III期胃癌的预后仍然较差，新辅助化疗越来越多地用于改善预后。治疗前的准确诊断对于制定针对不良预后亚组的适当治疗策略至关重要。本研究旨在使用以循环环状 RNA (circRNA) 为中心的生物学方法来提高治疗前胃癌诊断的准确性。我们使用两个基因表达综合数据集对 circRNA 表达谱进行了全面分析，以确定与 III 期相关的 circRNA 候选者胃癌。随后，我们通过实时聚合酶链式反应 (PCR) 在两个独立的临床队列中验证了这些 circRNA 生物标志物，该队列由 174 名胃癌患者和非疾病对照患者组成。全基因组 circRNA 分析确定了一组四种生物标志物，能够诊断经病理证实的 III 期 (pStage III) 胃癌。在训练队列（n = 83）中，开发了临床适用的四种 circRNA 组（AUC 0.81），并在独立临床队列（n = 82；AUC 0.76）中成功得到验证。为了评估临床实用性，我们将临床成像 (cStage) 与 circRNA 组合结合起来。在最初诊断为 cStage III 但后来确认为 pStage I/II 的患者中，86% 使用 circRNA 的分子生物学方法准确诊断。我们开发了基于 circRNA 的非侵入性液体活检，可以提高 pStage III 的诊断性能治疗前的胃癌。我们的 circRNA 模型可以提供复杂且个性化的方法来协助晚期胃癌患者的治疗计划。© 2024。外科肿瘤学会。

The prognosis remains poor for stage III gastric cancer, and neoadjuvant chemotherapy is increasingly used to improve outcomes. Accurate diagnosis prior to treatment is essential to develop appropriate treatment strategies for poor prognosis subgroups. This study aims to enhance the accuracy of pre-treatment gastric cancer diagnosis using a biological approach centered on circulating circular RNA (circRNA).We conducted a comprehensive analysis of circRNA expression profiles using two Gene Expression Omnibus datasets to identify circRNA candidates associated with stage III gastric cancer. Subsequently, we validated these circRNA biomarkers in two independent clinical cohorts comprising a total of 174 patients with gastric cancer and non-disease controls through real-time polymerase chain reaction (PCR).Genome-wide circRNA analysis identified a panel of four biomarkers capable of diagnosing pathologically confirmed stage III (pStage III) gastric cancer. In a training cohort (n = 83), a clinically applicable panel of four circRNAs was developed (AUC 0.81), which was successfully validated in an independent clinical cohort (n = 82; AUC 0.76). To assess clinical utility, we combined clinical imaging (cStage) with the circRNA panel. Among those initially diagnosed as cStage III but later confirmed as pStage I/II, 86% were accurately diagnosed using the molecular biological approach with circRNAs.We have developed a circRNA-based non-invasive liquid biopsy that can improve the diagnostic performance of pStage III gastric cancer before treatment. Our circRNA model could provide a sophisticated and personalized approach to assist in treatment planning for patients with advanced gastric cancer.© 2024. Society of Surgical Oncology.