Pulmonary Tuberculosis
SARAH M. LYON1 and MILTON D. ROSSMAN1
1
Pulmonary, Allergy and Critical Care Division, Department of Medicine,
Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104
ABSTRACT This review on pulmonary tuberculosis includes
an introduction that describes how the lung is the portal of entry
for the tuberculosis bacilli to enter the body and then spread to
the rest of the body. The symptoms and signs of both primary
and reactivation tuberculosis are described. Routine laboratory
tests are rarely helpful for making the diagnosis of tuberculosis.
The differences between the chest X ray in primary and
reactivation tuberculosis is also described. The chest computed
tomography appearance in primary and reactivation tuberculosis
is also described. The criteria for the diagnosis of pulmonary
tuberculosis are described, and the differential is discussed.
The pulmonary findings of tuberculosis in HIV infection are
described and differentiated from those in patients without HIV
infection. The occurrence of tuberculosis in the elderly and in
those patients on anti-tumor necrosis factor alpha inhibitors is
described. Pleural tuberculosis and its diagnosis are described.
Efforts to define the activity of tuberculosis and the need for
respiratory isolation are discussed. The complications of
pulmonary tuberculosis are also described.
INTRODUCTION
The lung is the most commonly affected organ in tu-
berculosis infection in the immunocompetent host, with
estimates of lung involvement in subjects with active
tuberculosis of 79 to 87% (1–3). Estimates of lung in-
volvement are similar in immunocompromised hosts,
such as those with human immunodeficiency virus (HIV)
infection, with studies from the 1980 to 1990s suggest-
ing that the rates of pulmonary involvement were on the
order of 70 to 92% (4–6). However, these individuals
are also more likely to have extrapulmonary disease as
well (7).
The lung is the portal of entry in the majority of cases
of tuberculosis (5, 6, 8). The first contact with the or-
ganism results in few or no clinical symptoms or signs.
Ordinarily, the tubercle bacillus sets up a localized in-
fection in the periphery of the lung, where it has been
deposited by inhalation. Body defenses appear to have
ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
little effect on the organism until the time of develop-
ment of tuberculin hypersensitivity (4 to 6 weeks). At
this time, mild fever and malaise develop, and occa-
sionally other hypersensitivity manifestations are noted.
In the majority of patients, no additional evidence of
tuberculosis develops, and the process is contained by
local and systemic defenses. Since the primary pulmo-
nary focus is usually subpleural, rupture into the pleural
space may result, with the development of a tuberculous
pleurisy with effusion. This is usually accompanied by
the classic but nonspecific symptoms of pleurisy. Local
spread to the hilar lymph nodes is a common occurrence,
and from there the disease spreads to other areas of
the body. It is this hematogenous dissemination of the
organism that results in the pulmonary and extrapul-
monary foci that are responsible for the major clinical
manifestations of tuberculosis. Radiographically, spread
is manifested by enlargement of the lymph nodes, with
later calcification of both the lymph nodes and the pa-
renchymal lesion. This is the classic Ghon’s complex and
is suggestive not only of an old tuberculous infection
but also of diseases such as histoplasmosis. Progressive
(reactivation) tuberculosis usually develops after a peri-
od of dormancy and arises from the sites of hematoge-
nous dissemination (9, 10).
Thus, the first infection with tuberculosis frequently is
clinically insignificant and unrecognized. In the majority
Received: 25 November 2016, Accepted: 2 December 2016,
Published: 10 February 2017
Editor: David Schlossberg, Philadelphia Health Department,
Philadelphia, PA
Citation: Lyon SM, Rossman MD. 2017. Pulmonary tuberculosis.
Microbiol Spectrum 5(1):TNMI7-0032-2016. doi:10.1128
/microbiolspec.TNMI7-0032-2016.
Correspondence: Sarah M. Lyon, Sarah.Lyon@uphs.upenn.edu;
Milton D. Rossman, Rossmanm@mail.med.upenn.edu
© 2017 American Society for Microbiology. All rights reserved.
1
Lyon and Rossman
of patients, the disease stays dormant either indefinitely
or for many years and when a breakdown occurs, it may
be secondary to a decrease in body immunity (Table 1).
CLINICAL PRESENTATION
Symptoms and Signs
Pulmonary tuberculosis frequently develops slowly,
without a definite date of onset. The disease has a wide
spectrum of manifestations ranging from skin positivity
with negative X rays to far advanced tuberculosis. Or-
dinarily, until the disease is moderately or far advanced,
as shown by changes on the roentgenogram, symptoms
are minimal and often attributable to other causes, such
as excessive smoking, hard work, pregnancy, or other
conditions.
Symptoms may be divided into two categories,
constitutional and pulmonary. The frequency of these
symptoms differs according to whether the patient has
primary tuberculosis or reactivation tuberculosis. Sub-
jects with primary tuberculosis are much more likely
to be asymptomatic or minimally symptomatic. See
Table 2 for a list of the most common symptoms and
their relative frequencies in representative case series
of both primary and reactivation tuberculosis. The con-
stitutional symptom most frequently seen is fever, low
grade at the onset but becoming quite marked as
the disease progresses. Characteristically, the fever de-
velops in the late afternoon and may not be accompa-
nied by pronounced symptoms. With defervescence,
usually during sleep, sweating occurs—the classic “night
sweats.” Other signs of toxemia, such as malaise,
irritability, weakness, unusual fatigue, headache, and
TABLE 1 Increased susceptibility to tuberculosis
Nonspecific decrease in resistance
Adolescence
Senescence
Malnutrition
Postgastrectomy state
Diabetes mellitus
Renal failure
Decrease in resistance due to hormonal effects
Pregnancy
Therapy with adrenocortical steroids
Decrease in local resistance
Silicosis
Decrease in specific immunity
Lymphomas
Immunosuppressive therapy
Sarcoidosis
Live-virus vaccination
HIV infection
Transplantation
2 ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
TABLE 2 Clinical symptoms of patients presenting with
active tuberculosis
% of patients affecteda
Symptom Primary Reactivation
Cough 23–37 42
Fever 18–42 37–79
Weight loss NR 7–24
Hemoptysis 8 9
a Estimates based on several studies (49). NR, not reported.
weight loss, may be present. With the development
of caseation necrosis and concomitant liquefaction of
the caseation, the patient will usually notice cough and
sputum, often associated with mild hemoptysis. Chest
pain may be localized and pleuritic. Shortness of breath
usually indicates extensive disease with widespread in-
volvement of the lung and parenchyma or some form
of tracheobronchial obstruction and therefore usually
occurs late in the course of the disease.
Physical examination of the chest is ordinarily of
minimal help early in the disease. At this stage, the
principal finding over areas of infiltration is one of fine
rales detected on deep inspiration followed by full ex-
piration and a hard, terminal cough (posttussive rales).
This sign is found particularly in the apexes of the lungs,
where reactivation disease has its onset in a large ma-
jority of patients. As the disease progresses, more ex-
tensive findings are present, corresponding to the areas
of involvement and type of pathology. Allergic mani-
festations may occur, usually developing at the time
of onset of infection. These include erythema nodosum
and phlyctenular conjunctivitis. Erythema induratum,
involvement of the lower leg and foot with redness,
swelling, and necrosis, probably represents a combi-
nation of local subcutaneous bacterial infection with
an allergic response and should not be confused with
erythema nodosum, the latter considered to be due to
circulating immune complexes with resultant localized
vascular damage. Initially, erythema nodosum occurs in
the dependent portion of the body and, if the reaction is
severe, may be followed by a more disseminated process.
Laboratory Examination
Routine laboratory examinations are rarely helpful in
establishing or suggesting the diagnosis (11). A mild
normochromic normocytic anemia may be present in
chronic tuberculosis. The white blood cell count is often
normal, and counts over 20,000/μl would suggest an-
other infectious process; however, a leukemoid reaction
may occasionally occur in miliary tuberculosis, but not
 Pulmonary Tuberculosis

