The Nonprofit Organizations Identifying New Therapeutic Uses for Existing Drugs

People in the U.S. have access to thousands of FDA-approved drugs, yet millions live with conditions that have no registered treatments and are faced with shortened life spans as a result.

While pharmaceutical companies invest billions of dollars and years of time developing new medications, the FDA typically only approves around 50 new treatments per year, which is inadequate to address the scale of the problem.

Drug repurposing offers a potential solution, particularly for patients with rare diseases, who are consistently overlooked by pharmaceutical companies because developing a treatment for a small number of patients is not financially viable.

Drug repurposing is the process of identifying new therapeutic uses for existing or investigational drugs. Although many repurposed drugs have been discovered by chance, including the well-known example of Viagra, researchers more commonly combine experimental and computational approaches to determine whether existing medicines can be redirected to treat other diseases.

Using an already approved drug to treat a currently untreatable disease is appealing because the safety, efficacy, and toxicity are already known. Having this information saves time and money—researchers estimate that repurposed drugs are generally approved 3–12 years sooner than novel treatments and cost 50–60% less. Furthermore, up to 75% of repurposed drugs gain market approval compared with less than 10% of new drug applications.

Despite these benefits, the focus within the pharmaceutical industry remains on new drugs, leaving not-for-profit companies to tackle the problem.

Creating a library of drugs for repurposing

In 2015, researchers at the Broad Institute of MIT and Harvard at Cambridge, Massachusetts were looking at how knocking out certain genes affected disease phenotypes. They also wanted to know how different drugs altered these phenotypes, but were surprised to learn that there was no single place where all the drugs were assembled for research purposes.

“We saw this as a missing opportunity,” said Alex Burgin, PhD, senior director of the Center for the Development of Therapeutics (CDoT) at the Broad Institute. “As a nonprofit, we said we’ve got to do this.”

In 2016, a philanthropic donor provided a multi-year gift that allowed the Broad Institute to establish its Drug Repurposing Hub—a publicly-available, curated and annotated library of FDA-approved drugs, clinical trial drugs, and pre-clinical compounds.

Burgin said that creating the hub was a much larger endeavor than initially thought; many of the compounds chosen for the library could not be bought and had to be synthesized from scratch, whereas others were extremely expensive.

To make the Drug Repurposing Hub useful, the compounds also had to be extensively annotated. This was time consuming because the information was not aggregated in one place.

At present, the library holds 6,801 unique drug compounds that have 670 therapeutic indications and 2,183 protein targets. Each compound is annotated with information on the mechanism of action, chemical structure, target proteins, and clinical indications. The level of detail within each annotation is what makes the Drug Repurposing Hub unique among other repurposing libraries that have since been created.

“There are other repurposing libraries out there,” said Jaime Cheah, PhD, director of collaborative screening at the CDoT and one of the managers of the Drug Repurposing Hub at the Broad Institute. “The power of ours is the heavy annotation.”

The ReFRAME library, curated by Calibr, a nonprofit drug discovery division of Scripps Research, the NIH’s National Center for Advancing Translational Sciences Pharmaceutical Collection, and REMEDi4ALL, a European platform for drug repurposing, are some of the other tools available to researchers.

Cheah stresses that as not-for-profit entities, they try not to compete. “The idea is we should be harnessing our power together, rather than being competitors,” she remarked. Burgin added: “I feel very lucky to be at the Broad, where we assembled the first library. I feel like we showed the world you need to do this, and now people in government and companies are doing it.”

The Drug Repurposing Hub creates high-throughput screening-ready plates that researchers can use to see if any of the drugs within the library are effective against their target protein or disease model system. The screening can be done by the researchers themselves or in collaboration with Cheah and team at the CDoT, who can help develop suitable biologic assays to test the compounds.

The primary screen will bring up an initial “hit list” with which the researchers can use in-depth annotation to check for connections within the list. If there are several hits from a single drug class, then a secondary screen can be used to expand the number of compounds tested within that class. Alternatively, the hits may have similar chemical structures, in which case Cheah and team will look for other similar chemicals that are not already in their library. Eventually, over 7,000 compounds are whittled down to a lead hit that can move toward clinical testing.

