{"paper_id":"1c7eb7b7-7c13-47f4-8743-3d0fa7b3e288","body_text":"Vol.:(0123456789)1 3\nArchives of Gynecology and Obstetrics (2022) 305:977–984 \nhttps://doi.org/10.1007/s00404-021-06200-w\nGYNECOLOGIC ENDOCRINOLOGY AND REPRODUCTIVE MEDICINE\nComprehensive characterization of endometriosis patients \nand disease patterns in a large clinical cohort\nSimon Blum1  · Peter A. Fasching1 · Thomas Hildebrandt1 · Johannes Lermann1 · Felix Heindl1 · Tilman Born1 · \nHannah Lubrich1 · Sophia Antoniadis1 · Karina Becker1 · Christine Fahlbusch1 · Katharina Heusinger1 · \nStefanie Burghaus1 · Matthias W. Beckmann1 · Alexander Hein1\nReceived: 15 April 2021 / Accepted: 17 August 2021 / Published online: 26 August 2021 \n© The Author(s) 2021\nAbstract\nPurpose In many diseases, it is possible to classify a heterogeneous group into subgroups relative to tumor biology, genetic \nvariations, or clinical and pathological features. No such classification is available for endometriosis. In our retrospective \ncase–case analysis we defined subgroups of endometriosis patients relative to the type and location of the endometriosis \nlesion and relative to basic patient characteristics.\nMethods From June 2013 to July 2017, a total of 1576 patients with endometriosis diagnosed at surgery were included in \nthis study. The patients’ history and clinical data were documented using a web-based remote data entry system. To build \nsubgroups, all possible combinations of endometriosis locations/types (peritoneal; ovarian endometriosis; deeply infiltrat-\ning endometriosis; adenomyosis) were used. Due to the variation in group sizes, they were combined into five substantial \nlarger groups.\nResults Age, pregnancy rate, and live birth rate were identified as characteristics that significantly differed between the five \npatient groups that were defined. No significant differences were noted in relation to body mass index, length of menstrual \ncycle, age at menarche, reason for presentation, or educational level.\nConclusion This study describes basic patient characteristics in relation to common clinical subgroups in a large clinical \ncohort of endometriosis patients. Epidemiological information about different clinical groups may be helpful in identifying \ngroups with specific clinical courses, potentially suggesting novel approaches to early detection and to surgical and systemic \ntreatment.\nKeywords Endometriosis · Case–case analysis · Classification\nIntroduction\nEndometriosis is a disease that is heterogeneous in relation \nto both symptoms and patterns of spread in the peritoneal \ncavity. Patients usually present with lower abdominal pain \nand dysmenorrhea and/or infertility, or may be asympto-\nmatic [1–3]. In the abdomen, endometriotic lesions can be \nfound either confined to the pelvis or spreading up to the \ndiaphragm [4, 5]. In some patients, endometriosis lesions are \nnot attached to the mesothelium superficially, but infiltrate \nadjacent structures or organs [4].\nFor many diseases, classification has helped to improve \ntreatment, as it enables subgroups to be identified that have \ndifferent types of risk for the most important outcomes, or \ndifferent responses to specific therapies. In many diseases, it \nis possible to classify a heterogeneous group into subgroups \nrelative to tumor biology, genetic variations, or clinical and \npathological features [ 6, 7]. This can make individualized \ntreatment possible, improving the disease-specific and over-\nall survival for the patients. No such classification is avail-\nable for endometriosis, although it is one of the most com-\nmon diseases in women of reproductive age and reduces \ntheir quality of life and ability to work [8].\n * Alexander Hein \n alexander.hein@uk-erlangen.de\n1 Department of Gynecology and Obstetrics, University \nEndometriosis Center for Franconia, Erlangen University \nHospital, Friedrich Alexander University of Erlangen-\nNuremberg, Universitätsstrasse 21–23, 91054 Erlangen, \nGermany\n\n978 Archives of Gynecology and Obstetrics (2022) 305:977–984\n1 3\nThere are only descriptive endometriosis classifications, \nbut they do not classify a heterogenous group into subgroups \nrelative to clinically relevant features. The most common are \nspecific classification systems, such as the revised Ameri-\ncan Society for Reproductive Medicine (rASRM) score [9 ] \nand the Enzian classification [10]. The latter classification \nsystem describes different locations for deeply infiltrating \nendometriotic lesions in general. The rASRM score is a \nweighted scoring system for assessing endometrial implants, \nplaques, endometriomas, and/or adhesions on the ovaries \nand/or peritoneum. None of the available classification sys-\ntems gathers together all the different types of endometriotic \nlesion [11]. It is not possible to derive subgroups from any of \nthese classifications that would be of meaningful diagnostic \nor therapeutic relevance.