{"paper_id":"d08fcd0f-5c10-440e-8273-dcf94b501301","body_text":"Deep colorectal endometriosis affects about 20% of women with endometriosis. It has recently been shown that deep and/or ovarian endometriosis is prevalent in up to 19% of women attending a general gynecology clinic. \n 1 \n  Colorectal deep endometriosis (DE) is commonly associated with more pain and adverse gastrointestinal (GI) outcomes. \n 2 \n ,  \n 3 \n  Although a large percentage of women with colorectal DE can be managed conservatively , 4 \n  several studies have demonstrated a significant short‐ but also long‐term decrease in pain symptoms following surgical resection of colorectal DE. \n 5 \n ,  \n 6 \n ,  \n 7 \n ,  \n 8\nThere is, so far, no real consensus on how to optimally treat colorectal endometriosis surgically, since the application of the main three techniques, namely, segmental colorectal resection, full thickness discoid resection (FTDR) and so‐called shaving, depends on various factors which may differ from unit to unit and surgeon to surgeon. \n 9 \n ,  \n 10 \n  These include infiltration length and depth of DE, presence of multifocal disease and surgical experience and personal preference. \n 9 \n ,  \n 10 \n  In general, the majority of gynecological and/or colorectal surgeons may prefer segmental resection for multifocal and large nodules exceeding at least 3–4 cm, whereas small nodules <3 cm and/or DE not deeply infiltrating the muscular layer would be treated with FTDR resection or partial thickness discoid resection, or so‐called shaving. \n 10 \n  On the other hand, some authors do successfully perform shaving procedures in large lesions \n 11 \n  and apply FTDR resection in larger lesions >3 cm in diameter in over 65% of cases with good results and low postoperative complication rates. \n 12\nTo date, the only randomized controlled trial comparing segmental resection vs FTDR was unable to demonstrate differences in complication rates and functional outcomes. \n 6 \n  There are, however, several studies and reviews suggesting that discoid resection techniques and especially partial thickness discoid resection (ie rectal shaving) may confer lower postsurgical complication rates compared to segmental resection. \n 9 \n ,  \n 13 \n  Taking into account anastomotic height as a main factor for risk of leakage and rectovaginal fistula, a retrospective multicenter study was unable to support this view when complication rates of FTDR using transanal staplers were compared with a modified nerve and vessel sparing technique for segmental resection (NVSSR) in women undergoing low resections for colorectal DE. \n 14 \n  Within this, a recent systematic review on differences in surgical techniques of segmental resection for colorectal DE concluded that data so far were inhomogeneous and insufficient to show a proven advantage of nerve and nerve plus vessel sparing approaches. \n 15 \n  The primary aim of the present study was to further investigate changes in pre‐ and postsurgical GI function following a modified nerve and vessel sparing segmental resection (NVSSR) method compared with FTDR. Furthermore, changes in health profiles, pain symptoms, fertility issues and major postsurgical complications were evaluated prospectively.\n\nThis was a prospective comparative cohort study in a consecutive cohort of women who were treated with surgery for symptomatic colorectal DE by either NVSSR or FTDR at the Hospital of St. John Vienne and Rudolfinerhaus Private Clinic Vienna from April 2017 to May 2022.The primary outcome was change in GI function, which was measured by the prevalence of lower anterior resection syndrome (LARS) and the GI function‐related quality of life (GIQLI) score. Secondary outcomes were changes in pain scales, ie visual analog scale (VAS) of dysmenorrhea, dyspareunia, dyschezia and dysuria, Endometriosis Health Profile‐30 (EHP‐30) items, prevalence of postsurgical complications classified according to Clavien–Dindo and postsurgical fertility outcomes (pregnancy rate, time to pregnancy, miscarriage rate, preterm and term delivery rate and ongoing pregnancy rate).