Main

Before the human immunodeficiency virus (HIV) epidemic, Kaposi's sarcoma had a much greater geographical variation in incidence than most other malignancies. It was as common in parts of sub-Saharan Africa, such as Uganda, as colon cancer is in Europe and the USA, representing up to 9% of all cancers in men (Oettlé, 1962; D'Oliveira and Torres, 1972; Templeton, 1981; Hutt, 1984; Cook-Mozaffari et al, 1998). Kaposi's sarcoma was also endemic, although much rarer, in counties around the Mediterranean, particularly in Italy, Greece and the Middle East, but was almost nonexistent elsewhere, except in immigrants from these endemic countries (Biggar et al, 1984; Grulich et al, 1992; Hjalgrim et al, 1996). The pattern of geographical variation in incidence correlates broadly with the worldwide distribution of Kaposi's sarcoma-associated herpesvirus (KSHV; human herpesvirus-8 (HHV-8)), which is considered to be a necessary cause of Kaposi's sarcoma (Chang et al, 1994, 1996; Gao et al, 1996; Simpson et al; 1996; Boshoff, 1999). Here, we review published data on the prevalence of infection with KSHV across Africa and make comparison with the distribution of Kaposi's sarcoma before and since the start of the HIV epidemic.

Methods

Studies published up to the end of March 2002 that looked for evidence of infection with KSHV among people living in Africa were identified through a Medline search, supplemented by an examination of references given in identified papers and by searching the index of relevant journals. No restriction was placed on the language of publication, and foreign language papers were translated. Only peer-reviewed journals were considered and no attempt was made to identify unpublished studies or to obtain unpublished data from published studies. Conference abstracts, which are often based on preliminary analyses, were excluded. We adopted the criteria that had been used within individual studies for evidence of KSHV infection. Many of the identified studies supplied prevalence estimates for children and adults, or estimates derived from the use of more than one KSHV assay–the full range of available estimates are shown in Table 1. Cumulative incidence (for ages 0–64) of Kaposi's sarcoma in men, prior to 1980, was obtained from a published report by Cook-Mozaffari et al (1998). Where regional rates within a given country were available, they were used to calculate a country-wide mean incidence rate. Brief details of the assays used in each study are included in the footnote to the table, but this report is not intended as a review of methods of detection of KSHV infection.

Table 1 Reported prevalence of evidence of infection with KSHV in Africa, in relation to the cumulative incidence of Kaposi's sarcoma in men in the years before the HIV epidemic
Full size table

Results and discussion

In all, we identified 28 studies that had measured the prevalence of KSHV infection in 16 African countries. Of these, three studies identified evidence of KSHV-DNA in peripheral blood cells (Chang et al, 1996; Kasolo et al, 1997; Belec et al, 1998), two used a lytic antibody assay alone (He et al, 1998; Lampinen et al, 2000), five used a latent antibody assay alone (Gao et al, 1996; Ariyoshi et al, 1998; Bestetti et al, 1998; Bourboulia et al, 1998; Sitas et al, 1999) and 18 used both a lytic and a latent assay (Lennette et al, 1996; Simpson et al, 1996; Mayama et al, 1998; Olsen et al, 1988; de Thé et al, 1999; Ablashi et al, 1999; Andreoni et al, 1999, 2002; Enbom et al, 1999; Gessain et al, 1999; Wilkinson et al, 1999; Engels et al, 2000; Rezza et al, 2000; Gaye Serraino-Diallo et al, 2001; Kakoola et al, 2001; Nuvor et al, 2001; Serraino et al, 2001; Wawer et al, 2001). The results are presented in Table1, together with the estimated cumulative incidence of Kaposi's sarcoma in men aged 0–64 years in the period before HIV infection (Cook-Mozaffari et al, 1998).

KSHV is common in countries such as Uganda and Cameroon, where Kaposi's sarcoma was relatively frequent, but the virus is also common in countries such as Botswana and the Gambia, where Kaposi's sarcoma rarely occurred before the spread of HIV infection. There was no evidence in any of the studies that the prevalence of KSHV (however it was measured) differed between men and women. Before the onset of the HIV epidemic in the 1980s, Kaposi's sarcoma showed extreme geographical variation in incidence even within the African continent. Narrow belts of relatively high incidence stretched westward across the former Zaire to the coast of Cameroon and southward down the rift valley to Malawi. In all of these areas, Kaposi's sarcoma was more common in men than in women (Cook-Mozaffari et al, 1998).

There is no evidence that the prevalence of KSHV has changed in Africa since the spread of HIV. Indeed, KSHV, a γ2 herpesvirus, is thought to have coexisted alongside Homo sapiens since their origin (reviewed by Hayward, 1999). Furthermore, the study by de Thé et al (1999) has shown that anti-KSHV antibodies were present in Africa prior to the HIV epidemic at levels similar to those seen today. Why then, was Kaposi's sarcoma not more common in African countries that have a high prevalence of KSHV and why, when the virus is equally prevalent in men as it is in women, was Kaposi's sarcoma so much more frequent in men? Both anomalies suggest the importance of further (as yet unknown) cofactors in the aetiology of the tumour.

Parts of Africa with a high prevalence of HIV and where Kaposi's sarcoma was relatively common even in the years before acquired immunodeficiency syndrome (AIDS) have seen an explosion in the incidence of the tumour. In the last 10–15 years, the incidence of Kaposi's sarcoma has increased about 20-fold in Uganda and Zimbabwe, such that it is now the most common cancer in men and the second most common in women (Wabinga et al, 1993, 2000; Bassett et al, 1995). As a result of the HIV epidemic, the incidence of Kaposi's sarcoma has also increased in countries where it was previously relatively rare, but where KSHV was prevalent. For example, between 1988 and 1996, the incidence of Kaposi's sarcoma has risen at least three-fold in South Africa and continues to increase as the HIV epidemic grows. Data from Johannesburg, South Africa, show that incidence rates of Kaposi's sarcoma have doubled in men, but have increased seven-fold in women, such that the sex ratio of 7 : 1 in males vs females in 1988 has now declined to only 2 : 1 (Sitas and Newton, 2000). The role of other cofactors in the aetiology of Kaposi's sarcoma may, therefore, be less relevant in the presence of HIV infection than they seem to have been for the development of classical Kaposi's sarcoma prior to the spread of HIV.