Six-month outcomes of HIV-infected patients given short-course fluconazole therapy for asymptomatic cryptococcal antigenemia.

OBJECTIVE

In HIV-infected adults in sub-Saharan Africa, asymptomatic cryptococcal antigenemia at the time of antiretroviral therapy (ART) initiation is associated with more than 20% increased mortality. Provisional recommendations for treatment of asymptomatic cryptococcal antigenemia are neither well substantiated nor feasible in many resource-poor settings. After hospitals in Tanzania implemented a programme providing serum cryptococcal antigen (CrAg) screening with 4-week intensive fluconazole treatment for CrAg-positive patients, we were asked to assess the impact of this programme on mortality.

DESIGN

In this retrospective operational research study, we documented 6-month outcomes of HIV-infected adults who had had CD4 cell counts less than 200 cells/μl at the time of starting ART and had been screened for cryptococcal antigenemia over a period of 15 months.

METHODS

We randomly selected three CrAg-negative patients, matched for ART start date, for every CrAg-positive patient who had been identified and treated with the 4-week intensive fluconazole course. The primary outcome was 6-month mortality in CrAg-positive and CrAg-negative groups.

RESULTS

Mortality of CrAg-positive HIV-infected adults who received short-course fluconazole was noninferior to CrAg-negative adults. At 6 months, 16 of 18 CrAg-positive and 46 of 54 CrAg-negative patients were alive [88.9% versus 85.1%, -3.9% absolute difference (one-sided 90% confidence interval +10.8%)]. No deaths in the CrAg-positive group seemed to be due to cryptococcal meningitis.

CONCLUSION

This study suggests that even short-course intensive fluconazole could reduce the mortality of patients with asymptomatic cryptococcal antigenemia. Further studies are needed to confirm if this dose is both optimal for patient survival and feasible for wide implementation in resource-poor settings where mortality of cryptococcal disease is highest.