Relation of Endometrioma and Deep Infiltrating Endometriosis Surgery on Ovarian Reserve: A Large Single-Centre, Cross-Sectional Study

The prevalence of endometriosis is approximately 10% overall among reproductive age women, whose most common symptoms are infertility and pelvic pain. However, the prevalence of endometriosis in infertile women is about 20-50% ( 1 ). Although the relationship between endometriosis and infertility has been well-established, the precise association between the two is not well-known ( 2 , 3 ). The connection between endometriosis and infertility is multifactorial, being the result of a series of events. Endometriosis not only affects the fallopian tubes, but also the quality and number of oocytes, and endometrial receptivity ( 4 ). Since minimal forms of the endometriosis disease, such as tubal factor infertility, have limited impact in women under 35, researchers mainly focus on moderate to severe cases like endometrioma (OMA) and deep infiltrating endometriosis (DIE), due to their significant effects on reducing fertility and ovarian reserve ( 5 ). OMA plays a role in reducing the ovarian reserve in women with endometriosis compared to their healthy peers. OMA surgery reduces the ovarian reserve and fertility and this post-surgical decrease in ovarian reserve persists regardless of the techniques used to minimise damage to the ovarian tissue during surgery. Therefore, it is recommended that patients with infertility save an appropriate number of embryos prior to OMA surgery to prevent reductions in ovarian reserve ( 6 - 10 ). There are numerous studies on OMA; however, studies that pertain to DIE and its surgical effects on ovarian reserve are limited, and the results are unclear. The prevalence of DIE is 6.5%. Its associations with superficial endometriosis, OMA, and pelvic adhesion are reported to be 61.3, 50.5, and 74.2%, respectively, in the literature ( 11 ). Pre-surgical data about these lesions are obtained from imaging methods, which exhibit significant heterogeneity. Of note, ultrasound is operator-dependent and different operators may use different terms for the same structures and locations. Therefore, the most reliable data in the DIE group are based on intra-operative findings. Additionally, DIE surgeries are extensive procedures that have more complications, particularly in cases of colorectal involvement compared to other gynaecologic pelvic surgeries ( 12 ). Although it is reported that surgery for DIE lesions is associated with an increased chance of spontaneous pregnancy, it is important to point out the limitations with these studies that include the absence of a control group; data obtained from more than one surgeon, which may cause variations in the surgical procedures; not all patients were infertile; simultaneous examination of spontaneous and in vitro fertilisation (IVF) pregnancies within the same group; and lack of mention regarding the decrease in ovarian reserve following these extensive pelvic surgeries, which is a significant outcome. Limited data exists on spontaneous pregnancy and infertility with DIE, and no reliable information or evidence-based protocols exist for the management of DIE lesions in infertile women ( 13 - 15 ). In this single-centre study, we investigated the rate of decreased ovarian reserve based on anti-Müllerian hormone (AMH) levels in three groups (OMA, OMA+DIE, and DIE) - prior to endometriosis surgery, and at four and eight months after surgery. The results of this case-control study could assist with more informed decisions in terms of surgery and preoperative fertility preservation in women with simultaneous infertility and DIE involvement.

