See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8095643

Fever of unknown origin

Article  in  The Journal of the Association of Physicians of India · May 2004

Source: PubMed

CITATIONS READS

2 2,976

2 authors, including:

Atul Gogia
Sir Ganga Ram Hospital
70 PUBLICATIONS   137 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Unusual complication of prolonged indwelling urinary catheter ‑ iatrogenic hypospadias View project

All content following this page was uploaded by Atul Gogia on 21 May 2014.

The user has requested enhancement of the downloaded file.

 Review Article

Fever of Unknown Origin

PK Agarwal*, A Gogia**

INTRODUCTION 4. HIV associated FUO - When temperature of > 38.3ºC (>

101ºF) on several occasions is found over a period of
I n 1868 Wundelich, a German clinician, published in
monograph form a convincing demonstration of the value
of measuring the body temperatures in various diseases.1
more than 4 weeks for our patient or more than three
days for hospitalized patients with HIV infection is called
HIV associated FUO. This diagnosis is considered if
Thus the thermometer became the first instrument of precision appropriate investigations over three days including two
to be used in medical practice. “Fever” came to be used to day of incubation of cultures reveals no source.
designate a certain form of forms of illnesses, but the challenge
remained in the identification of the cause of this fever. The criterion of range of fever has been challenged. Timing
and method of temperature monitoring is not mentioned in
DEFINITION most studies. When taking into account all these limitations,
fever may be defined as a morning temperature value of >
A uniform and clear definition of FUO is essential for the
37.2ºC (99.0ºF) or temperature of > 37.8ºC (100ºF) at any other
purposes of epidemiological study and for delineation of the
time during the day, when taken orally (Table 1).
clinical approach to these patients.
Fever of unknown origin (FUO) identifies a syndrome of PREVALENCE AND SPECTRUM OF DISEASE
fever that does not resolve spontaneously, in which the cause The cuases of FUO have traditionally been grouped into
remains elusive after an extensive diagnostic workup. four categories: infectious, malignant, inflammatory and
Petersdorf and Beeson2 first coined the term fever of unknown undetermined.2,5-13
origin in 1961 and explicitly defined it as (1) Temperature >
38.3ºC (101ºF) on several occasions (2) duration of fever of There are 11 series that include over 1000 patients that
more than 3 weeks and (3) failure to reach to diagnosis despite have reported the diagnostic entities that constitute FUO.
one week of inpatient investigations. In 1991, DT Durrack Grouping all the patients collected from 1952 until 1994 reveals
and AC Street 3 suggested two changes to the earlier that the spectrum of disease includes infections in 28% and
definition. Durrack and Street proposed four types of FUO. inflammatory diseases in 21%. Malignancies account for a
smaller proportion (17%). A cause is never identified in a
1. Classic FUO - When temperature > 38ºC (101ºF) recorded significant proportion (19%) of patients (Fig. 1).
on several occasions occurring for more than three weeks
in spite of investigations on three OPD visits or three The most common infectious causes documented in the
days of stay in hospital or one week of invasive literature are tuberculosis and intra-abdominal
ambulatory investigations is called classic FUO. abscesses.2,5,7,8,11,12 Most common malignancies are Hodgkin’s
disease and non-Hodgkin’s lymphoma.2,5,7,8,11,12 Temporal
2. Nosocomial FUO - When temperature more than 38.3ºC arteritis accounts for 16-17% of all causes of FUO in the
(> 101°F) is recorded on several occasions in a elderly.15,16
hospitalized patient who is receiving acute care and in
whom infection was not manifest or incubating on
admission is called nosocomial FO. Three days of
investigations including at least two days incubtion of
cultures, is the minimum requirement for this diagnosis.
3. Neutropenic FUO - When temperature of > 38.3ºC (101ºF)
on several ocasion is observed in a patient whose
neutrophil count is less than 500/microliter or is expected
to fall to that level in 1 or 2 days is called neutropenic
FUO. This diagnosis should be considered when
investigation including at least two days of incubation
of cultures. It is also called immunodeficient PUO.

