The effect of aging on the incidence of Kaposi's sarcoma among HIV-positive individuals with known dates of seroconversion

In the general population, the incidence of the most common types of cancer (i.e., carcinomas of the lung, breast, colon, stomach, prostate and bladder) increases with age among both males and females and in different geographic areas.1,2 This pattern is explained by the multistage model for carcinogenesis, which posits that cancer arises as a consequence of multiple genetic hits acquired with aging.3 Alternatively, it has been hypothesized that this age-related increase in incidence can be attributed, at least in part, to the diminishing immunologic function associated with aging. However, among persons with severe immunosuppression, such as those who are iatrogenically immunosuppressed after organ transplantation or persons infected with the human immunodeficiency virus (HIV), the incidence of the most common types of cancer is no higher than that among immunocompetent persons. By contrast, the incidence of both non-Hodgkin’s lymphoma (NHL) and Kaposi’s sarcoma (KS), which are rare in immunocompetent persons younger than 40 years of age, has been reported to be quite high among persons with severe immunosuppression.4,5 KS is a mesenchymal tumor that, until the appearance of the HIV epidemic, was rare in the general population of industrialized countries; it was generally found in elderly men of Mediterranean or Eastern European origin.6 With regard to the pathogenesis of KS, an infectious agent – the human herpesvirus 8 (HHV-8)—has been isolated in all forms of the disease,7,8 and epidemiologic, serologic and molecular evidence has proved that infection with HHV-8 is necessary for the development of KS. Nonetheless, it has been difficult to determine the magnitude of the association between KS and HHV-8 infection because the currently used serologic assays are limited in terms of their sensitivity and specificity9 and because other, as yet unknown, cofactors exist that are likely to play a role in the development of KS.10,11 Given that HIV-infected individuals have been shown to live longer since highly active antiretroviral therapy (HAART) was introduced,12 it is now possible to study thoroughly the relationship among immunosuppression, aging and cancer. In a study13 that investigated whether or not aging affects the risk of developing HIV-associated KS and lung cancer, using data from the AIDS-Cancer Registry Match Study, the incidence of KS appeared to increase through 39 years of age and to decline thereafter. However, no information was available from this study on the duration of HIV infection, the level of immunosuppression or the prevalence of HHV-8 infection (i.e., the main determinant of KS8). We investigated the relationship among HIV-induced immunosuppression, aging and KS, taking into account the duration of HIV infection, the level of immunosuppression and, for a subgroup of the study population, the prevalence of HHV-8 infection. This study used data from the Italian HIV-Seroconversion Study. The Seroconversion Study, which was begun in 1987, is a prospective incident cohort study of HIV seroconverters who belong to different HIV exposure categories and for whom it is possible to estimate the date of seroconversion with a certain degree of accuracy, specifically, persons for whom the results of a negative HIV test followed by a positive test within 24 months are available; the date of seroconversion is estimated as the midpoint in time between the most recent negative HIV test result and the first positive result. A detailed description of the Seroconversion Study has been published elsewhere.14,15 Residual sera samples for detection of antibodies against HHV-8 were available for some of the individuals enrolled as of June 1996 in the Seroconversion Study. A single HHV-8 serologic measurement was performed: the median time that had elapsed from the date of HIV seroconversion to the date of HHV-8 testing was 0.7 years (interquartile range: 0.4–1.6 years). HHV-8 seropositivity was determined by HHV-8 lytic antigens, using an immunofluoroscence assay with a body cavity B-cell lymphoma cell line. A detailed description of the methodology used has been reported elsewhere.16 For the present analysis, we applied the Kaplan-Meier method and the Cox proportional hazards model.17 The date of HIV seroconversion was considered time-zero of the analyses, and KS was considered the end point. In particular, we estimated the KS-free survival time as the time that had elapsed from the date of HIV seroconversion to the date of diagnosis of KS (independently of whether or not it was the first AIDS-defining disease), or the date of death or the end of the study (i.e., December 2000). The