Background: Gastric and esophageal adenocarcinomas are a leading cause of cancer death worldwide. Many of these patients (pts) present with locally advanced unresectable or metastatic disease and are treated with combination cytotoxic chemotherapy. Single agent P is FDA approved for patients with recurrent locally advanced or metastatic gastric or gastroesophageal junction adenocarcinoma (GEJ) whose tumors have a combined positive score (CPS) ≥ 1 after disease progression on or after two lines of therapy including fluoropyrimidine and platinum and her2/neu-targeted therapy (if indicated). More effective therapy is needed earlier in the disease trajectory. We conducted a single-arm phase II trial to establish the safety and efficacy of first-line C and O + P. Methods: Pts with previously untreated metastatic GE adenocarcinoma regardless of PDL-1 status received intravenous (IV) P 200mg with IV O 130mg/m² every three weeks and oral C 850mg/m² twice daily for 14 days on/7 days off. After the 6 patient safety cohort, pts first completed a biomarker cycle that included fresh tumor biopsy before P and one week after P before chemotherapy started. Archived FFPE tumor samples were also obtained from all pts with available tissue. The primary endpoint was progression free survival (PFS); secondary endpoints included response rate (RR) and overall survival (OS). Results: 36 pts were enrolled and 34 pts were evaluable for efficacy (1 pt withdrew for personal reasons before end of cycle 1 and 1 pt had immune-related toxicity during cycle 1 and was taken off study before any efficacy assessment). 9 pts (26%) had an esophageal primary, 18 pts (53%) had a GEJ primary and 7 pts (21%) had a gastric primary. Median PFS was 7.6 months [95% CI: 5.8 to 12.2], RR was 72.7% [95% CI: 57% to 88%], and median OS was 15.8 months [95% CI: 11.6 to NE]. 27 patients (81.8%) had decrease in disease burden (ranging from -19% to -100%). After > 18 months of follow-up, 5 patients remained in durable complete response (CR). Immune-mediated treatment related adverse events (TRAEs) included thyroid disorders (n=5; 14%), colitis (n=4; 11%), adrenal insufficiency (n=2; 5%), and type 1 diabetes (n=1). Sixteen patients (44%) experienced grade 3 or 4 TRAEs. There were no grade 5 TRAEs. Conclusions: The combination of C and O + P had acceptable safety and significant clinical activity. These promising results indicate that C and O + P merits further study as a first line option for patients with unresectable locally advanced or metastatic GE adenocarcinoma. Updated survival and correlative data will be presented. Clinical trial information: NCT03342937