Oncology · Medical Breakthrough · May 2026
How an 'Impossible' Idea Led to a Pancreatic Cancer Breakthrough
A drug once dismissed as fantasy is nearing regulatory approval — and it could be the most significant advance in cancer treatment in fifteen years. The new strategy also holds promise for lung and colon tumors. Here's how scientists discovered it.
The Deadliest of Diagnoses — and Why Decades of Hope Failed
Pancreatic cancer is, in the blunt arithmetic of oncology, one of the most devastating diagnoses a physician can deliver. A gland tucked deep in the abdomen — largely invisible to standard screenings and rarely symptomatic until it is far too advanced — the pancreas offers cancer an almost perfect hiding place. By the time most patients receive a diagnosis, the disease has already spread. And once it has metastasized, the numbers are unsparing: only three percent of those patients are alive five years later.
The statistics have barely shifted in decades. While other cancers — breast, prostate, certain leukemias — have seen dramatic improvements in survival rates, pancreatic cancer has remained stubbornly resistant. More than 50,000 Americans die from it each year, making it the third-leading cause of cancer death in the United States, trailing only lung and colon cancer. Worldwide, the toll runs to more than half a million people annually, a number that has been climbing as populations age.
The standard treatment has long been chemotherapy: a grueling regimen that often causes severe side effects and, in the case of metastatic disease, typically extends life by only a matter of weeks rather than months. Surgeons can sometimes remove early-stage tumors, but fewer than one in five patients are eligible for surgery when first diagnosed. The rest are left with options that offer little more than a slowing of the inevitable.
This is not for lack of effort. For decades, academic researchers and pharmaceutical companies alike have poured enormous resources into defeating pancreatic cancer. They ran clinical trial after clinical trial. They tested promising compounds that had worked against other malignancies. Nearly every time, the trials failed. Drugs that had shown extraordinary promise in laboratory models did almost nothing in human patients. The biology of the disease, it seemed, was simply too hostile, too alien, too armored against conventional attack.
Many in the field began to believe the disease might be fundamentally unconquerable — at least until the underlying molecular machinery could be understood and disabled in a wholly different way. The question was whether anyone would ever find that way. For a long time, the honest answer was: probably not. The obstacles appeared not merely difficult, but categorically impossible.
"The possibility is unlike, really, anything we've seen in pancreatic cancer in many years."
The gland's anatomy compounds the challenge. When a tumor develops in the pancreas, it is typically surrounded by dense scar tissue — a fibrous barrier that blocks chemotherapy drugs from reaching cancer cells in useful concentrations. The tumor microenvironment is deeply immunosuppressive, meaning the body's natural defenses are effectively silenced. And the cancer's genetic profile is so chaotic and heterogeneous that targeted therapies, which have worked brilliantly against mutations in breast and lung cancer, rarely find a clean handle to grip.
And yet, as of May 2026, all of that has changed. A drug called daraxonrasib, developed by a small Silicon Valley company called Revolution Medicines, is nearing regulatory approval by the Food and Drug Administration. In a late-stage clinical trial, patients who received the drug lived a median of over 13 months — compared with fewer than seven months for patients who received chemotherapy. It is the first drug in history to substantially extend the lives of pancreatic cancer patients. And it did so by attacking a target that the entire field of oncology had spent thirty years declaring impossible to hit.
The Greasy Ball — Understanding the KRAS Protein and Why It Seemed Unbeatable
To understand why daraxonrasib represents such a radical departure from everything that came before, it helps to understand what it targets — and why targeting it was supposed to be impossible.
At the heart of nearly all pancreatic cancers — as well as a substantial proportion of lung and colon cancers — lies a mutation in a gene called KRAS. This gene, discovered in the early 1980s by researchers at M.I.T. and Harvard, belongs to a family of genes called RAS genes, which were among the first human cancer-causing genes ever identified. The KRAS gene's normal job is to help cells regulate their growth. It directs the production of proteins — also called KRAS proteins — that act as molecular switches: turned on when a cell needs to divide, then switched off when the division is complete.
In cancer, however, the mutations in KRAS are oncogenic — they leave the switch permanently jammed in the "on" position. The protein relentlessly signals cells to grow and divide, regardless of whether such growth is needed or appropriate. The result is a tumor that proliferates without restraint, consuming resources, invading adjacent tissues, and eventually spreading throughout the body. KRAS mutations are present in roughly 90 percent of pancreatic cancers, and in large proportions of lung adenocarcinoma and colorectal cancers as well.
