{"paper_id":"f7803dc9-336b-42cc-b334-72e6afba454c","body_text":"Endometriosis is an inflammatory, estrogen-dependent, chronic disorder in\nfertile-aged women. Endometriosis is defined as the presence of endometrial glands\nand stroma outside the uterine cavity. Although endometriosis is considered as a\nbenign disease, it can cause severe chronic pain and infertility, and decrease the\nquality of life. 1  Pharmacological treatments are the standard treatment for\nendometriosis. 2 , 3 \nHowever, when deep infiltrating endometriosis (DIE) decreases the quality of life\nbecause of associated pain or due to dysfunction of the bowels, bladder or ovaries,\nthen surgical treatment is necessary. Indications for surgical management are\nfailure of medical management, the purpose of diagnosis, treatment of an adnexal\nmass or treatment of infertility. 4\nThe mini-invasive approach of laparoscopic or robotic-assisted laparoscopy is highly\nrecommended for endometriosis. 5  However, a disadvantage of surgery is that when removing DIE lesions,\ncomplications often occur affecting gastrointestinal, urinary or sexual functions.\nComplications after surgery of DIE include rectal fistula (0.3%–2%), bowel stenosis\n(2%) and bladder atony (4%–6%). 6 – 8  Therefore, the decision of\nsurgery with its risks, benefits and extension should be carefully considered and\ndiscussed with patients who have endometriosis.\nCurrently, even extensive radical operations of the bowels or urinary tract can be\nperformed mini-invasively. 9 , 10  A few studies compared laparoscopic or robotic-assisted\napproaches in the surgical management of endometriosis. 11 – 15  Robotic-assisted laparoscopic\nsurgery is associated with a longer operation time than laparoscopic\nsurgery, 12 , 16  but results are controversial. 11 , 14  The results of previous\nstudies regarding benefits of robotic-assisted laparoscopy over conventional\nlaparoscopy are somewhat heterogeneous. However, patients with features of a complex\npelvic situation, such as severe endometriosis, an increased body mass index or\nprior surgeries, might benefit from robotic-assisted surgery. 17\nIn our institution, robotic-assisted surgeries were initiated in 2016. This study\naimed to evaluate the results of mini-invasive surgery for DIE in a single tertiary\ninstitution. Specifically, we aimed to 1) compare outcomes after conventional or\nrobotic-assisted laparoscopic surgery in our institution and 2) evaluate the quality\nof life after surgery by a specific questionnaire.\n\nThis retrospective study investigated consecutive patients who had been operated on\nfor endometriosis-related pain between January 2014 and December 2017 in Kuopio\nUniversity Hospital. The Research Ethical Committee of Northern Savo approved the\nstudy protocol (1012/13.02.00/2018) and written informed consent was obtained from\nall patients.\nEndometriosis was diagnosed by laparoscopy or histologically in all patients. The\nstage of endometriosis was classified in accordance with revised American Society\nfor Reproductive Medicine classification. 18  Briefly, the stage of endometriosis is divided into the four stages of I\n(minimal), II (mild), III (moderate) and IV (severe). Data collected from medical\nfiles included prognostic, diagnostic and operative information, such as age, body\nmass index, operation date, preoperative symptoms, cancer antigen 125 (CA125) and\nhuman epididymis protein 4 (HE4) biomarkers, magnetic resonance imaging (MRI)\nfindings, previous operations due to endometriosis, Clavien–Dindo classification, 19  operation technique, operative areas, hormonal treatments and postoperative\ncontact with a clinic because of pain from endometriosis. The upper normal limit for\nCA125 levels is 35 kU/L and that for HE4 levels is 70 pmol/L in premenopausal women\nin our hospital laboratory.\nAll of the patients were also sent a questionnaire inquiring about their well-being\nin January 2019. This questionnaire included questions about pain postoperatively,\nif the operation caused any short- or long-term harm, alternative treatments they\nhad tried and their benefits, whether and how endometriosis was still affecting\ntheir daily lives and if the patients felt that the operation was beneficial and\ncaused some change in their quality of life. The numeric rating scale (NRS) from 0\nto 10 was used, where 0 indicates no pain and 10 the worst possible pain.