Tel.: 82-2-2228-2111, Fax: 82-2-313-8289
Microsatellite instability in sporadic gastric cancer: its prognostic role and guidance for 5-FU based chemotherapy after R0 resection
Article first published online: 19 NOV 2011
Copyright © 2011 UICC
International Journal of Cancer
Volume 131, Issue 2, pages 505–511, 15 July 2012
How to Cite
An, J. Y., Kim, H., Cheong, J.-H., Hyung, W. J., Kim, H. and Noh, S. H. (2012), Microsatellite instability in sporadic gastric cancer: its prognostic role and guidance for 5-FU based chemotherapy after R0 resection. Int. J. Cancer, 131: 505–511. doi: 10.1002/ijc.26399
- Issue published online: 22 MAY 2012
- Article first published online: 19 NOV 2011
- Accepted manuscript online: 26 AUG 2011 07:45AM EST
- Manuscript Accepted: 4 AUG 2011
- Manuscript Received: 26 NOV 2010
- Korea Healthcare technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea. Grant Number: A084474
- Yonsei University College of Medicine. Grant Number: 6-2011-0015
- microsatellite instability;
- gastric cancer;
- prognostic marker;
- disease-free survival
This study investigated whether MSI status can be used as a prognostic biomarker and whether it is helpful for predicting which patients will benefit from 5-FU based adjuvant chemotherapy. Between 2005 and 2008, an MSI status examination was performed in 1,990 gastric cancer patients who had undergone curative gastrectomy for gastric adenocarcinoma. MSI was analyzed by PCR amplification with fluorescent dye-labeled primers of mononucleotide markers (BAT25 and BAT26) and dinucleotide markers (D5S346, D2S123 and D17S250) specific to the microsatellite loci. Patients with MSI-H tumors accounted for 8.5% (n = 170) of the total study population. They tended to be older and female and to have distal tumor location, lower tumor stage, intestinal type of Lauren classification and differentiated histological type. The disease-free survival curves showed no significant differences between MSS/MSI-L and MSI-H patients at each stage of I, II, III and IV. In gastric cancer patients with stage II and III, 5-FU-based adjuvant chemotherapy showed better disease-free survival in the MSS/MSI-L group, but showed no benefits in the MSI-H group. By multivariate analysis, patients with MSS/MSI-L tumors benefited from 5-FU-based adjuvant chemotherapy in terms of tumor disease-free survival. MSI status in gastric cancer is not itself a prognostic indicator. However, it appears to be a possible guidance for the use of 5-FU-based chemotherapy in stage II and III gastric cancers after R0 resection.
Microsatellite instability (MSI) is characterized by novel sized alleles detected in microsatellite sequences found in the DNA of carcinoma tissue that is not present in normal constitutional DNA. Although MSI has been reported along with several types of cancers, it is thought to be an important mechanism of hereditary nonpolyposis colorectal cancer (HNPCC), which occurs by germline mutations in one of the DNA mismatch repair (MMR) genes such as MLH1, MSH2, MSH6 and PMS2. The clinicopathological characteristics of MSI+ colorectal cancers have been widely studied: proximal location, reduced lymph node metastasis and better overall survival rate.1 Recently, several studies analyzed the relationship between MSI and benefits from adjuvant 5-Fluorouracil (5-FU)-based chemotherapy in colon cancer.2, 3 These studies reflect that MSI status has an important role in clinical practices for deciding tumor characteristics, predicting prognosis and planning adjuvant treatment.
In gastric cancer, the role of MSI remains to be proven, and it is difficult to judge its value in clinical application. Many studies have also attempted to elucidate the clinical significances of MSI in gastric cancer, suggesting that gastric cancers with a high frequency of MSI are associated with specific clinicopathological characteristics: intestinal type differentiation, reduced lymph node metastasis and better survival rates. However, these findings are not consistent and each study yields controversial results. Some studies reported that the MSI-H phenotype was associated with better survival in a subset of gastric cancer patients, but another showed no differences in survival according to MSI.4–8 Moreover, to date, the relation between MSI and adjuvant chemotherapy in association with tumor recurrence and survival has not been well investigated.
