119 Differences in Patterns of TP53 and KRAS2 Mutations in a Large Series of Endometrial Carcinomas with or without Microsatellite Instability Elizabeth M. Swisher, M.D.1 Stacia Peiffer-Schneider, Ph.D.1 David G. Mutch, M.D.1 Thomas J. Herzog, M.D.1 Janet S. Rader, M.D.1 Alaa Elbendary, M.D.1 Paul J. Goodfellow, Ph.D.2 1 Department of Obstetrics and Gynecology, Washington University School of Medicine, St. Louis, Missouri. 2 Department of Obstetrics and Gynecology and Department of Surgery, St. Louis, Missouri. Presented at the Society for Gynecologic Oncologists’ 29th Annual Meeting, Orlando, Florida, February 7–11, 1998. Supported in part by Grant CA71754. The assembly and characterization of the endometrial carcinoma specimens investigated was supported by a grant from the American Cancer Society (RPG-96042-MGO). The authors thank Kathryn Trinkaus, Ph.D., and the Biostatistics Department of the Washington University Cancer Center for statistical assistance in evaluation of the significance of the difference in the number of doubly mutant tumors. Address for reprints: Washington UniversitySchool of Medicine, Department of Obstetrics and Gynecology, Box 8064, 1 Barnes-Jewish Hospital Plaza, St. Louis, MO 63110. Received April 1, 1998; accepted May 26, 1998. © 1999 American Cancer Society BACKGROUND. The correlation between tumor microsatellite instability (MSI) and the accumulation of mutations in KRAS2 and TP53 is uncertain. The authors evaluated the TP53 and KRAS2 genes for mutations in sporadic endometrial carcinomas with microsatellite instability (MSI) and matched MSI negative controls to determine whether defective DNA mismatch repair impacts the patterns of mutations in two genes known to be involved in endometrial tumorigenesis. METHODS. Twenty-five MSI positive endometrial tumors were matched for prognostic factors with 25 MSI negative tumors. Mutations in codon 12 and 13 of KRAS2 were assessed using a polymerase chain reaction (PCR) restriction assay. Mutations in codon 61 of KRAS2 and exons 5– 8 of TP53 were evaluated using PCR amplification and single strand conformation variant (SSCV) analysis. All variants were subjected to direct DNA sequencing. RESULTS. KRAS2 and TP53 mutations were identified with equal frequency in the MSI positive and MSI negative groups. For TP53, the authors identified 5 mutations (20%) in the MSI positive specimens compared with 8 (32%) in the MSI negative group. For KRAS2, there were identified 8 mutations (32%) in the MSI positive specimens compared with 7 (28%) in the MSI negative tumors. The mutational spectra evident from sequence analysis of TP53 and KRAS2 variants were similar between MSI negative and MSI positive tumors. MSI negative tumors were more likely to have mutations in both KRAS2 and TP53 than MSI positive tumors, which were rarely mutant in both genes (P ⫽ 0.046). CONCLUSIONS. Although the overall frequency of mutations in TP53 and KRAS2 is similar, MSI positive tumors are less likely to have mutations in both genes than MSI negative sporadic endometrial carcinomas. MSI positive and MSI negative endometrial carcinomas may arise through distinct genetic pathways. Cancer 1999;85:119 –26. © 1999 American Cancer Society. KEYWORDS: endometrial carcinoma, microsatellite instability, TP53, KRAS2, mismatch repair, replication error. T he genetic events that lead to the formation of endometrial carcinoma remain elusive. Like colorectal carcinoma, endometrial carcinoma often arises from a well-defined precancerous lesion (atypical endometrial hyperplasia). Although a genetic model for colorectal tumorigenesis is well established, only a few genes that have important roles in endometrial tumorigenesis are known. Tumor behavior and risk factors vary for different histologies of endometrial carcinoma. Thus, there may be more than one oncogenic pathway in endometrial tumorigenesis. A high proportion of sporadic endometrial carcinomas (15–25%) are characterized by instability at simple DNA repeat sequences 120 CANCER January 1, 1999 / Volume 85 / Number 1 called microsatellites.1–3 A defect in the DNA mismatch repair system leads to a failure to correct slipped-strand mispairing during replication, resulting in polyclonal replication errors or microsatellite instability (MSI).4,5 MSI positive cell lines with defective DNA mismatch repair accumulate mutations at a much higher