Predicting treatment outcome in cervical diseases using liquid-based cytology, dynamic HPV genotyping and DNA cytometry.

Background: In this study, our prospective experience with a multimodal follow-up protocol is summarized, with special emphasis on predicting the treatment outcome of cervical diseases. Materials and Methods: Liquid-based cytology samples (ThinPrep) from 209 women exhibiting the whole spectrum of human papilloma virus (HPV)-related cervical diseases were investigated by cytology, PCR-based HPV genotyping and DNA cytometry pre-surgery. The first control cytology and type-specific HPV tests were performed at 3 months post-surgery. Results: The success rate of surgery was 95% in eradicating high-grade cervical disease and 90% in eliminating the baseline HPV genotype. Treatment failure was significantly correlated with baseline cytology (p = 0.011), resection margin status (p = 0.016) and HPV positivity at 3 months post-surgery (p = 0.04). Multivariate logistic regression analysis showed that type-specific persistent HPV infection (p = 0.028), baseline cytology (p = 0.039) and histology (p = 0.065) were independent predictors of residual cervical neoplasias. Conclusion: Our results showed that our multimodal surveillance protocol may help to individually assess the anticipated clinical outcome of cervical diseases post-surgery.