dence of VAPP222 compared with the high incidence of paralytic
poliomyelitis in areas where wild polioviruses are circulating.
The attenuating substitutions of the Sabin strains are some-
times described as mutations impairing the specific determi-
nants of poliovirus neurovirulence. However, attenuation is a
specific phenotype of each Sabin strain. Poliovirus neuroviru-
lence, by contrast, is a much more complex property. Expression
of a neurovirulent phenotype requires the efficient function of
numerous steps in the natural life cycle of the virus and, thus,
the efficient expression of several viral genes.57,314 Impairment
in the expression of any of these poliovirus genes can reduce
replicative fitness and confer a more attenuated phenotype.314
However, such mutants may not be suitable candidates for live
virus vaccines. The Sabin strains contain specific genetic defects
that are not found among the highly diverse population of circu-
lating wild polioviruses. These exceptional defects are unstable
to replication in the human intestine, and variants with higher
replicative fitness are regularly selected. Other combinations of
substitutions can produce highly attenuated polioviruses poten-
tially suitable for vaccine use.315 However, the attenuating sub-
stitutions of the Sabin strains were shown to confer highly
favorable biologic properties, making these strains the best avail-
able candidates for licensure as oral poliovirus vaccines.
Functional basis for attenuation of the Sabin
OPV strains
The determination in the early 1980s of the complete genomes
of the Sabin strains and their neurovirulent parents (in the case
of types 1 and 3)63,303,316–318 and the availability of infectious
cDNA poliovirus clones302,305,319 opened the way for detailed
analysis of the biologic mechanisms for the attenuated phe-
notype of the Sabin OPV strains.320 The most detailed analy-
ses have been of the substitutions in the 5-UTR (A480G in
Sabin 1, G481A in Sabin 2, and C472U in Sabin 3) that are the
principal attenuating mutations of the respective Sabin strains
(Figure 28-6). These substitutions map to a specific domain of
RNA secondary structure (stem-loop region V) within the IRES
of the 5-UTR that is highly conserved among polioviruses and
related enteroviruses. The attenuating substitutions in the IRES
are not found in natural wild poliovirus isolates. The localiza-
tion of these important determinants to the IRES,321 the initia-
tion site for translation of the poliovirus polyprotein, suggested
that an important aspect of attenuation may involve a deficiency
in translation for the Sabin strains. Indeed, in vitro translation
experiments demonstrated that decreased efficiency in the ini-
tiation of translation was associated with the U472 substitution
in Sabin 3322,323 and the G480 substitution in Sabin 1.324,325 This
view was further supported by the finding that replication of the
Sabin strains was similar to that of neurovirulent polioviruses
in HeLa cells but was differentially reduced in neuroblastoma
cells.326,327 The reduced Sabin strain yields were associated with
lower efficiencies of translation in neuroblastoma cells.325–327
A possible molecular mechanism for the translational defi-
cit involves the polypyrimidine tract binding protein (PTB), a
cellular protein that interacts with the 5-UTR and facilitates
poliovirus translation.328 The attenuating mutations weaken
the interaction of PTB with the 5-UTR in neuroblastoma cells
but not in HeLa cells.328 Intact spinal cords from chick embryos
were found to have reduced levels of PTB, and translation of
Sabin 3 in the neuronal cells was less efficient than that of neu-
rovirulent parental Leon strain.329

Despite the strong experimental support for the hypothesis
that a preferential defect in translation in motor neurons is a
major contributor to the attenuated phenotype of the Sabin
strains, this view has been challenged by recent studies in mice
in which translation from the Sabin 3 IRES was equivalently
reduced in cells of neuronal and nonneuronal origin.330 It was
suggested that attenuation is determined after internal ribo-
some entry, that the Sabin strains have a reduced fitness for
replication in all cells, and that reduced efficiency of replication
in intestinal cells may permit the development of an immune
response before sufficient numbers of virus reach the spinal
cord and brain.330
The major attenuating determinants in the IRES are the
ones that have been most intensively studied, particularly that
of Sabin 3 because substitutions at position 472 have more pro-
nounced effects on the attenuated phenotype than do substi-
tutions at position 480 in Sabin 1 or 481 in Sabin 2. Results
of similar studies with Sabin 1 and Sabin 2 have usually been
in agreement with the Sabin 3 findings, but the precise contri-
bution of the G481A substitution to the attenuation of Sabin
2 is less clear.308,309,331 The capsid determinants may affect the
attenuated phenotype in various ways such as by interfering
with virus assembly,313 altering interactions with the poliovirus
receptor,304 or reducing capsid stabilities.71 It has been empha-
sized that the attenuated phenotype is composite, involving the
complex interaction of multiple determinants.332
How trivalent vaccine was developed
Viral replication in the gastrointestinal tract and seroconversion
were usually demonstrated in 80% to 100% of seronegative recip-
ients with a single dose of monovalent vaccines at dosage lev-
els of 105 median tissue culture infective doses (TCID50).333–337
However, when doses of 105 TCID50 of each poliovirus serotype
were mixed and administered as trivalent preparations, the repli-
cation and antibody production were consistently lower for some
types compared with the sequential administration of monova-
lent vaccines.338–344 This effect could be modified somewhat by
increasing the doses of each type ( 107 TCID50).345,346 In addi-
tion, these studies showed that tOPV of similar potency for each
serotype was associated with a predominance of poliovirus type
2 excretion and significantly higher type 2 antibody titers than
for poliovirus types 1 and 3. These early trials did not evaluate
the impact of increasing the quantity of one serotype or reducing
the quantity of another on seroconversion to all three serotypes
because the interference effect of type 2 often could be overcome
by administering three or more doses of the trivalent vaccine.
In 1961, a large study in Canada tested a “balanced” formula-
tion of tOPV (106 TCID50 for Sabin type 1, 105 TCID50 for Sabin
type 2, and 105.5 TCID50 for Sabin type 3).347 A single dose of this
balanced (10:1:3) vaccine was administered to nearly 24,000
people, including 106 previously seronegative subjects, 103
(97%) of whom seroconverted to all three serotypes. Although
one could conclude from this study that a single dose of OPV
may be sufficiently immunogenic for a routine program, only
triple-seronegative infants were included in this analysis of the
Canadian trial, so the results represent the best possible scenario
for inducing optimal levels of seroconversion because the infants
lacked maternally derived antibody that can interfere with
seroconversion. On the basis of these findings and an unpub-
lished study from Guam, the balanced formulation of OPV was
licensed in Canada in 1962 and in the United States in 1963.
Dosage and route
In the United States, before discontinuation of OPV in 2000,
the trivalent live attenuated OPV was formulated to contain at
least 105.5 TCID50 for poliovirus type 1, 104.5 TCID50 for type 2,
and 105.2 TCID50 for type 3.348,349 However, the US manufacturer
Poliovirus vaccine—live 28 617
routinely exceeded these minimum requirements, and an
evaluation by a WHO reference laboratory found potency levels
of 106.5, 105.4, and 106.3 TCID50 per dose of types 1, 2, and 3,
respectively (WHO unpublished data). The WHO requires the
following minimum TCID50 values for each vaccine poliovirus
serotype: 105.9 0.5 TCID50 for type 1, 105.0 0.5 TCID50 for
type 2, and 105.7 0.5 TCID50 for type 3. However, because of evi-
dence from an evaluation in Brazil that poliovirus type 3 immu-
nogenicity is not satisfactory in a trivalent formulation with
105.5 (300,000) TCID50, particularly in tropical countries,350 the
WHO's Global Advisory Group recommended in 1990 that the
type 3 component of OPV should be increased to 105.8 (600,000)
TCID50.351 The OPV purchased by the United Nations Children's
Fund (UNICEF) beginning with the 1992 to 1993 tender period
included this recommendation.
The recommended route of administration for OPV is oral,
by releasing the vaccine volume (0.5 mL) contained in single-
dose droppers into the oral cavity352 for vaccine manufactured
in the United States (before 2000, when OPV production was
discontinued) or by providing two drops ( 0.1 mL) of OPV con-
tained in multidose vials produced by many non-US manufac-
turers.353 Although in the early 1960s some manufacturers put
OPV into dragées, at present, only OPV produced in China is in
dragée form; before administration, the dragée must be ground
up and mixed with water.
Producers
At least 10 manufacturers around the world are producing OPV
using the Sabin vaccine seeds (Table 28-10). The list includes
producers in Belgium, China, France, Indonesia, Iran, Italy,
Japan, Mexico, Russia, and Vietnam. In addition, manufactur-
ers in Brazil, Egypt, India (several companies), Pakistan, and
Thailand have the capacity for filling and finishing bulk OPV
produced by one of the primary manufacturers.
Most of these manufacturers use seed strains of types 1 and
2 no more than two passages away from the WHO master seed
(SO + 1, ie, Sabin original plus one passage). There is more
variation in the type 3 seed used; most manufacturers are now
using the Pfizer RNA-derived seed (SOR + 1). The type 3 seed
used by one manufacturer in China, the Zhong III strain, has
been shown by oligonucleotide fingerprinting to be indirectly
derived from Sabin type 3.354 Vero cells and human diploid cells
are used by at least one manufacturer each for growing their
vaccine viruses; the others continue to use primary monkey
kidney cells. It is recommended that the cells for cultivation
be taken from monkeys bred in captivity. Like the cell cultures
used, the monkey colony should be shown to be free of extrane-
ous viruses and other pathogens.
Preparations available (including combinations)
Although mOPV formulations of the three poliovirus sero-
types were used widely in the early 1960s, they were replaced
by tOPV starting in 1963 in the United States and other coun-
tries. Exceptions include Hungary, which used all three types
of mOPV sequentially until the early 1990s,355 and South
Africa, which routinely used monovalent type 1 OPV (mOPV1)
in its routine immunization program until the early 1990s.356
The mOPV formulations, although used extensively in the
early 1960s, were no longer licensed in 2004. To improve the
immunogenicity of OPV in developing countries, the Advisory
Committee for Poliomyelitis Eradication, the principal tech-
nical oversight committee for the GPEI at the time, recom-
mended in 2004 the rapid development and large-scale use of
mOPV1.357 Currently, seven manufacturers and fillers (fillers
procure bulk OPV from an established manufacturer with bulk
production and then blend, fill, and release the product) have
developed and received regulatory approval for use of mOPV1
 618
Manufacturers and Fillers of Oral Poliovirus Vaccine, 2011*

SECTION TWO Licensed vaccines

bOPV, bivalent oral poliovirus vaccine; mOPV, monovalent oral poliovirus vaccine; tOPV, trivalent oral poliovirus vaccine.
*Since 2007, four manufacturers have discontinued production: Chiron Behringwerke, Marburg, Germany; Evans Medical Ltd, Medeva-Speke, Liverpool, England; TORLAK, Belgrade, Serbia; and Wyeth-Lederle Vaccines and
Pediatrics, Pearl River, NY). , products that are World Health Organization–prequalified for UNICEF procurement on behalf of the Global Polio Eradication Initiative.
†
For the fillers, the cell substrate depends on the bulk vaccine supplier.

administered at an earlier age (birth, 6, 10, and 14 weeks), an age
when the influence of maternally derived antibody on vaccine
virus replication and, hence, seroconversion is more distinct.209
Increasing the potency of OPV can correct some of these limi-
tations of the currently formulated OPV;402,407 this also can be
achieved by administering additional doses of OPV in the rou-
tine program or through mass campaigns.94
Monovalent or bivalent OPV can overcome some of the limita-
tions of tOPV, including the interference among the three Sabin
strains. Because the type 2 component of the tOPV is considerably
more immunogenic than the other components, vaccinees usually
seroconvert to type 2 first (because of interference with types 1 and 3
seroconversion). In contrast, with monovalent vaccines, a type-spe-
cific immune response can be targeted according to programmatic
needs. Monovalent OPV has been shown to be more immuno-
genic than tOPV in inducing a type-specific immune response in
industrialized and developing countries. A comprehensive review of
experiences with mOPV demonstrated that following a single dose,
median seroconversion rates in industrialized countries of 95%
(range, 90%-100%) to poliovirus type 1, 98% (range, 83%-100%) to
poliovirus type 2, and 94% (range, 70%-100%) to poliovirus type 3
were reported. In developing countries, the median seroconversion
rates following a single dose of mOPV were lower, 81% (range, 53%-
89%) to poliovirus type 1, 89% (range, 77%-93%) to poliovirus type
2, and 72% (range, 52%-80%) to poliovirus type 3.337 Considerable
variation in immunogenicity of OPV has been documented in
developing tropical countries, ranging from poor (eg, Northern
India) to excellent (eg, Indonesia) immunogenicity.408,409
Monovalent OPV became the vaccine of choice for supple-
mental immunization campaigns between 2005 and 2009 in
polio-endemic or polio-epidemic areas, including in previously
polio-free countries with recent importation of wild poliovirus.
The vast majority of these vaccines are formulated as mOPV1
corresponding to the programmatic need to control wild poliovi-
rus type 1. A total of more than 1 billion doses of mOPV1 were
administered during this period. A case-control study reported
recently that the efficacy of mOPV1 was approximately threefold
higher than that of the type 1 component of tOPV.408 Massive use
of mOPV1 has resulted in the virtual elimination of poliovirus
type 1 in the most difficult to control districts in western Uttar
Pradesh, India. The mOPV3 has been used selectively in areas
with imported or endemic circulation of poliovirus type 3. Finally,
type 2 immunogenicity is high in tOPV,410 but mOPV2 may be
needed for potential applications in controlling an outbreak of
type 2 circulating VDPV (cVDPV) before and after cessation of
OPV. A clinical trial in India confirmed that the immunogenic-
ity of bOPV was superior to types 1 and 3 compared with tOPV,
and not inferior to the respective monovalent OPVs. The cumu-
lative two-dose seroconversion for poliovirus type 1 was 90% for
mOPV1 and 86% for bOPV compared with 63% with tOPV, and
for type 3 poliovirus it was 84% for mOPV3 and 74% for bOPV
compared with 52% for tOPV.410 In 2010, more than a decade
since wild poliovirus type 2 has last been detected,249 bOPV
(types 1 and 3) has become the vaccine of choice to eradicate the
remaining chains of poliovirus transmission.411 The prelicensure
clinical trial demonstrated that bOPV immunogenicity was supe-
rior to tOPV and noninferior to that of the respective mOPVs.
The main advantage of bOPV is that it enhances individual and
population immunity simultaneously for poliovirus types 1 and
3, without any serious loss in immunogenicity compared with
the mOPVs. Figure 28-10 highlights the countries that have used
bOPV in 2010. Almost 900 million doses were used in 2010, the
first full year of bOPV availability to the GPEI.
Intestinal excretion and resistance to change
Mucosal immunity, moderated by locally produced secretory IgA,
is measured primarily by resistance to poliovirus replication and
excretion in the pharynx and intestine after challenge with mOPV
Poliovirus vaccine—live 28 623
Countries using bivalent types 1 and 3 oral poliovirus
vaccine (bOPV1) in 2010. Yellow-shaded countries used bOPV in 2010
in supplemental immunization activities.
or tOPV290,291 or by measuring directly the secretory IgA in stool or
pharyngeal specimens.412,413 After challenge, lower titers of virus
are excreted for significantly shorter periods among vaccinees com-
pared with unvaccinated children414 (Table 28-12). Secretory IgA
can be measured directly in stool, saliva, and breast milk to assess
the degree of mucosal immunity. These methods are difficult to
perform, require tedious standardization, and are rarely used.
Mucosal immunity may exist even when levels of serum anti-
body are negligible,292,415 although the degree of mucosal immu-
nity seems most closely correlated with the titer of homologous
humoral antibody: 416 the lower the titer, the more likely excre-
tion of challenge virus can be demonstrated. One study reported
that mucosal immunity may be strain-specific and that mucosal
immunity induced by one vaccine poliovirus strain may not induce
mucosal immunity against another strain.417 Intestinal mucosal
immunity induced by IPV is less effective against infection than
that induced by OPV, as measured by the proportion of vaccinees
excreting virus or the duration of excretion. However, the clinical
importance of these differences in reducing wild virus spread in
highly immunized populations in industrialized countries is not
clear because IPV seems to reduce shedding compared with no
vaccine and pharyngeal spread may be important in industrialized
countries, and IPV induces pharyngeal immunity.418,419 IPV and
OPV induce equivalent pharyngeal mucosal immunity based on
studies in industrialized countries that rely on vaccine virus chal-
lenge to assess mucosal immunity.420 In contrast, in developing
countries with inadequate hygiene and great potential for fecal-
oral spread of enteric viruses, the clear increase in mucosal (intes-
tinal) immunity induced by OPV over IPV would seem to offer
a major advantage to OPV in reducing the circulation of poliovi-
ruses. Secretory IgA has an important role in defense against polio-
virus infections,192,420,421 and all available evidence indicates that
the immune response after vaccination with OPV is similar to that
after infection with wild poliovirus.
Few studies have provided data on the persistence of mucosal
immunity. No data are available from developing countries about
the duration of mucosal immunity for polioviruses. Several stud-
ies have assessed resistance to oral challenge by vaccine viruses
years after the initial administration of OPV. One study reported
that children were completely resistant to intestinal infection
10 years after vaccination, unless prechallenge serum antibod-
ies were 1:8 or lower.416 Another study reported similar find-
ings on the relationship of humoral antibody and resistance to
excretion.422 A study in India reported that healthy contacts of
poliomyelitis cases, despite multiple doses of previous OPV, may
still be infected and excrete wild-type poliovirus, albeit in rela-
tively low frequency (<1%).422a Furthermore, another study from
India suggests that intestinal mucosal immunity is relatively
624 SECTION TWO

Intestinal Immunity in Vaccinated (OPV or IPV) and Naturally Immune and Susceptible Children

IPV, inactivated poliovirus vaccine; OPV, oral poliovirus vaccine; TCID50, median tissue culture infective dose.
*Excretion index is the proportion of children excreting challenge type 1 virus times the mean duration of excretion days times the titer of virus excreted.
Constructed from data in Fine and Carneiro,206 Global Programme for Vaccines and Immunization,614 and Ghendon and Sanakoyeva.368

short-lived (~6 months).422b No data are available on the long-
term persistence of secretory IgA for polioviruses. However, look-
ing at the mucosal immunity induced by another enterovirus,
echovirus type 6, might provide some insights. A study assessing
the falloff in secretory IgA titer to echovirus type 6 in the phar-
ynx and intestine reported no declines during a 4 year follow-up
period,423 although the possibility of boosting with echovirus 6 or
other enteroviruses in the follow-up period cannot be excluded.
When poliovirus is ingested, the virus has contact with pro-
teolytic enzymes such as trypsin, which may alter viral anti-
gens.421 Secretory and humoral antibody responses after OPV
include those against the new antigens associated with trypsin-
cleaved virus. Such antigens are not accessible in IPV, and, con-
sequently, the immune response after IPV is more limited.424
Evidence of OPV effectiveness
A large body of empiric and scientific evidence has accumulated
since the late 1950s that demonstrates the effectiveness of OPV
in preventing paralytic disease. The use of OPV was pioneered
in the former Soviet Union,425–427 and the approaches developed
in the Soviet Union led to rapid control or elimination of polio-
myelitis in many countries. The most prominent example of
the effectiveness of OPV is the success of the GPEI,39 including
the Western Hemisphere, and the WHO's Western Pacific and
European Regions, which were certified free of wild poliovirus
by Regional Certification Commissions in 1994, 2000, and
2002, respectively.428–430 OPV has curtailed epidemics and has
greatly reduced the incidence of poliomyelitis, often eliminat-
ing the pattern of expected seasonal increase in poliomyelitis
cases.49,2221,225,364,427,431–439 Two vaccine effectiveness studies were
conducted in the 1990s.210,440 The study in Oman estimated that
the effectiveness of three doses of OPV in preventing paralytic
disease was approximately 90%.210 Much of the earlier evidence
of small and large trials is contained in meeting reports.290,291
The ability of OPV to infect contacts of vaccine recipients
(ie, “contact spread”) and “indirectly vaccinate” these contacts
against poliomyelitis is considered by many to be another
advantage of OPV compared with IPV. OPV vaccine virus
spread has been demonstrated by prospective virologic stud-
ies and by serologic studies in industrialized and develop-
ing countries.398,401,441–444 Serologic surveys have shown that
the proportion of people who have antibodies is considerably
greater than would be expected by vaccination or by the circula-
tion of wild polioviruses. After OPV introduction in Yaoundé,
Cameroon, the incidence of paralytic poliomyelitis decreased

Seroconversion to poliovirus types 1, 2, and 3 between birth and 10 weeks of age among children not exposed compared with
children exposed secondarily to OPV mass campaigns, by vaccine group, Oman. eIPV, enhanced-potency IPV. * .05. (From World Health
Organization Collaborative Study Group on Oral and Inactivated Poliovirus Vaccines. Combined immunization of infants with oral and inactivated poliovirus vaccines:
results of a randomized trial in The Gambia, Oman, and Thailand. J Infect Dis 175[suppl 1]:S215-S227, 1997, with permission.)

Poliovirus antibody seroprevalence among unvaccinated inner-city preschool children, by age groups, Detroit and Houston, 1990
to 1991. P1, poliovirus type 1; P2, poliovirus type 2; P3, poliovirus type 3; 12-23 m, 12 to 23 months of age; 24-35 m, 24 to 35 months of age.
(Constructed from data in Chen RT, Hausinger S, Dajani AS, et al. Seroprevalence of antibody against poliovirus in inner-city preschool children: implications for
vaccination policy in the United States. JAMA 275:1639-1645,1996.)
by 85%, although only 35% of children 12 to 13 months old
received three doses of OPV.225 Among infants who received IPV
in Oman, the seroconversion rates were significantly higher
among infants whose study period coincided with a mass
OPV campaign that was conducted elsewhere in the country401
(Figure 28-11). A serologic survey among unvaccinated inner-
city children in the United States also demonstrated that a sub-
stantial proportion of the children are exposed secondarily to
vaccine viruses443 (Figure 28-12). In a study conducted in the
United Kingdom, infants received a dose of IPV at 2 months of
age. In the ensuing 1 month period and before any OPV was
administered, 11% of infants excreted poliovirus type 1 and
4% excreted poliovirus type 2 in stool specimens.444 These data
indicate that vaccine virus spreads easily from OPV recipients
to contacts in industrialized and in developing countries.
Correlate of protection
A comprehensive review of the correlate of protection against
poliovirus was published in 1995.22 Briefly, the data suggest
that any titer of homologous neutralizing antibody is protective
against paralytic consequences of poliomyelitis.
Duration of immunity
Because attenuated viruses contained in OPV are live viruses that
induce the same types of antibody as wild poliovirus does and
because wild virus infection is believed to induce lifelong immu-
nity, it has been reasoned by analogy that immunity induced by
OPV is also lifelong. In an isolated Eskimo population, immu-
nity induced by wild poliovirus was shown to have persisted for
at least 40 years in the absence of any further exposure during
the intervening period.149 The best evidence of the persistence
of vaccine-induced immunity is the absence of disease in adoles-
cents and adults who had been vaccinated previously with OPV
and the persistence of type-specific antibody assessed in popula-
tion-based surveys.445,446 However, interpretation of these data is
difficult because of potential repeated exposures to shed virus. In
prospective studies in which the same vaccinees were observed
for several years, antibodies to types 1 and 2 were found in more
than 90% of children and to type 3 in 83% to 95%.28,447–452 Data
from population-based studies of antibody seroprevalence con-
ducted among Army recruits in the United States in 1989 ( 95%
to poliovirus types 1 and 2 and 85% to poliovirus type 3),446
among school-age children in Massachusetts in 1981 ( 99% to
poliovirus type 1, 99% to poliovirus type 2, and 99% to polio-
virus type 3),445 and in The Gambia (88.1% to poliovirus type
Poliovirus vaccine—live 28 625
1 and 89.3% to poliovirus type 3 among 3- to 4-year-old chil-
dren)453 have demonstrated that poliovirus antibodies induced
by OPV persist for many years.
Adverse events
In the early 1990s, the Institute of Medicine reviewed adverse
events associated with childhood vaccines, including poliovirus
vaccines.454 The major adverse event associated with OPV is
VAPP. Shortly after licensure and widespread use of monovalent
OPV, cases with paralytic manifestations followed vaccination
with monovalent type 3 vaccines. These cases were considered
clinically consistent with poliomyelitis and were supported by
laboratory findings that did not exclude a possible causal rela-
tionship to the administration of oral vaccine. The report by the
Surgeon General describes the earliest cases of VAPP.455
In 1969, the WHO coordinated a collaborative study to obtain
data on the potential risks associated with OPV use. The findings
of the first 5 and 10 year follow-up studies were published.456,457
During the 5 year period from 1980 to 1984, 395 cases of acute
persisting spinal paralysis were reported from 13 countries with
a total population of 547 million.458 The risk of VAPP (in recipi-
ents or contacts of recipients) was less than 0.3 cases per million
doses of OPV distributed (or 1 case per 3.3 million doses), and
the average annual incidence of VAPP was 0.14 per 1 million
people (range, 0.0-0.33), excluding Romania. Romania reported
an average annual incidence of 2.7 per 1 million people.
Although some have challenged the existence of VAPP,28,459,460
believing that the cases of paralysis have different etiologic fac-
tors, the following evidence supports vaccine viruses as causative:
– Clinical syndromes are typical of poliomyelitis.
– Vaccine virus is frequently isolated from cases.
– History of exposure to vaccine is often obtained.
– Recipient and contact cases cluster after receipt of the
first dose of OPV. (One would expect virtually equal
numbers of cases after each dose if there were other
etiologic agents causing the illnesses.)
– Shed viruses have been shown to have mutated toward
neurovirulence.
– The incidence of VAPP is highest in immunodeficient
people with B-cell deficiencies, a group also at higher risk of
poliomyelitis from wild poliovirus.461 The completeness of
reporting of VAPP cases to the Centers for Disease Control
and Prevention (CDC) was estimated to be 81%.462
626 SECTION TWO Licensed vaccines
Between 1990 and 003, a total of 61 cases classified as VAPP
were reported in the United States, including 27 (44%) among
immunologically normal vaccine recipients, 10 (16%) among
immunologically normal contacts of vaccine recipients, 6 (10%)
among immunologically normal nonhousehold contacts, 16
(26%) among immunologically compromised OPV recipients
or contacts of OPV recipients, 1 indeterminate case (2%), and
1 imported case (2%).462,463 Figure 28-13 shows VAPP by year
from 1964 to 2003; no cases occurred after the United States
implemented an all-IPV policy in 2000. Since then, no VAPP
has been reported in the United States, with two exceptions: In
2005, the first case of imported VAPP in an immunologically
compromised person was reported. The person had never been
vaccinated because of a religious exemption and acquired the
poliovirus infection during travel to an OPV-using country in
Central or South America.464 The second case occurred in 2009
in an adult with long-standing common variable immunode-
ficiency whose infection probably occurred through household
contact with her child who had received OPV 12 years earlier.465
For the period 1990 to 1999, the risk of VAPP in the United
States was estimated as 1 case per 2.9 million doses of OPV dis-
tributed; for children receiving the first doses of OPV, the risk
was estimated as 1 case per 1.4 million children vaccinated463,466
(Table 28-13). The risk of VAPP is highest after the first dose of
OPV. Recipients of a first dose and their contacts had a 6.6-fold
higher risk of VAPP than did recipients of subsequent doses and
their contacts. People with immunodeficiency disorders are at
highest risk for VAPP. The risk of VAPP among immunocompro-
mised people is elevated to more than 3,200 times the risk for
immunocompetent people.467 Almost all cases occurred in people
Total reported paralytic poliomyelitis cases and vaccine-
associated paralytic poliomyelitis (VAPP), United States, 1964 to 2000
(Centers for Disease Control and Prevention).
Ratio of Number of Cases of Vaccine-Associated Paralytic Poliomyelitis per Million Doses of Oral Poliovirus Vaccine Distributed,
United States, 1980-1999
NA, Not available.
*First dose ratio to subsequent dose ratio.
†
Includes normal and immunologically abnormal cases.
Adapted from Alexander et al.463
with congenital or acquired immunodeficiency. Immunodeficient
people with VAPP primarily had abnormalities affecting the B-cell
system (humoral immunity), with agammaglobulinemia or hypo-
gammaglobulinemia most frequently associated with VAPP.467
With the exception of one VAPP case with immunodeficiency dis-
order, in all other cases, the precipitating event for the diagnosis
of immunodeficiency was the onset of paralytic disease. Poliovirus
type 3 is the virus most frequently isolated from immunocompe-
tent people with VAPP. In contrast, poliovirus type 2 is the most
common virus detected in immunodeficient people with VAPP.
Poliovirus type 1 is rarely isolated from cases with VAPP.222
The epidemiology of VAPP was summarized in the United
States for the 1990-2003 period. During this period, three differ-
ent vaccination schedules were in use: (1) almost exclusive use
of OPV for primary prevention of poliomyelitis, 1990-1996; (2)
use of a sequential schedule of first IPV then OPV, 1997-1999;
and (3) the introduction of an exclusive IPV schedule in 2000.
Overall, the VAPP estimates for the first period were consistent
with earlier estimates (1 case per 2.9 million doses distributed).
Although 13 VAPP cases occurred during the 1997-1999 period,
none occurred in recipients of the sequential IPV-OPV sched-
ule. Finally, as expected, no VAPP cases were reported following
the introduction of an all-IPV schedule, resulting in the final
step toward polio control, the elimination of VAPP.463
Romania and Hungary have consistently reported higher
rates of VAPP than other countries with well-developed sur-
veillance systems. Until recently, Romania and Hungary used
tOPV and mOPV, respectively, solely in campaigns.241,355 The
high risk of VAPP in Hungary was associated primarily with
the administration of mOPV3.355 The high risk of VAPP in
Romania can be attributed to provocation poliomyelitis (ie,
multiple intramuscular injections in the 30 days before par-
alytic manifestations).468 People with VAPP in Romania had
received, on average, 16.8 intramuscular injections, primar-
ily for antibiotics, in the 30 days before onset of paralysis.
Analysis of VAPP cases in the United States between 1980
and 1993 suggested that cases with a history of intramuscular
injections received an average of only 1.5 intramuscular injec-
tions in the 45 days before onset of paralysis; no clustering of
injections during the 45 day period was observed.469 In con-
trast, in Romania, most injections were received in the peri-
ods from 0 to 7 days and from 8 to 14 days before onset. It
seems unlikely that intramuscular injections contribute sub-
stantially to the VAPP burden in the United States. The risk
of VAPP in the Americas, where Latin America administered
large quantities of OPV in mass campaigns (NIDs) to eradi-
cate poliomyelitis, was similar to the VAPP risk reported in
the United States and other countries.470 An analysis in India
further refined the risk estimates for one large developing
country and demonstrated that the risk of VAPP is substan-
tially lower in this country despite large quantities of OPV
administered in mass vaccination campaigns and through rou-

tine immunization programs ( 700 million doses of OPV in
1999 alone).471 The risk among first-dose recipients was esti-
mated as 1 case per 2.8 million children (compared with a
first-dose recipient risk of 1 case per 1.4 million children in
the United States). The lower risk estimates in India may be
explained in part by the following: (1) The proportion with and
the titers of maternally derived antibody are high among new-
borns because wild poliovirus circulated widely until recently
and may have led to frequent virus exposure and boosting of
titers in the general population. (2) A birth dose of OPV is rec-
ommended for institutional births (thus inducing immunity
when the newborn is still under protection against poliomyeli-
tis from maternally derived antibody). (3) The routine vacci-
nation schedule calls for OPV doses at 6, 10, and 14 weeks of
age (again inducing immunity at earlier ages than in industri-
alized countries). Thus, high levels of maternal antibody and
OPV vaccination at an age when maternal antibody would be
expected to protect against VAPP seem to be the main reasons
why the observed risk of VAPP in India (and probably in other
countries under similar circumstances) seems to be lower than
that in industrialized countries.471
The global burden of VAPP has been estimated by WHO
between 250 and 500 cases globally each year, based on a total
incidence of 2-4 cases per million birth cohort.472
Immunodeficient children are subject to infection, fre-
quently fatal, by a wide variety of normally benign or avirulent
agents. Nonpolio enteroviruses may cause serious or fatal ill-
nesses in immunocompromised people.473,474 A prominent fea-
ture of such infections is the patient's inability to eradicate the
virus from the CNS; some patients continue to yield virus from
cerebrospinal fluid for up to 3 years.475 At the end of 2011, a
total of 45 persons with immunodeficiency disorders have been
shown to excrete poliovirus for prolonged periods of approxi-
mately 6 months or longer (see “General epidemiology”) since
the early 1960s. Two cases in this series deserve attention. A
patient with VAPP in the United States may have excreted virus
for approximately 7 years before the onset of paralytic disease in
1981,158,159 and a person from the United Kingdom who does
not have paralytic disease has excreted poliovirus for more than
20 years (as of 2011).160 In immunodeficient persons, poliovirus
infection, by a wild virus or a vaccine strain, may develop in an
atypical manner, with an incubation period longer than 28 days,
a high mortality rate after a long chronic illness, and unusual
lesions in the CNS.22,275,467,476
In principle, all clinical and environmental poliovirus iso-
lates that are related to the OPV strains are VDPVs. However,
derivatives of the Sabin OPV strains have been classified into
two broad categories: (1) “OPV-like” isolates that have close
sequence relationships ( 99% VP1 sequence identity) to the
original OPV strains, and (2) VDPV isolates that have sequence
properties ( 99% VP1 sequence identity from the parental
Sabin strains) indicative of prolonged replication of the vaccine
virus. The VDPV isolates, in turn, have been subdivided into
three groups, reflective of the differing conditions that lead to
their appearance: (1) immunodeficient VDPVs (iVDPVs), iso-
lated from immunodeficient patients who become chronically
infected after exposure to OPV; (2) cVDPVs require evidence
of transmission and neurovirulence (at least 2 cases with AFP)
and arise usually in communities with inadequate OPV cov-
erage; and (3) ambiguous VDPVs (aVDPVs) not known to be
associated with outbreaks of AFP cases or infections of patients
with known B-cell immunodeficiencies. The VDPVs are of par-
ticular interest because of their implications for current and
future strategies for global polio eradication.160,206,477
The vast majority of vaccine-related isolates are OPV-like.
Capsid sequences of OPV-like isolates (usually surveyed by VP1
sequencing) closely match those of the parental Sabin strains.
Because poliovirus genomes evolve rapidly (approximately
Poliovirus vaccine—live 28 627
1% per year),243 the extent of VP1 sequence divergence can
be used to approximate the duration of a poliovirus infec-
tion.159,162,166 However, the confidence intervals for these time
estimates are wide because of the stochastic nature of muta-
tion and because some VP1 sites in the Sabin strains are subject
to negative selection during replication of OPV in the human
intestine.308 Confidence intervals narrow when the sequencing
window is extended to include the complete capsid region or
the complete genome.159,260,261 The number of VP1 nucleotide
substitutions expected from the poliovirus molecular clock243 to
accumulate during infection of an immunologically normal pri-
mary vaccine recipient (assuming a maximum duration of infec-
tion of 8 weeks) is zero to three. Selected reversion may add
one or two additional VP1 substitutions during a normal infec-
tion. Thus, the 1% (equivalent to approximately nine nucleotide
substitutions in VP1) demarcation between OPV-like isolates
and VDPVs should include virtually all OPV-like isolates. For
cVDPV type 2, the GPLN recommended in 2010 to include
cases with isolates that demonstrated 0.6% or more sequence
divergence from the parental Sabin type 2 strain.The introduc-
tion of real-time PCR into the GPLN in the past few years seems
to have increased the sensitivity of detecting cVDPV type 2113
(see “Virologic testing”).
Many OPV-like isolates have recombinant genomes.166,478–481
The large majority of OPV-like recombinants were generated by
heterotypic genetic exchange among Sabin strains. OPV-like iso-
lates with non-Sabin strain sequences are rare.480,482 The abun-
dance of vaccine-related recombinants results from the trivalent
formulation of OPV and the likelihood that some recombinants
have higher fitness for replication in the human intestine than
the original OPV strains. Crossovers are most frequently found
in the noncapsid region, less frequently in the 5 -UTR, and least
frequently in the 3 -terminal sequences of the capsid region.483
Vaccine-related recombinants are more often associated with
types 3 and 2 than with type 1.478–480,484 Because capsid and non-
capsid sequences coevolve, the extent of VP1 sequence identity
to the corresponding Sabin OPV strain is generally reflective of
the evolutionary history of the overall genome.
It seems likely that many OPV-like isolates have recovered
the capacity for higher neurovirulence and possibly increased
transmissibility. The small number of substitutions controlling
neurovirulence in experimental animals were found to have
reverted among many OPV-like isolates, especially among iso-
lates of types 2 and 3.308,311,485,486 Because the critical and unsta-
ble attenuating mutations in the 5-UTRs of the Sabin strains
also affect fitness for virus replication in the human intestine,485
it seems possible that revertants at these sites would have a
higher fitness for person-to-person spread. However, spread is
normally limited by high OPV coverage.
It has long been recognized that immunodeficient patients
with defects in antibody production (especially patients with
CVID or X-linked agammaglobulinemia) could become chron-
ically infected when exposed to OPV.467 Unambiguous dem-
onstration that vaccine-related poliovirus isolates from
immunodeficient patients had unusual sequence properties
awaited the application of molecular tools, such as oligonucle-
otide fingerprinting487 and genomic sequencing,159,162,166 to polio-
virus diagnostics. As expected, the extent of sequence divergence
is related to the duration of the prolonged infection. Not all iso-
lates from immunodeficient patients would be classified as iVD-
PVs. Some isolates are specimens obtained early in the chronic
infection, and no subsequent specimens were obtained. In other
situations, the prolonged infections had resolved spontaneously
or the patient died of complications of the immunodeficiency
(including fatal poliomyelitis). However, some iVDPV isolates
are highly divergent ( 90% VP1 sequence identity to the paren-
tal OPV strain), suggesting that the chronic poliovirus infec-
tions had persisted for 10 years or more.160,306 Prolonged iVDPV
infections are independent events,222,467 and the isolates obtained
from such infection trace separate pathways of divergence from
628 SECTION TWO
the original OPV strains.159,162 Many of the iVDPV isolates are
recombinant, but, as with most OPV-like isolates, the recombi-
nant sequences of all known iVDPV isolates have been derived
from the Sabin strains. This pattern of recombination has per-
mitted the recognition of iVDPV isolates from their sequence
properties. The underlying reason for this pattern of recombina-
tion may be that chronically infected immunodeficient patients
are generally resistant to superinfection by other enteroviruses.
Prolonged infection of immunodeficient patients with
VDPVs is problematic because such infections cannot be pre-
vented by high OPV coverage. So far, all reports of persis-
tent iVDPV infections have been from countries with high to
intermediate levels of development, where the rates of OPV
coverage are high and the survival times of immunodefi-
cient patients may be extended by their access to appropriate
clinical management, except for a single case from India. The
survival rates for hypogammaglobulinemic patients are prob-
ably very low in developing countries at highest risk for polio-
virus spread. The population of chronic iVDPV excreters is
declining in developed countries because some patients have
died, some have cleared their infections, and no new iVDPV
infections have been found in countries that have shifted to
IPV. Although there is only rarely evidence of spread of iVDPVs
from immunodeficient patients to the wider community,160,222
the potential for such spread may be limited by high vaccine
coverage in the communities where immunodeficient patients
have extended survival times.
The cVDPVs have been recognized since 2000. A total of
17 cVDPV outbreaks have been recognized in 15 countries of
Africa, the Americas, and Asia (Table 28-14), including an out-
break in Egypt that was detected retrospectively. All of these
outbreaks were associated with few paralytic cases, with some
notable exceptions in Egypt, Indonesia, and Nigeria. However,
it seems that in some of these outbreaks, the case-to-infection
ratio must have been very low,488 but the clinical manifestations
of cVDPV seem to be similar to those caused by wild poliovi-
rus infection.489 In Egypt, type 2 cVDPV was isolated from 30
patients with polio during the years 1988 to 1993.258,490 The rate
and pattern of VP1 divergence from the Sabin type 2 OPV strain
suggested that all lineages were derived from an OPV infec-
tion that occurred around 1983. Phylogenetic analysis showed
that the cVDPVs circulated widely in Egypt along several
independent chains of transmission and established indepen-
dent foci of endemicity in separate communities. The circula-
tion of cVDPV ceased when OPV coverage rates increased. In
Hispaniola (Haiti and the Dominican Republic), an outbreak
of 21 confirmed cases (including two fatal cases) in 2000 to
2001 was associated with type 1 cVDPV.65 A more limited out-
break, associated with an independent-lineage type 1 cVDPV,
occurred in the Philippines in 2001.260,491 An outbreak of type
2 cVDPV occurred in Madagascar in 2002.492,493 More recently,
outbreaks of cVDPV were reported from China in 2004 (2 cases
with type 1 virus),494 Madagascar in 2005 (3 cases with type 2
virus), Indonesia in 2005 (46 cases with type 1 virus),495 and
Cambodia in 2005 (3 cases with type 3 virus).496 However, the
largest outbreak thus far is occuring in Nigeria. Between 2005
and 2010, a total of 333 cases of cVDPV2 were reported.256,489
In this outbreak, several independent emergences of cVDPV2
were found, but a single lineage predominates and has caused
the vast majority of cases ( 90%). The cVDPVs have recovered
the essential properties of wild polioviruses: (1) the capacity to
cause paralytic disease in humans and (2) the capacity for con-
tinuous person-to-person transmission. The common risk fac-
tors for cVDPV circulation were major gaps in OPV coverage,
providing large numbers of susceptible people for virus circula-
tion; environmental conditions favoring poliovirus spread; and
the prior eradication of the corresponding serotype of indige-
nous wild poliovirus (except in Indonesia where imported wild
poliovirus type 1 cocirculated with type 1 cVDPV in East Java
Province). Most cVDPV outbreaks have been caused by type 1
or type 2; the reasons for this distribution are unknown. The
majority of VAPP cases among immunodeficient persons are
associated with Sabin type 2, and it is possible that the emer-
gence of cVDPV type 2 may also occur in immunodeficient
persons. This could explain the type 2 contribution of cVDPV
since types 1 and 3 are rarely associated with prolonged excre-
tion in immunodeficient persons. The type 1 contribution may
be due to relatively higher contagiousness of type 1 in the face of
partial OPV-induced immunity in the affected communities. Of
the 17 cVDPV outbreaks reported globally since 2000, 9 were
due to cVDPV type 2, 6 were due to cVDPV type 1, and 2 were
due to cVDPV type 3. However, the vast majority of cases ema-
nate from cVDPV type 2 outbreaks (>95%).
All outbreak-associated cVDPV isolates, except those
from China,261 have been recombinants with related enterovi-
ruses.65,490,497 The China cVDPV isolates demonstrate that recom-
bination is not essential to the phenotypic reversion of the Sabin
strains because the main determinants of attenuation of all three
Sabin strains map to the 5 -UTR and capsid region sites303,306,308,316
and most of the crossovers in cVDPV isolates map to the non-
capsid region. The high frequency of back-mutation of the atten-
uating 5 -UTR sites during replication of OPV in the human
intestine484–486 suggests that back-mutations would usually pre-
cede recombination. Recombination within the capsid region is
infrequent,498 and few of the independent cVDPV isolates char-
acterized so far show evidence of any such recombination.65,492,497
The observed successive recombination events involving the non-
capsid region65 and the 5 -UTR sequences may not be associated
with any progressive increase in replicative fitness. Rather, recom-
bination with other enteroviruses is likely to be a random pro-
cess that occurs during mixed infection, the probability of which
increases with the total number of people infected. Thus, the com-
bination in a vaccine-related isolate of significant divergence of
capsid nucleotide sequences ( 1% from the parental OPV strains)
and recombination with other enteroviruses is a likely indicator of
VDPV circulation. VP1 sequence divergence of greater than 1% in
a cVDPV isolate would suggest that circulation had occurred for
at least a year, raising the possibility that immunization activities
during the preceding year had failed to stop VDPV circulation. A
comprehensive review of VDPV was published in 2005,499 and
updates are published regularly.500–502
The cVDPV outbreaks in developing countries challenge the
assumption that poliovirus endemicity can be restored only by
reintroduction of wild poliovirus, underscore the urgency of
reaching the goal of global polio eradication as quickly as pos-
sible, and have important implications for the end-game strat-
egy of polio eradication. The immediate priority is to eliminate
the remaining pockets of wild poliovirus endemicity in South
Asia and sub-Saharan Africa.39 At the same time, it is essential
to maintain high levels of vaccination coverage against polio-
viruses in all countries to prevent the spread of imported wild
polioviruses and to suppress the emergence of cVDPVs. Areas
at highest risk for emergence of cVDPVs are those where vacci-
nation coverage rates have declined, the competing wild polio-
viruses have been eliminated, and epidemiologic conditions had
previously favored wild poliovirus transmission. In response to
the recent cVDPV outbreaks, the WHO has recommended the
reinstatement of mass immunization campaigns to close the
immunity gap in areas where coverage through routine immu-
nization has been insufficient to prevent widening susceptibil-
ity to polio.503 The frequency of the mass campaigns shall be
determined by the rate of accumulation of susceptible people in
the population and the basic reproduction number for poliovi-
rus in the highest risk populations in each area.170,206 No addi-
tional measures have been recommended for the remaining
polio-endemic countries because activities currently planned for
elimination of the last pockets of types 1 and 3 poliovirus circu-
lation would also effectively prevent dissemination of cVDPVs.
 Outbreaks of Circulating Vaccine-Derived Poliovirus, 1988-2010

cVDPV, circulating vaccine-derived poliovirus; nt, nucleotide.

*Start of outbreak estimated from the molecular clock.
†
cVDPV was imported from Haiti into the Dominican Republic.
‡
Cases in Niger (2 in 2006, 1 in 2010) and Chad (1 in 2010) were linked to the outbreak in Nigeria. Cases continued in Nigeria into 2011.

Poliovirus vaccine—live
§
Multiple independent cVDPV emergences.

28
629

Genetic sequencing studies suggest that reversion of Sabin
strains to potentially more neurovirulent phenotypes occurs
commonly after OPV administration.483–311,486,562–567 Two
relatively small studies568,569 indicated that the use of a sequen-
tial schedule may not reduce the frequency of such mutations.
However, one larger study suggests that the use of a dose of IPV
before two or more doses of OPV may reduce the amount of
type 3 virus shed, the most common cause of VAPP, but prob-
ably will not influence the shedding of type 1 or type 2 viruses
or the extent of reversion.570 VAPP has been rarely reported after
sequential schedules using at least one dose of IPV before mul-
tiple doses OPV. In Hungary, a country that had a high risk of
VAPP during decades, the switch to a sequential schedule (with
one dose of IPV) was associated with the apparent elimination
of VAPP. However, there is debate whether one dose of IPV is
adequate to prevent VAPP.
Recommendations for vaccine use
Two major objectives of vaccination, protection at the young-
est possible age and minimum rates of attrition (ie, dropout)
between OPV doses, govern the development of routine vac-
cination schedules in industrialized and developing countries.
In each of these settings, an optimal balance must be found
between these objectives.94
Most industrialized countries, including many Western
European countries, have recommended schedules in the
past that relied exclusively on OPV for the prevention of
poliomyelitis. More recently, encouraged by progress of the
GPEI and by the desire to reduce or eliminate the burden of
VAPP, many of the high- and middle-income countries are
reevaluating their vaccination policy options. As of 2011, a
total of 56 countries and reporting entities rely exclusively on
IPV, 16 countries and reporting entities use sequential sched-
ules of both IPV and OPV (Figure 28-14). The rest of the world
is using OPV. The major differences in the recommended
schedules between industrialized and developing countries
include the following: (1) age at first dose, (2) vaccines used
for each dose, and (3) interval between doses. The recom-
mendations for poliomyelitis prevention in the United States
were revised most recently in 2000.545 A schedule relying only
on IPV is now recommended by the CDC and the American
Poliovirus vaccine—live 28 633
Academy of Pediatrics for primary poliovirus vaccination of
children in the United States. The schedule calls for three
doses of IPV administered at 2 and at 4 months and between
6 and 18 months. A preschool booster dose of IPV is recom-
mended after age 4 years. Similarly, in the United Kingdom, an
IPV-only schedule was introduced in 2004; doses are provided
at 2, 3, and 4 months of age526 (Table 28-17). The minimum
recommended interval between doses of OPV for routine vac-
cination varies; in the United States, 6 to 8 weeks were rec-
ommended. For the United Kingdom, intervals between OPV
doses of 4 weeks were preferred.526 The major advantages and
disadvantages of the three poliovirus vaccination schedules
are shown in Table 28-18.
Routine immunization for adults residing in the continen-
tal United States and many other industrialized countries
is not believed to be necessary because of the small risk of
exposure to wild poliovirus.466,525,526 However, adults who are
at increased risk because of contact with a patient infected
with wild poliovirus or who are working with polioviruses
and adults planning travel to an epidemic or endemic area
should be immunized. Parents and other household mem-
bers who do not have definite evidence of having been com-
pletely immunized should receive IPV at the time the child is
vaccinated.466,525
The WHO-recommended schedule that calls for the admin-
istration of four doses of OPV at birth, 6, 10, and 14 weeks of
age should be used for polio-endemic or recently polio-endemic
countries. This is particularly important in areas in which fre-
quent importation or endemic circulation of wild polioviruses
takes place and in which a majority of infants are exposed
to all three poliovirus types early in life.170,209,547,554
634 SECTION TWO Licensed vaccines

OPV is the vaccine recommended for most of the world except in limited situations. SES, socioeconomic status.
preeradication era of poliomyelitis in 2010. This paper offers
guidance to countries that are considering changes in routine
immunization schedules for the preeradication era of polio
eradication.575 The WHO recommendations for the choice
of vaccine (and schedule) are determined by the potential for
importation of poliovirus and the risk for transmission (cov-
erage, socioeconomic status, hygiene) following importation.
OPV is the vaccine recommended for most of the world except
in limited situations as shown in Figure 28-15.
Public health considerations
Epidemiologic results of vaccination
Surveillance data since 1980 suggest that continuing trans-
mission of indigenous wild poliovirus has been interrupted in
the United States, during a period when the country relied on
OPV for immunization.222,461 As part of the certification of the
Western Hemisphere as polio-free, all countries in the Americas,
including the United States, were certified free of indigenous
wild poliovirus in 1994 by an International Commission con-
vened by the Pan American Health Organization on the basis
of a detailed review of available data by national committees.428
Data on poliovirus circulation during 1988 to 2011 by WHO
region, a period during which the GPEI has been implemented
in all WHO regions, can be found in Figure 28-16. Progress has
been impressive. In 1988, 125 countries were polio-endemic; by
2012 only 3 countries in South Asia and Africa were consid-
ered polio-endemic (ie, had never interrupted indigenous virus
transmission: Afghanistan, Nigeria, and Pakistan, importation
with limited transmission in time and space does not consti-
tute reestablishment of endemicity). Egypt was removed from
the list of endemic countries at the end of January 2006 follow-
ing a 12-month period without detection of poliovirus despite
excellent surveillance performance from AFP surveillance and
from environmental surveillance. Globally, cases of poliomy-
elitis decreased more than 99% from an estimated 350,000 in
1988 to 1,349 in 2010, and 650 cases in 2011 (as of 13 March
2012). Wild poliovirus type 2 was last detected in October 1999
in Uttar Pradesh, India.39 Wild poliovirus type 3 was detected
in seven countries in 2010 (Afghanistan, Chad, India, Nigeria,
Niger, Mali, and Pakistan). Wild poliovirus type 1 was detected in
2005 in the same countries (except Niger) and in 13 additional
countries: Angola, Congo-Brazzaville, Democratic Republic of
the Congo, Kazakhstan, Liberia, Mauritania, Nepal, Russian
Federation, Senegal, Sierra Leone, Tajikistan, and Turkmenistan.
Poliovirus vaccine—live 28 635
Outside the four polio-endemic countries, importations origi-
nated from India or Nigeria (Figure 28-17). However, viruses
from India, for example, reinfected Angola, and the failure to ter-
minate transmission within Angola for 6 years after three Indian
importations led to export of virus from Angola directly or indi-
rectly to Burundi, Central African Republic, Congo-Brazzaville,
Democratic Republic of the Congo, Gabon, and Nambia. Four
countries (Angola, Chad, Democratic Republic of the Congo,
Sudan) are considered to have “reestablished transmission”, pre-
viously polio-free countries that experienced persistent poliovi-
rus transmission more than 12 months following new poliovirus
importation. The latest information on the status of the polio
eradication initiative can be found at www.polioeradication.org.
Between 1988 and 2010, the GPEI prevented more than
9 million children from the crippling consequences of polio-
myelitis and averted more than 1.5 million deaths, some as
a result of poliomyelitis prevention and most because vitamin
A supplements are administered to the same target children in
many polio-endemic countries during national and subnational
immunization campaigns.576,577
Disease control strategies
In 1988, the World Health Assembly, the governing body of the
WHO, resolved to eradicate polio globally by the year 2000.3
The global resolution followed the 1990 regional elimina-
tion goal established in 1985 by the countries of the Western
Hemisphere. The last case of poliomyelitis associated with wild
poliovirus isolation in the Americas was reported from Peru in
1991, and the entire hemisphere was certified free of indigenous
wild poliovirus by an International Certification Commission
in 1994.428
The following strategies to achieve polio eradication, devel-
oped in the Western Hemisphere, were adopted by the WHO
for worldwide implementation in all polio-endemic countries37:
– Achieving and maintaining high routine coverage in
infants younger than 1 year with at least three doses of
oral poliovirus vaccine (OPV3).
– Administering supplemental doses of OPV to all young
children (usually children younger than 5 years) during
NIDs to rapidly interrupt poliovirus transmission.
– Conducting “mopping-up” vaccination campaigns—
localized campaigns targeting high-risk areas where
poliovirus transmission is most likely to persist at low levels.
– Developing sensitive systems of epidemiologic
and laboratory surveillance, including establishing
surveillance of cases of AFP.
636 SECTION TWO

Wild poliovius (WPV) cases and importation routes,* worldwide 2008-2010. (Data from Centers for Disease Control and Prevention. Progress
toward interrupting wild poliovirus circulation in countries with re-established transmission: Africa, 2009-2010 MMWR Morb Mortal Wkly Rep 60:306-311, 2011.)
Control of poliomyelitis and the global eradication initiative are
greatly aided by well-functioning routine immunization programs
that deliver potent OPV to a high proportion of infants in the first
year of life. Indeed, there is a clear association between routine
coverage and length of transmission of poliovirus following an
importation into a previously polio-free country. Reported global
coverage with three doses of diphtheria-tetanus-pertussis (DTP3)
vaccine (assumed to reflect routine OPV3 coverage) among infants
younger than 1 year was 82% in 2009 (WHO/UNICEF best esti-
mate of coverage). All WHO regions reported a coverage of more
than 80% except for the African Region, where coverage improved
from 56% in 1990 to 72% in 2009 but continues to be less than
the coverage achieved in the other WHO regions.224 However,
these global and regional figures may mask substantial variation
in coverage reported among and within individual countries.
Mass campaigns with OPV—administered during NIDs—are
the only proven strategy to reduce widespread transmission of
wild poliovirus in endemic countries.578,579 The NIDs are con-
ducted twice annually for a short period (1-3 days) in which
one dose of OPV is administered to all children in the target
age group, usually children younger than 5 years, regardless of
prior vaccination history. A second dose is administered in the
same way after an interval of 4 to 6 weeks. The NIDs usually
take place during the low transmission season when conditions
are optimal to interrupt the few remaining chains of poliovi-
rus transmission. Most countries provide OPV during NIDs,
relying primarily on fixed sites, including vaccination clinics
Poliovirus vaccine—live 28 637
supplemented by a large number of temporary vaccination sites.
With the acceleration of eradication activities in 1999, many of
the remaining polio-endemic countries began using a strategy
relying on house-to-house administration of OPV during NIDs.
The NIDs are necessary in developing countries to rap-
idly increase immunity levels in the population to achieve
and surpass herd immunity threshold levels for poliomyeli-
tis and, hence, rapidly interrupt the transmission of poliovi-
rus. OPV administered in campaigns also seems to be more
immunogenic compared with OPV administered in the rou-
tine program,580,581 probably for the following reasons: (1) NIDs
are conducted during the low poliovirus transmission season
because this is the period when the fewest chains of poliovirus
transmission are maintained. (2) NIDs are conducted during
the low transmission season for other enteroviruses that may
interfere with poliovirus seroconversion.582 (3) The cold chain
can be better maintained for these short campaigns. (4) Massive
use of OPV probably also results in intensive secondary spread
of shed virus.401 Children residing in polio-endemic countries
using NIDs may receive 13 to 14 doses of OPV by the time they
reach their fifth birthday.94,407 These OPV doses are adminis-
tered by the routine program (three or four doses) and through
NIDs (two doses annually during the first 5 years of life). These
additional doses of OPV, administered during NIDs, should cor-
rect the lower immunogenicity of OPV commonly observed in
tropical areas.
In 2010, 308 supplementary immunization activities with
OPV were conducted in 49 countries, reaching more than
392 million people, the vast majority children younger than
5 years, administering approximately 2 billion doses of OPV.
Each of the endemic countries conducted at least six large-scale
638 SECTION TWO Licensed vaccines
supplemental immunization campaigns in 2010, while each of
the reinfected countries with transmission conducted a mini-
mum of 4 large-scale supplemental immunization campaigns in
2010. All countries used well-supervised, house-to-house vacci-
nation in part or all of the target area for supplementary immu-
nization activities to further increase the quality and reach the
highest possible coverage among children younger than 5 years.
In response to importation of wild poliovirus type 1 originat-
ing from Nigeria, coordinated NIDs were conducted across
Western, Central, and Eastern Africa in 2009 and 2010.37,583
The NIDs in India—vaccinating as many as 157 million chil-
dren in a single round—represent the largest mass campaigns
ever conducted.
To eliminate the last potential or known reservoirs of wild
poliovirus circulation, mopping-up vaccination campaigns are
conducted. These mopping-up campaigns usually target chil-
dren younger than 5 years with two doses of OPV separated by
an interval of 4 to 6 weeks. These campaigns include house-
to-house administration of OPV to reach any children who
may have been missed during NIDs. Mopping up is a critical
component to achieve interruption of the final chains of polio-
virus transmission in all polio-endemic countries. Risk areas
(often targeting more than 1 million children younger than
5 years to have the optimal impact), usually defined at county
or district levels, to be included in mopping up include those
with recent circulation of wild poliovirus (usually within the
last 3 years), low vaccination coverage, suboptimal surveillance,
large migrant or refugee populations, and common borders with
known poliovirus-endemic areas.
These supplemental immunization activities have been suc-
cessful in decreasing the number of reported poliomyelitis cases
globally from 35,251 (estimated to be approximately 350,000
cases) in 1988 (when the polio eradication target was adopted)
to 650 in 2011, a decrease of more than 99%39 (Figure 28-18).
Figure 28-19 displays the remaining AFP cases associated with
wild poliovirus isolation during a 12-month period in 2010.
A detailed review of the current status of the GPEI was pub-
lished in 1997584 with annual progress reports, the most recent
in 2010.39
Total reported acute flaccid paralysis (AFP) cases and confirmed poliomyelitis cases, 1988 to 2010 (World Health Organization).
Surveillance for cases of AFP and for wild poliovirus is critical for
guiding programmatic activities and for contributing to the even-
tual certification of polio-free status. Systems for AFP surveillance
have been established in all polio-endemic or recently endemic
countries. Surveillance relies on two complementary and mutu-
ally reinforcing components: (1) AFP case investigations and (2)
virologic studies of polioviruses obtained from clinical specimens.
The major reason for using a symptom (eg, AFP) rather than a
diagnosis (eg, poliomyelitis) is to ensure that the sensitivity of the
surveillance system can be maximized; all possible cases of polio-
myelitis, including those with atypical manifestations, will be
included in the surveillance system. In addition, AFP surveillance
helps to monitor the quality of surveillance even in the absence
of cases of poliomyelitis. In the last stages of the eradication pro-
gram, no cases of poliomyelitis (except for rare cases of VAPP)
would be expected to be detected. Thus, it would be impossible
to determine whether the absence of poliomyelitis cases repre-
sents “true” absence or deficiencies in surveillance. On the basis
of the experience in the Americas, in each population, a rate of
at least 1 case of nonpolio AFP per 100,000 population younger
than 15 years would be expected annually, and achievement of
such a rate would indicate adequate surveillance, defined as the
ability of the surveillance system to detect wild poliovirus circu-
lation resulting from indigenous transmission or virus importa-
tion, should it occur. A global network of 145 formally accredited
laboratories has been established to process all stool specimens
collected from AFP cases worldwide for virologic investigations
(Figure 28-20). The formal annual accreditation process relies on
six criteria (Table 28-19). Standard methods and reagents are used
to isolate virus in tissue cultures and perform intratypic differen-
tiation (ITD) as wild, Sabin-derived ( 1% genomic sequencing
divergence from Sabin parent strains), or vaccine-derived ( 1%
genomic sequencing divergence from Sabin parent strains) polio-
virus. For poliovirus type 2, 0.6% (6 nucleotide changes) qualify
isolate to be included as cVDPV. All wild poliovirus strains and all
VDPVs are sequenced in specialized network laboratories to guide
programmatic action. Based on additional virologic and epidemio-
logic data, VDPVs can further be classified as circulating (cVDPV)
or associated with prolonged replication in an immunodeficient
person (iVDPV) (see “Vaccine-derived polioviruses”).
640 SECTION TWO Licensed vaccines
Standards Used for the Annual Accreditation of Laboratories Participating in the World Health Organization Global Laboratory Network for
Poliomyelitis Eradication
ITD, intratypic differentiation; NA, not available; RRL, regional reference laboratory.
Confirmed Poliomyelitis Cases and AFP Surveillance Indicators, by World Health Organization Region, 2009 to 2010
AFP, acute flaccid paralysis.
*Rate per 100,000 children older than 15 years.
Data from World Health Organization.
All six regions of the WHO have achieved a rate of slightly
more than 1 nonpolio AFP case per 100,000 population younger
than 15 years in 2009 and 2010. Table 28-20 shows the AFP rates
for the different WHO regions in 2009 and 2010.39,232 In 2005, the
Advisory Committee on Poliomyelitis Eradication recommended
that the nonpolio AFP rate in polio-endemic countries should be
at least 2/100,000 persons younger than 15 years. All three polio-
endemic WHO regions reported a nonpolio AFP rate of more
than 3 in 2010 (African Region, 4.9; Eastern Mediterranean
Region, 5.0; and South East Asia Region, 10.2), greatly increas-
ing the sensitivity of poliomyelitis detection. All WHO regions
also surpassed the second most important quality indicator for
adequate surveillance (ie, the proportion of AFP cases from which
two stool samples had been obtained within 14 days of paralysis
onset). This indicator was 86% in 2010 globally. A comprehen-
sive list of performance indicators used to monitor the quality of
AFP surveillance can be found in Table 28-21.
Remaining obstacles to eradication
The major obstacle to achieve eradication is the failure to imple-
ment adequately the current eradication strategies. The eradi-
cation strategies have been successful in almost all countries
and regions within countries. However, the three remaining
polio-endemic countries (ie, those that never eliminated indig-
enous wild poliovirus transmission) pose special challenges. In
certain states in Northern Nigeria, routine and mass vaccina-
tion coverage with OPV is suboptimal. In Afghanistan, access in
the southeastern part of the country is limited owing to security
concerns. In Pakistan, access in the tribal territories (Federally
Administered Tribal Areas and Khyber Pakhtunkhwa) along
the border with Afghanistan remains a problem, and coverage
in some accessible areas remains suboptimal. And in Uttar Pradesh
and Bihar states in Northern India, the transmissibility conditions
for poliovirus are probably almost optimal, and the program has
only recently been reaching a high enough proportion of children
younger than 5 years to surpass the threshold for herd immunity
because the efficacy of OPV in this setting is lower than in other
tropical settings, and, on average, AFP cases had a history of receiv-
ing more than 10 doses of OPV. Innovative strategies and implemen-
tation of new tools must be tailored for each of these polio-endemic
areas to achieve eradication. In addition, access to and reaching
high immunization coverage among migrant and nomad popula-
tions is an issue for all of these polio-endemic countries.
Given that most importation in polio-free countries has orig-
inated in India or Nigeria, progress toward eradication in these
countries is critical to decrease the risk of further wild polio-
virus importation. Two reestablished transmission countries

Indicators of AFP Disease Surveillance and Laboratory
Performance
AFP, acute flaccid paralysis.
*Good condition means that, on arrival (1) there is ice or frozen icepacks or
a temperature indicator (showing 8 °C) in the container, (2) the specimen
volume is adequate ( 5 g), (3) there is no evidence of leakage or desiccation,
and (4) appropriate documentation (laboratory request/reporting form) is
completed.
(Chad, Democratic Republic of the Congo) and one polio-
endemic country (Pakistan) accounted for more than 80% of all
global cases in 2011 (as of June 30).
On the other hand, wild poliovirus type 3 was last isolated
in Asia on 19 February 2012 (only the third case since the begin-
ning of 2011 in Asia) (data as of 13 March 2012). Therefore,
wild poliovirus type 3 may be on the verge of elimination from
Asia. Similarly, 13 January 2012 marked a 12 month period
during which no polioviruses have been detected in India. India
has been removed subsequently from the list of polio-endemic
countries. Thus, substantial progress has been achieved in the
last 12 to 18 months, due in large part to the availability and
widespread use of bOPV and improved operations.
Given the progress toward interrupting wild poliovirus trans-
mission, preparations of the end game for the global GPEI have
accelerated in recent years. The goal of any eradication program
is for the preventive intervention to be no longer necessary (ie,
discontinuation of vaccination against poliomyelitis).585,586 The
priority in OPV cessation planning is to reduce to the extent
possible and manage the risks of paralytic disease caused by live
polioviruses among current and future generations of children.
Total elimination of these risks is not feasible.
There are important lessons from the successful small-
pox eradication effort, including the following expectations:
(1) that vaccination can be stopped, (2) that the virus must be
restricted to an absolute minimum number of facilities, and (3)
that manipulations with the virus will be regulated.587 However,
there are important differences in the vaccines used to eradi-
cate each of these diseases and in the political circumstances
that prevailed during the period in which each initiative was
conducted. The smallpox vaccine was made from the vaccinia
Poliovirus vaccine—live 28 641
virus rather than the smallpox virus itself and caused a very high
rate of severe side effects (ie, up to one severe adverse event per
25,000 doses administered),588 but it could not mutate back into
smallpox virus or establish endemicity through person-to-person
transmission. Furthermore, the disease was eliminated at a time
when concerns about the potential deliberate use of smallpox
virus to cause harm were much less than they are today.
In contrast, OPV, the live attenuated poliovirus vaccine used
in the eradication initiative, can mutate back into a poliovi-
rus that is biologically indistinguishable from wild poliovirus.
Because of this innate characteristic of the Sabin strains con-
tained in OPV, an end-game program of work has been defined
to ensure that OPV viruses will not simply replace wild polio-
viruses (after these wild polioviruses have been eradicated) and
reestablish global endemic and epidemic transmission of polio-
virus. Without addressing these issues successfully, the ulti-
mate goal of polio eradication will be elusive.
The end-game area of work has evolved greatly during the
past decade. Until the late 1990s, discussions on long-term
polio immunization policy were primarily driven by the human-
itarian and economic benefits of OPV cessation. In particular,
it was widely thought that once wild poliovirus had been inter-
rupted globally, the public health benefits of OPV would no
longer outweigh the estimated 250 to 500 cases of VAPP world-
wide that would continue to occur each year based on current
OPV vaccine utilization patterns.589 As progress toward global
polio eradication advanced in the late 1990s and the risk of
wild poliovirus importations declined, industrialized countries
began switching from OPV to IPV for routine childhood immu-
nization to avoid VAPP,545 despite higher costs for IPV.590
More recently, however, evidence emerged of a new OPV-
associated risk that is of even greater significance than VAPP
for long-term polio immunization policy. Specifically, it has
been shown that the Sabin strain polioviruses in OPV can
regain neurovirulence and the capacity to circulate and cause
outbreaks65 (see the discussion of cVDPV). Between 2000 and
2010, such cVDPVs caused a total of 16 outbreaks on three
continents (Africa, Asia, and the Americas) (Table 28-14).
One confirmed and other probable cVDPV-associated polio
outbreaks have been described retrospectively.258 Such out-
breaks demonstrate that after wild poliovirus eradication, the
continued use of OPV would generate cVPDVs on a regular
basis, the spread of which would eventually negate the eradi-
cation achievement.
An additional risk associated with continued OPV use after
global eradication is the generation of new long-term excret-
ers of VDPVs who might subsequently reinfect an increasingly
susceptible human population.499 Prolonged VDPV excretion
( 6 months) occurs only rarely and almost always in persons
with certain primary B cell–related immunodeficiency syn-
dromes (iVPDVs). With more than 45 years of OPV use, none
of the 45 iVDPVs documented as of January 2010 have led to
any known secondary cases of paralytic poliomyelitis, although
there has been at least one documented case of asymptom-
atic infection of contacts496 (see iVDPV) (Table 28-4). Acquired
immunodeficiency syndromes involving T cells, such as that
associated with HIV infection, do not seem to be associated
with prolonged poliovirus excretion or iVDPVs.532
Recognizing that paralytic poliomyelitis cases and out-
breaks would continue as long as there is routine OPV use,
expert committees since 1998 have recommended the even-
tual, simultaneous cessation of all routine OPV immuniza-
tion, as soon as possible after confirmation of wild poliovirus
eradication.477,591,592
Major risks associated with OPV cessation
While there are compelling benefits to eventually stopping
OPV, these must be weighed against the associated risks.
The risks can be considered in three categories: the live
642 SECTION TWO
attenuated poliovirus strains used for OPV production (eg,
Sabin strains), VDPVs, and wild polioviruses. In each case,
the risk is of human infection and subsequent transmission
following ingestion of the virus. Poliovirus survival in the
environment is finite; infectivity is lost during a period of
days, under hot dry conditions, to weeks or several months
under cool, moist conditions.204
The term “Sabin strain” poliovirus denotes one of the three
live attenuated viruses that Albert Sabin developed for the oral
poliovirus vaccine (see “Vaccine development”). Sabin strain
polioviruses are ubiquitous due to routine immunization pro-
grams that currently use approximately 2 billion doses of OPV
each year globally to vaccinate children in more than 150 coun-
tries. Such live attenuated polioviruses are present in a variety
of other settings owing to their use as seed viruses for OPV pro-
duction, as reference standards in vaccine quality assurance and
control testing, as controls for some polio diagnostic tests, and
for basic research in a number of laboratories and for teaching
purposes in some academic centers. While the vast majority of
OPV recipients will lead to time-limited (ie, 3-4 weeks) intesti-
nal replication and shedding of the virus, Sabin viruses rarely
can give rise to cVDPVs or generate an iVDPV. Of particular
importance for risk management planning is the danger posed by
the emergence of one or more cVDPVs immediately after coun-
tries stop using OPV, when population immunity will already be
starting to decline. Mathematical modeling suggests that even
with simultaneous cessation of OPV use worldwide, a 60% to
95% chance exists of at least one such outbreak occurring some-
where in the world during the 12 months immediately after ces-
sation, with that risk declining to 1% to 6% at 36 months and
much lower thereafter.591 To decrease the risks of cVDPV emer-
gence, simultaneous OPV cessation worldwide will be of critical
importance. Should cVDPV outbreaks occur, mOPV will be used
for rapid type-specific control and elimination. However, the use
of mOPV itself may be associated with generating cVDPVs. To
minimize this risk, high coverage with mOPV will be essential,
and other strategies such as supplemental IPV or use of antiviral
compounds may be needed in a ring campaign around the initial
mOPV-targeted areas.593 After global cessation of OPV, cohorts
susceptible to poliomyelitis will accumulate rapidly, and the risk
assessment would likely change.594
Since 2011, a process to define a “new polio endgame strat-
egy” is being led by a SAGE Working Group on Polio Vaccines.
This re-thinking of the endgame discusses the discontinuation
of Sabin type 2 from all routine OPV, associated with a switch
from tOPV to bOPV, and the introduction of a dose of IPV (frac-
tional or full dose) with DTP3 at or around age 14 weeks for risk
mitigation purposes. This switch should greatly reduce the risk
of VAPP and the emergence and spread of cVDPVs.
Retaining wild polioviruses in laboratories represents another
concern. These are currently used as seed viruses for the produc-
tion of IPV, in vaccine quality control and assurance testing, as
controls for some diagnostic tests, and in research laboratories.
Although wild polioviruses are no longer nearly as ubiquitous as
Sabin polioviruses, the consequences of an inadvertent or inten-
tional release in the post-OPV era pose a potentially far greater
threat. A recent consequence assessment suggests that whereas
transmission of a wild or Sabin strain virus may be self-limiting
if released into an IPV-vaccinated population in a moderate cli-
mate zone in the post-OPV era, a wild poliovirus released into an
unvaccinated population in a tropical developing country would
almost certainly result in a large-scale outbreak with a real risk
of eventually reestablishing endemic transmission globally.594
As a final concern, full-length poliovirus cDNA has been
synthesized by assembling oligonucleotides of plus and minus
strand polarity. The synthetic poliovirus cDNA was transcribed
by RNA polymerase into viral RNA that translated and rep-
licated in a cell-free extract, resulting in de novo synthesis of
infectious poliovirus,595 demonstrating that, even if poliovirus
Evolution of risk of poliovirus following discontinuation of
oral poliovirus vaccine (OPV) among high- and low-income countries.
has become extinct (complete destruction of all virus globally),
laboratories could create fully infectious polioviruses.
Figure 28-21 summarizes, based on current knowledge, the
expected evolution of these risks in low- and high-income coun-
tries during a probable 3 to 5 year “OPV cessation” phase fol-
lowing confirmation of global interruption of wild poliovirus
transmission and appropriate biocontainment of all poliovirus
stocks globally.
Risk reduction and management
A comprehensive approach must be taken to optimize the man-
agement of the risks associated with OPV cessation. The core
principles forming such an approach must be to: (1) simultane-
ously stop the routine use of OPV worldwide, while population
immunity is high, with the subsequent recall and destruction
of remaining OPV stocks; (2) reduce the number of procedures
performed with polioviruses to those that are essential in the
post-OPV era; (3) replace, where possible, wild polioviruses
with Sabin strain viruses for any procedure that must continue
to be conducted in the post-OPV era; (4) minimize the num-
ber of sites handling or storing polioviruses or potentially
poliovirus-infectious materials (wild viruses before OPV cessa-
tion; Sabin strains immediately thereafter), and limit these sites
to areas where the consequences of an inadvertent release could
be minimized; (5) institute processes to ensure that residual
poliovirus-containing sites fully implement appropriate bio-
containment and biosafety procedures; (6) maintain high-level
surveillance to identify and monitor iVDPVs and to detect an
inadvertent or intentional release of any poliovirus into the
human population; and (7) establish a stockpile of monovalent
OPVs with internationally agreed-on criteria for their use in
mounting type-specific outbreak responses to circulating polio-
viruses in the post-OPV era.
Minimizing the risks associated with OPV cessation would
require establishing international concurrence through a bind-
ing agreement. The WHO is exploring a process for an amend-
ment to an already existing treaty, such as the International
Health Regulations, to apply these core principles in all areas
of all countries in the world. Additional risk reduction or risk
management strategies may be required in areas that pose par-
ticular risks during or after OPV cessation. For example, OPV-
using areas with large, high-density populations and low routine
immunization coverage may be at high risk for generating cVD-
PVs and need special attention, such as maximizing the popula-
tion immunity before stopping OPV. Countries or communities
with facilities that continue to store or handle polioviruses in
the post-OPV era for IPV production and quality assurance pur-
poses, to provide international diagnostic services, or to conduct
specialized research may constitute an international biohazard

and require rather extraordinary measures to prevent or mini-
mize the consequences of inadvertent virus release.
These principles have been incorporated into a set of pre-
requisites for eventually stopping the routine use of OPV for
childhood immunization globally.596 Six prerequisites have
been defined and must be met before OPV vaccination can be
stopped, including assurance that (1) wild poliovirus transmis-
sion has been interrupted and containment of wild poliovi-
ruses has been achieved, (2) global surveillance and notification
capacity is maintained, (3) a global stockpile of monovalent
OPV and a response capacity have been established, (4) the IPV
vaccination requirements have been implemented in biohazard
settings,597,598 (5) cessation of OPV has been synchronized glob-
ally, and (6) Sabin polioviruses are contained.
Interruption of wild transmission and containment of
wild poliovirus
The first and most critical prerequisite before OPV can be dis-
continued is to ensure that wild poliovirus has been eradicated
globally. The processes for certification are well-established
(see “Certification of polio eradication”), and containment of
wild poliovirus is an integral part of certification. The objec-
tive of containment is to minimize the possibility that polio-
virus will be reintroduced into the community. Laboratories
holding poliovirus or potentially infectious materials need to
ensure safe handling and ultimately appropriate laboratory con-
tainment. At present, containment activities focus on a survey
of laboratories and inventory for wild poliovirus and potentially
infectious materials. Eventually, appropriate containment will
be needed for all polioviruses (Sabin, VDPVs, and wild poliovi-
rus), whether in laboratories or in vaccine production facilities.
Global surveillance and notification capacity
Surveillance strategies relying on AFP to detect all paralytic cases
will need to continue for at least 5 years after interruption of wild
poliovirus circulation (2-3 years before OPV cessation and 3 years
after detection of the last cVDPVs). At that point, the risk of unde-
tected cVDPV should be negligible. Long-term surveillance strat-
egies would then focus on the detection of events (ie, clusters of
paralytic cases), the prompt epidemiologic and virologic investiga-
tion of such events, and the institution of effective control mea-
sures. Future surveillance efforts will also focus on identifying
persons with iVDPVs. The use of environmental surveillance (ie,
sewage sampling) in selected areas to supplement AFP surveil-
lance is being considered. The study in New Zealand conducted
before and after the switch to universal IPV provides useful infor-
mation about the kinetics of vaccine virus disappearance.599 The
New Zealand study relied on three monitoring systems, pediatric
inpatient surveillance, AFP surveillance, and environmental sur-
veillance, to detect Sabin-like viruses before, during, and after the
switch from OPV to IPV. One month after the switch, no Sabin
viruses were detected from pediatric inpatient and AFP surveil-
lance but environmental surveillance detected Sabin-like viruses
until May 2002 (4 months after the switch). After May 2002,
Sabin-like viruses were isolated infrequently, and genetic studies
showed that these isolates represented recent importations rather
than continued circulation of Sabin-like viruses. Thus, in a devel-
oped, temperate-climate country, Sabin-like viruses did not cir-
culate for more than 4 months after the switch from OPV to IPV.
Vaccine stockpile and response capacity
This area of work has two major objectives: (1) short-term, to
rapidly control outbreaks of cVDPVs, should these occur dur-
ing a 3 to 5 year transition period after OPV cessation; and (2)
long-term, to minimize the consequences of poliovirus reintro-
duction into the community until polio immunization can be
reestablished globally. Many countries are not expected to insti-
tute routine immunization with IPV after discontinuation of
OPV. Therefore, in these countries, infants will no longer be vac-
cinated against polio, and population immunity to polioviruses
Poliovirus vaccine—live 28 643
will decrease over time. If poliovirus is reintroduced into these
countries, rapid control efforts using large quantities of mOPV
will be required for outbreak control.
The WHO is planning the establishment of a vaccine stockpile
relying primarily on mOPV to allow type-specific responses to the
reintroduction of one or more poliovirus serotypes, maximize the
immune response, and avoid introduction of unnecessary vaccine
poliovirus serotypes. Since 2004-2005, mOPV1 and mOPV3 have
been licensed in several countries and widely used (see the discus-
sion of mOPV1 development). More recently, mOPV2 has been
licensed by a manufacturer and a filler. The WHO is also assessing
the role of IPV in such settings, particularly to boost population
immunity in high-risk areas not yet affected by such an outbreak.
A stockpile of polio vaccine is necessary for the period after
OPV cessation. Many of the basic stockpile issues, such as
the size, the vaccine (mOPV) and the composition (filled vac-
cine vials, bulk, or a combination), number of storage sites, and
release protocols (eg, who retains the authority to release vac-
cine) are being refined.600 The primary purpose of a stockpile and
response capacity is to ensure that control efforts can be insti-
tuted immediately after detection of cVDPV at any time during
the 3 to 5 year period following OPV cessation. For the longer
term after OPV cessation, other options need to be explored, such
as restarting the production of polio vaccines. The immunogenic-
ity and the kinetics of the immune response of stockpile vaccines
become critical decision determinants. Monovalent (ie, type-
specific) OPVs seems to be the most appropriate stockpile vac-
cines337 for the following reasons: (1) Immunogenicity per dose
is substantially higher compared with tOPV or IPV (one dose
of monovalent OPV may provide an immune response to the
specific serotype that is equivalent to three doses of tOPV). (2)
Monovalent OPVs are faster in inducing a type-specific immune
response (one would not have to wait for administration of sev-
eral doses). (3) Their use would be directed against a specific sero-
type of poliovirus, without introducing unnecessary serotypes
of vaccine virus. Since December 2004, regulatory approval for
mOPV1, mOPV2, and mOPV3 has been obtained (Table 28-10).
IPV vaccination requirements
Guidelines have been prepared to facilitate national decision
making on OPV cessation. These guidelines include a discus-
sion of the risks and benefits of discontinuing OPV use (and
OPV production), the need for comprehensive discussions and
decision making at the national level, and the implications of
IPV introduction. In addition, the WHO has prepared position
statements on IPV for the preeradication and posteradication
eras that provide guidance.575,597,598 Furthermore, a Working
Group of the SAGE, the main technical oversight committee
for immunization at the WHO, is responsible for drafting rec-
ommendations for polio vaccine use in the posteradication era,
with an expected due date in 2012. The most recent IPV posi-
tion statement for the posteradication era introduces the follow-
ing: (1) the need for countries that plan to retain poliovirus after
eradication to establish and/or maintain high population immu-
nity against polioviruses through vaccination with IPV,601,602 and
(2) a one to two dose schedule of IPV for countries that do not
retain virus but want to maintain an immunity base against
poliovirus. However, recent work suggests that a one-dose frac-
tional IPV given in the second half of the first year (to reduce
the influence of maternally derived antibody on IPV immunoge-
nicity) may provide a low-cost option to induce and maintain a
basic immunity against all three poliovirus serotypes.
Absent from the core principles outlined is the use of IPV
for universal childhood immunization. Although universal IPV
childhood immunization has been proposed as a potential solu-
tion to the risk of cVDPV emergence at the time of OPV ces-
sation, mathematical modeling suggests that IPV would only
partially reduce the already small risk of a cVDPV in most coun-
tries.603 Routine infant immunization with IPV would not sub-
644 SECTION TWO Licensed vaccines
stantially mitigate the consequences of poliovirus reintroduction
in terms of spread in countries with low routine coverage, such
as much of sub-Saharan Africa,604 but would reduce the para-
lytic risk among IPV vaccinees.604a Consequently, countries must
decide at the national level whether to stop all polio vaccination
or switch to IPV based on whether there is a real or perceived
need to maintain population immunity against polioviruses
indefinitely. If high IPV routine coverage could be achieved every-
where, it is possible that the mucosal immunity induced by IPV
could have substantial influence on poliovirus transmission.
Policy makers in each country must balance their national
willingness to pay for IPV to maintain population immunity to
polioviruses against the financial, programmatic, and opportu-
nity costs of introducing IPV, the true costs of which may not be
immediately apparent, particularly for resource-poor areas.598 In
financial terms alone, UNICEF currently procures IPV at 5 times
the estimated break-even cost, the cost needed to achieve com-
parable immunity in populations with OPV vs IPV.605 Despite
anticipated reductions in the unit price of IPV for low-income
countries, the substantial opportunity costs associated with
the use of scarce health resources for that vaccine rather than,
for example, directing those resources to combat HIV, malaria,
tuberculosis, measles, and pneumococcal and rotavirus diseases,
will strongly influence decision making at the national level.
Based on their review of the implications, costs, and benefits of
IPV introduction for routine immunization, some low-income
countries have indicated that the advantages of stopping all polio
immunization currently outweigh the short-term risk of cVDPV
emergence and the longer term risks of poliovirus reintroduc-
tion. In contrast, some middle-income, OPV-using countries are
considering the introduction of routine infant immunization
with IPV as a transition strategy to maintain population immu-
nity against polio during the 3 to 5 year period of OPV cessation
and verification of the absence of cVDPVs worldwide.
Despite these rational arguments, it is imperative that the
WHO strengthen the scientific basis for IPV use in develop-
ing countries and devise innovative new strategies for IPV
affordable for use in these countries, including dose-reduction
schedules and dose-sparing approaches, to allow low-income
countries to decide on future IPV vaccination policies without
undue financial implications.606 The main objective is to pro-
vide a cost-neutral IPV option (cost-neutral with routine OPV)
so that IPV cost will not be a major consideration or obstacle for
countries that may want to use it.
Coordinated OPV cessation
To minimize the risk of emergence of and exposure to cVDPVs
during the period following OPV cessation, all countries will
need to stop the use of OPV during a relatively short period
(ideally, a few weeks) everywhere, and all must institute mecha-
nisms to ensure that OPV from throughout the health system
has been recalled and destroyed. There must be agreement that
no country will again use OPV, unless specifically endorsed by
the international community for control of an outbreak.
Sabin poliovirus containment
Shortly after global OPV cessation, Sabin polioviruses must
be contained. The WHO is preparing a Global Action Plan to
Minimize Facility-Associated Risk in the Post Eradication/OPV
Era, which will be subject to broad public comment in 2012.607
The communication of the end-game issues is another very
important objective of the GPEI. It is important to understand
the current thinking and to appreciate that it is a work in prog-
ress, and that new scientific data, programmatic experiences,
and expert advice all will feed into a program of work that allows
the adoption of policies that minimize the paralytic burden of
poliomyelitis to current and future generations of children.608,609
There have been voices that have questioned recently whether
eradication is feasible.610,611 However, if the program learned any-
thing during the past few years, it is that if the eradication efforts
are stopped, the world will have to live with hundreds of thou-
sands of cases of paralytic disease again because routine immu-
nization coverage in many parts of the world has not improved
substantially since 1988. Only the aggressive use of mass vac-
cination campaigns has led to control and elimination of polio
transmission in many parts of the world. The experiences from
the recent outbreaks, particularly in Yemen, where imported
poliovirus caused 479 cases in a country with suboptimal rou-
tine immunization, and interruption of mass vaccination cam-
paigns for more than 1 year present an example of how the
poliovirus can exploit the immunity gaps in the population.
Similar large-scale outbreaks occurred in Tajikistan174 and
Congo-Brazzaville175,200 in 2010 due to critical immunity gaps
against polioviruses. To achieve eradication, such importations
need to be controlled rapidly and, still better, prevented by elimi-
nating the reservoirs in polio-endemic countries.
A process that started with the constitution of an International
Commission for the Certification of Polio Eradication in the WHO
Region of the Americas in 1990 (which certified the entire Western
Hemisphere free of polio in 1994)428 is being replicated in each of six
WHO regions, guided by the Global Certification Commission.612
The Commission in the Americas defined four criteria on which
confirmation of poliovirus eradication could be assessed: (1) the
absence of virologically confirmed cases for a period of 3 years in the
presence of adequate surveillance; (2) the absence of detected wild
poliovirus in tests of stools from healthy children (eg, from the con-
tacts of cases of AFP being investigated) and, when indicated, from
waste water; (3) evaluation by a National Certification Committee
convened for that purpose in the country, eventually reporting to
the Regional Certification Commission; and (4) establishment of
appropriate measures to deal with wild poliovirus importations.
In recently polio-endemic countries, the certification process
will rely primarily on data from AFP surveillance; in countries
that have been free of poliovirus for many years (and that have
not implemented AFP surveillance), the process will evaluate
data from all relevant sources (ie, VAPP surveillance, virologic
surveillance, environmental surveillance [eg, Finland], adverse
events reporting systems). It is not clear what role environmen-
tal surveillance will have in the certification process of recently
polio-endemic countries. Ongoing evaluations in Egypt and
India have highlighted the strength of this form of surveillance
in confirming wild poliovirus circulation in areas of Egypt where
AFP surveillance has not detected cases and demonstrating fre-
quent episodes of importation of wild poliovirus into Mumbai
from an endemic state in Northern India.
In addition to the Americas, the process of certification has been
concluded in the Western Pacific Region (a region that includes
China, the world's most populous country) in 2000428 and the
European Region in 2002.430 The experience with certification in
these three regions has demonstrated that the criteria applied are
sound, and no wild poliovirus previously indigenous to any of these
three regions had reemerged from a certified region as of July 2006.
The Global Commission for the Certification of Polio
Eradication is responsible for certifying the world as free of wild
polioviruses. It also started discussions in 2001613 on the impli-
cations of the outbreaks of cVDPVs in Haiti and the Dominican
Republic65 for the global certification process but postponed a
decision about whether to extend its mandate and include cVD-
PVs as part of the process (see the discussion of cVDPV).
In 1997, the second meeting of the Global Certification
Commission added the requirement to effectively contain any
laboratory stocks of wild poliovirus as a condition for eventual
global certification.614 In contrast with the smallpox eradica-
tion program, in which the virus was restricted to a selected
group of laboratories, poliovirus is used in many laboratories

initial replication may occur in muscle cells surrounding the
wound, providing an amplification of the original inoculum.
However, experimental data show that CNS entry can occur
without any prior replication in the muscle.26,73 Another site
proposed for possible harborage of rabies virus before entry
into the CNS is the macrophage, from which the virus could
reactivate to cause disease,74 but the importance of replica-
tion in nonnerve cells to the pathogenesis of rabies remains
highly controversial.75 In any event, at some point the virus
enters nerve cells through nerve spindles of sensory nerves
or neuromuscular junctions of motor nerves.76 Rabies virus
G protein has sequences similar to certain neurotoxins,77–79
and there are many viral receptors, including the subunit
of the nicotinic acetylcholine receptors of the neuromuscular
junction,77–81 the neural cell adhesion molecule, the neuro-
trophin receptor, p75NTR, and, perhaps, certain lipoproteins
on the cell membrane. The virus then begins a journey to the
neuronal cytosol, where it replicates and spreads within the
CNS.82,83 Dietzschold and colleagues84 demonstrated that the
action of rabies VNAs is not exerted solely outside the cell.
In an animal model, the effectiveness of antibody was asso-
ciated with entry into the cell by endocytosis and inhibition
of viral transcription. Whether the antibody acts directly or
by signal transduction to inhibit viral protein synthesis is
unclear.85
Once in the neurons, the virus travels rapidly within the
axons at a rate of 8 to 20 mm/d in rodents, probably faster
in humans (15-100 mm/d). Experimental studies in rodents
showed that virus may reach the CNS in 3 to 5 days, where it
causes a widespread encephalitis.86
After establishment in the neurons of the brain, the virus
starts to move in the opposite direction, down the axons to rep-
licate in peripheral tissues, most notably in the nerve plexus
and acinar cells of the salivary glands, from which excretion in
saliva permits viral transmission by bite to maintain the cir-
cuit of infection.58 However, at the end stage of infection, other
extraneural tissues also are affected, including the heart, pan-
creas, adrenal glands, and gastrointestinal tract.87,88
Moreover, the finding of virus disseminated throughout the
body in patients who acquired rabies from virus-infected organ
transplants (see “Nonbite transmission”) raises questions about
the possibility of altered pathogenesis. The presence of virus in
nonneuronal cells may also explain cases with long incubation
periods.
The pathophysiology of the fatal outcome is not completely
understood.89 Although encephalitis is widespread, neuronal
destruction is not.90 Death probably results from the involve-
ment and dysfunction of brain centers controlling the cardio-
respiratory system. In general, the histologic presence of Negri
bodies parallels that of rabies virus antigen, although many
infected cells do not have these inclusions. Rabies virus anti-
gen is most prevalent in the periaqueductal gray matter and
the Purkinje cells of the cerebellum, but the quantity of rabies
virus does not correlate with severity of symptoms.91,92 Virulent
rabies virus variants may be more capable of evading host
innate immunity and of destroying neuronal processes.93,94 The
production of nitric oxide within the brain and general down-
regulation of host genes have been advanced as possible expla-
nations for brain dysfunction.26 Although the fatality rate in
rabies is extremely elevated, recovery has been documented in
some animals, such as the dog.95
The neural mechanisms that distinguish furious from par-
alytic rabies are not well understood, but research, including
electrophysiologic studies, suggests that denervation and ante-
rior horn cell dysfunction are prominent in the former, whereas
in paralytic rabies, there is inflammation and demyelination of
peripheral nerves.96,97 The mechanism causing demyelination
could involve autoimmunity or a bystander effect of immune
response to virus in the axons.5 Hemachudha and colleagues5
Rabies vaccines 29 649
emphasized the role of cellular immunity in the course of
rabies. They correlate strong T-cell responses and cytokine
secretion (in particular, interleukin-6) with early death and the
encephalitic form of the disease, and weak T-cell responses
with paralysis and longer survival. Whereas it is certain that
lack of an immune response to rabies virus, particularly anti-
bodies, leads to fatal disease, there is thus some evidence that,
on the contrary, some cellular responses are responsible for neu-
ral pathology.98,99
Evidence of a serologic response to rabies virus can be dem-
onstrated by a variety of laboratory techniques, including
mouse neutralization, fluorescent focus inhibition, indirect flu-
orescent antibody, plaque neutralization, immunolysis of rabies
virus–infected cells, and binding techniques using radioimmu-
noassay or enzyme-linked immunosorbent assay procedures.
Serum antibodies develop relatively late after natural infection
in humans. In persons without a history of vaccination, serum
antibodies are first detected on or about the 6th to 10th day of
illness and thereafter can rise rapidly to high levels. Antibodies
are also present in CSF later in the clinical course. Antibody
titers of CSF are higher than would be expected from seepage
into CSF from circulating blood, indicating local production.
Because vaccination does not ordinarily induce CSF antibod-
ies, the presence of CSF antibody titers supports the diagnosis
of clinical rabies.18,19
The absence of most detectable serum antibodies until
around the second week of illness (if at all) and of CSF anti-
bodies until approximately the third week of illness (when sig-
nificant systemic and neurologic problems occur) raises the
possibility that some of the clinical symptoms result from the
interaction of host antibodies with rabies virus–infected cells.82
In mouse experiments, neutralizing antibodies and infiltration
of cellular inflammatory cells were necessary to clear infection
with an attenuated rabies from the CNS,100 but this process
occurs too late in the usual situation.
Most persons who die of rabies develop specific antibodies
to the virus, particularly if they received PEP. However, that
response does not protect against a lethal outcome and might
possibly even contribute to the disease. Experimental find-
ings suggest that the inflammatory response is associated with
opening the blood-brain barrier, delivering antibody-producing
B cells to the CNS. Nevertheless, this response usually fails or
occurs too late to be therapeutic.
Diagnosis
The history of an associated bite from a known or suspected
rabid animal, coupled with the striking clinical manifesta-
tions, should provide a reasonably simple diagnosis of rabies.
However, such straightforward attributes are not always pres-
ent. Especially in absence of a documented exposure source,
clinical diagnosis of rabies requires differentiation from a wide
variety of diseases that can cause neurologic symptoms. Because
laboratory diagnosis may not be possible during the first week
of illness, presumptive diagnosis based on clinical symptoms
is important. As noted previously, the clinical symptoms that
may distinguish rabies from other forms of encephalitis are as
follows: pain and paresthesia near the site of exposure, hydro-
phobia, hypersalivation, hyperventilation, agitated behavior,
asymmetric or ascending paralysis, and aerophobia, all develop-
ing typically during a 2 to 10 day period. Detection of lesions
in the brainstem, hippocampus, and thalamus by magnetic
resonance imaging is suggestive of rabies.5 These symptoms,
signs, findings, and a history of exposure to a rabid animal
strongly suggest the diagnosis of rabies. Other diagnoses that
figure prominently in the differential diagnosis of furious rabies
include porphyria, tetanus, and drug intoxication; for paralytic
rabies, Guillain-Barré syndrome, poliomyelitis, and Japanese
650 SECTION TWO Licensed vaccines
encephalitis should be considered.13 More recently, West Nile
virus, enterovirus 71, and Nipah virus all should enter into
the differential diagnosis, as appropriate. Magnetic resonance
imaging shows hypersignals in the brainstem, hippocampus,
and gray and white matter in the cortex.16 Definitive diagno-
sis of rabies virus infection of humans and suspected animal
vectors depends on the detection and identification in infected
brain tissue of rabies virus antigens or suspected intracytoplas-
mic neuronal inclusions (Negri bodies); of viral nucleic acid by
reverse transcriptase—polymerase chain reaction (RT-PCR),101
on the presence of rabies virus-specific antibodies in the CSF,
or in the serum of unvaccinated patients; and on the isola-
tion and identification of the virus from brain tissue, saliva, or
other infected substances. The standard diagnostic technique
is to search for rabies virus antigens in brain tissue by fluores-
cent antibody staining. The brainstem, cerebellum, and thala-
mus provide some of the best samples.102 Identification of viral
nucleic acid by RT-PCR is useful, particularly if specimens are
in poor condition. Demonstration of Negri bodies has a variable
sensitivity and is of only historical interest, whereas virus isola-
tion is a procedure used for confirmation of other positive test
results. Isolation can be accomplished in mouse neuroblastoma
cell culture or by intracerebral inoculation of suckling mice.103
Although diagnostic procedures generally are initiated in tis-
sue specimens obtained postmortem, rabies virus infection also
can be identified in vivo during the extended course of the dis-
ease. In addition to its usefulness for brain biopsy specimens,
the fluorescent antibody staining technique enables detection of
viral antigens in impressions of cornea or in cryoscopic sections
of skin biopsy samples from the hairline of the neck, where
antigens can be detected in the nerves surrounding the hair
follicles.104
In addition, as shown in Table 29-1, RT-PCR on saliva pro-
vides a rapid and sensitive test for rabies as early as 5 days after
Antemortem Diagnostic Test Results for 20 Human Patients
with Rabies in the United States, 1980-1996*
*Data are from reverse transcriptase studies.
†
Two patients had earlier skin biopsies that were negative but became
positive on subsequent biopsy.
‡
One patient had an earlier test that was negative.
§
Latest negative result on day 24; median to positive result was 10 days.
Latest negative result on day 24.
CSF, cerebrospinal fluid; RT-PCR, reverse transcriptase–polymerase chain
reaction.
Data from Noah DL, Drenzek CL, Smith JS, et al. Epidemiology of human
rabies in the United States, 1980 to 1996. Ann Intern Med 128:922-930,
1998.
onset of symptoms.105,106 Improved sensitivity was obtained
when an RNA polymerase was included in the reaction and elec-
trochemiluminescence was used for detection of product. This
technique, called nucleic acid–based sequence amplification,
detected the rabies virus genome in saliva, concentrated urine
sediment, and CSF as early as 2 or 3 days after symptoms.107,108
Serial tests should be performed to increase sensitivity.108 A
study in Thailand showed high sensitivity (91%) of rabies virus
genome detection on the first day of hospitalization, particu-
larly in saliva.109 CSF and concentrated urine sediment speci-
mens were also useful. However, the genome was not detected
in all patients before death. A latex agglutination test for viral
antigen in dog saliva also has been developed.110 One report
found rabies virus–specific immune complexes in the CSF of
77% of patients with rabies 5 to 7 days after onset of disease.111
Between 1980 and 2005, more than 65 cases of human rabies
were diagnosed in the United States, but not all were diagnosed
before death, probably because a minority had a known history
of exposure to potentially rabid animals. This problem arises in
large part because many of the patients had been infected with
rabies virus variants from bats, possibly as a result of undocu-
mented or ignored bat bites.112
Epizootiology and epidemiology
Mammals
Rabies is a disease of domestic and wild mammals, particu-
larly dogs and related canid species, and raccoons, mongooses,
skunks, and bats. In areas in which domestic animal control
programs are not extensively developed, dogs and cats account
for most of the rabid animals reported and cause the major-
ity ( 90%) of human rabies exposures and deaths. After effec-
tive domestic animal rabies control programs in these areas, the
numbers of rabid dogs and cats markedly decrease, as illustrated
in the United States from the 1940s to the 1960s. Wildlife is
then recognized as the main reservoir of rabies virus. In the
United States since 1960, the majority of cases of animal rabies
have been in wildlife species, and most human rabies cases have
been secondary to bites by rabid wildlife, especially bats.11,113
Luckily, bats do not often transmit virus to dogs or cats.114 The
situation is similar in Western Europe, where domestic species
accounted for only 28% of 8,155 reported rabid animals. Of the
wild species, foxes accounted for 83% and raccoon dogs, 11%.115
Figure 29-1 is a composite map of the United States show-
ing the major carnivore reservoirs in each region.116,117 The
salient features are that skunks are the important vectors for
rabies in the western and central United States, whereas rac-
coon rabies dominates in the east. The original focus of raccoon
rabies was in the southeast, but a second front developed in the
mid-Atlantic and northeast owing to translocation of infected
animals, and now raccoon rabies is contiguous from Maine to
Florida and west to Ohio.118 Foci of red fox rabies are evident in
Canada and Alaska. Rabies in Puerto Rico is due to the mon-
goose. Rabies control is costly: In New York State alone, the cost
of rabies prevention in humans was calculated to be $2.3 mil-
lion per year.119 Table 29-2 lists rabies cases by species for the
United States in 2009.117
Table 29-3 lists the principal animal vectors of rabies
throughout the world.106 In Western Europe, foxes accounted
for up to 80% of rabid animals, but infection in raccoon dogs is
becoming more common.120,121 Molecular analysis of European
viruses suggests that rabies originated in dogs but now affects
red foxes and raccoon dogs.122 Infection in the latter animals
has gradually spread from east to west. Canine rabies is still
widespread in Asia, Africa, and parts of Latin America. Between
1993 and 2009, there was a significant drop in canine and