rate than MSI negative cell lines. Thus, MSI positive tumors have been said to have a mutator phenotype. It is presumed that this mutator phenotype will confer an oncogenic advantage to MSI positive tumor cells, a consequence of more rapid accumulation of mutations in genes that are important in tumorigenesis. Hereditary nonpolyposis colorectal carcinoma (HNPCC) is a familial cancer syndrome in which tumors frequently exhibit the MSI phenotype. Colorectal carcinoma and endometrial carcinoma are the first and second most common cancers in HNPCC. These patients carry germline mutations in 1 of 4 DNA mismatch repair genes (MSH2, MLH1, PMS1, and PMS2).6 –10 Many MSI positive sporadic colorectal carcinomas have somatic mutations in MSH2 or MLH1.11,12 On the other hand, the genetic basis for defective mismatch repair for the majority of MSI positive sporadic endometrial carcinomas is unknown.13 The genetic targets of defective mismatch repair in MSI positive tumors is poorly defined. MSI positive colorectal carcinomas have an increased incidence of frameshift mutations in the tumor growth factor (TGF)-␤2 receptor gene and in the BAX gene, which are important in the TP53-directed apoptotic pathway.14 –17 Recent reports by Tashiro et al. and Teng et al. found a higher rate of mutation of the PTEN gene in MSI positive endometrial carcinomas compared with MSI negative endometrial carcinomas.18,19 Some, but not all, studies have shown MSI positive colorectal carcinomas to have a lower rate of TP53 and KRAS2 mutations than MSI negative colorectal carcinomas.20 –22 We hypothesized that sporadic MSI positive endometrial carcinomas arise from genetically distinct pathways more often than MSI negative endometrial tumors, and this difference would be reflected in the pattern of mutations in genes commonly involved in endometrial tumorigenesis. Two of the genes most frequently implicated in endometrial carcinoma are the tumor suppressor gene TP53 and the protooncogene KRAS2, which are mutated in 10 –30% of cases.23–27 We performed a mutation analysis for these two genes in MSI negative and MSI positive tumors. We matched the MSI positive and MSI negative specimens for prognostic factors known to be important in endometrial carcinoma and which might impact the frequency of mutation in these two genes. Using DNA sequence analysis, we sought to determine whether the specific types of mutations differed between MSI positive and MSI negative tumors. MATERIALS AND METHODS Specimens Tumor samples were obtained at the time of total abdominal hysterectomy and bilateral salpingo-oophorectomy for primary endometrial carcinoma at Washington University Medical Center from 1993 to the present. The study protocol had been approved by the institutional review board at Washington University, and all patients gave informed consent prior to participation. Tissue specimens were frozen at -70°C and DNA was prepared using proteinase K and phenol extraction. High neoplastic cellularity for the tumor tissues used as the source of DNA was histologically confirmed by one pathologist. Normal DNA was extracted from peripheral blood as previously described.28 For a few cases, normal myometrium was used as a source of normal DNA in lieu of blood samples. Polymerase chain reaction (PCR)– based analysis for microsatellite instability was performed as previously described, using 7 simple sequence repeat markers (FABP3, D2S123, D3S1029, D5S346, D10S212, D17S261, and D18S55)2 along with the BAT-26 and MLH1(TA)nTn markers.13 Tumors were classified as MSI positive if they exhibited instability at two or more markers. Characteristics of the MSI positive specimens have been previously described.13 Clinical and pathologic information was used to match 25 MSI positive specimens with 25 MSI negative endometrial tumors that served as controls. The known prognostic factors in endometrial carcinoma of grade, histology, surgical stage, and patient race were matched between groups. Surgical stage was defined according to the International Federation of Obstetrics and Gynecology (FIGO) 1988 criteria minus information on cytology.29 Age was matched between groups as closely as possible after other prognostic factors were matched. TP53 Mutation Analysis Exons 5– 8 of the TP53 gene were amplified from genomic tumor DNA using primers derived from the flanking intronic sequence of each