Clinicopathologic Significance of Defective DNA Mismatch Repair in Endometrial Carcinoma

VOLUME 24 䡠 NUMBER 11 䡠 APRIL 10 2006 JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T Clinicopathologic Significance of Defective DNA Mismatch Repair in Endometrial Carcinoma Destin Black, Robert A. Soslow, Douglas A. Levine, Carmen Tornos, Shirley C. Chen, Amanda J. Hummer, Faina Bogomolniy, Narciso Olvera, Richard R. Barakat, and Jeff Boyd From the Departments of Surgery, Pathology, Epidemiology and Biostatistics, and Medicine, Memorial SloanKettering Cancer Center, New York, NY. Submitted September 6, 2005; accepted December 19, 2005. Supported by Grant No. R01 CA100272 from the National Institutes of Health, Bethesda, MD. Presented at the 41st Annual Meeting of the American Society of Clinical Oncology, Orlando, FL, May 13-17, 2005. Terms in blue are defined in the glossary, found at the end of this article and online at www.jco.org. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article. Address reprint requests to Jeff Boyd, PhD, Department of Surgery, Box 201, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, New York, NY 10021; e-mail: boydj@mskcc.org. © 2006 by American Society of Clinical Oncology A B S T R A C T Purpose Defective DNA mismatch repair is commonly present in sporadic manifestations of gastrointestinal, endometrial, and other cancers. The pathognomonic molecular manifestation of this repair defect is microsatellite instability (MSI). Here, we test the hypothesis that MSI predicts the clinicopathologic features of endometrial carcinoma. Patients and Methods A retrospective cohort of 473 patients treated for endometrial carcinoma at this institution was identified. All cases were reviewed by a gynecologic pathologist, and clinical information was abstracted from medical records. Using consensus criteria, DNA samples from nontumor and tumor tissue pairs were genotyped for MSI. Associations between MSI status and pathologic and clinical variables were assessed. Results Ninety-three (20%) of 473 tumors were MSI⫹. In the MSI⫹ tumor group compared with the MSI⫺ tumor group, the proportion of advanced compared with early-stage tumors was higher (92% v 81%; P ⫽ .01), as was the proportion of tumors of endometrioid compared with nonendometrioid histologic subtype (94% v 23%; P ⫽ .001), and the proportion of tumors with myometrial invasion compared with those with none (92% v 78%; P ⫽ .01). By multivariate analyses, disease-free survival (hazard ratio, 0.3; 95% CI, 0.2 to 0.7) and disease-specific survival (hazard ratio, 0.3; 95% CI, 0.1 to 0.8) were significantly improved in patients with MSI⫹ tumors. Conclusion In endometrial carcinoma, the presence of MSI was independently associated with a more favorable clinical outcome. 0732-183X/06/2411-1745/$20.00 DOI: 10.1200/JCO.2005.04.1574 J Clin Oncol 24:1745-1753. © 2006 by American Society of Clinical Oncology INTRODUCTION Endometrial carcinoma is the most common gynecologic malignancy in the United States, with 41,200 new cases and 7,350 deaths estimated in 2005.1 During the last 15 years, several important advances have been made in defining the molecular genetic alterations that contribute to endometrial tumorigenesis.2 Among the most well characterized of these is a defect in the DNA mismatch repair system, affecting approximately 20% of tumors, which occurs predominantly in the type I class of endometrial cancers.2 Defective mismatch repair was initially described in sporadic colorectal tumors and those associated with the hereditary nonpolyposis colorectal cancer (HNPCC) syndrome.3-5 The HNPCC syndrome is associated with autosomal dominant cancer predisposition characterized by the development of colorectal carcinoma at an early age, as well as tumors of the upper urologic, gastrointestinal, and female reproductive tracts.6 Endometrial carcinoma is the most common extracolonic cancer in HNPCC, with an estimated frequency of 40% to 60% in female mutation carriers.7-9 Germline mutation of one or another of the mismatch repair genes MSH2, MLH1, or MSH6 is believed to account for the great majority of HNPCC kindreds,10 whereas somatic promoter hypermethylation and silencing of the MLH1 gene appears to be the primary mechanism for loss of mismatch repair in endometrial and other sporadic cancers.11 In either case, failure of this system leads to the accumulation of single base-pair mismatches, as well as small insertions and deletions in tandem repeats known as microsatellites, which manifests as microsatellite instability (MSI). This phenomenon is readily quantitated by gel electrophoresis of polymerase chain reaction (PCR) products containing 1745 Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Black et al microsatellites, and the observation of mobility shifts present in tumor compared with normal DNA products from the same individual. A specific panel of microsatellite markers has been adopted through National Cancer Institute (NCI; Bethesda, MD) –sponsored consensus recommendations, and found to have high sensitivity for the detection of defective mismatch repair.12 In colorectal carcinoma, several studies have demonstrated that tumors exhibiting MSI have distinctive molecular and clinicopathologic profiles. These tumors are associated with preferential proximal