High-grade squamous intraepithelial lesion (CIN 2 and 3) excised with negative margins by loop electrosurgical excision procedure: the significance of CIN 1 at the margins of excision

The significance of cervical intraepithelial neoplasia grades 2 or 3 (CIN 2-3) at the margin of a loop electrosurgical excision procedure (LEEP) specimen as a predictor of recurrent or persistent disease is relatively controversial. Most studies have found the presence of high-grade dysplasia at the margins in this setting to be an adverse parameter. However, the significance of an isolated finding of CIN 1 at the surgical margins of LEEP specimens has not been rigorously studied. Herein, we evaluate the significance of finding CIN 1 at one or more margins of a LEEP specimen when these margins are negative for CIN 2-3. Consecutive LEEP specimens with a CIN 2-3 diagnosis (and whose margins were negative for the latter) were retrieved, and their slides were reviewed. The cases were classified as to whether CIN 1 was present at the margins. Follow-up cytologic and histologic data were compared between patients with and without CIN 1 at the margin. For patients with multiple follow-up samples, only the most severe abnormality (1 sample per patient) was counted. Seventy-three cases were evaluated, with CIN 1 present at the margin in 59 (81%, group 1) and absent in 14 (19%, group 2). Follow-up cytologic information was available in 60 patients, with a median follow-up of 19 months. The cytologic follow-up of groups 1 and 2 patients were not significantly different regarding the diagnostic frequency of any one of the Bethesda 2001 diagnostic categories (P >.05 for all, Fisher exact and chi(2) tests). Of the 33 patients who received a cervical biopsy and/or endocervical curettage during the follow-up period, 29 belonged to group 1 and 4 to group 2. The distribution of '' negative '' CIN 1 and CIN 2-3 diagnoses for group 1 were 16 (55%) of 29, 7 (24%) of 29, and 6 (21 %) of 29, respectively. Parallel figures for group 2 were 2 (50%) of 4, 2 (50%) of 4, and 0 (0%) of 4, respectively. Although CIN 2-3 was more frequent in group 1, none of the comparisons attained statistical significance. The current study is a small single-institution study, whose findings should not be extrapolated to cold-knife conizations, and need to be confirmed in larger data sets. However, our preliminary analysis suggests