\nIBM SPSS Statistics for Windows, version 27 (IBM Corp., Armonk, NY, USA) was used in\nstatistical analysis. Values are presented as mean ± standard deviation, unless\notherwise stated. The Kruskal–Wallis test followed by the Mann–Whitney test for\ncontinuous variables in multiple comparisons were used when appropriate. We used the\nchi-square test to analyze frequency tables. A p value of < 0.05 was considered\nsignificant.\n\nThe characteristics of the patients are shown in  Table 1 . All patients (n = 103)\nenrolled in this study experienced symptoms of pain. The most common symptoms\nwere pelvic pain, followed by dyspareunia and dyschezia. Two patients reported\nshoulder pinch. The symptoms were continuous in more than half of the patients,\nsymptoms occurred mainly during menstruation in approximately 40% and symptoms\nwere only experienced occasionally in the remaining 5% ( Table 1 ).\nClinicopathological characteristics of patients with endometriosis\n(n = 103)\nBMI, body mass index; MRI, magnetic resonance imaging; CA125, cancer\nantigen 125; HE4, human epididymis protein 4; IUD, intrauterine\ndevice; GnHR, gonadotrophin-releasing hormone.\nPreoperative hormonal treatments are shown in  Table 1 . Two thirds of the patients\nused combined oral contraceptives, approximately half were receiving\nprogesterone treatment, and 39% (n = 40) had a hormonal intrauterine device.\nAs shown in  Table 1 ,\n73 (71%) patients underwent MRI before surgery to evaluate the presence and\nlocation of possible DIE. Forty-two (58%) patients showed signs of retrocervical\nDIE in MRI. The majority (71%) of patients who had evidence of retrocervical DIE\nin preoperative MRI underwent retrocervical resection of DIE in surgery\n(p = 0.011 vs patients who did not undergo retrocervical resection).\nIn the majority (n = 43, 72%) of patients with CA125 level measurement,\npreoperative serum CA125 levels were above the normal limit. In contrast, serum\nHE4 levels were within the normal limit (n = 37).\nMost patients underwent conventional laparoscopy or robotic-assisted laparoscopy\n( Table 1 ). In\nthree patients, laparotomy was performed. The reasons for performing laparotomy\nwere poor lung dysfunction in one patient and complex adhesions in the abdominal\ncavity in two patients. No conversions to laparotomy were undertaken.\nAdditionally, five patients underwent vaginal hysterectomy because of\nendometriosis of the uterus, and in one patient, endometriotic tissue was\nremoved from a cesarean section scar.\nDetails of the operated areas are shown in  Table 2 . When we compared only\nconventional laparoscopy and robotic-assisted laparoscopy, significantly higher\nrates of parametrectomy (p = 0.036) and rectovaginal resections (p = 0.001) were\nperformed in robotic-assisted laparoscopy than in conventional laparoscopy.\nAdditionally, significantly more bowel operations were performed in\nrobotic-assisted laparoscopy than in conventional laparoscopy (p = 0.011). In\nparticular, the shaving technique was applied more frequently in\nrobotic-assisted laparoscopy than in conventional laparoscopy (p = 0.011) ( Figure 1 ).\nSurgical procedures and complications\nRobotic, robotic-assisted laparoscopy; BMI, body mass index; ns, not\nsignificant.\nPie charts showing the rates of bowel operations. Significantly more\nbowel operations were performed in robotic-assisted laparoscopy (a) than\nin conventional laparoscopy (b) (p = 0.011). Shaving was used\nsignificantly more often in robotic-assisted laparoscopy (a) than in\nconventional laparoscopy (b) (p = 0.011).\nThere were no significant differences in the rates of postoperative complications\nbetween conventional laparoscopy and robotic-assisted laparoscopy ( Table 2 ). The most\ncommon postoperative complications were urinary and genital infections,\nprolonged pain and short-term dysuria. Two patients had more severe\npostoperative complications (Clavien–Dindo IIIa and IIIb), with an abscess in\nthe pouch of Douglas in one patient and a rectovaginal fistula in one patient.\nOnly one intraoperative complication was observed, which was perforation of the\nrectum. This perforation was sutured immediately during surgery and no\npostoperative symptoms due to perforation were observed.\nAlmost half (44%, n = 45) of the patients returned the well-being questionnaire,\nwhich was sent to them after the operation. The median time between their\noperation and their answers to the questionnaire was 38 months (range: 14–61\nmonths). Detailed results of the questionnaire are shown in  Table 3 . The majority\n(n = 34, 76%) of the respondents had undergone conventional laparoscopy and nine\n(20%) had undergone robotic-assisted laparoscopy. Only one respondent had been\ntreated with laparotomy. Laparotomy was excluded from this assessment because of\nthe lack of answers from patients who had been treated with laparotomy.