colon location, poorly differentiated histology, and extensive peritumoral lymphocyte infiltration.3,4,13,14 In addition, studies have consistently demonstrated that colorectal cancers with MSI are associated with an improved prognosis compared with tumors with intact mismatch repair.9,15-18 In contrast, the clinicopathologic significance of MSI in endometrial cancer has not yet been clarified. From the period 1995 to 2004, we identified 13 publications19-31 that reported on a range of 42 to 259 unselected endometrial carcinoma cases that included data on MSI status and any correlation with clinicopathologic characteristics of the cases. The most consistent findings were related to tumor grade and histology, with five studies19,20,26,28,29 reporting higher grade in MSI⫹ tumors compared with MSI⫺ tumors and four studies22,23,27,30 reporting no difference, while tumors of endometrioid histology were overrepresented among MSI⫹ tumors in three studies21,22,24 and no correlation seen in one.27 Neither age at diagnosis nor extent of myometrial invasion were associated with MSI status in seven19-23,26,27 and four20,26,27,30 studies, respectively. Two studies examined race and tumor MSI status, with one finding no assocation,27 and a second finding a significantly increased incidence of MSI⫹ tumors, by multivariate analysis, in white patients.24 More conflicting are reported associations of tumor MSI status with stage and clinical outcome. Disease stage was not significantly associated with MSI status in seven studies,19-23,27,30 whereas MSI⫹ tumors were associated with advanced stage disease in two studies,28,29 and with lower stage disease, by multivariate analysis, in one study.24 With respect to clinical outcome, disease-free survival was not associated with MSI status in four studies,24,25,28,29 but MSI⫹ tumors were associated with worse disease-free survival, by multivariate analysis, in one study.26 In terms of overall survival, MSI⫹ tumors were associated with improved survival in one study,27 but worse survival (although not independent of tumor grade) in another study,20 and there was no correlation observed in six studies.23-25,28-30 The purpose of this study was to perform a comprehensive investigation of the possible associations of all of these clinical and pathologic variables with tumor MSI status in a large group of unselected patients with endometrial carcinoma. PATIENTS AND METHODS Study Participants This study was approved by the institutional review board of the Memorial Sloan-Kettering Cancer Center (New York, NY). Endometrial carcinomas and corresponding nontumor tissues were obtained from 473 patients, unselected for age or personal or family history of cancer, treated at this institution from 1992 to 2003. The 946 nontumor and tumor tissue samples consisted of snap-frozen tissue and peripheral blood lymphocyte samples from 190 patients, and paraffin-embedded tissue samples from 283 patients. Only patients with carcinomas were included in the study. All patients under1746 went exploratory laparotomy for diagnosis and staging followed by adjuvant radiation and chemotherapy if indicated. Standard post-treatment surveillance included serial physical examination with Pap smears. Recurrence of disease was confirmed by biopsy. Clinical information was abstracted from hospital records and included age at diagnosis, race, surgical stage, adjuvant treatment, and disease status. The patients were staged according to International Federation of Gynecology and Obstetrics (FIGO) criteria. When complete surgical staging had not been performed, a clinical stage was assigned using FIGO criteria from 1971. All tumors were subjected to additional review by a gynecologic pathologist for tumor grade (based on 1994 WHO criteria), histologic subtype, depth of myometrial invasion, and presence of lymphovascular invasion. DNA Isolation and Genotyping Procedures DNA was isolated from formalin-fixed, paraffin-embedded tissue specimens after the tumor area was identified by a pathologist on hematoxylin and eosin–stained sections. Tumor cells and nontumor cells were macroscopically dissected with a discoid-cleoid instrument (American Eagle Instruments, Missoula, MT). Genomic DNA was isolated using a DNeasy Tissue Kit (Qiagen, Valencia, CA). Tumor DNA from frozen tissue specimens and corresponding normal DNA from peripheral blood lymphocytes were isolated in a similar fashion. Paired DNA samples from endometrial carcinoma and nontumor tissues were genotyped for MSI using five microsatellite markers: BAT25, BAT26, D2S123, D5S346, D17S250. These markers were chosen on the basis of the consensus recommendation of the NCI workshop on the development of international guidelines for the evaluation of MSI in colorectal cancer.12 Table 1. Clinical and Pathologic Characteristics of Study Population Patients (n ⫽ 473) Characteristic Stage IA IB/IC II III/IV Tumor grade 1 2 3 Histology Endometrioid Serous Clear cell Undifferentiated Myometrial invasion None ⬍ 1/2 ⬎ 1/2 Complete Race White African American Other Unknown Lymphovascular invasion No Yes MSI status Negative Positive No. % 76 258 40 97 16 55 8 21 225 114 134 48 24 28 379 71 18 5 80 15 4 1 90 265 98 19 19 56 21 4 421 35 8 9 89 7 2 2 344 129 73 27 380 93 80 20 Abbreviation: MSI, microsatellite