exon. Primer sequences, reaction conditions, and PCR conditions have been detailed previously.30 Twenty-seven rounds of PCR cycling were performed, incorporating ␣32PdCTP (Amersham, Arlington Heights, IL). Each amplimer was denatured at 94°C for 2 minutes, chilled on ice, and electrophoresed through nondenaturing gels of 0.5⫻ MDE (J. T. Baker) with and without 5% glycerol at room temperature for 14 –20 hours at TP53 and KRAS2 in Endometrial Carcinomas with Microsatellite Instability/Swisher et al. 4-5 W. Gels were dried and products were visualized by autoradiography at -70°C. PCR amplification and single-strand conformation variant (SSCV) analysis was repeated on all variants and their corresponding normal cellular DNA to verify the somatic nature of the variation. KRAS2 Mutation Analysis KRAS2 codons 12 and 13 were evaluated by PCR amplification and restriction digestion as previously described.31 Briefly, the two assays use mutagenic primers, which introduce a new restriction site next to the target codon. Any missense mutation in the codon causes loss of the introduced restriction site. For codon 12, mutagenic primers introduce a restriction site for BstN I in wild-type alleles, resulting in a diagnostic 114 base pair fragment. For codon 13, mutagenic primers insert a Hae III restriction site, resulting in an 83 base pair fragment. The mutant allele is identified as a restriction fragment of different size by electrophoresis through a 6% polyacrylamide gel. Gels were stained with ethidium bromide, and products were visualized under ultraviolet illumination. All codon 12 and 13 KRAS2 mutants identified in this assay were then subjected to PCR amplification and SSCV analysis of exon 1 using primers and PCR conditions previously described.32 Variant bands were cut out from the SSCV gel, eluted, and reamplified. Direct DNA sequencing was performed following PCR purification as described below. KRAS2 codon 61 mutations were identified by PCR amplification and SSCV analysis of exon 2. Primers for exon 2 amplification were 5´-cccttctcaggattcctaca-3´ and 5´-agaaagccctccccagtcct-3´. PCR reactions were performed in 10 L containing 20 ng DNA, 0.07 mM each deoxy-NTP, 0.2 mM of each primer, 0.5 units Taq polymerase, 0.1 mCi ␣32PdCTP (Amersham), 10 mM Tris HCl (pH 8.5), 1.5 mM MgCl2, and 100 mM KCl. Twenty-seven rounds of PCR cycling were performed, with each cycle consisting of 30 seconds of denaturing at 94°C, 30 seconds of annealing at 59°C, and 30 seconds of extension at 72°C. SSCV analysis was performed in identical fashion to that described for TP53. All variants and the corresponding normal cellular DNA were subjected to repeat PCR amplification and SSCV analysis. Variants present in normal as well as tumor DNA were assumed to be polymorphisms, as KRAS2 germline mutations have not been described. These variants were not subject to sequence analysis. Only variants present in tumor samples and not found in corresponding normal DNA were sequenced. 121 Variant Analysis Variant TP53 and KRAS2 bands were excised from the dried SSCV gels. DNA was eluted, reamplified, and electrophoresed on a 1% low-melt agarose gel or purified with a PCR product purification kit (Qiagen, Chatsworth, CA). Templates were sequenced directly using the dideoxy chain termination method with radiolabeled primers as previously described,33 or using the fmol Sequencing Kit (Promega, Madison, WI). In cases of previously undescribed TP53 mutations, variant sequences were confirmed by sequencing the complementary strand or by restriction digest when appropriate. Statistics The statistical significance of differences between the MSI positive and MSI negative control groups was evaluated with the McNemar chi-square test for paired observations. RESULTS Table 1 describes the clinical features of the specimens investigated and summarizes the mutation data for the 25 paired MSI negative and MSI positive tumors. The race of the patient was matched exactly; 3 patients in each group were African American (12%) and the remaining patients were white. The median age (66 years) was the same in each group. International Federation of Gynecology and Obstetrics (FIGO) surgical stage was matched exactly between specimens, except for Case 1033, which was a Stage IB tumor (a tumor invading less than one-half the thickness of the myometrium) with only 1 mm invasion and was matched