{"paper_id":"48853fc3-fcbc-4564-b6fb-8360a51779e5","body_text":"CCNE1 copy-number gain and overexpression\nidentify ovarian clear cell carcinoma with a\npoor prognosis\nAyse Ayhan1,2,3,4, Elisabetta Kuhn5,R e n - C h i nW u6, Hiroshi Ogawa2, Asli Bahadirli-Talbott1,\nTsui-Lien Mao7, Haruhiko Sugimura 3, Ie-Ming Shih 1,8 and Tian-Li Wang 1,8\n1Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD, USA; 2Department of\nPathology, Seirei Mikatahara Hospital, Hamamatsu, Japan; 3Department of Tumor Pathology, Hamamatsu\nUniversity School of Medicine, Hamamatsu, Japan; 4Department of Molecular Pathology, Hiroshima\nUniversity School of Medicine, Hiroshima, Japan; 5Pathology Unit, Arcispedale S. Maria Nuova-IRCCS,\nReggio Emilia, Italy; 6Department of Pathology, Chang Gung Memorial Hospital and Chang Gung University\nCollege of Medicine, Taoyuan, Taiwan; 7Department of Pathology, National Taiwan University College of\nMedicine, Taipei City, Taiwan and 8Department of Gynecology and Obstetrics, Johns Hopkins Medical\nInstitutions, Baltimore, MD, USA\nOvarian clear cell carcinoma is a unique type of ovarian cancer, often derived from endometriosis, and advanced-\nstage disease has a dismal prognosis primarily due to the resistance to conventional chemotherapy. Previous\nstudies have shown frequent somatic mutations in ARID1A, PIK3CA, hTERT promoter, and amplification of\nZNF217; however, the molecular alterations that are associated with its aggressiveness remain largely unknown.\nThis study examined and compared cyclin E1 expression in endometriosis-related ovarian tumors, with the aim\nof determining the relationship between hTERT mutations and ARID1A expression and evaluating the effects of\nthese molecular alterations on patient survival. We performed immunohistochemistry on 207 tumors [clear cell\ncarcinoma (n = 120), endometrioid carcinoma ( n = 49), and seromucinous tumors ( n = 38)], followed by two-color\nfluorescence in situ hybridization ( n = 88) and compared with ARID1A expression and hTERT promoter\nmutations in the same samples. Cyclin E1 overexpression and CCNE1 copy-number gain occurred in 23.3%\nand 14.8% of ovarian clear cell carcinomas, respectively, but they were not detected in any of the other\nendometriosis-related tumors. All cases with CCNE1 copy-number gain demonstrated an intense cyclin E1\nimmunoreactivity ( Po 0.001). Cyclin E1 overexpression was positively correlated with hTERT promoter\nmutations ( P = 0.01), but not with the loss of ARID1A expression. A multivariate analysis revealed that CCNE1\noverexpression predicts poor overall survival, even after adjusting for stage and age. Specifically, CCNE1\noverexpression and copy-number gain were both correlated with a poor outcome in patients with stage I disease.\nMoreover, the subset with CCNE1 overexpression and ARID1A retention demonstrated the worst outcome. Our\nfindings suggest that gene copy-number gain and upregulation of CCNE1 occur in ovarian clear cell carcinoma\nand are associated with a worse clinical outcome, dictating the survival of early-stage patients, and that these\nmolecular alterations are unique to clear cell carcinoma among different types of endometriosis-related ovarian\nneoplasms.\nModern Pathology (2017) 30, 297–303; doi:10.1038/modpathol.2016.160; published online 21 October 2016\nOvarian cancer is a heterogeneous disease that\ncomprises at least five histological subtypes, which\nare characterized by distinct clinicopathological and\nmolecular features as well as by their cellular\norigins.\n1 Among them, ovarian clear cell carcinoma,\nendometrioid carcinoma, and seromucinous neo-\nplasms represent a unique group of diseases because\nthey are often associated with endometriosis.\nBecause most of these cancers arise from pre-\nexisting ovarian endometriotic cysts, it has been\nsuggested that the inflammatory and oxidative stress\nassociated with the endometriosis microenviron-\nment promote mutagenesis and facilitate tumor\nCorrespondence: Adjunct Professor Dr A Ayhan, MD, PhD or\nProfessor T-L Wang, PhD, Department of Pathology, Johns Hopkins\nMedical Institutions, 1550 Orleans Street, Baltimore, MD 21231, USA.