Comparison of two surgical methods for the treatment of CIN: classical LLETZ (large-loop excision of the transformation zone) versus isolated resection of the colposcopic apparent lesion – study protocol for a randomized controlled trial

The Evaluation of Clinical Outcome after Reduction of Conization Size (or ECO-ROCS) trial is a multicenter, prospective, parallel-group, patient-blinded, non-inferiority randomized controlled trial in which we are assessing oncological comparability of two conization procedures: (1) classical LLETZ, comprising excision of the dysplastic lesion, including resection of the transformation zone; and (2) excision of the dysplastic lesion only.

This study is being conducted in 13 study centers, all of which run a specialized dysplasia clinic (Munich (two), Hannover, Wolfsburg, Berlin, Kiel, Regensburg, Stralsund, Hamburg, Düsseldorf, Aachen, Ulm, Münster). The participating study centers and investigators are fully listed in the German Clinical Trials Register (DRKS00006169).

A sample size of 892 patients was calculated with the POWER procedure of SAS for Linux 9.3 software (SAS Institute, Cary, NC, USA) using the following parameters: non-inferiority margin of 5 %, α = 0.05, 80 % power and one-sided test. A rate of 90 % negative HPV test results 6 months after conization was taken as a reference [20] for both groups. Assuming a 10 % dropout rate, we aim for a sample size of 1000 patients.

The statistical inference regarding the primary outcome will be based on a one-sided 95 % confidence interval for the risk difference, calculated as the proportion of HPV-positive patients in the control group minus the corresponding proportion in the experimental group. If the upper limit of the confidence interval is less than the non-inferiority margin of minus 5 %, non-inferiority will be regarded as established.

Patients are being seen and diagnosed in specialized dysplasia outpatient clinics of the participating study centers. After they provide us their written informed consent, participants fulfilling all required inclusion and exclusion criteria are given a five-digit pseudonym (patient identifier), with the first two numbers indicating the study center and the latter three numbers indicating a sequential patient number. Patients are randomly assigned by their pseudonym with equal probability to either of the two treatment arms, using a 1:1 ratio. Randomization is carried out using the web-based Randoulette randomization service of the Institute for Medical Informatics, Biometry and Epidemiology, University of Munich, Germany. The platform offers a web-based form in which to enter patient identifier, sex, age and study center, as well as the strata of the patient. Access to the randomization service is provided to the two leading investigators at each study center. Patients are stratified by age (>30 or ≤30 years) and study center.

After the information is entered, the study arm is selected according to a predefined randomization scheme, in this case an unblended, stratified block randomization. To maintain allocation concealment, the exact length of blocks is kept secret by the Randoulette randomization service. The entered information, as well as the randomization results, are stored in the Randoulette database.

Approval of trial protocol and informed consent documents have been obtained primarily from the University of Munich Institutional Review Board (project number 275-14), followed by the responsible ethics committees and/or state chambers of physicians of the 12 other study centers (University of Munich Institutional Review Board, ethics committee of the Medical Chamber of Westphalia-Lippe, University of Regensburg Institutional Review Board, ethics committee of the Medical Chamber of Hamburg, ethics committee of the Hannover Medical School, State Medical Chamber of Lower Saxony, institutional review board of Berlin, ethics committee of the Medical Chamber of Schleswig-Holstein, ethics committee of the Heinrich-Heine University of Düsseldorf, institutional review board of Greifswald, University of Aachen Institutional Review Board and institutional review board of the University of Ulm).

All study staff commit themselves to the Declaration of Helsinki (2013 version), as well as to all pertinent national laws and the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) harmonized tripartite guidelines for Good Clinical Practice issued in June 1996 (http://​www.​ich.​org/​fileadmin/​Public_​Web_​Site/​ICH_​Products/​Guidelines/​Efficacy/​E6/​E6_​R1_​Guideline.​pdf) and the European Medicines Agency’s CPMP/ICH/135/95 guideline (July 2002; http://​www.​edctp.​org/​fileadmin/​documents/​EMEA_​ICH-GCP_​Guidelines_​July_​2002.​pdf). Important protocol modifications will be reported to all relevant parties. All participants complete informed consent forms prior to participation in the trial.

To exclude bias by unconsciously different behavior, the study is patient-blinded for the first 6 postoperative months. In case of complications, however, access to the surgical report is possible at any time.

Both surgical techniques are performed using a thin, low-voltage, electrified wire loop known as a loop electrosurgical excision procedure electrode. Procedures are performed by experienced surgeons at 13 German dysplasia centers under colposcopic control and are carried out as shown in Figs. 1, 2, 3 and 4. A dysplastic cervical lesion is identified after application of acetic acid and iodine solution.