Cases of Animal Rabies in the United States, 2010
From Blanton JD, Palmer D, Rupprecht CE. Rabies surveillance in the United
States during 2009. J Am Vet Med Assoc 239:773-783, 2011.
human cases in Latin America.123 However, dogs remained the
predominant vector for humans (65%), followed by bats (15%)
and cats (3%). Vaccination of pet dogs is an effective strategy for
protection of humans and has eliminated terrestrial rabies in
Great Britain, Iceland, Japan, and many other islands.82,124 An
immunization coverage of 60% to 70% is estimated to prevent
canine rabies outbreaks.125
In support of historical and epidemiologic observations in
developed countries, one Thai study confirmed that rabid dogs
and cats uniformly die within 10 days of the onset of illness.126
Foxes are important vectors in Canada, Alaska, and the for-
mer Union of Soviet Socialist Republics. Mongooses imported
into the Caribbean Islands now form a reservoir for rabies, and
Rabies vaccines 29 651
Principal Animal Vectors of Rabies
mongooses serve the same role in southern Africa and parts of
Asia. The vampire bat is a major threat to livestock in Latin
America and has been involved in many biting incidents in
humans. Rabid cattle also may excrete rabies virus in saliva.127
Like all mammals, rodents are susceptible to infection82 but
are infrequently rabid, and human transmission of disease by
these animals has not resulted. In Thailand, 95% of the ani-
mals involved in biting incidents are dogs, often younger than
6 months,128 with cats accounting for another 3%. The remain-
ing 2% include monkeys, civets, tigers, and other animals,
which testifies to the wide host range of rabies virus. Dogs ran-
domly captured in Thailand developed rabies within 1 month
in 3% to 4% of cases, and, interestingly, there was serologic evi-
dence suggestive of prior rabies virus exposure or prior vaccina-
tion in about 15% to 20% of dogs.129
Rabies in animals (excluding bats) is absent in many islands
such as the United Kingdom, Australia, and Japan. Much of
Western Europe is also now “rabies-free” in carnivores, although
rabies still occurs in central and Eastern Europe. From 2000
to 2005, there were 45 nonimported cases of human rabies in
Europe, but 44 occurred in Eastern Europe, and the remain-
ing case was due to a bat Lyssavirus.130 Most of the large Asian
countries (such as China), Africa, the former Soviet Union, and
some parts of South America still report considerable rabies in
domestic animals.
Travelers are at risk, as confirmed by the report of 10 cases of
rabies in the United States acquired abroad between 1990 and
2006.117 The risk of rabies exposure in Nepal was calculated to
be 5.7 per 1,000 person-years for expatriates and 1.9 per 1,000
person-years for tourists.131 The epidemiology of rabies has been
revolutionized by viral typing with monoclonal antibodies since
the late 1970s.132 Panels of these antibodies, directed against
epitopes specific to isolates from different animal species and
from different geographic locations, are used to identify viruses.
Thus, it is now possible to identify the source of an isolate from
humans or animals and to demonstrate that the infection was
transmitted far away in time and place. Significantly, genetic
sequencing of the viral genome after RT-PCR amplification
has supplemented greatly our knowledge of viral variability.133
Vaccine fixed virus strains differ in sequence from wild “street
viruses” by as much as 10% to 15% of nucleotides. Sequence data
suggest that many viruses in the Western Hemisphere and in
South Africa, and elsewhere, were imported from Europe.134,135
Bats
Bats are not incidental vectors of rabies virus, but rather prob-
ably the original hosts of Lyssavirus, which in some distant
past developed epizootiologic cycles in other mammals by spill-
over infections.35 More genomic studies have supported this
idea. Widespread infection of insectivorous bats throughout the
 SECTION TWO
United States was well documented in the 1950s. However, the
importance of bat rabies, particularly from the silver-haired bat
, to human transmission has become
more evident (see “Human rabies”). The eastern tricolored bat
136
() is important in the East and Midwest.
Although most rabid bats appear ill,137 an infected bat may act
normally.138,139 The rabies virus recovered from the silver-haired
bat is one variant that seems to be able to replicate better in epi-
dermal tissues.140
Experimental data also suggest that virus from silver-haired
bats is more pathogenic for mice than other bat rabies viruses.141
Analysis of natural transmission to other mammals supports
the idea that rabies viruses associated with these two species are
more infectious than viruses found in other bats.136 The viru-
lence of the viruses from silver-haired bats may relate to block-
age of the passage of immune cells to the CNS.142
Bats are also important reservoirs of rabies for wild animals
in the United States, as are vampire bats in South America.143,144
Although neutralizing antibody titers are not always raised in
humans against rabies-related viruses maintained by bats,67
there is good cross-neutralization of bat rabies virus variants by
standard cell culture vaccines (CCVs).145
Nonbite transmission
Nonbite transmission has been reviewed by Gibbons,146 who
found 27 cases of transmission by means other than bites and
17 other less well-documented cases. Of the total 44 cases, 18
were caused by improperly inactivated vaccine in Brazil, 8 were
from corneal transplants, 8 from contamination of skin whose
integrity had been impaired, 4 by aerosols created in laboratory
or bat-infested caves, and 6 from human to human. Among
the alleged historical causes of human-to-human transmission
were transplacental passage, lactation, kissing, intercourse, and
providing health care. Gibbons146 also found three reports of
transmission by human bites.
Transmission to transplant recipients from organ donors is
particularly worrying. In one case, transmission occurred to
four patients who received kidneys, liver, and an artery segment
from a donor with undiagnosed rabies.147 A similar scenario was
reported for German transplant recipients in 2005. Thus, these
cases demonstrate that rabies virus is disseminated throughout
the body, at least late in the disease, and that direct implanta-
tion of infected tissue may transmit rabies to humans. Deaths
of suspected encephalitis should be investigated for rabies or
other etiologic agents before tissues from the donor are used.
Infection by aerosol has been suspected in unique cave habi-
tats inhabited by millions of bats and implicated under certain
laboratory conditions.11,148 Other lyssaviruses potentially could
also be transmitted by aerosol.149 Unwitting corneal transplan-
tation from patients who died of rabies also has resulted in
transmission.11,150 As mentioned, human-to-human transmis-
sion by bite is possible but extremely rare.146,151,152
Human rabies
The epidemiology of human rabies follows closely the epizo-
otiology of animal rabies. The dog is the major global reser-
voir of rabies. In the United States alone, more than 1 million
dog bites occur each year,153 and the situation is worse in many
other parts of the world.154 Human rabies has been reported
from all continents except Antarctica, but the majority of cases
occur in countries where canine rabies is not well controlled.
The World Health Organization (WHO) estimate of humans
vaccinated for exposure to rabies exceeds 10 million annually.155
The WHO also estimated between 35,000 and 50,000 human
rabies cases in 1997, most of them in India. The annual inci-
dence of rabies deaths per 100,000 population has been calcu-
lated as about 2 in India, 0.01 to 0.2 in Latin America, and an
uncertain 0.0001 to 13 in Africa.151,156 Human rabies is most
common in people younger than 15 years, with about 40% of
cases found in children 5 to 14 years, but all age groups are sus-
ceptible. The majority of rabies victims are male. In one study
in the United States, the highest incidence of human rabies
PEP occurred among rural boys, primarily during the summer
months.157
In the United States, the recent salient fact has been the
emergence of bats as the leading transmitter to humans.
Excluding the four transplant recipients infected by a common
donor, between 1990 and 2001, bats were associated in most
diagnosed human rabies cases of presumed domestic origin.136
There were also other cases associated with nonindigenous
canine rabies viruses during the same period.158 No definitive
contact with bats could be elicited in the majority of cases,
and only a handful of patients had a documented history of a
bite reported before diagnosis, suggesting that memory lapses,
neural impairment, and rarely, ignored or unperceived bites in
some persons may have been responsible.159–161 A similar situ-
ation exists in Canada, and this phenomenon should be opera-
tive in any areas in which bats serve as reservoirs.152
Passive immunization
Antiserum alone may not prevent rabies and is not recom-
mended except in combination with vaccine (see “Serum and
vaccine prophylaxis”).
Active immunization
Prior approaches
Table 29-4 summarizes the history of the development of rabies
vaccines and lists currently available vaccines. Many fixed
rabies virus strains have been used for preparation of rabies vac-
cines. The presumed history of some of these strains is given
in Figure 29-2.
For more than 70 years after Pasteur's original work, only vac-
cines containing nerve tissue were available. Major modifications
in nerve tissue vaccine preparation were introduced by Fermi162
and by Semple,163 who used phenol to partially or completely
inactivate virus. Adverse reactions to rabies vaccines containing
brain tissue have been recognized since the time of Pasteur. In
addition to neurologic complications attributed to the presence of
myelinated tissue in the vaccine, fixed virus may be pathogenic
for humans, contrary to the “Pasteurian dogma”, although it took
75 years before it was proved that some cases of paralysis after
vaccination were caused by imperfectly inactivated vaccine virus.
Adult animal neural tissue vaccines are discouraged but still used
in some developing countries, such as Ethiopia.164
Myelin-free vaccines prepared from neonatal mouse brains
were introduced by Fuenzalida and colleagues165 in 1956 and
are still used in parts of Latin America. Introduction of the
duck embryo vaccine (DEV), prepared from virus propagated
in embryonated duck eggs,166 greatly reduced the number and
severity of postvaccinal reactions, but DEV was less immuno-
genic than the brain tissue vaccine. For mouse brain and DEV
vaccines, 14 to 23 daily inoculations were recommended, but
even this “heroic” dosage did not always protect against rabies
after severe exposure. Thus, there had long been a pressing need
for a highly immunogenic antirabies vaccine that could be used
safely and effectively at low doses for primary immunization
and for prevention after exposure.

but untreated for various reasons, it seemed that the postex-
posure vaccination of severely exposed persons with vaccine
alone conferred insignificant protection.
The efficacy of the combined use of vaccine and rabies serum
to improve these results was established through studies of the
WHO Committee on Rabies. The superior results obtained
experimentally by Habel and Koprowski261 were confirmed in
a field study in Iran in 1954. Of 5 patients severely bitten by
rabid wolves and treated with vaccine alone, 3 contracted rabies
and died, whereas of 13 patients similarly bitten by the same
wolf and treated with vaccine and serum, only 1 died. Cho and
Lawson262 performed an experiment on postexposure rabies vac-
cination. When dogs were inoculated in the femoral muscle,
vaccine alone was not protective because of the short incuba-
tion period in this experimental system, whereas serum alone
protected about 50% and serum together with vaccine pro-
tected 100%. Similar results were reported by Hanlon et al,263
who used monoclonal antibodies against rabies as a successful
alternative to RIG. Baer and Cleary264 showed that HRIG has
synergistic activity with HDCV in a mouse challenge model.
Thus, these investigations demonstrated again the importance
of rapid protection with antibodies and the synergy of serum
with postexposure vaccination.
The intended purpose of RIG is to provide rabies VNA before
an active response to the vaccine occurs. Isotyping of antibody
after vaccination suggests that IgG antibody may not appear for
14 days.204 Evidence that the need for simultaneous RIG and
HDCV is not merely theoretical is provided by reports of rabies
that occurred after administration of HDCV without RIG.265,266
The first commercially available rabies antisera were produced
in horses, but approximately 40% of adult recipients developed
serum sickness.267,268 A purified and pepsin-treated equine rabies
immune globulin (ERIG) was developed by the Swiss Serum
Institute and Sanofi Pasteur and was widely used in Asia, but not
in the United States. It is associated with only a 1% rate of serum
sickness reactions,269 after a recommended dose of 40 IU/kg.
Wilde and coworkers269 and Tantawichien et al270 in Thailand
carefully studied the use of ERIG. In their experience, the risk
of anaphylaxis is only 1 in 35,000 people, and only 1% to 1.6%
of recipients develop serum sickness. Moreover, although a skin
test with 0.02 mL of a 1:100 dilution of ERIG is recommended
before use, and the response is positive in 5% to 10% of patients,
a positive test response was not predictive of serum sickness.
They consider a risk of anaphylaxis to be present only if there
is a wheal of 10 mm or greater in diameter or if a wheal of 5 to
10 mm is accompanied by a flare of 20 mm or more. A study of
286 Indian patients treated with ERIG showed no cases of ana-
phylaxis.271 To ensure safety from contaminating equine viruses,
a new heat-treated and more purified ERIG has been developed,
which should give fewer reactions.272,273 The old and new ERIG
contain F(ab )2 fragments rather than the whole equine globulin.
The kinetics of ERIG fractions vary substantially between intact
and fragmented immunoglobulin molecules, with obvious dif-
ferences in experimental PEP studies of animals. Moreover, in
the process of substitution of new for old, ERIG supplies have
diminished, creating shortages in some areas.274
A recent review of PEP in the Philippines, in which purified
ERIG was used, followed up 144 patients exposed to laboratory-
confirmed rabid animals. There were two deaths of rabies: one
child with multiple deep wounds on the neck and another who
did not also undergo vaccination.275
HRIG was prepared originally by Bayer (formerly Cutter, and
now Talecris) and Sanofi Pasteur from the plasma of volunteers
immunized with HDCV to provide an antibody preparation that
would not produce serum sickness. The products sold in the
United States currently are Imogam Rabies-HT (Sanofi Pasteur)
and Bayrab (Talecris Biological Products). Administration of
HRIG has not been associated with reactions other than occa-
sional local pain and low-grade fever.276 The dose is 20 IU/kg,
Rabies vaccines 29 659
with as much as possible of the volume instilled at the site of
the bite, and the remainder, if any, given into the muscle at
a distant site. The WHO recommends that HRIG be used in
preference to ERIG for all severe multiple bites.170 The recom-
mended dose should not be exceeded, because too much passive
antibody may have a dampening effect on the active antibody
responses.277 For the same reason, the dose should not be
repeated. However, if the HRIG (or ERIG) was not given imme-
diately, it should still be given up to the seventh day after expo-
sure, by which time an active response to vaccination should
begin. After administration of HRIG, serum neutralizing anti-
bodies can be detected within 24 hours, reach a level of about
0.1 IU at 3 days, and decay with a half-life of about 21 days.278 In
Thailand, patients started receiving vaccine up to 5 days before
coming for RIG treatment, but this delay was not accompanied
by suppression of the active immune response.272
An interesting case report mentions a patient with furi-
ous rabies who became paralyzed after receiving intravenous
HRIG.57 An autoimmune mechanism involving deposition of
immune complexes on axons has been suggested but not proved.
Local infiltration of antibodies is crucial because serum lev-
els of VNAs achieved after intramuscular administration alone
may be low or negligible.272,273 Local infiltration provides addi-
tional VNAs at the site of contamination with the virus. To
have a sufficient volume for infiltration in the case of multiple
bites, particularly in children, the RIG may be diluted in nor-
mal saline.279,280 Rabies developed in five children in whom local
infiltration was not performed as recommended.281
Current recommendations of the Centers for Disease Control
and Prevention (CDC) Advisory Committee on Immunization
Practices call for the local infiltration of the entire HRIG dose, if
feasible. If not, the remainder should be given intramuscularly
at a site distant from the vaccine injection. Where HRIG is not
available, ERIG, such as in its new chromatographically puri-
fied and heat-treated form, should be used.274
For a discussion of rabies monoclonal antibodies, see the
“Monoclonal antibodies” section.
Results of immunization
Immune responses
Nerve tissue vaccines are not optimal with regard to safety or
efficacy.282
Extensive studies have been conducted of antibody responses
to CCVs.24,107,128,137–139,145,149–153,155,157,159,161,164–169,173,176,182,183,186,191,192,
194,197,199,201,207,208,219,228–231,233–237,283–297
Several key findings of those
studies are as follows:
1. The most important immune response to rabies
vaccines is production of relevant antibodies to the G
protein of the viral envelope.36,288 Antibody is normally
measured by neutralization, such as the inhibition of
fluorescent foci. However, the choice of the virus used
in the neutralization test is important because the use
of a virus homologous to the vaccine strain may yield
30% higher titers than do heterologous viruses.298
2. Typically, IgM antibodies appear after 4 days and
IgG antibodies by 7 to 14 days after the first dose of
vaccine.299
3. Vaccine doses given during the first 14 days prime
the immune system, but for preexposure vaccination,
at least one booster dose at 21 days or later seems
necessary for high and persistent titers.
4. Three doses of CCV delivered over 21 to 28 days induce
antibodies in 100% of healthy persons and can be used
as a preexposure regimen.
660 SECTION TWO
5. With intramuscular vaccination, the primary regimen
for postexposure use is four doses given during the first
2 weeks in healthy persons.
6. Intradermal regimens using multiple sites (two to four
or more) administered on four to five occasions between
days 0 and 30 also induce adequate antibody responses
and use much lower volumes of vaccine, which is an
economic advantage.
7. Antibody titers after vaccination with CCV are
usually greater than 10 IU/mL, significantly higher
than obtained in persons given nerve tissue vaccine.
Antibody induction exceeding 0.5 IU/mL is always
achieved, and this level correlates with protection.
8. It is not necessary to check rabies VNA titers after
routine preexposure or postexposure immunization
with CCV, unless the vaccinee is immunosuppressed
(see later), has received chloroquine with intradermal
vaccination, or has undergone a significant deviation
from standard recommendations.300 HIV-infected
patients may respond poorly to rabies vaccines301 and
should be monitored serologically.
9. Age is a factor in response; persons older than 50 years
may respond less well than younger persons, but all
seroconvert.302,303
10. The place of cell-mediated immunity in routine
prophylaxis remains uncertain. Clearly, although
cellular responses of the T-helper and the HLA-
restricted cytotoxic T-cell types clearly are produced,
direct proof of a useful function of the latter is lacking,
whereas VNAs are demonstrably effective.45,304–306
11. Certain host response genes are important, as vaccinees
of HLA groups B7 and DR2 show early and higher
antibody responses, whereas those of HLA group DR3
respond later and with lower levels.283
12. There is molecular mimicry of the nicotine receptor–
binding motif between rabies G protein and HIV
glycoprotein 120, which may induce antibodies to
HIV in rabies vaccinees.307,308 False-positive tests for
antibodies to HIV have been reported after rabies
vaccination,309,310 but the phenomenon seems to be
uncommon.311
13. Seroconversion with VNA titers greater than 0.5 IU/
mL was achieved in 33% of people receiving two doses
of vaccine but in 100% of recipients of three or more
doses.312
14. A preliminary study showed that vaccination elicited a
range of cytokine and chemokine responses but failed
to distinguish between intramuscular and intradermal
vaccination.313
The typical evolution of VNAs after vaccination is shown in
Tables 29-5 and 29-9.314
Effectiveness
The efficacy of Semple-type nerve tissue vaccine has been esti-
mated to be about 84% in India,315 although protection may be
lower after severe exposure. For CCVs, Nicholson316 estimates a
failure rate of 1 in 80,000 in developed countries and between 1
in 12,000 and 1 in 30,000 in developing countries.
During the development of HDCV, efficacy studies were
carried out in Iran, Germany, and the United States. In Iran,
45 people exposed in eight different incidents to eight proven
rabid wolves or dogs were given six doses of HDCV after expo-
sure. Antirabies serum also was given with the first vaccine
doses. The presence of rabies virus in the brain was confirmed
for all eight animals. In four animals tested, high virus titers
also were found in the salivary glands. All vaccinees developed
antibodies, and none developed rabies.176 In Germany, 63 per-
sons bitten by rabid animals were uniformly protected. The
accumulated experience of the US CDC was summarized in
1980, at which time 90 people exposed to rabid animals had all
survived after HDCV vaccination.317 Similar inferences can be
made for PCECV, especially in the United States. Between the
recall of HDCV in 2004 and its reintroduction during 2006,
PCECV was the only human rabies vaccine available for routine
PEP. These epidemiologic data are supported by multiple stud-
ies that demonstrate the effectiveness of PCECV in preexposure
prophylaxis or PEP scenarios in developing countries.
The HDCV produced by Berna was subjected to trial in 100
Thai patients exposed to proven rabid animals. All patients
were protected by the standard five-dose regimen, although the
GMT of rabies VNAs at 90 days after the first vaccination was
relatively low, at 2.57 IU.318
Understandably, there have been no placebo-controlled stud-
ies of the efficacy of rabies vaccines. The CCVs, including PVRV,
PCECV, PDEV, and PHKCV, have been accepted based on the
induction of VNAs and lack of failures after postexposure vacci-
nation.24,139,149–152,155,157,159,161,164,165,167,168,183,191–194,197,199,201,206,207,283,284
A recent analysis of rabies postexposure vaccination in the
United States showed an average of more than 23,000 courses
given per year, and although the decision to vaccinate was
often made without consulting public health subject matter
experts in rabies, no failures were seen.319 A review of 20 years
of use of PVRV revealed only one failure after a standard regi-
men, occurring in a child with multiple bites on the face.320
Postmarketing surveillance of an Indian-made HDCV also
showed protection.321
The efficacy of rabies vaccination has been well supported
using a variety of animal species as human surrogates. For
example, protection studies in mice showed that VNAs induced
by the G protein contained in HDCV neutralized 17 different
street rabies viruses.307
Persistence of immunity and booster doses
Rabies VNAs do not stay at elevated levels for long periods
after vaccination with the usual preexposure schedules. By
1 year, VNAs may fall to geometric mean levels between 1 and
3.5 IU,172,322,323 and by 2 years, they may fall below the minimum
acceptable level of 0.5 IU in approximately 15% to 20% of sub-
jects.324 However, B-cell memory is prolonged,325 and Thraenhart
and colleagues326 reported the presence of VNAs in the serum of
18 people vaccinated 2 to 14 years earlier. Antibodies reacting
with many virus proteins were still present, in addition to lym-
phocyte proliferation responses to the same proteins. A survey of
travelers to Nepal vaccinated against rabies during the preceding
5 years showed antibodies in 37 of 38.327
Booster doses of vaccine are efficient in restoring VNAs, with
100% of subjects showing a fivefold rise by day 7.322 Two booster
doses enhance the speed of the booster response.328 Therefore,
previously immunized persons who are again exposed to rabies
should receive two booster doses 3 days apart, without RIG.
Boosters can be given intramuscularly or intradermally, although
no preparation approved for intradermal use is available in the
United States. Studies195,329–331 have compared four-site intra-
dermal inoculations with the standard two-dose intramuscu-
lar inoculations in previously vaccinated volunteers undergoing
simulated reexposure to rabies and found that the intradermal
group had better booster responses. Vietnamese children vacci-
nated 5 years earlier were easily boosted.332 When maintenance
of VNA is desired after preexposure vaccination, as in laboratory
workers, a booster may be given at 1 year after primary vaccina-
tion, which invariably produces an anamnestic response (see
further discussion in “Indications for vaccination”). However,

because of allergic reactions to the HDCV used in the United
States (see “Reactions to cell culture vaccines”), routine boost-
ers are not recommended in the absence of definite exposure.
Laboratory workers who have continuous exposure to rabies
virus should have antibody levels checked every 6 months and
should receive a single booster immunization if the titer falls
below complete virus neutralization at a serum dilution of 1:5
(or 0.5 IU/mL).
Vaccination of immunosuppressed persons
Briggs and Schwenke333 compared the persistence of antibodies
in civilians and in Peace Corps volunteers who receive chloro-
quine for prophylaxis of malaria. At 1.5 to 2 years after primary
vaccination, adequate titers were found in 99% of civilians and
88% of Peace Corps volunteers who received the vaccine intra-
muscularly and in 93% of civilians and 64% of Peace Corps vol-
unteers who received the vaccine intradermally, suggesting that
chloroquine was immunosuppressive.
Numerous studies on HIV-infected patients have shown that,
although rabies vaccination is safe and generally effective,334,335
persons with CD4+ cell counts less than 300/mm3 or 15% of lym-
phocytes are less likely to respond with VNAs.334,336–339 Patients
receiving immunosuppressive drugs and persons with diabetes
also should be monitored for proper VNA responses.340,341 In
contrast, pregnant women respond normally to rabies vaccine,
and there is no suggestion of fetal risk.342,343
Malnourished children respond to PCECV rabies vaccine
with satisfactory antibody titers.344
A study of intradermal rabies vaccination in patients under-
going hemodialysis showed adequate immune responses,345 as
did intramuscular vaccination in pediatric solid organ trans-
plant recipients.346
Replication-defective vaccinia viruses carrying rabies and
genes have been shown to protect B cell–deficient mice, and
such future development might be of value in B cell–deficient
persons exposed to rabies.347
Protection against lyssaviruses other
than genotype 1
The ability of vaccines formulated with classical rabies virus
to protect against other Rhabdoviruses in the Lyssavirus genus
remains questionable. However, it seems that antibodies are
cross-neutralizing in vitro and that cross-protection is good
in animals challenged with lyssaviruses 5 through 7 by par-
enteral routes, although not after intracerebral inoculation.348
This is particularly comforting for prevention of rabies caused
by European bat lyssaviruses, although titers against EBLY
1 (lyssavirus 5) may be more variable.349 However, the results
for more distant Asian and African lyssaviruses (types 2-4 and
other new viral species) depend on relative genetic distance, with
lower protection for the most distantly related lyssaviruses.263
Correlates of immunoprotection
Preexposure vaccination with potent rabies vaccines leads to the
development of antibodies. Vaccination also induces produc-
tion of cytotoxic T cells, which have been shown to protect vac-
cinated mice in the absence of neutralizing antibodies. A high
level of cell-mediated cytotoxic activity can be maintained by
repeated inoculations of vaccine, and the presence of antibodies
does not interfere with the secondary stimulation of sensitized
lymphocytes.
Rabies vaccines 29 661
Although the exact mechanism of protection of humans
through postexposure vaccination is unknown, neutralizing
antibodies have a major role.350 The fact that only monoclo-
nal antibodies that interact with macrophages are effective in
protection of mice against disease may indicate that a complex
mechanism is involved in antibody protection after challenge.
Concentrated and inactivated rabies vaccine of tissue cul-
ture origin is able to induce high levels of circulating interferon
a few hours after its administration and can protect animals
from rabies infection if it is given shortly before or after chal-
lenge with virus. This interferon-induced protection, however,
is not specific because similar protection can be obtained with
concentrated vaccines produced from unrelated viruses, such
as influenza and Kern Canyon. However, only the rabies vac-
cine can protect when it is given several days before challenge
with rabies virus. The combined treatment with interferon or
interferon inducers in addition to rabies vaccine is more effica-
cious than vaccine alone in experimental animals, when treat-
ment is initiated several hours after challenge.351 Considering
the critical usefulness of antibodies, the role of interferon and
other cytokines in human prophylaxis is likely important, but
undefined.
In reality, when humans are bitten by a rabid animal, the
viral dose, route, and severity cannot be controlled in nature. As
such, there is no distinct “seroprotective” antibody level defined
for humans or other animals as an absolute predictor of out-
come. The international standard of 0.5 IU/mL is merely one
arbitrary level chosen as a representative value that defines a
given amount of VNA that can be measured in a serum neu-
tralization test by reference laboratories. However, no vaccine
is 100% efficacious, and no set level of VNA is acceptable as
a facsimile of direct outcome. In practice, after viral exposure,
previously vaccinated persons should receive immediate wound
care and booster doses of vaccine to induce a rapid anamnestic
response—and not rely on routine laboratory antibody testing,
or a predetermined titer, to decide a course of immunization
by medical providers. Similarly, in experimental testing of vac-
cine efficacy, animals having no detectable VNA may be pro-
tected against viral challenge, and animals with a titer even
more than 0.5 IU/mL can die. Clearly, the immune response
during rabies prophylaxis after viral infection is complex, and
VNAs are important, but understanding the basis for this rep-
ertoire should lead to knowledgeable intervention and the more
strategic development of improved biologics, rather than depen-
dence on measureable phenomenology or reliance on an anti-
body value alone.352
Treatment failures
Most human cases of rabies occur because no prophylaxis was
given. Clearly, modern rabies postexposure vaccination is highly
effective, but failures can occur. As rabies in the biting animal is
frequently untested and thus unknown, it is impossible to give a
figure for the rate of failure, but it is probably less than 1%. Most
failures involve inadequate PEP, such as delays between exposure
and initiation of care, and often involve severe exposures, such
as multiple bites and bites on the head and neck.
Thraenhart and colleagues283 reviewed 28 cases of rabies that
developed despite PEP with modern vaccines. In 90% of cases,
RIG had not been administered or had been administered incor-
rectly. Other errors included passive immunization more than
24 hours before vaccine, incorrect local wound cleansing, injec-
tion of vaccine into the buttocks instead of the deltoid, and late
initiation of immunization. Only two patients, both of whom
had severe facial injuries, were considered to have true prophy-
laxis failures. Wilde353 described 15 cases of vaccine failure in
detail, and others have been described by Shantavasinkul et al354
and Hemachudha et al.355
662 SECTION TWO Licensed vaccines
An “early death” phenomenon has been described in animals
exposed to rabies and then vaccinated. In other words, some
animals that are vaccinated but nevertheless develop rabies die
sooner than challenged unvaccinated animals. It is unclear as to
whether this phenomenon is mediated by antibodies, cytotoxic
T cells, or a combination of the two. Willoughby356 recently
reviewed this subject and concluded that early death has not
been convincingly demonstrated in humans, although there
was a hint of it in patients bitten by rabid bats.
A follow-up study of effectiveness after 15 million doses of
PCECV reported 47 suspected failures. All occurred in India
and Thailand, and in no case were WHO guidelines followed
completely.357
Other failures also may relate to deviations from standard
prophylaxis.225,358,359 However, a bite on the face with a large
inoculum of rabies virus may overwhelm PEP, especially if
delays or significant deviations from protocol occur.
Adverse reactions
Reactions to rabies vaccines containing animal
brain tissues
There are several types of reactions to rabies vaccines contain-
ing animal brain tissue. These reactions include general sys-
temic, local, and severe and fatal reactions.174
The various minor disorders that may develop during or after a
course of antirabies vaccination include fever, headache, insom-
nia, palpitations, and diarrhea. Sensitization to proteins contained
in older nerve tissue vaccines can cause a sudden shock-like col-
lapse, usually toward the end of the course of treatment.
Erythematous patches may develop approximately 7 to 10 days
after the beginning of antirabies vaccination. Local erythema
and swelling may also appear a few hours after vaccine injection
and fade in 6 to 8 hours.
A patient may have serious and often fatal illness after nerve
tissue vaccine. These accidents are of two types: (1) rage de
laboratoire, a disease induced by the noninactivated fixed virus
present in the vaccine, and (2) neuroparalytic accidents, which
present the greatest danger from rabies vaccination. All types of
vaccine containing adult mammalian nervous tissues exhibit
similar capacities for inducing neuroparalytic reactions. The
neuroparalytic accident usually develops between the 13th and
15th days of vaccination and may assume one of the following
forms:
1. Landry type. In this type of accident, the patient rapidly
becomes pyrexial and has pain in the back. Flaccid
paralysis of the legs begins, and within 1 day the arms
become paralyzed. Later the paralysis spreads to the face,
tongue, and other muscles. The fatality rate is about
30%; in the remaining 70%, recovery usually occurs
rapidly. The incidence varies between 1 in 7,000 and 1
in 42,000 recipients.360
2. Dorsolumbar type. Less severe than the Landry type,
this is the most common form of neuroparalytic
accident. Clinical features are explicable by the presence
of dorsolumbar myelitis. The patient may be febrile and
feel weak, with paralysis of the lower limbs, diminished
sensation, and sphincter disturbances. Typically, the
fatality rate does not exceed 5%.
3. Neuritic type. In this type of accident, the patient may
be pyrexial and usually shows a temporary paralysis of
the facial, oculomotor, glossopharyngeal, or vagus nerve.
4. Optic neuritis has also been reported.361
Neuroparalytic accidents are caused by allergic “encephalomy-
elitis”, attributable specifically to sensitization to adult nerve
tissue antigen (myelin basic protein).362 High-titered antibodies
to human myelin proteins and to gangliosides often are detected
in patients who are affected.363 The incidence of these reactions
to nerve tissue vaccine varies widely from 0.017% (1:6,000) to
0.44% (1:230) and is definitely lower in people receiving DEV
(1:32,000) and in people receiving properly manufactured vac-
cine of newborn rodent brain (1:8,000).
One study observed 1,392 Tunisian adults who were given
a Semple-type vaccine prepared by phenol inactivation of
rabies virus–infected lamb brains.364 In seven patients neuro-
logic complications developed, including paralysis or paresis in
five. Most of the patients had elevated cell counts in their CSF.
A rate of one neurologic complication in 200 vaccinations is
unacceptable and illustrates the desirability of replacing nerve
tissue vaccines with CCVs. Progress is ongoing, as demon-
strated by the progress in India, which switched to CCVs.
Reactions to cell culture vaccines
CCVs are widely accepted as well-tolerated rabies vaccines,
although reported reaction rates to primary immunization have
varied with the monitoring system. In large-scale testing of the
safety and immunogenicity of HDCV performed on American
veterinary students, adverse reaction rates observed in more
than 1,770 volunteers were as follows: significantly sore arm,
15% to 25%; headache, 5% to 8%; malaise, nausea, or both, 2%
to 5%; and allergic edema, 0.1%.172 In another study of postex-
posure vaccination, 21% had local reactions, 3.6% had fever, 7%
had headache, and 5% had nausea.365 The most common local
reactions are erythema, pain, and induration. These results
have been confirmed by more recent studies.185 When HDCV is
administered to children, in whom psychological overlay is pre-
sumably less than in adults, there are fewer complaints.
An observational study of 290 health care workers given
PCECV also showed 53% of systemic reactions, including
fatigue, malaise, headache, chills, and fever, and 5% of sub-
jects discontinued vaccination. However, there was no control
group.366 Data concerning the same vaccine were collected by
the VAERS [Vaccine Adverse Event Reporting System] passive
reporting system. Headache was the most prominent symptom
reported. Neurologic events did not seem to be related to the
vaccine, and other reported reactions were usually nonserious.367
A study comparing PDEV with PVRV and PCECV showed no
differences in reactions.368 PCECV was safely coadministered to
Thai children along with Japanese encephalitis vaccine.369
After licensure of HDCV in the United States and widespread
use, allergic reactions began to be reported, principally after
booster doses.370,371 The overall incidence of reactions was 11
in 10,000 (0.11%) vaccinees, but, after boosters, the incidence
rose to 6%.372 Anaphylactic type 1 (IgE) reactions occurred in
about 10% of the reported cases, all during the primary series (1
in 10,000 vaccinations), but the majority seemed to be type 3
hypersensitivity (IgG-IgM) reactions occurring 2 to 21 days after
booster doses (Table 29-10). These reactions have been attrib-
uted to antigenicity conferred on human albumin used as sta-
bilizer in the vaccine by the -propiolactone used to inactivate
the virus, which increases the capacity of the albumin to form
immune complexes.373–375

Signs and Symptoms in Three Cohorts Containing 255
Subjects Reporting Presumed Immune Complex-Type Hypersensitivity
Reactions* After Booster Immunization with Human Diploid Cell Rabies
Vaccine Given ID or IM
*Coombs and Gell type III.
†
Data are given as number (percentage).
ID, intradermally; IM, intramuscularly.
From Centers for Disease Control. Systemic allergic reactions following
immunization with human diploid cell rabies vaccine. MMWR Morb Mortal
Wkly Rep 33:185-188, 1984.
Fortunately, respiratory symptoms are mild, and there have
been no deaths. Antihistamines, epinephrine, and, occasion-
ally, steroids have been used in successful treatment of the reac-
tions, which resolved in 2 to 3 days.
Newer vaccines are produced by use of additional purification
steps to remove human albumin. Systemic reactions to booster
doses are uncommon with these vaccines.376,377 Twenty cases of
possible anaphylaxis after PCECV were reported to VAERS.367
Although five cases of CNS disease, including transient neuro-
paralytic illness of the Guillain-Barré type, have been reported
among the millions of persons given HDCV,378–382 this rate is
too low to be causally related to vaccination because the back-
ground incidence of such diseases is about 1 in 100,000 per
year. The low incidence after HDCV compares with a neuro-
logic complication rate of 1 in 1,600 people for nerve tissue
vaccine, 1 in 8,000 for suckling mouse brain vaccine, and 1 in
32,000 for DEV. A review of Guillain-Barré syndrome after vac-
cination described several cases after rabies vaccination, but at
rates consistent with coincidence.383
In Thailand, a switch from Semple-type vaccine to HDCV
resulted in a drop in the rate of neurologic complications from
1 in 155 to less than 1 in 50,000 treatments. At the same time,
the failure rate dropped from 1 in 2,000 to 1 in 25,000 treat-
ments without the use of RIG.384 If reactions occur after one
CCV, a switch can be made to another produced in a different
cell substrate without danger.
A case of bilateral neuroretinitis has been reported after
PCECV.385
Indications for vaccination
Preexposure vaccination
All high-risk professionals, such as veterinarians, hunters, trap-
pers, dog catchers, mail carriers, cavers, and laboratory workers
contemplating working with rabies virus, should be prophylacti-
cally immunized against rabies. The recommended regimens are
given in Table 29-7. After receiving the three-dose preexposure
regimen, persons who are continuously exposed to high concen-
trations of rabies virus in the laboratory should have VNA lev-
els checked every 6 months and should be given a booster dose
intramuscularly or intradermally if the titer is less than ade-
quate (complete virus neutralization at a serum dilution of 1:5,
Rabies vaccines 29 663
or 0.5 IU). Veterinarians and others exposed regularly to rabid
animals or rabies virus should be similarly checked every 2 years.
Peace Corps workers, missionaries, or others who remain
for long periods in rabies-enzootic countries certainly deserve
consideration for preexposure vaccination.386 A survey of trav-
elers revealed that, after an average of 17 days in Thailand,
1.3% and 8.9% had been bitten or licked by a dog, respectively,
and 0.5% had required rabies vaccination.387 Moreover, post-
exposure vaccination in developing countries is often compli-
cated by problems of availability of potent vaccine and RIG.388
However, a decision analysis concluded that routine preexpo-
sure vaccination would cost $275,000 per case averted and
that it should be individualized according to the traveler's
situation.389 Our recommendation would be to vaccinate the
travelers who will be staying in remote areas of hyperendemic
dog rabies for more than a few days, particularly children.
Travelers should be warned about the dangers of rabies and
other zoonotic diseases and educated about animal bite pre-
vention strategies.390
In countries where rabies is endemic and children are
frequently exposed to rabid animals (such as in Amazonia
because of nightly feeding from vampire bats), one might
contemplate prophylactic vaccination against rabies as part
of future pediatric immunization. Dog bites are a consider-
able problem in many areas of the world and accounted for
more than 5% of visits to the emergency department of a
Bangkok hospital, of which 55% were by children.391 So far,
prophylaxis has been restricted to the children of Westerners
going to live in areas enzootic for rabies,392 but clinical tri-
als in children have been performed in developing coun-
tries.393 A preliminary study was done with the PVRV vaccine
in Thailand, where two doses were given intramuscularly in
association with routine pediatric immunizations at 2 and
4 months of age.394 Seroconversion to rabies occurred in 100%
of infants, without significant interference with the other
vaccines. A similar study was done in Vietnam with three
doses of intradermal rabies vaccine at 2, 3, and 4 months of
age, with an adequate immune response (GMT, 12 IU/mL).395
An economic analysis concluded that preexposure vaccina-
tion becomes cost-effective when the dog-bite incidence is
between 2% and 30%.396
Preexposure booster vaccination
After preexposure vaccination, antibodies may decline sharply
within the first year of vaccination. If a booster dose is given 1 year
later, subjects segregate into two groups: “good” responders,
who develop a titer greater than 30 IU by 14 days after booster,
and “poor” responders, whose titers are lower. The former,
who represent 75% of subjects, may not need further booster
vaccination for 10 years or more, whereas the latter may
need more frequent boosters.397 This strategy may reduce
costs.398,399
Primary vaccination by the intradermal route may provide
less sustained immunity, but the intradermal route is an effec-
tive means of administering a routine booster.180
Postexposure prophylaxis
Guidelines and schedules for PEP are given in Tables 29-11
through 29-13.
The essential triad of postexposure rabies prophylaxis is
local wound care, vaccination, and antiserum administration.
Local treatment of bites and scratches consists of vigorous
washing with soap and water, ideally for at least 15 minutes.
Surgical suturing should be avoided for 7 days, if possible, but in
any case, RIG (see later) should always be administered before
suturing.400 Antibiotics and tetanus toxoid may be indicated to
prevent other infections.401
664 SECTION TWO
Rabies Postexposure Prophylaxis Guide for the United States
*During the 10-day observation period, begin postexposure prophylaxis at the first sign of rabies in a dog, cat, or ferret that has bitten someone. If the animal exhibits
clinical signs of rabies, it should be humanely killed immediately and tested.
†
The animal should be humanely killed and tested as soon as possible. Holding for observation is not recommended. Discontinue vaccine if immunofluorescence test
results of the animal are negative.
From Centers for Disease Control and Prevention. Human rabies prevention, United States, 2010: recommendations of the Advisory committee on Immunization
Practices (ACIP). MMWR,Mar.19, 2010 Vol.59/No. RR-2
Comparative Immunogenicity of Different Schedules*
*Data are given as geometric mean titer.
ID, intradermally; IM, intramuscularly.
Data from Warrell MJ, Riddell A, Yu LM, et al. A simplified 4-site economical
intradermal post-exposure rabies vaccine regimen: a randomised controlled
comparison with standard methods. PLoS Negl Trop Dis 2:e224, 2008.
doi:10.1371/journal.pntd.0000224.
A decision to give postexposure rabies vaccine should be
based on consideration of the following issues402,403:
1. Was the patient's skin broken by a bite or scratch, or
were mucous membranes contaminated? If not, no true
exposure has occurred. Seeing or touching a rabid animal
is not an exposure. The risk of rabies after a bite by a
rabid animal has been estimated at 5% to 80%, whereas
the risk after scratches is much less (0.1%-1%). The risk
after mucous membrane contact is extremely low.394
2. If the bite was by a dog or cat, is domestic animal rabies
found in the particular geographic area? Many areas of
the world, such as Oceania, Antarctica, and the United
Kingdom, are free of rabies in carnivores. In many cities
of the United States, even stray animals are unlikely to
be rabid.
3. Was the dog or cat vaccinated against rabies?
Vaccination diminishes the risk, but not completely.
Proper vaccination of pets should not be accepted at face
value. Documentation should be required to show at
least two vaccinations with a potent vaccine; one dose
of inactivated vaccine is insufficient to always warrant
efficacy, especially if administered in within the last
30 days.400,404
4. Is the biting animal a domestic dog or cat and available
for observation? If yes, vaccination may be postponed.
However, in rabies-enzootic areas, vaccination is
advisable, even if the domestic animal appears normal,
because of short incubation periods in humans and the
frequent difficulty in being certain of the identity of the
biting animal.405
5. If the bite was inflicted by a wild animal, was it a
species likely (eg, skunk or raccoon) or unlikely (eg,
squirrel or rat) to be rabid? Local conditions must be
taken into account; for example, in the United States,
rabies has been reported occasionally in larger bodied
rodents, such as woodchucks, a very unlikely finding
elsewhere. Rapid diagnostics, performed within 24 to
48 hours of exposure, will assist in PEP decisions.
6. Was the bite provoked or unprovoked? This criterion is
useful only in areas of low incidence and not where the
incidence of rabies in dogs is elevated.400 In Thailand,
even when the animal's behavior appeared normal, an
assessment of provocation did not correlate with the
presence of rabies at necropsy of the animal.406 Attempts to
play with wild animals should be considered provocative.
7. Was there exposure to a rabid human? Under these
circumstances, use the same criteria for vaccination
as with a biting animal. Only persons who were bitten
or scratched, who had mucosal exposure, such as with
mouth-to-mouth resuscitation, or who were otherwise
exposed to infectious saliva or nerve tissues need to be
vaccinated.404
8. Should RIG be administered? In naïve, previously
unvaccinated persons, RIG should always be given
in combination with vaccine for transdermal or
mucous membrane exposure (see “Serum and vaccine
prophylaxis”). The dose of HRIG is 20 IU/kg and that of
ERIG, 40 IU/kg. As much as possible of the RIG should
be inoculated locally, diluted if necessary in saline to
provide sufficient volume. If necessary, local anesthesia
can be provided with procaine-type compounds.407 The
remainder of the RIG, if any, should be given in the
deltoid or other muscles.
 Rabies vaccines 29 665

World Health Organization 2004 Recommendations for Rabies for Postexposure Prophylaxis

*Exposure to rodents, rabbits, and hares seldom, if ever, requires specific antirabies postexposure prophylaxis.
†
If an apparently healthy dog or cat in or from a low-risk area is placed under observation, the situation may warrant delaying initiation of treatment.
‡
This observation period applies only to dogs and cats. Except in the case of threatened or endangered species, other domestic and wild animals suspected as rabid
should be humanely killed and their tissues examined for the presence of rabies antigen using appropriate laboratory techniques.
§
Postexposure prophylaxis should be considered when contact between a human and a bat has occurred unless the exposed person can rule out a bite or scratch or
exposure to a mucous membrane.
From WHO Expert Consultation on Rabies. WHO Technical Report Series, 931, October 2005.

9. Is patient age important? The dose is the same
regardless of age; children tolerate vaccination well
and demonstrate excellent antibody response, as
demonstrated in different ethnic groups.296,408–410
10. What about bats? Rabies PEP is recommended for
all people with bite, scratch, or mucous membrane
exposure to a bat, unless the bat is available for
diagnostic testing and is negative for evidence of
rabies virus infection. If a report on the presence of
rabies by laboratory diagnosis is not available rapidly,
such as within about 3 days, PEP should be started.
The inability of care providers to elicit information
surrounding potential exposures may be influenced
by the sometimes tiny injury inflicted by a bat bite (in
comparison with lesions inflicted by larger carnivores)
or by circumstances that hinder accurate recall of
events. Therefore, PEP is also appropriate even in the
absence of a demonstrable bite or scratch when there is
reasonable probability that such contact occurred (eg,
a sleeping person awakens to find a bat in the room,
an adult witnesses a bat in the room with a previously
unattended child, a mentally deficient person, or an
intoxicated person). Fortunately, tests for VNA obtained
from humans after vaccination with HDCV or PCECV
show good neutralization of rabies virus associated
with bats.411
Unnecessary PEP can be avoided if circumstances are found
that make the chances of rabies virus exposure remote. When
exposures occur to animals in which rabies is not suspected,
large numbers of people often receive vaccine.412 A survey of
practice in 11 university-associated urban emergency depart-
ments revealed that rabies prophylaxis was administered inap-
propriately in 40% of cases.413 Exposure to bats in the absence
of known bites is a complicated situation, and consulta-
tion with experienced public health providers is warranted.
Tables 29-11 through 29-13 should be consulted as a guide
for the selection of PEP in the United States and elsewhere in
the world.404,414 Table 29-7 describes preexposure and postex-
posure regimens.
Postexposure booster vaccination
If a person has been vaccinated with a CCV and later exposed
again to rabies, two booster injections of a vaccine are recom-
mended because one may not always be sufficient.415 Persons
who received vaccine intradermally for their primary series are
particularly at risk for a slow response to booster.416 Another
suggestion from Thailand is to give four simultaneous intrader-
mal vaccinations as a single booster.417 However, subjects who
previously received CCV uniformly have long-lasting immuno-
logic memory,418 and there is no evidence that more than two
boosters are necessary. Persons who gave a history of vaccina-
tion with nerve tissue vaccines responded poorly to boosters
in 18% of cases419 and should, therefore, receive a full primary
regimen unless antibodies were previously shown to be present.
Contraindications to vaccination
Because rabies is a lethal disease, any contraindication to post-
exposure treatment should be considered carefully before dis-
qualifying a person for PEP.
Persons with histories of severe allergies are more prone to
having allergic reactions to rabies vaccine. When persons with
such allergies are vaccinated, prophylactic antihistamines
should be given and epinephrine should be available. If an aller-
gic reaction occurs, one may give an alternative vaccine of dif-
ferent tissue origin, for example, PVRV or PCECV in the case of
a reaction to HDCV. A similar strategy was applied in allergic
persons in whom brain tissue vaccine caused symptoms of CNS
involvement during the course of injections. Administration of
nerve tissue vaccine was interrupted immediately, and the series
was completed with vaccine produced in tissue other than brain.
However, only very severe reactions not controlled with pre-
medication may be grounds for interruption of rabies vacci-
nation. Treatment with steroids may control allergy but may
also inhibit VNA responses. Accordingly, VNA titers should
be determined after the last dose if steroids have been used.
Similarly, patients receiving immunosuppressive medications
666 SECTION TWO Licensed vaccines
for other diseases should have VNA levels checked after immu-
nization to verify an adequate response to the vaccine.414 If
titers are inadequate, a booster dose should be administered.
Pregnancy is not a contraindication to rabies vaccina-
tion.420,421 Follow-up of 202 Thai women vaccinated during preg-
nancy revealed no excess of medical complications or abnormal
births.422 Vaccination of the newborn is probably unnecessary
but has been carried out successfully.423
Public health considerations
Rabies vaccination of animals
From the beginning of his involvement with rabies, Pasteur rec-
ognized that indirect protection of humans could be achieved
effectively through the vaccination of dogs. Although dogs were
used to obtain most of the experimental data on protection
from 1884 to 1885, it was not until the early 1920s that a prac-
tical and successful canine vaccine was developed.
The first use of mass vaccination of dogs was a modified
Semple type prepared by Umeno and Doi424 in Tokyo in 1921.
It proved effective in controlling rabies in dogs in Japan and in
other countries that produced and used this type of vaccine.
The quality of vaccine improved greatly with the introduction
by Habel365 of a standard mouse potency test for the Semple-
type vaccine, which ensured the potency of the vaccines in mass
vaccination programs.
In 1945, Johnson425 demonstrated that a single dose of
a potent, phenol-inactivated vaccine protected dogs against
a challenge with street rabies virus for a period of more than
1 year. From 1945 on, this was virtually the only type of vac-
cine used for control of rabies in dogs, cats, and other domestic
animals.
A modified live virus vaccine was introduced by Koprowski426
in 1948. Successive passages of a strain of virus of human ori-
gin, first in 1-day-old chicks and later in embryonated hens'
eggs, resulted in the loss of pathogenicity for dogs, providing an
attenuated virus strain safe for dogs, called Flury LEP. Further
passages of Flury LEP virus in embryonated hens' eggs resulted
in a vaccine that was no longer virulent for adult laboratory ani-
mals yet was lethal for newborn mice, called Flury HEP. Both
Flury LEP and Flury HEP were given to many types of domestic
animals in different parts of the world. However, the live atten-
uated Flury virus vaccine was discontinued in the late 1970s in
the United States.
Another attenuated strain of rabies virus, ERA, was intro-
duced by Canadian workers in 1964.427 The ERA vaccine
was shown to provide excellent immunity, lasting for at least
3 years. However, several vaccine-induced cases of rabies in cats
and other animals resulted in the cessation of its routine use.
Inactivated rabies vaccines for animals prepared from brains
of newborn mice or from virus of cell culture origin were intro-
duced in the 1980s and are now in general use in Europe and
the Americas. However, only cell culture inactivated virus vac-
cines and live recombinant vaccines are licensed for domestic
animals in the United States.
Elimination of human rabies by control of canine rabies
through vaccination, sterilization, and education has been a
successful strategy, as demonstrated in South America and in
parts of Thailand.428,429 In areas where canine rabies has been
eliminated and the dog can be observed for 10 days by quali-
fied persons, immediate vaccination may not be necessary, to be
started only if the animal develops signs of rabies.
Oral vaccination of wildlife to prevent the spread of rabies
in terrestrial animals such as foxes, raccoons, and coyotes has
become possible. The SAD B19 attenuated strain has been placed
into baits for the vaccination of foxes.430 The virus is grown on
a baby hamster kidney cell line and is stable even at high envi-
ronmental temperatures. A dose of approximately 106 infectious
units immunizes 100% of foxes, which has allowed its wide
application in Germany and elsewhere in Europe.431,432 However,
because the SAD B19 strain retains residual pathogenicity for
rodents, a more attenuated strain called SAG2 was developed.433
Genetic engineering has been applied to rabies immuniza-
tion by the construction of a vaccinia virus recombinant con-
taining the gene for the G protein (V-RG).434 The recombinant
is placed in baits, and, on ingestion by animals, multiplies only
in the tonsils and the buccal area. Extensive tests conducted
in many species have confirmed the safety and efficacy of vac-
cination with this construct, and field tests in Europe, Canada,
and the United States have confirmed the promising labora-
tory results.435–439 Two human infections with the vaccinia
recombinant produced local lesions but no lasting effect.440
Widespread application of V-RG in Belgium starting in 1989
reduced fox rabies from 841 cases in that year to 2 cases in
1993. Concomitantly, a marked drop occurred in human expo-
sures requiring vaccination.441 The widespread use of these
oral vaccines has eliminated red fox rabies in Western Europe.
In North America, millions of doses of V-RG delivered by air
have blocked the spread of raccoon rabies from the eastern sea-
board, contained grey fox rabies in Texas, eliminated coyote
rabies from the Mexican border, and suppressed red fox rabies
in Southern Ontario.442–444
In the United States, two types of rabies vaccines are avail-
able for domestic animals: inactivated virus vaccines and
canarypox recombinant vaccines. Both types require boosters,
usually after 1 or 3 years.444
New rabies vaccines given orally have been shown to protect
dogs from challenge with rabies virus, equivalent to other rabies
vaccines.445 This may be of particular interest for developing
countries. Recombinant adenoviruses also have shown useful-
ness in the laboratory and in the field.
Rabies vaccination of humans
More than 10 million people are vaccinated each year after pre-
sumed exposure to rabies, particularly in Asia.446 It is difficult
to define precisely the effect of rabies vaccination on the inci-
dence of human rabies because the risk of the disease is variable.
Nevertheless, when untreated persons who were bitten by proven
rabid animals are followed up, disease rates of between 3% and
80% are observed,6,447 depending on the location and severity
of bites.448 In the United States, relatively few of the 30,000 to
40,000 people vaccinated annually are actually at risk for rabies.
However, the paucity of cases of rabies in persons given potent
vaccines together with antiserum argues that countless cases of
rabies are being prevented. In Texas in 1989, 34% of the skunks,
19% of the foxes, and 15% of the bats involved in biting inci-
dents were rabid; risk to humans is obvious.449
Control of dog rabies is jointly responsible with vaccination
of exposed humans for the current rarity of human rabies in
North America and Europe. In developing countries, even nerve
tissue vaccine reduces the incidence of rabies in vaccinees,447
but large numbers of exposed persons turn to indigenous meth-
ods of rabies control and never receive vaccine.450 Thus, great
need exists for inexpensive rabies vaccines and for health educa-
tion to aid in appropriate vaccine use.
The “fourth-generation” rabies vaccines
Several vaccines can be considered fourth-generation vaccines sub-
sequent to the first three generations listed in Table 4.451,452
Vaccinia and canarypox vectors containing the rabies G pro-
tein have been prepared and tested in humans.453–455 With both
 Rotavirus vaccines 30 681

and Bangladesh), three doses of RotaTeq were 51% (95% CI,

13%-73%) effective against severe rotavirus gastroenteritis. The
somewhat moderate efficacy of oral rotavirus vaccines in devel-
oping countries compared with that in industrialized coun-
tries is consistent with the experience with other oral vaccines
against polio, cholera, and shigellosis. Explanations include a
diminished immune response resulting from higher levels of
transplacental maternal antibody, immune and nonimmune
components of breast milk, micronutrient malnutrition (e.g.,
vitamin A and zinc), interfering gut pathogens such as toxin-
producing bacteria or parasites, and comorbidity in the infant
(e.g., HIV infection).

Effectiveness
A case-control RotaTeq vaccine effectiveness study was con-
ducted using active surveillance at a large pediatric hospital in
Houston, Texas, in 2008.253 Cases were children admitted or
seen in the ED with laboratory-confirmed rotavirus acute gas-
troenteritis and children presenting with acute respiratory ill-
ness; children with rotavirus-negative acute gastroenteritis
acted as the two sets of controls. Three doses of RotaTeq were
85% (95% CI, 55%-95%) and 89% (95% CI, 70%-96%) effective
for the two different control groups, respectively, in preventing
severe rotavirus gastroenteritis resulting in hospitalization or
ED care, and 100% (95% CI, 71%-100%) effective in prevent-
ing hospitalization resulting from rotavirus gastroenteritis. A
partial series also provided substantial protection, with one and
two doses being 69% (95% CI, 13%-89%) and 81% (95% CI,
13%-96%) effective, respectively, against severe rotavirus gas-
troenteritis resulting in hospitalization or ED care for the two
control groups combined. A follow-up study at the same hospi-
tal showed that vaccine effectiveness was sustained at a similar
level during the second year of life.254
In another study, active, prospective population-based
surveillance for acute gastroenteritis and acute respiratory
infection in inpatient and ED settings provided subjects for
a case-control evaluation of RotaTeq vaccine effectiveness
in three US counties from January 2006 through June 2009.
Children with laboratory-confirmed rotavirus acute gastro-
enteritis (cases) were matched by date of birth and onset of
illness to two sets of controls: children with rotavirus-
negative acute gastroenteritis and children with acute respi-
ratory infection. The effectiveness outcomes of one, two, and
three doses of RotaTeq against all rotavirus genotypes using the
controls with rotavirus-negative gastroenteritis were 74% (95%
CI, 37%-90%), 88% (95% CI, 66%-96%), and 87% (95% CI, 71%-
94%), respectively, and using the controls with acute respiratory
infection were 73% (95% CI, 43%-88%), 88% (95% CI, 68%-
95%), and 85% (95% CI, 72%-91%), respectively. The overall
effectiveness estimates were comparable during the first and
second years of life and against acute gastroenteritis caused by
different rotavirus strains.255
In a US study using an insurance claims database, three
doses of RotaTeq were 100% (95% CI, 87%-100%) effective in
preventing rotavirus-coded hospitalizations and ED visits, and
59% (95% CI, 47%-68%) effective against all-cause acute gas-
troenteritis requiring hospital or ED care.256 In the outpatient
setting, three doses of RotaTeq were 96% (95% CI, 76%-100%)
and 28% (95% CI, 22%-33%) effective against rotavirus-coded
gastroenteritis and all-cause acute gastroenteritis, respectively.
In Austria, RotaTeq was estimated to be 61% to 98% effective
in preventing rotavirus gastroenteritis hospitalizations.247 In
troenteritis
the
59% phitalizations
(95% CI,
requiring
47%-68%)
andhospital
ED effective
visits,
or and
ED
against
care. all-cause acute gas-
682 SECTION TWO Licensed vaccines
attributed to vaccination by blinded investigators. Sudden infant
death syndrome (SIDS) accounted for 17 of the 52 deaths; cases
were equally distributed between RotaTeq and placebo recipi-
ents ( = 8 and = 9, respectively).
About 11,700 children in the phase III trials were studied
in detail to assess other potential adverse experiences, such as
fever, diarrhea, and vomiting.232 Within 42 days of vaccination,
vaccinees had a small but statistically significant greater rate of
several symptoms than placebo recipients, including vomiting
(15% versus 14%), diarrhea (24% versus 21%), nasopharyngi-
tis (7% versus 6%), otitis media (15% versus 13%), and bron-
chospasm (1.1% versus 0.7%), all respectively. Among RotaTeq
and placebo recipients, the incidences of reported episodes of
fever (43% versus 43%, respectively) and hematochezia (0.5%
versus 0.3%, respectively) were similar. Although Kawasaki's
disease was reported in several children after administration of
RotaTeq in the phase III trial, this association was not found
after licensure.260
In the 7-day period after vaccination, vaccinees had a statis-
tically significant greater rate of diarrhea after dose 1 (10% ver-
sus 9%), after dose 2 (9% versus 6%), and after any dose (18%
versus 15%). Similarly, vaccinees had a statistically significant
greater rate of vomiting after dose 1 (7% versus 5%) and after
any dose (12% and 10%). The incidences of fever and irritabil-
ity during the 7-day period after any vaccine dose were similar
among RotaTeq and placebo recipients.
Bovine–human reassortant rotaviruses have also been gener-
ated that incorporate the genes for VP7 of either Gl, G2, G3,
or G4 human serotypes and 10 genes from the bovine UK
strain rotavirus, which (like strain WC3) is serotype P7[5]
G6.265–267 These reassortants are serotypically similar to the
G1, G2, G3, and G4 reassortants of bovine rotavirus WC3.
Phase I studies were performed on each of the four UK reas-
sortants individually in small numbers of infants (11 to 20
per G serotype) at doses up to 105.8 PFU per strain. The vac-
cine was well tolerated and was shed in feces at rates rang-
ing from 10% to 64% per group. Serum neutralizing antibody
responses to the human parent rotaviruses (representing G
types 1, 2, 3, and 4) ranged from 0% to 30% per serotype,
but the neutralizing antibody responses to the correspond-
ing UK parent virus were higher (30% to 82%). Only 50% of a
group of 14 infants given a G1 reassortant developed evidence
of any immune response (serum neutralizing antibodies or
serum IgA or IgG detected by ELISA) after the first dose of
vaccine. However, after a second dose, all exhibited evidence
of an immune response.266
Subsequently a quadrivalent vaccine containing G1, G2, G3,
and G4 reassortants of UK was administered in three doses
to 20 infants.266 No adverse events occurred after vaccination.
Nineteen (95%) of the infants developed neutralizing antibod-
ies to UK bovine rotavirus; neutralizing antibody response rates
to human rotaviruses of serotypes G1 through G4 ranged from
28% to 37%.267,268
The quadrivalent vaccine was tested for efficacy in a
placebo-controlled trial involving 256 infants given two doses
of 106 PFU (total) in Finland.268 The first and second doses
were given at mean ages of 95 days and 150 days, respectively.
Efficacy in the first year after vaccination was 59% against all
rotavirus gastroenteritis, and 90% against severe gastroenteri-
tis. The vaccine showed no protective efficacy in the second
rotavirus season after immunization. It has not been submit-
ted for licensure in the United States. Despite limited informa-
tion about the multivalent UK reassortant vaccine, the NIH
has now licensed it to seven non-US research groups includ-
ing investigators located primarily in less-developed countries
(including China, Brazil, and India). Certain investigators
believe that development and manufacturing costs of the two
vaccines will be so much lower in less-developed nations than
in developed nations that prices will be sharply reduced, and
the vaccine can be inexpensively distributed to the entire pop-
ulation. However, any such effort will have to meet stringent
WHO guidelines, to ensure safe production practices, and FDA
standards for clinical trials before it can be licensed. For an
impoverished country to meet all these requirements and pro-
duce an inexpensive vaccine would seem to be a significant
challenge.269
Naturally occurring bovine–human reassortant rotaviruses
were isolated in two sites in India—New Delhi (the 116E
strain)270 and Bangalore (the I321 strain).271 Newborns infected
with either strain were found to be protected against severe
disease on reinfection.127,272 The 116E strain is a human G9
isolate with a single VP4 gene segment from a bovine strain
(P[10]G9), and the I321 strain (P[11]G10) has a bovine back-
bone with two segments coding nonstructural proteins NSP1
and NSP3 from human strains. Both of these strains have
been developed for clinical trials. Clinical trials of strain
116E are further advanced than for strain I321; infants inocu-
lated with strain 116E develop a robust rotavirus-specific IgA
response.269,273,274
Because administration of oral rotavirus vaccines in the devel-
oping world may induce a lesser response than in higher-
income countries, and because reduced immunogenicity and
effectiveness in the developing world might be caused by pas-
sively transferred maternal antibodies in breast milk or interfer-
ence by parasites or toxin-producing bacteria in the gut, some
researchers have pursued inactivated whole rotaviruses admin-
istered parenterally. Preliminary studies in experimental ani-
mals have shown promise for this approach.275
Attenuated human rotaviruses
Rotavirus strains isolated from infants in the first month of life
have been studied as potential vaccines because infants in neo-
natal nurseries in several locales worldwide have been found to
be infected with rotavirus at a high prevalence but with little
gastroenteritis, and the infection is followed by resistance to
subsequent episodes of severe disease.80,276 Many of these new-
born strains share a similar P gene.39,276 These strains have
been largely ineffective as vaccine candidates.2,29,277–279 Only the
Australian strain RV3 is currently in development as a potential
vaccine.85,280,281
The attenuated human rotavirus strain P1A[8]G1 (Rotarix) is
now licensed in nearly 100 countries, including the European
Union. The vaccine was licensed and recommended for use in
the United States in October 2008. Rotarix is given by mouth in
two doses at 6 to 14 weeks and at 14 to 24 weeks of age.
Origin
Symptomatic or asymptomatic infection of young children
with a single circulating wild-type strain (serotype P1A[8]G1)
in Cincinnati provided 100% protection against subsequent
rotavirus disease caused primarily by the P1A[8]G1 serotype
over a 2-year period.81 A vaccine of this strain (strain 89-12)
obtained from the stool of one of the study subjects was
developed after 26 passages of the virus in primary AGMK
cells, and seven passages in serially passaged AGMK cells.282

Initial safety and immunogenicity studies with this vaccine
candidate indicated it was relatively safe and immunogenic
in children less than 4 months of age.282 Nearly every child
developed an immune response to the vaccine virus after two
doses.
Limited comparisons of the effects of dosage on the immu-
nogenicity of the 89-12 candidate vaccine have been con-
ducted.282 However, 19 of 20 infants inoculated with two doses
of 105 fluorescent-focus units (FFU) of 89-12 developed rotavi-
rus-specific antibodies and 16 did so after the first dose.
The strain 89-12 vaccine candidate was evaluated in an
efficacy trial in healthy 10 to 16 week old infants at four cen-
ters in the United States. Two doses of 105 FFU or placebo were
administered to 108 or 107 subjects, respectively.283 Low-grade
fever after the first dose was the only side effect more common
for the vaccine than for the placebo recipients (21 [19%] versus
5 [4.6%] subjects, respectively; .001). An immune response
to rotavirus was detected in 94% of vaccinees and in only 4% of
the placebo recipients. Rotavirus-specific serum IgA responses
were detected in 91.6% and anti-89-12 serum neutralizing anti-
bodies in 69.2% of vaccines.
During the first rotavirus season, rotavirus disease was
detected in 18 placebo recipients and two vaccine recipi-
ents (efficacy of 89%). Vaccine episodes were scored for
clinical severity using a 20-point system: in this system,
cases scoring greater than 8 were considered “severe” and
those greater than 14 as “very severe”. Ten placebo recip-
ients but no vaccine recipients presented for medical care
necessitated by rotavirus disease. In the second year, effi-
cacy decreased to 59%, but only one case of severe rotavi-
rus gastroenteritis occurred in vaccine recipients, whereas
10 severe cases occurred in placebo recipients. Vaccine effi-
cacy for 2 years was 76% against any rotavirus gastroenteri-
tis, 84% against severe disease, and 100% against very severe
disease. As G1 rotaviruses predominated during both years,
the efficacy against rotaviruses of other G or P types was not
determinable.
To produce a homogeneous vaccine, strain 89-12 was puri-
fied by a series of three endpoint dilutions in Vero cells and pas-
saged seven more times in these cells after it was sublicensed
by GlaxoSmithKline in 1997. The final product, now at passage
43, was called RIX4414 and is being sold under the trademark
name Rotarix.
Immunogenicity
Investigators inoculated 6 to 12 week old infants orally in
Finland with RIX4414 and found no increase in any elicited
symptoms including fever.284 Thus, it appeared that clonal
selection and increased passage in tissue culture produced a
more attenuated vaccine relative to its parental 89-12 strain. A
dose range effect was seen for the immune response in Finnish
infants, with seroconversion detected in 73% to 96% after the
second dose (Table 30-9).284
Clinical Studies of RotaRix in Infants
ND, not determined.
*Percent of infants who developed an immune response after two doses as determined by rotavirus-specific IgA, by enzyme-linked immunosorbent assay (ELISA).
Rotavirus vaccines 30 683
Evaluation of RIX4414 began in Asia shortly thereafter,
marking the initiation of what would become a global effort
to evaluate this vaccine.285,286 Unlike evaluation of most vac-
cines, evaluation of RIX4414 was performed in developed and
developing nations simultaneously (see Table 30-9). In a dos-
age range trial conducted in Singapore in 2,464 infants, there
was again no increase in fever, diarrhea, vomiting, or irrita-
bility detected, even at the highest dosage of 106.1 FFU, and
seroconversion rates determined by rotavirus-specific serum
IgA responses were 76% to 91% depending on the dose.286
Interference with the other routine childhood vaccines admin-
istered concomitantly was not observed. In a multicenter
study conducted in the United States, the vaccine appeared
to be safe in infants 5 to 15 weeks of age and did not inter-
fere with the immune responses to concomitantly adminis-
tered vaccines.287 Concomitant use of Rotarix with OPV was
reported to suppress the uptake of the rotavirus vaccine. This
effect was eliminated after the second dose.288 Vaccine-specific
immune responses assessed by serum IgA response after two
doses occurred in 81.5% to 88.0% of vaccinees, depending on
the size of the dose.
Efficacy
The first efficacy study of RIX4414 was conducted in Finland
during two rotavirus epidemic seasons, after oral admin-
istration of two relatively low doses of the vaccine (104.7
FFU). In this trial of 405 infants, the vaccine was well toler-
ated, with a seroconversion rate of 80% after the second dose
(see Table 30-9).289 Efficacy against any episode of rotavirus gas-
troenteritis was 73%, and efficacy against severe rotavirus dis-
ease was 90% in the first season and 85% after two seasons.
Because 35 of the 38 cases of rotavirus gastroenteritis were
associated with the same G1 serotype as the vaccine, this trial
did not yield evidence about the potential of this vaccine to
protect against heterotypic rotaviruses. In an initial placebo-
controlled efficacy trial with RIX4414 in Latin America, sero-
conversion after two doses of either 104.7, 105.2, or 105.8 FFU of
vaccine virus, based on serum rotavirus IgA responses, was a
disappointing 61% to 65%, and the dosage administered had
no significant effect on these responses.290 In spite of these low
detectable serum antibody response rates, protection against
severe rotavirus disease was 86% in vaccinees given the high-
est dosage. Equally important was the observation that vaccine
efficacy against non-G1 rotaviruses, most commonly P1AG9,
was similar to that against G1 serotypes, confirming the
hypothesis that a vaccine containing P1A could protect against
P1A-containing viruses of a different G serotype from that of
the vaccine.
A randomized, placebo-controlled safety, immunogenicity,
and efficacy trial was next performed with the RIX4414 vac-
cine (now called by its commercial name Rotarix) in more than
63,000 infants, primarily in 11 countries in Latin America (see
684 SECTION TWO
Table 30-9).291 The primary purposes of the study were to deter-
mine if the rates of intussusception were equivalent in vaccin-
ees and placebo recipients and, in a subset of 10,159 vaccinees
and 10,010 placebo recipients, to determine the efficacy of
the vaccine against severe rotavirus disease and hospitaliza-
tions. The efficacy of Rotarix against hospitalizations resulting
from rotavirus was 85% and reached 100% against the most
severe cases of rotavirus gastroenteritis. Overall efficacy against
severe rotavirus disease was 91.8%. Efficacy against any sever-
ity of disease caused by non-G1 strains (G3, G4, and G9) that
belonged to the same P1A[8] serotype and genotype as the vac-
cine was 87.3%. Although protection was only 41.0% against
G2[P4] strains in this trial, a meta-analysis of the results of this
trial combined with two previous trials of Rotarix289,290 allowed
the calculation of a protection rate against the G2[P4] strain of
67%. Thus, although this vaccine appeared to be less protective
against a rotavirus strain that differed serotypically in both its
VP4 and VP7 neutralization proteins from that of the vaccine,
it still provided protection. It is also important to note that pro-
tection against all cases of severe gastroenteritis or diarrhea-
associated hospitalizations of any etiology in vaccinees in this
trial was 40% or 42%, respectively, thus indicating that an effec-
tive rotavirus vaccine could dramatically reduce the number
of cases of severe gastroenteritis in young children. The final
results of the phase III efficacy study of Rotarix gave protection
rates against severe gastroenteritis caused by G1, G2, G3, G4,
and G9 strains of 96%, 86%, 94%, 95%, and 85% respectively.
In both Europe and Latin America, the efficacy of Rotarix has
been sustained in the second year of life.292
The first phase III trials of Rotarix in developing countries
have recently been completed.293 Rotarix efficacy ranged from
49% (95% CI, 19%-68%) in Malawi to 77% (95% CI, 56%-88%)
in South Africa. The number of severe rotavirus cases prevented
per 100 infants vaccinated was substantial—7 cases per 100
vaccinated in Malawi and 4 cases per 100 vaccinated in South
Africa—illustrating the large potential public health impact of
even a moderately efficacious vaccine.
Effectiveness
A case-control study of the effectiveness of Rotarix conducted
at seven hospitals in El Salvador found that the effectiveness of
two doses of vaccine against rotavirus diarrhea requiring hospi-
tal admission was 76% (95% CI, 64%-84%) and of a single dose
of vaccine was 51% (95% CI, 26%-67%) effective during a sea-
son when homotypic P1A[8]G1 predominated.294 In Brazil and
Belgium, G2[P4] strains have emerged in the setting of a reduc-
tion of rotavirus disease.295–298 Whether this represents normal
genotype fluctuation or true emergence of a strain less likely
to be covered by the vaccine will require further study. In sepa-
rate evaluations conducted in Brazil and in Australia, Rotarix
demonstrated reasonable effectiveness against severe rotavirus
disease caused by G2P4 rotavirus strains in the first year of
life.299,300 However, in both settings, the effectiveness against
G2P4 strains appeared to diminish significantly in the second
year of life.301a
Transmission of vaccine virus
In phase II or III studies of Rotarix administered at 106.5 to
106.8 cell-culture-infective-dose-50 (CCID50) per dose in various
countries, after dose 1, rotavirus antigen shedding was detected
by ELISA in 50.0% to 80.0% of infants at approximately day
7, 19.2% to 64.1% at approximately day 15, 0% to 24.3% at
approximately day 30, and 0% to 2.6% at approximately day
60.301 Shedding was lower after dose 2, and it was detected in
4.2% to 18.4% of infants at approximately day 7, 0% to 16.2%
at approximately day 15, 0% to 1.2% at approximately day 30,
and 0% at approximately day 45. Shedding of live rotavirus was
assessed in two studies in which Rotarix was administered at
106.5 CCID50 per dose. In both studies, stool samples that were
antigen positive by ELISA at approximately day 7 after dose
1 were tested subsequently for live virus by FFU assay. Live
virus was detected in six (46.2%) of 13 and 15 (45.5%) of 33
rotavirus-antigen-positive stools, for an estimated 26% of vac-
cinated infants shedding live virus at approximately day 7 after
dose 1. The potential for transmission of vaccine virus to con-
tacts was evaluated in a study of twin pairs conducted in the
Dominican Republic. In this study, one twin from each pair
received Rotarix and the other received placebo, and fecal spec-
imens were obtained thrice weekly for six consecutive weeks
after vaccination. In 80 infants who received placebo, the rota-
virus vaccine strain was detected in at least one fecal specimen
in 15 infants (19%).302
Adverse events
No excess of adverse effects in vaccinees compared with con-
trol infants was reported in any of the clinical trials just men-
tioned.303,304 However, the virus was shed in concentrations
detectable by ELISA in 60% to 80% of infants305,306 after dose 1
of vaccine and often after dose 2, supporting the possibility of
horizontal transmission to nonvaccinees.
Because of the association of RotaShield with intussusception,
a large phase III safety trial was conducted with approximately
63,000 subjects administered two doses of the Rotarix vaccine,
as mentioned previously. The primary endpoint of this large
trial was the occurrence of intussusception within 31 days after
each dose of vaccine.291 A secondary objective was to assess the
occurrence of serious adverse events, including intussusception,
during the entire study period (i.e., a median of 100 days after
the second dose). A total of 13 cases of intussusception were
identified during the 31 day window, and seven of these were in
placebo recipients. No clustering in the initial 1 to 2 weeks was
identified after either dose. During the entire duration of the
study, 16 cases of intussusception occurred in the placebo group
compared with nine in the vaccine group.
Two separate post-licensure studies have recently evaluated
the safety of Rotarix with regard to intussusception in Mexico.
One study, sponsored by the vaccine manufacturer, reported a
borderline significant 1.7 fold elevation of intussusception risk
during the 0 to 30 day period after the first Rotarix dose; most
of the reported cases were clustered during the first week after
vaccination, but a specific risk estimate for week 1 was not pro-
vided.307 The other evaluation found an approximately fivefold
elevated risk of intussusception in the first week after the first
vaccination dose. Because intussusception is relatively uncom-
mon at 2 months of age when most of the first doses of vaccine
are given, this increased risk translates into 41 excess hospital-
izations (1 per 52,000 vaccinated infants) for intussusception
annually among Mexican children, against a baseline of about
1,200 annual intussusception hospitalizations. A similar level of
risk of intussusception has also been reported for Rotarix among
Australian infants, although this is based on relatively few cases.260
As for RotaShield, the cause of the intussusception that
rarely follows administration of Rotarix is unknown. However,
one could postulate the following. Rotarix is a highly attenu-
ated version of a natural P1A[8]G1 strain obtained from a child
in Cincinnati in 1989. If Rotarix is a rare cause of intussus-
ception, it is possible that natural rotavirus infections are also
a rare cause of intussusception. If this is true, then one could
argue that rotavirus vaccines, which prevent natural rotavirus
infections, might also prevent intussusception caused by natu-
ral infection. The best way to test this hypothesis would be to
determine the incidence of intussusception in countries before
and after the extensive use of rotavirus vaccines. It is possible
that if natural rotavirus infections cause intussusception more
commonly than rotavirus vaccines, the incidence of intussus-
ception could decrease. These studies are ongoing.

Recommendations for use of RotaTeq
and Rotarix
Routine administration
Recommendations for the use of rotavirus vaccines have been
issued by several committees.308,309 The ACIP recommends rou-
tine oral immunization of US infants with either three doses of
RotaTeq administered at 2, 4, and 6 months of age, or two doses
of Rotarix administered at 2 and 4 months of age.232 The mini-
mum age for the first dose is 6 weeks, and the maximum age is
14 weeks and 6 days. Subsequent doses should be given with an
interval of at least 4 weeks, and all doses of vaccine should be
administered by age 8 months and 0 days. The first dose of vac-
cine was not recommended for infants more than 14 weeks and
6 days old, to avoid the age range at which idiopathic intussus-
ception occurs. It was also recommended that no dose of vac-
cine be administered to infants more than 8 months and 0 days
of age because of insufficient data on safety. It is unlikely that
these age restrictions will be loosened any time soon.
Rotavirus vaccines can be administered together with diph-
theria, tetanus, and pertussis (DTaP),
type B (Hib), inactivated polio (IPV), hepatitis B, and pneumo-
coccal conjugate vaccines. Rotavirus vaccine can be adminis-
tered to infants with mild illness. Infants with moderate to
severe illness should be vaccinated as soon as they recover from
the acute phase of the illness.231
The effect of simultaneous administration of both RotaTeq
and Rotarix with OPV on the immune response to each of the
vaccines has been evaluated.310–312 Neither rotavirus vaccine
interferes with the immune response to any of the three anti-
gens in OPV. OPV appears to interfere with the immune
response of the first dose of rotavirus vaccine, but this interfer-
ence is largely overcome after administration of the complete
vaccination series.
Breastfeeding
The efficacy of the vaccine is similar among breastfed and
non-breastfed infants.231 In the phase III pre-licensure study
of RotaTeq, protection against severe rotavirus-induced gas-
troenteritis was 100%, 95.4%, and 100% in children who were
never breastfed, sometimes breastfed, or exclusively breastfed,
respectively.257
Premature infants
Premature infants may be vaccinated if they are at least 6 weeks
of age, have been or are about to be discharged from the hospital
nursery, and are clinically stable. The ACIP considers the ben-
efits of rotavirus vaccination of premature infants to outweigh
the theoretical risks.231
Contraindications
Rotavirus vaccine should not be administered to persons who
have severe hypersensitivity to any component of the vaccine or
who have experienced a serious allergic reaction to a previous
dose. Latex rubber is contained in the Rotarix oral applicator,
so infants with a severe (anaphylactic) allergy to latex should
not receive Rotarix. The RotaTeq dosing tube is latex free. In
response to reported cases of vaccine-acquired rotavirus infec-
tion in infants with severe combined immunodeficiency dis-
ease (SCID) after rotavirus vaccine administration,313 both
RotaTeq and Rotarix are contraindicated in infants diagnosed
with SCID.314
Rotavirus vaccines 30 685
Precautions
Practitioners should consider the potential risks and benefits
when deciding whether to administer RotaTeq to the following
special populations.
Children and adults who are immunocompromised because of
congenital immunodeficiency, hematopoietic transplantation,
or solid organ transplantation sometimes experience severe,
prolonged, and even fatal rotavirus gastroenteritis.171–173
The safety and efficacy of RotaTeq have not been established for
infants with chronic gastrointestinal diseases such as congeni-
tal malabsorption syndromes, Hirschsprung's disease, short-gut
syndrome, or persistent vomiting of unknown cause. However,
the benefits of vaccination outweigh the theoretical risks in
infants with preexisting chronic gastrointestinal conditions
who are not undergoing immunosuppressive therapy.
Because infants with a history of intussusception might be at
higher risk of a repeat episode than other infants, and because
post-licensure data suggest that rotavirus vaccines might be a
very rare cause of intussusception, administration of rotavirus
vaccines to infants with a previous episode of intussusception
should be withheld.
Special Situations
Infants living in households with persons who have or are sus-
pected of having an immunodeficiency disorder or impaired
immune status may be vaccinated. Protection of the immuno-
compromised household member afforded by immunization
of young children in the household very likely outweighs the
small risk of transmitting vaccine virus to the immunocompro-
mised household member and any subsequent theoretical risk
of vaccine virus-associated disease. To minimize potential virus
transmission, all members of the household should use mea-
sures such as good hand washing after contact with the feces of
the vaccinated infant.231
Infants living in households with pregnant women may be vacci-
nated. Most women of childbearing age have preexisting immu-
nity to rotavirus; therefore, the risk of infection and disease from
potential exposure to the attenuated vaccine virus strains is very
low. In addition, immunization of young children would avoid
potential exposure of pregnant women to wild-type virus if the
unvaccinated infant suffers from rotavirus gastroenteritis.
The practitioner should not re-administer a dose of rotavirus
vaccine to an infant who regurgitates, spits out, or vomits dur-
ing or after administration of vaccine. The infant can receive
the remaining recommended doses of rotavirus vaccine at
appropriate intervals.231
If a recently vaccinated child is hospitalized for any reason,
no precautions other than routine universal precautions need
be taken to prevent the spread of vaccine virus in the hospital
setting.231
686 SECTION TWO Licensed vaccines
Post-licensure studies: detection of porcine
circoviruses (PCV) in rotavirus vaccines
In 2010, Eric Delwart and colleagues at the University of
California, San Francisco, using deep sequencing and microarray
technologies, found nucleic acids from PCV type 1 in Rotarix.315
Subsequent studies performed by the FDA, Merck, and
GlaxoSmithKline found evidence for live PCV type 1 in Rotarix
and genetic fragments but not live PCV type 2 in RotaTeq.316–318
Neither PCV-1 nor PCV-2 has been found to cause an immune
response or disease in humans.319 Consistent with that observa-
tion, children inoculated with Rotarix or RotaTeq in phase III tri-
als were not found to have PCV-1 or PCV-2 antibodies in serum.
The source of PCV in both vaccine preparations was probably
the porcine trypsin used to grow rotaviruses in vitro. Although
the presence of PCV does not pose a safety risk, both companies
have committed to removing PCV from their vaccines.
Public health considerations
After the withdrawal of RotaShield vaccine from the market,
there was deep disappointment that, despite the licensing of a
very effective rotavirus vaccine, public health advocates were
back at square one. A vaccine had been developed that had a rare
flaw in its safety profile: intussusception in infants. The situa-
tion was made worse by the facts that the mechanism for the
causation of intussusception was unknown and that the inci-
dence of intussusception attributed to the vaccine was extremely
low (about 1 in 10,000). This meant that establishing appar-
ent relative safety with a new and alternative vaccine could
be claimed only by blindly vaccinating very large numbers of
infants and observing their health for an extended period, with-
out the benefit of any biological markers known in the infant
hosts to determine whether the vaccine trial in progress was safe
(the number of infants tested by two companies that took up
the challenge [GSK and Merck] was more than 60,000 for each
company). That two companies pursued this product at extraor-
dinary cost and with absolutely no guarantee of success is just as
remarkable as the fact that each created, produced, and licensed
a new rotavirus vaccine using different biological approaches.
Several countries that have introduced rotavirus vaccines in
their national childhood immunization programs have already
seen remarkable declines in severe rotavirus gastroenteritis after
vaccine introduction. In the United States, data from a national
Number of positive and total rotavirus tests from 25 continuously reporting NREVSS
laboratories, by week of year, United States, June 2000–July 2010, 3 week moving average
Impact of rotavirus vaccines on the incidence of rotavirus disease in the United States: 2006–2010.
network of sentinel clinical laboratories showed that the tim-
ing and magnitude of the first two rotavirus seasons (2008 and
2009) after vaccine introduction were delayed by 6 to 15 weeks
and decreased in peak magnitude by 42% to 60% compared
with pre-vaccine years.320,321 Similarly, declines in hospitaliza-
tions and office visits for acute gastroenteritis were observed
during the 2008 season in many regions of the country.322–326
Remarkably, large declines were also observed among older
children who were not age-eligible for vaccination, suggesting
indirect benefits from reduced transmission of rotavirus in
the community (Figure 30-4). Similar observations have been
reported from Australia,327 as well as from several countries
in Latin America that are routinely vaccinating against rotavi-
rus.328 Furthermore, in Mexico, where use of Rotarix began in
2007 nationwide, all-cause diarrhea-related mortality fell from
an annual median of 18.1 deaths per 100,000 children before
vaccine to 11.8 per 100,000 children in 2008 (rate reduction,
35%; .001), providing the first evidence of impact of vacci-
nation on mortality from diarrheal disease.329
Against these remarkable health benefits from vaccina-
tion, post-licensure monitoring has also identified some poten-
tial safety concerns. In March 2010, the FDA temporarily
suspended use of Rotarix after the presence of an extraneous
PCV was identified in commercial vaccine lots. Parts of the
genome of PCV were also later identified in RotaTeq. The FDA
later resumed the use of Rotarix and continued to recommend
RotaTeq based on the three considerations: (1) both vaccines
have strong safety records, including clinical trials involv-
ing tens of thousands of patients as well as clinical experience
with millions of vaccine recipients, (2) there is no evidence that
PCVs pose a safety risk in humans, and they are not known
to cause infection or illness in humans, and (3) the benefits of
the vaccines are substantial and include prevention of death in
some parts of the world and hospitalization for severe rotavi-
rus disease in the United States. More recently, emerging post-
licensure data from Australia and from Latin America suggest
the possibility of a low-level risk of intussusception from the
current rotavirus vaccines. Although a risk has not been docu-
mented in ongoing safety monitoring in US infants, if a risk of
the magnitude seen in other settings exists, it would translate
to an excess of approximately one case of intussusception per
100,000 vaccinated US infants. The ACIP reviewed the data on
the demonstrated benefits of vaccination against the potential
risk in US infants and made no change to its recommendation
for continued vaccination of US infants against rotavirus.

