Background

Polycystic ovarian syndrome is the most common cause of anovulatory subfertility [1]. Weight reduction, lifestyle modification, and ovulation induction are the recommended initial management strategies [2, 3]. Laparoscopic ovarian drilling (LOD) has been suggested to induce ovulation in these women, especially those who fail to ovulate through ovulatory medications [4,5,6]. It has been suggested that the procedure is as effective as ovarian stimulation with exogenous gonadotropins [7], yet it does not increase multiple pregnancy rates or ovarian hyperstimulation syndrome (OHSS) rates. Many women may ovulate after LOD, yet they fail to conceive [8]. Those women may need to undergo IVF treatment in their pursuit for a baby.

Endometrial scratching is a procedure where the endometrium is subjected to physical trauma that caused injury to the functional layer of the endometrium mechanically [9,10,11,12]. It has been suggested that endometrial injury could improve IVF outcome in women with recurrent implantation failure after IVF [13]. Nonetheless, endometrial scratch has been also proposed to overcome subfertility in women with unexplained infertility [14]. Randomized controlled trials have also shown improvements of intrauterine insemination (IUI) results in women subjected to controlled endometrial injury prior to insemination [9, 10]. However, there were some other studies that have shown no benefit from the procedure [15, 16].

The aim of our study was to find out whether performing endometrial scratch at time of laparoscopic drilling would improve live birth rate in subfertile women with PCOS.

Patients and methods

Study design and participants

We conducted a parallel randomized controlled trial (RCT), approved by our university ethics committee. We approached all infertile women with anovulatory infertility due to PCOS referred for laparoscopic ovarian drilling in Mansoura University Teaching Hospitals in Mansoura, Egypt. Our hospital is a tertiary care center conducting between 600 and 700 laparoscopic surgeries per year for infertile women. The study was conducted during the period from April 2014 to April 2015 (last patient enrollment). Follow-up was continued for 9 months after laparoscopy. The last pregnancy was in December 2015. Last data collection was in September 2016. An informed written consent was obtained from all women who participated in the study.

Our inclusion criteria were women aged 20 and less than 39 and women with PCOS as diagnosed by Rotterdam criteria, fertile semen analysis according to WHO 2010, and bilateral tubal patency as demonstrated by hysterosalpingogram (HSG) [17, 18]. The exclusion criteria were suspected endometriosis, suspected uterine cavity anomaly or mass, associated male factor infertility, presence of endocrinopathy as thyroid dysfunction, and women subjected to endometrial curettage for any reason in the last 6 months.

Intervention

Women were admitted to our hospital 1 day before laparoscopic drilling. Women were randomized into two groups: group A (the intervention group) and group B (the control group). Randomization was through a computer-generated list of random numbers. Allocation of women to groups was through an opaque sealed envelope that had to be picked by a nurse in the operative theater. The surgeon was not blinded to the procedure while patients and data assessor were blinded to their allocation.

All women underwent a three-puncture laparoscopy procedure where laparoscopic ovarian drilling (LOD) was achieved. Ovarian drilling was performed through monopolar coagulation diathermy. Four punctures were performed. Each penetrates about 4 mm depth, using 40-W power that lasts for 4 s. In the intervention group (group A), endometrial scratching was performed at the end of laparoscopy by endometrial curette. The curette was introduced gently through the cervix up to the uterine fundus then withdrawn for 1 or 2 cm. One act of scratching was performed on the posterior wall of the uterus after the end of drilling. The obtained specimens were sent for histopathology. The control group (group B) had LOD only, and no endometrial scratch was performed.

Women in both groups were seen 3 months after laparoscopy and were asked whether they had a positive pregnancy test, still have oligomenorrhea, or had had regular periods. Women who had regular periods were subjected to folliculometry to confirm the establishment of ovulation while those with oligomenorrhea were subjected to ovulation induction with clomiphene citrate, tamoxifen, or letrozole. Women who did not respond to ovulatory oral medications were stimulated using exogenous gonadotropins using the low-dose step-up protocol with a 37.5 IU starting dose [19]. The primary outcome measure in this trial was live birth rate per woman randomized. Secondary outcome measures were clinical pregnancy rate, time to pregnancy, miscarriage rate, and multiple pregnancy rate. The study was registered in ClinicalTrials.gov with identifier number NCT02140398.

Definitions

Clinical pregnancy was defined as the presence of intrauterine gestational sac 1 or 2 weeks after positive pregnancy test in blood. Live birth was defined as the delivery of living fetus after 24 weeks gestation.