*Consultant; **PG Student, Department of Medicine, Sir Ganga

Ram Hospital, Rajinder Nagar, New Delhi 110 060. Fig. 1 : The percentage of patients with fever of unknown origin by
Received : 6.11.2003; Accepted : 1.3.2004 cause over the past 40 years.21

314 www.japi.org © JAPI • VOL. 52 • APRIL 2004

 Table 1 : Summary of definitions and major features of the four subtyps of fever of unknown origin (FUO)4
Classic FUO
o o
Definition > 38 C, 3 wk, > 2 visits
or 3 d in hospital
Patient location Community,
clinic or hospital
Leading causes Infections, inflammatory
conditions, cancer,
undiagnosed, habitual
hyperthermia
History emphasis Travel, contacts, animal and
insect exposure, medications,
immunizations, family
history, cardiac
valve disorder
Examination emphasis Fundi, oropharynx, temporal
artery, abdomen, lymph
nodes, spleen, joints, skin,
nails, genitalia, rectum or
postate, lower limb deep veins.
Investigation Imaging, biopsies,
emphasis sedimentation
rate, skin testes
Management Observation, outpatient
temperature chart,
investigations, avoidance of
empirical drug treatments
Time course of Months
disease
Tempo of Weeks
investigation
Nosocomial FUO Immuno-deficient FUO
o
> 38 C, 3d, not present > 38 C, 3d, negative culures
or incubating on alter 48 in < 1000
admission PMN/cumm
Acute care Hospital or clinic
hospital
Nosocomial infections, Majority due to infections,
postoperative complications, but cause documented
drug fever in only 40-60%
Operations and procedures, Stage of chemotherapy,
devices, anatomic drugs administered,
considerations, underlying
drug treatment immunosuppressive
disorder
Wound, drains, devices, Skin folds, IV sites,
sinuses, urine lungs, perianal area
Imaging, bacterial cultures CXR, bacterial cultures
Depends on situation Antimicrobial treatment
Weeks Days
Days Hours
HIV-related FUO
< 38oC, < 3w for
inpatients, HIV infection
confirmed
Community,
clinic or hospital
(HIV primary infection),
typical and atypical
mycobacteria, CMV,
lymphomas, toxoplasmosis,
cryptococcosis
Drugs, exposures, risk
factors, travel, contacts,
stag of infection.
Mouth, sinuses, skin, lymph
nodes, eyes, lungs
perianal area
Blood and lymphocyte
count; serologic test; CXR;
stool examination; biopsies
of lung, bone marrow, and
liver for cultures and
cytological tests, brain
imaging
Antimicrobial protocols,
vaccines, revision of
treatment regimens, good
nutrition
Weeks to months
Days to weeks

CAUSES OF FUO Table 2 : Final diagnosis in elderly compared with

younger patients with fever of unknown origin16,18
Of the many publications concerned with the etiology of
Diagnosis < 65 years > 65 years
FUO most have dealt with classic FUO rather than the other
n=152 n=201
subclasses listed previously. Almost all causes belong to
one of four general categories of disease: infections; Infections 33 (21%) 72 (35%)
inflammatory disorder or connective tissue disorders, Abscess 6 25
Endocarditis 2 14
neoplasms and undetermined conditions.17 A list of them is
Tuberculosis 4 20
given in Table 2. Viral infections 8 1
1. Infections Other** 13 2
Tumours 8 (5%) 37 (19%)
• Abscess - sinus, dental osteomyelitis, hepatic,
Hematologic 3 19
subhepatic, gall bladder, subphrenic, splenic, Solid 5 18
periappendiceal, perinephric, retro peritoneal, pelvic and Multi-system disease* 27 (17%) 57 (28%)
other sites - pyogenic (also amoebic) Miscellaneous 39 (26%) 17 (8%)
• Granulomatous - extra-pulmonary and miliary No diagnosis 45 (29%) 18 (9%)
tuberculosis, atypical mycobacteria infection, fungal *Rheumatic disease, connective tissue disorders, vasculitis (including
infection. temporal arteritis), polymyalgia rheumatica, and sarcoidosis;
**Inlucidng factitious fever (7 cases), habitual hyperthermia (5 cases)
• Intravascular - catheter-related endocarditis, and drug-induced fever (3 cases).
meningococcemia, gonococcemia, Listeria, Brucella, rat
bite fever, relapsing fever. fever, psittacosis.
• Viral, rickettsial and chlamydial - infectious • Parasitic - extra-intestinal amoebiasis, malaria, Kala azar,
mononucleosis, cytomegalovirus, HIV, viral hepatitis, Q toxoplamosis.

