IJC International Journal of Cancer

Transcription

1 IJC International Journal of Cancer The benefit of microsatellite instability is attenuated by chemotherapy in stage II and stage III gastric cancer: Results from a large cohort with subgroup analyses Soo Young Kim 1, Yoon Young Choi 1, Hyun Beak Shin 1, Ara Jo 2, Hyeji Choi 2, Sang Hyuk Seo 1, Hui-Jae Bang 1, Jae-Ho Cheong 1, Woo Jin Hyung 1, Sung Hoon Noh 1,2 and Ji Yeong An 1 1 Department of Surgery, Yonsei University College of Medicine, Seoul, Republic of Korea 2 Brain Korea 21 Plus Project for Medical Science, Yonsei University Health System, Yonsei University College of Medicine, Seoul, Republic of Korea We previously reported that the prognosis of microsatellite instability high (MSI-H) gastric cancer is similar to that of MSIlow/microsatellite stable (MSI-L/MSS) gastric cancer. The reason for this seemed to be related to the effects of chemotherapy. To verify this hypothesis, we expanded the study population and reanalyzed the prognosis of MSI-H gastric cancer. Data from 1,276 patients with Stage II and III gastric cancer who underwent gastrectomy with curative intent between January 2005 and June 2010 were reviewed. The prognosis of MSI-H tumors in comparison with MSI-L/MSS tumors was analyzed, according to the administration of chemotherapy and other clinicopathologic features. A total of 361 (28.3%) patients did not receive chemotherapy (MSI-H 5 47 and MSI-L/MSS 5 314), whereas 915 (71.7%) patients did receive chemotherapy (MSI-H 5 58 and MSI-L/MSS 5 857). The hazard ratio of MSI-H versus MSI-L/MSS was 0.49 (95% confidence interval: , p ) when chemotherapy was not received and 1.16 (95% confidence interval: , p ) when chemotherapy was received. In subgroup analyses, the prognosis of MSI-H was better in Stage III, women, with lymph node metastasis, and undifferentiated histology subgroups when chemotherapy was not received. However, in patients treated with chemotherapy, prognosis was worse for MSI-H tumors in Stage III, undifferentiated histology, and diffuse type subgroups of gastric cancer. In conclusion, MSI-H tumors were associated with a good prognosis in Stage II and III gastric cancer when patients were treated by surgery alone, and the benefits of MSI-H status were attenuated by chemotherapy. Microsatellite instability (MSI) is a form of genomic instability resulting from a deficient DNA mismatch repair system. MSI at high frequency (MSI-H) is a hallmark of hereditary nonpolyposis colorectal cancer (HNPCC) and has been observed in various sporadic tumors of the colon, stomach and endometrium. 1,2 The presence of MSI-H has been shown to be associated with a good prognosis. 3 5 Key words: gastric cancer, microsatellite instability, prognosis, chemotherapy S.Y.K. and Y.Y.C. contributed equally to this article and are joint first authors Grant sponsor: National R&D Program for Cancer Control, Ministry of Health and Welfare, Republic of Korea; Grant numbers: , ; Grant sponsor: National Research Foundation of Korea (NRF) grant from the Korean government (MSIP); Grant number: NRF DOI: /ijc History: Received 13 Aug 2014; Accepted 13 Jan 2015; Online 22 Jan 2015 Correspondence to: Ji Yeong An, M.D., Ph.D., Department of Surgery, Yonsei University College of Medicine, 50 Yonsei-Ro, Seodaemun-gu, Seoul, Republic of Korea, Tel.: [ ], Fax: 1[ ], ugids@naver.com In gastric cancer, various studies have reported that MSI- H tumors are associated with older age, female sex, intestinal type (Lauren classification), lower ptnm stage, and lower prevalence of lymph node metastasis. 1,6,7 MSI-H gastric cancer has also been shown to be more common in patients with multiple gastric cancers than in those with a solitary tumor. 