\nSome studies have investigated the relationship between \npatient characteristics and endometriosis [12], but no \nattempts have previously been made to correlate patient \ncharacteristics such as age at diagnosis, body mass index \n(BMI), length of menstrual cycle, age at menarche, reason \nfor consultation, pregnancy rates, live birth rates, or educa-\ntional level with subgroups of endometriosis patients. Some \nof these patient characteristics have been recognized as rep-\nresenting risk factors for endometriosis in general [13, 14].\nThe aim of this study was therefore to identify groups of \npatients with endometriosis and to investigate the character-\nistics of patients in different groups. This may be helpful for \nfurther specifying and individualizing endometriosis therapy \n[15], whether surgical and/or medical, and thus improving \nthe patients’ quality of life.\nMethods\nStudy population\nThis retrospective observational study was conducted from \nJune 2013 to July 2017 in the Department of Gynecol-\nogy and Obstetrics at Erlangen University Hospital. It was \ndesigned as a case–case analysis in which different groups \nof patients with the same disease were compared with each \nother. Patients who were diagnosed with endometriosis at \nlaparoscopy during this time span were eligible for inclu-\nsion. The time of the patient’s first operation in Erlangen \nUniversity Hospital during this period was defined as the \ntime of first presentation. A total of 1576 patients in whom \nendometriosis was diagnosed at surgery were identified. The \ndiagnosis date was the earliest date on which endometriosis \nwas first diagnosed at surgery. Of the 1576 patients, 356 had \nmissing data or an imprecise surgical diagnosis and were \nexcluded. A further 144 patients were excluded because \ninformation about the type and location of the endometrio-\nsis was missing. After all the exclusion criteria had been \napplied, the remaining study population was 1076 patients \n(Supplementary Fig. 1). All of the participants provided \nwritten informed consent and the medical faculty’s ethics \ncommittee approved the study.\nData acquisition\nThe data collected was obtained from the patients’ charts or \nfrom a structured questionnaire completed by patients. The \npatients’ history and clinical data were documented using a \nweb-based remote data entry system (electronic case report \nform, eCRF) [15] and were transferred to an MS Access \ndatabase for further variable extraction. The eCRF collects \ndata on 23 variables at registration, and data on at least 41 \nvariables were collected when the patient’s medical history \nwas being documented. In addition, at least 22 endometrio-\nsis-specific variables were recorded; if the patient underwent \nsurgery for endometriosis, data on a further 18 variables \nwere recorded. Detailed information was collected regard-\ning age at first diagnosis, body mass index, age at menarche, \nmenstrual cycle length, number of pregnancies and live \nbirths, educational level, marital status, ethnicity and current \nemployment, diseases apart from endometriosis, reason for \nconsultation, the way in which endometriosis was diagnosed, \ndetailed surgical information, grade and location of endo-\nmetriosis, and histological information. The basic patient \ncharacteristics are listed in Table  1. This procedure allows \nthe type of endometriosis diagnosis to be differentiated very \nprecisely, with the data showing whether the patient had \nsuperficial endometriosis, deeply infiltrating endometriosis \n(DIE), endometrioma, and/or adenomyosis uteri.\nDefinition of groups\nEndometriosis was diagnosed during laparoscopic sur -\ngery in 1076 patients. Preoperative a standardized vaginal \nexamination and ultrasound was performed. Endometrio-\nsis was diagnosed through histological examination of the \nspecimens. In cases where a histologic diagnosis was not \navailable, the diagnosis and determination of the spread of \nendometriosis, e.g., adenomyosis, was made by the surgeon \nbased on the preoperative and perioperative findings. The \nearliest date of an operation during which endometriosis was \ndiagnosed in each patient was used as the date of diagnosis. \nSince no clinically relevant subgroups can be defined with \nthe available classifications, we decided to define the sub-\ngroups based on the type and location of the