{"paper_id":"8149c84a-8eaa-4039-8ce6-68605047c7f3","body_text":"Loss of ARID1A/BAF250a-expression in\nendometriosis: a biomarker for risk of\ncarcinogenic transformation?\nEleftherios P Samartzis 1, Nicolas Samartzis 1, Aurelia Noske 2, Andre´ Fedier1,\nRosmarie Caduff 2, Konstantin J Dedes 1, Daniel Fink 1 and Patrick Imesch 1\n1Department of Gynecology, University Hospital Zurich, Zurich, Switzerland and 2Department of Pathology,\nUniversity Hospital Zurich, Zurich, Switzerland\nMutations of the tumor-suppressor gene ARID1A result in the loss of protein expression of the BRG-associated\nfactor 250a (BAF250a), a large subunit of transcription-regulating Human SWI/SNF complexes, which have an\nimportant role in the control of cell proliferation and tumor suppression. ARID1A mutations are particularly\nfrequent in endometriosis-associated ovarian clear cell and endometrioid carcinomas, and were recently\ndescribed as a possible key mechanism and early step in the transformation of endometriosis into cancer. Here,\nwe examined the immunohistochemical expression pattern of BAF250a in a tissue microarray including 74\nendometriosis and 30 endometrium samples. Ovarian cancer samples ( n ¼ 136) served as a control. Epithelial\nBAF250a expression was assessable in 90/104 (87%) and stromal BAF250a expression in 95/104 (91%) of the\nendometriosis, and endometrium cases due to lack of adequate tissue in some spots. Complete lack of\nBAF250a expression was observed in three endometriomas ( n ¼ 3/20, 15%) and one deep-infiltrating endo-\nmetriosis sample ( n ¼ 1/22, 5%), but in none of the peritoneal endometriosis ( n ¼ 0/16) and eutopic endometrium\nsamples ( n ¼ 0/30). A comparison of the mean immunoreactivity scores revealed a significantly lower\nexpression rate of BAF250a in endometriomas compared with normal endometrium ( Po0.0005), as well as\nperitoneal (P ¼ 0.003) and deep-infiltrating endometriosis ( P ¼ 0.02). Our data demonstrates that a complete loss\nof BAF250a expression is observable in some endometriotic lesions, especially in endometriomas. In addition,\nwe report that a partial loss of BAF250a expression is occurring in the form of cell clusters indicating a clonal\nloss of BAF250a expression in these cells. The loss of expression of the tumor-suppressor protein BAF250a in\nsome endometriomas possibly indicates a risk of malignant transformation in these cases, which could be of\nimportance in the determination of individual treatment strategies. However, its role and value as a prognostic\nparameter in endometriosis needs to be further studied.\nModern Pathology (2012) 25, 885–892; doi:10.1038/modpathol.2011.217; published online 3 February 2012\nKeywords:\nARID1A; BAF250a; endometriosis; ovarian carcinomas\nEndometriosis is an estrogen-dependent inflammatory\ndisease affecting 5–10% of women of reproductive\nage.\n1,2 It is defined by the presence of ectopic\nendometrium-like tissue, and is clinically character-\nized by chronic pelvic pain, dysmenorrhea, dyspar-\neunia, and infertility.\n1,2 Although endometriosis is a\nbenign disease, a clinical association with ovarian\ncancer, in particular ovarian clear cell carcinomas\nand endometrioid ovarian carcinomas, has repeat-\nedly been observed in the past years. 3–5 There is\ngrowing evidence, that endometriosis is linked to\novarian clear cell carcinomas and endometrioid\novarian carcinomas through distinct pathomechan-\nisms.\n5,6 One of them is possibly linked to mutations\nthat result in the subsequent loss of the tumor-\nsuppressor gene AT rich interactive domain 1A\n(ARID1A). Using whole-exome sequencing, recur-\nrent mutations in ARID1A were recently reported in\novarian clear cell carcinomas and endometrioid\novarian carcinomas by two independent groups.\n7,8\nWiegand et al 7 found mutations in ARID1A in 46%\nof 119 ovarian clear cell carcinomas and 30% of 33\nendometrioid ovarian carcinomas but in none of 76\nReceived 23 June 2011; revised 1 December 2011; accepted 9\nDecember 2011; published online 3 February 2012\nCorrespondence: Dr P Imesch, MD, University Hospital Zurich,\nDepartment of Gynecology, Frauenklinikstrasse 10, CH-8091\nZurich, Switzerland.