Document resource
Background Optimal screening strategies for anal cancer in men who have sex with men (MSM) remain debated, particularly regarding the integration of anal cytology and human papillomavirus (HPV) testing and their impact on referral burden to anal resolution anoscopy (HRA). Real-world comparative data on diagnostic performance of different screening algorithms are limited.Methods We retrospectively analyzed data from MSM undergoing first HRA. For each participant, the most recent anal cytology and HPV test performed within 12 months before (or on) HRA were collected. Histological high-grade squamous intraepithelial lesion (HSIL), defined as anal intraepithelial neoplasia grade 2 or higher (AIN2+), was the primary endpoint.Seven screening strategies were evaluated, grouped according to their biological and clinical rationale: human papillomavirus genotype 16 (HPV16); high-risk human papillomavirus (HR-HPV); low-grade squamous intraepithelial lesion or worse (LSIL+); atypical squamous cells of undetermined significance or worse (ASCUS+); ASCUS+ with HPV16; LSIL+ with HR-HPV; and ASCUS+ with HR-HPV (table 1). Diagnostic performance was assessed using complete-case analysis. Referral burden was estimated as the proportion of patients who would have been referred to HRA under each strategy and was additionally expressed as the number of HRA procedures required per HSIL detected. Differences in sensitivity between human immunodeficiency virus (HIV) strata were assessed using Fisher’s exact test.Results A total of 317 MSM were included, whose 70.3% were living with HIV; median age was 44 years (IQR:36-53). HR-HPV demonstrated the highest sensitivity for HSIL detection (0.87), with moderate specificity (0.52) and high negative predictive value. Cytology-only strategies were associated with higher referral proportions and lower specificity. Combined strategies (ASCUS+ and HR-HPV; LSIL+ and HR-HPV) reduced referral burden compared with HR-HPV alone while maintaining acceptable sensitivity. HPV16-based strategies were highly specific (approximately 0.85–0.90) but markedly less sensitive (approximately 0.30–0.35), missing a substantial proportion of HSIL cases. Referral rates varied considerably across algorithms, translating into significant differences in HRA workload and HRA-per-HSIL ratios.When stratified by HIV status, sensitivity estimates were comparable between MSM living with or without HIV across all strategies, and no statistically significant differences were observed.Conclusions In MSM undergoing first HRA, HR-HPV–based strategies maximized HSIL detection at the cost of increased referral burden, whereas HPV16-based approaches minimized HRA workload but compromised sensitivity. Combined cytology and HR-HPV algorithms provided a more balanced trade-off between detection and resource utilization. These findings support risk-adapted anal cancer screening strategies in MSM irrespective of HIV status.Abstract SC9 Table 1Diagnostic performance and referral burden of seven anal cancer screening strategies. Sensitivity and specificity were calculated using complete-case analysis with histological HSIL as the reference standard. N: the number of patients with available screening results included in performance analyses; HRA/HSIL ratio: the number of HRA procedures required to detect one HSIL case; MSM-WH: MSM living with HIV; MSM-WoH: MSM without HIV; Referred to HRA: the proportion of the overall study population who would have been referred to HRA under each strategy