Turning Abandoned Drugs into Breakthroughs

Episode Summary

Executive Summary: PureTech’s Eric Alenko explains the company’s hub-and-spoke model for systematically rescuing discontinued drugs by starting with unmet need, leveraging human pharmacology, and solving the specific liability that blocked development. He highlights disciplined go/no-go criteria, AI as an acceleration tool, and the Karuna/Cobenfy success as proof that de-risked assets can become major commercial wins.

Main Topics: PureTech’s hub-and-spoke innovation model (Priority: 5/5): PureTech creates therapeutics centrally, then spins them into independent companies that raise third-party capital while the hub shifts to a board/investor role. Systematic drug rescue from validated pharmacology (Priority: 5/5): The company focuses on drugs with human clinical evidence that were discontinued for fixable reasons, then works backward from the problem to a solution. Dispassionate go/no-go decision-making (Priority: 5/5): Alenko stresses defining success before seeing data, using pre-set criteria and killer experiments to avoid emotional bias and sunk-cost behavior. AI as a force multiplier, not a black box (Priority: 4/5): PureTech uses AI with Biooptic to accelerate identification of promising drugs and liabilities, but keeps the key scientific insight and solution design in-house. The xenomeline/Cobenfy story (Priority: 5/5): PureTech identified xenomeline as a schizophrenia opportunity, then solved its peripheral side effects by pairing it with another drug that blocks unwanted receptor activity. Capital efficiency and value creation (Priority: 5/5): The Karuna example shows how the model can produce outsized returns by building a dedicated company around a validated asset and attracting the right team and capital. Practical limits and licensing challenges (Priority: 3/5): While discontinued assets are visible, the hardest operational issue can be out-licensing them from big pharma, which often lacks a clear internal champion.

Key Arguments: PureTech’s model institutionalizes biotech creation instead of relying on serendipity or a single entrepreneur’s intuition. The company’s edge comes from focusing on validated human pharmacology and solving the precise reason a drug failed. Dispassion is essential: success criteria are defined before experiments so programs can be advanced or killed objectively. The hub-and-spoke structure lets PureTech redeploy resources when a program fails, improving capital efficiency. AI can speed up candidate and liability discovery, but it does not replace the scientific judgment needed to craft the solution. Cobenfy succeeded because PureTech used a simpler, bespoke solution to xenomeline’s peripheral side effects rather than pursuing a more traditional medicinal chemistry strategy. Independent spoke companies can attract specialist talent and capital better than a centralized internal program. Big pharma’s main challenge in rescue partnerships is often process and championing, not simply identifying shelved assets.

Data Points: PureTech investment in Karuna: $18.5 million - Approximate own capital invested by PureTech in the company that developed Cobenfy Cash proceeds from Karuna/Cobenfy: ~$1.1 billion - Equity sales, milestones, and royalty monetization received by PureTech tied to Karuna/Cobenfy BMS acquisition of Karuna: $14 billion - Bristol Myers Squibb acquired Karuna in 2024 Royalty rate: 2% - PureTech retains royalties above $2 billion in sales Schizophrenia treatment gap: 3 decades - Cobenfy was described as the first new drug to treat schizophrenia in adults in 30 years AI analogy scale: 100 analysts - Alenko compared AI to having the equivalent of a very large analyst team crunching data দ্রুত

Pivotal Quotes: "“It’s really hard to make someone understand something when their livelihood depends on them not understanding.”" — Eric Alenko: On the industry’s bias to continue weak programs and the need for objective kill criteria "“We really look at it as a force multiplier.”" — Eric Alenko: Describing how PureTech uses AI with Biooptic to speed up drug rescue work "“What we really want to understand is how much does the mechanism of action really hold water in terms of being proven in the context of the disease.”" — Eric Alenko: On selecting rescue candidates and judging whether the biology is sufficiently validated

Implications: The conversation suggests drug rescue can be a highly efficient innovation path when paired with disciplined selection, clear stop rules, and specialized company-building. For biotech, the model favors fewer, better bets over broad discovery pipelines.