\nIncluded were patients who underwent surgery for symptomatic DE and showed endometriotic involvement of the rectum and/or sigmoid colon involving at least the serosal and muscular layer confirmed by histological analysis. Patients with diagnosed or suspected malignancy, women with a history of colorectal surgery or psychiatric disorders, and virgins were excluded from the primary analysis of postoperative complications. Patients who were lost to follow‐up and who refused to participate in the evaluation of GI function were also excluded from the final analysis to evaluate changes in GI function, EHP‐30 items, pain scores, and fertility outcomes.\nAs depicted in Figure  1 , a total of 162 consecutive women who underwent surgical treatment for symptomatic DE, all of whom showed involvement of the rectum and/or sigmoid colon involving at least the serosal and muscular layer confirmed by histological analysis, were included in the primary analysis. A power analysis was performed to evaluate a minimum patient number regarding the primary outcome. Assuming a small effect between the two groups at two measurement points, a minimum of 90 patients was necessary to reach a power of 80%. Presurgical diagnosis of colorectal endometriosis was made by transvaginal sonography in accordance with the IDEA consensus by GH (expert level sonographer, EFSUMB level III). \n 16 \n  The severity of DE was evaluated according to the rASRM score as well as by the #ENZIAN classification. \n 17 \n ,  \n 18 \n  Clinical data including age, fertility status and previous surgical treatments and symptoms described by a numerical 10‐point analog rating scale for dysmenorrhea, dyspareunia, dyschezia and dysuria and quality of life/EHP‐30 \n 19 \n  as well as LARS and GIQLI scores were evaluated (Figure  1 , Tables  1 ,  2 ,  3 ). LARS and GIQLI are commonly accepted tools for the evaluation of postoperative bowel dysfunction in patients undergoing a low anterior resection. \n 20 \n ,  \n 21 \n  Women who failed to achieve a clinical pregnancy by ≥12 months of regular unprotected intercourse were considered infertile. All infertile women had preoperative infertility workup including ovulation studies, transvaginal sonography examination and semen analysis. The primary indication for surgery was quality of life reducing pain not responding to conservative treatments and/or infertility. Surgical findings were documented and a postoperative check‐up was performed at 4–6 weeks followed by a telephone survey at the time of re‐evaluation of data to evaluate changes in GI function (LARS, GIQLI), EHP‐30, pain and fertility outcomes. Postsurgical complications were classified according to Clavien–Dindo. \n 22\nFlow chart depicting patients eligible for evaluation of postsurgical complications and analysis of pain, fertility and gastrointestinal function outcomes.\nPatient characteristics, demographic data and intraoperative findings of 121 women included in the final analysis following nerve‐vessel sparing segmental and full thickness discoid resection for colorectal deep endometriosis. All values given as  n  (%) unless indicated otherwise.\nAbbreviations: DE, colorectal deep endometriosis; ENZIAN, ENZIAN score for deep infiltrating endometriosis; FA, adenomyosis; FB, bladder; FTDR, full thickness discoid resection; FU, ureter; n.a., not applicable; NVSSR, nerve‐vessel sparing segmental resection; rASRM, revised American Society of Reproductive Medicine.\nBold values indicate statistically significance.\nIndependent samples  t ‐test.\nFisher's exact test.\nChi‐square test.\nClavien–Dindo grade II–IV postsurgical complications of 162 women undergoing nerve‐vessel sparing segmental and full thickness discoid resection for colorectal deep endometriosis. All values given as  n  (%).\nAbbreviations: DE, colorectal deep endometriosis; FTDR, full thickness discoid resection; n.a., not applicable; NVSSR, nerve‐vessel sparing segmental resection.\nSurgical outcomes regarding LARS and pain scores following nerve‐vessel sparing segmental resection and full thickness discoid resection for colorectal DE.\nNerve‐vessel sparing segmental resection (NVSSR) ( n  = 98)\nMean ± SD\nFull thickness discoid resection (FTDR) ( n  = 23)\nMean ± SD\nAbbreviations: DE, colorectal deep endometriosis; FTDR, full thickness discoid resection; LARS, low anterior resection syndrome; NVSSR, nerve‐vessel sparing segmental resection.\nBold values indicate statistically significance.