\nE-mail: ayseayhanjp@gmail.com or tlw@jhmi.edu\nReceived 11 May 2016; revised 29 July 2016; accepted 30 July\n2016; published online 21 October 2016\nModern Pathology (2017) 30, 297 –303\n© 2017 USCAP, Inc All rights reserved 0893-3952/17 $32.00 297\nwww.modernpathology.org\n\ndevelopment.2 Compared with other types of ovarian\ncarcinoma, endometriosis-related ovarian cancers\nare characterized molecularly by aberrations in the\nPI3K–AKT pathway and frequent somatic mutations\ninvolving ARID1A, a tumor suppressor gene that\nparticipates in the SWI/SNF chromatin remodeling\ncomplex.\n3 Although endometriosis-related neoplasms\nshare several common features, they also differ in\nmany respects. For example, compared with endome-\ntrioid carcinoma, clear cell carcinoma is characterized\nby unique histopathological characteristics, resistance\nto platinum-based chemotherapy, presence of activat-\ning mutations in the hTERT promoter, and expression\nof specific biomarkers including HNF-1 β,N a p s i nA ,\nand α-methylacyl-coenzyme A racemase (AMACR,\nP504S).\n4–7 Genetically engineered mouse models have\ndemonstrated that the co-deletion of the ARID1A and\nPTEN genes induces ovarian endometrioid carcinoma\nin the murine ovary, whereas the deletion of ARID1A\nand the expression of a PIK3CA activating mutant\ngenerate ovarian clear cell-like tumors. 8,9\nIn a previous study, we identified DNA copy-\nnumber gain at the CCNE1 locus in a small set\nof ovarian clear cell carcinomas. 10 CCNE1 encodes\nthe cyclin E1 protein of the highly conserved\n‘cyclin’ family of proteins, which are involved in\ncell cycle regulation. Cyclin E1 forms a complex\nwith the regulatory subunit, cyclin-dependent\nkinase-2 (Cdk2), and promotes transition of the G1\nto S phase of the cell cycle. Although normal cells\nmaintain strict control of cyclin E1 activity, cancer\ncells can exploit the upregulation of this protein as\na way to promote tumor cell replication. Several\nstudies have reported that CCNE1 gene amplification\nor protein upregulation is associated with higher\ntumor grades and with a worse clinical outcome in\na variety of cancers.\n11–13 For example, increased\nDNA copy number and overexpression of CCNE1 are\nobserved in uterine serous carcinoma, a high-grade\nuterine cancer, but not in the more indolent uterine\nendometrioid carcinoma.\n14–16 Similarly, CCNE1\namplification characterizes ovarian high-grade, but\nnot low-grade, serous carcinomas.\n17–19 To further\nexplore the significance of CCNE1 in endometriosis-\nrelated ovarian cancers, we performed immuno-\nhistochemistry and fluorescence in situ hybridiza-\ntion (FISH) analyses of CCNE1.\nMaterials and methods\nTissue Samples\nThe paraffin-embedded tumor tissues were obtained\nfrom the Seirei Mikatahara Hospital and from the\nfile of The Johns Hopkins Hospital. In all, after\nexcluding those with neoadjuvant chemotherapy,\n120 anonymous cases of ovarian clear cell carcinoma\n(96 cases from tissue microarrays and 24 from\nwhole sections), 49 cases of ovarian endometrioid\ncarcinoma (31 from tissue microarrays and 18 from\nwhole sections), and 38 cases of seromucinous\nneoplasm (12 malignant, the remaining were border-\nline) were collected between 1995 and 2013. Ninety-\nfour patients with ovarian clear cell carcinoma\nunderwent clinical follow-up for up to 175 months,\nand all of the patients who were followed-up were\nfrom East Asian populations. All samples were\nanonymized, and the study was approved by the\ninstitutional review boards of each participating\nhospital. The selection criteria were based on the\navailability of archived tissue, tumors from patients\nreceiving similar treatments, and pre-existing follow-\nup data rather than on other clinical or demographic\ncharacteristics.\nImmunohistochemistry\nImmunohistochemistry was performed to assess the\nexpression levels of cyclin E1 in paraffin-embedded\ntissue sections that were obtained from 120 ovarian\nclear cell carcinomas, 49 endometrioid carcinomas,\nand 38 seromucinous neoplasms. In brief, after\ndeparaffinization and rehydration, the slides were\nplaced in citrate buffer (Vector Laboratories, Burlin-\ngame, CA, USA) for antigen retrieval. An anti-cyclin\nE1 antibody (Sigma-Prestige, St Louis, MO, USA)\nwas then applied at a dilution of 1:300. After incu-\nbation with the primary antibody at room tempera-\nture for 2 h, 3,3 ′-diaminobenzidine was used to\ndevelop immunoreactivity, which was detected by\nthe EnVision+System (Dako, Carpinteria, CA, USA).