A step toward the goal of ending placebo arms
Replacing traditional control arms with digital twins could address the ethical dilemma of administering a placebo to gravely sick patients. Speaking of the study, which was co-authored by Dr Yi-Bin Chen, Director for the Blood and Marrow Transplant Program at Massachusetts General Hospital, Gen Li, Ph.D. notes that the development is “a dream come true for most medical professionals who work with innovative pharmaceuticals,” he said, highlighting the predictive capabilities of the digital twin approach. “We’re allowing you to see your patients before you even start your trial.”
The Bone Marrow Transplantation study emphasizes that the carefully constructed digital twin arms could “eventually replace placebo/control arms to deliver time, cost, and ethical benefits.” In addition, digital twins can “dramatically reduce patient burden,” Li notes. In addition, the technology can also help overcome the persistent challenge of recruiting participants for clinical trials.
Faster, smarter trials
By streamlining the implementation of clinical trials, digital twins could curb both the time and resources required for drug development. “Traditionally, you put together a protocol, start recruiting patients, and then realize there are problems with your design,” Li said. “You then modify your design to get the patients you want. With our approach, you don’t have to do that.”
Phesi’s platform goes beyond simply creating digital twins. Dr. Li explains that its “accurate digital patient profile” can be used to identify “gaps and misalignments” in trial design, leading to more robust and efficient studies. The Bone Marrow Transplantation study noted that the digital twin approach could “objectively improve clinical development planning” and can lead to “protocol design optimization.”
Ensuring accuracy in digital twins
That doesn’t mean the transition is without challenges. Li acknowledges the inherent challenge of working with large, complex datasets. He describes Phesi’s strategy to optimize data quality: “Once we reach the digital twin stage, every single piece of data is human-verified and validated. All source documents are available for anyone who wants to check the quality of the source data we’re using in constructing that digital twin.”
“Our platform is a process that can yield implemented benefits for clinical development organizations along the way, with the end goal of using digital twins to replace external control arms,” Dr. Li explains.
Regulatory considerations for digital twins
While digital twins show promise, regulatory approval remains a crucial step. The Bone Marrow Transplantation study highlights the FDA’s increasing engagement with AI/ML in medicine, as evidenced by its “Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD) Action Plan.” Additionally, the study’s emphasis on using “real-world data” aligns with the growing regulatory trend of valuing real-world evidence in clinical decision-making.
Looking ahead, transparency and trust will be key to the widespread adoption of digital twin technology. “We’ve intentionally designed steps in this process to allow people to see that there are, obviously, flaws, errors, and problems,” Li notes. “This is a necessary component for people to build trust and confidence in the body of evidence we’re using.”
Filed Under: clinical trials, Drug Discovery, machine learning and AI, Regulatory affairs