Utility of urine and serum lateral flow assays to determine the prevalence and predictors of cryptococcal antigenemia in HIV-positive outpatients beginning antiretroviral therapy in Mwanza, Tanzania

BACKGROUND Detection of subclinical cryptococcal disease using cryptococcal antigen screening among HIV-positive individuals presents a potential opportunity for prevention of both clinical disease and death if patients with detectable cryptococcal antigen are identified and treated pre-emptively. Recently developed point-of-care cryptococcal antigen tests may be useful for screening, particularly in resource-limiting settings, but few studies have assessed their utility. METHODOLOGY The objectives of this study were to determine the prevalence and factors associated with cryptococcal antigenemia in HIV-positive patients with CD4(+) T-cell counts ≤200 cells/µL who were initiating ART, and also to evaluate the utility of the point-of-care urine lateral flow assay (LFA) cryptococcal antigen test using two different diluents, compared to gold standard serum antigen testing, as a screening tool. Urine and serum of outpatients initiating antiretroviral therapy at two hospitals in Mwanza were tested for cryptococcal antigen, and demographic and clinical characteristics were obtained using structured questionnaires and patients' files. Patients with asymptomatic cryptococcal antigenemia received oral fluconazole in accordance with World Health Organization recommendations. RESULTS Among 140 patients screened, 10 (7.1%) had asymptomatic cryptococcal antigenemia with a positive serum cryptococcal antigen. Four of these ten patients had CD4 counts between 100 and 200 cells/µL. The prevalence of cryptococcal antigen detected in urine using a standard (older) and a test (newer) diluent were 44 (31.4%) and 19 (13.6%), with Kappa coefficients compared to serum of 0.28 and 0.51 (p<0.001 for both). Compared to the new LFA diluent for urine cryptococcal antigen, the standard diluent had higher sensitivity (100% versus 80%) but lower specificity (74% versus 92%) using serum cryptococcal antigen as a gold standard. CONCLUSIONS Our findings suggest that HIV-positive outpatients with CD4 counts <200 cells/µL, rather than 100, should be screened for asymptomatic cryptococcal antigenemia given its association with mortality if untreated. Agreement of the urine LFA with the serum LFA was not sufficient to recommend routine screening with urine LFA.