Background
Phase 3, international, open-label, randomized. N=338 treated patients with previously treated KRAS G12C-mutated advanced/metastatic NSCLC. Divarasib is a next-generation oral KRAS G12C inhibitor, more potent and selective in vitro than sotorasib and adagrasib.
Results
Interventions and follow up: Arm A: divarasib 400 mg PO daily (n=167)
Arm B: sotorasib 960 mg daily (n=156) or adagrasib 600 mg BID (n=15) (n=171)
Primary endpoint: PFS by BICR
Secondary endpoints: OS, confirmed ORR, DOR, PROs, safety
mFollow up: 14.3mo
Results: PFS: 9.8 vs 5.7mo, HR 0.58 (95% CI 0.44–0.76), P<.001
OS: 17.7 vs 12.0mo, HR 0.56 (95% CI 0.41–0.78), P<.001
Confirmed ORR (n=328 evaluable): 58.4% vs 32.3%
Median DOR: 11.9 vs 6.9mo
Arm B: sotorasib 960 mg daily (n=156) or adagrasib 600 mg BID (n=15) (n=171)
Primary endpoint: PFS by BICR
Secondary endpoints: OS, confirmed ORR, DOR, PROs, safety
mFollow up: 14.3mo
Results: PFS: 9.8 vs 5.7mo, HR 0.58 (95% CI 0.44–0.76), P<.001
OS: 17.7 vs 12.0mo, HR 0.56 (95% CI 0.41–0.78), P<.001
Confirmed ORR (n=328 evaluable): 58.4% vs 32.3%
Median DOR: 11.9 vs 6.9mo
Adverse events
Grade 3–4 treatment-related AEs: 24.6% vs 34.5%
Specific toxicities: NR in abstract
Discontinuation: NR in abstract
Specific toxicities: NR in abstract
Discontinuation: NR in abstract
Conclusions
Divarasib significantly improved PFS and OS vs sotorasib/adagrasib in previously treated KRAS G12C NSCLC, with higher response rates and fewer grade 3–4 treatment-related AEs. First randomized evidence that a next-generation KRAS G12C inhibitor outperforms first-generation agents.
Key Limitations
Open-label; control arm was 92% sotorasib, so the comparison is mostly vs sotorasib rather than adagrasib. Median follow-up 14.3mo; OS is secondary and possibly confounded by subsequent therapy. Not compared against docetaxel or chemoimmunotherapy-era standards. Broader safety and PRO detail NR in abstract.
Clinical Context
Context: sotorasib and adagrasib are approved for previously treated KRAS G12C NSCLC but with modest PFS (~5–6mo), and CodeBreaK 200 showed no OS benefit vs docetaxel. Divarasib would raise the bar for this class. Regulatory status of divarasib is not claimed here.