The AMH levels of 508 patients in the three groups (OMA, OMA+DIE, and DIE) were examined and compared to 50 healthy controls prior to surgical in tervention, and at four and eight months after surgery. Out of 508 patients, 27.5% (156) were in the OMA group, 42.11% (235) in the OMA+DIE group, and 20.96% (117) in the DIE group. Table 1 shows the demographic data of the patients and the healthy group. There is no significant difference between the patients and the control group in terms of age (P=0.076) and body mass index (BMI) (P=0.109). Data normality were checked using the Kolmogorov-Smirnov and Shapiro-Wilk tests. Relationship between age and BMI with group Data are presented as mean ± SD and median (IQR). OMA; Endometrioma, DIE; Deep infiltrating endometriosis, BMI; Body mass index, SD; Standard deviation, IQR; The interquartile range, and *; Kruskal-Wallis test. ASRM scoring for all three groups and patients’ fertility status are mentioned in Table 2. There is no significant difference in the size of the OMA between groups one and two ( Table 2 ). The ASRM score in group two was higher than the other groups (P<0.001). All patients belonged to stage three or four endometriosis according to the ASRM scoring system for endometriosis. In the OMA group, 76.3% of the patients had a unilat eral cyst; in the OMA+DIE group, 53.2% had unilateral OMA (P<0.001). Table 3 shows the changes in AMH levels in the groups before surgery, and at four and eight months after surgery compared to the control group. The DIE group had lower baseline AMH levels compared to groups I and III, and the control group (P<0.0001). Serum AMH levels significantly decreased after surgery in all three groups (P<0.001). There were no significant differences in AMH levels at four and eight months after surgery in all three groups of patients (Tables 3 , 4 , Fig .1 , P=0.202). Relationship between fecundity variables with groups Data are presented as mean ± SD and median (IQR). OMA; Endometrioma, DIE; Deep infiltrating endometriosis, SD; Standard deviation, IQR; The interquartile range, *; Kruskal-Wallis test, and **; Mann-Whitney U test, The trend in reduction in AMH levels in the groups before surgery, and four and eight months after surgery. AMH; Anti-Müllerian hormone, OMA; Endometrioma, and DIE; Deep infiltrative endometriosis. Comparison of AMH levels in the study groups to the control group before surgery, and four and eight months after surgery Data are presented as mean ± SD and median (IQR). AMH; Anti-Müllerian hormone, OMA; Endometrioma, DIE; Deep infiltrating endometriosis, SD; Standard deviation, IQR; The interquar tile range, *; Kruskal-Wallis test, and **; Freidman’s test. The decrease in serum AMH levels after surgery was 49.84% in the OMA group, 62.20% in the OMA+DIE group, and 43.46% in the DIE group. The OMA+DIE group had the most significant decrease (Tables 3 , 4 ). In the OMA and OMA+DIE groups, larger cyst sizes correlated with a greater decrease in AMH levels by month eight in the OMA (r=-0.23, P=0.071) and OMA+DIE (r= 0.14, P=0.087) groups. Comparison of decrease in AMH levels at different time points after surgery Data are presented as mean ± SD. AMH; Anti-Müllerian hormone, OMA; Endometrioma, DIE; Deep infiltrating endometriosis, SD; Standard deviation, and *; Kruskal-Wallis test. P value within groups: <0.0001.

The effects of OMA surgery on reduction of ovarian reserve and the need for fertility preservation in these patients, especially in cases of simultaneous infertility, is well-documented. Nonetheless, there are no definitive recommendations for patients with DIE in this regard ( 17 ). The management of severe DIE in patients who desire to become pregnant is a focus of recent studies. However, most research is retrospective and non-comparative; hence, the role of fertility preservation before endome triosis surgery in women with DIE is not clear. Furthermore, the removal of DIE before an assisted reproductive technique procedure in patients with endometriosis-relat ed infertility has not been well established ( 17 ). Although some researchers reported a beneficial effect of DIE surgery on pregnancy outcomes with a 45% improvement in IVF results, there is no randomised study that confirms this improvement or the effect of surgery on spontane ous pregnancy rate in these patients ( 14 , 15 ). DIE surgery requires a highly skilled surgeon and cannot be routinely offered to all patients because of the inevitable complications. We divided 508 endometriosis patients into three groups (OMA, OMA+DIE, and DIE) and compared them with 50 healthy women. Based on our findings, the baseline AMH levels in the DIE and OMA+DIE groups were much lower than the OMA and control groups. The baseline AMH in the OMA and control groups was similar. The decrease in AMH levels in all three groups after surgery was statistically significant. There was no evidence of AMH recov ery in any of the surgical groups at