P-213 Prognostic biomarkers in patients with oesophago-gastric cancer treated with immunotherapy

Immune checkpoint inhibition is a new standard of targeted therapy in the management advanced oesophago-gastric cancer with or without chemotherapy. PD-L1 expression, microsatellite instability (MSI) and other potential biomarkers, such as tumor-infiltrating lymphocyte neutrophil-to-lymphocyte ratio (NLR) represent good prognostic predictors survival patients undergoing immunotherapy (IO). We aim to analyse outcomes, biomarkers clinicopathological patients’ features metastatic treated IO. performed an unicentric retrospective analysis (p) treatment IO (nivolumab pembrolizumab) alone combination chemotherapy (based on fluoropyrimidines platinum) from May/2020 Jan/2023. used X 2 test compare characteristics Kaplan Meier analyze overall (OS) among different subgroups. 17p were included. Age (median): 65 [42-83]. Males (76.5%). Tumour localization: stomach 10p (58.8%), gastroesophageal junction 1p (5.9%) oesophagus 6p (35.3%). Histology defined intestinal/diffuse adenocarcinoma 15p (82.4%) epidermoid 2p (17.6%). (58.8%) debuted stage 4. Surgery primary tumour was 7p (41.2%). Most frequent involvement lymph nodes (82.4%), liver (29.4%) peritoneum (29.4%). Deficiency mismatch-repair (MMR) proteins observed tested by combined positive score (CPS) 12p (70.6%). plus 7p. (70.6%) received 1L treatment. Of population, 70.6% had not progressed at 6.5 months (m) follow-up. Median OS 26m (95%CI;12.3-39.6). For those CPS>10 (8p), we longer survival; media (37.7m vs 18.6m; p=0.65). also noticed worse prognosis metastases; median (18.5m 26m; p=0.55). An objective response rate (ORR) 52.9% recorded. described grade 3 toxicity-related 3p In terms baseline analytical factors, derived NLR below value ( 17%); p=0.73, low relative eosinophil count (< 1.8%); p=0.58 absolute monocyte < 640/μL); p=0.33. Immunotherapy effective therapy, especially deficient MMR/MSI-high tumours high CPS, line our results. Other count, well absence metastases, could identify subgroup greater clinical benefit.