\nThe following simple dichotomous system was used\nfor the evaluation of CCNE1 expression: diffuse and\nintense immunoreactivity in 480% of tumor cells\nwere considered positive (abnormal or overex-\npressed), which was based on the agreement of three\npathologists (AA, EK, and IMS). Observers were\nblind to survival or FISH data.\nFluorescence In Situ Hybridization\nDNA copy number at the CCNE1 locus was deter-\nmined by FISH in 88 cases of ovarian clear cell\ncarcinoma (74 on tissue microarray and 14 on whole\ntissue sections). In brief, 4- μm thick sections were\ndeparaffinized in xylene, hydrated through graded\nethanol, and incubated with proteinase K at 37 °C for\n30 min. After washing in 2 × Aniara saline-sodium\ncitrate, the slides were placed in a denaturation\nsolution (70% formamide/2 × saline-sodium citrate\nat 75 °C for 5 min) and rinsed in 2 × saline-sodium\ncitrate. CCNE1/CEN19p FISH probe (cat# FG0013,\nAbnova, Taipei, Taiwan) was applied to the slides\nand cover slipped. DNA was denatured thorough\nincubation for 15 min at 80 °C and hybridization was\nperformed at 37 °C for 20 –24 h. After washing for\n20 min in 1.5 mol/l urea in 0.2 × saline-sodium\ncitrate, slides were drained, dehydrated through\ngraded ethanol, air-dried, counterstained with\n4′-6-diamidino-2-phenylindole (cat# P-36931,\nCCNE1 in ovarian clear cell carcinoma\n298 A Ayhan et al\nModern Pathology (2017) 30, 297 –303\n\nInvitrogen, Eugene, Oregon, USA) at a concentration\nof 500 ng/ml (Sigma) for 3 min at room temperature,\nwashed in water, mounted with ProLong Gold\nAntifade mounting medium (Molecular Probes,\nEugene, OR, USA), and imaged.\n15,19\nImage Analysis\nPhotomicrographs from individual cases were cap-\ntured using a Nikon 50i epifluorescence microscope\nequipped with fluorescence excitation/emission\nfilters for different fluorophores (Omega Optical) used\nfor CCNE1 FISH. Grayscale images were merged using\nNikon NIS-Elements software with an attached Photo-\nmetrics Cool snap EZ digital camera. The CCNE1 copy\nnumber per cell was analyzed in at least 50 non-\noverlapping nuclei. Tumor cell nuclei were consid-\nered to have copy-number gain if the ratio of CCNE1\nto centromeric probe (CEP19) was ≥ 2i n 420% of the\nanalyzed tumor cells (amplification) or if ≥ 4 CCNE1\ncopies were observed in ≥ 40% of tumor cells (high\npolysomy).15,19\nStatistical Analysis\nDisease-specific overall survival of patients with\novarian clear cell carcinoma with and without cyclin\nE1 expression and CCNE1 copy-number gain were\ncompared using the Kaplan –Meier method and the\nCox proportional hazards model. This was followed\nby univariate (log-rank rest) and multivariate (Wald\ntest) analyses to determine the significance. The\nvariables examined were stage of the disease, age of\nthe patient, CCNE1 copy-number gain, cyclin E1\noverexpression, ARID1A loss, and the presence of\nsomatic mutations in the hTERT promoter. Compar-\nisons of CCNE1 overexpression/copy-number gain and\nhTERT mutation or ARID1A status were performed\nusing the two-tailed Fisher ’se x a c tt e s t .AP-value of\n0.05 or less was considered statistically significant.\nResults\nThe purposes of our study were to evaluate cyclin E\nexpression in endometriosis-related ovarian neo-\nplasms, to assess the relationship of CCNE1 status\nwith other molecular characteristics, and to determine\nthe clinical significance of these molecular altera-\ntions. Among 120 ovarian clear cell carcinomas, 28\n(23.3%) showed cyclin E1 protein overexpression.