Pneumocystis carinii pneumonia is considered the most common opportunistic infection in patients infected with human immunodeficiency virus (HIV). Patients with CD4 lymphocyte counts of less than 200 cells per mm³ (200 × 10⁶ per L) are most commonly affected. Primary prophylaxis with several agents, including dapsone, inhaled pentamidine and trimethoprim-sulfamethoxazole, dramatically decreases the risk of this opportunistic infection. The advent of potent combination antiretroviral therapy significantly increases CD4 lymphocyte counts up to the 200 to 500 cells per mm³ (200 to 500 × 10⁶ per L) range and sometimes higher in a short period of time. Whether this resultant increase in immune response is adequate to discontinue primary prophylaxis is unknown. Furrer and colleagues conducted a prospective study to determine the safety of discontinuing primary prophylaxis for P. carinii pneumonia in a group of patients taking antiretroviral therapy.

Patients from the Swiss HIV Cohort Study, a prospective study with ongoing enrollment of HIV-infected adults, were eligible for inclusion if they were receiving primary prophylaxis against P. carinii pneumonia and were taking combination antiretroviral medications. In addition, their CD4 counts after treatment had to be at least 200 cells per mm³ and at least 14 percent of the total lymphocyte count for a minimum of 12 weeks. After prophylaxis was discontinued, CD4 counts were measured every three months, and prophylaxis was resumed if the absolute CD4 count or percentage fell below study criteria thresholds. Patients were also instructed to contact their physician if symptoms of pneumonia or toxoplasmic encephalitis developed. The primary end point of the study was a diagnosis of P. carinii pneumonia, and the secondary end point was a diagnosis of toxoplasmic encephalitis.

For control purposes, the incidence of P. carinii pneumonia and toxoplasmic encephalitis was recorded annually for four years in a cohort of patients. Patients were stratified by CD4 counts at ranges of less than 200, 200 to 500, and more than 500 cells per mm³. Patients in the control group were not treated with combination antiretroviral medications.

Of the 279 patients enrolled, 242 were available for follow-up at the end of the study. Median patient age was 37 years, and approximately two thirds of the group were men. Median CD4 counts increased from 325 cells per mm³ at baseline to 422 cells per mm³ at the end of the study. The median nadir CD4 count was 110 cells per mm³. Trimethoprim-sulfamethoxazole was used for primary prophylaxis in 85 percent of the patients. Approximately 89 percent were taking three or more antiretroviral drugs, including a protease inhibitor. A total of 121 patients had a positive Toxoplasma gondii IgG antibody titer, theoretically placing them at increased risk for toxoplasmic encephalitis.

No patients in the study group were diagnosed with P. carinii pneumonia or toxoplasmic encephalitis during the follow-up period. Follow-up ranged from three to 19 months (median: 11.3 months). No other acquired immunodeficiency disease-defining illnesses were diagnosed in these patients during this time. The incidence of P. carinii pneumonia was 6 per 100 patient-years and 16 per 100 patient-years for toxoplasmosis in patients in the control group who had CD4 counts of less than 200 cells per mm³.

The authors conclude that primary prophylaxis against P. carinii pneumonia can be safely discontinued in patients whose CD4 counts are more than 200 cells per mm³ and 14 percent of total lymphocytes for at least 12 weeks. If cell counts decrease, prophylaxis should be resumed. The study did not evaluate the effects of discontinuing secondary prophylaxis, which is an important issue in a number of ongoing studies for P. carinii pneumonia, as well as several other opportunistic infections, including cytomegalovirus, retinitis and Mycobacterium avium complex disease.