●BREAKING Larkspur Biosciences doses first patient in Phase 1 study of LRK-4189 (Dec 2025) ●First-in-class degrader of lipid kinase PIP4K2C ●Launched 2023 with $35.5M seed + Series A ●Founders: Cantley · Gray · Kuchroo ●Discovery engine: LarkX machine-learning platform ●Target: microsatellite stable colorectal cancer ●HQ: One Canal Park, Cambridge MA

YesPress Profile · Precision Oncology

Larkspur Biosciences

// destroying the proteins cancer can't live without

A Cambridge biotech that decided blocking cancer's survival switch wasn't enough. So it built drugs that delete the switch entirely.

Founded 2022 Cambridge, MA ~19 employees Series A · $35.5M Phase 1 (2025)

Above: the larkspur mark - a stylized "k" with a gradient that fades from deep violet to gold, like a flower caught mid-bloom. Fitting, for a company whose whole pitch is what happens when something opens up.

Dispatch from One Canal Park

Right now, a cancer cell is trying to survive. Larkspur is making sure it can't.

In December 2025, in a clinical site somewhere far less glamorous than the science behind it, a person swallowed the first dose of a drug called LRK-4189. The molecule has one job. Find a protein named PIP4K2C inside cancer cells and degrade it - not block it, not slow it, destroy it. What happens next is the entire bet of a 19-person company in Cambridge, Massachusetts.

Larkspur Biosciences is not large. It is not loud. It rents space at One Canal Park alongside other mission-driven biotechs, the kind of building where the coffee is fine and the ambition is not. But it is doing something most of oncology has spent two decades avoiding: treating the tumor's own survival kit as the target, and treating immune evasion as a problem you solve from inside the cancer cell rather than outside it.

The problem they saw

Some cancers don't fight the immune system. They simply hide from it.

Immunotherapy rewrote oncology by teaching the immune system to see tumors. For some patients it has been close to a miracle. For others - notably people with microsatellite stable (MSS) colorectal cancer - it does almost nothing. These tumors are "cold." They evade the immune response so thoroughly that releasing the brakes on T-cells changes very little.

The conventional response is to push harder from the outside: more checkpoint inhibitors, more combinations, more of the same. Larkspur looked at the same wall and asked a less fashionable question. What if the bottleneck isn't the immune system at all? What if it's the cancer cell's own internal machinery - the genes that let it adapt to stress, survive, and stay invisible?

The founders' bet

Three scientists, one unfashionable idea, and a degrader instead of a blocker.

Larkspur was co-founded by three researchers who do not need the work: Lewis Cantley, who discovered the PI3K pathway that anchors a generation of cancer biology; Nathanael Gray of Stanford, a prolific chemical biologist; and Vijay Kuchroo of Harvard, who holds more than 25 patents and has founded five biotech companies. Their shared conviction was that cancer cells depend on a hidden set of "fitness genes" - not the obvious growth drivers, but the support staff that keep a tumor alive when conditions turn hostile.

The second half of the bet was mechanism. Most cancer drugs inhibit a protein - they sit on it and slow it down. Larkspur builds degraders - small molecules that mark a target protein for destruction and let the cell's own disposal system clear it out. For a target like PIP4K2C, the team found that degradation, not inhibition, is what actually kills the cell. Block it and the cancer adapts. Remove it and the cancer collapses.

Fig. 1 — The difference between a brake and a delete key. One asks the protein to please slow down. The other removes the protein from the conversation entirely.

The product

A discovery engine, and the pipeline it found.

The engine is called LarkX - a machine-learning-enabled gene signature platform that hunts for pathways where cancer cells lean hardest on stress-adaptation tricks. It is, in effect, a way of asking thousands of genes a single question: which of you does the tumor truly need? The answers become degradation targets.

Phase 1 · Lead

LRK-4189

First-in-class targeted degrader of the lipid kinase PIP4K2C. Triggers intrinsic cancer cell death in microsatellite stable colorectal cancer and other solid tumors. First patient dosed December 2025.

Lead Optimization

Pin1 Degrader

First-in-class degrader targeting Pin1 to modulate fibroblast differentiation and re-activate immune cells in cancer and fibrosis.

Lead Discovery

TIM-1

Early program advanced as a joint venture with Alloy Therapeutics (Tiburon Therapeutics).

Platform

LarkX

Machine-learning gene-signature platform that identifies cancer cell-intrinsic stress adaptations and surfaces novel degradation targets.

Fig. 2 — A pipeline that fits on one card each. Lean is not a weakness when every program is first-in-class.

The short, fast history

From launch to first-in-human in about thirty months.

MAY 2023

Larkspur launches

Emerges publicly with $35.5M in combined seed and Series A financing led by Polaris Innovation Fund, 3E Bioventures Capital, and Takeda Ventures.

AUG 2025

LRK-4189 revealed at ACS Fall Meeting

Announces discovery of the first reported degrader of the lipid kinase PIP4K2C, with IND-enabling studies complete and Phase 1 guided for Q4.

