Abstract
Background/Aim: Endoscopic submucosal dissection (ESD) effectively treats selected early gastric cancers (EGCs). However, the association of microsatellite instability (MSI), Epstein-Barr virus (EBV), p53, and β-catenin status with clinicopathologic parameters in EGCs treated with ESD have not been well studied. Patients and Methods: We retrospectively collected 312 consecutive EGC cases treated with ESD from January 2021 to December 2023 at Kyungpook National University Chilgok Hospital. MSI polymerase chain reaction, EBV encoded RNA in situ hybridization, and p53 and β-catenin immunostaining were performed for all cases. Results: Among 312 EGC cases, there were 42 MSI-High (MSI-H) cases (13.5%), 13 EBV-associated gastric cancer (EBVaGC) cases (4.2%), 249 intestinal type cases (79.8%), and eight poorly cohesive carcinoma cases (2.6%). MSI-H was significantly associated with lymphovascular invasion (p=0.02), local recurrence (p=0.03), and synchronous tumors (p<0.001). More than half of EBVaGC cases showed submucosal invasion (61.5%, 8/13) (p=0.016). Consequently, non-curative ESD was more frequently found in EBVaGC than in other subtypes (p<0.001). Mutant p53 patterns and nuclear translocation of β-catenin were almost exclusively found in the intestinal type (p<0.001), without association with clinicopathologic parameters. Margin involvement was frequent in poorly cohesive carcinoma (p=0.003). Conclusion: We demonstrated that MSI-H and EBVaGC are strongly associated with clinicopathologic parameters and risk factors in EGCs treated with ESD. Molecular testing of gastric cancers should be considered before ESD for better patient management.
- Early gastric cancer
- endoscopic submucosal dissection
- microsatellite instability
- Epstein-Barr virus
- p53
- β-catenin
Gastric cancer is one of the most common malignancies worldwide, particularly more prevalent among East Asian patients (1, 2). Thanks to the national cancer screening program and advancements in endoscopic procedures, over 70% of gastric cancers in Korea are detected early at resectable stages (3). Endoscopic submucosal dissection (ESD), introduced in the early 2000s, has largely replaced surgical gastrectomy for early gastric cancers (EGCs) that meet the absolute or expanded indications for ESD (4). Several large cohort studies have demonstrated the effectiveness and safety of ESD. Compared to surgical procedures, ESD offers several advantages, including improved quality of life, preservation of gastric function, lower complication rates, and better short- and long-term clinical outcomes (3, 5, 6).
With advancements in sequencing technology, several research groups have proposed molecular classifications of gastric cancer (7, 8). The Cancer Genome Atlas (TCGA) group has categorized gastric cancer into four subgroups: microsatellite instability-high (MSI-H), Epstein-Barr virus (EBV)-associated gastric cancer (EBVaGC), chromosomally unstable, and genomically stable (9). The chromosomally unstable and genomically stable groups correspond to Lauren’s classification of intestinal and diffuse subtypes. Intestinal-type gastric cancer frequently harbors TP53 mutations. Additionally, the Wnt/β-catenin signaling pathway is known to be an early event in tumorigenesis in intestinal-type gastric cancer (10). MSI-H gastric cancer has a very high mutational tumor burden and rarely exhibits copy number alterations. EBVaGC displays distinct clinicopathologic characteristics, including upper location, male predominance, and infrequent TP53 mutations (9).
Although several groups have studied the clinicopathologic characteristics of EGCs treated by ESD (11-13), their molecular characteristics have not been well investigated. A subset of EGC patients treated with ESD undergo non-curative resection, requiring additional gastrectomy (14). During follow-up after ESD, some patients may experience local or distant recurrence (15, 16). The molecular characteristics of EGCs and their association with non-curative resection or local recurrence have not been well characterized.
In this study, we retrospectively investigated the status of MSI, EBV, p53, and β-catenin in EGCs resected via the ESD procedure. We further demonstrated their association with clinicopathologic parameters.