\nFourteen clear cell carcinomas were found to have\nadjacent normal-appearing endometriotic cysts that\ndemonstrated histologic continuity, but none of which\noverexpressed CCNE1 (Data not shown). Among 49\novarian endometrioid carcinomas and 38 seromuci-\nnous tumors, none displayed cyclin E1 protein over-\nexpression. To investigate whether cyclin E1 over-\nexpression was caused by gene copy-number gain, we\nperformed a CCNE1 dual-color FISH assay. Indeed,\n13 (14.8%) of 88 clear cell carcinomas available for\nFISH analysis displayed gene copy-number gain\n(Figure 1a). Most clear cell carcinomas (7/13) with\nCCNE1 copy-number gain revealed a CCNE1/CEP19\nratio of 2.0–2.9 (low level gain), whereas 3 carcinomas\ndisplayed a CCNE1/CEP19 ratio ≥ 3 (high level gain).\nThree cases showed high polysomy, that is, parallel\nincrease of CCNE1 copy number and CEP19 probe.\nAll 13 clear cell carcinomas withCCNE1 copy-number\ngain showed cyclin E1 protein overexpression; there-\nfore, CCNE1 copy-number gain and cyclin E1 over-\nexpression were significantly correlated ( Po0.001,\ntwo-tailed Fisher’s exact test). Among 20 cases with\npositive cyclin E1 staining, 7 did not qualify asCCNE1\ncopy-number gain by FISH using above criteria.\nCCNE1 FISH results and cyclin E1 expression status\nare summarized in Table 1. Based on re-evaluation by\nhematoxylin and eosin staining, clear cell carcinomas\nwith CCNE1 gain or overexpression did not show any\nunique morphological characteristics.\nSurvival analysis in 94 cases with complete follow-\nup demonstrated that cyclin E1 overexpression was\nsignificantly associated with poor overall survival,\neven after adjustment for clinical stage and patient age\n(age o60 vs ≥ 60 years) in the multivariate analysis.\nUnivariate and multivariate overall survival analysis\nby the Cox proportional hazards model is presented in\nTable 2 and is illustrated in Figure 1b. Specifically,\nshorter not only was cyclin E1 overexpression\ncorrelated with shorter overall survival when all\nstages were considered but both copy-number gain\nand/or overexpression were significantly associated\nwith poor survival of patients with stage I disease\n(Table 2; Figure 1c). As both gain and overexpression\nwere associated with poor prognosis, we further\ndetermined whether immunohistochemistry or FISH\nor both would be better to identify clinically relevant\nsubsets. There was no statistical difference in terms of\nsurvival between those cases ‘with overexpression\nand copy-number gain ’ and ‘with overexpression\nwithout copy-number gain ’ for both all stages\n(P = 0.1) and stage 1 ( P = 0.82; Figures 1b and c)\nalthough the case numbers were low.\nBecause the loss of ARID1A expression due to\ninactivating mutations is the most common mole-\ncular alteration known to date in ovarian clear\ncell carcinoma\n20 and because hTERT mutations\nare specific to this type of ovarian cancer, we deter-\nmined whether an association was present between\nthese two molecular alterations and CCNE1 upregu-\nlation. ARID1A expression and hTERT promoter\nmutation status in these cases were previously\nreported by us and were available for evaluation\nalong with CCNE1 status in 65 and 64 overlapping\ncases, respectively.\n21,22 The results demonstrated\nthat cases with CCNE1 upregulation typically did\nnot harbor mutations in the hTERT promoter\n(P = 0.01 and P = 0.04; Table 3). On the contrary,\nno significant association was observed between\nCCNE1 upregulation and loss of ARID1A expression.\nInterestingly, patients with CCNE1 upregulation and\nModern Pathology (2017) 30, 297 –303\nCCNE1 in ovarian clear cell carcinoma\nA Ayhan et al 299\n\nARID1A retention had the worst overall survival\ncompared with patients in the other groups\n(Figure 2). We did not observe any relationship\nbetween both ARID1A expression and hTERT pro-\nmoter mutations with overall survival (data not\nshown).\nDiscussion\nClear cell carcinomas belong to a specific group of\novarian tumors that are frequently associated with\nendometriosis, especially the lesions that present as\novarian endometriotic cysts (also known as endome-\ntriomas). Endometriosis-related ovarian neoplasms\ninclude clear cell carcinoma, endometrioid carci-\nnoma, and, less frequently, seromucinous tumors. In\nthis study, we demonstrated that CCNE1 gene copy-\nnumber gain and cyclin E1 overexpression are\nspecific to ovarian clear cell carcinoma compared\nwith other types of endometriosis-related neoplasms.