HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by KLAM, S. Articles by FERENCZY, A. Search for Related Content PubMed PubMed Citation Articles by KLAM, S. Articles by FERENCZY, A.

Comparison of Endocervical Curettage and Endocervical Brushing

From the Departments of Obstetrics and Gynecology, Pathology, Oncology, and Epidemiology, McGill University; the Departments of Obstetrics and Gynecology and Pathology, Sir Mortimer B. Davis–Jewish General Hospital; and the Departments of Obstetrics and Gynecology and Pathology, Royal Victoria Hospital, Montreal, Quebec, Canada.

Address reprint requests to: Alex Ferenczy, MD, Departments of Obstetrics and Gynecology and Pathology, McGill University and Sir Mortimer B. Davis–Jewish General Hospital, 3755 Côte Ste.-Catherine Road, Montreal, QC H3T 1E2, Canada

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Objective: To compare endocervical brushing with endocervical curettage with respect to diagnostic yield by histology and patient discomfort.

Methods: Nonpregnant women referred for colposcopy because of abnormal Papanicolaou test results were randomized to endocervical sampling with either a metal curette (endocervical curettage [ECC]) or an endocervical brush. Extensive endocervical canal brushing was performed. All samples were submitted for histologic study. Results were evaluated against the histologic findings in electroconization specimens in a masked fashion. Pain scores were recorded using Melzack’s Present Pain Intensity Scale.

Results: During the study period, 315 patients were randomized to the techniques: 157 to ECC and 158 to endocervical brushing. Of the 315 patients, 147 also underwent electroconization. Overall false-positive rates were 28.6% for endocervical brushing and 30.8% for ECC. False positives were due to contamination of the endocervical sample by lesional epithelium near the external os. The proportion of scanty specimens obtained by endocervical brushing (7.6%) was higher than that obtained by ECC (2.5%) (P = .041). One sample obtained by brushing was insufficient for diagnosis; none obtained by ECC were insufficient. There were no statistically significant differences in the median pain scores between the two groups.

Conclusion: The techniques were similar in terms of diagnostic yield and patient discomfort. Endocervical brushing had lower false-positive rates than those reported in the literature for cytologic analysis. Although ECC remains the method of choice for evaluation of the endocervical canal, brushing is an acceptable alternative.

In many colposcopic practices, evaluation of the endocervical canal is an integral part of the initial investigation in women referred because of abnormal Papanicolaou test results.^1–7 Many argue against the routine use of endocervical curettage (ECC) because of patient discomfort, relatively high cost, and false-positive and false-negative results.^4–6 Therefore, there has been a trend in recent years to use less painful methods such as sampling of the canal with an endocervical cytobrush. The literature is unclear with regard to whether endocervical brushing is advantageous. Previous studies yielded false-positive rates of 63–75%.^8–10 A potential result of these high rates is an increased number of unnecessary excisional procedures. If endocervical brushing were proved to be as accurate with respect to diagnostic yield as ECC but less painful, evaluating the endocervical canal using a brush technique might be considered an attractive alternative to ECC. The purpose of the present study was to compare conventional ECC with endocervical brushing using a new sampling technique, with respect to histologic evaluation of the endocervical canal and level of discomfort in patients referred for colposcopic evaluation because of abnormal Papanicolaou test results.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
This prospective, randomized clinical trial was carried out in a single referral center for colposcopy. The McGill University–Jewish General Hospital Ethics Research Committee approved the research. From July 1997 to June 1998, all nonpregnant women referred to the Colposcopy Clinic of the Sir Mortimer B. Davis–Jewish General Hospital, Montreal, Quebec, for evaluation because of abnormal Papanicolaou test results were invited to participate in the study. Each woman who consented underwent, in sequence, a repeat exoendocervical Papanicolaou test, colposcopic evaluation, endocervical canal sampling, and punch biopsies. Patients were randomized to endocervical sampling with either brushing or ECC. Specimens from both brushing and ECC were submitted for histologic evaluation. Patients were masked to the method used for endocervical sampling. The procedures were performed by three of the authors (JA, NM, AF). The histopathology laboratory was unaware of the source of the sample.

