Daraxonrasib — The Drug That Finally Cracks Pancreatic Cancer's 'Undruggable' Wall
For decades, pancreatic cancer has been one of the most lethal diagnoses a person could receive. A new class of drug is changing that.
Nearly 90 percent of pancreatic cancer cases carry a mutation in the KRAS gene, a protein that scientists had long described as "undruggable." Its surface is smooth and round, with no deep pockets for a drug molecule to latch onto — a biological design that seemed to have evolved specifically to evade pharmaceuticals. For 40 years, every attempt to block KRAS in pancreatic cancer failed in clinical trials.
That changed in 2026. Daraxonrasib, a next-generation KRAS inhibitor developed by Revolution Medicines, targets a broad range of KRAS mutations — not just the rare G12C variant, but the common G12D, G12V, and G13D mutations that drive most pancreatic cancers. In a landmark Phase 3 trial presented at ASCO 2026, daraxonrasib nearly doubled overall survival for patients with metastatic pancreatic cancer, from 6.7 months under standard chemotherapy to 13.2 months — a 60 percent reduction in the risk of death.
The drug works by trapping KRAS in its inactive, GDP-bound state, preventing it from flipping the "on" switch that drives uncontrolled cell division. Because it binds to a conserved region present across multiple KRAS variants, daraxonrasib covers roughly 80 percent of all KRAS mutations found in pancreatic tumors — a breadth no previous KRAS inhibitor has achieved.
The most common side effect is a prominent skin rash, affecting more than 86 percent of patients in the trial, but the rash is manageable and rarely leads to discontinuation. Researchers at UT Southwestern, Harvard, and the Salk Institute contributed to the foundational discoveries that made the drug possible.
For the estimated 67,500 Americans who will be diagnosed with pancreatic cancer in 2026, daraxonrasib represents the first real therapeutic advance in a generation — and a proof that the "undruggable" label is no longer permanent.