{"paper_id":"030c8956-a436-4bd7-a732-2eb2d719682f","body_text":"R E S E A R C H Open Access\nRisk factors for postoperative recurrence of\novarian endometriosis: long-term follow-up\nof 358 women\nXiao-Yan Li †, Xiao-Pei Chao †, Jin-Hua Leng * , Wen Zhang, Jun-Ji Zhang, Yi Dai, Jing-Hua Shi, Shuang-Zheng Jia,\nXiao-Xuan Xu, Si-Kai Chen and Yu-Shi Wu\nAbstract\nObjective: To explore the risk factors for the recurrence of endometrioma and the risk factors for the recurrence of\nendometriosis-related pain after long-term follow-up.\nMethods: This study retrospectively analyzed 358 women with endometriomas who had a minimum of 5-years\nfollow up after laparoscopic endometrioma excision, which was performed at Peking Union Medical College\nHospital from January 2009 to April 2013. All women were divided into recurrence group and nonrecurrence group.\nAnalysis was performed with regard to preoperative history, laboratory analysis, findings during surgery, and\nsymptoms during follow-up, including improvement and recurrence.\nResults: The cumulative incidence rates of recurrence from 5 to 10 years after surgery were 15.4, 16.8, 19.3, 22.5,\n22.5, and 22.5%, respectively. Significant differences were found between two groups in terms of age at surgery\n(RR: 0.764, 95% CI: 0.615 –0.949, p = 0.015), duration of dysmenorrhea (RR: 1.120, 95% CI: 1.054 –1.190, p < 0.001),\npresence of adenomyosis (RR: 1.629, 95% CI: 1.008 –2.630, p = 0.046), CA125 level (RR: 1.856, 95% CI: 1.072 –3.214,\np = 0.021) and severity of dysmenorrhea. The severity of dysmenorrhea (RR: 1.711, 95% CI: 1.175 –2.493, p = 0.005)\nand postoperative pregnancy (RR: 0.649, 95% CI: 0.460 –0.914, p = 0.013) were significantly correlated with\nendometrioma recurrence in the multivariate analysis. No significant associations were found between the\nrecurrence rate and gravida, parity, body mass index, infertility, leiomyoma presence, the size of ovarian\nendometrioma, the presence of deep infiltrating endome triosis, disease stage or postoperative medication.\nConclusions: The severity of dysmenorrhea and postoperative pr egnancy were independent risk factors for the\nrecurrence of ovarian endometriomas after surgery during the long-time follow up.\nKeywords: Ovarian endometriomas, Recurrence, Risk factors\nIntroduction\nEndometriosis is a chronic benign estrogen-dependent\ndisease. It is observed primarily in patients of reproductive\nage, and its prevalence in this population is estimated to\nbe 5–10%. Endometriosis is defined as the presence of ac-\ntive endometrial tissue outside the uterine cavity, usually\non the peritoneum of the minor pelvis, ovaries and fallo-\npian tubes and sometimes in extraperitoneal regions.\nBased on the locations of the lesions, the disease is\nclassified as peritoneal, ovarian or deep infiltrating endo-\nmetriosis [1, 2]. Endometriosis causes impaired quality of\nlife (QoL) for women of reproductive age, and has malig-\nnant clinical manifestation despite being a benign disease.\nOvarian endometriosis is the most common type, ac-\ncounting 17% and 44% of all endometriosis. Laparoscopic\nconservative surgery has been considered the gold stand-\nard treatment for ovarian endometrioma [ 3, 4]. However,\nsurgery may affect ovarian reserve function, and thus,\nsurgery, especially repeat surgery, is not recommended for\novarian endometrioma recurrence [ 5–7]. The recurrence\nrate following surgical intervention remains high, even for\n© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0\nInternational License ( http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and\nreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to\nthe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver\n(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.\n* Correspondence: lengjh_pumch@163.com\n†Xiao-Yan Li and Xiao-Pei Chao contributed equally to this work.\nDepartment of Obstetrics and Gynecology, Peking Union Medical College\nHospital, Peking Union Medical College & Chinese Academy of Medical\nScience, Shuaifuyuan No. 1, Dongcheng District, Beijing 100730, China\nLi et al. Journal of Ovarian Research           (2019) 12:79 \nhttps://doi.org/10.1186/s13048-019-0552-y\n\nthose who receive postoperative medical therapy.\nTherefore, endometrioma recurrence is one of the most\nimportant unresolved problems in the management of\nendometriosis. The recurrence rate of ovarian endome-\ntrioma after conservative laparoscopic surgery has been\nreported to be 29 –56% at 2 years and 43% at 5 years.\nWhen postoperative medical treatment was introduced,\nthe endometrioma recurrence rate substantially declined\nto 3–11% at 2 years and 6% at 5 years [ 8, 9]. A pooled ana-\nlysis of 23 studies estimated recurrence rates of 21.5% at 2\nyears and 40.0–50.0% at 5 years after primary surgery [ 10].\nUnfortunately, few studies have analyzed the determinants\nof the long-term recurrence rate for endometrioma be-\nyond 5 years after surgery. Therefore, the aim of this study\nwas to explore the risk factors for endometrioma\nrecurrence and the risk factors for the recurrence of\nendometriosis-related pain after long-term follow up.\nMaterials and methods\nEthical approval\nThis study was approved by the Ethics Committee of\nPeking Union Medical College Hospital. Written in-\nformed consent was obtained from all participants.