Efficacy of RotaTeq and Rotarix against Severe Rotavirus Gastroenteritis in the First Year of Life in Africa and Asia
In 2010, results of phase III trials of both RotaTeq and Rotarix
conducted in developing countries of Africa and Asia were
reported (Table 30-10). The efficacy of both vaccines in these tri-
als was moderate compared with that reported in earlier trials in
affluent settings. However, because of the greater baseline bur-
den of severe rotavirus in developing countries, the public health
benefits of vaccination in terms of absolute number of cases of
severe rotavirus disease prevented were substantial. Although
the exact reasons for the somewhat diminished performance of
rotavirus vaccines in developing countries are unclear, other live
oral vaccines such as those against polio, cholera, and typhoid
have also not worked equally well in populations in developed
and developing country settings. Many factors in infants in the
developing world—levels of maternal antibody, breastfeeding
practices, micronutrient malnutrition, the presence of interfer-
ing microorganisms, and companion diseases such as HIV and
malaria—can alter the ability of the infant to process a live oral
vaccine, lower the effective dose, and interfere with virus replica-
tion and impede a good immune response.330,331 Data from Mali
are limited due to the wide confidence intervals.
The current age restrictions on administering rotavirus vac-
cine dose 1—at the latest by 15 weeks of age, and the full series by
32 weeks of age—could exclude a substantial number of children
from vaccination in developing countries. In countries of Africa
and Asia, where greater than 85% of the estimated half a mil-
lion annual rotavirus deaths occur, delays in immunization are
common. A recent analysis of demographic and health surveys
showed that median coverage for the first dose of diphtheria, teta-
nus, and pertussis vaccines in 45 developing countries was 57%
by 12 weeks of age, rising to 80% by 5 months of age.332 For the
third dose, coverage rates were 27% and 65% by 5 and 12 months,
respectively. A scenario analysis was conducted to assess the
potential benefits of mortality reduction from rotavirus versus
the risk of fatal intussusception (assuming hypothetical risks of
Access the complete reference list online at http://www.expertconsult.com
4. Glass RI, Kilgore PE, Holman RC, et al. The epidemiology of rotavirus
diarrhea in the United States: surveillance and estimates of disease burden. J
Infect Dis 174(suppl 1):S5–S11, 1996.
6. Parashar UD, Holman RC, Clarke MJ, et al. Hospitalizations associated with
rotavirus diarrhea in the United States, 1993 through 1995: surveillance based on
the new ICD-9-CM rotavirus-specific diagnostic code. J Infect Dis 177:13–17, 1998.
16. Tallett S, MacKenzie C, Middleton P, et al. Clinical, laboratory and
epidemiologic features of a viral gastroenteritis in infants and children.
Pediatrics 60:217–222, 1977.
214. Rennels MB, Glass RI, Dennehy PH, et al. Safety and efficacy of high-
dose rhesus-human reassortant rotavirus vaccines: report of the national
multicenter trial. Pediatrics 97:7–13, 1996.
218. Rotavirus vaccine for the prevention of rotavirus gastroenteritis among
children: recommendations of the Advisory Committee on Immunization
Practices. MMWR Recomm Rep 48(RR-2):1–20, 1999.
Rotavirus vaccines 30 687
intussusception similar to those that have now been seen in some
post-licensure studies) when the first dose of the vaccine is strictly
administered by 12 weeks of age compared with a free strategy
with vaccine administered before 1 year of age.333 This analysis
showed that the free strategy would prevent an additional 54,087
rotavirus-associated deaths (total = 248,651) while potentially
resulting in an additional 1,226 intussusception deaths (total =
2,332). Thus, the impact of diminishing real life-saving benefits
of these vaccines through strict age restrictions will need to be
carefully considered in making future recommendations for vac-
cine use in settings with high rotavirus mortality.
Future development and introduction of rotavirus vaccines
will require substantial input from the international donor
community, including the Global Alliance for Vaccines and
Immunizations and the Bill and Melinda Gates Foundation, as
well as creative financing mechanisms such as the recently pro-
posed International Finance Facility. Many local vaccine makers
in China, India, Indonesia, and other settings have begun their
own programs for rotavirus vaccine development and introduc-
tion, and these efforts could provide an additional supply of vac-
cine at an affordable cost to ensure that they reach the poorest
children in the world who need them the most. In addition to
other candidate oral rotavirus vaccines, approaches such as inac-
tivated rotavirus vaccines are being pursued, and this could help
overcome some of the challenges to diminished performance of
oral rotavirus vaccines in developing countries and also poten-
tially avoid a risk of intussusception.334 Such new vaccines are
unlikely to be proposed for international use for many years.
Acknowledgements
The authors acknowledge Roger Glass and Richard Ward for
contributions to this chapter.
225. Centers for Disease Control and Prevention. Intussusception among
recipients of rotavirus vaccine: United States, 1998-1999. MMWR Morb
Mortal Wkly Rep 48:577–581, 1999.
227. Murphy TV, Garguillo PM, Massoudi MS, et al. Intussusception among
infants given an oral rotavirus vaccine. N Engl J Med 344:564–572, 2001.
248. Vesikari T, Matson DO, Dennehy P, et al. Safety and efficacy of a pentavalent
human-bovine (WC3) reassortant rotavirus vaccine. N Engl J Med 354:23–
33, 2006.
291. Ruiz-Palacios GM, Perez-Schael I, Veláquez FR, et al. Safety and efficacy of
an attenuated vaccine against severe rotavirus gastroenteritis. N Engl J Med
354:11–21, 2006.
322. Cortese MM, Tate JE, Simonsen L, et al. Reduction in gastroenteritis in
United States children and correlation with early rotavirus vaccine uptake
from national medical claims databases. Pediatr Infect Dis J 29:489–494,
2010.
Seroconversion Rates and Neutralizing Antibody Responses to Yellow Fever 17D Vaccine in Clinical Trials

Yellow fever vaccine

38
921
 922
Seroconversion Rates and Neutralizing Antibody Responses to Yellow Fever 17D Vaccine in Clinical Trials—cont'd

SECTION TWO
*
Seroconversion Rates and Neutralizing Antibody Responses to Yellow Fever 17D Vaccine in Clinical Trials—cont'd

Yellow fever vaccine

38
923
 924
Seroconversion Rates and Neutralizing Antibody Responses to Yellow Fever 17D Vaccine in Clinical Trials—cont'd

SECTION TWO Licensed vaccines

*

§§
Seroconversion Rates and Neutralizing Antibody Responses to Yellow Fever 17D Vaccine in Clinical Trials—cont'd

*Interval between vaccination and serologic testing.

†
Serum-dilution constant virus N test performed in suckling mice inoculated IC.
‡
Volunteers in the study with no preexisting yellow fever immunity.
§
Seventeen of 28 subjects had previously received an experimental vaccine against dengue type 2.
Seventy-eight percent of vaccinees had high N antibody titers ( 320).
¶
Includes all subjects in the study except pregnant women.
**Serum dilution, constant virus neutralization test in Vero cell cultures using cytopathic effect as the endpoint.
††
Vaccine bulk manufactured by Wellcome and formulated for sale by Evans Medical; Powderject Corp, and then Chiron Vaccines acquired this company.
‡‡
Formerly Connaught, then Pasteur Mérieux Connaught, then Aventis Pasteur, currently Sanofi Pasteur.
§§
Flavimun, derived from the 17D-213-77 secondary seed stock also used to produce the Robert Koch Institute vaccine, manufacture of which was discontinued.
¶¶
Lot produced from a new secondary seed lot; 17DD 102-84 is vaccine prepared from the previous secondary seed lot.
ALV, avian leukosis virus; BJS, BioJect system; cult, culture; DTP, diphtheria, tetanus, pertussis vaccine; GMT, geometric mean titer; HAV, hepatitis A virus; HBV, hepatitis B virus; IC, intracerebral [injection]; JI, jet injector; LLC-MK2,
monkey kidney cells; LNI, log neutralization index; MLD50, mouse LD50 performed per WHO requirements in 4- to 6-week-old animals; M = mouse neutralization test; N/A, not available; ND, not determined, but vaccine meets WHO
standards ( 1,000 MLD50); NSI, needle and syringe injection; NT, not tested; PFU, plaque-forming unit (titration in cell culture; cell type specified); PRNT, plaque reduction neutralization test with geometric mean serum dilution; PS,
porcine kidney cells; SDNT, serum dilution neutralization test; SMLD50, suckling mouse LD50; TCID50, median tissue culture infective dose; WHO, World Health Organization.

Yellow fever vaccine

38
925
926 SECTION TWO
the serum dilution–constant virus neutralization test has not yet
been agreed on by any regulatory agency. The LNI test may, how-
ever, be slightly less sensitive than the serum dilution–constant
virus plaque-reduction method and appears to be influenced by
antibody affinity, which can change over time. For measurement
of the immune response in an individual, the difference in titer
between pre- and postvaccination sera expressed as the LNI rep-
resents the neutralizing capacity of the serum.
The test is performed by mixing a volume of normal or test
serum (undiluted or at a very low dilution) with an equal volume
of a series of tenfold dilutions of virus over a range that includes
the virus titer endpoint, for example 10 2, 10 3, 10 4, 10 5, 10
6
. After incubation for a suitable period (eg, 1 hour at 37° C) the
serum-virus mixtures are titrated in cell culture by plaque assay.
The titer of virus (expressed as log10 PFU/mL) of the test series
is subtracted from the titer in the control series to determine the
LNI. For example, if the virus titer with normal serum is 5.0 log10
PFU/mL and with the test serum it is 2.0 log10 PFU/mL, the LNI is
3.0, reflecting neutralization of 3 logs of virus. If a prevaccination
serum is not available, the LNI is calculated using a normal serum
control. Because complement increases the sensitivity of the yel-
low fever neutralization test, and because complement levels may
be unstable in stored sera, it is preferable to heat-inactivate test
samples; some laboratories add complement (or a standard source
of fresh-frozen serum) to the virus-serum mixture.
The indirect fluorescent antibody, ELISA, hemagglutination-
inhibition, and complement-fixation methods have been used
to measure responses to yellow fever vaccine.368,603,658,662,663,692
When compared to the plaque-reduction neutralization test, a
commercial indirect fluorescent antibody assay (Euroimmun
AG, Lübeck, Germany), showed greater than 95% specificity
and sensitivity for detecting IgG antibodies following primary
yellow fever vaccination, and it appears to be suitable for rapid
determination of serologic response.368 There was a high (98%)
correlation in categorical response but no correlation between
indirect fluorescent antibody and neutralizing antibody titer.
Where critical decisions on protective immunity are required,
however, confirmation by neutralization test may be war-
ranted until sufficient data on sensitivity of indirect fluorescent
antibody tests accumulate. The hemagglutination-inhibition
test is less sensitive than neutralization and is complicated
by low specificity in persons with prior flavivirus exposure.693
The choice of yellow fever antigen might affect results; use of
17D viral antigen provided a more sensitive assay for detecting
vaccine-induced immunity than antigens prepared from the
French neurotropic virus or a wild-type (JSS) strain.692
Persons without prior flavivirus exposure generally do not
develop CF antibodies after administration of 17D vaccine.335,664
The complement fixation test has therefore been thought to
distinguish recent infection with wild-type virus from vaccine-
induced immunity. However, in one study, subjects with prior het-
erologous flavivirus immunity developed broadly cross-reactive
CF antibodies to yellow fever following 17D vaccination.664 The
complement fixation test might correlate with protein-specific
antibodies to NS1. When measured by ELISA or Western blot,
anti-NS1 antibodies are absent after primary 17D vaccination,
but they can be found in vaccinated persons who have preexisting
flavivirus immunity. IgG ELISA is insensitive and unsatisfactory
for detecting seroconversion to 17D vaccine.284,694
IgM antibody responses measured by indirect fluorescent
antibody and ELISA are further discussed in the section “Effects
on immune response”, later.
In the majority of cases, antibody titers are lower and their
appearance is delayed compared to natural infection with yel-
low fever virus,16,22,300 reflecting less virus replication and anti-
gen expression of the attenuated strain. The minimal protective
level of neutralizing antibodies induced by 17D vaccine has
been estimated by dose-response studies in rhesus monkeys
that were challenged after immunization with virulent yel-
low fever.556,607,648 An LNI, measured by plaque reduction, of
greater than 0.7 of that measured before challenge (20 weeks
after immunization) was strongly associated with protection
(Table 38-13). Of 11 vaccinated monkeys that succumbed to
challenge, 10 had a prechallenge LNI lower than 0.5 and one
had an LNI of 0.9. Clinical trials of 17D vaccine have shown
geometric mean LNI values greater than 2.2 measured within
a relatively brief interval (usually 1 month) after vaccination
(see Table 38-12). Antibody titers measured by serum-dilution
plaque-reduction tests have been more variable, and no cut-off
has been established correlating with protection.
Other experimental evidence supports the concept that min-
imal immunity is required for protection. Neutralizing anti-
bodies appear in the sera of rhesus monkeys on days 6 to 7
after inoculation of 17D virus. However, some monkeys survive
challenge performed after a shorter interval (1, 3, or 5 days) after
vaccination, despite the absence of detectable antibodies.19,275
Early protection may be mediated by a low-level specific anti-
body response (see “Kinetics of the immune response”, later),
or by innate immune responses.
For pivotal trials of a yellow fever 17D vaccine between
1999 and 2001, the FDA recognized neutralizing antibodies as
a surrogate for protective immunity against yellow fever and
accepted the LNI 0.7 as the minimum level for protection based
on published animal studies556,653,665 and data showing that pas-
sive immunization with antibodies is sufficient for protection
(see “Passive immunization and passive-active immunization”,
earlier). Because there was an accepted immune surrogate, it
was possible to use the seroconversion rate (using the LNI 0.7
cut-off) as the primary endpoint in pivotal trials.
Passive immunization is an established method for defin-
ing the level of antibody required to protect against virus chal-
lenge. Hamsters were injected with graded doses of homologous
yellow fever immune serum or control serum and challenged
24 hours after transfer. Before challenge, hamsters were bled
for measurement of passively acquired neutralizing antibody
(PRNT50). After challenge, hamsters were followed for illness,
death, and weight change. Full protection was observed in ani-
mals with PRNT50 titers 40 or greater, and partial protection
was observed in animals with passive titers of 10 to 20.695 These
are somewhat higher titers than established as the immune
correlate for encephalitis viruses (eg, Japanese encephalitis),
where a titer of 10 is considered protective in both mice and
humans.696 It is likely that higher neutralizing antibody titers
are required to protect against viscerotropic infections than are
required to protect the brain against neuroinvasion.
The neutralizing antibody response to 17D vaccine has been
evaluated in numerous studies since the vaccine was devel-
oped in the late 1930s. Theiler and Smith21 and Smith et al22
Neutralizing Antibody Response to Yellow Fever 17D
Vaccine Correlates with Protection Against Challenge with 5.0 log10 LD50
of Virulent Asibi Virus
Chi-square: .0001.
From Mason RA, Tauraso NM, Spertzel RO, et al. Yellow fever vaccine: direct
challenge of monkeys given graded doses of 17D vaccine. Appl Microbiol
25(4):539-544, 1973, with permission.
 Yellow fever vaccine 38 927

demonstrated appearance of neutralizing antibodies within

1 to 2 weeks after immunization. In a field study in Brazil
(1937-1938), 94.1% of 882 vaccinees seroconverted after vac-
cination.22 Subsequent clinical trials have confirmed the high
immunogenicity of yellow fever 17D vaccine, with development
of neutralizing antibodies exceeding 90% in nearly all studies
(see Table 38-12). Since the turn of the century, there have been
four large randomized efficacy (immunogenicity) trials in adults
or children, the purpose of which was to compare vaccines for
registration by local national control authorities615,653,665 or to
compare lots made from new secondary seeds and 17D and
17DD substrain vaccines.612 Response rates have been similar
for vaccines produced from the 17DD and 17D-204 (or 17D-
213) substrains, for vaccines formulated with or without stabi-
lizers, and for vaccines with and without avian leukosis virus.
Neutralizing antibody levels following 17D vaccine show a
degree of individual variability (Figure 38-24). However, when a
validated LNI assay was applied in a randomized clinical trial
setting, the 95% confidence limits were quite narrow: Among
seronegative subjects treated with YF-VAX (N 291), geometric
mean LNI was 2.20, 95% CI was 2.12, 2.28; for Arilvax-treated
subjects, geometric mean LNI was 2.08 and 95% CI was 1.97,
2.13.653 These data are similar to those reported in other trials
using the plaque-reduction neutralization method.612,615
Variability of the neutralizing antibody responses in
seronegative healthy adults in a phase III clinical trial of YF-VAX and
ARILVAX. LNI, log10 neutralization index. Acambis Inc., Clinical study report,
Age Protocol H-070-005, 1999; see Monath TP, Nichols R, Archambault WT, et al.
Comparative safety and immunogenicity of two yellow fever 17D vaccines (Arilvax
Some reports suggested that young children did not respond as and YF-VAX) in a phase III multicenter, double-blind clinical trial. Am J Trop Med
well as older persons to 17D vaccine or lost immunity more Hyg 66(5):533-541, 2002.
rapidly,697,698 but this was not confirmed in other contempo-
rary studies.662,699 Interestingly, one pediatric trial did show manufactured by the Institut Pasteur in Senegal, seroconver-
lower neutralizing antibody responses to YF-VAX and Arilvax sion rates 3 months after vaccination were 98.6% and 98.0%,
(90.6% and 94.9% seroconversion and geometric mean LNIs with mean 50% plaque reduction neutralization titers of 159
of 1.26 and 1.32, respectively)665 compared to a study of the and 130, respectively.666
same vaccines in adults (99.3% and 98.6% seroconversion A large double-blind clinical study comparing 17DD and
and geometric mean LNIs of 2.21 and 2.06, respectively) (see 17D-213/77 vaccines in children in Brazil is under way700 and
Table 38-12).653 The two trials used the same serologic assay the results are anticipated with interest. They can be historically
and testing laboratory, but they were conducted in different compared with a previous trial of similar design in adults.612
populations (in children in Peru and in adults in the United In a large clinical trial, there was no difference between
States). Although antibodies to dengue were present at base- young adult and elderly subjects ( 60 years) in the rate of sero-
line in approximately 15% of the pediatric subjects, there was conversion to 17D vaccines.701 Seroconversion rates were 97%
no effect of this covariate on seroconversion or on geometric to 100% in all ages, and the rare primary vaccine failures were
mean antibody titers.665 In a study of seronegative Ghanaian observed only in the younger age groups. Geometric mean LNIs
infants 6 and 9 months of age immunized with 17D vaccine were similar across age groups (Figure 38-25). At least with

Kinetics of the yellow fever 17D–specific CD8+

T-cell response by major histocompatibility complex (MHC) class I
tetramer staining in human subjects. The red line indicates the level
of 17D RNA in blood (mean SD). The black lines are responses of
individual subjects, showing the percentage of positive cells against an
immunodominant epitope on NS4B. From Akondy RS, Monson ND, Miller
JD, et al. The yellow fever virus vaccine induces a broad and polyfunctional human
memory CD8 + T cell response. J Immunol 183:7919-7930, 2009, with permission.
Reverse cumulative distribution of neutralizing antibody
titers (log10 neutralization index [LNI]) for persons 18 to 44 and older
than 60 years. From Monath TP CM, McCarthy K, et al. Yellow fever 17D vaccine
safety and immunogenicity in the elderly. Human Vaccines 1:207-214, 2005, with
permission.
928 SECTION TWO Licensed vaccines

respect to B cell response, there was no evidence for immuno-

logic senescence, possibly owing to the strong inherent adjuvant
activity of the live vaccine.
Sex
There does not appear to be a strong gender bias In one large
trial, the magnitude of the neutralizing antibody response was
statistically significantly higher in male than in female sub-
jects,653 although the seroconversion rates did not differ and the
difference in geometric mean LNI was not biologically relevant
( 0.3 log10) (Table 38-15). In a study in Brazil, the seroconver-
sion rate in male subjects (but not the antibody titer) was higher
than in female subjects.612 Similarly, Niedrig et al also reported
higher neutralizing and IgM antibody titers in male vaccinees.368
Naturally acquired yellow fever disease is more common in
men, a finding that cannot be explained by epidemiologic factors
(see “Risk factors” under “Adverse events”, later). Moreover,
postvaccinal encephalitis caused by French neurotropic vaccine
was higher in male than in female patients, suggesting that male
patients undergo a more active infection (see “Adverse events”,
later). However, it has been noted that men have greater serologic
responses to some nonreplicating vaccines (eg, pneumococcal
and meningococcal polysaccharide),702 so that the gender differ-
ence seen with yellow fever might represent an immunologic
mechanism unrelated to replication of the virus.
Race
European Americans were noted to have higher mean neutral-
izing antibody response than African Americans in two studies
(Table 38-15).653,667 The response was also statistically higher in
European Americans than in Latin Americans.653 The lethality Neutralizing antibody responses (mean SE) in rhesus
of wild-type yellow fever appears to be higher in whites than monkeys immunized with graded doses of yellow fever 17D vaccine.
*Percentage of monkeys positive (log neutralization index 0.7)
blacks. The trial data showing higher antibody responses in 20 weeks after immunization. **Suckling mouse median lethal dose
whites might reflect a higher level of susceptibility, virus repli- of 0.5 mL inoculum. Data from Mason RA, Tauraso NM, Ginn RK, et al. Yellow
cation, and antigen expression in this racial group. fever vaccine. V. Antibody response in monkeys inoculated with graded doses of
the 17D vaccine Appl Microbiol 23(5):908-913, 1972.
Vaccine dose
All yellow fever 17D vaccines contain a minimum of 3.0 log10 antibody response compared to the standard dose given subcu-
IU/dose.614,615 Vaccine potency of manufactured lots exceeds taneously, but the trial design did not elucidate whether this
this minimal limit by approximately 5- to 50-fold to account was due to a more effective route of inoculation or simply the
for losses upon storage. The infectivity assay underestimates fact that (as expected from the foregoing) 17D vaccine efficacy
the number of genome equivalents (measured by RT-PCR) by drops off only at very low doses. Thus, the minimum potency
approximately 1000-fold,623 and the difference can be composed requirements set by WHO for yellow fever vaccines ( 3.0 log 10
of both noninfectious (defective) and infectious virions that IU/0.5 mL) exceeds the 90% immunizing dose by 100-fold.
were not detected by the assay used. To further investigate the potential for dose sparing, Bio-
A dose-response relationship was observed in rhe- Manguinhos has embarked on dose-response clinical trials in
sus monkeys inoculated intramuscularly with 17D vaccine Brazil. It is anticipated that these studies will confirm that it is
(Figure 38-26).556,648 The dose at which 90% of the animals possible to reduce vaccine dose without affecting safety and immu-
developed a rise in LNI greater than 0.7 was about 1000 MLD50 nogenicity. Marginal dose reductions (eg, 2- to 10-fold) could have
(~ 4.0 log10 PFU) and the 90% protective dose against lethal a dramatic effect on vaccine supply. One strategy would be to
challenge was about 200 MLD50 (~ 400 PFU). The 50% immu- deploy such dose reductions in an epidemic emergency, where
nizing dose (ID50) was approximately 2 MLD50 (~ 20 PFU). The demand for vaccine outstrips supply, preferably by reducing vol-
relationship between PFU and MLD50 is approximately 10:1.634 ume of vaccine injected rather than by diluting vaccine in the
Dose-response in humans was first determined by Fox et al703 field. To reset the potency specifications for release of vaccine at
in 1943. The minimal dose resulting in seroconversion was a lower dose would require amending both national licenses and
between 14 and 140 MLD50 (presumed equivalent doses in IU WHO requirements, with data on long-term stability, statistical
and estimated to represent 140 and 1400 PFU, respectively). At noninferiority endpoints in clinical studies, and (possibly) follow-
a dose of 14 MLD50, 70% of the volunteers seroconverted. Large up studies on duration of immunity. A factor to be considered in
scale-field trials of various vaccine lots, some of which were of assessing a lower vaccine dose is dengue immunity, which can
suboptimal potency, indicated that administration of doses in reduce antibody titer to yellow fever (see “Interference with 17D
the range of 10 to 50 MLD50 immunized more than 85% of vac- vaccine caused by heterologous immunity”, later).
cinees. More recent dose-response studies (Table 38-14) indi- Interestingly, an inverse relationship between vaccine dose
cate that doses of 100 to 200 PFU result in the development of and antibody titer has been consistently observed.596,668,669,704
neutralizing antibodies in more than 90% of immunologically Smith et al704 found significantly higher antibody responses in
naïve persons vaccinated. In addition, an unpublished study subjects given doses of 5 to 50 MLD50 than in those given doses
of Arilvax (17D-204 vaccine previously manufactured in the of 500 to 5,000 MLD50. In monkeys, inoculation of large doses of
United Kingdom) found that 13 of 15 (93%) subjects given 200 17D virus resulted in earlier appearance of viremia, but viremia
PFU seroconverted and that all subjects seroconverted at a dose was inconsistent, lower in magnitude, and briefer in duration
of 2,000 PFU or more. Roukens et al640 gave 17D vaccine by the than after inoculation of diluted virus.551 Limited replication of
intradermal route at one-fifth dose and found no differences in the virus after the inoculation of very large doses might explain

In contrast, a study in Brazil of 433 women inadvertently
immunized with the 17DD vaccine during pregnancy or shortly
before conception showed a 98.2% seroconversion rate by neu-
tralization test.735 The majority of the subjects were in the
early stage of pregnancy (mean gestational age at immuniza-
tion, 5.7 weeks; SD, 4.9). The difference between this study
and that of Nasidi et al could be explained based on gestational
age, nutritional factors, or preexisting heterologous flavivirus
immunity (not defined). It is unlikely that difference in the vac-
cine substrain played a role.
HIV infection
Asymptomatic HIV infection with moderate immunosuppres-
sion (CD4+ count 200-499 cells/mm3 or 15% to 24% of cells
for children younger than 6 years) constitutes a precaution
for yellow fever vaccination, and risk should be outweighed
by benefit of protection against exposure to wild-type virus.736
Symptomatic AIDS or asymptomatic HIV with CD4+ counts
less than 200/mm3 constitute contraindications to 17D vac-
cination. Although safety is the primary concern, the accu-
mulated evidence suggests that asymptomatic HIV infection
reduces immune responses to yellow fever vaccine. There is
no evidence at present that 17D adversely affects chronic HIV
infection. The widespread use of 17D vaccine in endemic coun-
tries with high prevalences of HIV mandates a clear under-
standing of the interactions, but data are quite few at this point.
Moreover, the fraction of HIV infected persons who travel to
yellow fever endemic countries appears to be increasing.737
The number of subjects studies with HIV infection or immune
suppression and exposed to 17D is too small for assessing safety,
although the limited experience of retrospective and prospective
studies to date (which includes about 450 subjects, only 10 of
whom had CD4+ counts 200/mm3)736 suggests that vaccination
is not associated with serious adverse events. There is a single case
report of fatal YEL-AND developing in a patient with asymptom-
atic HIV infection and having a low CD4+ count (108/mm3).738
HIV infection has been shown to reduce immunologic respon-
siveness to a number of nonreplicating and live childhood vac-
cines. In the case of inactivated flaviviral vaccines, HIV infection
reduced the seroconversion rate to Japanese encephalitis739
and diminished the humoral and T-cell responses to tickborne
encephalitis vaccine.740 Some case reports indicate that vaccina-
tion of HIV-infected subjects who are not immunosuppressed
(CD4+ T lymphocytes 500/mm3) was followed by serocon-
version.741 Tattevin et al742 vaccinated 12 subjects with CD4+
counts as low as 240/mm3 without adverse effects; all developed
yellow fever neutralizing antibodies, and there were no signifi-
cant changes in CD4+ cell counts or viral loads. On the other
hand, 17D vaccine was administered to 33 adult travelers with
Subtle Host Factor Effects on the Log10 Neutralization Index in a Clinical Trial of Two Yellow Fever 17D Vaccines
*Niedrig et al368 also reported higher neutralizing (and IgM ELISA antibody) titers in male than in female vaccinees.
ANOVA, analysis of variance; ELISA, enzyme-linked immunosorbent assay; LNI, log neutralization index.
Data from Monath TP, Nichols R, Archambault WT, et al. Comparative safety and immunogenicity of two yellow fever 17D vaccines (ARILVAX and YELLOW
FEVER-VAX) in a phase III multicenter, double-blind clinical trial. Am J Trop Med Hyg 66(5):533-541, 2002.
Yellow fever vaccine 38 935
HIV infection and CD4+ cell counts greater than 200 per mm;743
only 70% responded by 1 month after vaccination, and one per-
son seroreverted between 1 and 3 months after vaccination.
A separate study found 93% (13/14) of HIV-infected persons
with no baseline immunity seroconverted following vaccination,
but the time needed to achieve seroconversion was prolonged.737
A retrospective cohort study found that significantly fewer
HIV-infected persons had yellow fever neutralizing antibodies
at 1 year after vaccination compared with vaccinated uninfected
persons (83% and 97%, respectively; = 0.01).744 The study
measured yellow fever neutralizing antibodies in HIV-infected
persons with median baseline CD4 cell count of 496 cells/mm3
(range, 72-1730 cells/mm3) and varying levels of HIV RNA
detected in their blood (52% had HIV RNA levels 50 cop-
ies/mL). Among HIV-infected infants in one developing nation,
only 17% developed neutralizing antibodies within 10 months
of yellow fever vaccine compared with 74% of HIV-uninfected
controls matched for age and nutritional status.745 In addition to
a suboptimal immune response in primary vaccine recipients, a
small study also found that only three of nine HIV-infected per-
sons had a booster effect following repeat vaccination.737
Antiretroviral therapy with resulting improvements in CD4+
counts is associated with an improved response to vaccines. If
feasible, patients starting on antiretroviral therapy should be
deferred for 17D vaccination until CD4+ counts exceed 350 to
400 per mm3 for several months and HIV RNA levels become
undetectable. The report of a case of YEL-AVD in a patient who
had a genetic polymorphism in the CCR5 chemokine recep-
tor indicating a possible defect in innate immunity259 raises
the theoretical concern about yellow fever vaccination of HIV-
infected subjects who are being treated with a CCR5-receptor
antagonists (eg, maraviroc).746 More data are required before a
recommendation can be formulated.
The results of these various studies indicate that HIV infec-
tion can impair the immune response to 17D vaccine. Because
both HIV and 17D viruses exhibit tropism for human lymphoid
cells,747 it is possible that HIV infection interferes with replica-
tion of 17D vaccine. This is supported by data that found HIV
viral load was more likely to determine immune response to the
vaccine than the CD4 count was.748 In vitro studies and clini-
cal trials comparing 17D viremia across HIV-infected and unin-
fected subjects would help to address the issues of how HIV
infection is affecting the immune response to the vaccine.
Other host-specific effects
Evaluation of neutralizing antibody responses in a large clinical
trial revealed some heretofore unknown and subtle host effects on
the immune response (Table 38-15).653 The mean LNI was higher
in smokers than in nonsmokers. There are few studies on the
936 SECTION TWO Licensed vaccines
interaction of smoking and the immune system, but most report
a suppressive effect, for example after hepatitis B vaccine.749 In the
case of 17D vaccine, it was postulated that the adaptive immune
response might be enhanced by the suppression of innate immune
responses (eg, NK cell function) attributable to smoking.
Adverse events
The 17D vaccines have been widely acknowledged as one of
the safest vaccines in use. More than 500 million persons have
been immunized, and the vaccine has a long record of tolerabil-
ity and safety.705,750,751 However, these vaccines have been under
close scrutiny as clinical, histopathologic, and virologic evidence
linked 17D vaccines to severe and previously unrecognized sig-
nificant adverse events. A clinical entity, identified as yellow fever
vaccine–associated viscerotropic disease (YEL-AVD), is reported
to occur at low incidence following the administration of different
17D vaccines, raising concerns about the safety of 17D vaccines.
Historical problems
Two significant events in the history of manufacture and use of
17D vaccine are of special interest. The first (described in “Seed
lot system” under “Yellow fever 17D vaccine”, earlier) was the
occurrence of postvaccinal encephalitis associated with uncon-
trolled passage of 17D substrains during the early years of vaccine
manufacture.596,598,705 The problem was resolved when stabiliza-
tion of passage level during vaccine manufacture was instituted
in 1941. The second event was the development of acute hep-
atitis in persons who received 17D vaccine, and it was recog-
nized as early as 1937 by Findlay and MacCallum.752 Cases were
reported in Brazil during the vaccination campaigns between
1938 and 1940 and, after careful study, were attributed to an
adventitious agent in the vaccine rather than to viscerotropism of
17D virus.753 In 1942, a massive outbreak of hepatitis appeared
in US military personnel immunized with 17D vaccine, resulting
in approximately 28,000 cases and 62 deaths.754 These reactions
were due to the use of pooled human serum (contaminated with
hepatitis virus) as a vaccine stabilizer. This practice was discon-
tinued in Brazil in 1940 and in the United States in 1943,755 and
the problem resolved. Subsequent retrospective serologic studies
confirmed that the responsible agent was hepatitis B virus.756
Common adverse events and clinical studies
Fever, headache, and backache, described as mild, were noted
since the earliest studies of 17D vaccine.21,22 During large-scale
field studies in Brazil in 1937 to 1938, Soper et al597 noted mild
systemic reactions in 5% to 8% of vaccinees. Among 2,457 per-
sons in Brazil from whom “reasonably accurate” clinical follow-
up was obtained by Smith et al,22 14.6% complained of headache
for 1 to 2 days, 10.2% developed pains in the body (usually accom-
panied by headache), 1.4% missed time from work (usually only
1 day) and 0.16% spent one or more days in bed. These reactions,
which were generally considered mild, occurred on the 5th to 7th
day after immunization. Local reactions at the site of inoculation
were not observed, and no systemic allergic reactions were noted.
Reactogenicity of 17D vaccine was monitored in twenty-one
clinical studies conducted between 1953 and 2008, including
one placebo-controlled study (Table 38-16).* Self-limited and
mild local reactions (erythema and pain at the inoculation site)
*References 604,613,653,663,665,666,668,671,674,675,677,681,
684–686,691,757–761.
and systemic reactions (headache, headache and fever, and
fever without symptoms) occurred in a minority of subjects 3
to 7 days after immunization. Reactogenicity in infants is no
greater (or perhaps less) than in adults; this conclusion was also
made during the early studies in Brazil in 1937 to 1938.22
When subjects were under daily surveillance using a diary card
to record local and systemic adverse events, a higher incidence of
adverse events was detected (Tables 38-17 and 38-18).668,674
In one study, 1,440 subjects were monitored, half of whom
received YF-VAX and half Arilvax.653 Safety was assessed
through a diary card and by clinic visits, recording all signs and
symptoms. There were no serious adverse events attributable to
either vaccine. Significantly more subjects in the YF-VAX group
(71.9%) experienced one or more nonserious vaccine-related
adverse events than in the Arilvax group (65.3%, 0.008) (see
Table 38-18). The difference was due to a higher rate of local
reactions of mild to moderate severity caused by YF-VAX.
The most common systemic side effects were headache, asthe-
nia, myalgia, malaise, fever, and chills. These systemic adverse
events are presumably caused by activation of T lymphocytes and
the release of cytokines, including interferons and TNF- during
the viremic period.291 These were generally mild or moderate, but
7% to 8% of all treatment-emergent events were severe and were
reported to interfere with normal activities. Rash was noted in
approximately 3% of the subjects and urticaria in 0.3%. There were
no cases of anaphylaxis, serum sickness, or other severe allergic
reactions. The incidence of nonserious adverse events was lower
in elderly persons than in younger subjects, and the difference was
statistically significant for headache, malaise, injection site edema,
and pain. Lower reactogenicity of 17D vaccine in elderly versus
young adult travelers was also noted in a report by Philipps et al.762
Injection site reactions occurred between days 1 and 5. Systemic
adverse events also occurred at highest incidence during days 1
to 5, but they continued between days 6 and 11. The mean
white blood cell count decreased slightly between baseline and
day 11, with a mild neutropenia. No subjects developed clini-
cally significant thrombocytopenia (0.5 10 5/L).
Elevations in AST were seen in 3.5% to 3.9% of subjects;
3.9% to 4.6% of subjects had an increase in ALT from normal
to abnormal range between baseline and day 11. In most cases,
levels were minimally elevated and levels returned to normal or
had decreased by day 31. Subjects with elevated serum enzymes
had no associated syndrome or constellation of other clinical
laboratory abnormalities. There was no relationship between
elevated serum enzymes and use of concomitant medications.
In a second randomized, double-blind phase III 17D vaccine
study, a total of 1,107 healthy children 9 months to 10 years of
age in Sullana (northern Peru) were recruited (see Table 38-18).
Safety was assessed through a diary card and by clinic visits,
recording all signs and symptoms. No treatment-related seri-
ous adverse events were reported. A similar proportion of sub-
jects in each group reported at least one adverse event: 441 out
of 738 (59.8%) for Arilvax versus 211 out of 369 (59.9%) for
YF-VAX. Most (591; 96.7%) of the adverse events were mild and
resolved without treatment.
These trials were comparative studies of two vaccines,
intended to show noninferiority against a comparator vac-
cine. Without a placebo group, it is difficult to ascribe sys-
temic adverse events and minor laboratory abnormalities to
the vaccines under study. One double-blind trial included a
placebo control.613 This study, conducted in Brazil, enrolled
1,087 subjects 15 to 68 years of age; 272 received a 17D-213
yellow fever vaccine, 543 received two lots of the 17DD vac-
cine, and 272 received placebo (see Table 38-16). The study
participants recorded on a diary card all signs and symptoms
for a 10-day period.
There were no severe or immediate adverse events, and four
hospitalizations within 30 days of vaccination were for reasons
unrelated to vaccination. Compared to placebo, an excess risk
 Yellow fever vaccine 38 965

endemic zone and was conducted at fixed vaccination centers.

In 1994, the responsibility passed to the National Program of
Immunization (COPNI), and in 1998 routine yellow fever vac-
cination of infants was introduced into the EPI. In the endemic
area, the vaccine is given at 6 months of age, whereas in the
coastal area the vaccine is to be given at 9 months of age, reflect-
ing the lower risk of exposure. Currently, all children in 13
states are immunized, along with children in selected (at-risk)
areas in an additional 14 states of Brazil; more than 90% cover-
age has been achieved by 18 months of age.872
Enhanced efforts to cover the population have occurred dur-
ing periods of increased viral activity or geographic expansion. An
example of the latter is the period of 1999 to 2000, when a large
epizootic epidemic swept Brazil, causing 192 human cases and
threatening densely populated areas on the fringe of the endemic
zone. During that interval more than 35 million doses of vac-
cine were administered in mass campaigns, and consideration
was given to introducing routine immunization in coastal areas
outside the endemic zone. Following another major epizootic
wave between 2007 and 2009, which involved states in south-
ern Brazil, additional mass campaigns were instituted. Indeed,
between 1999 and 2009, the interval covering these major expan-
sions of yellow fever in Brazil, 104 million doses were used.919
A continuing debate for policy-makers is the need for vac-
cination of the large population in the nonendemic coastal
region of Brazil at risk for urban yellow fever.920 The risk of seri-
ous adverse events, particularly YEL-AVD, has been judged to
outweigh the benefits,819 although it is acknowledged that the
introduction into the EPI would exclude higher-risk elderly per-
sons and gradually provide an immune barrier to urban yellow
fever. The dramatic clinical presentation of yellow fever and
improved surveillance for infectious diseases are important fac-
tors in making health care policy decisions of this sort, because
an urban outbreak of yellow fever would come to light quickly
and could be contained rapidly.
Other countries in the region have also implemented rou-
tine immunization in the EPI. The early adopters were French
Guiana (1967), Panama (1974), Trinidad and Tobago (1980),
and Brazil. Other countries introduced this policy between
1999 (Ecuador) and 2002 (Colombia). Countries with nation-
wide immunization programs include Colombia, French
Guiana, Guyana, Trinidad and Tobago, and Venezuela. Other
countries in South America perform immunization within the
endemic portions of the country only. Vaccine coverage in South
America varies between 70% and 99% within the endemic
region (Figure 38-30).921
A chronic problem in South America is the movement of
unimmunized persons from coastal regions, where immuni-
zation is not practiced, into the endemic zone. Improvements
in roads, increased settlement within the Amazon region,
and fluidity of human population movements hamper vacci-
nation of immigrants and migrant workers.446 Moreover, the
reinvasion of South America by has increased the
potential for urbanization of the disease and introduction into
coastal regions where immunization is not practiced. In addi-
tion to Brazil, other South American countries have officially
accepted introduction of yellow fever vaccine into the routine
EPI schedule.
The result of vaccination policies in South America is dif-
ficult to assess by historical comparisons with the prevaccine
era, because confounding events coincided with the introduc-
tion of wide-scale immunization, including the introduction
of active surveillance (using viscerotomy, initiated in 1930)
and the expansion of human settlements in endemic areas.
Nevertheless, historically, in countries like Brazil, which
instituted a program of immunization, the incidence of jun- The increased incidence of epidemic yellow fever in Africa,
gle yellow fever declined as vaccination coverage increased beginning in the mid-1980s27,437,923 and the recognition that the
(Figure 38-31),922 whereas in other countries, where coverage disease predominantly affects children,27,54 led to a reassessment
was lower (eg, Peru), large numbers of cases have occurred. of vaccination policy for Africa. In 1988, the joint UNICEF
966 SECTION TWO
A, Incidence of jungle yellow fever in Brazil, 1932-1988.
B, The cumulative number of doses of 17D vaccine administered
in the endemic area, 1932-1988. There is controversy surrounding
the events in Bolivia, and whether the cases that occurred in La Paz
could have acquired infection from contact with the jungle cycle.
Not included in the denominator were four children who had various
underlying conditions, including nephritis, Kawasaki's disease, recent
aseptic meningitis, and recent fever, vomiting, diarrhea, and/or use of
bronchodilators and corticosteroids. Data from Calheiros L. A febre amarela
no Brasil. Cinquentenario da introdução de Cepa 17 D no Brasil. Paper presented
at Simposio Internacional sobre Febre Amarela e Dengue. Rio de Janeiro:
Fundaçao Oswaldo Cruz; 74-85, 1988.
and WHO Technical Group on Immunization for the African
Region and the EPI Global Advisory Group recommended
that countries endemic for yellow fever incorporate 17D vac-
cine into the routine EPI schedule, either at 6 months of age
or at 9 months of age, together with measles vaccine.924,925 In
1990, this recommendation was reemphasized, with the addi-
tional suggestion that catch-up immunization of older children
is needed in countries at high risk. Surveys conducted between
1987 and 1990 indicated coverage of about 80% of infants by
1 year of age in the Gambia, Côte d'Ivoire, and Senegal but rates
of only about 40% in Burkina Faso, Chad, Mauritania, and the
Central African Republic.
By 1991, 14* of Africa's 33 countries at risk for yellow fever
had officially incorporated 17D vaccine in the EPI, but uptake
of the vaccine was poor in most countries, principally because
of lack of donor funding for purchase of vaccine. In 1992 (13
countries reporting), the overall coverage was 19%; in 1993 (12
countries reporting) coverage was 14%; in 1993 (11 countries
reporting), coverage was 29%.926 By 1995, rates were less than
50% in all countries, except the Gambia, the Central African
Republic, and Burkina Faso. The Gambia, which suffered a
major outbreak in 1978 to 197953 and responded with mass
immunization of children and adults, followed by sustained
high rates of coverage of infants in the EPI, is one of the few
African countries that has maintained a high level of vaccine
coverage against yellow fever.
Another technical consensus meeting was held at WHO in
1998, at which time 17 of the 34 countries endemic for yellow
fever had incorporated vaccine into the EPI, although cover-
age remained low. Vaccination coverage has gradually improved
over the last few years, in part because donor agencies, includ-
ing the Bill & Melinda Gates Foundation and the Global
Alliance for Vaccines and Immunization (GAVI), have pro-
vided funds for yellow fever vaccine purchases. Of interest was
the forecast for vaccine doses required under assumptions of
increased coverage rates. In 1997, only 4 million doses of vac-
cine were supplied by UNICEF to African countries, whereas
24 million doses would be required to achieve 80% coverage
of children living in high- and moderate-risk areas in the 34
affected countries, and 240 million doses would be required to
undertake preventive mass campaigns. Demands for vaccine
in response to outbreaks could severely stretch the capacity
of manufacturers. These scenarios should be compared to the
annual production capacity of approximately 30 million doses
from manufacturers serving the African market. Shortages of
vaccine for control of epidemics have been experienced, for
example during an outbreak in Guinea in 2000 and 2001,925,927
leading to a recommendation for a strategic stockpile of vaccine
to be held by UNICEF.
The Guinea case study led to an initiative whereby GAVI
encouraged and supported inclusion of yellow fever vaccine
in the EPI, encouraged catch-up campaigns, and provided a 6
million dose emergency stockpile of vaccine for rapid distribu-
tion in emergencies. By the end of 2004, the number of African
countries practicing routine infant immunization had risen
from 15 in 1999 to 23. GAVI has continued to support the
emergency use stockpile and increased it from 6 to 11 million
doses. Vaccine nearing its expiration date is used in the EPI or
for catch-up campaigns and replaced with fresh doses. In 2007,
$58 million was pledged by GAVI to perform catch-up yellow
fever immunization campaigns with the target of reaching 48
million people in West Africa.928
The purpose of the WHO Yellow Fever Initiative is to
ensure inclusion of yellow fever vaccine in the EPI at age
9 months and older and to implement preventive mass
immunization campaigns for older persons in high-risk
areas. The Initiative has been extremely successful; in the
interval between 2007 and 2009, preventive campaigns were
conducted in Togo, Senegal, Mali, Burkina Faso, Cameroon,
Sierra Leone, Liberia, and Benin; 41 million doses were
administered and coverage in the areas targeted was
90%.763,921,929 To guide the deployment of vaccine in catch-
up campaigns in Africa, WHO prepared a useful, practical
modeling approach to the identification of high-risk areas.930
The risk assessment model employs indicators of exposure to
*Angola, Burkina Faso, Cameroon, Central African Republic, Chad,
Côte d'Ivoire, Gambia, Ghana, Mali, Mauritania, Niger, Nigeria,
Senegal, Togo.
 SECTION TWO: Licensed vaccines
Zoster vaccine
Herpes zoster (HZ), also called shingles, is a dermatomal-vesicular
disease. (A dermatome is the area of skin innervated by one
sensory nerve. Figure 39-1 depicts the right fourth-fifth cervical
nerve [C4-C5] dermatome.) The cutaneous component of HZ
is usually preceded by prodromal pain in the affected derma-
tome, followed by the subsequent appearance of vesicular skin
lesions that are accompanied by pain that often persists after
the skin lesions heal.1
In 1892, Bokay made the seminal observation that hinted at
the cause of HZ. He reported five instances in which varicella
occurred in children exposed to adults with HZ, leading him
to query: “I would like to bring up the question of whether or
not the unknown infectious material of chickenpox could under
certain circumstances manifest itself, instead of a generalized
eruption, as a zoster eruption”.2
In the ensuing half century, clinical and histologic evi-
dence lent support to this prescient suggestion.3–6 In the
1950s, viruses isolated from varicella or HZ were recognized
to have similar properties in tissue culture. Convalescent
serum samples from varicella or HZ fixed complement using
antigen from vesicle fluid from varicella or HZ cases and
immunofluorescent techniques demonstrated that a new
antibody appeared in blood by day 3 of varicella that equally
stained cells infected with virus isolated from varicella or
HZ.7 Antibody appearing after varicella was subsequently
shown to neutralize viral isolates from varicella or HZ.
Weller and coworkers8 concluded that the same virus caused
both diseases and named it “varicella-zoster virus” (VZV).
The etiologic relationship between these two diseases was
elegantly proved when two isolates were obtained from an
immunocompromised child. The first isolate was obtained
when the child developed varicella, and the second isolate
was obtained several years later when HZ developed. DNA
extracted from the two isolates were identical by restriction
endonuclease mapping using multiple enzymes that focused
on known variable regions of the VZV genome.9
The molecular virology of VZV, a member of the Herpesviridae
family, is described in Chapter 37 on varicella vaccine.
Pathogenesis
Localization of latent VZV in sensory ganglia
The VZV and herpes simplex viruses are -herpesviruses that
cause human infections. An essential and unique feature of
-herpesviruses is their ability to become latent in neurons in
sensory ganglia during their initial (primary) infection of the
Myron J. Levin
39
host. The primary infection with VZV is manifest clinically as
varicella (chickenpox), which is endemic worldwide. Varicella
usually occurs in childhood in temperate climates, generally
as winter-spring epidemics.10,11 In countries with widespread
immunization against varicella, this pattern is disappearing. In
tropical climates, the prevalence of varicella remains high in
late adolescence or young adulthood. Because varicella vaccine
was not introduced into the United States until 1995, more
than 97% of adults have had varicella.12
Varicella zoster virus is transmitted by an airborne route as
droplet spread from the pharynx of an index case or from fomi-
tes from skin lesions of an index case of varicella or HZ (see
Chapter 37 on varicella vaccine). It is postulated that the ini-
tial replication of VZV occurs in an epithelial location in the
nasopharynx and subsequently spreads to adjacent lymphoid
tissue, thereby infecting memory CD4+ T cells, such as those
that are abundant in tonsillar lymphoid tissue.13–15 Dendritic
cells may have a central role in this process.16 Activated CD4+
T cells expressing cutaneous homing markers are preferentially
infected and able to deliver VZV to cutaneous epithelia within
a few days of infection.14,17 Infection of cells at the dermal-
epidermal junction leads to the most visible consequence of
VZV viremia, which is the characteristic vesicular lesions.18
The incubation period after exposure may be explained by the
time required for this initial cell-to-cell transmission, innate
immune responses that delay VZV replication, and the devel-
opment of specific immune pathogenesis that results in skin
lesions.15
The significance of the varicella lesions for the pathogen-
esis of HZ is that the termini of sensory axons are located
at the dermal-epidermal junction at the base of vesicles. It is
hypothesized that from this locus, VZV enters and ascends by
retrograde axonal transport to become latent in the soma of
neurons in sensory ganglia. This mechanism is suggested by
the clinical observation that the dermatomes most frequently
affected with HZ are those that have the highest density of skin
lesions during varicella.10 Moreover, early studies with the vari-
cella vaccine indicated that vaccine recipients were much more
likely to develop subsequent HZ with the vaccine strain virus
if they had a vaccine-related rash in the postvaccine period.19
In addition, vaccine-related HZ occurring in recipients of the
varicella vaccine most often appears in the dermatome where
the vaccine is administered.20 However, this is not always the
case, suggesting that viremia occurring after varicella vaccine
administration, as well as during varicella, could provide an
alternative or additional mechanism for VZV to gain access to
sensory ganglia. The presence of VZV in ganglia innervating the
gastrointestinal tract, as well as enteric ganglia, also indicates
that the viremia of varicella can seed autonomic ganglia.21,22
970 SECTION TWO Licensed vaccines

A, A man with herpes zoster in the right C4-C5 dermatome.

B, Dermatome map indicates the area of skin innervated by individual
sensory neurons; gradation indicates dermatomes involved in part A.

Nature of latent VZV DNA
Cranial and dorsal root sensory ganglia contain VZV DNA
detectable by polymerase chain reaction (PCR). This is true
of more than 90% of adult trigeminal ganglia and 70% of tho-
racic ganglia, reflecting the prevaccine epidemiology of varicella;
other cranial nerve and autonomic ganglia contain latent VZV
DNA.21–28 Approximately 2% to 5% of neurons in sensory gan-
glia contain latent VZV DNA; VZV DNA is not present in non-
neuronal satellite cells.29,30 The VZV DNA in neurons is latent
in the sense that infectious virus cannot be isolated from gan-
glia. The latent VZV DNA is present in the nuclei of neurons
in a circular form different from that present in intact virions.31
Episomal DNA also characterizes the physical state of latent
herpes simplex virus in sensory ganglia. The number of copies of
VZV DNA in a neuron is 1,000-fold lower than the burst size in
infected fibroblasts, indicating rapid shut off of VZV replication
in infected neurons.29,30,32 The VZV DNA serves as an active
template during latency from which five or six immediate-early
or early genes of VZV are transcribed during latency and their
gene products detected.25,33–36 A recent report describes five addi-
tional VZV-specific transcripts, including at least one late tran-
script.28 Some of the latency gene products, many of which are
strong transactivators, are not transported to the nucleus dur-
ing latency, whereas they are found in the nucleus during lytic
infection. It is hypothesized that this failure to translocate one
or more of the early gene products prevents the cascade of events
in the nucleus that normally leads to viral replication.34,37

Maintenance of latency by VZV-specific
immune responses
Although the mechanism of latency is not understood, there
is strong evidence that the maintenance of latency is closely
related to the level of VZV-specific immunity in the host, which
first develops at the time of or shortly after the appearance of
skin lesions with varicella.38 These immune responses include
polyclonal VZV-specific antibody and T cell–mediated immune
(VZV-CMI) responses, including CD4+ and CD8+ T-cell VZV-
specific effector and memory cells. These persist lifelong to
protect the host against subsequent cases of varicella and to
prevent HZ (see also Chapter 37 on varicella vaccine).39
The relationship between VZV-CMI and HZ is evident from
numerous and varied clinical observations. It was early recog-
nized that the age-specific incidence and severity of HZ was
greatly increased in patients with immune compromise result-
ing from underlying illness (eg, human immunodeficiency
virus [HIV] infection) or immunosuppressive therapies for
malignancy, autoimmune disease, or organ transplantation.40–44
Natural and iatrogenic experiments indicate that VZV-CMI is
necessary sufficient to maintain latency of VZV and pre-
vent HZ.45 For example, children born with isolated -globulin
deficiencies do not develop severe varicella or experience an
increased incidence or severity of HZ, whereas children with
severe combined immune deficiency often have severe morbid-
ity with VZV infections.46,47 The likelihood of chemotherapy-
treated patients with lymphoma developing HZ correlated with
the preservation or recovery of VZV-CMI and was not influ-
enced by the presence of anti-VZV antibody; anti–varicella
antibody before bone marrow transplantation was not predic-
tive of the occurrence of HZ.44,48 Moreover, recipients of allo-
geneic hematopoietic stem cell transplants, who have their
immune responses totally ablated and then receive replacement
therapy with intravenous -globulin (which has high levels of
anti-VZV antibody), nevertheless have a high prevalence of HZ,
often with severe manifestations. The risk of HZ in this set-
ting abates only with engraftment and return of the potential
for pathogen-specific CMI.49 The essential role of VZV-CMI
in maintaining latency is verified by the success of an exper-
imental, inactivated vaccine that boosted VZV-CMI (but not
anti-VZV antibody) after stem cell transplantation and thereby
reduced the frequency and severity of HZ.50,51
Clinical manifestations
Varicella zoster virus, because of the global endemicity of vari-
cella, is latent in many ganglia in most humans and has the
potential to reactivate intermittently in a subclinical manner.
This is suspected because unexplained increases in VZV-specific
antibody occur intermittently, and VZV DNA is intermittently
detected in the blood of asymptomatic immune-compromised
and immune-competent people.42,52–54 Presumably these ran-
dom events, , are of no consequence
because the VZV-CMI present when they occur is sufficient to
prevent propagation of VZV infection in the ganglion. The sub-
sequent boost in immunity that accompanies subclinical reac-
tivation may be a factor in maintaining VZV-CMI throughout
life (termed “endogenous boosting”). When an immune per-
son is exposed to someone with varicella or HZ, VZV-specific
immunity may also be boosted (termed “exogenous boost-
ing”).55–58 The accumulation of activated VZV-specific CD4+
T cells in older people may be a reflection of these phenomena.59
Herpes zoster is the outcome when reactivation occurs in
the absence of some (undefined) critical level of VZV-CMI. In
this scenario, the appearance of infectious VZV in a ganglion,
unchecked by the ambient VZV-CMI, leads to ganglionitis,
Zoster vaccine 39 971
during which many neurons and supporting cells are damaged
by the widespread infection and/or the intense inflamma-
tory response that follows.60–65 An alternative potential role of
VZV-CMI is to directly prevent (rather than limit) VZV reacti-
vation. There is insufficient information to distinguish between
these two possibilities. Although the inadequate immune
response to a VZV reactivation that culminates in HZ may
result from therapy-related or disease-related immune suppres-
sion, much more often it is the result of the decline in VZV-
CMI that accompanies the normal aging process.66–70 It is also
likely that the level of VZV-specific immunity, at any age, can be
temporarily blunted by changes in mental health,71 depression
(M.R. Irwin et al, unpublished data), stress,72 or intercurrent
infection with viruses that can alter CMI responses, such as
Epstein-Barr virus and cytomegalovirus.73,74 Trauma to a derma-
tome may also lower the threshold for symptomatic reactivation
in the ganglion innervating that dermatome.75 These additional
factors explain why children sometimes develop HZ.76,77 Race
may also influence the age-specific incidence of HZ.78 There is
uncertainty about familial trends in the incidence of HZ and its
complications.79,80
The ganglionitis and related neuronal damage that accom-
pany extensive VZV reactivation cause the neuropathic prodro-
mal pain in the dermatome where skin lesions subsequently
appear. Prodromal pain accompanies 70% to 80% of HZ cases
in older adults.1,81–83 The prodromal pain typically lasts 3 to
4 days, but may last a week or longer. Its character varies with
the patient: shooting, boring, aching, and throbbing are some
descriptors. The pain may be constant or intermittent. Intense
itching is common.84 The cause of this localized pain in the
involved dermatome initially is unclear to the medical provider,
often leading to a search for visceral disease suggested by the
location of the pain, such as myocardial infarction when a left
upper thoracic dermatome is involved; renal stone or inter-
vertebral disk disease when a lumbar dermatome is involved;
or intra-abdominal disease (eg, cholecystitis, appendicitis)
when right-side mid-lower thoracic dermatomes are involved.
Approximately 12% of medical costs for HZ in the United
States occur before the appearance of skin lesions.85,86
The duration of the prodrome represents the time required
for VZV to replicate in the ganglion and then descend the
involved nerve to the dermal-epidermal junction and for
subsequent VZV replication in the skin to induce the charac-
teristic rash. The appearance of skin lesions reveals the origin
of the prodromal pain and provides the diagnosis. Mild con-
stitutional symptoms are infrequent (10%-20%). The rash of
HZ in immunocompetent patients characteristically involves
a single dermatome; hence, the lesions do not cross the mid-
line except for minor individual variations in dermatomes of
segmental nerves (Figure 39-1).81,82 Several contiguous derma-
tomes may develop lesions, partly because the normal variation
in innervation gives the impression that adjacent dermatomes
are involved. Infection of epithelial cells causes a varicelliform
rash that is limited to the dermatome involved. After brief mac-
ular and papular stages, the characteristic rash (ie, vesicles on
an erythematous base) appears. New vesicles appear for 3 to
4 days in groups that tend to cluster where there are branches
of the involved cutaneous sensory nerve. Pustulation of vesi-
cles begins within 1 week, followed 3 to 5 days later by lesion
ulceration, crusting, or both. The lesions may evolve over a lon-
ger time or become hemorrhagic in patients with advanced age
or immune suppression.87 The rash is usually accompanied by
the same pain experienced during the prodrome, but this acute
phase pain can worsen, improve, or appear for the first time
after the skin lesions appear.88 The acute pain and the preceding
prodrome can be very severe, disabling, and a barrier to employ-
ment and activities of daily living. The nociceptive pain, which
accompanies extensive skin involvement, is additive to the
neuropathic pain. Itching may also increase in prominence.84,89
972 SECTION TWO
It is likely that, to a limited extent, VZV gains access to the
bloodstream, as demonstrated by VZV DNAemia during the
early stages of HZ.54,90,91 This is of little consequence to most
patients, but in immunocompromised patients, extensive vire-
mia in the absence of a vigorous immune response can result
in a disseminated form of HZ that includes severe, multiorgan
disease.87,92–94 Additional insight into the age-related decline in
VZV-CMI is gained from the observation that elderly patients
with HZ frequently have cutaneous VZV lesions at a distance
from the involved dermatome.87 This reflects their inability
to muster adequate VZV-CMI in a timely manner after VZV
reactivation, thereby permitting more extensive viremia,
unimpeded VZV replication at distant sites, or both.
In addition to the acute pain, complications of HZ occur in
approximately 8% of patients who are 50 to 59 years old and in
more than 12% of persons 70 years or older.86 Complications
include bacterial superinfection of skin (2%); segmental motor
nerve damage, including nerves of the face, limb weakness,
and other neurologic complications (3%-5%); and dysfunction
of bowel or bladder when sacral nerves are involved.82,86 The
presence of motor involvement may exceed 10% if this finding
is carefully sought by a neurologist.95 Motor deficits are rarely
permanent, although recovery is less likely to be complete in
older patients. The occurrence of motor deficits is consistent
with data that VZV often extends through the ganglionic root
to the central nervous system. For example, lymphocytic pleo-
cytosis and VZV DNA are commonly found in the cerebrospi-
nal fluid of patients with HZ, and magnetic resonance imaging
studies obtained during HZ often demonstrate inflammation
in the spinal cord at the level of HZ involvement.96 Spread to
the spinal cord has been confirmed histologically in patients
who had HZ when they died of unrelated diseases.26,61,97 While
these abnormalities are generally inconsequential, they prob-
ably explain the uncommon complication of transverse myelitis
and bilateral dermatomal involvement, occasionally seen in an
immune-compromised patient.98,99 VZV meningitis and menin-
goencephalitis can occur with HZ.26,100 Because the ophthalmic
branch of the trigeminal nerve is involved in 10% to 15% of
cases, damage to ocular structures is a common occurrence (5%
of HZ in older patients).86,101,102
The most common complication of HZ is the persistence
of significant pain for months after onset of the rash.88,89,103–105
The frequency of this postherpetic neuralgia (PHN) is strongly
influenced by the age of the patient. Postherpetic neuralgia has
been defined as pain persisting for an extended period after
the onset of HZ (ie, prodromal pain) or after the onset of HZ
rash. Thus, the age-specific frequency of PHN varies with its
Age-specific incidence of herpes zoster as a function of age.
definition. The postevent interval used to define PHN has been
between 30 and 90 days, and because the onset of rash is easily
recognized and remembered, most investigators now use this
starting point, and many now define PHN as pain being pres-
ent at 90 days after the rash appears.82,86,106 Age is the strongest
prognostic factor for the occurrence of PHN, which is uncom-
mon for people younger than 40 years but becomes common
when HZ occurs in people older than 50 years. The various
definitions have led to estimates of PHN ranging from 7% to
25% of HZ cases. Herpes zoster cases that occurred during a
prospective zoster vaccine trial in subjects at least 60 years old
resulted in significant pain lasting or beginning (sometimes
pain occurs late after the rash appears) at least 30 days after
rash onset in 30% of placebo recipients; pain was present for
60 days in 17% and for 90 days in 12%.106 In some patients,
PHN may last for a year or longer. Postherpetic neuralgia may
be intermittent or constant, stabbing or lancinating, deep
burning or throbbing, or allodynia, which is a painful sensation
resulting from an otherwise normal stimulation of the skin. It
is the third most common cause of chronic neuropathic pain
in the United States, with a point estimate of 500,000 cases
annually.107 Acute and chronic pain strongly influence the qual-
ity of life, with effects on physical, psychological, social, and
functional domains. These effects are particularly profound in
elderly patients.108–110
Epidemiology
Herpes zoster is of endogenous origin, resulting solely from
reactivation of latent VZV in sensory ganglia. As such, the at-
risk population is enormous because 95.5% of US-born adults
20 to 29 years old and more than 99.6% of persons older than
40 years have anti-VZV antibody.12 The risk of HZ may change
in subsequent years as a result of the 1995 recommendation for
universal immunization with the varicella vaccine. Although
recipients of the varicella vaccine can develop HZ with the vac-
cine strain VZV,20,111 the frequency and the severity of this occur-
rence are believed to be less than after natural infection.112–114
In the seminal study by Hope-Simpson,115 who recorded
16 years of HZ in his practice, the frequency of HZ corre-
lated closely with increasing age. His findings have been repli-
cated in many countries during an extended observation period
(Figure 39-2).86,116 Herpes zoster is 5- to 10-fold more likely to
occur after 60 years of age than in childhood. Thus, although
the frequency of HZ in the general population has been reported
to be 1.2 to 4.8/1,000 person-years, the frequency increases
(With permission from J. Pellissier and M. Brisson, Merck & Co, Inc.)

to 7.2 to 11.8/1,000 person-years for people at least 60 years
old.81,116,117 Using current census data and these age-specific
rates, it is likely that more than 1 million cases of HZ occur
annually in the United States. Of these, 45% to 50% of cases
occur in people 60 years or older; almost 20% occur in people
50 to 59 years old.86 The lifetime risk of HZ approaches 50%
for people who reach 80 years of age. The annual risk of HZ
in people older than 60 years is greater than 1.1 per 100 per-
sons.106 The severity of HZ also increases with age, as shown
by the greater frequency and duration of PHN with age and the
increase in complications with age.86,104,105,116–119
In addition to age, other strong risk factors repeatedly identi-
fied for PHN and other complications of HZ include the inten-
sity of the prodromal pain, the intensity of the acute pain, and
the extent of the rash.82,89,86,104,120 Each of these is probably a
marker for the inability of the patient to rapidly mobilize VZV-
CMI responses early after VZV reactivation occurs and thereby,
failure to limit virus-induced damage within the involved gan-
glion.121 Immune compromise from disease or immunosuppres-
sive therapy is another important risk factor, but this probably
contributes less than 10% of the societal burden of HZ.85,122
Treatment
The acute phase of HZ should be treated as soon as possible
with nucleoside analogs (eg, acyclovir, valacyclovir, famciclo-
vir).82,105 These are excellent antiviral drugs that inhibit VZV
in vitro and limit virus replication in vivo (ie, reduce dura-
tion of shedding). Their use is invariably suboptimal because
of the delay in appreciating the diagnosis, during which time
the ganglionitis proceeds. Nevertheless, six placebo-controlled
clinical trials with these antivirals administered within 72
hours of rash onset significantly limited new lesion formation,
shortened the time to healing, and decreased the severity and
duration of acute pain. Valacyclovir and famciclovir have bio-
availability superior to that of acyclovir and, therefore, can be
dosed less frequently, but the clinical superiority of one of these
drugs has not been established. The effect of antiviral drugs on
PHN is controversial. Three meta-analyses and some (but not
all) controlled trials failed to determine that antivirals reduced
the incidence of PHN, but some suggested that the duration
of prolonged pain was shortened for some patients.82,105,123–127
Positive results on acute and chronic pain were obtained under
ideal prospective trials conditions, which would be difficult to
achieve in routine practice, especially because fewer than 60%
of patients receive antiviral drugs within 72 hours after rash
onset, as mandated by the successful trials. Moreover, the pro-
portion of advanced elderly patients participating in the clinical
trials was generally lower than found in most practices. There
may be some value in beginning antiviral drugs even later
than the 72-hour window after rash onset recommended for
administration, especially in very old patients whose immune
response may be delayed.82
Additional treatment in the acute phase of HZ often requires
the use of strong analgesics for pain.82 Opioids are effective in the
early phase of HZ and are a mainstay for persistent pain.128–130
Corticosteroids shorten the duration of healing and acute pain
and hasten the return to a normal quality of life when used with
concomitant antiviral therapy, but their use must be weighed
against potential side effects.131,132 Corticosteroids do not prevent
PHN. Drugs that act on nerve transmission, such as gabapen-
tin and pregabalin, are under study for use in the acute phase
of HZ.128,133 Other than analgesics, the clinical impact, although
statistically significant, of drugs used to treat HZ in the acute
period is modest.
Significant PHN often requires multiple drugs.82,130,134
Strong analgesics, usually opioids, are most often requi-
red.82,128–130,135 Many patients get additional relief from gabapentin
Zoster vaccine 39 973
or pregabalin and another class of neural-active drugs, tricyclic
antidepressants.130,136,137 Some patients benefit from topical appli-
cation of lidocaine patches, which inhibit skin sensations that
are perceived as allodynia.138 Capsaicin is a topical therapy that
desensitizes receptors on sensory nerve axons and inhibits ini-
tiation of pain transmission. Because capsaicin initially causes a
local burning sensation, it is difficult to study in controlled trials,
but there are credible reports that topical capsaicin ameliorates
PHN in some patients.139 More complex treatment approaches
for unresponsive PHN can be provided by pain specialists.82
Despite these numerous therapies, the management of
HZ pain in the acute phase, especially PHN, is difficult. Only
about 50% of patients benefit from each of the therapies men-
tioned and then experience only partial relief. Moreover, many
of these drugs frequently produce disabling side effects, espe-
cially in elderly patients who are most likely to get HZ and
PHN. Some of the side effects, such as difficulty concentrat-
ing, difficulty with balance, and bowel or bladder dysfunction,
are preexisting problems for many older patients. Furthermore,
adverse drug effects are more common when combinations of
drugs are required for intractable pain, when they are metabo-
lized more slowly in older patients, and when they are used for
patients already receiving many other drugs with potent side
effects or with the potential for drug interactions. The therapy
for severe PHN is time-consuming for treating physicians and
is expensive.
Passive immunization
Passive protection is not successful, as indicated by the frequency
of HZ in patients after allogeneic hematopoietic cell transplan-
tation, even though these patients receive large amounts of
VZV-specific antibody, which is contained in the intravenous
immunoglobulin typically administered for many months after
transplantation. Furthermore, older people who frequently
develop HZ have high levels of VZV-specific antibody.66–68,140,141
Active immunization
Rationale for a zoster vaccine
When Hope-Simpson115 showed in his seminal monograph that
the frequency of HZ increased with age, he also suggested that
this was the consequence of an age-related decline in VZV-
specific immunity. His conclusion was consistent with the clin-
ical information available at that time, since corroborated, that
immune-compromising diseases and therapies are associated
with a large increase in the frequency and severity of HZ. We
now know that the essential component in protection against
HZ is VZV-CMI. Distinct humoral and T cell–mediated immu-
nity had not been defined when Hope-Simpson115 discerned the
relationship of age and HZ frequency. The hypothesized pro-
gressive decline in VZV-specific immunity (ie, VZV-CMI) has
been confirmed repeatedly, and this immunosenescence is now
considered to be one characteristic of the normal aging process
(Figure 39-3).66–68,70 The close relationship between the decline
in VZV-CMI and the increasing frequency of HZ with aging is
assumed to represent cause and effect. This relationship and
the continuing decline in VZV-CMI are maintained into the
sixth, seventh, and eighth decades.69 Anti-VZV antibody levels
do not decline with age.66–68,140,141 In addition, Hope-Simpson115
noted that second cases of HZ were rare in elderly patients, sug-
gesting that the occurrence of HZ sufficiently stimulated VZV-
specific immunity, in most cases, to prevent second attacks.
Many investigators have documented that second attacks of HZ
980 SECTION TWO
Elderly household contacts of VZV-susceptible
immune-compromised patients and pregnant women
It is important that elderly persons receive zoster vaccine if they
are in contact with people who might develop varicella because
HZ can cause varicella in VZV-naïve contacts.
Current research on vaccines to prevent HZ
The problem of immune senescence is likely to complicate
efforts to improve the efficacy of a vaccine to prevent HZ.
The administration of two doses of the current vaccine is not
Access the complete reference list online at http://www.expertconsult.com
69. Levin MJ, Oxman MN, Zhang JH, et al. Varicella-zoster virus–specific
immune responses in elderly recipients of a herpes zoster vaccine. J Infect
Dis 197:825–835, 2008.
82. Dworkin RH, Johnson RW, Breuer J, et al. Recommendations for the
management of herpes zoster. J Infect Dis 44(suppl 1):S1–26, 2007.
86. Yawn BP, Saddier P, Wollan P, et al. A population-based study of the incidence
and complication rates of herpes zoster before zoster vaccination. Mayo Clin
Proc 82:1341–1349, 2007.
87. Straus SE, Oxman MN. Varicella and herpes zoster. In: Freedberg IM, Eisen
AZ, Wolf L, et al, (eds) Fitzpatick's Dermatology in General Medicine,
New York, NY: McGraw-Hill; 2008. 2427–2450.
106. Oxman MN, Levin MJ, Johnson GR, et al. A vaccine to prevent herpes
zoster and postherpetic neuralgia in older adults. N Engl J Med 352:
2271–2284, 2005.
108. Schmader K. Herpes zoster in the elderly: issues related to geriatrics. Clin
Infect Dis 28:736–739, 1999.
significantly more immunogenic than a single dose in boosting
VZV-CMI responses.156 Much higher doses can be safely given,
but it is unclear that these will significantly improve VZV-CMI
responses (Levin et al, unpublished),190 and this may not be
commercially feasible. It is unknown if either approach will alter
the persistence of the protective response. A heat-inactivated
formulation of the current vaccine is under study for use in
highly immune-compromised patients (http://clinicaltrials.gov,
study NCT01229267).51 A recombinant vaccine containing
VZV glycoprotein E and the proprietary adjuvant ASO1b is the
subject of a large efficacy trial in elderly subjects. This vaccine
may also be of unique value in immune-compromised patients
(http://clinicaltrials.gov, study NCT01165229).
110. Schmader K, Johnson G, Ciarleglio M, et al. Effect of a zoster vaccine on
herpes zoster–related interference with functional status and health-related
quality of life measures in older adults. J Am Geriatr Soc 58:1634–1641,
2010.
115. Hope-Simpson RE. The nature of herpes zoster: a long-term study and a new
hypothesis. Proc R Soc Med 58:9–20, 1965.
158. Harbecke R, Oxman MN, Arnold BA, et al. A real-time PCR assay to
identify and discriminate among wild-type and vaccine strains of varicella-
zoster virus and herpes simplex virus in clinical specimens. J Med Virol
81:1310–1322, 2009.
166. Harpaz R, Ortega-Sanchez IR, Seward JF Advisory Committee on
Immunization Practices (ACIP), Centers for Disease Control, Prevention
(CDC). Prevention of herpes zoster: recommendations of the Advisory
Committee on Immunization Practices (ACIP). MMWR Recomm Rep
57(RR-5):1–30, 2008.
 SECTION TWO: Licensed vaccines
Combination vaccines
The combining of multiple antigens into a single vaccine
is not a new concept; combination vaccines have long been
a bedrock of immunization programs. During the past two
decades, we have transitioned from a world in which DTP
and MMR represented essentially the only combination vac-
cines in use to one in which even developing countries are
routinely using various modern combination vaccines provid-
ing protection against diphtheria, tetanus, pertussis, hepatitis
B, type b, or polio. The field of com-
bination vaccines has matured to the point at which a very
large number of combination vaccines is available and rela-
tively few remain in development. These efforts, undertaken
by many manufacturers and research entities worldwide, have
been driven by the recognition that the continual increase in
the number of effective childhood vaccines posed substantial
economic and logistic difficulties. Providing these vaccines as
separate injections requires multiple needle sticks, leading to
distressed parents, providers, and vaccinees. Scheduling addi-
tional vaccination visits to reduce the number of injections
per visit increases costs, burdens staff, and jeopardizes the
entire immunization program by increasing the likelihood of
missed vaccinations. The shipping, handling, storage, and
accountability of a plethora of vaccines are burdensome and
expensive and increase the possibility of error. These issues
have stimulated continuing efforts to provide new combina-
tion vaccines. However, the development and evaluation of
combination vaccines can pose numerous issues, as discussed
further in this chapter and reviewed elsewhere.1,2
The combination vaccines in common use two decades ago
included diphtheria and tetanus toxoids, available alone (DT
or Td) or with whole-cell (DTwP) pertussis vaccine; inactivated
(IPV) or live oral (OPV) trivalent poliovirus vaccine; and mea-
sles and rubella vaccine, available alone (MR) or with mumps
vaccine (MMR).
The first combination vaccine licensed in the United States
was trivalent influenza vaccine, approved in November 1945,
and the second was a hexavalent polysaccharide pneumococcal
vaccine, licensed in 1947.3 DTwP, although developed in 1943,
was not licensed until March 1948. IPV was licensed in 1955,
and the individual OPV serotypes were licensed from 1961 to
1962. Efforts to overcome the interference seen with simultane-
ous administration of three attenuated live polioviruses delayed
the licensure of trivalent OPV until June 1963. MMR and MR
were licensed in April 1971, and quadrivalent meningococcal
vaccine in 1978. Only brief mention is made in this chapter of
these traditional combination vaccines, which are discussed in
their respective chapters elsewhere in this text.
Most modern pediatric combination vaccines begin with
a DTwP or DT/acellular pertussis (DTaP) vaccine and add
Michael D. Decker
Kathryn M. Edwards
Hugues H. Bogaerts 40
such antigens as IPV, conjugate type b (Hib)
and hepatitis B (HepB). As development efforts for the DT(a)
P-based combinations have matured, some manufacturers
have turned their efforts toward developing so-called second-
shot or companion combinations designed to be given in coor-
dination with a DT(a)P-based combination, incorporating for
example conjugate pneumococcal (PnC) and conjugate menin-
gococcal (MnC) antigens. A third developmental stream has
been directed toward combination vaccines targeted prin-
cipally at travelers, typically based on HepB or hepatitis A
(HepA) components.
Our focus in this chapter is on the most relevant current
combination vaccines, which merge products such as IPV,
HepB, Hib, or meningococcal vaccines with each other or
with one or more of the aforementioned traditional combi-
nation vaccines. Readers interested in combination vaccines
of historical interest (eg, yellow fever/smallpox) or combi-
nations based on components no longer available (eg, Acel-
IMUNE, PRP-D) are referred to previous editions of this text.
Long-established combinations (eg, DTP) and those that
consist solely of multiple serotypes of the same organism
(eg, IPV; pneumococcus; meningococcus) are covered in their
respective chapters.
Terminology
The vaccine industry has undergone dramatic consolidation
during the past 20 years; long-established companies and new
biotechnology start-ups have been acquired or merged. These
changes render nomenclature problematic. For vaccines cur-
rently marketed, we will use the name of the current manufac-
turer, even when describing studies conducted by a predecessor
company. For products not presently marketed, we will use the
name of the company that produced them, even if that com-
pany now is owned by or operates under a successor name. To
further assist readers, Table 23-6 lists most current major vac-
cine manufacturers, along with the names of predecessor, com-
ponent, or acquired companies.
In this chapter, the use of a virgule to coordinate the names
of two vaccines (eg, DTwP/IPV) indicates a combination of the
two vaccines; a plus sign (eg, DTwP + IPV) indicates concurrent
but separate administration. When discussing a specific com-
bination vaccine, the use of a single virgule indicates that the
vaccine is supplied with the component following the virgule
premixed with the component preceding the virgule; the use of
a double virgule (eg, DTaP//Hib) indicates that the component
following the double virgule is reconstituted using the liquid
components that precede the double virgule.
982 SECTION TWO Licensed vaccines

Because so many different combination vaccines are avail-
able within some classes (eg, DTaP/Hib) and simple, unambigu-
ous generic names do not exist for the various products, trade
names are used whenever possible to refer to specific combina-
tion vaccines.
Table 40-1 outlines the current status of the newer combina-
tion vaccines, and Table 40-2 provides further details concern-
ing those that are licensed.
Principles of combined vaccines
Simultaneous vaccines
A combination vaccine consists of two or more separate
immunogens that have been physically combined in a sin-
gle preparation. This concept differs from that of simultane-
ous vaccines, which, although administered concurrently, are
physically separate (ie, injected at separate sites or given by
separate routes). Although some studies have shown altered
immune responses to various vaccines when they are given
concurrently with other vaccines but at separate sites, there
is no evidence that the efficacy of any vaccine recommended
for routine use in childhood is materially altered by concomi-
tant administration with any other vaccines recommended for
administration at the same age.4 Similarly, adverse events after
concomitant administration of multiple vaccines generally are
increased only modestly, if at all, compared with events after
the administration of the most reactogenic vaccine alone.
Thus, we will not review the many studies that have evalu-
ated simultaneous administration, except to the extent that
they provide reference data to which results from combined
vaccines can be compared.
Combination vaccines and the immune system
Antibodies recognize conformationally determined epitopes
on protein or polysaccharide antigens. Modification of an anti-
gen's B-cell epitopes during vaccine preparation may reduce
the ability of vaccine-induced antibody to bind to the patho-
gen. Consequently, the techniques used to produce an anti-
gen may have important implications for the immunogenicity
(and, presumably, the efficacy) of a vaccine containing that
antigen.5 This principle is illustrated by the results from the
Multicenter Acellular Pertussis Trial, which compared 13 acel-
lular pertussis (aP) vaccines and found that levels of antibody to
pertussis toxin correlated poorly with the quantity of toxoided
pertussis toxin present in the vaccines.6 For example, one of
the acellular vaccines contained a genetically inactivated per-
tussis toxin produced by recombinant technology. This vaccine
produced markedly higher antibody responses per microgram
than did the remainder of the evaluated vaccines, whose pertus-
sis toxin components were inactivated chemically rather than
genetically.6

Combination Vaccines Presently Available or Under Development*

aP, acellular pertussis vaccine (infant formulation); ap, acellular pertussis vaccine (adolescent/adult formulation); D, diphtheria toxoid vaccine (infant formulation); d,
diphtheria toxoid vaccine (adolescent/adult formulation); GSK, GlaxoSmithKline; HepA, hepatitis A vaccine; HepB, hepatitis B vaccine; Hib, conjugate
type b vaccine; IPV, enhanced inactivated trivalent poliovirus vaccine; MMR-V, measles, mumps, rubella, and varicella vaccine; Mn, meningococcal
conjugate vaccine (included serotypes are shown by following letter(s), eg MnC, MnCY); SII, Serum Institute of India; SP, Sanofi Pasteur; SPMSD, Sanofi Pasteur MSD;
SSI, Statens Seruminstitut; T, tetanus toxoid vaccine; wP, whole-cell pertussis vaccine. See Table 23-6 for information on prior names of pharmaceutical companies.
*
Products combining only multiple serotypes of a single pathogen are excluded, as are DT, Td, DTP, DTaP, Tdap, OPV, IPV, and MMR. Only manufacturers who
distribute their products internationally are listed; other manufacturers may produce some products (eg, DTP/IPV) for local or regional use. Some products represent
components derived from, or joint efforts of, more than one manufacturer; in such cases, their principal distributor is shown.
†
No discrimination is made between products distributed in combined form and those distributed in separate containers, for combination at the time of use.
‡
Licensed for the fourth (booster) dose only.
 Combination vaccines 40 983

Characteristics of Combination Vaccines Presently Available in Canada, Europe, or the United States
Studies Comparing Combined or Simultaneous Administration of DTwP/IPV, DTwP/Hib, and DTwP/IPV/HIB Vaccines for Primary
Immunization of Infants

AGG, pertussis agglutinins; D, diphtheria toxin; DCS, dual-chamber syringe; DTwP, diphtheria and tetanus toxoids and whole cell pertussis vaccine; DTwPc, DTwP
produced by Connaught Laboratories (Canada); DTwPe, DTwP produced by Wellcome Laboratories (England); DTwPf, DTwP produced by Sanofi Pasteur (France);
DTwPg, DTwP produced by Chiron Vaccines, Marburg, Germany; DTwPu, DTwP produced by Connaught Laboratories (US); DTwPv, DTwP produced by Wyeth
Laboratories; EU, ELISA units; FHA, filamentous hemagglutinin; IPV, inactivated poliovirus vaccine; OPV, oral poliovirus vaccine; PRP, polyribosylribitol phosphate; PRP-
D, PRP-diphtheria toxoid conjugate vaccine; PRP-HbOC, PRP-diphtheria CRM197 protein conjugate vaccine; PRP-OMP, PRP-meningococcal outer membrane protein
conjugate vaccine; PRP-T1 (ActHIB, AvP) and PRP-T2 (Hiberix, GSK), PRP-tetanus toxoid protein conjugate vaccine (lyophilized and reconstituted at time of use with diluent
or DTwP, unless indicated as DCS); PT, pertussis toxin; T, tetanus toxin; UK, United Kingdom; US, United States.
*A ratio less than 1 indicates that mean antibody levels were lower with the combined vaccine than with separate injections; a ratio higher than one, that levels were
higher with combined than separate injections. A blank cell indicates that the comparison was not possible or is not available.
†
Difference significant at .05.
‡
value not available. However, the rate of seroconversion (AGG 320) was significantly lower ( .05) in the combined group (79% vs 92%).
§
Agglutinin titers were not determined. However, ratios for antibody to pertactin and fimbrial antigens were 0.55 ( .05) and 0.74 respectively.
¶
Combined vaccine group was compared with UK historical control subjects, who received the same PRP-T on the same schedule, but a different DTwP.
¶
Serologic assays were performed only for the DTwP/PRP-HbOC group (PRP, 8.20 g/mL; D, 0.92 IU/mL; T, 7.52 U/mL; AGG, 110.1/dilution) and were said to be
“comparable to values reported…in other series.”
**Pertactin, 0.56.
990 SECTION TWO

Thus, reduced immunogenicity of Hib when given in combina-
tion with DTwP is most often of no clinical importance. Gold
et al62 demonstrated reduced levels of Hib, tetanus, and some
pertussis antibodies following DTwP/IPV/Hib compared with
separate vaccines, but those results may have been specific to
that combination.
Surveillance data provide further reassurance that use of the
combined DTwP//PRP-T does not reduce efficacy in comparison
with DTwP and PRP-T administered separately. In Chile, sur-
veillance for pertussis in matched areas that used DTwP alone
or DTwP//PRP-T found no significant difference in the rates of
pertussis in the two areas.80 In the area using DTwP//PRP-T,
efficacy against invasive Hib disease was more than 90%.
Surveillance in Canada found a continued low rate of invasive
Hib disease after the licensure and widespread use of DTwP/
PRP-T in that country, with no change in the extremely low rates
of vaccine failure.81,82 Similarly, there was no increase in inva-
sive Hib disease in the United States after the 1993 licensure
of DTwP//PRP-T and a similar product, DTwP/HbOC (PRP-
type b oligosaccharide conjugate; Tetramune,
Wyeth Lederle Vaccines & Pediatrics).83
None of the combinations that include DTwP and any con-
jugate Hib vaccine has been shown to be associated with mate-
rially increased adverse reactions compared with DTwP alone.
Typically, injection site reactions were slightly greater with the
combinations than with the DTwP alone, but they were less
than the aggregate of local reactions seen when separate vac-
cines were given at separate injection sites.
DTwP/HepB (with or without Hib)
A number of studies84–104 have evaluated combination vaccines
that incorporate DTwP, HepB, and, more recently, Hib com-
ponents (Table 40-4).84–92 In general, the addition of HepB to
DTwP resulted in significantly increased mean HepB antibody
levels and unchanged DTwP responses; the further addition of
Hib to the combination resulted in no consistent changes in
antibody responses.
DTwP/Hep B with or without Hib has been introduced in
an increasing number of countries, many of them developing
countries that use the Expanded Programme on Immunization
6-10-14 week schedule for their public vaccination program.

Studies Comparing Combined or Simultaneous Administration of Vaccines Containing DTwP and Hepatitis B Components, With or
Without Hib Components, for Primary Immunization of Infants

AGG, pertussis agglutinins; D, diphtheria toxin; DTwP, diphtheria and tetanus toxoids and whole cell pertussis vaccine; GSK, GlaxoSmithKline; HBs, hepatitis B
surface antigen; HepB, hepatitis B vaccine; MenAC, meningococcal polysaccharide serotypes A and C conjugated to tetanus toxoid; OMPC, outer membrane
protein complex of ; PRP, polyribosylribitol phosphate; PRP-T, PRP-tetanus toxoid protein conjugate vaccine; PT, pertussis toxin; T, tetanus toxin;
WBP, whole (a mixture of serotypes 1, 2, and 3 used in a solid-phase immunoassay).
*A ratio less than 1 indicates that mean antibody levels were lower with the combined vaccine than with separate injections; a ratio higher than 1, that levels were
higher with combined than separate injections. A blank cell indicates that the comparison was not possible or is not available.
†
DTwP/HepB: Tritanrix-HepB, GSK; 30 IU D, 60 IU T, 10 g HBs.
‡
DTwP/HepB/PRP-T: Tritanrix-HepB/Hiberix, GSK; 30 IU D, 60 IU T, 10 g HBs, 10 g PRP-T.
§
OmniHIB, GSK-distributed version of ActHIB.
¶
Difference significant at . 05.
¶
PRP-T: Hiberix, GlaxoSmithKline; 10 g PRP-T.
#
Including only subjects seronegative at birth.
**
DTwPm/HepB/PRP-OMP: Pentavax, Merck & Co; 30 Lf D, 6 Lf T, 5 g HBs, 7.5 g PRP conjugated to 125 g OMPC (liquid).
††
DTwPm/HepB: Quadrivax, Merck & Co; 30 Lf D, 6 Lf T, 5 g HBs. PRP-OMP: PedvaxHIB, Merck & Co: 7.5 g PRP conjugated to 125 g OMPC (liquid).
‡‡
HepB/PRO-OMP: Comvax, Merck & Co; 5 g HBs, 7.5 g PRP conjugated to 125 g OMPC (liquid). DTwP: CSL Ltd: 30 Lf D, 6 Lf T.
§§
DTwP: CSL Ltd, 30 Lf D, 6 Lf T. HepB: Merck & Co, 5 g HBs. PRP-OMP: Merck & Co, 15 g PRP conjugated to 250 g OMPC (lyophilized).
¶¶
DTwP/HepB: formulated by GSK using D and T manufactured at its facility in Hungary and wP manufactured by CSL Ltd.
##
DTwP/HepB/PRP-T: formulated by GSK using D and T manufactured at their facility in Hungary, wP manufactured by CSL Ltd, combined with Hiberix.
***
A new GSK formulation containing 10 g of PRP-T.
†††
DTwP/HepB/Hib: Quinvaxem, Crucell; 30 IU DT, 60 IU TT, 4 IU inactivated , 10 g HBs, and 10 g PRP conjugated to CRM197.
‡‡‡
DTwP: Serum Institute of India; 30 IU DT, 60 IU TT, and 4 IU .
§§§
PRP-T: Vaxem-Hib, Novartis Vaccines and Diagnostics; 10 g PRP conjugated to CRM197.

