Autopsy findings in a human immunodeficiency virus-infected population over 2 decades: influences of gender, ethnicity, risk factors, and time.

OBJECTIVES: To examine autopsy pathology in an urban population infected with the human immunodeficiency virus (HIV) and to determine if age at death and disease frequencies are associated with gender, HIV risk factors, ethnicity, and therapeutic era. DESIGN AND METHODS: Retrospective analysis of autopsy data from 394 HIV-infected adults. The population was divided into 3 therapeutic eras for analysis: group A, 1979-1986; group B, 1987-1995; and group C, 1996-2000. RESULTS: Women died at significantly younger ages than men (mean +/- SEM age, 38.9 +/- 1.0 years vs 42.5 +/- 0.64 years), even after adjustment for risk factors, ethnicity, and therapeutic era. This age discrepancy occurred despite a lower prevalence of arteriosclerosis, cachexia, and hepatitis B in women and no significant differences in the frequencies of other infectious diseases. Whites had a longer survival than patients of other ethnicities (mean age at death, 44.7 +/- 1.2 years for whites, 39.9 +/- 0.80 years for blacks, and 41.3 +/- 0.87 years for Hispanic individuals). Renal, cardiac, and splenic pathologies, Mycobacterium avium-intracellulare (MAI) infection, and cachexia were more common in blacks than in whites and/or Hispanic individuals, and cytomegalovirus and systemic lymphoma were more common in whites and Hispanic individuals than in blacks. Diseases associated with intravenous drug use were hepatitis C, cirrhosis, and tuberculosis; those with all sexual risk factors, cytomegalovirus infection, herpes simplex virus infection, and Pneumocystis carinii pneumonia; and those with homosexual risk, Kaposi sarcoma and MAI infection. The prevalence of many disease entities changed over time: compared with the other groups, group C had lower prevalences of many viral and fungal illnesses, MAI infection, systemic lymphoma, cachexia, and Kaposi sarcoma and higher prevalences of hepatitis, cirrhosis, arteriosclerosis, staphylococcal and streptococcal infections, and traumatic lesions. When the data were adjusted for changing demographic and risk composition, the only significant changes in disease frequency for period C were decreased prevalences of PCP and Kaposi sarcoma and increased prevalences of cirrhosis and arteriosclerosis. CONCLUSIONS: Significant gender- and ethnicity-related differences in age of death occurred in this HIV-infected population, and these differences were not explained by the frequencies of diseases. The lower prevalences of PCP and Kaposi sarcoma in group C are likely a reflection of the impact of potent therapies on causes of mortality. The higher prevalences of cirrhosis and arteriosclerosis suggest that entities not targeted by antiretroviral reconstitution of immunity will play an increasingly important role in HIV-related mortality in the future.