\nE-mail: patrick.imesch@usz.ch\nModern Pathology (2012) 25, 885–892\n& 2012 USCAP, Inc. All rights reserved 0893-3952/12 $32.00 885\nwww.modernpathology.org\n\nhigh-grade serous carcinomas. Likewise, Jones et al8\nobserved ARID1A-mutations in 57% of 42 ovarian\nclear cell carcinomas. Using immunohistochemistry\nit was demonstrated that ARID1A mutations result\nin the loss of the protein BRG-associated factor 250a\n(BAF250a), which is a large subunit of transcription-\nregulating Human SWI/SNF complexes and has an\nimportant role in the control of cell proliferation and\ntumor suppression.\n7,9,10 Recently, Guan et al11 found\nsomatic ARID1A mutations in 10/25 uterine endo-\nmetrioid carcinomas, and could also confirm a\nstrong correlation to loss of BAF250a expression\nby immunohistochemistry.\nFurther evidence for the importance of mutations\nin ARID1A in the pathogenesis of endometriosis-\nassociated carcinomas could be seen in two cases,\nwhere loss of BAF250a protein expression was obser-\nvable simultaneously in ARID1A mutated tumors\nand contiguous atypical endometriosis but not in\ndistant endometriosis.\n7 Loss of BAF250a expression\ncould therefore represent a key mechanism and\nearly step in the transformation of endometriosis\ninto cancer.7,8 The aim of this study is to examine if\nloss of BAF250a expression is present in benign\nendometriotic tissue. Therefore, we examined a\nlarge cohort of endometriotic lesions as well as a\ncontrol cohort of ovarian cancer by immunohisto-\nchemistry. As loss of expression of this tumor-sup-\npressor gene is regarded to be an early and crucial\nmechanism in endometriosis-associated ovarian\ncarcinomas\n7,11 it could be an important indicator of\na risk for carcinogenic transformation.\nMaterials and methods\nPatients and Tissue Specimens\nThe study was approved by the local ethical com-\nmittee (StV 27–2009 and KEK-ZH-2010-0174/0).\nARID1A (BAF250a) expression was examined by\nimmunohistochemistry in two tissue microarrays,\none consisting of study cases (ectopic endometriotic\ntissue and eutopic endometrium) and one of control\ncases (different ovarian carcinoma subtypes). Clin-\nical and pathological characteristics for the study\nand the control tissue microarray were collected\nfrom the clinical database of the gynecological\ndepartment and the pathology records of the\nUniversity Hospital of Zurich.\nThe study cases included 74 samples of 71 patients\nwith ectopic typical endometriotic tissue (19 peri-\ntoneal, 28 deep infiltrating and 27 ovarian) under-\ngoing laparoscopic endometriosis treatment as well\nas 30 samples of healthy eutopic endometrium from\npatients undergoing hysterectomies for predomi-\nnantly benign conditions such as uterine myomas\nor prolapsed uterus. All tissue samples originated\nfrom premenopausal patients. The diagnosis of\nendometriosis was initially done at the Institute of\nPathology of the University Hospital Zurich between\n2000 and 2010, and all included cases were reviewed\nfor this study by a gynecologic pathologist.\nThe control cases consisted of 136 primary epi-\nthelial ovarian carcinomas (66 serous, 24 clear cell,\n31 endometrioid and 15 mucinous) as well as 3\nsamples of the fallopian tube. These patients were\nsurgically treated at the University Hospital of\nZurich (Switzerland) between 1995 and 2005. The\nstage of tumors was assessed according to the\nInternational Federation of Gynecology and Obste-\ntrics (FIGO) staging system. The tumor grade and\nhistological subtype were defined according to the\nWHO classification 2003.\nTissue Microarray\nThe tissue microarrays were constructed using a\nsemiautomatic tissue arrayer (Beecher Instruments,\nWoodland, TX, USA) as previously described.\n12 Tissue\nsamples were fixed in 4% neutral buffered formal-\ndehyde and were embedded in paraffin. Routine\nhematoxylin/eosin-stained sections were performed\nfor histopathological evaluation. Areas involving\nendometriotic or tumor tissue were marked respec-\ntively. Samples were further selected under the term\nof tissue availability. T wo cylindrical tissue cores of\n0.6 mm diameter were punched out from each case\nof the corresponding paraffin-embedded blocks and\ninserted into the recipient blocks.