with a Stage IA tumor (with no invasion). The breakdown of the 25 MSI positive specimens by surgical stage is as follows: 5 Stage IA, 11 Stage IB, 6 Stage IC (invading greater than or equal to one-half the myometrial thickness without serosal penetration), 1 Stage IIA (involving endocervical mucosa), and 2 IIIC (involving pelvic or para-aortic lymph nodes). Nine tumors were ⱖStage IC (36%). In each group, all tumors were adenocarcinomas with endometrioid histology. MSI positive tumors were matched to MSI negative endometrioid carcinomas of the same grade, without regard to the presence of squamous elements. Grade was matched nearly exactly. Of 25 tumors in each group, 7 (28%) were FIGO Grade 1, 15 (60%) were FIGO Grade 2, and 3 (12%) were FIGO Grade 3. KRAS2 and TP53 mutations were identified in equal frequency in the MSI positive and MSI negative groups (Table 2). Of 25 specimens, 12 (48%) of the MSI positive and 10 (40%) of the MSI negative tumors had 2 3 2 3 3 1 2 2 2 2 2 2 2 2 1 2 2 1 1 2 1 1 1 2 2 37 82 1027 1033 1044 1061 1062 1064 1070 1078 1085 1086 1087 1089 1094 1096 1100 1102 1110 1112 1125 1128 1133 1141 1145 IA IC IB IB* IIIC IB IC IB IB IIIC IB IIA IC IB IB IB IC IA IB IC IB IA IA IC IA Tumor stage 244 144 273 181 274 5 8 5 8 Codon 7 Exon cag(gln) to tag(stop) insert A before cat cgc(arg) to cac(his) gtt(val) to ggt(gly) ggc(gly) to gtc(val) Nucleotide change GAT GTT CAC GAT AGC GGA 12 61 12 13 12 GTT GTT Nucleotide 12 12 12 Codon 69 1042 1117 1029 1111 1093 1072 1129 1058 1119 1028 38 1066 92 1065 1043 1059 1045 1126 1049 1081 1109 1046 1054 169 Specimen I.D. 2 3 2 3 3 1 2 2 2 2 2 2 2 2 1 2 2 1 1 2 1 1 1 2 2 Tumor grade Tumor stage Exon IA IC IB IA* IIIC IB IC IB IB IIIC IB IIA IC IB IB IB IC IA IB IC IB IA IA IC IA 5 138 280 173 272 8 8 5 splice 248 192 266 5 7 6 8 Codon gcc(ala) to gtc(val) aga(arg) to aaa(lys) gtg(val) to ggg(gly) gtg(val) to ggg(gly) intron 4 del A of AG cgg(arg) to tgg(trp) cag(gln) to tag(stop) gga(gly) to gaa(glu) Nucleotide change TP53 mutation KRAS2 mutation TP53 mutation 12 12 13 12 12 61 12 Codon GTT GTT GAC TGT GTT CAC GAT Nucleotide KRAS2 mutation MSI: microsatellite instability. a Each MSI positive tumor is matched with the MSI negative tumor to its right. Grade and stage for MSI negative specimens are identical to those given for their matched MSI positive specimens. The exception is 1033, with 1 mm invasion (Stage 1B*), which was matched with a Stage 1A* tumor 1029. Wild-type KRAS2 codons 12, 13, and 61 are GGT, GGC, and CAA, respectively. Tumor grade Specimen I.D. MSI negative tumors MSI positive tumors TABLE 1 Clinical Features and Mutation Data for MSI Positive Endometrial Tumors and Their Matched MSI Negative Controlsa 122 CANCER January 1, 1999 / Volume 85 / Number 1 TP53 and KRAS2 in Endometrial Carcinomas with Microsatellite Instability/Swisher et al. TABLE 2 Mutations in MSI Positive and MSI Negative Endometrial Tumors No. (%) of cases Gene Mutation type MSI positive MSI negative TP53 TP53 TP53 TP53 KRAS2 KRAS2 KRAS2 Transition Transversion Frameshift Total Transition Transversion Total 2 2 1 5 (20%) 5 3 8 (36%) 5 2 1 8 (32%) 5 2 7 (32%) MSI: microsatellite instability. a KRAS2 and/or TP53 mutation. For TP53, we identified 5 mutations (20%) in the MSI positive specimens compared with 8 (36%) in the MSI negative group. For KRAS2, we identified 8 mutations (36%) in the MSI positive specimens compared with 7 (32%) in the MSI negative tumors. Although the overall frequency of mutations in both KRAS2 and TP53 were similar, the distribution of these mutations varied between the two groups (Table 1). In the MSI positive group, significantly more tumors (5, 20%) were found to have mutations in both KRAS2 and TP53 than in the MSI negative group (1, 5%, P⫽ 0.046). Representative mutation analysis and DNA sequence analysis for TP53 and KRAS2 variants are demonstrated in Figures 1 and 2, respectively. The presence of a TP53 mutation was significantly associated with higher grade tumors. In 14 Grade 1 tumors, 1 TP53 mutation was identified (7.1%). In contrast, among 6 Grade 3 tumors, 4 TP53 mutations were identified (67%, P ⫽ 0.01, two-tailed). There was a trend toward higher frequency of KRAS2 mutations in higher grade tumors; 2 of 14 Grade 1 and 3 of 6 Grade 3 tumors exhibited KRAS2 mutations (P ⫽ 0.13). TP53 and KRAS2 mutation status did not correlate with surgical stage. The