Clinical significance of c-Met expression in patients with advanced gastroesophageal adenocarcinoma.

426 Background: The prognosis for advanced gastroesophageal adenocarcinoma (GEA) remains poor, highlighting the urgent need for novel therapeutic options. MET-targeted agents including antibody-drug conjugates are under clinical investigation in GEA. However, the prevalence and prognostic impact of c-Met expression across different treatment regimens remain unclear. Understanding c-Met expression and its correlation with treatment regimens and other biomarkers is crucial for refining patient selection and guiding therapeutic strategy. Methods: We retrospectively analyzed 400 patients (pts) with advanced GEA receiving first-line therapy at our institution. Immunohistochemistry was performed for c-Met, HER2, PD-L1, CLDN18, and FGFR2. Pts were classified as c-Met high vs. low using a threshold of ≥50% with ≥2+ staining. We assessed correlations between c-Met high and other biomarkers, as well as survival outcomes by treatment regimen: chemotherapy alone (CTx, n = 173), combination with immune checkpoint inhibitors (ICI, n = 130), and combination with molecular targeted agents (MTA, n = 97). Progression-free survival (PFS) and overall survival (OS) were assessed using Kaplan-Meier method, log-rank test and Cox proportional hazards models. Exploratory genomic characterization by next-generation sequencing was performed in 177 pts; transcriptomic analysis in 88 pts. Results: High c-Met expression was detected in 11.3% of pts by membrane staining th Ultraview detection kit, showing significant association with PD-L1 positivity (CPS≥5) (P < 0.05) but not other protein biomarkers. c-Met high pts exhibited a trend of shorter PFS in the overall cohort (HR = 1.34; 95% CI: 0.96-1.88; P = 0.086), most pronounced in MTA group (HR = 1.73; 95% CI: 0.95-3.14; P = 0.069). Multivariate analysis confirmed c-Met high as an independent worse prognostic factor for PFS (HR = 1.52; 95% CI: 1.05-2.20; P = 0.025). OS did not differ significantly in the overall cohort (HR = 1.11; 95% CI: 0.77-1.62; P = 0.575) but was notably worse in CTx group for c-Met high pts (HR = 1.89; 95% CI: 1.02-3.52; P = 0.040). Multivariate analysis in CTx cohort supported c-Met as an independent predictor of poor OS (HR = 3.30; 95% CI: 1.48-7.3; P = 0.003). Genomic analysis identified MET amplification in 2.3% (4/177). Transcriptomic analysis revealed significant correlations between c-Met high and elevated MET RNA expression (c-Met high vs. low, median TPM: 800 vs. 113, P = 0.003). Conclusions: c-Met high expression was significantly associated with worse PFS in general, and poor OS in pts treated with chemotherapy alone. Further investigation is warranted to elucidate mechanisms underlying poor outcomes in c-Met high GEA pts as well as expression correlation with recently approved c-Met CDx assay, which may help inform ongoing MET-targeted treatment strategies.