Endocervical crypt involvement by high-grade cervical intraepithelial neoplasia and its association with high-grade histopathological recurrence after cervical excision in women with negative excision margins: a systematic review and meta-analysis.

ABSTRACT

BACKGROUND: There is a growing body of evidence suggesting that endocervical crypt involvement by high-grade cervical intraepithelial neoplasia (CIN) may represent a risk factor for disease recurrence after cervical treatment. OBJECTIVES: To provide a systematic review and meta-analysis on whether endocervical crypt involvement by high-grade CIN on the excised cervical specimen is associated with high-grade histopathological recurrence during the follow-up of women after cervical excisional treatment. SEARCH STRATEGY We searched the Medline, Scopus, Central, and Clinical Trials.gov databases from inception till May 2023. SELECTION CRITERIA Studies that reported on women with a single cervical treatment with any method of excision for CIN2 or CIN3 lesion, negative excision margins, and whose recurrence was defined histopathologically were included. DATA COLLECTION AND ANALYSIS: Two reviewers independently evaluated study eligibility. We used the fixed effects model for meta-analysis. MAIN RESULTS: There were 4 eligible studies included in the present systematic review that evaluated 1088 women treated with either large loop excision of the transformation zone (LLETZ) or with cold knife conization (CKC). We found no significant association of endocervical crypt involvement by CIN2-3 with high-grade histopathological recurrence at follow-up after cervical excision (OR 1.93; 95% CI 0.51-3.35). The subgroup analysis of women with LLETZ cervical excision showed again no significant association with high-grade histopathological recurrence at follow-up (OR 2.00; 95% CI 0.26-3.74). CONCLUSION: Endocervical crypt involvement by high-grade CIN does not seem to be a risk factor for high-grade histopathological recurrence after cervical excision with negative excision margins.