Background: Immune checkpoint inhibitor (ICI) therapy has demonstrated efficacy in some patients (pts) with advanced gastric cancer (AGC) and is now approved in 1^st line combination with chemotherapy. Preclinical studies suggest interplay between angiogenesis and the immune system, with VEGF exerting effects on the anti-tumor immune response including antigen presentation, effector mechanisms and immune cell trafficking. We hypothesized that the combination of VEGF inhibition and ICI may be more beneficial than either agent alone. We designed an open-label prospective phase II trial to evaluate the efficacy of lenvatinib (Len) plus pembrolizumab (Pem) beyond first line therapy in ICI-naïve AGC. Methods: Eligible pts included advanced gastric or gastroesophageal junction adenocarcinoma regardless of PD-L1 status, with measurable disease per RECIST 1.1, ECOG ≤1, and progressive disease on ≥1 prior regimen, no prior ICI. Pts received an initial lead-in week of Len 20mg PO daily which was then continued in combination with Pem 200mg IV every 3 weeks from day 8. Therapy was continued until progression or intolerance. Imaging was performed every 6 wks for the first 12 wks and then every 9 wks. Primary end point was objective response rate (ORR). Using the Simon optimal 2 stage design, planned sample size was 29 patients (pts) with 80% power and one-sided alpha of 5%. The study would be considered positive if ≥6 pts achieved an objective response. Secondary endpoints included safety and tolerability, PFS and OS. Results: During the course of this study, ICI was approved in the 1st line setting and as a result of changing practice patterns, enrollment was prematurely suspended in 7/2022 with 24 of 29 planned pts. All enrolled pts were included for the safety analysis. All subjects were mismatch repair proficient, 15 pts (62.5%) were CPS positive (≥ 1), data was missing for 1 pt. 19 pts who completed at least 2 cycles were evaluable for response; 15 were treated in 2^nd line and 4 in 3^rd line. Five pts (21%) had a partial response, including 1 with CPS=0 and 1 with CPS=1. Median PFS and OS was 12.9 and 21.4 weeks for the entire cohort. The combination was generally tolerable without new safety signals. Dose reductions were required in 3 (13%) pts, overall compliance with len was 79% of planned doses across the study. Eleven (46%) pts experienced at least one G3 toxicity and one (4%) pt experienced G4 toxicity. Most common G3 toxicities included fatigue (27%), anorexia (18%), lymphopenia (18%), and hyperbilirubinemia (18%). Conclusions: Lenvatinib combined with pembrolizumab was well tolerated, responses were seen, but the study did not meet the primary endpoint in ICI naïve AGC. Given the changing landscape of care, ICI is now more broadly used in first line therapy, additional studies are warranted to determine the role of this combination in ICI exposed pts. Clinical trial information: NCT03321630.