Patients who were diagnosed with EGC between July 2009 and December 2013 by endoscopic screening and subsequent pathology at Chinese PLA General Hospital were included. The exclusion criteria were (1) other malignancies; (2) incomplete clinical or pathological data.
Patients were followed up by phone calls to determine survival status and cancer recurrence. The most recent follow-up was conducted in July 2020. All included patients were followed up for at least 6 years. Finally, 211 patients were included in the analysis. Additional file 1: Figure S1 shows the process of identification for participants. The study was conducted under the approval of the Ethics Board of the Chinese PLA General Hospital (ethics number: S2020-251-01), in accordance with the standards of the Declaration of Helsinki and Good Clinical Practice guidelines.
HER2 expression status was determined using immunohistochemistry (IHC). Rabbit anti-human HER2 antibody was used (Roche Diagnostics Japan. Tokyo, Japan). Two pathologists individually scored the HER2 expression. Conflicts were resolved by a third pathologist’s interpretation. A semiquantitative approach, recommended by the National Comprehensive Cancer Network guidelines, was used to classify the expression status of HER2. We classified HER2 expression status into 2 groups: negative (0 and 1 +) and overexpression (2 + and 3 +) . The definition for each was as follows: 0 = no membrane staining, or membrane responds to staining with < 10% of tumor cells; 1 + = faint staining or barely recognizable staining in more than 10% of tumor cells; 2 + = weak to moderate staining of the entire membrane in > 10% of tumor cells; 3 + = > 10% of tumor cells show a strong complete staining reaction in the entire membrane.
Pathologic characteristics and definitions
Pathologic characteristics included depth of infiltration, lesion location, lesion morphology, lesion differentiation, tumor size, and lymphovascular infiltration. Depth of infiltration was categorized as mucosal or submucosal; lesion location as upper third, middle third, or lower third based on the lesion location on the longitudinal axis of the stomach; and lesion morphology as elevated (types 0-I, 0-IIa, 0-I + IIa, 0-IIa + IIb, 0-IIa + IIc), flat (type 0-IIb), or depressed (types 0-IIc, 0-III, 0-IIc + IIa, and 0-III + IIa) . Histologically, tumors were characterized into 2 types: differentiated (papillary adenocarcinoma or moderately- and well-differentiated tubular adenocarcinoma) or undifferentiated (poorly differentiated tubular adenocarcinoma, signet-ring cell carcinoma, or mucinous adenocarcinoma). Tumor size was measured by the longest diameter of the lesion in the specimen and categorized into 2 groups: > 2 cm or ≤ 2 cm. Lymphovascular infiltration was classified into 2 groups: present or absent. In this study, three other clinical variables were also included: resection margin (positive or negative), current gastric ulcer (yes or no) and current H. pylori (Hp) infection when admitted to our hospital for EGC. Patients had previous Hp infection and have received successful Hp eradication therapy were classified as not having current Hp infection. Methodology of Hp diagnostic included serum antibody, rapid urease test, immunohistochemistry, 13C or 14C urea breath test, or stool antigen test.
Prognostic outcomes of this study included overall survival (OS), disease-specific survival (DSS), and tumor recurrence. OS and DSS was ascertained by asking the family members the detailed information recorded on the death certificate of the patients.The OS period was determined from the date of endoscopic treatment for EGC until the date of death from any cause. The DSS period was determined from the date of endoscopic treatment for EGC to the date of death due to tumor recurrence (patients experienced recurrent GC, and died from dyscrasia or massive hemorrhage due to GC, or died from malignant event caused by GC distant metastasis, including severe lung infection, cerebral edema or hernia, or hepatic failure). The recurrence pattern in this study included both metachronous recurrence and metastatic recurrence. The tumor recurrence period was defined as the time from endoscopic treatment for EGC to the date of detection of new gastric cancer. Tumor recurrence was identified using routine abdomen-pelvis computed tomography and endoscopy after endoscopic treatment.
We reported and compared the baseline clinicopathologic characteristics of included patients according to HER2 expression status, using the chi-square test or Fisher exact test for categorical data and the Student t test for continuous data. Multivariate Cox regression with stepwise backward selection was applied to identify factors associated with prognostic outcomes. For OS and DSS, 8 factors (age, sex, tumor size, depth of invasion, resection margin status, lymphovascular invasion, tumor differentiation, and HER2 expression status) were included in the model. For tumor recurrence, 2 additional factors (recent history of gastric ulcer and HP infection) were added in the model. These factors were chosen because they were considered potentially clinically relevant for each outcome. Survival curves for each prognostic outcome were generated using the Kaplan–Meier method and were compared using the log-rank test. Statistical analyses were performed using SPSS 26.0 for Windows (SPSS, Inc., Chicago, IL). Survival curves were graphed using GraphPad Prism 8 (GraphPad Prism version 8.1, GraphPad Software, La Jolla, CA). Two-tailed P values < 0.05 were considered statistically significant.