\nWilcoxon signed‐rank test.\nPaired‐samples  t ‐test.\nThe primary indication for surgery was quality of life reducing pain not responding to conservative treatments and/or infertility. All surgical procedures were performed by one main gynecological surgeon (GH) in a multidisciplinary team setting consisting of two colorectal surgeons (BD, TB) and one urological surgeon. All procedures were initially performed laparoscopically under general anesthesia using 4 ports with single‐shot antibiotic treatment i.v. 1 hour prior to surgery. The surgical steps of the modified segmental resection technique NVSSR and FTDR technique have previously been defined by the authors in detail. \n 7 \n ,  \n 14 \n  In our practice and in accordance with general recommendations on surgery for colorectal DE \n 10 \n  we usually perform FTDR in DE lesions up to 3 cm in length and NVSSR in lesions >3 cm or multifocal disease based on the findings of presurgical transvaginal sonography based on the IDEA consensus \n 16 \n  for diagnosis of DE and the MUSA statement \n 23 \n  on diagnosis of uterine adenomyosis aiming to minimize the length of the resected segment. Surgical findings were documented and a postoperative check‐up was performed at 4–6 weeks followed by a telephone survey at the time of re‐evaluation of data to evaluate changes in GI function (LARS, GIQLI), EHP‐30, pain and fertility outcomes. Postsurgical complications were classified according to Clavien–Dindo. \n 22\nSample size calculation (power analysis) was conducted with G*Power (University of Duesseldorf, Germany) for the primary outcome. All statistical analyses were performed using SPSS 28. Means and standard deviation as well as frequencies and percentages were presented as descriptive results. Differences between the two surgical groups were tested for significance using independent samples  t ‐tests for metric data, and differences between pre‐ and postsurgery were tested for significance using dependent samples  t ‐tests. The changes between pre‐ and postsurgery for metric variables (EHP‐30 and GIQLI) were additionally tested multivariate using analysis of variances (within/between/interaction) for differences in changes between the two surgery groups. For ordinal data VAS, Wilcoxon tests were calculated for comparison between pre‐ and postsurgery. Differences in categorical data were tested for significance using Chi‐square tests; for dichotomous data, Fisher's exact test was applied. Correlations between metric variables were calculated using Pearson correlation. A value of  P  ≤ 0.05 was considered statistically significant.\nThe study was approved by the local Institutional Review Board (IRB at Hospital St. John of God 2017/1‐BB) on April 5, 2017. Patient consent to be included in the study was obtained prior to surgery.\n\nIn all, 162 patients were included in the primary analysis regarding perioperative outcomes and complications. Out of these women, 125 underwent NVSSR (77.2%) and 37 FTDR. All cases of presurgical diagnosis of colorectal DE by transvaginal ultrasound were confirmed by histopathological analysis. A total of 121 patients were included in the final analysis regarding the primary study outcome; 20/162 patients (12.3%) were lost to follow‐up and 21/162 (12.9%) did not agree to take part in the evaluation of GI function parameters pre‐ and postoperatively (Figure  1 ). The mean follow‐up interval was 42.27 (± 17.59) months with a mean follow‐up period of 41.11 (± 18.72) months in the NVSSR group ( n  = 98) and 41.96 (± 11.85) months in the FTDR group ( n  = 23), and was the time point of outcome measurement, ie time of postsurgical evaluation equaled the follow‐up interval. Patient characteristics of the final cohorts, demographic data and severity of DE are presented in Table  1 . Further data on intra‐ and perioperative complications and outcomes are presented in Tables  1  and  2 . No significant differences were observed in overall grade III complication rates between NVSSR and FTDR (Table  2 ). In the pre‐ and postsurgical evaluation of GI function, EHP‐30 and pain parameters, 121 patients were eligible for analysis.