\nResults of the well-being questionnaire postoperatively\nRobotic, robotic-assisted laparoscopy.\nMost (91%) of the respondents felt that surgical treatment had relieved their\npain and 90% of the respondents thought that the operation had been beneficial.\nIn the laparoscopic and robotic-assisted groups, 88% of the respondents felt\nthat their quality of life had improved after surgery.\nTwo-thirds (62%) of the respondents who had laparoscopy and one-third of\nrespondents who had robotic-assisted laparoscopy reported that they were still\nexperiencing pain due to endometriosis less or more often than monthly. Patients\nwere asked in the questionnaire to score their pain by the NRS at the current\nmoment. The mean NRS value was 1.9 ± 1.2 at 1 year after surgery, 1.0 ± 0.8 at 2\nyears, 3.1 ± 1.0 at 3 years, 1.7 ± 1 at 4 years and 3.6 ± 0.8 at 5 years ( Figure 2 ). There were no\nsignificant changes in NRS values between time points or between laparoscopic or\nrobotic-assisted surgery.\nNumeric rating scale scores of postoperative pain after conventional\nlaparoscopy or robotic-assisted laparoscopy. The first 2 years of\nfollow-up included robotic-assisted laparoscopy and conventional\nlaparoscopy. After this time, only conventional laparoscopy was\nincluded\nApproximately half of the respondents described short-term adverse effects after\nthe operation ( Table\n3 ). Dysuria, pain, urinary and genital infections and catheterization\nwere the most commonly reported adverse effects. Moreover fever, drip leakage of\nurine, tingling feelings in the uterine area and less intense orgasms were\nreported in the questionnaires. The rate of short-term adverse effects was\nsimilar in patients who had conventional laparoscopy to those who had\nrobotic-assisted laparoscopy. Long-term dyschezia was reported by two\nrespondents and long-term dyspareunia by one respondent who had undergone\nconventional laparoscopy. Some respondents also described adhesion pain, pelvic\npain, menopausal symptoms and neuralgia in the scar area.\n\nWe found that robotic-assisted laparoscopy was a feasible method for resection of\nDIE, especially in the rectosigmoid area. Furthermore, pain and quality of life of\nthe patients were evaluated by asking them to fill in a questionnaire. Most of the\nresponders reported that surgery relieved their endometriosis-related pain and their\nquality of life had improved.\nWe report a single tertiary center experience of mini-invasive surgical treatment of\npainful endometriosis. Our patients represent a typical cohort of those who have\nendometriosis-related pain. 13  Pelvic pain, dyspareunia and dyschezia were the most common symptoms in our\npatients. Hormonal treatments were widely used preoperatively in most cases.\nAccording to the European Society of Urogenital Radiology, MRI is recommended as a\nsecond-line imaging technique preoperatively. 20  Transvaginal ultrasound and MRI achieve a similar accuracy in the diagnosis\nof DIE. 21  Currently, MRI imaging is a routine procedure before surgery to evaluate the\nlocation and extent of DIE being used in addition to transvaginal ultrasound in our\nhospital.\nCA125 levels are often elevated in patients with endometriosis. As expected, in our\ncohort, 72% of the patients had elevated CA125 levels. However, the benign nature of\nthese findings was confirmed because HE4 levels were normal in all of our patients\nand no ovarian carcinomas were diagnosed.\nKondo et al. reported complication rates in patients who underwent a rectal operation\nthat involved segmental resection, discoid excision or shaving. 22  They found that less complications were associated with shaving than with\nsegmental resection. In a large study by Mabrouk et al., the overall rate of\nshort-term postoperative complications was significantly higher in patients who\nunderwent segmental resection compared with those who underwent discoid excision or shaving. 8  Furthermore, segmental resection does not appear to achieve more long-lasting\nimprovement of symptoms compared with discoid resection or shaving. 23  Especially at the level of the low rectum, shaving is the recommended method\nto avoid injury of vascular and sympathetic and parasympathetic nerve\nbundles. 7 , 24 , 25  However, discoid excision or segmental resection is still an\noption to treat DIE at or above the sigmoid colon. 