APMIS 123: 215–222
© 2014 APMIS. Published by John Wiley & Sons Ltd. DOI 10.1111/apm.12345
Correlation between microsatellite instability-high phenotype and occult lymph node metastasis in gastric carcinoma JIWOON CHOI,1 SOO KYUNG NAM,2 DO JOONG PARK,3 HWAL WOONG KIM,4 HYUNG-HO KIM,3 WOO HO KIM1 and HYE SEUNG LEE2 1 Department of Pathology, Seoul National University College of Medicine, Seoul, Republic of Korea; Departments of 2Pathology, and 3Surgery, Seoul National University Bundang Hospital, Seongnam, Republic of Korea; and 4Department of Pathology, GoodMoonhwa Hospital, Busan, Republic of Korea
Choi J, Nam SK, Park DJ, Kim HW, Kim HH, Kim WH, Lee HS. Correlation between microsatellite instability-high phenotype and occult lymph node metastasis in gastric carcinoma. APMIS 2015; 123: 215–222. The aim of this study is to investigate the association of microsatellite instability (MSI) status with nodal status in gastric carcinoma (GC). MSI status was investigated in 623 consecutively resected GCs. To detect occult lymph node (LN) metastasis, immunohistochemistry (IHC) using antibodies against pan-cytokeratin was performed in 391 node-negative cases by initial histologic examination. MSI-high (MSI-H) phenotype was found in 68 GC cases (10.9%) and was significantly associated with increased patient age, antral location, intestinal type, absence of venous/perineural invasion, and expanding growth type (p < 0.05). When the nodal status was evaluated, the number of metastatic LNs of MSI-H tumors tended to be lower than that of microsatellite stable/MSI-low (MSS/L) tumors (1.49 3.15 vs 4.37 9.81; p = 0.052), but the MSI-H phenotype was associated with the presence of lymphatic invasion (p = 0.036) and IHCpositive occult LN metastasis (p = 0.007). By multivariate analysis, MSI-H phenotype was significantly associated with IHC-positive occult LN metastasis (Odds ratio, 2.654; p = 0.044). MSI status and occult LN metastasis were not prognostic factors by survival analysis. Our findings suggest that the relationship between MSI status and regional LN metastasis may have some clinical and biologic implications to be elucidated. Key words: Gastric cancer; microsatellite instability; lymph node metastasis; occult lymph node metastasis. Hye Seung Lee, Department of Pathology, Seoul National University Bundang Hospital, 173-82 Gumi-ro, Bundang-gu, Seongnam-si, Gyeonggi-do 463-707, Republic of Korea. e-mail: [email protected]
Gastric carcinoma (GC) is the fourth most common cancer and the second leading cause of cancer-related death in the world (1). The pathogenesis of GC is complex and multifactorial involving various genetic, epigenetic, and environmental factors (2, 3). Among these factors, microsatellite instability (MSI) is a form of genomic instability characterized by the accelerated accumulation of single nucleotide mutations and alterations in the length of simple, repetitive microsatellite sequences throughout the genome (4). The presence of MSI is determined by comparing the length of microsatellite repeats in tumor cells and normal cells, and three categories of MSI were defined:
MSI-high (MSI-H), MSI-low (MSI-L), and microsatellite stable (MSS) (4). MSI is caused by genetic or epigenetic inactivation of DNA mismatch repair genes (MLH1, MSH2, MSH6, and PMS2). In GCs, epigenetic silencing (promoter hypermethylation) of the MLH1 gene is the most common cause of MSI (5). Also, MSI GC with early onset may be associated with Lynch syndrome (HNPCC or hereditary non-polyposis colorectal cancer) which is an autosomal dominant disorder caused by a germline defect in one of the DNA mismatch repair genes. Currently, MSI-H GCs are known to have distinctive clinicopathologic features compared to MSS/MSI-L (MSS/L) GCs. The consistently reported features of MSI-H GCs are old age, antral
Received 7 May 2014. Accepted 17 October 2014
215
CHOI et al.
