Discrepancies between preoperative assessment and final pathological criteria in early gastric cancer.

BACKGROUND: Inappropriate selection of patients with early gastric cancer (EGC) for endoscopic submucosal dissection (ESD) may lead to non-curative resection, necessitating additional gastrectomy. Conversely, inappropriate selection for gastrectomy may result in overtreatment, adversely affecting patients' quality of life. Few have systematically evaluated the concordance between therapeutic indications under current Japanese guidelines and pathological criteria in EGC. To minimize noncurative resection risks while sparing unnecessary surgery for low-risk patients', we specifically assess the suitability of Japanese guidelines in non-Japanese populations. This work aims to optimize clinical practice by refining endoscopic treatment criteria for adoption beyond Japan. AIM: To evaluate EGC clinical decision accuracy by comparing therapeutic indication with postoperative pathological criteria and analyzing factors influencing discrepancies. METHODS: A retrospective analysis was conducted on 796 EGC cases diagnosed at Peking University Third Hospital between January 2010 and December 2022. Cases were categorized into three groups: Same-estimated (preoperative therapeutic indication with postoperative pathological criteria matched), underestimated (preoperative ESD indication but postoperative surgical criteria), and overestimated (preoperative surgical indication but postoperative ESD criteria). The rate of discrepancy and associated risk factors were assessed. RESULTS: The accuracy rates of preoperative evaluation for ESD and gastrectomy indications were 73.0% (321/430) and 76.0% (278/366), respectively. The overall discrepancy rate was 25.6% (204/796). Multivariate analysis identified tumor location in the upper-third stomach (odds ratio = 2.158, 95% confidence interval: 1.373-3.390, P = 0.001) was significantly associated with a higher likelihood of being underestimated and undifferentiated histologic type on preoperative biopsy (odds ratio = 2.005, 95% confidence interval: 1.036-3.879, P = 0.039) was more likely to be overestimated. Significant differences were observed in tumor diameter (P < 0.001), depth of infiltration (P < 0.001), ulcerative findings (P < 0.001), and histologic type (P < 0.001) between preoperative and postoperative evaluations. CONCLUSION: The accuracy of preoperative EGC indications is 74.4%. Upper-third stomach and undifferentiated histology are primary discrepancy predictors. Upper-third tumors are prone to underestimation, while undifferentiated tumors are prone to overestimation.