Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Gynecologic Oncology 140 (2016) 226–233 Contents lists available at ScienceDirect Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno Clinicopathologic implications of DNA mismatch repair status in endometrial carcinomas Ayumi Shikama, Takeo Minaguchi ⁎, Koji Matsumoto, Azusa Akiyama-Abe, Yuko Nakamura, Hiroo Michikami, Sari Nakao, Manabu Sakurai, Hiroyuki Ochi, Mamiko Onuki, Toyomi Satoh, Akinori Oki, Hiroyuki Yoshikawa Department of Obstetrics and Gynecology, Faculty of Medicine, University of Tsukuba, Tsukuba, Japan H I G H L I G H T S • 221 endometrial cancers were classified as sporadic or Lynch syndrome by MMR analyses. • LS correlated with favorable prognostic factors and sensitivity to adjuvant therapies. • Analyzing MMR and searching for LS may find patients with favorable survival and sensitivity to adjuvant therapies. a r t i c l e i n f o Article history: Received 23 August 2015 Received in revised form 24 November 2015 Accepted 27 November 2015 Available online 28 November 2015 Keywords: Mismatch repair deficiency Endometrial carcinoma Lynch syndrome Survival MLH1 promoter methylation a b s t r a c t Objective. Endometrial carcinoma is the most common malignancy in women with Lynch syndrome caused by mismatch repair (MMR) deficiency. We investigated the clinicopathologic significance of deficient MMR and Lynch syndrome presumed by MMR analyses in unselected endometrial carcinomas. Methods. We analyzed immunohistochemistry of MMR proteins (MLH1/MSH2/MSH6/PMS2) and MLH1 promoter methylation in primary endometrial carcinomas from 221 consecutive patients. Based on these results, tumors were categorized as sporadic or probable Lynch syndrome (PLS). Clinicopathologic variables and prognosis were compared according to MMR status and sporadic/PLS classification. Results. Deficient MMR showed only trends towards favorable overall survival (OS) compared with intact MMR (p = 0.13), whereas PLS showed significantly better OS than sporadic (p = 0.038). Sporadic was significantly associated with older age, obesity, deep myometrial invasion, and advanced stage (p = 0.008, 0.01, 0.02 and 0.03), while PLS was significantly associated with early stage and Lynch syndrome-associated multiple cancer (p = 0.04 and 0.001). The trend towards favorable OS of PLS was stronger in advanced stage than in early stage (hazard ratio, 0.044 [95% CI 0–25.6] vs. 0.49 [0.063–3.8]). In the subset receiving adjuvant therapies, PLS showed trends towards favorable disease-free survival compared to sporadic by contrast with patients receiving no adjuvant therapies showing no such trend (hazard ratio, 0.045 [95% CI 0–20.3] vs. 0.81 [0.095–7.0]). Conclusions. The current findings suggest that analyzing MMR status and searching for Lynch syndrome may identify a subset of patients with favorable survival and high sensitivity to adjuvant therapies, providing novel and useful implications for formulating the precision medicine in endometrial carcinoma. © 2015 Elsevier Inc. All rights reserved. 1. Introduction In the western world, endometrial carcinoma is the most common gynecologic malignancy. Based on differences in clinicopathologic characteristics, there are two subtypes of endometrial carcinoma. Type I tumors, estrogen-related, are caused by the unopposed estrogen and follow endometrial hyperplasias. This type of tumors is characterized by endometrioid histology, occurrence in perimenopausal, obese ⁎ Corresponding author at: Department of Obstetrics and Gynecology, Faculty of Medicine, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8575, Japan. E-mail address: [email protected] (T. Minaguchi). http://dx.doi.org/10.1016/j.ygyno.2015.11.032 0090-8258/© 2015 Elsevier Inc. All rights reserved. women, superficial myometrial invasion, and favorable prognosis. On the contrary, type II tumors, estrogen-unrelated, develop from atrophic endometrium. This type of tumors is generally characterized by nonendometrioid histology, occurrence in older, postmenopausal, thin women, deep myometrial invasion, and poor prognosis. Defect in specific signal transduction pathways are known to drive endometrial carcinogenesis. These pathways include the phosphatidylinositol 3-kinase (PI3K)-PTEN-Akt pathway, the p53 pathway, and the DNA mismatch repair (MMR) mechanism. The MMR system is a strand-specific DNA repair mechanism. The role of MMR gene is to maintain genomic integrity by correcting base substitutions mismatches and small insertion-deletion mismatches that are generated by errors in base pairing during DNA replication. Of A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 Table 1 Patient characteristics. Characteristic Number (n = 221) Median age (range) Median BMI (kg/m2 range) FIGO stage I Ia Ib Ic II IIa IIb III IIIa IIIc IV IVa IVb Histotype Endometrioid G1 G2 G3 Serous Adenosquamous Clear cell Poorly differentiated Undifferentiated Mixed epithelial Myometrial invasion N 1/2 Lymophovascular space invasion Carcinoma of the lower uterine segment 57.4 (26–84) 23.9 (17.0–43.9) Primary treatment Surgery Lymphadenectomy Lymph node sampling Lymph node not removed Adjuvant chemotherapy TC CAP Adjuvant radiotherapy Lynch syndrome-associated multiple cancer % 227 results of our molecular analyses. The current findings will provide novel and useful implications for the precision medicine in endometrial carcinoma. 