Patients and Methods
Study population. We retrospectively collected 312 consecutive EGC cases treated with ESD from January 2021 to December 2023 at Kyungpook National University Chilgok Hospital. Regular endoscopic surveillance was performed at 3 months, 6 months, 1 year, and annually after ESD. Clinicopathological data, including age, sex, gross appearance, tumor size, tumor location, tumor depth, lymphovascular invasion, lymph node metastasis, margin involvement, local recurrence, synchronous tumors, and gastrectomy after non-curative resection were obtained from the hospital’s electronic medical records. Local recurrence was defined as a tumor recurrence at the ESD site (16). Curative and non-curative resections were evaluated following Japanese ESD guidelines (17). The study was conducted in accordance with the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Kyungpook National University of Chilgok Hospital (No. KNUCH 2024-09-016). The requirement for written informed consent from the patients was waived because of the retrospective nature of the study.
Pathological evaluation. Gastric tumor specimens were preserved in 10% neutral-buffered formalin and embedded in paraffin blocks. The paraffin-embedded tissues were then cut into 4-micrometer sections and stained with hematoxylin and eosin for histological analysis. Pathologic evaluation was performed according to the Japanese classification of Gastric Cancer (18) and the WHO classification (19).
Immunohistochemistry and interpretation. Deparaffinization and rehydration of formalin-fixed, paraffin-embedded tissue sections were carried out using xylene and graded alcohol solutions. The sections were then incubated with primary antibodies targeting p53 (mouse monoclonal, clone DO7, dilution 1:300, NCL-p53-CM5p, Novocastra, Newcastle, UK) and β-catenin (mouse monoclonal, clone 14, dilution 1:400, Rocklin, CA, USA). Immunohistochemical staining for p53 and β-catenin was conducted on all cases as part of the diagnostic evaluation. Chromogenic detection was accomplished using the ultraView Universal DAB Detection Kit (Ventana Medical Systems, Oro Valley, AZ, USA) or EnVision FLEX/HRP (Agilent Technologies, Santa Clara, CA, USA), followed by counterstaining with hematoxylin. The interpretation of p53 immunoreactivity classified a “mutant” pattern when 60% or more of tumor cells exhibited strong nuclear staining or a complete absence of staining, while a “wild-type” pattern was attributed to cases where 1% to 59% of tumor cells showed nuclear p53 expression, as previously established (20). For β-catenin immunostaining evaluation, any percentage of nuclear translocation was considered nuclear translocation positive. Previous studies have reported a reduced membranous beta-catenin staining pattern in gastric cancers (21, 22). However, no consensus guideline exists for interpreting the intensity of β-catenin membranous staining. To enhance reproducibility and avoid equivocal results, we considered only the absence or near absence of membranous staining as indicative of reduced membranous staining.
Epstein-Barr virus-encoded RNA (EBER) in situ hybridization. EBER in situ hybridization was carried out on all specimens as part of the diagnostic assessment. To determine the EBV status, the INFORM EBV-encoded RNA probe (Ventana Medical Systems) was applied according to the manufacturer’s protocol. Positive controls, consisting of samples from patients with EBV-positive nasopharyngeal carcinoma, were included in each hybridization run.
MSI testing. MSI testing was conducted for all cases as part of the diagnostic process. MSI status was assessed through polymerase chain reaction (PCR) using five National Cancer Institute-recommended markers: BAT-26, BAT-25, D5S346, D17S250, and S2S123. Both tumor tissue samples and matched normal tissue were analyzed. PCR products were examined using a DNA sequencer (ABI 3731 Genetic Analyzer; Thermo Fisher Scientific, Waltham, MA, USA). Based on the revised Bethesda Guidelines (23), tumors exhibiting instability in two or more markers were classified as MSI-H, while those with one or no unstable markers were defined as microsatellite stable.
Statistical analysis. Associations between clinicopathological variables were analyzed using chi-square and one-way ANOVA tests. Fisher’s exact test was applied for variables with an expected frequency below 5. Statistical significance was defined as a p-value of <0.05. All statistical analyses were performed using R software (version 4.3.2; R Foundation for Statistical Computing, Vienna, Austria).
