Survival by sex and HIV status in patients with anal cancer in the USA between 2001 and 2019: a retrospective cohort study

Background The risk of anal cancer is increased among people with HIV, particularly men who have sex with men. Estimating survival by HIV status and sex and identifying groups at high risk is crucial for documenting prognostic differences between populations. We aimed to compare all-cause and anal cancer-specific survival in patients with anal cancer with and without HIV, stratified by sex, and to identify predictors of survival, stratified by HIV status. Methods In this retrospective cohort study, we used data from the HIV/AIDS Cancer Match Study of 13 population-based HIV and cancer registries throughout the USA. We included individuals aged 20-79 years diagnosed with invasive anal cancer between 2001 and 2019. To estimate associations between HIV status and both all-cause and anal cancer-specific mortality overall, we used Cox proportional hazards models, adjusting for year of and age at diagnosis, sex, race and ethnicity, histology, cancer stage, region, and treatment. We also calculated sex-specific adjusted hazard ratios (HRs). By HIV status, we identified characteristics associated with mortality. Models among people with HIV were further adjusted for AIDS status and HIV transmission risk group. Findings Between Jan 1, 2001, and Dec 31, 2019, 1161 (43 center dot 6%) of 2662 patients with anal cancer and HIV and 7722 (35 center dot 4%) of 21 824 patients without HIV died. HIV was associated with a 1 center dot 35 times (95% CI 1 center dot 24-1 center dot 47) increase in all-cause mortality among male patients and a 2 center dot 47 times (2 center dot 10-2 center dot 90) increase among female patients. Among patients with HIV, all-cause mortality was increased among non-Hispanic Black individuals (adjusted HR 1 center dot 19, 95% CI 1 center dot 04-1 center dot 38), people with AIDS (1 center dot 36, 1 center dot 10-1 center dot 68), people who inject drugs (PWID; 1 center dot 49, 1 center dot 17-1 center dot 90), patients with adenocarcinoma (2 center dot 74, 1 center dot 82-4 center dot 13), and those with no or unknown surgery treatment (1 center dot 34, 1 center dot 18-1 center dot 53). HIV was associated with anal cancer-specific mortality among female patients only (1 center dot 52, 1 center dot 18-1 center dot 97). Among patients with HIV, anal cancer-specific mortality was increased among patients with adenocarcinoma (3 center dot 29, 1 center dot 89-5 center dot 72), those with no or unknown treatment (1 center dot 59, 1 center dot 17-2 center dot 17), and PWID (1 center dot 60, 1 center dot 05-2 center dot 44). Interpretation HIV was associated with all-cause mortality among patients with anal cancer, especially women. Anal cancer-specific mortality was elevated among female patients with HIV. As screening for anal cancer becomes more widespread, examining the effects of screening on survival by HIV status and sex is crucial. Funding US National Cancer Institute Intramural Research Program. Published Elsevier Ltd.