in tuberculosis confined to the chest. Although a “left

shift” in the differential white blood cell count can occur
in advanced disease, these changes are neither specific
nor useful. Other nonspecific tests that may be elevated
in active tuberculosis include the sedimentation rate, α2-
globulins, and gamma globulin. The finding of pyuria
without bacteria by Gram stain is suggestive of renal
involvement. Liver enzymes (transaminases and alka-
line phosphatase) may occasionally be elevated prior to
treatment. However, this finding is usually due to con-
comitant liver disease secondary to other problems such
as alcoholism rather than to tuberculous involvement.
Since the drugs used in the treatment of tuberculosis
are often associated with hepatotoxicity, it is important
to quantitate any hepatic abnormalities prior to treat-
ment (12). On rare occasions, the serum sodium may
be depressed owing to inappropriate secretion of anti-
diuretic hormone. This occurs only in advanced pul-
monary tuberculosis.
A positive delayed hypersensitivity reaction to tuber-
culin or gamma interferon release assay indicates only
the occurrence of a prior primary infection (13).

Chest Radiography
The chest radiograph is the single most useful study for
suggesting the diagnosis of tuberculosis. The appearance
of the radiograph differs in primary (Fig. 1, 2, and 3) and
reactivation (14) tuberculosis.
FIGURE 1 Primary tuberculosis in an adult. Shown is a right
lower lobe infiltrate with bilateral hilar adenopathy.
FIGURE 2 Left upper lobe tuberculosis. Shown is a typical
fibronodular pattern of reactivation tuberculosis with linear
densities extending to the left hilum.
Primary Tuberculosis
As opposed to reactivation tuberculosis, which usually
involves the superior and dorsal segments, in primary
tuberculosis parenchymal involvement can happen in
any segment of the lung (15). In the primary infection
there is only a slight predilection for the upper lobes;
also, anterior as well as posterior segments can be in-
volved. The air space consolidation appears as a ho-
mogeneous density with ill-defined borders (Fig. 1), and
cavitation is rare except in malnourished or other im-
munocompromised patients. Miliary involvement at the
onset is seen in less than 3% of cases; it is most com-
monly seen in children under 2 to 3 years of age but can
also be seen in adults (Fig. 4).
Hilar or paratracheal lymph node enlargement is a
characteristic finding in primary tuberculosis. In 15%
of the cases, bilateral hilar adenopathy may be present.
More commonly, the adenopathy is unilateral. Uni-
lateral hilar adenopathy and unilateral hilar and para-
tracheal adenopathy are equally common. Massive hilar
adenopathy may herald a complicated course.
Atelectasis with an obstructive pneumonia may result
from bronchial compression by inflamed lymph nodes

ASMscience.org/MicrobiolSpectrum 3
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
Lyon and Rossman

FIGURE 3 Late changes of upper lobe tuberculosis. (A) Posterior-anterior chest radio-
graph with volume loss of the right upper lobe indicated by the elevated minor fissure.
Small cavities are not clearly visible, but there is endobronchial spread to the superior
segment of the right lower lobe, suggesting cavitary formation. (B) A CT scan of the same
patient that clearly demonstrates extensive bilateral cavitary disease.

or from a caseous lymph node that ruptures into a
bronchus. Obstructive “emphysema” or a localized hy-
perinflated segment at times precedes atelectasis. The
most common segments involved are the anterior seg-
ment of the right upper lobe and the medial segment of
the right middle lobe. Right-sided collapse is twice as
common as left-sided collapse. Residual bronchiectatic
changes may persist after the obstruction has cleared.
An isolated pleural effusion of a mild to moderate
degree may be the only manifestation of primary tuber-
culosis. However, the most common radiographic ap-
pearance of primary tuberculosis is a normal radiograph.
Reactivation Tuberculosis
Although reactivation tuberculosis may involve any lung
segment, the characteristic distribution usually suggests

4 ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
 Pulmonary Tuberculosis

FIGURE 4 Miliary tuberculosis. (A) Characteristic diffuse small nodules are seen in the
posterior-anterior radiograph. (B) CT scan of the lung in the same subject demonstrates
the diffuse small nodular disease.

the disease. In 95% of cases of localized pulmonary tu-
berculosis, the lesions are present in the apical or pos-
terior segment of the upper lobes or the superior segment
of the lower lobes (Fig. 2 and 3). The anterior segment
of the upper lobe is almost never the only manifest
area of involvement (16). Although some radiologists
attempt to describe the activity of a lesion on the basis of
its radiographic appearance, the documentation of ac-
tivity is best left to bacteriological and clinical evalua-
tion (Table 3). Too often a lesion described as inactive
or stable by radiography progresses to symptomatic
tuberculosis.
The typical parenchymal pattern of reactivation tu-
berculosis is of an air space consolidation in a patchy or
TABLE 3 Criteria for activity in pulmonary tuberculosis
Symptoms
Change in roentgenogram
Evidence of cavitation
Positive sputum by smear or culture
Response to therapeutic trial
Tree-in-bud opacities on CT chest imaging
confluent nature. Frequently, there are increased linear
densities to the ipsilateral hilum (Fig. 2). Cavitation is
not uncommon, and lymph node enlargement is rarely
seen. As the lesions become more chronic, they become
more sharply circumscribed and irregular in contour.
Fibrosis leads to volume loss in the involved lung. The
combination of patchy pneumonitis, fibrosis, and calci-
fication is always suggestive of chronic granulomatous
disease, usually tuberculosis.
The cavities that develop in tuberculosis are charac-
terized by a moderately thick wall, a smooth inner sur-
face, and the lack of an air-fluid level (Fig. 3). Cavitation
is frequently associated with endobronchial spread of
disease. Radiographically, it appears as multiple small
acinar shadows.
CT Findings in Pulmonary Tuberculosis
Computed tomography (CT) scans allow practitioners
to examine both the pulmonary parenchyma and the
lymph nodes in greater detail than can be done with
plain chest X ray alone. The chest CT for patients
with primary tuberculosis typically demonstrates lobar