8,9 Furthermore, a recent meta-analysis 10 demonstrated that MSI-H gastric cancer is associated with a better prognosis compared with MSI-low (MSI-L) and/or microsatellite stable (MSS) tumors. However, among the studies included in the meta-analysis, ours reported the most heterogeneous results, and a recent article from our institution 6 noted that the prognosis of MSI-H gastric cancer was similar to that of MSI-L/MSS gastric cancer. Before 2000, many studies failed to show a benefit from chemotherapy in gastric cancer, 11 and surgery alone remained the standard treatment for gastric cancer. 12 Populations from most of the studies included in the aforementioned meta-analysis 10 were treated before the era of chemotherapy for gastric cancer, and therefore, most of these studies may not have involved the use of chemotherapy. These observations led us to hypothesize that although the natural course of MSI-H in gastric cancer after surgery is associated with good prognosis, the resistance of MSI-H tumors to chemotherapy may have resulted in the lack of a Int. J. Cancer: 137, (2015) VC 2015 UICC

2 820 Prognosis of MSI-H gastric cancer What s new? The frequency of microsatellite instability (MSI) in gastric cancer is associated with prognosis, with high frequency (MSI-H) generally linked to good prognosis. Importantly, however, the ability of MSI frequency to predict patient outcome in gastric cancer may be influenced by chemotherapy. The authors of the present study tested that hypothesis by analyzing data on stage II and stage III gastric cancer patients who underwent gastrectomy. The prognostic benefits conferred by MSI-H status were found to be significantly reduced among patients who received chemotherapy following surgery. Certain clinicopathological characteristics also modified the association of MSI-H with gastric cancer prognosis. difference in prognosis between MSI-H and MSI-L/MSS gastric cancer noted in our previous report. 6 To verify this hypothesis, we expanded the study population and reanalyzed the prognosis of MSI-H according to the use of chemotherapy. Furthermore, the influence of known clinicopathologic features on the prognosis of MSI-H gastric cancer, compared with the prognosis of MSI-L/MSS gastric cancer, was evaluated through subgroup analyses. Material and Methods Enrolled patients Data from patients who underwent gastrectomy with curative intent for gastric cancer at Severance Hospital, Yonsei University Health System, between January 2005 and June 2010 were reviewed. This study was approved by the Institutional Review Board of our hospital ( ). MSI status was reported for 2,549 of these patients. Cancer staging was performed according to the 7th edition of the International Union Against Cancer classification (UICC). 13 Patients with prior gastric surgery, remnant stomach cancer or distant metastasis (such as peritoneal seeding or liver metastasis) were excluded. Because chemotherapy is not standardly performed for Stage I gastric cancer, patients diagnosed with Stage I cancer were also excluded. Patients with a family history of HNPCC were excluded, because the mechanism of mismatch repair deficiency is different in sporadic gastric cancer than in HNPCC. 14 An additional 26 patients with an unclear history of chemotherapy were also excluded. Thus, a total of 1,276 patients were included in the final analysis: in our previous report, 6 the number of patients with Stage II IV disease, according to the 6th edition of the UICC, 15 who underwent gastrectomy with curative intent was 796. Thus, applying the 7th edition of the UICC, an additional 480 patients with Stage II or III disease were enrolled in this study. Surgical treatment and postoperative management The standard strategy of lymph node dissection for clinically advanced gastric cancer was D2 lymph node dissection, in accordance with the guidelines of the Japanese Gastric Cancer Association. 