Novartis Bets on Antares as the Oncology Targeted-Therapy Race Turns Toward “Seeing” Undruggable Proteins

Many bottlenecks in cancer treatment do not stem from researchers being unaware that tumor cells have abnormalities, but from the fact that the most critical abnormal proteins are often difficult for drugs to touch. So-called “undruggable” targets have long been like the dark side of oncology: they may drive cancer cell growth, yet lack clear binding pockets, predictable ways to regulate them, or a sufficiently safe path for drug design.

Fierce Biotech reported that Novartis will pay Antares Therapeutics $105 million to advance a collaboration aimed at finding and validating these hard-to-attack cancer targets. The report’s headline said the core of the deal is to “illuminate” undruggable targets; in other words, the point is not for a single drug to immediately enter late-stage clinical development, but first to build a more actionable map in the murky areas of cancer biology.

Public information remains quite limited. Existing reports have not provided a full list of targets, specific tumor types, details of preclinical data, or the triggering conditions for the various milestones in the collaboration. Therefore, this amount should be understood as a major pharmaceutical company’s investment in an early-stage platform and R&D direction, rather than an endorsement of efficacy that has already been clinically proven.

This type of collaboration is drawing attention because oncology drug development is moving beyond “finding mutations” toward “judging which mutations can truly be exploited.” Gene sequencing has made cancer maps more detailed, but a large number of abnormalities remain at a level that is difficult to translate into drugs. If Antares’ platform can more accurately identify vulnerabilities that cancer cells depend on, Novartis may be able to secure a position before candidate targets become crowded.

However, multiple barriers still separate target discovery from patient benefit. Proteins that are difficult to drug are often difficult not simply because screening tools are lacking, but because protein structure, intracellular localization, toxicity windows, and tumor heterogeneity are involved. Early-stage platforms can improve hit rates, but they cannot replace step-by-step validation in animal models, human safety, and efficacy signals.

For Novartis, this deal extends large pharmaceutical companies’ recent interest in novel oncology target platforms: rather than only entering bidding wars once assets have matured, it can move earlier into the discovery end and obtain options for future pipelines. For Antares, the collaboration brings funding and drug development experience, while also placing the platform under more rigorous industry scrutiny.

The real answer will not appear in the deal value, but in whether subsequent candidate targets can be clearly validated, whether they can yield designable molecules, and whether they can cross the safety and efficacy thresholds before and after entering the clinic. This collaboration is a reminder that the next step in the cancer drug race may not be to pursue known targets faster, but first to see clearly the targets that have long been hiding in the dark.