TP53 and KRAS2 mutation spectra variants were similar in MSI negative and MSI positive tumors (Table 2). Two of 13 TP53 mutations (15%) were insertion or deletion (frameshift) mutations, with 1 occurring in each group. The percentages of transition and transversion mutations were likewise equally distributed between the two groups for both KRAS2 and TP53. The fraction of transition mutations occurring at CpG dinucleotides was also equal for both TP53 and KRAS2 between MSI positive and MSI negative tumors. Of the 15 KRAS2 mutations identified, 2 occurred in codon 13 and 2 occurred in codon 61. One KRAS2 codon 13 and one codon 61 mutation were found in 123 each group. The number of TP53 mutations identified in each exon examined was also similar between MSI negative and MSI positive tumors. DISCUSSION MSI is diagnostic for a mutator phenotype attributable to defective DNA mismatch repair. Mutation rates are increased for MSI positive cells and are presumed to confer an oncogenic advantage.4,5 The MSI phenotype has been noted in atypical hyperplasia adjacent to MSI positive endometrial carcinoma, and as such appears to be an early event in endometrial carcinoma development.34 Because MSI appears to be an early event in the development of a subset of endometrial carcinomas, the mutator phenotype is likely to influence the entire cascade of genetic events that contribute to tumor formation and progression. The genes that accumulate mutations in tumor cells with the mutator phenotype are, for the most part, unknown. MSI positive cancers could accumulate mutations in genes known to be involved in endometrial tumorigenesis at an increased frequency, and/or accumulate mutations in genes not involved in MSI negative tumors. We hypothesized that the MSI phenotype would influence the mutation spectra of genes known to be involved in endometrial tumorigenesis, in particular TP53 and KRAS2. We observed high mutation rates in both KRAS2 and TP53 in this study, but we did not detect a difference in either the frequency or the types of mutations in these genes. There was, however, a statistically significant difference in the proportion of tumors that were found to have mutations in both KRAS2 and TP53 (P ⫽ 0.046). Because TP53 mutations are more frequent in higher grade and stage endometrial carcinomas as well as in certain histologic types, we matched our MSI negative and MSI positive groups for these features in an effort to control for differences in mutation rates that could be attributed to these factors. The difference noted in this study in the distribution of KRAS2 and TP53 mutations may not have been evident if we had not closely matched our tumors for many prognostic features. Caduff et al. reported the results of a study of 10 MSI positive endometrial carcinomas investigated for TP53 mutations by immunohistochemistry and for mutations in codon 12 of KRAS2 using a PCR amplification and restriction digest assay.39 They found no difference in mutation rates in MSI positive tumors compared with a large group of unmatched MSI negative tumors. In our study, which relied on different mutation detection methods and matched MSI positive and negative tumors, we were able to detect a difference in the pattern of KRAS2 and TP53 mutations in both MSI posi- 124 CANCER January 1, 1999 / Volume 85 / Number 1 FIGURE 1. Representative SSCV and DNA sequence analysis for exon 8 of TP53 is shown. Arrows indicate variant conformers. N: normal (wild-type) DNA pattern; M: tumor specimens containing variant conformers representing TP53 mutations. DNA sequence analysis demonstrates the wild-type sequence AGA at codon 280 (arginine). Tumor 1058 demonstrates a G-to-A transition, resulting in the sequence AAA at codon 280 (lysine). FIGURE 2. Detection of KRAS2 mutations is shown. Arrows indicate restriction fragments or conformational variants associated with mutations. (A) Polymerase chain reaction amplification and restriction analysis to detect codon 12 and 13 mutations is shown. N: normal (wild-type) pattern; M: tumor specimen heterozygous for a codon 12 or 13 mutation. (B) Single strand conformation variant analysis and DNA sequence analysis (of the noncoding DNA strand) to detect KRAS2 codon 61 mutations is shown. N: normal cellular DNA; T: tumor DNA from specimen 1087. The wild-type sequence for codon 61 is CAA (glutamine). The conformational