2. Materials and methods 2.1. Patients and specimens 128 22 76 30 26 10 16 43 20 23 24 2 22 196 115 56 25 12 4 4 1 1 3 81 84 13 221 171 21 29 60 55 4 58 15 58 10 34 14 12 5 7 19 9 10 11 1 10 89 52 25 11 5 2 2 0 0 1 37 38 6 100 77 10 13 27 25 2 26 7 Abbreviations: FIGO = International Federation of Gynecology and Obstetrics; TC = paclitaxel and carboplatin combination; CAP = cyclophosphamide, doxorubicin, and cisplatin combination. the MMR genes, heterodimeric complexes of MSH2 and MSH6 recognize mismatch nucleotides, and heterodimeric complexes of MLH1 and PMS2 are then recruited to excise the mismatched nucleotides. MMR deficiency is detected as microsatellite instability (MSI) or loss of MMR protein expression by immunohistochemistry (IHC). In Lynch syndrome, MMR deficiency results from germline mutations of MLH1, MSH2, MSH6, and PMS2. Deficient MMR is also detected in sporadic cancer due to hypermethylation of MLH1 promoter. MMR deficiency is reported to contribute to oncogenesis in some types of malignancy, including colon, gastric and endometrial cancers. In patients with colon cancer, multiple studies have reported that deficient MMR is associated with favorable prognosis [1,2]. Furthermore, colon cancers with deficient MMR have been demonstrated to be more responsive to 5FU-based chemotherapy compared with tumors with intact MMR [1– 3]. In endometrial cancer, most studies reported on the association between MSI and clinical outcome and it is controversial whether MSI status improves patient survival [4–7]. A few studies reported on the association between MMR protein expression and clinical outcome, but prognostic significance of MMR protein expression is yet to be determined [8,9]. Moreover, germline mutation of BRCA1/2, another cancer predisposition gene, is reported to be associated with improved survival and high response to platinum-based chemotherapy in ovarian cancer [10,11]. Here, we have investigated the clinicopathologic significance of deficient MMR and presumed Lynch syndrome based on the The Ethical Committee of the University of Tsukuba Hospital approved the study protocol. All patients diagnosed with endometrial carcinoma, who were treated in the Department of Obstetrics and Gynecology at the University of Tsukuba Hospital between 1999 and 2009, were identified through our database. A total of consecutive 221 patients with endometrial carcinomas were included in the present study, and their medical records were reviewed. A median follow-up duration was 92 months (range, 3–181 months). All patients provided written informed consent. Staging was performed based on the criteria of International Federation of Gynecology and Obstetrics (FIGO, 1988). Endometrioid adenocarcinomas were subclassified into three grades (G1, G2, and G3) according to the FIGO criteria. Treatment of patients was described elsewhere [12]. Table 1 summarizes the patient characteristics. We categorized patients as sporadic or probable Lynch syndrome (PLS) based on the results of our MMR analyses as follows [13–15]. A patient with endometrial carcinoma with intact expression of all MLH1, MSH2, MSH6, and PMS2 proteins was considered as having a sporadic tumor. A patient with loss of MLH1 expression, but presence of MLH1 promoter methylation, was also considered sporadic. Patients with loss of MSH2, MSH6, or PMS2 expression were considered as PLS. Patients with tumors with loss of MLH1 and absence of MLH1 promoter methylation were also considered PLS. 2.2. Immunohistochemistry (IHC) IHC procedures were conducted as described previously [16]. Monoclonal antibodies used are MLH1 (BD Pharmingen, Franklin Lakes, NJ), MSH2 (EMD Millipore corporation, Billerica, MA, USA), MSH6 (Epitomics, Burlingame, CA, USA), and PMS2 (BD Pharmingen). The corresponding normal tissue provided an internal positive control, and cases already known as mutated were used for negative controls for MMR proteins. Loss of MMR protein expression was defined as the absence of nuclear staining in tumor cells in the presence of adjacent non-neoplastic cells with nuclear staining. Blinded for clinical and pathologic parameters, immunoreaction was assigned independently by two investigators (AS and TM), and any discrepancies were resolved by conferring over a multiviewer microscope. Tumors were classified as deficient MMR, if expression loss of one or more MMR proteins was detected. Fig. 1 shows examples of IHC staining for MMR proteins in endometrial carcinomas. PTEN and p53 expressions were examined and evaluated as previously described [12]. 