A 16-person Cambridge biotech going after the cancer mutations the rest of the field filed under undruggable - and it just put one of its drugs into people.
The company that named itself after the thing it hunts: druggable pockets nested inside mutations everyone else walked past.
At the 2026 AACR meeting - the year's loudest room for cancer science - a tiny Cambridge biotech walked up with data most observers expected to be a shrug. Early Phase 1 numbers usually are. Instead, Nested Therapeutics reported that its lead drug, NST-628, produced a 38% response rate and 85% disease control at the recommended dose in patients with melanoma who had already run out of options. Sixty-nine people had received it. The room paid attention.
That is the strange position Nested occupies right now: a sixteen-person outfit, founded in 2021, sitting on a single clinical asset and a development-stage second program - and yet talking about a class of cancer drivers the industry mostly gave up on. The company is not large. Its ambition is the opposite.
Here is the uncomfortable truth that started the company. Targeted cancer drugs are genuinely a triumph. They also reach a frustratingly narrow slice of patients. Map a tumor's mutations and you'll find a long list of culprits; only a short list has a drug pointed at it. The rest get the polite industry label "undruggable," which mostly means "nobody has found the handle yet."
The RAS/MAPK pathway is the cautionary tale. It is one of cancer's busiest signaling routes, mutated across a huge range of tumors. Drugs exist. They also have a habit of being outwitted - the pathway reactivates, the tumor adapts, the benefit fades. And many of these drugs never reach the brain, where cancers love to hide.
Nested's wager is that the handles are there - tucked into cryptic pockets, clustered around overlooked mutations, visible only if you bring structural biology, computational biophysics, chemistry and machine learning to the same table. Hence the name. The good targets are nested inside the ones everyone skipped.
The company was built inside Versant Ventures' discovery engine and stitched together a founding cast that reads like a who's-who of pathway biology. Co-founder Kevan Shokat is, among other things, the chemist widely credited with helping make KRAS - the textbook "undruggable" target - druggable. Arvin Dar of Memorial Sloan Kettering and Klaus Hoeflich round out the scientific founding trio. The bet they share is unfashionable: skip the crowded targets, go where the chemistry is genuinely hard.
To run it, they brought in Darrin Miles as CEO - a commercial operator, not a bench scientist. Miles spent fourteen years at Genentech, where he worked on the commercialization of Herceptin, and later served as chief commercial officer at Agios. The signal is deliberate: this is a science company that intends to ship product, not just publish.
Built within Versant Ventures around a thesis: find druggable pockets in mutations everyone else overlooked.
Emerges from stealth with $35M from Versant plus a $90M Series A led by Goldman Sachs Asset Management's life sciences division, with Foresite, Avidity, Cowen Healthcare and Section 32.
Appoints John A. Orwin as chair and adds new directors as the first program heads toward the clinic.
Preclinical data featured in AACR's "New Drugs on the Horizon" series.
Reports 38% response and 85% disease control at the recommended dose in pretreated NRAS / BRAF Class II/III melanoma. 69 patients dosed as of the Feb 1, 2026 cutoff.
Most pathway drugs try to block. NST-628 does something more devious: it acts as a molecular glue. Rather than degrading its target, it locks RAF and MEK together in a catalytically inactive complex - in plain English, it clamps the machinery shut and won't let it restart. That design directly attacks the pathway-reactivation problem that wears down earlier inhibitors.
Two more things make it unusual. It is pan-RAF, hitting multiple components of the RAS/MAPK pathway rather than a single node. And it is fully brain-penetrant - engineered to cross the blood-brain barrier, a bar most cancer drugs never clear. Behind it sits a second, development-stage program: a mutation-selective allosteric molecule aimed at the p53 pathway, latching onto a novel cryptic pocket.
Non-degrading, brain-penetrant pan-RAF/MEK molecular glue. Stabilizes RAF-MEK in an "off" state to prevent pathway reactivation in RAS/MAPK-driven solid tumors.
Mutation-selective allosteric small molecule engaging a novel cryptic pocket on one of cancer's most frequently mutated regulators.
Structural biology + computational biophysics + chemical biology + machine learning, pointed at mutation clusters to surface hidden, druggable pockets.
Widen precision oncology beyond the lucky few whose mutations already have a matching drug.
Early data is not approval, and a 69-patient readout is exactly that - early. But in heavily pretreated patients who had run out of standard options, the signal was hard to wave away.
NRAS & BRAF Class II/III melanoma, recommended dose // AACR 2026 // data cutoff Feb 1, 2026
Source: Nested Therapeutics AACR 2026 presentation. Figures reflect the recommended dose in a defined melanoma subgroup; patient-count bar is scaled to 100 for display. Early Phase 1 data - interpret with the usual caution.
Conviction is cheap; a $125M launch is not. Versant founded the company and seeded it with $35M. Goldman Sachs Asset Management's life sciences arm led the $90M Series A, joined by Foresite Capital, Avidity Partners, Cowen Healthcare Investments and Section 32. That is a serious table for a company this young, and it buys the one thing chemistry-hard biology needs most: time.
The mission underneath the financing is refreshingly narrow. Not "cure cancer." Find the driver mutations hiding in plain sight, build drugs that grip them, and widen the door so precision medicine reaches patients it currently leaves outside. "Limitless possibilities revealed" is the tagline. The work is the opposite of limitless - it is specific, slow, and unglamorous. Which is rather the point.
Return to that AACR stage. A sixteen-person company, a single clinical drug, a molecular glue that crosses into the brain and refuses to let cancer's signaling restart. The data is early. The caveats are real. Nobody is declaring victory.
But the thing that made the room pay attention wasn't the percentages. It was the premise. If the "undruggable" mutations are really just under-examined ones - if the handles have been there all along, nested inside the targets everyone skipped - then the map of treatable cancer is larger than the industry has been acting like it is. Nested Therapeutics has not proven that yet. It has done something almost as interesting: it has made the bet testable, and put it in front of patients.