MRK Pays $400M Upfront for SciBrunch KRAS G12D Drug

Merck (NYSE: MRK) signed an exclusive global license for SPR2015, a preclinical oral KRAS G12D(ON) inhibitor from privately held SciBrunch Therapeutics, paying $400 million upfront in a deal worth up to $2.13 billion. SciBrunch, a China-based biotech, is eligible for up to $1.73 billion more in development and commercial milestones across multiple indications.

The asset is a “molecular glue” designed to lock the KRAS G12D protein in its inactive, or “off,” state, a different mechanistic approach from the switch-pocket small molecules that dominate the field. KRAS G12D is the most common oncogenic RAS mutation, and it appears across pancreatic, colorectal, and lung cancers. There is no human data yet; the program is preclinical.

Mechanism: Why G12D(ON) and Why a Molecular Glue

KRAS was called undruggable for decades because it is a smooth, globular signaling switch with no obvious binding pocket. The G12D mutation keeps the switch stuck in its GTP-bound, active “on” state, driving relentless growth signaling. Two attack strategies exist: (RAS)OFF inhibitors, which catch the inactive GDP-bound state and keep the protein from re-activating, and (RAS)ON inhibitors, which bind the active state directly. The “ON” approach is considered the harder, potentially cleaner one — it hits the disease-driving form and can address the resistance mutations that arise under OFF pressure. Revolution Medicines’ Rasonque proved the ON concept in pancreatic cancer; SPR2015 extends the same target class with a different modality.

The molecular-glue design is the novelty. Instead of occupying a pocket, a glue induces or stabilizes a protein-protein interaction — recruiting SPR2015 to KRAS G12D in a conformation that locks it in the off state. Whether the modality translates into differentiated pharmacokinetics or safety can only be answered in humans, which is why this is preclinical option value. That said, Merck’s bet lands on the most prevalent RAS mutation in cancer: pancreatic tumors are roughly KRAS G12D-driven in a large share of cases, and pancreatic cancer remains one of the most lethal solid tumors with the fewest options.

What It Means for the Trade

The timing is not accidental. Merck’s move lands weeks after Revolution Medicines’ Rasonque (daraxonrasib) became the first RAS(ON) multi-selective inhibitor approved by the FDA, for previously treated metastatic pancreatic cancer. That approval validated the RAS-targeting thesis and pulled pharma capital into the space. Merck is now paying for option value on a differentiated approach while Keytruda’s 2028 patent cliff continues to loom over its oncology franchise — the drug that generates the bulk of Merck’s earnings loses exclusivity, and every pipeline dollar is pointed at finding the next anchor.

The structure signals how Merck is accounting for the bet. The company will take a $400 million pre-tax charge, roughly $0.13 per share, in its third-quarter 2026 results. For investors, the trade expression is one-sided. SciBrunch is private and China-based, founded by serial entrepreneur Dr. Tao Hu, so there is no tradable ticker there. Merck, at $148.78 a share, is the only listed vehicle through which the deal is investable, and a preclinical licensing agreement is a rounding error against a pharma of Merck’s size. The real trade relevance is thematic confirmation: big pharma is paying hundreds of millions upfront for preclinical RAS assets, which validates the platform valuations of the public RAS players — Revolution Medicines chief among them.

Risks and What Could Go Wrong

The competitive field is filling fast. Beyond Revolution Medicines, competitors targeting the same KRAS G12D node include Lilly with olomorasib, and a raft of preclinical G12D(ON) programs looking to differentiate on selectivity and oral bioavailability. Merck’s molecular-glue bet is a bet that the mechanism, not the specific scaffold, is what ultimately wins the class.

The risk is straightforward: this is a pre-human bet on a validated but crowded mechanism. Rasonque already owns the first-to-market position, and Merck’s molecular glue will have to distinguish itself in the clinic before it means anything commercially. Glue-type mechanisms carry their own pharmacology risks — off-target protein interactions are the modality’s known failure mode — and the pharmacokinetic and specificity hurdles that have challenged competing G12D(ON) inhibitors do not disappear because the chemistry is novel.

The next signal will be preclinical-to-clinical translation, likely years out: an IND filing, first-in-human data, and any evidence the glue achieves oral exposure at tolerable doses. Until then, the $2.13 billion headline is a series of gated milestones that may never be paid, and the $400 million is already spent.

Merck shares closed at $148.78. For context on the drug Merck is effectively chasing, see Rasonque’s August FDA approval. The terms are detailed in Merck’s announcement.