Endocervical brushing was performed using a new sampling technique. A standard cytobrush was used to take ten to 12 swipes of the entire length of the endocervical canal; the brush was rotated simultaneously to ensure that the entire canal had been sampled. The specimen then was pressed out from between the brush bristles onto lens paper, submitted, and processed in the same manner as the sample obtained by ECC. Endocervical curettage was performed twice with a Kevorkian metal curette using short, firm strokes from the lower uterine segment down to the external os, circumferentially. Efforts were made to collect the curetted sample within the endocervical canal and remove the curette’s jaw without tearing off lesional tissue that could be located close to or at the external os. Endocervical curettage samples were collected on a lens paper.

Both brush and ECC samples were fixed in formalin and sent to the Department of Pathology for histologic processing. Histologic interpretation of specimens was performed on a routine basis by one of the authors (AF) without knowledge of the method used. In our institution, a satisfactory endocervical specimen must contain at least five strips of epithelium, the shortest fragment containing at least ten endocervical cells in a field of view at 10x magnification.

As per our usual protocol, patients with biopsy-proven disease returned for the loop electrosurgical excision procedure (LEEP), and those with negative biopsy results underwent repeat Papanicolaou testing, colposcopy, and ECC. The endocervical brushing and ECC results were evaluated against the final histopathologic findings in the LEEP specimens. A false-positive result was recorded when an endocervical specimen contained several strips of squamous neoplastic epithelium and disease was absent in the endocervical canal. Those specimens contaminated by an exocervical lesion located close to or at the external cervical os contained one or two tiny fragments of squamous neoplastic epithelium. Thus, the determination of a false-positive result was based on the location of the squamocolumnar junction of the lesional tissue as determined by colposcopy and ultimately by histologic analysis of the LEEP cone specimens.

Physicians were asked to assess the perceived level of discomfort experienced by patients undergoing the procedures. In addition, patients were requested to rate their levels of discomfort during an interview with a nurse who was unaware of the groups to which they had been assigned. Both subjective and objective scores were recorded using the Present Pain Intensity Scale.¹¹

Data were entered into a computerized database for later extraction and analysis. Analyzed in this study were reason for referral for colposcopy (diagnosis made from Papanicolaou test results), patient age and parity, colposcopic interpretation, and findings from cytologic analysis, biopsies, and endocervical canal sampling. In the LEEP specimens, the location of the lesional tissue also was analyzed. In addition, both subjective and objective pain scores were entered into the database. Bethesda system terminology was used for describing cytologic specimens.

Block randomization was used for assigning interventions. A block size of four treatment assignments (two per arm) was used throughout to ensure that whenever the study was interrupted, the difference between arms would not be appreciable. The primary study hypothesis was that endocervical brushing might have better diagnostic value than ECC. The size of this study population was chosen to ensure that a relative improvement in the diagnostic accuracy of 23% could be detected with an error of .05 and a ß error of .20. This difference was based on calculations of diagnostic accuracy for both endocervical brushing and ECC using data from Andersen et al.⁸ Standard formulas were used to arrive at a total sample size of 138 patients, 69 subjects in each arm.¹² Assuming a 50% LEEP rate, twice as many patients (ie, at least 276 subjects) were required for this study. Intention-to-treat analysis was performed.

False-positive and false-negative rates were compared for brushing and ECC. The difference in the proportion of unsatisfactory histologic specimens obtained by brushing and ECC was examined using the ² test. Patient and observer median pain scores associated with each procedure were compared using the ² test.

	Results

Top Abstract Materials and Methods Results Discussion References

 
During the study period, 315 patients were randomized to endocervical sampling with either ECC or brushing; 157 patients were assigned to ECC and 158 to endocervical brushing, corresponding altogether to a response rate of 98%. The distribution of patient characteristics by study group is given in Table 1. There were no statistically significant differences between the two groups in categories of age, parity, diagnosis based on results of referral Papanicolaou testing, colposcopic findings, diagnosis based on results of repeat Papanicolaou testing, and cervical biopsy histologic findings. Endocervical histology results were also similar irrespective of the method used (Table 2).

Of the remaining 151 patients in the ECC group, 138 had no evidence of disease. In the brushing group, 139 patients had no evidence of disease on histologic examination of the endocervical specimen. Thus there were 13 patients in the ECC group and 14 in the brushing group with positive results reported as "presence of strips of neoplastic epithelium" in the endocervical specimen. Results for four (30.8%) of these 13 patients in the ECC group and four (28.6%) of these 14 patients in the brushing group were interpreted as false positive. The false-positive results were due to contamination of the endocervical specimen by lesional tissue located near or at the cervical os.