Aspinall et al100 have presented lot-consistency results for
Crucell's fully liquid DTwP/HepB/Hib vaccine (Quinvaxem,
developed with Novartis), which contains 10 g hepatitis B sur-
face antigen and 10 g PRP-CRM197. The 1-month postimmuni-
zation GMTs were 7.9 g/mL for Hib and 108 mIU/mL for HepB;
91% of subjects had HepB titers of 10 mIU/mL or more.100 Kanra
et al101 compared this combination with separately administered
DTwP/Hib (Quattvaxem) and HepB (Hepavax-Gene) vaccines
from the same manufacturer and found comparable adverse
event and seroconversion rates in the two groups. Compared
with separate injections, the combination elicited GMTs that
were significantly higher for HepB and significantly lower for
Hib, tetanus, and pertussis. Suarez et al102 evaluated use of the
pentavalent combination vs separate DTwP and Hib vaccines to
boost toddlers primed with DTwP/HepB//Hib vaccine and found
no material differences in antibody responses for the two groups.
Shantha Biotechnics has developed a fully liquid DTwP/HepB/
Hib vaccine (Shan5), for which it reports D, T, P, HepB, and
Hib seroconversion rates of 99.4%, 99.4%, 89.9%, 97.8%, and
98.3% respectively.103
Combinations based on acellular pertussis
vaccine
Overview
The development of numerous effective aP vaccines (see
Chapter 23) and their licensure in combination with diphtheria
and tetanus toxoids (DTaP) represented an important advance
that quickly stimulated efforts to combine DTaP with other rou-
tine vaccines of infancy, such as Hib, IPV, and HepB. Building
on the experience with DTwP combination vaccines, efforts
turned first to evaluating combinations of DTaP and conjugate
Hib vaccines, in light of their similar schedules, universal use
in developed countries, and lack of orally administered alterna-
tives. It was soon found that combining DTaP with Hib tended
to reduce, often markedly, the Hib antibody response. This dis-
covery slowed development of combinations based on DTaP/
Hib and stimulated development of alternative combinations
such as DTaP/IPV, DTaP/HepB, and DTaP/IPV/HepB. It also
prompted research into the clinical relevance of the reduced
response, which has resulted in Hib-containing pentavalent (eg,
DTaP/IPV/Hib) and hexavalent (eg, DTaP/IPV/Hib/HepB) com-
binations becoming accepted in Europe and some other jurisdic-
tions despite reduced Hib responses. In contrast, attention in
North America has focused on DTaP/IPV/HepB and on DTaP/
Hib-based combinations built on the Canadian-manufactured
DTaP5 (see Chapter 23 for details of nomenclature), which
apparently does not interfere materially with Hib responses.
The DTaP-based combinations most widely available world-
wide are produced by SP and GSK. SP markets products based
on the French DTaP2 (eg, Tetravac, Tetraxim) in Europe and
elsewhere and products based on the Canadian DTaP5 (eg,
Quadracel, Pentacel, Pediacel) in the Western Hemisphere,
Asia, and elsewhere. In the United States, a DTaP2/Hib combi-
nation (TriHIBit) based on the US-Japanese DTaP2 is marketed
for use as a fourth, booster dose. GSK markets a full range of
Infanrix (DTaP3)-based combinations worldwide, including the
only currently available hexavalent and various pentavalent and
quadrivalent combinations.
Adding IPV to DTaP, Tdap, or DTaP/Hib
The combining of IPV with DTaP, DTaP/Hib, or Tdap (adolescent-
adult formulation tetanus, diphtheria, and acellular pertussis
Combination vaccines 40 991
vaccine) has no consistent effect on antibody responses to the
included components, with few comparisons achieving statisti-
cal significance. Surveillance in Sweden conducted over many
years showed continued reductions in pertussis incidence
among the Swedish population, concomitant with the transi-
tion from DTaP to DTaP/IPV and DTaP/IPV/Hib.105
Although Tables 40-5 through 40-9 attempt to separate data
regarding DTaP/IPV, DTaP/Hib, DTaP/HepB, DTaP/HepB/Hib,
and DTaP/IPV/Hib combinations, in fact, most of the relevant
studies have evaluated several of these combinations simulta-
neously. In the few that have looked only at DTaP/IPV, pertussis
responses tend to be maintained or enhanced with the combi-
nation106–117 and poliovirus responses vary inconsistently; few of
these variations achieved statistical significance.
A number of key issues were illustrated by one of the first
published studies comparing combined vs separate administra-
tion of DTaP, IPV, and Hib. Finnish infants were immunized at
2 months with DTaP3 and then at 4 and 6 months with DTaP3,
IPV, and PRP-T given all separately, all combined, with the
DTaP3 and IPV combined, or with the DTaP3 and PRP-T com-
bined.106 As shown in Table 40-5,106,107 PRP antibody responses
were markedly reduced among infants receiving PRP-T in com-
bination (whether or not IPV was included in the combina-
tion), but not if the PRP-T was given separately (whether or not
the IPV was given separately). Pertussis responses varied little
(Table 40-6); poliovirus responses were reduced with the combi-
nation. The extent to which any of these results was due to the
nontraditional schedule, which included only two Hib doses,
is unclear. In a follow-up study, available participants were
boosted with DTaP3 and PRP-T at 24 months of age. Vaccines
were given separately to children who had been primed with
separate vaccines; children who had been primed with com-
bination vaccines were randomized to be given a booster with
separate or combined vaccines.107 Despite the large difference
in PRP antibody levels at 7 months of age, there was little differ-
ence in levels before the booster dose at 24 months of age. After
receiving a booster dose, all groups showed strong responses,
which were about twice as high among children primed with
separate vaccines. The groups primed with combined vaccine
had roughly equal responses to the booster dose, whether they
were given combined or separate vaccines as booster doses.
Antibody to PRP Among 120 Infants Given DTaP at 2
Months; Then DTaP, IPV, and Conjugate Hib Vaccine, Separately or
Together, at 4 and 6 Months106,107
DTaP3, diphtheria and tetanus toxoids and acellular pertussis vaccine
(Infanrix, GlaxoSmithKline); GMC, geometric mean concentration of
antibody; Hib, type b; IPV, inactivated poliovirus
vaccine; No., number of subjects providing serum samples for assay; PRP,
polyribosylribitol phosphate; PRP-T2, PRP-tetanus toxoid protein conjugate
vaccine (Hiberix, GlaxoSmithKline).
 992
Studies Evaluating Combined or Simultaneous Administration of DTaP and IPV Vaccines

SECTION TWO Licensed vaccines

Note: In the studies incorporating Hib, the Hib administration was not different between study groups (ie, Hib was given separately to all subjects or was part of the combination for all subjects).
aP, acellular pertussis vaccine; D, diphtheria toxin; DTaP, diphtheria and tetanus toxoids and acellular pertussis vaccine; FHA, filamentous hemagglutinin; FIM, fimbriae; GMC, geometric mean concentration of antibody; Hib,
type b; IPV, inactivated poliovirus vaccine; PRN, pertactin; PRP, polyribosylribitol phosphate; PRP-T, PRP-tetanus toxoid protein conjugate vaccine; PT, pertussis toxin; T, tetanus toxin.
*A ratio less than 1 indicates that mean antibody levels were lower with the combined vaccine than with separate injections; a ratio higher than 1, that levels were higher with combined than separate injections. A blank cell indicates
that the comparison was not possible or is not available.
†
aP2 = French 2-component aP (eg, Triavax or similar); aP3 = Infanrix; aP5 = Tripacel (see Chapter 21 for details of vaccines); PRP-T1 = ActHib (Sanofi Pasteur); PRP-T2 = Hiberix (GlaxoSmithKline); Tdap5 = Covaxis (Adacel)
(Sanofi Pasteur); Tdap5/IPV = Repevax (Sanofi Pasteur).
‡
Difference significant at .05.
§
DTaP and IPV were given at 2, 4, and 6 months; PRP-T was given at 3, 5, and 7 months. Polio antibody levels estimated from figures.
Only DTaP was given at 2 months, without IPV or PRP-T.

Studies Evaluating Combined or Simultaneous Administration of DTaP3 and HepB Vaccines.
D, diphtheria; DTaP3, GSK diphtheria and tetanus toxoids and acellular pertussis vaccine; EU, ELISA units; FHA, filamentous hemagglutinin; HepB, hepatitis B, PRN,
pertactin; PT, pertussis toxin; T, tetanus toxin.
*Units: HepB, mIU/mL; pertussis components, EU; diphtheria and tetanus, IU. A blank cell indicates that data are not available.
†
All vaccines produced by GlaxoSmithKline.
‡
Difference significant at .05.
A large randomized trial in Korea recently evaluated DTaP2/
IPV (Tetraxim) vs separate DTaP (supplied by Biken) and
IPV (supplied by SP) vaccines, given at 2, 4, and 6 months.108
Antibody responses to the polio and filamentous hemagglutinin
(FHA) components were significantly higher with the combina-
tion; adverse events did not materially differ.
In addition to the DTaP/IPV combinations presented in
Table 40-6,108–118 the aP component of Baxter's Certiva (a DTaP1
with a pertussis component that consists solely of pertussis
toxin; see Chapter 23) has been combined with DT and IPV
from Statens Seruminstitut (SSI) as DiTeKiPol.119–121 A compar-
ison of this DTaP1/IPV at 3, 5, and 12 months of age vs DT/
IPV at 5, 6, and 15 months plus wP at 5 weeks, 9 weeks, and
10 months found that the combination achieved protective
antibody levels for diphtheria, tetanus, and the polio serotypes.
Adverse reactions were similar for the DTaP1/IPV and DT/IPV
groups.
Although it is generally true that vaccines given simulta-
neously at separate injection sites do not interfere with each
other, such interactions are occasionally noted. Among sub-
jects given DTaP2//PRP-T (TriHIBit) along with IPV, sequential
IPV-IPV-OPV, or OPV at 2, 4, and 6 months, Rennels et al122
found that the GMTs (1.2, 1.3, and 3.1 g/mL, respectively) and
the proportions achieving PRP antibody responses of 1.0 g/mL
or greater (54%, 55%, and 79%, respectively) were reduced by
coadministration of IPV. In contrast, Daum et al123 found that
PRP antibody responses to the same DTaP2/PRP-T combination
did not significantly differ among groups coadministered OPV
or IPV at 2 and 4 months (GMTs, 4.0 vs 3.4 g/mL; proportions
= 1.0 g/mL, 77% vs 74%). The difference in results of these
two studies remains unexplained.
Combination vaccines 40 993
Tdap3/IPV and Tdap5/IPV combination vaccines (Boostrix-
IPV, GSK; Repevax, SP MSD) have been licensed for use as a
booster among previously primed persons (the lower limit of
licensure age varies by regulatory authority, from 3 to 10 years).
The combinations produce antibody responses that are similar
to those of separately administered vaccines and reaction rates
that do not materially differ between separate and combined
administration.116–118,124–129 SSI also produces a Tdap1/IPV com-
bination vaccine, DiTeKiPol Booster.
Theeten et al130 evaluated three doses of Tdap3, one dose of
Tdap/IPV followed by two doses of Tdap3, and three doses of
Td, administered on a 0-1-6 month schedule to adults 40 years
or older with no known T or d vaccination for at least 20 years.
Rates of adverse reactions and responses to the T and d com-
ponents did not differ among the groups; pertussis and polio
immune responses were as expected.130
Langley et al131 compared Tdap5 (Adacel) with DTaP/IPV
(Quadracel) as a fifth-dose booster for children 4 to 6 years old
and found that the Tdap group experienced fewer adverse reac-
tions but had comparable diphtheria and tetanus seroprotection
rates to the DTaP/IPV group.131 The GMTs were similar, except
for diphtheria, which, as expected, was higher in the DTaP group.
Adding HepB to DTaP or to DTaP/IPV
Combining HepB with DTaP or with DTaP/IPV generally pro-
duces somewhat higher DTaP and polio antibody responses
than are achieved with the same components given separately
on the same schedule. However, the HepB responses following
994 SECTION TWO

Studies Evaluating Combined or Simultaneous Administration of DTaP and Hib Vaccines

aP, acellular pertussis vaccine; D, diphtheria toxin; DTaP, diphtheria and tetanus toxoids and acellular pertussis vaccine; FHA, filamentous hemagglutinin; FIM,
fimbriae; GMC, geometric mean concentration of antibody; HbOC, PRP-diphtheria CRM197 protein conjugate vaccine; IPV, inactivated poliovirus vaccine; OPV, oral
poliovirus vaccine; PRN, pertactin; PRP, polyribosylribitol phosphate; PRP-D, PRP-diphtheria toxoid conjugate vaccine; PRP-T, PRP-tetanus toxoid protein conjugate
vaccine; PT, pertussis toxin; T, tetanus toxin; US, United States.
*A ratio less than 1 indicates that mean antibody levels were lower with the combined vaccine than with separate injections; a ratio higher than 1, that levels were
higher with combined than separate injections. A blank cell indicates that the comparison was not possible or is not available.
†
aP2f = French 2-component aP (eg, Triavax or similar); aP2u = Tripedia; aP3 = Infanrix; aP4 = ACEL-IMUNE or similar Takeda-type aP vaccine; aP5 = Tripacel or
equivalent 5-component aP vaccine (see Chapter 21 for details of vaccines). PRP-T1 = ActHib (Sanofi Pasteur); PRP-T2 = Hiberix (GlaxoSmithKline).
‡
Difference significant at .05.

the combinations typically are lower than seen with monova-
lent HepB, not because of interference, but because adminis-
tration schedules for combinations typically are more closely
spaced than are schedules for monovalent HepB. The magni-
tude of HepB antibody responses is directly correlated with the
time between doses and, in particular, the time between the sec-
ond and third doses. Accordingly, HepB responses are lower if
the HepB is administered (whether separately or in a combina-
tion) at, for example, 2, 4, and 6 months or 3, 4, and 5 months
rather than at, for example, 0, 1, and 6 months or 3, 5, and
11 months.
Table 40-7132–137 summarizes studies comparing the perfor-
mance of the GSK DTaP3/HepB vaccine (a building block of the
more advanced combinations currently offered by GSK) with
the separately administered components and also comparing
various administration schedules.132–137 The combination vac-
cine retained the immunogenicity profile of the separate com-
ponents and stimulated antibody concentrations associated
with protection with a variety of schedules.
A comparison of combined vaccine at ages 2, 4, and 6 months
vs a currently recommended schedule in the United States—
HepB at birth and 1 and 6 months of age and DTaP3 at 2, 4, and
6 months of age—found significantly higher antibody responses
for combined vaccine for every component except HepB, which
was significantly lower.135 However, the mean HepB antibody
with combined vaccine was nevertheless high (1,280 mIU/mL),
and 98% of subjects had levels greater than 10 mIU/mL, the
level considered protective.
Studies Evaluating Administration of Pentavalent or Hexavalent DTaP-based Combinations vs Separate Administration of One or More Contained Components

Combination vaccines
40
995
 996
SECTION TWO Licensed vaccines
Studies Evaluating Administration of Pentavalent or Hexavalent DTaP-based Combinations vs Separate Administration of One or More Contained Components—cont'd

DENV types when inoculated in mice and monkeys success-
fully raised neutralizing antibodies. Monkeys resisted challenge
with DENV-1 but not DENV-2.254,255
Exploring the safety of DNA vaccines
Although the DNA approach offers advantages, it also carries
unique risks.256 These include the theoretical risk of nucleic acid
integration into the host's chromosomal DNA to potentially inac-
tivate tumor suppressor genes or to activate oncogenes. This risk
appears to be well below the spontaneous mutation frequency
for mammalian cells.257,258 However, if a mutation resulting from
DNA integration is a part of a multiple hit phenomenon lead-
ing to carcinogenesis, it could be many years before this problem
becomes evident. Another concern is that foreign DNA might
induce anti-DNA antibodies, leading to autoimmune diseases
such as systemic lupus erythematosus. Studies in lupus-prone
mice, normal mice, rabbits, and people have not validated this
concern,259,260 and, in fact, DNA vaccines are being proposed as
an approach to the management of autoimmune diseases.261,262
Access the complete reference list online at http://www.expertconsult.com
3. Guzman MG, Halstead SB, Artsob H, et al. Dengue: a continuing global
threat. Nat Rev Microbiol 2010;8(12 Suppl.):S7–16.
5. Lum LC, Suaya JA, Tan LH, et al. Quality of life of dengue patients. Am J
Trop Med Hyg 2008;78:862–7.
10. Suaya JA, Shepard DS, Siqueira JB, et al. Cost of Dengue cases in eight
countries in the Americas and Asia: a prospective study. Am J Trop Med Hyg
2009;80:846–55.
11. Shepard DS, Coudeville L, Halasa YA, et al. Economic impact of dengue
illness in the Americas. Am J Trop Med Hyg 2011;84:200–7.
23. Halstead SB. Pathogenesis of dengue: challenges to molecular biology.
Science 1988;239:476–81.
Dengue vaccines 44 1051
Conclusions
Dengue virus infections occur regularly among military per-
sonnel assigned to combat or peacekeeping roles in tropical
countries. These infections are consistently among the lead-
ing causes of febrile diseases in tourists and expatriate residents
of tropical countries. Considerable morbidity and mortality is
inflicted on native populations of all social strata in endemic
areas. These medical and social impacts create a significant
market, as evidenced by the interest shown by major vaccine
manufacturers. As described here, several promising dengue
vaccine candidates are in preclinical and clinical development,
and one manufacturer has entered phase 3 testing. If the safety
concerns described can be surmounted, economic forces and
technologic advances should soon bring one or more dengue
vaccines to the market. It remains for the vaccine community
to develop and implement plans for the strategic use of dengue
vaccines by developing evidence-based policies to target high-
risk groups and decrease virus transmission.
29. Rothman AL. T lymphocyte responses to heterologous secondary dengue
virus infections. Ann N Y Acad Sci 2009;1171(Suppl. 1):E36–41.
51. Swaminathan S, Batra G, Khanna N. Dengue vaccines: state of the art.
Expert Opin Ther Pat 2010;20:819–35.
52. Murphy BR, Whitehead SS. Immune response to dengue virus and
prospects for a vaccine. Annu Rev Immunol 2011;29:587–619.
238. Halstead SB, Mahalingam S, Marovich MA, et al. Intrinsic antibody-
dependent enhancement of microbial infection in macrophages: disease
regulation by immune complexes. Lancet Infect Dis 2010;10:712–22.
1078 SECTION THREE Vaccines in development and new vaccine strategies
Current view about immunological correlates of protection against hepatitis C virus (HCV) infection and chronic hepatitis. NK,
natural killer.
homologous, but not heterologous, genotypes. Moreover, it has
been recently reported that previously infected chimpanzees
are not consistently protected against reinfection or persis-
tent infection after reexposure to the identical hepatitis C virus
strain.124
In humans, evidence of cross-protective immunity has been
reported in a prospective study of IVDUs from the United States.
Strikingly, the incidence of persistent viremia in IVDUs who
had recovered from a previous infection was 12-fold lower than
that in IVDUs who had not experienced a previous infection, as
shown using multivariate analyses.125 As seen also in the chim-
panzee, peak viral loads were substantially higher (by almost
2 log) in first-time infections compared with reinfections. In
addition, HIV coinfection produced persistent HCV infection
in all cases, indicating the role of the immune response in HCV
recovery.125 Interestingly, it has been recently reported that the
reduced risk of HCV persistence in IVDUs previously recovered
from HCV infection correlated with T-cell responses, and pro-
longed antigenic stimulation seems to be required to maintain
humoral responses.126
Collectively, these chimpanzee and human data provide
evidence for the existence of immunity to HCV and, impor-
tantly, suggest the existence of cross-protective immunity
within and between commonly occurring HCV genotypes.
It is important to note, however, that not all reinfections
in IVDUs are resolved without progression to chronicity,125
indicating that natural immunity to HCV is not complete
and not as effective as for the hepatitis A and B viruses. It
should also be mentioned that earlier studies in the chimpan-
zee model have concluded a lack of protective immunity to
HCV.127 This apparent contradiction may be due to the earlier
studies measuring immunity in terms of sterilizing immu-
nity, rather than as the ability to prevent the development of
chronic infection.
The current consensus is that the development of an
early, broad, and multispecific T-cell response during acute
viral infection is associated with spontaneous clearance
of infection. Similarly, protection from reinfection corre-
lates with the breadth and magnitude of a multifunctional
CD8+ T-cell response and also with the quality, functional
potency, and cytotoxic potential of HCV-specific CD8+ T
cells responding against diverse HCV core, E1, NS3, NS4,
and NS5 epitopes. It is also increasingly clear that the pres-
ence of neutralizing antibodies against the virus envelope
glycoproteins can be protective and is associated with the
rapid clearance of viremia.
Strategies for an HCV vaccine
The current strategies for the development of HCV vaccines
include approaches aimed at the following: (1) preventing ini-
tial infection (ie, providing sterilizing immunity), (2) preventing
the development of chronic viral persistence that occurs in the
majority of natural infections, and (3) clearing the virus in per-
sons who already have an established chronic HCV infection
(ie, therapeutic vaccines).128 Since the major clinical sequelae
of HCV infection stem from its chronic nature, a well-tolerated
vaccine that enhances spontaneous resolution after the acute
infection or assists antiviral drugs in clearing the virus would
represent an effective means for combating chronic liver dis-
ease. A therapeutic vaccine could also provide an option for
persons infected with drug-resistant HCV.
Several HCV vaccine candidates are currently in the pre-
clinical or early phases of clinical trials. Ideally, immunization
strategies against HCV should elicit an effective priming of
broadly cross-neutralizing, antienvelope antibodies and broad
HCV-specific Th1 CD4 and broad CD8 T-cell responses. This
condition is usually difficult to achieve through the use of sol-
uble proteins combined with conventional adjuvants, which
mainly results in the induction of antibodies and Th responses
to the vaccine antigen(s). However, substantial progress toward
obtaining an effective priming of CTL responses has been made
by using recombinant HCV proteins formulated as immuno-
stimulatory complexes (ISCOMs).129,130 ISCOMs are obtained
by combining the purified antigens with a particulate adjuvant
comprising cholesterol, phospholipid, and naturally occurring
saponins (ISCOMATRIX).131
Various forms of cDNA vaccines are also being explored to
elicit HCV-specific humoral and cellular immune responses to
encoded antigens that, by virtue of being newly synthesized in
the cytosol of transfected cells, can be particularly effective at
priming CD8+ CTLs. DNA vaccines can also include immu-
nostimulatory CpG-containing motifs capable of activating
antigen-presenting dendritic cells leading to stimulation of
innate immune responses (eg, the synthesis of type 1 interferons
and NK cells), as well as adaptive B- and T-cell responses to
vaccine antigens.
 Hepatitis C vaccines 48 1079

Finally, various live attenuated or defective viral vectors
expressing HCV genes are also being investigated. An improved
vaccine immunogenicity can result from more efficient expres-
sion and delivery of HCV antigens, including, in some cases,
the targeting of antigen-presenting cells. The use of various
prime-boost immunization modes and regimens is also being
explored to optimize vaccine immunogenicity and potency.
The possibility of using a killed or attenuated HCV vaccine
has not been investigated in detail owing to the historic inabil-
ity to propagate HCV in vitro and produce commercial levels
of virus. However, with the recent breakthrough in propagat-
ing HCV in cell culture, these avenues can now be formally
explored.
Vaccines based on adjuvanted recombinant
HCV proteins
A vaccine consisting of the recombinant gpE1/gpE2 heterodi-
mer derived from mammalian cells has been tested exten-
sively for efficacy in the chimpanzee model132–136 (Table 48-1).
When combined with oil/water-based adjuvants and used to
vaccinate naïve animals, this vaccine candidate elicited anti-
envelope antibodies and Th-cell responses to gpE1 and gpE2.
When the vaccinated animals were challenged experimentally
with homologous viral inocula, the highest responding ani-
mals (in terms of anti-gpE1/gpE2 antibody titers) were com-
pletely protected against infection.132 Using sensitive reverse
transcription–polymerase chain reaction assays, no viremia was
detected in serum or liver samples at any time after challenge in
these “sterilized” animals. This apparent sterilizing immunity
correlated directly with anti-gpE2 antibody titers that prevent
the binding of gpE2 or the virus itself to CD81.137 Furthermore,
although lower-responding animals became infected, the major-
ity underwent an abortive acute infection that did not result
in the persistently infected carrier state.132,135 Overall, these
data showed that the progression rate to chronic infection was
significantly lower in vaccinees than in unimmunized control
animals.133 These pioneering studies provided the first data sup-
porting the feasibility of a vaccine against HCV because they
indicated the capability of recombinant envelope glycoprotein
immunization to elicit sterilizing immunity or to prevent the
progression to chronic infection following an acute, transient
infection after viral challenge.
This work was subsequently extended to address the key
question of vaccine protection against challenge with a heter-
ologous 1a viral strain. (The vaccine strain, HCV-1, and the
challenge strain, HCV-H, are members of the 1a genotype that
predominates in the United States and occurs worldwide.)
Although none of these animals were sterilized against this
heterologous viral challenge, the majority failed to develop into
persistent carriers of the virus, unlike the situation with control
unvaccinated animals and for human infections133 (Table 48-1).
Based on these preclinical data, a clinical development program
was started aiming at assessing tolerability and immunogenicity
of the gpE1/gpE2 vaccine purified from Chinese hamster ovary
cell lines and mixed with the oil/water M59 adjuvant.138,139
Strong Th responses were elicited in the human volunteers
along with substantial anti–gpE1/gpE2 antibody titers that were
similar to those elicited in vaccinated and protected chimpan-
zees. Low but significant titers of antibodies that block the
binding of recombinant gpE2 to the HCV receptor component
CD81 were also shown to be present in many vaccinees.139
Additional analyses showed that most vaccinees developed anti-
bodies that could neutralize the infectivity of HCVpp bearing
envelope glycoproteins derived from the homologous 1a strain,
HCV-1.138 Further assays are in progress to assess the presence
of neutralizing antibodies against more diverse HCV isolates
and genotypes.133
In a study with a different vaccine candidate,140 a chimpan-
zee was immunized with insect cell–derived recombinant gpE1
and gpE2 derived from HCV strain N2 and a peptide spanning
the E2HVR1 region from HCV strain 6. The two envelope pro-
teins, which lacked the C-terminal transmembrane anchors,
were expressed and purified separately. The E2HVR1 peptide
was conjugated to keyhole limpet hemocyanin, and immuniza-
tions were performed with the Freund adjuvant. The conclu-
sion, derived from a series of immunizations and challenges
in the same animal with HCV strain 6, was that sterilization
against infection was dependent on high anti-E2HVR1 anti-
body titers rather than total anti-gpE2 or anti-gpE1 titers.140,141
The outcome of challenging with other viral strains was not
reported, however. It should also be noted that insect-derived
C-terminally truncated gpE2 has been shown to bind poorly to
the HCV receptor, CD81, compared with gpE2 derived from
mammalian cells.137 The gpE1/gpE2 heterodimer is generally
considered a more native reflection of the HCV virion than
either ectodomain alone.142
Expression of a core-gpE1-gpE2 gene cassette in insect cells
has been reported to result in the generation of 40 to 60 nm
viral-like particles within cytoplasmic cisternae.143 After partial
purification, these viral-like particles appeared to be immuno-
genic in mice144 and baboons145 and may, therefore, be consid-
ered an interesting tool for the design of a vaccine for humans,
provided that sufficient yields and purity could be attained.

Summary of Outcome of Chimpanzee Vaccination/Challenge Studies Performed With Adjuvanted Recombinant gpE1/gpE2 Formulations*

gp, glycoprotein; HCV, hepatitis C virus.

*These studies were performed during the course of many years. Typically, animals were immunized with 30-80 g gpE1/gpE2 in various oil/water adjuvants (MF59
MTP or MF75 MTP or MF59 CpG) at months 0, 1, and 6 approximately, followed by intravenous challenge 2-4 weeks later with 10-100 chimpanzee infectious
doses of HCV-1 or HCV-H77. Both strains belong to the 1a subtype that predominates in North America. Circulating levels of viremia were measured using reverse
transcription–polymerase chain reaction assays for HCV genomic RNA for at least 1 year postchallenge. Qualitative TMA assays were also performed in some animals.
(Reprinted with permission from Houghton M. Prospects for prophylactic and therapeutic vaccines against the hepatitis C viruses. Immunol Rev 239:99-108, 2011.)
†
P values (Fisher exact test) refer to chronic carrier rates between control animals and vaccinees.
‡
Note: In five vaccinees challenged with homologous HCV-1 virus, no viremia could be detected at any time postchallenge in plasma PBMNCs or liver biopsy
samples, and, thus, they were considered to have been sterilized.

recombinant gpE2 glycoprotein was more effective in eliciting
high anti-gpE2 titers than repeated DNA or protein immuniza-
tions.181 The immunogenicity of gpE2 DNA vaccines has been
confirmed in rhesus macaques.173 However, it must be under-
lined that small animals and even rhesus macaques are not the
optimal models to demonstrate the immunogenicity of DNA
vaccines in humans.
In the chimpanzee model, only one small DNA vaccine
study has been reported so far.182 To optimize immunogenicity,
the recombinant construct was designed by fusing the ectodo-
main of gpE2 (aa 384-715) to the C-terminal, transmembrane
region of CD4, to favor sequestration of the encoded gpE2 glyco-
protein to the outer cell surface. DNA was administered using
a bioinjector into the quadriceps at weeks 0, 9, and 23, followed
by experimental challenge with homologous monoclonal virus,
3 weeks later. Importantly, both vaccinees resolved their acute
infections quickly, whereas the control unvaccinated animal
became chronically infected following viral challenge. Although
humoral and cellular immune responses to the vaccine were
observed in only one animal of two, both vaccinated animals
displayed lower viral titers following challenge and developed
hepatitis earlier compared with control animals, as a result of
the primed immunity.182
Other DNA vaccines targeting different HCV gene products
have been demonstrated to be capable of priming specific CD4+
and CD8+ T-cell responses in small animal models. Examples
include a DNA vaccine encoding NS3, NS4, and NS5 that is
able to prime, in immunized mice, broad and specific antibody
responses, CD4+ Th and CD8+ CTL responses, also conferred
protection against a challenge with a syngeneic, SP2/0 myeloma
cell line expressing NS5 proteins.183 In this immunization
model, the coexpression of the GM-CSF gene has been shown
to augment the cellular immune responses directed against
HCV NS gene products.184
DNA vaccine approaches based on the HCV core gene have
also been investigated in preclinical species. As mentioned,
the core gene encodes the most conserved HCV protein that
contains relevant epitopes for T-cell responses. This strategy
has been effective in eliciting specific immune responses in
mice,185–188 and the coadministration of interleukin-2 or GM-CSF
was shown to augment the immunizing capacity of the recom-
binant DNA construct.185 Conversely, coimmunization with a
plasmid expressing interleukin-4 resulted in the induction of
a Th0 phenotype and a concomitant suppression of C-specific
CTLs.185 Experimental studies have shown that long-term
expression of the C gene can exert multiple pathogenic effects in
transgenic mouse models, including steatosis and hepatocellular
carcinoma,189,190 and this can also copromote cellular transforma-
tion in vitro.191 Therefore, it may be prudent to avoid the inclu-
sion of the C gene in a potential HCV DNA vaccine. In contrast,
periodic vaccination using a recombinant C polypeptide subunit
vaccine is unlikely to raise the same level of safety issues.
As of today, two DNA vaccine candidates have entered
human experimentation, both limited to a therapeutic rather
than a prophylactic setting. The first candidate, ChronVac-C,
is based on a codon optimized HCV nonstructural NS3/4A
DNA gene expressed under the control of the cytomegalovi-
rus immediate early promoter. The vaccine has now entered a
phase 1/2a clinical trial in HCV genotype 1–infected patients
to assess safety and immunogenicity.192 Interestingly, follow-
ing dosing of the vaccine by intramuscular electroporation, two
patients had moderate viral load reductions and development of
HCV-specific T-cell responses that coincided with the time of
the viral load reductions.192 Based on these preliminary results,
ChronVac-C will be further developed as an add-on to Peg-IFN
and ribavirin for the treatment of chronic hepatitis C.
A different group vaccinated HCV chronically infected
patients in a phase 1 study with a therapeutic vaccine (CIGB-
230) containing a combination of a DNA plasmid expressing
Hepatitis C vaccines 48 1081
HCV structural antigens and recombinant HCV core protein.
The patients, all nonresponders to previous antiviral treat-
ment, received 6 doses of vaccine 1 by intramuscular injection
at 4 week intervals. Neutralizing antibody against heterologous
viral pseudoparticles and HCV core–specific T-cell responses
developed in the majority of patients. Despite persistent vire-
mia, an improvement in liver histologic findings, with a reduc-
tion in fibrosis, was reported for nearly half of the vaccinated
patients,193 making the observed result difficult to interpret.
Finally, it is important to mention that vaccines based on
defective RNA have been shown to be very effective in elicit-
ing protection against flaviviral infections.194 In particular, they
confer good protective immunity in the absence of integration
into the host genome and may represent a promising approach
for the design of vaccines against HCV.
HCV vaccines delivered by viral vectors
The use of a defective or attenuated viral or bacterial vec-
tor to deliver vaccines can offer improved immunogenicity
as a result of a wide tropism of the vector, including that of
antigen-presenting cells, as well as being able to stimulate innate
immune responses that, in turn, stimulate adaptive immune
responses to the encoded vaccine antigens. Furthermore, the
use of a vector that is already used as a vaccine itself offers a
potential advantage with respect to safety, manufacturing, and
distribution issues. Finally, many vectors allow the insertion of
multiple genes, thus facilitating the induction of a broad, cross-
protective immune response, particularly useful against hetero-
geneous agents such as HCV.
One approach has been the use of an attenuated rabies viral
vector into which the HCV gpE1-gpE2-p7 gene cassette was
inserted or only the ectodomain of gpE2 linked to the CD4,
C-terminal transmembrane region, and cytoplasmic domain.
Virions expressing gpE1-gpE2-p7 were immunogenic in mice
eliciting CTL responses to gpE2.195 Similarly, defective Semliki
Forest virions containing the HCV NS3 gene produced long-last-
ing NS3-specific CTLs after one immunization in mice trans-
genic for human HLA-A2.1.196 As observed in HCV-infected
patients, the immune response was directed to one immuno-
dominant epitope within NS3. Defective, recombinant adenovi-
ruses expressing the HCV C–gpE1–gpE2 gene cassette have also
been shown to prime HCV-specific CTLs in mice immunized
intramuscularly, although the induction of anti–gpE1/gpE2 anti-
bodies required further immunization with purified gpE1/gpE2
glycoproteins.197 Replication-defective adenoviruses expressing
C and gpE1 also primed long-lasting, specific CTL responses in
mice.198 Recombinant canarypox viruses, expressing an HCV
gene cassette containing C-gpE1-gpE2-p7-NS2-NS3, elicited
HCV-specific humoral and cellular immune responses in mice,
although the optimum immunization regimen required first
priming with a plasmid DNA expressing the HCV genes before
boosting with the recombinant canarypox virus.199
A recent interesting approach has been the use of recom-
binant HCV polypeptides combined with various Th1-type
adjuvants and replication-defective alphaviral particles encod-
ing HCV proteins.200 In this study, mice were immunized with
defective chimeric alphaviral particles, derived from the Sindbis
and Venezuelan equine encephalitis viruses encoding the HCV
envelope glycoprotein gpE1/gpE2 heterodimer (E1E2) or non-
structural proteins 3, 4, and 5 (NS345), and strong CD8+ T-cell
responses but low CD4+ Th responses to these HCV gene
products were detected. In contrast, recombinant E1E2 glyco-
proteins adjuvanted with MF59 containing a CpG oligonucle-
otide elicited strong CD4+ Th responses but no CD8+ T-cell
responses. A recombinant NS345 polyprotein also stimulated
strong CD4+ Th responses but no CD8+ T-cell responses when
adjuvanted with ISCOMATRIX containing CpG. Optimal elici-
tation of broad CD4+ and CD8+ T-cell responses to E1E2 and
1082 SECTION THREE Vaccines in development and new vaccine strategies
NS345 was obtained by first priming with Th1-adjuvanted pro-
teins and then boosting with chimeric, defective alphaviruses
expressing these HCV genes. In addition, this prime-boost regi-
men resulted in the induction of anti-E1E2 antibodies capable
of cross-neutralizing heterologous HCV isolates in vitro. This
vaccine formulation and regimen may, therefore, represent
a promising development candidate in humans because it to
recapitulates all of the critical cellular and humoral immune
responses of an ideal vaccine regimen.
Chimpanzee studies have been performed in which naïve
animals were first immunized twice with attenuated adeno-
viruses (Ad5 with a deletion of the E1 gene) expressing the
HCV NS3, 4, and 5 genes from an HCV genotype 1b isolate.201
Subsequently, boosting of immune responses was achieved by
repeated electroporation of a naked DNA vaccine encoding the
same HCV NS genes. This regimen elicited HCV-specific and
multispecific CD4+ and CD8+ T cells in four of five vaccin-
ees. Following a delayed experimental challenge with a heter-
ologous 1a strain, the four animals showed a clear amelioration
of acute viremia and hepatitis followed by eradication of vire-
mia. In contrast, the weak responder to the vaccine exhibited
viremia and hepatitis similar to that shown in control unvac-
cinated animals and this vaccinee became a chronic carrier of
the virus, as in the case of four of five control unvaccinated
animals challenged with the same 1a viral strain (Figure 48-4).
This convincing study supports the concept of developing a
“T-cell vaccine” against HCV, although in view of the efficacy
of a vaccine comprising recombinant gpE1 and gpE2 glycopro-
teins in the chimpanzee challenge model reported above, the
ideal HCV vaccine may be one that elicits cross-neutralizing
anti–gpE1/gpE2 antibodies and broad, cross-protective cel-
lular immune responses to these gene products and those of
the HCV NS genes. Replication-defective adenovirus 6 and a
replication-defective chimpanzee adenovirus 3 expressing the
HCV NS 3, 4, and 5 gene cassette are currently being assessed
in phase 1 clinical trials exploring different prime-boost immu-
nization regimens in prophylactic and therapeutic settings.133
An additional trial is in progress exploring prime-boost regi-
mens using the chimpanzee adenovirus 3 vector and the highly
attenuated modified vaccinia Ankara (MVA) vector expressing
the same HCV gene cassette.133
Another group has used a replicating vaccinia virus express-
ing HCV structural and nonstructural genes to immunize four
naïve chimpanzees. Following experimental challenge with
homologous virus, all four vaccinees resolved their acute infec-
tions without progressing to the carrier state, whereas two
control animals developed chronic infection after the same
challenge,202 providing further support for the ability of prophy-
lactic vaccination to prevent the development of the chronic
carrier state.
In a different study,161,203 chimpanzees were vaccinated with
DNA plasmids encoding HCV core-E1-E2 and NS3 and subse-
quently boosted with recombinant MVA encoding core-E1-E2
and NS3 genes. This DNA-primed, MVA-boosted immuniza-
tion induced strong Th1- and Th2-cytokine responses together
with strong HCV-specific CD8+ T-cell response and high HCV-
specific antibody titers. A subsequent challenge with a heterolo-
gous virus demonstrated that the DNA-primed, MVA-boosted
immunization was associated with control of HCV viral levels
in the acute stage. However, despite control of HCV in plasma
and liver in the acute period, three of four vaccinated animals
developed persistent infection.
Last, TG4040 is a recombinant polyantigenic T-cell vaccine
based on an MVA that is advancing in clinical trials. TG4040
encodes the HCV NS3, NS4, and NS5B proteins.204 It was
originally tested in HLA31 class I transgenic mice and shown
to produce strong, long-lasting, cross-reactive, HCV-specific
CD8+ and CD4+ T cells. These responses could be readily
boosted by an additional dose of vaccine given after 6 months,
indicating that the vaccination protocol was able to induce
effective memory T-cell responses.204 The safety and biologi-
cal activity of TG4040 as a candidate therapeutic vaccine has
been evaluated in a phase 1 study in 15 treatment-naïve HCV
subjects.205 Of 15 patients, 6 received three weekly injections of
vaccine, and the remaining patients received a fourth injection
at 6 months. HCV-specific T-cell responses were detected in all
patients as early as 1 week after the first vaccination and were
maintained during the 6 month follow-up. Vaccination reduced
HCV viral loads by more than 1 log10, and the strongest vac-
cine-specific T-cell responses were observed in patients who
achieved the greatest viral load reductions.
Future directions
A few years ago, prospects for effective vaccination against
HCV were considered remote because of the high propen-
sity of this virus to promote chronic persistent infections,
evidence that convalescent humans and chimpanzees could
be readily reinfected following reexposure, and the consid-
erable genetic heterogeneity of this virus. Today, we can be
more optimistic for several reasons. First, we now know
that the spontaneous eradication of virus occurs in a signifi-
cant fraction of acute infections and is associated with spe-
cific immune responses to the virus. Recapitulation of such
immune responses by appropriate vaccination, therefore,
becomes a realistic option. Second, clear evidence for some
natural immunity has emerged in humans and chimpan-
zees. These studies have shown that convalescent humans
and chimpanzees are protected from chronic infection against
reexposure to virus in the majority of cases, even against
divergent viral strains.
Current approaches to HCV vaccination include the use
of recombinant envelope proteins to elicit neutralizing anti-
bodies and CD4+ T cells and various defective or attenuated
viral vectors to enhance priming of humoral and cellular (CD4
and CD8) immune responses to multiple HCV gene prod-
ucts expressed by the vector. It is likely that relevant strat-
egies for developing a vaccine will involve eliciting broad
humoral (anti–gpE1/gpE2) and cellular immune responses.
Several prophylactic or therapeutic vaccine candidates have
reached human clinical trials (Table 48-2). In the future, it
will be important to better determine correlates of protec-
tion, memory of vaccine-induced protection, and the ability to
cross-protect against diverse genotypes and to identify optimal
vaccine formulations.
Considering that HCV is the first cause of primary liver
cancer, if a prophylactic vaccine is successfully developed, an
important cause of global morbidity and mortality will be
controlled. Even in countries with a relatively low incidence
of infection, a prophylactic vaccine will be cost-effective
when used in the general population. Some issues surround-
ing clinical development of a prophylactic HCV vaccine,
however, remain to be resolved. Because of the relatively
low incidence of new infections in the developed world, it
is not trivial to identify the appropriate at-risk population
to enroll in an efficacy trial for a preventive HCV vaccine.
Moreover, the design of an efficacy trial to measure preven-
tion of chronic infection is not straightforward. As discussed,
a vaccine that allowed only a “transient infection”, while pre-
venting the development of chronic HCV infection, would
be as beneficial as one that provided sterilizing immunity.
Indeed, vaccine efficacy data from the chimpanzee challenge
model indicate that it is possible to prevent the progression
to chronic infection in vaccinees. However, this end point, in
the absence of robust correlates of immunity, makes it very
complicated to organize efficacy clinical trials. In acute HCV
infection, early treatment with Peg-IFN leads to high virological
1086 SECTION THREE Vaccines in development and new vaccine strategies
Diagnosis
Hepatitis E is diagnosed by detecting viral RNA (reverse tran-
scription–polymerase chain reaction [PCR]) in the serum and/
or feces during the incubation period or early acute phase of
disease or, more commonly, by demonstrating anti-HEV of
the IgM class or a rising titer of anti-HEV of the IgG class in
the serum during the late acute phase or convalescent phase
of the illness.36 Commercial enzyme-linked immunosorbent
assays for detecting such antibody have been developed but are
not licensed in the United States.
Treatment
Patients with acute hepatitis E should be treated symptom-
atically. Immunosuppressed patients chronically infected
with HEV have been successfully treated with interferon and
ribavirin.14,37,38
Epidemiology
Incidence and prevalence
Clinical hepatitis E has been reported with increasing frequency
in industrialized countries of North America and Europe, prob-
ably as the result of better and more widely used diagnostic
tests. Nevertheless, the incidence of hepatitis E in developing
countries of Asia, the Middle East, and North Africa remains
much higher (Figure 49-1). In contrast with anti-HAV preva-
lence, there is little anti-HEV in infants and young children in
developing countries and the peak prevalence of anti-HEV and
the peak incidence of clinical disease both occur among older
children and young adults.39 With few exceptions, the preva-
lence of anti-HEV seldom exceeds 40%.39–41 Surprisingly, the
prevalence of anti-HEV in industrialized countries may exceed
20%, even though very little clinical disease is observed.42–44
This may represent zoonotic transmission of the virus (see
subsequent text).
Risk groups
Seronegative children and adults are at general risk of disease,
and pregnant women are at special risk for severe or fulminant
hepatitis E in endemic regions. Recipients of blood in such
regions may also be at increased risk.45–47 People with possible
increased zoonotic exposure (exposure to animal vectors and
consumption of undercooked pork or wild game) may be at
special risk in industrialized countries;42,48 most such hepatitis
E cases occur in elderly people and HIV-infected patients; organ
recipients are at special risk of chronic HEV infections.
Modes of transmission and reservoirs of infection
In endemic countries, genotypes 1 and 2 predominate and con-
taminated water is the major source of infection. In North
America, South America, and Europe, genotypes 3 (also geno-
type 4 in China and Japan) predominate, suggesting that zoo-
notic spread of HEV is the principal mode of transmission in
these regions.23 Epidemiologic and direct evidence for transmis-
sion of genotype 3 HEV from swine and Sika deer has been
reported (reviewed by Takahashi et al49). Because human HEV
genotypes 1 and 2 may be more virulent than the zoonoti-
cally associated genotypes 3 and 4, virulence differences may
partially explain the difference in incidence and age at onset
of clinical hepatitis E between industrialized and developing
regions of the world.23

Significance as a public health problem
Although hepatitis E is much more widely distributed than
previously thought, significant disease is restricted to certain trop-
ical and subtropical regions of the developing world. At present,
the health burden imposed by HEV on industrialized countries is
being reassessed as more cases are being recognized.50,51
Passive immunization
Preexisting anti-HEV is associated with protection against
subsequent exposure to the virus.52 However, normal immune
globulin prepared in endemic countries failed to protect,
probably because of the relatively low titers of anti-HEV
(reviewed by Purcell and Emerson53). Direct demonstration of
the protective efficacy of anti-HEV has come from studies in
which higher-titered anti-HEV plasma or serum was infused
into naïve nonhuman primates, which were then challenged
with HEV.54,55 Thus, if an immune globulin preparation with a
sufficiently high titer of anti-HEV could be prepared, it would
probably be efficacious in preventing hepatitis E when adminis-
tered before exposure.
Active immunization
History of vaccine development
Among the HEV proteins, those encoded by ORF1 are nonstruc-
tural and not accessible to antibody. Thus, an ORF1 protein
vaccine would have to protect solely through cellular immune
mechanisms.56,57 Antibody to ORF3 is relatively short-lived and
genotype-specific.58 Although antibody to the ORF3 protein does
not neutralize HEV, immunization of nonhuman primates with
recombinant ORF3 provides some protection, probably through
cellular mechanisms.59,60 Thus, the capsid protein encoded by
ORF2 is the best candidate for a hepatitis E vaccine: Its sequence
is highly conserved, and antibody to it is long-lived and highly
cross-reactive among different HEV genotypes.58 Finally, anti-
body to the capsid protein neutralizes HEV in vitro and protects
nonhuman primates against hepatitis E.59,61,62
Since HEV grows poorly in cell culture, recombinant proteins
expressed in a variety of systems have been the principal source
of diagnostic and immunoprophylactic reagents.53 The most use-
ful have been expressed from or in insect cells
from baculovirus vectors. Both forms of expressed proteins have
been studied extensively and yield similar results. Full-length
expressed capsid proteins can be relatively hydrophobic and
insoluble and, therefore, are not useful reagents. Truncated
forms of the protein have been the most useful for diagnosis
and immunoprophylaxis, but small peptides have been of lim-
ited usefulness because the neutralization epitopes of the capsid
protein are conformational and encompass approximately 150
amino acids (aa 458–607) of this 660-amino-acid protein.61,63
This region of the capsid protein also contains elements that
are essential for homodimer formation and assembly into virus-
like particles (VLPs), both associated with eliciting neutralizing
antibodies64–67 (Figure 49-2). Homodimers, the smallest units
that can interact with neutralizing antibodies and stimulate
protection, spontaneously form VLPs that share the antigenic
characteristics of homodimers.66,68,69
The three-dimensional structure of the HEV VLP has
been solved by crystallographic and cryoelectron microscopic
examination of truncated capsid proteins.70–72 The 23 to 27 nm
subviral particles had a T = 1 icosahedral structure, whereas
a T = 3 icosahedral structure has been solved for full-size
HEV virions. However, the homodimer and VLP forms of the
Hepatitis E vaccines 49 1087
truncated capsid antigens, whether expressed from baculovirus
or have been highly antigenic and suitable for use as vac-
cines. Baculovirus-expressed vaccine candidates of 62 kDa (aa
112–636), 56 kDa (aa 112-607), and 53 kDa (aa 112-578) were
the residual products of spontaneous proteolytic cleavage in the
insect cell cultures.73–77 They underwent preliminary evaluation
as vaccines in vitro and in rhesus monkeys, and the 56 kDa
candidate, which combined the best qualities of immunoge-
nicity and stability, was chosen by GlaxoSmithKline (GSK),
Belgium, for clinical evaluation, and a vaccine lot was prepared
by Novavax, Inc. This vaccine was expressed in SF9 insect cells
from a baculovirus vector, purified by chromatography, and
packaged in five formulations containing 1 to 40 g antigen
plus 0.5 mg aluminum hydroxide in 0.5 mL.
Similarly, –derived vaccine candidates of 23 kDa (aa
394-606) and 30 kDa (aa 368-606) were evaluated, and the
larger, more immunogenic candidate was chosen for vaccine
development.64,69 A lot of vaccine was produced by Wantai
Biological Pharmaceutical Co, China, packaged in three
formulations of 5 to 20 g plus 0.8 mg of aluminum hydroxide
adjuvant in 0.5 mL, and designated “HEV 239 vaccine”.64
In other studies, a candidate hepatitis E vaccine consisting
of VLPs expressed in from truncated ORF2 of a genotype
4 HEV was combined with a commercial inactivated hepatitis
A vaccine and tested in mice for production of neutralizing anti-
bodies to the two viruses. Both components of the vaccine were
immunogenic, and the combined vaccine was at least as potent
as its individual components.78 The HEV component of the
vaccine had been shown previously to protect monkeys from
hepatitis, but the combined vaccine has not been subjected to
other preclinical or clinical trials.
Other approaches to vaccine development, including DNA
vaccines, chimeric virus particles, fusion proteins, encapsulated
proteins, oral vaccines, and cell-mediated vaccines have been
studied. However, their success has been limited, and none has
been fully developed for use in humans.79–88
Preclinical trials: safety, immunogenicity,
and efficacy
Following demonstration of immunogenicity in mice, the
baculovirus-expressed 56-kDa vaccine prepared for GSK was
evaluated in 44 rhesus monkeys.77 Two formulations of the
vaccine and two vaccination schedules were compared. There
were no untoward reactions in any of the animals. Both vaccine
formulations were highly immunogenic, and two doses of either
produced a similar level of antibody. One month following vac-
cination, the monkeys were challenged intravenously with
10,000 50% monkey infectious doses (ID50) of one of three chal-
lenge viruses: a genotype 1 strain (homologous to the vaccine),
a genotype 2 strain, and a genotype 3 strain.
None of the animals that received two doses of vaccine devel-
oped hepatitis, regardless of the challenge virus, whereas, 75%
of the monkeys receiving placebo (hepatitis A vaccine) devel-
oped hepatitis. Thus, a genotype 1–derived vaccine protected
against challenge with HEV genotypes 1, 2, and 3. The two-
dose regimens had a protective efficacy of 100% with a 95%
confidence interval (CI) of 65% to 100%, whereas the single-
dose regimen had a protective efficacy of 78% (CI, 29%-94%). As
in previous studies, protection correlated with antibody levels.
However, a similar vaccine failed to protect nonhuman primates
from infection or hepatitis when administered after challenge.89
In this respect, HEV vaccine differs from HAV vaccine.
Similarly, the –expressed 239 vaccine was tested for
safety, immunogenicity, and efficacy in 36 rhesus monkeys.64
Three vaccine formulations were administered at weeks 0 and
4; there were no untoward reactions. There was little difference
1088 SECTION THREE

Immunogenic proteins derived from open reading frame (ORF)2 (encoding the viral capsid) of hepatitis E virus (HEV) are depicted.
A, The full-length capsid protein, when expressed in insect cells, is processed at the amino- and carboxy-termini by vector (baculovirus)-derived
or insect cell–derived proteases in a cascade of truncated proteins, including B, D, and E.66,73 This protein, the candidate vaccine developed by
Genelabs,74 may also be a product of protease-mediated processing. Proteins C and D protected macaques from hepatitis E; protein E was only
partially protective in similar experiments.75,76,89 Protein D protected humans against hepatitis E in a phase 2 randomized, double-blind, placebo-
controlled trial.92 Regions of interest in the capsid protein expressed from insect cells are depicted in A: Amino acids 1 to 111 contain a signal-like
sequence (s.s.) that is the first cleavage product of the protease cascade; amino acids 125 to 601 are necessary for formation of virus-like particles
(VLPs);66 amino acids 459 to 607 comprise the smallest peptide that encompasses the HEV neutralization epitope (nt epitope) that can stimulate
neutralizing antibody when used as a vaccine and that can detect neutralizing antibody when used as an enzyme-linked immunosorbent assay
antigen.58,61,63 The full-length capsid protein, when expressed in : It is not suitable for vaccine use because important epitopes are
masked. H, A truncated form of HEV capsid protein formed homodimers and protected macaques from hepatitis E, but a slightly larger peptide
(G) dimerized, formed VLPs, was more immunogenic, and also protected macaques from hepatitis E.64 This latter peptide, designated HEV 239,
protected humans in a randomized, double-blind, placebo-controlled phase 3 trial.93 Regions of interest in the capsid protein expressed from
are depicted in F: Amino acids 368 to 394 are said to be necessary for formation of VLPs; amino acids 597 to 602 were found to
be necessary for dimerization (di) (see Li et al64). Apparent and real molecular masses, amino- and carboxy-termini, and regions of interest are
approximate because of differences between genotypes and inconsistencies reported among and within laboratories.

in the level of antibody following two doses of vaccine, regard-
less of the formulation. Three weeks after the second dose of
vaccine was administered, the monkeys were challenged with
the homologous genotype 1 virus, and one group was challenged
with the genotype 4 virus. Challenge doses were 107 and 104
genome equivalents, as measured by PCR (the ID50 titer was not
determined). There was no significant difference in the response
of monkeys receiving different doses of vaccine or challenged
with different genotypes of virus. The efficacy of the vaccine
against infection with the high-dose challenge was 75% (CI,
46.2%-90.9%). Vaccine efficacy against infection with the low-
dose challenge and protection against hepatitis with both chal-
lenge doses approached 100% (CI, 75.3%-99.8%). Thus, the 239
vaccine protected equally well against genotypes 1 and 4 of HEV.
Clinical trials: safety and immunogenicity
Phase 1 clinical trials of the GSK baculovirus vaccine were per-
formed by the Walter Reed Army Institute of Research in 88 US
Army volunteers, aged 18 to 50 years.90 It was administered in
four formulations (1, 5, 20, and 40 g) to 22 subjects each at
0, 1, and 6 months. All formulations of the vaccine were well
tolerated. The seroconversion rate was 67% for a 1-g formu-
lation and 89% to 95% following administration of the other
three formulations.
Subsequently, a second phase 1 evaluation was performed
in Royal Nepalese Army volunteers in Nepal, where hepatitis
E is endemic. Three doses were administered intramuscularly
to 22 Nepalese volunteers. The vaccine was well tolerated. One
month following the third dose, 100% of the volunteers had
developed anti-HEV.
The HEV 39 vaccine was tested in 457 seronegative
adult subjects and 155 seronegative students in China.91 The
adults were randomized, and two different 20-g dose regimens
were evaluated. The students received three-dose regimens
of four formulations (10, 20, 30, and 40 g). All doses of the
vaccine were well tolerated. Only the rate of total local reac-
tion was higher in the vaccinated groups than in the placebo
group. Vaccination with three doses of at least 10 g of vaccine
induced 100% seroconversion, and two 20-g doses induced
98% seroconversion.

Hepatitis E Vaccines Tested in Humans
aa, amino acids; CI, confidence interval.
Clinical trials: safety, immunogenicity,
and efficacy
A phase 2 trial of the GSK vaccine was carried out by the US
Army in seronegative members of the Nepalese military.92 Two
thousand volunteers stationed in the Kathmandu Valley ran-
domly received 20 g of the vaccine or a placebo at 0, 1, and
6 months in a double-blind study and were followed up for
2 years.
Adverse reactions occurred with similar frequency in the vac-
cine and placebo recipients. In the vaccine group, 100% had
seroconverted 1 month after the third vaccine dose. Among
participants, 87 definite cases of hepatitis E (84 icteric and 3 anic-
teric) were identified. There were 66 cases of hepatitis E among
participants who had received three doses of placebo, compared
with 3 cases among participants who had received three doses
of vaccine, for a vaccine efficacy of 95.5% (CI, 85.6%-98.6%).
Among participants who received only two doses of vaccine or
placebo, the vaccine efficacy was 85.7% (CI, 16.0%-98.2%). All
infections identified were caused by genotype 1 HEV.
Thus, the vaccine was highly efficacious in preventing clin-
ical hepatitis E in a region in which the disease is endemic.
All three cases of hepatitis in vaccinees occurred in the last
8 months of the 2-year trial, following 16 months of 100% pro-
tection, suggesting that such protection may not be long-lived.
A large phase 3 trial of the 239 vaccine (renamed Hecolin)
was carried out in China, under the sponsorship of Ziamen
Innovax Biotech.93 A total of 112,604 male and female partici-
pants, 16 to 65 years old, were randomly assigned to vaccine
or placebo groups without regard to preexisting antibody (47%
of a sampled subset were anti-HEV positive). Three doses of
vaccine (30 g) were administered intramuscularly to 86% of
the participants. Adverse events were few and mild. During
follow-up, 15 cases of hepatitis E were diagnosed, and all of
the cases were in the placebo group. Of 13 cases, 12 examined
were caused by genotype 4 HEV and 1 was genotype 1. Vaccine
Access the complete reference list online at http://www.expertconsult.com
23. Purcell RH, Emerson SU. Hepatitis E: an emerging awareness of an old
disease. J Hepatol 48:494–503, 2008.
34. Pavio N, Meng XJ, Renou C. Zoonotic hepatitis E: animal reservoirs and
emerging risks. Vet Res 41:46, 2010.
51. Lewis HC, Wichmann O, Duizer E. Transmission routes and risk factors for
autochthonous hepatitis E virus infection in Europe: a systematic review.
Epidemiol Infect 138:145–166, 2010.
Hepatitis E vaccines 49 1089
efficacy after three doses was 100% (CI, 72.1%-100.0%). Vaccine
efficacy after two doses was also 100% (CI, 9.1%-100.0%).
However, vaccinees were followed up for only 13 months, and
the duration of protection could not be determined.
Thus, although the two trials were quite different in design,
they both established a high protective efficacy of candidate
hepatitis E vaccines, whether expressed from baculovirus or
(Table 49-1). Based on analysis of the preclinical trials
of these two vaccines in rhesus monkeys,64,77 the level of anti-
HEV, expressed in WHO units per milliliter,94 that correlates
with 50% reduction in hepatitis is estimated to be less than 60
units, and the level of such antibody that correlates with 50%
reduction in infection is estimated to be less than 100 units.
Prospects for the future
Despite the success of these vaccines, they have not yet been
marketed. Although there is an apparent need for hepatitis E
vaccine in developing countries of Asia and Africa, the market
for such a vaccine in industrialized countries is perceived to
be quite small. However, this perception is changing as more
cases of hepatitis E are diagnosed in industrialized countries.
Certainly, if the cost of vaccine were low enough, it could be
widely used in endemic areas. It will be necessary to vaccinate
before adolescence, as most disease occurs in older children
and young adults. Attempts to halt epidemics by mass vacci-
nation are unlikely to be effective because most exposures to
the virus will probably have occurred by the time the epidemic
is recognized and because postexposure vaccination was not
protective for monkeys.89 Thus, a vaccination program that
would fit into World Health Organization vaccination sched-
ules of the Expanded Programme on Immunisation would seem
to be the most feasible. Based on experience with vaccination
against other enterically transmitted viruses, such as poliovi-
rus and hepatitis A virus, universal pediatric vaccination would
probably be necessary for effective control of hepatitis E.
92. Shrestha MP, Scott RM, Joshi DM, et al. Safety and efficacy of a recombinant
hepatitis E vaccine. N Engl J Med 356:895–903, 2007.
93. Zhu FC, Zhang J, Zhang XF, et al. Efficacy and safety of a recombinant
hepatitis E vaccine in healthy adults: a large-scale, randomised, double-blind
placebo-controlled, phase 3 trial. Lancet 376:895–902, 2010.
1102 SECTION THREE Vaccines in development and new vaccine strategies
HIV immunity
Diagrammatic representation of the evolution of virological
and immunological markers during the course of typical HIV-1 infection.
To date, there have been more than 25 years of HIV research, and there
has yet to be a successful vaccine or cure and will likely not be any
in the near future because there are still many areas that we do not
understand regarding interaction of HIV and the immune system.
A humoral immune response sets in and the patient becomes
seropositive, except in rare cases of rapid progression to
AIDS.195 The patient is less infectious than during the primary
phase of the disease because viral replication continues at a
more restrained pace. The major site of virus replication again
is the GALT, essentially the colon and rectum.196 Plasma viral
load, which peaked during the acute phase of infection at val-
ues of 106 to more than 107 viral RNA copies per mL, decreases
and remains at a relatively stable plateau (the set point) during
chronic infection. The set point differs from patient to patient
and can serve as an indicator of disease progression. It remains
high in “fast progressors”, in whom disease rapidly progresses
to AIDS, whereas it falls to relatively low values in “long-term
nonprogressors”, who contain the viremia spontaneously and
progress to disease slowly. Thus, patients with more than 105
HIV RNA copies per milliliter of plasma at set point are 10-fold
more likely to progress to AIDS during the following 5 years
than patients with fewer than 105 copies. In a few seropos-
itive persons, called “elite controllers” or “elite suppressors”
(ES), viral loads spontaneously fall below the limit of detection
by standard clinical assays (viral RNA copies, 50-75/mL).
However, even in these patients, virus replication continues at
a low pace, as judged from the observation that the and
genes of the resident HIV-1 strain undergo sequence changes
over time.197
HIV-1 rapidly evades neutralizing antibody responses dur-
ing the infection of a person.22,73 The virus also readily escapes
specific CTL responses.198–205 Of note is the fact that some
viral escape mutations in response to immune pressure come
at a fitness cost to the virus,206 which eventually reverts to
wild-type sequences after transmission to a new host.207–209
Selective pressures within microenvironments of different
anatomic compartments result in the emergence in various tis-
sues of distinct, independently evolving, replication-competent
quasi-species.210
The asymptomatic phase can last from a few months to
more than 25 years, after which symptoms occur, following
onset of opportunistic infections. At times, HIV infection can
also manifest as a purely cachectic disease, to which the name
“slim disease” has been given in Africa. It can also cause a neu-
rological syndrome such as acute encephalitis or progressive
dementia (AIDS dementia complex). The symptomatic phase
of the disease is accompanied by a return of p24 antigen-
emia, an accelerated decrease in the number of CD4+ T cells,
and a greatly increased virus load. The number of peripheral T
cells producing HIV-1 can reach 1 in 10, with production of up
to 1010 virions per day.211–214
Direct killing of CD4+ T cells by HIV-1 cannot, however,
explain the full spectrum of AIDS pathogenesis. Chimpanzee-
adapted strains of HIV-1 are cytopathic for chimpanzee CD4+
T cells, grow in chimpanzee macrophages-monocytes, and
replicate efficiently in chimpanzees, but usually do not cause
AIDS in the animal.215,216 Similarly, SIVmac infection in African
NHP, such as mandrills or sooty mangabey monkeys, remains
chronic and almost never progresses to AIDS.217–221 Natural
NHP hosts that are infected by SIVs in the wild show no evi-
dence of destruction of CD4+ T cells, although they maintain
high plasma viral loads222–224 and do not exhibit superior cellular
immune control of viremia.225
Interestingly, these SIV-infected African monkeys show no
evidence of persisting chronic immune activation226 or of apop-
totic CD4+ T lymphocytes, in contrast with HIV-1–infected
humans or SIVmac-infected macaques. These observations
strengthen the likely role of chronic immune activation due to
microbial translocation in the eventual collapse of the T-cell
compartment during infection.171,227–229 Hyperactivation over-
whelms the immune system and gradually causes a collapse
of immune protective mechanisms, leading to AIDS. This is
observed even in elite controllers, who show abnormal T-cell
activation levels leading to progressive CD4+ T-cell loss in the
absence of detectable viremia.230 In addition, other mechanisms
have been suggested to account for the depletion of CD4+ T
cells, such as cytolytic attack by HIV-1–specific CTLs of
bystander CD4+ T cells that have bound virus-shed circulating
gp120 molecules; apoptosis of bystander CD4+ T cells medi-
ated by circulating Tat or Nef proteins; apoptosis of bystander T
cells triggered by incomplete reverse transcripts during abortive
HIV infection;231 and molecular mimicry between the ectodo-
main of gp41 and IL-2, leading to the induction of anti–IL-2
antibodies that could block the interaction of the cytokine with
its receptor.232
People with AIDS are anergic to recall antigens, as seen by
lack of delayed-type hypersensitivity (skin test) in vivo and by
lack of Th cell proliferation in vitro. The unresponsiveness
to recall antigens has been interpreted as the consequence of
HIV infection of follicular DCs, which severely impairs the
ability of these cells to present antigens in vitro and triggers
the secretion of IL-10.233,234 Anergy was indeed found to cor-
relate with destruction of the lymph node follicles.235,236 People
with AIDS also have increased levels of tumor necrosis factor
(TNF)-; IL-1, IL-2, and IL-6; interferon (IFN)- and IFN- ; 2-
microglobulin; and neopterin, an interleukin produced by mac-
rophages in response to TNF-. 237
Not all exposures to HIV, however, lead to infection, and
not all HIV infections lead to AIDS. Only one third of infants
born to infected mothers acquire infection, heterosexual trans-
mission occurs approximately between 1 of 100 and 1 of 1,000
exposures,238–241 persons with hemophilia exposed to infected
blood products do not consistently get infected,242 and some
commercial sex workers and seronegative partners in serodis-
cordant couples seem to remain uninfected despite repeated
1104 SECTION THREE
The biological significance of such a high degree of variabil-
ity remains unknown. There are no clear genotype-serotype
relationships between subtype or CRF sequences and neutral-
izing antibody patterns. Presumably, a recombinant virus strain
that becomes predominant in a person has some type of selec-
tive advantage over the parental strains due to enhanced fitness
or immune evasion. When such a recombinant emerges as the
predominant strain in an epidemic in a population, this also
must be due to significant selective advantages. The fear that an
HIV strain would readily escape from too narrow an immune
response induced by a vaccine is in favor of the development
of vaccines that target multiple HIV antigens and, if possible,
multiple viral subtypes, so as to provide as broad as possible a
protection. This has also led to the development of fully syn-
thetic genes derived from theoretically defined “consensus”
or ancestral HIV envelope sequences,262–264 which were found
to be superior to many wild-type immunogens for inducing
cross-subtype T- and B-cell immune responses.265,266 Similarly,
“mosaic antigens” have been developed with sequences derived
from algorithms that closely match the sequences of
natural HIV-1 strains worldwide and maximize the overlap
between them.267–270
There is convincing evidence that HIV originated in
NHPs.271,272 Several independent crossing of the species barrier
from chimpanzees living in western central Africa led to HIV-1
groups M, N, and O in humans, whereas multiple independent
transmissions of SIV from West African sooty mangabeys led to
HIV-2.273–275 Thus, HIV-1 groups M, N, and O are phylogeneti-
cally close to SIVcpz, a commensal virus in chimpanzees (
276,277
), whereas HIV-2 is closely related to SIVsmm,
a commensal virus in sooty mangabey monkeys (
); the SIV, which causes AIDS in Asian rhesus macaques
(), is similarly closely related to SIVsmm from
sooty mangabeys. Recently identified HIV-1 group P is related
to SIVgor, a commensal virus in gorillas.249,278
All of these viruses belong in the same group of lentiviruses
that infect a wide range of NHP species in Africa,275,279,280 where
multiple cross-species transmissions in the wild have been doc-
umented.281–283 Molecular clock estimates indicate that HIV-1
group M probably established itself in the human population at
the beginning of the 20th century (1884-1924),284 group O in the
1920s (1890-1940),285 and group N in the 1960s (1948-1977).274
In fact, SIVs have been present in African primates for more
than 30,000 years, suggesting that humans most probably had
many chances of sporadic encounter with these viruses for mil-
lennia.286 Humans are still naturally exposed today to infection
by diverse SIVs, as illustrated by the high frequency and diversity
of lentiviruses that can be encountered in primate bushmeat in
southeastern Cameroon.287,288
Epidemiology and disease burden
HIV transmission occurs essentially through homosexual or
heterosexual intercourse, through injection of blood or blood-
derived products, and from mother-to-child during pregnancy,
at delivery, or through breastfeeding. Recently infected sub-
jects are potentially more infectious than infected persons in
the chronic phase of the disease: It has been suggested that
up to 50% of new HIV-1 infections are acquired from newly
infected persons.239,289–291 The importance of the acute infec-
tion period for subsequent transmission events is related to the
level of plasma viral load during the early phase of disease and
to virus-specific properties.238 Thus, virions isolated from SIV-
infected macaques at the time of primary infection are more
infectious than those from the chronic phase of infection,292
possibly because a greater number of defective virus particles
are circulating during the chronic than during the acute stage
of disease167 and/or because lack of protective antibodies at the
primary infection stage substantially increases virus infectivity.293
The difference in sexual behaviors makes MSM add to their
risk of infection because receptive anal intercourse has a rela-
tive transmission probability of 14.3 infections per 1,000 coital
acts, which is about 10 times higher than that of receptive vagi-
nal intercourse.294, 294a
Since 1981, more than 25 million people have died of AIDS.
HIV-1 is the fourth largest infectious killer in the world, with
an annual toll of about 2.0 million deaths,295 the great majority
of which occurs in sub-Saharan Africa. As of 2009, UNAIDS
estimated that 33.4 million people were living with HIV-1 infec-
tion worldwide, including 2.5 million children younger than
15 years. Of these 33.4 million people, 22.5 million lived in sub-
Saharan Africa and about 4.1 million in Asia. About 2.7 million
people become infected each year, 95% of whom live in develop-
ing countries.295 The number of new infections is equally dis-
tributed among men and women, but infection rates in young
African women are close to three times higher than those among
young men, reflecting the degree to which gender inequities are
driving the epidemic.
The HIV-1 epidemic expanded at an alarming rate in the
1990s in sub-Saharan Africa. The region accounts for nearly
75% of the global AIDS deaths; 80% of females living with
HIV are in sub-Saharan Africa. The highest infection rates
are found among commercial sex workers, truck drivers, and
seasonal migrant workers. Countless cases of mother-to-
child transmission have been documented. In some African
countries, overall prevalence in the adult population can
be greater than 10%, with figures reaching up to 28% in
some areas. Most severely hit are South Africa, with more
than 5.5 million infected people, Botswana, Mozambique,
Zimbabwe, Tanzania, and Ethiopia. In addition, sub-Saharan
Africa faces numerous wars and civil conflicts, produc-
ing large numbers of refugees who are at heightened risk of
contracting HIV. Through increased awareness campaigns,
education, and wider access to ART, many African countries
are fortunately seeing the incidence rate stabilize or decrease.
According to the 2010 annual report of the Joint United
Nations Programme on HIV/AIDS (UNAIDS), HIV incidence
declined by more than 25% during the last decade in 33 coun-
tries, 22 of them in sub-Saharan Africa. The incidence rate in
African children has also started to decrease owing to the sys-
tematic use of antiviral therapy in pregnant women, as exem-
plified by programs implemented in South Africa, Swaziland,
Botswana, and Namibia.
The estimated number of people living today with HIV-1 in
Asia and the Pacific region is more than 4 million, but the accu-
racy of the figure is questionable, in view of the fast pace at
which the epidemic has been expanding. Increasing sex trade,
use of illicit drugs, and rates of sexually transmitted infections
contributed to increased vulnerability in the region. Injection
drug use and heterosexual intercourse are the primary modes
of transmission, although improper blood donation practices in
China and unsafe injection practices in health care settings in
India and surrounding countries have resulted in hundreds of
thousands of infections. Substantial transmission also occurs
in MSM, with prevalence rates of 14% to 20% reported in male
homosexual communities in India, Cambodia, and Thailand.
Gender inequities have a major role in the epidemic as young
girls are frequently steered toward sex work by their families.
The female commercial sex industry facilitates transmission
in Thailand, Cambodia, India, Nepal, China, Vietnam, and
several other countries.
The estimated number of adults and children living with
HIV-1 in Latin America and the Caribbean at the end of 2008
was 1.6 million. While in some countries HIV infections
remained concentrated mainly in MSM and injecting drug
users (people who inject drugs [PWIDs]), others were experienc-
ing increasing rates of heterosexual transmission.

The eastern European countries continue to experience,
since the mid-1990s, a sharp increase in the number of new
HIV-1 infections, most of which occur among PWIDs, commer-
cial sex workers, and MSM. The number of HIV-infected per-
sons in the region nearly tripled between 2000 and 2009, from
500,000 to 1.4 million, of which an estimated 940,000 live in
the Russian Federation and 440,000 in Ukraine.296
An estimated 2.3 million people are living with HIV-1 in
industrialized countries (1.5 million in North America alone).
In France, the Netherlands, and Spain, from one third to three
quarters of new HIV-1 infections are concentrated among
migrants. For the general population, highly active ART
(HAART) has considerably reduced disease progression to AIDS
and transformed HIV infection from a deadly disease to a some-
what manageable chronic disease. However, success in treat-
ment and care is not matched by progress in prevention, and
new evidence of rising HIV-1 infection rates is emerging, par-
ticularly in marginalized communities. For example, 1 in 16
black men in Washington, DC, is HIV-infected, as is 1 in 10
MSM and 1 in 8 PWIDs in New York City.297 In several US
urban areas, the prevalence among MSM is as high as 30%. It is
estimated that half a million Americans became infected with
HIV in the 2000-2010 decade, including more than 55,000 in
2009.298 The US capital is beset by an HIV-1 epidemic that
rivals some in developing countries,299 with HIV rates in par-
ticular subsets of the heterosexual population as high as 5.2%
(6.3% in women vs 3.9% in men).300,301 HIV-1 infection rates
among young people in Western nations are three times higher
than they were in the early 2000s. HIV thus remains a major
health threat in these countries.
Immune correlates of protection
The immune correlates of protection against HIV infection in
humans remain unknown,302,303 mostly because of the lack of
natural protective immunity against HIV. Most available data
have been drawn from the study of chronically infected persons,
which does not imply that they would be associated with pro-
tection if elicited by a vaccine before infection.304 Other data
arose from the study of vaccine protection of macaques against
experimental SIV infection. Possible immune correlates of pro-
tection can be of four types: neutralizing antibodies, cellular
immune responses, nonneutralizing antibodies, and innate
immune responses.
Neutralizing antibodies
Neutralizing antibody responses to HIV-1 are unusually late,
arising several weeks after infection.305,306 In addition, most
antibodies elicited during HIV infection are nonneutralizing
or neutralize only a narrow range of circulating viral strains,73
and many target the highly variable loops in gp120 that act
as antigenic decoys, such as the V1 loop,307 the V2 loop,308 or
the V3 loop.309–314 Broad neutralization epitopes actually exist
on gp120 and gp41 (see later text), but they remain mini-
mally accessible to the immune system due to conformation-
dependent exposure on the molecule, hindrance by the “glycan
shield”, and masking by the hypervariable loops in gp120.315
Some infected persons, however, do develop broadly neutral-
izing antibodies.316–320 These appear with greater time after
infection, usually becoming evident, on average, 2.5 years after
infection, which might imply a need for antibody affinity matu-
ration,32,321 and in persons with higher viral plasma load, sug-
gesting a need for chronic antigen exposure.322–325 Up to 20% to
30% of subjects ultimately develop neutralizing antibodies with
a degree of breadth.326 Only about 1% of the donors, however,
designated as elite neutralizers, show unusually potent neu-
tralizing serum activity against a majority of clades.327 Broadly
Human immunodeficiency virus vaccines 51 1105
HIV-1–neutralizing antibodies have also been identified in rhe-
sus macaques infected with R5 SHIVSF162P4 and found to recog-
nize quaternary epitopes on gp120.328
Experiments in the macaque model have shown that SIV or
SHIV infection can be prevented by passive infusion of broadly
neutralizing antibodies329–335 or by adeno-associated virus (AAV)-
mediated continuous expression of immunoadhesins derived
from broadly neutralizing antibodies.336 Full protection against
SHIV89.6P infection in adult macaques was obtained by pas-
sive immunization with HIV-1 neutralizing monoclonal anti-
bodies 2 G12 and 2 F5 and anti–HIV-1 immunoglobulins332,337
or by monoclonal antibody 2 F5 or 4E10.338 Similarly, neona-
tal macaques could be protected from oral SHIV challenge by
passive immunization with broadly neutralizing monoclonal
antibodies b12, 2 G12, and 2 F5.330,339 Broadly neutralizing anti-
bodies are therefore not associated with control of viremia in
the host,323,340–342 but they can be protective against live virus
challenge.
The challenge remains, however, of translating these findings
into the design of immunogens that could elicit broadly neutral-
izing antibodies when incorporated into a vaccine.66,343–349
Cell-mediated immunity
Multiple studies have shown that CD8+ T-cell responses con-
trol SIV or SHIV replication and the level of the set point viral
load in infected rhesus macaques. The best evidence came
from the demonstration that experimental depletion of CD8+
T cells resulted in a rapid and dramatic increase in viremia
and accelerated the lethal outcome of the disease in the ani-
mals.350–353 Accidental reemergence of virus and progression to
AIDS in long-term protected, DNA-MVA vaccinated macaques
was observed at 4 years after challenge with SHIV-89.6P, coinci-
dent with a loss of T-cell function together with appearance of
escape mutations in CD8 epitopes.354 Transient depletion of
CD8+ T cells in rhesus macaques that controlled the replica-
tion of SIVmac for 1 to 5 years resulted in 100 to 10,000 fold
increases in viremia; control was reestablished when the CD8+
T cells returned, at which time some subsets of subdominant
CD8+ T cells expanded more than 2,500-fold above predepres-
sion level.355 Infection of female macaques with nonpathogenic
SHIV89.6 used as an attenuated live vaccine generated protec-
tion of the monkeys against uncontrolled SIV replication fol-
lowing vaginal challenge,356 but, again, depletion of CD8+ T
cells at the time of the challenge completely abrogated protec-
tion, demonstrating that it was the CD8+ T cells in the genital
tract that controlled viral replication and delayed progression
to AIDS.357
Efforts at elucidating the potentially protective immune
response of persons who remain resistant to HIV-1 infection358,359
have also helped shed light on the role of cell-mediated immu-
nity in protection.360 Elite controllers or ES seem to completely
control viral replication for up to 30 years or longer, maintain
viral loads below the level of detection by standard assays, and
most often do not show signs of disease progression.361 ES rarely
develop broadly neutralizing antibodies,320,325,362 but they show
potent cellular immune responses. On stimulation, their HIV-
specific CD4+ T cells proliferate and secrete multiple cytokines,
including IL-2, whereas cells from chronic progressors do not
proliferate and secrete only IFN- . The presence of strong virus-
specific CD4+ T-cell responses also distinguishes nonpatho-
genic HIV-2 infections from HIV-1 infections.363,364
Even more striking than CD4+ T cells, HIV-specific CD8+
T cells seem to have a key role in the control of viral replica-
tion in ES. This is demonstrated in part by the overrepresenta-
tion of certain HLA alleles such as HLA-B*57 and HLA-B*27
in ES cohorts365–367 and the fact that macaque monkeys with the
MHC class I Mamu-B*08 allele achieve the ES status. The cen-
tral role of HLA class I in controlling HIV-1 infection368–370 was
1106 SECTION THREE Vaccines in development and new vaccine strategies
highlighted by the International HIV Controllers Study, which
followed up a large cohort of people with HIV and showed
that HLA types B*57:01, B*27:05, and B*14 were protective,
whereas types B*35 and C*07 were associated with progression
to AIDS.371 Allele combinations like HLA-B*57:01-Cw0602,
HLA-B*27:05-Cw0102, and HLA-B*38:01-Cw1203 have the
strongest impact on nonprogression.372
While HIV-specific CD8+ T cells from patients with progressive
disease typically secrete only IFN- , those from ES are multifunc-
tional, secrete multiple cytokines,373–375 express CCL3 (MIP-1 )
and/or CD107a, undergo degranulation,376 and proliferate in
response to stimulation.377,378 They are also much more efficient
at granzyme B delivery and killing of infected CD4+ T cells.379,380
Moreover, they express perforin immediately on antigen-specific
stimulation, without the need for prior proliferation or the addi-
tion of exogenous cytokines.381 HIV Gag-specific T-cell responses,
especially those directed at conserved regions in Gag p24382 and
Gag p7, have been found to dominate in rectal mucosa of HIV
controllers, in magnitude and breadth,383 and often are more
abundant in gastrointestinal mucosa than in blood.384 Long-term
nonprogressors have more polyfunctional responses, higher-
magnitude and broader p24-specific proliferation, and higher
levels of IL-2 and TNF- production than do progressors.385
HIV-specific helper T-cell and CD8+ CTL responses to
multiple HIV epitopes have also been demonstrated in EU
persons.386,387 Some of the Nairobi EU sex workers seroconverted
after they took a break from sex work, which was associated with
a loss of HIV-specific CD8 responses, suggesting a waning of
HIV-specific immunity following reduced antigenic exposure.388
The conclusion that T cells, especially CD8+ T cells, can
limit virus replication and control viral set point viremia also
stems from vaccine experiments in the macaque/SIV model.
Indian rhesus macaques vaccinated intracolorectally with a pep-
tide prime/MVA boost adjuvanted with TLR agonists and IL-15
controlled virus replication after high-dose intra rectal SIV
challenge. In addition to a strong stimulation of A3G (see later
text), protection correlated with antigen-specific polyfunctional
CD8+ T cells.114 However, correlation was observed only above
a threshold of about 2% CD8+ T cells, below which no effect
of CD8+ T cells on viral load was observed. Efficient control
of SIV infection after repeated low-dose rectal challenge with
SIVmac239 was also observed in rhesus macaques vaccinated
with a replicating CMV vector expressing Gag, Rev, Tat, Nef,
and Env that efficiently induced effector memory CD4+ and
CD8+ tissue-resident T cells.353 Full protection against infec-
tion was observed in 13 of 24 monkeys, none of which showed
evidence of neutralizing antibodies, but all of which developed
CD4+ and CD8+ T cells in lymph nodes and mucosal tissues
(S.G. Hansen, personal communication).
A vaccine that would generate effector memory T cells at
mucosal target sites and high avidity, polyfunctional CTLs able
to inhibit viral replication through granzyme B–mediated kill-
ing of infected target cells should likely elicit long-term control
of HIV infection.38,389–393
Nonneutralizing antibodies
Nonneutralizing antibodies that bind to HIV antigens on the
surface of HIV-infected cells by their Fab fragments can recruit
by their Fc fragment innate immune cells that have an Fc
receptor (FcRs), including antigen-presenting cells), NK cells,
or monocyte-macrophages. Such FcR-mediated recruitment
of innate immune cells can lead to killing of the infected cell,
referred to as antibody-dependent cellular toxicity (ADCC),
or to inhibition of viral replication, referred to as antibody-
dependent cell-mediated virus inhibition (ADCVI).394,395
Studies in animal models have shown that protec-
tion of newborn monkeys against oral SIV infection by pas-
sive immunization with a nonneutralizing anti-SIV serum
correlated with ADCVI activity in the serum.396 Immunization
with a replicating recombinant adenovirus followed by boost-
ing with an Env subunit vaccine partially protected rhesus
macaques against mucosal challenge with SIV or pathogenic
SHIV in the absence of a neutralizing antibody response,397,398
and there was evidence that reduced viral loads at the acute and
chronic stages of infection significantly correlated with ADCC
and ADCVI.399–402 ADCC and ADCVI activities were, in turn,
directly correlated with antibody affinity, suggesting the impor-
tance of antibody maturation for functionality.
Most anti-gp41 antibodies are of the nonneutralizing type:
They are directed at the immunodominant C-C loop in gp41
(cluster I) or at epitopes located in the MPER segment (clus-
ter II). Cluster II antibodies, albeit nonneutralizing, mediate
ADCC and other Fc-mediated antiviral activities.403–405 A recent
study in which gp41-specific antibodies were cloned from the
memory B-cell compartment of HIV-1–infected persons showed
that clones targeting cluster II epitopes accounted for 49% of all
anti-gp41-reactive B cells.406 The gp120-depleted gp41 stumps
observed on the surface of HIV virions are probably in the trig-
gered, six-helix bundle form and may be the chief source of this
type of antibody response.407,408
Anti–HIV-1 broadly neutralizing antibodies may also be
mediators of ADCC, as suggested for antibody b12 whose infu-
sion in the native form protected 8 of 9 monkeys against high-
dose vaginal SHIV challenge, whereas a mutated version of
b12 whose Fc fragment had lost ability to bind to FcRs pro-
tected only 4 of 9 macaques.334,409 Another broadly neutraliz-
ing antibody, 2 G12, seems to much better protect in vivo than
would be expected from its neutralization ability in vitro.332,335
Modification of Fc fragment glycosylation of 2 G12 was capa-
ble of improving antiviral activity by enhancing ADCC and
ADCVI, as shown by comparing various glycoforms produced
in wild-type and glycoengineered plants.410 Similarly, the
in vitro neutralization potency of antibodies 2 F5 and 4E10 was
substantially increased when the cell line used for assessment
of HIV-1-neutralization was engineered to express FcRI 405,411
or, to a lesser extent, FcRIIb. 412 2 F5 at nanogram per milliliter
concentrations elicited ADCC against X4- and R5-HIV-infected
monocytes or monocytic cell line.413
In humans, the appearance of ADCVI in acutely infected
subjects occurs as early as the CTL response, ie, much earlier
than the neutralizing antibody response.305,394 Elite controllers
were shown to exhibit more potent ADCC activity than viremic
persons, whereas neutralizing antibody activity tended to para-
doxically be higher in viremic subjects.414 HIV-1 Env-specific
ADCC was detected in a number of early studies.404,415
Another important function of some nonneutralizing antibod-
ies, especially IgAs, seems to be the inhibition of transcytosis.
Penetration of HIV-1 through the mucosal barrier of the geni-
tal or the intestinal mucosa can involve uptake of the virus by
intraepithelial Langerhans cells or DCs, as observed in pluristrati-
fied squamous epithelia such as those of the vagina and exocervix,
anus, oral cavity, and esophagus. It can also happen by transloca-
tion of the virus across simple columnar epithelia such as those of
the endocervix, rectum, and intestinal tract by the mechanism of
transcytosis.416 Mucosal transcytosis-blocking antibodies, mostly
IgAs, have been detected in the cervicovaginal secretions of HEPS
women in discordant couples417 and in sex worker cohorts.418,419
HEPS persons, also known as EU persons,420,421 show no
HIV-1–specific plasma IgG422–424 but exhibit HIV-1–specific
mucosal IgGs and IgAs,422,425–428 with IgAs targeting gp41,
especially the Gal/Cer binding site in the MPER of gp41.429,430
A mucosal gp41-specific Fab IgA library was constructed, and
cervical B cells from these women were used to generate librar-
ies of gp41 MPER-specific monoclonal IgAs that were able to
inhibit HIV-1 transcytosis across an epithelial cell membrane
in vitro.431 IgAs with HIV-1-neutralizing activity could also be
recovered from plasma and saliva from HEPS persons.432,433 At