\nPatient population\nWe collected 358 women with endometriomas who had a\nminimum of 5 years of postoperative follow-up after under-\ngoing a laparoscopic cystectomy at Peking Union Medical\nCollege Hospital from January 2009 to April 2013.\nThe inclusion criteria were as follows: (1) the diagno-\nsis was confirmed by pathologists; (2) ultrasonography\nwas conducted to determine endometrioma recurrence\nat least 6 months after surgery; and (3) patients were\nobserved without postoperative medications or were\ntreated with postoperative gonadotropin-releasing hor-\nmone agonist (GnRHa) injections for 3 –6 cycles, with\nor without a Mirena levonorgestrel-releasing intrauter-\nine device (LNG-IUD). (4) The duration of follow-up\nwas at least 5 years. The exclusion criteria were as fol-\nlows: (1) age < 20 or > 45 years; (2) having underwent\nbilateral oophorectomy or hysterectomy; and (3) intra-\noperative conversion to laparotomy.\nMethods\nMedical charts were reviewed to collect data on age at sur-\ngery; body mass index; presence of adenomyosis or leio-\nmyoma; surgical history; symptoms of dysmenorrhea;\nparity; size of endometriom a; location of cysts; serum\nCA125 levels; operative time; intraoperative blood loss;\nAmerican Society for Reproductive Medicine (ASRM) stage\n[11]; postoperative medications ; postoperative pregnancy;\na n dr e c u r r e n c et i m e .T h es i z eo ft h eo v a r i a ne n d o m e t r i o m a\nwas defined as the largest diameter of cysts. Obliteration of\nthe pouch of Douglas was defined when there was any ad-\nhesion in the pouch.\nAll surgeries were performed by one expert laparos-\ncopist. In all cases, the aim of the surgical procedures\nwas to remove all visible implants of endometriosis,\ncomplete the opening of the pouch in the operation\nand perform lysis of adhesion. Endometriosis was\nstaged according to the classification of the ASRM. The\npresence, localization, and extent of typical powder-\nburn and subtle lesions, adhesions, and deep infiltrating\nimplants were recorded. Both ovaries were then com-\npletely mobilized. A sharp cortical incision was made\ninto the cyst, and a cleavage plane was identified. A\ntraumatic forceps and counter traction were used to\nstrip the endometrioma from the ovarian parenchyma.\nHemostasis was achieved by the selective application of\nbipolar coagulation. After the endometriomas were re-\nmoved, all remaining visible endometriotic lesions were\nexcised or fulgurated. Anatomical restoration was then\nachieved. Specimens underwent thorough histological\nanalysis. When complicatio ns occurred, laparoscopy\nwas converted into laparotomy. Laparotomic conver-\nsions were not included in the study.\nIn our center, all women are followed up according to\nan internal protocol. A standard gynecological examin-\nation and a transvaginal ultrasound are conducted before\nsurgery; at 3, 6, and 12 months after surgery; and then\nyearly after surgery. Menstrual-reproductive factors and\npain symptoms of pain are also evaluated. During the\nfollow-up visits, patients were asked whether dysmenor-\nrhea, pelvic pain or dyspareunia occurred during the\nfollow-up period. The time of their first appearance and\nthe intensity of the pain symptoms after laparoscopy\nwere also documented. Pain was rated on the basis of a\n10-cm visual analog scale (VAS), and the intensity was\ndivided into none (0), mild (1 –4), moderate (5 –7), or se-\nvere (8 –10). The presence of pain before surgery and\npain recurrence/occurrence were defined as the recur-\nrence of pain after a period of at least 3 months of relief\nafter surgery. Threshold points defining different sever-\nities of pain were chosen based on a previous correlation\nanalysis [ 12]. Endometrioma recurrence was determined\nusing ultrasonography. Recurrent endometrioma was\ndefined as the presence of a persistent ovarian cyst that\nhad a thin wall (with a diameter of at least 2 cm), regular\nmargins, a homogenous low echogenic fluid content\nwith scattered internal echoes and did not resolve after\nseveral successive menstrual cycles. An improvement in\npain was defined as a decrease in the VAS score for at\nleast two points. Persistent pain was defined as no\nimprovement or an increase in the VAS score of less\nthan 2 points after surgery. Recurrence was identified\nwhen the same VAS score as that before surgery was\nnoted at follow-up.\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 2 of 10\n\nStatistical analysis\nStatistical analysis was performed using the Statistics Pack-\nage for Social Sciences Version 22.0 (SPSS Inc., Chicago, IL,\nUSA). Quantitative variables were compared using t-tests\nand ANOVA with Bonferroni’s correction for multiple test-\ning. Fisher’s exact or chi-square tests were used to analyze\nqualitative variables. Potential risk factors ( p <0 . 2 ) w e r e\nidentified using univariate analysis and Cox ’sm u l t i v a r i a t e\nproportional hazard analysis. The hazard ratio (HR) and\n95% confidence interval (CI) were calculated as a measure\nof the risk of recurrence in each study. Significance was de-\nfined as p < 0.05.\nResults\nFor the total of 358 women, the mean age and BMI were\n33.2 ± 5.4 years and 21.2 ± 2.6 kg/m2. The median gravida,\nparity, and duration of dysmenorrhea were 1(0–10), 0(0–2)\nand 24 months (0 –360 months). The median CA125 level\nwas 54 U/ml (8.4 –754 U/ml). The mean operation time,\nbleeding volume and duration of follow-up were 66.4 ± 22.3\nmin, 5.4 ± 73.5 ml, and 84.2 ± 14.6 months. Before surgery,\ndysmenorrhea was absent in 23.7% (85/358) of cases, mild\nin 14% (50/358), moderate in 21.8% (78/358), and severe in\n40.5% (145/358). The infertility rate before surgery was\n18.4% (66/358). 