Background: VEGF blockade coupled with cytotoxic chemotherapy can promote an immune-permissive tumor microenvironment that augments response to PD-L1 inhibition. IMbrave151 (NCT04677504) is a randomized, double-blinded, global proof-of-concept Phase II study evaluating atezolizumab (atezo), bevacizumab (bev) and cisplatin and gemcitabine (CisGem) as first-line treatment for advanced biliary tract cancer (aBTC). Here we report updated clinical data and molecular correlates of response and resistance. Methods: Patients (pts) with previously untreated aBTC were randomized 1:1 to receive atezo (1200 mg every 3 weeks [q3w]) + bev (15 mg/kg q3w) or placebo (plb) + CisGem (Cis 25 mg/m² and Gem 1000 mg/m² on Days 1 and 8 q3w) for up to 8 cycles, followed by atezo (1200 mg q3w) + bev (15 mg/kg q3w) or plb until disease progression or unacceptable toxicity. The primary endpoint was progression-free survival (PFS). Secondary endpoints included overall survival (OS), objective response rate (ORR), duration of response (DOR), and safety. The final OS analysis was done when ≈90 deaths occurred or when all pts had ≥2 years of follow-up. Transcriptome analysis (n=98) and mutation profiling (n=103) were done on baseline tumor samples. This was a signal-seeking trial to estimate the treatment effect in each arm with no formal hypothesis testing. Results: In total, 162 pts were randomized to either atezo + bev + CisGem (atezo/bev; n=79) or atezo + plb + CisGem (atezo/plb; n=83). The updated PFS HR was 0.67 (95% CI: 0.46, 0.95) in favor of atezo/bev. Updated median PFS was 8.35 mo for atezo/bev and 7.9 mo for atezo/plb, with 6-mo rates of 78% and 63%, respectively. The updated OS HR was 0.97 (95% CI: 0.64, 1.47), with a median of 14.9 mo for atezo/bev and 14.6 mo for atezo/plb. Confirmed ORR was 26.6% for atezo/bev and 26.5% for atezo/plb, with median DORs of 10.28 (95% CI: 6.7, 16.7) and 6.18 mo (95% CI: 4.3, 6.7), respectively. The incidence of Grade 3/4 adverse events was 73% with atezo/bev and 74% with atezo/plb. High VEGFA gene expression was associated with an improved PFS (HR, 0.43; 95% CI: 0.22, 0.83) and OS (HR, 0.66; 95% CI: 0.31, 1.4) in favor of atezo/bev. Hepatocytes high gene signature also was associated with an improved PFS (HR, 0.47; 95% CI: 0.24, 0.92) and OS (HR, 0.66; 95% CI: 0.31, 1.4) in favor of atezo/bev. Pts with PI3k/AKT pathway mutations appeared to have worse OS than pts without mutations (HR, 3.7; 95% CI: 1.5, 9.1) with atezo/bev. Conclusions: Final analysis of the IMbrave151 study indicates a modest PFS benefit in intent-to-treat pts combining atezo with bev and chemotherapy. The trial is limited by small numbers and a non-comparative design. Exploratory analysis of correlative biomarkers suggests that high VEGF-A gene expression and hepatocytes high gene signature may be predictive markers of benefit with atezo/bev, warranting further investigation. Clinical trial information: NCT04677504.