four and eight months following surgery. Based on an extensive literature search, our study is the largest to date, involving the highest number of endometriosis cases, and uniquely investigates AMH as a marker to assess ovarian reserve both before and after surgery across three different groups of endometriosis patients. This study was conducted on a homogeneous population in terms of age and BMI. All surgeries were performed by a skilled surgeon, which eliminated the bias of the surgical technique. Limitations of this study include not utilising the ENZI AN classification system for imaging and operation note reports; this classification can determine the severity of the disease better than the ASRM scoring system. In addition, we did not assess fertility outcomes in all three surgical groups and compare them with the normal population ( 18 ). In line with our results, Ashrafi et al. ( 19 ) conducted a study on 125 patients who underwent endometriosis sur gery. Their patients were divided into three groups [OMA (n=40), DIE (n=58), and OMA+DIE (n=27)]. The results showed a decrease in ovarian reserve, with an increase in severity and ASRM score. In their study, the lowest post operative AMH level was observed in the OMA+DIE group (1.4 ± 1.4 ng/dl). Their study aimed to investigate the number of retrieved oocytes from all three endometriosis groups. They observed the lowest reserve in the OMA+DIE group (P<0.001); however, AMH levels be fore and after surgery were not compared. The ASRM scores in the groups were much less than our current study ( 19 ). Papaleo et al. ( 20 ) investigated the ovarian reserve in 51 patients after endometriosis surgery. The patients were divided into two groups: OMA (n=27) and OMA+DIE (n=24), all of whom had ASRM stages 3 and 4 endometriosis. Despite the similar size of OMA in both groups (4.6 ± 2.1 cm versus 4.2 ± 2.0 cm, P=0.04), the number of antral follicles under the monitor on the third day of the menstrual cycle for performing intracytoplasmic sperm injection was significantly lower in the second group. However, they did not mention the ovarian reserve prior to the surgery. Goodman et al. ( 21 ) compared AMH levels in 58 patients who underwent OMA surgery to 58 healthy con trols. They concluded that the reduced AMH levels after surgery depended on the initial AMH level as well as the size and laterality of the OMA. They reported a 52% AMH drop in case of bilateral OMA versus a 17.5% decline in unilateral ovarian involvement with OMA after surgery. This decrease continued up to six months after surgery, but it was not significant. There was no evidence of recovery in AMH levels during the follow-up period. DIE cases were not investigated in this study ( 21 ). Sarbazi et al. ( 22 ) examined 174 endometriosis patients in three groups: OMA (n=33), DIE (n=6), and OMA+DIE (n=135). They observed a general decrease in AMH levels compared to the baseline value of 2.8 ± 1.86 (ng/dl) before surgery to 1.76 ± 1.40 (ng/dl) after surgery; the results were not separately reported for each group ( 22 ). Considering the 6.5% prevalence of DIE, as the only form of the endometriosis disease in the literature, the number of articles is much lower than OMA and OMA+DIE. Nevertheless, in our study, we had an adequate number of cases that presented with DIE. The burnout theory in endometriosis was proposed for the first time in 2007 by Dolmans et al. ( 23 ) According to this theory, ovarian homeostasis is very important in causing follicle loss in cases of iatrogenic ovotoxicity. This theory also notes that any imbalance in ovarian homeostasis, even during normal and physiological condi tions, can lead to follicular burnout ( 23 , 24 ). Any process that accelerates follicle activation and reduces in follicle inhibition can lead to loss of ovarian reserve (e.g., the aging process). These processes include those causing toxicity in the ovary (OMA), either directly or through an indirect mechanism, and contribute to vascular infarction, hypoxia, increased oxidative stress, increased oxygen species in the pelvic environment, and even induction of the short term PTEN inhibition or PI3K activation in the genes of cortical ovarian neurons that lead to apoptosis of ovarian cells ( 25 - 27 ). According to the burnout theory, the presence of endometriosis lesions in the pelvis with an increase in pro-inflammatory cytokines and adhesion molecules, such as nitric oxide, tumour necrosis factor, interleukin (IL) 6, IL8, and interferon gamma, as well as proliferation and fibrosis induction, would lead to disruption of the blood supply to the ovaries. Simultaneously, with an increase in reactive oxygen species in the pelvis, these processes can lead to ovarian aging and rapid follicular activation and loss. This phenomenon occurs not only in the presence of OMA, but also in the presence of DIE, and leads to a decrease in ovarian reserve prior to surgery in these patients ( 28 - 31 ).