16 Adjuvant chemotherapy was standard for patients with Stage II and III cancer after curative resection unless the patient refused chemotherapy. In some instances, the patients were enrolled in a randomized controlled trial as a control group and, thus, did not receive chemotherapy. Clinicopathologic variables Clinicopathologic features (e.g., gender, age, tumor size, histologic type, Lauren classification, pt stage, pn stage, ptnm stage, number of retrieved lymph nodes and receipt of chemotherapy) were retrospectively reviewed from a prospectively designed database. Determining the survival of patients was based on our hospital records (time of their last visit), the database of the Korea National Statistical Office and telephone surveys. Survival data were updated since December The median follow-up period was 47 months. Histologic tumor type was classified into three categories: differentiated type, which included papillary, well or moderately differentiated tubular adenocarcinoma; undifferentiated type, which included poorly differentiated tubular adenocarcinoma, signet ring cell carcinoma and mucinous carcinoma; and others, which included unknown histologic type and lymphoepithelioma-like carcinoma. Patients were followed-up according to our hospital s protocol, which included evaluations every 3 months until 2 years after surgery and then every 6 months thereafter until 5 years after surgery. Evaluations included a physical examination, laboratory test, imaging and endoscopy. Definition of microsatellite status From all matched normal and tumor tissues, DNA was extracted for polymerase chain reaction amplification. Handling of all surgical specimens, sample preparation, polymerase chain reaction amplification and fragment analysis were performed as described previously by An et al. 6 For DNA amplification, 20 ll reaction solutions was used, containing 2 ll of 103 buffer (Roche, Mannheim, Germany), mmol/l of MgCl 2, 0.3 lm of each primer pair, 250 lm of deoxynucleotide triphosphates and 2.5 units of DNA polymerase (Roche). Amplification involved an initial step at 94 C for 5 min; 30 one-minute cycles at 94, 55, and 72 C (step by step); and final extension at 72 C for 10 min. Subsequently, 0.7 ll of the amplified samples were mixed with 0.3 ll of GeneScan 500 size Standard and 9 ll of HiDi Formamide in the ABI Prism 3100 Genetic Analyzer for fragment separation. Electrophoresis was initiated when the temperature was 60 C, and 16 capillaries (36 cm in length and

3 Kim et al. 821 Table 1. Baseline characteristics of enrolled patients with Stage II and III gastric cancers Patient or tumor characteristics Gender Mean 6 standard deviation or number (%) Male 871 (68.3) Female 405 (31.7) Age (years) Tumor size (mm) Histology Differentiated 428 (33.5) Undifferentiated 841 (65.9) Others 1 7 (0.6) Lauren classification Intestinal 519 (40.7) Diffuse 627 (49.1) Mixed 92 (7.2) Others 38 (3.0) pt stage Mucosa 6 (0.5) Submucosa 40 (3.1) Proper muscle 139 (10.9) Subserosa 359 (28.1) Serosa 699 (54.8) Adjacent organ 33 (2.6) Number of lymph node metastases pn stage pn0 234 (18.3) pn1 (1 2) 236 (18.5) pn2 (3 6) 305 (23.9) pn3 (7) 501 (39.3) p Stage II 465 (36.4) III 811 (63.6) Chemotherapy No 361 (28.3) Yes 915 (71.7) MSI status MSI-H 105 (8.2) MSI-L/MSS 1171 (91.8) Retrieved lymph nodes Including unknown histologic type and lymphoepithelioma-like carcinoma. Abbreviations: MSI, microsatellite instability; MSI-H, MSI-high; MSI-L, MSI-low; MSS, microsatellite stable. 50 lm in diameter) were used for the array. POP-4 was applied as a separation medium, and fluorescence was converted into digital information and sent to a workstation (the ABI Prism 3100 Data Collection software). Two mononucleotide repeat markers (BAT25 and BAT26) and three dinucleotide repeat markers (D5S346, D2S123 and D17S250) were used for estimating the status of MSI, as recommended by the National Cancer Institute consensus group. 17 When two or more mutated markers were identified, the tumor was classified as MSI-H. When MSI was demonstrated at only one marker, the tumor was classified as MSI-L. When there was no MSI, tumors were classified as MSS. Statistical analysis Statistical analysis was performed using SPSS version 20.0 for Windows (SPSS, Chicago, IL) and R software version using the packhv and survival packages. Categorical variables were analyzed using the v 2 test or Fisher s exact test, and continuous variables were compared using the independent t- test. Data are presented as mean 6 standard deviation. For survival analysis, a Kaplan Meier curve, log-rank test and Cox proportional hazards model were applied. For subgroup analysis, the role of MSI status on survival was evaluated using hazard ratio (HR) and its 95% confidence interval with and without adjustment for age, sex, pt stage and pn stage. p-values < 0.05 were considered statistically significant. Results Clinical characteristics of 1,276 patients with Stage II or III gastric cancer The baseline characteristics of 1,276 patients with Stage II and III gastric cancer are shown in Table 1. Of these, 871 (68.3%) were men and 31.7% were women, and their mean age was 58.0 years. The mean tumor size was 21.4 mm, and the mean number of retrieved lymph nodes and metastatic lymph nodes were 45.8 and 7.9, respectively. There were 361 (36.4%) patients who did not receive chemotherapy and 915 (71.7%) patients who underwent chemotherapy. Clinical characteristics of MSI-H and MSI-L/MSS gastric cancers according to chemotherapy status Table 2 shows the characteristics of the patient with MSI-H or MSI-L/MSS tumors according to their chemotherapy status. There were 915 (72%) patients who received chemotherapy and 361 (28%) patients who did not. There were 47 (13.0%) patients with an MSI-H tumor in the no chemotherapy group and 58 (6.3%) patients with an MSI-H tumor in the chemotherapy group. In both chemotherapy and no chemotherapy groups, the histologic type and Lauren classification were different between the MSI-H and MSI-L/MSS gastric cancers. In the no chemotherapy group, the age of the patients with an MSI-H tumor was higher than those with an MSI-L/MSS tumor. Also, pn stage and pathologic substage were different between MSI-H and MSI-L/MSS tumors. Prognosis of MSI-H compared with MSI-L/MSS Among patients with Stage II and III disease who did not receive chemotherapy, those with MSI-H tumors had

4 822 Prognosis of MSI-H gastric cancer Table 2. Baseline characteristics of MSI-H and MSI-L/MSS gastric cancer according to chemotherapy status No chemotherapy Chemotherapy MSI-L/MSS MSI-H MSI-L/MSS MSI-H Characteristics (n 5 314) (n 5 47) p-values (n 5 857) (n 5 58) p-values Gender Male 223 (71.0) 28 (59.6) 586 (68.4) 34 (58.6) Female 91 (29.0) 19 (40.4) 271 (31.6) 24 (41.4) Age (years) Tumor size (mm) Histology < Differentiated 118 (37.6) 30 (63.8) 254 (29.6) 26 (44.8) Undifferentiated 196 (62.4) 16 (34.0) 597 (69.7) 32 (55.2) Others 1 0 (0) 1 (2.1) 6 (0.7) 0 (0) Lauren classification Intestinal 139 (44.3) 33 (70.2) 316 (36.9) 31 (53.4) Diffuse 145 (46.2) 9 (19.1) 454 (53.0) 19 (32.8) Mixed 22 (7.0) 3 (6.4) 64 (7.5) 3 (5.2) Others 8 (2.5) 2 (4.3) 23 (2.7) 5 (8.6) pt stage Mucosa 1 (0.3) 0 (0) 5 (0.6) 0 (0) Submucosa 18 (5.7) 3 (6.4) 18 (2.1) 1 (1.7) Proper muscle 43 (13.7) 8 (17.0) 83 (9.7) 5 (8.6) Subserosa 134 (42.7) 23 (48.9) 181 (21.1) 21 (36.2) Serosa 115 (36.6) 12 (25.5) 544 (63.5) 28 (48.3) Adjacent organ 3 (1.0) 1 (2.1) 26 (3.0) 3 (5.2) pn stage pn0 113 (36.0) 22 (46.8) (10.7) 7 (12.1) 0.96 pn1 (1 2) 60 (19.1) 14 (29.8) 151 (17.6) 11 (19.0) pn2 (3 6) 70 (22.3) 7 (14.9) 215 (25.1) 13 (22.4) pn3 (7) 71 (19.7) 4 (8.5) 399 (46.6) 27 (46.6) p Stage II 186 (59.2) 38 (80.9) 224 (26.1) 17 (29.3) III 128 (40.8) 9 (19.1) 633 (73.9) 41 (70.7) Retrieved lymph nodes Including unknown histologic type and lymphoepithelioma-like carcinoma. Data are mean 6 standard deviation or number (%). Abbreviations: MSI, microsatellite instability; MSI-H, MSI-high; MSI-L, MSI-low; MSS, microsatellite stable. significantly longer survival than those with MSI-L/MSS tumors (p ) (Fig. 1a). However, overall survival in patients who received chemotherapy did not differ significantly between MSI-H and