\nRegarding evaluation of pre‐ and postsurgical GI function, presurgical scores of LARS (ie LARS‐like symptoms) were observed in 59/121 patients (48.7%; NVSSR 27/98, 27.6% minor LARS; 22/98, 22.4% major LARS; FTDR 6/23, 26.1% minor LARS; 4/23, 17.4% major LARS). These symptoms significantly decreased following surgery, leaving 15/121 women (12.3%; NVSSR 9/98, 9% minor LARS; 4/98, 4.1% major LARS; FTDR 2/23, 8.7% minor LARS, 0/23, 0% major LARS; Figure  2 , Table  3 ). As depicted, worsening of LARS symptoms was only observed in 5/121 patients (4.1%, Tables  4  and  5 ). When comparing patients undergoing NVSSR and FTDR, worsening of LARS symptoms was exclusively observed in NVSSR patients (5/98, 5.1%) with de novo LARS in 3/98 women (3.0%, Table  5 ). However, there were no significant differences between postsurgical GIQLI ( P  = 0.849) and LARS ( P  = 0.678) scores in both intervention groups). The results were evaluated for possible confounders including BMI, age and height of anastomosis between the two cohorts. No significant factors influencing the results were observed. In contrast, we observed a significant correlation between the lower height of surgical anastomosis and the occurrence of postsurgical LARS ( r  = −0.194;  P  = 0.035) in the whole study cohort but no association between the length of the resected segment and postinterventional LARS ( r  = 0.038;  P  = 0.713).\nSemiquantitative data on lower anterior resection syndrome (LARS) scores before and after surgery for colorectal endometriosis with either nerve‐vessel sparing segmental resection (NVSSR) or full thickness discoid resection.\nCross‐tabulation depicting the prevalence of LARS pre‐ and postsurgically in 121 women undergoing colorectal surgery for bowel endometriosis.\nAbbreviation: LARS, low anterior resection syndrome.\nCross‐tabulation depicting the prevalence of LARS pre‐ and postsurgically in 121 women undergoing colorectal surgery for bowel endometriosis with nerve‐vessel sparing segmental resection and full thickness discoid resection.\nAbbreviations: FTDR, full thickness discoid resection; LARS, low anterior resection syndrome; NVSSR, nerve‐vessel sparing segmental resection.\nFigure  3  depicts changes in cumulative GIQLI scores during follow‐up in two groups. All patients who were available for follow‐up exhibited a significant increase in GIQLI scores ( P  < 0.001) with no differences between the NVSSR and FTDR group ( P  = 0.542). Finally, fertility outcomes of infertile patients are shown in Table  6 .\nSemiquantitative data on gastrointestinal function‐related quality of life (GIQLI) scores before and after surgery for colorectal endometriosis with either nerve‐vessel sparing segmental resection (NVSSR) or full thickness discoid resection.\nLong‐term outcomes of 121 eligible patients (52/121 presurgical infertile patients) following nerve‐vessel sparing segmental resection and full thickness discoid resection regarding fertility following colorectal deep endometriosis. All values in  n  (%)\nAbbreviations: DE, colorectal deep endometriosis; FTDR, full thickness discoid resection; NVSSR, nerve vessel sparing segmental resection.\nFisher's exact test.\nChi‐square test.\nIndependent samples  t ‐test.\nAs shown in Table  3 , pain scores significantly decreased following surgery ( P  ≤ 0.05, Figure  4A–C ). Furthermore, EHP‐30 presurgical total scores decreased significantly postsurgically, with significantly lower subscale ratings regarding pain, control and powerlessness, emotional wellbeing, social support and self‐image and sexual relationships (Figure  5 , Table  7 ). Analysis of variance with repeated measures for the EHP scales showed significant improvement over time ( P  < 0.001) and no significant interaction ( P  = 0.314). This means that there was significant improvement in all scales in both groups. Table  7  shows the changes in EHP separately by group, based on the results for the  t ‐tests with effect sizes.\n(A–C) Semi‐quantitative data on pain symptoms before and after surgery for colorectal endometriosis with either nerve‐vessel sparing segmental resection (NVSSR) or full thickness discoid resection.\nSemiquantitative data on total Endometriosis Health Profile‐30 (EHP‐30) scores before and after surgery for colorectal endometriosis with either nerve‐vessel sparing segmental resection (NVSSR) or full thickness discoid resection.