17.9% (64/358) cases had concurrent leio-\nmyoma, 53.1% (190/358) had concurrent deep infiltrating\nendometriosis, and 39.9% (143/358) had concurrent\nadenomyosis. Abdominal and vaginal ultrasound revealed\nthe presence of bilateral ovarian endometriomas in 46.1%\n(165/358) of cases and unilateral ovarian endometriomas in\n53.9% (193/358) of cases (31.0%,111 left, 22.9%, 82 right).\nAt laparoscopy, ovarian endometrioma was confirmed\nwith a mean size of 5.4 ± 2.1 cm on the left side and\n5.3 ± 1.9 cm on the right. Leiomyomas were present in\n18% (64/358) of cases, deep infiltrating endometriosis\n53.1% (190/358), and aden omyosis 39.9% (143/358).\nOne hundred two patients had partial and 162 patients\nhad total cul-de-sac obliteration. According to the\nrASRM classification, 134 (37.4%) and 213 (59.5%)\ncases were stage III and stage IV. Postoperative therapy\nwas chosen on the basis of individual characteristics\nand careful counseling with the patient: 12 (3.4%)\nwomen had no therapy; 12 (3.4%) oral contraceptives\n(drospirenone and ethinylestradiol tablets for at least 1\nyear); 14 (3.9%) Mirena for least 1 year, 242 (67.6%) re-\nceived GnRHa for 3 –6 months, and 78 (21.8%) had\nG n R H af o l l o w e db yM i r e n a .\nAfter more than 5-years follow-up (median 83, 60 –\n120), we observed 9.4% (34/358) had pain recurrence,\n6.4% (23/358) cyst recurrence and 3% (11/358) for\nboth. The cumulative rate of endometrioma and/or\nendometriosis-related pain recurrence over 5 years\nwas 19.0%. Sixty-eight women were in the recurrence\ngroup and 290 cases in the nonrecurrence group.\nDysmenorrhea recurred in 34 women (68%), endome-\ntrioma recurred in 23 women (33.8%), and both\nrecurred in 11 patients (16.2%). The clinical charac-\nteristics of the different groups of endometriomas are\ns h o w ni nT a b l e1.\nUnivariate analysis and Cox ’s multivariate proportional\nhazard analyses were performed. Differences were found\nbetween two groups in terms of age at surgery (RR: 0.764,\n95% CI: 0.615–0.949, p = 0.015), duration of dysmenorrhea\n(RR: 1.120, 95% CI 1.054–1.190, p < 0.001), presence of ade-\nnomyosis (RR: 1.629, 95% CI: 1.008 –2.630, p = 0.046),\nCA125 level (RR: 1.856, 95% CI: 1.072 –3.214, p =0 . 0 2 1 )\nand extent of dysmenorrhea. The risk ratio of mild dysmen-\norrhea was 4.506 (95% CI: 1.413–14.368, p = 0.011), moder-\nate was 2.451 (95% CI: 1.433 –4.193, p = 0.001), and severe\ndysmenorrhea was 1.771 (95% CI: 1.254–2.502, p = 0.001).\nNo significant association was found between endome-\ntrioma and/or endometriosis-related pain recurrence rate\nand gravida, parity, BMI, infe rtility, leiomyoma presence,\novarian endometrioma size, deep infiltrating endometriosis\npresence, obliteration of thepouch of Douglas, disease stage\nand postoperative medication (Table 2). Only the extent of\ndysmenorrhea (RR: 1.711, 95% CI: 1.175 –2.493, p =0 . 0 0 5 )\nand postoperative pregnancy (RR: 0.649, 95% CI: 0.460 –\n0.914, p = 0.013) were significantly correlated with endome-\ntrioma and/or endometriosis-related pain recurrence in the\nmultivariate analysis (Table2).\nDiscussion\nThere are a variety of factors, both clinically and\nsurgically, that might be related to the risk of endome-\ntrioma and/or endometriosis-related pain recurrence.\nAccumulating evidence suggests that immune cells,\nadhesion molecules, extracellular matrix metallopro-\nteinase and pro-inflammatory cytokines activate/alter\nperitoneal microenvironment, creating the conditions\nfor differentiation, adhesion, proliferation and survival\nof ectopic endometrial cells [ 13–15]. In a study by\nTobiume et al., the rAFS score was an independent\nfactor associated with recurrence [ 16]. Chon et al. re-\nported that dysmenorrhea and ovarian cyst separations\nsignificantly affected the postoperative recurrence rate\n[17]. Selcuk et al. reported that the depth of penetration\nof the endometrial tissue into the ovarian cyst wall was\nan independent risk factor for recurrence [ 18]. Guzel et\nal. reported that the CA125 level, ovarian cyst size, and\nhistory of pelvic surgery affected the recurrence rate\n[19]. However, it is difficult to compare the results of\nthe above studies due to differences in the study popu-\nlation, the follow-up duration, and the definition of\nendometrioma and/or endometriosis-related pain re-\ncurrence. Notably, most of the above studies reported\nthe endometrioma and/or endometriosis-related pain\nrecurrence rate within 5 years after initial surgery. We\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 3 of 10\n\nTable 1 Baseline characteristics of the endometrioma and/or endometriosis-related pain recurrence and nonrecurrence groups\nCharacteristics Recurrence group ( n = 68) Nonrecurrence group ( n = 290) χ2 p value\nAge, y 31.8 ± 5.0 32.9 ± 5.2 1.457 0.146\nGravida 0 (0 –4) 1 (0 –10) 1.207 0.228\nParity 0 (0 –1) 0 (0 –2) 2.385 0.018\nBMI, kg/m2 20.9 ± 2.5 21.2 ± 2.6 0.814 0.416\nrAFS stage 3.093 0.377\nStage I 0 (0%) 1 (0.3%)\nStage II 3 (4.4%) 7 (2.4%)\nStage III 20 (29.4%) 113 (39.2%)\nStage IV 45 (66.2%) 167 (58.0%)\nLargest-diameter endometrioma, cm\nLeft 5.6 ± 1.9 5.4 ± 2.2 −0.707 0.481\nRight 5.7 ± 2.0 5.2 ± 1.8 −1.631 0.104\nCA125, U/ml 