\nThe study tissue microarray block was freshly cut\nin two-and-a-half micrometer sections and mounted\non superfrost slides (Menzel Gla ¨ser, Braunschweig,\nGermany). After antigen retrieval with CC1m- heat-\ninduced epitope retrieval, the slides were incubated\nwith a monoclonal mouse ARID1A antibody (AB-\nGENT, code: AT1188a), as previously validated,\n7,11,13\nwhich was diluted 1:200 in Ventana dilution buffer.\nAfter incubation for 1 h at room temperature, the\nstaining was further conducted with the Ventana\nBenchmark automated system (Ventana Medical\nSystems, Tuscon, AZ, USA). Slides were counter-\nstained with haematoxylin, dehydrated and mounted.\nThe expression analysis was performed by two\nauthors (AN, NS) and was evaluated according to\nthe percentage of positive cells and the staining\nintensity. Nuclear immunoreactivity was considered\nas a positive expression. Endothelial cells served as\npositive internal controls. All BAF250a negative\nendometriotic cases were confirmed on whole tissue\nsections. Negative controls were performed by\nreplacement of the primary antibody by non-reactive\nrabbit IgG (DAKO). Identical procedure was applied\nfor the control tissue microarray block.\nScoring and Statistical Analysis\nThe percentage of positive cells was scored as: 0\n(0%); þ (r10%); þþ (11–50%); þþþ (51–80%);\nand þþþþ (480%). The staining intensity was\ndefined as: 0* (negative), 1* (weak), 2* (moderate)\nARID1A/BAF250a deficiency in endometriosis\n886 EP Samartzis et al\nModern Pathology (2012) 25, 885–892\n\nand 3* (strong). For the immunoreactive score, 14 the\npercentage of positive cells and staining intensity\nwas multiplicated, resulting in a value between 0\nand 12.\nSPSS software (Version 19.0, SPSS, Chicago, IL,\nUSA) was used for the statistical evaluation. Generally,\nP-valueso0.05 were considered as significant. Values\nare indicated as percentage and absolute numbers,\nas mean and standard deviation ( ±s.d.) or as median\nand interquartile range (IQR), respectively. BAF250a\nexpression rates were compared by a w\n2 test for\ntrends. Immunoreactivity scores were compared by\na Mann–Whitney test or a Kruskal–Wallis test where\nappropriate. Scoring and statistical methods were\nidentical in the study and control cases.\nThe statistical significance of the association\nbetween BAF250a expression and clinico-patholo-\ngical parameters was assessed by a w\n2 test for trends,\na Fisher’s exact test or a Spearman’s r test (bivariat\ncorrelation analysis) where appropriate.\nResults\nPatients with ovarian endometriosis had a mean age\n(±s.d., range) of 35 ( ±7, 19–48) years, with peri-\ntoneal endometriosis 31 ( ±4, 25–38) years and with\ndeep infiltrating endometriosis 35 ( ±5, 25–42)\nyears. The average age of the patients with normal\nuterine endometrium samples was 40 ( ±3, 29–44)\nyears. Clinical endometriosis severity score follow-\ning the American Society for Reproductive Medicine\n(ASRM) was known for 58/71 (82%) endometriosis\npatients, with following distribution: ASRM-I, 4/71\n(6%); ASRM-II, 7/71 (10%); ASRM-III, 15/71 (21%);\nand ASRM-IV , 32/71 (45%). Three patients in the\nstudy group had a diagnosis of malignant disease.\nT wo of them were cervical carcinomas and were in\nthe eutopic endometrium group and one was a\ncervical carcinoma with simultaneous pulmonary\nadenocarcinoma and was in the deep-infiltrating\nendometriosis group. In particular, no ovarian\ncarcinomas were observed in the study group prior\nor during the study time. However, all endometrio-\nsis cases were from relatively recent time and hence\nno longer clinical follow-up was available.