instability. JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Mismatch Repair–Deficient Endometrial Cancers Fig 1. Representative examples of microsatellite instability (MSI) in endometrial carcinoma. Shown are autoradiographic images of polymerase chain reaction (PCR) products after gel electrophoresis, demonstrating MSI at the five microsatellite markers used in this study: (A) BAT 25, (B) BAT 26, (C) D17S250, (D) D2S123, (E) D5S346. N, PCR product from nontumor tissue DNA; T, PCR product from corresponding tumor DNA. The arrows to the right indicate product mobility shifts characteristic of MSI. Reactions for PCR amplification were carried out in a volume of 10 ␮L containing 50 ng of genomic DNA, and 200 ␮mol/L each deoxynucleotide triphosphate, 1.6 ␮mol/L each primer, 1 ␮L of 10⫻ PCR buffer (Applied Biosystems, Foster City, CA), and 0.1 ␮L of Gold Taq polymerase (Applied Biosystems). The forward primer was end-labeled with [␥-33P]. Amplification was then performed using a Perkin-Elmer (San Francisco, CA) 9600 thermal cycler with the following settings: 1 cycle at 95°C for 10 minutes, followed by 35 cycles at 95°C for 20 seconds, 55°C for 20 seconds, and 72°C for 30 seconds, and 1 cycle at 72°C for 7 minutes. PCR products were diluted 1:4 in denaturing loading buffer (95% formamide, 10 mmol/L EDTA [pH, 8.0], 0.02% xylene cyanol FF, and 0.02% bromphenol blue), heated at 95°C for 5 minutes, and placed on ice. Electrophoresis was performed in a 6% polyacrylamide gel for 2.5 hours at 80 W. Gels were dried and exposed to Hyperfilm MP autoradiography film (Amersham Corp, Louisville, CO) for 24 to 48 hours. Determination of MSI status was made according to the presence of mutant alleles in tumor DNA compared with nontumor tissue DNA. Tumors were classified as MSI⫹ if at least two (40%) of five markers displayed evidence of mutant alleles in tumor DNA when compared with nontumor tissue DNA.12 Statistical Analyses Comparisons between MSI status and categoric clinicopathologic characteristics were performed using the ␹2 test statistic. Test of association of MSI status and age was performed using the t test. Disease-free survival (DFS) was defined as time to progression of disease. Disease-specific survival (DSS) was defined as time to cancer-related death. To estimate DFS and DSS, competing risk methods were used to control for non– cancerrelated deaths. Competing risks analysis was performed using Gray’s method.32,33 The univariate hazard ratios were estimated using the competing risk model. For each end point, only those variables with a P value of less than .05 by Gray’s method in univariate analysis were included in the initial multivariate model. The final model was determined by backward stepwise regression procedures using the Cox proportional hazards competing risk model including only the significant variables. Statistical analyses were performed using SAS Software (SAS Institute, Cary, NC) and the cmprsk R package (http://biowww.dfci.harvard.edu/⬃gray). RESULTS Of 473 endometrial cancer patients included in the study, the mean age was 63 years (range, 27 to 89 years for those with MSI⫺ tumors and 34 to 88 years for those with MSI⫹ tumors), with a median follow-up time for living patients of 5 years (range, 0.02 to 12.2 years). Clinical and pathologic information associated with these patients is summarized in Table 1. Ninety-three (20%) of 473 tumors were classified as MSI⫹. A representative example of MSI in an endometrial carcinoma is illustrated in Figure 1. After controlling for those clinicopathologic variables associated with MSI status, we did not observe a disproportionate ratio of MSI⫹ to MSI⫺ cases in paraffin-embedded compared with frozen tissues. Table 2. Relationships Between MSI Status and Clinicopathologic Characteristics MSI⫺ (n ⫽ 380) Characteristic Age, years Mean Standard deviation Stage IA IB/IC II III/IV Tumor grade 1 2 3 Histology Endometrioid Serous Clear cell Undifferentiated Myometrial invasion None ⬍ 1/2 ⬎ 1/2 Complete Race White African American Other Lymphovascular invasion No Yes No. % MSI⫹ ( n ⫽ 93) No. % P .36 63 11 62 13 .01 71 197 31 81 91 76 78 84 7 61 9 16 9 24 22 16 182 85 113 81 75 84 43 29 21 19 25 16 292 68 17 3 77 96 94 60 7 3 1 2 23 4 6 40 83 205 76 16 92 77 78 84 7 60 22 3 8 23 22 16 336 30 6 80 86 75 85 5 2 20 14 25 281 99 82 77 63 30 18 23 .15 .001 .01 .66 .23 Abbreviation: MSI, microsatellite instability. 1747 www.jco.org Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Black et al Table 3. Univariate Disease-Free Survival Analysis Patients With Disease by 3 Years After Treatment Characteristic % Age (continuous) Race White/Asian African American Histology Endometrioid Serous Clear cell Undifferentiated Grade 1 2 3 Stage IA IB/IC II III/IV Myometrial invasion None ⬍ 1/2 ⬎ 1/2 Complete Lymphovascular invasion No Yes Other treatment None Radiation Radiation/chemotherapy Chemotherapy MSI status Negative Positive NA 95% CI Hazard Ratio 95% CI P 1.03 1.01 to 1.05 .003 .47 0.6 to 2.6 19 21 15% to 23% 7% to 35% 1.0 1.3 11 50 35 NR 7% to 14% 38% to 63% 12% to 59% 1.0 5.8 4.8 10.5 3 19 43 1% to 6% 2% to 27% 34% to 52% 1.0 7.2 19.5 6 8 14 63 3% to 9% 4% to 11% 2% to 25% 53% to 73% 1.0 NE 1.7 2.5 11 12 30 74 4% to 17% 8% to 17% 21% to 40% 52% to 97% 1.0 1.2 3.3 10.1 12 35 9% to 16% 26% to 44% 1.0 3.0 