99.23 ± 89.72 100.37 ± 228.61 0.038 0.970\nDysmenorrhea, VAS 7 (0 –10) 5 (0 –10) −3.018 0.003\nExtent of dysmenorrhea 15.420 0.001\nNone 4 (5.9%) 81 (27.9%)\nMild 10 (14.7%) 40 (13.8%)\nModerate 20 (29.4%) 58 (20.0%)\nSevere 34 (50.0%) 111 (38.3%)\nDuration of dysmenorrhea, m 84 (0 –360) 24 (0 –348) −3.514 0.001\nEndometrioma side 3.082 0.214\nLeft 16 (23.5%) 95 (32.8%)\nRight 20 (29.4%) 62 (21.4%)\nBilateral 32 (47.1%) 133 (45.9%)\nObliteration of cul-de-sac 3.307 0.191\nAbsent 13 (19.1%) 81 (27.9%)\nPartial 18 (26.5%) 84 (29.0%)\nComplete 37 (54.4%) 125 (43.1%)\nLeiomyoma 18 (26.5%) 46 (16.0%) 5.057 0.044 (0.044)\nDeep infiltrating endometriosis 41 (63%) 149 (51.4%) 1.758 0.224\nAdenomyosis 34 (50.0%) 108 (37.4%) 4.059 0.044 (0.044)\nPostoperative dysmenorrhea relief 55 (87.3%) 8 (90.0%) 1.770 0.413\nInfertility 2.408 0.300\nPrimary 11 (16.2%) 32 (11.0%)\nSecondary 6 (8.8%) 17 (5.9%)\nOperation 0.705 0.590\nCystectomy 65 (95.6%) 21 (72%).\nSalpingo-oophorectomy 3 (4.4%) 269 (92.8%)\nPostoperative pregnancy 24 (35.3%) 113 (39.0%) 0.611 0.894\nPostoperative medication 11.066 0.026\nNone 0 (0%) 12 (4.1%)\nOCP 1 (1.5%) 11 (3.8%)\nMirena 1 (1.5%) 13 (4.5%)\nGnRHa with LNG-IUD 9 (13.2%) 69 (23.8%)\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 4 of 10\n\nhave little information about the recurrence rate be-\nyond 5 years after surgery.\nThis study found that the extent of dysmenorrhea\nand postoperative pregnancy were risk factors accord-\ning to the multivariate analysis. To avoid multiple\ncollinearity, the presence of dysmenorrhea was not\nselected for the multivariate analysis. Several studies\nidentified dysmenorrhea as a possible risk factor for\nendometrioma recurrence [ 16, 20]. Our study also\nshowed an association between the presence of dys-\nmenorrhea after surgery and endometrioma and/or\nendometriosis-related pain recurrence in a univariate\nanalysis (Fig. 1). Mechanisms of dysmenorrhea have\nnot been uniformly determined; however, direct and\nindirect effects of focal bleeding from endometriotic\nimplants, actions of inflammatory cytokines in the\nperitoneal cavity, and irritation or direct infiltration of\nnerves in the pelvic floor are causes [ 21]. It is hy-\npothesized that cyclic recurrent microbleeding within\nendometriotic lesions with consequent inflammation\nmay be the cause of severe dysmenorrhea among\nwomen with endometriosis [ 22]. Therefore, the longer\nthe disease duration is, the deeper the lesions will\ninfiltrate. In our series, women who suffered dysmen-\norrhea over 6 months had higher endometrioma and/\nor endometriosis-related pain recurrence rate (Fig. 2).\nOver time, most patients with endometriomas had a\nhigher rASRM score and were categorized as having\nadvanced-stage endometriosis. In fact, most of\npatients (96.9%) in this study demonstrated stage III\nor IV endometriosis. Adhesions may also cause deep\npelvic pain associated with recurrent endometriosis,\nand postoperative dysmenorrhea also suggests\nendometrioma recurrence [ 23]. Because of extensive\nadhesions and inflammation, incomplete resection can\noccur in advanced stages of endometriosis. Incomplete\nsurgical removal of endometriomas located on an\nendometriotic lesion or adhesion only decreases the\nseverity of symptoms because only the removal of vi-\nsualized cystic lesions or separation of adhesions is\nperformed; Although peritoneal superficial lesions and\novarian endometriomas represent the majority of\nendometriotic implants within the pelvis, deep infil-\ntrating endometriosis and extrapelvic endometriosis\nare the most challenging conditions to face off.\nDespite sometimes medical therapy is enough to\nreduce symptoms and, in a large number of patients\na complete eradication, with nerve-sparing and vascu-\nlar sparing approach is needed to restore the normal\npelvic anatomy and its functions [ 24, 25]. In our\nstudy, all remaining visible endometriotic lesions were\nexcised or fulgurated after the endometriomas were\nremoved. Anatomical restoration was then achieved.\nHowever, surgery does not address the underlying\nmechanisms that are active and driving disease in the\npelvic cavity. This naturally leads to recurrence of\nendometriosis in various forms after a certain period\nof time.\nThe presence of adenomyosis is another risk factor for\nendometrioma and/or endometriosis-related pain recur-\nrence. It is well known that there is an overlap in the\npathogenesis of endometriosis and adenomyosis [ 22].\nDior et al. found that sonographic features of adenomyo-\nsis may be included as a component of the clinical as-\nsessment when attempting to predict the presence of\nsevere endometriosis [ 26]. We observed a statistically\nsignificant correlation between pelvic pain and the pres-\nence of adenomyosis (Fig. 3), in agreement with the pre-\nviously reported results [ 27, 28]. Dysmenorrhea is a risk\nfactor in the deep adenomyotic process with a high\ndensity of endometrial glands in the myometrium [ 29].\nPerello et al. observed a clear trend associating of the\npresence of dysmenorrhea with adenomyosis, without\nstatistical significance. Among the patients reporting\nthese symptoms, 95.5% described the intensity as severe\n[30]. Our study showed that the concurrence rate of\nadenomyosis was up to 50.0% in the recurrence group.