\nIn the control group with ovarian tumors, the\nmean patient age at the time of surgery was 59 ( ±14)\nyears, ranging from 20 to 87. The median follow-up\ntime of the surviving patients was 59 months. FIGO\nstages were known in 131 cases as follows: I, 33/131\n(25%); II, 10/131 (8%); III, 50/131 (38%); and IV , 38/\n131 (29%). Tumor grades were: G1, 22/136 (16%);\nG2, 58/136 (43%); and G3, 56/136 (41%). Residual\ndisease status was reported in 90 cases, thereof\no1 cm for 26 (29%) cases, 1–2 cm for 27 (30%) and\n42 cm for 37 (41%). Data on adjuvant chemotherapy\nwas known for all patients and in the majority\na platinum-based combination therapy was\nadministered.\nImmunohistochemical staining of BAF250a\n(epithelial and/or stromal) was assessable in 102/\n104 (98%) of the study cases (endometriosis and\nendometrium) due to the lack of adequate tissue in\nsome tissue microarray spots. Epithelial BAF250a\nexpression was assessable in 90/104 (87%) of the\ncases (endometriomas 21/27 (78%), peritoneal\nendometriosis 17/19 (90%),deep-infiltrating endome-\ntriosis 22/28 (79%) and eutopic endometrium 30/30\n(100%)). Stromal BAF250a expression was assessable\nTable 1 Epithelial and stromal BAF250a ( ARID1A) expression\nTissue type BAF250a epithelial Total\n0 + ++ +++ ++++\nEndometrium ( P ¼ 0.001)a\nBAF250a stromal\n00 1 0 0 1 2\n+0 0 0 1 0 1\n++ 0 0 0 1 8 9\n+++ 0 0 0 0 5 5\n++++ 0 0 0 0 13 13\nTotal 0 1 0 2 27 30\nEndometriosis (ovarian) ( P ¼ 0.002)\na\nBAF250a stromal\n03 0 0 0 0 3\n+0 0 0 1 0 1\n++ 0 0 1 0 2 3\n+++ 0 0 1 2 5 8\n++++ 0 0 0 0 5 5\nTotal 3 0 2 3 12 20\nEndometriosis (peritoneal) ( P ¼ 0.003)\na\nBAF250a stromal\n00 0 0 1 1 2\n+0 0 1 0 0 1\n++ 0 0 0 0 4 4\n+++ 0 0 0 0 3 3\n++++ 0 0 0 0 6 6\nTotal 0 0 1 1 14 16\nEndometriosis (deep) ( P ¼ 0.001)\na\nBAF250a stromal\n01 0 0 0 0 1\n+0 0 0 0 0 0\n++ 0 0 0 0 7 7\n+++ 0 0 0 0 9 9\n++++ 0 0 1 0 4 5\nTotal 1 0 1 0 20 22\nTotal (P ¼ 0.001)\na\nBAF250a stromal\n04 1 0 1 2 8\n+0 0 1 2 0 3\n++ 0 0 1 1 21 23\n+++ 0 0 1 2 22 25\n++++ 0 0 1 0 28 29\nTotal 4 1 4 6 73 88\n0, 0% of the cells stained positive; +, up to 10% of the cells; ++, 11–\n50% of the cells; +++, 51–80%; ++++, 4 80% of the cells.\nRepresentation of only the cases that were simultaneously evaluable\nfor epithelial and stromal BAF250a ( ARID1A) expression. Cases that\ncould only be evaluated for either epithelial or stromal expression due\nto lack of adequate tissue in the sample are not included in this table.\naw2 test.\nARID1A/BAF250a deficiency in endometriosis\nEP Samartzis et al 887\nModern Pathology (2012) 25, 885–892\n\nin 95/104 (91%) of the cases (endometriomas 24/27\n(89%), peritoneal endometriosis 16/19 (84%), deep-\ninfiltrating endometriosis 25/28 (89%) and eutopic\nendometrium 30/30 (100%)).\nComplete lack of epithelial as well as stromal\nBAF250a expression was observed in three endome-\ntriomas ( n ¼ 3/20, 15%) and one deep-infiltrating\nendometriosis sample ( n ¼ 1/22, 5%), and was con-\nfirmed by whole tissue sections. There were no\nnegative cases in peritoneal endometriosis and in\neutopic endometrium.\nEpithelial and stromal BAF250a expression corre-\nlated strongly in all endometriosis types as well as\nin eutopic endometrium. In the cases with hetero-\ngeneous BAF250a expression, loss of BAF250a\nexpression principally occurred in large groups of\nadjacent endometriotic cells, whereas other cell\ngroups in the same sample clearly showed homo-\ngeneous positivity for BAF250a. This phenomenon\nis also reported as ‘cell clusters’ or ‘clonal loss’ in\nthe discussion part of this work. The expression\npattern of BAF250a in epithelial and stromal\nendometriosis and uterine endometrium cells is\nreported in detail in Table 1. Representative images\nare shown in Figures 1a–d.\nThe median immunoreactivity score for epithelial\nand stromal BAF250a expression in endometriotic\nlesions was 4 and 3, respectively. The median immu-\nnoreactivity score for epit helial BAF250a expression\nin ectopic endometriosis samples and in eutopic\nendometrium were 4 (IQR, 2–6) and 6 (IQR, 4–8),\nrespectively ( P ¼ 0.008). Considering each endome-\ntriosis subgroup, median immunoreactivity score of\nepithelial BAF250a expression was 2 (IQR, 2–4) in\novarian endometriosis, 6 (IQR, 3.5–8) in peritoneal\nendometriosis and 4 (IQR, 3–6) in deep-infiltrating\nendometriosis. Expressionrates of epithelial BAF250a\nwere significantly lower in ovarian endometriosis\ncompared with eutopic endometrium ( Po0.0005),\nto peritoneal endometriosis ( P ¼ 0.003) and to deep-\ninfiltrating endometriosis ( P ¼ 0.02), as represented\nin Figure 2.