2.3. DNA extraction Genomic DNA was extracted from tumors and corresponding normal areas of formalin-fixed, paraffin-embedded endometrial tissues using blackPREP DNA kit (Analytik Jena, Jena, Germany) according to the manufacturer's instructions. For isolating DNA from normal tissues, we used FFPE tissue blocks (uterus, ovary, or omentum) containing only normal tissues without any tumor tissues, which was confirmed by checking corresponding slides by microscopy. 2.4. MSI analysis MSI status was analyzed using five fluorescence-labeled microsatellite markers, BAT25, BAT26, D2S123, D5S346 and D17S250, recommended by the National Cancer Institute [17]. Tumors showing allelic 228 A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 Fig. 1. IHC staining patterns of MLH1, MSH2, MSH6, PMS2 in endometrial carcinomas (×100). shift at two or more markers were classified as MSI-high, at one marker as MSI-low, and tumors with no allelic shift at any marker as microsatellite stable (MSS). 72 °C. The PCR products were resolved by electrophoresis on a 3% agarose gel and visualized by ethidium bromide staining and UV illumination. 2.5. Methylation-specific PCR (MS-PCR) 2.6. Statistical analyses Modification of genomic DNAs with sodium bisulfite was performed using the EpiTect kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. MS-PCR was performed as described elsewhere with slight modifications [18]. Bisulfite-modified DNAs served as templates using the primer sets specific for methylated and unmethylated CpG islands in the MLH1 promoter. PCR mixture contained 1 × PCR buffer (Perkin-Elmer, Foster City, CA), 2.5 mM MgCl2, 200 μM of each dNTP, 0.5 μM of each PCR primer, 0.5 U of AmpliTaq Gold (PerkinElmer), and 50 ng of DNA in a final volume of 25 μl. Amplification proceeded at 95 °C for 10 min, followed by 40 cycles at 94 °C for 45 s, 60 or 62 °C for 30 s, and 72 °C for 60 s and a final 5-min extension at Differences in proportions were evaluated by the Fisher's exact test. Kaplan-Meier survival curves were calculated and compared statistically using the log-rank test. Univariate and multivariate analyses for prognostic factors was conducted by the Cox proportional hazard model. For all analysis p-value b 0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 21. 3. Results In our IHC analyses and MLH1 MS-PCR in primary endometrial carcinomas from 221 unselected patients, deficient MMR was observed in 62 8 (89%) 144 (68%) 0.170 1 (11%) 36 (17%) 0.538 20 (83%) 132 (67%) 0.077 2 (8%) 35 (18%) 0.193 52 (84%) 100 (63%) 0.002 4 (6%) 33 (21%) 0.006 32 (80%) 120 (66%) 0.063 4 (10%) 33 (18%) 0.151 54 (78%) 98 (65%) 0.028 6 (9%) 31 (20%) 0.021 128 (66%) 24 (86%) 0.027 35 (18%) 2 (7%) 0.113 118 (65%) 12 (75%) 0.309 33 (18%) 2 (13%) 0.432 Abbreviations: IHC = immunohistochemistry; MSI = microsatellite instability; MS-PCR = methylation-specific PCR; MMR = mismatch repair; PLS = probable Lynch syndrome. 3 (19%) 7 (44%) 1 (6%) 0 (0%) 3 (19%) 0 (0%) 0 (0%) 13 (81%) 2 (13%) 6 (38%) 8 (50%) 0 (0%) 10 (63%) 3 (19%) 4 (25%) 79 (44%) 51 (29%) 35 (19%) 29 (16%) 64 (35%) 18 (10%) 40 (22%) 161 (89%) 22 (12%) 74 (41%) 70 (39%) 11 (6%) 99 (55%) 60 (33%) 8 (5%) 0.008 0.58 0.11 0.01 0.12 0.27 0.05 0.19 0.14 0.02 0.52 0.51 0.04 0.03 0.001 5 (18%) 8 (29%) 2 (7%) 0 (0%) 6 (21%) 1 (4%) 2 (7%) 23 (82%) 1 (4%) 5 (18%) 11 (39%) 2 (7%) 21 (75%) 4 (14%) 7 (25%) 83 (43%) 55 (29%) 35 (18%) 30 (16%) 67 (35%) 18 (9%) 43 (22%) 173 (90%) 24 (12%) 76 (39%) 73 (38%) 11 (6%) 107 (55%) 63 (33%) 8 (4%) 0.018 0.11 0.21 0.044 0.53 0.38 0.11 0.019 0.11 0.17 0.47 0.17 0.45 0.42 0.53 68 (45%) 39 (26%) 28 (18%) 16 (11%) 50 (33%) 12 (8%) 27 (18%) 130 (86%) 14 (9%) 52 (34%) 57 (38%) 11 (7%) 89 (59%) 45 (30%) 10 (7%) 20 (29%) 24 (35%) 9 (13%) 14 (20%) 23 (33%) 7 (10%) 18 (26%) 66 (96%) 11 (16%) 29 (42%) 27 (39%) 2 (3%) 39 (57%) 22 (32%) 5 (7%) 0.055 0.34 0.06 0.16 0.08 0.54 0.40 0.51 0.005 0.018 0.029 0.07 0.41 0.30 0.11 77 (43%) 53 (30%) 34 (19%) 27 (15%) 64 (35%) 16 (9%) 38 (21%) 160 (88%) 15 (8%) 60 (33%) 63 (35%) 13 (7%) 106 (59%) 53 (29%) 10 (6%) PLS n = 16 Sporadic n = 181 Sporadic n = 193 Unmethylated n = 152 229 Abbreviations: MMR = mismatch repair; MSI = microsatellite instability; PLS = probable Lynch syndrome; BMI = body mass index; MI = myometrial invasion; LVI = lymophovascular space invasion; LUS = carcinoma of the lower uterine segment; FIGO = International Federation of Gynecology and Obstetrics. Lynch syndrome-associated multiple cancers include colorectal, endometrial, gastric, small bowel, ovarian, pancreatic, ureter and renal pelvic, biliary tract, and brain (glioblastoma in Torcot syndrome), sebaceous gland adenomas, and keratoacanthomas in Muir-Torre syndrome. 