ASMscience.org/MicrobiolSpectrum 5
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
Lyon and Rossman
consolidation in association with mediastinal or hilar
adenopathy. While lymph node enlargement is very
common in children (95%) with primary tuberculosis,
lymph node enlargement in primary tuberculosis in
adults is less frequently seen (43%) (17). The consoli-
dation is usually well defined, dense, homogenous,
and confined to a segment or lobe. Middle and lower
lobe involvement is very common. Small cavities may be
appreciated on the CT that were not seen on plain chest
X ray (Fig. 3).
In reactivation tuberculosis, the apical and posterior
segments of the upper lobe and the superior segment
of the lower lobe are most commonly involved. Cavi-
tation is seen in over 50% of cases. The cavitation may
be multiple and usually involves a thick wall without
an air-fluid level. Cavitation is usually associated with
bronchogenic spread of the disease. This is radiograph-
ically seen as multiple ill-defined 5- to 10-mm nodules
that usually involve the dependent lung zone (17).
Centrilobular nodules or branching linear structures
(“tree in bud”) with or without bronchial wall thicken-
ing can be appreciated on thin sections. Controversy
exists over whether CT scans can reliably distinguish
active tuberculosis from latent infection, with several
authors arguing that findings such as the tree in bud
and/or areas of centrilobular nodules predict active dis-
ease (18, 19).
DIAGNOSIS
The diagnosis of tuberculosis often can be very difficult.
Some of the problems that occur are listed in Table 4. A
firm diagnosis of tuberculosis requires bacteriological
confirmation. It is important to remember that a positive
acid-fast smear is not specific for Mycobacterium tu-
berculosis. Other mycobacteria, both saprophytes and
potential pathogens, can be acid fast. Thus, culture of
M. tuberculosis is the only absolute way of confirming
the diagnosis.
Freshly expectorated sputum is the best sample to
stain and culture for M. tuberculosis. Sputum samples
24 h old are frequently overgrown with organisms of
TABLE 4 Diagnostic difficulties
Lack of organisms for culture
Slow growth of culture
Chest X-ray findings absent or misinterpreted
Biopsy material may not be specific
Decreased tuberculin sensitivity
Symptoms and signs of tuberculosis easily attributed to a preexisting
disease
6 ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
the mouth flora and are much less useful. If the patient is
not spontaneously producing sputum, induced sputum
is the next best specimen for study. It can be obtained
by having the patient breathe an aerosol of isotonic or
hypertonic saline for 5 to 15 min. If the patient cannot
cooperate to give a spontaneous sputum sample, a gas-
tric aspirate to obtain swallowed sputum may be useful.
This sample must be obtained in the morning before the
patient arises or eats.
For the majority of patients, the above-mentioned
procedures are successful in obtaining positive mate-
rial for culture. Smears of gastric contents for acid-fast
bacilli are of limited value because of the presence of
nontuberculous ingested acid-fast bacilli. In a few cases,
one may have to resort to bronchoscopy. For 41 pa-
tients proven to have tuberculosis, cultures of specimens,
taken during fiber-optic bronchoscopy, were positive in
39 cases (18, 19). Stainable mycobacteria were seen in
14 of the cases, and in 8 cases, granulomas were seen on
biopsy. Similar results have been obtained in another
study of 22 patients with proven mycobacterial dis-
ease and negative smears prior to bronchoscopy (20).
The local anesthetics used during fiber-optic broncho-
scopy may be lethal to M. tuberculosis, so specimens
for culture should be obtained using a minimal amount
of anesthesia. However, irritation of the bronchial tree
during the fiber-optic bronchoscopy procedure fre-
quently leaves the patient with a productive cough.
Thus, collection of the postbronchoscopy sputum can
provide another valuable source of diagnostic material.
In nine (13%) of the above-mentioned cases, the post-
bronchoscopy sputum was the only source of positive
material.
Nucleic acid amplification (NAA) testing can be
used for rapid diagnosis of M. tuberculosis complex in
respiratory specimens. Current CDC recommendations
suggest that at least one respiratory sample be tested
using NAA testing for smear-negative patients in whom
active pulmonary tuberculosis is considered (4). In
smear-positive samples, NAA testing can be used to dif-
ferentiate M. tuberculosis from nontuberculous myco-
bacteria and has a positive predictive value of more than
95% (21).
However, if the NAA test is positive but the smear
is negative, clinical judgment must be used in interpret-
ing the test. A recent review of Xpert MTB/RIF, loop-
mediated isothermal amplification, and simultaneous
amplification testing methods found a sensitivity and
specificity of 98% and 68% in the smear-positive sub-
group but only a sensitivity and specificity of 72%
and 93% in the smear-negative subgroup (22). Impor-

