吉西他滨、顺铂、白蛋白结合型紫杉醇和替雷利珠单抗（GENTIS）治疗晚期胆道癌患者的Ib/2期试验

Background: Immunotherapy with gemcitabine and cisplatin (Gem/Cis) has become the standard first-line treatment for advanced biliary tract cancer (BTC), but survival gains remain modest. Although nab-paclitaxel (nab-P) in combination with Gem/Cis showed promising efficacy in a phase II trial, it failed to demonstrate survival benefit in a subsequent phase III trial due to the toxicity associated with the high-dose regimen. Based on these findings, we designed the GENTIS trial to evaluate Gem/Cis in combination with tislelizumab (anti-PD-1) and optimized dose nab-P, aiming to enhance efficacy while maintaining tolerability in treatment-naïve patients with advanced BTC. Here, we present the phase 1b safety results of the GENTIS trial. Material and Methods: Patients with locally advanced or metastatic BTC were dosed in the GENTIS (NCT06893380). The phase 1b part followed a 3 + 3 dose-escalation design. All patients received tislelizumab 200 mg on day 1, gemcitabine 1000 mg/m², cisplatin 25 mg/m², and nab-P at 50-100 mg/m² on days 1 and 8 of a 21-day cycle. Planned dose levels of nab-P were Level −1 (50 mg/m²), Level 0 (75 mg/m²; starting dose), and Level +1 (100 mg/m²). Dose escalation proceeded from Level 0 to Level +1 if no dose-limiting toxicities (DLTs) were observed among the initial 3 patients at Level 0. DLTs were evaluated during the first cycle of study treatment. Results: Six patients were enrolled in phase 1b part: intrahepatic (50.0%), extrahepatic (16.7%), and gallbladder cancer (33.3%); most had metastatic disease (83.3%). All received at least one dose of the combination of Gem/Cis, nab-P, and tislelizumab: 3 at Level 0 (75 mg/m²) and 3 at Level +1 (100 mg/m²) of nab-P. Through the data cutoff of November 6, 2025, patients received 3-9 cycles, with all patients ongoing on treatment. No DLTs were observed. Treatment-related adverse events (TRAEs) of any grade occurred in 5 patients (83.3%), and grade 3 in 3 patients (50.0%); no grade 4-5 TRAEs were observed. The most frequent TRAEs were neutropenia in 3 patients (50.0%) and anemia, increased creatinine, and urticaria in 2 patients (33.3%). Serious adverse events occurred in one patient (16.7%; biliary tract infection), which was not treatment-related and resolved. Median follow-up was 3.8 months at the time of the data cutoff; the confirmed objective response rate was 66.7% (4 partial responses), and the disease control rate was 100% (including 2 stable diseases). As no DLTs were observed and long-term tolerability was considered, the RP2D of nab-P in combination with Gem/Cis and tislelizumab was 75 mg/m². Conclusion: The combination of Gem/Cis, tislelizumab and dose-optimized nab-P showed a manageable safety profile and promising clinical outcomes in patients with advanced BTC and is now being investigated in the phase 2 expansion at the RP2D of nab-P 75 mg/m².