Comparison of Patient- and Clinician-Collected Anal Cytology Samples to Screen for Human Papillomavirus–Associated Anal Intraepithelial Neoplasia in Men Who Have Sex with Men

BACKGROUND: Human papillomavirus (HPV)-associated anal cancer is increasing in prevalence and is more common among men who have sex with men and HIV-positive individuals than cervical cancer is among women in the United States. Cytology screening can detect the anal cancer precursor, anal intraepithelial neoplasia (AIN). Little is known about self-collected samples for AIN screening, and few community-based AIN estimates exist. OBJECTIVE: To compare the sensitivity of self-collected versus clinician-collected anal cytology specimens to detect biopsy-confirmed AIN and the prevalence estimate of AIN in a community sample. DESIGN: Cross-sectional study. Participants were mailed anal cytology self-collection kits with instructions. Clinicians repeated anal cytology and performed high-resolution anoscopy with biopsies as the diagnostic reference standard. SETTING: San Francisco, California. PATIENTS: Community-based sample of men who have sex with men. MEASUREMENTS: Prevalence of anal HPV and AIN. Sensitivity and specificity of self-collected and clinician-collected anal cytology specimens to diagnose AIN were calculated. RESULTS: Biopsy-proven AIN was diagnosed in 57% of HIV-positive and 35% of HIV-negative participants (P = 0.04), and 80% provided adequate self-collected specimens for interpretation. The sensitivity of cytology to detect AIN in HIV-positive men was 75% (95% CI, 51% to 93%) when self-collected and 90% (CI, 68% to 99%) when clinician-collected; respective values in HIV-negative men were 48% (CI, 26% to 70%) and 62% (CI, 38% to 82%). The specificity of cytology to detect AIN in HIV-positive men was 50% (CI, 22% to 78%) when self-collected and 64% (CI, 36% to 86%) when clinician-collected; respective values in HIV-negative men were 86% (CI, 71% to 94%) and 85% (CI, 72% to 93%). LIMITATIONS: The study sample was from a narrowly defined geographical area. Participants self-reported HIV status. CONCLUSION: In a community-based sample, a high proportion of HIV-positive and HIV-negative men who have sex with men have AIN. The sensitivity of cytology to detect AIN is higher for clinician-collected versus self-collected specimens and for HIV-positive versus HIV-negative men. The specificity of cytology to detect AIN is higher in HIV-negative versus HIV-positive men. However, the probability of AIN in a patient with a negative cytology result may not be low enough (23% for HIV-negative men and 45% for HIV-positive men with a patient-collected specimen) for clinicians to be comfortable recommending no anoscopy for those with a negative cytology result if done as a one-time test. These data raise the question of whether the optimal population screening strategy is cytology screening with anoscopy only for those who test positive or whether anoscopy should be recommended for everyone in these risk groups. Given limited resources and the limited number of clinicians trained in anoscopy, cytology screening may be the best current approach to identifying disease in the at-risk population.