tantly, a negative NAA result is insufficient to exclude
a diagnosis of pulmonary tuberculosis (4). Both posi-
tive and negative NAA results should be evaluated
within the individual clinical context and local labora-
tory capabilities.
In 2014 (3) in the United States, sputum culture con-
firmed the diagnosis in 77% of cases. In 16% of cases,
the diagnosis was confirmed by a clinical response to
therapy. Thus, in a significant number of cases, the di-
agnosis of tuberculosis is made in the absence of bacte-
riological confirmation.
Differential Diagnosis
Since tuberculosis today is a disease frequently present
in older individuals, as well as immigrants, one major
differential diagnosis is usually between tuberculosis
and carcinoma of the lung. An important concept to
remember is that carcinoma may cause a focus of tu-
berculosis to spread; thus, carcinoma of the lung and
tuberculosis may be present simultaneously. In cases
with the simultaneous presentation of carcinoma and
tuberculosis, the diagnosis of tuberculosis frequently is
made first, and the diagnosis of carcinoma is delayed
for several months. Thus, if radiograph and clinical
findings suggest carcinoma but the sputum has acid-fast
bacilli, further procedures to diagnose carcinoma may
still be indicated. Isolated involvement of the anterior
segment of the upper lobe, isolated lower lobe involve-
ment, or the presence of irregular cavities would sug-
gest carcinoma, and further diagnostic workup may be
indicated despite acid-fast bacilli in the sputum smear.
Any type of infectious or granulomatous disease
may be radiographically identical to tuberculosis.
Three broad categories of infectious disease must be
distinguished: those involving fungi (histoplasmosis,
coccidioidomycosis, and blastomycosis), bacteria (Pseu-
domonas pseudomallei), and atypical mycobacteria
(mainly Mycobacterium kansasii and Mycobacterium
intracellulare). Culture of the organism from the pa-
tient’s sputum is the best way to differentiate these dis-
eases, although titers of serum antibody to fungi are
also valuable. Common bacterial pneumonias are usu-
ally easily differentiated from tuberculosis. The local-
ized alveolar infiltrate on the chest radiograph and the
prompt response to antibiotic therapy usually differ-
entiate bacterial pneumonia from tuberculosis. When
in doubt, treatment for a bacterial pneumonia should
be given first and tuberculosis therapy withheld until
adequate sputum samples have been obtained and the
response to antibiotics determined. Lung abscesses can
usually be differentiated from tuberculous cavities by
ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
Pulmonary Tuberculosis
(i) prominent air-fluid level, (ii) more common lower
lobe distribution, and (iii) clinical findings (i.e., associ-
ated with seizures, alcoholism, dental caries, etc.).
While sarcoidosis is a noninfectious granulomatous
disease which can present with radiographic similari-
ties to tuberculosis, it does not typically cause cavitary
disease and is distinguished from tuberculosis on histo-
pathology by the presence of noncaseating granulomas
and the absence of acid-fast bacilli and negative culture
results.
TUBERCULOSIS AND AIDS
HIV infection led to the resurgence of tuberculosis in the
United States as well as internationally (23).
Several important differences have emerged about the
clinical presentation of subjects with tuberculosis with
and without HIV infection. As previously mentioned,
patients with HIV infection are more likely to present
with disseminated disease. Additionally, they tend to
have an increased number and severity of symptoms and
have a more rapid progression to death unless treatment
is begun (24–27).
Radiographic findings of tuberculosis in the context
of HIV infection have been found to correlate with the
degree of immunosuppression due to HIV itself (5).
Lower CD4 counts (i.e., >200/mm3) are associated more
often with hilar and mediastinal lymphadenopathy,
while higher CD4 counts are more frequently associated
with cavitation. Additionally, some studies have sug-
gested that HIV-infected subjects are more likely to have
nonapical infiltrates, pleural effusions, and miliary in-
filtrates (6).
In patients with HIV infection but with CD4 counts
greater than 300, the tuberculin skin test will be positive
in 50 to 80% of those with active tuberculosis. Once an
individual has developed AIDS, the tuberculin skin test
will be less likely to be positive, but reactivity may be
seen in as many as 30 to 50% of patients.
While in areas of low tuberculosis prevalence, a
positive tuberculosis skin test can support the diagnosis
of active disease in those with symptoms, tuberculosis
skin testing and interferon gamma release assays are
testing modalities for latent tuberculosis. These tests
are frequently negative in the setting of active tubercu-
losis disease, even in the absence of a chronic immuno-
suppressing condition, and have a minimal role in the
evaluation for active tuberculosis. Tuberculin skin test-
ing and interferon gamma release assays should not
be used to either make or exclude a diagnosis of active
M. tuberculosis infection in any patient population.
7
Lyon and Rossman
Patients with HIV are more likely to have smear-
negative pulmonary tuberculosis (28).Whenever an acid-
fast organism is identified, the assumption must be that
the organism is M. tuberculosis and treatment should be
initiated until definitive identification of the organism
occurs. DNA probes can be useful to distinguish between
M. tuberculosis and nontuberculous mycobacteria.
TUBERCULOSIS IN THE ELDERLY
Age is a risk factor for tuberculosis disease, and elderly
patients are also at increased risk of other comorbidities
such as diabetes mellitus and renal disease, which in-
creases the risk of tuberculosis. Some studies have begun
to suggest that not only is increasing age a risk factor
for the development of active tuberculosis (7) but also
the disease itself may present differently in the elderly,
making it harder to recognize and therefore diag-
nose. One study prospectively examined 93 consecutive
patients over the age of 60 admitted with pulmonary
tuberculosis to a hospital in South Africa (29). Among
these patients, “atypical” radiographic findings were the
norm rather than the exception. For instance, only 7%
had purely apical infiltrates, while 48% had middle and
lower lung zones only, and 46% had mixed infiltrates
between the upper and lower lung fields. A pleural re-
action was common (46%) and cavities were not (33%),
with half of those seen in the lower and middle lung
fields. In addition, the investigators found that systemic
abnormalities of routine blood work were common, in-
cluding anemia (66%), elevated erythrocyte sedimentation
rates (90%), hyponatremia (60%), and hypoalbuminemia
(83%). However, not all studies have confirmed these
differences (30). One recent meta-analysis of 12 studies of
tuberculosis found that the elderly were less likely to have
symptoms such as fever, sweating, hemoptysis, and cavi-
tary lung disease. They were more likely to have dyspnea
and significant comorbidities (31). In this study, the only
differences seen in radiographic patterns between young
adults and the elderly was an increased incidence of mili-
ary disease in the older population.
TUBERCULOSIS IN PATIENTS
TAKING TNF INHIBITORS
Tumor necrosis factor alpha (TNF-α) is a cytokine
which acts as a central mediator of inflammation and
immune regulation. Inhibition of TNF-α is now used for
the treatment of several diseases, including rheumatoid
arthritis, inflammatory bowel disease, juvenile rheuma-
toid arthritis, ankylosing spondylitis, and psoriatric
8 ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
arthritis. The following anti-TNF agents are currently
approved by the FDA: infliximab, a chimeric (human
and murine) monoclonal anti-TNF alpha antibody;
etanercept, a TNF-α antagonist; certolizumab pegol,
a pegylated Fab fragment of humanized monoclonal
antibody; and adalimumab and golimumab, human
anti-TNF monoclonal antibodies.
All TNF-α antagonists have the potential to predis-
pose patients to granulomatous infections (26). In reg-
istry studies, the monoclonal antibodies infliximab and
adalimumab carry a greater risk than etanercept, which
was felt to be due to their induction of complement-
mediated cell lysis. If, for example, lysis of a macro-
phage involved in the formation of a granuloma occurs,
this could destabilize the granuloma, resulting in dis-
semination of the infection. After FDA approval of in-
fliximab in 1998, an increasing frequency of tuberculosis
infections was noted in postmarketing surveillance. In
one study, 70 reported cases of tuberculosis after the
initiation of infliximab therapy were analyzed (27).
Forty-eight of the patients developed tuberculosis after
three or fewer infusions, with a median interval of 12
weeks. Forty had extrapulmonary disease, and 17 had
disseminated tuberculosis. Of these 70 patients, 12 died,
and at least four of these deaths were thought to be
directly related to tuberculosis infection. During the
same period, only nine cases of tuberculosis during the
treatment with etanercept were reported to the FDA.
However, by 2002, 25 cases of tuberculosis in patients
receiving etanercept were described. A total of 52% of
these, or 13, involved extrapulmonary disease, and 1 pa-
tient died secondary to the infection. Several tuberculosis
cases attributed to high-dose adalimumab therapy have
now also been reported.
Data on newer agents are limited. Several cases
of tuberculosis were seen in the treatment arm of clini-
cal trials of certolizumab pegol, while none occurred
in controls (32). Several important points have been
learned from these reports. TNF-α antagonists do in-
crease the incidence of active tuberculosis. Patients who
develop tuberculosis on TNF-α antagonist therapy usu-
ally do so after only 2 to 3 months of therapy and have
more extrapulmonary and disseminated disease than
do immunocompetent, HIV-negative patients. This may
reflect either a delay in the diagnosis of tuberculosis
or atypical presentations of tuberculosis in this subgroup
of patients. Current recommendations include screening
all patients with skin testing for tuberculosis and treat-
ment of any latent tuberculosis (more than 5 mm of
induration on purified protein derivative [PPD]) prior
to initiation of TNF-α antagonist therapy (19, 26). For

