ARC-9: Etrumadenant Regimen Versus Regorafenib in Refractory mCRC

Key Takeaways

• EZFB was associated with longer survival than regorafenib in the randomized third-line comparison.
• Response and disease control findings also favored EZFB over regorafenib.
• Adverse events and discontinuation of at least one study drug were more frequent with EZFB, while no treatment-emergent deaths or unexpected safety signals were reported.

In ARC-9 Cohort B, a randomized phase II trial in third-line unresectable metastatic colorectal cancer, EZFB was associated with longer overall survival than regorafenib. Median overall survival was 19.7 months with EZFB versus 9.5 months with regorafenib, with a hazard ratio of 0.37. EZFB combined etrumadenant, zimberelimab, mFOLFOX-6, and bevacizumab, while the comparison arm received third-line regorafenib. No unexpected safety signals were reported during the study treatment period.

ARC-9 Cohort B was an international, open-label, multicenter phase II trial conducted at 29 sites in South Korea, France, Italy, Spain, the United Kingdom, and the United States. Between September 21, 2021, and September 12, 2022, 112 patients were randomized 2:1 to EZFB or regorafenib, with 75 and 37 assigned, respectively. Eligible patients had third-line unresectable mCRC after progression during or after no more than two prior lines that included oxaliplatin and irinotecan, with prior anti-VEGF(R) or anti-EGFR therapy, ECOG 0 to 1, measurable disease, and no progression within 2 months of the last oxaliplatin dose in the metastatic setting. Median age was 58 versus 60 years, and most patients had liver metastases, prior oxaliplatin exposure in the metastatic setting, and MSS or MSI-L disease. The primary endpoint was investigator-assessed progression-free survival.

At a median survival follow-up of 20.4 months, progression-free survival was 6.2 months with EZFB and 2.1 months with regorafenib. The hazard ratio for progression or death was 0.27, with a 95% CI of 0.17 to 0.43 and nominal P < 0.0001. Overall survival also favored EZFB, with a hazard ratio of 0.37, a 95% CI of 0.22 to 0.63, and nominal P = 0.0003. Confirmed response rate was 17% with EZFB and 3% with regorafenib, while disease control rate was 63% and 8%, respectively. Median duration of response was 11.5 months with EZFB and was not evaluable with regorafenib because only one patient responded.

Among 74 EZFB-treated and 35 regorafenib-treated patients, any treatment-emergent adverse event occurred in 99% and 89%, respectively. Grade 3 or higher events occurred in 82% and 49%, and discontinuation of at least one study drug occurred in 60% and 17%. Four EZFB patients, or 5%, discontinued all study treatment versus 6 regorafenib patients, or 17%. Most discontinuations of at least one study drug in the EZFB arm were primarily related to oxaliplatin, and median treatment duration was 26 versus 7 weeks. Immune-mediated adverse events were reported in 16% and 6% of patients, with grade 3 or higher events in 5% and 0%, and no treatment-emergent adverse events led to death. No unexpected safety signals were reported, although toxicity and discontinuation of at least one study drug were more frequent with EZFB.

Exploratory analyses showed longer overall survival with EZFB across liver metastases, peritoneal metastases, KRAS categories, prior therapy response, and intervals since prior oxaliplatin. Among patients with baseline liver metastases, progression-free survival was 5.7 versus 2.0 months and overall survival was 19.7 versus 9.1 months. The subgroup hazard ratios were 0.19 for progression-free survival, with a 95% CI of 0.10 to 0.35, and 0.36 for overall survival, with a 95% CI of 0.20 to 0.66. Interpretation remains limited by the open-label, small phase II design, nominal P values without multiplicity control, 57% crossover, and uncertainty about individual component contributions. The phase II results favored EZFB, while confirmation was left to later investigation.