9 13 35 78 4% to 14% 9% to 18% 11% to 60% 65% to 92% 1.0 1.5 4.9 14.9 21 9 17% to 25% 2% to 15% 1.0 0.4 .001 3.7 to 9.0 2.2 to 10.5 3.4 to 32.4 .001 3.0 to 17.8 8.5 to 44.7 .001 1.1 to 2.6 2.1 to 2.9 .001 0.6 to 2.5 1.5 to 6.9 4.4 to 23.0 .001 2.0 to 4.6 .001 0.8 to 2.9 4.8 to 12.5 7.5 to 29.4 .005 0.2 to 0.8 Abbreviations: NA, not applicable; NR, not reached; NE, not estimable; MSI, microsatellite instability. The relationships between MSI and clinicopathologic features of the endometrial cancer cases are summarized in Table 2. Notable findings were that in the MSI⫹ tumor group compared with the MSI⫺ tumor group, the proportion of advanced (IB-IV) compared with early (IA) stage tumors was higher (92% v 81%; P ⫽ .01), as was the proportion of tumors of endometrioid compared with nonendometrioid histologic subtype (94% v 23%; P ⫽ .001), and the proportion of tumors with myometrial invasion compared with those with none (92% v 78%; P ⫽ .01). There were no significant differences between MSI status and age, tumor grade, race, or lymphovascular invasion. Using competing risks methodology to examine univariate predictors of DFS, all variables except race were significant predictors of DFS, as shown in Table 3. A highly significant difference in DFS was observed between patients with MSI⫹ tumors and those with MSI⫺ tumors, as shown in Figure 2 (P ⫽ .005); MSI⫹ status was a strong predictor of improved DFS. We next considered all clinicopathologic variables with a P value of less than .05 by univariate analysis to construct the initial multivariate model. After stepwise backwards regression procedures, high tumor grade, advanced surgi1748 cal stage, presence of myometrial invasion, and MSI⫺ tumor status remained significant predictors of progression in the multivariate DFS model (Table 4). With respect to DSS, competing risks methodology was used to show that all variables except race were significant predictors of DSS (Table 5). There was a significant difference in DSS when comparing patients with MSI⫹ and MSI–tumors, as shown in Figure 3 (P ⫽ .006); MSI⫹ status was also a strong predictor of DSS. All clinicopathologic variables with a P value of less than .05 by univariate analysis were then included in the initial multivariate model. After stepwise backwards regression procedures, high tumor grade, advanced surgical stage, and MSI–tumor status remained significant predictors of poorer DSS in the multivariate model (Table 6). The data shown in Table 2 suggest an association between surgical stage and MSI status. We observed a strong independent association (P ⫽ .01) between improved DSS and MSI⫹ tumors in patients with advanced stage disease (Fig 4). In contrast, there was no significant correlation between tumor MSI status and DSS in patients with early stage disease (P ⫽ .39). Additionally, we did not observe a difference in survival among patients with high-grade tumors according to MSI status, JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Mismatch Repair–Deficient Endometrial Cancers Fig 2. Disease-free survival by microsatellite instability (MSI) status. Solid black, disease progression, MSI–; solid red, disease progression, MSI⫹; dashed black, non– disease-specific death, MSI–; dashed red, non– diseasespecific death, MSI⫹. although this may have reflected inadequate power to detect such an association. DISCUSSION In this retrospective cohort of 473 endometrial carcinoma patients unselected for age or personal or family history of cancer, the proportion with tumors affected by defective mismatch repair was 20%, consistent with a mean of 25% reported in 13 other studies.19-31 We did not attempt to distinguish those cases with MSI resulting from somatic alteration of a mismatch repair gene from those with a germline mutation (ie, the HNPCC syndrome). The rate of germline mutation of DNA mismatch repair genes in unselected endometrial cancer patients remains unknown, although it may be inferred from the literature that this rate is substantially less than 5%. Further, there Table 4. Multivariate Disease-Free Survival Analysis Characteristic Grade 1 2 3 Stage IA IB/IC II III/IV Myometrial invasion None ⬍ 1/2 ⬎ 1/2 Complete MSI status Negative Positive Hazard Ratio 95% CI P .001 1.0 4.0 6.8 1.7 to 9.2 3.0 to 15.4 .001 1.0 NE 1.2 1.9 0.8 to 1.9 1.6 to 2.3 .01 1.0 0.7 1.2 2.1 0.4 to 1.5 0.6 to 2.5 0.9 to 4.9 .01 1.0 0.3 0.2 to 0.7 Abbreviations: NE, not estimable; MSI, microsatellite instability. is no evidence that the clinical and pathologic features of endometrial cancer cases from germline mutation carriers, other than a younger age of diagnosis, differ significantly from sporadic cases with MSI. Although the clinical and pathologic features commonly associated with colorectal cancers arising from defective DNA mismatch repair are relatively well established, the available data related to endometrial carcinomas affected by this molecular genetic alteration remain equivocal, as discussed in the Introduction. The present study represents the largest series of MSI-genotyped endometrial cancer patients reported to date, and revealed several significant clinicopathologic relationships. There were significant associations of MSI⫹ tumors with myometrial invasion and advanced pathologic stage. It is not readily apparent why myometrial