endometriosis. \nIn addition, the available classifications are too specific and \ndo not cover all types of endometriotic lesions. To build \nsubgroups without presuppositions, all possible combina-\ntions of endometriosis locations (peritoneal, yes/no; ovarian \nendometriosis, yes/no; deeply infiltrating endometriosis, yes/\nno; adenomyosis, yes/no) were used On the basis of these \n\n979Archives of Gynecology and Obstetrics (2022) 305:977–984 \n1 3\nfour criteria, which could either be present or not, 16 distinct \ngroups were formed. Each patient could be assigned to a \nspecific group. However, one group—with no endometrio-\nsis and with all four criteria negative (subgroup 16)—did \nnot exist (Supplementary Table 1). The 1076 patients were \nclearly allocated to one of the 15 subgroups. The subgroups \nhad widely varying sizes, ranging from two to 350 patients. \nDue to the variation in group sizes, it was not practicable to \ninvestigate all 15 subgroups and they were therefore com-\nbined into meaningful larger groups: 1—peritoneal endome-\ntriosis only; 2—peritoneal endometriosis and adenomyosis; \n3—adenomyosis only; 4—peritoneal and DIE-dominant; \nand 5—endometrioma-dominant and other findings (Table  2 \nand Supplementary Fig. 2 [ 16]). It may be hypothesized that \nthese different types need different treatment approaches. \nFurther research will be needed to validate these results and \nattempt to identify predictive factors to assist in the choice \nof therapy.\nStatistical considerations\nThe patients’ characteristics are presented as means with \nstandard deviation, or counts and percentages.\nOne-way analysis of variation (ANOVA) was performed \nfor the characteristics of age at first diagnosis of endometrio-\nsis, BMI, age at menarche, and length of menstrual cycle. \nFor categorical variables such as the number of pregnan-\ncies, number of live births, educational level, and main \nreason for presentation, a Chi-squared test was performed. \nThree categories were formed to examine pregnancy rates: \npatients without pregnancies, patients with one pregnancy, \nand patients with two or more pregnancies. The same pro-\ncedure was used for live birth rates. Educational level was \ndivided into two groups: patients with a university degree \nand patients without a university degree. The main reason \nfor consultation was classified into pain, infertility, or other \nreasons.\nAll of the tests were two-sided, and a P value < 0.05 was \nregarded as statistically significant. Calculations were car -\nried out using the R system for statistical computing (version \n2.13.1; R Development Core Team, Vienna, Austria, 2011).\nResults\nThe average age at which endometriosis was first diagnosed \nwas 33.9 years (SD ± 8.4). The study population was eth-\nnically very homogeneous. Most of the patients were of \nEuropean ethnicity (93.2%); only 5.4% were Asian, and the \nremainder (1.4%) were of other backgrounds. In all, 36.9% \nof the patients had a university degree. The patients’ mean \nBMI was in the upper normal range, at 24.1 (SD ± 5.1). \nThe mean length of the menstrual cycle was 28.5 days \n(SD ± 9.0). The mean age at menarche was 13.0 years. The \npatients’ main reasons for consultation were pain (48.5%), \nfollowed by infertility (27.5%), and 24% had other rea-\nsons for consultation. A total of 944 of the 1076 patients \n(87.7%) underwent their first operation during this study; \nTable 1  Patient characteristics\nThe table shows the patient characteristics oft the study population\nBMI body mass index\nPatient characteristics Mean or \nfrequency\nStandard \ndeviation or \npercent\nTotal 1076 100%\nAge at first diagnosis (y) 33.9  ± 8.4\nBMI at first presentation (kg/m2) 24.1  ± 5.1\nAge at menarche (y) 13.0  ± 1.5\nLength of menstrual cycle (days) 28.6  ± 9.0\nNumber of pregnancies at first presentation\n 0 670 62.7%\n 1 190 17.8%\n ≥ 2 209 19.6%\nNumber of live births\n 0 769 72.1%\n 1 144 13.5%\n ≥ 2 154 14.4%\nEducational level\n University 142 36.9%\n Other 243 63.1%\nEthnicity\n European 138 93.2%\n Hispanic American 0 –\n African 0 –\n Asian 8 5.4%\n Other 2 1.4%\nCurrent employment\n Employed full-time/part-time 82 59.4%\n Retired 3 2.2%\n Housewife 20 14.5%\n Student 28 20.3%\n Unemployed 5 3.6%\nMain reason for presentation\n Pain 520 48.5%\n Infertility 295 27.5%\n Other 258 24.0%\nPrevious surgery\n 0 944 87.7%\n 1 112 10.4%\n ≥ 2 20 1.9%\nMedical history\n No previous therapy 952 88.5%\n Previous therapy 98 9.1%\n\n980 Archives of Gynecology and Obstetrics (2022) 305:977–984\n1 3\n112 patients (10.4%) had had one previous operation, and \n20 patients (1.9%) had had more than one previous opera -\ntion. Most of the patients (n  = 952, 88.5%) had not had any \nprevious therapy. Ninety-eight patients (9.1%) had already \nreceived drug therapy. Only 79 of them were using hormo-\nnal therapy such as contraceptives, a hormonal intrauterine \ndevice (IUD) or gonadotropin-releasing hormone (GnRH) \nagonists. No information about previous therapy was avail-\nable for 26 patients (2.4%) (Table  1).