Prognostic factors of perioperative FLOT regimen in operable gastric and

Abstract

Background/aim: Perioperative FLOT regimen is a standard of care in locally advanced operable gastric and GEJ adenocarcinoma. We aimed to determine the efficacy, prognostic factors of perioperative FLOT chemotherapy in real-life gastric and GEJ tumors.

Materials and methods: The data of patients who were treated with perioperative FLOT chemotherapy were retrospectively analyzed from 34 different oncology centers in Turkey. Baseline clinical and demographic characteristics, pretreatment laboratory values, histological and molecular characteristics were recorded.

Results: A total of 441 patients were included in the study. The median of age our study population was 60 years. The majority of patients with radiological staging were cT3-4N(+) (89.9%, n = 338). After median 13.5 months (IQR: 8.5-20.5) follow-up, the median overall survival was NR (95% CI, NR to NR), and median disease free survival was 22.9 (95% CI, 18.6 to 27.3) months. The estimated overall survival at 24 months was 62%. Complete pathological response (pCR) and near pCR was achieved in 23.8% of all patients. Patients with lower NLR or PLR have significantly longer median OS (p = 0.007 and p = 0.033, respectively), and patients with lower NLR have significantly longer median DFS (p = 0.039), but PLR level did not affect DFS (p = 0.062). The OS and DFS of patients with better ECOG performance scores and those who could receive FLOT as adjuvant chemotherapy instead of other regimens were found to be better. NLR was found to be independent prognostic factor for OS in the multivariant analysis. At least one adverse event reported in 57.6% of the patients and grade 3-4 toxicity was seen in 23.6% patients.

Conclusion: Real-life perioperative FLOT regimen in operable gastric and GEJ tumors showed similar oncologic outcomes compared to clinical trials. Better performance status, receiving adjuvant chemotherapy as same regimen, low grade and low NLR and PLR improved outcomes in real-life. However, in multivariate analysis, only NLR affected OS.