Overall, 147 patients underwent LEEP cone biopsy, and the ECC and endocervical brushing results were evaluated against the final histopathologic findings in the LEEP specimens. Of these patients, 76 (51.7%) underwent ECC and 71 (48.3%) underwent brushing. The indications for LEEP were the same for both groups and included extensive squamous intraepithelial lesions (SILs), high-grade SILs, positive brushing or ECC findings, and inadequate colposcopy. Loop electroconization histology results were similar for both groups (Table 2).

Twelve samples (7.6%) from the brush group and four samples (2.5%) from the ECC group contained scanty (fewer than five strips of endocervical epithelium) specimens (Table 2). Yet only one sample from the brush group was insufficient for diagnosis. None from the ECC group were insufficient.

The level of patient discomfort (rated using the Present Pain Intensity Scale) perceived by the observer was statistically invariant irrespective of the technique used (Table 2). Similarly, there was no statistically significant difference between the two groups in the level of discomfort as described by the patient. Because there were so few responses in the extreme categories of the Present Pain Intensity Scale, the extreme descriptor levels were combined for the purpose of analysis. The median pain score for the observers was 1 in both groups, and for the patients this score was 2 in both groups. There were no differences in frequencies of pain scores between the two study groups (P = .289 for observers and P = .910 for patients).

The sensitivities and specificities of and negative and positive predictive values for ECC and endocervical brushing are listed in Table 3. It appears that ECC is a somewhat less sensitive technique than brushing for evaluating the endocervical canal, although the difference was of no statistical significance. False-negative rates were 3.6% and 2.1% for ECC and brushing, respectively; both patients with invasive cancer had negative endocervical specimens and had International Federation of Gynecology and Obstetrics stage IB1 disease without endocervical canal involvement.

	Discussion

Top Abstract Materials and Methods Results Discussion References

Endocervical cytologic analysis has been compared with standard ECC in three studies with designs similar to ours.^8–10 The results from both ECC and brush cytologic analysis were compared with the final pathologic findings in the cone biopsy and/or hysterectomy specimens. In all three studies, endocervical cytologic analysis was comparatively more sensitive than ECC in terms of detection of disease in the endocervical canal but yielded false-positive rates of 63–75%. More recently, Mogensen et al¹³ conducted a randomized clinical trial comparing endocervical cytologic analysis and ECC, both preceded by colposcopically directed biopsies of the ectocervix. Sensitivity and specificity for colposcopically directed biopsies plus endocervical cytologic analysis were 96% and 95%, respectively. The same figures for the ECC were sensitivity 82% and specificity 88%.

It appears that endocervical cytologic analysis can be used in diagnosing canal lesions and has less preparation-related cost than histologic analysis. However, its high false positivity, due to contamination from lesions close to the external os and errors in cytologic interpretation that lead to costly conizations, remains its weakness.

In the present study, the false-positive rate of 28.6% for endocervical brushing was significantly lower than that reported in the literature for endocervical cytologic analysis (63–75%).^8–10 One possible explanation for the higher rate of false positives with endocervical cytologic analysis is the greater difficulty involved in interpreting cytology specimens. However, when we compared the false-positive rates for brushing for histologic analysis with traditional ECC, this difference was not statistically significant.

There are three possible explanations for the higher rate of scanty specimens obtained with brushing than with conventional ECC. First, there is a learning curve associated with this new technique; two-thirds of (8 of 12) scanty specimens were obtained from the first 40 patients. Insufficiencies of specimens were due to difficulties in retrieving the sample from the endocervical canal and in transferring the specimen to the lens paper. This rate decreased to 3.4% for the remaining 118 patients and was not statistically different from that associated with ECC (2.5%). Endocervical curettage has been a routine procedure in our referral colposcopy center for the past 25 years. Second, the criteria used for defining a truly satisfactory specimen were stringent. Most specimens that failed to meet our arbitrary definition (five strips of endocervical epithelium) nevertheless contained sufficient tissue to establish that penetration of the endocervical canal had taken place. Also, the scantiness of the specimens did not prevent morphologic interpretation by the pathologist. Third, endocervical brushing yields less endocervical tissue than does ECC because it is less abrasive than the latter technique and the brush is more difficult to empty than is the jaw of the endocervical curette. Despite these drawbacks, only one specimen in the brush group was truly insufficient for diagnosis. This was not statistically different from the ECC group, in which none of the samples were insufficient for diagnosis.

