Digital Clinical Trials Show Promise in Accelerating Drug Repurposing for Heart Failure
A new study demonstrates the potential of using artificial intelligence and existing patient data to predict drug effectiveness, offering a faster and more cost-effective alternative to traditional clinical trials.
The pharmaceutical industry is increasingly exploring innovative methods to streamline drug development, and a recent study published in npj Digital Medicine in May offers a compelling glimpse into the future of digital clinical trials. Researchers at the mayo Clinic leveraged data from 59,000 patient electronic health records (EHRs),combined with refined computer modeling,to assess the potential of 17 existing drugs in treating heart failure.
Trials achieved an remarkable 89% accuracy in predicting whether drugs could improve key indicators of heart failure, according to Nansu zong, a biomedical informatician at Mayo Clinic and lead author of the study. “We found our digital clinical trials predicted whether or not the drugs could improve several heart failure prognostic markers with about 89% accuracy,” Zong stated.
This technology isn’t designed to replace traditional trials,but rather to act as a powerful screening tool. “I don’t think our intention will be to replace the actual clinical trial but to provide some signals so that it will be more effective and more efficient,” explained a senior official at Mayo Clinic Platform Informatics.The researchers envision a future where this model helps prioritize which drugs warrant further examination for new applications.
The team’s “ultimate goal” is to develop a predictive model capable of assessing the efficacy of any drug against a given disease, not just determining if it works, but how well it will work. This would allow researchers to make more informed decisions about whether to invest in costly clinical trials. “If you are low risk, low probability to achieve your expectation, maybe you’re going to change your drugs or you’re going to change your trials,” Zong added.
Expanding Access and Reducing Bias
Beyond cost savings, emulation offers the potential to include a more diverse patient population in drug research. Traditional clinical trials often struggle to represent older adults, minorities, and individuals with multiple health conditions (comorbidities). Emulation, by utilizing real-world data, can overcome these limitations and provide insights into drug effectiveness across a broader spectrum of patients.
The Mayo Clinic study distinguished itself from previous emulation efforts by incorporating drug-target prediction – an analysis that assesses the likelihood of a drug interacting with specific genes. this approach proved more accurate than relying solely on EHR data, achieving approximately 70% predictive accuracy with EHR-based emulation alone. However, researchers cautioned that it is “premature” to draw definitive conclusions about the method’s overall effectiveness at this stage.
Implications for Future Research
Experts in the field are optimistic about the potential of this technology. “The research offers a great direction for helping to design more targeted trials,” noted Jimeng Sun,a professor at the University of Illinois Urbana-Champaign and co-founder of Keiji AI,a generative AI platform for clinical research. Sun believes similar approaches could be applied to other complex diseases, such as oncology, provided sufficient EHR data is available.
However, challenges remain. Outcome measures must be consistently recorded in EHRs for the model to function effectively. Diseases with slow progression or unclear endpoints, like Alzheimer’s disease, may prove tough to assess using this method.
The mayo Clinic team is already exploring ways to expand the framework, including the use of artificial intelligence and synthetic data to evaluate new, untested drug candidates. “We are currently investigating the use of advanced simulation technologies to generate synthetic data, which may extend the framework to evaluating new drug candidates,” Zong said.This ongoing research signals a meaningful shift in how drugs are developed and tested, possibly ushering in an era of faster, more efficient, and more inclusive pharmaceutical innovation.