Therefore, in this large scale study, we conducted a retrospective cohort analysis to determine whether MSI status can be used as a prognostic marker, and whether it is helpful for predicting patients who will benefit from 5-FU-based adjuvant chemotherapy.
Material and Methods
We analyzed data from 1,990 patients who underwent curative gastrectomy for primary gastric cancer at Yonsei University College of Medicine between January 2005 and June 2008, in a prospectively designed database. All patients had a histologically confirmed adenocarcinoma of the stomach, and patients with a prior gastric surgery or neoadjuvant chemotherapy were excluded. In addition, patients with family history of HNPCC based on medical records were excluded because the mechanism of MMR deficiency is different between sporadic gastric cancer and HNPCC. MMR deficiency is caused by promoter methylation in sporadic gastric cancer, but it is caused by germline mutation in HNPCC.
All surgical specimens were delivered to the Department of Pathology. Normal and tumor DNA was extracted from the uncovered hematoxylin and eosin-stained 6 μm-thick sections from formalin-fixed, paraffin-embedded (FFPE) tissues. To ensure sufficient tumor percentage and to enrich the tumor cells, the uncovered hematoxylin and eosin-stained slides were evaluated by pathologists under light microscopy. Cases with >10% of tumor percentage were included in this study. Then our pathologist reviewed the tumor percentage and increased the tumor cell fraction >30% in all cases.
To extract normal DNA, the furthest proximal or distal tumor-free gastric FFPE tissues were used. This study was reviewed and approved by the Institutional Review Board of Severance Hospital, Yonsei University College of Medicine, and written informed consent was obtained from all patients before operation for tissue acquisition.
Surgical treatment and adjuvant chemotherapy
We performed D1 + β or D2 lymph node dissection on all patients according to the guidelines of the Japanese Gastric Cancer Association.9 D1 + β lymph node dissection was defined as a nodal dissection that included the perigastric area along the left gastric artery (No. 7), common hepatic artery (No. 8a) and celiac axis (No. 9). For distal gastric cancer, D2 lymph node dissection was defined as nodal dissection along the proximal splenic artery (No. 11p), hepatoduodenal ligament (No. 12a) and superior mesenteric vein (No. 14v). For upper gastric cancer, D2 lymph node dissection included the lymph nodes around the splenic hilum (No. 10) and the splenic artery (No. 11p, 11d). Adjuvant chemotherapy based on 5-FU (5-FU, 5-FU+Cisplatin, 5-FU+Adriamycin, 5-FU+Leucovorin) was performed on patients with stages II, III and IV after curative resection unless the patient disagreed or rejected the procedure.
Tumor staging and histological classification
Tumors were staged according to the sixth edition of the International Union against Cancer classification.10 The maximum diameter was recorded as the tumor size. Tumors was classified into two groups by histology: namely, the differentiated type, which included papillary, well or moderately differentiated adenocarcinoma and the undifferentiated type, which included poorly or undifferentiated adenocarcinoma, signet ring cell carcinoma and mucinous carcinoma.
Evaluation of clinicopathological variables and tumor recurrence
We retrospectively studied clinicopathological features (e.g., gender, age, tumor size, tumor location, number of tumors, resection extent, histological type, numbers of retrieved and metastatic lymph nodes and tumor stage), tumor recurrence and survival based on the information in the prospectively designed database. Our standard clinical practice includes evaluating patients every three months until 2 years after the operation and then every 6 months until 5 years after the operation, with physical examinations, laboratory tests, imaging and endoscopy. Recurrence was confirmed by tissue biopsy when possible.
