Background: Pathological tumor regression is a key prognostic marker in resectable gastric and gastroesophageal junction (GEJ) adenocarcinoma. Perioperative FLOT (fluorouracil, leucovorin, oxaliplatin, docetaxel) established the modern treatment backbone, and the recent MATTERHORN trial further intensified this paradigm with the addition of durvalumab. However, a substantial proportion of patients fail to achieve major pathological regression, and the prognostic magnitude of poor response in the FLOT era has not been systematically quantified. Methods: We performed a PRISMA 2020–compliant, PROSPERO registered (CRD420251150054) systematic review and meta-analysis. PubMed, Web of Science, Scopus, and Cochrane were searched through September 2025 for studies including patients with resectable gastric/GEJ adenocarcinoma treated with perioperative FLOT followed by curative surgery. Associations between poor pathological response and survival were quantified. Exploratory analyses evaluated completion of postoperative FLOT versus incomplete therapy. Random effects models were applied to pooled hazard ratios (HRs). Results: Twelve studies comprising 4201 resected patients were included in this systematic review; 1817 (43.2%) were classified as poor responders. Poor pathological response was strongly associated with inferior outcomes: pooled HR for overall survival (OS) was 2.73 (95% CI 2.18–3.43; I2=0.0%) and pooled HR for disease-free/recurrence-free survival (DFS/RFS) was 2.68 (95% CI 2.14–3.34; I2=51.2%). Effect direction was consistent across Becker, Mandard, and CAP grading systems. Exploratory analysis of three observational cohorts found an association between completion of postoperative FLOT and improved survival (HR 0.49, 95% CI 0.31–0.79); however, this comparison is inherently confounded by postoperative fitness and treatment selection and should not inform clinical decision-making. Conclusions: Poor pathological response after perioperative FLOT identifies a large, high risk subgroup with approximately threefold inferior survival despite R0 resection. These findings establish contemporary prognostic benchmarks and underscore the absence of prospective evidence guiding postoperative management in poor responders.
Massalha et al. (2026) studied this question.