\nPeritoneal endometriosis was found in 82.8% of the 1076 \npatients. Diagnoses of adenomyosis uteri were less common, \nat 37.5%, followed by deeply infiltrating endometriosis at \n27.8% and endometrioma at 27.0%.\nAdenomyosis was diagnosed in 404 patients. Most often, \nadenomyosis was diagnosed intraoperatively. Only 76 \npatients had a histology of adenomyosis after hysterectomy. \nThe 76 patients with histologically confirmed adenomyosis \nwere almost evenly distributed among the different groups. \nIn the group “Adenomyosis only” were 25 of 115 patients \nand in the group “Peritoneal endometriosis and adenomyo-\nsis” 22 of 142 patients with histologically confirmed adeno-\nmyosis. The remaining patients were distributed among the \nother groups. A correlation between histologically diagnosed \nadenomyosis and older age is possible, but due to the even \ndistribution of the histologically confirmed adenomyosis, it \nis rather unlikely.\nThe data showed that the mean age at the first diagno-\nsis of endometriosis differed significantly in the different \ngroups (P < 0.001). Patients with peritoneal endometriosis \nonly (group 1) were more than 3 years younger (32.5 years) \nthan patients with adenomyosis uteri (35.9 years) (Table  3; \nSupplementary Fig. 3). In comparison with the other groups, \nthe group with adenomyosis only had the lowest rate without \npregnancy (42.5%) and the lowest rate without a live birth \n(53.6%). Over 30.4% of patients with adenomyosis had at \nleast two live births, while 16.1% had one live birth. This \nTable 2  Division of the subgroups into five patient groups\nAll the 15 subgroups were combined into meaningful larger groups\nDIE deeply infiltrating endometriosis\nEndometriosis location in subgroups n (%) in subgroup Group n (%) in group\nPeritoneal yes/endometrioma no /DIE no/adenomyo-\nsis no\n350 (32.5) Group 1: Peritoneal endometriosis only 350 (32.5)\nPeritoneal yes/endometrioma no / DIE no/adenomyo-\nsis yes\n142 (13.2) Group 2: Peritoneal endometriosis and adenomyosis 142 (13.2)\nPeritoneal no/endometrioma no /DIE no/adenomyosis \nyes\n115 (10.7) Group 3: Adenomyosis 115 (10.7)\nPeritoneal yes/endometrioma no /DIE yes/adenomyo-\nsis no\n105 (9.8) Group 4: Peritoneal and DIE-dominant 275 (25.6)\nPeritoneal yes/endometrioma yes/DIE yes/adenomyo-\nsis no\n60 (5.6)\nPeritoneal yes/endometrioma no /DIE yes/adenomyo-\nsis yes\n57 (5.3)\nPeritoneal yes/endometrioma yes/DIE yes/adenomyo-\nsis yes\n53 (4.9)\nPeritoneal yes/endometrioma yes/DIE no/adenomyo-\nsis no\n99 (9.2) Group 5: Endometrioma-dominant and other 194 (18.0)\nPeritoneal no/endometrioma yes /DIE no/adenomyo-\nsis no\n40 (3.7)\nPeritoneal yes/endometrioma yes/DIE no/adenomyo-\nsis yes\n25 (2.3)\nPeritoneal no/endometrioma no /DIE yes/adenomyo-\nsis no\n14 (1.3)\nPeritoneal no/endometrioma yes /DIE no/adenomyo-\nsis yes\n6 (0.6)\nPeritoneal no/endometrioma yes /DIE yes/adenomyo-\nsis no\n4 (0.4)\nPeritoneal no/endometrioma yes /DIE yes/adenomyo-\nsis yes\n4 (0.4)\nPeritoneal no/endometrioma no /DIE yes/adenomyo-\nsis yes\n2 (0.2)\nTotal 1076 (100.0) All groups 1076 (100.0)\n\n981Archives of Gynecology and Obstetrics (2022) 305:977–984 \n1 3\nTable 3  Patient characteristics in the five patient groups\nThe table shows the patient characteristics per group. In our study we tested for differences between the five groups. The mean age at first diag-\nnosis of endometriosis differed significantly in the different groups (P < 0.001). There were also significant results for pregnancies (P < 0.001) \nand for live births (P < 0.001). There were no significant differences between the other patient characteristics\nBMI body mass index, DIE deeply infiltrating endometriosis\nGroup 1 \nPeritoneal only \nn = 350 (32.5%)\nMean (SD) or n (%)\nGroup 2 \nPeritoneal & \nadenomyosis \nn = 142 \n(13.2%)\nMean (SD) \nor n (%)\nGroup 3 \nAdenomyosis only \nn = 115 (10.7%)\nMean (SD) or n (%)\nGroup 4 \nPeritoneal & DIE \nn = 275 (25.6%)\nMean (SD) or n (%)\nGroup 5 \nEndometrioma \nn = 194 (18.0%)\nMean (SD) or n (%)\nTotal \nn = 1076 (100%)\nMean (SD) or n (%)\nAge at first diagno-\nsis (y)\n32.6 (9.1)\nn = 350\n33.3 (8.1)\nn = 142\n35.9 (9.9)\nn = 115\n34.3 (7.2)\nn = 275\n35.0 (7.5)\nn = 194\n33.9 (8.4)\nn = 1076\nBMI at first presen-\ntation (kg/m2)\n24.0 (5.5)\nn = 279\n24.5 (5.1)\nn = 90\n24.3 (4.7)\nn = 74\n24.0 (5.2)\nn = 211\n24.2 (4.7)\nn = 159\n24.1 (5.1)\nn = 813\nAge at menarche (y) 13.0 (1.4)\nn = 318\n13.0 (1.7)\nn = 137\n12.8 (1.6)\nn = 106\n13.1 (1.4)\nn = 255\n13.0 (1.5)\nn = 175\n13.0 (1.5)\nn = 991\nLength of menstrual \ncycle (days)\n28.1 (5.2)\nn = 183\n29.1 (8.5)\nn = 80\n30.4 (14.2)\nn = 58\n27.7 (3.3)\nn = 168\n29.4 (15.0)\nn = 111\n28.6 (9.0)\nn = 600\nNo. of pregnancies at first presentation\n 0 229 (66.2) 82 (58.2) 48 (42.5) 186 (67.6) 125 (64.4) 670 (62.7)\n 1 55 (15.9) 34 (24.1) 20 (17.7) 51 (18.5) 30 (15.5) 190 (17.8)\n ≥ 2 62 (17.9) 25 (17.7) 45 (39.8) 38 (13.8) 39 (20.1) 209 (19.6)\n Total 346 (100) 141 (100) 113 (100) 275 (100) 194 (100) 1069 (100)\nNo. of live births\n 0 259 (75.1) 102 (72.3) 60 (53.6) 211 (76.7) 137 (70.6) 769 (72.1)\n 1 44 (12.8) 18 (12.8) 18 (16.1) 36 (13.1) 28 (14.4) 144 (13.5)\n ≥ 2 42 (12.2) 21 (14.9) 34 (30.4) 28 (10.2) 29 (14.9) 154 (14.4)\n Total 345 (100) 141 (100) 112 (100) 275 (100) 194 (100) 1067 (100)\nEducational level\n University 52 (39.4) 15 (34.1) 8 (26.7) 42 (40.4) 25 (33.3) 142 (36.9)\n Other 80 (60.6) 29 (65.9) 22 (73.3) 62 (59.6) 50 (66.6) 243 (63.1)\n Total 132 (100) 44 (100) 30 (100) 104 (100) 75 (100) 385 (100)\nEthnicity\n European 43 (91.5) 16 (94.1) 8 (88.9) 44 (95.7) 27 (93.1) 138 (93.2)\n Hispanic Ameri-\ncan\n0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)\n African 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)\n Asian 3 (6.4) 1 (5.9) 0 (0.0) 2 (4.3) 2 (6.9) 8 (5.4)\n Other 1 (2.1) 0 (0.0) 1 (11.1) 0 (0.0) 0 (0.0) 2 (1.4)\n Total 47 (100) 17 (100) 9 (100) 46 (100) 29 (100) 148 (100)\nCurrent employment\n Employed full-\ntime/part-time\n22 (50.0) 10 (71.4) 7 (50.0) 27 (71.1) 16 (57.1) 82 (59.4)\n Retired 2 (4.5) 0 (0.0) 0 (0.0) 0 (0.0) 1 (3.6) 3 (2.2)\n Housewife 6 (13.6) 1 (7.1) 4 (28.6) 3 (7.9) 6 (21.4) 20 (14.5)\n Student 13 (29.5) 3 (21.4) 2 (14.3) 7 (18.4) 3 (10.7) 28 (20.3)\n Unemployed 1 (2.3) 0 (0.0) 1 (7.1) 1 (2.6) 2 (7.1) 5 (3.6)\n Total 44 (100) 14 (100) 14 (100) 38 (100) 28 (100) 138 (100)\nMain reason for presentation\n Pain 166 (47.7) 73 (51.4) 67 (58.3) 129 (46.9) 85 (44.0) 520 (48.5)\n Infertility 87 (25.0) 45 (31.7) 24 (20.9) 85 (30.9) 54 (28.0) 295 (27.5)\n Other 95 (27.3) 24 (16.9) 24 (20.9) 61 (22.2) 54 (28.0) 258 (24.0)\n Total 348 (100) 142 (100) 115 (100) 275 (100) 193 (100) 1073 (100)\n\n982 Archives of Gynecology and Obstetrics (2022) 305:977–984\n1 3\nmeans that 46.5% of the patients with adenomyosis had \nat least one child, in comparison with the other groups, in \nwhich fewer than 30% had at least one live birth (Supple-\nmentary Figs. 4 and 5). There were also significant results \nfor pregnancies (P  < 0.001) and for live births (P  < 0.001) \n(Table  3). In view of the high rates of pregnancy and live \nbirths, infertility as the main reason for consultation was \nrare in the group with adenomyosis. However, there were \nno significant differences between the five groups in rela-\ntion to this parameter (P  = 0.62). There were also no sig-\nnificant differences between the groups in relation to body \nmass index, educational level, length of menstrual cycle, or \nage at menarche.\nDiscussion\nIn this retrospective case–case study, basic patient charac-\nteristics were associated with the location of endometrio-\nsis by developing subgroups of endometriosis lesions. In \nfive defined groups, significant differences were found in \nthe patient’s mean ages at first diagnosis of endometriosis. \nSignificant differences were also found between pregnancy \nrates and also between live birth rates in the groups.\nIt has been shown in previous studies that there is a risk \nof developing endometriosis only during women’s repro-\nductive years. In earlier studies, the highest risk was found \nto be at 44 years of age [17], with another peak between 25 \nand 29 years. The highest risk reported in recent studies was \nbetween 25 and 29 years. The risk was found to decrease \nafter the age of 44 in all of the studies [12, 18, 19]. In the \npresent study, the mean age at onset was 33.9 years. There is \na known delay between the first symptoms and the diagnosis \nof endometriosis at surgery. In addition, the time of the first \noperation is influenced by socio–economic factors such as \neducation, access to special medical facilities, or age at the \nonset of symptoms [8 , 20, 21]. This can delay the surgical \ndiagnosis of endometriosis. However, as the present study \ninvolved a case–case analysis, the mean age may be influ-\nenced, but not the differences between the groups.