Primary prophylaxis for Pneumocystis carinii pneumonia: a randomized trial comparing cotrimoxazole, aerosolized pentamidine and dapsone plus pyrimethamine.

ABSTRACT

OBJECTIVE: To compare the efficacy and tolerance of monthly aerosolized pentamidine versus cotrimoxazole versus dapsone plus pyrimethamine to prevent the initial episodes of Pneumocystis carinii pneumonia (PCP) in HIV-infected patients. DESIGN: An open randomized clinical trial. PATIENTS AND METHODS: HIV-infected patients (n = 331) with CD4 cell counts < 200 x 10(6)/l or with AIDS but without a history of PCP or cerebral toxoplasmosis (CT) were randomized to receive pentamidine (300 mg every 4 weeks), cotrimoxazole (160/800 mg 3 days a week) or dapsone plus pyrimethamine (100 and 25 mg weekly). If immunoglobulin G (IgG) antibodies to Toxoplasma were present, patients in the first two groups were randomized further to 25 mg pyrimethamine per week or to no treatment. RESULTS: The mean follow-up was 313 days (range, 30-670 days). The three groups were homogeneous for age, sex, risk group for HIV infection, initial CD4 cell count and mean follow-up. PCP developed in 16 patients, with an estimated cumulative probability of 5.3% at 1 year of follow-up. The PCP rate per year of observation, using an intention-to-treat analysis, was 5.6% [95% confidence interval (CI), 0.9-10.3], 3% (95% CI, 0-6.3) and 8.3% (95% CI, 2.8-13.8) in the groups treated with pentamidine, cotrimoxazole and dapsone plus pyrimethamine, respectively (P > 0.05). Moderate or severe side-effects were observed in one patient on pentamidine, 10 on cotrimoxazole and nine on dapsone plus pyrimethamine (P < 0.05); the study drug had to be discontinued in no, 10 and six patients, respectively (P < 0.05). Neither cotrimoxazole alone nor pyrimethamine combined with dapsone or cotrimoxazole prevented initial episodes of toxoplasmosis among patients with IgG antibodies to Toxoplasma gondii. CONCLUSIONS: Low-dose thrice-weekly cotrimoxazole or weekly dapsone plus pyrimethamine was not significantly worse (differences > 15% would have been detected with 90% certainty) than monthly aerosolized pentamidine in preventing a first episode of PCP in patients at high risk, but aerosolized pentamidine was better tolerated.