If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Clinical Medicine, Aarhus University, Incuba Skejby, Building 2, Palle Juul-Jensens Boulevard 82, 8200 Aarhus N, DenmarkDepartment of Urology, Aarhus University Hospital, Palle Juul-Jensens Boulevard 99, 8200 Aarhus N, Denmark
Department of Clinical Medicine, Aarhus University, Incuba Skejby, Building 2, Palle Juul-Jensens Boulevard 82, 8200 Aarhus N, DenmarkDepartment of Obstetrics and Gynecology, Aarhus University Hospital, Palle Juul-Jensens Boulevard 99, 8200 Aarhus N, Denmark
Pregnancy rate after surgery for bowel endometriosis is 60.6%.
Complications do not seem to affect live birth rate or time to pregnancy.
Pregnancy rate after surgery and expectant management are comparable.
The aim of the study was to investigate pregnancy and live birth rate after surgical resection of rectosigmoid deep infiltrating endometriosis (DIE), and study if complications affect these rates.
Historical case series. 193 patients with rectosigmoid DIE and pregnancy intention undergoing a rectosigmoid resection for DIE from January 2009 to May 2019. All operations were performed at the Department of Obstetrics and Gynecology, Aarhus University Hospital, Denmark. Surgical and fertility outcome data were obtained through patient files. Anonymized data was analyzed statistically. Normally distributed continuous variables are stated as means, categorical data as percentages and time to pregnancy as Kaplan-Meier failure function. Live birth rates stratified on complications were tested with chi2 test.
117 patients became pregnant postoperatively with a pregnancy and live birth rate of 60.6% and 53.9%, respectively. 39 patients (20.2%) became pregnant spontaneously and 78 patients (40.4%) by intrauterine insemination or assisted reproductive technologies. Median time to pregnancy after surgery was 12.4 months (range: 0.4–58). Clavien-Dindo complication grade III (none grade IV) was registered among 16.6%. These patients had pregnancy and live birth rates of 50%, not statistically significantly different from those without complications.
Postoperative pregnancy and live birth rates after resection of rectosigmoid endometriosis in this study are in line with conservative treatment, when comparing with the literature. Interestingly, complications (Clavien-Dindo grade III) did not affect live birth rate or time to pregnancy.
]. Deep infiltrating endometriosis (DIE) is defined as endometriotic subperitoneal invasion exceeding 5 mm in depth and can be located uterosacrally, in the intestine, vagina, bladder and ureter (most to least frequent) [
]. However, besides gonadotropin-releasing hormone (GnRH) agonist in relation to fertility treatment, hormonal therapy has no place for patients with pregnancy intention, because of the contraceptive effect.
Assisted reproductive technologies (ART) is generally accepted as infertility treatment among women with endometriosis, and shows similar live births rates to women without endometriosis undergoing ART [
], as well as if complications affect pregnancy and live birth rate.
The aim of this study was to evaluate pregnancy and live birth rates after surgical treatment for colorectal DIE as well as if they are affected by complications.
Material and methods
Single-center 10-year experience involving 193 patients who had surgery for DIE and pregnancy intention from 1st of January 2009 to 31st of May 2019. The study was conducted at Department of Obstetrics and Gynecology at Aarhus University Hospital (AUH), one of two tertial referral centers for DIE in Denmark. Primary outcome was live birth rate according to Harbin consensus [
] and pregnancy rate, not including miscarriages. Rates were stratified after conception mode. Pregnancy was defined as ongoing clinical pregnancy by positive pregnancy test or live fetus in gestational week six to eight visualized by ultrasound. Secondary outcomes were time to pregnancy, postoperative complications, and stoma.
Surgery was performed as minimally invasive laparoscopy with rectosigmoid segment or disc resection, depending on size and accessibility of the infiltration. Endometriosis in other locations was surgically excised as well. Surgery was performed as joined venture between a gynecologist and general surgeon. During the study period, the specialized team consisted of four gynecologist and seven colorectal surgeons. Postoperative follow-up was scheduled at four months, one and two years after surgery. Follow-ups reviewed pain and fertility status and were carried out at the outpatient clinic or by phone.
The patients were identified by The Danish Civil Registration System (CPR). Inclusion criteria were ICD-10 code for rectosigmoid endometriosis (DN805C) and procedure code for laparoscopic rectosigmoid resection (KJGB01). Exclusion criteria were no postoperative pregnancy intention and age above 41 years at date of surgery, as this is the age limit for publicly funded fertility treatment in Denmark.
Data collection and management
Data were obtained from patient files and included the following: Demographic information, previous surgery, perioperative surgical information, complications, hospitalization time, magnetic resonance imaging (MRI) information on the extend of the bowel infiltrate and other locations of endometriosis as well as pre- and postoperative pregnancy and fertility. For each patient an electronic case report form (eCRF) was created in the secure web-based database platform REDCap® [
]. This platform provides an audit trail of all activities regarding all eCRF. Data entry was maintained by two of the authors (MM and MR).