MSI-L/MSS tumors (p ) (Fig. 1b). Kaplan Meier analysis showed that the prognosis of MSI-H tumors was better than that of MSI-L/MSS tumors in Stage II and III gastric cancer when chemotherapy was not received, but the difference was not statistically significant (Figs. 2ac). In the chemotherapy group, the curves for MSI-H tumors were similar to or worse than those of MSI-L/ MSS tumors (Figs. 2bd). Subgroup analyses were performed to further assess the effect of MSI status according to various clinicopathologic features with respect to whether chemotherapy was received. Each unadjusted and adjusted (by sex, age, pt stage and pn stage) HR of MSI-H tumors compared with MSI-L/MSS tumors was analyzed using a Cox proportional hazards model for the overall population and for various subgroups: Stage II, Stage III, male, female, with or without lymph node metastasis, differentiated or undifferentiated type and intestinal or diffuse type. In the no chemotherapy group, the prognosis of MSI-H tumors was significantly better not only for

5 Kim et al. 823 Figure 1. Kaplan Meier curves for the prognosis of MSI-H and MSI-L/MSS tumors in Stage II and III gastric cancer. (a) No chemotherapy group and (b) chemotherapy group. Figure 2. The prognosis of MSI-H compared with MSI-L/MSS gastric cancer in Stage II and III cancer. (a) Stage II no chemotherapy group, (b) Stage II chemotherapy group, (c) Stage III no chemotherapy group and (d) Stage III chemotherapy group.

6 824 Prognosis of MSI-H gastric cancer Table 3. Subgroup analyses of the prognosis of MSI-H versus MSI-L/MSS Stage II and III gastric cancers No chemotherapy Chemotherapy Case number 1 Unadjusted HR p-values Adjusted HR 2 p-values Case number 1 Unadjusted HR p-values Adjusted HR 2 p-values Overall 47/ ( ) ( ) / ( ) ( ) Stage II 38/ ( ) ( ) / ( ) ( ) Stage III 9/ ( ) ( ) / ( ) ( ) Male 28/ ( ) ( ) / ( ) ( ) Female 19/ ( ) ( ) / ( ) ( ) LN negative 22/ ( ) ( ) / ( ) ( ) LN positive 25/ ( ) ( ) / ( ) ( ) Differentiated 30/ ( ) ( ) / ( ) ( ) Undifferentiated 16/ ( ) ( ) / ( ) ( ) Intestinal 33/ ( ) ( ) / ( ) ( ) Diffuse 9/ ( ) ( ) / ( ) < ( ) < MSI-H/MSI-L or MSS. 2 Adjusted by age, sex, pt stage, and pn stage. Abbreviations: LN, lymph node; HR, hazard ratio. the overall population but also for the Stage III, female, with lymph node metastasis and undifferentiated type subgroups (adjusted HRs of 0.422, 0.197, 0.188, and 0.273, respectively; Table 3). In all subgroups of patients who did not receive chemotherapy, the unadjusted and adjusted HRs of MSI-H tumors, compared with MSI-L/MSS tumors, were less than 1. In the chemotherapy group, the prognosis of MSI-H tumors was worse than that of MSI-L/MSS in the Stage III, undifferentiated type and diffuse type gastric cancer subgroups (adjusted HRs of 1.610, and 2.946, respectively). In most subgroups of patients who received chemotherapy, the unadjusted and adjusted HRs of MSI-H tumors, compared with MSI-L/MSS tumors, were higher than 1. Discussion The results showed that MSI-H gastric cancer is associated with good prognosis when patients do not receive chemotherapy after gastrectomy with lymph node dissection. However, among patients who received chemotherapy, the prognosis of MSI-H gastric cancer was similar to that of MSI-L/MSS tumors. This finding implies that the natural oncologic course of Stage II and III MSI-H gastric cancer is better than that of MSI-L/MSS tumors after gastrectomy with lymph node dissection. Moreover, the results suggest that the presence of chemotherapy acted as a confounding factor in our previous report, influencing the role of MSI status as a prognostic marker for gastric cancer, 6 because the effect of chemotherapy differed according to MSI status. Thus, the study results support the conclusions of the recent meta-analysis 10 that reported that the prognosis of MSI-H gastric cancer is better than that of MSI-L/MSS tumors. The reason why MSI-H gastric cancer exhibits better prognosis has not yet