0.012 11 (28%) 10 (25%) 3 (8%) 3 (8%) 9 (23%) 3 (8%) 7 (18%) 36 (90%) 10 (25%) 21 (53%) 21 (53%) 0 (0%) 22 (55%) 14 (35%) 5 (13%) 2 (5%) 35 (20%) 0.002 0.21 0.006 0.037 0.012 0.34 0.22 0.41 0.24 0.48 0.18 0.24 0.22 0.47 0.03 0.005 73 (46%) 48 (30%) 33 (21%) 26 (16%) 60 (38%) 15 (9%) 35 (22%) 140 (88%) 16 (10%) 59 (37%) 57 (36%) 11 (7%) 89 (56%) 49 (31%) 7 (4%) 36 (86%) 116 (64%) 15 (24%) 15 (24%) 4 (6%) 4 (6%) 13 (21%) 4 (6%) 10 (16%) 56 (90%) 9 (15%) 22 (35%) 27 (44%) 2 (3%) 39 (63%) 18 (29%) 8 (13%) 0.021 Age ≥ 60 Pre-menopause Null Parity BMI N 30 Hypertension Hyperlipidemia Diabetes Endometrioid (vs. Non-endometrioid) G3 MI N 1/2 LVI LUS FIGO stage I FIGO stage III-IV Lynch syndrome-associated multiple cancer 2 (5%) 35 (19%) MSI-H n = 40 0.034 P-value 32 (82%) 120 (66%) Intact n = 159 P-value Deficient n = 62 Positive p53 MSI P-value MMR Negative PTEN Clinicopathologic variables Expression Table 3 Relationships between clinicopathologic variables and results of MMR analyses. Table 2 Results of IHC, MSI and MLH1 MS-PCR. MLH1 Negative (n = 39) Positive (n = 182) PMS2 Negative (n = 42) Positive (n = 179) MSH2 Negative (n = 9) Positive (n = 212) MSH6 Negative (n = 24) Positive (n = 197) MMR Deficient (n = 62) Intact (n = 159) MSI MSI-H (n = 40) MSI-L + MSS (n = 181) MLH1 promoter Methylated (n = 69) Unmethylated (n = 152) Sporadic/PLS Sporadic (n = 193) PLS (n = 28) Sporadic/PLS (+MSI) Sporadic (n = 181) PLS (n = 16) Methylated n = 69 MSI-L + MSS n = 181 P-value MLH1 promoter P-value Sporadic/PLS PLS n = 28 P-value Sporadic/PLS (+MSI) P-value cases (28%) and MLH1 promoter methylation in 69 cases (31%) (Table 2). MLH1 promoter methylation was significantly associated with loss of MLH1 expression by IHC (p = 2.2E-15; Table S1). We also conducted MSI analysis in all cases. Deficient MMR was found to be significantly associated with MSI (p = 5.6E − 26; Table S1). Together, these results support the validity of our IHC analyses. Comparison between expressions of MMR proteins and those of PTEN/p53 tumor suppressors showed that MMR deficiency was significantly associated with negative PTEN and inversely with positive p53 (p = 0.002 and 0.006; Table 2). Expressions of MLH1, PMS2, and MLH1 methylation also showed the similar significant correlations with PTEN and p53 (Table 2). We next correlated our IHC results to clinicopathologic variables. MMR deficiency was found to be significantly associated with age b 60 and Lynch syndrome-associated multiple cancer, and inversely with null parity, BMI N30, and hypertension (p = 0.002, 0.03, 0.006, 0.037, and 0.012; Table 3). MLH1 methylation was found to be significantly associated with age b 60, BMI N 30, and endometrioid histology (p = 0.018, 0.044, and 0.019; Table 3). Subsequently we examined the prognostic significance of MMR and MSI status in our survival analysis. Patients with deficient MMR showed only a tendency towards better overall survival than those with intact MMR without statistical significance (p = 0.13; Fig. 2A). MSI status did not show any correlation with overall survival (data not shown). Aiming to next investigate the possible clinicopathologic difference between sporadic and Lynch syndrome-associated endometrial carcinomas, we categorized tumors according to the results of MMR protein expression and MLH1 promoter methylation. Consequently, 28 cases (13%) were classified as PLS, and 193 cases (87%) were classified as sporadic. Comparison with clinicopathologic variables showed that sporadic group was significantly associated with older age (≥ 60), BMI N 30, deep myometrial invasion (N 1/2), and advanced FIGO stage (III-IV) (p = 0.008, 0.01, 0.02 and 0.03; Table 3), while PLS group was significantly associated with early FIGO stage (I) and Lynch syndromeassociated multiple cancer (p = 0.04 and 0.001; Table 3). Subsequent survival analyses revealed that PLS group had significantly better overall survival than sporadic group (p = 0.038; Fig. 2B). Interestingly, the trend towards favorable overall survival of PLS group against sporadic 0.043 0.17 0.17 0.07 0.14 0.20 0.022 0.28 0.61 0.51 0.27 0.38 0.37 0.19 0.010 A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 230 A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 group was stronger in advanced-stage disease than in early-stage disease (Fig. 2C, D; hazard ratio, 0.044 [95% CI 0–25.6] vs. 0.49 [0.063– 3.8]; Table S2). In order to identify the mechanism underlying this prognostic difference between sporadic and PLS groups, we further conducted subset analyses on disease-free survivals (DFS) in patients who received adjuvant therapies. In the patients receiving adjuvant therapies, PLS group showed trends towards favorable DFS compared to sporadic group, whereas no such trend was observed in patients