Permission to carry out this study and review patient files has been granted by Central Region Denmark (no. 1–16-02–345-16, 8th June 2016) and the Danish Patient Safety Authority (no. 3–3013-2894/1, 21st March 2019). No permission from the Committee on Health Research Ethics was necessary due to study design.
Prior to statistical analysis, data were anonymized. Normally distributed continuous variables are stated as means with standard deviation (SD) or 95% confidence intervals (95% CI). Data, not normally distributed, were presented as median and range. Categorical data are shown as percentages. Kaplan-Meier failure function illustrates time to pregnancy. Time to pregnancy limit was set at five years for patients not obtaining pregnancy. Pregnancy and livebirths rates stratified on complications were tested with Chi2 test according assumptions related to populations size in each cell. Statistical analyses were performed in Stata 16 (StataCorp, College Station, TX, USA).
The final study population for analysis consisted of 193 patients (Fig. 1). If evaluation of ovaries and salpinges during surgery indicated possibility for spontaneous pregnancy, patients were encouraged to try for spontaneous conception. Some patients, however, decided to continue directly to ART.
Table 1 presents baseline characteristics of study population. The department’s guideline requires a preoperative BMI of 30 or lower as a preventative measure to decrease risk of peri- and postoperative complications. Nevertheless, 14 patients had surgery with above BMI 30, and three of them had a postoperative complication. The majority, 107 patients, were in the normal BMI range. “Other” infertility reasons were single civil status (three patients), polycystic ovary syndrome (one patient), uterus didelphys (one patient) and previous bilateral salpingo-oophorectomy (one patient). Six patients had combined decreased tubal passage and male factor infertility. Five patients had one ovary.
MRI was the primary imaging modality for DIE diagnosis and surgical planning (Table 2). Only five patients did not get an MRI scan prior to surgery. Not all MRIs are stated with exact measurements but just with the statement, that a rectosigmoid nodule is present.
The main part, 188 (97.4%) surgeries were performed laparoscopically. Segmental resection was performed among 149 patients (77.2%) and the remaining 22.8% had a disc excision. No patients were treated with shaving only. Median duration of operation was 178 min (range: 35 – 454), and median hospitalization time was 6 days (range: 2 – 35).
Live birth rate was 53.9% (104/193), and the majority (61.5%) obtained pregnancy after ART. Table 3 provides information on conception mode and pregnancy. Specific information on the number of patients who intended spontaneous conception postoperatively prior to ART is not available, but 126 patients were treated with ART, 67.5% of whom got pregnant. 88.9% of all pregnancies resulted in live birth. Pregnancy rate for women with MR-verified adenomyosis was 64.3% (18/28).
Table 3Postoperative fertility and pregnancy characteristics.
Pregnancy rate (%)
Time to pregnancy, median in months (range)
12.2 (0.4 – 58)
Twin pregnancies (%)
Pregnancy rate in relation to mode of conception (%)
Women with primary infertility preoperatively (n = 148) had a postoperative pregnancy rate of 64.9% with 34.4% as spontaneous pregnancies. Patients with secondary infertility prior to surgery (parity of one or two, n = 45) had a postoperative pregnancy rate of 46.7% with 28.6% as spontaneous pregnancies. Mean age of nullipara and multipara was 31.4 years (95% CI: 30.8 – 32.0) and 32.7 years (95% CI: 31.7 – 33.8), respectively.
Time to pregnancy is shown in Fig. 2, Fig. 3. Median time to spontaneous pregnancy was 9.8 months (range: 0.4 – 44) and 12.6 months (range: 1.8 – 58) for ART pregnancies. After operation 67 patients got medical intervention with median duration time of 7 months (range: 1 – 53). 59.7% (40/67) became pregnant, 72.5% by ART, 16.7 months (range: 3.8 – 49.5) after treatment termination.
Postoperative complications occurred for 38 patients (19.7%). Five had two simultaneous complications after surgery. Specific types of complications are shown in Table 4. One “Other” complication was rotation of stoma. Eleven patients (5.7%) had Clavien-Dindo Classification [
] grade III anastomotic leakage with need of surgical intervention. Complication rates were 15.9% (7/44) and 16.8% (25/149) for disc and segmental resection, respectively. Forty-one patients (21.2%) got a temporary stoma, either during primary surgery or secondary in relation to a complication. One patient (0.5%) got a permanent stoma.
Table 4Postoperative complications.
Types of complications (all Clavien-Dindo grades) (%)
38 13 12 5 2 3 1 1 1
19.7 6.7 6.2 2.6 1.0 1.6 0.5 0.5 0.5
Complication grade in regard to the Clavien-Dindo Classification:
Days until diagnosis of complication, median (range)
5.5 (0 – 49)
During primary surgery
During secondary surgery
Days until reversed stoma, median (range)
110 (25 – 913)
Among patients with one or two postoperative complications 16 patients (50%) became pregnant, seven by spontaneous conception and nine by ART. Median time to pregnancy 10.7 months (range: 2.6 – 58.0) for Clavien-Dindo III. Complications did not statistically significantly affect pregnancy and live birth rates (Table 5).