© JAPI • VOL. 52 • APRIL 2004 www.japi.org 315

2. Collagen vascular diseases Table 3 : Examples of subtle physical findings having
specific significance in patients with fever of unknown
• Collagen vascular diseases - rheumatic fever, systemic
origin
lupus erythematosus, rheumatoid arthritis particularly
Still’s disease, vasculitis (all types) Body site Physical finding Diagnosis
• Granulomtous - sarcoidosis, granulomatus hepatitis, Head Sinus tenderness Sinusitis
Crohn’s disease Temporal artery Nodules, reduced Temporal arteritis
pulsation
• Tissue injury - pulmonary emboli, sickle cell disease, Oropharynx Ulceration, tender Diseminated
hemolytic anemia. tooth histoplasmosis,
3. Neoplasia periapical abscess
Fundi or conjunctiva Choroid tubercle, Disseminated
• Leukemias, Hodgkin’s and non-Hodgkin’s lymphoma,
petechiae, granulomatosis,*
acute leukemia, myelodysplasic syndrome. Roth’s spot endocarditis,
• Carcinoma - kidney, pancreas, liver, GI tract, lung Thyroid Enlargment, Thyroditis
especially when metasatic tenderness
Heart Murmur Infective
• Atrial myxomas endocarditis
• Central nervous system tumors Abdomen Enlarged iliac Lymphoma,
crest; lymph nodes, endocarditis,
4. Miscellaneous
splenomegaly disseminated
• Vascular - Haematoma, thrombosis, recurrent pulmonary granulomatosis
embolism, aortic dissection, femoral aneurysm, post- Rectum Perirectal fluctuance, Abscess
myocardial infarction syndrome tenderness;
prostatic tenderness, Abscess
• Drug fever fluctuance
• Endocrinal - Subacute (de Quervain’s) thyroiditis, Genitalia Testicular nodule Periarteritis nodosa
hyeprthyroidism, adrenal insufficiency, primary Epididymal nodule Disseminated
hyeprparathyroidism, hypothalamic hypopituitarism, granulomatosis
Sweet’s syndrome, familial Mediterranean fever, and Lower extremities Deep venous Thrombosis or
tenderness thrombophlebitis
familial Hibernian fever, hyperimmunoglobulin D
Skin and nails Petechiae, splinter Vasculitis, endocarditis
syndrome. hemorrhages,
• Hepatic - Cirrhosis, chronic active hepatitis, alcoholic subcutaneous
hepatitis. nodules, clubbing

