Does endometrial scratching increase the rate of spontaneous conception in couples with unexplained infertility and a good prognosis (Hunault > 30%)? Study protocol of the SCRaTCH-OFO trial: a randomized controlled trial

Study objective & design

The aim of this study is to evaluate whether endometrial scratching increases the chance of live birth in women with unexplained infertility and a good prognosis to conceive spontaneously (Hunault > 30%). Furthermore, the study seeks to determine endometrial characteristics correlating with (un) successful implantation and to develop endometrial organoids.

The SCRaTCH-OFO study is a multicentre randomized controlled trial (RCT) coordinated by the University Medical Centre Utrecht (UMC Utrecht, The Netherlands), which will be performed within the Dutch Consortium for Healthcare Evaluation and Research in Obstetrics and Gynecology. A prospective cohort study is embedded within the multicentre RCT, in which the endometrial biopsy will be banked for determination of characteristics associated with reproductive success or failure and for development of endometrial organoids that will open possibilities for further research into endometrium related disorders. Some of the study methods of this trial will be comparable to those previously described in the study protocol of our other scratching trial, in which scratching is performed in women undergoing IVF/intracytoplasmic sperm injection (ICSI) cycles [16].

Study setting

Study endpoints

The primary endpoint is cumulative live birth rate, of which the status of ‘ongoing pregnancy’ should be achieved within 12 months after randomization, and is defined as the delivery of at least one live fetus ≥24 weeks of gestation. Secondary endpoints include 1) ongoing pregnancy rate, defined as the detection of a fetal heartbeat on ultrasound at a gestational age of 10–12 weeks; 2) clinical pregnancy rate, defined as a gestational sac visualized on ultrasound; 3) miscarriage rate, defined as a demise of an intrauterine pregnancy confirmed by ultrasound or histology; 4) biochemical pregnancy loss, defined as a spontaneous pregnancy demise after serum or urinary bèta-hCG had been detected, without an ultrasound evaluation; 5) multiple pregnancy rate; 6) time to pregnancy, defined as the time from randomization till a positive pregnancy test; however, only pregnancies that will reach the ‘ongoing’ status will be included in the analysis; 7) progression to IUI or IVF treatment; 8) pregnancy complications; 9) complications of endometrial scratching (e.g. abnormal bleeding, abnormal vaginal discharge, abdominal pain, fever); 10) costs; 11) endometrial tissue parameters associated with reproductive success or failure, such as endometrial gene expression profiles (EGPs).

Sample size calculation

A total of 792 women will be included in the RCT of which we expect 150 to be included in the embedded study. Based on previous studies, the estimated difference in live birth rate is at least 10% between the patients with and without endometrial scratch (live birth rate respectively 45% vs. 35%).To detect such a difference with 80% power, 396 patients are needed per study arm, resulting in 792 patients in total. This number takes into account an estimated dropout rate of 5%.

Recruitment, consent and randomization

Eligible women will be given oral and written information about the study by their gynaecologist or fertility physician. Subsequently, these women will receive additional counselling by the investigator or research nurse after 1 week to allow the women to make an informed decision on trial participation. To participate in the embedded study, patients will give broad consent to the use of their endometrial sample and data (i.e. for storing tissue in the Biobank to determine EGPs and to develop endometrial organoids). After obtaining written informed consent, the patient will be randomly allocated to either the scratch procedure or no scratch procedure. Randomization is performed by a third party with a web-based randomization program using random blocks with block sizes of 2, 4 and 6. Due to the nature of the intervention, patients and physicians are not blinded and a sham intervention will not be performed. A survey among women participating in a RCT comparing endometrial scratching to a placebo procedure, showed that participants allocated to the scratching arm were not adequately blinded due to the discomfort they experienced during the procedure, whereas the sham intervention group was in fact less likely to guess their trial allocation (unpublished thesis by Lensen). However, the guess of trial allocation was not associated with the probability of committing protocol violations or the pattern and timing of intercourse (unpublished thesis by Lensen). Furthermore, a sham procedure could expose the control group to some degree of endometrial scratching, thereby possibly interfering with the study outcome.