receiving no adjuvant therapies (Fig. 2E, F; hazard ratio, 0.045 [95% CI 0–20.3] vs. 0.81 [0.095–7.0]; Table S2). We moreover compared the trends of PLS towards favorable DFS between patients receiving adjuvant chemotherapies alone and those receiving adjuvant radiotherapies alone, but both were found to be similar (Fig. 2G, H; hazard ratio, 0.043 [95% CI 0–66.2] vs. 0.043 [0–1303.2]; Table S2). Our survival analyses showed that, in addition to the factor of PLS against sporadic, factors of age b 60, endometrioid histology, G1, MI ≤ 1/2, absent LVI, FIGO stage I/II, and no adjuvant therapy received were also significantly associated with favorable overall survival (p = 0.001, 2.2E−9, 1.8E− 5, 5.7E− 7, 2.9E−5, 9.5E− 11, and 5.0E− 9 by the log-rank test). Among these significant prognostic factors, subsequent multivariate analysis exhibited that endometrioid histology, early FIGO stage, and no adjuvant therapy received were found to be significant and independent for favorable overall survival (p = 1.9E−4, 0.02, and 0.017; Table 4). 4. Discussion Our IHC analyses showed that deficient MMR was significantly associated with negative PTEN and inversely correlated with positive p53 (Table 2), being consistent with previous publications [19−21]. However, deficient MMR was found to be significantly associated inversely with null parity, obesity, and hypertension (Table 3). These associations are generally uncommon in type I endometrial cancer where mutant PTEN and wild-type p53 are frequent, suggesting that the characteristics of MMR-deficient endometrial carcinoma are distinct from those of typical type I tumor. This may be because more diverse range of oncogenes and tumor suppressor genes than just PTEN is targeted by MMR deficiency. By contrast, MLH1 methylation was significantly associated with obesity and endometrioid histology (Table 3), probably reflecting sporadic origin. We next examined the prognostic significance of MMR and MSI status. MMR deficiency determined by IHC alone showed only trends towards better overall survival compared with intact MMR without statistical significance (Fig. 2A). MSI status did not exhibit any correlation with overall survival (data not shown). Therefore, we further examined the clinicopathologic significance of PLS determined by MMR IHC plus MLH1 promoter methylation, as germline mutation of BRCA1/ 2, which is another cancer predisposition gene, is reported to be associated with improved survival and high response to platinum-based chemotherapy in ovarian cancer [10,11]. Sporadic group was significantly associated with older age, obesity, deep myometrial invasion, and advanced FIGO stage, while PLS group was significantly associated with early FIGO stage and Lynch syndrome-associated multiple cancer (Table 3). Subsequent survival analyses revealed that overall survival of PLS group was significantly better than that of sporadic group (Fig. 2B). These findings suggest that favorable prognosis of PLS group may be due to superficial myometrial invasion and early stage, possibly reflecting the indolent biological tumor behavior. If the favorable prognostic impact by PLS group is mainly attributed only to early stage and superficial myometrial invasion, the both prognostic effects by early stage or superficial myometrial invasion should be diluted in advanced-stage diseases than in early-stage diseases. However, our subset survival analyses showed that the tendency of PLS group towards favorable survival was observed in advanced-stage (III-IV) diseases, while no such trend was observed in early-stage (I-II) diseases (Fig. 2C, D; Table S2), suggesting that factors other than stage and depth of myometrial invasion are also contributing to favorable prognosis. As an attempt to identify other mechanisms of the prognostic difference between PLS and sporadic groups, we next conducted subset analyses on DFS in patients who received adjuvant therapies. Interestingly, the trend of PLS group towards better DFS was observed in patients receiving adjuvant therapies, whereas no such trend was observed in patients receiving no adjuvant therapies (Fig. 2E, F; Table S2). These observations implicate that the prognostic difference between PLS and sporadic groups may be also attributed to higher sensitivity to adjuvant therapies in PLS group. It can be theoretically thought that tumors with deficient MMR are more vulnerable to genotoxic therapies than MMR-intact tumors. Sporadic group must include much less MMR-deficient tumors than PLS group, possibly explaining the prognostic impact of PLS over sporadic (Fig. 2B). Likewise, advanced-stage diseases include more patients receiving adjuvant therapies compared to early-stage diseases, possibly explaining the prognostic difference of PLS between the advanced and early-stage diseases (Figs. 2C, D; Table S2). Regarding MMR deficiency