invasion and advanced stage would correlate with MSI⫹ tumor status, the latter of which is an independent predictor of improved prognosis, when myometrial invasion and advanced stage are both independent predictors of poor prognosis. However, this observation is analogous to that made with respect to MSI⫹ colorectal carcinomas, which also have an improved prognosis compared with MSI–tumors, and are more likely to present with a significantly greater depth of tumor invasion.17 Among the other clinicopathologic features examined, there was a strong association of MSI with endometrioid histology, a relationship that has been noted consistently in studies that examined this characteristic.21,22,24 This is one factor that contributes to the general consensus that MSI⫹ endometrial carcinomas are appropriately classified as type I tumors.2 No significant associations were observed between MSI status and age, tumor grade, race, or lymphovascular invasion. Although the data on colorectal carcinomas clearly indicate an improved prognosis for patients with MSI⫹ tumors, there are fewer data on the impact of this molecular alteration on prognosis from endometrial cancer, and those that have been reported are equivocal. In this study, multivariate analyses indicated that the presence of tumor MSI was an independent predictor of improved DFS and DSS. For DSS, the substantial reduction in hazard ratio associated with MSI⫹ tumor status appeared to be driven by a strong effect on advanced-stage cases; when the data were analyzed separately according to early or advanced stage, tumor MSI status lost prognostic significance for early-stage tumors. Not unexpectedly, other wellestablished risk factors for endometrial cancer also achieved independent prognostic significance in this study, including tumor grade, stage, and presence of myometrial invasion for DFS, and tumor grade and stage for DSS. These data on improved prognosis associated with MSI⫹ tumors warrant a discussion of the mechanistic rationale for this phenomenon. There are at least two potential, and possibly overlapping, genetic explanations for this association. Genetic instability is believed to represent a hallmark of most human cancers, and MSI (also known as MIN) is one of two well characterized forms of genetic instability, the other being chromosomal instability (CIN), which leads to aneuploidy.34,35 These defects in the maintenance of genetic stability occur in a mutually exclusive fashion with high frequency in colorectal and other cancers, including endometrial carcinoma. Unlike cancers affected by CIN that develop aneuploidy, cancer cells that display MSI invariably have a diploid or near-diploid karyotype, but a nucleotide mutation rate two to three orders of magnitude higher than in mismatch repair proficient cells.36-39 Although some form of genetic instability is a 1749 www.jco.org Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Black et al Table 5. Univariate Disease-Specific Survival Analysis Patients With Disease-Specific Death by 3 Years Characteristic % Age (continuous) Race White/Asian African American Histology Endometrioid Serous Clear cell Undifferentiated Grade 1 2 3 Stage IA IB/IC II III/IV Myometrial invasion None ⬍ 1/2 ⬎ 1/2 Complete Lymphovascular invasion No Yes Other treatment None Radiation Radiation/chemotherapy Chemotherapy MSI status Negative Positive NA 95% CI Hazard Ratio 95% CI P 1.07 1.04 to 1.11 .001 12 16 9% to 15% 3% to 29% 1.0 1.4 0.7 to 3.1 .36 7 32 35 NR 4% to 9% 20% to 43% 8% to 63% 1.0 5.0 4.3 18.3 3.0 to 8.3 1.9 to 9.9 4.8 to 70.8 1 13 29 0% to 3% 6% to 19% 21% to 37% 1.0 6.4 17.0 2.6 to 15.8 7.4 to 39.0 4 5 11 40 1% to 6% 2% to 8% 1% to 22% 30% to 50% 1.0 NE 1.6 2.4 1.0 to 2.5 2.0 to 2.8 7 8 21 42 2% to 13% 4% to 11% 13% to 30% 19% to 65% 1.0 1.3 3.6 10.6 0.6 to 3.0 1.5 to 8.5 4.2 to 26.5 7 24 5% to 10% 16% to 32% 1.0 3.6 2.3 to 5.8 8 8 23 50 4% to 12% 4% to 11% 0% to 47% 33% to 67% 1.0 1.5 2.5 11.8 0.7 to 2.8 0.7 to 8.8 5.7 to 24.4 14 5 10% to 17% 0% to 10% 1.0 0.3 0.1 to 0.8 .001 .001 .001 .001 .001 .001 .006 Abbreviations: NA, not applicable; NR, not reached; NE, not estimable; MSI, microsatellite instability. likely requirement for cells to overcome the many intrinsic cellular barriers to neoplastic progression, the accumulated MSI-induced damage to the genome that ensues after many cancer cell divisions may ultimately rise above the threshold for cell viability, leading to the activation of apoptotic pathways or perhaps increased susceptibility to particular antineoplastic therapeutic agents.40 Thus, cancer cells with defective mismatch repair may behave in a more indolent fashion and/or ultimately lose their evolutionary fitness for survival before the host dies as a result of disease. A related hypothesis is that MSI⫹ endometrial cancers acquire a unique spectrum of additional somatic molecular genetic alterations that contributes to a less aggressive biologic behavior in this class of tumors. Among the well-characterized somatic mutations that occur in endometrial tumorigenesis, mutation and/or overexpression of the TP53 gene, hallmarks of type II endometrial cancers and associated with a poor prognosis, are consistently low or absent in MSI⫹ endometrial carcinomas.22,27,41-43 Conversely, mutational inactivation of the PTEN tumor suppressor