A room full of cancer doctors rose to their feet at the 2026 American Society of Clinical Oncology (ASCO) conference, applauding the full results of a new cancer trial [1][2] — one with the potential to open a new treatment strategy for as many as 20% of all cancers.
The trial involved a novel treatment for pancreatic cancer, one of the deadliest cancers there is. It is the only major cancer where fewer than one in five people are still alive five years after diagnosis [3].
And here is the remarkable part. The target this drug goes after was, for forty years, written off as impossible to hit — a gene so slippery that researchers gave it a nickname: undruggable [4]. Yet that same gene drives about nine in ten pancreatic cancers [5], and it belongs to a family of genes, called RAS, that is mutated in nearly one in five cancers of any kind [6]. This previously "undruggable" pathway sits behind a huge share of human cancer.
Table of Contents
- The Gene No Drug Could Touch
- A Glimpse of Hope — And the Wrong Door
- A New Strategy, and the Verdict
- What This Breakthrough Really Means
- References
The Gene No Drug Could Touch
If the stakes are this high, why did it take forty years to find a drug? This pathway was first identified all the way back in 1982. Scientists studying cells from a human bladder tumour found something that rewrote cancer biology: a single typo in the DNA — one wrong letter out of three billion — was enough to turn a normal gene into one that drives cancer [7]. That gene belonged to a small family called RAS [4]. It encodes a class of proteins that control cell growth, division, and survival [8].
One member of that family, KRAS, is the engine of pancreatic cancer — switched on in more than nine out of ten cases [5]. And pancreatic cancer is about as brutal as cancer gets: only around thirteen percent of patients are alive five years after diagnosis [3].
There is a fundamental problem when it comes to designing targeted therapies for KRAS. Its protein is almost perfectly smooth — like a featureless marble. There is no pocket on its surface, no keyhole for a drug to fit into [9]. So for decades, the attempts failed, and the field hung that label on it: undruggable [4].
So how do you drug a protein with no keyhole?
A Glimpse of Hope — And the Wrong Door
The first crack in the "undruggable" reputation came in 2013, from a lab that refused to accept the surface was truly smooth. A scientist named Kevan Shokat and colleagues found that one specific version of the mutation — called G12C — briefly opens up a tiny pocket that had never appeared in any previous picture of the protein [9]. Suddenly, there was a keyhole. Chemists designed drugs that slip into that pocket and lock the protein in its "off" state [9].
The first of these drugs, sotorasib, was approved in 2021 — the first approved drug to directly target the G12C version of the KRAS mutation [10]. A second, adagrasib, followed the next year [11]. For lung cancer, where the G12C version appears more frequently, these represented genuine breakthroughs.
But there is a crucial limitation. That pocket only opens on the G12C version. And pancreatic cancer almost never has G12C — it is overwhelmingly driven by a different mutation, called G12D. Among KRAS-mutated pancreatic tumours, G12C accounts for just one percent [8]. Researchers had cut a key — but for a door the cancer that needed it most almost never has.
A New Strategy, and the Verdict
The idea that changed everything — and earned the applause at ASCO — was to stop trying to pick each individual mutation's lock. Instead of forcing a drug onto the smooth surface of the KRAS protein, scientists approached the problem from an entirely different direction. The new drug grabs a protein the cell already makes in abundance — called cyclophilin A — and binds to it first, forming a new molecule. That new molecule then clamps onto RAS proteins while they are switched on, shutting down the signal they would otherwise send for cells to keep multiplying [5].
And because it works through that borrowed protein instead of a mutation-specific pocket, it is not fussy about which version of the mutation it faces. It shuts down G12D — the dominant pancreatic version — along with the others, and even normal RAS that is stuck in the "on" state [5].
The drug is called daraxonrasib. The trial that put it to the test was the one that received the standing ovation at ASCO 2026. It enrolled five hundred patients across six countries, all with metastatic pancreatic cancer that had already progressed after chemotherapy — about as difficult a setting as oncology has. Half received daraxonrasib as an oral pill; half received further chemotherapy, the current standard of care [5].
The primary endpoint was overall survival: how long did patients with this deadly cancer live? Those in the chemotherapy group lived a median of 6.6 months. The patients on daraxonrasib lived a median of 13.2 months [5].
An additional six months of life may not sound dramatic in isolation — but the oncology community's response reflects what this trial represents. It demonstrates that researchers can drug the active state of RAS itself — proof of concept that this gene, written off for four decades, can finally be effectively targeted [5]. That is why the room stood up.
What This Breakthrough Really Means
KRAS is not only a pancreatic gene. The same gene drives about half of colorectal cancers and roughly a third of lung cancers [8]. And the RAS gene family, of which KRAS is a part, stands behind nearly 20% of all cancers [6]. A drug that can reach the active state of RAS, regardless of the exact mutation, could matter far beyond the pancreas — and daraxonrasib is already being tested in lung cancer [12].
It is also being combined with other therapies, and this is where the science becomes particularly compelling. Daraxonrasib does not have to work alone. A company called Tango Therapeutics recently reported the first results of pairing it with a second drug that attacks the cancer through a completely different vulnerability: many pancreatic tumours are missing a gene called MTAP, and that loss leaves them dependent on another protein to survive — so researchers block that one as well. Two independent attacks at once. In a small, early trial of heavily pre-treated patients, the combination shrank tumours in 11 of 12 people, and nine in ten were still progression-free at six months [13]. The data are very early, but the signal is striking.
For a gene that spent forty years labelled undruggable — one that sits behind roughly one in five cancers — these results mark a turning point. The dominoes are beginning to fall.
References
1. https://x.com/DrSamuelBHume/status/2061225858384248845/video/1
4. https://www.nature.com/articles/s41392-021-00780-4
5. https://www.nejm.org/doi/full/10.1056/NEJMoa2605555
6. https://pmc.ncbi.nlm.nih.gov/articles/PMC7367715/
7. https://www.nature.com/articles/300149a0
8. https://www.cancerbiomed.org/content/22/7/762
9. https://www.nature.com/articles/nature12796
