Dr. David Fajgenbaum on Huberman Lab: How Old Drugs Cure

There are roughly 4,000 FDA‑approved drugs for about 4,000 diseases, but ~18,000 human diseases exist and most drugs act on 20–30 targets or pathways, not just the one they’re approved for. Because 80% of these drugs are now generic and no longer profitable to develop, almost no one is systematically looking for new indications. This creates a massive missed opportunity where effective treatments sit unrecognized on pharmacy shelves while patients are told their diseases are ‘untreatable’.

Once a drug goes off patent, no single company has sufficient financial incentive to invest in large trials, guideline lobbying, and physician education for new uses. Findings like lidocaine injection around breast tumors reducing mortality ~29% at five years (large RCT in India) or aspirin lowering colon cancer recurrence in certain genotypes remain poorly implemented. It isn’t that companies are hiding these drugs; the system simply doesn’t reward anyone for doing the hard push once profit margins are pennies per dose.

Rather than hunting disease by disease, Every Cure builds biomedical knowledge graphs and machine learning models to score how likely every drug is to help every disease. They then triage the highest‑ranking pairs for lab work, clinical data mining, and, where justified, clinical trials. This has already yielded nine active programs and 14 repurposed drug successes (e.g., sirolimus for Castleman disease, PD‑1 inhibitors for angiosarcoma, DFMO for Bachmann‑Bupp syndrome, TNF inhibitors for DADA2). The aim is a “master list” of validated new uses that can be pushed into practice ahead of crisis moments.

Fajgenbaum stresses three concrete steps: (1) connect with the disease-specific organization (e.g., Castleman Disease Collaborative Network, ALS Association, Michael J. Fox Foundation) which often aggregates global treatment experiments; (2) identify and, if possible, see the true world expert for the condition; (3) ask structured questions such as, “What are the standard options? Are any drugs being used elsewhere or off‑label? Is there emerging evidence for other treatments?” In some cases, simple PubMed searches have uncovered life‑saving options that were sitting untried for years.

Many of Fajgenbaum’s and Every Cure’s early wins were Hail Mary off‑label tries—e.g., PD‑1 inhibitors for angiosarcoma based on one 2013 paper, myeloma drugs for POEMS syndrome, or high‑dose sirolimus for his own Castleman disease. These successes prove the concept but also highlight risk: off‑label drugs can cause serious side effects or death, and a single bad outcome can set back entire therapeutic areas. Thus, Every Cure emphasizes rigorous evidence review, mechanistic plausibility, and physician-patient decision-making rather than speculative one-off experiments.