Adjunctive Dexamethasone in HIV-Associated Cryptococcal Meningitis

Abstract

Background: Cryptococcal meningitis associated with human immunodeficiency virus (HIV) infection causes more than 600,000 deaths each year worldwide. Treatment has changed little in 20 years, and there are no imminent new anticryptococcal agents. The use of adjuvant glucocorticoids reduces mortality among patients with other forms of meningitis in some populations, but their use is untested in patients with cryptococcal meningitis.

Methods: In this double-blind, randomized, placebo-controlled trial, we recruited adult patients with HIV-associated cryptococcal meningitis in Vietnam, Thailand, Indonesia, Laos, Uganda, and Malawi. All the patients received either dexamethasone or placebo for 6 weeks, along with combination antifungal therapy with amphotericin B and fluconazole.

Results: The trial was stopped for safety reasons after the enrollment of 451 patients. Mortality was 47% in the dexamethasone group and 41% in the placebo group by 10 weeks (hazard ratio in the dexamethasone group, 1.11; 95% confidence interval [CI], 0.84 to 1.47; P=0.45) and 57% and 49%, respectively, by 6 months (hazard ratio, 1.18; 95% CI, 0.91 to 1.53; P=0.20). The percentage of patients with disability at 10 weeks was higher in the dexamethasone group than in the placebo group, with 13% versus 25% having a prespecified good outcome (odds ratio, 0.42; 95% CI, 0.25 to 0.69; P<0.001). Clinical adverse events were more common in the dexamethasone group than in the placebo group (667 vs. 494 events, P=0.01), with more patients in the dexamethasone group having grade 3 or 4 infection (48 vs. 25 patients, P=0.003), renal events (22 vs. 7, P=0.004), and cardiac events (8 vs. 0, P=0.004). Fungal clearance in cerebrospinal fluid was slower in the dexamethasone group. Results were consistent across Asian and African sites.

Conclusions: Dexamethasone did not reduce mortality among patients with HIV-associated cryptococcal meningitis and was associated with more adverse events and disability than was placebo. (Funded by the United Kingdom Department for International Development and others through the Joint Global Health Trials program; Current Controlled Trials number, ISRCTN59144167.).

Figure 1. Enrollment, Randomization, and Follow-up

CM denotes cryptococcal meningitis, GI gastrointestinal, and HIV human immunodeficiency virus.

Figure 2. Survival among All Patients and According to Continent

Shown are Kaplan–Meier survival estimates for all patients (Panel A) and for those in Africa (Panel B) and Asia (Panel C) during the 6 months of follow-up. By 10 weeks (the cutoff for the primary outcome), 106 of 224 patients (47%) in the dexamethasone group and 93 of 226 (41%) in the placebo group had died. At 6 months, the estimated risks of death were 57% and 49%, respectively.

Figure 3. Quantitative Fungal Counts in Cerebrospinal Fluid (CSF)

Shown are the CSF quantitative fungal counts in the dexamethasone group (Panel A) and placebo group (Panel B). Study day 1 corresponds to the day of randomization. All recorded CSF quantitative counts are shown, including those in patients who subsequently died. The gray lines indicate data for individual patients, and the solid line shows scatterplot smoothing based on local regression. The decrease in CSF fungal counts, as measured in colony-forming units (CFU) per milliliter, during the first 14 days was significantly slower among patients in the dexamethasone group than among those in the placebo group.