Epidemiology of AIDS-related Kaposi's sarcoma in Europe over 10 years

Objectives: To determine the incidence and risk factors associated with Kaposi's sarcoma (KS) occurrence as an AIDS-defining condition or after the diagnosis of AIDS. Design: Multicentre retrospective cohort study of AIDS in Europe database from 52 clinical centres in 17 European countries. Methods: Patients' charts (n = 6546) were reviewed and collected in the database of the AIDS in Europe Study Group from 1979 to 1989. At the time of AIDS diagnosis 1394 patients had KS, whereas an additional 525 others developed KS after AIDS diagnosis. Univariate analysis and development of multivariate models determined factors associated with KS occurrence. Results: Frequency of KS as an AIDS-defining condition significantly declined over time (P < 0.0001). In our cohort of patients, home-/bisexual men from central Europe with CD4 cell counts > 150x10(6)/l were statistically more likely to develop KS at the time of AIDS diagnosis (P< 0.0001). For patients with an AIDS diagnosis other than KS, the probability of developing KS during the follow-up was 10 and 24% after 12 and 36 months, respectively. Variables significantly associated with a further KS development were transmission group, central European residence, previous herpes simplex infection other than ulcers, and low CD4 cells (< 150x10(6)/l). Previous zidovudine therapy had no influence on KS appearance. For patients who developed KS subsequent to AIDS diagnosis, there was no significant decline of the incidence over the 10-year time period. Conclusions: This large cohort study clearly shows that demographic data such as sex, transmission group and region of Europe have a major influence on KS development. It also suggests that KS as an AIDS-defining disease occurs earlier in the course of the chronic HIV infection than other opportunistic diseases. Reasons for geographical variations and its declining frequency as an initial AIDS diagnosis remain undetermined.