\nThere have only been a few studies examining endo-\nmetriosis groups. In one study, no significant difference \nwas found between patients with DIE and patients with \nperitoneal endometriosis or endometrioma [22]. In another \nstudy, a significant difference was found between the \naverage age of women diagnosed with endometriosis and \nwomen diagnosed with adenomyosis uteri. Patients with \nendometriosis were younger than those with adenomyosis \nuteri [23]. The present study examined whether there were \nany differences between five defined groups. The mean age \nat first diagnosis of endometriosis differed significantly in \nthe group analysis. Patients with adenomyosis uteri were \n3 years older than patients with peritoneal endometriosis \nonly. In the present study, there were two diagnostic meth-\nods of identifying adenomyosis uteri—either through his-\ntology or the surgeon’s impression. However, hysterec-\ntomy was avoided in young patients, and this may have \ninfluenced the high mean age in patients with adenomyosis \nuteri. Nevertheless, the mean age in the group with perito-\nneal endometriosis and adenomyosis uteri was the second \nlowest in all the groups. It is not therefore expected that \nthe results may have been biased as a result of the diag-\nnostic methods used.\nThe data showed significant differences between the \ngroups in relation to pregnancy rates and also live birth \nrates, and suggest that patients with adenomyosis have a \nhigher pregnancy rate and live birth rate than patients in \nthe other four groups. In previous studies, adenomyosis in \nparticular is considered to be a cause of infertility and mis-\ncarriage when assisted reproductive techniques are used [24, \n25]. Other studies have primarily examined the influence of \nendometriosis on infertility [26]. However, there are no data \non the pregnancy rate and live birth rate with spontaneous \nconception. In addition, no studies were found in the litera-\nture that have compared pregnancy rates or live birth rates \nbetween different groups with endometriosis. The average \nrates for women without pregnancy and without live births \nwere high. However, comparison with a control group was \nnot available and it is therefore not possible to draw any con-\nclusions regarding infertility in comparison with a healthy \npopulation.\nThe odds of developing endometriosis have been reported \nto be lower among women with histologically confirmed \nendometriosis who had a large versus lean body size [27]; \na higher body mass index may also be associated with a \nlower risk of endometriosis [28]. In a subgroup analysis, \nobese women were found to be more likely to have a surgi-\ncal diagnosis of adenomyosis [23]. In the present study, the \ngroups with adenomyosis uteri had the highest BMI, but the \ndifferences were not significant.\nWith regard to the menstrual cycle, current studies \nassume that frequent menstrual bleeding is a risk factor for \nendometriosis. Early age at menarche, shorter menstrual \ncycles and thus more frequent period bleeding, longer bleed-\ning periods, and few pregnancies appear to be risk factors \n[29]. It has previously been reported that age at menarche \ndid not differ between patients who underwent hysterectomy \nwith a diagnosis of adenomyosis and those who did not have \na diagnosis of adenomyosis [30].\nThere is not known to be any association between endo-\nmetriosis and educational level [8], but no subgroup analyses \nhave as yet been published.\nNo differences were found between the groups with \nregard to the main reasons for consultation. These data are \nconsistent with data showing that there is no association \nbetween the extent of endometriosis and pain [31]. On the \n\n983Archives of Gynecology and Obstetrics (2022) 305:977–984 \n1 3\nother hand, there is some evidence that adenomyosis uteri \nis associated with pain [32].\nCase selection in a hospital-based study may be biased by \nthe variety of health care that is provided in the hospital con-\ncerned. Women seeking help for pelvic pain might increase \nthe numbers of cases of pain-inducing endometriosis, and \na hospital providing specialized health care for infertility \npatients might have larger numbers of patients. In the pre-\nsent study, the hospital is a center for all types of treatment, \nso that a bias toward one of these groups seems unlikely. \nPatients who underwent surgical therapy were selected. The \nstudy group included therefore does not correspond to the \nnormal distribution of endometriosis patients and thus does \nnot represent an adequate epidemiological picture.