Follow-up after LLETZ: a study of 682 cases of CIN 2-CIN 3 in a single institution.

ABSTRACT

OBJECTIVE: To evaluate the importance of resection margins in the risk of persistent/recurrent lesions and to investigate other factors such as detection of high-risk HPV, which could potentially predict persistent/recurrent disease before patients engage in follow-up. STUDY DESIGN: 682 women with a histologically confirmed diagnosis of CIN 2-3 treated by loop electrosurgical excision procedure (LEEP) were included, between January 2000 and December 2006. Age, high-risk HPV detection determined by Hybrid Capture II and cone margins were evaluated as possible predictors of persistent/recurrent disease. RESULTS: The mean age at diagnosis was 37.8 years (range 18-73). The mean follow-up period was 39.9 months (SD 25.8). 6.6% of patients (45/682) were lost to follow-up. 64.7% of patients (441/682) had clear margins in the specimen and 20.1% of patients had positive surgical margins (137/682). In 8.6% of patients (59/682) the resection margins were uncertain. Positive endocervical sweep was found in 10.8% of cases (73/682). Residual/recurrent disease was demonstrated by colposcopy-guided biopsy in 13.9% of patients (88/637); 77.3% (68/88) of them developed CIN 1 while only 22.7% (20/88) developed high-grade premalignant lesions or carcinomas during the follow-up. We found significant differences in the frequency of persistent/recurrent disease depending on the status of margins 24.8% of cases with positive margins vs 11.1% of cases with negative margins (p<0.0001). Multivariate analysis showed that only post-treatment high-risk HPV detection and status of the cone margins were significantly predictive of persistent/recurrent disease (OR 4.1, 95%CI 2.4-7.3, p<0.0001 and OR 2.7, 95%CI 1.5-4.7, p=0.001; respectively). CONCLUSION: The combination of histological examination of resection margins plus post-treatment tests for HPV detection would help to classify LEEP-treated patients into categories at different risk of recurrence.