location and intestinal type by Lauren classification (6–8). However, the regional lymph node (LN) status and patient prognosis according to MSI status remains unclear, whereas MSI-H colorectal cancers are less prone to lymph node (LN) metastasis and have more favorable prognosis than MSS tumors (9, 10). Meanwhile, upon routine pathologic examination of resected GC specimens, minute metastatic tumors, which are present as a single cell or a cluster of cells, are occasionally found within the dissected LNs. These occult LN metastases often go undetected by routine examination using hematoxylin-eosin staining, and this lack of detection may lead to recurrence of tumors in node-negative GC patients who had undergone curative surgery (11, 12). However, to date, the prognostic impact of occult LN metastases on patient survival in GCs remains controversial (13). Moreover, it has not been studied whether the occult LN metastasis status of GCs differs according to the MSI status. In this study, we investigated the prevalence of the MSI-H phenotype in 623 consecutively resected GC cases and its correlation with various clinicopathologic features, particularly relevant to regional LN metastasis. We also analyzed the clinicopathologic variables associated with occult LN metastasis. In addition, we evaluated the prognostic significance of MSI status and occult LN metastasis. MATERIALS AND METHODS Patients A total of 646 consecutive patients with gastric adenocarcinoma who underwent curative resection and pathologic examination at the Seoul National University Bundang Hospital between August 2010 and September 2011 were enrolled. Among them, specimens were available in 623 GC cases (96.4%), thus a total of 623 cases were included in this study. The demographic and clinicopathologic parameters were evaluated by reviewing medical charts and pathologic records. Tumor staging was based on the pathologic TNM classification according to the 7th edition of the AJCC Cancer Staging Manual (14). None of the patients had fulfilled the revised Bethesda criteria for Lynch syndrome (15). Clinical outcomes were followed up from the date of surgery until October 2013 [median (range), 28 (2–39) months]. This study was approved by the institutional review board of Seoul National University Bundang Hospital.
Microsatellite analysis In all 623 cases, tumor and corresponding non-neoplastic gastric mucosa tissue were obtained from formalin-fixed and paraffin-embedded tissue samples by manual microdissection using a scalpel blade. The size of the microdis-
216
sected tumor was 1.0 9 0.5 9 0.01 cm, and the percentage of tumor cells exceeded a minimum of 60% of the total number of cells. DNA was extracted from the microdissected tissues with InstaGene Matrix (Bio-Rad Laboratories, Hercules, CA, USA). The extracted DNA was amplified by polymerase chain reaction (PCR) with fluorescent dye-labeled primers targeting five microsatellite markers as recommended by Bethesda guideline on MSI (BAT25, BAT26, D2S123, D5S346, and D17S250). Fragment analysis was performed with ABI-3130xl genetic analyzer (Applied Biosystems, Foster City, CA, USA) and GeneMapperâ software (Applied Biosystems). According to the Bethesda guideline on MSI, tumors were classified as MSI-H when at least two of the five markers displayed novel bands, MSI-L when additional alleles were observed with one of the five markers, and MSS when all microsatellite markers examined displayed identical patterns in both tumor and normal tissues (4).
Immunohistochemistry Immunohistochemical staining of pan-cytokeratin (panCK) was performed for all dissected regional LNs from 391 GC patients in whom regional LN metastasis could not be identified by initial pathologic examination. The total number of examined LNs was 17 548. A single serial section (4 lm) was cut from each paraffin block, deparaffinized, and dehydrated. Immunostaining of the fixed tissues was performed with antibodies against pan-CK (mAb anti-human cytokeratin clone MNF116, DAKO, Glostrup, Denmark) at 1:100 dilution using a streptavidin peroxidase procedure, following an antigen retrieval process using microwave heating. According to the 7th edition of the AJCC Cancer Staging Manual, LN metastases detected by immunohistochemistry can be subdivided into three groups (14): Isolated tumor cells (ITCs), when the largest diameter was no larger than 0.2 mm, micrometastases when larger than 0.2 mm but their largest diameter was no larger than 2.0 mm, and macrometastases when their largest diameter was larger than 2.0 mm. The tumor deposits in regional LNs with histologic evidence of malignant activity, such as gland formation or stromal reaction were defined as micrometastases even if their largest dimension was smaller than 0.2 mm (Fig. 1). ITCs were subdivided into four groups (single cell, multiple individual cells, single cluster, and multiple clusters), as previously reported (16). In pathologic staging, ITCs were classified as pN0(i+), and micrometastases were classified as pN1mi. Micrometastases were found in seven of 391 cases (1.8%), which were restaged into pN1.
