R. Newton et al., A case-control study of human immunodeficiency virus infection and cancer in adults and children residing in Kampala, Uganda, INT J CANC, 92(5), 2001, pp. 622-627
Uganda offers a unique setting in which to study the effect of human immuno
deficiency virus-l (HIV-l) on cancer. HIV-l is prevalent there, and cancers
which are known to be HIV-associated, such as Kaposi's sarcoma and Burkitt
's lymphoma, are endemic. Adults residing in Kampala, Uganda, presenting wi
th cancer in city hospitals were interviewed and had an HIV test. Of the 30
2 adults recruited, 190 had cancers with a potentially infectious aetiology
(cases). The remaining 112 adults with tumours not known to have an infect
ious aetiology formed the control group. In addition, 318 children who were
also Kampala residents were recruited and tested for HIV: 128 with cancer
(cases) and 190 with non-malignant conditions (controls). HIV seroprevalenc
e was 24% in adult controls and 6% in childhood controls. The odds of HIV s
eropositivity among cases with specific cancers (other than Kaposi's sarcom
a in adults) were compared with that among controls, using odds ratios (ORs
), estimated with unconditional logistic regression. All ORs were adjusted
for age (<5, 5-14, 15-19, 30-44, 45+) and sex and, in adults, also for the
number of lifetime sexual partners (1 or 2, 3-9, 10+). In adults, HIV infec
tion was associated with a significantly (p < 0.05) increased risk of non-H
odgkin's lymphoma COR 6.2, 95% confidence interval (CI) 1.9-19.9, based on
2 1 cases] and conjunctival squamous-cell carcinoma (OR 10.9, 95% CI 3.1-37
.7, based on 22 cases) but not with cancer at other common sites, including
liver and uterine cervix. In children, HIV infection was associated with a
significantly increased risk of Kaposi's sarcoma (OR 94.9, 95% CI 28.5-315
.3, based on 36 cases) and Burkitt's lymphoma (OR 7.5, 95% CI 2.8-20.1, bas
ed on 33 cases) but not with other cancers. The pattern of HIV-associated c
ancers in Uganda is broadly similar to that described elsewhere, but the re
lative frequency of specific cancers, such as conjunctival carcinoma, in HI
V-infected people differs.