Pancreatic cancer's undruggable target falls, and the median moves from 6.7 to 13.2 months
The FDA cleared daraxonrasib on Aug. 26 for patients whose chemotherapy has already failed. The median gain is 6.5 months, and the price has not been disclosed.
The Scientist · Science desk
What happened
- The FDA approved daraxonrasib on Aug. 26, 2026, for metastatic pancreatic cancer.
- The label reaches only patients who have already been through chemotherapy, which makes it a second-line option.
- In the registrational trial, median overall survival was 13.2 months on the drug against 6.7 months on standard chemotherapy.
- KRAS mutations drive more than 90% of pancreatic tumors, and the protein had been written off as undruggable for decades.
- More than 86% of patients in the study developed a prominent skin rash, with mouth sores, diarrhea, nausea and vomiting also frequent.
Why it matters
- What it limits constraint Nothing in this approval moves diagnosis earlier. With no screening test and few early symptoms, the drug only ever gets a late-stage window on a patient already carrying spread disease.
- What it costs cost The benefit accrues as months on a daily oral pill, so the patients who do best generate the longest bills, and nobody outside the company can size that yet.
- What doesn't add up contradiction Two effect sizes for the same trial are in public, and the larger one is the one likely to be quoted; which is right decides how much a clinician should promise.
- What it sets up precedent A licensed drug against a pocketless target turns molecular-glue chemistry into a fundable default rather than a long shot, and makes combination programs the obvious next ask.
Next story · Invest · Oct 6
Lotte Biologics wins Alvotech biosimilar work for its Syracuse plantConfirmed2 publishers
The gain is a factor of 1.97 at the median [13], and it is a second-line number, so the clock starts after the disease has already outrun chemotherapy [2]. Put that against the baseline. Of patients diagnosed with metastatic pancreatic cancer between 2015 and 2021, about 97% were dead inside five years [8], which implies roughly 3% five-year survival for metastatic disease, against about 13% across all stages of the cancer with the worst five-year figure of any major type [9][22]. A median that crosses one year does not touch a five-year number. The tail might: Scientific American reports that some patients on the drug lived for years, which very few people with late-stage disease do [21].
Two effect sizes are circulating for the same 500-patient trial [7]. The oncologist writing in The Conversation puts the reduction in risk of death at 60% [23]. If survival were exponential, the ratio of medians would be the hazard ratio, and 6.7 divided by 13.2 gives 0.51, a 49% reduction [24]. The distance between 49% and 60% is not a rounding artifact. Either the curves separate more than the medians show, which would fit the long-survivor tail, or the two figures describe different endpoints or populations.
The mutation selection did less work than the framing implies. In the mutation-positive group the chemotherapy arm ran to 6.6 months and in the full trial to 6.7, while the daraxonrasib arm was 13.2 months in both [3][4][14]. Most enrolled patients carried the driver mutations [7], so this readout cannot show that testing identifies who benefits.
What actually got corrected is narrower than the slogan. The obstacle was a surface described as exceptionally smooth, lacking the molecular pockets a drug needs to grip [19]. Daraxonrasib never grips it: the pill binds cyclophilin A, a protein-folding helper, and that complex latches onto active KRAS and cuts the growth signal [20]. The failed assumption was about binding sites, not about the target.
Fewer patients stopped daraxonrasib for severe side effects than stopped chemotherapy, and they reported better quality of life with less pain [16], which for a population this sick is close to the whole point. Andrew Ko, a gastrointestinal oncologist at the University of California, San Francisco, told STAT it is the biggest development in pancreas cancer in decades [10]. That is a statement about sequencing options rather than cure rates. The regimen is two pills a day [5], and Revolution Medicines has not said what they cost, nor answered Scientific American's request for comment [11].
What to watch
- Publication of the hazard ratio and the full survival curves, which would settle whether the 60% death-risk reduction survives scrutiny.
- A list price for Rasonque, and whether payers or the label restrict it to patients with confirmed KRAS mutations.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence60
- Adoption
- Not scored
- Hype gap+20
- Incentives
- Not scored
- Confidence65
Perspective Coverage
3 publishers- Builder
- Builder 27%
- Operator
- Operator 58%
- Investor
- Investor 15%
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
A drug approved on Aug. 26, 2026, shuts down the key protein that drives pancreatic cancer, nearly doubling survival for patients with advanced disease; with the FDA's August 2026 approval, eligible patients can receive it for metastatic pancreatic cancer.
