Design: Phase 3, international, open-label randomised trial (Krascendo 1)
Population: 338 patients with previously treated KRAS G12C-mutated advanced NSCLC
Primary outcome: Progression-free survival by blinded central review
Effect: PFS 9.8 vs 5.7 months (HR 0.58, 0.44–0.76); OS 17.7 vs 12.0 months (HR 0.56)

Krascendo 1 was an international, open-label phase 3 trial in 338 patients with previously treated KRAS G12C-mutated advanced NSCLC. They received divarasib, a next-generation oral KRAS G12C inhibitor, or a first-generation inhibitor (156 sotorasib, 15 adagrasib).

Median progression-free survival by blinded central review was 9.8 against 5.7 months (HR 0.58, 95% CI 0.44 to 0.76), and overall survival 17.7 against 12.0 months (HR 0.56, 0.41 to 0.78). Confirmed response was 58.4% against 32.3%, lasting a median 11.9 against 6.9 months. Grade 3 or 4 treatment-related adverse events were less common with divarasib (24.6% vs 34.5%).

Head-to-head trials within a drug class are uncommon, and this one shows a clear gain on both efficacy and tolerability. The comparator was mostly sotorasib, so the comparison with adagrasib rests on few patients. Divarasib is investigational, and access will depend on approval.

Test for KRAS G12C in non-squamous NSCLC as part of comprehensive molecular profiling.
Divarasib improved progression-free and overall survival against first-generation inhibitors.
Grade 3–4 treatment-related toxicity was lower with divarasib than the comparators.
The adagrasib comparison rests on only 15 patients.

Why it matters

It shows that within-class improvements in targeted therapy can extend survival, not just response.

The statistics, in plain English

Both hazard ratios exclude 1 by a clear margin, so the improvements are unlikely to be chance. Open-label design can bias response assessment, which is why central blinded review of progression matters here.