Results
Study cohorts. Detailed clinicopathologic characteristics of the patient cohort are presented in Table I. The median age of the patients was 68 years (range=40-89 years). Of the total, 234 (75.0%) were male, and 78 (25.0%) were female. The median follow-up duration was 23 months (range=2-36 months). Local recurrence was found in four cases. Curative resection was achieved in 82.1% (256/312). For the 56 non-curative resection cases, additional gastrectomy was performed in 45 patients. The remaining 11 patients either refused the surgery or were transferred to another hospital. Among the 45 cases that underwent additional gastrectomy, residual gastric cancer was found in eight cases. Regional lymph node metastasis was found in six cases (Table II). No clinicopathologic parameters were statistically associated with residual tumor cells or lymph node metastasis in the additional gastrectomy specimens. However, lymph node metastasis in the gastrectomy specimens tended to be related to lymphovascular invasion in the previous ESD specimens (p=0.08).
Baseline characteristics of the patient cohort.
Pathologic evaluation of additional gastrectomy specimens after non-curative endoscopic submucosal dissection.
Molecular characteristics of gastric cancers treated with ESD and their association with clinicopathologic parameters. We categorized gastric cancers into MSI-H (n=42, 13.5%), EBVaGC (n=13, 4.2%), and non-MSI-H/non-EBVaGC (n=257, 82.4%). We further divided non-MSI-H/non-EBVaGC into intestinal type (n=249, 79.8%) and poorly cohesive carcinoma (n=8, 2.6%) according to histology. Mixed adenocarcinomas were categorized based on the dominant histology. This division is similar to the TCGA classification (9), allowing us to best reflect their biological behavior. Representative photos of each subtype are shown in Figure 1.
Representative images according to gastric cancer subtypes. (A) H&E image of intestinal-type gastric cancer. Tumor cells show irregular, branching and distorted tubular structures. (B) H&E image of poorly cohesive carcinoma. Discohesive tumor cells show abundant intracytoplasmic mucin. (C) EBV-associated gastric cancer (histologically, gastric carcinoma with lymphoid stroma) is characterized by irregular sheets and tubules of tumor cells embedded within dense lymphocytic infiltration. In this image, tumor cells with dense lymphoid stroma invade into the submucosa. (D) EBER in situ hybridization result corresponding to H&E image (C). (E) MSI-High type gastric cancer. The tumor cells infiltrate into the submucosa, with frequent evidence of lymphatic invasion. Original magnifications: (A), (C): ×50, (B), (D): ×100, (E): ×20.
Clinicopathologic characteristics according to each subtype are summarized in Table III. MSI-H tumors were statistically significantly associated with elevated gross appearance (p=0.002), local recurrence (p=0.03), and synchronous tumors (p<0.001). Lymphovascular invasion was also frequently observed compared to other subtypes (p=0.02). EBVaGCs were exclusively found in male sex (p<0.001). Notably, more than half of EBVaGCs presented with submucosal invasion (61.5%, 8/13) (p=0.016). Consequently, non-curative ESD was frequently found in EBVaGC compared to other subtypes (p<0.001). Poorly cohesive carcinoma was associated with female sex (p<0.001) and middle location (p<0.001). Intestinal subtype was associated with male sex (p<0.001) and lower location (p<0.001).
Clinicopathologic characteristics of early gastric cancers according to subtypes.
On immunostaining, mutant pattern p53 (p<0.001) and nuclear translocation of β-catenin (p<0.001) were predominantly found in the intestinal type. Of note, nuclear translocation of β-catenin almost always accompanied mutant-pattern p53 staining (41/45, 91.1%) (p<0.001). Reduced β-catenin staining tended to be associated with mutant-pattern p53 staining, although not statistically significant (p=0.086) (Figure 2). Mutant pattern p53 or aberrant β-catenin staining pattern was not associated with any clinicopathologic parameters and risk factors investigated.
Representative images of p53 and β-catenin immunostaining. (A) Diffuse, strong nuclear p53 (mutant pattern p53) immunostaining pattern was found in tumor cells. (B) β-catenin staining in the same tumor. Nuclear translocation is observed. (C) A complete negative (mutant pattern p53) staining pattern was identified. (D) In the same tumor, reduced membranous β-catenin staining is observed. Original magnifications: (A)-(B): ×100, (C)-(D): ×50.