One of the biggest success stories from the Drug Repurposing Hub is the discovery of a potential treatment for a rare genetic disease that leads to kidney failure. MUC1 kidney disease is caused by the intracellular accumulation of misfolded MUC1 protein due to a mutation in the MUC1 gene. Working with the CDoT, Anna Greka, MD, PhD, director of the Kidney Disease Initiative at the Broad Institute, and colleagues found a compound that was originally developed for something completely different. Fortunately, it had the ability to degrade and remove the misfolded MUC1 protein.

It was not initially clear how the compound, named BRD4780, was acting, but the researchers ultimately discovered that it was targeting a protein called TMED9. BRD4780 works by preventing misfolded MUC1 from getting trapped in TMED9 cargo receptor-containing vesicles and re-routing it for lysosomal degradation. The role TMED9 cargo receptors play in how cells deal with toxic misfolded proteins was previously unknown, but the discovery has important implications for other diseases like retinitis pigmentosa, which is also caused by misfolded proteins. Greka and team are now working toward bringing a TMED9-targeting compound to clinical trials.

Burgin says that this case highlights what a powerful research tool the Drug Repurposing Hub is. Finding previously unknown targets leads to information about biological pathways that the researchers might have had no prior understanding of. That knowledge could then help with the development of a drug in future.

“That’s why I like to remind everybody, even if you don’t repurpose a drug, you learn about the disease, you learn about the biology, you learn which pathways are involved,” he remarked.

Harnessing the power of AI to connect drugs
with diseases

Another not-for-profit organization hoping to unlock connections between existing drugs and currently untreatable diseases is Every Cure. For Every Cure, the mission is personal. In 2010, the company’s co-founder and president, David Fajgenbaum, MD, was a healthy third-year medical student when he became critically ill with Castleman disease, a rare condition in which benign growths form in lymph node tissue. The disease can weaken the immune system and cause it to attack and shut down vital organs. After 6 months in the hospital, Fajgenbaum was told by doctors that there was nothing more they could do—he was fighting for his life.

In a last-ditch attempt to save him, Fajgenbaum was given seven-agent chemotherapy and survived. However, he then relapsed and nearly died four more times in following 3 years. With his medical background, he realized that the only way to find a treatment would be to try and repurpose a drug that already existed, rationalizing that many diseases that share the same underlying mechanisms can be treated with the same drug.

He started experimenting on his own blood cells, eventually finding the immunosuppressant drug sirolimus, also known as rapamycin, to be a potential candidate. He tested the drug on himself and has now been in remission for 10 years while using sirolimus.

By Fajgenbaum’s side throughout this period was his university roommate, and now CEO and co-founder of Every Cure, Grant Mitchell, MD. He said that after the initial shock of Fajgenbaum being told there were no further treatments, they thought that maybe they could do something about it.

“We said don’t worry, ‘they’re’ working on it, with this nebulous concept of ‘they,’ like a medical factory in the sky that’s tinkering away, just in the way that you expect it to. But we started researching and looking around, and there really was no ‘they.’ Then we had this moment where we looked at each other and said we could be ‘they,’ why not?”

Finding sirolimus lit a spark in Fajgenbaum and Mitchell to do more. “We looked at the situation and realized that particular drug had been sitting there for 30 years, just waiting while 1,000s of people died from this disease, when all they had to do was take this little pill,” said Mitchell. Together the friends questioned how many more drugs were waiting to save lives and why were they were not being researched.

Despite regularly discussing how they could address the problem on a larger scale, it wasn’t until 2018 when advances in artificial intelligence (AI) and data science made it a realistic prospect. By this time, Fajgenbaum had become an expert in repurposing drugs in the lab. Mitchell had been working in machine learning, leading teams that were pioneering the use of large medical record databases to inform drug development.

In 2022, with their third co-founder, rare disease specialist Tracey Sikora, they launched Every Cure as a nonprofit initiative that would use an innovative AI approach to find the most promising drug-repurposing opportunities across all 3000+ unique FDA-approved drugs and all 22,000+ diseases.

The technology is based on biomedical knowledge graphs that integrate all available data on the drugs, diseases, and targets and builds connections between them. Predictive algorithms can then harness the knowledge within the graphs to find new links between drugs and diseases that have yet to be tested. All 66 million possible combinations of drug and disease are given a normalized score from 0 to 1 that allows researchers accessing the data to focus on which combinations are most likely to work.

Every Cure will also test these relationships in house, prioritizing not only drug–disease pairs that have the highest scores, but also those that are going to relieve the most suffering and have the greatest impact on the biggest number of people.

