Follow-up after treatment of cervical intraepithelial neoplasia by human papillomavirus genotyping.

OBJECTIVE: To assess the use of human papillomavirus genotyping in cervical intraepithelial neoplasia posttreatment follow-up.

STUDY DESIGN: Prospective observational study. Ninety women underwent cytologic testing and human papillomavirus genotyping at the follow-up visit after conization. Cones were retrospectively genotyped. A second cytologic follow-up was performed.

RESULTS: Margin status and presence of cervical intraepithelial neoplasia 3+ in the cone were poor predictors of treatment outcome (sensitivity, < 50%; diagnostic odds ratio, <or= 2.5). Presence of high-/intermediate-risk human papillomavirus types predicted 100% of residual high-grade squamous intraepithelial lesion/cervical intraepithelial 2+ at a specificity of 73%. Testing only 13 high-risk types showed equal sensitivity but higher specificity (86%; P < .01). Persistent high-risk human papillomavirus infection (13 types) detected high-grade residual disease with a sensitivity of 60% at a very high specificity (95%), resulting in a positive predictive value of 43%, which exceeded the positive predictive values of all other criteria.

CONCLUSION: Testing for high-risk human papillomavirus identified all recurrent/residual high-grade cervical intraepithelial neoplasia. Focusing on women with persistent human papillomavirus types through genotyping substantially increased positive predictive value but at a loss in sensitivity.