DEC 2025

First patient dosed

Begins dosing in the Phase 1 first-in-human study of LRK-4189 - on the timeline it set itself.

Read the data

Early tolerability, pharmacokinetics and pharmacodynamics will tell Larkspur whether the degrader thesis holds in humans.

Fig. 3 — Biotech rarely hits its own dates. Larkspur said Q4 2025 and dosed in Q4 2025. Note it for later.

The proof, so far

The numbers are small. The conviction behind them isn't.

Larkspur is pre-revenue and pre-approval - which is exactly what a clinical-stage biotech should be at this point. The relevant scoreboard isn't sales. It's capital raised, programs advanced, and timelines met. Here is the financing that got them to the clinic.

Launch financing — May 2023 ($35.5M total)

Combined seed + Series A, by reported lead and syndicate role

Total raised

$35.5M

Programs

4 active

Headcount

~19 people

Lead in clinic

Phase 1

Bars are scaled for comparison, not to a common unit. Capital, programs, headcount and stage tell one story together: a lot of science per dollar.

The money came with conviction attached. Polaris led, with Amy Schulman taking the board chair; Takeda Ventures took a board seat through Jiaping Gu; 3E Bioventures, Creacion Ventures, Med-Fine Capital and Cornell University filled out the round. For a first round, that is a serious set of names betting on a contrarian thesis.

"First and foremost, we were attracted to Larkspur because of its leadership team. They have what it takes to transform a fresh and important perspective on cancer immunity into reality."

— Amy Schulman, Managing Partner, Polaris Partners (Board Chair)

Who's running it

An all-female leadership team in an industry that rarely has one.

Worth saying plainly, because biotech leadership rosters rarely read this way. The company that three eminent male academics founded is run, day to day, by women - led by a CEO who built immune-modulation programs at Novartis and a CSO with two decades of medicinal chemistry leadership at GSK.

Catherine Sabatos-Peyton, Ph.D.

CEO · ex-Novartis (NIBR); led discovery of a first-in-class anti-TIM-3 antibody

Krista Goodman, Ph.D.

Chief Scientific Officer · 20 years at GSK, Global Head of Flexible Medicinal Chemistry

Emily Corse, Ph.D.

EVP, Biology & Translation

Amy Schulman, J.D.

Board Chair · Polaris Partners

And the founders behind the science: Lewis C. Cantley (discoverer of PI3K), Nathanael Gray (Stanford; board member), and Vijay K. Kuchroo (Harvard; 25+ patents, five companies).

Fig. 4 — The org chart most people don't expect, attached to the founding pedigree most people would kill for.

The mission

Outsmart the bottleneck. Start inside the cell.

Larkspur states its purpose without flourish: develop first-in-class protein degraders targeting cancer cell fitness genes to destroy tumors more effectively. Underneath that is a worldview. Cancer's most durable advantage isn't speed - it's adaptability. Tumors survive because they have backups, redundancies, stress responses. Larkspur's whole approach is to find the one component the backups can't replace, and delete it.

If that sounds modest for a company tackling colorectal cancer, that's the point. The team would rather under-promise and dose patients on schedule than narrate a revolution it hasn't run yet.

Why it matters tomorrow

If degraders work where blockers stalled, the map of "untreatable" shrinks.

Microsatellite stable colorectal cancer is one of the harder problems in solid tumors - common, deadly, and largely deaf to the immunotherapy that transformed other cancers. A first-in-class degrader that makes those cold tumors kill themselves would matter far beyond a single indication, because the same logic - find the fitness gene, degrade it, watch the cell fail - is a template, not a one-off. LarkX exists to keep finding the next target.

None of that is proven yet. Phase 1 is the start of a long road, and most drugs that begin it don't finish. But the thesis is testable, the team is credible, and the company did the unglamorous thing of meeting its own deadline. In biotech, that is its own kind of signal.

Back to that first dose.

The patient who swallowed LRK-4189 in December 2025 didn't make headlines. But inside their cells, a 19-person company's entire argument was being put to the only test that counts. Larkspur spent two and a half years and $35.5 million getting one molecule into one human body to ask a single question: when you take away the protein cancer needs, does cancer fail?

The flower in the logo is finally open. Now everyone watches to see what blooms.

Field notes

Six things worth knowing

The company is named after the larkspur flower; its logo is a stylized "k" with a violet-to-gold gradient, like a bloom caught mid-open.
Larkspur doesn't block its target protein - it destroys it, betting degradation kills cancer cells where inhibition fails.
Co-founder Lewis Cantley is the scientist who discovered PI3K, one of the most-studied pathways in all of cancer biology.
Co-founder Vijay Kuchroo holds more than 25 patents and has founded five biotech companies.
Its discovery engine, LarkX, hunts for "cancer fitness genes" - the survival kit tumors can't live without under stress.
Larkspur shares its Cambridge address, One Canal Park, with other mission-driven biotech startups.