Discussion
In this study, we demonstrated that the molecular characteristics are closely associated with clinicopathologic parameters and risk factors in EGC treated with ESD. EBVaGCs morphologically correspond to carcinoma with lymphoid stroma, characterized by irregular tubules and sheets of tumor cells embedded with lymphocyte-rich stroma (24). Park et al. suggested EBVaGC as a possible candidate for ESD based on the lower rate of lymphatic invasion and lymph node metastasis (25). However, EBVaGC was frequently associated with submucosa invasion and non-curative resection in this study. A few previous studies demonstrated deeper submucosal invasion of EBVaGC compared to a control group in submucosa-invasive gastric cancer (26, 27), consistent with the results of the present study.
MSI-H type tumors can be found in diverse tumor types, frequently occurring in the stomach, colon and uterus tumors (28). In line with the present study, frequent lymphovascular invasion has been previously reported in MSI-High-type tumors (29, 30). The propensity for synchronous and metachronous tumors in MSI-H gastric cancers has also been reported (31, 32). Additionally, we suggest that local recurrence may happen more frequently in MSI-H gastric cancer compared to other subtypes. However, the number of observed recurred cases in the study was small.
EBV-positivity and MSI-H gastric cancer treated with ESD might thus lead to non-curative resection and additional gastrectomy. Although EBV-positivity and MSI-H status seem to be a warning sign for curative ESD, interestingly, EBVaGCs and MSI-H gastric cancer are usually considered to have favorable clinical outcomes with good responses to immune checkpoint inhibitors (24, 33, 34). More extensive cohort studies should be followed to understand the short- and long-term clinical outcomes of EBVaGC and MSI-H gastric cancer. Feasibility of endoscopic submucosal dissection in these subtypes should be carefully considered.
Most mutant pattern p53 staining was found in intestinal-type gastric cancer, consistent with previous studies (7, 9). Although the prognostic significance of TP53 mutation has been reported in several tumor types (35-37), TP53 mutational status was not associated with any clinicopathologic parameters in this study. However, the significance of this result can be limited due to the lack of survival analysis. Interestingly, nuclear translocation of β-catenin staining predominantly accompanied mutant pattern p53 staining in this study. Conversely, nuclear translocation of β-catenin without mutant pattern p53 staining was rarely found in EGCs in this cohort. This is consistent with the fact that the Wnt/β-catenin pathway involves early tumorigenesis and requires TP53 mutation as a late event during gastric cancer carcinogenesis (10). Several previous reports demonstrated that reduced β-catenin immunostaining can be associated with a worse prognosis in gastric cancer (21, 22). However, reduced β-catenin staining was not associated with any clinicopathologic parameters in this study.
Study limitations. Since ESD is selectively performed in poorly cohesive carcinoma according to the expanded ESD criteria, our cohort has only a few poorly cohesive carcinoma cases; their clinicopathologic characteristics were not comprehensively investigated. Due to the short follow-up period, survival analysis was not performed. While MSI EBV testing in combination with histologic evaluation and immunohistochemistry can successfully recapitulate TCGA molecular classification (38), next-generation sequencing might be needed to best illustrate the genomic characteristics of early gastric cancer.
Conclusion
We have demonstrated that MSI-High type and EBVaGC can be associated with adverse clinicopathologic parameters in EGCs treated with ESD. Molecular testing of gastric cancers should be considered before ESD for better patient management.
Footnotes
Authors’ Contributions
MSK conceived and designed the manuscript. JHK and JYC drafted the manuscript. JYC and HJK analyzed previous articles on gastric cancer and endoscopic submucosal dissection. JHP, HJK, and MSK reviewed and revised the manuscript carefully. All Authors have read and approved the final manuscript.
Funding
This work was supported by a Biomedical Research Institute grant, Kyungpook National University Hospital (2023).
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
- Received September 7, 2024.
- Revision received September 18, 2024.
- Accepted September 19, 2024.
- Copyright © 2024 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.