KRAS was identified as one of the first human cancer genes in 1982. For nearly forty years, the field called it undruggable. Revolution Medicines inverted the approach, targeting the protein in its active state rather than its inactive one, and doubled survival in pancreatic cancer.
A drug called daraxonrasib doubled survival in patients with metastatic pancreatic cancer. The Phase 3 trial enrolled 500 patients who had failed prior chemotherapy. Median overall survival was 13.2 months versus 6.7 months for standard treatment, a hazard ratio of 0.40. The results, published in the New England Journal of Medicine, represent the largest survival benefit ever recorded in previously treated pancreatic cancer.
Revolution Medicines, the company that built daraxonrasib, traded at $34 per share a year ago. It trades near $149 after a broad selloff took ten percent off the highs. The market capitalization is roughly $30 billion. The company has zero revenue.
The gene that daraxonrasib targets, KRAS, was identified as one of the first human cancer genes in 1982. Within a decade, researchers established that KRAS mutations drive roughly 90 percent of pancreatic cancers, 34 percent of non-small cell lung cancers, and 45 percent of colorectal cancers. KRAS accounts for 90 percent of all RAS mutations, which themselves appear in about 30 percent of all human cancers. It is the single most common oncogenic driver in human biology.
For nearly four decades, every pharmaceutical company that tried to drug KRAS failed. The protein's surface is smooth and featureless. There are no obvious pockets where a small molecule can bind. By the 2000s, the consensus had hardened into a label: undruggable. Grant applications stopped mentioning direct KRAS inhibition. Careers moved to downstream targets. The most common cancer driver on earth sat unaddressed because the field agreed it could not be addressed.
The label was wrong. Not because KRAS is easy to drug, but because everyone was looking at the wrong version of the protein. KRAS toggles between two conformations: an active GTP-bound form that drives tumor growth and an inactive GDP-bound form that does not. For forty years, drug designers targeted the inactive form because that was where the structural data existed. The active form was dismissed as too transient, too slippery to bind.
Revolution Medicines inverted the approach. Their RAS(ON) platform targets the active, GTP-bound conformation. The shape change that occurs when KRAS switches on creates a binding pocket that is invisible in the off state. Daraxonrasib locks into this pocket and blocks signaling across multiple KRAS mutations simultaneously. Sotorasib, the first KRAS inhibitor approved by the FDA in 2021, works only on a single mutation called G12C. Daraxonrasib is multiselective. It works on the mutations that matter most in pancreatic cancer, where G12C is rare and G12D and G12V dominate.
The gap between 1982 and 2021 was not a hardware problem. The protein did not change. The chemistry did not suddenly improve. What changed was the question. Instead of asking how to bind KRAS when it is off, someone asked how to bind it when it is on. The answer had been there the entire time, in a conformational pocket that only exists in the state everyone had been ignoring. Undruggable was never a property of the target. It was a property of the approach.
The market is pricing daraxonrasib for pancreatic cancer. That is the Phase 3 result and the FDA pathway. Pancreatic cancer is roughly 67,500 new US cases per year. But KRAS mutations appear in 34 percent of the 229,000 annual lung cancer diagnoses and 45 percent of colorectal cancers. The addressable mutation spans three of the five most common cancers and exceeds 200,000 patients annually in the US alone.
Pancreatic cancer was the proof of concept in the hardest tissue. Blood supply is poor. Drug delivery is worst. Five-year survival has been stuck near 13 percent for decades. If daraxonrasib works there, the pharmacology should translate to tumors with better vasculature and higher drug exposure. Revolution Medicines at $30 billion prices one indication. The mutation is in three.
Sotorasib, limited to a single G12C mutation in lung cancer, generates less than $400 million in annual revenue and faces growing competition in its narrow niche. A multiselective inhibitor that works across the dominant KRAS mutations in the dominant cancer types is a different category of asset. The competitive moat is mechanism, not molecule. Targeting the active state solves the problem that targeting the inactive state could not.
The uncomfortable question is how many other targets sit behind labels that describe the approach rather than the biology. The word undruggable entered the literature, entered the grant review process, entered the training of a generation of oncologists. It became a fact about the world rather than a fact about the tools. Forty years and millions of lives separated the framing error from its correction.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)