gene, generally associated with type I endometrial cancers and an improved prognosis, is more common in 1750 Fig 3. Disease-specific survival by microsatellite instability (MSI) status for all patients. Solid black, disease-specific death, MSI–; Solid red, disease-specific death, MSI⫹; dashed black, non– disease-specific death, MSI–; dashed red, non– disease-specific death, MSI–. JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Mismatch Repair–Deficient Endometrial Cancers Table 6. Multivariate Disease-Specific Survival Analysis Characteristic Grade 1 2 3 Stage IA IB/IC II III/IV MSI status Negative Positive Hazard Ratio 95% CI P .001 1.0 3.6 7.0 1.4 to 9.6 2.9 to 17.1 .001 1.0 NE 1.2 1.9 0.7 to 2.1 1.5 to 2.3 .02 1.0 0.3 0.1 to 0.9 Abbreviations: NE, not estimable; MSI, microsatellite instability. endometrial cancers with MSI.27,41,43-45 Mutational activation of the KRAS oncogene, which has proven more difficult to classify with respect to relative incidence in type I versus type II endometrial cancers, is similarly subject to inconsistent correlation with MSI status; three studies found no correlation,20,43,45,46 whereas two reported a higher frequency in MSI⫹ tumors.42,47 Mutation in the CTNNB1 gene, encoding ␤-catenin, is consistently reported not to correlate with MSI status in endometrial cancers.45,48,49 Finally, it is axiomatic that MSI⫹ endometrial cancers have a high frequency of inactivating frameshift mutations in genes with coding region microsatellite repeats (eg, TGF␤RII, IGFIIR, BAX), and that this may represent the major mechanism through which MSI contributes to tumorigenesis.43,50 However, there are a large number of such genes in the genome, and it is not clear which of these, when mutated, play a causal or bystander role in tumorigenesis.51 A statistical meta-analysis of MSI⫹ human cancers revealed that most of the putative targets of MSI are probably bystanders, and that for endometrial carcinoma, the most highly-ranked Real Common Targets of MSI are TAF1B, AIM2, and SLC23A1.52 In addition, the ATR (ataxia telangiectasia and Rad3-related) gene was recently shown to undergo frameshift mutations in MSI⫹ endometrial tumors, with the functional consequence of abrogation of ATR-dependent Chk1 phosphorylation and cell cycle arrest following DNA damage.53 REFERENCES 1. Jemal A, Murray T, Ward E, et al: Cancer statistics, 2006. CA Cancer J Clin 56:106-130, 2006 2. Boyd J, Berchuck A: Oncogenes and tumorsuppressor genes, in Hoskins WJ, Perez CA, Young RC, et al (eds): Principles and Practice of Gynecologic Oncology. Philadelphia, PA, Lippincott Williams & Wilkins, 2004, pp 93-122 3. Ionov Y, Peinado MA, Malkhosyan S, et al: Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 363:558-561, 1993 4. Thibodeau SN, Bren G, Schaid D: Microsatellite instability in cancer of the proximal colon. Science 260:816-819, 1993 5. Aaltonen LA, Peltomaki P, Leach FS, et al: Clues to the pathogenesis of familial colorectal cancer. Science 260:812-816, 1993 Fig 4. Disease-specific survival by microsatellite instability (MSI) status for advanced stage disease patients. Solid black, disease-specific death, MSI–; Solid red, disease-specific death, MSI⫹; dashed black, non– disease-specific death, MSI–; dashed red, non– disease-specific death, MSI⫹. With respect to DNA content and global gene expression profiles, all studies that have examined the correlation of ploidy with MSI status consistently find that MSI⫹ tumors are predominantly diploid,30,43,45,49,54 and conversely, that aneuploid endometrial cancers do not exhibit MSI, consistent with the MSI (MIN)/ CIN dichotomy observed in colorectal carcinoma.34,35 The unsupervised analysis of global gene expression profiles of early stage endometrioid endometrial carcinomas reveals two distinct classes of tumors that segregate according to MSI status, implicating the alterations of discrete molecular pathways in these two categories of tumors.55 In summary, the existing literature together with data from this report provide sufficient evidence to warrant the inclusion of a majority of MSI⫹ tumors in the type I category of endometrial carcinomas and further, to conclude that MSI⫹ endometrial tumors have a unique biologic behavior compared with MSI–tumors. The classification of MSI as an independent prognostic variable is one of a very small number of instances in which a specific molecular marker has achieved this standard in endometrial carcinoma, and this finding warrants further study with respect to translational potential in the clinical management of this disease. 