Statistical analysis In order to determine the correlation between MSI status and clinicopathologic parameters, the categorical variables were compared using the v2 test or Fisher’s exact test and the continuous data were compared by the Mann–Whitney U test. Logistic regression analysis was performed to determine the predictors of occult LN metastasis. Kaplan–Meier method was used to estimate the survival and the differences between survival curves were determined using the log-rank test. The results were considered to be statistically significant when p values
© 2014 APMIS. Published by John Wiley & Sons Ltd
NODAL STATUS IN MSI-H GASTRIC CARCINOMA
A
B
C
D
Fig. 1. Representative photomicrographs of MSI-high (MSI-H) and Microsatellite stable (MSS) gastric cancers with occult lymph node metastasis. (A) MSI-H gastric cancer showing intestinal histologic type and lymphatic invasion (hematoxylineosin stain, 9200). (B) Occult lymph node metastasis in MSI-H gastric cancer (pan-cytokeratin immunohistochemistry, 9400). (C) MSS gastric cancer showing diffuse histologic type without lymphatic invasion (hematoxylin-eosin stain, 9200). (D) Occult lymph node metastasis in MSS gastric cancer (pan-cytokeratin immunohistochemistry, 9400). were less than 0.05. All statistical analysis was conducted using SPSS 19.0 statistical software program (SPSS, Chicago, IL, USA).
RESULTS Clinicopathologic characteristics of MSI-H gastric cancers
Of the 623 consecutive GC cases, 513 (82.3%), 42 (6.7%), and 68 cases (10.9%) were MSS, MSI-L, and MSI-H, respectively. Table 1 summarizes the clinicopathologic characteristics of GCs according to the MSI status. The MSI-H phenotype was significantly associated with old age (p < 0.001), distal location (p < 0.001), intestinal type based on Lauren’s classification (p < 0.001), less infiltrative gross type in both early GC and advanced GC (p = 0.026 and 0.024, respectively), expanding growth based on Ming’s classification (p < 0.001), the absence of venous invasion (p = 0.015), the absence of perineural invasion (p = 0.017), and the absence of stromal desmoplasia (p = 0.003). MSI-H GCs were more likely to be tubular or papillary adenocarcinomas (p < 0.001) and well-to-moderately differentiated type tumors (p < 0.001). On the other hand, there was no significant association between the MSI-H phenotype and female gender (p = 0.434), mucinous histology (p = 0.380), stromal lymphoid follicles (p = 0.112), invasion © 2014 APMIS. Published by John Wiley & Sons Ltd
depth (p = 0.657), distant metastasis (p = 0.563), or pTNM stage (p = 0.166). The percentage of stage III–IV cases was lower in MSI-H GCs than those of MSS/L GCs although the difference was not statistically significant (MSS/L, 26.7%; MSI-H, 16.2%; p = 0.061). Survival analysis suggested that MSI-H GCs showed longer disease-specific survival (DSS) and progression-free survival (PFS) than those of MSS/ L GCs, but the differences were not statistically significant (p = 0.159 and 0.286, respectively) (Fig. 2). MSI-H phenotype and regional LN metastasis
There was no statistically significant difference in the presence of regional LN metastasis between MSI-H and MSS/L GCs (p = 0.115). However, when we analyzed the subgroups of tumor invasion, the MSI-H phenotype was significantly associated with the absence of LN metastasis in pT3/T4 GCs (MSS/L, 19/178; MSI-H, 8/20; p = 0.002), but not in pT1/T2 GCs (data not shown). In addition, MSI-H GCs tended to have fewer metastatic LNs than MSS/L GCs, with borderline statistical significance (p = 0.052) (Table 2). When we analyzed the subgroups of tumor invasion, MSI-H GCs had fewer metastatic LNs compared to MSS/L GCs in pT3/T4 cases (p = 0.002). In contrast, pT1/T2 GCs did not show any difference in the number of metastatic LNs according to the MSI status (p = 0.901).
217
CHOI et al.
