IPM Take
This is what mature precision oncology should demand. Once a target has multiple medicines, another single-arm response rate is not enough. The relevant question becomes whether a new targeted drug is actually better than the targeted treatments already available.
Executive Summary
The Phase III KRASCENDO-1 trial randomised 338 patients with previously treated KRAS G12C-mutated advanced or metastatic NSCLC to divarasib or investigator-selected sotorasib/adagrasib. Median progression-free survival was 9.8 versus 5.7 months, with a hazard ratio of 0.58. Median overall survival was 17.7 versus 12.0 months, with a hazard ratio of 0.56. Confirmed response occurred in 58.4% versus 32.3%, while grade 3–4 treatment-related adverse events were less frequent with divarasib.
Why it matters
- Patients / advocates: Medicines aimed at the same mutation are not necessarily equivalent.
- Clinicians: KRAS G12C treatment may become a comparative-efficacy decision rather than simple target matching.
- HTA bodies / payers: Head-to-head evidence supports more sophisticated value assessment.
- Industry: Being first against a molecular target no longer guarantees long-term dominance.
The KRASCENDO-1 Phase III paper in the New England Journal of Medicine marks a useful maturation of the KRAS story. Sotorasib and adagrasib helped prove that KRAS G12C was druggable after decades in which RAS was treated as one of oncology’s most frustrating targets. Divarasib has now moved the question forward by being tested not against chemotherapy or historical control, but directly against those established KRAS G12C inhibitors. It improved progression-free survival, overall survival, response rate and treatment tolerability in the same trial.
That head-to-head design matters because molecular oncology eventually has to move beyond the excitement of matching a mutation to any active drug. Once several therapies exist for the same alteration, patients need comparative evidence on depth and duration of response, survival, toxicity, resistance patterns and quality of life. The mutation identifies a group of potentially relevant medicines; it does not make those medicines clinically interchangeable.
The same principle should influence reimbursement and guideline policy. Health systems often struggle when several high-cost targeted medicines sit inside the same molecular category, particularly if access decisions were made at different times using different evidence standards. A direct randomised comparison gives HTA bodies and payers something far more useful than parallel single-arm datasets. KRAS G12C has progressed from undruggable to druggable to competitive, and that is a healthier place for precision oncology to be.