\nSurgical outcomes regarding quality of life according to endometriosis health profile‐30 following nerve‐vessel sparing segmental resection and full thickness discoid resection for colorectal DE.\nNote : 0 indicates the best health status and 100, worst health status; mean ± SD, paired samples  t ‐tests,  d  = Cohen's  d : >0.20 = small, >0.50 = medium, >0.80 = large effect). Bold values indicate statistically significance.\nAbbreviations: DE, colorectal deep endometriosis; EHP‐30, endometriosis health profile; FTDR, full thickness discoid resection; LARS, low anterior resection syndrome; NVSSR, nerve‐vessel sparing segmental resection.\n\nThe present study demonstrates that NVSSR and FTDR for colorectal DE confers a reduction of GI function impairment reflected by a postsurgical decrease of LARS‐like symptoms, low rates of de novo LARS and increased GIQLI scores. Both techniques exerted a significant beneficial effect on pain reduction and increase in quality of life and exhibited low rates of severe complications. We observed an overall grade III complication rate of 4.3% without significant differences in the two cohorts. Anastomotic leakage and fistula development were observed respectively in 1.6% and 1.6% in patients undergoing NVSSR, which is low when compared with the published rates of these complications in women undergoing segmental resection. \n 13 \n  Interestingly, impairment of GI function as reflected by minor or major LARS‐like symptoms and low GIQLI scores, was observed with a high prevalence already present presurgically in both cohorts: NVSSR (27.6% minor LARS in 27/98 patients; 22.4% major LARS in 22/98 patients; GIQLI 95, 45–139, and FTDR (26.1% minor LARS in 6/23 patients; 17.4% major LARS in 4/23 patients; GIQLI 96 (42–126). This is in line with previous works observing the presence of major LARS‐like symptoms before surgery in women with colorectal DE in 18% \n 2 \n ,  \n 24 \n  to 45% of patients \n 25 \n  and low GIQLI scores between 87 \n 26 \n  and 92 \n 6 \n  points, possibly caused by the presence of colorectal DE and associated adhesions. \n 27\nWe were able to observe a relevant decrease in LARS scores and increase in GIQLI scores after a median follow‐up period of 40 months in both cohorts, with a prevalence of major LARS in only 4.1% and 0% and median GIQLI scores of 128 and 123 points in NVSSR and FTDR cohorts, respectively. Worsening of LARS symptoms was exclusively observed in NVSSR patients (5/98, 5.1%) with de novo LARS in 3/98 (3.0%) women. However, we did not observe significant differences in postsurgical GI function impairment reflected by LARS or GIQLI between NVSSR and FTDR cohorts. Although some authors do not support this observation, \n 28 \n  favoring conservative surgery over segmental resection in terms of digestive function outcomes, a retrospective analysis of women undergoing low NVSSR or FTDR did not observe clear benefits of conservative surgery regarding the prevalence of postsurgical LARS in these patients. \n 14 \n  In line with these findings, a recently published meta‐analysis of bowel function in patients following radical and conservative surgery for colorectal DE concluded that FTDR or partial thickness discoid resection may confer fewer events of frequent bowel movement and constipation but equal rates of defecation pain, anal incontinence, and minor and major LARS. \n 29\nTo date, only a few studies evaluated the changes of GI function parameters pre‐ and postsurgically. Riiskjaer et al. \n 25 \n  as well as Casas et al. \n 24 \n  were unable to observe changes in LARS‐like symptoms and postsurgical LARS in women following bowel resection for colorectal DE. Similarly, constipation reflected by Knowles–Eccersley–Scott symptom scores exceeding 10 points remained unchanged with increased overall GIQLI scores of <100 (inferior to 100 for patients suffering from GI diseases) to 111 (97–135) and 121 (99–128) following FTDR or segmental resection in the prospective randomized controlled trial by Roman et al. \n 6 \n  The results of the present work do not support the observation of unchanged LARS symptoms following colorectal surgery with decreased prevalence of LARS but increased GIQLI scores. Regarding overall changes in GIQLI, similar results were observed by Roman et al. \n 6 \n  The low rates of LARS and their postsurgical decline in our cohorts may indeed be due to a beneficial effect of the exclusion of all tissues lateral to the rectal wall which can be accomplished in both NVSSR and FTDR procedures. On the other hand, factors such as low height of anastomosis were observed to correlate with increased risk of LARS, supporting the influence of this factor. As a consequence, studies on the occurrence of LARS postsurgically should include this information to be comparable with other data. Future trials will need to support this hypothesis.