patients undergoing consideration for TNF-α antagonist
therapy who have been exposed to Mycobacterium
bovis BCG vaccination, a positive tuberculosis skin test
and a negative interferon gamma release assay may or
may not reflect a false-positive skin test. Patients with
either a positive skin test or a positive interferon gamma
release assay, as well as those with evidence of prior
tuberculosis infection on chest radiography or known
prior close tuberculosis contact who have negative
laboratory testing, should be viewed as being at high
risk for latent tuberculosis infection. Clinicians should
strongly consider initiation of treatment for latent tu-
berculosis infection in these patients prior to starting
TNF-α antagonist therapy. At least 1 month of treatment
for latent tuberculosis infection is recommended prior
to TNF-α antagonist initiation (32).
PLEURAL EFFUSIONS DUE
TO TUBERCULOSIS
Pleural effusion is a relatively uncommon manifestation,
particularly of primary tuberculosis, occurring in only
3% of clinical cases. Those with HIV coinfection appear
to have a higher rate of pleural disease (33). Tuberculous
pleural effusions are almost always due to rupture of
subpleural foci of tuberculosis, which may not be evi-
dent radiographically. The effusions in tuberculosis are
unilateral and mild to moderate in extent. The presence
of bilateral effusions in tuberculosis usually means a
miliary spread. The natural course of a tuberculous
pleural effusion is to gradually resorb and frequently
disappear completely or with minimal changes on the
chest radiograph.
Tuberculous pleural effusions must be differentiated
from effusions due to congestive heart failure, carci-
noma, and other types of infections. Pleural fluid protein
is most useful for differentiating tuberculous effusions
from transudates (34). Almost without fail, the pleural
fluid protein in tuberculosis will be greater than 4 g/dl
(exudate), whereas it is most unusual for congestive
heart failure fluid to have protein levels this high (tran-
sudate). The differentiation of carcinomatous from tu-
berculous effusions is more difficult. Both may appear
exudative, with high levels of lactic dehydrogenase and
protein in the pleural fluid. In tuberculous effusions,
the differential count of the cells in the pleural fluid
usually does not contain any mesothelial cells. A low
pleural fluid glucose (less than 30 mg/dl) is common in
tuberculosis and rare in carcinoma. Similarly, elevated
adenosine deaminase is frequently found in tuberculous
effusions but is rare in carcinoma (35). Pleural biopsies
ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
Pulmonary Tuberculosis
are recommended in cases where pleural fluid evalua-
tion is nondiagnostic.
It is usually easy to differentiate tuberculous effusion
from bacterial effusions, since bacterial effusions usually
contain a predominance of neutrophils, whereas tuber-
culous effusions are predominantly lymphocytic (36,
37). However, early in the course of tuberculous effu-
sions, neutrophils may be seen. A Gram stain of the pleu-
ral fluid and a culture of the fluid, sputum, and blood will
usually establish the etiologic agent of bacterial effu-
sions. More difficult is the differentiation of a viral pleu-
ral effusion from an effusion due to tuberculosis. When
patients have a positive 5-tuberculin unit tuberculin test
(induration greater than 10 mm), their effusions should
be presumed to be tuberculous until proven other-
wise. If a patient has an undiagnosed exudative effusion
and a negative tuberculin test, the tuberculin test should
be repeated within 2 weeks, since it is not uncommon
for patients with tuberculous effusions to have an ini-
tially negative tuberculin test (38). Pleural biopsies and
mycobacterial cultures of the pleural fluid and biopsy
should be performed in all cases. NAA testing of pleural
effusion in HIV-uninfected patients with a moderate to
high suspicion of tuberculosis infection has a high spec-
ificity but a low sensitivity (39, 40). Of note, NAA testing
of pleural fluid and other nonrespiratory secretions is
not approved by the FDA and is considered “off-label.”
The diagnosis of pleural effusions due to tuberculosis
can be accomplished by appropriate studies of the pleura
and fluid in approximately 80% of cases (14, 41). These
studies involve evaluation of the character of the fluid,
which is an exudate with a prominent lymphocytosis.
Smears of the fluid usually are negative for tubercle
bacilli, but positive cultures are found in more than half
of the cases. Repeated thoracenteses with culture of large
quantities of fluid combined with centrifugation may
also increase bacteriological yield. When one combines
the histological examination and culture of the pleural
biopsy specimen with study of the fluid, one has the
highest rate of diagnosis (80%). Pleural biopsy demon-
strates granulomas in 50 to 97% of cases, and cultures
are positive in 40 to 80% of cases. Caseating granulo-
mas even in the absence of acid-fast bacilli on smear or
culture are considered adequate for diagnosis of pleural
tuberculosis (42). Noncaseating granulomas can be seen
on histologic examination of the pleura in sarcoidosis,
as well as fungal infections and rheumatoid pleuritis.
For each undiagnosed pleural effusion, in addition to
the studies for tuberculosis, studies for malignant cells,
fungi, and bacteria should be performed on the aspirated
fluid and any biopsy material. At times, a video-assisted
9
Lyon and Rossman

