— Next-generation KRAS inhibitor proves mettle versus approved agents
October 26, 2026 • 4 min read
A next-generation KRAS G12C inhibitor significantly improved survival outcomes versus first-generation drugs for previously treated advanced or metastatic non-small cell lung cancer, a phase III trial showed.
Median progression-free survival (PFS) reached 9.8 months with divarasib, as compared with 5.7 months for the group treated with sotorasib (Lumakras)opens in a new tab or window or adagrasib (Krazati)opens in a new tab or window, the two FDA-approved agents for KRAS G12C-mutant disease (HR 0.58, 95% CI 0.44-0.76, P<0.001).
And divarasib reduced the risk for death by 44% during the trial, with a median overall survival (OS) of 17.7 versus 12 months with the older drugs (HR 0.56, 95% CI 0.41-0.78, P<0.001), reported Ferdinandos Skoulidis, MD, PhD, of the University of Texas MD Anderson Cancer Center in Houston.
Divarasib "represents a new standard of care" for KRAS G12C-mutant advanced or metastatic lung cancer, Skoulidis said at the European Society for Medical Oncology (ESMO) congressopens in a new tab or window in Madrid.
The so-called Krascendo 1 trial marks the first randomized study to directly compare a next-generation KRAS G12C inhibitor head to head against the first-generation drugs. Safety of divarasib was "favorable" compared with sotorasib or adagrasib, Skoulidis said, with fewer grade 3 or higher treatment-related adverse events (TRAEs) and serious TRAEs with the novel agent.
Findings of the trial were published concurrently in the New England Journal of Medicineopens in a new tab or window.
ESMO-invited discussant Benjamin Besse, MD, PhD, of Gustave Roussy in Villejuif, France, called the results "very strong," but questioned the rationale for not allowing crossover from the control group "to the magic drug."
"It's mostly the regulators and the agencies," he observed, noting that FDA and others typically want to see a clear signal of benefit. But it means oncologists have to tell their patients that if they're assigned to the control arm, they cannot access the newer drug "because some people want to look at your survival on the old drug," said Besse.
That's not "something we can tolerate anymore," he said -- a declaration that was met with a round of applause.
Skoulidis noted that sotorasib and adagrasib -- which have been approved for previously treated KRAS-mutant lung cancers since 2021 and 2022, respectively -- have "modest" efficacy.
"Divarasib is a next-generation, inactive state, selective KRAS G12C inhibitor that was engineered to exhibit greater on-target potency and selectivity compared with sotorasib or adagrasib," he explained.
Krascendo 1opens in a new tab or window randomized 338 patients with previously treated KRAS G12C-mutant advanced or metastatic non-small cell lung cancer in a 1:1 ratio to either the divarasib arm or the control arm, where most (91%) received sotorasib. At baseline, participants had a median age of 66 years, over three-fourths were white, 61% had an Eastern Cooperative Oncology Group performance status score of 1, and 29% had a history of central nervous system involvement.
The primary endpoint was PFS, with landmark analyses showing that 43.9% of the divarasib group were alive without disease progression at 12 months versus 24% of the control group. OS rates at that point were 67.1% and 49.2%, respectively.
Confirmed responses occurred in 58.4% of the divarasib group (complete in 4.3%) and in 32.3% of the control group (complete in 3.6%), with median durations of response of 11.9 months and 6.9 months, respectively.
In terms of safety, divarasib was associated with a lower rate of grade ≥3 TRAEs compared with the control arm (24.6% vs 34.5%) and a lower rate of serious TRAEs (7.8% vs 15.2%). TRAEs leading to treatment discontinuation occurred in 4.8% of the divarasib-treated patients and 11.7% of the controls.
The most common TRAEs in the study were gastrointestinal and hepatotoxic in nature.
The percentage with hepatotoxic effects was lower with divarasib than with sotorasib or adagrasib (26.9% vs 39.2%), while the percentage with gastrointestinal toxicities -- including diarrhea, nausea, and vomiting -- was higher with divarasib (91% vs 68.4%). But "these were mostly grade 1, manageable with standard supportive measures and reversible," Skoulidis said.
In his discussion, Besse said a new method of grading toxicity is needed to "capture chronic treatment burden."
"We have the grade, but not the duration," he said. "Is the diarrhea there for 10 months or is the diarrhea here for just 2 days and then the patients have a normal life?"
Treatment-related instances of QT prolongation were reported in two of the control-arm participants, and three participants in control group had pneumonitis -- neither of those adverse events were observed in the divarasib arm.
Adverse events that led to death occurred in 6% of the divarasib group and 4.1% of the control group, but none were considered treatment-related by investigators.