Prevalence and clinical predictors of Pneumocystis colonization among HIV-infected men.

Abstract

The epidemiology and transmission of Pneumocystis are poorly understood. The incidence of colonization, or detection of organisms without signs of disease, has been debated, and risk factors for colonization are largely unknown. To determine the rate of Pneumocystis colonization among HIV-infected patients at autopsy and analyze associated clinical variables. Subjects were selected from the Multicenter AIDS Cohort Study. Subjects who died from causes other than Pneumocystis pneumonia and consented to autopsy were included in analysis. DNA was extracted from lung tissue, and nested PCR was performed to detect the presence of Pneumocystis. Clinical data were obtained from the Multicenter AIDS Cohort database. Univariate and multivariate analyses were performed to determine predictors of Pneumocystis colonization. Pneumocystis DNA was detected in 42 of 91 (46%) subjects by nested PCR. Clinical variables such as CD4 cell count, use of Pneumocystis prophylaxis or antiretroviral drugs, and history of previous Pneumocystis pneumonia were not related to risk of colonization. Multivariate analysis demonstrated that cigarette smoking was related to an increased risk of colonization [odds ratio (OR), 4.5; 95% confidence interval (CI), 1.27-15.6; P = 0.02] and risk also varied by city of residence (OR, 0.12; 95% CI, 0.03-0.45; P = 0.002 for living in Los Angeles). This study found a high rate of Pneumocystis colonization among HIV-infected patients. We also identified cigarette smoking and city of residence as novel, independent risk factors for colonization. The role of subclinical colonization in disease transmission and the effects of Pneumocystis colonization on the lung require further study.