The degree of discomfort associated with endocervical brushing is generally believed to be less than that associated with ECC.^4–10 In fact, ECC-associated discomfort was the single most important reason for conducting a number of clinical trials to find an alternative to ECC for evaluation of the endocervical canal.^8–10 We found no differences in either subjective or objective pain scores between endocervical brushing and ECC. However, observers’ ratings of patient discomfort were universally lower than patients’ ratings. In the majority of cases the discomfort was well tolerated, and in none of the cases was either procedure discontinued because of excessive pain.

	Footnotes

 
PII S0029-7844(00)00836-X

Received October 6, 1999. Received in revised form January 7, 2000. Accepted January 20, 2000.

	References

Top Abstract Materials and Methods Results Discussion References

 
1. Townsend DE, Richart RM. Diagnostic errors in colposcopy. Gynecol Oncol 1981;12(Suppl):S259–64.[Medline]

2. Hatch KD, Shingleton HM, Orr JW, Gore H, Soong SJ. Role of endocervical curettage in colposcopy. Obstet Gynecol 1985;65: 403–8.[Abstract/Free Full Text]

3. Ferenczy A. Endocervical curettage has no place in the routine management of women with cervical intraepithelial neoplasia: Debate. Clin Obstet Gynecol 1995;38:644–8.[Medline]

4. Noller KL. Endocervical curettage: A technique in search of an indication? Clin Obstet Gynecol 1995;38:649–52.[Medline]

5. Urcuyo R, Rome RM, Nelson JH. Some observations on the value of endocervical curettage performed as an integral part of colposcopic examination of patients with abnormal cervical cytology. Am J Obstet Gynecol 1977;128:787–92.[Medline]

6. Spirtos NM, Schlaerth JB, d’Ablaing G 3rd, Morrow CP. A critical evaluation of the endocervical curettage. Obstet Gynecol 1987;70: 729–33.[Medline]

7. Helmerhorst TJM. Clinical significance of endocervical curettage as part of colposcopic evaluation: A review. Int J Gynecol Cancer 1992;2:256–62.[Medline]

8. Andersen W, Frierson H, Barber S, Tabarah S, Taylor P, Underwood P. Sensitivity and specificity of endocervical curettage and the endocervical brush for the evaluation of the endocervical canal. Am J Obstet Gynecol 1988;159:702–7.[Medline]

9. Weitzman GA, Korhonen MO, Reeves KO, Irwin JF, Carter TS, Kaufman RH. Endocervical brush cytology: An alternative to endocervical curettage? J Reprod Med 1988;33:677–83.[Medline]

10. Hoffman MS, Stratton S Jr, Gordy LW, Gunasekaran S, Cavanagh D. Evaluation of the cervical canal with the endocervical brush. Obstet Gynecol 1993;82:573–7.[Medline]

11. Melzack R. The short-form McGill Pain Questionnaire. Pain 1987; 30:191–7.[Medline]

12. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: Wiley, 1981.

13. Mogensen ST, Bak M, Dueholm M, Frost L, Knoblauch NO, Praest J, et al. Cytobrush and endocervical curettage in the diagnosis of dysplasia and malignancy of the uterine cervix. Acta Obstet Gynecol Scand 1997;76:69–73.[Medline]

This article has been cited by other articles:

				L. A. Boardman, H. Meinz, M. M. Steinhoff, W. W. Heber, and J. Blume A Randomized Trial of the Sleeved Cytobrush and the Endocervical Curette Obstet. Gynecol., March 1, 2003; 101(3): 426 - 430. [Abstract] [Full Text] [PDF]

				B. J. Monk, M. Cogan, J. C. Felix, N. M. Lonky, J. S. Bentz, C. J. Marshall, R. M. Cestero, L. R. Rowe, and S. A. Lonky A Stiff Bristled, Spiral-Shaped Ectocervical Brush: A Device for Transepithelial Tissue Biopsy Obstet. Gynecol., December 1, 2002; 100(6): 1276 - 1284. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by KLAM, S. Articles by FERENCZY, A. Search for Related Content PubMed PubMed Citation Articles by KLAM, S. Articles by FERENCZY, A.