Geographical patterns of Kaposi’s sarcoma, nonHodgkin lymphomas, and cervical cancer associated with HIV infection in five African populations

Abstract

The objective of this study is to describe the most recent geographical patterns of incidence of AIDS-related cancers, Kaposi's sarcoma (KS), nonHodgkin lymphoma (NHL), and cervical cancer in North African and subSaharan African populations. Data were extracted for the period 1998-2002 from five African population-based cancer registries: Kyadondo, Harare, Setif, Sousse, and Gharbiah. Age-standardized rates were calculated using the African standard population; a comparison was made between these populations by computing the standardized incidence ratio and 95% confidence intervals. The KS rate was found to be significantly higher in men than in women, and higher in Harare (women: 26.3/100,000; men: 50.4/100,000) and Kyadondo (women: 23.6/100,000; men: 30.2/100,000) than in the North African sites for both sexes (<0.3/100,000). In addition, the KS rate in women from Harare was similar to that for Kyadondo. Gharbiah presented the highest rates for NHL (women: 7 per 100,000; men: 11.9/100,000) for both sexes. We observed that Harare and Kyadondo had similar age-specific incidence in the high-risk age group for HIV/AIDS (15-49 years), and these rates were 4.5-fold higher in subSaharan populations than those in the North African sites. Thus, it was observed that the pattern of HIV prevalence is variable with the lowest prevalence in North African countries, intermediate prevalence in Uganda, and the highest prevalence in Zimbabwe. Our findings show that the incidence of NHL and cervical cancer, considered to be HIV/AIDS-related cancers, does not follow the pattern of HIV prevalence in the five studied African populations. Thus, the highest NHL incidence rate in both sexes in Gambia may be explained, at least in great part, by the highest hepatitis C virus prevalence observed there. Indeed, factors other than HIV infection likely contribute to their geographical patterns.