Expert Commentary

Is endocervical curettage really useful in assessing mildly abnormal cytology?

Author and Disclosure Information



Yes—in women over 40. This analysis of data from the multicenter ASCUS-LSIL Triage Study (ALTS) found endocervical curettage (ECC) to be of questionable value as an ancillary diagnostic technique to colposcopically directed biopsy in women under age 40. In women age 40 or older, however, the sensitivity of colposcopic biopsy decreased and the sensitivity of ECC increased, so ECC may be of value in assessing mildly abnormal cytology in this population.


Routine ECC as part of the colposcopy exam is controversial. Many colposcopists perform ECC with every colposcopy in nonpregnant patients. The rationale is that it will increase the sensitivity of the overall exam. A recent study by Pretorius and colleagues lends support to this view. The authors diagnosed high-grade dysplasia—CIN 2,3—based on ECC specimens alone in 20 of 364 women (5.5%) who had satisfactory colposcopy exams.1

The opposing view holds that squamous dysplasia arises at the squamocolumnar junction and does not “skip” over apparently normal endocervical tissue to restart de novo within the canal. The proponents of this argument maintain that ECC adds little other than cost and discomfort when colposcopy is “satisfactory” or “adequate.”

Solomon and associates have contributed to this debate with their analysis of data from the ALTS trial, in which ECC was performed on 1,119 women, 41 (3.7%) of whom were diagnosed with CIN 2 or worse. In 10 of those women (0.89%), the diagnosis was based solely on the ECC.

More sensitive in older women

As in other analyses of colposcopy from the ALTS trial,2 the sensitivity of colposcopy with biopsy was disappointingly low: 72.5%. Solomon and colleagues estimate that ECC adds an additional 3% overall to sensitivity. When stratified by age, however, this marginal increase was higher in women age 40 or older than it was in younger women—13% and 2.2%, respectively. In the Pretorius study already mentioned,1 in which 5.5% of women were diagnosed with CIN 2 or worse solely on the basis of ECC, the mean age of the population was 42.

Weaknesses of the study

The study by Solomon and associates had several limitations, most of which were adequately discussed by the authors. ECC was performed at the discretion of the colposcopist; indications were not standardized. Moreover, the number of women age 40 and older was relatively small, and the study was not powered to address the issue of ECC and age.

Improvement in sensitivity is small but welcome in an older population

Older women are at higher risk of CIN 3 and unsatisfactory colposcopy and tend to have less accurate colposcopic impressions. Anything that can add to the sensitivity of colposcopy in this population is welcome. The authors point to the low utility of ECC in women under 40 and wisely stop short of recommending it in older women—although they come closer to such a recommendation in the abstract than in the body of the paper. This analysis of ALTS data certainly adds to the discussion of ECC, but more studies are clearly needed.

In the meantime, endocervical sampling with curettage or brush has an established role in colposcopy. It should be performed in women whose endocervical canal cannot be assessed, i.e., those with unsatisfactory colposcopy in whom an excision procedure is not otherwise planned. Endocervical sampling is also recommended as part of conservative management of women with high-grade squamous intraepithelial lesions on cytology but no cervical intraepithelial neoplasia or worse on biopsy. Other indications include women with low-grade squamous intraepithelial lesions or atypical squamous cells of undetermined significance in whom no lesion is identified, and those with atypical glandular cells on cytology.3

Next Article: