Background: Biliary tract cancers (BTC) have a dismal prognosis. Recently, standard of care changed by adding durvalumab (D) to chemotherapy after the TOPAZ-1 study demonstrated an extended and durable survival benefit (median OS: 12.8 vs 11.5 months; 24-month OS rate: 23.6% vs 11.5%). Extrapolating from IMbrave150 and HIMALAYA trials in liver cancer, we combined an anti-VEGF monoclonal antibody with the STRIDE regimen (Single Tremelimumab Regular Interval Durvalumab) in advanced BTC. Methods: This is a phase II single center trial of unresectable or metastatic BTC with a safety run-in included (NCT03937830). In schedule A, the first 6 patients (pts) received single dose T 300mg with D 1,150mg flat dose + bevacizumab (B) 7.5mg/kg/dose every 21 days until disease progression or unacceptable toxicity. In schedule B, the next 8 pts did not start B until cycle 2 (C2); all other scheduling and dosing remained the same. Paired tumor tissue and blood were collected before and on treatment for correlative studies. Primary endpoint was 6-month PFS, and secondary endpoints were safety and feasibility, OS, and best ORR according to Response Evaluation Criteria (RECIST 1.1). Results: We enrolled 14 pts as of data cutoff (1/11/23) with 8 patients alive (median follow-up 4.3 mo, 0.7-17.7). Accrual is expected to be completed by June 2023 with 20 evaluable pts; interim analysis presented early due to vital unexpected results with different dosing schedules. Updated data will be presented at ASCO along with 6-month PFS and correlative studies. We treated 6 pts with schedule A and observed ORR 17% (PR x 1, SD x 4, PD x1), mPFS 8.5 mo (96% CI: 1.8 - NE), and mOS not reached (range: 3.0 - 17.7 mo). 50% of pts experienced G3 or G4 immune-related adverse events (irAE), including overlapping myositis-myocarditis-myasthenia gravis syndrome, thyroiditis, colitis, and hepatitis; these pts received systemic steroids and were taken off treatment. irAE occurred early after initiation of treatment with most presenting within 40 days. Based on the high rate of severe irAE, we changed to schedule B and treated 8 pts. We observed ORR 0% (SD x 1, PD x 5, unevaluable x 2), mPFS 2.6 mo (96% CI: 0.9 - 3.9), and mOS 4.7 mo (95% CI: 0.9 - 4.7). There were no G3 or G4 irAE; there was 1 death from tumor-related bleed that occurred prior to restaging in C2 and may have been worsened by bevacizumab. Clinical response appears to diminish with treatment schedule change, albeit small numbers limit interpretation (Cochran-Armitage trend test, p = 0.0671). Conclusions: The combination of anti-VEGF with anti-PDL1 and anti-CTLA4 appears to induce a strong immune response in advanced BTC with more severe and earlier irAE. In contrast, historical data showed limited benefit with D+T alone in BTC. We observed that modifying dosing schedule may impact outcomes by reducing efficacy when the three drugs are not given together. Clinical trial information: NCT03937830.