\nUnresolved adenomyosis may consequently lead to pain\nrecurrences after pelvic surgery. Therefore, correct iden-\ntification of coexisting pathological conditions for DIE\nand adenomyosis is necessary for the development of ef-\nfective surgical protocols [ 30].\nIt was reported that age at surgery is associated with\nthe postsurgical endometrioma recurrence rate, and the\nreason is unclear. We postulate that endometriosis is a\nhormone dependent disease. Thus, the higher circulation\nestrogen levels in younger women may produce a more\naggressive form of endometriosis; therefore, the younger,\nthe more likely to recur [ 18]. Seo et al. [ 31] reported\nthat, at an average follow-up time of 29 months with no\npostoperative medication, the cumulative endometrioma\nrecurrence rates were 43.3% for patients aged 20 –29\nTable 1 Baseline characteristics of the endometrioma and/or endometriosis-related pain recurrence and nonrecurrence groups\n(Continued)\nCharacteristics Recurrence group ( n = 68) Nonrecurrence group ( n = 290) χ2 p value\nGnRHa without LNG-IUD 57 (83.8%) 185 (63.8%)\nDuration of follow-up 83 (60 –120) 85 (60 –116) −0.459 0.646\nAbbreviations: BMI Body mass index, CA-125 Cancer antigen 125, cm centimeter, GnRHa Gonadotropin-releasing hormone agonist, LNG-IUD Levonorgestrel-\nreleasing intrauterine device, m month, OCP Oral contraceptive pills, VAS Visual analogue score\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 5 of 10\n\nyears, 22.5% for 30 –39 years, and 10.2% for 40 –45 years.\nIn our study, the cumulative endometrioma and/or\nendometriosis-related pain recurrence rates were 24.2%\nfor patients aged 20 –30 years, 17.7% for 31 –40 years,\nand 7.9% for 41 –45 years, which was consistent with the\nliterature. Our study also found that 33.5 years is a cutoff\nvalue in the ROC age analysis (Fig. 4a). Patients who are\nyounger than 33.5 years may have a higher risk for\nTable 2 Univariate and multivariate analysis of risk factors in the endometrioma and/or endometriosis-related pain recurrence and\nnonrecurrence groups\nFactor Univariate analysis Multivariate analysis\nRelative risk 95%CI P value Relative risk 95%CI P value\nAge at surgery (per 5 years) 0.764 0.615 –0.949 0.015 0.738 0.520 –1.048 0.090\nGravida\n0 1.000\n1 0.673 0.370 –1.224 0.194\n2 (and above) 0.757 0.557 –1.029 0.075\nParity\n0 1.000\n1 0.478 0.261 –0.876 0.017 0.002 0.000 –1.222 0.974\n2 0.216 0.011 –4.254 0.314\nBMI 0.820 0.478 –1.407 0.470\nInfertility 1.527 0.882 –2.644 0.131\nExtent of dysmenorrhea 1.711 1.175 –2.493 0.005\nNone 1.000\nMild 4.506 1.413 –14.368 0.011\nModerate 2.451 1.433 –4.193 0.001\nSevere 1.771 1.254 –2.502 0.001\nDuration of dysmenorrhea (per year) 1.120 1.054 –1.190 < 0.001 1.026 0.944 –1.114 0.548\nExtent of dysmenorrhea 2.534 1.408 –4.563 0.002\nVAS (0-4)\nVAS (5-10)\nDyspareunia 1.734 1.049 –2.866 0.032 1.081 0.511 –2.285 0.839\nChronic pelvic pain 0.959 0.503 –1.828 0.898\nDyschezia 0.735 0.268 –2.018 0.550\nPresence of leiomyoma 1.661 0.969 –2.848 0.065 1.713 0.825 –3.555 0.149\nPresence of adenomyosis 1.629 1.008 –2.630 0.046 1.446 0.719 –2.908 0.301\nPresence of deep infiltrating endometriosis 1.380 0.849 –2.243 0.194 1.118 0.582 –2.149 0.737\nCA125 level (45 U/ml) 1.856 1.072 –3.214 0.021 1.020 0.965 –1.078 0.491\nrARSM stage 1.204 0.774 –1.873 0.409\nBilateral involvement 1.046 0.650 –1.684 0.853\nLarge dimeter (per 10 mm) 1.016 0.944 –1.093 0.674\nPostoperative medication\nNone 1.000\nGnRHa 21.818 0.125 –3810.382 0.242\nOCP 4.057 0.019 –856.392 0.610\nMirena 2.741 0.048 –155.496 0.624\nGnRHa+Mirena 1.991 0.375 –9.730 0.436\nPostoperative pregnancy 0.938 0.655 –1.344 0.728 0.649 0.460 –0.914 0.013\nAbbreviations: BMI Body mass index, CA-125 Cancer antigen 125, GnRHa Gonadotropin-releasing hormone agonist, OCP Oral contraceptive pills, VAS Visual\nanalogue score\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 6 of 10\n\nrecurrence (Fig. 4b). Our previous study showed that\nlaparoscopic cystectomy for endometrioma with an age\ngreater than 35 years may decrease the remaining ovar-\nian reserve [ 32]. Therefore, a decrease in the rate of\nendometrioma and/or endometriosis-related pain recur-\nrence also results in a reduction in ovarian reserve func-\ntion. 33 to 35 years of age is a “dilemma window ”, and\nwe should pay more attention to patients in this age\ngroup when performing surgery.\nPreventive use of medications after operation is rec-\nommended for patients with a high risk of recurrence\n[33–36]. Many studies have investigated factors deter-\nmining the recurrence of endometrioma and pain\nafter surgery [ 16, 19, 20]. The mechanism of oral\ncontraceptive pills (OCPs) r educing ovarian endome-\ntrioma recurrence is unclear . OCP increases apoptosis\nand decreases cell proliferati on in eutopic endometria,\nwhich can decrease both recurrence from small\nFig. 1 Kaplan-Meier curves presenting the cumulative rate of recurrence according to the severity of dysmenorrhea. There were significant\ndifferences between the four groups according to the log-rank test analysis ( χ2 = 11.487, p = 0.001)\nFig. 2 a The ROC analysis of duration of dysmenorrhea in recurrent endometrioma patients. Area under curve: 0.689, Cut-off value: 5.5 months;\nb Kaplan-Meier curves presenting the cumulative rate of recurrence according to duration of dysmenorrhea (< 6 months or ≥ 6 months). There\nwere significant differences between the two groups according to the log-rank test analysis ( χ2 = 22.352, p < 0.001)\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 7 of 10\n\nendometriotic foci not seen at surgery and de novo\ndisease development [ 33]. Some reports suggest that\novarian endometrioma can de velop from ovarian folli-\ncles or the corpus luteum and that consequent inhib-\nition of ovulation may decrease the risk of\nendometrioma development [ 36–38]. Regardless of\nthe mechanism, the present and previous studies sug-\ngest that postoperative medical treatment is known to\ndelay but not completely prevent recurrence. Vercel-\nlini et al. [ 39]r e p o r t e dt h a tp o s t o p e r a t i v eu s eo f\nGnRHa could only prolong the recurrence interval\nbut could not improve the overall recurrence rate.