Table 1. Clinicopathologic characteristics of 623 consecutive Characteristics Total (%) (n = 623) Age (mean SD) (y) 60.5 12.9 Sex Male 425 (68.2) Female 198 (31.8) Tumor size 3.9 2.8 (mean SD) (cm) Location Lower 1/3 370 (59.4) Middle 1/3 137 (22.0) Upper 1/3 112 (18.0) Remnant 4 (0.6) Gross type (EGC) 0-I (protruded) 14 (4.0) 0-IIa (elevated) 60 (17.1) 0-IIb (flat) 45 (12.8) 0-IIc (depressed) 226 (64.4) 0-III (excavated) 6 (1.7) Gross type (AGC) Borrmann 1 8 (3.0) Borrmann 2 34 (12.6) Borrmann 3 201 (74.4) Borrmann 4 27 (10.0) WHO classification Tubular/papillary 448 (71.9) Mucinous 12 (1.9) Poorly cohesive 146 (23.4) Mixed 11 (1.8) Other 6 (1.0) Histologic grade Differentiated 288 (46.2) Undifferentiated 335 (53.8) Lauren’s classification Intestinal 303 (48.6) Diffuse 294 (47.2) Mixed/indeterminate 26 (4.2) Lymphatic invasion Absent 377 (60.5) Present 246 (39.5) Venous invasion Absent 535 (85.9) Present 88 (14.1) Perineural invasion Absent 431 (69.2) Present 192 (30.8) Ming’s classification Expanding 213 (34.2) Infiltrative 410 (65.8) Desmoplasia Absent 513 (82.6) Present 108 (17.4) Stromal lymphoid follicles Absent 581 (93.4) Present 41 (6.6) Depth of invasion pT1/T2 425 (68.2%) pT3/T4 198 (31.8%) Lymph node metastasis Absent 384 (61.6) Present 239 (38.4)
218
gastric cancers according to the MSI status MSS/L (%) MSI-H (%) (n = 555) (n = 68) 59.5 12.9 68.6 9.6 381 (68.6) 174 (31.4) 3.9 2.8
44 (64.7) 24 (35.3) 3.8 2.7
p value
© 2014 APMIS. Published by John Wiley & Sons Ltd. DOI 10.1111/apm.12345
Correlation between microsatellite instability-high phenotype and occult lymph node metastasis in gastric carcinoma JIWOON CHOI,1 SOO KYUNG NAM,2 DO JOONG PARK,3 HWAL WOONG KIM,4 HYUNG-HO KIM,3 WOO HO KIM1 and HYE SEUNG LEE2 1 Department of Pathology, Seoul National University College of Medicine, Seoul, Republic of Korea; Departments of 2Pathology, and 3Surgery, Seoul National University Bundang Hospital, Seongnam, Republic of Korea; and 4Department of Pathology, GoodMoonhwa Hospital, Busan, Republic of Korea
Choi J, Nam SK, Park DJ, Kim HW, Kim HH, Kim WH, Lee HS. Correlation between microsatellite instability-high phenotype and occult lymph node metastasis in gastric carcinoma. APMIS 2015; 123: 215–222. The aim of this study is to investigate the association of microsatellite instability (MSI) status with nodal status in gastric carcinoma (GC). MSI status was investigated in 623 consecutively resected GCs. To detect occult lymph node (LN) metastasis, immunohistochemistry (IHC) using antibodies against pan-cytokeratin was performed in 391 node-negative cases by initial histologic examination. MSI-high (MSI-H) phenotype was found in 68 GC cases (10.9%) and was significantly associated with increased patient age, antral location, intestinal type, absence of venous/perineural invasion, and expanding growth type (p < 0.05). When the nodal status was evaluated, the number of metastatic LNs of MSI-H tumors tended to be lower than that of microsatellite stable/MSI-low (MSS/L) tumors (1.49 3.15 vs 4.37 9.81; p = 0.052), but the MSI-H phenotype was associated with the presence of lymphatic invasion (p = 0.036) and IHCpositive occult LN metastasis (p = 0.007). By multivariate analysis, MSI-H phenotype was significantly associated with IHC-positive occult LN metastasis (Odds ratio, 2.654; p = 0.044). MSI status and occult LN metastasis were not prognostic factors by survival analysis. Our findings suggest that the relationship between MSI status and regional LN metastasis may have some clinical and biologic implications to be elucidated. Key words: Gastric cancer; microsatellite instability; lymph node metastasis; occult lymph node metastasis. Hye Seung Lee, Department of Pathology, Seoul National University Bundang Hospital, 173-82 Gumi-ro, Bundang-gu, Seongnam-si, Gyeonggi-do 463-707, Republic of Korea. e-mail: [email protected]
