Adjunctive dexamethasone not only failed to decrease mortality, it actually induced more disability and adverse events than placebo in patients with HIV-associated cryptococcal meningitis, according to a report published online Feb. 11 in the New England Journal of Medicine.

Glucocorticoids are inexpensive, readily available, and relatively safe for patients with central nervous system infections, and they are widely used for HIV-associated cryptococcal meningitis in regions where the burden of the infection is highest. Glucocorticoids are even recommended for this indication in some international guidelines, as they are thought to reduce intracranial pressure and inflammatory complications. But evidence of their usefulness from randomized controlled trials is sparse, said Dr. Jeremy Day of Oxford University Clinical Research Unit, Wellcome Trust Major Overseas Programme Vietnam, Ho Chi Minh City, and his associates.

They performed a double-blind, randomized trial to compare the effectiveness and safety of adjunctive dexamethasone against placebo, which involved 451 patients treated for 6 weeks at 13 hospitals in Indonesia, Laos, Thailand, Vietnam, Malawi, and Uganda. The study participants also received standard antifungal therapy including amphotericin B and fluconazole, as well as antiretroviral therapy and Pneumocystis prophylaxis with trimethoprim-sulfamethoxazole.

The trial was stopped prematurely when the safety committee found “dexamethasone was causing harm across key outcomes, including fungal clearance, adverse events, and disability outcomes.” Consequently, the study didn’t have the statistical power to show an effect of dexamethasone on the primary outcome measure – mortality at 10 weeks after randomization (4 weeks after study treatment ended).

Dexamethasone did reduce intracranial pressure more rapidly than placebo, but this effect didn’t translate into improved survival. Mortality at 10 weeks was 47% for the 224 patients assigned to dexamethasone and 41% for the 226 patients assigned to placebo, a nonsignificant difference. However, the drug’s effects changed over time: Hazard ratios for death were 0.77 at days 1-22 but rose to 1.94 at days 23-43 and to 2.50 at days 44-71. By 6-month follow-up, mortality risk showed a trend toward harm with dexamethasone, and was 9% higher in the intention-to-treat population and 11% higher in the per-protocol population analyses, Dr. Day and his associates said ( N. Engl J Med. 2016 Feb 11;374[6]: doi:10.1056/NEJMoa1509024 ). The rate of disability or death was significantly higher with dexamethasone than with placebo at both 10 weeks and 6 months, with odds ratios for a good outcome of only 0.42 and 0.49, respectively. Infections or infestations developed in 48 patients (21%) taking dexamethasone but only 25 (11%) of those taking placebo. Gastrointestinal disorders (13% vs. 7%), renal or urinary disorders (10% vs. 3%), and cardiac disorders (4% vs. 0%) also were significantly more frequent with dexamethasone, as were episodes of hyperglycemia, hypercreatinemia, hyperkalemia, and hyponatremia.

This study was supported by the United Kingdom Department for International Development, the Wellcome Trust, and the U.K. Medical Research Council through the Joint Global Health Trials Program. Dr. Day and his associates reported having no relevant financial disclosures.