\nIn addition, women whose clear cell carcinomas\ndisplay cyclin E1 overexpression have a shorter\noverall survival. Remarkably, when patients with\nstage 1 clear cell carcinoma were analyzed, CCNE1\nupregulation indicated a poor prognosis. This is\nsignificant because clear cell carcinomas can follow\nan aggressive clinical course even for those who\ninitially present with early-stage disease. Further-\nmore, patients with clear cell carcinomas with cyclin\nE1 overexpression and no concomitant loss of\nARID1A expression exhibit the worst clinical out-\ncome. The above results, if they can be confirmed in\nlarger cohorts, may have an impact on the manage-\nment of patients with clear cell carcinoma.\nOvarian clear cell carcinoma has been tradition-\nally classified as a type I ovarian neoplasm because\nmost patients with this subtype present at early\nstages and also because these tumors arise from\nendometriotic cysts or adenofibromas in a stepwise\nfashion. In addition, these tumors harbor molecular\ngenetic changes, including somatic mutations in\nARID1A and PIK3CA, which are commonly shared\nby other type I ovarian cancers; however, these\ntumors do not frequently have TP53 mutations.1 The\nfinding that CCNE1 copy-number gain in clear cell\ncarcinoma, but not in other type I neoplasms,\nTable 1 Correlation between CCNE1 copy number by fluores-\ncence in situ hybridization and Cyclin E1 immunohistochemistry\nin 88 ovarian clear cell carcinomas\nCCNE1 no\ncopy-number\ngain\nCCNE1\ncopy-number\ngain Total\nNegative cyclin E1 IHC 68 (77%) 0 68 (77%)\nPositive cyclin E1 IHC 7 (8%) 13 (15%) 20 (23%)\nPo 0.001, two-tailed Fisher ’s exact test.\nFigure 1 All clear cell carcinomas with CCNE1 copy-number gain\nhave cyclin E1 overexpression. A strong correlation was observed\nbetween CCNE1 copy-number gain and cyclin E1 overexpression,\nP o0.001. Increased DNA copy number at the CCNE1 locus by\nFISH; diffuse and strong CCNE1 expression by immunohistochem-\nistry (inset) ( a). CCNE1 overexpression and/or copy-number gain\nwas correlated with overall survival. P = 0.004 ( b). Cyclin E1\noverexpression and/or CCNE1 copy-number gain predicted poor\nsurvival of patients with stage I clear cell carcinoma ( P = 0.003) (c).\nBoth overexpression and copy-number gain related to poor\nprognosis and there was no statistical difference in terms of their\ncontribution to survival in both overall and stage 1 cases ( b and c,\nP = 0.1 and P = 0.82).\nModern Pathology (2017) 30, 297 –303\nCCNE1 in ovarian clear cell carcinoma\n300 A Ayhan et al\n\nsuggests that a subset of clear cell carcinomas may be\nmolecularly close to type II ovarian cancer. The\nabsence of CCNE1 gene copy-number gain and the\nlack of hTERT promoter mutations in ovarian\nendometrioid carcinoma 3,21 further indicate distinct\nmolecular pathways that are involved in the\ndevelopment of ovarian clear cell and endometrioid\ncarcinomas.\nThe amplification and subsequent overexpression\nof CCNE1 have been reported as one of the major\nmolecular mechanisms that promote the progression\nof ovarian high-grade serous carcinoma, the most\nTable 2 Univariate and multivariate overall survival analysis for patients of all stages and for those with stage 1 ovarian clear cell\ncarcinoma by the Cox proportional hazards model\nCancer stage Variables\nUnivariate analysis Multivariate analysis\nHR\n95% CI\nP-value HR\n95% CI\nP-value\nLH LH\nAll stages CCNE1 copy-number gain 1.42 0.61 3.29 0.41 —— — —\nCyclin E1 overexpression 2.16 1.11 4.18 0.02 4 1.42 11.3 0.009\nStage 3.29 2.39 4.52 o0.001 4.12 2.45 6.92 o 0.001\nAge 460 years 3.65 1.7 7.83 o0.001 6.62 1.82 24.2 0.004\nARID1A loss 0.72 0.31 1.67 0.44 0.68 0.25 1.85 0.45\nhTERT mutation 1.15 0.42 3.13 0.78 0.55 0.14 2.15 0.39\nStage I CCNE1 copy-number gain 3.77 1.18 12.1 0.02 5.38 1.24 23.4 0.02\nCyclin E1 overexpression 5.27 1.67 16.7 0.002 —— — —\nAge 460 years 2.67 0.33 21.3 0.34 6.6 0.61 71.3 0.12\nARID1A loss 0.61 0.18 2.1 0.44 0.43 0.09 1.89 0.26\nhTERT mutation 1.75 0.45 6.82 0.41 0.57 0.1 3.08 0.51\nStatistically significant P-values are depicted in bold.