Synthekine and the Art of Asking Less of IL-2

Interleukin-2 has the social instincts of someone who sends an invitation to the entire address book. The protein can help rally immune cells against cancer, but its message reaches cells that may cause serious harm. For drug developers, the seduction is obvious: keep the useful instruction, lose the indiscriminate broadcast. Synthekine was founded to see whether molecular engineering could make that distinction real.

The company’s scientific founder, Stanford structural biologist K. Christopher Garcia, has spent a career looking at the shapes made when cytokines meet their receptors. Those shapes matter. A small alteration to a protein’s binding surface can change which cells hear it and how loudly. Synthekine licensed work from Stanford, raised $82 million in a 2020 Series A, and set out to turn that molecular view into medicines.

The trouble with a powerful message

Cytokines are small proteins that help immune cells communicate. IL-2, one of the best known, can expand T cells that attack tumors. An IL-2 drug, aldesleukin, has long shown that the pathway can matter in cancer. It has also shown the price of turning that pathway on broadly: severe toxicity, including capillary leak syndrome, restricts its use. Synthekine’s premise was that the trouble lay partly in the audience. If a drug spoke more clearly to tumor antigen-activated T cells and less to other lymphocytes, perhaps the therapeutic window would widen.

STK-012 embodies that idea. It is an engineered, alpha/beta-biased IL-2 partial agonist. The name is a thicket; the design is easier to picture. Activated T cells carry a high-affinity IL-2 receptor with three subunits. Natural killer cells and some other lymphocytes rely more on a different receptor arrangement. Synthekine tuned the cytokine’s binding and signaling so the intended T cells receive a stronger instruction while broad activation is limited. “Partial” describes the controlled signal. In this case, restraint is the design brief.

The company is trying to turn a broadcast into a more selective conversation.What receptor bias is meant to accomplish

The approach also distinguishes Synthekine from a simple increase in dose, and from some earlier IL-2 engineering efforts that avoided the alpha receptor subunit. Its candidate is designed to retain a path to the activated T cells that express that subunit. It remains an investigational claim about mechanism and clinical benefit; the attractive drawing of a receptor cannot tell a patient what will happen.

The moment a platform became a question

At launch, Synthekine had three engineering routes. It could alter natural cytokines into partial agonists. It could assemble surrogate agonists that bring receptor pieces together in new geometries. And it could pair an engineered cytokine with an engineered receptor in a cell therapy, a kind of private channel between drug and cell. That variety helped attract $107.5 million in a 2021 Series B and $100 million in a 2023 Series C. It also created the classic platform company’s problem: each promising direction asks for its own expensive proof.

The orthogonal route reached patients. STK-009, an altered IL-2, was paired with SYNCAR-001, a CD19-directed CAR-T cell carrying the matching receptor. The combination entered Phase 1 for blood cancers. But Synthekine’s current pipeline places STK-009, SYNCAR-001 and SYNCAR-002 under “Legacy Programs” and says they are no longer in active internal development. They are available for partnering. That statement does not say the science failed. It does show a company choosing where to spend its own next dollar.

The sharper bet is STK-012. In November 2025, the first patient was dosed in SYNERGY-101, a global randomized Phase 2 trial in first-line PD-L1-negative nonsquamous non-small cell lung cancer. Patients receive the current pembrolizumab-and-chemotherapy regimen with or without STK-012. The planned enrollment is about 105. Its primary endpoint is objective response rate; progression-free survival and safety are among the secondary measures. It is the sort of comparison that a platform presentation, however clever, can never supply.

What the early data can say

At AACR in April 2026, Synthekine presented updated results from an earlier study of STK-012 combined with pembrolizumab, pemetrexed and carboplatin. Among 36 patients evaluable for efficacy, 50% had an objective response. Thirty-two of the 36 were PD-L1-negative. In a group of 18 patients whose tumors carried at least one alteration in STK11, KEAP1 or SMARCA4, the reported response rate was 61%. Those numbers deserve attention because these tumor features are associated with resistance to current chemoimmunotherapy.

There is a second set of numbers, less suited to a triumphant poster. Among 39 patients evaluable for safety, rash or dermatitis and nausea each occurred in 51%, and fatigue in 46%, according to the company. It reported no dose-limiting toxicities and no discontinuations attributed to STK-012 in that group. Such findings are useful for judging whether the engineered signal behaves as intended. They are still early findings, gathered in a small study where every patient received several treatments at once.

The tempting comparison is with historical response rates of 23% to 32% for standard chemoimmunotherapy in PD-L1-negative disease, figures cited by Synthekine. But historical patients differ from trial patients in ways a graph cannot erase. SYNERGY-101 exists to make the comparison inside one randomized study. If the two arms look alike, the beautifully selective receptor design will have met a stubborn clinical fact. If they differ convincingly, Synthekine will have evidence that the molecular distinction matters beyond the laboratory.

Two partners, two kinds of value

Synthekine’s business has another strand: letting large drugmakers use parts of its engineering system. In 2021, Merck agreed to fund research using Synthekine’s surrogate cytokine agonist platform, with rights to develop candidates for as many as two cytokine targets. The deal included an upfront payment of undisclosed size, possible milestones of up to $525 million per target, and potential royalties. That headline milestone number is conditional, not cash collected.

In 2024, Sanofi made a more concrete commitment to selective IL-10 agonists for inflammatory disease: $40 million upfront, with additional milestones and royalties possible. Synthekine does defined early research; Sanofi takes responsibility for later development and commercialization. The arrangement lets a small company pursue more biology than it could finance alone.

Merck appears again in a different role. A February 2026 clinical trial agreement has Merck supply Keytruda, its pembrolizumab drug, for SYNERGY-101. Synthekine retains commercial rights to STK-012. One Merck deal monetizes a discovery platform; the other helps test Synthekine’s own lead asset. The distinction explains the company’s model better than a single revenue estimate would. Its present customers are pharmaceutical collaborators and clinical investigators; patients are the intended beneficiaries if a candidate eventually gains approval.

A narrower company, a harder test

The rest of the active pipeline has not vanished. STK-026, an IL-12 partial agonist for solid tumors, is listed at the stage before an investigational new drug application. An IL-22 program, STK-038, addresses obesity and inflammatory disease earlier in development. There is an IL-10 effort with Sanofi. These programs keep the original platform idea alive while STK-012 carries the most immediate clinical burden.

In August 2026, Susan Molineaux became CEO, succeeding Debanjan Ray, who had led Synthekine since its founding and remains an adviser. Molineaux has run oncology drug companies through later stages of development. The appointment makes sense at a company whose central story has become trial execution rather than the elegance of its discovery machinery.

There is a lesson here for anyone building from a powerful technology. Synthekine made several tools, financed several programs, then put a more specific claim before a randomized test. The piece a reader can borrow is the sequence: understand the mechanism, choose the patient group where it should matter most, and design the comparison that can embarrass the hypothesis. IL-2 has never lacked ambition. Synthekine’s more interesting wager is that, sometimes, asking an immune signal to do less may let a medicine do more.