Secondary Logo

Institutional members access full text with Ovid®

Dore Gregory J.; Li, Yueming; Grulich, Andrew E.; Hoy, Jennifer F.; Mallal, Simon A.; Mijch, Anne M.; French, Martyn A.; Cooper, David A.; Kaldor, John M.
Original Papers: Epidemiology and Social: PDF Only

Objective:To explore trend in cumulative incidence of Kaposi's sarcoma (KS) and the level of immunodeficiency at KS diagnosis among people with AIDS in Australia.

Setting:Three hospital-based HIV units.

Study population:Retrospective cohort of 2580 people diagnosed with AIDS over the period 1983–1994, representing 45% of cases of AIDS in Australia over this period.

Methods:Data including date and CD4 T-lymphocyte count of KS diagnosis was abstracted from medical records. KS occurring as both an initial and subsequent AIDS illness was included. Three subcohorts were defined based on interval of AIDS diagnosis: 1983–1987, 1988–1990, 1991–1994. Cumulative risk estimates for KS development were calculated by the Kaplan–Meier method.

Results:KS was diagnosed in 716 people (27.8%), and in 451 (63%) of these as the initial AIDS illness. There was a decline over time in cumulative incidence of KS (P < 0.0005); the cumulative risk of KS at 1 year after AIDS diagnosis declined from 35% for those diagnosed with AIDS during 1983–1987 to 25% for 1991–1994. This decline was not due to a decline in homosexual HIV exposure category, and was independent of CD4 T-lymphocyte count at AIDS. In multivariate analysis independent risk factors for KS development were year of AIDS diagnosis (P = 0.003), male homosexuality (P = 0.003), and CD4 T-lymphocyte count at AIDS greater than 150×106/l (P = 0.02). A decline in median CD4 T-lymphocyte count at KS diagnosis was seen, from 67×106/l in 1984–1987 to 20×106/l for 1991–1994 (P < 0.0005).

Conclusion:The decline in incidence and later occurrence of KS suggest several hypotheses, including declining prevalence or reduced virulence of a KS cofactor.

© Lippincott-Raven Publishers.