RESUMO
OBJECTIVE: To determine mortality associated with a new cancer diagnosis in an urban, predominantly African-American, HIV-infected population. DESIGN: Retrospective cohort study. METHODS: All HIV-infected patients diagnosed with cancer between 1 January 2000 and 30 June 2010 were reviewed. Mortality was examined using Kaplan-Meier estimates and Cox proportional hazards models. RESULTS: There were 470 cases of cancer among 447 patients. Patients were predominantly African-American (85%) and male (79%). Non-AIDS-defining cancers (NADCs, 69%) were more common than AIDS-defining cancers (ADCs, 31%). Cumulative cancer incidence increased significantly over the study period. The majority (55.9%) was taking antiretroviral therapy (ART) at cancer diagnosis or started afterward (26.9%); 17.2% never received ART. Stage 3 or 4 cancer was diagnosed in 67%. There were 226 deaths during 1096 person years of follow-up, yielding an overall mortality rate of 206 per 1000 person years. The cumulative mortality rate at 30 days, 1 year, and 2 years was 6.5, 32.2, and 41.4%, respectively. Mortality was similar between patients on ART whether they started before or after the cancer diagnosis but was higher in patients who never received ART. In patients with a known cause of death, 68% were related to progression of the underlying cancer. CONCLUSION: In a large cohort of urban, predominantly African-American patients with HIV and cancer, many patients presented with late-stage cancer. There was substantial 30-day and 2-year mortality, although ART had a significant mortality benefit. Deaths were most often caused by progression of cancer and not from another HIV-related or AIDS-related event.