KRASCENDO-1 provides the first randomized phase 3 evidence that one KRAS G12C inhibitor can outperform others within the same drug class.

The study randomized 338 patients with previously treated, locally advanced or metastatic KRAS G12C-mutated non-small cell lung cancer (NSCLC) to divarasib 400 mg once daily or investigator-selected sotorasib or adagrasib.

Median progression-free survival (PFS) by blinded independent central review was 9.8 months with divarasib and 5.7 months with control. The HR for progression or death was 0.58 (95% CI, 0.44–0.76; P<0.0001). Twelve-month PFS was 44% versus 24%.

At the interim overall-survival (OS) analysis, median survival was 17.7 versus 12 months. The HR for death was 0.56 (95% CI, 0.41–0.78; P=0.0005). Twelve-month survival was 67.1% with divarasib and 49.2% with control.

Confirmed objective response was 58.4% versus 32.3%, with median response durations of 11.9 and 6.9 months.

Dr. Pasi A. Jänne, a thoracic oncologist and senior vice president for translational medicine at Dana-Farber Cancer Institute, described divarasib as a next-generation inhibitor that binds KRAS G12C more potently and selectively. Based on the efficacy and safety results, he said it “will likely be the new standard of care” in this setting.

Does This Replace Sotorasib and Adagrasib?

For a KRAS G12C inhibitor-naïve patient whose disease has progressed after standard first-line therapy, the answer is likely yes, once divarasib receives regulatory approval.

The trial demonstrated superiority rather than noninferiority, and it did so across PFS, OS, response rate, and response duration. The survival benefit is especially relevant because prior randomized trials of sotorasib and adagrasib primarily established PFS improvements over docetaxel.

However, the control group was heavily weighted toward sotorasib. Of 171 control patients, 156 received sotorasib and only 15 received adagrasib. KRASCENDO-1 therefore offers strong evidence against a predominantly sotorasib comparator, but less precision regarding a direct comparison with adagrasib.

Divarasib remains investigational. Until approval, it is available only through clinical trials, even if the data support a future preferred position.

Toxicity Requires a More Nuanced Reading

Grade 3 or 4 (high-grade symptoms) treatment-related adverse events occurred in 24.6% of patients receiving divarasib and 34.5% receiving sotorasib or adagrasib. Serious treatment-related events occurred in 7.8% and 15.2%, respectively.

Treatment-related toxicity led to discontinuation in 4.8% with divarasib and 11.7% with control. That difference occurred despite median treatment exposure of 11.3 months with divarasib and 5 months with the comparison drugs.

The overall severe-toxicity profile favored divarasib. The symptom profile did not uniformly do so.

Gastrointestinal events were more frequent with divarasib but were predominantly grade 1 (low-grade symptoms):

• Diarrhea: 66% versus 38%
• Nausea: 53% versus 20%
• Vomiting: 39% versus 8%

Hepatotoxicity was less frequent with divarasib, occurring in 26.9% compared with 39.2%. ALT elevation occurred in 15% versus 23%, and AST elevation in 14% versus 21%.

Clinicians should therefore avoid telling patients that divarasib simply causes fewer side effects. A more accurate explanation is that low-grade gastrointestinal symptoms are common, but severe toxicity, serious events, hepatic toxicity, and treatment discontinuation were less frequent.

Dr. Jänne emphasized the practical meaning of fewer severe events: patients may require fewer emergency visits and hospitalizations. Early recognition and treatment of diarrhea, nausea, and vomiting will remain important.

What the Trial Does Not Answer

Prior KRAS G12C or pan-RAS inhibitor treatment was an exclusion criterion. KRASCENDO-1 does not support switching to divarasib after progression on sotorasib or adagrasib.

Resistance may involve secondary KRAS alterations, bypass signaling, or histologic transformation. Whether divarasib can overcome selected resistance mechanisms requires separate prospective evidence.

Known untreated or active central nervous system (CNS) metastases were also excluded. Patients with treated, stable CNS disease could participate, but intracranial efficacy remains insufficiently defined. Adagrasib’s prospective CNS data may still influence individual treatment decisions until dedicated divarasib results mature.

The study was open-label, and the OS analysis remains interim. Longer follow-up will clarify the durability of the survival advantage.

Biomarker Testing Is the Immediate Practice Issue

KRAS G12C is present in approximately 12% to 14% of NSCLC. The mutation must be identified before any of these treatments can be considered.

“Everyone should be getting biomarker testing,” Dr. Jänne said. For clinicians, that means broad next-generation sequencing at diagnosis of advanced nonsquamous NSCLC, rather than waiting until first-line treatment fails.

Testing early preserves tissue, prevents delays, and identifies co-alterations that may affect prognosis or trial eligibility.

What Comes Next?

KRASCENDO-1 supports divarasib as the preferred first KRAS G12C inhibitor after progression on initial systemic therapy, pending regulatory approval.

It should not yet be extrapolated to first-line treatment. KRASCENDO-2 is separately evaluating divarasib plus pembrolizumab against pembrolizumab plus chemotherapy in previously untreated advanced disease.

What clinicians should watch next is straightforward: the FDA timeline, mature OS, CNS-specific data, outcomes separated by control drug, and resistance studies after prior G12C inhibition.

Taken together, KRASCENDO-1 shows that meaningful differences exist within the KRAS G12C inhibitor class. Potency, tolerability, and duration of target suppression appear to have translated into longer survival, not just a better response rate.