\nThere is no consensus regarding whether the LNG-\nIUD could reduce the endometrioma recurrence rate\neither [ 40]. Notably, Jee et al. [ 41]r e p o r t e dt h a ta l -\nthough postoperative GnRH agonist treatment does\nFig. 4 a The ROC analysis of age at surgery in recurrent endometrioma patients. Area under curve: 0.413, Cut-off value: 33.5 yrs.; b Kaplan-Meier\ncurves presenting the cumulative rate of recurrence according to age at surgery (< 33 yrs. or ≥ 33 yrs). There were significant differences between\nthe two groups according to the log-rank test analysis ( χ2 = 22.352, p < 0.001)\nFig. 3 Kaplan-Meier curves presenting the cumulative rate of recurrence according to the presence of adenomyosis. There were significant\ndifferences between the two groups according to the log-rank test analysis ( χ2 = 4.113, p = 0.043)\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 8 of 10\n\nnot reduce objective disease recurrence in stage III/IV dis-\nease, GnRHa delays the time of recurrence, as indicated by\nVercellini et al. [39]. In our study, we also failed to observe\na benefit for postoperative medication in preventing endo-\nmetrioma and/or endometriosis-related pain recurrence.\nPatients with advanced-stage endometriosis may prefer to\ntake postoperative medication. This may partially explain\nwhy a statistically significant difference was not reached in\nterms of postoperative medication between patients with\nand without recurrence. We found that the estimated cu-\nmulative endometrioma and/or endometriosis-related pain\nrecurrence rates at 1 to 10 years after surgery were 2.2, 5.3,\n9.2, 12.0, 15.4, 16.8, 19.3, 22.5, 22.5, and 22.5%, respectively.\nThese relatively low endometrioma and/or endometriosis-\nrelated pain recurrence rates may be due to most patients\nhaving postoperative medication applied in our study. The\nrecurrent rate increased yearly until 8 years after surgery.\nMost ovarian endometriomas are composed of an\nextraovarian pseudocystic s tructure with no cystic wall\nbut are surrounded by fibrosis with underlying ovarian\ncortical follicles [ 42]. Therefore, the inner surface of an\nendometrioma is lined by endometriosis with variable\npenetration into the surrounding fibrosis. The mean\ncyst wall thickness varied between 1.2 and 1.6 mm.\nEndometriosis tissue covers the inner aspect of the cyst\nfor approximately 60% of its surface with a mean depth\nof penetration of 0.6 mm [ 43]. With the aging of the\nendometrioma, infiltration and invasion of the ovarian\ninterstitial, resulting in the incomplete removal of\nendometriomas [ 42]. Concordantly, the univariate ana-\nlysis in our study shows that the duration of dysmenor-\nrhea is significantly associated with endometrioma and/\nor endometriosis-related pain recurrence (Fig. 2).\nIt is clear that the prevalence of recurrent endometrioma\nvaries according to whether there is a successful pregnancy\nafter surgical treatment of the endometrioma [44, 45]. Koga\net al. demonstrated that patients with postoperative preg-\nnancies had much lower rates of recurrence, which indi-\ncates that a subsequent pregnancy may have a protective\neffect on endometrioma recurrence [ 46]. In our study,\npostoperative pregnancies also prevented endometrioma re-\ncurrence (OR: 0.649, 95%CI: 0.460–0.914, p = 0.013).\nThe merits of our study were the long time follow-up of\nmore than 5 years, all with detailed record clinical data,\nthe same experienced surgeon, and the large sample size.\nHowever, there are some limitations. This was a retro-\nspective study, and it may contain biases with regard to\npatient characteristics; women with severe forms of endo-\nmetriosis may prefer to receive preoperative or postopera-\ntive medical treatment, and this is a single center study.\nRecurrent endometrioma was defined by ultrasound as\nthe presence of a persistent ovarian cyst and did not re-\nsolve after several successive menstrual cycles. It depends\nmainly on the skill and experience of radiologists. These\nlimitations may have resulted in an under or over estima-\ntion of the associations in our study.\nIn conclusion, we conducted a long-term follow-up\nstudy for more than 5 years. The rate of endometrioma\nand/or endometriosis-related pain recurrence increased\nyearly in the first 8 years. Our research revealed that the\nextent of dysmenorrhea and postoperative pregnancy are\nindependent risk factors for endometrioma and/or\nendometriosis-related pain recurrence. For patients who\nhave serious dysmenorrhea, we should pay more attention\nto their risk for endometrioma and/or endometriosis-\nrelated pain recurrence and apply individual management\nto achieve better efficacy.