A standardized minimum diagnostic workup should be completed before fever of unknown origin (FUO) is diagnosed in ambulatory immunocompetent adults, according to a state-of-the-art review published in Clinical Infectious Diseases.
William F. Wright, DO, MPH, of the Division of Infectious Diseases, Department of Medicine, Johns Hopkins University School of Medicine in Baltimore, and colleagues reviewed contemporary evidence on the diagnosis and management of FUO. The authors said inconsistent definitions and evaluation strategies have complicated both clinical care and comparisons across studies and argued that newer qualitative criteria offer a more useful framework.
Definitions of FUO have evolved since the original 1961 criteria, which required more than 3 weeks of illness and no diagnosis after 1 week of hospital evaluation. Later definitions incorporated outpatient evaluation and minimum testing requirements.
Updated criteria define a minimum workup
Under recommendations developed by an expert consensus panel using a modified Delphi process, FUO in immunocompetent adults is defined as an unexplained fever of at least 38.3°C (100.9°F) on more than three occasions over more than 3 weeks. The diagnosis also requires completion of a standard minimum diagnostic evaluation.
The review was limited to immunocompetent adults because prolonged unexplained fevers in immunocompromised patients have a different spectrum of potential causes.
The minimum workup includes a complete blood count with differential, comprehensive metabolic panel, erythrocyte sedimentation rate, C-reactive protein, ferritin, thyroid-stimulating hormone, rheumatoid factor, antineutrophil cytoplasmic antibodies and antinuclear antibodies. Microbiologic evaluation includes at least three spaced sets of blood cultures incubated for 5 days, urinalysis with reflex urine culture, HIV screening and tuberculosis testing. Recommended imaging includes abdominal ultrasonography plus either chest radiography or computed tomography of the chest, abdomen and pelvis.
The authors recommended classifying potential causes into five groups: infections, noninfectious inflammatory disorders, oncologic diseases, miscellaneous conditions and undiagnosed illnesses. They stressed, however, that FUO does not usually signal an exotic disease. “Most patients are not suffering from unusual or rare diseases; instead, they more likely exhibit atypical manifestations of common illnesses,” they wrote. The likely causes also vary with age and geography.
That geographic variation was particularly evident among infectious causes. Across 19 prospective studies involving 2,667 patients, including 832 with infections, Mycobacterium tuberculosis complex accounted for 34.3% of infections and was the most commonly reported infectious cause in four of six World Health Organization regions. Brucellosis accounted for 9.7%, endocarditis for 7.5%, abscesses for 7.3% and herpesvirus infections for 7.2%. No eligible contemporary studies in that meta-analysis came from Africa or the Americas.
The authors recommended including tuberculosis screening in the minimum FUO evaluation regardless of geography. When available, they favored an interferon-gamma release assay over tuberculin skin testing because it provides an objective result in one visit and is not affected by previous Bacillus Calmette-Guérin vaccination. However, they cautioned that a negative screening result does not exclude tuberculosis.
Diagnostic clues can help — and mislead
A thorough history and physical examination remain central to the evaluation. Travel and geographic history, animal and food exposures, work environment, medications and previous illnesses can help direct subsequent testing. The authors also emphasized identifying potential diagnostic clues from the history, examination, laboratory results and imaging rather than ordering additional testing indiscriminately.
But those clues require careful interpretation. In one study cited in the review, an average of 15 potential diagnostic clues were identified per patient, but only 19% contributed to the eventual diagnosis and 81% proved misleading. Even when clues lead to dead ends, the authors called following them “the most crucial strategy” for determining the source of FUO.
PET/CT may play a larger role after initial evaluation
The review also examined the role of 18F-fluorodeoxyglucose positron emission tomography/computed tomography, or 18F-FDG PET/CT, in patients who have already met the updated FUO criteria.
Pooled data from nuclear medicine guidance and meta-analyses showed diagnostic yields of 84% to 98% for 18F-FDG PET/CT. The technique can localize metabolically active lesions associated with infection, inflammation or malignancy that may not be apparent on conventional testing.
The authors did not recommend adding 18F-FDG PET/CT to the initial minimum workup. Rigorous studies directly comparing it with currently recommended conventional imaging are lacking, and cost and limited access remain barriers. The Delphi panel instead reached strong consensus that 18F-FDG PET/CT should be used after a patient fulfills all criteria for FUO.
The authors also urged restraint with empiric treatment. “A fundamental principle of FUO management is that therapy should be withheld whenever possible until the cause has been determined so treatment can be tailored to a specific diagnosis,” they wrote. Antimicrobials, corticosteroids or anti-inflammatory agents should generally be reserved for patients in whom other diagnostic approaches have failed or those who are too ill for treatment to be delayed.
Exceptions include suspected temporal arteritis when treatment is needed to prevent blindness or stroke, as well as hemodynamic instability or newly identified immunocompromise or neutropenia. Empiric treatment may also be appropriate when tuberculosis is strongly suspected in a high-prevalence setting but cannot be confirmed with appropriate testing.
Ultimately, the authors said FUO evaluation should remain individualized even within the standardized framework. “A personalized clinician approach, informed by findings and the patient’s therapy tolerance, is essential,” they wrote, adding that effective management typically requires a coordinated multidisciplinary strategy.
The authors reported having no relevant conflicts. The work and publication were supported by the Sherrilyn and Ken Fisher Center for Environmental Infectious Diseases.