been clearly defined, although it could be related to peritumoral lymphocyte infiltration. In colorectal cancer, peritumoral lymphocyte infiltration (Crohn s-like inflammation), a feature of MSI-H colorectal cancer, and inflammation are related to good prognosis, as increased cytotoxic T-lymphocyte counts induce greater neoplastic cell apoptosis In gastric cancer, gastric carcinoma with lymphoid stroma (lymphoepithelioma-like carcinoma) has been reported as being related to good prognosis, 21 and there are reports that MSI-H gastric cancer exhibits greater lymphoid stromal reactions than other types. 22,23 During subgroup analyses, the prognosis of MSI-H gastric cancer in Stage III, female, lymph node positive and undifferentiated type cases was statistically better than that of MSI-L/ MSS when chemotherapy was not administered. These findings are consistent with the results of the meta-regression in the previous meta-analysis 10 : A trend of better prognosis for MSI-H gastric cancer was noted among studies with a higher proportion of female, metastatic lymph node, and advanced stage (III and IV according to the UICC 6th edition 15 ) cases. Conversely, the prognosis of undifferentiated and diffuse type MSI-H gastric cancers was worse than that of MSI-L/MSS tumors when chemotherapy was administered (Table 3). This implies that chemotherapy for MSI-H gastric cancer does not improve, and may even worsen, prognosis in specific

7 Kim et al. 825 subgroups, such as diffuse type gastric cancer. However, there is a possibility that the positive results from the subgroup analyses reflect a Type I error because of the multiple analyses. Chemotherapy, applied before and/or after surgery, is now a routine procedure for advanced gastric cancer, and it seems to attenuate the benefits of MSI-H. Consequently, use of MSI status as a prognostic marker in clinical practice may be unwarranted. Notwithstanding, it could still be of use as a biomarker for achieving personalized therapy in gastric cancer. Our previous results 6 showed that MSI status may be of use as guide for the administration of chemotherapy, and the present results show that prognosis can differ according to the use of chemotherapy and the presence of certain clinicopathologic features of gastric cancer. Nevertheless, our results should be validated in other populations, and the biological rationale thereof should be investigated. Additionally, future studies about the prognosis of MSI-H gastric cancer should consider the effect of chemotherapy thereon. This study has limitations. The design was retrospective, and there were a variety of reasons why patients did not receive chemotherapy. Also, in patients who received chemotherapy, the chemotherapeutic regimens were heterogeneous: the regimens were oral 5-fluorouracil (FU) based only, 5FU 1 platinum, 5FU 1 anthracycline, 5FU 1 leukovorin, 5FU 1 taxane, etc. In addition, the number of patients in each subgroup analysis was limited and may have led to a Type II error. However, this study included the largest cohort to date, and future studies are needed to assess the validity of the present results. In conclusion, we confirmed that the prognosis of MSI-H gastric cancers is superior to that of MSI-L/MSS tumors for patients who undergo gastrectomy with lymph node dissection without chemotherapy. However, we also showed that the benefits of MSI-H are attenuated if chemotherapy is received. Thus, the role of MSI as a prognostic marker for gastric cancer may no longer be justified in the current era of administering chemotherapy for gastric cancer, and future studies should consider the effects of chemotherapy on the prognosis of MSI-H gastric cancers. References 1. Lee HS, Choi SI, Lee HK, et al. Distinct clinical features and outcomes of gastric cancers with microsatellite instability. Mod Pathol 2002; 15: Furlan D, Casati B, Cerutti R, et al. Genetic progression in sporadic endometrial and gastrointestinal cancers