\nAcknowledgments\nWe appreciate the stuff at Peking Union Medical College Hospital for their\ndiligent clinical work and precise data recording about the cases we\nreported in this article.\nAuthors’contributions\nXY-L and JH-L developed the idea for the project. The study was designed\nby XY-L and XP-C, and JJ-Z, Y-D, JH-S, SZ-J, and WZ performed the data\nanalysis and takes full responsibility for the integrity of the data. XX-X, YS-W,\nWZ and XP-C drafted the manuscript with inputs and critical discussion from\nJH-L and XY-L. The final version has been approved by all authors.\nFunding\nOur work was supported by the National Key R&D Program of China (No.\n2017YFC1001200), National Natural Science Foundation of China (No. 81501237),\nand 2016 PUMCH Science Fund for Junior Faculty (No. PUMCH-2016-2.2).\nAvailability of data and materials\nThe dataset supporting the conclusions of this article is included within the\narticle and its additional files.\nEthics approval and consent to participate\nThis study was approved by the Ethics Committee of Peking Union Medical\nCollege Hospital.\nConsent for publication\nNot applicable.\nCompeting interests\nThe authors declare that they have no competing interests.\nReceived: 11 April 2019 Accepted: 13 August 2019\nReferences\n1. Chapron C, Vercellini P, Barakat H, Vieira M, Dubuisson JB. Management of\novarian endometriomas. Hum Reprod Update. 2002;8(6):591 –7.\n2. Muzii L, Di Tucci C, Di Mascio D, Di Feliciantonio M, Capri O, Pietrangeli D,\net al. Current management of ovarian endometriomas. Minerva Ginecol.\n2018;70(3):286–94.\n3. Luu TH, Uy-Kroh MJ. New developments in surgery for endometriosis and\npelvic pain. Clin Obstet Gynecol. 2017;60(2):245 –51.\n4. Deckers P, Ribeiro SC, Simoes RDS, Miyahara C, Baracat EC. Systematic\nreview and meta-analysis of the effect of bipolar electrocoagulation during\nlaparoscopic ovarian endometrioma stripping on ovarian reserve. Int J\nGynaecol Obstet. 2018;140(1):11 –7.\n5. Mehdizadeh Kashi A, Chaichian S, Ariana S, Fazaeli M, Moradi Y, Rashidi M, et al.\nThe impact of laparoscopic cystectomy on ovarian reserve in patients with\nunilateral and bilateral endometrioma. Int J Gynaecol Obstet. 2017;136(2):200–4.\n6. Ozaki R, Kumakiri J, Tinelli A, Grimbizis GF, Kitade M, Takeda S. Evaluation of\nfactors predicting diminished ovarian reserve before and after laparoscopic\ncystectomy for ovarian endometriomas: a prospective cohort study. J\nOvarian Res. 2016;9(1):37.\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 9 of 10\n\n7. Muzii L, Achilli C, Lecce F, Bianchi A, Franceschetti S, Marchetti C, et al. Second\nsurgery for recurrent endometriomas is more harmful to healthy ovarian tissue\nand ovarian reserve than first surgery. Fertil Steril. 2015;103(3):738–43.\n8. Lee DY, Bae DS, Yoon BK, Choi D. Post-operative cyclic oral contraceptive\nuse after gonadotrophin-releasing hormone agonist treatment effectively\nprevents endometrioma recurrence. Hum Reprod. 2010;25(12):3050 –4.\n9. Seracchioli R, Mabrouk M, Frasca C, Manuzzi L, Montanari G, Keramyda A, et al.\nLong-term cyclic and continuous oral contraceptive therapy and endometrioma\nrecurrence: a randomized controlledtrial. Fertil Steril. 2010;93(1):52–6.\n10. Guo SW. Recurrence of endometriosis and its control. Hum Reprod Update.\n2009;15(4):441–61.\n11. Revised American Society for Reproductive Medicine classification of\nendometriosis: 1996. Fertil Steril 1997; 67(5):817 –821.\n12. Porpora MG, Pallante D, Ferro A, Crisafi B, Bellati F, Benedetti Panici P. Pain\nand ovarian endometrioma recurrence after laparoscopic treatment of\nendometriosis: a long-term prospective study. Fertil Steril. 2010;93(3):716 –21.\n13. Lagana AS, Vitale SG, Salmeri FM, Triolo O, Ban Frangez H, Vrtacnik-Bokal E, et al.\nUnus pro omnibus, omnes pro uno: a novel,evidence-based, unifying theory for\nthe pathogenesis of endometriosis. Med Hypotheses. 2017;103:10–20.\n14. Vitale SG, Capriglione S, Peterlunger I, La Rosa VL, Vitagliano A, Noventa M,\net al. The role of oxidative stress and membrane transport systems during\nendometriosis: a fresh look at a busy corner. Oxidative Med Cell Longev.\n2018;2018:7924021.\n15. Lagana AS, Garzon S, Franchi M, Casarin J, Gullo G, Ghezzi F. Translational\nanimal models for endometriosis research: a long and windy road. Ann\nTransl Med. 2018;6(22):431.\n16. Tobiume T, Kotani Y, Takaya H, Nakai H, Tsuji I, Suzuki A, et al.\nDeterminant factors of postoperative recurrence of endometriosis:\ndifference between endometrioma and pain. Eur J Obstet Gynecol\nReprod Biol. 2016;205:54 –9.\n17. Chon SJ, Lee SH, Choi JH, Lee JS. Preoperative risk factors in recurrent\nendometrioma after primary conservative surgery. Obstet Gynecol Sci. 2016;\n59(4):286–94.\n18. Selcuk S, Cam C, Koc N, Kucukbas M, Ozkaya E, Eser A, et al. Evaluation of\nrisk factors for the recurrence of ovarian endometriomas. Eur J Obstet\nGynecol Reprod Biol. 2016;203:56 –60.\n19. Guzel AI, Topcu HO, Ekilinc S, Tokmak A, Kokanali MK, Cavkaytar S, et al.\nRecurrence factors in women underwent laparoscopic surgery for\nendometrioma. Minerva Chir. 2014;69(5):277 –82.\n20. Han S, Lee H, Kim S, Joo J, Suh D, Kim K, et al. Risk factors related to the\nrecurrence of endometrioma in patients with long-term postoperative\nmedical therapy. Ginekol Pol. 2018;89(11):611 –7.\n21. Lee DY, Kim HJ, Yoon BK, Choi D. Factors associated with the laterality of\nrecurrent endometriomas after conservative surgery. Gynecol Endocrinol.