\nImmunohistochemical analysis was successful in\n129/136 ovarian carcinomas owing to the lack of\nFigure 1 BAF250a expression in endometriosis. Immunohistochemical staining of ARID1A/BAF250a in endometriosis compared with\nnormal endometrium. ( a and b): Examples of complete loss of BAF250a expression in epithelium and adjacent stroma of ovarian\nendometrioma ( a, /C2 200) as well as deep-infiltrating endometriosis ( b, /C2 200). ( c) Cluster-like epithelial deficiency of BAF250a\nexpression in an ovarian endometrioma ( /C2 200). ( d) Strong nuclear BAF250a expression in normal endometrium ( /C2 200).\nARID1A/BAF250a deficiency in endometriosis\n888 EP Samartzis et al\nModern Pathology (2012) 25, 885–892\n\ntumor tissue in some tissue microarray spots. We\nobserved a completely negative expression of BAF250a\nin 29 (23%) of the 129 carcinomas. Regarding the\ndifferent carcinoma subtypes 22% ( n ¼ 5/23) of the\novarian clear cell carcinomas, 46% ( n ¼ 13/28) of\nthe endometrioid ovarian carcinomas, 27% (n ¼ 4/15)\nof the mucinous and 11% ( n ¼ 7/63) of the serous\novarian carcinomas stained completely BAF250a\nnegative. Representative images are shown in\nFigures 3a–d. Epithelium of the fallopian tube\nstained BAF250a positive in all evaluable cases\n(n ¼ 3). We found that negative BAF250a expression\nwas significantly associated with clinico-pathologi-\ncal factors like FIGO stage, tumor- and nodal stage,\nendometrioid histological subtype, intraoperative\nresidual disease and patient age but not with tumor\ndifferentiation (grade) as indicated in Table 2.\nDiscussion\nTransformation of endometriosis into cancer is a rare\nevent.15 The pathogenesis, however, is still not well\nunderstood.5,16 Historically, Sampson17 reported the\nfirst case of malignant transformation of endome-\ntriosis into an ovarian carcinoma. In the past years\nthere is growing evidence for the relevance of a\npathogenic link between these two diseases. 6\nSeveral studies indicated that in women with\novarian endometriosis there is an increased risk for\nthe development of epithelial ovarian cancer, espe-\ncially the histological subtypes of ovarian clear cell\ncarcinomas and endometrioid ovarian carcino-\nmas.4,16,18–22 Brinton et al 18 reported a 4.2-fold risk\nfor ovarian cancer in cases of long-standing history\nof ovarian endometriosis in a cohort of 20 686 women\nhospitalized with endometriosis. Melin et al 19 con-\nfirmed an increased risk for ovarian cancer in a\ncohort of 64 492 women in cases with long-standing\nendometriosis. Kobayashi et al4 reported 46 cases of\novarian carcinomas in a large prospective popula-\ntion-based Japanese study with an ovarian endome-\ntriosis cohort of 6398 women and a follow-up of\nup to 17 years. It was shown that women with\novarian endometriomas in this cohort had a ninefold\nincreased risk to develop ovarian cancer.3,4 As shown\non the other hand by Vercellini et al20 in 556 patients\nthat underwent surgery for ovarian cancers, the fre-\nquency of simultaneous endometriosis was also\nsignificantly higher with 26.3, 21.1 and 22.2% in\ncases with endometrioid ovarian carcinomas, ovarian\nclear cell carcinomas or mixed subtypes, respec-\ntively, compared with 3.6–5.6% in serous, mucinous\nand miscellaneous ovarian cancers. In another study\nwith 127 patients with primary ovarian car-\ncinomas endometriosis rates of 70% were reported\nfor ovarian clear cell carcinomas (30/43) and 43%\nfor endometrioid ovarian carcinomas (3/7), whereas\nonly 7% of the serous carcinomas (4/60) and none of\nthe mucinous carcinomas (0/17) showed simulta-\nneous endometriosis. Moreover, the authors described\n29 cases of atypical endometriosis in this cohort and\ndemonstrated evidence for the transition from atypical\nendometriosis to carcinoma in 23 cases. 