variant from tumor 1087 has a CAC (histidine) at codon 61 (GTG on the noncoding strand). tive and MSI negative tumors. The fact that more MSI negative tumors have mutations in both genes, whereas MSI positive tumors tend to have mutations in only one of these genes, supports the hypothesis that endometrial tumors with defective mismatch repair arise via a different genetic pathway or pathways than endometrial tumors with intact mismatch repair. Several groups have compared KRAS2 and TP53 mutation frequency between MSI positive and MSI negative colorectal carcinomas. Methods for the detection of MSI and for mutation analysis have differed among studies.5 Some authors have reported an inverse correlation between the MSI phenotype and KRAS2 and/or TP53 mutation rates.20 –22 Other groups found no difference in KRAS2 and TP53 mutation rates between MSI positive and MSI negative colorectal carcinomas.35–37 Konishi et al. noted an inverse correlation with MSI phenotype and TP53 and KRAS2 mutations in tumors from Japanese HNPCC patients. However, this correlation did not hold for sporadic MSI positive colorectal carcinomas.38 It is difficult to determine whether inherited and sporadic tumors with defective DNA mismatch repair differ in how they accumulate mutations. Not all studies specified whether the MSI positive colorectal carcinomas were sporadic or familial. In studies in which only sporadic TP53 and KRAS2 in Endometrial Carcinomas with Microsatellite Instability/Swisher et al. colorectal carcinomas were investigated, there was no association between the frequency of TP53 and KRAS2 mutation and the MSI phenotype.36 –38 Thus, sporadic MSI positive colorectal carcinomas with somatic mutations in DNA mismatch repair genes may progress through different oncogenic pathways than HNPCC tumors with constitutional mutations in mismatch repair genes. If inherited and sporadic colorectal carcinomas differ in how they accumulate mutations, it might explain some of the discordant results regarding TP53 and KRAS2 mutation frequency and MSI phenotype in colorectal carcinomas. The observed frequency of TP53 and KRAS2 mutations in this study (26% and 30%, respectively) is high. Other studies have noted TP53 and KRAS2 mutations in 9 –30% of endometrial carcinomas.23–27,40 The high mutation frequency that we observed is consistent with the nature of the tumors we investigated and the sensitive mutation detection methods used. TP53 mutation is correlated with higher grade in our study, as has been noted previously by other investigators.23,24 Poorly differentiated endometrial carcinomas are more likely to have defective mismatch repair,3 and, as expected, our series includes a high percentage of Grade 2 and 3 tumors. We believe that the composition of our tumor panel explains the high incidence of TP53 mutations in this study. Most studies of KRAS2 in endometrial carcinoma have been restricted to analysis of codon 12.27,32,39,41 Greater than one-fourth of the mutations (4 of 15) that we observed occurred in codons 13 and 61, increasing our total KRAS2 mutation rate from 22% to 30%. To the best of our knowledge, KRAS2 mutations in codon 61 have not previously been reported in endometrial carcinoma. Other investigators studying endometrial carcinoma have used dot blot oligohybridization or direct DNA sequencing to identify codon 61 mutations.26,42 The PCR and SSCV methods we used may be more sensitive in identifying codon 61 mutations. One codon 61 mutation was found in the MSI positive and 1 in the MSI negative tumors. The uncommon codon 13 and 61 mutations in KRAS2 seem to be distributed equally between the cancers with intact and those with defective mismatch repair. Only a few genetic targets of defective mismatch repair in MSI positive cancers have been identified. Colon carcinomas from HNPCC patients have a very high rate of mutation in the TGF-␤ type II receptor gene.14 –16 The majority of these mutations involve insertion or deletion in a string of 10 adenosine residues, a typical target of defective mismatch repair. The recently discovered PTEN tumor suppressor gene has been shown to be mutated in a very high percentage of MSI positive endometrial carcinomas (12 of 14, 125 85%), in contrast to MSI negative endometrial carcinomas (4 of 12, 33%).18 PTEN mutations in MSI positive endometrial carcinomas include both insertion and deletion