• Allergic - Milk protein allergy, hypersensitivity
penumonitis, extrinsic allergic alveolitis, metal fume fever,
polymer fume fever, idiopathic hypereosinophilic
syndrome.
• Nervous system - Complex partial status epilepticus,
cerebrovascular accident, brain tumour, encephalitis.
• Others - Anomalous thoracic duct, psychogenic fever,
habitual hyperthermia, factitious illness, shunt nephritis,
malacolapkia, Kawasaki’s syndrome, Kikuchi’s syndrome,
mesenteric fibromatosis, inflammatory pseudotumour
Castleman’s disease, Vogt-Koyanagi-Harada syndrome,
Graucher’s disease, Schnitzler’s syndrome, FAPA
syndrome (fever, aphthous stomatitis, pharyngitis,
adenitis), Fabry’s disease, cholesterol emboli, silicone
embolisation, teflon embolisation, lymph node infarction,
sickle-cell disease vasoocclusive crisis, anhidrotic
ectodermal dysplasia, cyclic neutropenia, Brewer’s yeast
ingestion, Hamman-Rich syndrome.
EVALUATION OF PATIENT OF FUO
In FUO, there is no diagnostic gold standard agaisnt which
other diagnostic tests may be measured. Final diagnoses are
determined in a number of ways, including a comprehensive
history, repeated physical examination (Table 3) and a host of
laboratory investigations.
*Includes tuebrculosis, histoplasmosis, coccidioidomycosis, sarcoidosis
and syphilis.
The initial evaluation of the patient with PUO typically
includes, comprehensive history and physcal examination.
Thorough history is important and this should include
information about alcohol intake, medications, occupational
exposures, pets, travel, familial disorders and previous
illnesses. The specific findings that have led to a diagnosis
in FUO are numerous and diverse. The yield of physical
examination is not recorded in most studies of FUO. The
yield may be high is suggested by two studies reporting that
in pediatric patients about 60% had abnormal findings that
contributed to a diagnosis.19,20 The following table shows
how a complete and meticulous physical examination can
lead to a diagnosis.
Further in the history careful attention is to be paid to the
host factors.
1. Age
2. Sex - autoimmune diseases are more common in female,
specific disease like pelvic inflammatory disease or male
specific disease like prostatitis.
3. Residence present and past - some diseases are more
prevalent in a particular area - Kala azar in Bihar.

316 www.japi.org © JAPI • VOL. 52 • APRIL 2004

4. Past infections are sometimes responsible for reactivation Table 4 : Recommendations for diagnostic testing in
or other effects. FUO21
5. Immunization and medications. Maneuver Effectiveness Level of Recommendation
6. Exposure to pets, other animals, contacts persons. evidence

