Kaposi's sarcoma and sexually transmitted disease

Abstract

Kaposi's sarcoma (KS) is the most common neoplasm among people infected with human immunodeficiency virus (HIV). A compelling body of evidence suggests that KS is caused by a novel human herpesvirus, referred to as Kaposi's sarcoma-associated herpesvirus (KSHV). The purpose of this investigation was to characterize possible modes of KSHV transmission by examining associations between KS and prior infection with selected pathogens among people with the acquired immunodeficiency syndrome (AIDS).The HIV/AIDS Reporting System (HARS) was used to identify a cohort of 3,873 residents of thirteen counties in western Washington state who were diagnosed with AIDS during the period 1982--92. Incident cases of KS among cohort members were documented through HARS and by matching the cohort to the Cancer Surveillance System, a population-based cancer registry. Data from communicable disease registries were linked to the cohort to identify individuals previously infected with treponema pallidum, hepatitis-B virus (HBV), hepatitis-A virus (HAV), salmonella spp., salmonella spp., giardia spp., campylobacter spp., and entamoeba histolytica .As documented in earlier studies, the risk of KS in this cohort was far greater among homosexual and bisexual men than other HIV risk groups. By Cox proportional hazards model, simultaneously adjusted for mode of HIV transmission, age at AIDS diagnosis, and year of AIDS diagnosis, we observed a modest, positive association between KS that occurred following the diagnosis of AIDS and prior infection with treponema pallidum (proportional hazards (PH) = 1.53, 95 percent confidence interval (CI) = 1.18--1.99). We also observed a modest, positive association between KS and prior infection with any of six enteric pathogens examined in this study (PH = 1.21, 95 percent CI = 0.86--1.70), and individually with five of the six enteric pathogens considered. However, most of the positive associations were modest and were based on a small number of infected individuals. Modest, inverse associations between KS and prior HAV and HBV infection were also observed.Although some associations observed in this investigation may have been underestimated due to incomplete surveillance in the participating registries, prior infections with these pathogens were not strong or consistent indices of the likely mode of KSHV infection.