21 Another\nstudy reported an endometriosis rate of 37% in a\nlarge series of 221 patients with endometrioid ovarian\ncarcinomas.23\nEpidemiological data indicates an early onset of\nendometriosis-associated ovarian carcinomas such\nas ovarian clear cell carcinomas and endometrioid\novarian carcinomas.3,4,24 In western countries ovarian\nclear cell carcinomas account for 12% of epithelial\novarian carcinomas and endometrioid ovarian car-\ncinomas for 11%, whereas serous carcinomas are the\nmost common subtype with 68%.25 In Asian countries\nas in Japan, the percentage of ovarian clear cell\ncarcinomas is even higher with more than 20%, 26\nwhich is remarkable as studies report a higher pre-\nvalence of endometriosis in Asian women. 27 Ovar-\nian clear cell carcinomas are relatively resistant to\nconventional platinum- or taxane-based therapy and\nare associated with a poor prognosis especially in\nmore advanced stages. 3,26,28 Identification of biomar-\nkers for an increased risk of malignant transforma-\ntion in endometriosis could improve early detection\nof endometriosis-associated epithelial ovarian can-\ncers. Lack of BAF250a, as a potential early event in\nthe carcinogenesis, could possibly serve as screen-\ning for endometriotic lesions at high-risk for the\ndevelopment of ovarian clear cell carcinomas or\nendometrioid ovarian carcinomas. 29\nWe used immunohistochemical staining of BAF250a\n(ARID1A) as a surrogate marker for ARID1A muta-\ntions as previously validated, 7,11,13 as the small\nendometriotic samples obtained after laparoscopic\nbiopsies would not yield a sufficient amount of DNA\nfor screening ARID1A mutations after microdissection\nFigure 2 Immunoreactivity score in endometriosis and endome-\ntrium. The immunoreactivity score 14 for BAF250a was signifi-\ncantly lower in ovarian endometriosis compared with eutopic\nendometrium (Po0.0005), to peritoneal endometriosis ( P ¼ 0.003)\nand to deep-infiltrating endometriosis ( P ¼ 0.02), as represented\nin this chart diagram (calculated with a Kruskal–Wallis test).\nARID1A/BAF250a deficiency in endometriosis\nEP Samartzis et al 889\nModern Pathology (2012) 25, 885–892\n\nusing Sanger sequencing. We found complete lack-\ning of BAF250a expression in three of the examined\nendometriomas and in one deep-infiltrating endo-\nmetriotic lesion. Moreover, we observed that in most\nof the partly negative BAF250a cases lack of expres-\nsion occurred in the form of grouped cell clusters.\nAccording to Guan et al ,11 this cluster-like negative\nexpression pattern indicates a clonal loss of\nBAF250a in the concerned cells and happens only\nin ARID1A mutated clones.\nRecent studies support that serous ovarian tumors\narise from implantation of epithelium (benign or\nFigure 3 BAF250a expression in ovarian carcinomas. Immunohistochemical staining of ARID1A/BAF250a in ovarian carcinomas.\n(a) Example of strong nuclear expression in a high-grade serous carcinoma ( /C2 200). ( b) Moderate intense expression in a mucinous\ncarcinoma ( /C2 200). (c–e) Loss of expression in an endometrioid carcinoma ( c, /C2 200), in a clear cell carcinoma ( d, /C2 200), in a mucinous\ncarcinoma ( e, /C2 200) and a serous carcinoma ( f, /C2 200). Note that only stromal cells show immunoreactivity.\nARID1A/BAF250a deficiency in endometriosis\n890 EP Samartzis et al\nModern Pathology (2012) 25, 885–892\n\nmalignant) from the fallopian tube. 6 This theory also\nsupports the fact that ovarian clear cell carcinomas\nand endometrioid ovarian carcinomas represent a\ndifferent entity and that ovarian endometriotic\ninclusion cysts (endometriomas) are regarded as\nprecursor of these tumors, originating from retro-\ngrade menstruation.\n6 There is evidence that eutopic\nendometrium in women with endometriosis exhi-\nbits intrinsic molecular abnormalities such as\nactivation of oncogenic pathways.\n1,6 Some epide-\nmiological studies demonstrated a lower risk for\novarian clear cell carcinomas and endometrioid\novarian carcinomas after tubal ligation or hyster-\nectomy.\n30–32 A recent proposition for an ovarian\ncancer risk-assessment tool integrated endometrio-\nsis as well as no tubal ligation each as one of totally\neight conditions associated with ovarian cancer.