The clinical significance of a negative loop electrosurgical cone biopsy for high-grade dysplasia

Abstract

Objective: We sought to estimate the incidence and clinical significance of a negative therapeutic loop electrosurgical excision procedure (LEEP) and to evaluate patient specimens for limiting histologic features associated with a negative LEEP.

Methods: We identified 674 patients with biopsy-confirmed high-grade cervical dysplasia who were treated with LEEP from 1991 through 2001. The results of these LEEP procedures were reviewed for the absence of dysplasia or the presence of cervical intraepithelial neoplasia stages 1-3. Computerized pathology files of patients were then reviewed through July 2002 to determine whether dysplasia recurred. Slides of negative LEEP specimens were reviewed to confirm the absence of dysplasia and to search for histologic features that may have limited our interpretation of the specimen.

Results: Ninety-three (14%) of LEEP specimens reviewed were completely negative for dysplasia. Clinical follow-up was available on 75 of the 93 patients, with a median follow-up time of 2 years. Eighteen (24%) patients had subsequent positive follow-up, including carcinoma (n = 2), high-grade squamous intraepithelial lesions (n = 8), low-grade squamous intraepithelial lesion (n = 6), and atypical squamous cells of undetermined significance (n = 2). Patients with negative LEEPs had a recurrence rate similar to patients with positive LEEPs (24% versus 27%). Limiting histologic features were more commonly identified in negative LEEPs as compared with LEEPs containing dysplasia (16% versus 5%, P <.001).

Conclusion: A negative LEEP is not an uncommon finding, occurring in 14% (95% confidence interval 11-17%) of specimens at our institution. Negative LEEPs are more likely to contain histologic features that limit pathology interpretation. A negative LEEP is not a reassuring finding and was associated with a recurrence rate similar to those of a positive LEEP. Both negative and positive populations should be carefully followed.