754. PATHOLOGICAL RESPONSE PREDICTS EFFICACY OF ADJUVANT FLOT CHEMOTHERAPY IN GASTROESOPHAGEAL CANCER: AN INTERNATIONAL COHORT STUDY

Abstract Background Perioperative FLOT-based chemotherapy is the current standard of care for all patients with locally-advanced gastroesophageal cancer. However, many patients are unable to complete all planned adjuvant FLOT cycles due to treatment related toxicity and diminished performance status after radical surgery. Within this context, it is unclear whether adjuvant FLOT will improve oncological outcomes for patients with cancers that either don’t respond or respond exceptionally well to neoadjuvant FLOT. To facilitate clinical decision-making, we examined whether pathological response to neoadjuvant FLOT can guide its adjuvant use. Methods Prospectively collected data from 43 centres across 12 countries was analysed. Patients with non-metastatic gastroesophageal adenocarcinoma who received neoadjuvant FLOT followed by radical surgery between 2017-2022 were included. Pathological response was assessed using validated tumour regression grading (TRG) systems. All TRG were trichotomized into minimal responders (MR=worse TRG category), complete responders (CR=pathological complete response), and partial responders (PR=all TRG categories between MR and CR). Survival outcomes of patients who did and did not receive adjuvant FLOT within these cohorts were compared using Kaplan Meier and Cox regression analysis. These findings were validated using propensity score matched analysis. Results 1887 patients (MR n=459, CR n=221, PR n=1207) were evaluated. The median follow-up was 25.5 months. In the MR group, there was no DFS (HR=1.03, 95%CI 0.78-1.36) or adjusted-OS (HR=0.96, 95%CI 0.70-1.30) difference between those who did (n=272) and did not (n=187) receive adjuvant FLOT. In the CR group, there was no DFS (HR=0.88, 95%CI 0.41-1.85) or OS (HR=0.69, 95%CI 0.31-1.54) difference between those who did (n=136) and did not (n=85) receive adjuvant FLOT. In the PR group (adjuvant FLOT n=847, no adjuvant FLOT n=360), adjuvant FLOT conferred a significant DFS (HR=0.68, 95%CI 0.55-0.86) and OS (HR=0.55, 95%CI 0.44-0.69) benefit. Conclusions Pathological response to neoadjuvant FLOT predicts therapeutic efficacy to adjuvant FLOT. Only partial responders benefit from adjuvant treatment. Our findings challenge the current doctrine that perioperative FLOT should be applied to all patients regardless of pathological tumour response.