Summary of 664. Are Thousands of Medical Cures Hiding in Plain Sight?

Overview of 664. Are Thousands of Medical Cures Hiding in Plain Sight? (Freakonomics Radio)

This episode examines drug repurposing — finding new uses for existing medicines — as a fast, low‑cost way to treat diseases that currently have no approved therapies. Through personal stories (a brain‑eating amoeba and a physician‑scientist’s near‑death and recovery), interviews with researchers, FDA staff, and economists, the show lays out how science, AI, registries, and market‑shaping incentives could unlock many effective treatments already sitting on pharmacy shelves — if the economic and adoption barriers can be solved.

Main stories and examples

• Balamuthia (a rare, often fatal brain‑eating amoeba)
  • No FDA‑approved treatments exist. A clinician used nitroxoline (a 50‑year‑old UTI drug approved in Europe) under an emergency IND to treat patients; some survivors resulted.
• David Fagenbaum / Castleman disease
  • Fagenbaum (physician‑scientist, cofounder of EveryCure) relapsed with Castleman disease and used stored blood and lymph node samples to identify mTOR pathway hyperactivation.
  • He repurposed sirolimus (rapamycin) — originally discovered from soil on Rapa Nui and used as an immunosuppressant — producing dramatic recoveries in some patients. Sirolimus helps about 20–25% of Castleman patients.
• Lidocaine and breast cancer
  • A randomized trial of 1,600 patients showed peri‑tumor lidocaine injection before surgery correlated with a 29% reduction in mortality. Despite large potential benefit, it hasn’t been adopted widely due to lack of commercial incentive.
• System successes & scale
  • EveryCure reports discovering 14 drug‑disease matches and claiming well over 1,000 lives helped through repurposed drugs.
  • CureID (FDA/partners) is a public registry for clinician, patient, and literature reports about repurposed treatments.

How drug repurposing works (science + tech)

• Biological signal → drug match
  • Example: detect pathway (mTOR) activated in disease → identify drugs that inhibit that pathway (sirolimus).
• Computational approach
  • EveryCure uses biomedical knowledge graphs and AI to score all drug × disease pairs (~4,000 drugs × ~18,000 diseases) for likely efficacy.
  • The model rates candidates (0–1); high scores are human‑reviewed by MDs/PhDs and prioritized for lab/clinical follow‑up.
• Evidence pipeline
  • Case reports and registries (CureID) generate hypotheses.
  • Preclinical validation (lab samples, proteomics).
  • Clinical trials or compassionate/emergency use follow when justified.

Major barriers

• Economic disincentives
  • Most repurposing candidates are generics; pharmaceutical firms lack monopoly profit incentives to fund trials because others can then sell the same cheap drug.
  • Patenting a new use is often unenforceable in practice: prescriptions don’t specify indication and generics can substitute.
• Adoption lag
  • On average, it takes ~17 years for medical advances to reach routine practice; absence of marketing and incentives for generics slows uptake.
• Reporting & data problems
  • Clinicians have limited incentive/time to enter case data in registries; scaling registries like CureID is challenging.
• Regulatory & institutional resistance
  • FDA and medical communities are cautious; repurposing data often seen as hypothesis‑generating rather than definitive evidence for approval.

Policy and funding proposals discussed

• Advanced Market Commitments / Pull funding (Market Shaping Accelerator)
  • Historical success: pneumococcal vaccine AMCs (Gavi + Gates) helped drive lower prices and broader access; estimated to save hundreds of thousands of lives.
  • Proposed repurposing model: federal payers (Medicare/Medicaid/NIH) promise outcome‑based rewards to firms that prove new uses for generics. Payer would measure annual adoption and pay a portion of estimated health‑care savings (pay‑for‑impact).
  • Rough cost estimate: ~ $1 billion per successful repurposing opportunity (policy design still being developed; conversations with HHS ongoing).
• CureID (FDA)
  • A public, open registry and app to collect clinician/patient reports and published case data to accelerate hypothesis generation and evidence aggregation.
  • Current scale: ~700 clinician submissions, ~650 patient/caregiver submissions, ~5,000 literature cases, and ~115,000 EMR‑extracted cases (mostly for acute COVID‑19).
• EveryCure’s strategy
  • Combining lab work, AI screening, and targeted clinical programs. Has raised >$100M in philanthropic and federal support to scale repurposing efforts for 15–25 conditions over five years.

Key data & soundbites

• Rough universe: ~18,000 known human diseases; only ~25% have an FDA‑approved treatment.
• Drug pipeline: ~4,000 drugs considered for repurposing analysis.
• Cost/time contrast:
  • New drug development: $1–2 billion and 10–15 years.
  • Repurposing: ~1% of new‑drug development cost (far cheaper and faster).
• Adoption lag: ~17 years for medical innovations to be widely used.
• Sirolimus efficacy: helps ~20–25% of Castleman disease patients.
• Notable quotes:
  • “For a disease that had a 90% fatality rate, to have two patients survive … was pretty remarkable.” — Heather Stone (FDA), on nitroxoline use for balamuthia.
  • “These drugs are already at the pharmacy … and they’re cheap.” — David Fagenbaum (EveryCure).
  • “If you come up with a brand‑new use for an existing pharmaceutical compound, you can get a patent … the problem in practice is that it’s completely unenforceable.” — Chris Snyder (economist, Market Shaping Accelerator).

Takeaways and recommended next steps (for policymakers, clinicians, and the public)

• For policymakers:
  • Consider scalable pull‑funding mechanisms (outcome‑based rewards) to create financial incentives for repurposing generics.
  • Support registries and better data linkage to measure real‑world impact and speed adoption.
• For researchers and funders:
  • Invest more in repurposing research (experts suggested current funding is a fraction of 1% — possibly should be much larger given ROI).
  • Combine AI screening with targeted lab validation and pragmatic trials.
• For clinicians and patients:
  • Contribute to CureID and similar registries to build the evidence base.
  • Stay alert to high‑quality trials showing large benefits for low‑cost interventions (e.g., peri‑tumor lidocaine).
• For philanthropists and foundations:
  • Funding can be catalytic (EveryCure’s $100M shows philanthropic capital can scale programs that government funding has neglected).

Bottom line

Repurposing existing drugs is a scientifically plausible, cost‑effective route to treat many unmet medical needs. The bottleneck is largely economic and institutional: the lack of incentives for companies to run trials on generics and the slow uptake of evidence. Practical solutions discussed include public pull‑funding models that pay for impact, better registries (CureID), AI‑driven screening (EveryCure), and targeted philanthropic and government support to bridge the incentive gap. If implemented at scale, these changes could unlock many low‑cost, life‑saving treatments hiding in plain sight.