Statistical analysis

We estimated that the pregnancy rate after laparoscopic ovarian drilling was around 50% [20]. The intervention was suggested to boost pregnancy rate up to 70%. We calculated that we will need to study 93 experimental subjects and 93 control subjects to be able to reject the null hypothesis that the failure rates for experimental and control subjects are equal with a study power (probability) of 80%. The type I error probability associated with this test of this null hypothesis is 0.05 [21]. To compensate for dropouts, we calculated that we needed to randomize 210 women. We used SPSS 15 program. We adopted the intention-to-treat analysis.

Results

From 250 infertile women with PCOS, who were admitted to Obstetrics and Gynecology Department and were planned for laparoscopy, 220 cases meet eligibility criteria of the study. Two hundred ten infertile women with PCOS agreed to participate (Fig. 1).

Fig. 1
figure 1

Flow chart for the study recruitment and analysis process

Full size image

There was no difference in women randomized to each group as regards to age, BMI, FSH, and duration of infertility (Table 1).

Table 1 Patients’ characteristics
Full size table

There were no significant differences in ovulation rate, miscarriage rate, cumulative pregnancy rate, or live birth rate in women in group A compared to women in group B (Table 2).

Table 2 Clinical outcomes
Full size table

Discussion

The results of this randomized trial have shown no benefit of endometrial curettage in women with anovulatory infertility undergoing LOD in terms of pregnancy rate or live birth rate.

Up to our knowledge, there were no published similar trials to compare our results. However, the effect of endometrial scratching in overcoming infertility has been examined in different platforms. Endometrial scratch was suggested to improve pregnancy rates in women undergoing IVF, intrauterine insemination (IUI), or even practicing timed intercourse (TI). Women with unexplained subfertility have shown to have benefitted from the procedure without any further intervention [13]. Endometrial scratching may also boost live birth rate in women undergoing IUI using partners’ semen [9, 10]. Indeed, some published systematic reviews have identified an increase in pregnancy rates in subfertile women undergoing the intervention [12, 22].

Many studies had shown a potential benefit of the procedure in improving fertility performance in women with previous IVF failure undergoing a fresh cycle [13, 14]. However, the most recent systematic review has reported no effect of the procedure on women undergoing IVF treatment with a previous one failed IVF cycle [23].

However, there is still uncertainty about the value of the procedure in women undergoing it for the first time [23,24,25].

Many theories have been proposed to explain the favorable effect of endometrial injury. Endometrial healing, increased implantation mediators, and gene expression theories all had some biological plausibility [26, 27]. In our trial, the population was subfertile PCOS women in whom the apparent factor was anovulation. It seems, based on our findings, that controlled trauma to the endometrium did not make a positive effect to this group of patients.

Our results have shown almost 50% spontaneous ovulation rate after LOD. This is consistent with the reported 70–80% ovulation rate in the literature [8]. Women who failed to have spontaneous ovulation following LOD were subjected to ovulation induction [28, 29]. We used clomiphene citrate, tamoxifen, letrozole, and exogenous gonadotropins to induce ovulation in these women. Currently, there is no evidence of a difference in pregnancy rates between clomiphene citrate and tamoxifen in ovulation induction in women with PCOS; hence, we categorized women taking clomiphene and tamoxifen together. This is unlike the current growing evidence that the use of letrozole might be associated with higher pregnancy rates in ovulation induction for women with PCOS [30, 31]

In our study, we observed no evidence of a difference in miscarriage rate between the two groups. The overall miscarriage rate in pregnant women with PCOS varies up to 50% [8]. A Cochrane systematic review did not find any significant differences in the miscarriage rates in women treated by LOD compared to those adopted other medical modalities [8].

We thought that one of the strength points in our study is that it is a randomized trial with a predetermined sample size calculation. Although we calculated the sample size based on an assumption that pregnancy rates after LOD are 50%, in our series, the overall pregnancy rate was less than this. This might be one of the points that turned out to be the weakness point in this trial. Besides, our follow-up period was extended to only 9 months after drilling. There have been reports that the effect of drilling extends up to 1 year. Moreover, many clinicians consider any manipulation in the uterus could provoke an effect similar to the intentional scratching, and in our series, a uterine manipulator was inserted in all cases whether subjected to curettage or not. The implication of our results to the clinical world of subfertility is that we do not recommend doing the procedure for this population at time of laparoscopic drilling. However, it should be clear that our study was examining the effect of endometrial scratch in PCOS women undergoing drilling. We do not know whether the procedure might improve pregnancy rates in PCOS women undergoing ovulation induction through ovulation induction medications alone or not. Extrapolating the evidence from this trial to other interventions done prior to recommending laparoscopic drilling may be inappropriate.

Conclusion

To conclude, endometrial scratch at time of LOD for subfertile anovulatory women with PCOS may not lead to an improvement of live birth rates.