Pathos AI acquires majority stake in DeuterOncology for AI-sourced cancer drug
What's the deal? Pathos AI, a New York-based clinical-stage company, has acquired a majority stake in Belgium's DeuterOncology to advance DO-2, a third-generation MET kinase inhibitor for patients with MET-altered cancers. The deal marks one of the first clinical-stage oncology acquisitions sourced, evaluated, and executed through an AI-powered drug development platform.
Pathos used its Foundry platform — which continuously analyses clinical datasets, regulatory filings, conference proceedings, and real-world evidence — to flag DO-2 as a top-ranked candidate in late 2025. The entire process from identification to investment decision took a fraction of the time required by traditional due diligence.
"Foundry evaluates every asset purely on its merits — mechanism, pharmacokinetics, clinical signal, and probability of success," said Iker Huerga, chief executive officer of Pathos AI. "DO-2 scored at the top of our models."
Why now? MET inhibitors are an established treatment class for MET-altered non-small cell lung cancer (NSCLC), but every approved agent carries peripheral edema rates of 62–82%, frequently forcing dose reductions or treatment discontinuation.
DO-2's deuterated structure and "fast on / fast off" binding kinetics deliver potent MET inhibition for 8–12 hours per day — enough for antitumour activity without the sustained endothelial damage that drives chronic edema. In a Phase 1 study of 28 patients, DO-2 achieved 100% tumour shrinkage in all evaluable MET exon 14 skipping NSCLC patients (10 out of 10), with a peripheral edema rate of just 5%. It recorded zero Grade 4 adverse events and requires only a 60mg once-daily oral dose. Patent exclusivity runs to December 2040.
What could go wrong? Phase 1 results from 28 patients, while striking, are a small sample. Larger trials could reveal new safety signals or less consistent efficacy. The AI-driven sourcing model is also unproven at scale — Pathos itself calls this one of the first such acquisitions, meaning there is limited precedent for how well algorithmic deal-sourcing translates into clinical and commercial success.
The signal: Pathos AI's move from algorithmic asset screening to outright acquisition compresses a process that typically takes large pharma months of relationship-driven scouting into weeks. Dealroom classifies Pathos as a corporate investor, underscoring that the company is not simply licensing or partnering but deploying capital directly — a model that, if validated by DO-2's later-stage trials, could pressure traditional biopharma acquirers to adopt similar data-driven sourcing or risk losing access to undervalued clinical assets altogether.
Read more: globenewswire.com