\n33\nARID1A mutations are supposed to have a key\nrole in the formation of endometriosis-associated\novarian carcinomas and are probably occurring\nearly, already at the stage of endometriosis.\n7 Ovarian\ncancer arises from the inclusion of epithelial cells\ninto the ovarian stromal tissue. 6 It seems evident\nthat cluster-like ARID1A mutated clones in such\ninclusion cysts could therefore have a key role in the\nmalignant transformation of these cysts. 6,7,11 We\ndemonstrate that complete loss of BAF250a expres-\nsion occurs in a minority of endometriosis cases,\nmainly in endometriomas, without any evidence of\nmalignant diagnosis at the moment of the operation.\nAdditionally, we demonstrate that cluster-like loss\nof BAF250a expression can be shown in some\nfurther cases, implicating a clonal BAF250a expres-\nsion deficiency.\n11 By these results, we corroborate\nthat BAF250a expression deficiency in endometrio-\nsis is not restricted to tumor contiguous lesions\nand that loss of BAF250a expression can be obser-\nved already in benign endometriotic lesions, and\ncould therefore indicate a risk of malignant trans-\nformation.\nHowever, many questions need still to be\nanswered, principally about the clinical consequences\nand the reasonability of a BAF250a screening in\n(ovarian) endometriosis. Further prospective and\nlarger clinical studies are needed to answer these\nquestions.\nDisclosure/conflict of interest\nThe authors declare no conflict of interest.\nAcknowledgements\nThis study was supported by Grants from the EMDO-\nand Hartmann-Mueller-Foundations. We thank\nSilvia Behnke and Martina Storz for their excellent\ntechnical assistance. We also thank Prof B Seifert,\nPhD, Institute for Biostatistics University of Zurich,\nSwitzerland, for the statistical support in this study.\nReferences\n1 Bulun SE. Endometriosis. N Engl J Med 2009;360:\n268–279.\n2 Giudice LC, Kao LC. Endometriosis. Lancet 2004;364:\n1789–1799.\n3 Anglesio MS, Carey MS, Ko ¨ bel M, et al. Clear cell\ncarcinoma of the ovary: a report from the first Ovarian\nClear Cell Symposium, June 24th, 2010. Gynecol Oncol\n2011;121:407–415.\n4 Kobayashi H, Sumimoto K, Moniwa N, et al. Risk of\ndeveloping ovarian cancer among women with ovarian\nendometrioma: a cohort study in Shizuoka, Japan. Int J\nGynecol Cancer 2007;17:37–43.\n5 Ness RB. Endometriosis and ovarian cancer: thoughts\non shared pathophysiology. Am J Obstet Gynecol\n2003;189:280–294.\n6 Kurman RJ, Shih IM. The origin and pathogenesis of\nepithelial ovarian cancer: a proposed unifying theory.\nAm J Surg Pathol 2010;34:433–443.\nTable 2 Association of BAF250a ( ARID1A) expression with\nclinico-pathological features in ovarian carcinomas\nCharacteristics All cases\nn (%)\nBAF250a\nexpression n (%)\nP-valuea\nNegative Positive\nAll carcinomas 129 (100%) 29 (23) 100 (78)\nAge at diagnosis 0.004\np60 years 63 21 (33) 42 (67)\n460 years 66 8 (12) 58 (88)\nHistological type 0.003\nClear cell 23 5 (22) 18 (78)\nEndometrioid 28 13 (46) 15 (54)\nSerous 63 7 (11) 56 (89)\nMucinous 15 4 (27) 11 (73)\nTumor stage ( n ¼ 128) 0.008\nPT1 31 10 (32) 21 (68)\nPT2 15 7 (47) 8 (53)\nPT3 82 12 (15) 70 (85)\nNodal stage ( n ¼ 94) 0.045\n0 60 18 (30) 42 (70)\n1 34 4 (12) 30 (88)\nFIGO stage ( n ¼ 123) 0.005\n1 30 11 (37) 19 (63)\n2 9 5 (56) 4 (44)\n3 46 8 (17) 38 (83)\n4 38 4 (11) 34 (90)\nTumor grade 0.201\n1 23 8 (28) 15 (65)\n2 51 12 (24) 39 (77)\n3 55 9 (16) 46 (84)\nResidual disease ( n ¼ 87) 0.006\nR0 25 10 (40) 15 (60)\nR1 27 6 (22) 21 (78)\nR2 34 2 (6) 32 (94)\naw2 test.\nARID1A/BAF250a deficiency in endometriosis\nEP Samartzis et al 891\nModern Pathology (2012) 25, 885–892\n\n7 Wiegand KC, Shah SP , Al-Agha OM, et al. ARID1A\nmutations in endometriosis-associated ovarian carci-\nnomas. N Engl J Med 2010;363:1532–1543.\n8 Jones S, Wang TL, Shih IM, et al. Frequent mutations\nof chromatin remodeling gene ARID1A in ovarian clear\ncell carcinoma. Science 2010;330:228–231.\n9 Wang X, Nagl Jr NG, Flowers S, et al. Expression of\np270 (ARID1A), a component of human SWI/SNF\ncomplexes, in human tumors. Int J Cancer 2004;\n112:636.