\n2013;29(11):978–81.\n22. Fauconnier A, Chapron C. Endometriosis and pelvic pain: epidemiological\nevidence of the relationship and implications. Hum Reprod Update. 2005;\n11(6):595–606.\n23. Kucukbas M, Kurek Eken M, Ilhan G, Senol T, Herkiloglu D, Kapudere B.\nWhich factors are associated with the recurrence of endometrioma after\ncystectomy? J Obstet Gynaecol. 2018;38(3):372 –6.\n24. Lagana AS, Vitale SG, Trovato MA, Palmara VI, Rapisarda AM, Granese R, et al.\nFull-thickness excision versus shaving by laparoscopy for intestinal deep\ninfiltrating endometriosis: rationale and potential treatment options. Biomed\nRes Int. 2016;2016:3617179.\n25. Raffaelli R, Garzon S, Baggio S, Genna M, Pomini P, Lagana AS, et al.\nMesenteric vascular and nerve sparing surgery in laparoscopic segmental\nintestinal resection for deep infiltrating endometriosis. Eur J Obstet Gynecol\nReprod Biol. 2018;231:214 –9.\n26. Dior UP, Nisbet D, Fung JN, et al. The Association of Sonographic Evidence\nof Adenomyosis with Severe Endometriosis and Gene Expression in Eutopic\nEndometrium. J Minim Invasive Gynecol. 2019;26(5):941 –8.\n27. Genc M, Genc B, Cengiz H. Adenomyosis and accompanying gynecological\npathologies. Arch Gynecol Obstet. 2015;291(4):877 –81.\n28. Aleksandrovych V, Basta P, Gil K. Current facts constituting an understanding\nof the nature of adenomyosis. Adv Clin Exp Med. 2019;28(6):839 –46.\n29. Mehasseb MK, Bell SC, Pringle JH, Habiba MA. Uterine adenomyosis is\nassociated with ultrastructural features of altered contractility in the inner\nmyometrium. Fertil Steril. 2010;93(7):2130 –6.\n30. Perello MF, Martinez-Zamora MA, Torres X, Munros J, Balasch Cortina J,\nCarmona F. Endometriotic pain is associated with Adenomyosis but not\nwith the compartments affected by deep infiltrating endometriosis. Gynecol\nObstet Investig. 2017;82(3):240 –6.\n31. Seo JW, Lee DY, Yoon BK, Choi D. The age-related recurrence of\nendometrioma after conservative surgery. Eur J Obstet Gynecol Reprod Biol.\n2017;208:81–5.\n32. Shi J, Leng J, Cui Q, Lang J. Follicle loss after laparoscopic treatment of\novarian endometriotic cysts. Int J Gynaecol Obstet. 2011;115(3):277 –81.\n33. Vercellini P, Somigliana E, Daguati R, Vigano P, Meroni F, Crosignani PG.\nPostoperative oral contraceptive exposure and risk of endometrioma\nrecurrence. Am J Obstet Gynecol. 2008;198(5):504 e1 –5.\n34. Zhu S, Zhu Y, Liu Y, Zhang H. Comparison of outcomes of different\npostoperative hormone therapy in the treatment of ovarian endometriosis:\na brief report. Adv Ther. 2018;35(6):857 –63.\n35. Reis FM, Petraglia F, Kim MK. Postoperative Levonorgestrel-releasing\nintrauterine system insertion after gonadotropin-releasing hormone agonist\ntreatment for preventing Endometriotic cyst recurrence: a prospective\nobservational study. F1000Research. 2018;25(1):39 –43.\n36. Seo JW, Lee DY, Yoon BK, Choi D. The efficacy of postoperative cyclic Oral\ncontraceptives after gonadotropin-releasing hormone agonist therapy to\nprevent Endometrioma recurrence in adolescents. J Pediatr Adolesc\nGynecol. 2017;30(2):223–7.\n37. Takamura M, Koga K, Osuga Y, Takemura Y, Hamasaki K, Hirota Y, et al.\nPost-operative oral contraceptive use reduces the risk of ovarian\nendometrioma recurrence after laparoscopic excision. Hum Reprod.\n2009;24(12):3042 –8.\n38. Muzii L, Di Tucci C, Achilli C, Di Donato V, Musella A, Palaia I, et al.\nContinuous versus cyclic oral contraceptives after laparoscopic excision of\novarian endometriomas: a systematic review and metaanalysis. Am J Obstet\nGynecol. 2016;214(2):203 –11.\n39. Vercellini P, Crosignani PG, Fadini R, Radici E, Belloni C, Sismondi P. A\ngonadotrophin-releasing hormone agonist compared with expectant\nmanagement after conservative surgery for symptomatic endometriosis. Br J\nObstet Gynaecol. 1999;106(7):672 –7.\n40. Kim MK, Chon SJ, Lee JH, Yun BH, Cho S, Choi YS, et al. Postoperative\nLevonorgestrel-releasing intr auterine system insertion after\ngonadotropin-releasing hormone agonist treatment for preventing\nEndometriotic cyst recurrence: a prospective observational study.\nReprod Sci (Thousand Oaks, Calif). 2018;25(1):39 –43.\n41. Jee BC, Lee JY, Suh CS, Kim SH, Choi YM, Moon SY. Impact of GnRH agonist\ntreatment on recurrence of ovarian endometriomas after conservative\nlaparoscopic surgery. Fertil Steril. 2009;91(1):40 –5.\n42. Gordts S, Campo R. Modern approaches to surgical management of\nendometrioma. Best Pract Res Clin Obstet Gynaecol. 2019.\n43. Muzii L, Bianchi A, Bellati F, Cristi E, Pernice M, Zullo MA, et al. Histologic\nanalysis of endometriomas: what the surgeon needs to know. Fertil Steril.\n2007;87(2):362–6.\n44. Liu X, Yuan L, Shen F, Zhu Z, Jiang H, Guo SW. Patterns of and risk factors\nfor recurrence in women with ovarian endometriomas. Obstet Gynecol.\n2007;109(6):1411–20.\n45. Ashrafi M, Sadatmahalleh SJ, Akhoond MR, Talebi M. Evaluation of risk\nfactors associated with endometriosis in infertile women. Int J Fertil Steril.\n2016;10(1):11–21.\n46. Koga K, Takemura Y, Osuga Y, Yoshino O, Hirota Y, Hirata T, et al. Recurrence\nof ovarian endometrioma after laparoscopic excision. Hum Reprod. 2006;\n21(8):2171–4.\nPublisher’sN o t e\nSpringer Nature remains neutral with regard to jurisdictional claims in\npublished maps and institutional affiliations.\nLi et al. Journal of Ovarian Research           (2019) 12:79 Page 10 of 10","source_license":"CC0","license_restricted":false}