Study procedure

In the intervention group, an endometrial scratch will be performed once before starting 6 months of timed intercourse, during the luteal phase of the natural cycle. Women are instructed to use urinary ovulation prediction tests that detect the putative luteinizing hormone (LH) surge (Ovulady, Clindia Benelux B.V., The Netherlands). The scratch will be performed 6 or 7 days after a positive urine ovulation test by a dedicated gynaecologist or fertility physician. If scheduling on this day is impossible then the scratch can also be planned 5 or 8 days after a positive ovulation test. Women will receive instructions that they need to make sure that they are not pregnant in the scratch cycle (no sexual intercourse or sexual intercourse with appropriate contraceptive methods (e.g.condoms)). After cleansing of the cervix with betadine, an endometrial biopsy catheter will be introduced into the cervix up to the uterine fundus. In most cases this will be a Pipelle de Cornier® biopsy catheter, but the type of biopsy catheter may vary per participating centre. The piston will be drawn back to the end of the biopsy catheter. Then the catheter is slowly retracted while rotating over several ranges of 360 degrees during 1 to 2 min. The protocol differs slightly for the subgroup of whom endometrial tissue will be stored. Firstly, a sterile gown, gloves and cap will be worn during the procedure in order to minimalize RNase contamination. Secondly, cleaning of the cervix will not be performed with betadine but with sterile saline solution (sodium chloride 0.9%) because the outcomes of RNA analysis could be affected by betadine. In order to minimalize RNase contamination, sterile gloves are changed for new gloves directly after the procedure. The endometrial tissue will be divided into three equal parts and each part will be stored in a Tissue Sampling Storage Tube (3 ml, Cat. No. 68–4000-00, Fluid X). In the coordinating centre (UMC Utrecht) and in one of the participating centres (Diakonessenhuis), two tubes will be snap-frozen in liquid nitrogen and one tube will be used to collect tissue for slow freezing of cells for organoid development. In the other tissue-storing centres, three tubes will be snap-frozen. The process of snap-freezing is performed within maximally 3 min from taking the biopsy. Subsequently, the snap-frozen tissue will be stored in freezers at − 80 °C until later determination of RNA profiles. Due to rapid developments in the field of genetics, the suited gene sets and techniques will be determined at the time of tissue analysis. The endometrial tissue intended for development of endometrial organoids will first be slow-frozen in medium containing dimethylsulfoxide (DMSO) and then stored in liquid nitrogen at − 196 °C until later formation of endometrial organoids. For this latter procedure the technique published by Boretto et al. (2017) [17] will be used.

Both the intervention and control group will undergo at least 6 months of timed intercourse from the moment of randomization. An overview of the study design and a study time schedule can be found in respectively Fig. 1 and Fig. 2.

Follow-up

The follow-up duration is 12 months from the moment of randomization. Each centre registers whether a patient has conceived during the follow-up period. When a woman has become pregnant, she will undergo ultrasounds at 7–8 weeks and 10–12 weeks of gestation, after which she will continue her prenatal visits with her midwife or gynaecologist. The outcome of her pregnancy will be obtained after her due date by a structured questionnaire.

Data on all pregnancies and further fertility treatment(s) within 12 months after randomization will be obtained by a structured questionnaire. If available, this information will also be extracted from the electronic patient file.

Monitoring

Monitoring will be performed by a monitor from the Dutch Consortium for Healthcare Evaluation in Obstetrics and Gynaecology. Furthermore, we have installed a Data and Safety Monitoring Board (DSMB) from the Dutch Consortium for Healthcare Evaluation and Research in Obstetrics and Gynecology. Roles of the DSMB include monitoring for evidence of treatment harm, such as serious adverse events, and advising on continuation or termination of the trial. Interim analyses have not been planned.

Data collection and data analysis

Data will be collected in a web based registration system. Database cleaning will consist of internal consistency checks and identification of database entries outside expected ranges. Analysis will primarily be conducted according to the intention-to-treat principle. In case many women allocated to the intervention group ended up not having the scratch procedure (or vice versa), a per-protocol analysis may also be performed to clarify whether there may have been an underestimation of the magnitude of effect. The researchers performing the data collection and data analysis are not blinded to group allocation.

SPSS Statistics (IBM Corporation, Armonk, NY, United States), R (R Foundation, New Zealand) and Excel (Microsoft, Washington, United States) will be used to perform the statistical analysis. The primary and secondary outcome variables (except the costs, of which analysis is described below) will be compared between the treatment arms and expressed as relative risk with 95% confidence interval. A p-value less than 0.05 will be considered to indicate a statistically significant difference.

Cost-effectiveness analysis

A cost-effectiveness analysis covering a period of 12 months will be performed parallel to the clinical trial for economic evaluation, which is designed from both a healthcare and societal perspective. This cost-effectiveness analysis will be based on live birth rate and average costs per patient, taking into account any additional fertility treatments patients undergo during these 12 months. The impact on patient’s life will be expressed in quality-adjusted life years (QALY). A decision model will be used to evaluate the optimal strategy, taking into account the time to pregnancy, (in) direct costs and estimated QALY. Additionally, the incremental cost-effectiveness ratio (ICER) and long-term costs, such as delivery and perinatal costs, will be determined. Sensitivity analysis will define robustness of the results.