7. Work environment/home environment. The Duke Specificity 99% Fair Very high
criteria specificity in
8. Drug, underlying disease, cardiac valve disorder, patients with FUO
previous surgeries including splenectomy. Abdominal CT Diagnostic yield 19% Fair High diagnostic
Sensitivity 71% yield
9. Family history of TB or rarely hereditary cause of fever
Specificity 71%
like familial Mediterranean fever should be sought. Tc99m Specificity 93%-94% Fair High Specificity
Clinical features nuclear scan Sensitivity 40%-75% poor sensitivity
Tracer of choice
Fever has been characterized by magnitude and frequency in FUO
and specific fever patterns have been ascribed to many of Gallium 67 Specificity 70%-78% Fair Poor specificity
the causes of FUO.22 Unfortunately in most cases series, the scan Sensitivity 54%-67% and sensitivity
height, pattern or duration of fever did not relate to diagnosis. ESR, C-reactive No evidence to
The response of fever to naproxen sodium may be helpful in protein No study make
MRI, echocardio recommendations
that the fever due to solid tumours and many rheumatological
for or agaisnt it
diseases (most notably Still’s Disease) usually subside Liver biopsy Diagnostic yield Fair High diagnostic
promptly while fever due to other causes may persist.23,24 14%-17% yield and minimal
Laboratory Investigations toxicity
Bone marrow Diagnostic yield Fair Very low
While planning a diagnostic workup for a patient of FUO cultures 0%-20% diagnostic yield
it will be useful to remember that most often the cause of FUO Laparotomy/ Poor High yield in pre-
is a common disorder presenting in atypical rather than an Laparoscopy CT era
exotic disease presenting in its typical form. Empiric No study
Therapy
Diagnostic workup of fever of unknown origin19
A. Comperhensive history
13. Echocardiography in case of cardiac murmur.
B. Physical examination
14. Venous duplex scan of lower limbs.
C. Laboraotry tests (Table 4)
15. Radionuclide scans - Indinium and 67Gallium scans
1. Complete blood count and differential
16. Invasive procedures-
2. Erythrocyte sedimentation rate
Bone marrow aspiration and biopsy (also imprint
3. Blood film reviewed by the hematopathologist smear and culture/serology)
including malarial parasite and malarial serology.
Lymph node FNAC and biopsy
4. Routine blood chemistry (including lactate
dehydrogenase, bilirubin, and liver enzymes) Liver biopsy/thoracoscopy (where indicated)
5. Urinalysis and microscopy Splenic aspirate (where indicated)
6. Blood (x3) and urine cultures (and other sterile CT guided FNAC of mass/lymph node
compartment if clinically indicated e.g. joints pleura, Laparoscopy
CSF) Bronchoscopy and transbronchial biopsy
7. Chest radiogrpah Further evaluation if any abnormalities detected in the
8. Abdominal ultrasonography above tests.
9. Skin Testing-tuberculin skin test Non-invasive test yield diagnosis in about 25% of cases
10. Serology - a) Human immunodeficiency virus, of FUO. Imaging studies have been used to localize
cytomegalovirus IgM antibody; heterophil antibody abnormalities as a preamble to more definitive (invasive)
test (if consistent with mononucleosis - like testing.
syndrome), Q - fever serology (if exposure factor The Duke’s criteria have a very high specificity (99%) in
exist), hepatitis serology (if abnormal liver enzyme patients with FUO and suspected infective endocarditis, and
test result), angiotensin converting enzyme, thus should be used to identify endocarditis as the cause of
b) PCR/NASBA for tuberculosis, hepatitis, CMV. FUO. When the initial investigations are not helpful in
11. Collagen makers - antinuclear antibodies, rheumatoid identifying a cause, the clinician should then proceed to
factors, ENA, pANCA, c ANCA, complement levels. imaging. These should include a CT of the abdomen and a
Technetium based nuclear scan. A CT of the abdomen has a
12. CT abdomen/chest high diagnostic yield (19%)25 and carries a low risk. Two fair-

