Effect of the introduction of screening for cancer precursor lesions on anal cancer incidence overtime in people living with HIV: a nationwide cohort study

Background Incidence of anal cancer is high in people living with HIV, particularly in men who have sex with men (MSM). Screening for and treatment of precursor lesions might prevent progression to anal cancer in people living with HIV. We examined trends in incidence of and mortality after anal cancer diagnosis in people living with HIV, including the effect of screening from 2007 onwards, in the Netherlands.Methods In this observational cohort study, we analysed data from the ongoing open nationwide Dutch AIDS Therapy Evaluation in the Netherlands (ATHENA) cohort. We included all consenting adults living with HIV and identified all primary anal squamous cell carcinoma. We reported temporal trends in incident anal cancer cases from Jan 1, 1996, to Dec 31, 2020, and all-cause and anal cancer-related mortality in individuals diagnosed with anal cancer. Multivariable Poisson regression was used to explore risk factors for incident anal cancer and multivariable Cox regression was used to explore risk factors for anal cancer-related mortality.Findings Among 28 175 individuals in HIV care (59middot7% MSM), 227 primary anal cancer cases were diagnosed. Despite the increasing average age of the cohort, crude incidence rates of anal cancer in MSM declined slowly over time, from 107middot0 (95% CI 75middot7-147middot0) per 100 000 person-years in 1996-2005 to 93middot7 (75middot3-115middot0) per 100 000 person-years in 2013-20 (p=0middot49). Crude incidence rates in men who do not have sex with men (non-MSM) and women were generally lower than in MSM, but increased slightly over time, from 51middot08 (95% CI 20middot54-105middot25) to 67middot82 (40middot83-105middot91; p=0middot52) per 100 000 person-years in non-MSM and from 8middot09 (0middot20-45middot06) to 24middot95 (10middot03-51middot40; p=0middot29) per 100 000 person-years in women. The age-adjusted incidence rate in MSM in 2013-20 was significantly lower (rate ratio 0middot62 [95% CI 0middot41-0middot92]) compared with in 1996-2005. Changes in risk factors (less smoking, cumulative exposure to CD4 count of <200 cells per mu L, and plasma HIV-1 RNA of >1000 copies per mL) mostly explained the decrease in anal cancer risk over time in MSM. 3866 (23middot0%) of 16 819 MSM participated in anal cancer screening at least once. TNM tumour staging was more favourable (Cochrane-Armitage test for trend p=0middot033) in individuals diagnosed during screening. Crude anal cancer-associated 5-year mortality in people living with HIV decreased from 30middot4% (1996-2005) to 18middot3% (2013-20; odds ratio 0middot48; p=0middot070). Anal cancer-related mortality was 3middot7% (95% CI 0middot5-23middot5) in all men who had been screened and 24middot0% (95% CI 18middot1-31middot3) in men who had not been screened (p=0middot023). In men, screening participation (hazard ratio [HR] 0middot31, p=0middot051) and cumulative exposure to CD4 counts of less than 200 cells per mu L (HR 1middot11 per year; p=0middot0022) were independently associated with anal cancer-related mortality.Interpretation As anal cancer incidence is slowly declining in MSM but not in non-MSM and women, health-care professionals should not focus only on MSM for anal cancer prevention. Men diagnosed with anal cancer during screening had improved survival, probably because they were diagnosed at an earlier disease stage. Next to preventing anal cancer, these data are an important justification to screen those most at risk of anal cancer.