We report 100 cases of Kaposi's sarcoma (KS) in children under 15 years of age treated at the Uganda Cancer Institute in the 6-year period 1989-1994. The incidence of childhood KS has risen more than 40-fold in the era of AIDS, and 78% of 63 cases tested were seropositive for HIV-1. There were 63 boys and 37 girls. The median age was 4 years and the median age of onset was 33 months. Tumour distribution was lymphadenopathic and muco-cutaneous, with 2 major patterns: pattern I, oro-facial dominant (79%); and pattern II, inguinal-genital dominant (13%). A newly described herpes-like virus is implicated as the cause of KS (KSHV), and DNA sequences of this virus were present in all of 8 childhood cases tested. If KSHV is a direct cause of KS, this tumour distribution in children suggests mucosal routes of virus entry, possibly during birth or breast feeding. The dramatic increase of childhood KS implies that the prevalence of causative factors is rising in Uganda.
PIP: During 1989-94 clinicians had treated 100 cases of histologically or clinically confirmed Kaposi's sarcoma (KS) in children under 15 years of age (median age, 4 years) at the Uganda Cancer Institute in Kampala. The male/female ratio was 1.7/1. 78% of the children tested positive for HIV, while 69% of all mothers were HIV positive. 81% of the mothers with an HIV-infected child also tested HIV positive. There were seven discordant child-mother pairs among the 34 pairs tested for HIV. Five children were HIV positive; yet their mothers were HIV negative. There were no child-mother pairs with KS, even though the prevalence of KS in HIV-seropositive adults (7-10%) would predict two mothers with KS. 39% of patients had an infection in a site later involved with KS. The most prominent KS-related findings were lymphadenopathy and mucocutaneous lesions. Scientists have implicated a newly described herpes-like virus as the cause of KS (KSHV). All eight children tested for DNA sequences of this virus had these DNA sequences in archival tissues. The areas of the body most affected by KS were orofacial (79%) and inguinal/genital (13%) areas, sites that favor the entry of a causative agent, e.g., KSHV. The median age of onset of KS symptoms was 33 months for HIV-positive, child-mother pairs and 34 months for HIV-negative, child-mother pairs, suggesting that exposure to a potentially infectious agent occurred at birth or early infancy. It was 36 months for those of unknown HIV status. The high concordancy of child-mother HIV seropositivity suggests that the KS agent could be spread perinatally or during breast feeding. The dramatic increase in the incidence of KS in children (40-fold increase) since the emergence of AIDS suggests that the prevalence of causative factors is also increasing.