or standard thoracotomy with pathologic examination
of the pleura for tuberculosis, fungi, malignant cells, and
bacteria may be necessary to establish the cause of a
pleural effusion (43).
In the absence of a diagnosis and the presence of
a compatible pleural effusion, consideration should be
given to a trial of chemotherapy on the basis of a pre-
sumptive diagnosis of tuberculosis. Many patients with
an undiagnosed pleural effusion later develop progres-
sive parenchymal tuberculosis if they are not treated.
However, a therapeutic trial in pleural tuberculosis is
not as helpful diagnostically as one in pulmonary tu-
berculosis, since the natural course of pleural tubercu-
losis is toward resolution.
However, in contrast to tuberculous pleurisy, which
does not require surgery, tuberculous empyema is usu-
ally accompanied by a thick pleural peal and requires
surgical drainage or decortication (Fig. 5) in addition to
antituberculosis therapy.
bations and remissions, it is important to determine if the
disease is “healed,” quiescent, or progressive. Decisions
concerning infectiousness and the need for chemotherapy
depend on this evaluation. The bases for these decisions
are (i) clinical signs of infection (fever, weight loss, cough,
sputum, etc.), (ii) progressive X-ray changes, and (iii) a
positive sputum smear or culture. An improving X-ray
study is also presumed to represent prior active tuber-
culosis. In the appropriate setting, the presence of any
one of these findings is an indication for full therapy.
Therapy for pulmonary or pleural tuberculosis is
discussed elsewhere. Corticosteroids are used as an ad-
junct to specific chemotherapy only in the most severe
cases of active pulmonary tuberculosis. In the patient
who is in danger of dying from tuberculosis, cortico-
steroids can be lifesaving by causing a rapid deferves-
cence, symptomatic improvement, and weight gain.
However, the routine use of corticosteroids has been
shown to have no effect on the late effects of pulmonary
or pleural tuberculosis.

ACTIVITY Predicting Who Has Active Tuberculosis

Table 3 lists criteria for activity in tuberculosis. Since Several recent studies have focused on predicting
tuberculosis is a chronic disease with multiple exacer- which subjects admitted to the hospital with suspicion

FIGURE 5 Tuberculous empyema. Posterior-anterior (A) and lateral (B) chest radiographs
demonstrate a left lower lobe effusion.

10 ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
 Pulmonary Tuberculosis

of tuberculosis should be placed immediately in respi- on exam (protective), and upper lobe consolidation
ratory isolation versus those who may not need it, given into a single score that could be used to determine who
its high cost. Initial efforts to stratify subjects on the should be isolated. In their derivation cohort, a cutoff
basis of risk factors had inadequate sensitivity from value of 1 or higher could be used to determine who
a public health perspective (44). However, two groups to isolate (98% sensitivity and 46% specificity). When
have published prediction rules that may be more clini- prospectively applied to a new cohort, the model had
cally useful (Table 5). The first assigned points were similar sensitivity (95%) and specificity (35%) and would
based on immigration status, history of BCG immu- have avoided unnecessary respiratory isolation for 35%
nization, HIV status, homelessness, compatible clinical of patients (47).
symptoms, and compatible chest X rays to define a risk
score (45). In their derivation cohort, this scoring sys-
tem had a sensitivity of 100% and a negative predictive COMPLICATIONS
value of 100%. When this scoring system was tested in Although a relatively uncommon complication of tuber-
two other retrospective groups, the sensitivity was not culous infection, the development of a pneumothorax
as good (91%) but was within an acceptable range. requires rapid attention. One of the postulated theories
A second scoring system was developed at New York of etiology is the rupture of a cavity that then connects
University (46). This system factored in tuberculosis risk the tracheobronchial tree with the pleural space, creating
factors (recent immigration, institutionalization, and a bronchopleural fistula. In this occurrence, contami-
history of tuberculosis exposure) and chronic symptoms nation of the pleural space with caseous material results
(weight loss, malaise, weakness and night sweats for in spread of the infection to the pleura and should be
3 or more months), positive PPD, shortness of breath corrected immediately because of the tendency to pro-
(protective), low- or high-grade temperatures, crackles duce pleural fibrosis with expansion failure.
A second possible mechanism is the development
TABLE 5 Predicting active pulmonary tuberculosis
of a submucosal bronchiolar lesion with air trapping in
an acinus or subsegment that causes the development of
No. of a bleb. Rupture of this bleb allows air to enter the pleu-
Study Risk factor points ral space, but often without tuberculous infection of the
Tattevin et al. (45) Immigrant from Eastern or Southern 1 pleura. However, both occurrences should be treated
Europe, South America, or French
Guyana with rapid expansion of the lungs by tube suction to
BCG immunization >10 yr earlier 1 avoid the possibility of further infection and fibrosis of
No HIV infection 5 the pleura with trapping of the lung. A bronchopleural
Homelessness 1
Compatible symptoms 6
fistula may persist after these episodes of pneumothorax
Compatible chest X ray 7 and, especially if untreated, often results in major prob-
Immigrant from sub-Saharan Africa, 2 lems owing to the tuberculous infection complicated by
North Africa, Haiti, Southeast Asia
No BCG 2
secondary invaders (“mixed” empyema).
No HIV infection 5 Minor endobronchial disease is a common occurrence
Homelessness 1 in tuberculosis but usually involves the distal bronchi.
Typical clinical symptoms 12
Typical chest X ray 14
Resected lung specimens frequently show either ulcera-
Total, 100% sensitivity ≥18 tion or stenosis of the draining bronchioles or bron-
Wisnivesky et al. (46) TB risk factorsa or chronic 4 chi. Bronchial stenosis of significance may occur in the
symptomsb major bronchi but is rare. At times, it results from in-
Positive PPD 5
Shortness of breath −3
volvement of the central lymph nodes draining into the
Temp (°C) lobar bronchi, with caseation, ulceration, and fibrosis.
<38.5 0 Since fibrosis due to tuberculosis tends to contract and
38.5–39.0 3
>39.0 6
aggravate the stenosis, resection of the involved lung
Crackles noted during examination −3 segment may be required after chemotherapy has pro-
Upper lobe consolidation 6 duced inactivity of the acute inflammatory reaction.
Total, 98% sensitivity 1 The same endobronchial processes may result in
aTuberculosis (TB) risk factors include recent immigration, recent institutionali- bronchiectasis due to destruction of the bronchial wall.
zation, and known TB exposure.
bTB chronic symptoms include weight loss, malaise, weakness, and/or night
This usually is distal and frequently is in the upper lobes.
sweats for 3 or more months. The so-called “dry” bronchiectasis (without sputum)