mutations in repeats as well as single base substitutions.18,19 It is unclear why this gene seems to be a particular target for mutation in endometrial carcinomas with defective mismatch repair and whether this predilection for PTEN mutations applies to other MSI positive human cancers. Endometrial carcinomas with MSI are less likely than MSI negative tumors to have mutations in both TP53 and KRAS2. This difference in the mutation pattern of these two genes supports the hypothesis that at least a subset of endometrial tumors with defective DNA mismatch repair arise via a different oncogenic pathway than tumors without MSI. An improved understanding of the genetic targets of defective mismatch repair will clarify how the MSI phenotype contributes to oncogenesis in endometrial carcinoma. REFERENCES 1. Burks RT, Kessis TD, Cho KR, Hedrick L. Microsatellite instability in endometrial carcinoma. Oncogene 1994;9: 1163– 6. 2. Peiffer SL, Herzog TJ, Tribune DJ, Mutch DG, Gersell DJ, Goodfellow PJ. Allelic loss of sequences from the long arm of chromosome 10 and replication errors in endometrial cancers. Cancer Res 1995;55:1922– 6. 3. Kobayashi K, Sagae S, Kudo R, Saito H, Koi S, Nakamura Y. Microsatellite instability in endometrial carcinomas: frequent replication errors in tumors of early onset and/or of poorly differentiated type. Genes Chromosomes Cancer 1995; 14:128 –32. 4. Parsons R, Li G-M, Longley MJ, Fang W-H, Papadopoulos N, Jen J, et al. Hypermutability and mismatch repair deficiency in RER⫹ tumor cells. Cell 1993;75:1227–36. 5. Bocker T, Diermann J, Friedl W, Gebert J, Holinski-Feder E, Karner-Hanusch J, et al. Microsatellite instability analysis: a multicenter study for reliabilty and quality control. Cancer Res 1997;57:4739 – 43. 6. Bronner CE, Baker SM, Morrison PT, Warren G, Smith LG, Lescoe MK, et al. Mutation in the DNA mismatch repair gene homologue hMLH1 is associated with hereditary nonpolyposis colon cancer. Nature 1994;368:258 – 61. 7. Leach FS, Nicolaides NC, Papadopoulos N, Liu B, Jen J, Parsons R, et al. Mutations of a mutS homolog in hereditary nonpolyposis colorectal cancer. Cell 1993;75:1215–25. 8. Nicolaides NC, Papadopoulos N, Liu B, Wei Y-F, Carter KC, Ruben SM, et al. Mutations of two PMS homologues in hereditary nonpolyposis colon cancer. Nature 1994;371:75– 80. 9. Fishel R, Lescoe MK, Rao MRS, Copeland NG, Jenkins NA, Garber J, et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 1993;75:1027–38. 10. Papadopoulos N, Nicolaides NC, Wei Y-F, Ruben SM, Carter KC, Rosen CA, et al. Mutation of a mutL homolog in hereditary colon cancer. Science 1994;263:1625–9. 126 CANCER January 1, 1999 / Volume 85 / Number 1 11. Herfarth K-F, Kodner IJ, Whelan AJ, Ivanovich JL, Bracamontes JR, Wells SA Jr., et al. Mutations in MLH1 are more frequent than in MSH2 in sporadic colorectal cancers with microsatellite instability. Genes Chromosomes Cancer 1997; 18:42–9. 12. Liu B, Nicolaides NC, Markowitz S, Wilson JKV, Parsons RE, Jen J, et al. Mismatch repair gene defects in sporadic colorectal cancers with microsatellite instability. Nat Genet 1995; 9:48 –55. 13. Kowalski LD, Mutch DG, Herzog TJ, Rader JS, Goodfellow PJ. Mutational analysis of MLH1 and MSH2 in 25 prospectivelyacquired RER⫹ endometrial cancers. Genes Chromosomes Cancer 1997;18:219 –27. 14. Lu S-L, Zhang W-C, Akiyama Y, Nomizu T, Yuasa Y. Genomic structure of the transforming growth factor-␤ type II receptor gene and its mutations in hereditary nonpolyposis colorectal cancers. Cancer Res 1996;56:4595– 8. 15. Markowitz S, Wang J, Myeroff L, Parsons R, Sun L, Lutterbaugh J, et al. Inactivation of the type II TGF-␤ receptor in colon cancer cells with microsatellite instability. Science 1995;268:1336 – 8. 16. Parsons R, Myerhoff LL, Liu B, Wilson JKV, Markowitz SD, Kinzler KW, et al. Microsatellite instability and mutations of the transforming growth factor-␤ type II receptor gene in colorectal cancer. Cancer Res 1995;55:5548 –50. 17. Rampino N, Yamamoto H, Ionov Y, Li Y, Sawai H, Reed JC, et al. Somatic frameshift mutations in the BAX gene in colon cancers of the microsatellite mutator phenotype. Science 1997;275:967–9. 18. Tashiro H, Blazes MS, Wu R, Cho KR, Bose S, Wang SI, et al. Mutations in PTEN are frequent in endometrial carcinoma but rare in other common gynecological malignancies. Cancer Res 1997;57:3935– 40. 