Background:  Cryptococcal disease remains a significant cause of morbidity and mortality in HIV-infected individuals in tropical settings, despite the introduction of ART. No large trial of fluconazole as primary prophylaxis has been done in Africa.

Methods:  We performed a prospective, double-blind randomized controlled trial comparing 200-mg fluconazole 3x per week to identical placebo in rural Uganda. We enrolled 1519, ART-naïve adults with CD4 counts <200. Patients were excluded if baseline cryptococcal antigen (CrAg) was positive. Trial drug was commenced and participants actively referred to local providers for ART. Participants were reviewed every 8 weeks and encouraged to attend if unwell. After enrolment, 1335 (88%) participants started ART at a median time of 74 days. Trial drug was continued until CD4 counts rose to 200 (median 197 days IQR 160 to 339). Primary end points were invasive cryptoccocal disease and all cause mortality. Survival and time to end points were assessed in intention to treat Kaplan-Meier survival analyses. Cox’s proportional hazard regression models analyses were used to adjust for the effect of ART and CD4 count.

Results:  In all, 760 participants received fluconazole and 759 received placebo:  19 participants developed definite cryptococcal disease, 1 on fluconazole, and 18 on placebo (log rank X² 15.36 p = 0.0001) with an adjusted Hazard Ratio (aHR) of 0.053 (95%CI 0.007 to 0.4, p = 0.004); 12 developed cryptococcus before starting ART (11 on placebo) and 7 while on ART (all placebo); fluconazole was effective both pre-ART (log rank X²  = 8.02, p = 0.0046) and post-ART (log rank X² = 7.45, p = 0.0064); 3 participants died of cryptococcal disease. There was no difference in all cause mortality between fluconazole (n = 100) and placebo (n = 98) arms. Fluconazole reduced esphageal Candida (aHR = 0.14, 95%CI 0.067 to 0.30, p = <0.0001), oropharyngeal and vaginal Candida (aHR = 0.15, 95%CI 0.097 to 0.22, p = <0.001). The frequency of elevated transaminases (>5x upper limit) was similar in both arms (aHR = 0.94, 95%CI 0.65 to 1.35 p = <0.47).

Conclusions:  Fluconazole proved to be safe and effective prophylaxis against cryptococcal disease, both before starting ART and during early ART. Cryptococcal infection was less common than in routine practice, probably because of the exclusion of those with positive CrAg. The case fatality rate was low because of rapid diagnosis and treatment. In programmatic use, fluconazole has a potential for even greater benefit.