Microsatellite markers and polymerase chain reaction
DNA was extracted from all matched normal and tumor tissues for PCR amplification. To determine tumor MSI status, we used microsatellite markers consisting of two mononucleotide repeat markers (BAT25 and BAT26) and three dinucleotide repeat markers (D2S123, D5S346 and D17S250), as recommended by the National Cancer Institute (NCI) consensus group. Fifty nanograms of DNA were amplified in 20 μl of reaction solution containing 2 μl of 10× buffer (Roche, Mannheim, Germany), 1.7–2.5 mmol/L of MgCl2, 0.3 μM of each primer pairs, 250 μM of deoxynucleotide triphosphate and 2.5 units of DNA polymerase (Roche). Amplification was performed with an initial denaturation step at 94°C for 5 min; followed by 30 cycles of 1 min at 94°C, 1 min at 55°C and 1 min at 72°C and a 10-min final extension at 72°C.
Sample preparation and fragment analysis
For fragment separation, 0.7 μL of the amplified samples were combined with 0.3 μL of the GeneScan 500 Size Standard and 9 μL of HiDi Formamide on the ABI Prism 3100 Genetic Analyzer. Electrophoresis was initiated when the Genetic Analyzer oven temperature was equal to 60°C. Application of 15 kV across all 16 capillaries of the array allowed simultaneous electrokinetic injection of DNA fragments for a duration of 5 s. The capillaries were 36 cm in length and measured 50 μm in diameter. POP-4 was employed as the separation medium. Raw data was detected on a real time basis as the fluorescence is converted into digital information and sent to the workstation to be processed by the ABI Prism 3100 Data Collection software. The data is displayed as an electropherogram that plots base-pair size of the DNAs displayed on the x-axis against peak signal intensity (peak height) in relative fluorescent units (rfu) on the y-axis.
Scoring of MSI
When interpreting the results of the MSI analysis, shifted signal intensities occupying more than half of the tumor over non-shifted signal intensities was considered significant. If cases show questionable signal intensity compared to noise level and tumor percentage, MSI analysis was repeated using tumor DNA from other tumor areas. When there was a shift in microsatellites at two or more markers, it was classified as microsatellite instability-high (MSI-H). When microsatellite instability was demonstrated at only one marker, it was classified as microsatellite instability-low (MSI-L). Tumors were classified as microsatellite stable (MSS) when there was no microsatellite instability. Of the five markers, BAT 26 had the highest frequency of positivity in tumours with high microsatellite instability because it is associated with the highest incidence of high microsatellite instability and the lowest incidence of low microsatellite instability (Table 1).
Statistical analysis was carried out using SPSS® version 18.0 for Windows® (SPSS, Chicago, IL). Categorical variables were compared using the chi-square or Fisher's exact test, and continuous data were compared by Mann-Whitney test. Data are presented as mean ± standard deviation (SD). Disease-free survival curves were calculated in months based on the length of time between primary surgical treatment and final follow-up or recurrence using the Kaplan–Meier method. The log-rank test was used for univariate association between the MSI-H and MSS/MSI-L groups in terms of various clinicopathologic variables. Examination for independent prognostic factors was carried out using the Cox proportional hazards regression model. p values of <0.05 were considered as significant.
Clinical characteristics and microsatellite instability status in 1,990 gastric cancer patients
There were 1,722 patients with MSS, 98 patients with MSI-L and 170 patients with MSI-H. Because the clinicopathological features between patients with MSS and MSI-L were more similar than those between patients with MSI-L and MSI-H (data not shown), we divided patients into two groups; MSI-H tumors and MSI-L or MSS tumors. The clinicopathological features of patients with MSI-H tumors were compared with those with MSI-L or MSS tumors, as shown in Table 2. Patients with MSI-H tumors tended to be older (p < 0.001) and female (p < 0.001), and to have distal tumor location (p < 0.001), intestinal type of Lauren classification (p < 0.001), differentiated histological type (p < 0.001) and lower tumor stage (p = 0.001).
Relation between MSI status and disease-free survival
During 30.2 months of median follow-up, 298 patients showed tumor recurrence. The disease-free survival curves showed no significant differences between MSS/MSI-L and MSI-H patients at stages I, II, III and IV (Fig. 1).
