Pneumocystis carinii pneumonia (PCP) is a significant cause of morbidity and mortality in patients infected with human immuno-deficiency virus (HIV). Currently, the treatment of choice for PCP is trimethoprim-sulfamethoxazole; however, many patients cannot tolerate this drug. Common alternatives include dapsone and pentamidine, but these agents are less effective and are also associated with a high rate of intolerance. Atovaquone is a newer agent that has been shown to be effective in treating PCP. However, its role as a prophylactic agent has not been previously evaluated. El-Sadr and colleagues from the AIDS Clinical Trials Group and the Community Program for Clinical Research on AIDS compared the effectiveness of atovaquone with that of dapsone in preventing PCP in patients with HIV infection who were intolerant to trimethoprim-sulfamethoxazole.

Two groups of HIV-infected patients 13 years of age and older were eligible for this multicenter, open-label, randomized trial. The primary prophylaxis group included patients who had a CD4+ lymphocyte count of less than 200 per mm³ (200 × 10⁶ per L) or less than 15 percent of the total lymphocyte count. The secondary prophylaxis group included patients who had a previous history of PCP. Patients were randomized to receive either 100 mg of dapsone daily or 1,500 mg of atovaquone suspension daily. Patients in the dapsone group who had CD4+ counts of less than 100 per mm³ (100 × 10⁶ per L) and a positive serology for toxoplasma were also encouraged to take weekly pyrimethamine and leucovorin. If adverse events or PCP developed in any patient, he or she could be switched to the alternative study drug. The primary end point of the study was the development of PCP, as confirmed by histology or cytology of the organism in tissue or sputum specimens. Probable causes included clinical, radiographic and blood gas evidence consistent with PCP and the exclusion of other diagnoses. Secondary end points included drug intolerance necessitating its discontinuation, PCP pneumonia resulting in death, or confirmed or probable toxoplasmosis.

A total of 1,057 patients enrolled in the study. Most patients were men approximately 38 years of age with baseline median CD4+ counts of 60 per mm³ (60 × 10⁶ per L). Approximately 20 percent of the patients were black, and 65 percent were white. Almost 50 percent of the patients had an established diagnosis of AIDS, and 28 percent had a prior history of PCP. At baseline, only 59 percent of the patients were receiving antiretroviral therapy; however, by the 24-month follow-up visit this number had increased to 93 percent.

At the end of the study, the mean follow-up time for all patients was 27 months and at this time, 617 of those originally enrolled were still alive. Of these, 289 were still taking the study-supplied dapsone or atovaquone, 214 were taking the originally prescribed drug and 75 had switched to a different treatment. A total of 257 patients had either confirmed or probable PCP, including 122 patients in the atovaquone group and 135 in the dapsone group. The cumulative percentages for PCP development over the two-year study period were 27 percent in the atovaquone group and 30 percent in the dapsone group. Toxoplasmosis developed in only seven patients, with essentially no difference between treatment groups. Of the 440 patients who died during the course of the study, 233 were in the atovaquone group and 208 in the dapsone group. The number of patients who had to stop taking medication because of intolerance was similar between groups. However, hypersensitivity reactions and anemia were more common in the dapsone group.

The authors conclude that atovaquone and dapsone are equally effective in preventing PCP in patients with HIV infection. Atovaquone appears to be better tolerated and may be the preferred drug in patients who cannot take trimethoprim-sulfamethoxazole.

editor's note: This study supports a role for a fourth drug as prophylaxis against PCP in patients whose CD4+ counts are less than 200 per mm³ (200 × 10⁶ per L) or in patients with a prior history of PCP. However, in the era of combination antiretroviral therapy, the best “prophylaxis” is a three-drug regimen directed against the virus.—j.t.k.