According to a practicing rheumatologist, fever doesn't always mean an infection and hoofbeats don't always mean a horse.
Adult-onset Still's disease (AOSD) is rare and "there really isn't a set test to identify it," said Christopher R. Morris, MD, of Arthritis Associates in Kingsport, Tennessee.
"There is a lack of disease awareness. Patients can go for weeks and months with symptoms ... Nobody bothers to put all of the pieces together to give them a diagnosis," he told 鶹ý.
Patients with this systemic inflammatory disorder usually present with achy swollen joints, as well as fever (quotidian fever greater than 100° F) and a salmon-colored rash over the trunk and arms that is mildly itchy, which both peak in the evening. Bloodwork typically reveals elevations in white blood cell count with a left shift, erythrocyte sedimentation rate, and C-reactive protein. Serum ferritin level is also extremely elevated (10,000 ng/mL), which can help clinicians in making a diagnosis of AOSD.
Morris explained that, absent specific serologic, radiologic, or pathologic tests, AOSD is often diagnosed by exclusion. In most cases, clinicians see the presenting symptoms and suspect an infection. When the patient does not respond to antibiotics and an infectious organism is not identified, clinicians may begin to suspect cancer. This suspicion may be reinforced by the fact that most patients with AOSD have elevated liver function tests, leading to concerns that cancer may have metastasized to the liver.
As a result, many patients undergo an extensive evaluation with whole-body imaging and histopathologic examination to exclude infections and malignancies prior to receiving a diagnosis of AOSD.
In most cases, a rheumatologist is consulted once infection and cancer have been ruled out. Morris noted that the combination of elevated liver function tests and elevated ferritin in particular are most suggestive of AOSD, and clinicians who suspect AOSD should run these two tests.
Nilanjana Bose, MD, MBA, a rheumatologist at Lonestar Rheumatology in Houston, told 鶹ý that AOSD is "very rare even for rheumatologists ... It's just something to keep on the differential."
AOSD can resemble rheumatic fever, Morris said. Other mimics include rheumatoid arthritis (RA) and systemic lupus erythematosus.
Bose explained that some patients present with a more RA-like phenotype with warm, stiff joints, and some patients may not have a rash at presentation. She also noted that patients with AOSD are often sicker than patients with RA -- so much sicker, said Morris, that often rheumatologists encounter them for the first time in the hospital after they have been seen first by an infectious disease specialist and then by an oncologist.
In this situation, rheumatologists should keep in mind that macrophage activation syndrome (MAS) is common in patients with AOSD and may cause systemic complications and death.
In order to diagnose AOSD prior to hospitalization, Morris suggested that a primary care physician who suspects AOSD make a referral to a rheumatologist, order ferritin and liver function tests, and have the patient keep a record of the timing of fevers and rashes.
While no diagnostic criteria or biomarkers have been formally validated for AOSD, and the validation of such biomarkers is unlikely, considering the low incidence of disease, research has suggested that glycosylated ferritin (GF) may be an accurate diagnostic test for AOSD, even in patients without elevated ferritin.
suggested an optimal GF cutoff value of 16% for AOSD diagnosis, yielding a specificity of 89% and a sensitivity of 63%. This is lower than the level that had been previously recommended as a diagnostic criterion for AOSD (≤20%), which is more suggestive of hemophagocytic lymphohistiocytosis (HLH) in the setting of hematologic malignancies.
GF may also serve as a useful tool for AOSD prognosis in that patients with lower GF levels tend to have a worse prognosis in terms of intensive care unit admission and mortality. Likewise, GF may be able to identify patients in remission, with a threshold of >50% having a negative predictive value of 100% for AOSD.
Disclosures
Morris and Bose reported no relevant conflicts of interest.