Abstract
OBJECTIVE: JC virus (JCV) seropositivity is a risk factor for progressive multifocal leukoencephalopathy (PML) in patients on natalizumab. Accordingly, the JCV serological antibody test is of paramount importance in determining disease risk. METHODS: We tested the accuracy of the JCV serum antibody test by comparing the results of JCV serology to JCV viruria and viremia in 67 patients enrolled in a single-center, retrospective cohort study. Bodily fluids (urine and blood) were assessed for JCV DNA by real time quantitative polymerase chain reaction 6 to 47 months (mean = 26.1 months) before JCV antibody testing. In 10 individuals, blood and urine samples were obtained on 2 separate occasions at 6-month intervals. RESULTS: Forty (59.7%) of the 67 patients were JCV seropositive. Of 27 JCV seronegative patients, 10 (37%) had JCV viruria. Urine JCV DNA copy numbers were significantly higher in the seropositive group (mean log copy number = 5.93, range = 1.85-9.21) than the seronegative group (mean log copy number = 2.41, range = 1.85-5.43; p = 0.0026). Considering all body fluid test results, 50 (74.6%) of the 67 patients were previously infected with JCV. INTERPRETATION: The false-negative rate of the JCV serology in this study was 37%; therefore, JCV serostatus does not appear to identify all patients infected with JCV. Thus, a negative JCV antibody result should not be conflated with absence of JCV infection. This discordance may be important in understanding JCV biology, risk for PML, and PML pathogenesis.