AI-Powered Digital Twins Give Clinical Trials a 75-Year Upgrade

The intersection of generative AI and computational medicine is not just a convergence of buzzwords. It signals a paradigm shift in clinical research, and it could be coming in large part through the use of AI-powered digital twins, which in healthcare are simple in theory, but transformative in practice.

[contact-form-7]

Drug development has traditionally been an industry defined by high costs, slow timelines and steep failure rates. Today, that reality is changing. With an explosion of interest in generative AI and computational medicine, digital twins are no longer theoretical. They are a tool increasingly adopted by pharmaceutical sponsors to streamline trials, especially in high-need areas like neuroscience, thanks to their potential to cut trial durations, reduce reliance on placebo groups and fast-track life-saving drugs to market.

“We’re not just tweaking the system,” Jon Walsh, founder and chief scientific officer at Unlearn, explained during a discussion hosted by PYMNTS CEO Karen Webster. “We’re re-architecting it.”

Digital twin technology, Walsh believes, can help level the playing field in healthcare.

“Not all doctors have access to all of the tools and not all doctors have the same capabilities,” Webster noted during the conversation.

Walsh agreed. “This is as much a technical problem as it is a social one. Better infrastructure, better use of medical records, better education for doctors. It’s very dangerous to build tools that doctors can’t understand and can’t trust.”

A New Kind of Patient

The concept of digital twins originated in aerospace and manufacturing, where engineers used simulations to monitor and optimize the performance of physical systems like jet engines. In healthcare, the idea is far more ambitious: model not just a machine, but a human being. This entails integrating multiscale data like genomic, phenotypic, imaging, clinical and behavioral data into a computational construct that can mimic how an individual might progress if they had received a placebo instead of treatment.

“Patients in trials can preferentially receive treatment over a placebo,” Walsh explained. “Typically the way sponsors choose to utilize this would be to reduce the sample size of the placebo arm so you can give more people treatment. That’s more motivating for patients.”

The impact is twofold: trials are faster and more ethical. Patients are less likely to receive an inactive treatment, and sponsors get to market sooner.

According to Walsh, shaving even six months off the clinical development timeline has ripple effects across the healthcare economy: earlier access for patients, longer patent lives for sponsors and lower overall development costs.

But innovation alone isn’t enough to disrupt an industry where regulators rightly demand rigorous validation. Rather than circumvent the rules, Unlearn itself chose to operate within them, a decision Walsh said has paid off.

“More and more we’re seeing that sponsors say, ‘Well, I can see that my peers are doing this,’” he explained. “The concerns have shifted from regulatory risk to operational risk — ‘Can you integrate with the complex machinery of my trial?’”

So far, Unlearn’s strongest traction has come in neuroscience, particularly in Alzheimer’s and ALS, diseases with small patient populations and high mortality rates.

“That’s where we see really the fastest path to making an impact for patients and getting drugs to market faster,” said Walsh.

This is a significant breakthrough for diseases that often struggle to attract funding due to their limited market sizes. As Webster pointed out, ALS only affects around 5,600 people per year. The economics rarely make sense for traditional trials.

Walsh agreed: “Reducing the number of participants required for clinical trials and enabling people to run, in some cases, more heterogeneous trials that are easier to recruit for makes it cheaper for people developing drugs.”

That, in turn, can help encourage innovation in neglected or high-risk areas of medicine.

Asked about the timeline for transformation, Walsh is cautiously optimistic.

“I would say more like 10 years than 75,” a reference to the fact that clinical trials haven’t changed much at all in the 75 years that randomized trials have been a part of the drug development process. “We’ve already seen a fast ramp in adopting AI technologies even over the last three or four years.”

Redesigning the Clinical Trial

One of the most consequential shifts digital twins can impact is their effect on decision making in phase II trials, where the cost-to-benefit ratio is at its lowest.

“Across the board, it’s extremely expensive and the ROI is low,” Walsh explained. “What we think of is using digital twins to help you get to quick decisions that are still confident: yes, this drug doesn’t work, we should stop. Or yes, it does work, and we should accelerate planning for phase III.”

In a field where more than 85% of drugs fail to move from phase I to approval, making better decisions earlier isn’t just smart science — it’s smart business.

The implications go beyond cost and speed.

“If you’ve run many trials in an indication, why are you running so many randomized trials where patients have to donate their time to get a placebo?” Walsh asked. “It’s incumbent on sponsors to make good decisions, but if you can use this kind of technology to run single-arm trials… that’s a pathway.”

Such a shift would require broader regulatory acceptance, but it’s already happening in oncology, where life-threatening conditions and limited treatment options make single-arm designs more common.

“We’re really advocating for looking at this holistically,” Walsh said. “As an entity, we run thousands of clinical trials a year. How do you leverage that data to make better decisions?”

Beyond the trial stage, Walsh sees potential to bring digital twin technology into clinical practice itself. Imagine a future where a patient’s treatment is predicted and prescribed based on an AI model that knows their likely response better than even the most experienced doctor.

“Our long-term mission is to advance AI to eliminate trial and error in medicine,” he said. “Right now, a doctor gives you a sequence of treatments based on guidelines. But wouldn’t it be great if you could get the third-line treatment that’s right for you first?”

 

The post AI-Powered Digital Twins Give Clinical Trials a 75-Year Upgrade appeared first on PYMNTS.com.