and chemosensitivity, Bertagnolli et al. reported on colon cancer that defect MMR, determined by MSI testing or MMR IHC, was significantly associated with improved DFS in stage III patients treated with adjuvant irinotecan, fluorouracil, and leucovorin [22]. Zaanan et al. reported that defect MMR was significantly associated with improved DFS in patients with stage III colon cancer receiving adjuvant 5-FU–oxaliplatin chemotherapy [23]. Sinicrope et al. reported on colon cancer that MMR-deficient tumors with suspected germline mutations were associated with improved DFS after 5-FU-based treatment compared with sporadic tumors where no benefit was observed [1]. As for endometrial cancer, Kato et al. have recently reported that response rate of first-line platinum-based chemotherapy in evaluable cases was slightly higher in MMR-deficient cases (determined by MMR IHC; 67% vs. 44%, p = 0.34), and MMR-deficient cases had significantly better progression-free and overall survival compared with MMR-intact cases [24]. In the current study, however, MMR deficiency in itself showed only a tendency towards favorable overall survival without statistical significance (Fig. 2A). This discrepancy may be due to the difference of adjuvant treatment strategies between institutions. Regarding MMR and sensitivity to adjuvant radiotherapy in endometrial cancer, Resnick et al. reported a significant increase in overall (p = 0.003) and progression-free (p = 0.004) survival in a subgroup of patients with MMR-deficient, non-endometrioid tumors treated with adjuvant radiotherapy compared to those with intact MMR [25]. Further studies are warranted to clarify the prognostic significance of MMR status in endometrial carcinoma. In addition to the factor of PLS against sporadic, our survival analyses showed that younger age, endometrioid histology, G1, superficial myometrial invasion, absent LVI, early FIGO stage, and no adjuvant therapy received were also significant prognostic factors for favorable overall survival. In the subsequent multivariate analysis, PLS against sporadic was not left significant (Table 4), keeping in line with our above findings that PLS showed significant correlations with other prognostic factors, i.e. younger age, superficial myometrial invasion, and early stage (Table 3). There are two publications which adopted the same criteria as ours to predict Lynch syndrome among unselected endometrial cancers. Bruegl et al. reported 10.5% [13] and Buchanan et al. reported 8.3% [15] as predicted Lynch syndrome, both frequencies being equivalent to ours (13%). The latter also examined germline MMR mutation, and sensitivity, specificity, positive predictive value, and negative predictive value were reported as 100%, 94.5%, 36.2%, and 100%, respectively. Accordingly, all of the germline MMR mutation carriers in the current study are supposed to be categorized in the PLS group. Therefore, although positive predictive value may be not high, we consider that the clinicopathologic associations of the current study are significant. We also examined incorporating MSI status into the criteria to predict PLS as follows [26,27]. Patients with MSI-L/MSS and intact all MMR expression, and patients with MSI-H, MLH1 loss, and MLH1 methylation A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 231 232 A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 References Table 4 Multivariate analysis of favorable prognostic factors for overall survival. Prognostic factor HR 95% CI P-value PLS (vs. Sporadic) Age b 60 (vs. ≥60) Endometrioid (vs. Non-endometrioid) G1 (vs. Others) MI ≤1/2 (vs. N1/2) LVI absent (vs. present) FIGO stage I/II (vs. III/IV) AdjTx not received (vs. received) 2.0E-06 0.94 0.23 0.61 1.01 0.93 0.35 0.19 0–∞ 0.49–1.80 0.11–0.50 0.29–1.32 0.44–2.32 0.47–1.82 0.15–0.85 0.05–0.74 0.97 0.85 0.0002 0.21 0.97 0.82 0.02 0.017 Abbreviations: HR = hazard ratio; 95% CI = 95% confidence interval; PLS = probable Lynch syndrome; MI = myometrial invasion; LVI = lymphovascular space invasion; FIGO = International Federation of Gynecology and Obstetrics; AdjTx = adjuvant therapy. were both considered sporadic. Patients with MSI-H and MSH2/MSH6/ PMS2 expression loss, and patients with MSI-H, MLH1 loss, and unmethylated MLH1 were both considered PLS. Consequently, 181 cases were classified as sporadic, and 16 cases (8%) were classified as PLS, designated as Sporadic/PLS (+MSI) (Table 2). Significant associations of PLS (+MSI) with Lynch syndrome-associated multiple cancer remained (p = 0.01; Table 3), but those with early stage or superficial myometrial invasion were lost (p = 0.37 and 0.51, respectively; Table 3). PLS (+MSI) showed only trends towards favorable overall survival compared to sporadic (+ MSI) (p = 0.19, data not shown). Whether or not MSI status should be incorporated into the criteria for PLS is beyond