Gastric carcinoma (GC) is the fourth most common cancer and the second leading cause of cancer-related death in the world (1). The pathogenesis of GC is complex and multifactorial involving various genetic, epigenetic, and environmental factors (2, 3). Among these factors, microsatellite instability (MSI) is a form of genomic instability characterized by the accelerated accumulation of single nucleotide mutations and alterations in the length of simple, repetitive microsatellite sequences throughout the genome (4). The presence of MSI is determined by comparing the length of microsatellite repeats in tumor cells and normal cells, and three categories of MSI were defined:
MSI-high (MSI-H), MSI-low (MSI-L), and microsatellite stable (MSS) (4). MSI is caused by genetic or epigenetic inactivation of DNA mismatch repair genes (MLH1, MSH2, MSH6, and PMS2). In GCs, epigenetic silencing (promoter hypermethylation) of the MLH1 gene is the most common cause of MSI (5). Also, MSI GC with early onset may be associated with Lynch syndrome (HNPCC or hereditary non-polyposis colorectal cancer) which is an autosomal dominant disorder caused by a germline defect in one of the DNA mismatch repair genes. Currently, MSI-H GCs are known to have distinctive clinicopathologic features compared to MSS/MSI-L (MSS/L) GCs. The consistently reported features of MSI-H GCs are old age, antral
Received 7 May 2014. Accepted 17 October 2014
215
CHOI et al.
location and intestinal type by Lauren classification (6–8). However, the regional lymph node (LN) status and patient prognosis according to MSI status remains unclear, whereas MSI-H colorectal cancers are less prone to lymph node (LN) metastasis and have more favorable prognosis than MSS tumors (9, 10). Meanwhile, upon routine pathologic examination of resected GC specimens, minute metastatic tumors, which are present as a single cell or a cluster of cells, are occasionally found within the dissected LNs. These occult LN metastases often go undetected by routine examination using hematoxylin-eosin staining, and this lack of detection may lead to recurrence of tumors in node-negative GC patients who had undergone curative surgery (11, 12). However, to date, the prognostic impact of occult LN metastases on patient survival in GCs remains controversial (13). Moreover, it has not been studied whether the occult LN metastasis status of GCs differs according to the MSI status. In this study, we investigated the prevalence of the MSI-H phenotype in 623 consecutively resected GC cases and its correlation with various clinicopathologic features, particularly relevant to regional LN metastasis. We also analyzed the clinicopathologic variables associated with occult LN metastasis. In addition, we evaluated the prognostic significance of MSI status and occult LN metastasis. MATERIALS AND METHODS Patients A total of 646 consecutive patients with gastric adenocarcinoma who underwent curative resection and pathologic examination at the Seoul National University Bundang Hospital between August 2010 and September 2011 were enrolled. Among them, specimens were available in 623 GC cases (96.4%), thus a total of 623 cases were included in this study. The demographic and clinicopathologic parameters were evaluated by reviewing medical charts and pathologic records. Tumor staging was based on the pathologic TNM classification according to the 7th edition of the AJCC Cancer Staging Manual (14). None of the patients had fulfilled the revised Bethesda criteria for Lynch syndrome (15). Clinical outcomes were followed up from the date of surgery until October 2013 [median (range), 28 (2–39) months]. This study was approved by the institutional review board of Seoul National University Bundang Hospital.