- [2]
The FDA approval makes the medication available to people with metastatic pancreatic cancer who have already been through chemotherapy; STAT describes it as a second-line treatment.
- [3]
Patients with advanced pancreatic cancer who received Rasonque as a second-line treatment achieved a median overall survival of 13.2 months, compared with 6.7 months for patients offered standard chemotherapy.
- [4]
In the patients with the specific cancer-driving mutations, median overall survival was 13.2 months in the daraxonrasib group versus 6.6 months in the chemotherapy group.
- [5]
The medication is called daraxonrasib, is manufactured by Revolution Medicines, will be sold as Rasonque, and takes the form of a two-pills-a-day regimen.
- [6]
The medicine is the first to attack a genetic cause of pancreatic cancer, and the approval was supported by a clinical trial that read out earlier this year.
- [7]
In the phase 3 trial, 500 patients, most of whom had specific cancer-driving gene mutations and all of whom had previously treated metastatic pancreatic cancer, were assigned to receive daraxonrasib or chemotherapy.
- [8]
For patients diagnosed with metastatic pancreatic cancer between 2015 and 2021, about 97% died within five years of diagnosis.
- [9]
Pancreatic cancer has the lowest five-year survival rate of any major cancer, at around 13 percent.
- [10]
Andrew Ko, a medical oncologist specializing in gastrointestinal cancers at the University of California, San Francisco, said: "It will be transformative in the way we treat pancreas cancer. It's the biggest development we've had in pancreas cancer in decades."
- [11]
It is unclear how much the drug will cost, and Revolution Medicines did not immediately respond to Scientific American's request for comment.
- [12]
Pancreatic cancer is deadly in part because there are no effective screening tests and it rarely causes noticeable symptoms early, so by the time signs such as jaundice or abdominal pain appear the cancer has often already spread.
- [13]
The reported medians differ by 6.5 months, a ratio of 1.97 between the daraxonrasib and chemotherapy arms.
- [14]
The daraxonrasib median is identical at 13.2 months in the mutation-positive subgroup and the full trial population, while the chemotherapy arms differ by 0.1 month.
- [15]
The most common side effect is a prominent skin rash, which affected more than 86% of patients in the study; patients also frequently had stomatitis, painful swelling and sores inside the mouth, as well as diarrhea, nausea and vomiting.
- [16]
Patients taking daraxonrasib were far less likely to stop treatment due to severe side effects compared with chemotherapy, and they had improved quality of life with reduced pain.
- [17]
More than 90% of pancreatic tumors are driven by mutations in the KRAS gene, whose protein acts as a growth switch stuck in the on position when mutated.
- [18]
For decades, scientists considered KRAS to be undruggable, and successfully targeting the central mechanism behind most pancreatic cancers was considered impossible.
- [19]
The surface of the KRAS protein is exceptionally smooth, lacking the molecular pockets that standard drugs require to bind to it and turn the switch off.
- [20]
Daraxonrasib is taken daily by mouth and, instead of binding KRAS directly, attaches to cyclophilin A, a molecule that helps fold proteins; the resulting complex binds active KRAS and shuts down its ability to signal cancer cells to multiply.
- [21]
Some patients who received the drug lived for years, something very few people with late-stage pancreatic cancer do.
- [22]
A 97% five-year mortality rate implies roughly 3% five-year survival for metastatic pancreatic cancer, against about 13% for pancreatic cancer across all stages.
- [23]
Daraxonrasib reduced the risk of death for metastatic pancreatic cancer patients by 60%, according to the gastrointestinal oncologist writing in The Conversation.
- [24]
Under an exponential survival assumption, the ratio of medians implies a hazard ratio of 0.51, a 49% reduction in risk of death, below the 60% figure reported in The Conversation.
- [25]
The oncologist author expects an increase in clinical trials of combination therapies pairing KRAS inhibitors with other drugs to prevent resistance, plus numerous next-generation inhibitors in development.
Sources
3 independent publishers whose own reporting we read for this story.
- scientificamerican.comU.S. approves breakthrough pancreatic cancer drug, heralding a new era for an intractable disease
1 article · August 26, 2026
- statnews.comSTAT+: FDA approves new pancreatic cancer drug expected to usher in new era of treatment
3 articles · August 27, 2026
- theconversation.comFDA approves breakthrough pancreatic cancer drug – an oncologist explains how daraxonrasib nearly doubles survival
1 article · August 26, 2026
Topics and entities
Topics
- Pancreatic Cancer
- FDA Drug Approvals