ASMscience.org/MicrobiolSpectrum 11
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
Lyon and Rossman
often is the result of prior pulmonary tuberculosis and
may manifest itself chiefly as low-grade hemoptysis.
Empyema due to tuberculosis may result uncom-
monly from a primary infection with an associated tu-
berculous pleural effusion. However, the latter usually
clears; empyema is more common later in the disease,
associated with debility and loss of resistance to infection
(Fig. 5). It is usually a part of a progressive, extensive
parenchymal infection with caseation and cavitation, the
presumed sources for pleural contamination.
After treatment of extensive tuberculosis, the patient
is often left with open, healed cavities as well as with
areas of bronchiectasis. Colonization of these areas may
occur with a variety of infectious agents. Usual oro-
respiratory flora may produce the syndrome of “wet”
bronchiectasis, i.e., with sputum production. Other my-
cobacteria may be recovered during the development of
inactivity and were at one time considered to be a sign of
healing. The presence of other pathogenic mycobacteria
brings up the possibility of a dual infection, especially
when found early in the disease.
Aspergillus species are common in badly damaged
lung areas, especially those that are cavitary. In England,
a prospective study (48) revealed that 25% of clinically
healed tuberculosis patients who had residual cavities
developed positive precipitins to Aspergillus species
and 11% had demonstrable cavitary “balls,” presumed
to be aspergillomas, or “fungus balls.” Three years later,
these numbers had risen to 34% and 17%, respectively.
This high incidence may be due in part to the increased
incidence of Aspergillus noted in the United Kingdom,
both in the environment and as an infective agent,
probably as a result of the more humid environment.
Massive hemorrhage, a dramatic event occurring in
advanced cases of tuberculosis, is frequently terminal.
Mild hemoptysis itself is very common in acute in-
fection and not infrequently calls the attention of an
otherwise unconcerned patient to the presence of serious
disease. Rupture of a mycotic aneurysm of a branch of
the pulmonary artery (Rasmussen’s aneurysm) has been
well publicized as a cause of death; an aspergilloma
may be associated with severe and fatal hemorrhage.
However, less well-defined major hemorrhages may also
occur.
Resection of the involved area has been the most
widely used method of control; unhappily, many pa-
tients die before this can be accomplished, and often
(as in the case of aspergillomas) the areas are multiple,
thus not lending themselves to excisional therapy. The
extensive disease found in these patients often contra-
indicates surgery, since functional lung tissue necessary
12
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
for survival must often be removed along with the dis-
eased area at the time of surgery.
During the acute infectious phase of the disease,
two interesting complications have been reported, the
syndrome of inappropriate antidiuretic hormone excre-
tion (SIADH) and a reset osmostat (49). Both manifest
themselves by abnormally low sodium. However, the
former is associated with all of the clinical and renal
abnormalities associated with SIADH. A reset osmostat
is characterized by decreased serum osmolality without
clinical symptoms and the obligatory renal salt wasting
found in SIADH (50). Both conditions disappear with
control of the infection; however, they should be dif-
ferentiated from each other since SIADH requires met-
abolic control.
Septic shock is an uncommon complication of pul-
monary tuberculosis infection, although it is seen more
frequently in those with HIV coinfection and is associ-
ated with a high mortality rate (51).
REFERENCES
1. Farer LS, Lowell AM, Meador MP. 1979. Extrapulmonary tuberculosis
in the United States. Am J Epidemiol 109:205–217.
2. Arango L, Brewin AW, Murray JF. 1973. The spectrum of tuberculosis as
currently seen in a metropolitan hospital. Am Rev Respir Dis 108:805–812.
3. CDC. 2015. Reported Tuberculosis in the United States, 2014. De-
partment of Health and Human Services, CDC, Atlanta, GA. https://www
.cdc.gov/tb/statistics/reports/2014/.
4. Centers for Disease Control and Prevention. 2009. Updated guidelines
for the use of nucleic acid amplification tests in the diagnosis of tubercu-
losis. MMWR Morb Mortal Wkly Rep 58:7–10.
5. Shafer RW, Kim DS, Weiss JP, Quale JM. 1991. Extrapulmonary tu-
berculosis in patients with human immunodeficiency virus infection.
Medicine (Baltimore) 70:384–397.
6. Perlman DC, el-Sadr WM, Nelson ET, Matts JP, Telzak EE, Salomon
N, Chirgwin K, Hafner R, The Terry Beirn Community Programs for
Clinical Research on AIDS, The AIDS Clinical Trials Group. 1997. Var-
iation of chest radiographic patterns in pulmonary tuberculosis by degree
of human immunodeficiency virus-related immunosuppression. Clin In-
fect Dis 25:242–246.
7. Shafer RW, Edlin BR. 1996. Tuberculosis in patients infected with
human immunodeficiency virus: perspective on the past decade. Clin In-
fect Dis 22:683–704.
8. Small PM, Schecter GF, Goodman PC, Sande MA, Chaisson RE,
Hopewell PC. 1991. Treatment of tuberculosis in patients with advanced
human immunodeficiency virus infection. N Engl J Med 324:289–294.
9. Glassroth J, Robins AG, Snider DE, Jr. 1980. Tuberculosis in the 1980s.
N Engl J Med 302:1441–1450.
10. Mayock RL, MacGregor RR. 1976. Diagnosis, prevention and early
therapy of tuberculosis. Dis Mon 22:1–60.
11. Farer LS, Lowell AM, Meador MP. 1979. Extrapulmonary tubercu-
losis in the United States. Am J Epidemiol 109:205–217.
12. MacGregor RR. 1975. A year’s experience with tuberculosis in a
private urban teaching hospital in the postsanatorium era. Am J Med
58:221–228.
13. Garibaldi RA, Drusin RE, Ferebee SH, Gregg MB. 1972. Isoniazid-
associated hepatitis. Report of an outbreak. Am Rev Respir Dis 106:357–
365.
ASMscience.org/MicrobiolSpectrum