So scientists knew the target. They knew it was there. They knew it was driving the cancer. The problem was that they couldn't seem to touch it.
The KRAS protein has a structure unlike many drug targets. Most proteins that drugs successfully disable have pockets, grooves, or clefts in their surfaces — physical features where a drug molecule can dock, anchor, and interfere with the protein's function. The KRAS protein, by contrast, appeared almost featureless: smooth, globular, slick. Scientists often referred to it as "the greasy ball." A drug that needed to bind to its surface had nowhere obvious to grab.
Why KRAS was called "undruggable"
Most cancer drugs work by docking into surface pockets of their target protein — like a key entering a lock. The KRAS protein appeared to have no such pockets. It was small, smooth, and featureless, with a near-perfect binding affinity for GTP (guanosine triphosphate) that made competitive inhibition nearly impossible. Drug companies spent billions across the 1980s and 1990s trying to find workarounds, all without success. By the mid-2000s, a broad consensus had formed: KRAS was simply undruggable.
The discovery of the RAS family of oncogenes in 1982 had triggered enormous excitement in the pharmaceutical industry. If cancer was being driven by a specific mutated protein, surely blocking that protein would stop the cancer. Drug companies launched massive research programs. Academic laboratories around the world oriented their work around RAS. Enormous sums of money were committed. And then, one by one, the programs failed.
"Everyone ran away from KRAS," said Channing Der, a pioneering KRAS researcher now at the University of North Carolina. "Very prominent members of the field argued this is lunacy, that this is crazy." The protein became something of a cautionary tale in oncology — proof that identifying a cancer driver was not the same as finding a way to stop it. By the early 2000s, most pharmaceutical companies had quietly abandoned their KRAS programs. Funding dried up. Talented scientists moved to more tractable targets. The field learned to work around KRAS rather than against it.
What remained was a small community of believers — researchers who continued to poke at the problem not because anyone expected them to succeed, but because they believed the problem was too important to abandon. Among them was Kevan Shokat, a biochemist at the University of California, San Francisco, whose work would eventually crack the problem open.
Cracking the Code — The Science of Molecular Glues and the Road to a Drug
Kevan Shokat was not convinced that KRAS was impenetrable. Where others saw a featureless surface, he saw a hypothesis: that perhaps the greasy ball wasn't quite as smooth as it appeared, and that the right molecule, applied in the right way, might find a foothold. He spent five years systematically screening 500 different molecules, looking for anything that might wedge into the KRAS protein's surface.
The screening was painstaking, dispiriting work — the kind of research that produces reams of negative results before any positive signal appears. But in 2013, Shokat found what he was looking for: a small molecule that could wedge into a previously unnoticed crevice in the KRAS protein. The crevice was subtle, almost invisible in standard structural analyses. It hadn't shown up in previous drug discovery efforts because researchers had been looking for the wrong thing — large, obvious pockets rather than the small, transient grooves that sometimes form in a protein's surface as it flexes and moves.
The molecule Shokat found didn't immediately become a drug. It was more of a proof of concept — a demonstration that the greasy ball could be touched after all, that the "undruggable" label was a verdict of the available tools rather than a verdict of nature. His landmark finding, published in the journal Nature in 2013, re-energized the field. Other researchers who had given up on KRAS began returning to the problem. Drug companies that had quietly shut down their programs began reopening them. A new wave of creativity swept through the field.
Around the same time, Greg Verdine, a scientist at Harvard, was pursuing a parallel avenue of investigation. Verdine was interested not just in KRAS, but in a broader question: how do you design drugs against proteins that conventional chemistry cannot reach? He was particularly fascinated by molecules found in nature that seemed to perform chemical operations that synthetic chemists couldn't replicate. Among these were what he called "molecular glues" — naturally occurring compounds that could stick two proteins together that would never normally interact.
Verdine's insight was that a molecular glue could be engineered specifically to disable KRAS. Rather than trying to directly block the protein — which was so difficult — what if a drug could be designed to stick to a nearby protein and then use the combined, enlarged surface area to wrap around KRAS and switch it off? It was a conceptual leap: instead of a key going into a lock, the approach was more like building a new lock around the key.
"Every time there was an advance, it led to another dumping of dogma and finding out that what everybody assumed was true was actually not true."
