e16036 Background: Anti-PD-1/PD-L1 therapy has changed the therapeutic landscape of ESCC. However, the efficacy of anti-PD-1 therapy alone remains modest. Cabozantinib, a multikinase inhibitor with immune modulatory effect in preclinical studies, in combination with atezolizumab has been investigated in various cancer types. We conducted a preclinical study to investigate the efficacy of cabozantinib plus anti-PD-L1 therapy using immunocompetent mouse ESCC models. Methods: From esophageal tumors developed in C57BL/6 mice following chronic exposure to 4-nitroquinoline 1-oxide, we generated several ESCC clones, which could be successfully propagated by in vitro culture for more than 25 passages. The viability of mouse ESCC clones in in vitro cell culture treated with various concentrations of cabozantinib (Adooq Bioscience LLC, Irvine, CA, US) was evaluated by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay. We used a subcutaneous mouse model by implanting 5 x 106 murine ESCC cells (clone 104) in male 5-week-old C57BL/6 mice to evaluate the in vivo therapeutic efficacy of anti-PD-L1 therapy (10F.9G2 clone [a rat anti-mouse PD-L1 antibody], Bio X Cell, Lebanon, NH, US) by intraperitoneal (ip) injection, cabozantinib by oral gavage, or the combination of both agents. Tumor volumes were measured twice per week using bilateral calipers. Following two weeks of treatment and an additional week of observation, all mice were sacrificed for tumor harvest. Tumor weights were determined and compared among different treatment groups. Results: Cabozantinib, up to 30 μM, induced only mild growth suppression (13.8 to 26.6%) of 3 mouse ESCC cell clones (clones 001, 103, and 104) in vitro. In the sc tumor model implanted with mouse ESCC clone 104, the anti-PD-L1 monotherapy (10mg/kg, twice per week) did not affect the tumor growth ( P = 0.2000), suggesting that the sc tumor of clone 104 is an anti-PD-L1 primary resistant model. On the contrary, cabozantinib (10 to 30 mg/kg/day, 5 days per week, for 2 weeks) induced a dose-dependent inhibition of tumor growth (tumor weight decreasing from 22.7 to 84.4%) in the sc tumor model of clone 104. The combination of anti-PD-L1 plus cabozantinib induced a more significant suppression of tumor growth than either single treatment alone, and cabozantinib at the dose of 20 mg/kg exhibited more pronounced enhancing effect than at the dose of 10 mg/kg. Conclusions: This preclinical study demonstrated a synergistic antitumor effect of combining anti-PD-L1 therapy with cabozantinib in immunocompetent murine models of ESCC. A phase II trial is ongoing to investigate the efficacy of cabozantinib plus atezolizumab in advanced ESCC (NCT05007613) (This work was supported by NTUCCS-112-06, NSTC112-2314-B-002-150-, NSTC112-2314-B-002-268-, MOHW111-TDU-B-221-114006, and MOHW112-TDU-B-221-124006).