Long term 5-year survival of persons with cryptococcal meningitis or asymptomatic subclinical antigenemia in Uganda

Abstract

Data presented previously as an abstract at the 2011 CUGH Global Health Conference in Montreal, Canada on 15 Nov 2011. The long-term survival of HIV-infected persons with symptomatic cryptococcal meningitis and asymptomatic, subclinical cryptococcal antigenemia (CRAG+) is unknown. We prospectively enrolled 25 asymptomatic, antiretroviral therapy (ART)-naïve CRAG+ Ugandans with CD4<100 cells/mcL who received pre-emptive fluconazole treatment (CRAG+ cohort) and 189 ART-naïve Ugandans with symptomatic cryptococcal meningitis treated with amphotericin (CM cohort). The 10-week survival was 84% (95%CI: 70-98%) in the CRAG+ cohort and 57% (95%CI: 50%-64%) in the CM cohort. The CRAG+ cohort had improved five-year survival of 76% (95%CI: 59%-93%) compared to 42% (95%CI: 35%-50%) in the CM cohort (P = 0.001). The two cohorts had similar immunosuppression pre-ART with median CD4 counts of 15 vs. 21 CD4/mcL in the CRAG+ and CM cohorts, respectively (P = 0.45). Despite substantial early mortality, subsequent 5-year survival of persons surviving 6-months was excellent (>88%), demonstrating that long term survival is possible in resource-limited settings. Pre-ART CRAG screening with preemptive fluconazole treatment and improved CM treatment(s) are needed to reduce AIDS-attributable mortality due to cryptococcosis which remains 20-25% in sub-Saharan Africa.

Figure 1. Long term survival among HIV-infected persons with cryptococcosis in Uganda.

The Kaplan-Meier survival curve displays the long term survival of 25 asymptomatic persons who tested positive for serum cryptococcal antigen (CRAG+) treated with fluconazole and ART compared to 189 symptomatic patients with cryptococcal meningitis (CM) treated with amphotericin B induction and then fluconazole consolidation therapy and then ART. Diamonds represent censored data. Persons were right-hand censored if they were still living at the time of analysis (n = 82), if they transferred their care to another clinic (n = 7), or if they were lost to follow up (n = 9). Survival table showing survival rate, number of deaths, number censored, and number of survivors at yearly intervals for each cohort.