Microsatellite analysis In all 623 cases, tumor and corresponding non-neoplastic gastric mucosa tissue were obtained from formalin-fixed and paraffin-embedded tissue samples by manual microdissection using a scalpel blade. The size of the microdis-
216
sected tumor was 1.0 9 0.5 9 0.01 cm, and the percentage of tumor cells exceeded a minimum of 60% of the total number of cells. DNA was extracted from the microdissected tissues with InstaGene Matrix (Bio-Rad Laboratories, Hercules, CA, USA). The extracted DNA was amplified by polymerase chain reaction (PCR) with fluorescent dye-labeled primers targeting five microsatellite markers as recommended by Bethesda guideline on MSI (BAT25, BAT26, D2S123, D5S346, and D17S250). Fragment analysis was performed with ABI-3130xl genetic analyzer (Applied Biosystems, Foster City, CA, USA) and GeneMapperâ software (Applied Biosystems). According to the Bethesda guideline on MSI, tumors were classified as MSI-H when at least two of the five markers displayed novel bands, MSI-L when additional alleles were observed with one of the five markers, and MSS when all microsatellite markers examined displayed identical patterns in both tumor and normal tissues (4).
Immunohistochemistry Immunohistochemical staining of pan-cytokeratin (panCK) was performed for all dissected regional LNs from 391 GC patients in whom regional LN metastasis could not be identified by initial pathologic examination. The total number of examined LNs was 17 548. A single serial section (4 lm) was cut from each paraffin block, deparaffinized, and dehydrated. Immunostaining of the fixed tissues was performed with antibodies against pan-CK (mAb anti-human cytokeratin clone MNF116, DAKO, Glostrup, Denmark) at 1:100 dilution using a streptavidin peroxidase procedure, following an antigen retrieval process using microwave heating. According to the 7th edition of the AJCC Cancer Staging Manual, LN metastases detected by immunohistochemistry can be subdivided into three groups (14): Isolated tumor cells (ITCs), when the largest diameter was no larger than 0.2 mm, micrometastases when larger than 0.2 mm but their largest diameter was no larger than 2.0 mm, and macrometastases when their largest diameter was larger than 2.0 mm. The tumor deposits in regional LNs with histologic evidence of malignant activity, such as gland formation or stromal reaction were defined as micrometastases even if their largest dimension was smaller than 0.2 mm (Fig. 1). ITCs were subdivided into four groups (single cell, multiple individual cells, single cluster, and multiple clusters), as previously reported (16). In pathologic staging, ITCs were classified as pN0(i+), and micrometastases were classified as pN1mi. Micrometastases were found in seven of 391 cases (1.8%), which were restaged into pN1.
Statistical analysis In order to determine the correlation between MSI status and clinicopathologic parameters, the categorical variables were compared using the v2 test or Fisher’s exact test and the continuous data were compared by the Mann–Whitney U test. Logistic regression analysis was performed to determine the predictors of occult LN metastasis. Kaplan–Meier method was used to estimate the survival and the differences between survival curves were determined using the log-rank test. The results were considered to be statistically significant when p values
© 2014 APMIS. Published by John Wiley & Sons Ltd
NODAL STATUS IN MSI-H GASTRIC CARCINOMA
A
B
C
D
Fig. 1. Representative photomicrographs of MSI-high (MSI-H) and Microsatellite stable (MSS) gastric cancers with occult lymph node metastasis. (A) MSI-H gastric cancer showing intestinal histologic type and lymphatic invasion (hematoxylineosin stain, 9200). (B) Occult lymph node metastasis in MSI-H gastric cancer (pan-cytokeratin immunohistochemistry, 9400). (C) MSS gastric cancer showing diffuse histologic type without lymphatic invasion (hematoxylin-eosin stain, 9200). (D) Occult lymph node metastasis in MSS gastric cancer (pan-cytokeratin immunohistochemistry, 9400). were less than 0.05. All statistical analysis was conducted using SPSS 19.0 statistical software program (SPSS, Chicago, IL, USA).
