BlossomHill
Therapeutics

In a lab off Science Center Drive, a small team is doing something most of biotech long ago outsourced to robots: designing cancer drugs by hand, on purpose, starting from the disease and working backward to the molecule.

BlossomHill Therapeutics is fifty-some scientists, two drugs in the clinic, and roughly $257 million in the bank. No products on pharmacy shelves. No revenue worth printing. What it has instead is a track record - the people here have already designed three medicines the FDA approved - and a conviction that the next leap in oncology comes from thinking harder, not screening faster.

It is, in other words, a bet on intelligence over brute force. The company even says so out loud: human intelligence, creative thinking, proven drug design. The kind of mission statement that sounds like marketing until you notice the founders have done it before.

Targeted therapy works. Then the cancer adapts.

Here is the uncomfortable truth about precision oncology: the better a drug is at hitting its target, the harder the tumor works to change the target. A patient with EGFR-mutant lung cancer starts a targeted therapy, responds beautifully, and then - months or a couple of years later - the cancer returns wearing a different mask. A new mutation. A bypass route. A hideout in the brain where many drugs can't follow.

Resistance is not a bug in targeted therapy. It is the plot. Every clean response carries the seed of its own rematch, and the rematch is the part that kills people. The field has spent two decades playing whack-a-mole: a drug for one mutation, then another drug for the mutation that defeated the first.

BlossomHill's wager is that you can see the rematch coming and design for it from the start - molecules engineered to stay selective across mutations, to reach the brain, and to grip targets at angles older drugs couldn't.

They sold the last company. Then they did it again.

J. Jean Cui, Ph.D. is the kind of chemist whose name shows up in the fine print of medicines that actually shipped. She designed crizotinib, lorlatinib and repotrectinib - three FDA-approved cancer drugs. If you want a short biography of a drug designer's career, "three approvals" is a long one.

Her co-founder, Y. Peter Li, Ph.D., M.B.A., runs the business side. Together they built Turning Point Therapeutics, took it public, and sold it to Bristol Myers Squibb in 2022. Most people would have called that a career. They called it a warm-up.

In 2020 they founded BlossomHill on a deceptively simple premise: design first. Don't screen a million compounds and hope. Understand the biology of the disease, then draw the molecule that fits it - including the resistance you haven't met yet. It is slower upfront and, when it works, much faster downstream.

Two molecules, one philosophy.

BlossomHill's pipeline is small and deliberate. Both lead candidates are macrocyclic - a ring-shaped chemistry that can wrap around targets that flatter, more conventional molecules slide right off. The shape is the strategy.

BH-30643 is the headline. It is designed to be "OMNI" - active across the spectrum of EGFR mutations rather than one at a time - and CNS-active, so it can chase tumors into the brain. BH-30236 is the quieter, stranger bet: instead of blocking one oncogene, it interferes with how cancer cells splice their RNA, a mechanism most drugs never touch.

Build the next leap, not the next me-too.

Plenty of biotechs promise to cure cancer. BlossomHill makes a narrower, more credible promise: to design medicines that hold up when the disease fights back. Its stated mission is to build on the team's record of FDA-approved precision medicines and create the next leap forward in oncology - and, increasingly, in autoimmune disease, where the same design-first chemistry applies.

That's the part skeptics should weigh. The company is wholly-owned and pipeline-deep, not a single-asset lottery ticket. It has the founders, the funding, and the chemistry. What it hasn't shown yet is the only thing that ultimately counts in this business: a drug that works in people, for years, after the cancer tries to take the lock apart.

Back to that lab off Science Center Drive.

The chemist is still arguing with the stubborn molecule. But the argument has changed. In 2020 it was theoretical - a thesis on a whiteboard, two founders, and a memory of three drugs that worked. In 2026 the thesis has a number attached: patients dosed, a trial running, data on the way.

If BH-30643 does what it's designed to do - stay selective across mutations, follow cancer into the brain, hold the line where older EGFR drugs broke - it won't just be another entry in the precision-oncology catalog. It will be evidence that you can anticipate the rematch and design for it, before the cancer ever changes the locks.

That's the whole bet. Not a miracle. A better-engineered fight. And for once, the people making it have already won this kind of fight before.