Although the presence of OMA is a pivotal factor for predicting reduced ovarian reserve in endometriosis patients, the role of DIE should not be ignored. It is important to diagnose DIE in patients with infertility who do not have OMA. Conducting target ultrasonography to detect pelvic DIE lesions and fertility preservation before surgery in these patients is of utmost importance.

This cross-sectional study was conducted in the Obstet rics and Gynaecology's Department of Shiraz University of Medical Sciences (Shiraz, Iran) after receiving approval from the Ethics Committee of Shiraz University of Medical Sciences (IR.SUMS.MED.REC.1398.580). The study population consisted of women who were referred to the tertiary care centre for laparoscopic endometriosis surgery from June 2018 to December 2022. Participants signed an informed consent form prior to entering the study. The inclusion criteria comprised women with confirmed cases of endometriosis based on histopathological samples who underwent surgery for the following indica tions: endometriosis-related pain unresponsive to medication; tubal involvement in cases of infertility; complete family planning; inability to use hormonal drugs; need for pathology samples; unwillingness to receive medical treatment; or preference for treatment with the lowest recurrence rate. The exclusion criteria consisted of women over the age of 40 with a previous history of endometriosis surgery, any history of chemotherapy or radiotherapy, autoimmune diseases, simultaneous adenomyosis, or uterine leiomyoma. We gathered clinical and biological data from all patients admitted to the operation room for their endome triosis surgery. General information and medical history were recorded through face-to-face interviews conducted by a specialist physician during the initial visit. The diagnosis of endometriosis was confirmed through surgical detection and histological analysis. Serum samples were collected during the month prior to the surgery, as well as four and eight months after surgery to measure AMH levels. A commercial ELISA kit (AMH Gen II ELISA; Beckman Coulter, Inc., Brea, CA, USA) was utilised for this measurement. Each calibrator, control or test sample was prepared by mixing one part of the sample with five parts of AMH Gen II assay buffer, with no dilution factor required for this preparation method. Samples that ex ceeded the highest calibrator were diluted using a sample diluent and retested according to the kit’s instructions. The intra-assay and inter-assay coefficients of variation were 5.4% and 5.6%, respectively. The interventions were performed by a physician with extensive expertise in endometriosis management. Additionally, the conservative laparoscopy was performed using mechanical instruments and electrosurgery. Surgery for DIE was performed according to previously reported surgical procedures ( 12 ). The adhesions were sectioned with micro scissors. For OMA, the ovaries were completely mobilised, the cysts were evacuated and rinsed with normal saline, and excised by countertraction applied to the pseudocapsule and normal gonadal cortex with atraumatic micro-forceps. Hemostasis was achieved using selective bipolar coagulation. The disease was staged according to the American Society for Reproductive Medicine (ASRM) classification ( 3 ). The patients were divided into three groups according to localisation of the endometriosis during surgical staging ( 16 ): (I) OMA (n=156), (II) OMA+DIE (n=235), and (III) DIE (n=117). For OMA, the recorded surgical data included size, number, and location (unilateral or bilat eral) of the endometriosis. DIE included ureteral endometriosis, rectovaginal endometriosis, and rectosigmoid endometriosis. The baseline AMH levels of all the patients were compared to 50 healthy women of the same age (control group). The normality of quantitative variables was assessed using the Kolmogorov-Smirnov test. Quantitative vari ables were reported as mean ± standard deviation (SD; median and interquartile range), and qualitative vari ables were reported as numbers and percentages. The Kruskal-Wallis test was utilised to assess the relationship between quantitative and qualitative (multivariate) variables. The Freidman test was used to evaluate quantitative variables over time. Data normality were checked using the Kolmogorov-Smirnov and Shapiro Wilk tests. The data were analysed using the Statistical Package for the Social Sciences 20.0 (SPSS, IBM Corp., Armonk, NY, USA).