KRAS is the most frequently mutated oncogene there is — and for forty years it was the textbook definition of "undruggable." Its surface is smooth, with no obvious pocket for a molecule to grab, and it clings to its fuel — the energy-rich molecule GTP — with a tenacity no ordinary drug could outcompete. Generations of chemists tried and failed, until the target became a kind of cautionary emblem: the most important oncogene in cancer, and the one no one could touch. Then, around 2018, the wall came down.
From the Knitify Pharma Demand Radar — our running read on where pharmaceutical demand is heading. We publish the observations; the method behind them stays in-house.
Why KRAS resisted for forty years
To understand why demand behaves the way it does here, you have to understand why the target was so stubborn. KRAS is a molecular switch: it flips between an "on" state bound to GTP and an "off" state bound to GDP, relaying growth signals into the cell. It normally switches itself off by chewing up its own GTP, a slow but reliable self-braking action. The cancer-driving mutations sabotage that brake — the protein loses its ability to turn itself off and stays locked "on," telling the cell to grow and divide without pause.
The chemistry of that switch is what made it so hard to drug. The molecule's surface is remarkably smooth — there is no deep, well-shaped cavity of the kind medicinal chemists normally exploit to anchor a small molecule. The one obvious binding site, where GTP itself docks, is essentially unusable: KRAS binds GTP with picomolar affinity — an almost impossibly tight grip — and the inside of a living cell is flooded with GTP. A competitive drug would have to displace that fuel against an overwhelming excess of it. That combination — a featureless surface plus an unbeatable native ligand — is precisely why, for four decades, "undruggable" and "KRAS" were nearly synonymous.
The trick that cracked the wall
The breakthrough was a trick of chemistry rather than brute force. Researchers stopped trying to compete at the fuel pocket and went looking for a new one. A specific mutation, G12C, swaps a single amino acid for a cysteine — a residue with a reactive sulphur atom. Under the right conditions the protein opens a shallow, previously unnoticed groove near that cysteine, and a molecule can be designed to slip in and form a permanent, covalent bond to the sulphur. Instead of competing with GTP, the drug simply nails the protein shut, catching it in its "off" state for good.
This is what makes the approach so specific: the covalent warhead only reacts with the mutant cysteine, which healthy cells do not carry, so the drug largely spares normal tissue. That insight produced the first approved KRAS drugs — sotorasib and adagrasib — turning the field's most famous dead end into a validated, drug-bearing target almost overnight.
The signal: the follow-on wave is building
The insight that G12C exposes a reactive cysteine a covalent molecule can latch onto permanently produced those first two drugs — and adagrasib is now on our accelerating list. [1]
But the G12C mutation is only one slice of KRAS-driven cancer, and the signal we care about isn't really about the two approved drugs. It's about what a fallen wall unleashes. Once the target was proven druggable, chemists fanned out in every direction at once: after the other, more common mutations; toward non-covalent binders that don't need a reactive cysteine; toward pan-RAS agents that aim to hit many mutations with a single molecule; and toward degraders that don't just switch the protein off but mark it for destruction entirely. One cracked target became a whole field, seemingly in the span of a few years.
The scale of the prize keeps the race relentless. KRAS mutations drive a large share of the deadliest cancers, so even a partial win against any one variant is commercially enormous. The approved G12C drugs are already pushing into new tumor types and combinations, while an entire second rank of programs lines up behind them. Each new mechanism is not a substitute for the last but an addition to it — and every addition is a fresh body of molecules that has to be designed, made, and supplied.
One mutation, many cancers — and why G12D matters
KRAS is not a single problem but a family of them, and the specific letter of the mutation matters enormously. Different cancers favor different variants. G12C, the one that fell first, is comparatively common in lung cancer, which is part of why the first drugs landed there. But in pancreatic and colorectal cancer — two of the hardest tumors in oncology — the dominant driver is more often G12D, a substitution that does not offer a convenient reactive cysteine to attack.
That is why so much follow-on energy points at G12D: it is a larger patient population and a more urgent unmet need, yet it demands a fundamentally different chemical strategy, because the covalent trick that worked for G12C simply doesn't apply. Solving G12D means building non-covalent binders that hold a smooth surface through shape and affinity alone — a harder problem, and therefore an even richer source of new molecules. Beyond the individual variants sit the pan-RAS ambitions, which try to address the whole mutant family at once. Each branch is a separate campaign, with its own chemical matter, intermediates, and manufacturing demands.
Resistance writes the next chapter
If the biology of KRAS explains why the field is broad, the biology of resistance explains why it is durable. The early clinical experience carried a sobering lesson: single-agent responses, while real, were often modest and rarely lasting. Cancers under pressure from a G12C drug find ways to escape — by acquiring new mutations in KRAS itself, by reactivating the same growth pathway through a neighboring node, or by leaning on parallel routes the drug never touched. Resistance is not a rare failure mode here; it is close to a certainty, and it arrives fast.
That inevitability is exactly what guarantees successive generations of chemistry. It has pushed the field hard toward combinations — pairing a KRAS inhibitor with an antibody against EGFR, for instance, to shut down a common escape route, a strategy especially important in colorectal cancer. It also motivates the next chemical generation outright: more potent binders, agents that block the protein in both its states, and degraders that remove the target so reactivating mutations have nothing left to bind. Each wave of resistance doesn't close the market; it reopens it, calling for another round of molecules to be invented and made.
From one program to a decade — and who feels it downstream
This is the pattern worth internalizing. An "undruggable" target that becomes druggable doesn't add a single program to the roster — it validates an entire disease axis and sets off a race. Every company with a stake in the most common oncogene in cancer now has a reason to build chemistry against it, and the resistance that inevitably follows guarantees successive generations. For a supplier, the demand signal from a newly-cracked target is among the most durable there is: it compounds for years, across mutations, mechanisms, and modalities. The first molecule leaves a sharp, accelerating trail — and the wave behind it is where the real volume lives.
Read from the perspective of the people who have to make things, the implications ripple outward. For API suppliers and specialty-chemical makers, a validated target signals a coming wave of new scaffolds, novel warheads, and building blocks that suddenly need to be sourced at scale — often exotic chemistry that didn't exist as a commercial product a year earlier. For contract manufacturers and CDMOs, a decade-long axis of related programs is exactly the durable, repeat-order demand worth building capacity around, because the chemistry keeps evolving rather than arriving once. And for business-development teams, the moment a wall visibly falls is the moment to move — before the follow-on field is crowded and the obvious partnerships are taken. The loudest signal, the first approvals, is the least useful to act on, because by then everyone sees it; the value is in reading what must come next. A cracked target is not an event — it is the opening of a schedule.
The caveats: a fallen wall is not a finished building
None of this should be mistaken for a straight line up and to the right. The honest reading of KRAS carries real caveats. Single-agent activity was more modest than the headlines suggested, durability has been a persistent disappointment, and the combinations meant to fix that bring their own toxicity problems. The hardest variants, G12D chief among them, remain genuinely difficult, and a non-covalent solution is not guaranteed to be as clean or potent as the covalent one before it.
A validated axis is also no promise that every program on it will succeed. Crowded fields produce many more failures than winners; being fifth or tenth into a mechanism is a very different bet than being first. The durable demand signal is about the field as a whole — the sustained, compounding need for new chemistry against a target the industry can no longer ignore — not a prediction that any single follow-on molecule will reach the market. A wall that falls opens the ground for building. It does not, by itself, decide which buildings will stand.
Editorial commentary from the Knitify Pharma Demand Radar. Not investment, legal, regulatory, or medical advice.