\nTable 3 Correlation between hTERT mutations and CCNE1 status in ovarian clear cell carcinomas\nTotal hTERT wild type hTERT mutation P-value\nNegative cyclin E1 IHC 45 (69%) 38 (58%) 7 (11%) 0.01\nPositive cyclin E1 IHC 20 (31%) 12 (19%) 8 (12%)\nCCNE1 no copy-number gain 51 (78%) 43 (66%) 8 (12%) 0.04\nCCNE1 copy-number gain 14 (22%) 7 (11%) 7 (11%)\nStatistically significant P-values are depicted in bold.\nFigure 2 The relationship between CCNE1, ARID1A, and hTERT mutation status in 64 and 65 overlapping patients with clear cell\ncarcinomas ( a). Patients with both CCNE1 overexpression and ARID1A retention have the worst survival. No association was found\nbetween overall survival (either all patients or stage 1 patients) and hTERT mutation or ARID1A loss ( b).\nModern Pathology (2017) 30, 297 –303\nCCNE1 in ovarian clear cell carcinoma\nA Ayhan et al 301\n\ncommon and aggressive type of ovarian cancer. 19,23\nHere we demonstrated that clear cell carcinomas\nalso harness this molecular pathway to promote\naggressive behaviour, and that women whose clear\ncell carcinomas display CCNE1 copy-number gain or\noverexpression exhibit a dismal outcome compared\nwith those without CCNE1. It should be noted that\nthe FISH signals of the CCNE1 locus that were\nobserved in clear cell carcinomas were mostly\ndiscrete countable signals, probably representing a\ndouble minute type amplicon, whereas the CCNE1\namplification detected in high-grade serous carci-\nnoma often exhibited as aggregated clusters of\nsignals, most likely representing a homogenously\nstaining region type amplicon. The above findings\nsuggest that different mechanisms are involved.\nCyclin E1 has been established as a critical factor\nin the regulation of cell cycle progression, and\noverexpression of cyclin E1 has been reported to\nresult in genomic instability. Cyclin E1 can inappro-\npriately initiate DNA replication and centrosome\nduplication, which leads to chromosomal instability,\naneuploidy, and eventually tumorigenesis.\n24,25 As\nthe prognosis for advanced-stage ovarian clear\ncell carcinoma is poor, 26,27 there is an unmet need\nto identify clear cell carcinoma patients who are at\nan increased risk for the development of progressive\ndiseases despite intensive platinum-based che-\nmotherapy. Remarkably, the copy-number gain can\noccur in the early stages of the disease and it can also\npredict poor disease outcome. In agreement with this\nfinding, we also found that either immunohisto-\nchemistry or FISH was similar in performance to\nidentify patients with better and worse survival.\nFurthermore, those ovarian clear cell carcinomas\nwith CCNE1 upregulation (gene copy-number gain\nand/or overexpression) and ARID1A retention is\nassociated with worse disease outcome even though\ntheir morphological features are indistinguishable\nfrom those of the other clear cell carcinoma types.\nClinically, women presenting stage I ovarian clear\ncell carcinomas may be candidates for chemotherapy\nbecause some of them may progress even after their\ntumors are resected. Therefore, it would be of great\ninterest for the future clinical studies to determine\nwhether cyclin E1 and ARID1A immunohisto-\nchemistry would help identify those with aggressive\nclinical behavior and they may warrant a more\nintensive therapy. For example, clear cell carcino-\nmas exhibiting CCNE1 upregulation and ARID1A\nexpression may benefit from target-based therapy.\nFuture research is needed to comprehensively\nelucidate the molecular landscape that accounts for\nthis dismal outcome in this subset of clear cell\ncarcinomas.\nAlthough this study provides several new insights\ninto the pathogenesis of clear cell carcinoma and\noffers novel potential biomarkers to identify a subset\nof clear cell carcinomas with the most aggressive\nclinical behavior, several limitations of this study\nshould be discussed. First, all the tumors that were\nanalyzed in the current study were derived from\npatients from Japan, where the percentages of clear\ncell carcinoma are higher than in Western countries.