Introduction

Just over a year ago, Joseph Coates faced a life-altering decision: would he prefer to spend his final days at home or in a hospital? At just 37 years old, living in Renton, Washington, Coates was grappling with a rare blood disorder known as POEMS syndrome. This debilitating condition caused him numbness in his extremities, an enlarged heart, and failing kidneys. His deteriorating health had reached a point where he was too weak to undergo a stem cell transplant, one of the few potential life-saving treatments available.

Feeling hopeless, Coates succumbed to despair. "I gave up," he admitted, believing that his fate was sealed. However, his girlfriend, Tara Theobald, refused to accept this grim outlook. In an act of determination, she reached out via email to Dr. David Fajgenbaum, a physician in Philadelphia they had met at a rare disease conference a year prior.

Dr. Fajgenbaum promptly responded with an unconventional treatment proposal—a unique combination of chemotherapy, immunotherapy, and steroids that had never been tried for Coates’s specific condition. Remarkably, Coates showed signs of improvement within a week and was healthy enough for a stem cell transplant within four months. Today, he is in remission, thanks in large part to a treatment plan devised with the assistance of artificial intelligence (AI).

The Role of AI in Drug Repurposing

The application of AI in pharmaceutical research is transforming how scientists approach drug repurposing—the process of finding new therapeutic uses for existing medications. This method is not entirely new, but the use of machine learning is dramatically accelerating the ability to discover effective treatments for rare diseases. Institutions like Dr. Fajgenbaum's team at the University of Pennsylvania are leading efforts to uncover novel applications for drugs related to difficult-to-treat conditions, from aggressive cancers to rare neurological disorders.

Donald C. Lo, former head of therapeutic development at the National Center for Advancing Translational Sciences (NCATS), stated that there's an abundance of previously overlooked medication that could be repurposed for various diseases. "We've lacked a systematic approach to explore this treasure trove," he explained, pointing out that repurposed drugs could be a game-changer, particularly given that many of these medications are already approved and readily available.

Challenges in Treating Rare Diseases

Rare diseases, classified by the National Institutes of Health as those affecting less than 200,000 people in the U.S., number in the thousands, impacting millions globally. Alarmingly, over 90% of these diseases currently have no approved treatment. Pharmaceutical companies often shy away from investing in research and development for new drugs targeting small patient populations due to limited financial incentives.

The Potential of Drug Repurposing with AI

Dr. Marinka Zitnik, an associate professor at Harvard Medical School, highlighted the encouraging potential of drug repurposing using AI, augmenting previous discovery methods. "AI puts rocket boosters on that," she noted, emphasizing its role in rapid data analysis and research advancements.

Drug repurposing is not a novel concept in the pharmaceutical industry. Well-known examples include minoxidil, initially developed for hypertension but later used for hair loss, and Viagra, originally a cardiac medication that became synonymous with erectile dysfunction treatment. Even the diabetes drug semaglutide has gained popularity for its weight loss effects.

Personal Stories of Drug Repurposing

Dr. Fajgenbaum's encounter with drug repurposing began as a personal mission to save his life. Diagnosed with a rare form of Castleman disease at 25, he turned to existing drugs after traditional treatments failed him. His perseverance led him to sirolimus, a generic medication initially intended to prevent kidney transplant rejection, which successfully kept his Castleman disease in remission for over a decade.

Establishing a nonprofit called Every Cure, Dr. Fajgenbaum aims to harness machine learning for broader research into drug options. His ambition is complemented by global efforts in drug repurposing technologies at prominent institutions such as Stanford and several research facilities in Japan and China.

Promising Results from AI-Driven Models

AI-driven models are already yielding promising results. For instance, medications like amphetamines used for ADHD have benefitted patients with rare genetic disorders. Similarly, Parkinson's drugs have shown to improve the mobility of patients with related conditions. These discoveries illuminate the multidimensional properties of medications often dismissed as side effects, potentially unlocking new treatment avenues.

At the University of Pennsylvania, Dr. Fajgenbaum's AI platform evaluates approximately 4,000 drugs against 18,500 diseases, assigning efficacy scores to each treatment based on the probability of success. Researchers then delve into the most promising candidates to conduct tests or partner with physicians to deploy these drugs in clinical scenarios.

Financial Barriers to Drug Repurposing

While pharmaceutical companies are keen to develop entirely new drugs, the discovery of new uses for low-cost generic drugs remains less appealing from a profit standpoint. Once a drug's patent expires, competition can significantly devalue it, making the pursuit of repurposing less attractive financially.

Support for Drug Repurposing Initiatives

To bolster these initiatives, Every Cure is backed by over $100 million from notable sources, including TED’s Audacious Project. This funding will help support clinical trials aimed at validating the efficacy of repurposed drugs.

"This is one example of AI that we should embrace with open arms," remarked Dr. Grant Mitchell, a co-founder of Every Cure. "It holds the promise of helping countless individuals."

Success Stories and Future Perspectives

In a striking case, a patient who had nearly exhausted treatment options responded positively to adalimumab, a drug previously thought to be ineffective for his condition. His subsequent recovery was even published in the prestigious New England Journal of Medicine.

Last month, a revitalized Joseph Coates returned to Philadelphia with his girlfriend to express gratitude to Dr. Fajgenbaum. The transformation in Coates was evident; he was in excellent health, features a strong physique, and is now focused on enjoying life to the fullest.

The world of drug discovery and repurposing is on the cusp of a revolution, with AI presenting unprecedented opportunities for finding effective treatments and bringing hope to those grappling with rare and challenging diseases.