the scope of the present study. In conclusion, we have demonstrated here that categorizing patients into PLS and sporadic based on MMR analyses exhibited significantly better overall survival for PLS group compared to sporadic group. This prognostic impact for favorable survival of PLS over sporadic may be attributed to higher sensitivity to adjuvant therapies as well as indolent tumor characteristics such as early stage and superficial myometrial invasion. Accordingly, advanced-stage patients of PLS group may benefit more from adjuvant therapies compared to sporadic group. Altogether, our observations suggest that analyzing MMR status and searching for Lynch syndrome may identify a subset of patients with higher sensitivity to adjuvant therapies and favorable survival in endometrial carcinomas. Besides, recent emerging evidence suggests that differentially methylated microRNAs serve as markers for distinguishing MMRdeficient sporadic from Lynch-associated tumors and for early progression to tumorigenesis in endometrium and colon [28]. We believe that further genetic and epigenetic MMR analyses will provide useful information for formulating precision medicine in endometrial carcinomas. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ygyno.2015.11.032. Conflict of interest statement The authors have no conflict of interest to disclose. Acknowledgments This study was partially supported by the Grant-in-Aid for Scientific Research (No. 24592503) from the Ministry of Education, Culture, Sports, Science, and Technology, Tokyo, Japan. [1] F.A. Sinicrope, N.R. Foster, S.N. Thibodeau, S. Marsoni, G. Monges, R. Labianca, et al., DNA mismatch repair status and colon cancer recurrence and survival in clinical trials of 5-fluorouracil-based adjuvant therapy, Journal of the National Cancer Institute 103 (2011) 863–875. [2] H. Elsaleh, D. Joseph, F. Grieu, N. Zeps, N. Spry, B. Iacopetta, Association of tumour site and sex with survival benefit from adjuvant chemotherapy in colorectal cancer, Lancet 355 (2000) 1745–1750. [3] C.M. Ribic, D.J. Sargent, M.J. Moore, S.N. Thibodeau, A.J. French, R.M. Goldberg, et al., Tumor microsatellite-instability status as a predictor of benefit from fluorouracilbased adjuvant chemotherapy for colon cancer, N. Engl. J. Med. 349 (2003) 247–257. [4] J.B. Basil, P.J. Goodfellow, J.S. Rader, D.G. Mutch, T.J. Herzog, Clinical significance of microsatellite instability in endometrial carcinoma, Cancer 89 (2000) 1758–1764. [5] D. Black, R.A. Soslow, D.A. Levine, C. Tornos, S.C. Chen, A.J. Hummer, et al., Clinicopathologic significance of defective DNA mismatch repair in endometrial carcinoma, J. Clin. Oncol. 24 (2006) 1745–1753. [6] I. Zighelboim, P.J. Goodfellow, F. Gao, R.K. Gibb, M.A. Powell, J.S. Rader, et al., Microsatellite instability and epigenetic inactivation of MLH1 and outcome of patients with endometrial carcinomas of the endometrioid type, J. Clin. Oncol. 25 (2007) 2042–2048. [7] H.J. Mackay, S. Gallinger, M.S. Tsao, C.M. McLachlin, D. Tu, K. Keiser, et al., Prognostic value of microsatellite instability (MSI) and PTEN expression in women with endometrial cancer: results from studies of the NCIC Clinical Trials Group (NCIC CTG), Eur. J. Cancer 46 (2010) 1365–1373. [8] H. Arabi, H. Guan, S. Kumar, M. Cote, S. Bandyopadhyay, C. Bryant, et al., Impact of microsatellite instability (MSI) on survival in high grade endometrial carcinoma, Gynecol. Oncol. 113 (2009) 153–158. [9] I. Ruiz, M. Martin-Arruti, E. Lopez-Lopez, A. Garcia-Orad, Lack of association between deficient mismatch repair expression and outcome in endometrial carcinomas of the endometrioid type, Gynecol. Oncol. 134 (2014) 20–23. [10] D.S. Tan, C. Rothermundt, K. Thomas, E. Bancroft, R. Eeles, S. Shanley, et al., “BRCAness” syndrome in ovarian cancer: a case–control study describing the clinical features and outcome of patients with epithelial ovarian cancer associated with BRCA1 and BRCA2 mutations, J. Clin. Oncol. 26 (2008) 5530–5536. [11] K. Alsop, S. Fereday, C. Meldrum, A. deFazio, C. Emmanuel, J. George, et al., BRCA mutation frequency and patterns of treatment response in BRCA mutation-positive women with ovarian cancer: a report from the Australian Ovarian Cancer Study Group, J. Clin. Oncol. 30 (2012) 2654–2663. [12] A. Akiyama-Abe, T. Minaguchi, Y. Nakamura, H. Michikami, A. Shikama, S. Nakao, et al., Loss of PTEN expression is an independent predictor of favourable survival in endometrial carcinomas, Br. J. Cancer 109 (2013) 1703–1710. [13] A.S. Bruegl, B. Djordjevic, B. Batte, M. Daniels, B. Fellman, D. Urbauer, et al., Evaluation of clinical criteria for the identification of lynch syndrome among unselected patients with endometrial cancer, Cancer Prev. Res. 7 (2014) 686–697. [14] A.S. Bruegl, B. Djordjevic, D.L. Urbauer, S.N. Westin, P.T. Soliman, K.H. Lu, et al., Utility of MLH1 methylation analysis in the clinical evaluation of lynch syndrome in women with endometrial cancer, Curr. Pharm. Des. 20 (2014) 1655–1663. [15] D.D. Buchanan, Y.Y. Tan, M.D. Walsh, M. Clendenning, A.M. Metcalf, K. Ferguson, et al., Tumor mismatch repair immunohistochemistry and DNA MLH1 methylation testing of patients with endometrial cancer diagnosed at age younger than 60 years optimizes triage for population-level germline mismatch repair gene mutation testing, J. Clin. Oncol. 