\nIt remains unclear whether the pathogenesis and\nmolecular landscape differ among individuals of\nvarious ethnic backgrounds because the previous\ncomprehensive genomic analysis was not performed\nin this population. Second, as discussed above, we\ndid not observe high CCNE1 DNA copy-number gain\n(so-called ‘homogenously staining region type\namplicon’) in this set of clear cell carcinomas as in\nuterine serous carcinomas.\n15,19 The CCNE1 copy-\nnumber gain in some of the tumors may be attributed\nto a high polysomy event rather than to the actual\nduplication of the discrete CCNE1 locus. However,\nwe observed a strong correlation between DNA copy-\nnumber gain and immunoreactivity in clear cell\ncarcinomas, which suggests that CCNE1 copy-\nnumber gain has a biological significance instead of\njust a bystander effect. Third, in this study, we found\nthat clear cell carcinomas with either overexpression\nof CCNE1 or CCNE1 copy-number gain tend to not\nharbor hTERT promoter (activating) mutations. Why\nthis occurs is not known, but one explanation is that\ncyclin E1 upregulation that is due to either increased\ngene copy number or transcriptional activation may\nparticipate in the maintenance of telomere length\nthrough hTERT promoter activation. This is because\nhTERT expression can be reactivated in tumors that\nharbor hTERT promoter activating mutations.\n28,29\nIndeed, it has been reported that, in breast cancer,\nthe highest telomerase activity was found in carci-\nnomas with cyclin E1 overexpression.\n30 Further\nstudies are required to determine whether cyclin\nE1 also contributes to the maintenance of telomeres\nin ovarian clear cell carcinomas and to determine the\nmechanisms involved.\nIn summary, we found that CCNE1 copy-number\ngain and overexpression occurred in ovarian clear\ncell carcinoma, but not in ovarian endometrioid or\nseromucinous carcinomas. CCNE1 copy-number\ngain and overexpression showed a high tendency\ntowards mutual exclusivity with hTERT promoter\nmutations, but not with the loss of ARID1A expres-\nsion. CCNE1 copy-number gain and overexpression\nconfer a poor prognosis, especially in the subset with\nARID1A retention. These results suggest the impor-\ntance of CCNE1 in the progression of ovarian clear\ncell carcinoma and support cyclin E1 as a possible\ntherapeutic target in ovarian clear cell carcinoma.\nAcknowledgments\nThis project was supported, in part, by SRF, MEXT\n(S-001), the National Cancer Center Research and\nDeveloping Fund, and A-MED, from Japan, and\nRichard W. TeLinde GynecologicPathology Program\nat Johns Hopkins Medical Institutions.\nModern Pathology (2017) 30, 297 –303\nCCNE1 in ovarian clear cell carcinoma\n302 A Ayhan et al\n\nDisclosure/conflict of interest\nThe authors declare no conflict of interest.\nReferences\n1 Kurman RJ, Shih Ie M. The dualistic model of ovarian\ncarcinogenesis: revisited, revised, and expanded. Am J\nPathol 2016;186:733 –747.\n2 Yamaguchi K, Mandai M, Toyokuni S, et al. Contents of\nendometriotic cysts, especially the high concentration\nof free iron, are a possible cause of carcinogenesis in\nthe cysts through the iron-induced persistent oxidative\nstress. Clin Cancer Res 2008;14:32 –40.\n3 Wu RC, Wang TL, Shih IM. The emerging roles of\nARID1A in tumor suppression. Cancer Biol Ther 2014;15:\n655–664.\n4 Kobel M, Kalloger SE, Carrick J, et al. A limited panel of\nimmunomarkers can reliably distinguish between clear\ncell and high-grade serous carcinoma of the ovary.\nAm J Surg Pathol 2009;33:14 –21.\n5 Fadare O, Zhao C, Khabele D, et al. Comparative\nanalysis of Napsin A, alpha-methylacyl-coenzyme A\nracemase (AMACR, P504S), and hepatocyte nuclear\nfactor 1 beta as diagnostic markers of ovarian clear cell\ncarcinoma: an immunohistochemical study of 279\novarian tumours. Pathology 2015;47:105 –111.\n6 Kandalaft PL, Gown AM, Isacson C. The lung-restricted\nmarker napsin A is highly expressed in clear cell\ncarcinomas of the ovary. Am J Clin Pathol 2014;142:\n830–836.\n7 Skirnisdottir I, Bjersand K, Akerud H, et al. Napsin A as\na marker of clear cell ovarian carcinoma. BMC Cancer\n2013;13:524.\n8 Guan B, Rahmanto YS, Wu RC, et al. Roles of deletion\nof Arid1a, a tumor suppressor, in mouse ovarian\ntumorigenesis. J Natl Cancer Inst 2014;106:dju 146.