\nOne major limitation of the study is that pregnancy rates \nand live birth rates may be influenced by other patient char-\nacteristics, such as the main reason for consultation or mean \nage. The present data are only able to show associations, but \nnot causality.\nAnother limitation is the type of diagnostic method used \nto identify adenomyosis uteri through clinical aspects and \nthe surgeon’s impression. The diagnosis of adenomyosis \nuteri is clinically challenging, as there is no consensus on \nthe best imaging features for a nonsurgical diagnosis of \nadenomyosis [33]. Nevertheless, standardized preoperative \nultrasound was performed in our study in 1045 of the 1076 \ncases. Adenomyosis was suspected in 135 patients. All of \nthe patients underwent a preoperative vaginal examination. \nThe gold standard of hysterectomy cannot be applied in this \npopulation. In relation to international guidelines, it was \ndecided that the surgeon should make the intraoperative \ndiagnosis of an endometriosis on the basis of the medical \nhistory and preoperative clinical findings [34]. Histological \ndiagnosis of the disease or longer-term follow-up data might \nbe helpful for refining the groups.\nIn conclusion, this study has added to the evidence that \ndifferent subgroups exist among endometriosis patients. It \nmay be hypothesized that these different types need differ -\nent treatment approaches. Further research will be needed \nto confirm these results and attempt to identify predictive \nfactors to assist in the choice of therapy,\nSupplementary Information The online version contains supplemen-\ntary material available at https:// doi. org/ 10. 1007/ s00404- 021- 06200-w.\nAcknowledgements The contribution of Simon Blum to this publica-\ntion was performed in partial fulfillment of the requirements for obtain-\ning the degree of Doctor of Medicine. Parts of the research published \nhere have been used for his doctoral thesis at the Medical Faculty of \nFriedrich Alexander University of Erlangen–Nuremberg (FAU).\nAuthor contributions SB: protocol/project development, data col-\nlection or management, data analysis, manuscript writing/editing. \nPAF: protocol/project development, data collection or management, \ndata analysis, manuscript writing/editing. TH: manuscript editing. \nJL: manuscript editing. FH: manuscript editing. TB: manuscript edit-\ning. HL: manuscript editing. SA: manuscript editing. KB: manuscript \nediting. CF: manuscript editing. KH: manuscript editing. SB: proto-\ncol/project development, manuscript writing/editing. MWB: project \ndevelopment, manuscript editing. AH: protocol/project development, \nmanuscript writing/editing.\nFunding Open Access funding enabled and organized by Projekt \nDEAL.\nDeclarations \nConflicts of interest There were no conflicts of interest.\nEthics approval The medical faculty’s ethics committee approved the \nstudy.\nOpen Access This article is licensed under a Creative Commons Attri-\nbution 4.0 International License, which permits use, sharing, adapta-\ntion, distribution and reproduction in any medium or format, as long \nas you give appropriate credit to the original author(s) and the source, \nprovide a link to the Creative Commons licence, and indicate if changes \nwere made. The images or other third party material in this article are \nincluded in the article’s Creative Commons licence, unless indicated \notherwise in a credit line to the material. If material is not included in \nthe article’s Creative Commons licence and your intended use is not \npermitted by statutory regulation or exceeds the permitted use, you will \nneed to obtain permission directly from the copyright holder. To view a \ncopy of this licence, visit http://creativecommons.org/licenses/by/4.0/.\nReferences\n 1. Fauconnier A, Chapron C (2005) Endometriosis and pelvic pain: \nepidemiological evidence of the relationship and implications. \nHum Reprod Update 11(6):595–606\n 2. Schliep KC et al (2015) Pain typology and incident endometriosis. \nHum Reprod 30(10):2427–2438\n 3. Burghaus S et al (2019) Standards used by a clinical and scientific \nendometriosis center for the diagnosis and therapy of patients with \nendometriosis. Geburtshilfe Frauenheilkd 79(5):487–497\n 4. Nisolle M, Donnez J (1997) Peritoneal endometriosis, ovarian \nendometriosis, and adenomyotic nodules of the rectovaginal sep-\ntum are three different entities. Fertil Steril 68(4):585–596\n 5. Audebert A et al (2018) Anatomic distribution of endometriosis: a \nreappraisal based on series of 1101 patients. Eur J Obstet Gynecol \nReprod Biol 230:36–40\n 6. Curtis C et al (2012) The genomic and transcriptomic architec-\nture of 2,000 breast tumours reveals novel subgroups. Nature \n486(7403):346–352\n 7. Prat J (2012) Ovarian carcinomas: five distinct