this time, there is no demonstration that topically or mucosally
delivered or passively administered transcytosis-blocking IgAs
can prevent experimental mucosal infection in NHPs.
Innate immunity
Viral infections usually result in a strong activation of the innate
immune system that precedes the development of adaptive
immune responses.434,435 Lentiviral infections are no exception.
Innate immunity responses are among the key determinants of
the outcome of HIV-1 infection.109,112,113
Thus, SIV infection of African green monkeys or sooty
mangabeys was found to induce a strong but rapidly controlled
innate response.436,437 The activation of innate immune cells,
essentially plasmacytoid DCs, proceeds through recognition
of HIV-1 or SIV by Toll-like receptors (TLRs), especially TLR7
and TLR8,438,439 and leads to the production of antiviral cyto-
kines,435,440 including IFN-. 441–443 TRIM22 and type I IFN
were found to be key players during primary SIV infection.444
IFN- , in turn, upregulates intrinsic host restriction factors,
such as human APOBEC3G (hA3G), which can edit the HIV-1
early reverse transcripts through cytidine deamination, leading
to degradation of the provirus DNA103,105 and tripartite motif
(TRIM) proteins, including TRIM-5. 445
A striking inverse correlation was found between A3G mRNA
levels and HIV viral loads, and a positive correlation was found
between A3G mRNA levels and CD4 cell counts in HIV-1–
infected subjects.446,447 Mucosal immunization of macaques
against SIV led to persistent expression of A3G,110 and elevated
A3G was demonstrated to mediate resistance of monocyte-derived
DCs to HIV-1 infection.111,448 The use of TLR agonists and IL-15
as vaccine adjuvants in macaques had a remarkable effect on pro-
tection against mucosal SIV challenge: The adjuvants alone with-
out vaccine antigen elicited enhanced A3G mRNA expression in
DCs, monocyte-macrophages, and CD4+ T cells, resulting in a
significant decrease in the set point plasma and colon tissue viral
loads following rectal SIV challenge.114
Another member of the APOBEC3 family that might be involved
in innate immunity against HIV-1 is APOBEC3B, as judged by the
fact that a naturally occurring deletion in the APOBEC3B gene
was significantly associated with increased susceptibility to HIV-1
infection, progression to AIDS, and viral set point.449
Acute HIV-1 infection also results in the activation and
expansion of CD3 NK T cells450,451 that, in combination with
appropriate HLA class I ligands, are associated with better con-
trol of HIV-1 replication and slower disease progression.452–455 NK
cells make up half of the IFN- –producing cells in HIV-1 peptide-
stimulated PBMCs from HIV-infected persons.456 NK activity
is increased in long-term non-progressors457,458 and in EU sub-
jects.458,459 The expression of certain alleles of killer immunoglob-
ulin receptors (KIRs) on NK cells, such as KIR3DS1 or 3DL1, in
combination with HLA-Bw4801, was associated with the slowest
rate of progression to AIDS in large cohort studies.453
NK-cell activity was found to be enhanced in EU subjects in
the absence of target stimulation, providing them with a better
armed immune response to resist HIV infection.455 NK activ-
ity is also increased in acute SIV infection of African NHPs,
which control disease progression, but not in rhesus macaques,
which develop full-blown disease.460 The variety of NK cells and
NK cell receptors makes their study complex, but the discovery
of innate memory in NK cells may offer interesting perspec-
tives for vaccine development.461,462 Cytokines such as IL-21
and IL-15 may help induce or improve NK-cell responses.463
Paradoxically, however, NK cells might also play an enhancing
role in HIV pathogenesis. In the initial weeks post-infection,
the great majority of dying cells are uninfected CD4+ T cells.
These cells were found to express on their surface NKp44L, a
cellular ligand for an NK cell receptor that renders them sus-
ceptible to autologous NK cell killing. A motif of HIV-1 gp41
Human immunodeficiency virus vaccines 51 1107
called “S3” was identified that binds to gC1qR on CD4+ T cells
and triggers the exposure of NKp44L. These findings throw new
light on the mechanisms of CD4+ T-cell depletion and might
lead to new strategies to prevent it.463a, 463b
Several other factors of innate immunity also have a role in the
control of lentivirus infection, such as p21, an inhibitor of cyclin-
dependent kinase present in CD4+ T cells;464,465 T-bet, a master
transcription factor that promotes the expression of IFN- , gran-
zyme B, and perforin by CD8+ T cells;466 IL-27;467 and TRIM-5 ,
which controls macaque susceptibility to SIVsmE66061,63 and was
found to be associated with protection from HIV-1 infection in
some sex worker cohorts.468 The p21 inhibitor seems to reduce the
susceptibility of activated CD4+ T cells and human macrophages
to HIV-1 replication through inhibition of transcription of the pro-
viral DNA. It also can block reverse transcription. The expression
of p21 and T-bet is upregulated in CD4+ T cells from elite control-
lers, and experimentally knocking out p21 substantially increases
cell susceptibility to HIV-1.469 Another antiviral innate immunity
factor is TRIM21, which neutralizes virus-antibody complexes
inside the cell by binding to antibodies that remain attached to
virus particles and targeting the virus-antibody complex to the pro-
teasome, thus mediating the intracellular disabling of the virus.470
It has become evident that innate immune mechanisms
contribute to control of HIV, but it is not clear if and how that
response could be harnessed in vaccination.113,359,471
Candidate HIV vaccines: preclinical
development
The development of an HIV-1 vaccine is a formidable chal-
lenge. An effective HIV vaccine should be able to induce dura-
ble immunity and prevent the acquisition of infection and/or
reduce virus replication in infected persons so as to slow disease
evolution and reduce viral transmission.471–475 The first phase 1
trial of an HIV vaccine was conducted in the United States in
1987 using a gp160 subunit vaccine. Since then, more than 40
vaccines have been tested in more than 80 phase 1/2 clinical tri-
als involving more than 10,000 healthy human volunteers.36,37
HIV-1 vaccines that were developed and tested in humans were
first based on induction of neutralizing antibodies, then on
induction of cellular immune responses.40,476–480 Recent stud-
ies have shown that robust mucosal immunity, high-avidity
polyfunctional T cells, and broadly neutralizing antibodies are
potentially important factors of protective immunity against
HIV-1,481,482 but they also highlighted the need to design vac-
cines that can induce more potent immune responses than
those elicited by natural infection because the latter provides
inadequate protective immunity.35,38,483 As classical vaccine
strategies based on attenuated live virus or whole inactivated
virus show severe limitations, most efforts to develop HIV vac-
cines have focused on newer vaccine approaches. This section
will discuss vaccine development efforts in animal models,
whereas vaccine clinical trials are reviewed in a later section.
Most of the efforts to develop and evaluate HIV vaccines
are borne by public sector research agencies (eg, the National
Institutes of Health, the Centers for Disease Control and
Prevention (CDC), and the Walter Reed Army Institute of
Research, in the United States; the French Agence Nationale de
Recherches sur le Sida; and the Karolinska Institute in Sweden),
international development agencies (eg, USAID, UK-DFID and
European international development donors), and nongovern-
mental organizations (eg, International AIDS Vaccine Initiative
[IAVI], New York, NY; Bill and Melinda Gates Foundation,
Seattle, Washington), together with initiatives in the World
Health Organization, UNAIDS, and the European Union. The
HIV Vaccine Trial Network, established by the NIAID in 2000,
with 25 clinical sites in 4 continents, represents a major resource
1108 SECTION THREE
for clinical HIV vaccine research. Complementary clinical tri-
als partnerships have been developed by the US Military HIV
Research Program in Africa and Thailand and by IAVI in Africa
and India. The European Union has created the European and
Developing Countries Clinical Trials Partnership with the aim to
help developing countries build capacity for testing the efficacy
of new drugs, microbicides, and vaccines. Major efforts have also
been made by the vaccine industry at developing and testing can-
didate HIV vaccines, including phase 2b/3 clinical trials.
In 2003, recognizing the enormity of the scientific challenge,
leading investigators and major funding organizations formed
the Global HIV Vaccine Enterprise, an alliance of independent
organizations committed to working together to accelerate the
development of an HIV vaccine.484 A strategic scientific plan
was drafted in 2005 and updated in 2010.485 The plan aims
at accelerating clinical trials of promising vaccine candidates,
bridging the gaps between basic and clinical research, facilitat-
ing exchange of reagents and data, enhancing global collabora-
tions, and fostering the development of young scientists and
clinicians in the field of HIV vaccines.
Animal models
HIV-1 can be transmitted to chimpanzees486,487 and pig-tail
macaques (), but neither of these models can
reproduce the pathology of AIDS in humans. HIV-1 also can rep-
licate in immunodeficient transgenic mice engrafted with human
fetal lymphoid tissues (SCID-hu mice) or with adult human
peripheral blood leukocytes (hu-PBL-SCID mice)488 or in human-
ized immunodeficient mice transplanted with human cord blood
hematopoietic stem cells,489–491 but this is not an easy model to
work with. In contrast, SIVmac is easily transmitted to and read-
ily causes AIDS in rhesus macaques, especially from India.492
Infection of macaque monkeys with SIVmac recapitulates the
pathogenic effects of HIV-1 in humans and provides the most reli-
able animal model for testing antiviral therapies and candidate
vaccines against HIV-1 available.493–496 As with HIV-1, the first cel-
lular targets for SIV are the CD4+/CCR5+ memory T cells in the
lymphoid tissue of the genital tract and in the GALT, which has
unique anatomical and functional features that make it a major
reservoir for virus sequestration, persistence, and ongoing repli-
cation.196 The percentage of CD4+ T cells in the lamina propria
of the gut of SIV-infected macaques dramatically decreased from
67% in uninfected animals to 6% at 21 days postinfection.158,497–502
Plasma viral loads at the peak of primary SIV infection and
at the set point during chronic SIV infection in macaques par-
allel those observed in HIV-1–infected humans.503 Some ani-
mals maintain high viral loads and quickly progress to AIDS, as
human fast progressors, while others contain viremia spontane-
ously and progress to disease slowly, as HIV-1–infected human
long-term non-progressors, or maintain barely detectable
viral loads without treatment, such as human controllers.504
Several monkey MHC class I haplotypes, such as Mamu-A*01,
Mamu-B*08, and Mamu-B*17, have been identified that cor-
relate with “elite control” of viral loads and relative resistance
of the animals to progression to SIV-induced AIDS,505,506 as
observed in humans with haplotypes HLA-B*27, HLA-B*57, or
HLA B*58 that correlate with long-term control of viremia in
HIV controllers and elite controllers.366,507–513
There is only, however, a limited number of SIV isolates
that can be used to test protection across genetically diverse
virus strains, a major obstacle for any field test of an HIV
vaccine. Also, because of a complete difference between the
SIV and HIV-1 Env antigenicity, the SIV/macaque model
does not allow evaluation of the role of neutralizing anti-
bodies in vaccine-induced protection.514 To alleviate this dif-
ficulty, replication-competent chimeric SIV/HIV viruses,
called SHIVs, were engineered that combine the and
genes from HIV-1 with the and genes from
SIVmac and replicate to high titers in cynomolgus and rhe-
sus monkeys.515,516 Serial passages of these hybrid viruses in
monkeys can lead to the emergence of stable pathogenic SHIV
strains able to induce CD4+ T-cell depletion and AIDS in the
animals, such as SHIV89.6P, an X4 virus,517,518 or SHIVSF162P3, an
R5 virus.519 The relative transmissibility of an R5 clade C SHIV
(SHIV -1157ipd3N4) via different mucosal routes was found to
parallel the relative risk of sexual transmission in humans,
rectal challenge requiring the least virus, followed by vaginal
and then oral routes,520 as observed in humans.521–524
The SHIV/macaque model has been widely used for preclini-
cal testing of candidate HIV vaccines. Paradoxically, however,
highly virulent X4 SHIVs, such as SHIV89.6P, have turned out
to be relatively easy to control by vaccination, casting doubts
on their validity as a model for HIV vaccine efficacy, especially
when the end point is control of infection.195,495,514,525
Monkey models suffer from the fact that a very high dose of
virus is typically used to challenge the animals in vaccine protec-
tion efficacy experiments so as to achieve 100% take of the virus
in the placebo group after a single exposure. These high doses
are by far superior to the amount of virus in natural exposures
in humans.526,527 The difficulty was eliminated by using repeated
low-dose mucosal challenges, after it was shown that low-dose
SIV challenges (10-50 TCID50) resulted in the same viral and
immunological kinetics of infection as a high-dose challenge.528
Another limitation of monkey models stems from the usually low
number of animals used in vaccine experiments, whose results,
therefore, often lack sufficient statistical significance.
Live attenuated vaccines
The observation that -deleted mutants of SIV were attenu-
ated and could serve as a live vaccine to elicit protection against
challenge with pathogenic SIV in rhesus macaques served as a
model in favor of a live attenuated HIV vaccine approach.91,93,529
The result was confirmed by using as a vaccine a molecular
clone of SIVmac that had been attenuated by a 12-bp deletion
in the gene, C8, which provided complete protection against
infection with a full-length, pathogenic molecular clone, J5.92
Deletion of , alone or in combination with deletions in
and in the long terminal repeat region,530 showed no drastic
effect on virus growth in cell culture.
SIV was actually found to establish a lifelong, persistent,
low-grade infection that protects monkeys against disease follow-
ing challenge with pathogenic virus but generally does not protect
them against infection. The actual correlates of protection remain
incompletely deciphered,95 but evidence favors the induction of a
long-lasting CTL response.531,532 Similarly, nonpathogenic SHIV89.6
established persistent infection in monkeys that, following muco-
sal challenge with SIV, provided protection against uncontrolled
virus replication, but not against infection. A clear role for CD8+
T cells in protection was evidenced.356,357 Macaques previously
exposed to low doses of HIV-2 also demonstrated a virus-specific
cellular immune response and were protected against uncontrolled
virus replication when challenged intrarectally with SIV.533
SIV is actually not totally attenuated and may still
cause AIDS when administered orally to infant monkeys.534 It
also could revert to full-size virulence in some animals. 535
Additional deletions or mutations could further attenuate its
virulence but at the expense of protective efficacy. Efforts to
make the virus safer included the construction of replication-
competent SIV viruses that expressed the human IFN-
gene536 or the herpes simplex thymidine kinase gene.537
Still, the prospect of using a live attenuated HIV vaccine
in humans has deep safety concerns owing to the high muta-
tion rate of the virus, and, thus, the approach was not tested
in humans. A few persons infected with HIV-1 for long periods
without any sign of AIDS were found to be infected with -
deleted viruses,538,539 but these persons eventually experienced

late decline in CD4 T-cell counts,540,541 and their disease slowly
evolved toward AIDS.542 As a result, there is no development of
live attenuated HIV vaccines ongoing as a strategy of prevention
of HIV-1 infection.
Whole inactivated vaccines and virus-like particles
The best way to inactivate HIV or SIV without destroying the
antigenicity of the viral envelope glycoproteins was found to be
through mild oxidation with aldithriol-2 to remove the essen-
tial Zn2 + ions from the Gag and Int proteins74 or alkylation with
N-ethyleneimine543 or through UV inactivation of the viral RNA
with the help of psoralen.544 The immunogenicity of the resulting
preparations was modest, perhaps owing to the low number of gly-
coprotein spikes per virion.55 Immunization with an inactivated
SIV vaccine was unable to protect monkeys against SIV infection,
but the animals showed decreased viremia after challenge.545
When coexpressed in insect cell cultures using a baculovirus
vector or in mammalian cell cultures using a poxvirus vector,
the HIV or SIV Gag and Env proteins spontaneously assembled
to form virus-like particles (VLPs) that were made only of the
viral core and envelope proteins.546,547 SIV VLPs were tested as
a vaccine in NHPs by a variety of routes, including the nasal
route,548 but their immunogenicity was found to be modest. No
broadly neutralizing antibody was elicited by HIV-1 VLPs in
mice or guinea pigs, even when the sequence of gp41 or that of
the gp41 MPER was included in the construction.549
Hybrid VLPs were also engineered by fusing the MPER from
gp41 to the N-terminus of the hepatitis B surface antigen550 or
to bovine papillomavirus L1 capsid protein, which spontane-
ously forms VLPs.551 Peptides with the sequence of the gp120
V3 loop or of a known CTL epitope, P18, were also fused to
BPV-L1 VLPs.552 None of the resulting chimeric VLPs were
found to elicit more than low titers of HIV-specific antibodies,
and none could elicit HIV-1 neutralizing antibodies.
Subunit vaccines
Subunit protein vaccines were initially developed based on
monomeric HIV-1 gp120 or gp140 (gp160 deleted of the trans-
membrane and intracytoplasmic domains of gp41) that were pre-
sented in their soluble form or as related synthetic peptides (V3
loop) using alum as an adjuvant. These vaccines were found to
induce neutralizing antibodies that targeted the V3 loop in gp120
and protected chimpanzees against challenge with a homolo-
gous or near-homologous HIV-1 X4 strain, but not against chal-
lenge with a distant virus strain or with a primary R5 HIV-1
isolate.553–558 Passive immunization with a monoclonal antibody
targeting the V3 loop was also shown to protect chimpanzees
against homologous X4 virus challenge.559 Similarly, in the SHIV/
macaque model, vaccines based on gp120 were able to provide
protection against homologous X4 SHIV challenge,560 but not
against challenge with a heterologous SHIV with a different enve-
lope specificity.561 In contrast, immunization of macaques with
SIVmac gp120 formulated with a variety of adjuvants, liposomes,
or immunostimulating complexes562 was unsuccessful at provid-
ing protection against SIV challenge.563 SIVmac is an R5 virus.
Recombinant soluble HIV-1 envelope proteins have been the
target of intensive investigation, as one of the major obstacles
in HIV vaccine development remains the inability of current
immunogens to elicit broadly neutralizing antibodies. Numerous
approaches have been explored,31,564 including the following:
– Trimeric gp140 molecules (gp120 + the ectodomain
of gp41) stabilized by the addition of heterologous
trimerization domains at the C-terminus of the gp41
sequence565,566 and similar trimers with a deletion of the
hypervariable V2 loop to better expose the neutralization
Human immunodeficiency virus vaccines 51 1109
epitopes that overlap the CD4-binding site.308,567,568
Systemic immunization or a combination of systemic
and mucosal immunization with the V2 gp140 subunit
vaccine emulsified with water-in-oil adjuvant MF59
protected female rhesus macaques against intravaginal
challenge with a homologous SHIV.569
– Trimeric gp140 molecules internally stabilized by a
disulfide bond between gp120 and gp41 (SOS proteins).
Their immunogenicity was tested using priming with a
DNA plasmid encoding the SOS protein and repeated
boosting with the purified SOS trimers. This regimen
resulted in high titers of neutralizing antibodies against
virus strains of the same clade but only low levels of
cross-clade neutralizing antibodies.570,571
– Oligomeric gp140 molecules covalently coupled to
synthetic mimics of the CD4 receptor or “single-chain
gp120 derivatives” covalently linked to the first two
domains of the human or simian CD4 molecule to induce
in the Env molecule the same conformational changes that
take place at the time of virus entry and to expose potential
neutralization epitopes that overlap the coreceptor-binding
site.572,573 The single chain antigen was found to elicit
in macaques neutralizing antibodies to CD4-induced
epitopes,18,574 which were able to accelerate control of
viremia on challenge with SHIV162P3.575
– A cocktail of Env immunogens derived from
globally prevalent HIV strains or, more simply, from
representatives of each of the different clades576–578 in
the hope that multivalent Env vaccines will broaden the
spectrum of neutralizing antibodies.579,580
– Fully synthetic genes derived from theoretically defined
consensus or “ancestral” HIV envelope sequences262–264
that were found to be superior to many wild-type
immunogens for inducing cross-subtype T- and B- cell
immune responses and antibodies that neutralized a
spectrum of HIV Env pseudoviruses.265 The M-group
CON-S and a clade B-based CON-B Env protein581,582
each elicited antibodies with good titers and breadth of
responses against tier-1/HIV-1 viruses but not against
isolates that are more difficult to neutralize.266
– Mosaic antigens with sequence derived from in silico
algorithms to closely match the sequences of natural HIV-1
strains worldwide and maximize overlap between them.267,268
Despite these many efforts and unabated attempts at designing
mimics of the viral envelope spike or specifically tailored Env
preparations,315 the quest for immunogens able to efficiently elicit
high-titer broadly neutralizing antibodies has remained unsuc-
cessful,66,345,347–349,583 and continued efforts to find an optimal
immunogen design and delivery are ongoing (see later text). The
formulation of gp120 or gp140 under the form of immunostimu-
lating complexes with saponin derivative Quil A or in combina-
tion with water-in-oil emulsions containing QS21 or muramyl
dipeptide derivatives or with specific adjuvant formulations such
as MF59 resulted in the induction of high gp120-binding anti-
body titers with neutralizing activity against closely matched X4
virus strains but not against primary HIV-1 isolates.584–587
Induction of 2 F5- or 4E10-like fusion-blocking antibodies by
immunization with recombinant oligomeric gp41 molecules or
formulations of the gp41 MPER has similarly been mostly a fail-
ure. Displays of constrained 2 F5 epitope in the context of various
protein scaffolds induced high antibody titers against the primary
amino acid sequence of the epitope but failed to elicit neutraliz-
ing antibodies against HIV-1 isolates,549,588,589 confirming the need
to adequately mimic the native conformation of the epitope590,591
and the likely importance of the lipid membrane and hydropho-
bic context for the binding of 2 F5 and 4E10 antibodies to their
respective epitopes.592–594 The challenges of eliciting broadly neu-
tralizing antibodies against HIV are further detailed subsequently.
1110 SECTION THREE Vaccines in development and new vaccine strategies
Subunit vaccines based on the accessory protein Tat have also
been developed in the hope that immune responses directed at
antigens such as Tat that are expressed very early in the virus
replication cycle might lead to containment, if not abortion, of
infection.595,596 Eight genetic variants of Tat have been identi-
fied.597 Anti-Tat antibody and Tat-specific CTLs have been cor-
related with reduced viremia and slow progression to AIDS.598,599
Conflicting results have been reported following immuniza-
tion of rhesus macaques with Tat protein, formalin-inactivated
Tat (Tat toxoid), or Tat peptides. These vaccines elicited little
or no protection against challenge with SIVmac239, SHIV33,
or SHIV89.6P.600–603 In contrast, immunization of cynomolgus
macaques with Tat protein or DNA-encoding Tat elicited con-
trol of infection following challenge with SHIV89.6P.604–607 A Tat-
Nef fusion protein that was combined with a gp120 subunit
vaccine in AS02 adjuvant was found to prevent the development
of disease in rhesus macaques for at least 2.5 years following
challenge with SHIV89.6P, especially when presented as a DNA
vaccine formulation.608,609 HIV Tat vaccines might, thus, show
better efficacy when combined with other HIV antigens.610–614
DNA vaccines
In view of the lack of success of vaccine designs to elicit
broadly neutralizing antibodies for prevention of HIV-1 infec-
tion,342 HIV vaccine development has turned toward vaccines
that would induce T-cell responses, especially a CD8+ CTL
response aimed at controlling HIV infection. Vaccines that
stimulate the T-cell arm of the immune response are not
expected to protect against infection, but rather to protect
against appearance of disease by reducing viral load, slowing
CD4+ T-cell decline, and preventing, or at least retarding, the
occurrence of symptoms.615 Naked DNA vaccines and live vec-
tored recombinant vaccines have been developed based on this
strategy.36–38,616 Reduction of viral loads in HIV-infected persons
should also result in lower probability of virus transmission to
seronegative partners.
DNA vaccines (also referred to as genetic vaccines) are bac-
terial plasmids that carry the gene(s) encoding one or several
desired antigens and that, once injected into the muscle or
the dermis of an animal, express the desired antigen(s) in situ,
leading to induction of an antigen-specific cellular immune
response.617–620 However, the immune potency of DNA vac-
cines in humans and NHPs has generally been mediocre,
inducing weak Th-cell responses and little if any antibody and
CD8+ T-cell response.621,622 Improved immunogenicity was
obtained through the use of synthetic genes with optimized
codons;623 the insertion of engineered B7 costimulatory mol-
ecules, the coexpression of cytokines such as IL-12, IL-15, or
GM-CSF to serve as vaccine adjuvants; and the simple repeti-
tion of vaccine injections three successive times,624–628 which
resulted in greatly enhanced expansion of polyfunctional
memory CD8+ T cells. A DNA vaccine expressing mosaic
sequences of potential T-cell epitopes from HIV-1 M group
Env was also found to induce CD8+ T-cell responses with sig-
nificantly greater breadth than similar vaccines with natural
sequences.269,270,392
Among strategies for further improvement of DNA vac-
cine potency629 are novel means of vaccine delivery, such as
needleless injections (the Biovector device);630,631 formulation
of the DNA onto microparticles, as done using cationic mic-
roparticles,632 or on gold beads administered by gene gun;633,634
delivery of the DNA transcutaneously with the Nanopatch
microprojection device;635–637 and the use of electroporation
and an electroporation device.638–642 Combined effects of IL-12
and electroporation greatly enhanced the potency of DNA vac-
cination in macaques.639 Optimized plasmid DNAs encoding
the majority of SIV proteins and delivered to macaques by
electroporation elicited strong cellular and humoral responses
against the vaccine Gag, Pol, Env, Nef, and Tat antigens and
decreased viremia in the acute and chronic phases of infection
following challenge with SIVmac.643 Similar protection results
were observed against a SHIV clade C challenge using DNA
plasmids expressing SIV Gag and Pol and HIV-1 clade C Env
that were delivered by electroporation.644
A topical plasmid DNA vaccine formulated in a nanoparticle
with polyethylenimine-mannose and glucose, DermaVir, was
developed to deliver DNA-encoded antigens into Langerhans
cells.645,646 DermaVir formulated with HIV-1 Gag plasmid in the
presence of an IL-15-encoding plasmid significantly enhanced
the induction of Gag-specific central memory T cells, as mea-
sured by IFN- ELISPOT. 647
Still, the best immunogenicity results with DNA vaccines
have been obtained using a prime-boost strategy, ie, deliver-
ing the DNA vaccine as a prime, later to be followed by a live
recombinant vaccine booster immunization.648 Thus, immu-
nization of macaques with a DNA-Env vaccine followed by a
polyvalent Env formulation was able to elicit a more diverse
antibody response, with higher antibody avidity and improved
cross-clade neutralizing activity, albeit only against relatively
easy to neutralize isolates of HIV-1, together with cross-subtype
T-cell responses.
Live recombinant (vectored) vaccines
Live recombinant vaccines are made of a live viral or bacterial
vector that was engineered to express a variety of exogenous
antigens in the cytoplasm of target cells, in this case HIV-1 or
genes. The antigens are eventually broken down in the protea-
some of the transduced cell and presented as externalized pep-
tides in the context of MHC class I molecules, thus priming
for a CD8+ T-cell response. Alternatively, antigen-presenting
cells can phagocytose transduced cells that become apop-
totic and present antigenic peptides to CTLs through cross-
presentation. If the vaccine contains too few antigens, virus
escape mutants with mutations in T-cell epitopes can readily
emerge.505 Vaccine impact on control of viral loads has been
the best with vaccines that express multiple HIV-1 proteins
including Env.649,650
Numerous vectors have been engineered into live recombi-
nant vaccines651–674 (Table 51-2). Most of these vectors had pre-
viously been made replication-defective by deletion of essential
genes, such as the E1A gene in adenoviruses or a series of genes
in the New York vaccinia virus strain (NYVAC). Other vectors
are basically unable to replicate in primates, such as canarypox
(ALVAC), fowlpox, or the attenuated vaccinia virus strain modi-
fied vaccinia Ankara (MVA). Replication defectiveness allows
for greater safety, but also results in greatly decreased immu-
nogenicity of the candidate vaccine, necessitating a consider-
able increase in the dose (number of particles of recombinant)
of the vaccine.
One of the major challenges for live recombinant vaccine
immunogenicity is the immune response to the vector, which
dampens the take of the vaccine and subsequent expression of
the vectored antigens: This not only implies refraining from
using the same vectored vaccine for repeated immuniza-
tions, but also applies to the situation of preexisting immune
responses to the vector in the general population. This is espe-
cially true for Ad5, against which the human population shows
a high prevalence of neutralizing antibodies.
Live recombinant vaccines mostly elicit T-cell responses,
especially CD8+ T-cell responses, but different vaccine vectors
delivering the same antigen can elicit CD8+ T cells with differ-
ent fine specificities.675 It was recently observed that multiple
innate immune pathways may contribute to the immunogenic-
ity of the Ad5-vectored vaccines.676 MVA recombinants were

more than 3 years later, in 2002, amid media coverage, fears
of vaccine side effects, and declining sales. Several articles
have reviewed the factors involved.152–155 Some of the reasons
included the low risk of Lyme disease in most parts of the
country, the need for booster injections every or every other
year, the relatively high cost of this preventive approach com-
pared with antibiotic treatment of early infection, and a tepid
endorsement by the public health community. However,
the major reason was probably the theoretical, although
never proved, concern that in rare cases, vaccination might
trigger autoimmune arthritis. The Lyme disease countercul-
ture, which promotes an antiscience ideology of long-term
antibiotic therapy for “chronic Lyme disease”,32 latched on
to this idea. Although most patients with so-called chronic
Lyme disease have other illnesses, particularly pain and
fatigue syndromes,32 they said that vaccination made their
Lyme disease worse. One law firm filed a class action suit
in December 1999, and individual suits began to flood
in.153 The bad publicity greatly reduced sales. In July 2003,
GlaxoSmithKline, now the maker of LYMErix, settled the
class action suits.153 The final agreement included over $1
million in legal fees for prosecuting attorneys, but it provided
no financial compensation to the vaccine “victims”.
Future considerations
OspA Vaccination for Global Use
The OspA Lyme disease vaccines developed in the 1990s
should be considered first-generation vaccines. A goal of future
vaccines would be to provide longer-term protection and
broader protection against variant strains of . The
development of a Lyme disease vaccine for use in Europe has
been more difficult than in the United States because all three
pathogenic species of cause the infec-
tion there.
Baxter Healthcare in Vienna, Austria, has developed a
multivalent OspA Lyme borreliosis vaccine (mv rOspA LB
vaccine), containing three recombinant OspA antigens,
which is designed to provide protection against almost all
strains associated with human dis-
ease worldwide. For this development, knowledge of protec-
tive OspA epitopes was used to design recombinant OspA
molecules combining the protective properties of two OspA
antigens in a single molecule. In a proof-of-principle study,
a single recombinant OspA containing protective elements
from two different OspA serotypes (1 and 2) was able to
induce antibody responses that protected mice against infec-
tion with either (serotype-1) or
(OspA serotype-2).156 In mice, as little as 0.03 g of
this rOspA, when administered in a two-dose immunization
schedule with aluminum hydroxide as adjuvant, was suffi-
cient to provide protection against the species targeted. In
this investigation, the sequence of the epitope
(OspA163–175) that has partial sequence homology with human
LFA-1 was replaced with a corresponding (non-hLFA-1)
sequence from a serotype-2 OspA molecule. Others have
noted that recombinant OspA, which lacks the predicted
cross-reactive epitope, is able to elicit OspA antibodies in
mice that are effective in generating protective type-specific
immunity.157 Active protection studies with Baxter's multi-
valent recombinant OspA vaccine (unpublished data) dem-
onstrated significant protection of immunized mice relative
to controls against challenges with virulent
and strains. In addition,
the vaccine induced antibodies to OspA epitopes expressed
on the surface of live as well as growth-inhibiting
Lyme disease vaccines 52 1131
antibodies. These functional antibodies recognized not only
the six OspA types targeted by the vaccine but also other
genospecies of including strains of
and
further indicating that the vaccine has the potential
to prevent Lyme disease worldwide. Phase I/II clinical trials
to evaluate the safety and immunogenicity of the mv rOspA
LB vaccine, with and without adjuvant, and to identify the
optimal dosage level and formulation is in progress. Antibody
persistence and the response to a booster vaccination will be
investigated in a subset of subjects.
Other altered OspA proteins
Other altered OspA proteins have been developed as vaccine
candidates. In one study, the entire 255-amino-acid ectodo-
main of OspA was inserted into the major B-cell epitope of
the hepatitis B virus (HBV) capsid protein.158 This fusion pro-
tein induced an antibody response to the LA-2 protective epit-
ope similar to that of recombinant lipidated OspA. In another
study, a C-terminal fragment of OspA was generated that lacked
approximately 45% of the residues from the N-terminus.159
Although this manipulation reduced conformational stabil-
ity and vaccine efficacy compared with that of the full-length
protein, the immunogenicity of the fragment was restored by
replacing amino acid residues in buried salt-bridges with resi-
dues that promoted hydrophobic interactions.
OspC vaccine candidates
In an effort to provide protection against the range of European
strains, Baxter Healthcare in Vienna initially researched a mul-
tivalent OspC vaccine. A potential advantage of this approach
is that strains express OspC on trans-
mission and entry into mammalian hosts.160 Therefore, anam-
nestic immune responses in the host could provide protection,
rather than depending solely on the maintenance of high lev-
els of circulating OspA antibodies, which kill spirochetes in
the tick while it feeds. However, a major disadvantage is the
greater variability of OspC compared with OspA. The com-
pany developed a 14-valent OspC vaccine, which was given in
three 100- or 140- g doses. In phase I trials, the major prob-
lem was vaccine-induced erythema and swelling at the injec-
tion site, particularly after the third injection, which occurred
in up to half of persons. Consequently, this approach was
discontinued. Subsequently, one research group performed a
series of comprehensive seroprofiling studies to select OspC
types that have the most cross-reactive immunodominant epi-
topes. They found that combinations of as few as two OspC
proteins identified all patients who had OspC antibodies.161
Another research group developed a tetravalent OspC-based,
recombinant, chimeric vaccinogen, with linear epitopes pre-
sented by OspC types A, B, K, and D.162 When mice were
vaccinated with this construct, they developed complement-
dependent bactericidal antibodies against strains expressing
each of the OspC types.
OspA immunization of rodent reservoirs
Another novel approach is being researched in which mouse
reservoirs of are immunized with OspA. In one
study, wild white-footed mice ( ), the pre-
ferred host for nymphal ticks, were immunized
by subcutaneous injection with either recombinant OspA or
a negative control antigen in experimental or control grids.163
In two large-scale experiments, nearly 1,000 mice were vacci-
nated. Over both field trials, the authors estimated that about
69% of the mouse population was immunized more than once,
about 55% of all mice in each experimental group seroconverted
 Malaria vaccines 53 1135

Vaccines targeting the sexual stages well tolerated and immunogenic in all age groups and to have an
acceptable safety profile.68–71,73–75,80–90 Estimates of vaccine effi-
Transmission-blocking vaccines rely on human antibody and cacy against clinical malaria measured during the pediatric phase
complement taken up by mosquitoes during a blood meal 2 program are shown in Table 53-1. Because of the complexity
(Figure 53-1). Within the mosquito, sexual-stage parasite of falciparum .5malaria
79mi05705
as a vaccine
.6cm .TJ ation
aT with
ationRTS,S
wndemic
should
sest685
theoretically
0 demwredu
0y
surface antigens or mosquito midgut binding sites can theo- and severity of clinical malaria disease.
retically be targeted to prevent successful development of infec-
tious sporozoites.55–57 This type of vaccine would not prevent
disease in a vaccinated person, but it would prevent the fur-
ther spread of malaria.58,59 If effective, such a vaccine could
be particularly useful in areas of low-moderate transmission
intensity or as an important component of a multiantigen
vaccine.

Vaccines in late-stage development

Currently, there are no registered malaria vaccines. A large
number of malaria vaccine candidates have been tested preclin-
ically or in phase 1/2 clinical trials,60–63 but only one is currently
in late-stage development. GlaxoSmithKline Biologicals (GSK),
in partnership with the PATH Malaria Vaccine Initiative,
is conducting a large-scale phase 3 program in sub-Saharan
Africa of the RTS,S vaccine.64 The program was designed in
consultation with the WHO to demonstrate that RTS,S can
provide significant durable protection against clinical malaria,
including severe malaria, in infants and young children
when administered via the existing Expanded Programme on
Immunization (EPI).
RTS,S was developed by GSK and the Walter Reed Army
Institute of Research (WRAIR) in a collaboration that spanned
more than two decades.65 The vaccine is based on the circum-
sporozoite (CS) protein of , a highly conserved
major sporozoite surface protein that contains B- and T-cell epit-
opes that seem important in protection.66 The CS protein under-
goes important conformational and structural modification as it
transits from the mosquito to the mammalian host, permitting
adherence and invasion of hepatocytes.67 RTS,S consists of a chi-
meric virus-like particle protein construct expressing 19 NANP
repeats of the central region and the entire C-terminal flanking
region (amino acids 207-395) fused to the hepatitis B virus sur-
face antigen, HBsAg.68 The fusion protein is coexpressed in yeast
cells with HBsAg, yielding a product that is composed of 25% of
the CS-HBsAg fusion protein and 75% HBsAg.
A series of pivotal clinical trials conducted at WRAIR that
incorporated an experimental malaria sporozoite challenge
demonstrated an absolute requirement for adjuvantation of
RTS,S with the immunostimulants monophosphoryl lipid
A and QS21 ( , a purified saponin deriva-
tive).68–71 Two related adjuvant formulations were studied subse-
quently during phase 2, AS02 in which the immunostimulants
are combined with an oil-in-water emulsion, and AS01, in
which the immunostimulants are combined with liposomes.72
Based on superior immunogenicity, an equivalent safety pro-
file and a trend toward better efficacy, the AS01 formulation
was ultimately selected for phase 3.71,73–75 RTS,S induces high
levels of antibodies against CS that inhibit sporozoite infection
of liver cells and high levels of antibodies against HBsAg. It also
induces CMI responses against CS that promote the destruc-
tion of schizont-infected liver cells.76–79 By preventing or signifi-
cantly decreasing the load of parasites emerging from the liver,
vaccination with RTS,S should theoretically reduce the rates
and severity of clinical malaria disease.42 It should be noted
however that there is yetyet no
no identified
identified immunological
immunologicalcorrelate
correlate
of RTS,S-induced protection.
protection.
The RTS,S vaccine was studied in more than 4,000 subjects
before phase 3, including malaria-naïve and malaria-experi-
enced adults, as well as young children and infants in malaria-
endemic settings. The vaccine has been shown to be generally
 1136
SECTION THREE
Estimates of VE Reported During Phase 2 Trials Following Three Doses of RTS,S Administered at 1-Month Intervals to Children and Infants Living in Malaria-Endemic Regions of
Sub-Saharan Africa

CI, confidence interval; ND, not done; VE, vaccine efficacy.

*Staggered with Expanded Programme on Immunization (EPI) vaccines.
†
Coadministration with EPI vaccines.

and viral vectors,62,96 but while there are numerous reports of
such vaccines conferring protection in animal models, to date,
none have been shown to confer robust and reproducible pro-
tection against falciparum malaria in humans.92 Following the
lead of HIV vaccine developers, a number of prime-boost reg-
imens are in development,97–100 the most advanced of which
uses a recombinant chimpanzee adenovirus 6 vector expressing
multiple preerythrocytic-stage antigens (ME-TRAP) followed by
boosts with modified vaccinia Ankara (MVA) vector expressing
the same antigens.101 Experimental challenge studies indicate
that this vaccine can induce partial protection against infection
and it is currently in exploratory field trials in African children.
Vaccines against are also beginning to enter clinical
development.102,103 The leading vaccine candidates have targeted
the antigens for which homologous protection has been seen
in human or animal models of falciparum malaria, including
the vivax forms of MSP-1, CSP, TRAP, AMA-1, and the Pvs
25 and Pvs 28 ookinete antigens. In addition, several unique
vivax asexual-stage antigens, including the Duffy binding pro-
tein have received considerable focus.104 As the life cycle
cannot currently be maintained in the laboratory, developing a
vivax sporozoite challenge model for use in vaccine challenge
trials is much more complex. Fortunately, good progress in
being made in this area, which is critical for development of
vivax vaccines.105
Another innovative strategy relies on the large-scale manual
dissection and purification of radiation-attenuated sporozo-
ites from mosquitoes and delivering them by intramuscular
or subcutaneous injection.106 The basis of this vaccine strat-
egy is the high level of protection seen in human volunteers
immunized by the bites of large numbers of malaria-infected
irradiated mosquitoes, but the translation of this approach to
a more practical delivery method has so far yielded disappoint-
ing results in experimental challenge studies in humans.107
The reasons for this unexpected failure in the clinic are not
Access the complete reference list online at http://www.expertconsult.com
33. Good MF, Doolan DL. Malaria vaccine design: immunological
considerations. Immunity 33:555–566, 2010.
58. Targett GA, Greenwood BM. Malaria vaccines and their potential role in the
elimination of malaria. Malar J 7(suppl 1):S10, 2008.
62. Bruder JT, Angov E, Limbach KJ, et al. Molecular vaccines for malaria. Hum
Vaccin 6:54–77, 2010.
63. Anders RF, Adda CG, Foley M, et al. Recombinant protein vaccines against
the asexual blood stages of . Hum Vaccin 6:39–53,
2010.
64. Birkett AJ. PATH Malaria Vaccine Initiative (MVI): perspectives on the status
of malaria vaccine development. Hum Vaccin 6:139–145, 2010.
66. Cohen J, Nussenzweig V, Nussenzweig R, et al. From the circumsporozoite
protein to the RTS, S/AS candidate vaccine. Hum Vaccin 6:90–96, 2010.
Malaria vaccines 53 1137
clear, but could include lower than expected viability of spo-
rozoites injected by syringe, failure of sporozoites injected
directly into tissues to properly home to lymph nodes and
the liver, and unidentified factors introduced during sporo-
zoite purification and processing that lead to a significantly
loss in vaccine potency. To eliminate the requirement for
irradiation attenuation, genetically attenuated sporozoites are
being developed, and whole-parasite blood-stage vaccines are
also being explored,108,109 but these strategies face additional
challenges in establishing a balance between safety from
malaria breakthroughs and induction of adequate immunoge-
nicity. These strategies remain highly experimental and face
enormous hurdles for large-scale manufacturing, formulation,
and delivery.110
In the midterm (5-10 years), improvements on RTS,S may
be feasible by using it in a prime-boost strategy with vaccines
that specifically prime for better cell-mediated responses.111
A recombinant attenuated adenovirus 35 expressing the CS
protein given as a priming dose is under clinical evaluation to
determine whether it can confer higher levels of vaccine effi-
cacy than RTS,S/AS01 alone. Other prime-boost regimens can
be envisioned, including, for example, combining RTS,S/AS01
with MVA, measles, or other adenovirus vectors expressing the
CS protein or other preerythrocytic antigens.101
Finally, a global scientific effort is underway to understand
what will be necessary to eliminate and eventually eradicate
malaria. This dialogue has led to the concept of "vaccines that
interrupt malaria transmission" (also called VIMT), which
may include components from one or more of the parasite
stages described in this chapter.112,113 It is likely that a signifi-
cant proportion of future funding for malaria vaccine develop-
ment will focus on this important long-term goal and that such
vaccines are developed within a framework that will position
their use in context with other new tools being developed to
support the malaria eradication agenda.
91. Vekemans J, Ballou WR. malaria vaccines in
development. Expert Rev Vaccines 7:223–240, 2008.
93. Schwartz L, Brown GV, Genton B, et al. A review of malaria vaccine clinical
projects based on the WHO Rainbow Table. Malaria J 11:11, 2012 doi:
10.1186/1475-2875-11-11.
95. World Health Organization. Initiative for Vaccine Research: The Rainbow
Tables. 2010. http://www.who.int/vaccine_research/links/Rainbow/en/index.
html. Accessed May 2011.
101. Hill AV, Reyes-Sandoval A, O'Hara G, et al. Prime-boost vectored malaria
vaccines: progress and prospects. Hum Vaccin 6:78–83, 2010.
 SECTION THREE: Vaccines in development and new vaccine strategies
Noninfectious disease vaccines
54 Philippe Saudan
Martin F. Bachmann
Introduction
Prophylactic vaccination against viral and bacterial pathogens is
one of the main reasons for the roughly 30-year increase in life
expectancy during the twentieth century, and it represents the
most effective intervention in medical history.1 With the con-
trol of many infectious diseases, our aging population is now
most susceptible to chronic diseases such as Alzheimer's, can-
cer, diabetes, or cardiovascular diseases, which account for more
than half of the global disease burden.2 In the last decade, pas-
sive immunotherapy with monoclonal antibodies (mAb) target-
ing self proteins has proven to be effective for the treatment
of different chronic diseases. However, the need for frequent
administrations, high cost and the induction of neturalizing
antibodies are major hurdles for their widespread use. Hence, it
is tempting to replace some of the passive vaccination strategies
by active vaccination to prevent and/or treat chronic diseases.
The therapeutic B-cell vaccines against noninfectious dis-
eases described in this chapter are designed to induce strong
antibody responses. With this respect, they are similar to most
conventional antiviral or antibacterial vaccines, which also
mostly rely on induction of protective antibody responses.3,4
General principles underlying their development and an over-
view of targets and therapeutic areas in which they have been
explored are provided in this chapter. In light of the unattain-
able task of providing a full summary on the subject, we have
decided to focus on vaccines that have been studied in humans.
There also is a vast literature on T-cell–based cancer vaccines,
which are discussed in Chapter 42.
Tolerance, T-cell help, and how to induce
antibodies against self
Tolerance
Self antigens generally do not induce antibody responses even
upon administration with adjuvants because of a set of check-
points the immune system has evolved to avoid autoreactive
responses. This unresponsiveness toward self, often called toler-
ance, can occur on the T- and/or B-cell level. Anti-self immune
responses are prevented by different mechanisms including
deletion (elimination of cells), anergy (functional silencing), or
ignorance (immune-competent lymphocytes are present in the
periphery but are not stimulated, since they do not encounter
the self antigen in an immunogenic context). Whereas autore-
active T-cells are efficiently deleted or silenced in the thymus,
B-cell tolerance is mainly induced in the bone marrow. B- and
T-cell tolerance also occurs in the periphery. A key feature
for the deletion of B cells is cross-linking of the B-cell recep-
tor (BCR). Hence, cells specific for highly expressed membrane
proteins are particularly efficiently eliminated while B cells spe-
cific for soluble antigens typically escape negative selection and
remain ignorant.5 In contrast, because of the expression of most
antigens in the thymus and the transport of soluble antigens to
the thymus, T-cell tolerance, in particular Th-cell tolerance, is
a much more efficient process. Thus, while most autoreactive
T cells are eliminated, a largely normal B-cell repertoire exists
for secreted self antigens. However, since naïve B cells critically
depend on T-cell help to proliferate, undergo isotype switching,
and differentiate into long-lived plasma cells, deletion or silenc-
ing of autoreactive T cells is sufficient to avoid self-specific
B-cell responses and maintain B-cell unresponsiveness.
Nevertheless, low-affinity autoreactive T cells can escape the
thymic and peripheral tolerization processes and may be activated
by administration of the antigen in the presence of strong adjuvants
such as complete Freund's adjuvant.6 One example is A -specific
T cells induced upon immunization of Alzheimer`s patients with
aggregated A peptides in a saponin adjuvant (QS21). While a rea-
sonably high proportion of patients produced A -specific IgG anti-
bodies, a small fraction of patients (about 6%) developed aseptic
meningoencephalitis, apparently due to induction of A -specific
T cells. Although not formally proven, it is generally believed
that A -specific CD4 T cells lead to local inflammation in the
brain upon stimulation with their cognate antigen A . Therefore
bypassing or circumventing T-cell tolerance, rather than breaking
it, should be the method of choice for the induction of antibodies
against self antigens. This is best achieved by fusing a strong for-
eign T-cell epitope to the antigen or, even more effective, conjugat-
ing the antigen to a foreign carrier molecule.
Induction of B-cell responses
For a B cell to generate a long-term response, two signals are
needed—activation of the BCR and stimulation by antigen-
specific Th cells. Through covalent linkage of an antigen to a
carrier molecule, B cells expressing a BCR specific for the anti-
gen will get carrier-specific T-help (Figure 54-1).7 Carrier mol-
ecules successfully used for marketed glycoconjugate vaccines
are pathogen-derived proteins like diphtheria toxoid (DT), tet-
anus toxoid (TT), meningococcus B outer membrane protein
C (OMPC), or protein D. Other clini-
cally tested carriers include toxin B,
exoprotein A, and keyhole limpet hemocyanin (KLH). Recently,
virus-like particles (VLPs) have gained considerable interest
for vaccines against noninfectious diseases. They share many

How B cells and TH cells collaborate and a way to
circumvent TH cell tolerance. A, After immunization with a foreign
antigen, specific B cells bind the antigen, inducing an initial activation
signal. In addition, B cells endocytose the antigen and display
antigen-derived peptides on their major histocompatibility complex
(MHC) class II molecules. B cells are not usually able to activate
naive TH cells, and dendritic cells are an essential intermediate in
the activation of T cells. Dendritic cells also take up the antigen and
process it for presentation in association with MHC class II molecules.
In contrast to B cells, dendritic cells are able to trigger the activation
of naive TH cells. These activated TH cells recognize antigen-derived
peptides on B cells, leading to successful T- and B-cell collaboration:
B cells proliferate, produce antibody and undergo isotype-switching.
B, In the absence of TH cells due to tolerance, antibody responses
remain abortive. C, It is possible to circumvent TH-cell tolerance by
coupling the self-antigen to a foreign protein or peptide carrier. B cells
specific for the self-antigen are able to take up the antigen plus the
linked carrier and present carrier-derived peptides on their MHC class
II molecules. As there is no TH-cell tolerance to the carrier protein,
a regular antibody response against the self-molecule will now be
mounted. (Image and figure legend is reproduced with permission of
Bachmann and Dyer7)
Noninfectious disease vaccines 54 1139
of the features of whole viruses but are not infectious. Their
particulate nature, their highly ordered repetitive structure, and
the Th cell epitopes are the source of their remarkable immuno-
genicity.7–11 Bacteriophage Q -derived VLPs, which are readily
produced in large quantities under good manufacturing prac-
tices in , have been extensively tested as immu-
nological carriers in animals and humans. Using chemical
cross-linkers, antigens can be directionally conjugated to their
surface and presented in a highly ordered fashion to B cells. As a
result BCRs are efficiently cross-linked, a key feature for induc-
ing potent immune response.12 Moreover, during self assembly
in , host RNA, which is a potent TLR7/8 agonist, is pack-
aged into the VLPs and acts as a built-in, adjuvant-enhancing
B-cell response and inducing class switching.13 In light of these
virtues, bacteriophage VLPs have been successfully used as car-
riers to induce potent immune responses against self molecules
and haptens in various preclinical14–22 and clinical studies.23–26
Longevity of the response and general safety
considerations
Ideally high antibody titers against a given target would be main-
tained by a classical boosting schedule. In contrast to classical
vaccines, where the memory response is boosted upon subse-
quent exposure to the pathogen, patients suffering from a chronic
disease or drug of addiction are not naturally exposed to the anti-
gen in an immunogenic form. This is because self antigens are
seen by B cells in the absence of specific Th (which has been
delivered by the carrier). Hence, regular booster immunization
may be required to maintain antibody titers at therapeutic lev-
els. The waning of the antibody response in the absence of con-
tinued boosting also has an advantage from a safety perspective,
particularly for the targeting of self molecules. In this case lifelong
neutralization may not be desired or required, and therapy could
be halted or at least passively stopped upon improvement of the
health conditions or in cases of antibody-mediated adverse events.
For B-cell vaccines different toxic effects may arise and need
to be addressed in detail in preclinical models and early phase
clinical trials in a target-specific manner.27 Such adverse effects
include generation of immune complexes with soluble target
molecules or binding to self antigens expressed on tissue and
subsequent induction of local inflammation by activation of anti-
body-dependent effector cells. In general, rare proteins may cause
fewer problems than abundant proteins, since only low amounts
of immune complexes can be formed. Soluble proteins are gener-
ally preferred over membrane-bound targets due to the absence
of antibody-dependent cellular cytotoxicity (ADCC), which may
only be observed with membrane proteins. Moreover, many
membrane proteins are receptors, therefore, antibodies may be
antagonistic (prevention of binding to the ligand) or agonistic
(activating the receptor through cross-linking). This further com-
plicates the matter, since it is virtually impossible to exclude the
induction of the undesired agonistic antibody with a vaccination
approach. Finally, safety issues associated with the neutralization
of the target molecule also have to be considered. This impor-
tant aspect is best addressed in studies with target-specific mAbs,
decoy receptors, or receptor antagonists. Valuable information
can also be obtained from rare human deficiencies for the target
molecule and mice rendered deficient for the target molecules.
Targets for therapeutic B-cell vaccination
In accordance with the better safety profile of soluble anti-
gens, most therapeutic vaccination approaches in humans have
employed this approach. Meanwhile 17 self proteins have been
targeted in humans by 26 different vaccines in various indications
(summarized in Table 54-1). Many of these molecules are also
1140 SECTION THREE Vaccines in development and new vaccine strategies
Clinically Tested B-cell Vaccines Targeting Self Antigens
*Opossum human hybrid constant domain of IgE.
-, chemically conjugated, Ang, angiotensin; DT, diphtheria toxoid; inact., inactivated; NSC: Nonsmall cell; Qb, Qbeta; RA, rheumatoid arthritis; SLE, systemic lupus
erythematosus.
targets for small molecule pharmaceuticals or biopharmaceuticals
and as such have been validated as appropriate therapeutic targets.
A second category of B-cell vaccines against noninfectious dis-
eases targets foreign antigens and can be divided into desensiti-
zation therapies for allergies (not discussed here) and vaccines
against drugs of abuse.28,29 The latter aim at inducing antibodies
that sequester the drugs in the blood, thereby preventing posi-
tive reinforcement action in the brain. In particular, vaccination
against smoking cessation has been extensively pursued, mainly
with two vaccines: Nabi's Nicotin-exoprotein A conjugate (NicVax)
partnered with GlaxoSmithKline and Cytos Biotechnology's
Nicotin-Qb conjugate (NicQb or Nic002) partnered with
Novartis. In a first phase II study with NicQb, the tertile with the
highest antibody levels was found to have a significant increase in
continuous abstinence from month 2 to 6 compared to placebo.24
However, in a subsequent study performed by Novartis, the pri-
mary end point was not achieved, and an additional phase II study
to improve the efficacy of anti-nicotine immunotherapy is ongo-
ing (NCT01280968). Likewise, Nabi, with its NicVax, reported a
significant increase in continuous abstinence between week 19
and 26 in the highest antibody responders (30% with the highest
antibody titers) compared to placebo.30 This vaccine was tested
in two phase III trials (NCT01178346 and NCT01102114).
Results were, however, negative and Nabi stopped the further
development of the vaccine.
Although the vaccination approach shows great promise from
the early efficacy trials, the remaining challenge is the induction
of sufficiently high antibody titers over an extended period in the
whole-study population.
Alzheimer's vaccines
Alzheimer's disease (AD) is a progressive neurodegenerative con-
dition that affects 26 million people worldwide and is a major
cause for dementia.31 With increased life expectancy this disorder
is becoming a large social and economic burden.32,33 Currently
only symptomatic treatments are available, and in view of the
alarming epidemiological data, therapeutic strategies to prevent,
mitigate, or delay the onset of dementia are extensively pur-
sued.34 One of the most promising therapeutic concepts is active
or passive immunization against amyloid plaques.35
The major pathological hallmarks of AD are plaques formed
by aggregated amyloid- (A 1-42), a degradation product of the
amyloid precursor protein (APP), and neurofibrillary tangles
(NFTs). Biochemical and genetic studies of APP and presenilin
1 and 2 mutations, all of which enhance amyloid deposition

and are associated with early onset familial AD,36 strongly sup-
port the central role A plays in the development of the disease.
In 1999, Schenk et al37 reported that active A immuno-
therapy reduces A pathology in APP transgenic mice. In these
studies, immunization with aggregated A peptides combined
with strong adjuvants resulted in reduced plaque burden37 and
improved mental performance.38 This finding was confirmed
in several transgenic models of AD with active and passive
immunization.39 In light of these results Elan/Wyeth initiated
a number of clinical studies with a vaccine consisting of A 1-42
(AN1792) administered with adjuvant QS21. In phase I trials
A -specific antibodies could be found in more than half of the
treated AD patients without notable adverse events.40
Subsequently, AN1792/QS21 vaccine was tested in a phase II
trial with 372 AD patients to evaluate the efficacy of the treat-
ment. However, treatment in this trial had to be discontinued
prematurely after 6% of the vaccine recipients developed aseptic
meningoencephalitis.41 Subsequent analysis revealed that the side
effects did not correlate with antibody titers.41–43 Moreover, infil-
trating T cells were found in the brain of one diseased patient,
which strongly suggested that A -specific T cells were the cause
for the side effects observed.44–47 Considering that this vaccine did
not contain a carrier molecule, induction of an IgG response had
to rely on breaking self tolerance to induce A 1-42-specific CD4+
T-helper cells (see Figure 54-1), which presumably triggered
meningoencephalitis in some of the patients. This was most
likely achieved by the combination of fibrillary A 1-42 aggregates
with the strong adjuvant QS21. Intriguingly, individuals who had
the highest antibody response in the phase II trial showed a sig-
nificant slower decline in some tests of cognitive function rela-
tive to controls, and despite the severe safety issues observed with
AN1792, this vaccine provided the first clinical evidence that
active vaccination against A may be beneficial for the treatment
of AD.45,48,49 Nevertheless, these cases of meningoencephalitis led
to a significant setback in the field of vaccination against AD. With
the identification of T cells as the main suspects for the observed
inflammatory reaction in the brain, subsequent approaches were
designed to avoid inducing T-cell immunity directed against A .
Second-generation vaccines currently undergoing clinical test-
ing are designed to minimize the risk of inducing autoreactive T
cells (Table 54-2). They are characterized by short N-terminal A
peptides linked to foreign carrier proteins that provide T-help for
the efficient induction of IgG responses. The follow-up program
for AN1792 uses a vaccine consisting of the N-terminal seven
amino acids of A covalently linked to the diphtheria toxin cross-
reactive mutant (CRM197) protein carrier, which is called ACC-
001. Although this peptide is smaller than the minimum length
of a T-cell epitope, strong A 42-specific antibody responses were
induced in transgenic mice that prevented plaque deposition and
improved cognitive function. This compound is currently subject
to intense clinical testing as documented by the nine ongoing or
planned phase II trials.
Likewise Novartis, using VLP technology licensed from Cytos
Biotechnology, has embarked on the development of a vaccine
for AD that targets A . This vaccine, CAD106, consists of A 1-6
covalently coupled to bacteriophage Q VLPs (Qb). Different
preclinical studies revealed that this vaccine induced high anti-
body responses in the absence of additional adjuvant in mice,
rabbits, and monkeys without inducing A -specific T cells.
Immunization with CAD106 prevented amyloid cortical plaque
deposition in both APP23 and APP24 transgenic mouse mod-
els. Moreover, immunization resulted in a reduction of further
amyloid deposition in aged mice with advanced pathology in
the brain.50 Available reports from the first studies in man with
this compound suggest that the vaccine is well tolerated and
induces A -specific immune responses in the majority of the
patients without inducing meningoencephalitis.51–53
Merck & Co. is testing a vaccine consisting of an N-terminal
A peptide presumably conjugated to OMPC (V950). Increasing
Noninfectious disease vaccines 54 1141
doses of V950 formulated in aluminum adjuvant with or without
Iscomatrix adjuvant are being tested in a multistage phase I trial.
Two other vaccines, AC-24 (AC Immune) and UB-311 (United
Biomedical), have achieved preclinical proof-of-concept54,55 and are
in early clinical testing. Whereas AC-24 contains A 1-15 embedded
within a liposomal surface, UB-311 consists of A 1-14 covalently
linked to the UBITh peptide and is administered together with a
mineral salt adjuvant.
AFFiRiS is exploring short peptides mimicking the
N-terminus of A as immunogens to avoid the induction of auto-
reactive T cells. Two of these affitopes (AD01, AD02) apparently
also showed promising preclinical results and were reported to
be safe and well tolerated in phase I trials.56,57 AD02 has pro-
ceeded to phase II clinical testing where the main focus is set on
identifying an optimal dosage for this drug. Exclusive rights to
this program have recently been acquired by GlaxoSmithKline.
The added value of these affitopes (they are not under patent
protection by other groups so there is freedom to operate with
yet another A -specific vaccine) needs to be explored further,
since the aforementioned vaccines using peptides smaller than 8
amino acids all fail to induce T cells because the chosen peptides
are too small to be presented on MHC class I and II molecules.
Moreover, at least seven passive AD immunotherapies are in
clinical development.35 The most advanced mAbs are bapineu-
zumab (Elan/Wyeths) and solanezumab (Eli Lilly's), which are
being tested in several thousand mild-to-moderate AD patients
in phase III trials. With this intense clinical development there
is accumulating evidence that immunotherapy directed against
A 42 has a clinical impact. Key questions to be answered include:
will improvements in cognitive function be clinically relevant,
will the disease course be slowed or even reversed, and will there
be an acceptable risk/benefit ratio. Current anti-A therapies are
developed in AD patients with mild-to-moderate disease. Based
on preclinical studies and the limited data from clinical trials,
progression of the disease may be delayed or halted at best in this
patient population. In fact, A immunotherapy might likely be
most effective in preventing or slowing the disease if patients are
immunized at very early stages of disease onset. Further clinical
validation with particular focus on the safety of this approach,
together with the identification of early bio and genetic markers
for the disease, may help develop these compounds as prophylac-
tic vaccines for the genetically predisposed middle-aged or elderly
and maximize the chance to combat this devastating disease.
Recently hyperphosphorylated tau, the main component of
NFTs, has gained attention as an immunotherapeutic target for
AD, and reduction in insoluble tau and improved behavioral
performance were observed in transgenic mice immunized with
phospho tau peptides.58–62 Hence, vaccines targeting hyperphos-
phorylated tau may enter clinical trials soon.
Hypertension vaccines
Persistent hypertension is one of the risk factors for coronary heart
disease and stroke. The prevalence of hypertension in the adult
population is estimated to be approximately 30%63 and, despite
effective available treatments such as the angiotensin converting
enzyme inhibitors (ACEis) and angiotensin II type I receptor block-
ers, only 30% to 50% of hypertensive patients in the United States
have adequate control of blood pressure (BP).64 This can be attrib-
uted to the absence of symptoms resulting in poor compliance
with prescribed medication. Active vaccination aimed at inducing
long-lived antibody responses against key regulators of BP such as
angiotensin I or II represents an alternative approach to treating
hypertension, and a regimen with a few injections per year would
considerably simplify treatment and improve compliance. The
mediators from the renin-angiotensin-aldosterone system (RAAS)
that regulate BP are attractive targets for active immunotherapy.
It is important to note that BP is regulated at different levels.
Even efficient blocking of the RAAS with ACEis does not result in
 1142
Clinical Trials with B-cell Vaccines Against Alzheimer's Disease. All Studies are Performed in Mild to Moderate AD patients (if not otherwise stated) and Look at Immunogenicity, Safety & Tolerability

SECTION THREE Vaccines in development and new vaccine strategies

plus Different mostly Exploratory Read Out

b
Clinical Trials with B-cell Vaccines Against Alzheimer's Disease. All Studies are Performed in Mild to Moderate AD patients (if not otherwise stated) and Look at Immunogenicity, Safety & T olerability
plus Different mostly Exploratory Read Out.—cont'd

*Extension studies, n.d., not dislosed, dose level only included in Ad1, Adjuvant 1; Ad2, Adjuvant 2; ld, low dose; md, mid dose; IM, Iscomatrix; PET, positron emission tomography; ADAS-Cog, Alzheimer's Disease Assessment Scale-
Cognitive Behavior; DAD, Disability Assessment for Dementia, NTB, Neuropsychiatric Test Battery; MMSE; Mini-Mental State Examination.

Noninfectious disease vaccines

54
1143
 SECTION THREE
hypotension, providing comfort for the safety of this approach for
this indication.65 Several vaccination approaches have been explored
in the last 60 years and been the subject of recent reviews.65–67
Antibodies have the potential to block the RAAS at various levels.
The first target explored in the 1950s was renin, which converts
angiotensinogen to angiotensin I. However, hog renin used as an
immunogen in clinical testing failed to induce a cross-reactive
antibody response against human renin. Years later, a similar
preclinical approach using strong adjuvants, although efficient in
reducing BP in hypertensive marmosets, proved to be lethal due
to induction of autoimmune disease of the kidney. The observed
lethal glomerulonephritis could be attributed to immune com-
plexes formed with renin, which is an abundant membrane-
bound protein of the kidney.68
In contrast to renin, the other investigated targets (angiotensin
I and II) of the RAAS are small soluble peptides without this com-
plication. They are not membrane bound or large enough to cause
immune complex formation. Early clinical testing by Protherics
with angiotensin I conjugated to KLH (PMD3117) formulated
in alum led to the induction of an angiotensin-specific antibody
response. However, antibody levels were not sufficient to block
angiotensin I and reduce BP.69 To address low immunogenicity
this vaccine was subsequently formulated in CovaccineHT adju-
vant and administered three times three weeks apart. This study
had to be terminated due to adverse effects (NCT00702221).
Cytos Biotechnology also set out to target angiotensin II
with a therapeutic vaccine. This vaccine (AngQb) consists of
angiotensin II N-terminal fused to a CysGlyGly linker for the
directional chemical conjugation to Qb VLPs. Vaccination with
AngQb led to a significant BP reduction in spontaneous hyper-
tensive rats and was shown to be safe, well tolerated, and immu-
nogenic in healthy normotensive volunteers in a phase I study.23
Antibody levels peaked 2 weeks after injection and subsequently
declined to reach background levels in a few months. In a sub-
sequent double-blind placebo controlled phase II study AngQb
demonstrated clinical efficacy in mild-to-moderate hyperten-
sive patients.26 Patients receiving 300 g of AngQb formulated
in alum three times (weeks 0, 4, and 12) displayed a reduced
( 9/ 4 systolic/diastolic mmHg) ambulatory daytime BP 2 weeks
after the third immunization. The strongest effect of vaccina-
tion was observed on early morning BP in this treatment group
where reductions of 25 mmHg for systolic and 13 mmHg for
diastolic BP were observed. This is of particular interest since
early morning BP surge has been linked to an increase in car-
diovascular events. In an attempt to further improve the clini-
cal efficacy of AngQb, higher doses at increased frequency were
evaluated in further phase II clinical studies (NCT00701649
and NCT00710372). Whereas this new regimen induced higher
antibody titers, the beneficial effect on BP was smaller than in
the previous study. The reduced efficacy correlated with reduced
affinity of the antibodies induced with this regimen. In conclu-
sion, active immunization against angiotensin II can result in
significantly reduced BP in hypertensive patients. The complex
interplay between antibody quantities and their quality will have
to be the subject of further investigation to develop an optimal
treatment concept for vaccination against hypertension.
Anti-cytokine vaccines
With increasing success of mAb therapies against different cyto-
kines such as TNF- , IL-1 , and INF- , several therapeutic vacci-
nation strategies have been applied to this class of molecules. Many
of these trials are ongoing and only sparse information is available
in the public domain. However, Neovacs recently announced pre-
liminary positive results in an open-label phase I/IIa study with
their TNF-kinoid in Crohn's disease70 in which clinical remission
(absence of symptoms) in almost half of the treated patients was
observed. Based on these results a double-blind phase II study in
66 Crohn's disease patients (NCT01291810) has been initiated
and first results will be available by the end of June 2012. The
same drug is also currently being tested in rheumatoid arthritis
patients in a double-blind phase II study (NCT01040715), which
will be completed in the course of 2012. Interestingly, Neovacs
reported promising positive results (reduction of symptom score)
in more than half of the patients that had mounted a TNF-specific
immune response.71 Cytos also reported a temporary nonsignifi-
cant improvement of disease symptoms with its first-generation
TNF peptide Qb vaccine for psoriasis.72 Extensive serological
studies revealed that the peptide-based vaccine induced strong
antibodies against the peptide, but reactivity to soluble TNF was
low at best. Hence, to induce better immune responses against
the native cytokine, vaccines based on the whole protein are now
being considered for further development. Finally, Pharmexa pro-
duced a TNF protein in which they had engineered a TT-derived
T-cell epitope. Although immunogenic in humans, the induced
antibodies only recognized a partially denatured form of TNF and
the development of this vaccine was abandoned.73
Results from ongoing trials in this field and further larger
placebo-controlled studies will need to be reviewed to get a bet-
ter understanding if such vaccines are able to compete with
existing mAb therapies.
Fertility management
As a result of the population explosion and the risk associated
with unwanted pregnancies in underdeveloped countries sig-
nificant efforts have been made to identify convenient, inex-
pensive contraceptive methods. In the less developed world,
unwanted pregnancies due to unavailability of contraceptive
methods or due to poor compliance are still a major cause of
death of young women. One of the approaches for new con-
traceptive methods that has been evaluated is therapeutic vac-
cines. The molecules explored for this purpose are involved in 1
gamete production (GnRH, FSH) and include sperm or oocyte
antigens or a hormone (human chorionic gonadotropin; hCG)
essential for the establishment and maintenance of early preg-
nancy.74 hCG conjugated to TT is the most extensively inves-
tigated target that has achieved clinical proof-of-concept in
a phase II trial conducted in 148 women. This vaccine was
highly effective in individuals that had reached antibody con-
centrations above 50 ng/mL. Importantly, fertility was restored
when the antibodies had fallen below 35 ng/mL demonstrating
the reversibility of the effect. Overall this study demonstrated
the feasibility of the approach. However, with the current vac-
cine and regimen only 80% of the patients reached effective
antibody levels and frequent injections were needed to maintain
protective antibody titers. This may be due to a low responder
rate for this therapeutic intervention; hence, successful con-
traceptive vaccines await the development of more immuno-
genic vaccine formulations that can be administered at lower
frequency and still achieve protective antibody concentrations.
Cancer vaccines
In addition to the myriad of vaccination strategies aimed at induc-
ing T cells (Chapter 42), several attempts were made to develop
B-cell vaccines against cancer. Two main approaches have been
pursued. One approach is to induce antibodies against tumor
surface antigen (eg, sialyl-Tn, STn, carbohydrate associated with
MUC1 or Her2), which bind to tumor cells and may ultimately
lead to the elimination of the tumor cells by ADCC. A vaccine
targeting mucin, STn-KLH, showed promising immunogenic-
ity in early trials and was subsequently tested in phase III trials
in more than 1,000 breast cancer patients. However, since no
significant improvement in time to disease progression and over-
all survival compared to the KLH control arm of the study was
observed, the program was discontinued by Merck/Oncoythyreon.
Her2AutoVac was also able to induce a Her2-specific immune

response that was particularly high when administered together
with QS21. However, the absence of clinical improvement of the
health condition in breast cancer patients in a phase II study led
to the early termination of the clinical trial.
A second approach is the neutralization of hormones
involved in cancer growth as potential targets of the immune
response. The most advanced candidate was an anti-gastrin
vaccine (Insegia/G17DT) that was tested in a phase III trial in
pancreatic cancer patients. Upon missing the primary end point
in this study, further development of this vaccine was aban-
doned. Similarly the further development of a vaccine targeting
GnRH in prostate cancer was not pursued.
Overall, antibodies against the target antigens were usu-
ally induced in these clinical studies; however, these antibodies
failed to prevent cancer progression. One of the complicating
factors for the development of these vaccines is that they had to
be clinically tested in late-stage cancer patients where clinical
benefit is extremely difficult to achieve. In addition, thorough
analysis of the amount and the quality of induced antibodies is
still missing in most if not all cases.
Conclusion—outlook
In the last decade a large effort was made to adapt the vaccine
concept to chronic diseases that are a major health threat to
modern societies. Our improved immunological understanding
Access the complete reference list online at http://www.expertconsult.com
2. WHO. Global health risks: mortality and burden of disease attributable to
selected major risks. 2009.
9. Jennings GT, Bachmann MF. The coming of age of virus-like particle
vaccines. Biol Chem 2008;389:521–36.
24. Cornuz J, Zwahlen S, Jungi WF, et al. A vaccine against nicotine for smoking
cessation: a randomized controlled trial. PLoS One 2008;3:e2547.
26. Tissot AC, Maurer P, Nussberger J, et al. Effect of immunisation against
angiotensin II with CYT006-AngQb on ambulatory blood pressure: a
double-blind, randomised, placebo-controlled phase IIa study. Lancet
2008;371:821–7.
27. Jennings GT, Bachmann MF. Immunodrugs: therapeutic VLP-based vaccines
for chronic diseases. Annu Rev Pharmacol Toxicol 2009;49:303–26.
Noninfectious disease vaccines 54 1145
together with the development of recombinant technologies
has allowed scientists to harness this successful therapeutic
modality to target self molecules which, if overproduced, play
central roles in the etiology of disease. However, none of these
approaches has become a marketed product, and further chal-
lenges remain to fully explore this therapeutic concept.
In this context it is important to remember that the
amounts and often affinities of antibodies required to success-
fully neutralize and inhibit self antigens or drugs of abuse are
orders of magnitude higher than those required for the pro-
tection against viral infection. Hence, the major challenges to
enable this approach for future treatments will include induc-
ing antibodies in sufficient amounts and of appropriate affin-
ity. Determination of the right combination of quantity and
affinity of antibody needed for therapeutic efficacy is complex.
Many of the specific requirements for the antibody responses
will be target specific and primarily dictated by individual con-
centration, size, and quaternary structure of the disease tar-
get. For example, whereas monovalent cytokines may critically
depend on high-affinity antibodies to be efficiently eliminated,
antibodies of lower affinity may be sufficient to clear aggregates
such as A .
Affinity and quantity of induced antibody responses can be
influenced by the use of different adjuvants, vaccine carriers,
and the regimen used. Target-specific optimization of these
three parameters may lead to successful treatments for chronic
disease in the future.
29. Moreno AY, Janda KD. Immunopharmacotherapy: vaccination strategies
as a treatment for drug abuse and dependence. Pharmacol Biochem Behav
2009;92:199–205.
34. Mangialasche F, Solomon A, Winblad B, et al. Alzheimer's disease: clinical
trials and drug development. Lancet Neurol 2010;9:702–16.
37. Schenk D, Barbour R, Dunn W, et al. Immunization with amyloid-beta
attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature
1999;400:173–7.
66. Do TH, Chen Y, Nguyen VT, et al. Vaccines in the management of
hypertension. Expert Opin Biol Ther 2010;10:1077–87.
82. Talwar GP, Singh O, Pal R, et al. A vaccine that prevents pregnancy in
women. Proc Natl Acad Sci U S A 1994;91:8532–6.
 SECTION THREE: Vaccines in development and new vaccine strategies
Respiratory syncytial virus and
55 parainfluenza virus vaccines
Ruth A. Karron
Respiratory syncytial virus (RSV) is the most important cause
of viral acute lower respiratory tract illness (ALRI) in infants
and children worldwide.1-3 In the United States, it is estimated
that approximately 70,000 to 126,000 infants are hospital-
ized annually with RSV pneumonia or bronchiolitis, and that
the rate of hospitalization for bronchiolitis has increased since
1980.4 RSV also accounts for a substantial proportion of out-
patient visits in US children younger than 5 years.3 Globally,
RSV is the most frequent cause of ALRI in children and was
estimated to have caused between 66,000 and 199,000 deaths
in 2005.5 Although traditionally regarded as a pediatric patho-
gen, RSV also can cause life-threatening pulmonary disease in
human stem cell transplant (HSCT) recipients. The elderly are
also at risk for severe RSV disease,7–12 and 14,000 to 62,000
RSV-associated hospitalizations of the elderly occur annually in
the United States.8
Human parainfluenza viruses types 1, 2, and 3 (HPIV1,
HPIV2, and HPIV3, respectively) are also important respiratory
pathogens in infancy and early childhood: 25% of persons in
this age group will develop clinically significant HPIV disease.13
In recent studies of the inpatient burden of HPIVs, these viruses
were collectively associated with approximately 7% of hospital-
izations for respiratory or febrile illness in US children younger
than 5 years.14 In an outpatient study involving young children
with illnesses associated with a positive viral culture, HPIVs
were recovered from 18% of outpatients with upper respiratory
illness (URI), 22% with ALRI, and 64% with croup.15 HPIV1
and HPIV2 are the principal causes of croup, which occurs
primarily in children 6 to 48 months of age. HPIV3 causes bron-
chiolitis and pneumonia predominantly in children younger
than 12 months.16–20 Similar to RSV, HPIV3 can also cause
severe ALRI in immunocompromised patients.21,22
Although the importance of RSV and HPIVs as respiratory
pathogens has been recognized for more than 40 years, several
obstacles have hindered vaccine development. For both RSV
and HPIV3, the peak of severe disease occurs in very young
infants, who may not respond adequately to vaccination because
of immunologic immaturity and/or suppression of the immune
response by maternally derived antibody.1,23–25 Because serious
RSV disease can occur in high-risk persons who have experi-
enced previous RSV infection as well as RSV-naïve infants, it is
likely that more than one type of RSV vaccine will be needed to
immunize all of those who would benefit from vaccination. In
addition, an RSV vaccine must not potentiate naturally occurring
RSV disease, as was observed with the formalin-inactivated RSV
(FI-RSV) vaccine (see “Past experience: formalin-inactivated
respiratory syncytial virus vaccine”).26–28
This chapter describes recent efforts to develop safe and
effective RSV and HPIV vaccines for populations at risk.
Background
Clinical disease
Reinfection with RSV and the HPIVs occurs throughout life,
although disease manifestations differ. In young children, RSV
and HPIV3 infections are associated with a spectrum of respira-
tory illness, ranging from mild upper respiratory illness (URI)
to life-threatening bronchiolitis and pneumonia. In infants
less than 6 weeks old, poor feeding and lethargy may predom-
inate, and apnea may occur in the absence of other respira-
tory signs or symptoms.29 HPIV1 and HPIV2 may cause mild
URI, but are also the principal causes of croup.15,16 Otitis
media occurs frequently in children with RSV and HPIV infec-
tions, and RSV is the principal cause of viral otitis media in
children.30 In children hospitalized with RSV in wealthy coun-
tries, mortality is estimated to be approximately 1%, although
it is several-fold higher in infants who are premature31,32 or who
have chronic lung disease,33 congenital heart disease,34 or pri-
mary immunodeficiency disorders.35,36 RSV is the leading viral
cause of hospitalization and reduction in pulmonary function
for children with cystic fibrosis.37–39 In Native American chil-
dren, the risk of hospitalization for RSV is three to five times
greater than that of the general US pediatric population.40,41
The potential link between RSV infection in early child-
hood and subsequent development of reactive airway disease
remains an open question. Infants hospitalized with RSV infec-
tion frequently have demonstrable abnormalities in pulmonary
function for years after the initial event, although they may ulti-
mately become symptom free.42 However, it is not clear whether
RSV causes abnormal airway function, or is only one of many
triggers in susceptible persons.42,43 Ultimately, an intervention
that minimizes the severity of RSV disease (such as universal
use of an RSV vaccine in infancy, or studies of RSV monoclonal
antibody use in term infants) may help to resolve this issue.
In healthy young adults, RSV and HPIV infections are gener-
ally associated with mild URI.44 Elderly adults attending adult
day care or in long-term care facilities and who are infected with
RSV may have rhinorrhea (67% to 92%), cough (90% to 97%),
fever (20% to 56%), and/or wheezing (6% to 35%).10 Pneumonia
occurs in up to 10% of these persons.11,45,46
Immunocompromised patients, particularly those with
hematologic malignancies and those undergoing HSCT or lung
transplant, are at high risk for severe RSV and HPIV3 disease.21,22,47
In these patients, URI precedes LRI, and the presence of
rhinorrhea, sinusitis, and/or otalgia are clinical features that
may help to distinguish between RSV and cytomegalovirus

pneumonia.6,10,37,48–50 The severity of disease depends on the
type and magnitude of immunosuppression, with rates of RSV
pneumonia of up to 75% in leukemic patients51 and of HPIV3
or RSV pneumonia in 24% to 79% of HSCT patients.21,52 For
HSCT recipients, infection that occurs preengraftment is associ-
ated with the highest risk of pneumonia and death, but mortality
is still high in those who develop pneumonia postengraftment.52
HPIV3 also causes ALRI in lung transplant recipients and has
reported associations with both acute allograft rejection and later
development of bronchiolitis obliterans.53 In the United States,
prior infection with human immunodeficiency virus (HIV) does
not appear to increase the morbidity and mortality associated
with RSV or HPIV infections in children, although prolonged
viral excretion has been reported.54 However, in South Africa,
RSV and HPIVs have been reported to be associated with greater
morbidity and mortality in HIV-infected than HIV-noninfected
children, although the contribution of other illnesses cannot be
excluded.55,56
The viruses
RSV and the HPIVs are all members of the family Paramy-
xoviridae. These viruses have genomes composed of a single
strand of negative-sense RNA that is tightly associated with
viral protein to form the nucleocapsid. The viral envelope is
composed of a plasma membrane-derived lipid bilayer that con-
tains virally encoded transmembrane proteins. A viral poly-
merase that transcribes genomic RNA into messenger RNA is
packaged within the virion.
RSV is a member of the genus Pneumovirus and is composed
of 15,222 nucleotides that encode three transmembrane surface
proteins (F, G, and SH), two matrix proteins (M and M2), three
nucleocapsid proteins (N, P, and L), and two nonstructural pro-
teins (NS1 and NS2). The surface fusion (F) and attachment (G)
glycoproteins are the only viral components that induce RSV
neutralizing antibody and therefore are important targets for
vaccine development. The F protein, in combination with pro-
teins G and SH, is responsible for fusion of the viral envelope
with the host cell membranes and for the characteristic syncy-
tium formation in cell culture. Its genome is highly conserved
between RSV groups. Along with the F protein, the G protein
mediates attachment to the host cell surface and is largely
responsible for the antigenic diversity observed between RSV
groups (see later in this section). A secreted form of the G pro-
tein that lacks the N-terminal signal/anchor region also is pro-
duced, and it is reported that up to 80% of the G glycoprotein
released from cells 24 hours after infection is in this secreted
form.57 The function of the secreted G protein is not known,
although it may serve as a decoy for RSV neutralizing antibody58
or as an inhibitor of innate immune responses.59 The NS1 and
NS2 proteins suppress a key component of host innate immune
defense by blocking induction (NS1) or signaling (NS2) of type I
and type III interferons (IFNs).60,61
Cross-neutralization studies have shown that RSV iso-
lates can be classified into two groups, designated A and B.62
Although RSV A and B strains differ in all 10 viral proteins, the
G glycoprotein shows the greatest divergence between groups,
with only 53% amino acid homology between prototype RSV A
and B viruses.63 Group A RSV infection may cause more severe
disease than group B RSV, although this has not been defini-
tively established.64,65 The impact of this antigenic dimorphism
is not completely understood, but young children experiencing
their second RSV infection frequently are reinfected with virus
from the same group.66
HPIV1 and HPIV3 are members of the genus Respirovirus,
and HPIV2 and HPIV4 are members of the genus Rubulavirus.
Each HPIV is composed of approximately 15,500 nucleotides
that encode two transmembrane surface proteins (F and HN), a
matrix protein (M), and three nucleocapsid proteins (N, P, and
Respiratory syncytial virus and parainfluenza virus vaccines 55 1147
L). Additional accessory proteins (C, V, D, and I) are made by
various HPIVs; these proteins are not essential for viral rep-
lication but in some instances inhibit interferon induction.67
The surface fusion (F) and hemagglutinin-neuraminidase
(HN) glycoproteins are the only viral components that induce
neutralizing and hemagglutination-inhibiting (HI) antibody (in
the case of HN) and are important targets for vaccine develop-
ment. The HPIV HN protein is partially analogous to the RSV
G protein in that it also mediates attachment to the host cell
surface. However, the neuraminidase present in HN cleaves
sialic acid residues on the virus and nearby cell surface pro-
teins late in infection to facilitate release of progeny virions. As
with the RSV F protein, the HPIV F protein is responsible for
fusion of the viral envelope with the host cell membranes and
for the characteristic syncytium formation in cell culture; for
most HPIVs, this process is HN-dependent in that HN must
bind to its receptor for fusion to occur.68
Unlike RSV, HPIV1, 2, and 3 each contain a single antigenic
group. HPIV4 viruses fall into two antigenic subgroups, A and
B, but the clinical significance of this is unclear.
Epidemiology
By 2 years of age, almost all children will have been infected with
RSV, and approximately 50% will have been infected twice.69
Similarly, at least 60% of children will have been infected with
HPIV3 by 2 years of age, and approximately 80% will have been
infected by 4 years of age.70,71 Infection with HPIV1 or 2 usu-
ally occurs somewhat later, but by 5 years of age most children
have been infected with HPIV2, and more than 75% have been
infected with HPIV1.70,71 Reinfection with each of these viruses
can occur throughout life and is usually symptomatic; however,
RSV and the HPIVs generally do not cause LRI in immunocom-
petent adults and healthy older children.44
RSV epidemics occur yearly during late fall, winter, and early
spring in temperate climates and in the rainy season in some,
but not all, tropical climates.72,73 RSV group A and group B
viruses cocirculate during epidemics, though one may predomi-
nate.74–77 In temperate climates, yearly HPIV3 epidemics occur
in the spring and summer. HPIV1 epidemics occur during the
fall of odd-numbered years, and although HPIV2 epidemics occur
annually in the fall, HPIV2 can also be isolated at other times
of the year.78,79 HPIV4 is more difficult to detect by viral culture
than HPIV1, 2, and 3 and so has only recently been described in
epidemiologic studies of respiratory viral illness. In these studies,
HPIV4 has been associated with URI and LRI in children and
accounts for 1% to 10% of parainfluenza-associated illnesses.80–82
Humans are the only known reservoir for RSV and the HPIVs.
Spread of RSV and the HPIVs from contaminated nasal secre-
tions occurs via large droplets rather than small-particle aero-
sols, so close contact with an infected person or contaminated
environmental surface is required for transmission.67,83 RSV
can persist as a fomite on hard surfaces for several hours,84,85
and for this reason is an important cause of nosocomial respi-
ratory illness, particularly on pediatric wards.85 Nosocomial
HPIV3 infections have also been described in hospitalized
children19 and immunocompromised patients,73 and transmis-
sion of HPIV3 has been described among immunocompromised
patients in an outpatient department.86
Mechanisms of immunity and correlates
of protection
Virus-specific immune responses are largely responsible for
protection against RSV-associated LRI and recovery from
RSV infection. Immunity to RSV is mediated via humoral
1148 SECTION THREE Vaccines in development and new vaccine strategies
and cellular effectors, including serum antibody (acquired as
a result of infection or maternally derived in young infants),
secretory antibody, and major histocompatibility complex class
I–restricted cytotoxic T lymphocytes. The RSV F glycoprotein
also may elicit innate immune responses via Toll-like receptors
and CD14.87,88 Natural immunity to RSV is incomplete, and
reinfection occurs throughout life, as has been demonstrated
by epidemiologic studies69,89 and challenge studies in healthy
young adults.90 Healthy older children and adults, however,
usually are protected against RSV-associated LRI. In general,
humoral immune responses (secretory and serum antibod-
ies) appear to protect against infection of the upper and lower
respiratory tract, respectively, while cell-mediated responses
directed against internal proteins appear to terminate infec-
tion. RSV-specific cytotoxic T lymphocytes (CTLs) have been
detected in infants recovering from RSV infection91 and may
contribute to short-term protection, such as reinfection dur-
ing the same RSV season. Although the adoptive transfer of
primed T cells will halt RSV replication in immunodeficient
mice, the adoptive transfer of RSV-specific cytotoxic T lym-
phocytes also may potentiate disease,92 suggesting that there
may be an immune component to RSV illness. CD4 Th2–
associated responses have been associated with eosinophilia,
increased mucus production, and delayed viral clearance in
some small animal models,93–95 and may also contribute to the
pathogenesis of RSV disease.96
The role of local immunity in the protection of the upper
respiratory tract against RSV is suggested by experimental data
from studies in cotton rats97–99 and adult volunteers,100 and by
observational data from infants.101 In adults, the presence of
secretory neutralizing antibody, but not serum antibody, cor-
related with protection of the upper respiratory tract against
RSV infection.100 In infants, the development of immunoglobu-
lin A (IgA) in nasal secretions correlated temporally with viral
clearance following natural infection.101
RSV replicates exclusively in the respiratory epithelium. For
this reason, serum neutralizing antibody does not prevent clini-
cally apparent infection, as it does for pathogens that produce
viremia, such as measles and varicella. However, high titers
of RSV serum neutralizing antibody protect the lower respira-
tory tract against RSV infection or disease, as has been dem-
onstrated by animal studies,97–99 epidemiologic observations
in infants and young children,102–104 and clinical trials of RSV
hyperimmune globulin and monoclonal antibodies in high-risk
infants (see “Passive immunization against RSV”).
Primary infection with RSV does not always elicit an
immune response that will protect the lower respiratory tract
because RSV-associated LRI can occur in young children expe-
riencing their second episode of RSV.69,104 Young infants often
develop levels of neutralizing antibody and F and G glycoprotein
antibodies to RSV that are only 15% to 25% of those observed
in older children.105 The suboptimal response of young infants
to primary infection with RSV has important implications for
vaccine development because it suggests that more than one
dose of vaccine will likely be needed to induce adequate levels
of RSV serum neutralizing antibody in this population.
The immunologic requirements for protection against severe
RSV disease are not clearly defined. Although it is known that
a high level of serum neutralizing antibody (titer 1:200, as
measured in a plaque reduction neutralization assay) is suffi-
cient,106,107 it is not known whether it is necessary. In adults and
older children who have been previously infected with RSV, it is
reasonable to expect that an effective RSV vaccine will boost lev-
els of serum neutralizing antibodies. However, it is conceivable
that vaccination may protect young RSV-naïve infants against
severe RSV, even if high titers of neutralizing antibodies are
not induced. In these infants, “challenge” with a second dose
of a live attenuated vaccine may help to predict whether some
protection has been conferred by the first dose (see “Biologically
derived live attenuated vaccines”).108 Ultimately, correlates of
protection against severe disease in young infants will have to
be evaluated in the context of efficacy trials.
Much of our understanding of the host response to the HPIVs
is by inference from what is known about RSV. Serum neutraliz-
ing antibody appears to provide partial protection against HPIV3
infection and illness in infants and children.17,109 Experimental
infection of adults with either HPIV1 or HPIV2 indicated that
the levels of local secretory IgA antibodies correlated best with
resistance to infection and upper respiratory tract disease.110,111
The importance of class I and class II MHC-restricted, virus-
specific T cells in recovery from HPIV infection has not been
formally established but is suggested by studies of chimeric
HPIV3-HPIV1 vaccines in hamsters112 and by the experience of
immunodeficient persons (see “Clinical disease” ).
Passive immunization against RSV
Based on experimental data in animals97–99 and epidemiologic
data in humans104 suggesting that RSV neutralizing antibody
protected against LRI, two products containing high titers of
RSV neutralizing antibody were developed for clinical adminis-
tration. The first of these, RSV Immune Globulin Intravenous
(RSV-IGIV; RespiGam [MedImmune, Gaithersburg, MD]), led
to significant reductions in the rate and severity of LRI when
administered to high-risk infants at doses of 750 mg/kg.106
The titer of RSV neutralizing antibody achieved in infants
who received this dose of RSV-IGIV was comparable to that
demonstrated to protect the lungs of cotton rats against RSV
infection in earlier studies.106,107 The protective effect of RSV-
IGIV in these young infants was confirmed by a subsequent
placebo-controlled trial.113 More recently, a monoclonal neu-
tralizing antibody directed against the RSV F glycoprotein
(palivizumab/Synagis; MedImmune, Gaithersburg, MD) also
was shown to reduce the risk of hospitalization for RSV disease
in premature infants and infants with chronic lung disease.114
Because palivizumab is 50 to 100-fold more potent than RSV-
IGIV, it can be administered intramuscularly in monthly doses.
In infants with cyanotic heart disease, RSV-IGIV was associ-
ated with an increased incidence of adverse events,106,115,116
but a phase 3 study demonstrated that palivizumab was safe
and effective in preventing hospitalizations for RSV in chil-
dren with congenital heart disease.117 Recently, the American
Academy of Pediatrics (AAP) released revised guidelines for
the use of palivizumab in high-risk infants and children, based
on expected burden of disease and cost-effectiveness consider-
ations. According to these guidelines, infants born at less than
32 weeks’ gestation, or with chronic lung disease of prematu-
rity or hemodynamically significant congenital heart disease,
should receive up to five monthly doses beginning in November
and ending in March, depending upon birth month (slightly
different guidance is provided for regions of Florida that expe-
rience an earlier RSV season). Children with chronic lung
disease of prematurity or hemodynamically significant
congenital heart disease may receive palivizumab prophylaxis
during a second season (until 24 months of age), depending
upon the severity of their underlying illness. However, infants
born between 32 and 35 weeks gestation are now recommended
to receive a maximum of three doses of palivizumab if they are
less than 3 months old at the start of the RSV season and have
at least one of two risk factors: attendance at day care and/or a
sibling younger than 5 years living in the home. Detailed rec-
ommendations are provided in an AAP policy statement.116
Phase 1 trials also showed that palivizumab was well toler-
ated in a small number of HSCT recipients.118 Although the
prophylactic efficacy of RSV-IGIV and palivizumab have been
established, neither appears to ameliorate RSV disease when
given therapeutically.119,120

A more potent RSV monoclonal antibody (mAb), designated
motavizumab (MEDI-524), was also developed by MedImmune-
AstraZeneca (Wilmington, DE). Compared with palivizumab,
motavizumab has increased neutralizing activity in vitro and
in vivo, and it has been shown to decrease RSV replication in
the upper as well as the lower respiratory tract in preclinical
studies.121–123 Two phase 3 efficacy trials of motavizumab have
been conducted. In a noninferiority trial, 6,635 preterm infants
were randomized to receive motavizumab or palivizumab.
Infants and children who received motavizumab had a 26%
relative reduction in RSV hospitalization and a 50% relative
reduction in medically attended lower respiratory illnesses
(MALRIs) compared with those who received palivizumab.124
Motavizumab was also evaluated in a randomized, double-
blind, placebo-controlled trial in healthy Native American
populations (www.clinicaltrials.gov [trial NCT00121108]).
In this study of more than 2,000 infants, motavizumab was
found to reduce RSV hospitalizations by 83% and outpatient
MALRIs by 71%.125 As of this writing, however, MedImmune-
AstraZeneca no longer intends to develop motavizumab for
prophylaxis against serious RSV disease.
Vaccine development
General considerations
Successful RSV and HPIV vaccines should prevent serious
RSV or HPIV-associated LRI in those at risk. The primary
target populations for amelioration of disease through RSV
vaccination are young infants and the elderly, although tod-
dlers and preschool children could also benefit from RSV
immunization. As with RSV, the primary target population
for HPIV3 immunization is the young infant. For HPIV1 and
HPIV2, the target population includes slightly older children,
though it would be advantageous to initiate immunization in
the first year of life. In the case of RSV, it is likely that dif-
ferent vaccines will be needed for the various target popula-
tions: nonreplicating vaccines may be useful in the elderly,
in high-risk older children, and for maternal immunization,
but live virus vaccines are likely to be required for RSV-naïve
infants. For RSV vaccines, induction of neutralizing antibody
responses would be highly desirable,96 and induction of a bal-
anced CD4 and CD8 cell response might also be beneficial.
Currently, a number of novel strategies are being used to
develop new experimental RSV vaccines, including the use of
Sendai virus126 and alphavirus127,128 vectors and of intranasally
delivered nanoemulsions of vaccine antigens. However, only
those vaccines currently being evaluated in clinical trials will
be considered here.
Past experience: formalin-inactivated respiratory
syncytial virus vaccine
In the early 1960s, a formalin-inactivated RSV vaccine (FI-RSV)
was prepared and tested in infants and children. This vaccine,
designated lot 100, was administered as two or three intramus-
cular doses, separated by 1 to 3 months, to infants and children
between 2 months and 7 years of age.26–28,129 Lot 100 not only
failed to protect against wild-type (wt) RSV disease, but induced
an exaggerated clinical response to wt RSV infection in infants
who were RSV naïve before vaccination. Many vaccinees were
hospitalized with LRI; in one study, the hospitalization rate
of vaccinees approached 80% compared with 5% in control
vaccinees.26 Tragically, two infants who received lot 100 died
following wt RSV infection, one at 14 months and the second
at 16 months of age.26 RSV was readily isolated from the lower
respiratory tracts of these infants.
Respiratory syncytial virus and parainfluenza virus vaccines 55 1149
The mechanisms responsible for the FI-RSV vaccine-associ-
ated disease enhancement are still not completely understood.
However, data obtained from lot-100 recipients129–131 and from
studies in rodent models132–135 have led to the hypothesis that
children vaccinated with FI-RSV remained susceptible to infec-
tion with wt RSV because vaccination produced inadequate
levels of serum neutralizing antibodies and did not induce local
immunity. Once infected with wt RSV, virus was not read-
ily cleared because FI-RSV did not prime for a CD8+ cytotoxic
T-cell response, and the viral infection produced a direct cyto-
pathic effect in the lower respiratory tracts of these infants.
In addition, immunization with FI-RSV primed for a type 2
helper T-cell–like response, with increased local production of
interleukin-4 (IL-4), IL-5, and IL-10; an influx of lymphocytes
and eosinophils; the release of additional mediators, and resul-
tant inflammation and bronchoconstriction.136–139 Immunization
with FI-RSV may have also induced low-avidity antibody, lead-
ing to immune complex deposition.140,141
The clinical experience with FI-RSV and the information
gleaned from animal models of disease enhancement suggest
key features of an RSV vaccine for seronegative infants. The
vaccine should induce protective levels of neutralizing antibody,
as well as CD8 RSV-specific cytotoxic T cells, and a pattern of
CD4 response similar to that evoked by wt RSV. A live atten-
uated vaccine is most likely to exhibit these characteristics.23
Nonreplicating vaccines could be used to boost immunity in
older children and adults at risk, or to provide passive protec-
tion to the infant through maternal immunization.142
Subunit RSV vaccines
RSV F and G, the viral glycoproteins that induce neutralizing
and protective antibodies (reviewed by Crowe and Collins1),
have been evaluated as potential candidate vaccines. Subunit
vaccines are most likely to be useful for immunization of the
elderly and high-risk children, and might also be used for
maternal immunization. Vaccines that have been previously
evaluated in clinical trials include purified F glycoproteins;143–148
copurified F, G, and M proteins;149 a chimeric RSV FG fusion
protein vaccine, and BBG2Na, a peptide from the G glycopro-
tein conjugated to the albumin-binding domain of streptococcal
protein G.150–153 As of this writing, the only RSV subunit vaccine
currently being evaluated in clinical trials is an RSV F protein
particle vaccine developed by Novavax (Rockville, MD) (www.
clinicaltrials.gov [trial NCT01290419]), which is currently
undergoing phase 1 evaluation in healthy adults (Table 55–1).
Previous RSV F subunit vaccines designated purified F pro-
tein (PFP)-1, PFP-2, and PFP-3, were evaluated in healthy adults,
in children more than 12 months old with and without chronic
underlying pulmonary disease (chronic lung disease of prematu-
rity or cystic fibrosis), in institutionalized and ambulatory elderly
subjects, and in pregnant women.144–148,149,154,155 The PFP vaccines
were well tolerated in these populations: acute reactions were min-
imal and enhanced disease was not observed.144–148,154 However,
the neutralizing antibody responses to these RSV F vaccines were
suboptimal.144–148,154–156 Knowledge of the structure and antigenic
properties of the RSV F protein157–159 may allow the development of
more stable and more immunogenic RSV F vaccine preparations.
Live attenuated respiratory syncytial virus and
human parainfluenza virus vaccines
Live attenuated vaccines may offer several advantages over
nonreplicating vaccines, especially for RSV and HPIV-naïve
infants and young children. Intranasal immunization with a
live, attenuated vaccine should induce both systemic and local
immunity and may therefore protect against URI as well as
LRI. Also, the immune response to a live vaccine should closely
resemble the response to natural infection and therefore not
1150 SECTION THREE Vaccines in development and new vaccine strategies
Respiratory Syncytial Virus (RSV) and Human Parainfluenza Virus (HPIV) Vaccines Evaluated in Clinical Trials Since 2000
Includes mutants with multiple mutations and/or deletions of nonessential genes. rA2cp248/404
and rA2cp530/1009 NS2 have been evaluated in clinical trials.
attenuating; AZ, AstraZeneca; cold passage; NIH, National Institutes of Health.
produce enhanced disease on exposure to wt virus.23 Similar
to other live attenuated intranasal respiratory virus candidate
vaccines,24,25 live intranasal RSV and HPIV3 candidate vaccines
have been shown to replicate in young infants in the presence
of maternally acquired antibody.108,161 This feature will be criti-
cal for the success of live attenuated RSV and HPIV vaccines in
young infants. Multiple doses of these vaccines will probably be
required for young infants.
Several strategies for the development of a live attenuated
RSV vaccine were originally explored, including the creation of
host-range mutants, cold-passaged () mutants, and temper-
ature-sensitive () mutants (which are unable to grow at high
temperatures). In brief, these candidate vaccines were either
162,163
underattenuated (RSV and RSV -1) and transmissible
or overattenuated (RSV -2). 23,138,164 Of importance, enhanced
disease was not observed when infants who received RSV -1
or RSV were naturally infected with wt RSV. 162,165 A series
of live attenuated RSV A candidate vaccines were derived
from further attenuation of RSV through chemical muta-
genesis. Several of these candidate intranasally administered
vaccines were evaluated in phase 1 clinical trials in adults and
children,108,166 and one (248/404) was evaluated in infants
as young as 1 month old.108 The 248/404 vaccine was highly
attenuated in these infants but caused nasal congestion that in
some instances interfered with feeding and sleeping.108 Although
the 248/404 vaccine was not sufficiently attenuated for young
infants, evaluation of this candidate vaccine provided important
SH, rA2cp248/404/1030 SH, rA2cp NS2, rA2cp248/404 NS2,
information regarding replication and immunogenicity of a live
attenuated RSV vaccine in the presence of maternal antibodies,
phenotypic stability of a vaccine, and preliminary evidence
of protection against illness following wt RSV infection.108
The ability to recover infectious virus from complementary
DNA (cDNA) clones of RSV167 (Figure 55-1) has provided
insight into the genetic basis of attenuation of biologically
derived vaccines and hastened the development of additional
live attenuated RSV candidate vaccines through the use of
recombinant technology.58,168 Mutations present in RSV and
six of its derivatives were inserted into wt RSV singly and
in combination, and the majority of attenuating mutations
were found to occur in the polymerase gene, with a notable
exception being the 404 mutation in the gene start signal
(see Figure 55–2).58,168–170 Using this information, attenuating
mutations from biologically derived vaccines have been com-
bined to produce further attenuated candidate vaccines.58,168,171
In addition, deletion () of a nonessential gene (
160
or ) in combination with known attenuating
and mutations has also been used to produce highly attenuated
vaccines (Figure 55-2).172,173 Two of these candidates, designated
rA2cp248/404 SH and rA2cp248/404/1030 SH, were evaluated
in young children.174 rA2cp248/404 SH was not more attenu-
ated than its biologic parent, 248/404. However, addition
of the 1030 amino acid–point mutation to rA2cp248/404 SH
yielded rA2cp248/404/1030 SH, which was highly attenuated
 Respiratory syncytial virus and parainfluenza virus vaccines 55 1151

A plasmid containing the full-length

RSV antigenome complementary DNA (cDNA) and four “helper” plasmids containing the RSV and (open reading frame 1) genes are
transfected into Vero cells. The RSV antigenome and helper genes, under control of the bacteriophage T7 RNA polymerase promoter, are expressed
following infection with MVA-T7 pol virus. Infectious RSV is recovered, amplified, and then biologically cloned by terminal dilution. The cloned virus
is amplified in Vero cells to make Good Manufacturing Practices vaccine lots for use in phase 1 clinical trials. This strategy was used to produce
the recombinant RSV strains that were previously evaluated in clinical trials. More recently, recombinant RSV strains have been generated for future
clinical trials using MVA-free expression systems. (Courtesy Drs. Valerie Randolph and Christopher Park, Wyeth Vaccines, Pearl River, NY.)