In this study, we found a pregnancy rate of 60.6% and a live birth rate of 53.9% after surgical treatment for colorectal DIE. This was, interestingly, not affected by complications, neither did it prolong time to pregnancy.
Our study reports a high live birth rate, which could be explained by exclusion of early miscarriage before gestational age of 12 weeks. In spite of this, when looking at study population size and the extent of the surgery, our pregnancy rate is comparable to similar studies. Other studies with complete DIE surgery show great variation in overall postoperative pregnancy rates from 42.9% to 81% [
]. The high pregnancy rate of 81% may be explained by fewer patients included compared to our study and a longer follow-up for all patients. A review by Cohen et al. found a pregnancy rate of 46.9% for patients surgically treated for bowel endometriosis [
] reported a live birth rate of 44% after surgery for DIE, but the study population was smaller (n = 25) and only 17 had bowel endometriosis (six underwent segment resection and 11 rectal shaving). A review found a postoperative spontaneous pregnancy rate of 31.4% [
]. Hence, segmental resection of bowel endometriosis does not seem to affect spontaneous pregnancy significantly.
Our results are comparable with expectant management, even though it is close to the lower end of the interval. Explanations could be, that 21.8% of our population had an additional cause of infertility, which made spontaneous conception impossible. Hence, these confounding infertility etiologies suggest that ART as well as surgery could be the reason for pregnancy. Patients who previously underwent ART, went directly into ART postoperatively instead of intending spontaneous conception. Furthermore, focus on motivating and encouraging the patients, who were candidates for spontaneous conception, to try for at least 12 months prior to ART have not been routine practice throughout all 10 years. Eight patients were not able to obtain pregnancy with ART preoperatively but got pregnant postoperatively. This relatively small number is not sufficient to state, that surgery for DIE improves fertility in itself.
According to cumulative pregnancy rate in Fig. 3, spontaneous pregnancies seem to be achieved quicker compared to ART pregnancies, which is in accordance with Roman et al. [
]. This could possibly be due to latency related to ART referral and downregulation as part of ART.
Some patients needed medical therapy to manage DIE postoperatively. This group had a median time to pregnancy of 16.7 months after treatment cessation, indicating a delay. This could be due to a delay in obtaining a regular menstrual cycle after medical endometriosis treatment with anticontraceptive.
A Clavien-Dindo grade III complication rate of 16.6% is higher than we expected. It reflects the complicated nature of these operations. It might also represent a learning curve phenomenon during the study period. But according to Vidal F el. al., the surgeons experience does not seem to affect fertility outcome [
We know that the first five study years were associated with more segmental resections and a higher percentage of stomas at primary surgery. Additionally, unpublished data from our latest DIE surgeries, after the study period, reveal a lower complication rate regarding anastomotic leakage and ureter lesions. The majority had a segmental resection, which indicates a severe degree of bowel involvement. To compare, disc excision was related to smaller infiltrates and lower complication rates. This is in line with existing literature describing fewer complications with a more conservative surgical approach [
]. On the other hand, Roman et al. showed no significant difference in complication risk between conservative surgical approach and segmental colorectal resection, apart from increased risk of bowel stenosis after resection [
The strengths of our study include the single-center specialized setting. The endometriosis team at our department is highly experienced in treatment of DIE. Furthermore, our study population is large compared to other retrospective and prospective studies [
]. Due to the great availability of public fertility treatment in Denmark, medical records regarding ART were easily obtainable for the majority of patients included in our study, hence securing sufficient data.
Our study is limited by the retrospective design not having a reference group. Additionally, it is a single center study which affects generalizability. At referral, the patients were severely affected by pain symptoms related to DIE. Therefore, it would not have been ethically acceptable not to offer these patients surgery and await spontaneous pregnancy instead.
Information on pain and other endometriosis-associated symptoms would have been valuable. A previous study from our department showed significant improvement in pain and quality of life among 175 women one year after bowel resection for rectosigmoid endometriosis [
]. Smoking habits and alcohol consumption is missing and would have been beneficial, as these factors can impact disease and fertility status as well as surgical complications.
Informing patients with bowel endometriosis and pregnancy intention regarding treatment choice is not easy. But our results indicate that the risk of complications does not seem to affect either chance of pregnancy, live birth or time to pregnancy. Hence, this implies that patients suffering from bowel endometriosis should not stay away from surgery because of fear of complications affecting their chances of pregnancy. However, it is always important to inform these patients thoroughly preoperatively and take into account every patient’s individual preference and priority in order for her to give informed consent on a solid basis. To elicit definitively whether surgery for bowel endometriosis improves fertility, RCT is needed and this is in progress at present.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Immune-inflammation gene signatures in endometriosis patients.