23  In our study, we preferred shaving in accordance with recommendations. In the\nrobotic-assisted laparoscopic group, shaving was used in 78% of patients who had\nundergone a bowel operation and no segmental resection was performed. The management\nof bowel endometriosis depends on the number of lesions, and their depth of\ninvasion, size and circumferential involvement. 17 , 26  Therefore, selection of the\nsurgical technique needs to be tailored to each individual patient.\nTo date, there are only limited data comparing management of rectosigmoid DIE between\nrobotic-assisted laparoscopy and conventional laparoscopy. The LAROSE trial, which\nwas a randomized, multicenter trial, compared the treatment of endometriosis between\nrobotic-assisted laparoscopy and conventional laparoscopy. 11  This trial was not able to detect any differences in perioperative outcomes\nor the operative time between the robotic-assisted procedure and laparoscopy.\nNonetheless, patients who required bowel resection were excluded from this trial.\nResults from other smaller mainly retrospective studies were heterogeneous. Some of\nthese studies reported longer operation times with robotic-assisted procedures than\nwith laparoscopy, while other studies found benefits from robotic-assisted\nsurgery. 12 , 14 , 16  Ercoli et al. showed that robotic-assisted laparoscopic\nnerve-sparing rectal nodulectomy appeared to be a feasible and safe approach in\ntreating isolated retrocervical–rectal DIE. 27  Recently, intravenous indocyanine green and near-infrared radiation imaging\nwere reported to have an additional benefit in rectosigmoid endometriosis in\nassessing the blood supply of the bowel after resection. 28  These techniques might also be helpful in separating endometrial nodules from\nhealthy tissue. However, intraoperative near-infrared radiation imaging can be used\nduring conventional or robotic-assisted laparoscopy. 29\nIn our study, there was no difference in the rate of complications between patients\nwho had robotic-assisted laparoscopy or conventional laparoscopy. However, two\npatients who had robotic-assisted laparoscopy had Clavien–Dindo grade III\npostoperative complications. This complication rate was acceptable because these\npatients had complicated DIE in the pelvis. Our results are also in line with a\nrecent pilot study that compared robotic-assisted and conventional laparoscopy in\ntreating colorectal endometriosis. 30  In our cohort, all of the laparotomies were performed when robotic-assisted\nsurgeries were not available in our institution. In the current study, no\nconversions to laparotomy were performed, which suggested the feasibility of using\nrobotics. However, a multidisciplinary robotic team is necessary to operate on\npatients with rectosigmoid or urinary tract DIE.\nApproximately 50% to 80% of patients with endometriosis consider a surgical treatment\nto be beneficial for endometriosis-related pain during the first 2 years after the\noperation. 31 , 32  However, after 2 to 5 years, 36% of surgically treated patients\nmight need to undergo a new operation 33  These results are in line with the present findings. In the present study,\nmost of the patients with endometriosis-related pain reported less pain and an\nimprovement in their quality of life after surgery. According to the NRS scores,\nduring the first 2 years after surgery, the patients’ pain symptoms were less\nintense, but subsequently, a trend towards higher NRS scores was observed. Notably,\nin this study, the first 2 years of follow-up included patients who had undergone\neither robotic-assisted or laparoscopic surgery, but the later evaluation included\nonly those who had been treated with conventional laparoscopic operations.\nThere are some limitations to this study. First, our study was retrospective and the\nnumber of patients was limited. There might have been some bias because the more\ncomplex cases were routinely operated on using robotic-assisted techniques after\n2016. These patients had a shorter follow-up time than patients who were operated on\nbefore this time. Second, our questionnaire of well-being has not been validated.\nThird, we had no information on the quality of life before the patients had surgery.\nFurthermore, pain was the only symptom that we evaluated.\nIn conclusion, the present study suggests that robotic-assisted laparoscopy is a\nfeasible method to resect DIE. Mini-invasive surgical treatment also improves the\nquality of life in the majority of patients suffering from endometriosis-related\npain. Further prospective investigations of mini-invasive treatment of patients with\nbowel endometriosis are warranted.","source_license":"CC0","license_restricted":false}