In both veterinary and human patients, fever may indicate infectious, inflammatory, immune-mediated, or neoplastic disease. In most cases, the history and physical examination reveal the cause of the fever, or the fever resolves spontaneously or in response to antibiotic therapy. However, in a small percentage of patients, the cause of fever is not readily apparent, and the problem becomes persistent or recurrent. These patients are said to have fever of unknown origin (FUO).
In human medicine, classic FUO is defined as fever >101°F (38.3°C) on several occasions over a period >2–3 wk with no diagnosis established after 3 outpatient visits or 3 days in the hospital. There is no recognized definition of this syndrome in veterinary medicine, making it difficult to determine its true prevalence. FUO is probably less prevalent now than in the past because of improved diagnostic technology (eg, imaging, molecular diagnostic tests).
Body temperature is regulated by the hypothalamus. This area of the brain acts as a thermostat to maintain temperature as close as possible to a normal set-point. The hypothalamus receives input from internal and external thermoreceptors, and it activates physiologic and behavioral activities that influence heat production, heat loss, and heat gain.
Hyperthermia refers to any increase in body temperature above the normal range. Fever is a particular form of hyperthermia in which the heat loss and heat gain mechanisms are adjusted to maintain body temperature at a higher hypothalamic set-point; thus, fever is essentially a regulated hyperthermia. In nonfebrile cases of hyperthermia (eg, heat stroke, exercise-induced hyperthermia, malignant hyperthermia, seizure), body temperature is increased by abnormal and unregulated heat loss, heat gain, or heat production, and the hypothalamic set-point is not altered. Depending on their severity, these conditions can potentially result in body temperatures ≥106°F (41.1°C). In comparison, most patients with true fever have body temperatures in the range of 103°–106°F (39.5°–41.1°C).
An increase in the hypothalamic set-point may be initiated by exogenous pyrogens, which include drugs, toxins, and viral or bacterial products (eg, endotoxin). These pyrogenic stimuli lead to the release of cytokines, termed endogenous pyrogens, from inflammatory cells. Ultimately, locally synthesized prostaglandin E2 in the hypothalamus is responsible for increasing the set-point, resulting in fever.
FUO may be defined as fever that does not resolve spontaneously in the period expected for self-limited infection and for which a cause cannot be found despite considerable diagnostic effort. This excludes patients that respond to antibiotic therapy (and do not relapse) and patients in which the cause of fever is determined from initial history, physical examination, or laboratory tests, or in which fever resolves spontaneously.
Infectious, immune-mediated, and neoplastic disease are the most common causes of FUO in dogs. In a study of 101 dogs with fever, 22% had immune-mediated diseases, 22% primary bone marrow abnormalities, 16% infectious diseases, 11.5% miscellaneous conditions, 9.5% neoplasia, and 19% genuine FUO. In cats, the cause is more likely to be infectious, but there are fewer published data on feline cases than on canine cases. In a case series of horses with FUO, 43% had infectious disease, 22% neoplasia, 6.5% immune-mediated disease, 19% miscellaneous causes, and in 9.5% the cause was not determined. In farm animals, the most likely causes of FUO are infectious or inflammatory diseases such as pneumonia, peritonitis, abscesses, endocarditis, metritis, mastitis, polyarthritis, and pyelonephritis.
The key to diagnosis of FUO is to develop and follow a systematic plan that allows for the detection of both common and uncommon causes of fever. The plan should always include repetition of relevant tests, because findings can change over time. Owners should be informed that diagnosis of FUO may require considerable time and patience and may demand more advanced or expensive diagnostic tests. Nevertheless, simple and inexpensive tests may also reveal diagnostic clues that eventually point to the cause of the fever. In one retrospective study of fever in dogs, radiography, cytology, and bacterial or fungal cultures of tissues or fluids were found to be the most useful diagnostic tests.
A staged or tiered approach to diagnosis can assist in choosing appropriate tests. The first stage should include history, physical examination, ophthalmic and neurologic examinations, CBC, fibrinogen, serum chemistry profile, urinalysis and urine culture, feline leukemia virus and feline immunodeficiency virus tests (cats), and usually thoracic and abdominal radiographs in small animals. Any medications that could induce fever should be discontinued. In the second stage, some first-stage tests may be repeated (particularly the physical examination), and additional specialized tests are performed. These may be dictated by abnormal findings in the first stage of testing or may be determined by consideration of the most common known causes of FUO. Tests included in this stage include blood cultures, arthrocentesis, abdominal ultrasonography, lymph node aspiration, aspiration of other organs or masses, analysis of body fluids (eg, fluid from body cavities, milk samples, reproductive tract secretions), rectal cytology, fecal culture, echocardiography (in the presence of a murmur), long-bone and joint radiographs, contrast radiographs, serology, and molecular diagnostic tests. In the third stage, earlier tests may be repeated again, as well as additional specialized procedures. These procedures are most likely to be chosen on the basis of previous findings but may also be considered when all previous testing has been unrewarding. Examples include echocardiography (in the absence of a murmur), dental radiographs, bone marrow aspiration, bronchoscopy and bronchoalveolar lavage, CSF analysis, CT, MRI, laparoscopy, thoracoscopy, biopsies, exploratory surgery, or trial therapy.
