It can take 10 years and cost pharmaceutical research labs more than $2 billion to develop, trial and receive federal approval for a new drug, according to the Tufts Center for the Study of Drug Development. In times of crisis, however, humanity can’t wait.
Drug repurposing of existing compounds for new medicinal uses could shorten drug discovery timelines. By using artificial intelligence to sift through the multitudes of compounds and their known treatments, researchers could piece together new therapies that save lives.
A Compelling Case
Perhaps the most famous story of drug repurposing comes from the pharmaceutical company Pfizer. In the 1990s, they developed sildenafil, a compound to treat high blood pressure and chest pain due to heart disease. During clinical trials, researchers discovered that the drug was better at inducing erections than treating heart problems, reports Quartz. The little blue pill Viagra was born.
Other lesser known drugs originally developed to treat one ailment were also repurposed for something else. The National Center for Biotechnology Information reports that bupropion, originally used for depression, helped people stop smoking and thalidomide, used to morning sickness, is now used for multiple myeloma.
Harnessing AI Tools
Artificial intelligence algorithms are being used to speed up that search. A range of companies from better known Google and IBM to startups like Insilico Medicine, Recursion Pharmaceuticals, BenevolentAI and Pharnext are harnessing the tools of AI to scrutinize existing drugs as well as anonymous patient data to find patterns that may point to new therapies, reports Fortune. Researchers use AI in a number of different ways for drug repurposing. Algorithms can be trained to read through troves of published research papers to sort through high-resolution medical images, analyze genomic profiles, screen massive chemical libraries, and search the genetic profiles of cells and patient information, for example. The analysis may explore specific interactions between drug molecules and cells to zero in on potential compounds that could show promise for a new disease.
Pharnext, for instance, used AI to dig through millions of records and find three existing drugs — baclofen, a muscle relaxant; naltrexone, used to treat opioid dependence; and sorbitol, a glucose-based laxative — to develop a drug treatment for a rare neurodegenerative condition called Charcot-Marie-Tooth disease (CMT). Because the three drugs were already in use, Pharnext was able skip the Phase I trials that the Food and Drug Administration normally requires to evaluate a drug’s safety and toxicity at different dose levels, according to Fortune. Using the same technique, the company is now trialing a drug for Alzheimer’s and another for amyotrophic lateral sclerosis, or ALS.
Using AI-based drug repurposing has the potential to shave time and money off drug discovery efforts. But progress is still in the early stages. Even if drug candidates can skip Phase I trials, they still need to pass through Phases II through IV. Few new AI-based drug candidates have made it that far, according to Nature. If the FDA’s process for testing and approval could be streamlined, that could change, but overhauling a federal regulatory system takes time and truly speedy drug development is likely years away. At the very least, AI can improve the quality of the data fed into drug discovery models, show where information is lacking and help researchers focus their efforts. Humans, for now, remain on the front lines, racing against the clock to produce therapies that can save lives.
Are you interested in all things related to technology? We are, too. Check out Northrop Grumman career opportunities to see how you can participate in this fascinating time of discovery.