Role of zidovudine antiretroviral therapy in the pathogenesis of acquired immunodeficiency syndrome-related lymphoma.

The role of zidovudine and other antiretroviral agents in the pathogenesis of acquired immunodeficiency syndrome (AIDS)-related lymphomas has been somewhat controversial. In an attempt to elucidate the precise role of antiretroviral agents in the subsequent development of AIDS-related lymphoma, we performed a population-based, case-control study of human immunodeficiency virus (HIV)-seropositive patients with intermediate- or high-grade lymphoma in Los Angeles County, California, in which information regarding use of antiretroviral medications was ascertained. Diagnostic biopsy material was reviewed to confirm intermediate-or high-grade lymphoma. A structured interview, conducted with all cases and controls, included information about use of zidovudine and other antiretroviral agents. A total of 112 HIV-infected homosexual/bisexual men with lymphoma were matched to 112 homosexual/bisexual men with asymptomatic HIV infection; 49 of the lymphoma cases were also matched to 49 additional controls with AIDS, as defined by conditions other than lymphoma. Positive histories of zidovudine use were reported by 44 (39%) lymphoma cases, 24 (21%) asymptomatic HIV controls, and 21 (42%) AIDS controls. The average duration of zidovudine use up to 12 months before lymphoma diagnosis was 19.0 +/- 13.0 months (mean +/- SD) for the lymphoma cases, 12.6 +/- 10.5 months for the asymptomatic controls, and 11.0 +/- 7.1 months for the AIDS controls. When comparing the 49 HIV-positive lymphoma cases with their 49 matched AIDS controls, all of whom were diagnosed with AIDS during the same time period, the matched relative odds of lymphoma associated with prior use of zidovudine was 0.43 (95% confidence interval [CI] = 0.17 to 1.12). In comparing all 112 lymphoma cases with 49 AIDS controls, the unmatched relative odds of lymphoma associated with zidovudine use was 0.93 (95% confidence interval = 0.47 to 1.83). One lymphoma case and no AIDS control cases had a history of didanosine use; no lymphoma case or AIDS control cases had taken zalcitabine. We conclude that zidovudine is not associated with an increased risk of development of lymphoma among HIV-infected homosexual or bisexual men.