19. Teng DHF, Hu R, Lin H, Davis T, Iliev D, Frye C, et al. MMAC1/PTEN mutations in primary tumor specimens and tumor cell lines. Cancer Res 1997;57:5221–5. 20. Ionov Y, Peinado MA, Malkhosyan S, Shibata D, Perucho M. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993;363:558 – 61. 21. Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1995;260:816 –9. 22. Bocker T, Schlegel J, Kullmann F, Stumm G, Zirngibl H, Epplen JT, et al. Genomic instability in colorectal carcinomas: comparison of different evaluation methods and their biological significance. Gastroenterology 1996;179:15–9. 23. Kohler MF, Berchuck A, Davidoff AM, Humphrey PA, Dodge RK, Iglehart JD, et al. Overexpression and mutation of p53 in endometrial carcinoma. Cancer Res 1992;52:1622–7. 24. Enomoto T, Fujita M, Inoue M, Rice JM, Nakajima R, Tanizawa O, et al. Alterations of the p53 tumor suppressor gene and its association with activation of the c-K-ras-2 protooncogene in premalignant and malignant lesions of the human uterine endometrium. Cancer Res 1993;53:1883– 8. 25. Duggan BD, Felix JC, Muderspach LI, Tsao J-L, Shibata DK. Early mutational activation of the c-Ki-ras oncogene in endometrial cancer. Cancer Res 1994;54:1604 –7. 26. Ignar-Trowbridge D, Risinger JI, Dent GA, Kohler M, Berchuck A, McLachlan JA, et al. Mutations of the Ki-ras oncogene in endometrial carcinoma. Am J Obstet Gynecol 1992; 167:227–32. 27. Sasaki H, Nishii H, Takahashi H, Tada A, Furusato M, Terashima Y, et al. Mutation of the Ki-ras protooncogene in human endometrial hyperplasia and carcinoma. Cancer Res 1993;53:1906 –10. 28. Lahiri DK, Nurnberger JIA. A rapid non-enzymatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res 1991;19:5444. 29. FIGO stages: 1988 revision. Gynecol Oncol 1989;35:125–7. 30. Herfarth KK-F, Wick MR, Marshall HN, Gartner E, Lum S, Moley JF. Absence of TP53 alterations in pheochromocytomas and medullary thyroid cancers. Genes Chromosomes Cancer 1997;20:24 –9. 31. Lin S-Y, Chen P-H, Wang C-K, Liu J-D, Siauw C-P, Chen Y-J, et al. Mutation analysis of K-ras oncogenes in gastroenterologic cancers by the amplified created restriction method. Am J Clin Pathol 1993;100:656 – 89. 32. Fujimoto I, Shimizu Y, Hirai Y, Chen J-T, Teshima H, Hasumi K, et al. Studies on ras oncogene activation in endometrial carcinoma. Gynecol Oncol 1993;48:196 –202. 33. Donis-Keller H, Dou S, Chi D, Carlson KM, Toshima K, Lairmore TC, et al. Mutations in the RET proto-oncogene are associated with MEN-2A and FMTC. Hum Mol Genet 1993;2:851– 6. 34. Jovanovic AS, Boynton KA, Mutter GL. Uteri of women with endometrial carcinoma contain a histopathological spectrum of monoclonal putative precancers, some with microsatellite instability. Cancer Res 1996;56:1917–21. 35. Aaltonen LA, Peltomaki P, Mecklin JP, Jarvinen H, Jass JR, Green JS, et al. Replication errors in benign and malignant tumors from hereditary nonpolyposis colorectal cancer patients. Cancer Res 1994;54:1645– 8. 36. Ilyas M, Tomlinson IPM, Novelli MR, Hanby A, Bodmer WF, Talbot IC. Clinicopathological features and p53 expression in left-sided sporadic colorectal cancers with and without microsatellite instability. J Pathol 1996;179:370 –5. 37. Watatani M, Yoshida T, Kuroda K, Ieda S, Yasutomi M. Allelic loss of chromosome 17p, mutation of the p53 gene, and microsatellite instability in right- and left- sided colorectal cancer. Cancer 1996;77(Suppl):1688 –93. 38. Konishi M, Kikuchi-Yanoshita R, Tanaka K, Muraoka M, Onda A, Okumura Y, et al. Molecular nature of colon tumors in hereditary nonpolyposis colon cancer, familial polyposis and sporadic colon cancer. Gastroenterology 1996;111:307– 17. 39. Caduff RF, Johnston CM, Svoboda-Newman SM, Poy EL, Merajver SD, Frank TS. Clinical and pathological significance of microsatellite instability in sporadic endometrial carcinoma. Am J Pathol 1996;148:1671– 8. 40. Ito K, Watanabe K, Nasim S, Sasano H, Sato S, Yajima A, et al. Prognostic significance of p53 overexpression in endometrial cancer. Cancer Res 1994;54:4667–70. 41. Ito K, Watanabe K, Nasim S, Sasano H, Sato S, Yajima A, et al. K-ras point mutations in endometrial carcinoma: effect on outcome is dependent on age. Gynecol Oncol 1996;63: 238 – 46. 42. Enomoto T, Fujita M, Inoue M, Nomura T, Shroyer KR. Alteration of the p53 tumor suppressor gene and activation of the c-K-ras-2 protooncogene in endometrial adenocarcinoma from Colorado. Anat Pathol 1993;103:224 –30.