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


Like other retrospective studies, ours has its limitations. This study is limited by the lack of follow-up in almost one half of the patients who underwent an ECC at the time of excisional biopsy of the cervical transformation zone during the interval studied. Because a repeat excisional biopsy or hysterectomy was not done on every patient, it is possible that the number of patients with persistent/recurrent disease is higher than what we identified. Some will argue that the rate of persistent disease after a LEEP should be less than that after a conization due to the thermal damage of the surgical bed. However, all of our patients had cauterization of the surgical bed regardless of whether they underwent a LEEP or a cold knife conization. Despite these findings, our findings are consistent with those reported by others.

We found that 100 (66%) of our cases had negative endocervical and/or ectocervical margins. In a retrospective study of 95 patients who had either a diagnostic LEEP (n = 45) or cold knife conization (n = 50), both margins were negative in 46% of those who underwent LEEP and in 48% of those who underwent cold knife conization.[17] Murdoch and colleagues[18] reported negative margins in 56% of 721 patients who underwent LEEP. Dietrich and colleagues[19] reported negative margins in 77% of their LEEP specimens. The specimen margins are reported as clear in 59% to 92% of women undergoing a cold knife conization.[1,7,8,20,21,22,23,24,25]

The endocervical margin was positive in 20% of our patients. Johnson and colleagues[26] reviewed 702 consecutive medical records of patients who underwent either a LEEP or cold knife conization. The endocervical margin was positive in 18% of the patients for which follow-up information was available. In a prospective randomized study comparing LEEP to cold knife conization, the endocervical margin was involved in 18% of those who underwent a LEEP and in 27% of those undergoing a cold knife conization.[27]

Our study showed that 27% of the patients had persistent/recurrent disease, which is similar to that reported by others. Giacalone and colleagues[28] randomized 66 women to undergo a LEEP (n = 28) or cold knife conization (n = 38). They followed their patients with colposcopy and cervical cytology. They reported histologically confirmed persistent dysplasia in 6 (21%) of the patients who had previously undergone a LEEP and in 4 (10.5%) patients in the cold knife conization group. Duggan and colleagues[27] randomized 180 patients to undergo a LEEP (n = 89) or cold knife conization (n = 91) and observed no difference in the rate of lesion clearance or disease recurrence between the 2 groups. Johnson and colleagues,[26] in a large retrospective study, found no difference in the rate of persistent or recurrent dysplasia following LEEP or cold knife conization. Dietrich and colleagues[19] reported 29% persistent cervical cytologic abnormality within 3 months after a LEEP. Abnormal cytology at a 3-month cervical smear was reported by Paterson-Brown and colleagues[8] among 27% of patients following a cold knife conization with a positive margin and in 9% of cases with a negative margin. The rate of persistent disease is much higher when defined as disease in a hysterectomy specimen obtained within 6 weeks of the conization.[4,12,29] In our study, only 3 of 33 patients who underwent a hysterectomy within 12 months of the excisional biopsy had no pathologic findings in the hysterectomy specimen. Fifteen (45%) had CIN 2/3 or cancer. This high rate of persistent/recurrent disease is expected because patients are selected to undergo hysterectomy after the conization based on factors that will predispose them to have persistent/recurrent disease (eg, high-grade dysplasia, cancer, positive margins). We opted to be inclusive and use all types of follow-up data in an effort to avoid that bias, but were aware that it could result in a falsely low rate of persistent/recurrent disease (eg, false-negative colposcopy and cytology findings).

As in other studies, by univariate analysis we found a statistically significant association between persistent/recurrent disease after excisional biopsy of the cervix and the histopathologic diagnosis in the excisional biopsy specimen, positive endocervical or ectocervical margin, and positive ECC.

We did not find an association between persistent/recurrent disease after cervical conization and age. However, Lu and colleagues[30] in a retrospective study of 120 women with CIN 3 who underwent a hysterectomy within 6 months of an excisional biopsy, found that residual disease was present in 56.5% of patients 50 years of age or older and in 29% of those younger than 50. Age greater than 50 years was also found to be associated with persistent/residual disease among 449 women who underwent conization for CIN 3 and who had at least 1 cytologic/histologic follow-up within 1 year of conization.[31]

The retrospective nature of our study did not allow us to look at the size of the cervical lesion or endocervical gland involvement as possible predictors of persistent/recurrent disease. Demopoulos and colleagues[32] found that both positive margins and endocervical gland involvement were highly significant predictors of persistent or recurrent disease in 245 women who had undergone a cold knife conization for CIN 3. Lu and colleagues[30] found that the rate of persistent disease in a hysterectomy done within 6 months of a cervical conization was 48% in women whose excisional biopsy specimen showed more than 2 quadrants of involvement and 26% with 1- or 2-quadrant disease. Others have also shown a higher risk of persistent disease within 6 months of treatment for CIN among women with large lesions.[33]

We did not exclude any of the excisional biopsy specimens (based on grade of CIN) in order to be able to evaluate the impact that the histopathologic diagnosis had on predicting persistent/recurrent disease. We found that as the severity of the disease in the cervical conization specimen increases so does the percentage of patients who have persistent/recurrent disease. Persistent/recurrent disease was found in 33% of patients who had CIN 2/3 in the excisional biopsy specimen and in 54.5% of those who had carcinoma. By univariate analysis, Chang and colleagues[29] found an association between the severity of cervical neoplasia and residual disease in a postcone hysterectomy and between the severity of the cervical neoplasia and the rate of positive cone margins. The rate of residual disease was 0, 23%, 23.8%, and 43.8% in patients with CIN 1, CIN 2/3, stage IA1 cancer, or stage IA2 cancer, respectively.

Persistent/recurrent disease was found in 50% of patients with positive endocervical and/or ectocervical margins, but only in 15% of those whose margins were negative. Many have shown a significantly higher rate of residual disease after a positive cone margin compared to a negative margin.[12,26,29] The rate of persistent or recurrent cervical neoplasia is highest when both the endocervical and ectocervical conization margins are positive, followed by a positive endocervical margin only, and lowest when only the ectocervical margin is positive.[22,24] In our multivariate stepwise regression analysis that included endocervical and ectocervical margins, severity of the neoplasia, and the ECC results, endocervical margin, and severity of the neoplasia best predicted persistent/recurrent disease.

Persistent/recurrent cervical neoplasia was found in 80% of patients with a positive ECC, but only in 18.5% with a negative ECC, and 20% with an ECC with insufficient tissue for diagnosis. There was not enough tissue for a histopathologic diagnosis in 19% of the ECC specimen, a finding reported by others.[14,33]

Persistent/recurrent disease was found in all 13 patients who had both a positive ECC and endocervical margin. This is consistent with the findings reported by Husseinzadeh and colleagues.[12] However, on the multivariate stepwise logistic regression analysis, endocervical margin status coupled to severity of neoplasia significantly predicted the occurrence of persistent/recurrent disease. Adding the results of the ECC did not add incremental value to their ability to predict persistent/recurrent cervical neoplasia.

Aware of the limitations of our retrospective study, the data suggest that the rate of persistent/recurrent cervical neoplasia is low in patients whose conization specimen shows CIN 1. As others have shown, the rate of persistent/recurrent disease is high when the conization specimen has CIN2/3 or cancer and the endocervical margin is involved. On multivariate stepwise regression analysis the ECC at the time of the conization offered no incremental value for predicting persistent/recurrent disease. ECC at the time of excisional biopsy of the cervical transformation zone does not need to be routinely performed.


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.