\n9 Chandler RL, Damrauer JS, Raab JR, et al. Coexistent\nARID1A-PIK3CA mutations promote ovarian clear-cell\ntumorigenesis through pro-tumorigenic inflammatory\ncytokine signalling. Nat Commun 2015;6:6118.\n10 Jones S, Wang TL, Shih Ie M, et al. Frequent mutations\nof chromatin remodeling gene ARID1A in ovarian clear\ncell carcinoma. Science 2010;330:228 –231.\n11 Keyomarsi K, Tucker SL, Bedrosian I. Cyclin E is a\nmore powerful predictor of breast cancer outcome than\nproliferation. Nat Med 2003;9:152.\n12 Nakayama N, Nakayama K, Shamima Y, et al. Gene\namplification CCNE1 is related to poor survival and\npotential therapeutic target in ovarian cancer. Cancer\n2010;116:2621–2634.\n13 Rosen DG, Yang G, Deavers MT, et al. Cyclin E\nexpression is correlated with tumor progression and\npredicts a poor prognosis in patients with ovarian\ncarcinoma. Cancer 2006;106:1925 –1932.\n14 Kuhn E, Wu RC, Guan B, et al. Identification of mole-\ncular pathway aberrations in uterine serous carcinoma\nby genome-wide analyses. J Natl Cancer Inst 2012;104:\n1503–1513.\n15 Kuhn E, Bahadirli-Talbott A, Shih Ie M. Frequent\nCCNE1 amplification in endometrial intraepithelial\ncarcinoma and uterine serous carcinoma. Mod Pathol\n2014;27:1014–1019.\n16 Cassia R, Moreno-Bueno G, Rodriguez-Perales S, et al.\nCyclin E gene (CCNE) amplification and hCDC4\nmutations in endometrial carcinoma. J Pathol 2003;201:\n589–595.\n17 Bedrosian I, Lu KH, Verschraegen C, et al. Cyclin E\nderegulation alters the biologic properties of ovarian\ncancer cells. Oncogene 2004;23:2648 –2657.\n18 Karst AM, Jones PM, Vena N, et al. Cyclin E1\nderegulation occurs early in secretory cell transforma-\ntion to promote formation of fallopian tube-derived\nhigh-grade serous ovarian cancers. Cancer Res 2014;74:\n1141–1152.\n19 Kuhn E, Wang TL, Doberstein K, et al. CCNE1\namplification and centrosome number abnormality in\nserous tubal intraepithelial carcinoma: further evidence\nsupporting its role as a precursor of ovarian high-grade\nserous carcinoma. Mod Pathol (e-pub ahead of print 22\nJuly 2016; doi:10.1038/modpathol.2016.101).\n20 Guan B, Wang TL, Shih Ie M. ARID1A, a factor that\npromotes formation of SWI/SNF-mediated chromatin\nremodeling, is a tumor suppressor in gynecologic cancers.\nCancer Res 2011;71:6718–6727.\n21 Wu RC, Ayhan A, Maeda D, et al. Frequent somatic\nmutations of the telomerase reverse transcriptase\npromoter in ovarian clear cell carcinoma but not\nin other major types of gynaecological malignancy.\nJ Pathol 2014;232:473 –481.\n22 Ayhan A, Mao TL, Seckin T, et al. Loss of ARID1A\nexpression is an early molecular event in tumor\nprogression from ovarian endometriotic cyst to clear\ncell and endometrioid carcinoma. Int J Gynecol Cancer\n2012;22:1310–1315.\n23 Nakayama K, Nakayama N, Jinawath N, et al. Amplicon\nprofiles in ovarian serous carcinomas. Int J Cancer\n2007;120:2613–2617.\n24 Pundir S, Vu HY, Solomon VR, et al. VR23: a\nquinoline-sulfonyl hybrid proteasome inhibitor that\nselectively kills cancer via cyclin E-mediated centro-\nsome amplification. Cancer Res 2015;75:4164 –4175.\n25 Kawamura K, Izumi H, Ma Z, et al. Induction of\ncentrosome amplification and chromosome instability\nin human bladder cancer cells by p53 mutation and\ncyclin E overexpression. Cancer Res 2004;64:4800–4809.\n26 Pather S, Quinn MA. Clear-cell cancer of the ovary-is\nit chemosensitive? Int J Gynecol Cancer 2005;15:\n432–437.\n27 Takano M, Sugiyama T, Yaegashi N, et al. Low res-\nponse rate of second-line chemotherapy for recurrent\nor refractory clear cell carcinoma of the ovary:\na retrospective Japan Clear Cell Carcinoma Study. Int\nJ Gynecol Cancer 2008;18:937 –942.\n28 Horn S, Figl A, Rachakonda PS, et al. TERT promoter\nmutations in familial and sporadic melanoma. Science\n2013;339:959–961.\n29 Huang FW, Hodis E, Xu MJ, et al. Highly recurrent\nTERT promoter mutations in human melanoma.\nScience 2013;339:957 –959.\n30 Landberg G, Nielsen NH, Nilsson P, et al. Telomerase\nactivity is associated with cell cycle deregulation in\nhuman breast cancer. Cancer Res 1997;57:549 –554.\nModern Pathology (2017) 30, 297 –303\nCCNE1 in ovarian clear cell carcinoma\nA Ayhan et al 303","source_license":"public-domain-us","license_restricted":false}