Relation between MSI status and benefit of 5-FU based chemotherapy
The effect of 5-FU-based adjuvant chemotherapy according to MSI status was analyzed in patients with stages II, III and IV gastric cancers. In stages II and III gastric cancer patients, 5-FU-based adjuvant chemotherapy improved disease-free survival of the MSS/MSI-L group (p = 0.022 and 0.020 in stages II and III), but showed no benefit for the MSI-H group (p = 0.843 and 0.108 in stages II and III) (Fig. 2). In stage IV gastric cancer, 5-FU-based adjuvant chemotherapy did not improve disease-free survival in either the MSS/MSI-L or the MSI-H group. However, because the number of patients with MSI-H stage IV gastric cancer was only seven, it was difficult to compare survival curves in the two groups.
To assess how 5-FU based chemotherapy and MSI affected survival, we performed multivariate analysis using the Cox proportion hazard model. When adjusted for sex, histological differentiation, Lauren type and tumor stage, the interaction between MSI and adjuvant chemotherapy and the tumor stage remained independent prognostic factors for disease-free survival (Table 3). Although tumor stage had the strongest statistical power for predicting tumor recurrence, patients with MSS/MSI-L tumors benefited from 5-FU based adjuvant chemotherapy in disease-free survival (HR = 0.664, p = 0.008).
Although the clinicopathological characteristics of MSI-H tumors are distinct in colon cancer, they have remained unclear and inconsistent in gastric cancer. Because there is no consensus panel to determine MSI in gastric cancer, MSI phenotype evaluation has been performed using different types of microsatellite sequences. In this study, we used five markers that are accepted as a panel for MSI determination in colon cancer, and the incidence of MSI-H in gastric cancer was 8.5%, which is similar to the results of previous eastern studies (8.2–9.5%)4, 7, 8 but lower than those of western studies (16–25.2%).5, 6, 11 This difference may be related to the difference in the molecular background of gastric carcinogenesis between eastern and western peoples and may be attributed to different numbers and types of markers used in evaluating the MSI phenotype in gastric cancer. After all, a consensus for MSI markers in gastric cancer is needed before clinical use of MSI status can increase.
Until now, the most important prognostic factor in gastric cancer has been tumor stage. Our study showed that MSI-H phenotype itself is not a prognostic marker. At each stage, I, II, III and IV, MSI-H tumor did not decrease tumor recurrence. The role of MSI as a prognostic marker in gastric cancer has been controversial in previous reports, which enrolled about 240–510 patients. In some reports, MSI status did not influence survival and tumor recurrence at all, which is similar to our results.7, 8 The authors insisted that MSI-H did not demonstrate any significant benefits in terms of survival and further studies should be conducted with more cases and long-term follow-up. However, in other studies, MSI-H phenotype showed better survival in subgroups of gastric cancer such as advanced cancer, stage II or showing instability at all five markers.5, 6 They suggested that MSI status can be a marker for selecting a particular subset of tumors with less aggressive biological behavior and a more favorable prognosis. From our large-scale study with 1,990 cases of R0 resection, it appears to be premature to apply MSI status to clinical practice for predicting prognosis and selecting subgroups for favorable survival.
The efficacy of 5-FU in patients with MSI-H colon cancer has been investigated in several studies. Defective DNA MMR might be one mechanism for tumor resistance to 5-FU and a reason that MMR deficient colon cancer cells are not readily killed by 5-FU.12–15 However, in vitro studies and clinical data do not accord well together. Some authors found that fluorouracil-based adjuvant chemotherapy benefited patients with stage II or stage III colon cancer with MSS and MSI-L phenotype but not those with tumors exhibiting MSI-H.2, 3 They suggested that adjuvant 5-FU-based chemotherapy may not be useful in stage II and III MMR deficient colorectal cancer. In contrast, other authors have published evidence that patients with MSI+ tumors gained a survival advantage from 5-FU chemotherapy.16, 17 Moreover, some authors were concerned that it appears to be irresponsible and premature to regard MSI status as a marker that guides the use of 5-FU treatment in colon cancer.18 They claimed that exclusion of MSI-H colon cancer patients from receiving 5-FU chemotherapy cannot yet be justified, considering the available data.
