The NYTimes has an excellent piece by Kate Morgan on AI discovering new uses for old drugs:
A little over a year ago, Joseph Coates was told there was only one thing left to decide. Did he want to die at home, or in the hospital?
Coates, then 37 and living in Renton, Wash., was barely conscious. For months, he had been battling a rare blood disorder called POEMS syndrome, which had left him with numb hands and feet, an enlarged heart and failing kidneys. Every few days, doctors needed to drain liters of fluid from his abdomen. He became too sick to receive a stem cell transplant — one of the only treatments that could have put him into remission.
“I gave up,” he said. “I just thought the end was inevitable.”
But Coates’s girlfriend, Tara Theobald, wasn’t ready to quit. So she sent an email begging for help to a doctor in Philadelphia named David Fajgenbaum, whom the couple met a year earlier at a rare disease summit.
By the next morning, Dr. Fajgenbaum had replied, suggesting an unconventional combination of chemotherapy, immunotherapy and steroids previously untested as a treatment for Coates’s disorder.
Within a week, Coates was responding to treatment. In four months, he was healthy enough for a stem cell transplant. Today, he’s in remission.
The lifesaving drug regimen wasn’t thought up by the doctor, or any person. It had been spit out by an artificial intelligence model.
AI is excellent at combing through large amounts of data to find surprising connections.
Discovering new uses for old drugs has some big advantages and one disadvantage. A big advantage is that once a drug has been approved for some use it can be prescribed for any use–thus new uses of old drugs do not have to go through the lengthy and arduous FDA approval procedures. In essence, off-label uses have been safety-tested but not FDA efficacy-tested in the new use. I use this fact about off-label prescribing to evaluate the FDA. During COVID, for example, the British Recovery trial, discovered that the common drug, dexamethasone could reduce mortality by up to one-third in hospitalized patients on oxygen support that knowledge was immediately applied, saving millions of lives worldwide:
Within hours, the result was breaking news across the world and hospitals were adopting the drug into the standard care given to all patients with COVID-19. In the nine months following the discovery, dexamethasone saved an estimated one million lives worldwide.
New uses for old drugs are typically unpatentable, which helps keep them cheap—but the disadvantage is that this also weakens private incentives to discover them. While FDA trials for these new uses are often unnecessary, making development costs much lower, the lack of strong market protection can still deter investment. The FDA offers some limited exclusivity through programs like 505(b)(2), which grants temporary protection for new clinical trials or safety and efficacy data. These programs are hard to calibrate—balancing cost and reward is difficult—but likely provide some net benefits.
The NIH should continue prioritizing research into unpatentable treatments, as this is where the market is most challenged. More broadly, research on novel mechanisms to support non-patentable innovations is valuable. That said, I’m not overly concerned about under-investment in repurposing old drugs, especially as AI further reduces the cost of discovery.
The post AI Discovers New Uses for Old Drugs appeared first on Marginal REVOLUTION.