A, Mutations in the RSV genome

are identified as point mutations (arrows) or as gene deletions (). Point mutations are further identified as mutations induced by serial cold-
passage () or as temperature-sensitive ( ) mutations induced by chemical mutagenesis (identified numerically). The number assigned to the
mutation indicates the clone number of the virus in which the mutation was first identified. B, The 15 mutations in 45 that are thought to be
important for conferring attenuation are identified at the site of the mutation (arrows). Those mutations that are known to produce virus that is cold
adapted (), temperature-sensitive (), or attenuated in nonhuman primates ( ) are indicated by the corresponding abbreviation.
 SECTION THREE
in infants as young as 1 to 2 months and moderately immuno-
genic.174 This candidate vaccine, now designated MEDI-559, is
currently in phase 1/2a trials in infants and children ages 1 to
23 months (www.clinicaltrials.gov [trial NCT00767416]).
A second set of engineered candidate vaccines, rA2cp NS2,
rA2cp248/404 NS2, and rA2cp530/1009 NS2, were also
evaluated in clinical trials.175 These vaccines all had dele-
tion of the interferon antagonist gene in common, which
could theoretically enhance innate and adaptive immune
responses.176 rA2cp NS2 was overattenuated for adults, and
rA2cp248/404 NS2 and rA2cp530/1009 NS2 were overatten-
uated for children at the doses studied.175 Nevertheless, these
studies demonstrated that deletion of attenuates RSV infec-
tion in humans so that generation of alternative -deletion
mutant candidate vaccines may be feasible. Similarly, deletion of
the gene, which also inhibits IFN- and IFN- induction,175
may yield an attenuated rRSV candidate vaccine.177 An addi-
tional candidate, designated RSV M2 2, lacks the RSV
gene. Compared with wild-type RSV, RSV M2 2 has increased
accumulation of intracellular viral mRNA and increased expres-
sion of RSV proteins,177 which might enhance vaccine-induced
immunity. This vaccine virus is attenuated in nonhuman
primates,177 and, as of this writing, is about to enter phase 1
clinical trials.
Recombinant technology also provides the opportunity for
creation of chimeric viruses containing the RSV F and G surface
glycoprotein, with one or more internal genes provided by related
respiratory viruses. The potential advantages and disadvantages
of these vectored PIV3 vaccines are discussed in the following sec-
tions. A chimeric vaccine containing the bovine PIV3 backbone,
human PIV3 surface glycoproteins, and RSV F glycoproteins,
designated MEDI-534, was evaluated in a phase 1 trial in 49 RSV
and HPIV3 doubly seronegative children aged 6 to 24 months
(www.clinicaltrials.gov [trial NCT00493285]). Children were
randomized to receive 3 doses of vaccine or placebo. All chil-
dren who received the highest dose of vaccine (106 tissue culture
infective dose [TCID]50) were infected with vaccine virus, and
seroconversion to RSV and HPIV3 occurred in 50% and 100%,
respectively, of recipients of this vaccine dose.178 Evaluation of
this candidate vaccine is continuing (see Table 55–1).
As described below, live attenuated vaccines for HPIV1, HPIV2,
and HPIV3 have been developed and are being evaluated in
clinical trials.
HPIV3 vaccines
Two biologically derived HPIV3 candidate vaccines have been
evaluated in clinical trials. HPIV3 45 ( 45) was derived from
the JS wild-type strain of HPIV3 by 45 passages in primary African
green monkey kidney cells at low temperatures.179 The 45
virus contains 20 point mutations (five of which are silent) that
differentiate it from the JS wild-type strain. Three of these muta-
tions occur in the L (polymerase) protein of 45 and contrib-
ute substantially to the attenuation and temperature-sensitive
phenotypes of this candidate vaccine.180,181 The second candi-
date, bovine parainfluenza virus type 3 (BPIV3), is a host-range
variant that is closely related to HPIV3.182 BPIV3 and HPIV3
are 25% related antigenically by cross-neutralization assays.
Both 45 and BPIV3 have been evaluated in phase 1 and
phase 2 trials in adults, HPIV3-seropositive children, HPIV3-
seronegative children, and infants as young as 1 month (45) or
2 months (BPIV3).161,183–186 Both candidates were overattenuated
in adults and in seropositive children but were highly infectious
in seronegative children and infants.161,183–186 There were no sig-
nificant differences in the incidence of respiratory or febrile ill-
nesses among seronegative vaccinees and placebo recipients.
Otitis media occurred more frequently among seronegative
children vaccinated with HPIV3 45 in phase 1 trials than
among placebo recipients, but this was not observed in infants
vaccinated with HPIV3 45, 161 nor in phase 2 trials of HPIV3
45 in seronegative children. 186 A recombinant form of the
45 vaccine, designated 45, has been developed and is cur-
rently being assessed in phase 1 and 2 trials in HPIV3-naïve
children to determine the optimal dosage and dosing interval
(www.clinicaltrials.gov [trials NCT01021397, NCT01254175,
and NCT01150799]).
Although both HPIV3 45 and BPIV3 elicited hemagglutination-
inhibiting antibody responses against HPIV3 in most vaccinated
seronegative children, the magnitude of the response to HPIV3
was lower in children who received BPIV3, which is consistent
with the limited antigenic relatedness of the bovine and human
PIV3 HN.182 For this reason, recombinant bovine/human PIV-3
candidate vaccines containing the HPIV3 and genes and
one or more BPIV3 internal genes have been developed (see
“HPIV1 and HPIV2 vaccines” and “PIVs as vaccine vectors”).
These chimeric human/bovine PIV3 candidate vaccines contain
one or more internal genes from BPIV3 and the genes encoding
the protective antigens HN and F from HPIV3, either through
replacement of the and genes of BPIV3 with their counter-
parts from HPIV3,187 or through replacement of individual “inter-
nal” protein genes of HPIV3 with their counterparts from BPIV3.
Two of these vaccines, rHPIV3-NB and rB/HPIV3, are in phase
1 clinical trials (www.clinicaltrials.gov [trial NCT00366782]). In
addition, the MEDI-534 vaccine described above is a chimeric
bovine/human PIV3 construct that expresses RSV F.
HPIV1 and HPIV2 vaccines
Sendai virus (SeV), the murine PIV1, has been evaluated as a
HPIV1 vaccine candidate in nonhuman primates and healthy
adults. SeV was highly immunogenic and protective against
HPIV1 challenge when administered intranasally to African
green monkeys and was well tolerated and immunogenic in
1
three of nine adult human recipients.188 The replication of
SeV was not significantly restricted in African green monkeys
and chimpanzees compared with HPIV1,189 which raises the
possibility that it might not be satisfactorily attenuated in
infants and children. However, Sendai virus is currently being
evaluated as an experimental intranasal vaccine in young chil-
dren (www.clinicaltrials.gov [trial NCT00186927]).
Live attenuated candidate vaccines for HPIV1 and HPIV2
have also been developed based on importation of known attenu-
ating mutations from related respiratory viruses such as HPIV3,
BPIV3, and RSV. rHPIV1 84/ 170/942A contains mutations in
the and genes and is highly restricted in replication in
adults and HPIV1-seropositive children. Evaluation of this vaccine
in HPIV1-seronegative children is in progress (www.clinicaltrials.
gov [trial NCT00641017]). Similarly, rHPIV2-15 C/948 L/ 1724,
an HPIV2 vaccine candidate, contains a mutation in the extra-
genic leader region of the genome, and an amino acid substi-
tution and a deletion in the gene. rHPIV2-15 C/948 L/ 1724
is highly restricted in replication in adults, and evaluation in
HPIV2-seropositive children is in progress (www.clinicaltrials.
gov [trial NCT01139437]).
PIVs as vaccine vectors
PIVs also have been evaluated as vectors for expressing the
protective antigens of related PIVs and other pediatric respiratory
viruses, such as RSV (see discussion of MEDI-534, “Genetically
engineered [complementary DNA–derived] live attenuated RSV
vaccines” section) and human metapneumovirus (HMPV), to
develop bivalent or multivalent vaccines. There are several
potential advantages to PIV vector–based immunization.
First, RSV and HMPV do not replicate efficiently in vitro and
readily lose infectivity. In addition, studies of live attenuated
RSV candidate vaccines suggest that only those that are highly
restricted in replication are likely to be sufficiently attenuated

for young infants,174 and these highly attenuated viruses may
not provide a sufficient antigenic stimulus to induce protec-
tive immunity. In contrast, the PIVs replicate to high titer, do
not readily lose infectivity, and provide higher levels of antigen
expression than the native attenuated viruses. In addition, the
vectors themselves (HPIV1 and HPIV3) are needed vaccines,
can be administered intranasally, and might be used alternately
for prime/boost strategies, thereby diminishing the problem of
restriction of replication induced by previous immunity to the
vector. Thus, the use of PIVs as vectors for the delivery of the
protective antigens of RSV and HMPV might simplify vaccine
manufacture and delivery. However, a drawback of the PIV vec-
tor strategy is that only one or two antigens of the vaccine target
are expressed, and thus, the full complement of viral antigens
is not present.190,191
Access the complete reference list online at http://www.expertconsult.com
3. Hall CB, Weinberg GA, Iwane MK, et al. The burden of respiratory syncytial
virus infection in young children. N Engl J Med 2009;360:588–98.
5. Nair H, Nokes DJ, Gessner BD, et al. Global Burden of acute lower
respiratory infections due to respiratory syncytial virus in young children: a
systematic review and meta-analysis. Lancet 2010;375:1545–55.
14. Weinberg GA, et al. Parainfluenza virus infection of young children: estimates
of the population-based burden of hospitalization. J Pediatr 2009;154:694–9.
27. Kapikian AZ, Mitchell RH, Chanock RM, et al. An epidemiologic study of
altered clinical reactivity to respiratory syncytial (RS) virus infection in children
previously vaccinated with an inactivated RS virus vaccine. Am J Epidemiol
1969;89:405–21.
106. Groothuis JR, Simões EAF, Levin MJ, et al. Prophylactic administration of
respiratory syncytial virus immune globulin to high-risk infants and young
children. N Engl J Med 1993;329:1524–30.
116. Committee on Infectious Diseases. From the American Academy of
Pediatrics: Policy statements: modified recommendations for use of
Respiratory syncytial virus and parainfluenza virus vaccines 55 1153
Conclusions
Over the past several years, recognition of the burden of
illness associated with RSV and HPIV infections in children
has increased,4,5,14 which has reinvigorated interest in the
development of vaccines against these pathogens. In particu-
lar, substantial new efforts have been made to develop RSV
vaccines, and many of these candidates are in late preclinical
or early clinical development stages. Although many obsta-
cles to the development of RSV and HPIV vaccines for target
populations still exist, it is hoped that the application of new
approaches and/or novel technologies will lead to the develop-
ment of safe and effective vaccines against these important
respiratory pathogens.
palivizumab for prevention of respiratory syncytial virus infections.
Pediatrics 2009;124:1694–701.
154. Paradiso PR, Hildreth SW, Hogerman DA, et al. Safety and immunogenicity
of a subunit respiratory syncytial virus vaccine in children 24 to 48 months
old. Pediatr Infect Dis J 1994;13:792–8.
158. Swanson KA, Settembre EC, Shaw CA, et al. Structural basis for
immunization with postfusion respiratory syncytial virus fusion F
glycoprotein (RSV F) to elicit high neutralizing antibody titers. Proc Natl
Acad Sci U S A 2011;108:9619–24.
174. Karron RA, Wright PF, Belshe RB, et al. Identification of a recombinant
live attenuated respiratory syncytial virus vaccine candidate that is highly
attenuated in infants. J Infect Dis 2005;191:1093–104.
190. Collins PL, Murphy BR. New generation live vaccines against human
respiratory syncytial virus designed by reverse genetics. Proc Am Thorac Soc
2005;2:166–73.
 SECTION THREE: Vaccines in development and new vaccine strategies
Parasitic disease vaccines
56 Peter J. Hotez
Jeffrey M. Bethony
Parasitic diseases caused by helminths and unicellular eukary-
otes (protozoa) are major causes of human disease in the
less-developed nations of the tropics. When measured in
disability-adjusted life years (DALYs, or the number of life
years lost from premature death or disability), the combined
burden of disease caused by human parasites is approximately
100 million DALYs, a value that exceeds the DALYs from diar-
rheal diseases, lower respiratory infections, or human immu-
nodeficiency virus or acquired immunodeficiency syndrome
(HIV/AIDS).1 Attempts to develop vaccines against these organ-
isms have been hampered by the difficulties in maintaining the
organisms in the laboratory. With a few exceptions, in vitro
culture methods have not been adequate; frequently, labora-
tory animals are necessary to maintain the complex life cycles
of these parasites. This problem has thwarted efforts to scale
up the production of large numbers of parasites to develop live
attenuated or killed vaccines.
In the past decade, investigators in academic, private, and
government laboratories have started to apply modern biotech-
nology to the study of parasites. Genome projects have been
completed for several neglected parasitic protozoa of global
public health importance,2–4 and the transcriptomes of several
important parasitic helminths have also undergone analysis.5–7
However, for several reasons, many of these molecular studies
have not yet translated into major vaccine development initia-
tives. First, most of the parasite genomes are not as amenable
to reverse vaccinology approaches as bacterial genomes, pri-
marily because of the absence of simplified eukaryotic expres-
sion systems that can simultaneously express hundreds of
properly folded antigens.8 In addition, there is a dearth of sim-
plified animal models for parasite vaccine testing.8,9 Finally,
there has been little commercial interest in the development
of protective antigens as human vaccines for neglected tropi-
cal diseases.9,10
In the past decade, new funds from private philanthropies
such as the Bill & Melinda Gates Foundation, together with
support from some European governments, have somewhat
reenergized attempts to manufacture vaccines for parasitic dis-
eases. These funds have stimulated the formation of product-
development public-private partnerships (PD-PPPs).10 Aside
from the malaria vaccines (see Chapter 53), five human vac-
cines are currently undergoing or are about to enter clinical
development. These include four anthelmintic vaccines (one
for hookworm infection and three for schistosomiasis) and an
antiprotozoan vaccine for leishmaniasis (Table 56-1). There is
also a several-decade history of first- and second-generation
leishmaniasis vaccines composed of live and killed parasites,
respectively.11,12 In addition, several veterinary vaccines are
under development, which could be used to interrupt zoonotic
transmission to humans. These include vaccines for bovine
schistosomiasis, pig cysticercosis, canine echinococcosis, and
canine leishmaniasis.
Anthelmintic vaccines
Helminth infections are among the most prevalent infections
of humans.13 Approximately 807 million, 604 million, and
576 million individuals worldwide harbor the soil-transmitted
nematodes and
hookworms in their intestines, respectively,13,14 and an addi-
tional 400 million or more people are infected with the
schistosomes (primarily
and ).15 The World
Health Organization (WHO) estimates that, together, the soil-
transmitted helminths and schistosomes account for 40% of
the global morbidity resulting from tropical infectious dis-
eases, exclusive of malaria (see Chapter 53). In much of the
developing world, soil-transmitted helminth infections and
schistosomiasis exhibit their highest prevalence and inten-
sity of infection (worm burden) among children and teenag-
ers. Children who are chronically infected with heavy worm
burdens develop deficits in physical, intellectual, and cogni-
tive growth.14 In addition, hookworm and schistosomiasis
are important health problems in pregnancy and contribute
to increased maternal morbidity and mortality as well as low
birthweight and prematurity.16
Among the global efforts to control soil-transmitted helminth
infections and schistosomiasis are concerted efforts to adminis-
ter albendazole or mebendazole and praziquantel, respectively,
on a yearly or twice-yearly basis. Because soil-transmitted hel-
minth infections and schistosomiasis primarily affect school-
age children, these anthelmintic drugs often are administered
as part of school-based health and education efforts.14 However,
although albendazole and praziquantel are highly effective at
removing current infections, children reacquire new helminth
infections several months after treatment.17 This potentially
limits the usefulness of anthelmintic agents as a school-based
public health control measure, particularly in light of new evi-
dence for the diminished efficacy of anthelmintic drugs with
frequent use or the possible emergence of anthelmintic drug
resistance.18,19 Moreover, single-dose mebendazole has been
shown to be highly ineffective as a treatment for hookworm
infection and trichuriasis.20
As an alternative approach to control, efforts are now under-
way to develop anthelmintic vaccines. The goals of anthel-
mintic vaccination are different from those of conventional

Human Antiparasitic Disease Vaccines Entering, or in, Clinical Development
APR, recombinant aspartic protease (hemoglobinase); GLA, glucopyranosyl lipid adjuvant; GST, glutathione-S-transferase; IDRI, Infectious Disease Research
Institute (Seattle, WA); INSERM, French National Institute for Health and Medical Research; MPL, monophosphoryl lipid; PD-PPP, product-development public-private
partnerships.
antiviral and antibacterial vaccination, as it is unlikely that
immunization with defined antigens will elicit sterilizing
immunity against complex metazoan organisms. Instead, the
most important goal is to reduce the worm burden to below
the disease-causing threshold—to immunize against the dis-
ease rather than the infection. For example, the goal is to
reduce the hookworm burden to below the threshold that
results in significant intestinal blood loss that would ulti-
mately lead to anemia, or to reduce the schistosome worm
burden below the threshold that results in significant egg
deposition and subsequent granuloma formation in the liver,
intestines, and bladder.21
Still another approach to vaccination against disease would
be to directly block the action of parasite-induced pathogenic
processes. In the case of hookworm, this would require block-
ing parasite-derived virulence factors that cause blood loss, and
for schistosomiasis, blocking egg deposition.22 It is likely that
anthelmintic vaccination will not be used in isolation but in
conjunction with other control efforts, including conventional
chemotherapy.23
Hookworm infection
Human hookworm infection is a leading cause of anemia and
malnutrition among children in developing countries,24 with the
greatest number of cases in sub-Saharan Africa, Asia, and tropi-
cal regions of the Americas.14 is the pre-
dominant hookworm world wide, with
responsible for most of the remaining cases. Humans become
infected when third-stage larvae penetrate the skin and undergo
extraintestinal migration in the vasculature and reach the heart
and lungs. Migration of the hookworm larvae to the lung is
associated with a mild pneumonitis, with the larvae ascending
the airways and reaching the larynx before they are coughed and
swallowed. The larvae molt twice in the intestine to become
adult hookworms, which invade tissue and cause blood loss.
Unlike other soil-transmitted helminth infections, high worm
burdens with hookworms occur in both children and adults
(including pregnant women),24 so pediatric deworming has no
impact on adult populations or on reducing the transmission
dynamics of the infection.
Parasitic disease vaccines 56 1155
Early attempts at developing a hookworm vaccine relied
on the observation that numerous small doses of living third-
stage infective larvae (L3) of the dog hookworm
could confer resistance against challenge hookworm
infections.25 Resistance was measured by reductions in the
worm burden, the worms' size, and their fecundity (egg count).
Immunity was never sterilizing, but efficacy was between 60%
and 70%. Later it was noted that larger doses of living L3 could
be administered over shorter time periods if they were first dam-
aged by ionizing radiation. This provided the basis for commer-
cial radiation-attenuated hookworm L3 vaccines that could be
administered to dogs in two doses, with an efficacy of 90%. The
canine hookworm vaccine was marketed in the eastern United
States during the 1970s, but ultimately it failed as a commercial
veterinary product.25
Although it is not feasible to develop human anti-hookworm
vaccines using living L3 (damaged or otherwise), stud-
ies have been done to identify, isolate, clone, express, and
test vaccine antigens from L3 that can reproduce the reduc-
tion in worm burdens afforded by the live vaccines.26–32 One
promising class of antigens to emerge from these investiga-
tions is the -secreted proteins (ASPs), which
are released by host-stimulated L3 and contain amino acid
sequences homologous to those of the major antigens from
insect venoms.27,28 In preclinical studies, it was shown that
ASP-2 is an immunodominant antigen associated with the
irradiated L3 vaccine,26,29,30 and that it is protective against
challenge infections with animal hookworms.29–31 Moreover,
there is an association between human anti-ASP-2 antibody
responses and a reduction in the risk of having heavy hook-
worm infection.29 The ASP-2 from ( -ASP-2)
was expressed in yeast and selected for subsequent process
development and pilot manufacture under current good man-
ufacturing practices.30,32 These studies were conducted by
the Human Hookworm Vaccine Initiative, a PD-PPP based
at the Sabin Vaccine Institute. Formulated with Alhydrogel,
the -ASP-2 hookworm vaccine underwent phase 1 test-
ing for safety and immunogenicity in the United States.
It was shown that the vaccine is safe and immunogenic in
a population of healthy adult volunteers without previous
hookworm infections,33 but subsequent clinical studies in a
1156 SECTION THREE Vaccines in development and new vaccine strategies
hookworm-endemic area of Brazil revealed that this vaccine
can result in generalized urticaria among a subset of chroni-
cally infected individuals with high titers of prevaccination
IgE to larval antigens including ASP-2. Therefore, the larval
antigen program was halted in favor of the development of a
different class of antigens.22
An alternative approach to hookworm vaccination and one
currently under development by the Sabin Vaccine Institute
takes advantage of the fact that adult hookworms ingest
blood.22,24 Recently, several proteolytic enzymes required for
the parasite's ability to digest hemoglobin were shown to line
the brush border membrane of the hookworm's gastrointesti-
nal tract.34 Vaccination with a recombinant aspartic protease (in
this case, a hemoglobinase), APR-1, was shown to reduce para-
site load and blood loss after hookworm infection in dogs.35 By
site-directed mutagenesis, an enzymatically inactivated hemo-
globinase from ( -APR-1) was also shown to
induce neutralizing antibodies against multiple hookworm spe-
cies and protect dogs against heterologous hookworm infec-
tion.36 In addition, a unique glutathione-S-transferase from
( -GST-1) was shown to exhibit a unique
heme- and hematin-binding function and is hypothesized to
be required for parasite heme-detoxification; a recombinant
-GST-1 expressed in yeast is highly protective in laboratory
animals.37,38
These studies provide the scientific basis of an adult
hookworm-vaccine antigen development program which has
entered phase 1 clinical testing. Ultimately, the final human
hookworm vaccine will probably be composed of -GST-1 and
-APR-1, combined and formulated with Alhydrogel (and pos-
sibly a second immunostimulant such as a synthetic Toll-like
receptor [TLR4] agonist known as GLA [glucopyranosyl lipid
A]), with a vaccination strategy that may be linked either to
the Expanded Programme on Immunization or to anthelmintic
chemotherapy in preschool or school-age children.22 A recent
cost-effectiveness study confirms the potential economic value
of a hookworm vaccine.39
Schistosomiasis
Schistosomes are snail-transmitted, water-borne parasitic
platyhelminths (order Trematoda). High rates of infection occur
near bodies of fresh water such as tributaries of the Nile River
in Egypt and the Dongting and Boyang lakes in China,40,41 with
the largest number of cases occurring in Africa. Revised esti-
mates indicate that more than 400 million cases may occur
in Africa,15 with roughly two thirds of the cases resulting from
the cause of urogenital schistosomiasis, and
one third from the cause of intestinal schistosomi-
asis.42 Africa's schistosomiasis accounts for more than 90% of
the world's cases, with most of the remainder caused by
in Brazil and elsewhere in Latin America and about 1%
caused by the complex (including
and ) in East Asia. 40
The human schisto-
somes are typically distinguished by their unique snail vectors,
by location in the host vasculature, and by egg morphology.
Members of the complex also have important
domestic animal reservoir hosts (pigs, cattle, water buffaloes).
Asexual reproduction of the parasites occurs in the freshwa-
ter snail intermediate hosts that release large numbers of free-
swimming, infective larval schistosomes, known as cercariae,
into the water (Figure 56-1). The cercariae invade human skin,
lose their tail, and then (now called schistosomulae) spend the
next few weeks migrating through the bloodstream and lungs
until they reach the liver. Here they differentiate into male and
female schistosomes. Male and female worm pairs migrate
through the portal vasculature until they reach their final des-
tination in the mesenteric or bladder venules. The worm pairs

release eggs, which exit from the body in feces or urine and
then hatch in fresh water. Most of the morbidity associated
with schistosomiasis occurs when the eggs fail to exit from the
definitive human host. Trapped in the intestinal or bladder wall,
or in the liver as they are swept up by the portal circulation, the
eggs elicit granulomas and host fibrosis. In the liver, Symmer's
pipestem fibrosis from chronic or
infection leads to portal hypertension and hepatosplenomeg-
aly; in the bladder, eggs stimulate formation of
multiple granuloma, resulting in hematuria and an obstructive
uropathy leading to chronic urinary tract infections, hydrone-
phrosis, and kidney failure.40 Chronic bladder fibrosis from the
eggs also results in squamous cell carcinoma
of the bladder;43 the eggs also cause granuloma formation in
the uterus and cervix, leading to female genital schistosomiasis,
which is an important risk factor for HIV/AIDS in Africa.44 In
addition, chronic schistosomiasis is associated with a wide vari-
ety of other sequelae, especially in children, including anemia,
chronic pain, undernutrition, growth failure, and cognitive defi-
cits. The full burden of disability-related outcomes in endemic
schistosomiasis is only beginning to be fully appreciated.45
A number of different approaches have been taken to design
antischistosomal vaccines. As in other systems, the adminis-
tration of radiation-attenuated (RA) cercariae results in the best
protection to date in mice. According to Wilson and Coulson,46
the development of the Th1/Th2 paradigm for T helper (Th) cell
differentiation was extremely influential in interpreting the RA
vaccine model. The two major patterns of cytokine synthesis
correlate with the induction of cell-mediated (Th1) or humoral
(Th2) immunity, thus providing a possible explanation for the
separate and often reciprocal regulation of these responses in
chronic schistosomiasis infection. This dichotomy in Th cell
differentiation, together with subsequent identification of
T-regulatory (Tr) cells and Th3/Tr1 (downregulatory interleukin
[IL]-10 and transforming growth factor beta [TGF- ] produc-
tion) extensions, provides the framework for the mechanisms
operating in the RA vaccine model. It is clear that under appro-
priate conditions, RA vaccine can elicit a partially protective
Th1 or Th2 response. If such responses were directed against
different parasite stages (larval and adult) and at different sites
(skin, lung, mesentery), they might act additively or even syner-
gistically to provide a very high level of protection. It is also pos-
sible (as Wilson and Coulson warn) that these responses to the
RA vaccine might actually prove mutually exclusive, in which
case it would be necessary to opt for the one that gave the maxi-
mal level of protection.
A second important step forward would be the identification
and cloning of the antigens that mediate the respective Th1 and
Th2 mechanisms induced by the RA vaccine.46 The task would
then be to find ways of formulating the relevant antigens to
invoke the desired responses, perhaps by means of two distinct
and sequential vaccination procedures. In short, the RA vaccine
model offers the possibility that, by manipulating the cytokine
milieu, we can bias the immune response in a protective direc-
tion. More specifically, if vaccination deflects the Th response
in either a Th1 or Th2 direction, it will produce a measure of
protection.46
Vigorous attention has been focused on potential vaccine
antigens from the schistosomula stages. Vaccination with
stage-specific schistosomula surface antigens has resulted in
protection (reductions in worm burden or egg counts) that
is usually less than 40%.47 Somewhat better protection has
been achieved using antigens shared between schistosomula
and adult schistosomes, including parasite-derived myosin
(63 kDa), paramyosin (97 kDa), triose phosphate isomerase
(28 kDa), glutathione-S-transferases (GSTs; 26 and 28 kDa), a
fatty acid–binding protein (14 kDa), and a 23-kDa surface pro-
tein. To date, most of the vaccine studies are conducted in mice
challenged with either or . However,
Parasitic disease vaccines 56 1157
because Asian schistosomiasis is associated with significant
animal reservoirs, including cattle, water buffaloes, and pigs,
several veterinary vaccine trials have been conducted in China
in these hosts, with a goal of developing an
transmission-blocking vaccine.47 To date, the recombinant
antigen-based vaccine trials have been conducted with
numerous adjuvants, including Freund's complete adjuvant,
alum, bacille Calmette-Guérin (BCG), saponin, Quil A, and
. Because of the suggested role of interferon
(IFN)- and Th1 cellular immune responses in mediating pro-
tection, there have been some efforts to bias the host cyto-
kine profile at the time of vaccination. Some success has been
reported in this regard using IL-12 as an adjuvant, and with
DNA immunizations.41,47 The molecular basis of immunity
with the schistosomula antigens is largely unstudied, except
for new evidence that paramyosin may function as an Fc recep-
tor for the parasite, allowing it to adsorb host immunoglob-
ulin to shield it from the host immune response. Therefore,
an anti-paramyosin immune response may somehow interfere
with the parasite's immune escape mechanisms.48
A second approach to vaccination against schistosomiasis
has been to target the fecundity of female adult schistosomes
so as to diminish egg excretion into target host organs. Success
with this approach has been reported by immunizing mice,
nonhuman primates, and large-animal reservoir hosts, includ-
ing pigs and water buffaloes, with 26-kDa and 28-kDa vari-
ants of a glutathione-S-transferase.47,49 If levels of egg excretion
can be reduced by vaccination of animal reservoir hosts, this
approach might be sufficient to reduce transmis-
sion.47 Alternatively, it has been proposed that egg deposition
and granuloma formation in the human host might be manipu-
lated by immunotherapy, because granuloma formation around
the schistosome egg depends heavily on host cytokines, includ-
ing tumor necrosis factor and Th2-associated cytokines.
During the late 1990s, the WHO's Special Programme for
Research and Training in Tropical Diseases (WHO/TDR) ini-
tiated murine trials of six vaccine candidates for

t
A 28-kDa GST, as noted earlier49,50

t
A 97-kDa paramyosin51

t
An irradiated larvae-associated vaccine antigen, the 62-kDa
IrV-5,52 which is a derivative of a 200-kDa molecule with
extensive homology with human myosin

t
A 28-kDa triose phosphate isomerase (TPI)53

t
A 23-kDa integral membrane antigen (Sm-23) that is part
of a superfamily of proteins that includes CD9 and TAPA-1,
first described in hematopoietic cells54

t
A 14-kDa fatty acid–binding protein that was thought to
also have protective immune cross-reactivity with the liver
fluke, 55
None of these candidate proteins reached the requirement
set by WHO/TDR of generating a 40% or better reduction in
challenge-derived worm burdens relative to nonimmunized
controls.56 However, it was recommended that work should
continue on these antigens, including progression to clinical
or veterinary trials. Recently, a membrane-spanning
surface protein (tetraspanin), known as -TSP-2, has been
identified that results in 60% to 70% worm burden reductions
in mice and was selectively recognized by IgG1 and IgG3 anti-
bodies in naturally resistant individuals, but not in chroni-
cally infected individuals living in endemic areas of Brazil.57
Furthermore, schistosomes treated with double-stranded
RNA encoding -TSP-2 display a vacuolated and disrupted
tegument and cannot fully develop in mice.58 The extracel-
lular domain of -TSP-2 has been expressed in yeast, and,
under the auspices of the Sabin Vaccine Institute, this anti-
gen will soon undergo clinical testing in Brazil as a vaccine
formulated with Alhydrogel (and possibly GLA as a second
1158 SECTION THREE Vaccines in development and new vaccine strategies
immunostimulant) for intestinal schistosomiasis.22 Additional
membrane-spanning and tegumental antigens may also enter
clinical testing in the coming years, as might -14, a fatty
acid–binding protein.59
For urogenital schistosomiasis, a group at the Institut Pasteur
in Lille, France, has produced the 28-kDa GST from
(Sh28GST) in yeast under current good manufacturing
practices and has embarked on clinical testing.49 Sh28GST was
selected on the basis of protection studies in primates vaccinated
with Sm28GST,49 and studies in humans, which showed that
IgG3 to Sh28GST correlates with an age-dependent decrease
60
in egg output in infection. Among human
volunteers, two subcutaneous injections of 100 g of recombi-
nant Sh28GST in alum resulted in an immune response com-
posed predominantly of IgG1 and IgG3.49 The anti-Sh28GST
antibody responses were shown to neutralize GST enzymatic
activity in vitro. Studies are underway to evaluate the clinical
efficacy of this vaccine, known as Bilhvax, in infected patients
in sub-Saharan Africa, with an aim to determine if there is syn-
ergy with praziquantel chemotherapy.
Larval tapeworm infections: cysticercosis and
echinococcosis
The pork tapeworm, is a major parasite of vet-
erinary and medical importance. When humans serve as an
intermediate host, can cause an important human
infection known as neurocysticercosis. Because transmission of
from pigs to humans can be interrupted, neurocysti-
cercosis has been identified as a potentially eradicable disease,
and increasing attention is being placed on efforts to control
transmission of the parasite. One way to control the disease is to
prevent infection in pigs by vaccinating them, thereby breaking
the parasite's life cycle and removing the source of infection for
humans. Several approaches to development of vaccines against
are being examined, 61
one of which is the applica-
tion of recombinant oncosphere antigens. Two different onco-
sphere antigens, designated TSOL18 and TSOL45, have been
evaluated, each of which has been shown to induce complete or
near complete protection against experimental challenge infec-
tion in four separate vaccine trials in pigs.61 Investigations have
begun toward characterizing various aspects of this vaccine
before undertaking controlled field trials. In a pilot field trial
conducted in pigs in Cameroon, a TSOL18 vaccine, composed
of 200 g of the recombinant antigen expressed in
and adjuvanted with Quil A (together with the anthelmin-
tic drug oxfendazole), was effective in eliminating
transmission to pigs.62 This pig vaccine–linked chemotherapy
approach may ultimately be effective in eliminating or eradi-
cating cysticercosis. 63
Efforts are also underway to
develop canine vaccines for the adult stages of
as a means to interrupt transmission of the disease
to humans and prevent hydatid cysts.64
Antiprotozoan vaccines
Significant progress has been made over the past 5 years in
the development of first- and second-generation vaccines for
leishmaniasis, and amebiasis. Both diseases are major causes
of mortality and morbidity in developing regions of Africa,
Latin America, and Asia. In contrast, preventive vaccine devel-
opment efforts for Chagas disease (American trypanosomia-
sis) has been curtailed because of successful vector control
efforts in the southern cone of South America,65 and because
of theoretical concerns about the role of autoimmunity in the
pathogenesis of Chagas heart disease. More current thinking
highlights the importance of parasitic persistence in Chagas
disease pathogenesis.66 New efforts are being considered for the
development of a therapeutic Chagas disease vaccine. The phe-
nomenon of antigenic variation67 has largely thwarted vaccine
development for African sleeping sickness (African trypanoso-
miasis). (Antimalaria vaccines are discussed in Chapter 53.)
Leishmaniasis and Chagas disease
species are flagellated kinetoplastid protozoan para-
sites transmitted by the bite of a female sandfly. There are cutane-
ous (CL), mucocutaneous (MCL), and visceralizing (VL) forms of
human leishmaniasis. CL and MCL in the Western Hemisphere
(predominantly in Central and South America) are caused by
members of the and
complex, whereas CL in the Eastern Hemisphere (pre-
dominantly in India, central Asia, and parts of Africa and the
Middle East) is caused by members of the
and complex. VL, also known as kala-azar, is
caused by in India, Bangladesh, Nepal, and
China; by in central Asia, North Africa, and
southern Europe; and by in Latin America.
Members of the complex are important emerging
opportunistic pathogens in patients with AIDS. It is estimated
that the worldwide prevalence of leishmaniasis is 12 million cases,
with approximately 350 million people at risk for infection.68
Leishmanization, the ancient Middle Eastern and cen-
tral Asian practice of injecting live parasites as a
means of actively immunizing against disease, predates vacci-
nation and may be as old as variolation. It relies on inoculating
a person with live parasites using a thorn or other sharp instru-
ment to artificially create a cutaneous lesion. Alternatively, an
arm or other body part is deliberately exposed to the bite of
sandflies.11 The procedure causes mild disease (typically 1 to
2 cm in diameter) at an unexposed site, frequently the arm or
buttocks, that heals over a period of 3 to 4 months. Successful
leishmanization prevents the possibility that a cosmetically dis-
figuring lesion (“Delhi boil” or “oriental sore”) might appear on
the face. Leishmanization strategies were developed in Israel
by Saul Adler at Hebrew University and in the former Soviet
Union.12 One such leishmanization “vaccine” is registered in
Uzbekistan.11 In the 1980s, leishmanization was reintroduced
as an emergency public health measure during the Iran-Iraq
war, when approximately 2 million people were inoculated.69
The Iranian leishmanization program was subsequently halted
when it was learned that up to 3% of the induced lesions lasted
for more than a year; in some instances, the lesions never
healed, even with subsequent antiparasitic antimonial che-
motherapy.69 To date, leishmanization is the only vaccine with
proven efficacy in humans.12 Several live attenuated strains of
12
are also under development as potential vaccines.
Among the hurdles that potentially block further refinement
of leishmanization and related practices are the difficulties in
standardizing the virulence of the vaccine and the constant risk
of developing severe and persistent lesions, especially in indi-
viduals with HIV/AIDS and other immune deficiencies.12,68
In an effort to establish a safer anti- immuniza-
tion procedure, Iran launched a national vaccine development
program during the early 1990s. A first-generation vaccine
comprising killed parasites was prepared under
good manufacturing practices by the Razi Vaccine and Serum
Research Institute (Hessarak, Iran). The cell bank for the vac-
cine was derived from the same strain of used previ-
ously for the wartime leishmanization program. The
parasites were killed by autoclaving, and they were then formu-
lated with BCG. In a randomized, double-blind trial, the safety
and efficacy of a single injection of the vaccine was compared
with a BCG control in 3,637 schoolchildren.69 Children with
a prior history of CL or those with a positive skin test reac-
tion against leishmanial antigen were excluded from the study.

The Iranian trial demonstrated the vaccine's safety, although its
efficacy was not apparent until 6 months after the injection. It
has been suggested that there was some protective effect from
BCG in the immediate postvaccination period. Also of inter-
est was the observation that protection was better among boys
who, because of Iranian dress and customs, presumably are
exposed to a greater number of sandfly bites than girls.69 Based
on these results, additional trials are underway to evaluate
multiple injections of the killed vaccine, as well as new formu-
lations with alum and IL-12. A trial to evaluate the killed vac-
cine against is also underway in Sudan. In South
America, killed and vac-
cines are being produced by BIOBRAS (Brazil) and Instituto de
Biomedicina (Venezuela), respectively. In reviewing the data
from several clinical trials of vaccines composed of whole-par-
asite antigens, Coler and Reed68 conclude that the efficacy var-
ies from 0% to 75% against CL, but it is only 6% against VL. A
phase 3 clinical trial in Ecuador and Colombia showed that the
vaccine was safe but not efficacious.12 A meta-
analysis concluded that killed whole-parasite vaccines do not
confer significant protection against human leishmaniasis.70
Earlier studies on whole-organism leishmania vaccines have
stimulated efforts by several laboratories to reproduce vac-
cine protection using second-generation recombinant subunit
antigens. Several antigens have been identified on the basis
of their surface localization, recognition by T-cell clones, and
immunoscreening using patient sera.68 A number of candidates
have shown promise in mice, particularly when formulated
with IL-12 or other costimulators of Th1 immunity. In some
cases, vaccination with antigens (either alone or in combina-
tion) encoded by plasmid DNA elicits more potent and dura-
ble immunity. Another promising approach has been to develop
vaccines composed of sandfly salivary proteins.71
Development of a human leishmaniasis vaccine has been
advanced most by a group at the Infectious Disease Research
Institute (IDRI) in Seattle, Washington. The IDRI group has
developed a polyprotein composed of three priority candidate
antigens, thiol-specific antioxidant (TSA),
Stress-Inducible Protein 1 (LmSTI1), and
elongation initiation Factor (LeIF), fused in tan-
dem.68 In a patient with refractory ML, these antigens had pre-
viously been shown to stimulate an immune response (when
administered with granulocyte-macrophage colony–stimulating
factor [GM-CSF]) that resulted in resolution of the condition.72
The polyprotein is known as Leish-111f, and as a vaccine it
achieves long-lasting protection in Balb/c mice when formu-
lated with a proprietary monophosphoryl lipid A known as
MPL-SE.68 In preclinical testing, the Leish-111f + MPL-SE vac-
cine has been shown to be safe in five animal species.68 The
clinical development plan is to perform safety and efficacy stud-
ies in therapeutic and prophylactic applications, with multiple
species of in several countries. The human vaccine
is being evaluated for prophylactic and therapeutic indications
in clinical trials in South America, India, and Sudan. Safety and
immunogenicity was demonstrated in a phase 1, double-blind,
dosage-escalation study at each dosage of the protein (10, 20,
and 40 g), and therapeutic trials of ML and CL are underway
in Peru and in Brazil and Colombia, respectively, to evaluate
the vaccine in combination with standard chemotherapies.68
For adult patients with ML in Peru, the Leish-111f + MPL-SE
vaccine, when used in combination with sodium stibogluco-
nate, was safe, was well tolerated, and induced both humoral
and cell-mediated immune responses.73 In a phase 1 trial for
the prevention of CL in healthy Colombian adult volunteers
with no history of leishmaniasis, the Leish-111f + MPL-SE
vaccine was shown to be safe and immunogenic,74 and among
44 adult patients with CL in Brazil, the vaccine was safe and
immunogenic and shortened the time to cure when used in
combination with meglumine antimoniate chemotherapy.75
Parasitic disease vaccines 56 1159
For VL, both Leish-111f + MPL-SE and a sterol
24-C-methyltransferase formulated in adjuvant are also under
development.76,77 In parallel with these studies, efforts to mine
genomes to identify new candidates for ML, CL,
and VL are underway,78,79 and some additional success has been
reported in the development of transmission-blocking vaccines
based on immunization with sandfly salivary gland proteins.80,81
In regions where zoonotic transmission of leishmaniasis
from dogs is common, there has been some additional interest
in developing a canine transmission-blocking vaccine.82,83 The
fucose-mannose ligand (FML) antigen of in com-
bination with saponin (Leishmune) was shown to induce up
to 97% protection against zoonotic leishmaniasis.83 The Leish-
111f + MPL-SE vaccine and a second Leish-110f + MPL-SE
vaccine are also under development as therapeutic vaccines for
canine visceral leishmaniasis.84,85
Chagas disease (American trypanosomiasis), like leish-
maniasis, is another kinetoplastid of great public health and
economic importance. An estimated 8 to 9 million people
are infected in Latin America, where it is a leading cause of
disabling cardiomyopathy,86 and up to several hundred thou-
sand imported infections are in the United States and Europe.
A recent cost analysis indicates that a cost-effective Chagas
disease vaccine would have substantial economic benefit.87
Several experimental vaccine candidates, including DNA vac-
cines, are under development as preventive and therapeutic vac-
cines,78,88–94 but none have progressed to clinical trials. In mice,
protective immunity is associated with strong Th1 responses
to defined antigens. Because of the ques-
tionable benefit of currently available drugs for the treatment
of seropositive individuals in order to delay or prevent Chagas
cardiomyopathy, there is an urgent need for new approaches,
including immunotherapy or therapeutic vaccines that stimu-
94
late –specific Th1 immunity.
Amebiasis
Amebiasis secondary to intestinal and hepatic infection with
is prevalent throughout the develop-
ing nations of the tropics. Human infection occurs through the
ingestion of parasite cysts and subsequent invasion by tropho-
zoites. Amebic colitis results from ulcerating mucosal lesions
that result from chemical invasion caused by the release of
parasite-derived proteases and hyaluronidases. Metastatic spread
to the liver occurs when the trophozoites gain access to the
afferent circulation that drains the colon into the portal vein.95
Evidence from a cohort of Bangladeshi children suggests that
mucosal IgA antibody directed against the major 170-kDa ame-
bic adherence lectin correlates with resistance to reinfection by
.96 Additional studies have shown that the native lectin
protects gerbils against challenge infections.97 These
observations have stimulated efforts to develop the amebic lectin
or its peptide derivatives as a first-generation vaccine.97,98 To date,
no human testing of a vaccine for amebiasis has been initiated.99
Toxoplasmosis
Toxoplasmosis is one of the most common infections of
humans, with some estimates suggesting that 6 billion people
100
have been infected with . Currently, the
highest global prevalence and incidence may occur in South
America,100 although there is a dearth of information on the
true global disease burden of human infection. In the
United States, an estimated 1 million new infections occur each
year, resulting in an estimated 20,000 cases of retinal infec-
tion and 750 deaths.100 Several research groups have developed
promising approaches for a human toxoplasmosis vaccine.
Because, like and , is an intracel-
lular parasite, such efforts are primarily directed at stimulating
 SECTION THREE
–specific Th1 immunity and generating long-lived IFN-
–producing CD8 + T cells.101
In preclinical studies, several approaches have shown prom-
ise. They include viral vectors and constructs capable of express-
102
ing antigens, DNA vaccinations,103 live attenuated
parasites,104 and pairing recombinant antigens with Th1-
stimulating adjuvants.104 Transmission-blocking approaches
from zoonotic reservoir hosts including cats and livestock used
Access the complete reference list online at http://www.expertconsult.com
22. Hotez PJ, Bethony JM, Diemert DJ, et al. Developing vaccines to combat
hookworm infection and intestinal schistosomiasis. Nat Rev Microbiol
8:814–826. doi:10.1038/nrmicro2438 2010.
49. Capron A, Riveau G, Capron M, et al. Schistosomes: the road from
host-parasite interactions to vaccines in clinical trials. Trends Parasitol
21:143–149. doi:10.1016/j.pt.2005.01.003 2005.
59. Tendler M, Simpson AJ. The biotechnology-value chain: development of
Sm14 as a schistosomiasis vaccine. Acta Trop 108:263–266. doi:10.1016/j.
actatropica.2008.09.002 2008.
63. Lightowlers MW. Eradication of cysticercosis: a role
for vaccination of pigs. Int J Parasitol 40:1183–1192. doi:10.1016/j.
ijpara.2010.05.001 2010.
68. Coler RN, Reed SG. Second-generation vaccines against leishmaniasis.
Trends Parasitol 21:244–249. doi:10.1016/j.pt.2005.03.006 2005.
73. Llanos-Cuentas A, Calderon W, Cruz M, et al. A clinical trial to evaluate the
safety and immunogenicity of the LEISH-F1+MPL-SE vaccine when used
in combination with sodium stibogluconate for the treatment of mucosal
leishmaniasis. Vaccine 28:7427–7435. doi:10.1016/j.vaccine.2010.08.092 2010.
for food have also been proposed.104,105 There is an urgent need
to better define the target product profile for toxoplasmosis
vaccines with respect to target populations—that is, prevent-
ing congenital infections versus producing therapeutic vaccines
for immunocompromised individuals—and to gain consen-
sus on lead candidate antigens, as well as to determine how
to best overcome genetic restrictions to the human immune
response.106 No human vaccine trials are underway.
74. Velez ID, Gilchrist K, Martinez S, et al. Safety and immunogenicity of a
defined vaccine for the prevention of cutaneous leishmaniasis. Vaccine
28:329–337. doi:10.1016/j.vaccine.2009.10.045 2009.
75. Nascimento E, Fernandes DF, Vieira EP, et al. A clinical trial to evaluate the
safety and immunogenicity of the LEISH-F1+MPL-SE vaccine when used in
combination with meglumine antimoniate for the treatment of cutaneous
leishmaniasis. Vaccine 28:6581–6587. doi:10.1016/j.vaccine.2010.07.063
2010.
77. Kumar R, Goto Y, Gidwani K, et al. Evaluation of ex vivo human
immune response against candidate antigens for a visceral leishmaniasis
vaccine. Am J Trop Med Hyg 82:808–813. doi:10.4269/ajtmh.2010.09-
0341 2010.
78. Dumonteil E. Vaccine development against and
species in the post-genomic era. Infect Genet Evol 9:1075–1082.
doi:10.1016/j.meegid.2009.02.009 2009.
91. Gupta S, Garg NJ. Prophylactic efficacy of TcVac2 against
in mice. PLoS Negl Trop Dis 4:e797. doi:10.1371/journal.pntd.0000797
2010.
1
 SECTION THREE: Vaccines in development and new vaccine strategies
vaccines
More than a century ago, the Scottish surgeon Alexander
Ogston first noticed an association between skin abscesses
and an organism that formed structures resembling clusters of
grapes when seen by light microscopy.1 The organism he saw,
, is now known to be responsible for
many serious community-acquired and nosocomially acquired
infections and several more recently recognized toxin-mediated
diseases. It is also the most frequently isolated bacterial
pathogen from patients with hospital-acquired infections. The
defining characteristics of this species are the production of the
extracellular enzyme coagulase and protein A. Staphylococci
lacking coagulase are grouped under the designation coagulase-
negative staphylococci (CoNS). Once thought to be nonpatho-
genic constituents of human bacterial flora, the approximately
16 species of CoNS isolated from humans are now also known
to cause nosocomial infections and urinary tract infections
in children, catheter-related infections, prosthetic joint infec-
tions, endocarditis, and septic episodes in premature infants
and other immunocompromised hosts. Although they can be
important pathogens in certain settings, vaccine development
has been focused primarily on , which is therefore the
subject of this chapter.
Ecology
Staphylococci usually have a symbiotic relationship with their
human hosts. Asymptomatic colonization occurs
intermittently in children and adults: 10% to 40% are asymp-
tomatically colonized.2–6 Nasal carriage is widely believed to be
a risk factor for invasive disease.7,8 In addition, the skin, hair,
nails, axillae, perineum, rectum, and vagina (in about 10% of
menstruating women) may be colonized.9–11 Children have
somewhat higher colonization rates than adults. Three patterns
of colonization by have been observed: about 20% of
the population is persistently colonized, 60% is intermittently
colonized, and about 20% is never colonized.12 Hospital person-
nel, persons with chronic skin conditions, implanted medical
devices, those who indulge in intravenous substance abuse,13
patients with type 1 diabetes, patients undergoing hemodialy-
sis, and patients with AIDS have a higher rate of asymptomatic
colonization than do persons in the general commu-
nity.2 CoNS isolates colonize human skin universally,14 and
multiple strains may be isolated from the same person.
An epidemic of infections in the US with onset in
the mid- to late 1990s has intensified interest regarding this
important pathogen. The emergence of several genetic
backgrounds circulating in the community as methicillin-resis-
tant clones, so-called community-acquired methicillin-resistant
Robert S. Daum
57
(CA-MRSA), has exponentially increased the burden
of disease in many areas.15 is now the most frequent
pathogen isolated from children.16 The rate of methicillin-
resistant asymptomatic colonization has increased and was as
high as 22% in children in Texas.17 Virulence determinants in
these CA-MRSA strains have been recognized that are appar-
ently new to the species,18 increased in frequency, or increased
in expression.19 This increase in the clinical burden of
disease associated with this epidemic has initiated discussions
about the possible need for a universal immunization strategy.
The clinical spectrum
is a major cause of morbidity and mortal-
ity. It is the most virulent member of the genus
Unlike other members of the genus, is well endowed
with a variety of virulence factors (Figure 57-1).
Superficial and invasive infections occur in pre-
viously healthy persons. Infections of the skin range from
impetigo to abscess formation, cellulitis, or lymphadenitis, par-
ticularly of the cervical lymph nodes in children or adjacent to
an infectious focus. also may cause several important
ocular infections, including conjunctivitis, preseptal cellulitis,
and endophthalmitis. is an important cause of endo-
carditis. It is responsible for 25% to 35% of endocarditis cases.20
Its clinical manifestations may be particularly severe, and the
infected heart valves may have been previously normal, espe-
cially when the mitral or aortic valves are involved. Pericarditis
may be an isolated syndrome or may accompany endocarditis.
Hematogenous seeding of a bone or joint may result in osteo-
myelitis, septic arthritis, or even bursitis. is a cause
of several respiratory tract infections, including an occasional
otitis media and pneumonia. The latter may be a severe, nec-
rotizing process with high mortality. Central nervous system
infections are infrequent and usually involve an assisted portal
of entry for the organism, such as extension from an infected
sinus, a dermal sinus, or a meningomyelocele. Central nervous
system infectious syndromes also include subdural and epidural
empyemas. A spinal epidural abscess adjacent to the dura also
may be caused by . may rarely infect the uri-
nary tract or be recovered from the urine of a patient with high-
grade bacteremia or a renal abscess.
frequently complicates surgical pro-
cedures in which the integument was breached and medi-
cal interventions in which indwelling foreign bodies are used
in management. The insertion of plastic, metal, or Gore-Tex
devices provides an opportunity for to persistently
adhere. Success in removing the organisms from the inserted
1162 SECTION THREE Vaccines in development and new vaccine strategies
Diagram of virulence factors of
N Engl J Med 339:520-532, 1998.)
device has posed a great challenge, even when the isolate is sus-
ceptible to the antibiotic used. Thus, certain patients, such as
those needing hemodialysis, those with indwelling venous cath-
eterization, indwelling intravascular Gore-Tex patches, artificial
prostheses, or cerebrospinal fluid flow diversionary devices, are
at high risk for infections.
A variety of veterinary infections are caused by as
well. For example, mastitis in cows is an economically impor-
tant disease in dairy ruminants. Horses, chickens, dogs, and
cats are among the animals known to be colonized or infected
with with documented transmission to humans from
horses.21 Mounting evidence suggests that pigs may represent
a new important reservoir for CA-MRSA strains. ST398 is the
most commonly reported sequence type among livestock in
Europe and the United States. Transmission to pig-farm work-
ers has occurred, although invasive disease has rarely been
documented.
Because of its extensive propensity to cause disease in
humans and animals, has been a subject for investi-
gations related to veterinary vaccine development.
is also the cause of a number of tox-
inoses with clinical manifestations that reflect the effects of one
or more elaborated and released exotoxins. Examples include
toxic shock syndrome (TSS), scalded skin syndrome, and food
poisoning.
Attention has been called to septic illnesses caused by
with multiorgan failure that likely reflect the activity
of one or more toxins. Examples include so-called severe
sepsis,22 an illness with a sometimes fulminant course, purpura
fulminans,23 and the Waterhouse-Friderichsen syndrome.24
The problem of antimicrobial resistance in
S. aureus: epidemic community-acquired
methicillin-resistant disease
Before the introduction of antimicrobials in the early 1940s, the
mortality rate of invasive infections was about 90%;
this rate decreased markedly following the introduction of peni-
cillin G into clinical practice.25 Almost immediately, however, a
few isolates were noted to be resistant to penicillin.26–28
(Modified from Lowy FD. Medical progress: infections.
By the end of the 1940s, most strains isolated from hospital-
ized patients were penicillin-resistant;29 currently, resistance
rates exceed 90%, rendering penicillin virtually ineffective in
treating infections. The mechanism of resistance
is the elaboration of a -lactamase, usually encoded by a
transposon borne on a plasmid; there is cross-resistance with
other -lactams that are susceptible to -lactamase hydrolysis.
In the early 1960s, methicillin, representative of a new class
of semisynthetic -lactam antibiotics that are relatively resis-
tant to hydrolysis by staphylococcal -lactamases, was mar-
keted. However, resistance to it was immediately recognized.30,31
Such resistance is still referred to as MRSA (methicillin-resistant
), although methicillin is no longer in clinical use.
MRSA isolates survive -lactam exposure by elaborating a
peptidoglycan-synthesizing enzyme called penicillin-binding
protein (PBP) 2a that, in concert with native PBP2, allows
peptidoglycan synthesis despite the presence of a -lactam com-
pound. This resistance mechanism conveys cross-resistance
to all -lactam antibiotics, including cephalosporins with the
exception of the newly licensed ceftaroline.31a Since their
recognition, the prevalence of MRSA isolates was slowly but
relentlessly increasing,32–34 but has now probably plateaued.
Institutional prevalence rates exceeding 45% are not uncom-
mon.35 A new development has been the recognition of epidemic
community-acquired MRSA infections in children and adults
in many regions of the United States, elsewhere in North
America, Western Europe, Asia, Australia, and South America
who lack traditional MRSA risk factors such as frequent
contact with health care facilities.36 The MRSA isolates
responsible for these community-acquired infections are less
resistant to multiple antibiotics than their hospital counter-
parts; the genetic backgrounds of these isolates differ from
their hospital counterparts, and the isolates contain novel,
37,38
smaller chromosomal elements called SCC IV or V
that may be more transferable than the larger SCC
elements found in hospital MRSA isolates. CA-MRSA isolates
have been associated with dramatic upsurges in the prevalence
of infections in general and rates of MRSA infections
in particular.39,40
Vancomycin, a glycopeptide antibiotic, until recently had
been the only agent to which MRSA isolates had remained
uniformly susceptible. This susceptibility certainty eroded,

however, with the recognition of glycopeptide intermediate-
resistant (minimal inhibitory concentration of van-
comycin, 2 and 16 g/mL) isolates in Japan,41 the United
States,42–45 and several other countries.46–51 Often, patients with
glycopeptide intermediate-resistant isolates have
been receiving dialysis and, therefore, have prolonged low-
level serum vancomycin levels, an in vivo environment simi-
lar to that used to select for vancomycin-intermediate resistant
mutants in the laboratory,52,53 and have failed vancomycin
treatment. More recently, higher level vancomycin-resistant
isolates have been isolated from patients in the United
States54 with minimal inhibitory concentrations of vancomycin
of more than 16 g/mL. These isolates bear genes from vanco-
mycin-resistant enterococcal strains that produce a structurally
altered peptidoglycan precursor that does not bind vancomycin.
The occurrence of these isolates represents important events
that will greatly impact health care throughout the world.55–57
The ever-evolving resistance mechanisms in suggest
that no antimicrobial therapy strategy will be unmet with a
resistance strategy response by the staphylococcus. The situa-
tion with antibiotic resistance in is certainly one issue
that has spurred vaccine development.
Resistance to other antimicrobials active against
such as, linezolid,58,59 daptomycin,60 and mupirocin61 have all
been identified as clinical concerns. This continuing saga of
antimicrobial resistance in and the slowing of the
development of new antimicrobials is reminiscent of the clini-
cal alarms sounded by resistance of
to ampicillin and chloramphenicol and resistance of
to penicillin which suggested that
infections caused by these important pathogens had become
increasingly difficult to treat. In both instances, the deploy-
ment of an effective vaccine muffled many of the increasing
concerns. Effective prevention outpaced new concerns about
therapy.
Microbiology and pathogenesis
Staphylococci are hardy aerobic or facultatively anaerobic gram-
positive bacteria that can persist in distressed environments
such as acidic conditions, high sodium concentrations, and
wide temperature variations; they can survive on fomites, in
dust, and on clothing from several days to more than a week.
Once infection is established, local and systemic
effects result from direct invasion, hematogenous dissemina-
tion, and/or toxin release.
Though previously the organisms were thought to be nonen-
capsulated, it is now clear that many clinical isolates
and at least some CoNS isolates have a polysaccharide capsule,
sometimes called a microcapsule because of its thinness and
adherence to the bacterial cell surface. The prevalence of two
capsular polysaccharide types (5 and 8) in most collections of
clinically important human and veterinary isolates suggests
an important role for these polysaccharides in pathogenesis,
although, to date, its nature is uncertain.
The cell wall of is composed of capsular polysac-
charide, peptidoglycan, lipoteichoic acid (LTA), ribitol wall tei-
choic acid, and numerous surface proteins. Protein A is an
important surface protein to which the Fc region of the IgG
molecule binds.62–65 IgG antibody binding to the staphylo-
coccal cell surface in this nonphysiologic manner decreases
the efficiency by which isolates are opsonized and
phagocytosed.66–69
The virulence of is due to a combination of many
elaborated virulence proteins that include extracellular prod-
ucts, such as -, 70 -, 71 -, 72 and - 73 hemolysins (or toxins);
leukocidins;74 proteases;75 lipase76 deoxyribonuclease; a fatty
acid-modifying enzyme;77 and hyaluronidase.78 Coagulase is
Staphylococcus aureus vaccines 57 1163
an extracellular protein that binds host prothrombin to form
staphylothrombin; this, in turn, activates thrombin and results
in the formation of fibrin from fibrinogen.79 A small amount
of coagulase remains cell-bound. Clumping factors are distinct
cell surface proteins that bind fibrinogen, thereby producing
the typical clusters of staphylococci when mixed with plasma.80
The pathogenic roles of coagulase and the clumping factors
have been uncertain; they may protect against host defenses
by causing localized clotting, and clumping factors may aid in
adherence to traumatized skin, endothelial structures, and for-
eign surfaces. Recognition of this role for clumping
factors has prompted investigation into their use as potential
vaccine antigens, as reviewed subsequently.
-Hemolysin, the best studied of the exotoxins, hemolyzes
erythrocytes, necroses skin, and causes the release of cyto-
kines and eicosanoids that can produce shock. It is lethal when
injected into animals,81 and mutants lacking -toxin are less
virulent.82 However, pathogenic isolates that do not
produce -hemolysin have been identified. -Toxin is a sphin-
gomyelinase that damages membranes rich in this lipid, but its
role in pathogenesis is uncertain. Although most human iso-
lates do not express -toxin, its elaboration by a majority of
isolates associated with mastitis in cows suggests a potentially
important pathogenic role, at least in that entity.83 -Toxin and
leukocidins are synergohymenotropic toxins that damage mem-
branes of certain cells as a result of in-concert action by several
elaborated proteins. Interestingly, the transcription of the indi-
vidual protein components is separate.72,84 The high percentage
of -leukocidin-producing isolates from necrotizing skin infec-
tions suggests the importance of this toxin in the production
of dermonecrosis. The Panton-Valentine leukocidin has been
found in cases of severe community-acquired pneumonia in
children; the majority of those cases were fatal, and all were
marked by hemorrhagic necrosis of the lung.85 Controversy per-
sists as to the role of this toxin in pathogenesis.
Other toxins are produced by specific isolates that
mediate certain clinical syndromes. For example, epidermolytic
(exfoliative) toxins A and B cause sloughing of skin that occurs
in staphylococcal scalded skin syndrome.86 The mechanism
of epidermal splitting is unknown, although the toxins have
esterase and, possibly, protease activity.87 The clinical charac-
teristics of toxic shock syndrome (TSS) can be attributed to the
actions of two different types of toxins: TSS toxin-1 (TSST-1)
and certain enterotoxins. TSST-1 can be elaborated by
and is responsible for most TSS cases and nearly all of cases
associated with menses; antibody to TSST-1 is protective,88 but
patients with TSS often do not make antibody during conva-
lescence from TSS. The family of enterotoxins may
cause vomiting and diarrhea when ingested and is responsible
for staphylococcal food poisoning; these enterotoxins also may
cause TSS when their entry is via a nongastrointestinal route.
Enterotoxins B and C account for about half the TSS cases not
associated with menses. The enterotoxins and TSST-1 can have
superantigen activity, that is, they can stimulate T cells nonspe-
cifically, without specific antigen recognition.89 This nonspecific
antigen T-cell stimulation results in release of immunomodula-
tors, particularly cytokines, an event that produces the clinical
picture of TSS. Superantigens bind to class II MHC receptors
but recognize only the V element of the receptor, resulting in
nonspecific polyclonal T-cell activation.90
Locally, organisms may invade or necrose tissue
and evoke a potent inflammatory response that largely affects
the interaction of bacteria and polymorphonuclear leukocytes.
Abscess formation is common, with a necrotic center consist-
ing of pus and a fibrin wall that makes penetration by antibiot-
ics difficult. The infection may spread locally by the formation
of sinus tracts and secondary abscesses. Hematogenous dissem-
ination and infection of distant bones, joints, cardiac valves, or
other tissues may result.
1170 SECTION THREE Vaccines in development and new vaccine strategies

Bacteriology, virulence factors, and pathogenic
mechanisms
An assortment of secreted and surface-expressed molecules
mediates the four pathogenic processes that characterize acute
infection: adhesion, immune evasion, inflammation, and tis-
sue invasion (Figure 58-1). The virulence factors that are being
explored as possible vaccine antigens will be highlighted.
GAS adhere to host epithelial cells in the pharynx, cervix,
or skin by binding to plasma and matrix proteins,32 forming
microbial aggregates and producing a biofilm that protects
against host defenses and nutrient deprivation.33 Among the
best described adhesins are lipotechoic acid, M protein,34 the
hyaluronic acid capsule,35 serum opacity factor,36 and fibronec-
tin binding proteins such as SfbI37 and FBP54.38 R28 is found in
GAS strains associated with puerperal fever and is thought to
promote binding of GAS to cervical epithelium.39
In 2005, the GAS pilus was first identified and has subse-
quently been shown to contribute to adhesion and biofilm forma-
tion.33,40 Interestingly, these filamentous structures protruding

Major surface structures of GAS. The main categories of proteins that have been exploited in vaccine development are shown. Most
proteins are anchored at the C terminus via a highly conserved LPTP consensus motif followed by a hydrophobic region that spans the cell
membrane and terminates with a short charged tail within the cytoplasm. After translocation to the surface, the LP TP motif is cleaved and the mature
molecule is anchored into peptidoglycan. The majority of proteins are multifunctional, usually with domains that bind molecules found in extracellular
matrix, such as fibronectin (SfbI), but also domains that exhibit enzymatic activity, such as C5a peptidase, or have other functions. In the case of
proteins with multiple repeats, such as M protein, each repeat domain (designated as A, B, C, and D) has a different function. The pepsin cleavage site
of M protein is marked with a crooked arrow. Extracellular proteins that exhibit toxin and/or superantigen properties include streptococcal pyrogenic
exotoxin (Spe) A and C, which have been targeted for vaccine development, as well as anti-streptolysin O (ASO), streptokinase, and hyaluronidase.

from the cell surface bear the T antigens described by Lancefield
and thus present a potential new avenue for vaccine develop-
ment. All GAS isolates tested thus far carry and express the pilus
genetic locus.33
After colonization, GAS persists in the host by evading
immune surveillance. GAS deploys a multipronged approach to
disarm the innate immune response. C5a peptidase (also called
SCPA, or surface-bound C5a peptidase) inactivates a chemo-
kine of the complement system41,42 while SpyCEP (also called
ScpC or Spy 1416), a cell envelope serine protease, cleaves
the neutrophil attractant interleukin-8 (IL-8).43 Extracellular
DNAases degrade chromatin webs produced by neutrophils to
entrap GAS bacteria.44 The hyaluronic capsule promotes resis-
tance to host defense peptides and to phagocyte killing within
the extracellular neutrophil traps45,46 and streptococcal inhib-
itor of complement inhibits complement-mediated lysis.47 M
protein, encoded by , impedes neutrophil phagocytosis in
the absence of type-specific antibody by binding plasma pro-
teins and inhibiting opsonization via the alternate complement
pathway.48–52 The antigenic heterogeneity of the amino termi-
nus of M protein further enables GAS to escape recognition by
specific antibody.
GAS elaborates several extracellular products, including
streptolysin O, deoxyribonuclease B, streptokinase, and hyal-
uronidase, which induce tissue liquefaction and facilitate inva-
sion.53 A secreted serine carboxylic esterase (sse, Spy 1718)
contributes to invasive skin infection and efficient systemic dis-
semination of GAS from the skin.54
M protein and the streptococcal pyrogenic exotoxins (SPEs)
have potent pro-inflammatory properties.55,56 The SPEs form
a family of superantigens that induce profound immunologic
changes57,58 as well as fever, tissue damage, and clinical mani-
festations of endotoxic shock.59–61 SPE A and C, formerly known
as erythrogenic toxins because of their association with the rash
of scarlet fever, are bacteriophage encoded and contain distinct
T-cell receptor and class II MHC binding sites.62 In the United
States, STSS has been associated with SPE A–producing strains.
The pathogenesis of the nonsuppurative postinfectious
sequelae of GAS is not well understood. ARF follows approxi-
mately 3% of episodes of untreated GAS pharyngitis under epi-
demic conditions,63 but is considerably less frequent in endemic
situations.64 Observations of aboriginal populations experienc-
ing high rates of ARF have led some investigators to suggest
that skin infections lead to ARF, but this remains unproven.65
Strains causing pharyngitis and ARF bear different M types and
other antigenic and virulence properties than those that cause
impetigo/pyoderma, although some overlap exists.28,66 A proven
association between skin strains and ARF could alter the neces-
sary composition of M protein–based vaccines.
Appreciation of the structure of M protein helps to explain
its function as the major virulence factor of GAS and also the
major protective antigen.67 These -helical coiled-coil fibrillar
rods consist of two polypeptide chains anchored in the cell wall
peptidoglycan by a C-terminus LPTP motif (Figure 58-1).68
Each polypeptide contains up to four internal repeating seg-
ments (labeled A-D). The C repeats contain epitopes that are
exposed on the surface of the cell wall and are largely conserved
among different M types.69 During infection, opsonic antibod-
ies that confer protection are directed predominantly at the N
terminus,70 a highly variable region encoded by the gene
that constitutes the basis for Lancefield's serotyping classifica-
tion of GAS. Epitopes found within the B repeats and adjacent
segments of the A and C repeats cross-react with human anti-
gens and have been proposed as a mechanism in the pathogen-
esis of ARF. Moreover, a segment within the B-repeat region
bears homology with known superantigens.71
The concept of molecular mimicry has been invoked to
explain the pathogenesis of ARF and PSGN. The M protein
bears antigenic similarities with human fibrillar proteins, such
Streptococcus group A vaccines 58 1171
as cardiac muscle myosin and molecules of the extracellular
matrix of joints and kidneys.72–75 A mechanism for carditis is
postulated whereby GAS infection stimulates cross-reactive
antibodies that bind to the valvular endothelium resulting in
inflammation75 and infiltration by CD4+ T cells that recognize
GAS M protein and cardiac antigens.76,77 Antibodies evoked
by the immunodominant epitope of group A carbohydrate,
-acetylglucosamine, cross-react with GM1 gangliosides on the
surface of neuronal and valvular endothelial cells and have been
proposed in the pathogenesis of Sydenham's chorea78 and car-
ditis,75 respectively. Seroreactivity with components of the cell
membrane, SPE B, and other GAS antigens has been suggested
in the etiology of PSGN.53,79
It is important to note that antibodies that cross-react with
human tissues and somatic components of GAS are found in
many healthy children, so their presence alone does not explain
the nonsuppurative sequelae of GAS.80,81 Further elucidation
of the pathogenesis of nonsuppurative sequelae will facilitate
rational development of a safe GAS vaccine.
Treatment and prevention with antimicrobials
In the 1950s, Denny et al.82 demonstrated that a course of peni-
cillin prevents more than 90% of ARF episodes (primary preven-
tion) if initiated within 9 days of the onset of GAS pharyngitis.
Furthermore, the observation that nearly 30% of patients with
rheumatic carditis experience additional valvular injury when
reinfected with GAS83 prompted recommendations to admin-
ister prophylactic penicillin during the years when the risk of
exposure to GAS is greatest (secondary prevention). In contrast,
although antibiotics speed recovery of skin lesions, they do not
appear to prevent glomerulonephritis following either pharyn-
gitis or impetigo. Severe invasive infections are treated with a
combination of penicillin and an antibiotic that inhibits pro-
tein synthesis such as clindamycin plus surgical debridement
of necrotic tissue.84 Intravenous immunoglobulin therapy has
been advocated as ancillary therapy for STSS based on its abil-
ity to neutralize superantigen toxins, but its efficacy remains
inconclusive.85
There are well-recognized limitations to the treatment of
GAS disease. Nearly one third of the episodes of ARF are pre-
ceded by a GAS infection that does not receive medical atten-
tion.86 In many areas of the world, resources are not available
for effective primary and secondary prevention. Even in the face
of adequate therapy, the outcome of invasive disease may be
poor.87,88 Development of a safe, effective, and affordable GAS
vaccine has the potential to overcome many of these barriers.
Active immunization
Feasibility of developing a safe and efficacious
GAS vaccine
The feasibility of developing a GAS vaccine depends on ade-
quate resolution of several safety and immunogenicity con-
cerns. The principal safety concern is the theoretic risk that a
GAS vaccine, like natural infection, might trigger an immuno-
pathological event such as ARF. Thus an important advance for
the development of M protein–based vaccines was the discovery
in the 1980s that type-specific amino-terminal regions of the
M protein elicited the strongest bactericidal immune responses
and could be separated from the potentially harmful cross-
reactive epitopes.89–93
Evidence suggests that there is immunity to natural GAS
infection, so it is reasonable to expect that a vaccine could be
constructed to do the same. The declining incidence of GAS
pharyngitis with age implies the acquisition of immunity
1172 SECTION THREE
following natural exposure.94 Lancefield95,96 established the role
of type-specific immunity in preventing challenge infections
in animals and later showed that humans develop long-last-
ing type-specific immunity following natural infection. Beachey
et al97 protected mice against M24 infection using serum from
human volunteers immunized with a type-specific peptide frag-
ment of type 24 M protein. The most direct evidence comes
from clinical trials performed by Fox et al98–100 (described in the
following section) demonstrating that vaccinated volunteers
were significantly protected against challenge with virulent
GAS bearing the homologous M protein.
Knowledge of the precise immune responses that mediate
clinical protection could be exploited in vaccine development,
but these remain incompletely elucidated. The findings of
Lancefield and Fox bespeak M type–specific immunity, i.e., that
protection is mediated by antibodies recognizing the amino-
terminal region of M protein. While it is presumed that such
antibodies must possess bactericidal activity, a direct correla-
tion between bactericidal antibodies and clinical immunity
has not been observed consistently.98 On the other hand, some
investigators argue that immune responses evoked by repeated
exposure to GAS antigens that are shared among different M
types would be required to account for the observed age-related
decreases in GAS pharyngitis, given the large number of M
types in circulation.101
Even if a safe and effective vaccine was developed, the feasi-
bility of introduction and uptake will be driven by factors such
as cost-benefit, cost-effectiveness, public perception, and mar-
ket forces in a given country. The Institute of Medicine per-
formed a cost-benefit analysis of GAS vaccination in the United
States in which it was estimated that the annual health care
costs for GAS in the United States reach $493 million.102 If a
vaccine were provided to 95% of infants and conferred 75% effi-
cacy, the committee predicted a net cost savings of $314 mil-
lion. Although similar cost:benefit analyses are not yet available
for developing countries, the World Health Organization esti-
mated that RHD accounted for 5.9 million disability-adjusted
life years during 2002.103
Historical perspective: clinical GAS vaccine trials
The history of modern GAS vaccine development dates back
to the 1940s, when approximately 4,000 young adults were
injected with whole-cell vaccines (M3, M17, and M19) inac-
tivated by heat or ultraviolet light.104 These vaccines produced
reactions, some of which were severe, but did not prevent GAS
disease. In the 1960s, clinical trials involving children and
adults were initiated using parenteral vaccines consisting of
partially purified M protein or cell wall preparations contain-
ing high M protein content.98,100,105–110 The crude preparations of
whole cell or cell wall were unacceptably reactogenic and incon-
sistently immunogenic, presumably because the toxicity lim-
ited the total amount of protein that could be administered.
However, when injected with mineral oil adjuvant, partially
purified M protein induced bactericidal antibodies in 10 of 16
subjects.105 These findings suggested that an effective vaccine
based on M protein may be an effective strategy if a safe and
reliable delivery system could be identified.
The 1960s saw an unfortunate turn of events. Investigators
conducting a clinical trial involving a partially purified M3
protein reported an apparent increase in the attack rate of
ARF among vaccinated children compared to historical con-
trols.111,112 The vaccine regimen consisted of increasing doses
(up to 1 mg per dose) administered (for the most part) subcuta-
neously at approximately weekly intervals for 18 to 33 weeks.
Recipients were siblings of ARF patients. The ARF episodes
appeared 37 days to 23 months after the previous vaccina-
tion and occurred among subjects who were experiencing high
rates of intercurrent GAS pharyngitis. Whether these cases
were caused by vaccination remains uncertain. Nevertheless, a
federal regulation was issued in 1979 that endured until 2006
and prohibited licensure of GAS organisms or their derivatives
for human administration.113
Meanwhile, sentinel studies conducted by Fox et al
clearly demonstrated that a vaccine could induce protective
immunity.98–100 These investigators immunized seronegative
volunteers with three monthly injections of purified M1, M3,
or M12 protein98–100 given either subcutaneously with alum
or via aerosol to the nose and pharynx. One to two months
after immunization, a cohort of vaccinated and unvaccinated
subjects was challenged by painting virulent GAS bearing the
homologous M type onto their tonsils and pharynx. Volunteers
were evaluated clinically and microbiologically for 6 days and
then treated with penicillin. Both parenteral and mucosal vac-
cines prevented GAS pharyngitis, although the level of pro-
tection achieved statistical significance only for M1 vaccine.
Furthermore, mucosal vaccination significantly prevented
acquisition of throat infection.
Around this same time Beachey et al97 discovered techniques
for creating highly purified M protein fragments as vaccine anti-
gens. They demonstrated that M24 peptide adsorbed to alum
was well tolerated and elicited bactericidal antibodies in volun-
teers. None of the sera contained antibodies that cross-reacted
with myocardial tissue. This trial provided further proof that
the use of well-defined M protein epitopes could be safe and
elicit type-specific immunity.
Current prospects for a GAS vaccine
The advent of molecular biology enabled the design of a new gen-
eration of GAS vaccines with rational, well-defined components
and allowed clinical trials to resume after a hiatus of nearly
30 years (Table 58.1). Dale et al114 constructed a recombinant
fusion protein containing N-terminus fragments from six GAS
M types of clinical and epidemiologic importance (1, 3, 5, 6, 19,
and 24) and excluding tissue cross-reactive epitopes. A phase 1,
proof-of-principle trial was conducted to evaluate three dosage
levels (50, 100, and 200 g), with 10 subjects per group. 115 This
intramuscular vaccine appeared safe, well tolerated, and did not
induce antibodies that cross-react with human tissues. Three
spaced doses of 200 g (at 0, 4, and 16 weeks) elicited a fourfold
or greater rise in antibody by ELISA in 57 (95%) of the 60 pos-
sible assays (ie, 6 antigens times 10 subjects). Postvaccination
rises in functional (bactericidal or opsonophagocytic) antibody
were common.
In an effort to develop a broadly protective vaccine, Dale
et al116 constructed a second-generation vaccine containing
amino-terminal peptides of 26 M types. The vaccine includes
protective M protein epitopes from serotypes accounting for
79% to 90% of cases of pharyngitis and invasive disease from the
civilian and military populations in various locations including
the United States, Canada, Mexico, and Israel (Table 58.2)116–122
while serotype coverage in less-developed regions of the world
has been estimated to be 32% to 68%.123 Thirty subjects received
a 200 g dose of vaccine intramuscularly at 0, 4, and 16 weeks.
Local and systemic reactions were common but tended to be
mild. No findings suggestive of ARF were seen. As shown in
Table 58.2, 17 of the 26 M antigens (65%) elicited a fourfold or
greater antibody response by ELISA in at least 80% of subjects,
many of whom exhibited bactericidal activity. The proportion of
subjects responding to the M types responsible for most severe
infections in the United States in recent history (1, 3, 5, 6, and
18)20 varied (Table 58.2). In addition, a mucosal 26-valent for-
mulation is under investigation, stimulated by encouraging
results immunizing mice intranasally with the hexavalent vac-
cine along with mucosal adjuvant.124 A 30-valent vaccine has

GAS Candidate Vaccine Antigens that Evoke Immune
Responses in Humans After Natural Infection and Demonstrate
Efficacy in Animal Models
Fibronectin binding proteins
*This is the only candidate to have been tested in clinical trials.
Response of 30 Volunteers to Immunization with 200 g
Doses of 26-valent GAS Vaccine at 0, 4, and 16 Weeks
From McNeil et al., Clin Infect Dis 41:114-1122, 2005002E
since been constructed that targets 98% of pharyngitis cases
in the United States and Canada, 90% of invasive disease in
the United States, and 78% of invasive disease in Europe.125
Accordingly, coverage of strains from developing countries is
broadened. Moreover, evidence that the vaccine evokes cross-
Streptococcus group A vaccines 58 1173
opsonic bactericidal antibodies against a significant percentage
of nonvaccine M serotypes tested to date may further enhance
the vaccine's spectrum of efficacy.
The search for GAS antigens that are conserved among a large
number of GAS serotypes identified an epitope in the C-repeat
region of M protein to which persons from highly endemic com-
munities seroreact.126 These antibodies have opsonic potential
in naturally infected persons.127 In animal models, mucosal vac-
cines comprising this region induce serum and mucosal anti-
bodies that do not cross-react with human heart tissue and
confer protection against challenge with both homologous and
heterologous GAS M types.128 The oral commensal bacteria
was evaluated in animals as mucosal
vector for surface expression of the conserved fragment of M
protein.129 An vector strain bearing no GAS anti-
gens was shown to be well tolerated and able to achieve short-
term colonization when delivered orally and nasally to healthy
adults.130 These studies were conducted to guide the design of
future phase I trials to evaluate an live vector vaccine
expressing the conserved region of GAS M protein. However,
the GAS antigen-bearing strain has not been constructed.
Good et al identified the peptide J14 and its derivative J8 as
vaccine candidates that contain minimal conserved murine pro-
tective B-cell epitopes in the C-repeat region within a non-M
protein helix-forming sequence (to maintain conformational
integrity), but lack heart cross-reactive T-cell epitopes. Several
formulations that vary by antigen content, adjuvant, dosing
schedules, routes of administration, and vehicles have been eval-
uated for protective efficacy against lethal infection in mice.131–
134
Parenteral immunization with J8 conjugated to diphtheria
toxoid (DT) in concert with adjuvant (alum or SBAS2) protected
mice against systemic (intraperitoneal) challenge.135 Intranasal
administration of J8-DT along with cholera toxin B subunit pro-
tected mice against mucosal challenge.136 Other investigators
demonstrated the protective efficacy of J14 conjugated to key-
hole limpet hemocyanin administered parenterally in a mouse
model of soft tissue challenge infections but not following chal-
lenge via the intraperitoneal or intranasal routes.137 The J8-DT
vaccine will soon be tested in phase 1 clinical trials in humans.138
Several virulence antigens not borne by the M protein that are
surface expressed or secreted among M types have undergone
preclinical testing but have not yet reached clinical trials. Those
discussed here are immunodominant (ie, composed of anti-
gens that elicit immune responses following natural infection),
shared by multiple M types, evoke both homologous and heter-
ologous protective immunity in animal models when delivered
in a formulation that could be used in humans, and have no
known epitopes that cross-react with human tissues. It is likely
that a broadly protective GAS vaccine will require a combina-
tion of antigens.
C5a peptidase has long been considered an attractive vaccine
antigen.139,140 In mice, intranasal immunization with recom-
binant enzymatically inactive protein prevents colonization
following intranasal challenge in a serotype-independent fash-
ion141 and enhances clearance of GAS from the nasopharynx
of infected mice, while an adjuvanted subcutaneous formula-
tion promotes clearance from both nasopharynx and lung.142
The carboxylic esterase sse has two antigenic variants, together
encompassing at least 10 M serotypes. The M types comprising
each sse variant share more than 98% amino acid homology.
Accordingly, vaccination with recombinant sse protects mice
against lethal subcutaneous infection with virulent M1 and M3
strains and inhibits GAS invasion of mouse skin tissue.143
1174 SECTION THREE Vaccines in development and new vaccine strategies
Several fibronectin binding proteins also show promise as
vaccine antigens. The most investigated of these is SfbI,144,145
an antigen that possesses intrinsic adjuvanticity.146 Intranasal
immunization with either purified recombinant SfbI or
DT-conjugated polypeptides encompassing T- and B-cell epit-
opes of the SfbI protein protected mice against lethal challenge
with homologous and heterologous GAS strains.147–150 FBP54
also induces heterologous protection when delivered by subcu-
taneous or mucosal inoculation.151 The putative cervical epithe-
lial cell adhesion R28 was protective in passive immunization
murine experiments39 and thus has potential as an immunogen
for prevention of puerperal fever.
Vaccine-mediated prevention of STSS and severe invasive
disease has focused on SPEs as antigens, in particular SPE A.
Vaccination with SPE A protects rabbits not only from STSS,
but also from necrotizing fasciitis-myositis manifestations.152
SPE A and C toxoids and recombinant antigens lacking regions
thought to mediate T-cell receptor or major histocompatibility
complex class II interaction lose their superantigenic properties
yet confer strong protection against lethal challenge.153–155
The GAS cell wall polysaccharide is surface exposed, con-
served, and immunoreactive, and abundantly expressed in most
GAS M serotypes, forming the basis of the group A antigen used
in rapid tests to diagnose GAS pharyngitis. Synthetic oligosac-
charide vaccines conjugated to the mutant DT CRM197 have
been constructed. When compared to purified native polysac-
charide glycoconjugates, several oligosaccharide formulations
administered parenterally with alum elicited comparable titers
of opsonophagocytic antibody and similar protection against
lethal injection of wild-type GAS in mice.156 Nonetheless,
potential safety issues related to possible cross-reactivity of the
vaccine antigens with valvular endothelial and neuronal tissues
will have to be addressed.75,78
More recently, the search for novel vaccine candidates has
utilized proteomic techniques, reverse vaccinology, and genome-
wide selection of surface-expressed antigens, taking advantage
of several genome sequences.157–159 The discovery of the GAS
pilus is one example. A combination of recombinant pilus pro-
teins administered intraperitoneally with Freund's adjuvant
conferred protection against lethal mucosal GAS challenge in
a mouse model.40 Since there are only 21 known Lancefield T
serotypes, these findings raise the possibility for broader strain
coverage with fewer vaccine antigens compared with M protein–
based vaccines. It has been estimated that 12 types would be
needed for a vaccine to protect against more than 90% of GAS
strains from the United States and Europe.160 The 15 pilus back-
bone protein variants identified to date account for 18 of the 21
Lancefield T types. T typing is needed from understudied areas
of the world to exclude heretofore unappreciated genetic diver-
sity.29 Another interesting protein discovered using genomic
techniques is the antiphagocytic serine protease SpyCEP.161,162
SpyCEP is conserved among strains, evokes antibody after
natural infection, and induces protective immunity in animal
models. Nine promising candidates for a GAS vaccine were
identified using antigenome technology that involves the use of
genome libraries to select, using a stepwise approach, potential
vaccine antigens that are surface expressed, conserved, reactive
with convalescent human serum, required for virulence of GAS,
and induce heterologous protection in animal challenge experi-
ments.159 Bioinformatic analysis was performed to ensure lack
of homology with human proteins. Sequence conservation was
assessed with published GAS strain sequences and an interna-
tional collection of clinical isolates. Spy 1536, which appears to
mediate binding to a broad array of human extracellular matrix
proteins, emerged as a promising candidate. Interestingly,
SpyCEP was one of the nine candidates selected by this ardu-
ous screening.
Indirect effects
Viral infections, in particular varicella and influenza, may pre-
dispose to invasive GAS infection. The ability of vaccines that
prevent viral infections to diminish the occurrence of severe
GAS disease should not be underestimated.163–165
Future challenges
Although promising, M type–specific vaccines must address the
antigenic variability posed by the approximately 80 M serotypes
and 170 distinct types with over 750 subtypes that have
been identified. Less-developed and transitional countries show
a much higher diversity and a different hierarchy of circulating
types compared to that seen in the Americas (excepting
Hawaii166), Europe, and Japan, resulting in low coverage rates
for the proposed 26-valent vaccine. Moreover, -type dis-
tributions can change over time,167 rendering a widely applica-
ble M type–specific vaccine even more difficult to achieve.163–171
Careful analyses will be required during later stages of clinical
development to determine the level of efficacy and serotype cov-
erage that would make a type-specific GAS vaccine desirable for
public health use.
Challenges face each stage of the clinical development of a
GAS vaccine, many due to the need to exclude the theoretic
risk of a vaccine triggering ARF.172,173 Preclinical assessments
are handicapped by the lack of a reliable animal model for ARF,
but should at least provide assurance that the candidate does
not contain B- and T-cell epitopes that cross-react with host
tissues and does not possess superantigenic properties. Recent
clinical trials provide a template for intensive safety evalua-
tions, but more expedient tools would be useful for larger phase
2 and phase 3 trials. Whether phase 3 trials must be designed
to exclude a potential rare occurrence of ARF remains to be
determined.
The choice of primary clinical end points for phase 3 trials
has engendered debate; considerations such as all pharyngitis,
medically attended pharyngitis, impetigo (particularly in aborig-
inal populations), invasive infections, primary ARF, and recur-
rent ARF each have relative merits. Other considerations for
phase 3 study design include the potential confounding effects
of herd immunity and high asymptomatic colonization rates.
Implementation of vaccine programs will face other complex
174,175
issues such as emergence of newly characterized types
and the possibility that nonvaccine serotypes may replace those
contained in the vaccine via “immune escape”.176,177 Finally, it
will be critical to determine a level of vaccine efficacy that would
sufficiently assure practitioners that they can safely abandon
the practice of culturing throats, empirically treating pharyngi-
tis with antibiotics, and/or recommending secondary antibiotic
prophylaxis in individuals with ARF.