6. Lynch HT, Lynch JF: Lynch syndrome: History and current status. Dis Markers 20:181-198, 2004 7. Vasen HF, Wijnen JT, Menko FH, et al: Cancer risk in families with hereditary nonpolyposis colorectal cancer diagnosed by mutation analysis. Gastroenterology 110:1020-1027, 1996 8. Dunlop MG, Farrington SM, Carothers AD, et al: Cancer risk associated with germline DNA mismatch repair gene mutations. Hum Mol Genet 6:105-110, 1997 9. Aarnio M, Sankila R, Pukkala E, et al: Cancer risk in mutation carriers of DNA-mismatch-repair genes. Int J Cancer 81:214-218, 1999 10. Lynch HT, de la Chapelle A: Hereditary colorectal cancer. N Engl J Med 348:919-932, 2003 11. Peltomaki P: Deficient DNA mismatch repair: A common etiologic factor for colon cancer. Hum Mol Genet 10:735-740, 2001 12. Boland CR, Thibodeau SN, Hamilton SR, et al: A National Cancer Institute workshop on microsat- ellite instability for cancer detection and familial predisposition: Development of international criteria for the determination of microsatellite instability of colorectal cancer. Cancer Res 58:5248-5257, 1998 13. Kim H, Jen J, Vogelstein B, et al: Clinical and pathological characteristics of sporadic colorectal carcinomas with DNA replication errors in microsatellite sequences. Am J Pathol 145:148-156, 1994 14. Dolcetti R, Viel A, Doglioni C, et al: High prevalence of activated intraepithelial cytotoxic T lymphocytes and increased neoplastic cell apoptosis in colorectal carcinomas with microsatellite instability. Am J Pathol 154:1805-1813, 1999 15. Sankila R, Aaltonen LA, Jarvinen HJ, et al: Better survival rates in patients with MLH1-associated hereditary colorectal cancer. Gastroenterology 110: 682-687, 1996 16. Lynch HT, Smyrk T: Colorectal cancer, survival advantage, and hereditary nonpolyposis colorectal carcinoma. Gastroenterology 110:943-954, 1996 1751 www.jco.org Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Black et al 17. Gryfe R, Kim H, Hsieh ET, et al: Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med 342: 69-77, 2000 18. Lukish JR, Muro K, DeNobile J, et al: Prognostic significance of DNA replication errors in young patients with colorectal cancer. Ann Surg 227:51-56, 1998 19. Kobayashi K, Sagae S, Kudo R, et al: Microsatellite instability in endometrial carcinomas: Frequent replication errors in tumors of early onset and/or of poorly differentiated type. Genes Chromosomes Cancer 14:128-132, 1995 20. Caduff RF, Johnston CM, Svoboda-Newman SM, et al: Clinical and pathological significance of microsatellite instability in sporadic endometrial carcinoma. Am J Pathol 148:1671-1678, 1996 21. Catasus L, Machin P, Matias-Guiu X, et al: Microsatellite instability in endometrial carcinomas: Clinicopathologic correlations in a series of 42 cases. Hum Pathol 29:1160-1164, 1998 22. Tibiletti MG, Furlan D, Taborelli M, et al: Microsatellite instability in endometrial cancer: Relation to histological subtypes. Gynecol Oncol 73: 247-252, 1999 23. Wong YF, Ip TY, Chung KH, et al: Clinical and pathologic significance of microsatellite instability in endometrial cancer. Int J Gynecol Cancer 9:406-410, 1999 24. Basil JB, Goodfellow PJ, Rader JS, et al: Clinical significance of microsatellite instability in endometrial carcinoma. Cancer 89:1758-1764, 2000 25. MacDonald ND, Salvesen HB, Ryan A, et al: Frequency and prognostic impact of microsatellite instability in a large population-based study of endometrial carcinomas. Cancer Res 60:1750-1752, 2000 26. Fiumicino S, Ercoli A, Ferrandina G, et al: Microsatellite instability is an independent indicator of recurrence in sporadic stage I-II endometrial adenocarcinoma. J Clin Oncol 19:1008-1014, 2001 27. Maxwell GL, Risinger JI, Alvarez AA, et al: Favorable survival associated with microsatellite instability in endometrioid endometrial cancers. Obstet Gynecol 97:417-422, 2001 28. Baldinu P, Cossu A, Manca A, et al: Microsatellite instability and mutation analysis of candidate genes in unselected Sardinian patients with endometrial carcinoma. Cancer 94:3157-3168, 2002 29. Muresu R, Sini MC, Cossu A, et al: Chromosomal abnormalities and microsatellite instability in sporadic endometrial cancer. Eur J Cancer 38:18021809, 2002 30. Orbo A, Eklo K, Kopp M: A semiautomated test for microsatellite instability and its significance for the prognosis of sporadic endometrial cancer in northern Norway. Int J Gynecol Pathol 21:27-33, 2002 31. Pijnenborg JMA, Dam-de Veen GC, de Haan J, et al: Defective mismatch repair and the development of recurrent endometrial carcinoma. Gynecol Oncol 94:550-559, 2004 32. Gray RJ: A class of k-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 16:1141-1154, 1988 33. Fine JP, Gray RJ: A proportional hazards model for the sub-distribution of a competing risk. J Am Stat Assoc 94:496-509, 1999 34. Lengauer C, Kinzler KW, Vogelstein B: Genetic instability in colorectal cancers. Nature 386: 623-627, 1997 35. Rajagopalan H, Nowak MA, Vogelstein B, et al: The significance of unstable chromosomes in colorectal cancer. Nat Rev Cancer 3:695-708, 2003 36. Parsons R, Li GM, Longley MJ, et al: Hypermutability and mismatch repair deficiency in RER⫹ tumor cells. Cell 75:1227-1236, 1993 37. Bhattacharyya NP, Skandalis A, Ganesh A, et al: Mutator phenotypes in human colorectal carcinoma cell lines. Proc Natl Acad Sci U S A 91:63196323, 1994 38. Eshleman JR, Lang EZ, Bowerfind GK, et al: Increased mutation rate at the hprt locus accompanies microsatellite instability in colon cancer. Oncogene 10:33-37, 1995 39. Boyer JC, Umar A, Risinger JI, et al: Microsatellite instability, mismatch repair deficiency, and genetic defects in human cancer cell lines. Cancer Res 55:6063-6070, 1995 40. Cahill DP, Kinzler KW, Vogelstein B, et al: Genetic instability and Darwinian selection in tumours. Trends Cell Biol 9:M57-M60, 1999 41. Risinger JI, Hayes K, Maxwell GL, et al: PTEN mutation in endometrial cancers is associated with favorable clinical and pathologic characteristics. Clin Cancer Res 4:3005-3010, 1998 42. Lax SF, Kendall B, Tashiro H, et al: The