Epidemiologic characteristics such as vaccination, parasite control, exposure to vectors, and travel history should always be reviewed. The response to previous medications should be determined, as well as the presence of illness in other animals or people. Owners should be questioned carefully about specific clinical signs, because these may help localize the source of the fever. The physical examination should be detailed; always include fundic, neurologic, and rectal examinations; and repeated frequently.
The CBC and chemistry changes in animals with FUO are often nonspecific but may suggest further diagnostic tests. For example, bile acids assay may be indicated in an animal with changes suggestive of hepatic dysfunction. The CBC should always be accompanied by blood smear evaluation to detect parasites or morphologic changes.
Thoracic and abdominal radiographs are useful screening tools for the early localization of fever. Skeletal radiographs and contrast radiographs may subsequently be considered, depending on initial findings. For example, myelography may be used to investigate back pain. The use of techniques such as CT or MRI is determined by the results of initial diagnostic testing or by consideration of the body system of interest, eg, MRI is particularly useful to evaluate the CNS. Advanced imaging with nuclear scintigraphy or positron emission tomography is used in human patients with FUO but is not yet widely reported in veterinary medicine.
Abdominal ultrasonography may reveal a source of fever in the abdomen, such as neoplasia, peritonitis, pancreatitis, or abscesses. The thoracic cavity, limbs, and retrobulbar areas may also be examined by ultrasound. Echocardiography is indicated at the early stages of evaluation of the FUO patient with a murmur. This may aid in the detection of endocarditis, although this diagnosis should also be based on signalment, characteristics of the heart murmur, and blood culture results.
Bone marrow cytology and histology should be evaluated in any animal with unexplained CBC abnormalities. Bone marrow disease is a common cause of FUO in small animals; therefore, bone marrow aspiration and biopsy, if possible, should also be included in the second stage of diagnostic testing in these patients. When obtaining bone marrow aspirates from cats, a sample should be saved for possible molecular diagnostic testing for feline leukemia virus.
Because immune-mediated polyarthritis is a common cause of FUO in dogs, arthrocentesis of multiple joints is included in the second stage of diagnostic testing in this species, even if the joints are normal on palpation. Some dogs with steroid-responsive meningitis-arteritis also have concurrent immune-mediated polyarthritis; therefore, arthrocentesis should be performed in dogs with spinal pain. Infectious polyarthritis is more commonly recognized in large animals, in which arthrocentesis is an important diagnostic test.
Blood cultures are recommended in all animals with unexplained fever. The techniques used should allow the collection of adequately large volumes of blood under aseptic conditions. If the size of the animal allows collecting more than one set of samples for blood culture, using appropriately sized aerobic and anaerobic bottles increases the sensitivity and specificity of the test. Special enrichment culture methods may be considered for certain organisms, eg, Bartonella spp.
Tests available for the diagnosis of infectious diseases include assays for detection of antibodies or antigen in blood, body fluids, or tissues. Molecular diagnostic tests detect nucleic acid, with PCR being the most common in this category. Selection of these tests should be based on the signalment, clinical signs, and epidemiologic characteristics of the animal. Interpretation of test results requires an understanding of disease prevalence, vaccination history, and sensitivity and specificity of the test. When requesting PCR-based assays, it is important to use laboratories that have quality management programs that address test performance and consistency, and control for sample contamination.
Fine-needle aspirates are safe and simple to obtain from effusions, masses, nodules, organs, tissues, and body fluids. Fluids should be examined cytologically and also submitted for microbiologic or molecular diagnostic testing. Tissue biopsies are generally obtained in the second or third stages of diagnostic testing, after clinical signs or initial diagnostic tests have localized the fever. When biopsies are obtained, sufficient samples should be submitted for histopathology, appropriate culture (aerobic and anaerobic, fungal, mycoplasmal, mycobacterial, etc), molecular diagnostics, and special stains. If exploratory surgery is performed, biopsies should be obtained from several sites.
In some FUO cases a specific diagnosis is not reached, or diagnostic testing is discontinued, leading to consideration of therapy in the absence of a diagnosis. Options include antibiotics, antifungal agents, and anti-inflammatory or immunosuppressive therapy (usually with corticosteroids). Trial therapy may resolve the clinical signs or confirm a presumptive diagnosis, but it is also associated with significant risk. Before pursuing a therapeutic trial, the owner should be informed of the potential risks and should be committed to careful monitoring of the animal for an appropriate length of time. The therapeutic trial should be based on a tentative diagnosis and should define the parameters to be followed and the criteria used to determine treatment success or failure. If an animal is likely to be referred for in-depth investigation of FUO, trial therapy should not be started because it may affect the results of further testing.
In true fever, the increase in body temperature is regulated; therefore, cooling methods such as water baths work against the body’s own regulatory mechanisms. It is also likely that fever itself has some beneficial effects, particularly in infectious diseases. However, fever can lead to anorexia, lethargy, and dehydration. Thus, animals with FUO may benefit from IV fluid therapy or antipyretic medications. Examples include NSAIDs such as aspirin, carprofen, ketoprofen, and meloxicam in small animals, and flunixin meglumine or phenylbutazone in large animals.