Access the complete reference list online at http://www.expertconsult.com
40. Mora M, Bensi G, Capo S, et al. Group A produce pilus-like
structures containing protective antigens and Lancefield T antigens. Proc
Natl Acad Sci U S A 2005;102:15641–6.
100. D'Alessandri R, Plotkin G, Kluge RM, et al. Protective studies with group A
streptococcal M protein vaccine. III. challenge of volunteers after systemic
or intranasal immunization with Type 3 or Type 12 group A . J
Infect Dis 1978;138:712–8.
111. Massell BF, Honikman LH, Amezcua J. Rheumatic fever
following streptococcal vaccination. Report of three cases. JAMA
1969;207:1115–9.
115. Kotloff KL, Corretti M, Palmer K, et al. Safety and immunogenicity of a
recombinant multivalent group a streptococcal vaccine in healthy adults:
phase 1 trial. JAMA 2004;292:709–15.
119. McNeil SA, Halperin SA, Langley JM, et al. Safety and immunogenicity of
26-valent group a vaccine in healthy adult volunteers. Clin
Infect Dis 2005;41:1114–22.
Streptococcus group A vaccines 58 1175
128. Bessen D, Fischetti VA. Influence of intranasal immunization with synthetic
peptides corresponding to conserved epitopes of M protein on mucosal
colonization by group A streptococci. Infect Immun 1988;56:2666–72.
135. Olive C, Clair T, Yarwood P, et al. Protection of mice from group A
streptococcal infection by intranasal immunisation with a peptide vaccine
that contains a conserved M protein B cell epitope and lacks a T cell
autoepitope. Vaccine 2002;20:2816–25.
142. Park HS, Cleary PP. Active and passive intranasal immunizations with
streptococcal surface protein C5a peptidase prevent infection of murine
nasal mucosa-associated lymphoid tissue, a functional homologue of human
tonsils. Infect Immun 2005;73:7878–86.
156. Kabanova A, Margarit I, Berti F, et al. Evaluation of a Group A
synthetic oligosaccharide as vaccine candidate. Vaccine 2010;29:104–14.
159. Fritzer A, Senn BM, Minh DB, et al. Novel conserved group A streptococcal
proteins identified by the antigenome technology as vaccine candidates for a
non-M protein-based vaccine. Infect Immun 2010;78:4051–67.
 SECTION THREE: Vaccines in development and new vaccine strategies
Streptococcus group B vaccines
59 C. Mary Healy
Carol J. Baker
In 1938, Fry1 reported three cases of fatal puerperal sepsis
caused by group B -hemolytic streptococci (GBS). Sporadic
reports appeared until the 1970s, when GBS sepsis and men-
ingitis in young infants increased dramatically,2–4 followed by
recognition of invasive GBS infections in pregnant women and
in nonpregnant adults; the latter infections occasionally were
severe and fatal.5,6 Explanations for the emergence of GBS as a
prominent human pathogen remain unknown. Despite imple-
mentation of intrapartum antibiotic prophylaxis (IAP) to pre-
vent peripartum maternal morbidity and early-onset neonatal
disease,7 GBS infections continue to occur, albeit at a reduced
incidence.8,9 Among infants 7 to 89 days old, incidence, case-
fatality rates, and neurodevelopmental sequelae have remained
stable for decades.9–11 Finally, invasive GBS disease is increas-
ingly impacting adults with underlying medical conditions and
elderly people.12–14
Background
Clinical description and complications
Invasive GBS disease in infants is characterized by age at
onset. Early-onset GBS (EOGBS) occurs before age 7 days
( 90% have signs of infection within 12 hours of birth), fre-
quently manifesting as sepsis without a focus (80%-95%),
pneumonia (10%-15%), or meningitis (5%-10%). Clinical find-
ings are nonspecific. Severity of illness ranges from multiorgan
failure in some, to healthy-appearing infants who are evalu-
ated and treated empirically because of maternal GBS coloni-
zation and inadequate duration of IAP before delivery. Focal
findings except pneumonia are rare. In contrast, late-onset
GBS (LOGBS) occurs in infants aged 7 to 89 days old, more
often manifests with fever, poor feeding, and lethargy and less
commonly with overwhelming sepsis. Meningitis is more fre-
quent with LOGBS (21% 35%), and approximately 30% of
cases have permanent sequelae (eg, hearing loss, developmen-
tal delay).10,11 Other manifestations include septic arthritis,
osteomyelitis, cellulitis and adenitis, and other less common
foci of infection.15–17 Late-late onset GBS occurs in infants older
than 89 days, (most often in preterm infants whose corrected
gestational age is younger than 3 months) and accounts for
approximately 20% of cases beyond the first 6 days of life.17,18
Recurrent GBS is reported in an estimated 0.5% to 3% of
GBS cases.17,19 Invasive GBS is uncommon after age 6 months
unless an underlying immunodeficiency is present.20
Lower vaginal and rectal colonization with GBS in pregnant
women is a prelude to maternal invasive disease and EOGBS
in the infant. The risk of invasion of amniotic fluid or the
infant bloodstream is related directly to the density of mater-
nal colonization and virulence of the organism.21–23 Maternal
GBS infection manifests during pregnancy as asymptomatic
bacteriuria, urinary tract infection, or pyelonephritis and dur-
ing the peripartum period as bacteremia, chorioamnionitis or
postpartum endometritis; these infections can result in spon-
taneous abortion, preterm delivery, and stillbirth. Two thirds of
GBS disease cases in adults are not associated with pregnancy;
rates in nonpregnant adults have increased twofold to fourfold
during the last 25 years.14 Skin and soft tissue infections are the
most common manifestations of GBS infection in this popula-
tion, but urinary tract infection, pneumonia, bacteremia, bone
and joint infection, endocarditis, meningitis, peritonitis, and
catheter-associated infections are described. The incidence of
GBS is directly related to increasing age and coexistent morbid
conditions.12–14,23,24
Bacteriology and diagnosis
GBS are aerobic gram-positive diplococci that are divided into
10 serotypes (Ia, Ib, and II though IX) based on their capsu-
lar polysaccharides (CPS). Many strains have surface proteins
(eg, - and -C, R1-4, rib, alpl-3, Sip, pilus islands 1, 2a, and
2b) that allow further characterization for epidemiologic stud-
ies. They form gray-white, mucoid, 3 to 4 mm diameter colo-
nies on sheep blood agar medium and produce a narrow zone of
hemolysis. Infection is diagnosed by isolation of GBS from a
usually sterile site.17
Pathogenesis
Virulence factors characterized as important for attachment
and invasion in human infection are CPS, -hemolysin, C pro-
teins, and pilus-like protein structures. Experimental models of
infection demonstrated that antibodies to CPS, -hemolysin,
- and -C proteins, and pili are protective.25–28 C proteins and
pili allow epithelial cell adherence, and -hemolysin and other
factors lyse epithelial and endothelial cells, leading to direct
tissue damage to and spread through host tissues. Immune
clearance of GBS is resisted by CPS and other virulence factor–
mediated effects at the cellular and molecular levels, leading
to inhibition of neutrophil recruitment, blocking of opsono-
phagocytosis and impairment of phagocytosis, and oxidative
burst killing. Finally, cell wall lipoteichoic acid, cell wall pepti-
doglycan, and -hemolysin/cytolysin trigger cytokine activation
and release.17 High-virulence clonal complexes such as ST-17
of type III GBS are also described; the latter has a tropism for
meninges and is typically found only among invasive cells.29,30

Treatment
Penicillin is the drug of choice for treatment of systemic infec-
tion.31,32 GBS also is susceptible to other -lactam antibiot-
ics and to vancomycin. GBS isolates with point mutations in
penicillin-binding proteins and reduced -lactam susceptibil-
ity have been reported from the United States and Japan; the
clinical significance of these is unclear.33–36 Resistance to eryth-
romycin and clindamycin is reported in approximately 32%
and 20% of isolates, respectively. While typically resistant to
aminoglycosides, synergistic killing of GBS occurs in vitro and
in vivo when gentamicin is combined with penicillin or ampi-
cillin.17 Treatment duration depends on the severity of illness
and sites involved, but a minimum of 10 and 14 days is recom-
mended for bacteremia and meningitis, respectively, and longer
for more complicated infections.12,32 Despite advances in inten-
sive care and prompt initiation of antimicrobials, GBS infant
disease overall mortality (5%) and morbidity exceeds that of
type b and
before the use of routine infant immunizations to prevent
these infections in infants.37,38
Prevention
Preventive strategies target neonatal EOGBS disease. IAP was
adopted in the United States in 1996 following reports that
intravenous ampicillin or penicillin G given intrapartum to
GBS carriers prevented EOGBS in their infants. Two proposed
methods, based on culture screening of pregnant women or
the presence of one or more risk factors predisposing to neo-
natal GBS disease, resulted in IAP being offered to approxi-
mately 25% of women at delivery.39–41 In 2002, the Centers for
Disease Control and Prevention recommended universal cul-
ture-based IAP based on its superior efficacy (relative risk of
EOGBS, 0.46 [95% confidence interval, 0.36-0.6]) compared
with the risk-factor strategy.42 These recommendations were
updated in 2010 (Table 59-1) to incorporate new diagnostic
methods to detect GBS colonization and revised algorithms
for screening and prophylaxis of women and management of
newborns.43 IAP remains an interim strategy while vaccine
development proceeds.7
Recommendations for IAP* to Prevent Perinatal GBS Disease†

t

1SFWJPVT
JOGBOU
XJUI
JOWBTJWF
(#4
EJTFBTF



t

(#4
CBDUFSJVSJB
JO
UIF
DVSSFOU
QSFHOBODZ



t

1PTJUJWF
(#4
TDSFFOJOH
DVMUVSF
JO
DVSSFOU
QSFHOBODZ
VOMFTT
QMBOOFE

DFTBSFBO
EFMJWFSZ
QFSGPSNFE
XJUIPVU
MBCPS
PS
SVQUVSF
PG
NFNCSBOFT


t

(#4
DVMUVSF
TUBUVT
OPU
LOPXO
BOE
BOZ
PG
UIF
GPMMPXJOH

o
%
FMJWFSZ
BU



XL
HFTUBUJPO



o

3VQUVSF
PG
NFNCSBOFT



I



o

*OUSBQBSUVN
GFWFS


×$
×'


o

*OUSBQBSUVN
/""5


QPTJUJWF
GPS
(#4


IAP, intrapartum antibiotic prophylaxis; GBS, group B .
*Vaginal and rectal GBS cultures are recommended at 35-37 wk gestation for all pregnant women (unless there is a history of GBS bacteriuria in the current
pregnancy or a previous infant with GBS disease). For IAP, intravenous penicillin G preferred, ampicillin acceptable, at onset or rupture of membranes and every 4 h
until the infant is delivered. For penicillin-allergic women at low risk for anaphylaxis, intravenous cefazolin every 8 h is preferred.
†
Adapted from Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: revised guidelines from CDC, 2010. MMWR
Recomm Rep 59(RR-10):1-36, 2010.
‡
If intrapartum nucleic acid amplification test (NAAT) is negative but any other intrapartum risk factor is present, IAP is indicated.
Streptococcus group B vaccines 59 1177
Epidemiology
Incidence and prevalence
GBS colonize mucous membranes, leading to intermittent
or chronic carriage. The most common sites of colonization
in adults are the rectum, lower vagina (women), and urethra
(men). Isolation rates increase 10% to 15% in pregnant women
if rectal cultures are assessed in addition to lower vaginal
specimens. Differences between ascertainment techniques make
direct comparisons of reported prevalence impossible, but GBS
carriage rates are approximately 10% in prepubertal children,
30% in women of childbearing age regardless of pregnancy sta-
tus, 25% in men, and 22% in healthy elderly adults.17,44
The incidence of EOGBS in the United States decreased dra-
matically following the introduction of IAP, from 1.7 per 1,000 live
births in 1993 to 0.26 to 0.37 per 1,000 live births between 2005
and 2009.45 The incidence of LOGBS has remained unchanged at
0.25 to 0.6 per 1,000 live births. Late-late onset disease numbers
are increasing in association with increased survival of preterm
infants.9,45 IAP has reduced invasive infection in pregnant women
to 0.12 per 1,000 deliveries.14 Invasive GBS infections are increas-
ing in nonpregnant adults (4.1-7.2 per 100,000 population), espe-
cially in elderly people (25.4/100,000).12,13 GBS infections in all
age groups are overrepresented in black populations.10,14,45 In
Europe, where national surveillance and IAP are not standard of
care, reported neonatal EOGBS rates are 0.3 to 0.4 per 1,000 live
births.46–52 There was a fivefold reduction in Australasian rates for
EOGBS as a result of IAP during the 1990s.53
In the United States and Europe, invasive disease in preg-
nant women and infants is caused predominately by five CPS
types, III, Ia, V, Ib, and II, in decreasing order of prevalence
(Table 59-2).14,47,50,51,54–56 In adults and elderly people, serotype
data from the United States and Canada demonstrate that the
same five CPS types cause more than 95% of disease, although
type V replaces III as most common.12,14,23,56
Before 1990, serotype V was rarely found in colonizing or
invasive strains of GBS. Serotypes VI and VIII predominate in
Japan among colonized persons and have been reported sporadi-
cally in Europe.57,58

t

1SFWJPVT
QSFHOBODZ
XJUI
(#4
DPMPOJ[BUJPO



t

1SFWJPVT
QSFHOBODZ
XJUI
(#4
CBDUFSJVSJB



t

1MBOOFE
DFTBSFBO
EFMJWFSZ
JO
UIF
BCTFODF
PG
MBCPS
PS

SVQUVSF
PG
NFNCSBOFT
SFHBSEMFTT
PG
(#4
DPMPOJ[BUJPO



t

/FHBUJWF
WBHJOBM
BOE
SFDUBM
(#4
TDSFFOJOH
DVMUVSFT
BU


XL
HFTUBUJPO


1178 SECTION THREE Vaccines in development and new vaccine strategies
Serotype Distribution of GBS Isolates Causing Invasive Infection and GBS CPS Conjugate Vaccine Responses in Healthy Adults
*B  *B55
*C *C55
**55
***55
 755
CI, confidence interval; CPS, capsular polysaccharide; GBS, group B ; GMC, geometric mean concentration; TT, tetanus toxoid.
*Data from references 54–56 and 14 (the remaining isolates are nontypable by serologic methods).
†
Data from references 12,13,23,54,56, and 14 (the remaining isolates are nontypable by serologic methods)
‡
Data from references 54,76,77,79–81.
Groups with additional risk of GBS
In the United States, the incidence of EOGBS and LOGBS in
African American infants is two to three times higher than in
white infants.10,14,43 Infants of Hispanic ethnicity also have higher
incidence rates.59 This may be due to reduced access to prenatal
care, higher rates of preterm birth, or higher colonization rates
attributable to vaginal hygiene practices. Among adults, blacks
have twice the rate of invasive GBS infection as whites, and dis-
ease rates rise in elderly people.13,14,60 Nursing home residents
have a greater risk than do community-dwelling persons (72.3
vs 17.5/100,000).61 After controlling for age, diabetes mellitus,
cirrhosis, stroke, breast cancer, decubitus ulcer, and neurogenic
bladder, all increase risk for invasive GBS infection.13,14,62
Disease burden
Case-fatality rates of 3% to 20% are reported for term and pre-
term infants with EOGBS, respectively. Case-fatality rates for
LOGBS is lower (1%-6%), but late-onset GBS meningitis is
associated with permanent neurologic sequelae in 30% to 50%
of survivors.10,11,14 In adults, invasive GBS infection is associated
with a 15% case-fatality rate, but this rate approaches 25% or
higher in elderly people and people with underlying medical con-
ditions.13 In most countries, including the United States, CPS
types Ia, Ib, II, III, and V account for approximately 85% to 95%
of invasive disease isolates.54
Active immunization
History of vaccine development
Lancefield's observations demonstrated protection against
lethal GBS infection in a mouse model by antibodies to the
CPS of GBS. In the 1970s, Baker and Kasper63 reported that
mothers of GBS-infected infants had very low serum III
CPS-specific antibodies at delivery, a finding confirmed and
extended for serotypes Ia and Ib.64,65 Human serum contain-
ing sufficient concentrations of Ia, Ib, II, III, and V CPS-
specific IgG promotes efficient opsonization and phagocytosis
of homologous strain in vitro and provides protection from
lethal experimental infection in vivo. Once the CPS of Ia, II,
and III GBS were purified and immunochemically defined,
an effective polysaccharide vaccine seemed imminent. It was
hoped that active immunization of pregnant women would
elicit protective levels of antibodies that would prevent mater-
nal disease and, through placental transport, protect their
 

  

 



 

infants from EOGBS and LOGBS. However, studies of uncou-
pled GBS polysaccharide vaccines in healthy adults were dis-
appointing. The first purified CPS (type III) was tested in
1978 and was followed by Ia and II. While well tolerated, CPS
Ia, II, and III elicited variable immune responses, 40%, 88%,
and 60%, respectively.66 Nearly 90% of subjects had low pre-
immunization CPS-specific antibodies, reflecting immuno-
logical naïveté, and this predicted no or less robust immune
responses to CPS vaccines. In the few subjects with higher
initial CPS-specific IgG, brisk, reliable immune responses
occurred.
The single maternal immunization study performed using
III CPS vaccine demonstrated that it was well tolerated and
had comparable immunogenicity to that in nonpregnant adults
(63% immune response).67 Among infants of maternal vac-
cine responders, 90% had presumed protective levels of CPS-
specific IgG in their serum samples at 2 months of age that
promoted opsonophagocytosis and killing of type III GBS
in vitro. Furthermore, these functionally active specific antibod-
ies persisted in the majority of infants through age 3 months.
While disappointing, these early studies demonstrated that
vaccine prevention of GBS disease was feasible and that improved
immunogens were needed.
The success of polysaccharide-protein conjugate vaccine
technology in preventing type b and
infections in infants heralded more hope to control
GBS.68,69 In addition to the expectation that polysaccharide-
protein glycoconjugates would generate T cell–dependent
responses associated with immunological memory and long-
term protection, antibodies induced by protein antigens are
predominantly of the IgG1 subclass and are actively and pas-
sively transported by the placenta, resulting in higher infant
antibody concentrations than those induced by polysaccharides
(predominantly IgG2 subclass).70–73 Theoretically, a GBS CPS-
protein glycoconjugate would induce higher maternal CPS-
specific IgG concentrations, favor an IgG1 subclass response,
thereby improving placental transport, and increase likelihood
of protection of mother and infant.
Constituents of candidate conjugate vaccines
Most candidate GBS conjugate vaccines for clinical trials
were developed at the Channing Laboratory, Harvard Medical
School (Boston, Massachusetts). The first GBS glycoconjugate
vaccine was the III CPS covalently linked to monomeric teta-
nus toxoid (TT) using reductive amination coupling chemis-
try.74,75 Monovalent Ia, Ib, II, and V CPS and one bivalent II/
III CPS were produced using the same technology with TT
as the carrier protein for all but one V CPS conjugate lot in
Technologies for making new vaccines 60 1183
1184 SECTION THREE Vaccines in development and new vaccine strategies

Status of Development of Representative Human Vaccines Made by Different Technologies—cont'd


( = 2,036), again in Guatemala and Mexico, the trial's pri-
mary target endpoint of greater than 60% efficacy against mod-
erate to severe ETEC diarrhea was not met, finding only about
35% protection. Nor was there an effect on the frequency of
all causes of diarrhea. However, there was a 60% reduction in
the incidence of LT-positive diarrhea of all degrees of severity,
along with a significant reduction in duration and severity of all
diarrhea causes. The patch also induced measurable immune
responses and was well tolerated.270,272
In a smaller phase 2 trial in India ( = 723),271 the LT patch
also did not reach its targeted endpoint, perhaps because of a
low attack rate (about 1%) for LT-positive ETEC. As a result of
these two trials, Intercell discontinued work on the LT patch
for traveler's diarrhea but still pursues its use with the Skin
Preparation System device for other applications.
Applying the Intercell LT patch near the site of injection
of parenteral influenza vaccine (an application referred to as a
vaccine-enhancement patch) was found to improve hemagglu-
tination inhibition (HI) titers in the serum and mucosa of both
young and aged mice,273,274 and to increase the HI titer or show
an improving trend for adult volunteers older than 60 years.275
In May 2011, a partnership between Intercell and GSK276 began
enrolling 300 volunteers for a study to compare the patch with
AS03 adjuvant in boosting responses to pandemic H5N1 influ-
enza vaccine.277
In preclinical studies of other applications, use of LT or a
structurally similar cholera toxin as cutaneous adjuvants
resulted in improved immune responses or challenge protec-
tion in animal models for tetanus,278 anthrax,279,280 malaria,281
282
and Shiga toxin–producing strains of
enterohemorrhagic . 283
In regard to safety, early clinical trials found no serious reac-
tions,267 but pruritus and maculopapular rash at the patch site
were found in 13%,275 74%,266 or 100%268 of patients exposed
to LT-containing patches for 6 hours, and in one study 17% of
rashes progressed to vesicle formation.268 Delayed-type hyper-
sensitivity contact dermatitis was observed when using recombi-
nant colonization factor.266 Later clinical trials found the LT patch
to be “well tolerated and consistent with previous studies”.270, 272
Microrasps
Other methods take advantage of low-cost fabrication tech-
niques adapted from the microelectronics industry to convert
silicon, metal, or other material into arrays of micrometer- to
millimeter-size microrasps designed to abrade the stratum cor-
neum (as distinct from creating holes in it; see “Poking and
piercing”, later).31,35,39–41,99,284 One example is the
(MEA, also known as Onvax), an investigational technol-
ogy that scrapes the skin before or after topical application of the
antigen or therapeutic agent.84,285 The MEA consists of a square
or round chip containing about 1-cm2 area of silicon or plas-
tic microprojections mounted on a finger-held applicator.22,101,247
Preclinical studies of the MEA device using mice inoculated
with hepatitis B surface antigen (HBsAg) or DNA plasmids encod-
ing firefly luciferase found similar or greater immune responses or
light emission, respectively, compared with control IM and exper-
imental ID injections. Anthrax rPA with alum or CpG adjuvants
applied with the MEA device to mouse skin produced equivalent
or better immune responses than IM controls (although not as
good as an ID microneedle), whereas immune responses and chal-
lenge survival were significantly less among MEA-immunized rab-
bits compared with IM controls.101 Among cynomolgus monkeys
vaccinated by six “swipes” of the MEA, with SC and 34-gauge,
microneedle-based ID controls, all animals seroconverted to an
investigational recombinant Japanese encephalitis vaccine.105
Those vaccinated by swiping the MEA through a drop of vaccine
already on the skin showed neutralizing antibody responses in
the same range as the SC controls, whereas applying vaccine after
the abrasion appeared to be less effective.
Alternative vaccine delivery methods 61 1207
A clinical trial of the MEA measured transepidermal water
loss (TEWL) as a surrogate indicator for removal of the stra-
tum corneum after each of five consecutive swipes across the
same site on the volar forearm of volunteers. Projection heights
of 100, 150, and 200 m showed steadily increasing rates of
TEWL, with the tallest projections producing the greatest water
loss. Control swipes with fibrous and sandpaper ECG pads
showed little or no TEWL.285 A human trial, however, in which
rabies vaccine was applied before or after four “rubs” of the
device over four separate deltoid skin sites did not detect any
immune response after three dosings on days 0, 7, and 21.99
Shaving and brushing
The razor and the brush can also remove layers of the stratum
corneum. In a clinical trial of adenovirus vectors encoded to
express influenza hemagglutinin antigen, the abdominal skin
of 24 adults was shaved with a disposable, twin-blade razor,
followed by “gentle brushing with a soft-bristle toothbrush for
30 strokes” and application of the antigen with an occlusive
Tegaderm patch.286 Two doses 28 days apart at the highest dos-
age level produced fourfold rises in HI titer with 67% of the
cutaneous vaccines (there was no control group receiving con-
ventional parenteral delivery of either the recombinant vaccine
vector or a licensed inactivated influenza vaccine). Occasional
mild erythema at the abdominal site was reported in 61% and
rash or itching in 39% of patients.
This same research team,287 studying mice, substituted an
electric trimmer for shaving but otherwise used similar brush-
ing to demonstrate that topical application of nonreplicating
vectors overproducing antigens for and
were immunogenic.288,289 Control animals demon-
strated that depilation alone had little effect; what made the
difference was the mild brushing, which produced minimal irri-
tation (Draize score, 1).290 Others studying Japanese encephali-
tis vaccine in an animal model supplemented skin shaving with
a commercial depilatory cream, followed by occlusion of the site
with an impermeable covering.291 The practicality of such steps
in routine immunization of humans is uncertain.
As with cutaneous vaccination in general, a diverse terminology
is applied to microscopic projections for perforating the superfi-
cial skin to deliver the drug.31,41,43,247,248,292,293 In addition to the
most common term terms such as
, and
have been used. This chapter uses for the
broad category of all such projections shorter than 1,000 m,
reserving for those of 1 mm or longer, whether solid
or hollow (see “Mini-needles” and “Microrasps”, earlier). The fol-
lowing sections divide microneedles into functional subcategories.
Uncoated microneedles
Earlier, we described methods in which vaccine or drug is
applied to the site after it is prepared. The 3M Corporation294
developed an uncoated microneedle device to prepare the skin
by perforating it. Although not licensed (or even intended) for
vaccine or drug, its of micronee-
dles appeared on the US market in 2011 as a “pretreatment
method for professional medical or cosmetic dermatologists to
create microchannels in the skin” (see Figure 61-3B).295,296 Each
application creates 351 holes through the stratum corneum
into the epidermis.297,298 Other investigational technologies for
uncoated microneedles are the MicroCor,299,300 the Functional
MicroArray patch,301 and the Micro-Trans.302
Coated solid microneedles
A common strategy pursued by a number of commercial and
academic teams to carry antigen across the stratum corneum
is to coat it onto solid microscopic projections, which are held
1208 SECTION THREE Vaccines in development and new vaccine strategies
for variable periods of time in the epidermal layer while antigen
or other drug elutes and diffuses.27,30–32,35,39–43,246–248 To date, only
limited published data have demonstrated suitability for human
vaccination, in contrast to therapeutic drugs.
One example of drug-coated microneedles that appears
closest to marketing approval is the investigational Zosano
Pharma platform (formerly known as Macroflux) (see
Figure 61-3C).303 Its titanium projections vary from 225 to 600 m
in height and are packed into an area of 1 to 2 cm2 at densities
from 140 to 650 tines per square centimeter. They are inserted
by a spring-mounted applicator and held in place by an adhesive
patch. The most advanced applications for these microneedles are
delivery of parathyroid hormone to treat osteoporosis,304 already
studied clinically, and erythropoietin to treat anemia.
Regarding vaccine applications,305 a graph from a human
study of Zosano Pharma's ZP-Flu influenza vaccine patch,
applied for 5 or 10 minutes onto the skin, trended toward
increased titers and seroprotection compared with an IM con-
trol injection306,307 (no further details were provided, nor could a
public clinical trials registration be found).
A hairless guinea pig model was used to study ovalbumin
on the patch's microneedles as a representative, large antigenic
protein.305,308 It was administered in two doses 4 weeks apart.
It induced post-booster titers comparable to those of control
IM, SC, and ID Mantoux-style injections at higher dosages,
and it surpassed IM and SC routes at lower dosages. Other pre-
clinical studies of the system demonstrated delivery of oligo-
nucleotides309 and the peptide hormone desmopressin.310 The
company reports animal work with tetanus, diphtheria, Lyme
disease, and hepatitis B (DNA) vaccine antigens.
Another coated-microneedle platform is the
(sMTS),26,311–315 from 3M.294 Its
drug-coated pyramidal projections vary from 250 to 750 m
height, in arrays of 300 to 1,500 microneedles mounted on an
adhesive patch at a density of 1,300 per square centimeter. 315–318
Application to the skin is by a manual finger-thumb
device294,297,319–321 (see Figure 61-3E) or by a spring-powered appli-
cator, shown elsewhere.22 Coatings of the microneedles are said
to hold up to 0.5 mg of active pharmaceutical ingredient.
In a rabbit model, coatings of tetanus toxoid and alum adju-
vant in various ratios induced antibody levels an order of mag-
nitude higher than the presumed protective threshold ( 0.2 IU),
using just a fraction of the standard IM dosage.322 Ovalbumin as
a surrogate vaccine applied to hairless guinea pigs by sMTS using
the applicator was reported to induce antibody, as
measured by enzyme-linked immunosorbent assay, equivalent
to that induced by IM-needle injection.319 A second study using
hairless guinea pigs compared three doses of 1.5 g of HBsAg by
sMTS ID and by IM injection; at 8 weeks, after two doses, sero-
conversion was 100% and GMT was 158 for the ID route, and
20% and GMT 0 for the IM route.320 After dose 3, seroconversion
for IM rose to 80% and GMT to 34, while the ID route remained
at 100% and GMT rose to 410. In swine, a model virus-like pro-
tein (HBsAg) demonstrated dosage sparing via sMTS compared
with antigen delivered by IM control route.321
Experimental placement of the sMTS microneedles device
on human volunteers found it to be “well-tolerated” and “non-
intimidating and not painful”.315 A more recent public registra-
tion described a safety trial without antigen.296 Otherwise, no
further clinical data were found in public registries or reports.
The Georgia Institute of Technology (GA Tech),323 a pio-
neering center for microneedle technology, has worked with
Emory University to conduct numerous studies of coated
microneedles247,324 in animal models for cutaneous vaccine
delivery. In a series of murine studies using solid metal
microneedles coated with inactivated influenza viruses, cuta-
neous vaccination induced robust immune responses—often
better than equivalent dosages in controls injected by the SC
route—as well as protection against lethal viral challenge.325–334
When coated with BCG, the same microneedle platform (see
Figure 61-3D) was highly immunogenic in guinea pigs, with
robust cell-mediated responses in lungs and spleen compara-
ble to those with Mantoux injection.335 Similarly, plasmid DNA
antigen for hepatitis C, coated on 500-m-long needles, primed
specific cytotoxic T lymphocytes in vaccinated mice more read-
ily than did typical “gene gun” delivery336,337 or conventional
needle.338 Inactivated rotavirus vaccine—developed to avoid the
inhibitory effect of breast milk on live, oral vaccines339—was
coated onto this microneedle platform and found immunogenic
in an animal model.340
For most of these formulations prepared at GA Tech, a key
ingredient of the carboxycellulose matrix of the dried coating
was trehalose, one of several sugars, including sucrose, that
have been found useful in protecting protein antigens from
damage by drying and freezing, and thereby improving vaccine
thermostability.341
Another center for microneedle research, in Australia,342
developed a novel nitrogen gas jet-drying method for coating
antigen onto silicon that overcomes the challenges of dip-coating
closely spaced projections,324,343,344 but it still elutes within 2 to
3 minutes upon skin entry (see Figure 61-3F). It has achieved
1/30th to 1/100th dosage sparing compared with the IM route
in a mouse model for influenza.345,346 Other antigens studied
with good results in murine models with this platform—called
the Nanopatch and recently transferred to industry347—include
human papillomavirus,348 herpes simplex type 2,349,350 and the
West Nile and chikungunya viruses.351
Coulman and coworkers studied nanoparticles and DNA
plasmids expressing -galactosidase and fluorescent proteins
applied to the epidermal surface of ex vivo human breast skin
donated at mastectomy.352 After applying the microneedles to
the skin for 10 seconds, they were able to verify epidermal pene-
tration and gene expression by a variety of histologic and photo-
metric means. Later work by this Welsh group reported decreased
pain in clinical studies with 180-m and 280-m microneedles
compared with the 25-gauge conventional needle,353 as well
as morphologic changes suggestive of immune activation in
human Langerhans cells after intradermal injection of influ-
enza virus–like particles into excised human skin.354 This group
also found that both public and private immunization provid-
ers were positive, in focus-group discussions, toward micronee-
dles as a change from conventional needle-syringe delivery.355
Research on and development of coated microneedles for vacci-
nation are also underway by many other groups.292,293,302,356
Dissolving microneedles
An elegant strategy to decrease risk from intentional reuse
of, or inadvertent contact with, used microneedles is for the
sharps to dissolve in the skin with hydration, thus releasing
the antigen.32,43,357–361 The most common matrix for dissolvable
microneedles hard enough to penetrate skin is carboxymethyl-
cellulose, “generally recognized as safe” for parenteral delivery
by the FDA, among other compounds.357,358 Chu and Prausnitz
molded arrowhead-shaped antigen carriers of blended polyvinyl
alcohol and polyvinylpyrrolidone, and mounted them on metal
shafts (Figure 61-4A).358 The lower corners of the “arrows”
act as barbs to keep the carrier in the skin when the patch is
removed, which is done immediately. From the same group
at GA Tech and Emory, Sullivan and coworkers encapsulated
inactivated influenza vaccine virus into biocompatible polymer,
which dissolved within minutes after its application to mouse
skin (see Figure 61-4B).359 Robust antibody and cellular immune
responses provided complete protection from lethal challenge.
Several sugars, such as trehalose, sucrose, and maltose, have
been found to be key ingredients in stabilizing and maintaining
the potency of antigen during the process of forming dissolv-
able microneedles,341,362,363 but thermostability studies have not
yet been reported to assess whether such formulations would

the immune response will be focused to the foreign transgene
rather than to gene products synthesized endogenously by the
poxvirus. In addition, as seen with rAd, the concern of antivec-
tor immunity remains for this virus as well, although it may be
a lesser concern for canarypox vectors.
Poxvirus vectors show thermostability, an ability to incorpo-
rate a large foreign transgene, a lack of persistence or genomic
integration, and a demonstrable success in smallpox eradica-
tion. However, the difficulties in manufacturing virus in high
yields from primary CEFs, as well as their antigenic complex-
ity, reactogenicity and poor immunogenicity, have limited their
usefulness in human trials. Whether additional modifications
of these vectors can be made to facilitate human trials remains
unknown. If such modifications of the vector platform can be
achieved, this vector may have an opportunity to contribute to
the development of a variety of successful vaccines.
Adeno-associated viruses
The adeno-associated viruses (AAVs) were defined initially by
their presence as “helper” viruses that facilitated the propaga-
tion of wild-type adenovirus in cell culture. In contrast to the
large genome sizes of rAd and vaccinia vectors, this virus is
much more limited in size, with an insert size of approximately
5 kb. Like other replication-defective viruses, these particles can
be produced in packaging lines that provide complementary
structural proteins made constitutively by the cell rather than
the virus. A variety of serotypes have been defined,237 and an
HIV vaccine expressed in AAV2 has been analyzed in phase 1
human studies, without evidence of strong immunogenicity.
Alternative serotypes, including AAV1, are currently under
development and may be assessed both alone and in prime-
boost combinations for efficacy in humans. These vectors have
also been used recently to deliver recombinant antibody genes
that protect against viral infection,238 raising the intriguing pos-
sibility that gene-based antibody delivery might be used to gen-
erate protective immunity.
Vectors in development
Alphaviruses are negative-stranded RNA viruses that can be
modified to express foreign recombinant genes without produc-
ing pathogenic infections often seen with prototypes such as
Venezuelan equine encephalitis virus,239,240 Sindbis virus,241,242
and Semliki Forest virus. Replication-defective herpes simplex
virus (HSV) can be produced using packaging cell lines similar
to those described for replication-defective rAd5, AAV, or alpha-
virus vectors. These vaccines have been developed not only to
deliver foreign genes as potential immunogens but also to be
vectors against HSV itself, including both HSV1 and HSV2.243
More recently, vesicular stomatitis virus, dengue virus type 4,
yellow fever virus, and alphavirus have been modified to express
heterologous viral genes for vaccines for infectious disease tar-
gets including HIV, West Nile virus, filoviruses, CMV, and other
pathogens.244–251
Cell substrates
The progress of more recent viral vectors has depended on the
development of appropriate packaging cell lines and cell sub-
strates for viral production. Changes in regulatory requirements
that allowed the advancement of transformed cell lines for virus
production have proved invaluable in facilitating this effort. For
recombinant adenoviral production, the PERC6 and GV11 cell
lines have supported production of clinical-grade adenovirus
type 5, and these have progressed into trials for HIV and are
under study for other infectious agents, such as Ebola virus,
The development of gene-based vectors for immunization 62 1241
Marburg virus, tuberculosis, and malaria. Once approved, these
cell lines can be used for diverse vectors, and the PERC6 cell
line has now been used to develop a number of vaccines, includ-
ing those for West Nile and influenza viruses. In these latter
cases, the propagated virus is subsequently inactivated before
administration to humans.
For the generation of replication-defective viral vectors, these
cell lines allow the production of vectors that can be used in
human vaccine studies. Of the viruses developed for such vac-
cines, representative members, summarized in Figure 62-1B,
include recombinant Ad, poxviruses, measles, Venezuelan
equine encephalitis virus, and AAV, all of which have progressed
into human trials. The development of transformed and con-
tinuously propagatable cell lines, in contrast to the previous
standard, avian leucosis–free primary CEFs, represents a major
advance in vaccine production technology, largely because these
cell lines facilitate the production of replication-defective viral
vectors in stably transfected cell lines. Such lines also offer
potentially improved yields and stable production capacity.
The development of these lines has taken years to implement
because of regulatory concerns regarding adventitious agents,
tumorigenicity, and other safety and consistency consider-
ations. Oversight and evaluation of the strengths and limita-
tions of these cell substrates continues,252 based on guidelines
created several years ago,253,254 with an increasing number of
such lines becoming better characterized and available.
Bacterial vaccine vectors
Because many infectious agents replicate at mucosal mem-
branes and transit through the gastrointestinal tract for primary
infection, the ability to elicit effective immune responses at
these sites is desirable. A variety of bacteria are able to replicate
at mucosal sites of natural infection, and it has been proposed
that attenuation of these microorganisms and modification to
facilitate the delivery of antigen might allow the development
of improved vaccines to protect against pathogens that enter
through the mucosa. Development of live bacterial vectors has
therefore focused both on their ability to induce mucosal IgA
responses and on cytolytic T-cell responses at mucosal sites.
The delivery of antigens into mammalian cells to stimulate
antibody responses does not require the types of novel gene-
based vaccines summarized in this chapter. On the other hand,
the synthesis of proteins in mammalian cells delivered by bac-
terial vectors has the potential to induce the cellular immunity
that is the goal of many gene-based viral and nonviral vaccines.
These approaches have been reviewed in detail elsewhere255–257
and are summarized briefly here.
Among the live bacterial vectors used for antigen delivery,
there are mucosal pathogens that have been attenuated, includ-
ing strains of
and
In addition, there are commensal strains
such as lactobacilli, and staphylococci
that have been used for the induction of humoral and cellular
responses. For gene-based vaccination, has
been a particular focus of research. This gram-positive intra-
cellular pathogen has been studied as a model for understand-
ing class I MHC-restricted immune responses. These responses
are normally seen against the bacterial proteins or coexpressed
antigens. This microorganism uses a specialized system to
introduce proteins into cells and facilitate processing and pre-
sentation through MHC class I, and different mutations have
been used to develop attenuated strains that retain the abil-
ity to deliver antigens. Similarly, bacterial strains
are intracellular pathogens that become restricted to the endo-
somal compartment of eukaryotic cells, where they are resis-
tant to lysis.258 A variety of mutations have been introduced into
to generate several different live vaccine carriers,
1242 SECTION THREE
and these vaccine prototypes have undergone further devel-
opment for vaccine delivery. Among the other bacterial carri-
ers, Calmette-Guérin (BCG) has been a widely used
bacterial vaccine; for example, this organism has been used to
express HIV antigens.259,260 In some instances, expression of
mammalian genes has required modification of codons more
consistent with the host cell type, which has improved immu-
nogenicity. At present, however, the ability of such microorgan-
isms to induce cellular immunity is limited.
An area of intense interest has been the use of live bacterial
vectors for the delivery of DNA vaccines. In this instance, the
aim is for the bacteria to deliver plasmid DNA into the cyto-
plasm of infected cells; organisms such as and
have been used for this purpose.261,262 In addition, attenuated
has been evaluated for these purposes and has
shown some promise in both infectious disease and tumor
models in experimental animals.263–265
Although the use of such bacterial vectors has been attrac-
tive in theory, it has been more difficult to reduce this method
to practice. Among the concerns is the possibility of rever-
sion or reactogenicity of these potentially pathogenic bacteria
to wild type forms, the stability of the recombinant bacteria,
and the possibility that preexisting immunity from exposure
to natural pathogens may limit their infectivity. A variety
of host genetic factors can modulate the immune response
induced by the bacterial carrier, and variability in the innate
immune responses to such pathogens may limit their consis-
tency in vivo. Finally, perhaps the most challenging problem
has been the ability to effect a gene transfer from bacteria into
mammalian cells. It is likely that very specialized transport
pathways are required for the successful implementation of
this technology, and additional improvements will be neces-
sary to improve the efficacy of this approach, which remains
limited in its present form.
Clinical applications of gene-based vector
technology
Although substantial work has progressed in animal models
of vaccine efficacy, the ultimate value of gene-based vaccina-
tion has yet to be shown in human studies. Several trials using
the poxvirus technology have advanced into clinical evaluation.
These include canarypox, MVA, and NYVAC, which have been
Access the complete reference list online at http://www.expertconsult.com
3. Tang DC, DeVit M, Johnston SA. Genetic immunization is a simple method
for eliciting an immune response. Nature 1992;356:152–4.
5. Ulmer JB, Donnelly JJ, Parker SE, et al. Heterologous protection
against influenza by injection of DNA encoding a viral protein. Science
1993;259:1745–9.
6. Wang B, Ugen KE, Srikantan V, et al. Gene inoculation generates immune
responses against human immunodeficiency virus type 1. Proc Natl Acad Sci
U S A 1993;90:4156–60.
58. McConkey SJ, Reece WH, Moorthy VS, et al. Enhanced T-cell
immunogenicity of plasmid DNA vaccines boosted by recombinant modified
vaccinia virus Ankara in humans. Nat Med 2003;9:729–35.
87. Nabel GJ, Nabel EG, Yang Z-Y, et al. Direct gene transfer with DNA-
liposome complexes in melanoma: expression, biologic activity, and lack of
toxicity in humans. Proc Natl Acad Sci U S A 1993;90:11307–11.
150. Plautz GE, Yang Z, Wu B, et al. Immunotherapy of malignancy by in vivo
gene transfer into tumors. Proc Natl Acad Sci U S A 1993;90:4645–9.
evaluated in various phase 1 to 3 human studies. Because the
production technology for poxviruses is well known, and good-
manufacturing-practice procedures for amplification of these
viruses followed protocols similar to those developed for vac-
cinia virus, the path into clinical studies has been relatively
straightforward, as have the several trials of MVA, which has
been evaluated both as a vaccine for HIV (alone and in prime-
boost combinations) and as a potentially safer next-generation
vaccine for smallpox.
Additionally, DNA vaccines have undergone phase 1 testing
for a variety of infectious diseases, including Ebola virus, West
Nile virus, the SARS coronavirus, and influenza virus. Proof-of-
concept efficacy studies with these viruses have been performed
first in animal models with either DNA or in prime-boost com-
binations. In such studies, impressive protection has been dem-
onstrated.155,266 Based on these findings, several phase 1 trials
have been completed for Ebola, SARS, and West Nile virus dis-
ease targets.103,104,267
In the case of influenza, both naked DNA and DNA adju-
vanted with gold microparticles (by biolistics) have advanced
into clinical testing. Of particular interest is the development
of prime-boost strategies to stimulate the production of broadly
neutralizing antibodies to influenza viruses, demonstrated ini-
tially in mice, ferrets, and monkeys.268 Phase 1 studies testing
this concept in humans have revealed that even a single injec-
tion of a DNA vaccine can prime for an effective traditional vac-
cine boost against the H5N1 virus. This regimen also showed
that more broadly neutralizing anti-stem antibodies can be elic-
ited by vaccination in humans.269
It is likely that licensure of a gene-based vaccine remains
several years in the future. Recently, two DNA vaccines have
been approved for veterinary use, including a DNA vaccine for
West Nile virus in horses, developed by Fort Dodge,270 and a
DNA vaccine for infectious hematopoietic necrosis virus, devel-
oped by Merieux for use in farm-raised fish. An additional vac-
cine is being developed against viral hemorrhagic septicemia
virus in farmed salmon. In these studies, a single injection of
microgram amounts of DNA induces rapid and long-lasting
immune protection.271 A recombinant yellow fever vaccine has
advanced into efficacy studies as well.272 The precedent set by
these studies provides hope that additional gene-based vaccines
will become available for human use and may contribute to the
development of protective immunity for a variety of challeng-
ing infectious diseases that have thus far eluded the grasp of
vaccine-induced immunity.
160. Letvin NL, Mascola JR, Sun Y, et al. Preserved CD4 + central memory
T cells and survival in vaccinated SIV-challenged monkeys. Science
2006;312:1530–3.
162. Buchbinder SP, Mehrotra DV, Duerr A, et al. Efficacy assessment of a
cell-mediated immunity HIV-1 vaccine (the STEP Study): a double-blind,
randomised, placebo-controlled, test-of-concept trial. Lancet 2008;372:1881–93.
166. Casimiro DR, Chen L, Fu TM, et al. Comparative immunogenicity in rhesus
monkeys of DNA plasmid, recombinant vaccinia virus, and replication-
defective adenovirus vectors expressing a human immunodeficiency virus
type 1 gag gene. J Virol 2003;77:6305–13.
236. Rerks-Ngarm S, Pitisuttithum P, Nitayaphan S, et al. Vaccination with
ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand. N Engl J Med
2009;361:2209–20.
269. Ledgerwood JE, Wei CJ, Hu Z, et al. VRC 306 Study Team. DNA priming
improves influenza vaccine immunogenicity: two phase 1 open label
randomised clinical trials. Lancet Infect Dis 2011;11:916–24.
 SECTION FOUR: Vaccination of special groups
Vaccination of
immunocompromised hosts
During the past decades, the number of immunocompromised
people has increased rapidly; they are vulnerable to several
infections against which vaccines exist. New vaccines have been
developed or are in development that might result in important
advantages in reducing infectious disease morbidity and possi-
bly mortality such as against cytomegalovirus (CMV), the “troll
of transplantation”, and papillomaviruses since immunosup-
pressed people are at an increased risk of human papillomavi-
rus (HPV) complications.1,2 In 2009, a new strain of influenza
A H1N1 appeared and rapidly developed into a pandemic.
Immunosuppressed persons were recognized early during the
pandemic to be at risk for severe complications.3–8 New vaccines
were developed and introduced quickly, although data were lim-
ited about safety and efficacy. The need for increased knowledge
regarding vaccines in immunocompromised hosts is, therefore,
paramount.
The issue is also complex because the background and char-
acteristics of immunosuppressed states differ between differ-
ent patient categories. Children with immunodeficiencies have
defects based on the inherited genetic defects. People with can-
cer can be immunosuppressed by their original disease, such
as hematological malignancies, and/or by different anticancer
therapies. In transplant recipients, the highest risk for infection
usually is early after transplantation. Persons undergoing allo-
geneic hematopoietic stem cell transplantation (HSCT) are the
most deeply immunocompromised group through several differ-
ent mechanisms, which change with time after transplantation.
Solid organ transplant recipients can be immunosuppressed
before transplantation, but the main cause for the immunosup-
pressed state is the T-cell suppressive agents given to prevent and
treat graft rejection. Several drugs and combinations are used
today in transplant recipients, and the drugs used might influ-
ence the risk for infections and the response to vaccinations.
Vaccination of immunocompromised persons is important
from two viewpoints. Obviously, the need to protect a patient
against serious infections is the primary goal. However, the pub-
lic health point of view is also significant because it is impor-
tant not to have an increasing number of persons vulnerable
to infectious agents such as poliovirus. Both aspects require an
analysis of risks and benefits for individual persons.
Patients with cancer
Cancer chemotherapy and radiotherapy have been substantially
intensified during the past decades, and many new types of ther-
apy, including monoclonal antibodies and targeted anticancer
drugs such as tyrosine kinase inhibitors, have been introduced.
Per Ljungman 63
The number of studies assessing vaccination responses in
persons receiving these modern therapies is limited. Most stud-
ies have been performed in persons with hematological malig-
nancies, while studies of immune responses in adults and
children with solid tumors are limited. Many persons with
non-Hodgkin lymphoma (NHL) or chronic lymphocytic leu-
kemia (CLL) will have received rituximab or alemtuzumab as
part of their therapy. Both of these monoclonal antibodies circu-
late for a long time after infusion and can influence vaccination
responses.9,10 All of these limitations must be considered when
interpreting the recommendations shown in Table 63-1.
Nonlive vaccines
Pneumococci are important causes of infection in persons
with hematological malignancies. In children with leukemia,
the 14-valent vaccine gave suboptimal antibody responses.11
Similar results were found with the 23-valent vaccine in per-
sons with multiple myeloma, among whom fewer than 40%
obtained “protective” antibody levels after vaccination.12 The
response in persons with Hodgkin lymphoma varies with the
time relative to therapy when immunization is performed.
Persons who receive immunizations after chemotherapy and/or
radiotherapy have a severely impaired antibody response.13,14 In
contrast, good responses can be obtained if immunizations are
performed before therapy is initiated.14,15 The response in chil-
dren with Hodgkin disease was poorer if immunizations were
performed after splenectomy.16 However, antibody responses
can be elicited in persons who have undergone splenectomy
who have NHL or Hodgkin lymphoma.17 Repeated vaccinations
with polysaccharide vaccine given before and after splenectomy
to persons with Hodgkin lymphoma induced repeated antibody
responses and were not associated with serious side effects.18
People with lymphoma who have inadequate antibody
responses to the polysaccharide vaccine have an increased risk
for severe pneumococcal infections.19 Molrine et al20 showed that
one dose of a pneumococcal conjugate vaccine gave suboptimal
responses in persons who had been treated for Hodgkin disease.
Persons with CLL are at increased risk for severe pneumococcal
infection. Although persons with CLL respond to pneumococcal
conjugate vaccine, the response rates are lower than in control
subjects and worse late in disease.21 Repeated doses are prob-
ably necessary to achieve a stable protective immune response.
Chan et al22 showed that priming with a 7-valent conjugated
pneumococcal vaccine (PCV7) could improve the response to
the 23-valent polysaccharide vaccine in persons with previously
treated Hodgkin lymphoma.
1244 SECTION FOUR Vaccination of special groups
Recommendations for Immunizations in Persons With Cancer
CLL, chronic lymphocytic leukemia; Hib, type b; MMR, measles, mumps, and rubella.
Immunization of persons with lymphoma and CLL against
pneumococcal infections is recommended as early as possi-
ble after diagnosis and preferably before chemotherapy and/or
radiotherapy is initiated.
Severe infections with type b (Hib) in
patients with cancer are less common than pneumococcal infec-
tions. However, children with leukemia are at a greater than
sixfold increased risk for Hib infection compared with healthy
children. Immunization with a conjugated Hib vaccine resulted
in lower antibody responses than in healthy children, and a
booster dose was ineffective. Longer duration and intensity of
antileukemic chemotherapy were associated with an inadequate
response to immunization.23,24 Children undergoing therapy for
solid tumors also have lower than normal responses to vaccina-
tion with an Hib vaccine.25 Immunization with a conjugated
Hib vaccine is indicated in children with cancer, preferably early
during anticancer chemotherapy.
Limited data are available regarding the risk for severe Hib
infections and the response to vaccination in adults with can-
cer. Patients with multiple myeloma had antibody responses
comparable to those of healthy adults.12 Molrine et al20 showed
that 99% of persons with Hodgkin lymphoma in whom therapy
had been discontinued responded to vaccination. In contrast,
persons with CLL respond rather poorly to Hib conjugate vac-
cine, with response rates around 50%.26,27
The morbidity and mortality of influenza vary in different
types of cancer, with the most severe consequences occurring
in persons with acute leukemia undergoing induction chemo-
therapy.28 The morbidity in adults with other types of hema-
tological malignancies and in persons with solid tumors is
probably lower, although good epidemiologic data are lacking.29
Kempe et al30 reported that children with cancer were more
likely to have influenza, but their symptoms had a duration
similar to that in healthy control subjects, and, although the
study sample was small, there were few clinical complications.
Earle31 showed that in persons with colorectal cancer, persons
who were immunized had fewer chemotherapy interruptions
and were more likely to survive for a year. However, many of
the patients are elderly and are likely to have other diseases,
strengthening the indication for vaccination. The uptake of
the vaccination recommendation is varying, and Ring et al32
showed a lower vaccination rate in UK patients receiving che-
motherapy for cancer than in the healthy elderly population.
Studies of influenza vaccination in children with acute lym-
phoblastic leukemia (ALL) show that the proportion of children
reaching protective antibody levels varied between 45% and
100% for the different influenza subtypes in the vaccines.33–36
A Cochrane review regarding influenza in children with cancer
summarized that immune responses were weaker in children
receiving chemotherapy than in children who had completed
chemotherapy and in healthy control subjects.37 In a study of
children with solid tumors or lymphoma who were given one or
two doses of vaccine, the overall results regardless of chemother-
apy were rather poor, with only 38% achieving protective anti-
body levels to all three influenza strains in the vaccine.38 It has
been reported that an antibody titer protective in healthy con-
trol subjects failed to prevent influenza in 24% of children with
cancer, but the possibility that the severity of the infection was
reduced in the cases of vaccine failure cannot be excluded.30,39
Vaccination against the pandemic H1N1 influenza A was stud-
ied in 54 children with different types of malignancies.40 The
overall seroconversion rate was 44%, which was significantly
lower in univariate analysis in children with hematological
malignancies compared with children with solid tumors and
children with ongoing chemotherapy. In multivariate analysis,
children with ongoing chemotherapy had lower responses
( = .05) and there was a trend for lower responses in children
with hematological malignancies ( = .10).
The data regarding vaccination efficacy in adults with solid
tumors are limited. In a study of subjects with lung cancer, the
vaccination response was similar to that of healthy control sub-
jects.39 In a recent study, patients with breast cancer vaccinated
during ongoing chemotherapy had significantly lower responses
than healthy control subjects.41 This study also addressed whether
it matters when during chemotherapy cycles the vaccine is given;
a tendency was found that patients responded better if vaccinated
early rather than late after a given chemotherapy course.41
In contrast, most published studies show that adults with
hematological malignancies respond inadequately to vaccina-
tion. In a study including subjects with multiple myeloma, the
response rate to one dose of vaccine was only 19%.12 Similar
results have been seen in subjects with lymphoma.42 Brydak

and Calbecka,43 in a small study including a mixed group of
subjects with hematological malignancies, showed response
rates of 10% for influenza A (H1N1), 35% for influenza A
(H3N2), and 70% for influenza B. In a study of subjects with
NHL, 32 were vaccinated, half with ongoing therapy (type not
specified) and half without therapy.44 The immune responses
were similar between groups, although the responses tended to
be stronger in subjects without ongoing therapy but were lower
than in healthy control subjects regarding response rates and
strength of responses.44 Another larger study from the same
group gave similar results comparing previously treated (within
2 months of vaccination) with untreated patients and control
subjects.45 One possible way to improve the results is to give
repeated doses of vaccine. Adults with lymphoma receiving a
two-dose schedule showed responses of approximately 30%
after one dose and approximately 45% after two doses of vac-
cine.46 However, two recent studies failed to show an improve-
ment by addition of a second dose for subjects with various
hematological malignancies10 or CLL.47 Preliminary data sug-
gest that vaccination results for people having received mono-
clonal antibodies for lymphoma are poor.10 The addition of
granulocyte-macrophage colony-stimulating factor (GM-CSF)
does not improve vaccination results.48
de Lavallade et al9 reported results of two doses of pandemic
H1N1 vaccine in subjects with chronic myeloid leukemia and
B-cell malignancies and compared with one dose given to healthy
control subjects. The study show improved seroconversion rates
after the second compared with the first dose (68% in B-lymphoid
malignancies; 95% in chronic myeloid leukemia), albeit lower
than in healthy control subjects after both doses. The response
rates were better in subjects with a more than 12-month inter-
val between the end of chemotherapy and vaccination, and no
subject who received maintenance rituximab responded to vac-
cination. They also showed that vaccination was able to induce a
T-cell response in approximately one third of patients compared
with approximately 50% in control subjects.9
Influenza vaccination with the trivalent inactivated vac-
cine is recommended for immunocompromised persons,
including people with cancer.49,50 No data exist with the live
attenuated vaccine, and it is not recommended for immuno-
compromised persons.49,50 However, it must be recognized that
protective effectiveness is likely to be low in persons at high-
est risk of severe complications of influenza. Thus, other pre-
ventive strategies are needed, including antiviral drugs. Elting
et al28 reported that most influenza infections in persons with
acute leukemia undergoing chemotherapy were nosocomially
acquired.28 Immunization of family members and hospital staff
is, therefore, recommended.
The protection against tetanus, diphtheria, and poliovirus is
frequently low in persons with cancer undergoing chemother-
apy. Hammarström et al51 have shown that 41% of persons
without transplant but with acute leukemia were not pro-
tected against tetanus. Risk factors for loss of immunity were
ALL vs acute myelogenous leukemia, more advanced disease,
and increasing age. People with lymphoma are also likely to
have deficient immunity. In children treated for cancer, the
immunities against tetanus, diphtheria, and poliovirus were
lower than what would be expected in an age-matched pop-
ulation.52,53 The loss of specific immunity is related to the
intensity of chemotherapy. Children with high-risk ALL were
more likely to become nonimmune to tetanus and diphtheria
than persons with low-risk or standard-risk ALL.54 In con-
trast, Nordoy et al55 reported that treatment of low-grade
NHL with radioimmunotherapy did not influence specific
immunity to tetanus.
Vaccination of immunocompromised hosts 63 1245
Immunity to tetanus toxoid, diphtheria toxoid, and poliovi-
rus is frequently deficient in persons with cancer, and immu-
nizations should be considered. The responses to diphtheria
toxoid and tetanus toxoid vaccinations in adults with cancer
have not been systematically studied. Most children respond
well to vaccination after chemotherapy for malignancies.53,56
However, children treated for high-risk ALL had poor responses
(22% protected against tetanus; 56% against diphtheria).54
Stenvik et al57 reported on 14 children with leukemia who
were given a booster dose of vaccine, with 12 of 14 respond-
ing. No data exist regarding the vaccination of adults with can-
cer. Despite the absence of data, it seems logical to recommend
an investigation of poliovirus immunity and booster immuniza-
tion for persons with cancer traveling to areas of the world still
endemic for poliovirus. The oral poliovirus vaccine can induce
paralytic disease in immunocompromised persons. Furthermore,
the vaccine strain might be transferred from healthy persons,58
so the use of an inactivated vaccine is recommended for family
members and health care providers caring for cancer patients.
Hepatitis B virus (HBV) infection is a major cause of morbidity
in many parts of the world. The currently available vaccines are
plasma-derived or produced through DNA recombinant tech-
nology. Several studies have been performed regarding the effi-
cacy of HBV vaccination.59–64 The results are summarized in
Table 63-2. All studies cited in the table report results in chil-
dren, and, to my knowledge, no study has reported results in
adults with cancer or hematologic malignancy.
Vaccinations could be considered against pertussis and menin-
gococci, although no studies have been published in persons
with cancer about the frequency or severity of these infections.
The number of long-term survivors after cancer treatment
is increasing. The survivors might be at increased risk for late
complications, such as those driven by HPV infections. The
existing papillomavirus vaccines have not been studied in per-
sons with cancer but could be considered, especially in survi-
vors after cancer during childhood.2
Live vaccines
Primary varicella infections cause high mortality in children with
cancer. The existing vaccine is live, attenuated, and based on the
Oka strain.65 The vaccine was shown to be effective and safe in
children with leukemia in remission.66 The seroconversion rate
was 88% after one dose and 98% after one to two doses. The rate
of varicella infection in vaccinees was 8%; all infected children
had mild disease.66 The frequency of side effects from the vaccine
is low, and breakthrough vaccine disease can usually be treated
effectively with acyclovir.66,67 The varicella-specific immunity is
stable for at least 5 years after immunization. However, reinfec-
tions in previously vaccinated children with cancer have been
reported68 and might become an increasing problem with more
adults with cancer or leukemia having been vaccinated during
childhood. The risk for herpes zoster after vaccination is lower
compared with that in people who had natural varicella dis-
ease.69,70 In a small randomized study, varicella vaccine was given
to children with newly diagnosed cancer before starting chemo-
therapy. There was a high rate of seroconversion, and no severe
side effects were found, but the study sample was small.71
Household exposure to varicella is associated with more
severe varicella disease in secondary cases. An option would
be to immunize healthy seronegative family members when
1250 SECTION FOUR
after HSCT.84,150 During the aforementioned outbreak in Brazil,
patients were vaccinated 1 year after HSCT without serious side
effects.151 The reported effect of vaccination varies among differ-
ent studies, with a seemingly higher response rate in adults than
in children.84,104,151–153
Rubella vaccine could be indicated in females who have the
potential for becoming pregnant. Existing data indicate that
rubella vaccine can be given without severe side effects 2 years
after BMT in patients without chronic GVHD or ongoing immu-
nosuppression.84 The effectiveness of the vaccine is high.84,153
The same risks exist with the vaccine against mumps. The risk
for severe infections with mumps virus in HSCT recipients is
likely low, although a case report of a fatal infection has been
published.154 Furthermore, the efficacy of mumps vaccination to
induce long-lasting serologic immunity has been shown to be
rather poor, at least in children.153 Vaccination with MMR is rec-
ommended for children and seronegative adults. A second dose
of MMR is recommended for children younger than 9 years.95,96
Other vaccines that can be considered in allogeneic HSCT recip-
ients are BCG and yellow fever. The possible benefits must be
weighed against the risk for side effects. Yellow fever is a life-
threatening infection primarily occurring in South America and
southern and central Africa. Rio et al155 described three patients
who were immunized 5 years after HSCT without severe side
effects, and this experience has since been expanded to 25
patients without serious side effects (B. Rio, personal communi-
cation). Immunization could be considered for persons who live
in or must visit areas where yellow fever is endemic. The BCG
vaccine can cause severe infections in persons with depressed
T-cell function and is not recommended for HSCT recipients.95,96
Autologous hematopoietic stem cell
transplantation recipients
In autologous HSCT recipients, there is no immunologic dispar-
ity between the graft and the host, and the immune regeneration
is faster than after allogeneic HSCT. Autologous HSCT recipients
are usually not prone to severe infections that are preventable
by immunization during the early phase after transplantation.
Several studies have shown that autologous transplant recipients
will also lose protective immunity to tetanus, poliovirus, and mea-
sles during long-term follow-up.83,156–159 Many patients undergo-
ing autologous HSCT for lymphoma will have received rituximab
or alemtuzumab. Both of these antibodies circulate for a long time
after infusion and can influence responses to vaccination.9,10
Nonlive vaccines
Influenza virus immunization with the inactivated vaccine
is recommended starting at 4 to 6 months after autologous
HSCT.95,96 However, the response to vaccination is likely to be
suboptimal early after transplantation.121,122,124 In one study, no
patient responded if the immunization was performed earlier
than 6 months after autologous BMT.121 No data exist with the
live attenuated vaccine, and it should, therefore, not be used.
Autologous HSCT recipients are less prone than allogeneic
recipients to the development of severe infection with Hib or
pneumococci. The infections are more likely to occur early after
transplantation when the response to immunization is poor. The
response to a single dose of pneumococcal polysaccharide vac-
cine is poor regardless of the stem cell source87,160 and remains
decreased compared with the response in control subjects for
several years after transplantation for lymphoma.159 Vaccination
with the PCV7 was shown to be effective and induced protective
antibody levels in more than 60% of vaccinees.161 Vaccination
before stem cell harvest resulted in significantly higher anti-
body levels at all studied time points up to 12 months after
HSCT. Vaccination with PCV13 is recommended starting at 3
to 6 months after autologous HSCT.95,96,114,115
Autologous HSCT recipients have an increased risk, compared
with the healthy population, for loss of protective immunity to
poliovirus,157,159 diphtheria,159 and tetanus.156 Reimmunization
with repeated doses of inactivated poliovirus vaccine, diph-
theria toxoid, and tetanus toxoid effectively restores protec-
tive immunity.83,157,159,160 There was no difference in response
in autologous bone marrow or peripheral blood stem cell graft
recipients.160 The immune response early after autologous
HSCT can be improved by immunizing the patient before the
stem cell harvest followed by tetanus toxoid given repeatedly
after autologous HSCT.162 Vaccination with reduced-dose per-
tussis vaccine (Tdap) failed to induce immune responses in
most autologous HSCT recipients.163 Immunization with teta-
nus toxoid, diphtheria toxoid, and inactivated poliovirus is rec-
ommended after autologous HSCT.95,96
Live attenuated vaccines
Children who have been immunized to measles before autol-
ogous HSCT frequently become seronegative during follow-
up, but adults who experienced natural measles disease before
autologous HSCT usually remain immune.158 The risk for side
effects after immunization seems low.158 MMR vaccination is
recommended for children and seronegative autologous recipi-
ents not earlier than 24 months after HSCT.95,96 Two doses are
recommended for children younger than 9 years.
No studies have been done with the varicella vaccine in sero-
negative autologous HSCT recipients. In seropositive patients
undergoing autologous HSCT for lymphoma, a randomized
study was performed with heat-inactivated varicella vaccine
given in four doses: within 30 days of vaccination and 30, 60,
and 90 days after vaccination. During the 12 months posttrans-
plantation, the rate of zoster was 13% in vaccinees and 33%
in unvaccinated patients, giving an efficacy of 61%.164 The live
attenuated vaccine was given in a pilot study to autologous
HSCT recipients 3 months after transplantation. No side effects
were seen, and there was a trend for improved cell-mediated
immune response after the vaccination.165
Solid organ transplant recipients
The need for immunization in solid organ transplant (SOT)
recipients can arise from three factors, each causing a suppres-
sion of the immune system: the immunosuppressive activity
of the underlying disease (eg, chronic renal failure), rejection
of the organ graft, and the immunosuppressive therapy given
after transplantation. Immunizations can be given to candidates
Immunogenicity of Hepatitis B Vaccines in HIV-1–Infected Children and Adults

Vaccination of human immunodeficiency virus–infected persons

64
1261
 1262
SECTION FOUR Vaccination of special groups
Immunogenicity of Hepatitis B Vaccines in HIV-1–Infected Children and Adults—cont'd

HAART, highly active antiretroviral therapy.

*Follow-up study of children previously described by Diamant et al, 1993.184
Immunogenicity of type b and Pneumococcal Conjugate Vaccines in HIV-1–Infected Children and Adults

Vaccination of human immunodeficiency virus–infected persons

64
1263
 1264
Immunogenicity of type b and Pneumococcal Conjugate Vaccines in HIV -1–Infected Children and Adults—cont'd

SECTION FOUR
HAART, highly active antiretroviral therapy; GMT, geometric mean antibody titer.
*Population-based study.

vaccine can induce rapid increases in antibody in some HIV-1–
infected children.67–70
Few studies of immune responses to Hib vaccine have been
conducted among children receiving HAART. In one study of 18
previously vaccinated children, 78% had detectable antibodies
after HAART initiation.14
Pneumococcal polysaccharide and polysaccharide-
protein conjugate vaccines
Responses to pneumococcal conjugate vaccines among chil-
dren not receiving HAART were high, ranging from 63% to 93%
by serotype, but were generally lower than among uninfected
children (Table 64-4). The quality of antibodies was deficient
among HIV-1–infected children in South Africa,64 and signif-
icant waning of antibody concentrations was found 5 years
after vaccination, such that the proportion with adequate lev-
els ranged only from 5% to 24%.71 Provision of a booster dose
improved responses for only three of seven serotypes.71 Vaccine
efficacy was lower in HIV-1–infected South African children
(65%) and decreased over time (39%) compared with unin-
fected children (83%).72,73
Among children receiving HAART, responses to pneu-
mococcal vaccines varied by serotype, ranging from 29% to
100%, with some evidence of waning immunity.70,74 The
benefits of a booster dose are unclear because studies have
reported conflicting results.70,74–76 Immune responses by
immunological status and viral load suppression were also
inconsistent.64,74–76 Children in whom HAART was initiated
in infancy had responses similar to uninfected children and
better responses than children in whom HAART was initi-
ated later in childhood.20 Another study of early and deferred
HAART among infants found similar responses between the
two groups and in comparison with uninfected children but
lower levels of functional antibodies among infants in the
deferred HAART group.77
Response to 23-valent pneumococcal polysaccharide vaccine
is poorer in HIV-1–infected than in uninfected persons but
better in persons receiving HAART.78–80 However, antibody
responses declined even in persons receiving HAART.81 The
23-valent pneumococcal vaccine reduced the risk of pneumo-
coccal disease in 305 HIV-1–infected adult vaccine recipients
receiving HAART in Taiwan82 and the risk of pneumonia in a
cohort of 23,255 HIV-1–infected adults in the United States.83
However, in HIV-1–infected Ugandan adults, the 23-valent
polysaccharide pneumococcal vaccine was found to be inef-
fective in preventing first episodes of invasive pneumococcal
disease.84
The antibody response to a glycoprotein conjugate pneu-
mococcal vaccine was better than that of the polysaccha-
ride vaccine in HIV-1–infected persons, except for adults
with CD4 T-lymphocyte counts less than 200 cells/mm3.85–87
However, the antibody response to specific pneumococcal
polysaccharide serotypes varies, with some serotypes elicit-
ing poor antibody responses in HIV-1–infected persons with
CD4 T-lymphocyte counts less than 200 cells/mm3.88 In a
placebo-controlled trial of seven-valent pneumococcal vac-
cine in predominantly HIV-1–infected adults in Malawi, all
of whom had prior invasive pneumococcal disease, vaccine
efficacy was 74%.89 Revaccination with conjugate pneumo-
coccal vaccine resulted in a higher proportion of HIV-1–
infected adults with protective antibody levels than adults
Vaccination of human immunodeficiency virus–infected persons 64 1265
vaccinated with polysaccharide vaccine, but this difference
was not sustained after 6 months.90
Meningococcal polysaccharide and conjugate
vaccines
Few data are available regarding responses to meningococcal
polysaccharide vaccine in HIV-1–infected children. Response
rates in a recent trial (IMPAACT P1065) of the immunogenicity
of a meningococcal polysaccharide diphtheria toxoid conjugate
vaccine among 319 HIV-1–infected adolescents 11 to 24 years
old, the majority of whom were receiving HAART, were 86% for
serogroup A, 55% for serogroup C, 72% for serogroup W-135,
and 73% for serogroup Y, although some level of immunity also
was detectable at baseline.91 Response rates were positively cor-
related with better immunological status and inversely with
HIV-1 viral load. The Advisory Committee on Immunization
Practices recommends a primary series of two doses of the
meningococcal conjugate vaccine for HIV-infected adolescents
11 through 18 years old.92
Rabies vaccine
Antibody responses to an eight-site intradermal rabies vacci-
nation regimen using the purified chick embryo cell rabies
vaccine resulted in good antibody responses among 27 HIV-
infected adults regardless of CD4 T-lymphocyte count,93
although another study suggested poor antibody responses to
the human diploid cell rabies vaccine in HIV-infected persons
with low CD4+ T-lymphocyte counts.94 Antiretroviral therapy
may improve the immunogenicity of rabies vaccine but not to
levels found in uninfected persons.95
Human papillomavirus vaccines
The human papillomavirus vaccine is recommended for
HIV-1–infected and uninfected children.6 The quadriva-
lent human papillomavirus vaccine was found to be safe and
immunogenic in a study of 126 HIV-1–infected children 7 to
12 years old.96
Japanese encephalitis vaccine
HIV-1–infected children have a poor response to mouse brain–
derived, inactivated Japanese encephalitis vaccine, but a
higher proportion responded to revaccination while receiving
HAART,97,98 and levels of protective antibodies persisted for at
least 3 years.99 A cell culture vaccine for Japanese encephalitis
is available although no data are available on its use in HIV-
infected persons.
Safety of nonreplicating vaccines
Nonreplicating vaccines are not associated with increased risks
of complications in immunocompromised persons. However,
a study of HIV-1–infected Ugandan adults found an increased
incidence of pneumonia in recipients of 23-valent pneumococ-
cal polysaccharide vaccine compared with unvaccinated HIV-1–
infected adults.84 The authors hypothesized that immunization
resulted in destruction of polysaccharide-responsive B-cell
clones, but no specific data to support this hypothesis were pro-
vided and this observation has not been corroborated in other
studies. Surprisingly, 6 year follow-up confirmed excess all-
cause pneumonia, but an overall survival advantage persisted
among vaccinated persons.100

Status of the National Vaccine Injury Compensation Program as of May 1, 2012
*Includes attorney fee awards. Some adjudicated claims above have not yet been processed for payment.
than $200 million in annual receipts against awards totaling
$189.8 million the previous fiscal year.) In addition, the licen-
sure of multiple-antigen combination vaccines makes difficult
the task of estimating the risk of adverse events for each anti-
gen. With passage of the Taxpayer Relief Act of 1997, the vac-
cine excise structure was revised, setting a 75-cents-per-“dose”
(disease prevented) rate on all covered vaccines under the pro-
gram, including the three vaccines added by rulemaking. The
effective date of coverage was August 6, 1997.
VICP cases since 1986
Without question, the biggest controversy (and challenge) in
adjudicating cases filed in the VICP has been determining vac-
cine causation. The legislative history underlying the NCVIA
noted the lingering controversy about what is and is not vac-
cine-caused. As science evolves on vaccine-injury causation,
the VIT would serve as an interim compromise mechanism to
facilitate recovery by persons presumed to be injured as a result
of vaccination. For the first 8 years after the VICP was created,
the vast majority of petitions filed alleged a VIT injury (“Table
claims”), which confers the presumption that the injury, signifi-
cant aggravation, or death, were vaccine-related. A VIT-based
adjudication was straightforward: A petition would usually say
that the injured person experienced the signs and symptoms
described in the VIT, as defined by the Qualifications and Aids
to Interpretation (QAIs), within a VIT time frame, regardless
of whether these allegations were substantiated or contradicted
by the medical record. Accordingly, in the earlier years of the
VICP, the special masters and the courts concentrated almost
exclusively on developing a body of law establishing the quality
and quantity of evidence needed to prove an injury included on
the VIT. Since that time, the majority of petitions have shifted
to allege injuries not found on the VIT (“Off-table claims”), in
which petitioners are required to prove that the vaccine actually
caused the injury.
Most special masters place more weight on the medical
records most contemporaneous to the vaccination than on
contradictory evidence assembled later, after memories fade or
when litigation is contemplated.32 In circumstances in which
the contemporaneous medical records were contradictory or did
not exist, the special masters relied on the credibility of the
family and the expert witnesses to determine whether a VIT
injury occurred. The appellate courts gave deference to spe-
cial masters' credibility determinations, as long as they were
not arbitrary or capricious.33 The courts also addressed ques-
tions concerning whether special masters must strictly apply
and interpret the definitions in the QAIs, which accompany the
VIT. In
34
and
35
the US Court of Appeals for the
Federal Circuit found that because special masters have wide
Legal issues 77 1491
latitude in adjudicating claims, they are not required to apply
the QAIs in a mechanical manner; rather, they have reasonable
discretion when applying them.
36
is the landmark Table injury case
regarding the meaning of “significant aggravation” under the
statute. Significant aggravation is important in VICP cases
because if the onset of signs or symptoms following vaccination
is not the first evidence of the alleged vaccine-related condition,
to receive compensation, the petitioner must prove that the
vaccine significantly aggravated a condition present before vac-
cination. The NCVIA defines significant aggravation as “any
change for the worse in a preexisting condition which results
in markedly greater disability, pain, or illness accompanied by
substantial deterioration of health”.37 According to the legisla-
tive history, an example would be a child whose seizure fre-
quency increased from one per month before vaccination to one
per day after vaccination.38 The special master in
denied compensation on the basis that the child was born with
a brain disorder that was responsible for the encephalopathy
that appeared after receipt of the third DTP vaccination. The
Court of Federal Claims agreed, but the US Court of Appeals
for the Federal Circuit reversed the decision based on a differ-
ent interpretation of “first symptom or manifestation of onset”,
writing that “first” did not necessarily mean that other clinical
signs could not appear before vaccination.39
The US Supreme Court unanimously reversed the Federal
Circuit's decision, finding that the Federal Circuit misread the
NCVIA. However, Justice O'Connor, in a concurring opinion,
noted that the Federal Circuit opinion did not address the issue
of significant aggravation of a preexisting condition, but instead
only addressed the issue of “first symptom or manifestation of
onset”. The case was remanded to the Federal Circuit, which
set forth a test for Table injury-significant aggravation claims,
specifying that the special master must (1) assess the person's
condition before the administration of the vaccine, (2) assess
the person's current condition, and (3) determine if the person's
current condition constitutes a significant aggravation of the
person's condition before vaccination within the meaning of the
statute; and, if the special master determines there has been
a significant aggravation, then he or she must (4) determine
whether the first symptom or manifestation of the significant
aggravation occurred within the period prescribed by the VIT
for the injury.40 Although involved a Table injury,
the courts have applied the significant aggravation test in off-
Table claims.41
A clarification of the fourth step of the test was
announced in
42
The Court of Federal Claims clarified that
the special master, in determining the first symptom or mani-
festation of onset of the significant aggravation, must take into
account the preexisting condition. The Court found that the
definition of the term should more prop-
erly be understood as “any change for worse in [
1492 SECTION FIVE
] which results in a markedly greater dis-
ability, pain, or illness accompanied by substantial deterioration
in health”. Furthermore, the Court indicated that the four-step
test articulated in does not merge the significant
aggravation and sequela inquiries. Rather, to obtain compensa-
tion for sequelae, a petitioner must establish that the aggrava-
tion of the Table injury caused them, without reliance on any
Table presumption.
The parameters of the defense also have
been tested. Under the NCVIA, if petitioners show proof of a
Table condition occurring within the specified time, they are
entitled to a presumption of vaccine causation unless the gov-
ernment can show greater evidence of an alternative cause
or factor unrelated. The US Court of Appeals for the Federal
Circuit generally used a strict interpretation when it applied
the NCVIA's definition of the term: A factor unrelated may
not include “any idiopathic, unexplained, unknown, hypo-
thetical, or undocumentable” condition.43 However, the Federal
Circuit has considered whether the government was required
to specifically identify a virus it alleged was a factor unrelated.
In
44
the Federal Circuit held that so long as the govern-
ment could prove that the petitioner did in fact have a viral
infection, the government need not identify the type of viral
infection with specificity.
The original NCVIA allowed the Secretary of HHS to modify
or amend injuries listed in the VIT and QAIs (see “Modifying
the VIT”). In petitions filed after the first set of VIT changes
took effect, the focus of most claims changed from proving a
VIT injury to proving that the injury was caused-in-fact by the
vaccine. This shift occurred for several reasons. First, HHS
removed residual seizure disorder and shock collapse from the
VIT. Second, the QAIs for encephalopathy were revised so they
more accurately reflected the medical view of significant acute
and chronic neurologic injuries, instead of the normal, tran-
sient side effects common to some childhood vaccines.45
Third, the new vaccines added to the VIT in recent years have
had few associated VIT injuries. Since 1997, only two conditions
were added for the nine “new” vaccines. Moreover, of more than
500 HBV claims, only one or two involve anaphylaxis or anaphy-
lactic shock, the only VIT condition listed for this vaccine.
Fourth, and arguably most significantly, the licensure of acel-
lular pertussis vaccines for primary use in infants and the tran-
sition to an IPV-only schedule all but eliminated DTP vaccine
and OPV filings, the former of which comprised a significant
percentage of alleged injury and death claims.
Current cases, therefore, establish a struggle by the special
masters and the courts to strike an appropriate balance between
(1) denying compensation of claims based merely on a temporal
relationship between the injury and the vaccine and a lack of
apparent alternative causes and (2) the need to fulfill the VICP's
mandate to provide compensation “generously” and “fairly”,
even in the absence of full scientific knowledge about poten-
tial vaccine adverse events. Early in the history of the VICP, the
court determined in
that causation-in-fact was shown when
there is “proof of a logical sequence of cause and effect show-
ing that the vaccination was the reason for the injury. A repu-
table medical or scientific explanation must support this logical
sequence of cause and effect”.46 The US Court of Appeals for
the Federal Circuit clarified this issue somewhat by holding that
the scientific evidence need not rise to the level of being a sci-
entific certainty.44 Still, there were questions. For example, spe-
cial masters considered whether animal studies could be used,
how much weight should be given to case reports, and whether
the petitioner had to rely on any published or peer-reviewed evi-
dence in pursuing a VICP claim.47,48
The standard of proof for causation-in-fact has been adjusted
and clarified over time. In
49
the US Court of Appeals for the
Federal Circuit found that the petitioners need not show that
the vaccination was the sole or even the predominant cause of
the injury or condition. Rather, the petitioner need only show
that the vaccination was at least a "substantial factor" in causing
the condition and was a "but for" cause. In
50
the Court held
that petitioner's burden was to show by preponderant evidence
that the vaccine brought about her injury by providing: “(1) a
medical theory causally connecting the vaccination and the
injury; (2) a logical sequence of cause and effect showing that
the vaccination was the reason for the injury; and (3) a showing
of proximate temporal relationship between vaccination and
injury”. The decision held that there is no requirement that
petitioners provide scientific literature to support their theory
of causation—medical opinion or medical records can suffice
under the law. The Court stated that while the case involved
“a sequence hitherto unproven in medicine, the purpose of the
NCVIA's preponderance standard is to allow the finding of cau-
sation in a field bereft of complete and direct proof of how vac-
cines affect the human body”. It further stated that “close calls
regarding causation are resolved in favor of injured claimants”.
The Federal Circuit expanded the holding in
51
in which it criticized the special master for inad-
equately considering the opinions of treating physicians. The
Court held that “medical records and medical opinion testi-
mony are favored in vaccine cases, as treating physicians are
likely to be in the best position to determine whether 'a logical
sequence of cause and effect show[s] that the vaccination was
the reason for the injury'”.
The Federal Circuit continues to refine the body of law regard-
ing causation-in-fact. In
52
the Federal Circuit clarified that
when the petitioner has proven all three prongs, the
petitioner is not required to eliminate other potential causes
in order to prevail. With regard to the nature of the petitioner's
evidence, in
53
the Federal Circuit noted that in proving a
medical theory of causation, the petitioner is neither required
to rely on epidemiologic studies nor to show that the theory has
general acceptance in the medical community. When petition-
ers offer medical literature or epidemiologic evidence, it is not
to be evaluated through the lens of a laboratorian, but instead
from the vantage point of the preponderance standard.
Most notably in however, was the decreased level
of deference given to the special master's credibility findings.
Under prior case law, the Federal Circuit asserted that it is not
its role “to reweigh … whether the special master correctly eval-
uated the evidence” or to “examine the probative value of the
evidence or the credibility of the witnesses”, because “these are
all matters within the purview of the fact finder”. Nevertheless,
in a shift away from deference, the Federal Circuit in
rejected the special master's credibility findings of the petition-
er's expert witness, remarking that special masters may not cloak
their evaluation of the evidence in the guise of a credibility find-
ing. Recently, however, in
54
the Federal Circuit renewed its
deference to special masters and reasserted that special mas-
ters are expected and entitled to make reliability and credibility
determinations of expert witnesses. Moreover, although peti-
tioners are not required to offer proof in the form of epidemio-
logic studies or well-established medical experience, it does not
mean that special masters are precluded from assessing the reli-
ability of testimony from expert witnesses.
In litigating these difficult, fact-intensive, causation-in-fact
cases, several special masters have indicated that a more expe-
ditious resolution could be obtained by combining several cases
in a consolidated or "omnibus" proceeding. To date, numerous

consolidated and omnibus proceedings have been initiated, and
several have concluded. The results of some of these proceed-
ings (eg, hepatitis B vaccine and demyelinating diseases) have
been controversial given the scientific literature.55–57
One such omnibus proceeding involved tuberous sclerosis
complex (TSC), a genetic disorder characterized in many cases
by classical skin lesions, growths (tubers), and other structural
changes in the brain, causing seizures and mental retardation.
Many of the claims alleging seizure onset within a VIT time
frame were litigated as a group to determine whether DTP vac-
cine significantly aggravated the preexisting TSC condition.
Although neurologists once believed that DTP vaccine might
aggravate TSC by triggering the early onset of seizures and,
therefore, result in more frequent seizures and associated greater
mental retardation, advances in magnetic resonance imaging in
the 1990s indicated otherwise.58,59 Based on these new data, the
presiding special master ruled that the number and presence of
tubers in the cerebral cortex, not the timing of DTP vaccination,
ultimately determines clinical outcome. Several peer-reviewed
articles by HHS consultants came from research generated by
these efforts.60–62 The special master's decision was affirmed by
the Court of Federal Claims and the US Court of Appeals for the
Federal Circuit in
63
64
and
65
By far the omnibus proceeding that has garnered the most
scrutiny is the Omnibus Autism Proceeding (OAP). Interest
and concern about vaccines and autism started with the pub-
lication of a 1998 case series by Wakefield and colleagues66 in
the suggesting that MMR vaccine was associated with
autism. Little attention was paid in the United States until the
following year when a congressionally mandated FDA review of
mercury in biological products showed some vaccinated infants
were receiving ethylmercury in excess of one federal safety
guideline established for methylmercury, the more extensively
studied form of organic mercury. Attention then became focused
on thimerosal, an ethylmercury compound used for decades
in the formulation of many routinely administered childhood
vaccines to prevent bacterial and fungal contamination.67 As
a result of the FDA review, beginning in 2001, individual and
class action lawsuits were filed in state courts alleging autism or
autism spectrum disorder caused by thimerosal-containing vac-
cines and/or the MMR vaccine, which does not contain the pre-
servative. Multiple large, well-controlled epidemiologic studies
all failed to find an association between thimerosal-containing
vaccines and autism.68–71 The IOM concluded in its report that
the body of scientific evidence favors rejection of a causal rela-
tionship between MMR and thimerosal-containing vaccines
and autism.72 A 2011 IOM report reached the same conclusion
for MMR vaccine and autism, and did not cover thimerosol-
containing vaccines as part of its review.72a Even still, some par-
ents of children with developmental disabilities argue that their
children's condition is vaccine-related, especially when physi-
cians are unable to provide a specific cause other than idiopathic.
The VICP began receiving claims alleging autism or autism
spectrum disorder caused by MMR or thimerosal-containing
vaccines starting in late 2001. As the number of claims grew to
nearly 5,000, the OSM established the OAP to adjudicate the
cases in a consolidated manner.73
The petitioners put forth six test cases to represent two sep-
arate theories of causation. Under the first theory (“Theory 1”),
petitioners argued that MMR vaccine and thimerosal-containing
vaccines cause autism. Under the second theory (“Theory 2”),
petitioners argued that thimerosal-containing vaccines alone
cause autism. The OSM conducted extensive evidentiary hear-
ings reviewing 939 medical articles, 50 expert reports, and tes-
timony from 28 experts. The hearings produced more than
5,000 pages of transcripts and 7,000 pages of past hearing briefs.
Legal issues 77 1493
The three presiding special masters issued decisions in favor of
the government in all six cases with respect to both theories.74
Petitioners appealed the three Theory 1 cases to the Court of
Federal Claims, without success.75 Petitioners further appealed
two of the three Theory 1 cases to the Court of Appeals for the
Federal Circuit.76 In both cases, the Federal Circuit ruled in favor
of the government, upholding the lower determinations that
petitioners had not proven their theory of causation by a pre-
ponderance of the evidence. Petitioners did not pursue appeals
in the Theory 2 cases.77
The OSM has now turned its attention to the nearly 5,000
remaining individual autism claims in which petitioners are
determining whether to dismiss their petitions or continue with
their petitions on an individual basis. Petitioners who choose to
continue will be required to file their theories of causation and
evidentiary support not introduced in the test cases as part of
the OAP.
In 2011, the Federal Circuit issued an decision hold-
ing that the Vaccine Act’s statute of limitations begins to run
on the date of occurrence of the first symptom or manifestation
of onset of the injury, reversing a decision below that concluded
the limitations period does not run until the medical commu-
nity at large understands and recognizes a causal relationship
between the vaccine and the injury.77a
Modifying the VIT
The Secretary of HHS has the authority to modify or amend
injuries listed in the VIT (and QAIs) in consultation with the
ACCV and after opportunity for public comment. Such changes
apply only to cases filed after the effective date of the changes.
Later in 1993, Congress provided a mechanism for the Secretary
to add to the VIT “new” vaccines recommended by CDC for rou-
tine administration to children based on the effective date of the
excise tax for that vaccine.30 Table 77-5 summarizes changes to
the VIT through 2011.
Separate efforts by the VICP to modify the initial statutory
VIT and QAIs began with publication of the two congressionally
mandated IOM reviews in 1991 and 1994, respectively.78–81 With
a few exceptions, the approach by the VICP was straightforward:
If the IOM concluded that there was evidence that a condition
was “causally related”, it was added to the VIT or left on the
VIT. However, if there was no proven evidence of an association,
it was removed. Publication of final rules in 1995 and 1997 set-
ting forth these modifications was the final step in a 3- to 4-year
process for each set of changes involving ad hoc scientific panel
reviews by the NVAC, mandatory ACCV reviews, and a 180-day
public comment period, including a public hearing.
The first set of changes, effective March 10, 1995, was the
most controversial. Hypotonic-hyporesponsive episode (HHE)
and residual seizure disorder were removed under DTP vac-
cines, and the definitions of encephalopathy and residual sei-
zure disorder were modified in the QAIs.45
For the most part, DTP injury claims were filed on behalf of
children and adults with chronic encephalopathy of unknown
cause manifesting as developmental delay or the onset of sei-
zures during the first year or two of life. Overall, no specific
cause is ever determined for as many as 40% of the cases, with
most thought to be due to migrational abnormalities of fetal
brain development or metabolic or “genetic” conditions not
identifiable by current technology.82 For petitioners to success-
fully pursue a VIT claim, the designated condition and time
frame must be satisfied, and, if so, there must not be evidence
of a factor unrelated to the vaccine, which has to be of known
cause and not idiopathic.
As for seizures, nearly half of the DTP filings involved the
initial onset of seizures in an otherwise healthy infant. Significant
 1494
National Vaccine Injury Compensation Program (VICP) Summary of Modifications to the Vaccine Injury Table and Qualifications and Aids to Interpretation

SECTION FIVE Public health and regulatory issues

Legal issues 77 1495
 1496
SECTION FIVE
National Vaccine Injury Compensation Program (VICP) Summary of Modifications to the Vaccine Injury Table and Qualifications and Aids to Interpretation—cont'd

* Time intervals for immunocompetent/immunodeficient individuals who receive OPV. Contact cases have no time limit.
†
In accordance with section 2114 of the National Childhood Vaccine Injury Act of 1986, as amended, the Secretary must add vaccines recommended by the Centers for Disease Control and Prevention for routine administration to
children to the Vaccine Injury Table (Table) as covered vaccines with an effective date of coverage of the effective date of the vaccines’ corresponding excise tax. Sometimes, such vaccines are added to the New Vaccines category of
the Table upon publication of a notice before they are added as separate categories on the Table. Also, note that a new vaccine is covered by the VICP or when a new injury/condition is added to the Table, individuals must file claims
that do not meet the VICP general filing deadlines within 2 years from the date the vaccine or injury/condition is added to the Table for injuries or deaths that occurred up to 8 years before the Table change.
††
No condition has been identified requiring inclusion on the Table, therefore compensation for alleged injuries must be pursued on a causation in fact basis.
The following parenthetical abbreviations found within the Federal Register Citation column (FR), (N), and (IFR) indicate final rule, notice, and interim final rule, respectively.