RESULTS Clinicopathologic characteristics of MSI-H gastric cancers
Of the 623 consecutive GC cases, 513 (82.3%), 42 (6.7%), and 68 cases (10.9%) were MSS, MSI-L, and MSI-H, respectively. Table 1 summarizes the clinicopathologic characteristics of GCs according to the MSI status. The MSI-H phenotype was significantly associated with old age (p < 0.001), distal location (p < 0.001), intestinal type based on Lauren’s classification (p < 0.001), less infiltrative gross type in both early GC and advanced GC (p = 0.026 and 0.024, respectively), expanding growth based on Ming’s classification (p < 0.001), the absence of venous invasion (p = 0.015), the absence of perineural invasion (p = 0.017), and the absence of stromal desmoplasia (p = 0.003). MSI-H GCs were more likely to be tubular or papillary adenocarcinomas (p < 0.001) and well-to-moderately differentiated type tumors (p < 0.001). On the other hand, there was no significant association between the MSI-H phenotype and female gender (p = 0.434), mucinous histology (p = 0.380), stromal lymphoid follicles (p = 0.112), invasion © 2014 APMIS. Published by John Wiley & Sons Ltd
depth (p = 0.657), distant metastasis (p = 0.563), or pTNM stage (p = 0.166). The percentage of stage III–IV cases was lower in MSI-H GCs than those of MSS/L GCs although the difference was not statistically significant (MSS/L, 26.7%; MSI-H, 16.2%; p = 0.061). Survival analysis suggested that MSI-H GCs showed longer disease-specific survival (DSS) and progression-free survival (PFS) than those of MSS/ L GCs, but the differences were not statistically significant (p = 0.159 and 0.286, respectively) (Fig. 2). MSI-H phenotype and regional LN metastasis
There was no statistically significant difference in the presence of regional LN metastasis between MSI-H and MSS/L GCs (p = 0.115). However, when we analyzed the subgroups of tumor invasion, the MSI-H phenotype was significantly associated with the absence of LN metastasis in pT3/T4 GCs (MSS/L, 19/178; MSI-H, 8/20; p = 0.002), but not in pT1/T2 GCs (data not shown). In addition, MSI-H GCs tended to have fewer metastatic LNs than MSS/L GCs, with borderline statistical significance (p = 0.052) (Table 2). When we analyzed the subgroups of tumor invasion, MSI-H GCs had fewer metastatic LNs compared to MSS/L GCs in pT3/T4 cases (p = 0.002). In contrast, pT1/T2 GCs did not show any difference in the number of metastatic LNs according to the MSI status (p = 0.901).
217
CHOI et al.
Table 1. Clinicopathologic characteristics of 623 consecutive Characteristics Total (%) (n = 623) Age (mean SD) (y) 60.5 12.9 Sex Male 425 (68.2) Female 198 (31.8) Tumor size 3.9 2.8 (mean SD) (cm) Location Lower 1/3 370 (59.4) Middle 1/3 137 (22.0) Upper 1/3 112 (18.0) Remnant 4 (0.6) Gross type (EGC) 0-I (protruded) 14 (4.0) 0-IIa (elevated) 60 (17.1) 0-IIb (flat) 45 (12.8) 0-IIc (depressed) 226 (64.4) 0-III (excavated) 6 (1.7) Gross type (AGC) Borrmann 1 8 (3.0) Borrmann 2 34 (12.6) Borrmann 3 201 (74.4) Borrmann 4 27 (10.0) WHO classification Tubular/papillary 448 (71.9) Mucinous 12 (1.9) Poorly cohesive 146 (23.4) Mixed 11 (1.8) Other 6 (1.0) Histologic grade Differentiated 288 (46.2) Undifferentiated 335 (53.8) Lauren’s classification Intestinal 303 (48.6) Diffuse 294 (47.2) Mixed/indeterminate 26 (4.2) Lymphatic invasion Absent 377 (60.5) Present 246 (39.5) Venous invasion Absent 535 (85.9) Present 88 (14.1) Perineural invasion Absent 431 (69.2) Present 192 (30.8) Ming’s classification Expanding 213 (34.2) Infiltrative 410 (65.8) Desmoplasia Absent 513 (82.6) Present 108 (17.4) Stromal lymphoid follicles Absent 581 (93.4) Present 41 (6.6) Depth of invasion pT1/T2 425 (68.2%) pT3/T4 198 (31.8%) Lymph node metastasis Absent 384 (61.6) Present 239 (38.4)
218
gastric cancers according to the MSI status MSS/L (%) MSI-H (%) (n = 555) (n = 68) 59.5 12.9 68.6 9.6 381 (68.6) 174 (31.4) 3.9 2.8
44 (64.7) 24 (35.3) 3.8 2.7
p value
