BMS Built a Smarter Cancer Drug. Chemo Still Won.

So what went wrong?

Why Colorectal Cancer Keeps Outsmarting Targeted Drugs

Colorectal cancer is a uniquely stubborn opponent. In lung cancer, KRAS G12C inhibitors have shown real promise because those tumors tend to be heavily "addicted" to the KRAS mutation. Block it, and the cancer struggles.

Colorectal tumors play a different game. They have a rich network of backup signaling pathways, especially through EGFR and related receptors like HER2, FGFR, and MET. Block one route, and the cancer finds another. It's like playing whack-a-mole with molecular biology.

Adding cetuximab was supposed to shut down the main backup route (EGFR). And it helped with response rates. But tumors adapted in other ways: amplifying copies of the KRAS gene itself, developing new mutations that block the drug from binding, or switching to entirely different survival pathways like PI3K.

Researchers who tracked patient tumors over time found polyclonal resistance, meaning multiple different escape routes popping up simultaneously. KRAS G12C gene amplification was the most consistent resistance mechanism, steadily increasing over the course of treatment.

The bottom line: colorectal cancer cells are master escape artists, and blocking two doors isn't enough when the building has a dozen exits.

What This Means for BMS (and Krazati's Future)

Krazati plus cetuximab received accelerated FDA approval in June 2024 for previously treated KRAS G12C colorectal cancer. That approval was based on earlier, smaller data showing a 34% objective response rate and duration of response. But accelerated approvals come with strings attached; they require a confirmatory trial proving real-world benefit.

KRYSTAL-10 was that confirmatory trial. And it failed.

BMS acknowledged the miss but emphasized that the data "supports the clinical activity" of the combination. The company said it's "discussing the data and potential next steps with regulatory authorities." Translation: the label is now in jeopardy. The FDA could restrict it, or in theory, pull the CRC indication entirely.

Commercially, this shrinks the opportunity. CRC was already a small slice of the KRAS G12C market compared to lung cancer. Now even that slice looks uncertain.

The Competition Is Circling

The timing couldn't be worse for BMS. Amgen's sotorasib plus panitumumab (a similar KRAS G12C/EGFR combo) already showed a progression-free survival win over standard therapy in its own CRC trial, CodeBreaK 300. That gives Amgen stronger phase III evidence in this space.

Then there's Roche. Its next-generation KRAS G12C inhibitor, divarasib, posted a jaw-dropping 62.5% response rate when combined with cetuximab in early CRC data. That's nearly double what adagrasib achieved. Roche has a phase III CRC trial underway and a head-to-head NSCLC study comparing divarasib directly against sotorasib or adagrasib.

And lurking further out is Revolution Medicines, which is taking a fundamentally different approach. Instead of targeting KRAS only when it's in its inactive state (the "off" position), Revolution's drugs hit KRAS while it's active. Their pipeline includes agents for G12C, G12D (far more common in GI cancers), and even a pan-RAS inhibitor. The company plans to launch pivotal combination trials in 2026.

The Bigger Lesson for Precision Oncology

KRYSTAL-10 is a cautionary tale for the entire field. Precision oncology has delivered remarkable wins in certain cancers. But the assumption that "targeted beats chemo" isn't universal.

In colorectal cancer, the biology is messier. Co-mutations matter. The tumor microenvironment matters. Resistance pathways are redundant and adaptive. A high response rate can be a mirage if tumors quickly find workarounds.

The trial also raises questions about how the FDA handles accelerated approvals in small, genetically defined patient populations. Approving drugs based on response rates alone, then watching confirmatory trials fail, erodes confidence in the system. Expect this case to come up in future regulatory policy discussions.

For patients with KRAS G12C colorectal cancer, the combo isn't useless. A 47% response rate still matters, especially for someone who can't tolerate more chemo or needs rapid tumor shrinkage. But as a replacement for chemotherapy in the second-line setting? The data say no.

Sometimes the old workhorse still wins the race.