32 (2014) 90–100. [16] A. Abe, T. Minaguchi, H. Ochi, M. Onuki, S. Okada, K. Matsumoto, et al., PIK3CA overexpression is a possible prognostic factor for favorable survival in ovarian clear cell carcinoma, Human Pathology 44 (2013) 199–207. [17] C.R. Boland, S.N. Thibodeau, S.R. Hamilton, D. Sidransky, J.R. Eshleman, R.W. Burt, et al., A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer, Cancer Research 58 (1998) 5248–5257. [18] T. Kanaya, S. Kyo, Y. Maida, N. Yatabe, M. Tanaka, M. Nakamura, et al., Frequent hypermethylation of MLH1 promoter in normal endometrium of patients with endometrial cancers, Oncogene 22 (2003) 2352–2360. [19] G.S. Nelson, A. Pink, S. Lee, G. Han, D. Morris, T. Ogilvie, et al., MMR deficiency is common in high-grade endometrioid carcinomas and is associated with an unfavorable outcome, Gynecol. Oncol. 131 (2013) 309–314. [20] B. Djordjevic, B.A. Barkoh, R. Luthra, R.R. Broaddus, Relationship between PTEN, DNA mismatch repair, and tumor histotype in endometrial carcinoma: retained positive expression of PTEN preferentially identifies sporadic non-endometrioid carcinomas, Modern Pathology: An Official Journal of the United States and Canadian Academy of Pathology, Inc. 26 (2013) 1401–1412. [21] M. Huang, B. Djordjevic, M.S. Yates, D. Urbauer, C. Sun, J. Burzawa, et al., Molecular pathogenesis of endometrial cancers in patients with Lynch syndrome, Cancer 119 (2013) 3027–3033. Fig. 2. Kaplan-Meier curves for survival according to the results of MMR analyses in patients with endometrial carcinomas. A, OS of patients with deficient MMR (n = 62) vs. intact MMR (n = 159) in the whole group of patients; B, OS of PLS group (n = 28) vs. sporadic group (n = 193) in the whole group of patients; C, OS of PLS group (n = 4) vs. sporadic group (n = 63) in the subset of advanced-stage (FIGO stage III-IV) patients; D, OS of PLS group (n = 24) vs. sporadic group (n = 130) in the subset of early-stage (FIGO stage I-II) patients; E, DFS of PLS group (n = 5) vs. sporadic group (n = 91) in the patients who received any adjuvant therapy; F, DFS of PLS group (n = 23) vs. sporadic group (n = 102) in the patients who received no adjuvant therapy; G, DFS of PLS group (n = 2) vs. sporadic group (n = 36) in the patients who received adjuvant chemotherapy alone; H, DFS of PLS group (n = 3) vs. sporadic group (n = 34) in the patients who received adjuvant radiotherapy alone. A. Shikama et al. / Gynecologic Oncology 140 (2016) 226–233 [22] M.M. Bertagnolli, D. Niedzwiecki, C.C. Compton, H.P. Hahn, M. Hall, B. Damas, et al., Microsatellite instability predicts improved response to adjuvant therapy with irinotecan, fluorouracil, and leucovorin in stage III colon cancer: Cancer and Leukemia Group B Protocol 89803, J. Clin. Oncol. 27 (2009) 1814–1821. [23] A. Zaanan, J.F. Flejou, J.F. Emile, G.G. Des, P. Cuilliere-Dartigues, D. Malka, et al., Defective mismatch repair status as a prognostic biomarker of disease-free survival in stage III colon cancer patients treated with adjuvant FOLFOX chemotherapy, Clin. Cancer Res. 17 (2011) 7470–7478. [24] M. Kato, M. Takano, M. Miyamoto, N. Sasaki, T. Goto, H. Tsuda, et al., DNA mismatch repair-related protein loss as a prognostic factor in endometrial cancers, J. Gynecol. Oncol. 26 (2015) 40–45. [25] K.E. Resnick, W.L. Frankel, C.D. Morrison, J.M. Fowler, L.J. Copeland, J. Stephens, et al., Mismatch repair status and outcomes after adjuvant therapy in patients with surgically staged endometrial cancer, Gynecol. Oncol. 117 (2010) 234–238. 233 [26] C.H. Leenen, M.G. van Lier, H.C. van Doorn, M.E. van Leerdam, S.G. Kooi, J. de Waard, et al., Prospective evaluation of molecular screening for lynch syndrome in patients with endometrial cancer b/= 70 years, Gynecol. Oncol. 125 (2012) 414–420. [27] P.J. Goodfellow, C.C. Billingsley, H.A. Lankes, S. Ali, D.E. Cohn, R.J. Broaddus, et al., Combined microsatellite instability, MLH1 methylation analysis, and immunohistochemistry for lynch syndrome screening in endometrial cancers from GOG210: an NRG Oncology and Gynecologic Oncology Group Study, J. Clin. Oncol. (2015). [28] S. Kaur, J.E. Lotsari, S. Al-Sohaily, J. Warusavitarne, M.R. Kohonen-Corish, P. Peltomaki, Identification of subgroup-specific miRNA patterns by epigenetic profiling of sporadic and lynch syndrome-associated colorectal and endometrial carcinoma, Clin. Epigenetics 7 (2015) 20.