First-line iparomlimab and tuvonralimab (QL1706) or iparomlimab (QL1604) + bevacizumab (BEV) for unresectable hepatocellular carcinoma (HCC): Updated results from the phase Ib/II DUBHE-H-106 study.

579 Background: Novel treatment options for unresectable HCC are needed. Iparomlimab and tuvonralimab are anti-PD-1 and anti-CTLA-4 antibodies, respectively. The DUBHE-H-106 study aims to assess safety and efficacy of first-line QL1706 or QL1604 + BEV for HCC. Preliminary data have been reported on 2023 ASCO Annual Meeting. Here, we report updated results. Methods: This study consists of three cohorts. Systemic therapy-naive adult patients (pts) with HCC, ≥ one measurable untreated lesion per RECIST v1.1, BCLC stage B–C, Child-Pugh score ≤ 7, not amenable to or progression after locoregional therapy, ECOG performance status of 0–1 were eligible. In Cohort A, six pts received QL1706 5 mg/kg + BEV 15 mg/kg Q3W. If ≤ two pts had dose-limiting toxicities (DLT), another six pts would be enrolled. If ≤ three of twelve pts had DLT, the safe dose of BEV would be determined, and eight more pts would be enrolled. Otherwise, enrollment of another dose group (QL1706 5 mg/kg + BEV 7.5 mg/kg Q3W) would initiate, using the same procedure. If number of DLT exceeded the criteria, further dose reduction of BEV or study termination would be discussed. Then 40–60 pts were randomized 1:1 to Cohort A or B. In Cohort B, pts received QL1604 200 mg + BEV (safe dose) Q3W. Enrollment of Cohort C would initiate according to the preliminary results of Cohort A and B. Pts received QL1706 7.5 mg/kg + BEV (safe dose) Q3W, using the same procedure in Cohort A. If ≤ three of twelve pts had DLT, 8–28 more pts would be enrolled. Results: Between Jun 2021 and Dec 2023, Cohort A, B, and C included 50, 26, and 40 pts, respectively. Baseline data were balanced in each cohort. All pts were in the safety set. No DLT was reported. Incidences of adverse events (AE) were similar in three cohorts. Efficacy evaluable set included 47, 26, and 37 pts in three cohorts. Numerically, QL1706 + BEV showed better efficacy compared to QL1604 + BEV, and higher response and 12-month progression-free survival (PFS) rates were found in Cohort C vs Cohort A. Detailed results were shown in Table. Conclusions: First-line QL1706 or QL1604 + BEV showed acceptable toxicities and promising efficacy for unresectable HCC. QL1706 7.5 mg/kg + BEV 15 mg/kg Q3W may have better anti-tumor activity and were recommended for trials in future. Clinical trial information: NCT05603039 . Endpoints Cohort A Cohort B Cohort C Treatment-related AE (TRAE) 43 (86%) 25 (96%) 37 (92%) Grade ≥ 3 TRAE 24 (48%) 14 (54%) 16 (40%) Serious TRAE 11 (22%) 9 (35%) 10 (25%) Immune-related AE 27 (54%) 9 (35%) 22 (55%) TRAE leading to dose interruption 23 (46%) 16 (62%) 16 (40%) Objective response, n (%; 95% CI) 18 (38%; 25%–54%) 6 (23%; 9%–44%) 16 (43%; 27%–61%) Disease control, n (%; 95% CI) 35 (74%; 60%–86%) 18 (69%; 48%–86%) 30 (81%; 65%–92%) Median PFS (95% CI), months 7.0 (3.1–9.6) 5.4 (2.4–11.0) 7.0 (4.2–not evaluable) 12-month PFS rate (95% CI) 26.8 (14.7–40.4) 24.4 (9.9–42.1) 40.9 (24.3–56.9)