Clinical Problem: Fever in the Patient with Rheumatic Disease: Underlying Disease or Superimposed Infection
Pearl: Rheumatoid arthritis rarely causes high-grade fever.
Comment: Although low-grade fevers in the range of 37.5آ°C often accompany active rheumatoid arthritis (RA), high-grade fevers are rare. Only 5% of cases manifest fever >38آ°C, and less than 1% have temperatures >38.3آ°C. Therefore, high-grade fever in a patient with well-established RA should prompt an investigation for an underlying cause (eg, infection) other than RA. In a patient with the new onset of inflammatory arthritis, the presence of high-grade fever is an argument against the diagnosis of RA or even a complication of RA, such as rheumatoid vasculitis
Pearl: Flare of a single joint in an RA patient signals serious concern about septic arthritis.
Comment: Most flares of RA are polyarticular. When signs of new, increased inflammation affect only one joint, infection should be strongly considered. The most common cause of septic arthritis in RA is Staphylococcus aureus. Absence of fever does not exclude infection because only 50% of patients with septic arthritis present with fever. In the RA patient with only one "active" joint, arthrocentesis should be performed to exclude infection before intensifying anti-inflammantory therapy.
Pearl: Always consider tuberculosis when a febrile illness develops in a patient treated with an anti–tumor necrosis factor agent.
Comment: Because tumor necrosis factor- (TNF-) is required for an intact host immune response to Mycobacterium tuberculosis, treatment with anti–TNF- agents greatly increases the risk of TB. Although infliximab has been most closely associated with the greatest risk of TB in studies, the risk is likely a class effect shared by all anti–TNF- agents. Most cases, caused by reactivation of infection, have developed within weeks of starting anti–TNF- therapy. There is almost certainly an increased risk of disease from primary exposure as well, however.
In the setting of anti–TNF- therapy, tuberculosis can be acute, is often disseminated or extrapulmonary, and has an atypical histopathology (TNF- is required for granuloma formation). Anti–TNF- therapy also carries an increased risk of infection with other intracellular pathogens, including Listeria monocytogenes, and fungi, such as Histoplasma capsulatum and Coccidioides immitis.
Pearl: When trying to distinguish between infection and active disease in a patient with SLE, the presence of rigors favors infection.
Comment: One of the great quandaries of hospital medicine for patients with SLE is determining whether an acute change is caused by infection or a flare of the disease. The presence of rigors clearly favors infection. Other clues may come from the complete blood count: many SLE patients have a baseline tendency toward neutropenia (particularly lymphopenia) and thrombocytopenia. Elevations in either of these two blood counts raises the likelihood of infection.
Pearl: Fever that develops in a patient with SLE treated with high-dose glucocorticoids is due to infection until proven otherwise.
Comment: Infection is a leading cause of morbidity and mortality in SLE, particularly in the setting of immunosuppression with high-dose glucocorticoids. Although SLE itself can cause fever >39آ°C, fever due to SLE most often occurs in the setting of clinically active disease (particularly serositis) and usually responds to glucocorticoid therapy. Fever that develops after high-dose glucocorticoids have been started should be attributed to SLE only after a vigorous search for infectious causes, including opportunistic infections.
Pearl: When a patient with SLE remembers the precise hour that the disease "flared," the patient more likely has an acute infection.
Comment: Most flares of SLE develop over days or weeks. Infections tend to present more abruptly. Therefore, a patient who says, "my disease flared at 10 o'clock" probably has an infection.
Pearl: Apparent "flares" of rheumatic disease that occur while the patient is taking cyclophosphamide are almost always caused by a superimposed opportunistic infection rather than by activation of the underlying disease.
Comment: Over the last 25 years at one large medical center, not one patient transferred to that center because of "refractory vasculitis" has had active vasculitis: every one of the transferred patients has had an opportunistic infection.