PCR detection rates of high risk human papillomavirus DNA in paired self-collected urine and cervical scrapes after laser CO2 conization for high-grade cervical intraepithelial neoplasia

Objective. To compare the PCR detection rates of high risk human papillomavirus DNA in self-collected urine and cervical scrapes during follow-up of patients treated for HG-CIN by laser CO2 conization. Patients and methods. 52 women who submitted to laser conization for HG-CIN were enrolled into this prospective follow-up study receiving liquid-based cytology and HR-HPV testing by PCR assay on self-collected urine and cervical scrapes before and at 3, 6 and 12 months after treatment. Diagnostic accuracy and predictive values for treatment failure were evaluated for both urinary and cervical HPV testing and follow-up cytology. Results. 3 cases (5.8%) of recurrent HG-CIN occurred during follow-up. Positive margins and HR-HPV persistence resulted to significant risk factors for recurrence (p=0.01). The overall concordance on HR-HPV detection between paired urine and cervical samples was 96.6% and discord trend between agreement rates during follow-up were excluded by overall fixed-effect index (OR 1.03; 95% CI 0.62-1.70). No difference was observed comparing the three- and six-month cumulative sensitivity and NPV for recurrent disease of urinary and cervical HPV detections, with an increase of 5.6% in specificity associated with urinary testing. Conclusions. PCR detection of HR-HPV in paired urine and cervical samples during follow-up revealed an excellent concordance, suggesting a potential equivalent role of the two methods within post-treatment follow-up. In our experience HPV testing on self-collected urine was more sensitive than cytology and more specific than cervical HPV detection to predict treatment failure. Larger studies are needed to definitively establish the role of urine-based HPV testing during follow-up. (c) 2008 Elsevier Inc. All rights reserved.