Randomized clinical evaluation of self-screening for anal cancer precursors in men who have sex with men

Cytojournal. 2006 Mar 20;3:4. doi: 10.1186/1742-6413-3-4.

Abstract

Background: Self-collection of anorectal swab specimens could greatly facilitate the completion of prerequisite studies and future implementation of anal cancer screening among men who have sex with men (MSM). We therefore compared self- versus clinician- collection procedures with respect to specimen adequacy for cytological evaluation, concordance of paired cytological results, and concordance of cytological with biopsy results.

Methods: Paired self- and clinician- collected anorectal Dacron swabs for liquid-based (Thin Prep) cytological evaluation were collected in random sequence from a mostly HIV-1 seronegative cohort of young MSM in Vancouver. Slides were reviewed by one cytopathologist. Presence of any cytological abnormality (atypical squamous cells of uncertain significance, ASCUS, or above) prompted referral for high-resolution anoscopy and possible biopsy.

Results: Among 222 patient-clinician specimen pairs, most were adequate for cytological evaluation, though self-collected specimens were less likely to be so (83% versus 92%, McNemar's test p < 0.001). Cytological abnormalities, noted in 47 (21%) of self-collected and 47 (21%) of clinician-collected specimens (with fair agreement, kappa = 0.414) included, respectively: ASCUS (5%, 5%), and low-grade (13%, 13%) and high-grade (3%, 3%) squamous intraepithelial lesions. Among 12 men with biopsy-confirmed high-grade neoplasia, most had abnormal cytological results (including 6 patient and 9 clinician swabs) but few (2 patient and 1 clinician swab) were high-grade.

Conclusion: Self-collection of anorectal swab specimens for cytologic screening in research and possibly clinical settings appears feasible, particularly if specimen adequacy can be further improved. The severity of biopsy-confirmed anorectal disease is seriously underestimated by cytological screening, regardless of collector.