At a company he founded called Warp Drive Bio, Verdine and his team developed a strategy centered on cyclophilin — a protein naturally present in human cells. The drug would first bind to cyclophilin, then use the combined surface to envelop and disable KRAS before drifting away and moving on to attack another KRAS protein. The mechanism was novel, elegant, and deeply counterintuitive: a drug that didn't stay attached to its target, but acted more like a catalyst, working in repeated cycles.
In 2018, Revolution Medicines — a small Silicon Valley company that had originally focused on anti-infective drugs — acquired Warp Drive Bio and doubled down on its KRAS platform. Revolution's chemists then took a decision that made company leaders nervous: their compound would target KRAS proteins not only in cancerous cells, but in healthy cells too, switching the protein from its "on" state to "off" regardless of whether the cell was malignant. This was a calculated gamble. Similar approaches in animal experiments had been lethal — killing mice because the drug disrupted normal cellular signaling throughout the body.
But the animal experiments with daraxonrasib told a different story. "We started shrinking tumors in animals, and seeing that the animals seemed to be just fine," said Dr. Mark Goldsmith, Revolution's chief executive. The drug had somehow managed to devastate cancer cells while mostly sparing normal tissue — a distinction that should not, by existing theory, have been possible. Scientists are still working to fully understand why, but the leading hypothesis involves differences in how heavily cancer cells depend on KRAS signaling compared with normal cells. Cancer cells that have been driven for years by a permanently active KRAS mutation become so reliant on that signal that its removal is catastrophic. Normal cells, which use KRAS only intermittently, can tolerate its temporary suppression.
From Laboratory to Patient — The Human Stories Behind the Breakthrough
The journey from a compelling laboratory result to a treatment that actually works in human beings is always uncertain — and in oncology, it is littered with the wreckage of drugs that looked miraculous in mice and failed catastrophically in people. When Revolution Medicines began its first human safety trial in 2022, the scientists and physicians involved were cautiously optimistic at best. They had every reason to worry.
The patients enrolled in that first trial were those with advanced cancers who had exhausted other options. They were, in the blunt language of clinical medicine, running out of time. What the researchers observed in those first months was startling. Tumors began to shrink. The side effects — rash, diarrhea, fatigue, nausea, raw and split fingertips — were real and sometimes significant, but they were manageable. Nobody was dying from the drug itself. "We started seeing shrinking tumors, and side effects that were manageable," Goldsmith recalled.
Dr. Anirban Maitra, a pancreatic cancer specialist now at New York University, attended a medical conference where early daraxonrasib trial data were presented. What he heard stopped him cold. The drug was blocking KRAS in both cancerous and normal cells — and yet patients were not experiencing the organ damage that every model had predicted. "How is this possible?" he recalled thinking. "How are these patients not dying?" The question was rhetorical, but it pointed to something fundamental: the drug was doing something that the field's theoretical framework had insisted it could not do.
For patients themselves, the results were not theoretical. Rhea Caras, a retired lawyer from Palos Verdes Estates, California, was diagnosed with metastatic pancreatic cancer in early 2023 and told she most likely had months to live. After a first round of grueling chemotherapy, her oncologist told her about an experimental drug he was excited about — data he had seen presented at a conference in Europe. She enrolled in a mid-stage daraxonrasib trial. More than two years later, she is still taking her pills every day.
"I'm pretty sure I would not be alive still but for this drug," she said. "I'm living a pretty good life, and I did not expect that." Now 67, she deals with side effects — fatigue, nausea, digestive problems — but her cancer has shrunk. She is planning a trip to Hawaii with her family. The horizon, which not long ago seemed to end weeks away, has receded into years. "I think I could die of something else," she said.
Vicky Stinson, another patient on the drug, hiked a seven-mile mountain trail in Colorado in October 2025 despite her Stage III diagnosis. She had been warned she had "months — not years — to live." She chose not to accept that prognosis. On daraxonrasib, her cancer showed no growth for 13 months. When the drug eventually stopped working for her and she was forced back onto chemotherapy, she described the transition not just medically but emotionally — the pill had been "so easy" compared with what came before.
The pill is not a cure, and the scientists who developed it are the first to say so. Eventually, for most patients, daraxonrasib stops working. Cancer cells find ways to evolve around the block, developing new mutations that allow them to grow despite the drug's presence. Some patients do not respond at all. The side effects are real and for some patients severe. But in a disease that previously offered so little, the shift from weeks to months — from months to years — is not incremental. It is transformational.
"It's the beginning, not the end."