© JAPI • VOL. 52 • APRIL 2004 www.japi.org 317

 quality studies show that technetium-based scans have a
high specificity but are insensitive. Leg Doppler imaging
should be considered the next step in identifying deep vein
thrombosis as a potential reversible and easily treatable
cause.8,18 A temporal artery biopsy should be considered in
elderly patients with FUO. There is fair evidence to suggest
that a liver biopsy has a high diagnostic yield (14%-17%)26,27
with minimal toxicity. Bone marrow cultures are of low yield
(0%-20%) and are not recommended in immunocompetent
patients with FUO.21
Invasive procedures
Diagnosis of fewer than half the cases of FUO has resulted
from excisional biopsy, needle biopsy, or laparotomy. Most
FUO patients undergo at least one of these procedures, even
though the diagnostic yield is moderate - i.e. 2.8-4.6 biopsies
per final diagnosis achieved.28 The yield from biopsies in the
operating theatre or under CT guidance is greater than that
of bedside biopsy procedures.28 The only biopsy that may
often be rewarding in the absence of prior localizing
information is temporal artery biopsy in elderly patients with
a very high ESR.16
Exploratory laparotomy in the absence of localizing
features is unusual these days. Laparoscopy, including
laparoscopic liver biopsy, is a less traumatic alternative. It is
most helpful when other features point to abdominal disease
and has had a yield of only 20% when such features are
absent.28
MANAGEMENT OF FUO
A fundamental principle in the management of classic FUO
is that therapy should be delayed until the cause of fever has
been determined so that the therapy can be tailored to a
specific diagnosis. Non-specific treatment is rarely curative
and has the potential to delay diagnosis. Non-specific
treatment is rarely curative and has the potential to delay
diagnosis. At the same time diagnostic delay adversely affect
the prognosis in intra-abdominal infections, military TB,
disseminated fungal infection and recurrent pulmonary
embolism. Diagnosis of PUO may require repeated
examination, interviewing the patient and investigations.
Therapeutic trials
In appropriate clinical settings, therapeutic trials,
employing agents with limited spectrum of activity like anti-
tubercular drugs may be accepted. It is particularly helpful in
cases where there is a history of prolonged low-grade fever
with evening rise along with raised ESR, a positive tuberculin
test with or without loss of appetite and weight.
In suspected temporal arteritis to prevent vascular
complications like blindness empirical corticosteroids may
be given.
In other situations empirical treatment with anti-
inflammatory agents, aspirin, corticosteroids, antibiotics, or
anti-neoplastic agents should be desisted.
318 www.japi.org
View publication stats
PROGNOSIS
Outcomes of patients with FUO are a function of the
underlying cause. Overall, 12-35% of patients will die from
FUO-related causes. 52% to 100% of patients with a final
diagnosis of malignancy will die within five years. Mortality
is much less if an infection is identified as the cause of FUO
(8%-22%).2,5,8,10,11 Therefore, the best predictor of survival is
disease category, with malignancy incurring the highest
mortality. The prognosis of patients with FUO in whom a
cause cannot be identified is excellent.2,8,11 Most of these
patients have spontaneous recovery (51%-100%) and only a
small proportion have persistent fever (0%-30%). The 5-year
mortality rate of undiagnosed FUO was only 3.2%. The
prognosis of nosocomial fever of unknown origin varies
according to underlying disese. The short-term prognosis of
neutropenic fever of unknown origin is excellent with over
90% to empirical antibiotic therapy.
REFERENCES
1. Wunderlich CA. On the temperature in diseases. London,
New Syndenham Society, 1871.
2. Petersdorf RG, Beeson P. Fever of unexplained origin, report
of 100 cases. Medicine (Baltimore) 1961;40:1-30.
3. Durrack DT, Street AT. Fever of unknown origin reexamined
and redefined. Current Clin Tropical Infec Dis 1991;11:35-51.
4. Durrack DT. Fever of unknown origin. In Oxford textbook
of medicine. Weatherhall DJ, Ledingham JGG, Warrell DA
(edit). 3rd edition Oxford Medical Publications; 1995;1015-
16.
5. Kazanjian PH. Fever of 86 patients treated in community
hospitals. Clin Infect Dis 1992;15:968-73.
6. De Kleijn EM, van der Meer JW. Fever of unknown origin
(FUO): report on 53 patients in a Dutch university hospital.
Neth J Med 1995;47:54-60.
7. Barbad FJ, Vazquez J. Pyrexia of unknown origin: changing
spectrum of disease in two consecutive series. Postgrad Med
J 1992;68:884-87.
8. Larson EB, Featherstone HJ, Petersdorf RG. Fever of
undetermined origin: diagnosis and follow up of 105 cases,
1970-1980. Medicine (Baltimore) 1982;61:269-92.
9. Barbado FJ, Vazquez JJ, Pena JM, et al. Fever of unknown
origin: a survey on 133 patiens. J Med 1984;15:185-92.
10. Deal WB. Fever of unknown origin: analysis of 34 patients.
Postgrad Med 1971;50:182-88.
11. De Kleijn EM, Vandenbrouchke JP, van der Meer JW. Fever
of unknown origin (FUO), IA : prospective multicentre study
of 167 patients with FUO using fixed epidemiological criteria.
Medicine (Baltimore) 1997;76:392-400.
12. Knockaert DC, Vanneste LJ, Vanneste SB, Bobaers HJ. Fever
of unknown origin in the 1980s: an update of the diagnostic
spectrum. Arch Intern Med 1992;152:51-55.
13. Smith JW. Southwestern Internal Medicine Conference; fever
of undetermined origin: not what it used to be. Am J Med Sci
1986;292:56-64.
© JAPI • VOL. 52 • APRIL 2004