Nivolumab combination therapies in patients with advanced gastric and gastroesophageal junction cancer: the phase II FRACTION gastric cancer study

Background: Nivolumab-based therapies are efficacious with acceptable safety in patients with gastric cancer (GC) and gastroesophageal junction cancer (GEJC). Novel nivolumab-based combination immunotherapies may offer enhanced efficacy in these indications. FRACTION-GC was a signal-seeking, randomized, open-label, phase II adaptive-design trial assessing efficacy and safety of nivolumab in combination with ipilimumab [cytotoxic T lymphocyte antigen-4 (CTLA-4) antibody], relatlimab (lymphocyte-activation gene 3 antibody), or IDO1i (BMS986205, an indoleamine-2,3-dioxygenase-1 inhibitor) in patients with unresectable, advanced/metastatic GC/GEJC. Patients and methods: Previously treated patients with GC/GEJC were randomized to receive nivolumab + ipilimumab, nivolumab + relatlimab, or nivolumab + IDO1i across two tracks: anti-programmed death-(ligand) 1/anti-CTLA-4-naïve (track 1) and -experienced (track 2). Primary endpoints were objective response rate (ORR) by investigator per RECIST v1.1, duration of response, and progression-free survival (PFS) rate at 24 weeks. Secondary endpoint was safety. Results: Eighty-one patients in track 1 and 81 in track 2 received one combination therapy. With a median follow-up of 50.2 months, ORR [95% confidence interval (CI)] by investigator for nivolumab + ipilimumab, nivolumab + relatlimab, and nivolumab + IDO1i in track 1 was 4% (0.1% to 21.9%), 5% (0.1% to 24.9%), and 13% (4.4% to 28.1%), and for track 2 was 9% (1.1% to 28.0%), 6% (0.7% to 18.7%), and 0% (0% to 15.4%), respectively. PFS rate at 24 weeks (95% CI) was 24% (11% to 39%) for nivolumab + IDO1i track 1, 17% (16% to 32%) for nivolumab + relatlimab track 2, and not estimable for other treatment arms. Grade 3/4 treatment-related adverse events were reported in 22%, 5%, and 18% of patients receiving nivolumab + ipilimumab, nivolumab + relatlimab, and nivolumab + IDO1i in track 1 and in 35%, 11%, and 18% of patients in track 2, respectively. No treatment-related deaths were reported. Conclusions: While ORR did not meet prespecified expansion criteria in any treatment arm, the safety profile of the combinations was manageable. FRACTION-GC represents a novel adaptive protocol for testing multiple combination immunotherapies.