numbers of seizure claims were being compensated using the stat-
utory version of the VIT originally put into law. Often the fever
commonly associated with DTP vaccine would serve as the trig-
gering event for the onset of seizures in a child predestined to
have epilepsy. Or a Table seizure condition would be found in a
child with cryptogenic (idiopathic) infantile spasms if the onset of
myoclonic seizures was within 3 days of DTP vaccination. This
resulted in compensated cases despite controlled epidemiologic
studies showing the condition to be non–vaccine-related.78,79
Claims alleging death due to DTP vaccine also proved chal-
lenging in the program's early years. Approximately half were
attributed to sudden infant death syndrome (SIDS), with com-
patible histories and forensic findings in accordance with the
1989 National Institutes of Health consensus definition.83
However, because the cause of SIDS remains unknown, these
cases were viewed as “idiopathic” by the court and could not
be used by the government to prove that a factor unrelated
to the vaccine caused the death. Furthermore, based on testi-
mony concerning events preceding the death, the Court some-
times concluded that an encephalopathy or HHE was present
before death and awarded compensation. Recently, however, the
Federal Circuit has said that special masters may consider evi-
dence of SIDS in weighing the strength of the petitioner's case.84
The only exception in using the IOM conclusions was
encephalopathy/encephalitis under DTP vaccine, which had
been proposed for removal but was left on the Table in response
to advice from the ACCV. The ACCV argued that claims of acute
encephalopathy of unknown etiology within 3 days of DTP vac-
cination should continue to receive a presumption of causation
but that the definition in the QAIs needed to be more clinically
precise. A subsequent 1994 analysis by the IOM85 of a 10-year
follow-up to the British National Childhood Encephalopathy
Study tried to answer, but fell short of answering, the ultimate
question of whether DTP vaccine causes permanent brain dam-
age,86 despite the fact that many investigators over time had
failed to find a link based on epidemiologic and other scientific
evidence.87–93 The changes to the VIT and the QAIs for DTP
have proved controversial and have been a topic addressed at
congressional oversight hearings.
The second set of changes to the VIT proved far less contro-
versial and was based, in large part, on the 1994 IOM report
covering the five remaining original VICP vaccines, as well as
Hib, HBV, and VZV vaccines, which were also added to the
VICP because all were recommended by the CDC for routine
administration to children. The other modifications, effective
March 24, 1997, included the addition to the VIT of thrombo-
cytopenia for measles-containing vaccines and brachial neuritis
for tetanus-containing vaccines, the removal of residual seizure
disorder under MMR vaccines, and the creation of a provisional
or placeholder category on the VIT for vaccines newly taxed
and satisfying the CDC recommendation requirement.94 Only
after notice to the public on publication in the Federal Register
would the newly added vaccine have a separate and distinct list-
ing on the VIT, including the addition of a Table condition when
appropriate.
As with the 1995 rulemaking, proposed VIT changes devel-
oped by the Secretary paralleled the IOM conclusions for or
against causation with two exceptions: the Secretary did not
add GBS following OPV and following tetanus-containing
vaccines. The IOM's OPV conclusion was based largely on a
Finnish study following a national OPV campaign.95 A subse-
quent US study (therefore not considered by the IOM) showed
no evidence of an increase in GBS following OPV administra-
tion.96 Further doubt was cast when one of the Finnish study's
coauthors wrote a letter noting that it was not their intention
to claim that OPV was causally related to GBS because the data
could be interpreted more than one way.97
The question of GBS and tetanus-containing vaccines was
even more difficult to decide. The IOM conclusion was based
Legal issues 77 1497
on case reports, particularly of one person who experienced
three episodes of GBS, each within weeks following tetanus
vaccination. The fact that he had other non–vaccine-related
episodes made him immunologically unique. Population stud-
ies, in contrast, showed no evidence that GBS incidence is
higher in people receiving tetanus vaccine compared with the
background rate.80,81 Because there was no evidence of increased
incidence overall, it was thought that petitions based on GBS
should continue to require proof of causation, and, therefore,
GBS should not be added to the VIT. However, persons who
experience more than one episode of GBS temporally related to
immunization will no doubt have strong causation arguments
under the VICP. In fact, the program has compensated a claim
involving two episodes of GBS, both within weeks of receiving
tetanus-containing vaccines at ages 5 and 15 years.98
Since 1997, the VIT has been modified further (Table 77-5).
The general category of rotavirus vaccines was added effective
October 21, 1998,99 following licensure of Rotashield, a tetravalent
rhesus-based rotavirus vaccine. Rotashield was subsequently with-
drawn from the market by the manufacturer in October 1999,
after epidemiologic studies confirmed an association between the
vaccine and cases of intussusception, a potentially life-threatening
bowel obstruction that occurs in infants.100–102 In July 2002, a final
rule was published adding intussusception as a listed injury to the
VIT under a second category of rotavirus vaccines (ie, live, oral,
rhesus-based).103 In addition, pneumococcal conjugate vaccines
were made a separate and distinct category on the VIT. The sec-
ond category of rotavirus vaccines (i.e., live, oral, rhesus-based),
with the associated injury of intussusception, was removed from
the Vaccine Injury Table, effective November 10, 2008.107 This
was a technical change since the 3-year filing deadline for all such
claims had long since passed given that the Rotashield vaccine
was withdrawn from distribution in the United States in 1999.
The general category of rotavirus vaccines with “no condition
specified” was not affected.
Lastly, beginning in 2004, hepatitis A and trivalent influenza,
meningococcal (conjugate and polysaccharide) and human papil-
lomavirus (HPV) vaccines were added to the VIT in the place-
holder category for “new” vaccines recommended by CDC for
routine administration to children.104–106 However, the vaccines
were subsequently moved to separate and distinct categories in
the VIT, effective July 22, 2011.106a Finally, the 2011 IOM review
reported the results of its reviews concerning the evidence with
respect to casual associations between 12 VICP-covered vaccines
(in various combination) and 158 adverse events. This review
is likely to lead to additional changes to the VIT, including the
addition of new injuries.72a
Medical review of claims
Nearly all claims filed under the VICP had some clinical out-
come in temporal relation to vaccination, varying from the nor-
mal, expected side effects of crying, fever, and local swelling to
much more serious acute and chronic illness. The claims rep-
resent a database of possible vaccine-related events, although
only a small percentage of serious outcomes was thought to be
caused by vaccines after VICP medical staff review. Some of the
more relevant clinical diagnoses are reviewed here.
Diphtheria, tetanus, and pertussis vaccines
Thousands of petitions filed for DTP vaccine for the most part
focused on outcomes after the primary series of immunization
in children younger than 12 months. Just over half of the injury
claims involved the initial onset of seizures in various intervals
following vaccination, ranging from hours to several weeks or
longer. Claims involving idiopathic epilepsy (31%) were the larg-
est group, followed by claims involving infantile spasms (14%)
1498 SECTION FIVE Public health and regulatory issues
and a small percentage with different seizure types (eg, absence,
complex partial, psychomotor). Another significant category was
the 20% of claimants who experienced developmental delay as
their initial neurologic sign, with medical records rarely showing
any significant effects following vaccination. However, parents
pointed to prolonged, inconsolable crying or extreme lethargy
as a basis for assuming DTP-related outcomes. Only 11% of
children demonstrated clinical signs of encephalopathy and, of
those children, approximately half had an unknown etiology.
The remaining claims comprised metabolic or genetic disorders
and a variety of other non–vaccine-related conditions. Only a
handful of HHE cases were identified on medical records review.
Approximately 50% of claims alleging DTP-related death
had medical records and autopsy findings consistent with SIDS.
The next most frequent diagnostic category for death cases
involved the 10% of patients with long-standing convulsive dis-
orders (patients who had severe, long-term seizures), followed
by patients with acute encephalopathy (9% [4% with known
etiology and 5% with unknown etiology]) and developmental
delay onset (7%); the remaining claims showed various non–
vaccine-related conditions (eg, choking, sepsis, congenital heart
disease). Four deaths were due to anaphylaxis.
Just over 400 DTaP claims had been filed by January 2011,
which make up 17% of filings during the past 5 years and 6%
of claims overall. Similar types of neurologic conditions have
been alleged to be caused by the acellular pertussis vaccine,
but in far smaller numbers than the whole-cell DTP product.
Postmarketing surveillance of serious adverse events following
DTaP continues to confirm the better safety profile vs the whole
cell product.108,109 Tdap claims, on the other hand, comprised
3% during the same 5-year period, with alleged injuries simi-
lar to those seen with other tetanus-containing vaccine claims.
Tetanus-containing vaccines
Claims involving tetanus-containing vaccine not includ-
ing a pertussis component (ie, DT, Td, or TT) comprise 3%
of filings overall. Brachial neuritis, a condition listed on the
VIT, is the most frequently alleged or documented condition.
After that come GBS, chronic inflammatory demyelinating
polyneuropathy, multiple sclerosis, acute encephalopathy of
unknown etiology, and other central and peripheral nervous sys-
tem disorders. Nearly all are adult vaccine recipients. Litigating
GBS claims has been particularly challenging. Few were decided
before the IOM report concluding that tetanus-containing vac-
cines could cause GBS if the onset was within 5 days to 6 weeks
following immunization.80,81 In one case in which compensa-
tion was not provided, the court found that, although Td can
cause GBS, one could not conclude that it did cause GBS in any
particular case. The decision was appealed but was affirmed by
the US Court of Appeals for the Federal Circuit.110 Since this
decision, however, the court has compensated some GBS claims
based on the IOM report. Other claims have been compensated
on the basis of a litigative risk settlement.
MMR vaccines
Overall, MMR vaccine is implicated in 13% of claims, most of
which are events associated with immunization during the sec-
ond year of life. Natural measles infection is known to cause
acute encephalopathy/encephalitis. Although the IOM found
the evidence to be “insufficient” to determine whether measles
vaccine can cause acute encephalopathy, there is some evidence
of clustering of encephalopathy/encephalitis cases whose onset
is 8 to 10 days following measles-containing immunization,
based on an analysis of 48 VICP claims.111 Only rarely are bio-
logical markers demonstrated to support a genetic, metabolic,
or vaccine cause. In only one VICP case was cerebrospinal fluid
measles antibody, a biological marker for measles causation,
found in a previously healthy 2 year old girl who received the
first dose of MMR and varicella vaccine.
New-onset seizures following MMR vaccine was alleged in
some claims. Nearly all were febrile seizures occurring in the
7 to 14 day time frame during which fever may accompany
replication of vaccine virus. The cases that fell within the VIT
onset interval were usually recommended for entitlement by
VICP staff. Most such claims were adjudicated before residual
seizure disorder was removed from the VIT.
Another common diagnosis was thrombocytopenic purpura,
a VIT condition with prescribed onset of 7 to 35 days following
vaccination. Most involved children with documented throm-
bocytopenia within this interval. In some cases, additional epi-
sodes of decreased platelets occurred in the ensuing months or
years. Many, however, showed normalization of their platelet
count before the 6 months of continued effects required by the
NCVIA.
Subacute sclerosing panencephalitis (SSPE) claims have also
been identified, more than half of them deaths. Early hearings
resulted in a few being compensated because of the initial lack
of administrative resources to defend claims. Subsequent tes-
timony by experts in infectious disease persuaded the court of
the lack of any data supporting a causal connection, noting the
tremendous decrease in SSPE incidence since licensure of the
vaccine decades ago. Since then, claims of SSPE-related injury
or death have been dismissed.
Claims alleging injury from rubella vaccine comprise 2% of
claims overall. Most are filed on behalf of adult women who
received vaccine in the postpartum period or for employment. In
the majority of cases, acute and chronic arthritis or arthralgia devel-
oped 7 to 90 days (median, 14 days) after rubella immunization.
Neuropathy and fibromyalgia were the other diagnoses. A previ-
ous inventory of 113 nonacute arthritis claims filed starting in
1988 showed 28% with chronic arthropathy or other generalized
complaints similar to those of chronic fatigue syndrome and 30%
with a variety of non–vaccine-related conditions.
Although 13% to 15% of susceptible (serology-negative)
women may experience transient arthritis after the currently
used rubella vaccine, with a higher percentage (up to 40%)
reporting some type of musculoskeletal complaint, it is less
clear what role, if any, the vaccine has in causing recurrent or
chronic arthropathy (ie, arthralgia or arthritis).78,79
Based on the 1991 IOM finding that chronic arthritis may
be caused by rubella vaccine, a special master held a hearing to
determine criteria necessary for proof of causation. Following
the special master's publication of these guidelines, the VICP
added chronic arthritis to the VIT in the 1995 final rule45 and
incorporated several of the court's criteria in the 1997 final
rule.94 The main difference between the criteria adopted by
the VICP and those adopted by the court has been the latter's
willingness to compensate arthralgia, a subjective symptom
that is difficult to assess, unlike observable signs of arthritis.
Adjudications to date have resulted in 70 claims being compen-
sated and 123 being dismissed. Additional research published
since the IOM report, including retrospective case reviews and
a prospective double-blind study, has produced mixed results. If
rubella vaccine does cause chronic arthropathy, it would seem
that the incidence is rare.112–114
Varicella vaccine
Varicella vaccine, which was licensed in March 1995, became a
covered vaccine in August 1997, after enactment of the excise
tax. By 2006, the VICP had received at least 62 claims in which
varicella vaccine was administered alone (17) or concomitantly
with MMR (12), MMR and other vaccines (24), and other vac-
cines (9). The group was composed of 29 females and 33 males;
48 were children between 12 and 18 months of age, 8 were
older children, and 6 were adults. The onset interval following

vaccination with varicella vaccine alone ranged from 12 hours
to 27 days. Neurologic conditions were the predominant injury
category (eg, ataxia, seizures, transverse myelitis, and hemipa-
resis), followed by hematologic (eg, chronic thrombocytopenia),
endocrine (eg, diabetes mellitus), skin (eg, scarring, erythema
multiforme major, vasculitis), and other conditions. Various
neurologic disorders have been reported to be associated with
primary VZV (wild-type) infections, including meningoenceph-
alitis, encephalitis, transverse myelitis, and acute cerebellar
ataxia.115 Vasculitis and acute cerebellar ataxia also have been
reported following receipt of varicella vaccine.116 All of these
conditions were reflected in the claims filed, with an onset
interval ranging from 4 to 46 days following vaccination. The
vaccine-strain genotype by polymerase chain reaction was iden-
tified in skin lesions of two persons.117 Since 2006, 21 MMRV
and 2 VZV claims have been filed, reflecting a shift to the com-
bination product following licensure in 2005.118
Because there is no VIT condition listed for varicella vac-
cine, petitioners can be eligible for compensation only by prov-
ing causation or that the vaccine significantly aggravated a
preexisting condition. Some claims have been compensated if
the nature of the clinical manifestations and the onset inter-
val following vaccination mirrored what has been reported with
wild-type (natural) VZV infection.
HBV vaccine
Almost 700 HBV vaccine claims have been filed as of January
2011, which represent 8% of all VICP claims. Medical review
shows the following diagnostic categories: neurologic, includ-
ing encephalopathies and central/peripheral demyelinating
conditions (30%); rheumatologic, including various arthritic
conditions (20%); and autoimmune conditions, including dia-
betes mellitus (20%).The remaining 30% of the HBV vaccine
claims involved various allegations, such as mercury-induced
autism, chronic fatigue, anaphylaxis/allergic, skin, and hema-
tologic conditions. Vaccine recipients range in age from birth to
69 years, with 46% pediatric claims (younger than 18 years) and
23% for children younger than 2 years.
In general, there are no serious adverse events linked to the
recombinant vaccine other than acute anaphylaxis.80,81 In April
2002, the IOM Vaccine Safety Review Committee published its
findings regarding HBV vaccine and neurologic disorders. The
IOM concluded that there is evidence against a causal relation
between HBV vaccine and multiple sclerosis (and relapse) and
insufficient evidence for other central or peripheral nervous sys-
tem demyelinating conditions.57
More than half of the HBV vaccine claims were filed in 1999
when the 2-year window for filing retroactive claims expired.
Working with the DOJ and the two petitioners' counsel who
filed the majority of cases, the Court agreed to group the claims
into diagnostic categories as a means of managing the large
volume. After many years of attempting to convene an expert
panel to deal with the constellation of cases and matrix of vari-
ous injuries, the Court abandoned the concept and the cases
were reassigned in early 2006 to three newly hired special mas-
ters. As a result, the claims are increasingly being assessed indi-
vidually, not necessarily in relation to other claims involving
similar allegations of injury.
In addition to individual hearings, there have been several omni-
bus proceedings using test cases with similar diagnoses to help
move the backlog of claims. A special master in the Hepatitis B–
Neurological Demyelinating Omnibus Proceeding119 ruled that HBV
vaccine can cause transverse myelitis,119 GBS,120 multiple sclero-
sis,121 and chronic inflammatory demyelinating polyneuropathy.122
However, another special master ruled differently,123 consistent
with the government's position based on scientific evidence.124–126
Another consolidated hearing resulted in the special master ruling
that the HBV vaccine did not aggravate petitioner's type 1 diabetes
Legal issues 77 1499
mellitus.127 In another proceeding, the special master ruled that the
petitioners never offered a medical theory causally connecting the
thimerosal contained in HBV vaccine to sudden death.128 Finally,
a split decision on appeal followed the special master's decision in
five cases involving allegations of HBV and autoimmune hepati-
tis. In this proceeding, the same theory was offered and the same
experts testified, but the facts of each of the cases were presented
separately and the special master wrote a separate decision on each
of the cases. The special master ruled in favor of HHS in all five
cases, but only two were affirmed on appeal.129,130 The three cases
reversed were on appeal to the US Court of Appeals for the Federal
Circuit at the time of this writing.131–133
Since 1999, some claims outside the aforementioned group-
ings have been adjudicated with varied results. Aside from ana-
phylaxis or anaphylactic shock, the only condition listed on the
VIT for HBV vaccine, all HBV claims are being adjudicated on
a causation-in-fact basis.
Polio vaccines
Polio vaccine–related claims totaled more than 500 of the pre-
1988 case filings. Most involved vaccines given in the 1950s
and 1960s, when polio was common in the United States.
IPV is a killed virus vaccine and, if properly manufactured,
is not associated with paralytic poliomyelitis. It is reasonable
to expect that some persons acquired natural poliomyelitis in
temporal relationship to receipt of IPV, particularly if only one
or two doses were given. The 1955 Cutter incident, in which an
estimated 260 cases of paralytic poliomyelitis were attributed
to residual live virus in the vaccine, is well documented.134 No
similar instances of killed vaccine–related poliomyelitis were
publicly identified, although there is evidence that at least one
lot of Wyeth vaccine given during this same period had infec-
tive amounts of live virus.135 Of 266 IPV claims, nearly all have
been dismissed by the court, with none being found to be sec-
ondary to Cutter vaccine administration. Only 13 claims are
for vaccines administered since 1979, when the last indigenous
case of wild poliovirus transmission occurred in the United
States.136
OPV claims were approached much differently because para-
lytic polio is known to be a rare complication in vaccine recipi-
ents and in contacts and, therefore, is listed in the VIT. Most
claims arising since the early 1960s with confirmed paralytic
polio were compensated by the program. More than half of the
claims had evidence of poliomyelitis. Naturally, it is difficult to
know which cases were actually vaccine-related when polio was
present in the community. Claims involving other conditions
such as transverse myelitis and GBS have been rejected by the
court based on the lack of proof of causation. Of the 306 claims
filed alleging OPV-related injury (or death), 157 were compen-
sated and 149 were dismissed.
Rotavirus vaccine
Claims alleging intussusception from the rhesus-based
rotavirus vaccine (Rotashield) started appearing in 2000.
Altogether, 31 claims were filed alleging injury, 24 of which
alleged intussusception. Patients with intussusception ranged
in age from 2 to 11 months. Nearly three quarters had onset
within 2 weeks of immunization and about half resulted from
the first dose. Nearly all required surgical reduction. There
was one death.
Some claims filed before the final rule adding intussuscep-
tion as a VIT condition were compensated on a causation-in-
fact basis, a burden that was made easier with epidemiologic
studies confirming an association.100–102 Once epidemiologic
studies confirmed an association between rotavirus vaccine
and intussusception in the first 2 weeks after immunization,
the VICP began the process for revising the VIT. A 2002 final
1500 SECTION FIVE
rule added intussusception to the VIT under a second category
of rotavirus vaccines (ie, live, oral, rhesus-based), with an onset
interval of 0 to 30 days following immunization.103 The wider
interval was chosen to provide a generous presumption of cau-
sation. Once intussusception was added to the VIT, petition-
ers otherwise barred under the general statute of limitations
benefited from a presumption of causation in claims involving
intussusception and had 8 years of retroactive coverage from the
effective date of the VIT change, with a 2-year window in which
to file their claims.
Ironically, the benign outcome of most cases of intussus-
ception proved challenging for the VICP. Until October 2000,
the NCVIA required all claimants to establish that the residual
effects of an injury persisted for more than 6 months after vac-
cination or that a death occurred.
Because most patients with intussusception recover completely
within days, some petitioners might be denied compensation under
that standard. However, the Children's Health Act of 2000 amended
the NCVIA to permit payment of compensation in claims in which
the effects of the injury last less than 6 months if the petitioner
demonstrates that the vaccine-related illness, disability, injury, or
condition “resulted in inpatient hospitalization and surgical inter-
vention”.137 Thus, under current law, infants who experienced intus-
susception following a rotavirus vaccine and did not have residual
effects for more than 6 months qualified for compensation if their
injury resulted in inpatient hospitalization and surgery.
With licensure of two new rotavirus vaccines in 2006 and
2008, the program once again began receiving claims alleging
intussusception, although large clinical trials at that time for
each vaccine had not shown any association.134a,134b By January
2011, 16 injury claims had been filed. Of the four claims
adjudicated, two were dismissed and two received compensa-
tion. At the time of publication, new evidence from post-mar-
keting studies outside the US suggested intussusception was
causally associated with Rotarix and Rotateq, although much
more rarely than was the case with Rotashield.134c,134d
HAV vaccines
HAV vaccines were added to the VICP in 2004. Since then, fewer
than 3% of nonautism claims filed alleged injury from HAV vac-
cine, solely or in conjunction with other vaccines. Petitioners
ranged in age from 1 to 58 years, with about half younger than
18 years. No patterns in claimed injuries emerged on medical
analysis of the cases, which included liver-related conditions
(hepatitis, cholangitis), neurologic disorders (seizure disorder,
acute disseminated encephalomyelitis, transverse myelitis, GBS),
and various immunologic conditions. There is no VIT condition
listed for HAV vaccines, nor is there is any serious adverse event
causally associated with the vaccine, other than acute anaphy-
laxis.138 The majority of cases heard before the Court have been
dismissed or settled based on legal and not scientific issues.
Meningococcal vaccines
Meningococcal polysaccharide and conjugate vaccines were
added to the VIT on February 1, 2007. Fewer than 1% of claims
filed since then have alleged injury from these vaccines, and
none involved the polysaccharide product. More than half of
the claimants had at least one other vaccine coadministered.
The great majority (70%) of MCV4 claims involved demyelin-
ating conditions, with GBS found in nearly 40% of claims after
medical review and more than half of claims alleging neurologic
injury. Other demyelinating disorders found were transverse
myelitis, chronic inflammatory demyelinating polyneuropathy,
and multiple sclerosis.
No specific injury related to meningococcal vaccines is listed
in the VIT, and most meningococcal vaccine claims have yet to
be adjudicated. A report by the CDC in 2006 suggested a small
increase in the risk of GBS among meningococcal conjugate
vaccine recipients.139 However, two large unpublished postli-
censure safety studies presented to the Advisory Committee
on Immunization Practices in June 2010 found no evidence of
increased GBS risk following vaccination.140
Influenza vaccines
Trivalent influenza vaccines (TIV, LAIV) were added to the
VICP in July 2005. Just over 200 claims were filed by July 2007
at the 2-year deadline for filing claims dating back 8 years when-
ever a new vaccine is added to the VIT. Since then, nearly half
of the total VICP claims filed allege influenza vaccine injury,
with increasing numbers being filed each year. Moreover, since
2007, nearly 60% of VICP claims filed are for adults, which is a
decided shift from the first 15 years of the program when most
claims were filed on behalf of children.
As of January 2011, 590 petitions had been filed claiming an
injury caused by an influenza vaccine. Only 12% were for peti-
tioners younger than 18 years. GBS comprised 55% of the claims.
In fact, claims alleging GBS are now the most common injury/
vaccine claim filed annually. Other demyelinating conditions
made up 15% of influenza claims, which included transverse
myelitis, chronic inflammatory demyelinating polyneuropathy,
and acute disseminated encephalomyelitis. The remaining 30%
involved various allegations, such as encephalopathy, seizures,
brachial neuritis, and a range of neuromuscular conditions.
Eight claims involved a shoulder injury related to vacci-
nation, an injury apparently from the unintentional injec-
tion of antigenic material into synovial tissues resulting in an
immune-mediated inflammatory reaction. A case series of 13
VICP cases is described in a 2010 report.141 Most claims were
conceded by the HHS.
No specific injury related to influenza vaccine is listed in the
VIT, and the HHS has only concluded that a small number of
influenza claims demonstrated injury causation by a prepon-
derance. Such claims include anaphylaxis,142 a syncopal episode
resulting in a motor vehicle accident with severe injuries, and
shoulder injury related to vaccine administration. However, it
is not uncommon for influenza claims to receive compensation
on the basis of a litigative risk settlement because of the evolved
causation standard.
A new challenge faced the VICP after increased rates of GBS
were detected following the 2009-2010 monovalent influenza pan-
demic campaign. A separate program within HHS is responsible for
alleged injuries or deaths related to use of the monovalent product
(see the discussion of the Countermeasures Injury Compensation
Program [CICP] in “Ongoing challenges”]. However, with incor-
poration of the H1N1 pandemic virus into the 2010-2011 triva-
lent vaccine, GBS claims are to be expected, perhaps in significant
numbers. Adjudication outcomes for these claims will no doubt
depend in part on final analysis of the active surveillance data from
the H1N1 monovalent pandemic vaccine.143
HPV vaccines
HPV vaccines were added to the VICP in February 2007.
By the end of 2010, 117 claims had been filed, all of which
involved adolescent girls or young women, except for one
middle-aged man. The majority of injury claims alleged neu-
rologic injuries, with medical review confirming 13 claims
with GBS, 4 with transverse myelitis, and 3 with acute dis-
seminated encephalomyelitis. Other neurologic categories
included headaches and seizures. The second largest group
(25%) was rheumatologic, with illnesses, such as rheuma-
toid arthritis, systemic lupus erythematosus, fibromyalgia,
and undifferentiated connective tissue disease. Other diag-
nostic categories included gastrointestinal, hematologic, and
endocrinologic conditions. Five cases involved syncope with

secondary trauma. Four of these cases resulted in compen-
sation (through a concession of entitlement by HHS or set-
tlement between the parties). One claim involved an injury
from the injection process or shoulder injury related to vac-
cine administration and was conceded.141
More than 20% of claims had mental health issues that con-
tributed to the alleged injury or significantly influenced the con-
tinued illness. Affective disorder was the most common mental
health illness that predated the vaccination or was diagnosed or
became significantly worse in the weeks to months after vac-
cination, particularly with the HPV series administered over
6 months. Given the adolescent population, initial manifesta-
tion or diagnosis of mental health disorders was not unexpected.
There were eight claims of vaccine-related death. All were
females, ages 13 to 21 years. All had autopsies, with seven per-
formed by medical examiners. Medical review with academic
pathology review of autopsy material found (convincing) evi-
dence of a cause of death unrelated to the vaccine in six cases.
There is no VIT condition listed for HPV vaccines, nor is
there is any serious adverse event causally associated with the
vaccine, other than acute anaphylaxis.144 Only a small number
of claims had been adjudicated by the time of this writing.
Ongoing challenges
During the last decade, the greatest challenges to the VICP have
been the autism-related litigation and adjudicating causation-in-fact
claims. Both depend heavily on the availability of credible and
timely scientific evidence. Only through a comprehensive vaccine
adverse event surveillance and research program can true vs coinci-
dental reactions following vaccination be distinguished. Rotavirus
vaccine and intussusception is an example of how epidemiologic
research led to expedited compensation once intussusception was
found to be vaccine-related. At the same time, safety studies allow
for the debunking of erroneous theories of vaccine causation, assist-
ing the court in dismissing claims not shown to be vaccine-related.
The creation of a special IOM committee in 2001 to perform inde-
pendent, expedited scientific reviews of current and emerging vac-
cine safety hypotheses (see discussion of the IOM Immunization
Safety Review Committee in Chapter 76, Vaccine Safety) was an
important step in ensuring continued evaluation of the safety of
vaccines. Eight reports issued from 2002 through 2004 have been
helpful to the VICP and Court in clarifying the current scientific
literature on vaccine causation. At the same time, securing ade-
quate funding for vaccine safety research overall remains one of the
greatest challenges at a time of budget deficits and competing pri-
orities for vaccine program funding. However, a new IOM com-
mittee reviewed the medical and scientific literature to assess the
causal relationship, if any, between 78 adverse events (158 adverse
event-vaccine combinations) and 12 VICP-covered vaccines. The
consensus report was released in August 2011.145 The report, which
is greater in scope than the sentinel reports of the 1990s, may result
in additional modifications to the VIT by the Secretary.
Criticism by petitioners and their attorneys led to over-
sight by Congress in the VICP's early years. Government
Accountability Office (GAO) reports issued in 1999 and 2000
addressed the VICP process and trust fund, respectively.146,147
The VICP adjudication process was judged to be easier than
the traditional tort system, but not as streamlined as Congress
had originally intended. Moreover, although there seemed to be
a scientific basis for the VIT changes made by HHS, the GAO
found some inconsistencies in applying results of the IOM
reviews. The trust fund and its huge, ever-increasing size was
the focus of the second effort by GAO. Potential solutions varied
depending on stakeholder perspective, with petitioners wanting
more claims compensated by decreasing the burden of proof,
vaccine companies wanting the tax reduced to lessen cost, and
Legal issues 77 1501
researchers calling for a portion of incoming revenues annually
to pay for vaccine safety research and surveillance. In the end,
the GAO's only recommendation was that HHS publish a clear
method for future changes to the VIT to help ensure that such
changes are perceived as fair.
Also in 2000, a subcommittee of the House Government
Reform Committee issued its own report on the VICP.148 The
bipartisan report recommended the following: (1) a review of the
VIT to ensure the inclusion of current science, (2) the increased
use of speedy informal dispute resolution, and (3) the develop-
ment of an alternative standard for non-VIT cases in response
to concerns that the burden of proof for proving causation under
the NCVIA was too difficult to meet. It is the last recommenda-
tion that is proving the most challenging.
A proposal by the American Academy of Pediatrics to use a
more relaxed burden-of-proof standard for adjudicating non-VIT
injuries was reviewed by the ACCV in 2001.149 No recom-
mendation was forthcoming, nor has language addressing the
causation standard been included in proposed congressional
amendments to the VICP during the past decade. For the most
part, bills included process improvements, changes to the stat-
ute of limitations for injury and death claims, and language
aimed at reducing the filing of lawsuits in the civil system with-
out first exhausting remedies within the VICP.150,151
Concerns that helped lead to the creation of the VICP (eg,
fears regarding liability and the availability of compensation in
relation to adverse events resulting from products) have sur-
faced with regard to vaccines not covered by the VICP (because
they are not routinely administered to children) and to other
countermeasures used to prevent and treat particular diseases.
Some interested parties have suggested adding more
selective-use vaccines to the VICP, such as pneumococcal poly-
saccharide, used primarily in adults. An NVAC review in 1996
found little evidence at that time to support the need, based on
a lack of liability concerns by manufacturers and health care
providers.152 More recently, others have pointed to the need to
cover clinical trials in pregnant women for vaccines to prevent
respiratory syncytial virus, cytomegalovirus, and group B strep-
tococcal infections, both of which cause significant morbidity
and mortality in very young infants.153 Citing liability con-
cerns as a key barrier to licensure of new vaccines for mater-
nal immunization, the American College of Obstetricians and
Gynecologists was spearheading efforts in 2005 to add this cat-
egory for coverage under the VICP.154
Despite the successful global eradication of smallpox in
1980, terrorist attacks in 2001 led to concerns that terrorists
might have access to the smallpox virus and might attempt to
use it against Americans. As part of President Bush's plan to
protect the American population from the threat of a smallpox
attack, liability protections were enacted to protect covered per-
sons in relation to their involvement in the federal smallpox
vaccination campaign. Covered persons included manufactur-
ers of smallpox vaccines, health care entities, and health care
providers under whose auspices the vaccine was administered
and state and local governments, including first responders.
The liability protections, which were enacted through sec-
tion 304 of the Homeland Security Act of 2002 and section 3
of the Smallpox Emergency Personnel Protection Act of 2003
(SEPPA),155,156 were meant to ensure that the smallpox vaccine
and other related smallpox countermeasures were available, if
necessary, to protect the public health.
Under the liability provisions, no lawsuit for liability for inju-
ries or deaths arising out of the administration of a smallpox
vaccine or other smallpox-covered countermeasure named in
a declaration could be maintained against covered persons.
Instead, such claims had to be pursued against the United
States as a substitute defendant. Such claims were to proceed
under the Federal Tort Claims Act, which requires plaintiffs to
demonstrate negligence or another cognizable tort. Although
1502 SECTION FIVE Public health and regulatory issues
these liability protections were broad, they were subject to
certain limitations. For example, if a covered person failed to
cooperate with the government in the defense of a claim, the
government would not be liable for any damages resulting from
that person's act or omission. Also, if the United States made a
payment on a claim that was based on bad acts or omissions by
a covered person (eg, gross negligence), the United States was
entitled to recover funds from the covered person.
These liability protections were triggered in January 2003,
when the Secretary issued a declaration in which he concluded
that a potential bioterrorist incident made advisable the admin-
istration of particular smallpox countermeasures (including
smallpox vaccines, products designed to prevent or treat
smallpox, and products designed to control or treat adverse
events resulting from smallpox vaccination).157 The Secretary
recommended that certain categories of persons (including first
responders and members of state, local, and HHS-approved
smallpox response teams), receive the covered countermeasures.
As a companion to these liability protections and as incentive
for the targeted persons to receive smallpox countermeasures,
the SEPPA authorized the Secretary to establish the SVICP to
provide medical, lost employment income, and death benefits to
eligible persons (including smallpox vaccine recipients; persons
who came into contact with a vaccinated person or with another
person with whom the vaccinated person had contact, known as
“vaccinia contacts”; and the estates and survivors of otherwise
eligible deceased persons).158 Among the eligibility requirements,
requesters had to demonstrate an injury that resulted from a
covered countermeasure. Like the VICP, requesters could dem-
onstrate this by showing that the injured person sustained an
injury included in an injury table promulgated by HHS within
the table time frame or by demonstrating that the person's injury
was actually caused by the vaccine or other covered countermea-
sure. Unlike the VICP (in which claims are litigated in the US
Court of Federal Claims), the SVICP was a purely administra-
tive program operated by the HHS. Persons eligible to file SVICP
claims had to exhaust their remedies with this program before
filing any action against the United States. We are not aware of
any post-SVICP claims filed against the United States. The regu-
lations governing the SVICP, including the tables of injuries, are
available at 42 CFR part 102. A total of 64 persons applied for
benefits with the SVICP, and 20 persons received compensation.
The SVICP is no longer operational and was terminated in 2008
owing to a lack of funding, among other factors. The smallpox
vaccine is now a covered countermeasure under the CICP.
More recently, concerns about a potential pandemic and the
possibility that the threat of liability would inhibit manufactur-
ers from producing vaccines and other countermeasures needed
to prevent or treat a pandemic or other emergency event (and
would inhibit health care professionals and others from distrib-
uting and administering such countermeasures) led to the enact-
ment of the Public Readiness and Emergency Preparedness Act
(PREP Act), within the Department of Defense Appropriations
Act, in 2005.159 The PREP Act offers extremely broad liability
protections to manufacturers and other persons and entities
(eg, distributors, program planners, vaccine administrators) for
all claims of loss resulting from the creation, distribution, and
administration of certain products if the Secretary issues a dec-
laration determining that a disease or other health condition
or threat constitutes a public health emergency or that there is
a credible risk that the disease, condition, or threat may in the
future constitute such an emergency and recommending that
particular countermeasures should be administered or used. In
such a declaration, the Secretary must explain his or her find-
ings and the parameters of the attendant liability protections,
including the category of diseases or conditions for which he
or she recommends the administration of the covered coun-
termeasure, the applicable periods, the pertinent population or
populations, and the applicable geographic areas.
A very narrow exception to these liability protections would
require a party to prove that an otherwise covered party engaged
in willful misconduct and that such misconduct caused death
or a serious physical injury. “Willful misconduct” is defined as
an act or omission taken intentionally to achieve a wrongful
purpose, knowingly without factual or legal justification, and in
disregard of a known or obvious risk that is so great as to make
it highly probable that the harm will outweigh the benefit.160
The PREP Act provides that this standard is more stringent
than any form of negligence or recklessness. Other defenses are
also available to covered persons.
The Secretary invoked the liability protections of the PREP Act
through a series of declarations published in 2008 and 2009 and
amended thereafter.161 These declarations extend to several coun-
termeasures meant to prevent or treat pandemic influenza and
other unrelated countermeasures (those relating to anthrax, botu-
lism, smallpox, and acute radiation syndrome). Although mono-
valent pandemic influenza vaccines are covered by one of these
declarations, trivalent influenza vaccines administered annu-
ally are excluded (and are subject to the NCVIA and the VICP).
Although several vaccines were included among the covered coun-
termeasures listed in the declarations, a broad variety of other
countermeasures (eg, antibiotics, antivirals) were also covered.
The PREP Act provides for the establishment of a program
to provide compensation to certain covered persons who sus-
tain serious physical injuries (or death) as a result of a covered
countermeasure designated under a declaration of the Secretary.
This program, which was established in 2010, is known as the
CICP. The structure of and benefits available under this pro-
gram are very similar to the SVICP. Like the SVICP, the CICP
is an exclusively administrative program run by the HHS. In
addition, requesters for benefits must satisfy the requirements
for a condition listed on a countermeasures injury table or prove
that a covered countermeasure caused a specific injury. Under
the CICP, the Secretary must rely on compelling, valid, reliable
medical and scientific evidence to add injuries to an injury table
or to determine that a countermeasure caused a specific injury.
The regulations governing the CICP, including any tables of
injuries, are codified at 42 CFR part 110.162 As of January 2011,
423 requests had been filed with the CICP, nearly all alleging
injury from the 2009-2010 H1N1 pandemic influenza vaccine.
The remaining 2% involved antiviral drugs, a respiratory pro-
tective device, anthrax vaccine, and smallpox vaccine.
Vaccine liability since 1986
The passage of the NCVIA in 1986 did not mark an immedi-
ate decline in the number of reported vaccine liability decisions
in state and federal courts (largely owing to the length of time
until cases reach a final, reported resolution and the fact that
retrospective claimants had until January 1991 to file claims
with the VICP). However, since 1992, the number of reported
vaccine liability cases has declined sharply. This decrease in
reported cases suggests that most VICP petitioners have chosen
not to pursue traditional tort litigation after the VICP process.
There are few reported federal and state court cases in
which a claimant filed a traditional tort claim against a vaccine
manufacturer or a vaccine administrator after being denied com-
pensation under the VICP. In many of these cases, the courts
concluded that the claims must be dismissed as barred by the
applicable statute of limitations (eg, state statute of limitations)
because the underlying VICP claim was time-barred or because
the plaintiffs failed to file a timely VICP claim within the limita-
tions period.163–170 Other reported post-VICP cases address other
issues. In one case, a state court dismissed a medical malpractice
action filed against a vaccine administrator by a pre-1988 peti-
tioner because of the petitioner's acceptance of a VICP award.171

An informal survey of vaccine companies revealed fewer than
a dozen nonautism lawsuits were filed for all VICP-covered vac-
cines from 2005 through 2010, mirroring the downward trend
in civil litigation for pertussis-containing vaccines since VICP
inception (Figure 77-1).172 In addition, the number of lawsuits
filed alleging autism had decreased significantly by January
2011 (Daniel Thomasch, Orrick Herrington and Sutcliffe;
personal communication, June 2011).
As a general rule, as long as petitioners properly exhaust
their remedies under the VICP, they may pursue civil actions
against vaccine administrators and manufacturers. The NCVIA
imposes certain limitations on post-VICP claims that may be
filed against vaccine manufacturers. No such protections exist
for vaccine administrators. Until recent years, these post-VICP
limitations on civil actions filed against vaccine manufacturers
attracted little attention.
However, one such limitation was recently brought into the
spotlight as different interpretations of its language led to a
post-VICP case, concerning a seizure disorder alleged to be the
result of a DTP vaccine, that was recently decided by the US
173
Supreme Court, The NCVIA provides
that “[n]o vaccine manufacturer shall be liable in a civil action
for damages arising from a vaccine-related injury or death
associated with the administration of a vaccine after the effec-
tive date of this part if the injury or death resulted from side
effects that were unavoidable even though the vaccine was prop-
erly prepared and was accompanied by proper directions and
warnings”.174 The issue before the Supreme Court was whether,
in post-VICP litigation, this provision prohibits claimants from
bringing “design defect” claims against manufacturers in civil
court. In these claims, the plaintiff alleges that the design of the
vaccine was defective; a safer alternative design was available;
the manufacturer failed to design its vaccine in conformity with
the safer alternative; and as a result, the claimant was injured.
Before the Supreme Court's decision, lower courts had dif-
fered in their interpretations of this language. The Supreme
Court of Georgia concluded in
(a thimerosal-neurologic damage claim) that design
defect claims could proceed on a case-by-case basis until a
court determined that the particular vaccine was unavoid-
ably unsafe.175 The US Court of Appeals for the Third Circuit
Nonautism DTP/DTaP lawsuits filed against US companies.
Legal issues 77 1503
176
reached a different result, in In that
case, family members filed a petition with the VICP arguing
that their child sustained a seizure disorder as the result of a
DTP vaccine. After their VICP claim was dismissed, the family
filed a design defect claim against the vaccine manufacturer in
civil court. On appeal, the Third Circuit dismissed their claim,
concluding that the statutory language preempted all post-VICP
design defect claims against vaccine manufacturers. The family
sought certiorari with the US Supreme Court.
Given the split in judicial interpretations and the significance
of the issue, the Supreme Court granted certiorari to clarify
whether this provision of the NCVIA preempts all design defect
claims in post-VICP litigation filed against vaccine manufacturers.
Many stakeholders filed amicus briefs on both sides, emphasiz-
ing the significance of the Supreme Court's anticipated ruling in
the case.177 The federal government filed an amicus brief arguing
that the NCVIA preempts all design defect claims against vac-
cine manufacturers in post-VICP litigation.178 The government
was given the opportunity to provide its views during oral argu-
ment, in which it emphasized the significant public health impli-
cations of the case and the important role of the government in
ensuring that licensed vaccines are safe and effective.
In an opinion delivered by Justice Scalia, the Court held that
the NCVIA preempts design defect claims against manufac-
turers in civil court.173 The plain text suggests that a vaccine's
design is not open to question in a tort action. In addition, the
structure of the NCVIA reinforces the interpretation that design
defect claims are preempted: The NCVIA provides for compre-
hensive federal agency improvement of vaccine design and for
federally prescribed compensation as an alternative means for
achieving the aims of design defect tort suits.
Beyond the issue of any limitations on design defect claims,
the NCVIA addresses “duty to warn” claims with respect to vac-
cine manufacturers (but not administrators) by providing that
“[n]o vaccine manufacturer shall be liable in a civil action for
damages arising from a vaccine-related injury or death associ-
ated with the administration of a vaccine after the effective date
of this subpart solely due to the manufacturer's failure to pro-
vide direct warnings to the injured party”.179
Before the and cases arose, the largest
development in non-VICP vaccine litigation had been an upsurge
1504 SECTION FIVE
in cases, beginning in 2001, in which parents alleged that cov-
ered vaccines (the MMR vaccine, which never contained the
preservative thimerosal, and/or childhood vaccines containing
thimerosal) caused their children's autism. Plaintiffs argued that
they were not required to exhaust their remedies within the VICP
based on their argument that the preservative thimerosal was an
adulterant. This argument hinged on the NCVIA's definition of
a “vaccine-related injury or death”, which excludes any injury or
death associated with an adulterant or contaminant intention-
ally added to a covered vaccine.180 The federal government's posi-
tion, filed in one of these cases, was that thimerosal was not an
adulterant because it was part of the vaccine formulation when
the products were licensed and approved for use by the FDA.181
Courts have found that such claims must be filed with the VICP
and all remedies there exhausted before any such claim may
be brought outside the program.182,183 Thus, unless a claim fits
within one of the limited statutory exceptions (see “Other situ-
ations in which suits are not barred by the NCVIA”), it must be
filed with the VICP before tort remedies can be pursued.
In recent years, several cases were also filed in state court
in New Jersey alleging that a contaminant in the OPV (simian
virus 40 or SV40) caused cancerous tumors.184 Manufacturers
have argued that the SV40 was not “intentionally added” to the
vaccine and, thus, that jurisdiction should be before the VICP.
These cases are still in litigation.
Risk to health care professionals
Civil actions against health care providers are rare, and few
have been pursued since the VICP was modified to cover vac-
cine administrators because claimants must go through the
VICP before filing most actions against a vaccine administrator.
However, if a claimant were to reject the VICP decision, he or
she might file a tort action against the health care provider on
such grounds as failure to adequately warn, negligent vaccine
administration, or negligent postvaccination care.
It should be noted that the courts routinely reject allegations
regarding failure to warn when the administrator provides a warn-
ing commensurate with the contents of the package insert. The
NCVIA bolsters these state court findings because it presumably
sets the standard regarding the warning for the administrator.
The NCVIA states that “each health care provider who adminis-
ters a vaccine set forth in the Vaccine Injury Table shall provide to
the legal representatives of any child or to any other individual to
whom such provider intends to administer such vaccine a copy of
the information materials developed [by the Secretary of Health
and Human Services]”.185 The standard of care presumably also
would include a requirement that the provider verify the respon-
sible person's review of the materials and/or that the provider has
shown due care in the exercise of medical judgment in immuniz-
ing the person. Assuming the standard of care is met, health care
providers should face little potential liability.
Other situations in which suits are not barred
by the NCVIA
The NCVIA covers only the losses sustained by the injured per-
son; thus, all loss of companionship claims by the parents, chil-
dren, or spouse of an injured person are not compensable. The
family members may file these so-called derivative claims in
state court, if they can be pursued in court under state law, but
the claims will be successful only if the plaintiff can prove that
the underlying injury or death was related to the vaccine and
prove a theory of liability. The plaintiff may not rely on a finding
under the VICP.186 Up until 2001, with the thimerosal litigation,
families rarely filed such claims. The issue first came to light
187
in The husband and daughter
of a woman who contracted vaccine-related polio filed a claim
for loss of consortium and emotional distress in Massachusetts
State Court after the woman accepted a VICP award. The case
was removed to federal court, and, on appeal, the US Court of
Appeals for the First Circuit found that the Schafers' claim was
not precluded by the acceptance of the VICP award. The Court
of Federal Claims' interpretation of the NCVIA was consistent
with the Schafer holding. In
188
a young man's mother filed a
state wrongful death claim against a health care provider alleg-
ing that the provider negligently left her son unattended in the
bathtub. The young man had a seizure in the tub and drowned.
Ms Abbott filed the state claim in her capacity as mother of the
decedent, and she accepted a settlement from the health care
provider. A few months later, the mother filed a VICP death
claim in her capacity as administrator of her son's estate. Here,
she alleged that the death was a sequela of the vaccine-related
seizure disorder. The claim was dismissed by the special master
because Abbott had already recovered in state court; however,
the US Court of Appeals for the Federal Circuit reversed, hold-
ing that the previous settlement covered only damages for inju-
ries to the young man's beneficiaries. The VICP award would be
for the young man's estate, and Ms Abbott was an appropriate
person to represent the estate. The HHS conceded the case, and
Ms Abbott received a death benefit award.
Other civil litigation related to autism claims has been brought
by plaintiffs. These cases include class action lawsuits filed on
behalf of the parents, children, or spouse of the injured person
seeking their own damages for loss of companionship or consor-
tium or services and loss of earnings, for example. Other law-
suits do not allege specific injuries, but instead demand the costs
of medical monitoring to determine whether thimerosal-related
injuries will develop in the future, generally alleging $1,000 or
less in damages.189 Because the persons in the purported class
action lawsuits are seeking $1,000 or less per person to pay for
“medical monitoring” (ie, future tests to determine if the child is
developing an injury), plaintiffs argue they are not covered by the
NCVIA.190 By May 2006, all medical monitoring lawsuits had
been dismissed (Randy Moss, Wilmer Cutler Pickering Hale and
Dorr; personal communication, February 2007).
Claims alleging that a provider willfully breached his or her
fiduciary duty also may not be covered by the VICP. In
191
parents claimed that a vaccine
administrator fraudulently concealed their child's vaccine-
related injury, and, as a result of this concealment, the time for
filing a VICP claim had passed. The court held that the VICP
does not cover claims for breach of fiduciary duty, so the par-
ents' claim against the administrators could continue in state
court. However, the claim for the underlying vaccine-related
injury must be filed under the VICP.
International compensation programs
Nineteen industrialized countries provide some form of com-
pensation for injuries (or deaths) following vaccination.190a
Germany and France were the first to institute programs in the
1960s, followed by Austria, Denmark, Japan, New Zealand,
Sweden, Switzerland, United Kingdom and Northern Ireland
in the 1970s; Taiwan, Finland, US and Quebec in the 1980s;
Italy, Norway and Republic of Korea in the 1990s; and Hungary,
Iceland and Slovenia since 2000. More than anything else, these
compensation programs were instituted in the belief that govern-
ments have a special responsibility to persons injured by properly
manufactured and administered vaccines used in public health
programs. Most are managed administratively through the
national government, including decisions on eligibility and the
amount of compensation. Eligibility may depend on the recipi-
ent's age, citizenship, or residency status; the category of vac-
cine (eg, recommended, compulsory); the location in which it is
administered (public vs private ambulatory setting); or satisfying

certain time frames for filing a claim. Because few vaccine-related
injuries have a clinical or laboratory marker, proving actual cau-
sation is difficult. Causation decisions are usually based on the
balance-of-probabilities standard of “more likely than not”. All
countries require that the effects be long-lasting (eg, 6 months),
and nearly all provide coverage for medical costs, disability pen-
sions, noneconomic damages (pain and suffering), and death
benefits. No compensation programs outside the United States
reimburse attorneys' fees and costs. Funding is generally from the
national treasury, with some programs receiving support from
lower governmental entities or vaccine manufacturers.
Interest has been expressed in expanding vaccine injury com-
pensation to the developing world (C.J. Clements, S. Oleja, P. Fife.
Vaccine injury compensation: an international perspective; unpub-
lished data, 2006). The World Health Organization, the United
Nations International Children's Emergency Fund, and other
major organizations have yet to adopt a policy in this regard,
which is understandable given the fact that the necessary eco-
nomic resources and health infrastructure are not often present.
That said, promoters emphasize the inherent responsibility of gov-
ernment/society to ensure that persons who may be harmed by
vaccines are cared for properly, whether through compensation, as
in the model in industrialized countries, or the availability of basic
medical services and support. They see this effort as achieving one
or both of these goals over time.
The power to compel vaccination
US courts generally have been deferential to public health judg-
ments that mandatory vaccination is required. In the leading
193
case, the US Supreme Court upheld
a Massachusetts statute that empowered each local board of
health to require vaccination of the inhabitants “if, in its opinion,
it is necessary for the public health or safety”. Under this author-
ity, the city of Cambridge required all residents to receive the
smallpox vaccination. Although children whose physicians deter-
mined that such vaccination was medically contraindicated were
exempted from this requirement, a similar exemption was not
available for similarly situated adults. Jacobson, who had appar-
ently experienced adverse reactions to the vaccine, refused vac-
cination and was fined. The Supreme Court rejected Jacobson's
argument that the Massachusetts law was constitutionally
invalid and, instead, held that the law was a reasonable and
proper exercise of the state's police power to protect the public
health and safety. The court rejected Jacobson's offer to present
evidence undermining the medical efficacy and safety of vacci-
nation. Instead, the court took judicial notice (without hearing
evidence) of the fact that the people of the state of Massachusetts
held a common belief (which was maintained by high medical
authority) that vaccination was a preventative of smallpox, deter-
mined that the legislature was entitled to rely on this theory,
and concluded that the legislature was not compelled to commit
such a public health and safety matter to the final decision of a
court or jury. This early case is significant insofar as the Supreme
Court recognized the importance of vaccination to the American
public and deferred to the authority of state lawmakers to develop
and impose vaccination requirements on the general population.
Decisions of contemporary courts continue to be supportive
of mandatory vaccination requirements and often cite
in deciding lawsuits challenging immunization mandates. In
194
the health
department sought an injunction that excluded unimmunized
children from school. The health department issued an emer-
gency rule barring unimmunized children from attending school
because of an outbreak of measles in the county. The injunction
was sought to enforce the health department's rule. The trial court
granted the injunction, and the appellate court affirmed. The
court rejected the families' argument that the health department
Legal issues 77 1505
had no authority to exclude children unless there had been a con-
firmed case of measles in the particular school, holding instead
that the health department had the authority to exclude children,
even in the absence of a serologically confirmed case.
Contemporary vaccination laws are not as sweeping as the
law involved in States generally require that chil-
dren receive certain recommended vaccines to attend schools
and child care centers. There is some variation among state
laws as to which vaccines are required for school entry. These
laws are dynamic and often incorporate new vaccines. The
CDC's National Immunization Program maintains a Web
site with current information on state immunization require-
ments for attendance in schools and child care centers.195 The
recommended age of administration for most of these vaccines
is during the first 2 years of life, and the vaccines are admin-
istered to a high percentage of children at the recommended
age.196 State laws requiring vaccination for school entry may
persuade reluctant parents to vaccinate at the recommended
times because vaccination will be required eventually.
The laws that require vaccination before school entry are
congruent with the grant program operated by the US Public
Health Service.197 If a state or local government participates in
that federally funded immunization program, it must, among
other requirements, have a “plan to assure that children begin
and complete their immunizations on schedule” and a “plan to
systematically immunize susceptible children at school entry
through vigorous enforcement of school immunization laws”.198
Although all states allow medical exemptions to school
immunization laws, the scope and application of these laws
vary by state.199,200 In a case concerning medical exemptions,
the Wyoming Supreme Court held that the state of Wyoming's
Health Department could waive state immunization require-
ments for a particular child if such vaccination was medically
contraindicated for that child but that the department could not
require the child or his or her physician to provide a reason for
such a contraindication to immunizations.201
Many states have also voluntarily granted persons exemp-
tions from mandated vaccination requirements based on
religious or personal beliefs.195 Parents have filed numerous
lawsuits challenging the fact that states did not allow reli-
gious and/or personal exemptions or challenging the scope or
application of the specific exemptions in state law. In a case
in which a mother challenged the constitutionality of a West
Virginia law that allowed no religious exemptions, a federal
judge concluded that states need not provide religious exclu-
sions from their compulsory immunization laws.202 Although
the court in 203
struck a religious exemption
from a state statute on the grounds that it violated the Equal
Protection Clause of the Fourteenth Amendment, most courts
have recognized the right of a state to allow religious exemp-
tions. Instead, the courts have focused on the scope of the
exemption. For example, in a fed-
eral district court judge found unconstitutional an Arkansas
law that granted religious exemptions to the state's general
school immunization requirements only to persons who were
members or adherents of a church or religious denomina-
tion recognized by the state but not to others whose objec-
tions to immunization were also grounded in sincere religious
beliefs.204 The judge found that the religious exemption sec-
tion of the Arkansas law violated the Establishment and
Free Exercise Clauses of the First Amendment and the Equal
Protection Clause of the Fourteenth Amendment to the US
Constitution, noting that the primary effect of this exemp-
tion provision was to “inhibit the earnest beliefs and practices
of those individuals who oppose immunization on religious
grounds but are not members of an officially recognized reli-
gious organization”. That case was appealed. However, the
Arkansas legislature broadened the exemption to encompass
philosophical and religious exemptions while the appeal was
pending, rendering the issues on appeal moot.205
1506 SECTION FIVE Public health and regulatory issues
In the Matter of Christine M.,206 a state court held that it
was child neglect for an otherwise responsible parent to refuse
to permit his 2 year old child to be vaccinated for measles during
a measles outbreak in New York City absent a demonstration of
sincere grounds for a religious exemption. However, the court
refrained from ordering immediate vaccination on the grounds
that the vaccination would be required when the child entered
school and that the measles outbreak had ended by the time of
the decision. In another case, New York's Administration for
Children's Services sought a court order to immunize two fos-
ter children over the objections of their foster mother. Applying
New York law, the judge conducted a searching inquiry
concerning the nature of the foster mother's objections to deter-
mine whether they were religious or philosophical. Although
the mother conceded that the tenets of Judaism and Breslov
Hasidism did not necessarily preclude vaccination, she argued
that her interpretation and strict adherence thereto did. Finding
that the foster mother's opposition to immunization was rooted
in genuine and sincerely held religious beliefs, the judge deter-
mined that she qualified for New York's religious exemption
and denied the agency's request for a court order directing the
immunization of the foster children.207 In another case, the US
Court of Appeals for the Second Circuit upheld a lower deter-
mination that the constitutional rights of a mother, who failed
to demonstrate that she sincerely held religious objections to
immunizations, were not violated by the school district policy
requiring persons requesting religious exemptions to complete
individualized questionnaires concerning the nature of their
objections.208 The Wyoming Supreme Court held that the state
was not authorized to hold hearings to judge the sincerity of a
parent's religious beliefs before a religious waiver to state immu-
nization requirements would be granted.209 The court did not
reach the issue of whether such a requirement would violate
the parents' constitutional right to the free exercise of religion.
To date, approximately one third of states allow philosophical
exemptions, and nearly all (48) allow religious exemptions.210
There are no federal governmental vaccination require-
ments for the general adult population in the United States.
However, certain segments of the adult population (eg, college
students and health care professionals) may be subject to par-
ticular vaccination requirements imposed by schools or employ-
ers. Members of the armed forces are also required to receive
vaccines in certain circumstances. In one case, military per-
sonnel subject to orders to receive the anthrax vaccine brought
an action against the FDA, the HHS, and the Department of
Defense, challenging the FDA's determination that the anthrax
vaccine was effective against inhalation anthrax and seeking
temporary and permanent relief from Department of Defense
anthrax vaccination program. A federal court granted the defen-
dants' motion to dismiss, finding in part that the FDA did not
act arbitrarily or capriciously and that the agency applied its
expertise and found that the vaccine was effective for immuni-
zation against anthrax.211
State and employer vaccination mandates for health care
workers have garnered significant attention in recent years.212
213
In a New York trial court upheld health
department regulations that required hospital employees and
medical staff to have current rubella immunizations. The law-
suit was filed by a staff physician who argued that the vaccination
requirement was a violation of the Fourth Amendment protec-
tion against unreasonable searches and seizures. More recently,
several provider groups, including those representing hospital
personnel, sought a temporary restraining order to prevent the
enforcement of a New York State regulation requiring health
care workers with direct patient contact to get the pandemic
H1N1 and seasonal influenza vaccines. The regulations permit-
ted medical exemptions only for medical contraindications.214 A
state court judge granted a temporary restraining order against
the application of the regulation to New York State health care
workers, and the governor announced the suspension of the influ-
enza shot mandate for health care workers due to vaccination
shortages before the order could go into effect. Thus, the lawsuit
was dismissed as moot by the New York Supreme Court.212
Summary, conclusions, and future
implications
The VICP has largely satisfied the public policy imperatives
driving passage of the NCVIA almost 25 years ago. First, per-
sons found to have sustained vaccine-related injuries (through a
presumption of causation under the Table or by demonstrating
causation for off-VIT claims) receive generous compensation
in a streamlined process. Second, the availability of compensa-
tion under the NCVIA has generally given plaintiffs a sufficient
incentive to abandon the pursuit of tort remedies against vac-
cine manufacturers and administrators.
Claims against vaccine manufacturers are significantly reduced
compared with the litigation manufacturers faced at the inception
of the VICP. Although claims against health care providers are more
difficult to track, there is no indication that their liability experience
is any different from that of manufacturers. Until the new wave of
court cases pertaining to autism, few cases alleging vaccine-related
injuries were filed outside the VICP in recent years. The reduc-
tion in the number of such civil actions can largely be explained by
the fact that the NCVIA requires petitioners to file for compensa-
tion under the NCVIA before pursuing outside litigation for alleged
vaccine-related injuries or deaths. With regard to post-1988 claims,
petitioners have two options once the VICP process is concluded:
(1) accept the court's judgment and the compensation awarded (if
any); or (2) reject the court's judgment, with the option of pursuing
a civil action in state or federal court (under limitations set forth in
the NCVIA). Virtually all VICP petitioners, even petitioners who
were not awarded compensation under the VICP, have chosen the
former option. Even among the small number of VICP petition-
ers who have chosen the latter option (a group consisting almost
entirely of persons who were not awarded compensation under the
VICP), most have chosen not to pursue further civil remedies. The
fact that most VICP petitioners choose to accept the court's judg-
ment and forgo additional civil litigation can largely be explained
by the VICP's generous compensation awards, which are devised in
an effort to create a lifetime stream of benefits for the injured party,
and the general understanding that the VICP imposes a lesser bur-
den on litigants than that imposed in traditional civil litigation.
One remarkable feature of the VICP is that all petitioners, includ-
ing petitioners for whom compensation is denied, receive attorneys'
fees and costs so long as their claims were brought in good faith
and on a reasonable basis. A new challenge facing the VICP is the
proper standard for judging causation-in-fact claims, an issue that
is certain to generate a lively debate among the VICP's stakeholders
and to result in an increasing number of judicial opinions as new
alleged vaccine-injury relationships are examined.
The reported decisions are consistent with the conclusion
that vaccine manufacturers and administrators are generally
shielded from any significant liability risk. However, the pro-
tection is not absolute. Because the NCVIA does not preclude
a person who is otherwise ineligible to file a claim under the
VICP (eg, family members of injured persons) from pursuing
civil litigation, liability exposure remains a possibility in states
where such suits may be brought. Many of the recently filed
thimerosal-autism claims include requests for relief by family
members of injured persons, who are not entitled to file under
the NCVIA. This relatively recent surge in non-VICP litigation
threatens to undermine many of the goals underlying the cre-
ation of the VICP. However, the recent Supreme Court deci-
sion in holding that injured claimants are
prohibited from bringing design defect claims against vaccine

manufacturers in civil court, strengthens the success of the
VICP as an alternative to the tort system.
A measure of the success of the VICP is the development and
administration to children of new vaccines to prevent childhood
diseases. Annual investigational new drug requests to the FDA
for vaccines, a necessary step in beginning testing in human
subjects, have stayed relatively steady the past 10 years (Center
for Biologics Evaluation and Research, unpublished data, 2010).
Since the creation of the VICP, the following vaccines have
been recommended for routine administration to children by
the CDC and included in the VICP's coverage: HBV, HAV, Hib,
varicella, rotavirus, pneumococcal conjugate, seasonal (triva-
lent) influenza, meningococcal, and HPV. The fact that since
creation of the VICP such vaccines have successfully traversed
the long, arduous licensing process to become routinely admin-
istered vaccines demonstrates that the scientific community is
able to develop innovative vaccine products to combat child-
hood diseases under the current system.
Another benefit of the VICP has been that liability concerns
no longer lead to destabilization of the marketplace. The mar-
ket experienced some vaccine supply shortages from 2000 to
2004, but such shortages did not seem attributable to liabil-
ity factors.215 Although many new vaccines are relatively expen-
sive (reflecting the manufacturers' need to recoup the costs of
research and development), vaccine prices today reflect public
and private sector purchase trends and the effects of inflation,
rather than liability concerns or repercussions. Thus, the exis-
tence and general success of the VICP in meeting the public
policy objectives underlying the NCVIA have led to a current
system in which innovative vaccine products and technolo-
gies are developed, the vast majority of American children are
immunized against a growing number of diseases that pose a
threat to their health, and the small number of persons who
sustain adverse events as the result of such immunizations are
able to receive appropriate compensation through the VICP.
Despite its successes, the VICP has been criticized for sev-
eral perceived weaknesses. For example, one criticism is of the
fact that the liability protections provided by the NCVIA are not
absolute (eg, persons with claims of $1,000 or less need not pur-
sue claims under the VICP, vaccine manufacturers and adminis-
trators may be liable if a person exhausts his or her remedies in
the VICP and then opts to pursue civil litigation outside of the
VICP).215 Another criticism is of the fact that the VICP extends
only to certain vaccines recommended for routine administra-
tion to children, rather than all vaccines. Thus, for vaccines that
are not covered by the VICP (eg, vaccines against Lyme disease,
vaccines not recommended for routine administration to chil-
dren that are administered to pregnant women), the NCVIA's
expedited no-fault process and attendant liability protections
are unavailable. The NCVIA's statute of limitations (for inju-
ries, 3 years of the date of the first symptom or her manifes-
tation of onset of the injury) also has been criticized. Because
this period begins running from the first symptom rather than
from diagnosis or any attribution to vaccines, some believe the
NCVIA's statute of limitations is particularly unfair as applied
to certain illnesses with early manifestations that may be subtle.
The fact that claims for damages sustained by persons close to
an injured person (eg, loss of consortium claims by an injured
party's spouse) cannot be pursued in the VICP has been criti-
cized. The conflict inherent in litigation may also be seen as a
weakness of the VICP when compared with other compensation
programs that are purely administrative. Depending on one's
vantage point, some of these perceived weaknesses may be seen
as strengths (eg, the fact that the liability protections afforded
by the NCVIA are not absolute may be seen as a weakness by
persons who see vaccines as an important but vulnerable part of
Access the complete reference list online at http://www.expertconsult.com
Legal issues 77 1507
the public's health but as a strength by injured parties who want
to file suit against persons who they believe responsible for the
damages sustained).
An increasing focus has been given to the issue of vaccine
safety during recent years. Through media involvement and
Internet access, the ability for misleading information to be
disseminated has increased. As a result, effective vaccine risk
communication becomes imperative and must include a recog-
nition of the difficult and confusing nature of risk assessment
and decision making for some vaccines.216 Such communica-
tion is particularly critical now, at a time when most parents
do not have an independent recollection of the devastation
caused by the infectious diseases prevented by childhood vac-
cines. Interwoven into vaccine safety awareness are the issue
of state mandates for immunization and calls for all states to
provide exemptions. Vaccination levels among American chil-
dren remain remarkably high and of vaccine-preventable dis-
eases, generally low. Moreover, states198 seem to have struck a
fair balance in implementing school and child care immuniza-
tion requirements against the rights and interests of persons
who object to immunization on various grounds through the
allowance of exemptions. Courts generally continue to uphold
the states' police powers to mandate immunization for school
entry, while allowing them to provide exemptions for medical,
religious, or philosophical reasons.
Our nation's experience with the public's perception con-
cerning immunization and other countries' experiences with
the same issue make clear that uncertainty and fear often
come together when information is lacking. Although most
parents seem to believe in the wisdom of vaccination and use
their health care providers as a key source of guidance in deci-
sion making, immunization must never be taken for granted.
Although the federal and state immunization programs,
together with the VICP, have done a remarkable job of ensur-
ing that our nation's children will be vaccinated against danger-
ous diseases and although a national surveillance system and
several IOM reports confirm the general safety of immuniza-
tions given routinely in this country, public perceptions con-
cerning the risks of vaccines remain an important challenge for
vaccine administrators to address. The recent trend of litiga-
tion concerning thimerosal raises this to a degree that had not
been present in recent years. Given the outcome of the omni-
bus autism proceedings, the VICP has shown to be able to ade-
quately address this new phenomenon. Finally, the ability of
the VICP to successfully fulfill its mandate and to further its
statutory purposes seems strong, particularly given the near-
universal consensus among the VICP's stakeholders (ie, physi-
cians, manufacturers, lawmakers, attorneys, and members of
the public, including parents of injured children) that a suc-
cessful national immunization program can be achieved only
with a compensation program and liability safeguards in place.
Recent experiences with legislation enacted in preparation for a
potential smallpox attack or the emergence of a pandemic flu
or other public health emergencies have emphasized the neces-
sity of providing clear liability protections and fair compensa-
tion to ensure the success of vaccination programs across the
United States.
Acknowledgments
We thank the following persons for their assistance in preparing the
manuscript: Rosemary Johann-Liang, MD, Robert E. Weibel, MD,
Barbara Shobach, MD, Thomas Ryan, MD, Ward Sorensen, and
Carole Marks from the VICP, HHS; David Benor, JD, and Margaret
McNamara from the Office of the General Counsel, HHS.
 SECTION FIVE: Public health and regulatory issues
Ethics
78 Arthur L. Caplan
Jason L. Schwartz
Vaccines present a spectrum of unique ethical considerations
compared with those associated with other medical interven-
tions. Central to these differences are the role of vaccines in dis-
ease prevention rather than treatment, the concurrent interest
of vaccination policy in improving the health of communities
and individual people, and the focus on children in many vac-
cination programs. Formal discussions of ethical issues related
to vaccination were once concentrated largely on aspects of
clinical research and specific topics regarding vaccine safety and
financing. Recent events, such as the arrival of vaccines against
human papillomavirus (HPV) and the 2009-2010 H1N1 influ-
enza pandemic, have demonstrated the need for a more com-
prehensive and proactive examination of vaccine ethics. Such
an approach offers important insights into all aspects of the
vaccine life cycle, from the earliest stages of research through
deployment in national and global immunization programs.
Ethical issues unique to vaccination
Ethics of prevention vs ethics of treatment
Unlike pharmaceuticals and most other medical interventions,
vaccines are intended to prevent disease rather than treat it.
The ethics of prevention differ greatly from the ethics of treat-
ment. Chief among these differences are the disparate mean-
ings and interpretations of "risk" in each context.
When treating disease, risks are largely confined to two broad
categories: the risks associated with a particular intervention
and the risks of doing nothing. These risks are weighed against
the potential benefits of a specific treatment and compared with
the risk-benefit profiles of potential alternatives. Decision mak-
ing is invariably complicated by the uncertainty associated with
any assessment of risks and benefits. Nevertheless, a treatment
decision is made based on the evaluation of all options in light
of the risks and potential benefits of each, all compared with the
consequences of doing nothing.
When aiming to prevent disease by means of vaccination,
assessing the risk of inaction is more challenging. Here, the
physiological consequences of a disease in an individual per-
son must be considered the specific likelihood
of acquiring that disease. As a result of successful vaccina-
tion programs, the incidence of many vaccine-preventable
diseases is extremely low in the United States and other devel-
oped nations. This complicates efforts to convey the continued
necessity of vaccines to parents, many of whom may have never
seen or experienced the diseases being prevented. Globally, the
risk of disease is subject to wide geographical, national, and
occupational variability. In all cases, high rates of vaccination
among communities are widely credited with preserving the
low incidence of many vaccine-preventable diseases.1
The benefit of high vaccination rates highlights a key dif-
ference between the ethics of treatment and the ethics of pre-
vention. Prevention, particularly the prevention of infectious
diseases through vaccination, has important implications for
individual people and communities. While considerations rela-
tive to justice and the allocation of scarce resources have ethical
relevance to some treatment decisions, rarely are they primary
factors in how care is delivered.
As a form of prevention, vaccination requires the juxtaposi-
tion of individual autonomy with societal best interests. These
values are often in alignment, but for some persons, they pres-
ent competing arguments regarding the decision to vaccinate.
Central to this debate is herd immunity, the additional protec-
tion against a disease that is a result of high vaccination rates in
a community. Some persons believe that the benefits of vaccina-
tion are outweighed by the associated risks, however small they
may be. This is a particularly likely scenario for the diseases
that have virtually disappeared in wealthy nations due in part to
successful vaccination programs. The people choosing not to be
vaccinated would still receive some protection because of herd
immunity, creating the potential ethical problem of responding
to "free riders".
Free riders are people who share in the benefits of vaccina-
tion programs without personally assuming any vaccine-related
risks or costs, such as the risk of adverse events or the time and
expense required to receive vaccinations. Recent vaccine-
preventable disease outbreaks among unvaccinated persons
suggest that personal reliance on herd immunity for disease
protection is often inadequate.2,3 Such decisions also increase
the overall risk of disease in communities, particularly for the
members too young to receive a vaccine or unable to do so
because of medical contraindications.4 People who knowingly
benefit from the actions of others without sharing in the bur-
den or cost of providing that benefit act in an unethical manner.
Target populations and the routinization of vaccines
The significance of healthy children as the largest target pop-
ulation for vaccines cannot be overstated. Young children are
ethically vulnerable because they cannot exercise their own
autonomy to mediate issues of risk and benefit. With a major-
ity of vaccinations in the United States recommended for
children in the first 24 months of life, vaccine recipients are
typically among the most ethically vulnerable of populations.
While striving to make decisions that are in a child's best inter-
ests, parents and guardians must navigate a sea of at times

conflicting information related not only to vaccines, but also
to all aspects of a child's medical care.5,6 The heated, heavily
publicized debates about the safety and value of vaccinations
in recent years have added to the challenges faced by parents
aiming to make responsible and informed decisions about their
children's health.
Ironically, communication efforts may be hampered as a
result of how routine childhood vaccination has become in
many parts of the world. Amid competing demands on health
care providers, efforts to explain the continued importance of
vaccination to parents may suffer. Absent unique concerns
raised by parents or providers, communication about vaccines
may be left instead to a few short questions and answers or reli-
ance on government-required printed information statements.6
The routinization of vaccines is amplified by state laws in the
United States requiring vaccination against many diseases as a
condition of school or day-care attendance. This atmosphere of
familiarity partly explains the widespread alarm generated by
reports of vaccine-related safety concerns, regardless of whether
the concerns are confirmed or only alleged.
Ethical considerations in the vaccine life
cycle: an overview
While the preceding discussion highlights some of the unique
attributes of vaccines and vaccine decision making, their
research, development, and regulation are similar in form
and function to those of pharmaceuticals and other medical
interventions. In the following sections, we provide an over-
view of relevant ethical considerations at various stages of the
vaccine life cycle, a period that begins with the earliest basic
research and extends through licensure and all dimensions of
national and international vaccine production and distribution
programs.
Research and development
Throughout the modern history of vaccines, the successful
development of new discoveries has relied on collaborations
between academic medicine and the pharmaceutical indus-
try. More recent additions to these collaborations have been
smaller biotechnology companies that typically conduct early
vaccine development before licensing their successful prod-
ucts to larger manufacturers. While financial support for
these activities has traditionally come from a combination of
investment funds and government-sponsored research awards,
philanthropic groups have become increasingly interested in
vaccine development in recent years. Particularly for diseases
more common in the developing world, these philanthropic
entities often bring substantial financial resources to support
their research interests.
Such a diverse group of research entities and funders all
but assures a collection of differing research priorities, objec-
tives, motives, and measures of success. While all contributors
share the general aim of developing safe and effective vaccines,
conflicts may arise regarding how best to achieve this goal.
If unchecked, such conflicts could impede progress toward
novel vaccines, waste limited financial resources, and fail to
respect the contributions of human research subjects. Large
research partnerships have the potential to advance public
health in ways that would occur much more slowly, if at all, if
attempted by individual entities. While respecting their obli-
gations to shareholders, boards, or other overseers, all contrib-
utors to vaccine research and those who invest in their work
should remain keenly aware of the morally distinguishing abil-
ity of their work to save lives, prevent suffering, and improve
global health.
Ethics 78 1509
Research partnerships become even more ethically complex
when including clinical research in developing nations. This is
increasingly important as more vaccine development efforts tar-
get diseases common in the developing world, often through
collaborations in which Western researchers have a leading
role. These activities bring needed expertise and capabilities to
nations often lacking robust medical research infrastructures.
During this process, however, local researchers and health min-
istries ought to be involved in a meaningful way in all aspects
of clinical research taking place in their country. The benefits
of such relationships are many. For example, they may provide
additional knowledge and training for local health officials,
expertise that could benefit communities long after research
has concluded. Meaningful contributions from local officials
also create an additional layer of protection to ensure that the
generosity of volunteers is not exploited by research that might
have been deemed unethical had it been proposed for popula-
tions in developed nations.
No topic related to vaccine research has generated more con-
troversy on ethical grounds than the designs of clinical trials,
particularly those in the developing world.7–11 As an example,
one notable debate during the past 15 years has centered on HIV
vaccine trials in the developing world, particularly the level and
duration of care that ought to be provided to research subjects
who become infected during trials.12–17 Options range from life-
long treatment with the latest in antiretroviral medications—
the norm in many developed countries but highly uncommon
elsewhere—to whatever the typical treatment is in the coun-
try where the trial took place. Such a level of care is often well
below that of the research sponsors’ home countries. It may be
nothing. Because of the significant consequences of this debate
to the feasibility of future research, attempts at reaching con-
sensus on the question of which standards should prevail have
largely failed.12,14 This long-standing and still unresolved debate
provides a useful example of the value of robust discussions
of ethical considerations controversies develop and deci-
sions cannot be undone.
Vaccine development can never be immune from the twin
pressures of personal advancement and corporate profitability.
Nevertheless, the world community is best served by vaccine
research programs that match these concerns with a continued
acknowledgment of the enormous suffering that can be averted
by vaccines and respect for the individual people and communi-
ties volunteering to assist in clinical research.
Finally, concern for justice requires that the populations in
which a vaccine candidate is studied mirror as closely as possi-
ble the groups expected to receive the licensed product. The use
of mentally handicapped children as primary sources of vaccine
research subjects in the 1950s and 1960s exemplifies how this
principle was violated in the past.18 The children often lived in
overcrowded institutions with sanitary conditions that placed
them at increased risk for the diseases for which potential vac-
cines were being tested. New vaccines would have been particu-
larly valuable to the children, who also provided a convenient,
accessible study population for researchers.
However, mentally handicapped children in state institu-
tions are among the most vulnerable of populations, for whom
clinical research is ethically appropriate only in extremely lim-
ited circumstances and always with strict oversight. In pursuit
of vaccines that would benefit society broadly, the past use of
these children as a primary study population often violated this
fundamental tenet of research ethics.
Today, the opposite extreme has become common.
Manufacturers have a strong disincentive to include among
their research subjects members of potentially high-risk popu-
lations, such as pregnant women or children with intellectual
disabilities. Without complete data on safety and efficacy in all
populations to whom a vaccine will be administered, patients,
parents, and policymaking bodies have inadequate information
1510 SECTION FIVE Public health and regulatory issues
on which to base their decisions regarding vaccination in
these groups. The vaccine research and regulatory communi-
ties should remain cognizant of past exploitation of vulnerable
research subjects but strive to conduct clinical research in ways
that protect subjects while providing as complete a picture as
possible of a vaccine's safety and efficacy profile.
Licensure and safety monitoring
Vaccines are subject to oversight and regulation by a variety of
entities in every country where they are available.19 It is initially
determined whether a new vaccine should be licensed, and if
so, the populations for whom it should be recommended. In
the United States, these responsibilities belong to groups within
the Food and Drug Administration (FDA) and the Centers for
Disease Control and Prevention (CDC), respectively. For the
life of the vaccine thereafter, activities are undertaken by these
groups in collaboration with its manufacturer to monitor its
safety and efficacy.20 These processes have generated consider-
able controversy in recent years, threatening to damage pub-
lic confidence in vaccination.21 Since the success of vaccination
programs depends on earning and maintaining public trust,
there are several points in vaccine regulation at which ethical
considerations are relevant to public policy.8,22
A primary concern of critics of vaccine policy in the United
States is the potential for conflicts of interest among govern-
ment advisors and researchers who have ties, financial or oth-
erwise, to vaccine manufacturers. The very nature of vaccine
development presents unique challenges in avoiding even the
appearance of a conflict. Any researcher working on novel vac-
cine candidates eventually partner at some point with
industry owing to the infrastructure needed for large-scale clini-
cal testing and development. To exclude all such researchers for
this reason would be to forfeit a wealth of expertise and wis-
dom on vaccine science and policy, ignoring internationally
respected leaders in vaccine science.
However, the importance to vaccination programs of main-
taining public trust demands that those contributing to policy
exercise particular caution and care regarding their professional
and financial relationships. Most advisory bodies have clear pol-
icies regarding the disclosure of potential conflicts of interest.23
Even when confident that financial relationships would have
no impact on their actions, individuals should remain particu-
larly attentive to how such interests might affect the manner in
which the decisions to which they contribute may be perceived.
Transparency, minimization of personal gain, divestiture, and
disclosure are crucial principles that work to counteract the per-
ception of conflict of interest influencing decision making.
Public attention to conflicts of interest among policymakers
and their expert advisors is often linked to reports of vaccine
safety concerns. The public health and regulatory communi-
ties should respond vigorously to reports of vaccine-associated
adverse events, even if they initially seem unlikely. Passive sur-
veillance programs such as the Vaccine Adverse Event Reporting
System in the United States are valuable tools in this effort, but
their limitations as hypothesis-generating mechanisms should
be clearly conveyed to the public and the media.24 Emerging pat-
terns of possible safety problems should be explored thoroughly,
with the results of these analyses promptly communicated to
the public. Even when evidence suggests that a reported safety
concern is unfounded, experience has shown it is unlikely that
any such assurance will allay the worries of all. Open, expedi-
tious, and detailed examinations of possible vaccine-associated
safety concerns are nevertheless essential to maintaining over-
all confidence in vaccination programs and their oversight.21,25
It is inevitable that a small number of persons will genuinely
be harmed as a result of vaccines. Since vaccination benefits
communities, victims are entitled to compensation for their
suffering in these rare cases. In the United States, the Vaccine
Injury Compensation Program is the principal mechanism for
resolving such issues.26 By means of a no-fault system with
funds collected via a tax on every vaccine dose, the system pro-
vides compensation without exposing manufacturers to sub-
stantial financial risk.
This program has faced scrutiny in recent years regarding how
claims are resolved and the scope of its coverage. The Omnibus
Autism Proceeding and the case of heard
by the US Supreme Court in 2010 represented challenges to
aspects of the federal government's system for identifying and
compensating for vaccine-related injuries.27,28 Given the con-
comitant challenges of compensating victims fairly, accurately
distinguishing correlation from causation regarding adverse
events, and ensuring that manufacturers remain committed to
developing and manufacturing vaccines, the current design of
the Vaccine Injury Compensation Program is a generally fair
way of resolving these ethical obligations.
Vaccine supply, access, and financing
The highly publicized influenza vaccine shortage in 2004 and
many other shortages for recommended childhood vaccines
underscore the vulnerability of vaccine supply worldwide.29
With limited manufacturers and often only a single licensed
product available in a country, vaccines are highly susceptible
to large fluctuations in supply and availability owing to unfore-
seen events.30–32 In the United States, CDC stockpiles of recom-
mended vaccines are maintained to provide a temporary buffer
in case of production or supply problems, but these efforts pro-
vide only limited protection.33,34
A spectrum of economic and business considerations help
explain why manufacturers have left the vaccine market and
why the remaining manufacturers are not eager to develop new
products to compete with older vaccines that, in general, are
able to meet demand.30,31 Increasing the number of vaccine
manufacturers and developing new vaccines for common dis-
eases would help to ensure a more resilient vaccine supply land-
scape better insulated against the shortages that have become
increasingly common.
Even when vaccine supplies are adequate, vaccination rates
reflect many of the same racial and ethnic disparities in access
present throughout medicine.35,36 While the underlying causes
of these conditions are debated, several programs seek to elim-
inate vaccine cost as an obstacle to childhood vaccination in
the United States, most notably the Vaccines for Children and
Section 317 grant programs for uninsured and underinsured
children.37 These programs make vital contributions to child-
hood vaccination rates, but aspects of their administration gen-
erate ethical dilemmas for state health departments and health
care providers.
The introduction of newly licensed and recommended vac-
cines, often with far higher prices than older products, has
outpaced the growth of Section 317 funds available for vacci-
nation programs in many states.38 This has forced state health
officials to decide which CDC-recommended vaccines to make
available to children when unable to afford them all. If required
to ration state public health funds available for public-sector
vaccine purchase, state policymakers should make decisions
informed by the best available data, state-specific epidemiol-
ogy, and comparative cost-effectiveness of each option under
consideration. At the same time, advocates for vaccination
have an obligation to raise awareness of inequities in vaccine
financing and access.
Private health insurance plans generally provide coverage for
CDC-recommended vaccines for children and adults, but they
may require substantial patient copayments for some vaccines
or fail to reimburse health care providers fully for related admin-
istrative costs. As a result, financial barriers remain for many
patients and providers even when private insurance is available.

According to the Institute of Medicine, an estimated 11 million
children and 59 million adults have private insurance that does
not adequately cover costs related to immunization.39
As part of the 2010 Affordable Care Act, new insurance plans
must provide coverage for any vaccine recommended for rou-
tine use in children or adults by the CDC Advisory Committee
on Immunization Practices (ACIP).40 No copayment, coinsur-
ance, or deductible can be required from the insured patient.
This provision for vaccines is part of a spectrum of preventive
services required of new private insurance plans. Over time,
these requirements are likely to bring transformative benefits to
efforts already underway to promote prevention. How this leg-
islation will ultimately reshape existing programs for promoting
and financing childhood vaccination remains uncertain.
In comparison with children and insured adults, uninsured
adults face far greater financial obstacles to vaccination. With
an increasing number of vaccines recommended for adults but
no program similar to Vaccines for Children, vaccine affordabil-
ity for uninsured adults is likely to become a growing concern
despite ongoing attempts at health insurance reform. Already,
there is evidence suggesting the detrimental effect of vaccine
cost concerns on uptake of herpes zoster (shingles) vaccine for
adults 60 years or older and HPV vaccine for young adults older
than 18 years.41 Federal and state governments, in collabora-
tion with vaccine manufacturers, should identify strategies that
reduce these financial obstacles to adult vaccination. The laud-
able patient assistance programs recently introduced by sev-
eral vaccine manufacturers may provide a foundation for more
expansive efforts in this area.
Vaccination mandates
Governments worldwide use a variety of approaches to promote
high vaccination rates among their citizens.37,42,43 The United
States is generally unique in its reliance on federal recommen-
dations coupled with state school-entry vaccination require-
ments as central to the success of vaccination efforts. While
specific requirements vary among states, all mandate that chil-
dren receive a series of vaccinations as a condition of attending
public school or state-licensed day-care facilities.44 Every state
allows for exemptions based on medical grounds, and nearly
all also accept religious or philosophical reasons, although not
every state includes all three types of exemptions.45
No topic related to vaccine ethics in the United States is more
frequently debated than state vaccination mandates. The debate
reflects a common tension in public health policy between indi-
vidual (or parental) autonomy and the public good. Vaccines are
mandated not primarily to protect the health of any particu-
lar individual, but to ensure that the transmission of diseases
through communities is limited to the greatest extent possible.
School-entry requirements are seen by public health officials as
essential to maintaining vaccination rates sufficiently high to
preserve herd immunity.46 This provides additional protection
against vaccine-preventable diseases to all members of a com-
munity, including those too young to receive vaccines or unable
to do so because of medical contraindications.
For persons whose views on medical ethics are guided by the
primacy of patient autonomy, it is understandable why US vac-
cination mandates are so contentious. However, few contempo-
rary ethical models place autonomy absolutely above all other
considerations. Instead, respect for autonomy is typically one
of several factors that should be examined in light of other rel-
evant considerations as part of ethical deliberation and decision
making. There is a compelling argument that the lives saved
and suffering prevented by vaccination outweigh the potential
infringement on personal autonomy created by school man-
dates. While one may be free to make medical decisions that
place his or her own health at risk, he or she may not jeop-
ardize the health of others, a consequence of low vaccination
Ethics 78 1511
rates within a community. Even in an autonomy-oriented cul-
ture such as the United States, there are ethical reasons to place
some limits on individual choice.
An ethically preferable scenario would be maintaining cur-
rent high rates of vaccination without needing the force of man-
dates to do so. Absent evidence that this is attainable in the
United States, the current policy is sound. Mandates serve as
a "safety net", a valuable tool to call attention to the impor-
tance of vaccines and help direct government and public health
resources to vaccination efforts.44 Exemption policies provide a
ready alternative in nearly all states for persons whose personal
beliefs do not coincide with protecting public health. Combined
with incomplete enforcement by state health departments or
local school districts, current policies fall far short of true com-
pulsion. They are instead best understood as presumptive or
default approaches to vaccination.
While national rates of nonmedical exemptions remain quite
low, recent trends—particularly within select communities—are
of deep concern to advocates of vaccination.47 Overall, however,
a large majority of Americans continue to believe in the value of
vaccination for themselves and their children despite occasional
controversies regarding the necessity or safety of vaccines.
Special topics in vaccine ethics
Human papillomavirus vaccines
Vaccines against HPV have raised a spectrum of ethical and
policy considerations since their arrival in 2006. Initially,
some critics expressed concern that HPV vaccination might
negatively influence decisions made by teenagers about sex-
ual behavior, increase the likelihood of promiscuity, and pro-
vide a false sense of security as to protection against sexually
transmitted infections.48 Evidence suggests that HPV is a very
small factor in decisions about sexual behavior among teenag-
ers, however.49 Since at least 30% of cervical cancers are caused
by HPV strains not included in current vaccines, education
remains essential to convey the importance of continued vigi-
lance regarding cervical cancer screening as part of comprehen-
sive prevention programs.
Shortly after the licensure of the first HPV vaccine in the
United States in 2006, many states introduced legislation
that would add the vaccine to those required for school atten-
dance.50 An executive order by Texas Governor Rick Perry in
early 2007 that mandated the vaccine for sixth-grade girls gen-
erated considerable controversy nationwide.50 The executive
order was overturned by the Texas legislature, and, as of early
2012, only Virginia and the District of Columbia have adopted
HPV vaccine requirements. However, extremely liberal exemp-
tion clauses mean that these "mandates" are little more than
"opt-out" policies.51
There is broad agreement that vaccine requirements should
be considered only after a new vaccine is well established. This
includes stable financing and supply arrangements, evidence of
long-term safety, and successful educational initiatives for par-
ents and health care providers.52 While well intentioned, early
efforts to mandate HPV vaccination were premature, and the
negative attention they generated may have been a distraction
to overall HPV vaccine educational efforts.
The approval of two similar but distinct HPV vaccines in
the United States and many other countries presents addi-
tional ethical challenges with respect to the prevention of geni-
tal warts.53 While both vaccines provide comparable protection
against two HPV strains responsible for a majority of cases
of cervical cancer, the quadrivalent vaccine manufactured by
Merck and Co. also includes protection against two additional
strains of HPV that are the most common causes of genital
1512 SECTION FIVE Public health and regulatory issues
warts in both sexes. Genital warts are not fatal but require med-
ical evaluation and treatment. Commentators have noted the
significant emotional impact on quality of life owing to genital
warts, a condition more common among adolescents and young
adults.54 Public and private sources of vaccine financing must
carefully consider how to assess the benefits of both HPV vac-
cines and evaluate the importance of genital wart prevention
relative to the far more publicized role of these vaccines in pre-
venting cervical cancer.
Related ethical considerations are raised by the vaccination
of males against HPV. In 2009, the quadrivalent vaccine was
licensed by the FDA for use in males for the prevention of geni-
tal warts.55 An expanded indication for the prevention of anal
cancer in both sexes was added in 2010.56 Potential indirect ben-
efits of male HPV vaccination include the additional reduction
in cervical cancer incidence that would result from targeting a
reservoir for the virus. Economic modeling of male vaccination
efforts was generally unfavorable in 2009-2010, suggesting that
concentrated attention to improving vaccination coverage in
females was the preferred strategy for HPV-related disease pre-
vention. However, encouraging both sexes to receive the vaccine
appeals to fairness and simplifies promotional efforts. It would
also symbolize the shared responsibility of men and women in
the prevention of cervical cancer and other sexually transmit-
ted infections.53
The ACIP initially opted against a routine recommendation
for male HPV vaccination, instead adopting in 2009 a “per-
missive use” statement that acknowledged that the vaccine
was available for persons who want it. Based on new informa-
tion about the effectiveness of the vaccine and additional eco-
nomic modeling data, the panel revisited its guidance in 2011.
Its amended recommendations now endorse routine use of the
quadrivalent vaccine in males aged 11 or 12 years and in unvac-
cinated males up to 21 years old.57
Health care providers
Health care providers have a twofold role in the success of
disease-prevention efforts through vaccination. Considerable
evidence points to the importance of recommendations from
physicians and other providers in influencing the decisions of
parents regarding vaccines.58 Amid conflicting information and
contentious debates about the safety, effectiveness, and value of
vaccines, health care providers can help patients (or more often
parents or guardians) make decisions about vaccination based
on the best available evidence. Doing so requires sustained
attention by providers to new information about vaccines—
particularly vaccine safety—and the time and willingness to
engage parents with concerns or uncertainty. Some pediatri-
cians have chosen to decline to care for patients whose parents
decline to receive the CDC-recommended vaccination sched-
ule. This approach is not endorsed by the American Academy of
Pediatrics, and it raises the concern that not all of these parents
will seek or find alternative care for their children.59
Health care providers can also demonstrate the value of
vaccination by ensuring that they are personally up-to-date on
recommended vaccines. Beyond the symbolic value of this action,
vaccination of providers protects patients, particularly in hospital
settings where diseases like influenza are easily transmitted.
Maintaining high vaccination rates against seasonal
influenza among health care providers has proven to be a con-
siderable challenge, despite the vaccine being recommended for
this group since 1981.60 A variety of programs, including those
involving incentives for compliance and appeals to professional
duty, have failed to yield adequate vaccination rates, prompt-
ing an increasing number of health care facilities to mandate
annual influenza vaccination as a condition of employment.61
In light of the consistent failure of voluntary approaches to
improve influenza vaccination rates among health care workers,
mandatory vaccination policies are an appropriate response,
particularly in light of the increased susceptibility for infection
among many patients in hospitals and related settings.60
Vaccination in the developing world
Special ethical considerations related to vaccines in the devel-
oping world extend beyond the research issues discussed pre-
viously. A particular challenge is ensuring that new vaccines
are introduced against diseases that are most prevalent or most
severe in those nations but uncommon or mild in developed
countries. Owing to the limited profitability of such vaccines,
manufacturers are often reluctant to invest in these efforts.62,63
Much of this work is therefore supported by private philanthro-
pies, nonprofit entities, and public-private partnerships. These
efforts should continue to be encouraged so that the benefits of
vaccination may be more equitably distributed among all pop-
ulations. For vaccines developed by corporate manufacturers,
work should continue to develop financing arrangements that
deliver existing products to populations in the developing world
that often would benefit most from them.
When organizing vaccine distribution programs, particular
respect should be paid to cultural traditions and social customs
specific to communities in the developing world. In locations
where health care infrastructures are radically different from
those of wealthy nations, successful vaccination efforts depend
on embracing these differences, valuing the input of commu-
nity leaders, and striving to develop programs that gain wide-
spread support.
Finally, efforts should be undertaken to better understand
the concept of consent in the context of developing world vacci-
nation programs. Informed consent in the strict Western sense
may not always be attainable, nor may it be a reasonable expec-
tation owing to the varying structures of communities and fam-
ilies present throughout the world. However, vaccination efforts
should remain faithful to the spirit of informed consent, with
those administering vaccines taking steps to ensure that recipi-
ents receive much more from vaccination programs than the
vaccine dose alone.
Eradication campaigns
The global eradication of smallpox, certified by the World
Health Organization in 1980, remains one of the foremost tri-
umphs in the history of public health.64 By means of a coordi-
nated, extensive vaccination campaign, a disease that had been
the cause of untold suffering and death for centuries was effec-
tively eliminated. The successful eradication of smallpox added
to the enthusiasm for other eradication campaigns against
vaccine-preventable diseases, including those already under-
way and others hypothesized at the time. This enthusiasm has
continued, despite as yet insurmountable challenges in adding
additional diseases to the list of those eradicated.
Measles and polio have long been among the most prom-
inent targets of eradication efforts, and eradication is now
mentioned as a target for malaria, a disease for which a partially
effective vaccine recently has been developed.65 In recent years,
most attention has been directed toward the potential eradica-
tion of polio, a goal that seems tantalizingly close in light of
the limited number of identified cases (approximately 2,000/
year) and the few countries where the disease remains endemic
(four).66 The unique characteristics of poliovirus have presented
significant challenges for further reductions in the incidence of
the disease, despite laudable attention and investment in the
effort, much of it supported by philanthropies including The
Bill and Melinda Gates Foundation and Rotary International.67
The obstacles facing polio eradication have prompted some
observers to suggest that a better overall strategy for global
health is to maintain current levels of control while redirecting