\n10 Reisman D, Glaros S, Thompson EA. The SWI/SNF\ncomplex and cancer. Oncogene 2009;28:1653–1668.\n11 Guan B, Mao TL, Panuganti PK, et al. Mutation and\nloss of expression of ARID1A in uterine low-grade\nendometrioid carcinoma. Am J Surg Pathol 2011;35:\n625–632.\n12 Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue\nmicroarrays for high-throughput molecular profiling of\ntumor specimens. Nat Med 1998;4:844–847.\n13 Maeda D, Mao T-L, Fukayama M, et al. Clinicopatho-\nlogical significance of loss of ARID1A immunoreactiv-\nity in ovarian clear cell carcinoma. Int J Mol Sci\n2010;11:5120–5128.\n14 Weichert W, Roske A, Gekeler V , et al. Association of\npatterns of class I histone deacetylase expression with\npatient prognosis in gastric cancer: a retrospective\nanalysis. Lancet Oncol 2008;9:139–148.\n15 Somigliana E, Vigano’ P , Parazzini F , et al. Association\nbetween endometriosis and cancer: a comprehensive\nreview and a critical analysis of clinical and epide-\nmiological evidence. Gynecol Oncol 2006;101:\n331–341.\n16 Vlahos NF , Kalampokas T, Fotiou S. Endometriosis\nand ovarian cancer: a review. Gynecol Endocrinol\n2010;26:213–219.\n17 Sampson JA. Endometrial carcinoma of the ovary,\narising in endometrial tissue in that organ. Arch Surg\n1925;10:1–72.\n18 Brinton LA, Gridley G, Persson I, et al. Cancer risk after\na hospital discharge diagnosis of endometriosis. Am J\nObstet Gynecol 1997;176:572–579.\n19 Melin A, Spare ´n P , Persson I, Bergqvist A. Endome-\ntriosis and the risk of cancer with special emphasis on\novarian cancer. Hum Reprod 2006;21:1237–1242.\n20 Vercellini P , Parazzini F , Bolis G, et al. Endometriosis\nand ovarian cancer. Am J Obstet Gynecol 1993;169:\n181–182.\n21 Ogawa S, Kaku T, Amada S, et al. Ovarian endome-\ntriosis associated with ovarian carcinoma: a clinico-\npathological and immunohistochemical study.\nGynecol Oncol 2000;77:298–304.\n22 Kumar S, Munkarah A, Arabi H, et al. Prognostic\nanalysis of ovarian cancer associated with endome-\ntriosis. Am J Obstet Gynecol 2011;204:63e1–63e7.\n23 Lim MC, Chun KC, Shin SJ, et al. Clinical presentation\nof endometrioid epithelial ovarian cancer with con-\ncurrent endometriosis: a multicenter retrospective\nstudy. Cancer Epidemiol Biomarkers Prev 2010;19:\n398–404.\n24 Aris A. Endometriosis-associated ovarian cancer: a\nten-year cohort study of women living in the Estrie\nRegion of Quebec, Canada. J Ovarian Res 2010;3:2.\n25 Ko¨ bel M, Kalloger SE, Huntsman DG, et al. Differences\nin tumor type in low-stage versus high-stage ovarian\ncarcinomas. Int J Gynecol Pathol 2010;29:203–211.\n26 Itamochi H, Kigawa J, Terakawa N. Mechanisms of\nchemoresistance and poor prognosis in ovarian clear\ncell carcinoma. Cancer Sci 2008;99:653–658.\n27 Jacoby VL, Fujimoto VY, Giudice LC, et al. Racial and\nethnic disparities in benign gynecologic conditions\nand associated surgeries. Am J Obstet Gynecol 2010;\n202:514–521.\n28 Sugiyama T, Kamura T, Kigawa J, et al. Clinical\ncharacteristics of clear cell carcinoma of the ovary: a\ndistinct histologic type with poor prognosis and\nresistance to platinum-based chemotherapy. Cancer\n2000;88:2584–2589.\n29 Birrer MJ. The origin of ovarian cancer—is it getting\nclearer? N Engl J Med 2010;363:1574–1575.\n30 Rosenblatt KA, Thomas DB. Reduced risk of ovarian\ncancer in women with a tubal ligation or hysterectomy.\nThe World Health Organization Collaborative Study of\nNeoplasia and Steroid Contraceptives. Cancer Epide-\nmiol Biomarkers Prev 1996;5:933–935.\n31 Miracle-McMahill HL, Calle EE, Kosinski AS, et al.\nTubal ligation and fatal ovarian cancer in a large pro-\nspective cohort study . Am J Epidemiol 1997;145:349–357.\n32 Hankinson SE, Hunter DJ, Colditz GA, et al. Tubal\nligation, hysterectomy, and risk of ovarian cancer.\nA prospective study. JAMA 1993;270:2813–2818.\n33 Vitonis AF , Titus-Ernstoff L, Cramer DW. Assessing\novarian cancer risk when considering elective oophor-\nectomy at the time of hysterectomy. Obstet Gynecol\n2011;117:1042–1050.\nARID1A/BAF250a deficiency in endometriosis\n892 EP Samartzis et al\nModern Pathology (2012) 25, 885–892","source_license":"CC0","license_restricted":false}