Roche's KRAS data suggest potential edge over Amgen and Mirati, but the jury remains out

Roche has provided fresh evidence that its KRAS inhibitor divarasib is a threat to Amgen and Mirati. The results of an early-phase monotherapy clinical trial suggest divarasib may improve on the response rate and durability of its more advanced rivals, but the jury remains out for now. 

As Roche has pumped money into divarasib despite ceding a big head start to its rivals, management has pointed to the superior potency and selectivity seen in in vitro tests to justify investing in the asset. The company began to support its argument with more clinical data last year, presenting monotherapy results in non-small cell lung cancer (NSCLC) and colorectal cancer at events in the back half of 2022.

Now, Roche has published an updated look at the NSCLC and colorectal cancer data in The New England Journal of Medicine. The authors of the paper framed the results in the context of the data from Roche’s rivals. 

“Divarasib appears to show numerically more responses and longer progression-free survival among patients with either NSCLC or colorectal cancer than those observed with existing single-agent KRAS G12C inhibitors,” the authors wrote. “However, conclusions drawn from cross-trial comparisons must be interpreted cautiously.” 

The response rates are in line with the earlier data on divarasib. In NSCLC, Roche saw responses in 53.4% of the 60 patients enrolled in its phase 1 trial. The response rate in colorectal cancer slipped to 29.1%, mirroring the differences between the indications seen in studies of other KRAS drugs. Roche saw higher NSCLC and colorectal cancer response rates, 56.4% and 35.9%, respectively, at the 400-mg dose.

As the authors of the paper note, the response rates compare favorably to the data on KRAS inhibitors from Amgen and Mirati. Amgen linked sotorasib, sold as Lumakras, to a 37% response rate in its NSCLC phase 2 and a 10% response rate in colorectal cancer. Mirati’s adagrasib, sold as Krazati, achieved a 43% response rate in NSCLC and a 19% response rate in colorectal cancer. 

Roche appears to have an edge in terms of durability of response, too. Median progression-free survival at the 400-mg dose came in at 13.7 months in NSCLC and 6.9 months in colorectal cancer. Neither Amgen nor Mirati passed the seven-month mark in NSCLC, although Roche’s apparent edge is far smaller in colorectal cancer. 

It is too early to tell whether the numerical differences reflect a real-world advantage for Roche, but the Swiss drugmaker has seen enough to invest more in divarasib. A phase 3 trial in NSCLC is underway.