diseases with dif-\nferent origins, genetic alterations, and clinicopathological fea-\ntures. Virchows Arch 460(3):237–249\n 8. Nnoaham KE et al (2011) Impact of endometriosis on quality of \nlife and work productivity: a multicenter study across ten coun-\ntries. Fertil Steril 96(2):366-373.e8\n 9. American Society for Reproductive Medicine (1997) Revised \nAmerican Society for Reproductive Medicine classification of \nendometriosis: 1996. Fertil Steril 67(5):817–821\n 10. Haas D et al (2013) The rASRM score and the Enzian classifi-\ncation for endometriosis: their strengths and weaknesses. Acta \nObstet Gynecol Scand 92(1):3–7\n\n984 Archives of Gynecology and Obstetrics (2022) 305:977–984\n1 3\n 11. Johnson NP et al (2017) World Endometriosis Society con-\nsensus on the classification of endometriosis. Hum Reprod \n32(2):315–324\n 12. Missmer SA et al (2004) Incidence of laparoscopically confirmed \nendometriosis by demographic, anthropometric, and lifestyle fac-\ntors. Am J Epidemiol 160(8):784–796\n 13. Burghaus S et al (2011) Risk factors for endometriosis in a german \ncase-control study. Geburtsh Frauenheilk 71(12):1073–1079\n 14. Shah DK et al (2013) Body size and endometriosis: results from \n20 years of follow-up within the Nurses’ Health Study II prospec-\ntive cohort. Hum Reprod 28(7):1783–1792\n 15. Burghaus S et al (2016) The international endometriosis evalu-\nation program (IEEP Study)—a systematic study for physicians \nresearchers and patients. Geburtsh Frauenheilk 76(8):875–881\n 16. Hulsen T, de Vlieg J, Alkema W (2008) BioVenn, a web applica-\ntion for the comparison and visualization of biological lists using \narea-proportional Venn diagrams. BMC Genom 9:488\n 17. Velebil P et al (1995) Rate of hospitalization for gynecologic \ndisorders among reproductive-age women in the United States. \nObstet Gynecol 86(5):764–769\n 18. Houston DE et al (1988) The epidemiology of pelvic endometrio-\nsis. Clin Obstet Gynecol 31(4):787–800\n 19. Sangi-Haghpeykar H, Poindexter AN 3rd (1995) Epidemiol-\nogy of endometriosis among parous women. Obstet Gynecol \n85(6):983–992\n 20. Staal AH, van der Zanden M, Nap AW (2016) Diagnostic Delay \nof Endometriosis in the Netherlands. Gynecol Obstet Invest \n81(4):321–324\n 21. Soliman AM, Fuldeore M, Snabes MC (2017) Factors associ -\nated with time to endometriosis diagnosis in the United States. J \nWomens Health (Larchmt) 26(7):788–797\n 22. Parazzini F et al (2008) Risk factors for deep endometriosis: a \ncomparison with pelvic and ovarian endometriosis. Fertil Steril \n90(1):174–179\n 23. Templeman C et al (2008) Adenomyosis and endometriosis in the \nCalifornia Teachers Study. Fertil Steril 90(2):415–424\n 24. Maheshwari A et al (2012) Adenomyosis and subfertility: a sys-\ntematic review of prevalence, diagnosis, treatment and fertility \noutcomes. Hum Reprod Update 18(4):374–392\n 25. Vercellini P et al (2014) Uterine adenomyosis and in vitro ferti-\nlization outcome: a systematic review and meta-analysis. Hum \nReprod 29(5):964–977\n 26. Harb HM et al (2013) The effect of endometriosis on in vitro fer -\ntilisation outcome: a systematic review and meta-analysis. BJOG \n120(11):1308–1320\n 27. Farland LV et al (2017) Associations among body size across \nthe life course, adult height and endometriosis. Hum Reprod \n32(8):1732–1742\n 28. Liu Y, Zhang W (2017) Association between body mass \nindex and endometriosis risk: a meta-analysis. Oncotarget \n8(29):46928–46936\n 29. Saha R, Marions L, Tornvall P (2017) Validity of self-reported \nendometriosis and endometriosis-related questions in a Swedish \nfemale twin cohort. Fertil Steril 107(1):174-178.e2\n 30. Vercellini P et al (1995) Adenomyosis at hysterectomy—a study \non frequency-distribution and patient characteristics. Hum Reprod \n10(5):1160–1162\n 31. Koninckx PR et al (1991) Suggestive evidence that pelvic endo-\nmetriosis is a progressive disease, whereas deeply infiltrat-\ning endometriosis is associated with pelvic pain. Fertil Steril \n55(4):759–765\n 32. Perello MF et al (2017) Endometriotic pain is associated with \nadenomyosis but not with the compartments affected by deep \ninfiltrating endometriosis. Gynecol Obstet Invest 82(3):240–246\n 33. Andres MP et al (2018) Transvaginal ultrasound for the diagnosis \nof adenomyosis: systematic review and meta-analysis. J Minim \nInvasive Gynecol 25(2):257–264\n 34. Dunselman GA et al (2014) ESHRE guideline: management of \nwomen with endometriosis. Hum Reprod 29(3):400–412\nPublisher's Note Springer Nature remains neutral with regard to \njurisdictional claims in published maps and institutional affiliations.","source_license":"CC0","license_restricted":false}