Predicting Persistent/Recurrent Disease in the Cervix After Excisional Biopsy

Sanjay M. Ramchandani, MD; Karen L. Houck, MD; Enrique Hernandez, MD; John P. Gaughan, PhD

In This Article

Material and Methods

This retrospective study was registered with the Institutional Review Board. We reviewed 152 cases of cervical conization (cold knife conization or LEEP) performed at Temple University Hospital between January 1, 1992 and July 31, 2001 for which an ECC was obtained and follow-up information was available. During the interval studied, we retrieved 281 cases that had an ECC at the time of cervical conization, but follow-up information was not available in 129 (46%). Indications for cold knife conization or LEEP included evaluation and management of abnormal cervical cytology in patients with unsatisfactory colposcopic evaluation (defined as inability to visualize entire lesion or entire transformation zone on colposcopy, poor correlation between cytology and colposcopically directed biopsies, abnormality involving more than 2 quadrants of the surface of the cervix), dysplasia or cancer present in the ECC specimens at the time of colposcopy, microinvasive carcinoma on colposcopically directed biopsy, or for the treatment of cervical intraepithelial neoplasia (CIN) 2 or 3. Approximately one half of the 152 patients underwent LEEP and the other half underwent cold knife conization. With rare exceptions, gynecology residents under direct supervision of the same gynecologic oncologist performed the procedures. Colposcopy was not routinely performed at the time of the cold knife conization or LEEP. The surgical bed was routinely electrocauterized after LEEP or cold knife conization and Monsel solution was also applied. There were no changes in clinical practice during the years included in the study. All conization specimens were processed in a standard fashion. At least 12 sections from each cone or LEEP specimen were submitted for microscopic study. All slides were reviewed by a staff pathologist of the Department of Pathology, Temple University Hospital. Data were obtained by manual review of patient charts. Pathology reports were obtained from the Temple University Hospital medical information computer system. Data abstracted included age, race, gravidity, parity, type of procedure (cold knife conization or LEEP), ECC specimen findings, and histopathologic diagnoses to include status of the ectocervical and endocervical margins. The highest degree of abnormality on the conization specimen was recorded. Severity of disease on the conization specimens was classified as CIN 1, CIN 2/3, microinvasive carcinoma, and invasive cancer. The ectocervical and endocervical margins were classified as either positive (margins involved with dysplasia or cancer) or negative (margins uninvolved with dysplasia or cancer). ECC was classified as either positive (if it contained dysplastic squamous epithelium or cancer), negative or tissue insufficient for histopathologic diagnosis.

Follow-up information up to 12 months after the procedure (to include hysterectomy, repeat conization, colposcopically directed biopsies, ECC, and/or cytology) was abstracted. Patients who did not undergo a hysterectomy or repeat excisional biopsy were asked to return for colposcopy and cervical cytology every 4 months until 3 normal consecutive examinations or an abnormality was documented. Patients were considered to have persistent/recurrent disease if after the cervical conization there was any evidence of CIN or carcinoma on a subsequent hysterectomy specimen, repeat conization, cervical biopsy after conization, endocervical curettage after conization, or cervical cytology after conization. For patients with persistent/recurrent cervical neoplasia, the most severe degree of neoplasia was recorded.

Univariate logistic regression was used to estimate the relationship of findings from ECC specimens, endocervical margin status, ectocervical margin status, degree of dysplasia on conization, and age with the occurrence of persistent/recurrent disease. The odds ratio (an estimate of the relative risk of persistent/recurrent disease) and the 95% confidence interval were calculated for each variable. The variables (ECC specimen findings, endocervical margin status, ectocervical margin status and degree of dysplasia on conization) were then combined in a multivariate stepwise logistic regression to estimate the incremental risk of persistent/recurrent disease for each factor independent of (adjusted for) the other factors. The relative risk of persistent/recurrent disease was also estimated using these variables. A P value less than or equal to .05 was the significance level used for all analyses.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.