In the lesion-only resection, the dysplastic area is removed with a safety margin of 2 mm. In the classical LLETZ, the lesion, including the transformation zone, is resected. Because the outer border of the transformation zone cannot be determined colposcopically, the following definition for resection of the transformation zone was stated after expert consensus. After identification of the squamocolumnar junction, an area of 4 mm surrounding the junction is defined as the transformation zone. In the event that the iodine-negative zone is greater than the defined area, the iodine-negative area is considered the transformation zone and should be removed as well.

Postoperatively, cone size is measured before fixation by submergence in a fluid-filled cylindrical vial using Archimedes’ principle. Afterward the specimen is evaluated histologically for the grade of dysplasia and achievement of resection margins.

According to German guidelines, the first follow-up examination is performed after an interval of 6 months. A Pap smear, HPV test and colposcopy are performed. In cases of colposcopically suspect lesions, a biopsy is taken. A second postoperative control is carried out 12 months after conization.

The primary endpoint is the rate of negative high-risk HPV tests after 6 months, because a negative HPV test is generally regarded as evidence of successful treatment. The negative predictive value for a negative HPV test after conization is between 92 % [16] and 100 % [22‐26], and successful treatment usually leads to the elimination of the virus [27‐30]. A negative HPV test therefore excludes CIN persistence or recurrence with high probability. In contrast, repeated positive HPV tests can be taken as suspicious of a continuation of dysplasia or as an indicator of recurrent dysplasia.

Because it is possible, from a clinical point of view, that patients who have received the lesion-only operation simply need more time for the viral infection to be eliminated, HPV is tested again in the postoperative control after 12 months. Further secondary endpoints are Pap smear results in both postoperative controls, biopsy results (if applicable) and resection margins.

It should be noted that, even with classical conization, resection margins are positive or only questionably negative in up to 25 % of cases.

If there is CIN 1 detectable at the cutting margins, recurrence or persistence is expected in only 0 % to 5 % of participants. In the case of CIN 3 at the surgical margins, recurrence can be expected in 20 % to 25 %. If non in sano resection happens at the endocervical margin, risk increases to 30 %. If both, endo- and ectocervical resection margins are affected with CIN 3, risk is highest with 50 % [31]. Importantly, in those cases where immediate reconization is indicated because of positive surgical margins, no dysplasia was detected in over 80 % of cases.

For this reason, an immediate reconization usually is not indicated, but should be assessed according to individual risk and discussed with the patient. Only after further histological confirmation of CIN during the first postoperative control after 6 months will reconization be recommended.

Thus, the initially conservative approach in terms of a checkup after 6 months is also justified in the case of an increased incidence of positive surgical margins. This can be seen especially in light of the expected reduction of complications associated with surgery.

An interim analysis will be performed at the postoperative 6-month control of the first 250 patients included. The primary outcome is the result of HPV tests. If the rate of positive HPV tests after 6 months in patients who have undergone lesion-only surgery is higher by more than 15 % than among patients who have had classical LLETZ, the study will be terminated. Because the interim analysis was included exclusively for safety reasons and thus does not involve a test of the non-inferiority study hypothesis, the probability of a type I error (that is, a false-positive study result) should not be affected.

Because HPV infection is the cause of dysplasia, further analyses to investigate HPV infection are performed in a subset of 250 patients. A particular focus is on the genotyping of HPV infection because often multiple infections with different HPV types are present simultaneously, especially in young patients. Thus, there is the risk that, despite eradication of the lesion-causing HPV type, the conventional HPV test result may continue to be positive because infection with another HPV type is present without clinical consequence.

To make a statement with respect to a more valid clinical consequence of HPV infection, detection of transcriptional activity of E6 and E7 is analyzed. Expression of viral oncoproteins E6 and E7 plays a crucial role in HPV-mediated carcinogenesis, as it interferes with the cell cycle, causes cell cycle progression (E7) and at the same time prevents apoptosis (E6). This occurs through the inactivation of oncosuppressor proteins p53 and retinoblastoma protein (pRb). pRb in particular prevents cells from entering S-phase and thus has antiproliferative effects [32‐37]. As opposed to the mere detection of HPV DNA, testing for the transcriptional activity of E6 and E7 can allow us to make a statement about the biological activity of the infection.

The described additional analyses are performed at five study centers (Munich, Hannover, Wolfsburg, Berlin and Regensburg).