Coformulated quavonlimab and pembrolizumab (pembro) in combination with lenvatinib (lenva) as first-line (1L) therapy for patients (pts) with advanced hepatocellular carcinoma (HCC): Phase 2 KEYSTEP-004 study.

484 Background: Combinations of a PD-1/L1 inhibitor with a CTLA-4 or VEGF inhibitor have been approved as 1L treatment options for pts with advanced HCC. The multicenter, phase 2 KEYSTEP-004 study (NCT04740307) was conducted to evaluate the coformulation of quavonlimab (CTLA-4 inhibitor) and pembro (PD-1 inhibitor) in combination with lenva (multi–tyrosine kinase inhibitor) as 1L treatment for pts with advanced HCC. Methods: Pts were ≥18 y of age, had histologically or radiographically confirmed HCC with no prior systemic therapy, and were not amenable to curative therapy; had Child-Pugh class A; and ECOG PS 0 or 1. The study consisted of a safety lead-in phase (approximately 6-20 pts) to evaluate the tolerability of the recommended phase 2 dose of lenva with a possibility of dose modification based on dose-limiting toxicities (DLTs) and an efficacy expansion phase (approximately 110 pts). In the safety lead-in phase, pts received a fixed-dose coformulation of quavonlimab 25 mg and pembro 400 mg IV Q6W in combination with lenva 8 mg (body weight [BW] <60 kg) or 12 mg (BW ≥60 kg) PO QD. Allocation to the efficacy expansion phase was allowed if the doses of lenva were determined to be tolerable. Primary end points were ORR per RECIST v1.1 by BICR and safety. Secondary end points included DOR, DCR, and PFS per RECIST v1.1 by BICR and OS. Results: No DLTs were observed in the 6 pts enrolled in the safety lead-in phase. 115 pts in the safety lead-in phase or the efficacy expansion phase were treated with the fixed-dose coformulation of quavonlimab and pembro plus lenva 8 or 12 mg. 94 pts (81.7%) were male, median age was 64 y (range, 32-83), and 80 pts (69.6%) had AFP ≤400 ng/mL. As of the June 22, 2023, data cutoff date, median follow-up (time from first dose to database cutoff date or death) was 7.7 mo (range, 0.9-22.1), and 89 pts (77.4%) discontinued treatment, most commonly due to progressive disease (39.1%). Confirmed ORR was 37.4% (95% CI, 28.5-46.9), with 5 CRs (4.3%) and 38 PRs (33.0%). DCR was 79.1% (95% CI, 70.6-86.1). Median DOR was 10.6 mo (range, 1.2+ to 23.1+). Median PFS and OS were 8.2 mo (95% CI, 6.2-10.2) and 22.1 mo (95% CI, 15.5-not reached), respectively; 12-mo PFS and OS rates were 32.2% and 65.2%, respectively. Treatment-related adverse events (TRAEs) occurred in 109 pts (94.8%); most common (≥25%) were hypertension (35.7%) and diarrhea (27.8%). Grade 3-5 TRAEs occurred in 66 pts (57.4%); most common (≥5%) were hypertension (8.7%) and immune-mediated enterocolitis (7.0%). 8 pts (7.0%) died due to AEs; 2 were considered treatment related (acute respiratory failure and subarachnoid hemorrhage). Conclusions: The coformulation of quavonlimab and pembrolizumab in combination with lenvatinib demonstrated manageable safety; further investigation is warranted to understand the efficacy of this triplet as 1L treatment for advanced HCC. Clinical trial information: NCT04740307 .