The drug has been fast-tracked for review by the FDA, and Revolution has received agency approval to offer expanded access to patients while formal approval is pending. Dozens of other companies, energized by Revolution's results, have launched their own KRAS-targeting programs. There are now more than 50 similar drugs in various stages of clinical testing, targeting KRAS mutations not only in pancreatic cancer but in lung adenocarcinoma and colorectal cancer as well — together the three leading causes of cancer death worldwide.
Beyond the Pancreas — Implications for Lung, Colon, and the Future of Cancer Medicine
The significance of the daraxonrasib breakthrough extends far beyond pancreatic cancer, important as that achievement is. KRAS is not a target unique to a single malignancy. It is, in fact, one of the most commonly mutated oncogenes in all of human cancer — present in roughly 30 percent of all cancers, with particularly high rates in the three deadliest tumor types.
In pancreatic ductal adenocarcinoma, which accounts for roughly 90 percent of all pancreatic cancers, KRAS mutations are found in approximately 90 percent of cases. But KRAS mutations also drive roughly 30 percent of lung adenocarcinomas and between 40 and 45 percent of colorectal cancers. Lung cancer kills more Americans than any other malignancy — more than 130,000 people per year. Colorectal cancer kills approximately 50,000. Together with pancreatic cancer, these three diseases account for somewhere between a quarter and a third of all cancer deaths in the United States.
The molecular glue mechanism developed for daraxonrasib, which targets KRAS in its active "on" state through an intermediary protein, is broadly applicable. The same mechanism that shuts down a KRAS mutation in a pancreatic tumor cell is capable of doing so in a lung tumor cell or a colon tumor cell. Several trials are already underway testing daraxonrasib and similar compounds specifically in patients with KRAS-mutant lung and colon cancers. Early results have been encouraging.
The broader implications extend beyond these three tumor types. Because the KRAS targeting mechanism is based on a novel principle — using molecular glues to engineer protein-protein interactions that don't exist in nature — it opens up an entirely new toolkit for drug discovery. Proteins that were previously considered undruggable because they lacked obvious surface pockets are no longer categorically off-limits. The approach could potentially be adapted to target other challenging proteins across a wide range of diseases, not only cancer.
Some scientists are already comparing the moment to the arrival of imatinib — the drug sold as Gleevec — in 2001, which transformed chronic myelogenous leukemia from a frequently fatal disease into one that most patients could live with for decades. Others draw comparisons to the introduction of checkpoint immunotherapy drugs in the early 2010s, which opened up long-term remission as a realistic outcome in some previously lethal cancers. Now, some predict that the KRAS approach could be comparably significant — perhaps the most important advance in cancer treatment in fifteen years.
The scale of the opportunity
KRAS mutations are found in an estimated 30% of all human cancers. They are the most common oncogenic driver in three of the four leading causes of cancer death worldwide. In the United States alone, roughly 250,000 people per year are diagnosed with a KRAS-mutant cancer. If daraxonrasib and its successors can deliver even a modest improvement in median survival across all three tumor types, the public health implications — measured in millions of life-years — would be extraordinary.
Marina Pasca di Magliano, a researcher at the University of Michigan who spent her career studying the microenvironment of pancreatic tumors, reflected on what it means to have lived through the transition. "Almost everybody thought that it was going to be impossible to make drugs against KRAS," she said. The consensus was so strong, so broad, so deeply embedded in the institutional culture of cancer research, that challenging it required not just scientific insight but a willingness to appear professionally foolish — to risk careers on an idea that the field had already decided was dead.
The lesson the scientific community is drawing is not only that KRAS can be targeted, but that dogma in biology is particularly dangerous. "Every time there was an advance, it led to another dumping of dogma and finding out that what everybody assumed was true was actually not true," said Adrienne Cox of the University of North Carolina. The KRAS story is a case study in how scientific consensus can calcify around a negative conclusion — the assumption that something is impossible — and how that calcification, however understandable, can blind a field to real possibilities for years or even decades.
Robert Weinberg, the M.I.T. scientist who made one of the original landmark discoveries about RAS oncogenes in 1982, is now 83. In a recent interview, he marveled that it had taken 44 years for patients to benefit from that work — and that he had lived long enough to see it happen. "It would have been nice," he said with characteristic dryness, "if the Good Lord had sent us down something easier to drug." The understatement of a lifetime. And now, finally, the payoff.
"It's the beginning, not the end," said Dr. Elizabeth Jaffee of Johns Hopkins. For the hundreds of thousands of patients diagnosed every year with KRAS-driven cancers, that beginning — long deferred, long doubted — has finally arrived.
0 Comments