Evaluation of colposcopy after the addition of human papillomavirus testing to the Turkish cervical cancer screening program

Abstract

Abstract Objective To evaluate colposcopy performance following the human papillomavirus (HPV) DNA screening program in Turkey. Methods Women aged 30–65 years are screened for cervical cancer every 5 years, with individuals positive for HPV 16 and/or 18 or other high‐risk HPV types with abnormal cytology referred for colposcopy. Both HPV test and cytology are obtained at the same visit. If HPV is negative, cytology will not be assessed. However, if HPV is positive, both cytology and HPV genotyping will be performed. Colposcopy‐require was defined as HPV 16/18 positivity or abnormal smear results with any hrHPV positivity, and the remaining patients (normal smear with hrHPV positivity other than HPV 16/18) were grouped as colposcopy non‐required. National data on colposcopy outcomes and unnecessary performance rates in February 2018–2019 were evaluated via a questionnaire. Results A total of 9808 patients were included, divided based on colposcopy requirement: 5751 (58.6%) patients required colposcopy and 4057 (41.4%) did not. Unnecessary colposcopy was performed on 90.1% of the non‐required group (3657 of 4057 patients). In the colposcopy‐required group, 4455 patients (79.9%) underwent punch biopsy; 3194 (57.1%), endocervical curettage (ECC); and 421 (7.5%), “see and treat” in the non‐required group, the results were 2790 (76.3%), 1957 (53.2%), and 211 (5.7%), respectively. A total of 746 cervical intraepithelial neoplasia (CIN)‐3 isolates were detected, including 702 using existing screening and triage with 94.1% sensitivity (702/746). Multiple biopsies were taken in 69.8% (n = 3110) of patients from the colposcopy‐required group and 63.7% (n = 1777) from the non‐required group. The ECC samples included 19 cervical cancers and 212 ≥CIN‐3 lesions in the colposcopy‐required group, and four cancers and 41 ≥CIN‐3 lesions in the non‐required group. The proportion of ≥CIN‐3 lesions detected by ECC only was 4.7% (35 of 746 ≥CIN‐3 lesions). Conclusion Our results showed high rates of unnecessary colposcopies, and a high percentage of multiple and random punch biopsies and ECC.

Keywords

• cervical cancer
• colposcopy
• human papillomavirus