with high microsatellite instability. J Pathol 2002; 197: Merok MA, Ahlquist T, Royrvik EC, et al. Microsatellite instability has a positive prognostic impact on stage II colorectal cancer after complete resection: results from a large, consecutive Norwegian series. Ann Oncol 2013; 24: Popat S, Hubner R, Houlston RS. Systematic review of microsatellite instability and colorectal cancer prognosis. J Clin Oncol 2005; 23: Black D, Soslow RA, Levine DA, et al. Clinicopathologic significance of defective DNA mismatch repair in endometrial carcinoma. J Clin Oncol 2006; 24: An JY, Kim H, Cheong JH, et al. Microsatellite instability in sporadic gastric cancer: its prognostic role and guidance for 5-FU based chemotherapy after R0 resection. Int J Cancer 2012; 131: Beghelli S, de Manzoni G, Barbi S, et al. Microsatellite instability in gastric cancer is associated with better prognosis in only stage II cancers. Surgery 2006; 139: Miyoshi E, Haruma K, Hiyama T, et al. Microsatellite instability is a genetic marker for the development of multiple gastric cancers. Int J Cancer 2001; 95: Lee HS, Lee BL, Kim SH, et al. Microsatellite instability in synchronous gastric carcinomas. Int J Cancer 2001; 91: Choi YY, Bae JM, An JY, et al. Is microsatellite instability a prognostic marker in gastric cancer?: a systematic review with meta-analysis. J Surg Oncol 2014; 110: Nakajima T, Nashimoto A, Kitamura M, et al. Adjuvant mitomycin and fluorouracil followed by oral uracil plus tegafur in serosa-negative gastric cancer: a randomised trial. Gastric Cancer Surgical Study Group. Lancet 1999; 354: Cunningham D, Chua YJ. East meets west in the treatment of gastric cancer. N Engl J Med 2007; 357: Sobin LH, Compton CC. TNM seventh edition: what s new, what s changed: communication from the International Union Against Cancer and the American Joint Committee on Cancer. Cancer 2010; 116: Leite M, Corso G, Sousa S, et al. MSI phenotype and MMR alterations in familial and sporadic gastric cancer. Int J Cancer 2011; 128: Sobin LH, Wittekind CH, eds. TNM classification of malignant tumors, 6th edn. New York: Wiley- Liss, Japanese Gastric Cancer Association. Japanese gastric cancer treatment guidelines 2010 (ver. 3). Gastric Cancer 2011;14: Bol CR, Thibodeau SN, Hamilton SR, 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 Res 1998; 58: 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 1999; 154: Phillips SM, Banerjea A, Feakins R, et al. Tumour-infiltrating lymphocytes in colorectal cancer with microsatellite instability are activated and cytotoxic. Br J Surg 2004; 91: Prall F, Duhrkop T, Weirich V, et al. Prognostic role of CD81 tumor-infiltrating lymphocytes in stage III colorectal cancer with and without microsatellite instability. Hum Pathol 2004; 35: Watanabe H, Enjoji M, Imai T. Gastric carcinoma with lymphoid stroma. Its morphologic characteristics and prognostic correlations. Cancer 1976; 38: Grogg KL, Lohse CM, Pankratz VS, et al. Lymphocyte-rich gastric cancer: associations with Epstein-Barr virus, microsatellite instability, histology, and survival. Mod Pathol 2003; 16: Kim H, An JY, Noh SH, et al. High microsatellite instability predicts good prognosis in intestinaltype gastric cancers. J Gastroenterol Hepatol 2011; 26: Group G, Paoletti X, Oba K, et al. Benefit of adjuvant chemotherapy for resectable gastric cancer: a meta-analysis. JAMA 2010; 303: Bang YJ, Kim YW, Yang HK, et al. Adjuvant capecitabine and oxaliplatin for gastric cancer after D2 gastrectomy (CLASSIC): a phase 3 openlabel, randomised controlled trial. Lancet 2012; 379: Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med 2006; 355: Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy after surgery compared with surgery alone for adenocarcinoma of the stomach or gastroesophageal junction. N Engl J Med 2001; 345: Sakuramoto S, Sasako M, Yamaguchi T, et al. Adjuvant chemotherapy for gastric cancer with S- 1, an oral fluoropyrimidine. N Engl J Med 2007; 357: