Pathos AI acquires majority stake in DeuterOncology's MET inhibitor

Pathos AI has acquired a majority stake in DeuterOncology, a Belgium-based oncology company whose lead asset, DO-2, is a deuterated third-generation MET kinase inhibitor targeting MET-altered cancers. The New York-based acquirer said the deal was driven entirely by its proprietary Foundry platform, an AI system that continuously analyses large-scale clinical datasets to surface what it calls undervalued oncology assets.

The transaction represents a live test of an emerging model in drug development: replacing relationship-driven deal-sourcing with algorithmic asset screening. Pathos said the entire process — from Foundry flagging DO-2 as a top-ranked candidate to management signing off on the acquisition — was completed in a fraction of the time required by conventional due diligence. DO-2 joins three other portfolio decisions Foundry is said to have driven in the first quarter of 2026 alone.

The asset

DO-2's distinguishing feature is its deuterated molecular structure combined with what DeuterOncology describes as "fast on / fast off" binding kinetics. The company argues this design delivers meaningful MET inhibition for eight to twelve hours per day — enough for antitumour activity — while avoiding the sustained endothelial damage that underpins the high peripheral oedema rates seen with approved MET inhibitors. Current approved agents in MET-altered non-small cell lung cancer, including tepotinib and capmatinib, carry reported peripheral oedema rates of 62–82%, often necessitating dose reductions or discontinuation.

In a Phase 1 dose-escalation study across eight clinical sites in the Netherlands, Belgium, and France, DO-2 was evaluated in 28 patients. The company reported 100% tumour shrinkage across all ten evaluable MET exon 14 skipping NSCLC patients, zero Grade 4 adverse events, and a peripheral oedema rate of 5%. The regimen was administered as a 60 mg once-daily oral dose. Patent exclusivity is stated to run to December 2040.

While these figures are striking, the cohort size is small — ten evaluable patients for the efficacy read — and Phase 1 studies are not designed or powered to establish efficacy. Investors and clinicians will need substantially larger, controlled datasets before drawing conclusions about DO-2's competitive standing.

Market context and competitive positioning

The MET-altered NSCLC space has attracted sustained investment over the past decade. Beyond tepotinib (Merck KGaA) and capmatinib (Novartis), next-generation MET-directed programmes are in development at several companies seeking to address tolerability limitations, making the competitive landscape increasingly crowded even at the third-generation tier. Savolitinib, partnered with AstraZeneca — which is also a Pathos AI partner — has shown activity in MET-driven disease, adding a degree of strategic complexity to Pathos's positioning.

The broader question raised by the deal is whether AI-native asset identification can reliably outperform traditional deal-making. Pathos's Foundry platform operates through thousands of AI agents drawing on the Pathos Oncology Foundation Model, continuously re-ranking assets and informing portfolio decisions. The company is one of a growing cohort of AI-first oncology developers — alongside others building on foundation models trained on clinical and genomic data — that argue systematic screening reduces the role of network bias in asset selection. Whether this translates to better clinical outcomes remains, for now, an open empirical question.

Chief executive Iker Huerga said Foundry "evaluates every asset purely on its merits — mechanism, pharmacokinetics, clinical signal, and probability of success." Pathos's existing pipeline includes clinical-stage programmes in metastatic castration-resistant prostate cancer in addition to MET-altered NSCLC.

The next material milestone for DO-2 will be the design and initiation of a Phase 2 study. Pathos has not disclosed a timeline, regulatory engagement plan, or the financial terms of the DeuterOncology stake acquisition.