Just over a year ago, doctors confronted then 37-year-old Joseph Coates with a daunting decision: did he want to spend his final days in the hospital or die at home?
Coates was severely ill with POEMS syndrome, leading to numbness in his extremities, an enlarged heart and failing kidneys. Every few days, his doctors needed to drain excess fluid from his abdomen; and he was too sick for a potentially life-saving stem cell transplant.
“I gave up,” Coates shared with The New York Times. “I just thought the end was inevitable.”
However, his girlfriend, Tara Theobald, refused to give up. Instead, she reached out to Dr. David Fajgenbaum, a physician in Philadelphia whom they had met at a rare disease conference a year prior. He responded promptly and, in collaboration with the University of Pennsylvania, provided an innovative treatment plan that combined chemotherapy, immunotherapy and steroids – methods that had not previously been used for Coates’s condition.
Within a week, Coates showed improvement, and after four months, he was healthy enough for a stem cell transplant.
He is now in remission.
This life-saving drug treatment wasn’t created by Fajgenbaum, or any other doctor for that matter. It was generated by an artificial intelligence program.
ChatGPT became wildly popular in late 2022, marking the start of a new era for artificial intelligence. Since then, AI has quickly advanced and became a major topic in important technology announcements for both hardware and software.
Still, the growing interest in AI has led to serious discussions about its safety and ethics. Supporters believe it could bring about new innovations that go beyond what humans can do. But at the same time, critics worry that developing AI without proper control could have catastrophic consequences.
Although the tech world hasn’t created an all-powerful AI yet, current models are demonstrating impressive skills, and are being successfully used in healthcare in ways that are improving how medicines are used to enhance patient care.
A prime example of this is how AI models are being implemented to help find new uses for existing drugs. These models take drugs that have already been approved for specific diseases and explore new ways to utilize them for rare conditions that don’t have good treatment options.
In this article, #AAMC24 speaker @DavidFajgenbaum discusses the potential of machine learning in drug repurposing. His journey, from curing his own rare disease to empowering others, demonstrates the hope AI can bring to patients. https://t.co/XOPdARLq4s
— AAMC (@AAMCtoday) March 22, 2025
Drug repurposing – the practice of using already available medications for new treatments – is not a new concept. For instance, Minoxidil was first created to help with high blood pressure, but now it’s used to treat hair loss. Viagra was initially designed for heart issues, but it is now mainly used to treat erectile dysfunction. Semaglutide, a medication for diabetes sold under brands like Ozempic and Wegovy, has gained popularity for helping people lose weight.
Still, machine learning is speeding up this process and could offer more options for people with rare diseases, limited treatments or who simply don’t have time on their side.
Thanks to advanced technology, existing drugs are being quickly modified to treat serious problems like rare and aggressive cancers, life-threatening inflammatory diseases and complex neurological issues. And often, these treatments are showing good results.
Although there are only a few success stories at the moment, they’ve made researchers wonder: How many other cures might be right in front of us?
There is a “treasure trove of medicine that could be used for so many other diseases. We just didn’t have a systematic way of looking at it,” Donald C. Lo, the former head of therapeutic development at the National Center for Advancing Translational Sciences and a scientific lead at Remedi4All, a group focused on drug repurposing, tells The Times.
“It’s essentially almost silly not to try this, because these drugs are already approved. You can already buy them at the pharmacy.”
According to reporting by The Times, using this kind of AI could help millions of people around the globe who are dealing with rare and hard-to-treat illnesses, even if these diseases only affect small amounts of people.
The National Institutes of Health (NIH) defines rare diseases as those that impact fewer than 200,000 individuals in the United States. Unfortunately, these numbers don’t usually incentivize pharmaceutical companies to develop new treatments for these diseases. They argue that repurposing existing medications is not very lucrative. Plus, while artificial intelligence is aiding in the discovery of new uses for approved drugs, the NIH notes that companies frequently lack the motivation to pursue these opportunities, as drug patents typically last about 20 years.
Experts point out that one of the best benefits of using AI for drug repurposing is its potential to streamline tasks that humans are already capable of performing.
Chris Smith, a senior writer at Boy Genius Report, an online resource for technology and science, explains that while it is possible to look online for existing medications that could address a rare disease by analyzing their side effects from prior uses, this method can be quite time-consuming, and patients may not have the luxury of waiting for answers.
“Essentially, you’d be looking for side effects in patients who take those drugs for the condition the medicine is effective for,” he writes. “But it would take humans a lot of work to comb the available literature.”
Smith also points out that by utilizing AI models that are specifically designed for the medical field, the process can be greatly expedited, allowing for the discovery of potential treatments that doctors might miss.
Ultimately, it’s crucial to remember that doctors are the ones who need to check and decide if they should use the drug combinations suggested by artificial intelligence. Fortunately, in Coates’s situation, they opted for the AI-recommended treatment, which worked well; and only a year after his near-death crisis, he is feeling much better.
Last month, Coates and his girlfriend visited Dr. Fajgenbaum in Philadelphia to thank him for his help. Coates appeared healthy, smiling, and had even put on muscle since their last appointment.
Although he had slightly injured his ankle while exercising that morning, he reported feeling “just fine.”