The first life to be saved with a repurposed drug discovered by Every Cure’s AI platform was reported in May 2023. A 50-year-old patient, also with life-threatening Castleman disease, was given adalimumab as a treatment of last resort following its identification by AI-guided discovery. The monoclonal antibody adalimumab gained FDA-approval in 2002 and is typically used to treat severe inflammatory conditions such as rheumatoid arthritis and Crohn’s disease. It had never been used for Castleman disease before, yet within a few days of taking it, the patient’s organs regained function, his symptoms subsided, and he went into remission.

Other promising candidates identified by Every Cure for repurposing include metreleptin for anorexia, folinic acid for autism spectrum disorder, anakinra for sepsis, and bosutinib for the motor neuron disease ALS. Investigating these connections will allow the company to further improve the predictions generated by the algorithm.

“The outcome of clinically testing repurposed drugs flows back into the algorithm, [which] then validates, refines, and improves it because it learns whether it is making accurate predictions or not,” said Mitchell. “So not only does this approach, in this nonprofit way, help patients the fastest, it also advances the field of data-driven drug discovery, which is an amazing thing that we didn’t set out to do but are now in a position to do.”

He points out that this contrasts with how AI-driven drug discovery companies can work in the private sector, because once their platform identifies a candidate, they have to shift focus and finance to developing the drug without knowing if the AI platform was accurate.

Mitchell hopes that within five years, Every Cure’s platform will be “the most comprehensive tool available for drug repurposing research.” He says it “should be an engine that generates life-saving opportunities every year, indefinitely,” and believes that if the company can use it to repurpose one to five drugs per year, it could be almost as big as some of the largest pharmaceutical companies.

He is also extremely proud of the team that Every Cure have built and feels that the potential for huge societal impact attracts “people that are willing to dive in on this really hard problem and try and solve it for the for the sake of everyone.”

Working toward a shared goal

Although Every Cure clearly has an excellent in-house team, bringing together the information required to create the AI platform is a huge undertaking that requires collaboration. At present, Every Cure researchers are integrating around 185 unique publicly available data sets into their knowledge graph and then enhancing it with additional proprietary data sets, such as large collections of medical literature, and multiomics information from research institutions building their own patient registries.

Every Cure’s website lists more than 30 key partners, including the NIH and the Broad Institute. In April this year, the AI-based drug development company BioPhy announced a partnership with Every Cure, giving Every Cure access to BioPhy’s predictive AI engine BioLogicAI.

BioLogicAI was also built using knowledge graphs and is designed to virtually guide promising therapeutics through clinical trial phases by continually evaluating the likelihood of success based on multiple inputs such as the mechanism of action and trial design.

Dave Latshaw II, PhD, CEO and co-founder of BioPhy, said that it makes sense for the two companies to work together as they have similar goals but are not in competition. BioPhy has been up and running for five years and Latshaw says that the company’s knowledge and experience will help Every Cure “skip some of the pitfalls they haven’t run in to yet.” In turn, as Every Cure’s system grows and improves, that knowledge can be used to advance BioPhy’s systems.

There is likely to be increasing focus on drug repurposing in the next 5 to 10 years. “While there’s all sorts of interesting new therapeutic modalities being developed, it’s also getting harder and harder to develop new drugs,” says Mitchell. He therefore hopes that regulators and legislators think about ways to incentivize research to identify new uses for existing drugs. “That would be a real benefit to the industry, and it’d be a real benefit to patients,” he says.

Read more:

1. Novel Drug Approvals for 2024, FDA.

2. Hernandez JJ, Pryszlak M, Smith L, et al. Giving drugs a second chance: overcoming regulatory and financial hurdles in repurposing approved drugs as cancer therapeutics. Front Oncol 2017; 7: 273.

3. Drug repurposing: Approaches and methods

4. Drug Repurposing Hub

5. Dvela-Levitt M, Kost-Alimova M, Emani M, et al. Small molecule targets TMED9 and promotes lysosomal degradation to reverse proteinopathy. Cell 2019; 178: 521–535.e23

6. Every Cure

7. BioPhy

Laura Cowen is a freelance medical journalist who has been covering healthcare news for over 10 years. Her main specialties are oncology and diabetes, but she has written about subjects ranging from cardiology to ophthalmology and is particularly interested in infectious diseases and public health.