User perception of endocervical sampling: A randomized comparison of endocervical evaluation with the curette vs cytobrush

Objectives To evaluate whether the endocervical brush (ECB) is better accepted by patients and health care providers for endocervical evaluation when compared to the endocervical curette (ECC), without a decrease in the quality of sampling. Methods Two hundred patients with cervical dysplasia were randomized at the colposcopy clinic of the University Hospital of Geneva into two groups according to technique. Patients and physicians’ preference regarding the technique as well as the quality of samples were assessed. ECB samples were analyzed using both cytological (cell block) and histologic analysis, while ECC samples were analyzed using standard histologic analysis. Results Of the 200 patients, 89 were randomized to ECC, 101 to ECB and 10 were excluded due to incomplete information or cervical stenosis. Physicians preferred ECB against ECC, classifying it more frequently as an easy technique (94.1% vs.61.4%, p<0.001). Physicians more frequently evaluated the ECB as little or not uncomfortable for patients (28.7% vs.10.2%, p<0.001), though patients themselves didn’t express a preference for either technique. From a quality standpoint, the brush allowed for a better quality of samples, with a lower rate of inadequate samples (2.0% vs 14.3%, p = 0.002) and greater amount of material. Conclusion Endocervical sampling using ECB seems to be easier to perform and provides better quality samples. ECB can therefore be an acceptable alternative to ECC in standard practice. Trial registration ClinicalTrials.gov NCT01435590


INTRODUCTION
In the last decade significant advances in the diagnosis of cervical cancer and preinvasive disease have been made. One of the fundamental steps in diagnosis is colposcopy with the evaluation of the endocervical canal. This is usually done by endocervical curettage (ECC), but due to its high rate of false positive and false negative results and important patient discomfort, efforts have been made to find alternative techniques.

OBJECTIVE
The objective of this study is to compare the adequacy of specimen of the endocervical brush versus the endocervical curette in the evaluation of the endocervical canal during routine colposcopy done in patients with suspicion of dysplasia. The second end point of this study is to evaluate patient and doctor discomfort with these two methods.

ACTUAL KNOWLEDGE IN THE DOMAIN
In the last decade, no more than 10 original articles have been written comparing endocervical curette with endocervical brush. The design of these studies has been very variable in terms of recollection techniques and analysis of specimen.

Recollection technique
Boardman [1], and Tate [2] used a sequential method for recollection of material, either by using first the curette and then recollecting the material with a brush or the opposite, while Mogensen [3] and Klam [4] used either one or the other. They all found that endocervical sampling with a cytobrush had similar sensitivity and specificity to traditional endocervical curettage. The sensitivity and specificity with the brush was even higher when using a sleeved cytobrush.

Analysis of specimen
Boardman [1] did a different analysis for the brush and the curette (cytology for the first and histology for the second) while Klam [4], Gibson [5], Maksem [6] and Mogensen [3] all did histological analysis of both samples. This is very important, because it has been demonstrated that when analyzed as cytology, cytobrush technique tends to have high false positive rate. Of all these groups, only Maksem [6] recollected their brush specimen into a liquid fixative process; all others fixed the material onto a slide.
Klam [4] et al evaluated the discomfort associated with each technique, and came to the conclusion that there was no difference.

STUDY DESIGN
This is a prospective randomized study, whose target population is all women consulting our colposcopy clinic. The eligibility of each patient will be analyzed based of the following characteristics: Information will be given to the patients before commencing their colposcopy, and informed consent will be obtained. Once the colposcopist has decided that a further evaluation of the endocervical canal is needed, women will then be randomized to either the ECC or the ECB groups. ECC will be performed maximum 3 times with a Novak metal curette using short, firm strokes from the lower uterine segment down to the external os, circumferentially. Samples will then be collected in CARNOY.
Endocervical brushing will be performed by taking 12 swipes of the entire length of the endocervical canal while rotating simultaneously the brush. The specimen will then be collected and fixed with Thin Prep.
Cytological and histological interpretation of the ECB specimens will then be performed, while for the EEC specimens only histological interpretation will be done. The adequacy of the specimens will be based on the quantity of endocervical cells present (< or ≥ 20 endocervical cells), the quantity of histological material (< or ≥ 3 epithelial stripes) and the quality of histological material (absence or presence of lamina propia).
To evaluate the degree of patient discomfort patients will be requested to complete a questionnaire once the exam has been completed. The main acceptance variables will be degree of helplessness, pain (on visual analog scales), willingness to undergo the test again, and overall satisfaction. Once the exam has been completed, the doctor performing the sampling will answer a questionnaire indicating his/her perception of the exam (patient's pain, technical difficulty performing the exam, doctors acceptability of the exam).

Statistical analysis
The statistical analysis will be based on the comparison of the performance of both tests. The main outcome variable will be the proportion of a satisfactory sample in each group. A satisfactory specimen will be defined by the pathologist as indicated in the paragraph "study design". To demonstrate a 20% difference (with 80% power and a risk of type 1 error set at 5%) in the quality of the material obtained with the brush, we will need a total of 180 patients.

ETHICAL ISSUES
The follow up of patients will not differ from guidelines already in use. The risk of inadequate sampling requiring a second exam has been evaluated as equal for both techniques when analyzing the citology with the slide technique analysis.

FINANCIAL CONSIDERATION
The cost of the different techniques used in this study is 93 francs for the analysis of the material obtained with the brush and 101 to 130 francs for the analysis of the material obtained with the curette, depending on the diagnsosis. Both of these exams will be performed only if indicated during her colposcopic exam, and only one of these techniques will be used. The cost of these exams is covered by normal health insurance, and the actual cost for the patients will depend on their "franchise". The study will include no extra charge for the HUG.

EXPECTED CONTRIBUTION
With this study we hope to demonstrate that endocervical sampling with cytobrush is superior to sampling with endocervical curette in term of diagnosis and patient comfort and doctors acceptability.