5-FU is also the mainstay adjuvant chemotherapy in gastric cancer. There have been few studies elucidating whether the efficacy of 5-FU-based chemotherapy can be affected by MMR status. In one study, the usefulness of MSI as a prognostic marker and a predictor of response to adjuvant chemotherapy with 5-FU in 240 gastric cancer patients was evaluated.7 They showed that there was no survival difference between the MSI-H and MSS/MSI-L groups, and that survival after 5-FU-based chemotherapy did not correlate with MSI status. However, their study was not sufficient to perform multivariate analysis corrected by tumor stage. In our study, 5-FU-based adjuvant chemotherapy did not gain disease-free survival benefits in stage II and III gastric cancer with MSI-H. Patients with MSS/MSI-L gastric cancers gained potential disease-free survival benefits from 5-FU based adjuvant chemotherapy. Upon multivariate analysis, the combination of MMR status and 5-FU chemotherapy was a significant prognostic factor for disease-free survival. Patients who had MSS/MSI-L gastric cancer and were treated with 5-FU-based chemotherapy had a higher probability of disease-free survival.
However, the heterogeneity of combined chemotherapeutic agents with 5-FU can be considered to be a confounding factor in evaluating the efficacy of 5-FU in relation to MSI status. At our institute, cisplatin, adriamycin or leucovorin was used in combination with 5-FU. Although there was no difference in tumor recurrence according to the kinds of combined chemotherapeutic agents (data not shown), this is a weak point of retrospective analyses. In addition, because of the small number of MSI-H gastric cancer patients, survival analysis in MSI-H tumors appears to be insufficient to conclude the correlation between 5-FU based chemotherapy and tumor recurrence. Although multivariate analysis confirmed the significance of 5-FU-based chemotherapy in MSS/MSI-L gastric cancer, these two points are still critical limitations of this study. Nevertheless, since most of 5-FU-based adjuvant chemotherapy did not show a significant improvement of survival in gastric cancer patients, this study is an important development towards a new aspect of adjuvant chemotherapy. Furthermore, we will continue to find reliable genetic or biological markers that may identify gastric cancer patients who would respond to chemotherapy.
In this study, we included gastric cancers with at least 10% tumor percentage for MSI analysis. We did not mean that the only 10% tumor was enough for scoring MSI and the tumor percentage was usually more than 20–40%. The process of tumor cell fraction done in all cases by experienced pathologists seems to be important for increasing tumor cell fraction more than 30%. In addition, we do not have the data of correlation between the MSI status and tumor percentage. However, we suppose that there would be no correlation between the two parameters.
MSI is an indicator of genetic instability, and accumulation of gene mutations due to MSI may play an important role in the mechanism of malignant transformation. The progressive accumulation of MSI in areas of intestinal metaplasia and dysplasia may be an important molecular event in multi-step gastric carcinogenesis and may contribute to gastric cancer development.19, 20 This is supported by several studies that showed that MMR deficit may provide an important pathway for intestinal type gastric cancer development.21, 22 Although we did not evaluate known cancer–related target genes such as TGFβRII, BAX, hMSH3 and MSH6,23 a significant association between MSI-H status and some clinical features (older age, distal location, intestinal type and well/moderately differentiated histology) suggests the necessity of assessing MSI status in potential preneoplastic lesions of gastric cancer and analyzing candidate target genes in MSI-H gastric cancers.
In conclusion, MSI status cannot itself be used as a prognostic marker yet, but it can be used to predict the efficacy of 5-FU-based chemotherapy especially in stage II and III gastric cancer. This study is important to the development of a new aspect of adjuvant chemotherapy and further investigation is needed to elucidate the role of MSI in predicting chemotherapeutic responses. A prospective study for evaluating the correlation of the efficacy of 5-FU-based chemotherapy and MSI status in gastric cancer would help clarify the role of MSI in gastric cancer.
- 10TNM classification of malignant tumors, 6th edn. New York: Wiley-Liss, 2002., .