14. Holden M, Dubin MR, Diamond PH. 1971. Frequency of negative
intermediate-strength tuberculin sensitivity in patients with active tuber-
culosis. N Engl J Med 285:1506–1509.
15. Fraser RG, Pare JAP. 1977. Diagnosis of diseases of the chest, p 731–
764. WB Saunders, Philadelphia, PA.
16. Weber AL, Bird KT, Janower ML. 1968. Primary tuberculosis in
childhood with particular emphasis on changes affecting the tracheo-
bronchial tree. Am J Roentgenol Radium Ther Nucl Med 103:123–132.
17. Cardinale L, Parlatano D, Boccuzzi F, Onoscuri M, Volpicelli G, Veltri
A. 2015. The imaging spectrum of pulmonary tuberculosis. Acta Radiol
56:557–564.
18. Lee KS, Hwang JW, Chung MP, Kim H, Kwon OJ. 1996. Utility of CT
in the evaluation of pulmonary tuberculosis in patients without AIDS.
Chest 110:977–984.
19. Hatipoğlu ON, Osma E, Manisali M, Uçan ES, Balci P, Akkoçlu A,
Akpinar O, Karlikaya C, Yüksel C. 1996. High resolution computed to-
mographic findings in pulmonary tuberculosis. Thorax 51:397–402.
20. Wallace JM, Deutsch AL, Harrell JH, Moser KM. 1981. Bronchos-
copy and transbronchial biopsy in evaluation of patients with suspected
active tuberculosis. Am J Med 70:1189–1194.
21. Moore DF, Guzman JA, Mikhail LT. 2005. Reduction in turnaround
time for laboratory diagnosis of pulmonary tuberculosis by routine use of
a nucleic acid amplification test. Diagn Microbiol Infect Dis 52:247–254.
22. Yan L, Xiao H, Zhang Q. 2016. Systematic review: comparison of
Xpert MTB/RIF, LAMP and SAT methods for the diagnosis of pulmonary
tuberculosis. Tuberculosis (Edinburgh) 96:75–86.
23. CDC. 2004. Reported Tuberculosis in the United States, 2003. US
Department of Heatlh and Human Services, CDC, Atlanta, GA.
24. Chaisson RE, Schecter GF, Theuer CP, Rutherford GW, Echenberg
DF, Hopewell PC. 1987. Tuberculosis in patients with the acquired
immunodeficiency syndrome. Clinical features, response to therapy, and
survival. Am Rev Respir Dis 136:570–574.
25. Handwerger S, Mildvan D, Senie R, McKinley FW. 1987. Tuber-
culosis and the acquired immunodeficiency syndrome at a New York City
hospital: 1978–1985. Chest 91:176–180.
26. Louie E, Rice LB, Holzman RS. 1986. Tuberculosis in non-Haitian
patients with acquired immunodeficiency syndrome. Chest 90:542–545.
27. Selwyn PA, Hartel D, Lewis VA, Schoenbaum EE, Vermund SH,
Klein RS, Walker AT, Friedland GH. 1989. A prospective study of the risk
of tuberculosis among intravenous drug users with human immunodefi-
ciency virus infection. N Engl J Med 320:545–550.
28. Reid MJ, Shah NS. 2009. Approaches to tuberculosis screening and
diagnosis in people with HIV in resource-limited settings. Lancet Infect
Dis 9:173–184.
29. Stead WW, Lofgren JP. 1983. Does the risk of tuberculosis increase in
old age? J Infect Dis 147:951–955.
30. Morris CDW. 1989. The radiography, haematology and biochemistry
of pulmonary tuberculosis in the aged. Q J Med 71:529–536.
31. Korzeniewska-Kosela M, Krysl J, Müller N, Black W, Allen E,
FitzGerald JM. 1994. Tuberculosis in young adults and the elderly. A
prospective comparison study. Chest 106:28–32.
32. Furst DE, Breedveld FC, Kalden JR, Smolen JS, Burmester GR, Sieper
J, Emery P, Keystone EC, Schiff MH, Mease P, van Riel PL, Fleischmann
R, Weisman MH, Weinblatt ME. 2007. Updated consensus statement
on biological agents for the treatment of rheumatic diseases, 2007. Ann
Rheum Dis 66(Suppl 3):iii2–iii22.
ASMscience.org/MicrobiolSpectrum
Downloaded from www.asmscience.org by
IP: 128.122.230.132
On: Mon, 13 Feb 2017 04:24:51
Pulmonary Tuberculosis
33. Frye MD, Pozsik CJ, Sahn SA. 1997. Tuberculous pleurisy is more
common in AIDS than in non-AIDS patients with tuberculosis. Chest
112:393–397.
34. Pérez-Guzmán C, Vargas MH, Torres-Cruz A, Villarreal-Velarde H.
1999. Does aging modify pulmonary tuberculosis? A meta-analytical re-
view. Chest 116:961–967.
35. Light RW, Macgregor MI, Luchsinger PC, Ball WC, Jr. 1972. Pleural
effusions: the diagnostic separation of transudates and exudates. Ann
Intern Med 77:507–513.
36. van Keimpema AR, Slaats EH, Wagenaar JP. 1987. Adenosine de-
aminase activity, not diagnostic for tuberculous pleurisy. Eur J Respir Dis
71:15–18.
37. Valdés L, San José E, Alvarez D, Valle JM. 1996. Adenosine deami-
nase (ADA) isoenzyme analysis in pleural effusions: diagnostic role, and
relevance to the origin of increased ADA in tuberculous pleurisy. Eur
Respir J 9:747–751.
38. Yam LT. 1972. Diagnostic significance of lymphocytes in pleural
effusions. Am Rev Respir Dis 105:458–460.
39. Trajman A, da Silva Santos Kleiz de Oliveira EF, Bastos ML, Belo
Neto E, Silva EM, da Silva Lourenço MC, Kritski A, Oliveira MM. 2014.
Accuracy of polymerase chain reaction for the diagnosis of pleural tu-
berculosis. Respir Med 108:918–923.
40. Denkinger CM, Schumacher SG, Boehme CC, Dendukuri N, Pai M,
Steingart KR. 2014. Xpert MTB/RIF assay for the diagnosis of extra-
pulmonary tuberculosis: a systematic review and meta-analysis. Eur
Respir J 44:435–446.
41. Berger HW, Mejia E. 1973. Tuberculous pleurisy. Chest 63:88–92.
42. Gopi A, Madhavan SM, Sharma SK, Sahn SA. 2007. Diagnosis and
treatment of tuberculous pleural effusion in 2006. Chest 131:880–889.
43. Scerbo J, Keltz H, Stone DJ. 1971. A prospective study of closed
pleural biopsies. JAMA 218:377–380.
44. Bock NN, McGowan JE, Jr, Ahn J, Tapia J, Blumberg HM. 1996.
Clinical predictors of tuberculosis as a guide for a respiratory isolation
policy. Am J Respir Crit Care Med 154:1468–1472.
45. Tattevin P, Casalino E, Fleury L, Egmann G, Ruel M, Bouvet E. 1999.
The validity of medical history, classic symptoms, and chest radiographs
in predicting pulmonary tuberculosis: derivation of a pulmonary tuber-
culosis prediction model. Chest 115:1248–1253.
46. Wisnivesky JP, Kaplan J, Henschke C, McGinn TG, Crystal RG.
2000. Evaluation of clinical parameters to predict Mycobacterium tu-
berculosis in inpatients. Arch Intern Med 160:2471–2476.
47. Wisnivesky JP, Henschke C, Balentine J, Willner C, Deloire AM,
McGinn TG. 2005. Prospective validation of a prediction model for iso-
lating inpatients with suspected pulmonary tuberculosis. Arch Intern Med
165:453–457.
48. Research Committee of the British Thoracic and Tuberculosis Asso-
ciation. 1970. Aspergilloma and residual tuberculous cavities—the results
of a resurvey. Tubercle 51:227–245.
49. Mayock RL, Goldberg M. 1964. Metabolic Considerations in Disease
of the Respiratory System. WB Saunders, Philadelphia, PA.
50. Chung DK, Hubbard WW. 1969. Hyponatremia in untreated active
pulmonary tuberculosis. Am Rev Respir Dis 99:595–597.
51. Kethireddy S, Light RB, Mirzanejad Y, Maki D, Arabi Y, Lapinsky S,
Simon D, Kumar A, Parrillo JE, Kumar A, Cooperative Antimicrobial
Therapy of Septic Shock (CATSS) Database Group. 2013. Mycobacte-
rium tuberculosis septic shock. Chest 144:474–482.
13