frequency of p53, K-ras mutations, and microsatellite instability differs in uterine endometrioid and serous carcinoma. Evidence of distinct molecular pathways Cancer 88:814-824, 2000 43. MacDonald ND, Salvesen HB, Ryan A, et al: Molecular differences between RER⫹ and RER– sporadic endometrial carcinomas in a large populationbased series. Int J Gynecol Cancer 14:957-965, 2004 44. Tashiro H, Blazes MS, Wu R, et al: Mutations in PTEN are frequent in endometrial carcinoma but rare in other common gynecological malignancies. Cancer Res 57:3935-3940, 1997 45. Koul A, Willen R, Bendahl PO, et al: Distinct sets of gene alterations in endometrial carcinoma implicate alternate modes of carcinogenesis. Cancer 94:2369-2379, 2002 46. Sakamoto T, Murase T, Urushibata H, et al: Microsatellite instability and somatic mutations in endometrial carcinomas. Gynecol Oncol 71:53-58, 1998 47. Duggan BD, Felix JC, Muderspach LI, et al: Microsatellite instability in sporadic endometrial carcinoma. J Natl Cancer Inst 86:1216-1221, 1994 48. Mirabelli-Primdahl L, Gryfe R, Kim H, et al: Beta-catenin mutations are specific for colorectal carcinomas with microsatellite instability but occur in endometrial carcinomas irrespective of mutator pathway. Cancer Res 59:3346-3351, 1999 49. Machin P, Catasus L, Pons C, et al: CTNNB1 mutations and beta-catenin expression in endometrial carcinomas. Hum Pathol 33:206-212, 2002 50. Gurin CC, Federici MG, Kang L, et al: Causes and consequences of microsatellite instability in endometrial carcinoma. Cancer Res 59:452-466, 1999 51. Perucho M: Tumors with microsatellite instability: Many mutations, targets, and paradoxes. Oncogene 22:2223-2225, 2003 52. Woerner SM, Benner A, Sutter C, et al: Pathogenesis of DNA repair-deficient cancers: A statistical meta-analysis of putative Real Common Target genes. Oncogene 22:2226-2235, 2003 53. Lewis KA, Mullany S, Thomas B, et al: Heterozygous ATR mutations in mismatch repairdeficient cancer cells have functional significance. Cancer Res 65:7091-7095, 2005 54. Atkin NB: Microsatellite instability. Cytogenet Cell Genet 92:177-181, 2001 55. Risinger JI, Maxwell GL, Chandramouli GV, et al: Gene expression profiling of microsatellite unstable and microsatellite stable endometrial cancers indicates distinct pathways of aberrant signaling. Cancer Res 65:5031-5037, 2005 ■ ■ ■ Authors’ Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. Author Contributions Conception and design: Destin Black, Douglas A. Levine, Amanda J. Hummer, Jeff Boyd Financial support: Richard R. Barakat, Jeff Boyd Administrative support: Richard R. Barakat, Jeff Boyd Provision of study materials or patients: Robert A. Soslow, Douglas A. Levine, Carmen Tornos, Shirley C. Chen, Narciso Olvera, Richard R. Barakat Collection and assembly of data: Destin Black, Robert A. Soslow, Carmen Tornos, Shirley C. Chen, Faina Bogomolniy, Narciso Olvera Data analysis and interpretation: Destin Black, Robert A. Soslow, Douglas A. Levine, Carmen Tornos, Shirley C. Chen, Amanda J. Hummer, Faina Bogomolniy Manuscript writing: Destin Black, Amanda J. Hummer, Jeff Boyd Final approval of manuscript: Destin Black, Robert A. Soslow, Douglas A. Levine, Carmen Tornos, Shirley C. Chen, Amanda J. Hummer, Faina Bogomolniy, Narciso Olvera, Richard R. Barakat, Jeff Boyd 1752 JOURNAL OF CLINICAL ONCOLOGY Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved. Mismatch Repair–Deficient Endometrial Cancers GLOSSARY HNPCC (hereditary nonpolyposis colorectal cancer): Accounting for 5% of all colorectal cancer, HNPCC arises due to a hereditary defect in one of the mismatch repair genes. Cancer of the uterus (ie, womb or endometrium), gastrointestinal tract (ie, stomach, small intestine, or pancreas), urinary system (ie, kidney or ureter), and female reproductive organs (ie, ovary) may be cancers seen in some HNPCC families. A family history, colon examination, and genetic testing for mutations in DNA mismatch repair genes are helpful in identifying HNPCC. Mismatch repair: One of four major pathways of DNA repair in mammalian cells. Mismatch repair recognizes and corrects errors in DNA replication leading to single base-pair mismatches or insertions/ deletions in small repetitive tracts of DNA known as microsatellites. Microsatellite instability: Microsatellites are repeating Oncogene: Oncogenes are normal cellular genes whose protein products may function in any number of critical molecular pathways (and are referred to as proto-oncogenes when functioning in a nonmutated state) that are subject to activating, or gain-of-function mutations, which stimulate tumorigenesis. units in DNA of 1–5 basepairs that are ubiquitous, abundant, and repeated several times in eukaryotic genomes. The presence of microsatellites is associated with genomic instability, giving rise to mutations that involve the addition or subtraction of one or two repeat units. Tumor suppressor gene: The products of tumor suppressor genes normally function to inhibit aspects of the neoplastic phenotype, and are subject to inactivating, or loss-of-function mutations, which contribute to tumorigenesis. 1753 www.jco.org Downloaded from jco.ascopubs.org on December 31, 2015. For personal use only. No other uses without permission. Copyright © 2006 American Society of Clinical Oncology. All rights reserved.