\nFinally, overall postsurgical pregnancy and live birth rates were satisfiable in both cohorts in presurgically infertile women with 58% and 56% in NVSSR patients and 56% and 100% in FTDR patients, underlining a beneficial effect of surgery on natural and artificial conception rates.\nThe present study has some limitations. First and foremost, the loss to follow‐up rate reached 12.3%, and 12.9% of women declined study participation. We therefore we cannot fully exclude a selection bias‐associated effect, since we do not know the outcome data of these women. Secondly, although a randomized study design is generally more favored, when it comes to surgical treatment decisions for bowel endometriosis, factors such as lesion localization and the extent of colorectal infiltration take precedence. Our surgical practice to perform FTDR in lesions up to 3 cm in length and NVSSR in larger lesions may also not be generalizable, since other units mainly perform partial thickness/shaving procedures and try to avoid segmental resection approaches. Thus, the nature of the patient population and the specific surgical techniques required, precluded the possibility of implementing randomization. Thirdly, the sample size of conservative surgeries, ie FTDR, is low compared with segmental resection cases; this can be explained by our surgical policy regarding the selection of patients undergoing NVSSR and FTDR. To clearly show or exclude a possible difference regarding outcomes of one technique over the other with respect to anastomotic leakage of fistula rates, several 100 patients in each arm would be necessary to reach adequate power. Therefore, the observations of this work, ie no relevant differences in major complication rates between the two cohorts, should be interpreted with caution. Finally, we included patients with different extent and location of colorectal DE. Extensive low rectovaginal DE may exert different GI function impairment compared with sigmoidal DE.\nThe strength of this work is that we can exclude surgical bias, since all procedures were performed by one surgeon (GH) and his team. Secondly, we are able to provide a sufficient follow‐up interval and the evaluation of pre‐ and postsurgical parameters of GI function. Adequate studies providing this information are sparse.\n\nThe results of this work suggest a decrease of GI impairment and related symptoms when compared with presurgical values. In addition, worsening of GI impairment or de novo GI impairment is relatively uncommon. Our data do not clearly support the general notion that local excision with transanally applied stapling devices, ie FTDR techniques, are superior in terms of complication rates and GI function outcomes. We suggest that evaluation of presurgical GI function is relevant to interpret post‐interventional GI function parameters and scores adequately. Differences in segmental resection techniques such as omission of resection of autonomous nerves, fat and feeding vessels may also be beneficial for anastomotic healing and vegetative function in women who undergo this procedure. To support this, future prospective multicenter studies with larger patient numbers are clearly needed.\n\nProject development was performed by GH. Data collection was performed by GH, DP, AR, BD, AB and JM. Statistical analysis was conducted by BS and ED. The first draft of paper was written by GH, ED; all authors commented on the paper. All authors read and approved the final article.\n\nProfessional statistical analysis was supported by a fund of the OEGEO (Österreichische Gesellschaft für Endokrinologische Onkologie).\n\nThe authors have stated explicitly that there are no conflicts of interest in connection with this article.","source_license":"CC0","license_restricted":false}