The European Medicines Agency (EMA) has qualified the artificial intelligence (AI)-driven approach of Unlearn for conducting smaller and quicker clinical trials.
In a final favourable qualification opinion, the agency offered a regulatory framework for the usage of the company’s TwinRCT solution in Phase II and III trials.
The three-step PROCOVA technique serves as the base for TwinRCTs and defines how to employ patient-specific prognostic scores obtained from digital twins to lower trial sizes while minimising the rate of Type-1 errors.
TwinRCTs merge AI, predictive digital twins, and new statistical approaches to carry out trials with fewer patients in the control arm.
This is considered to be the first time that a regulatory agency backed a machine learning-based approach for cutting down sample size in pivotal trials.
The EMA also noted in the final opinion that the Committee for Medicinal Products for Human Use (CHMP) ‘qualifies PROCOVA’ and that the proposed methods could aid gains in power and/or reductions in sample size in Phase II and III trials with continuous outcomes.
Unlearn was established to use AI to remove trial and error in medicine. Its TwinRCT solution now decreases control arm sizes by up to 35%, allowing more trial subjects to receive the experimental therapy.
TwinRCTs substantially reduce timelines, aiding in expediting the pace at which new lifesaving treatments reach patients.
Unlearn founder and CEO Charles Fisher said: “The EMA’s adoption of our novel PROCOVA procedure using digital twins is a historic milestone on the path toward achieving that ultimate goal.”
“PROCOVA is a key part of our mission to innovate in regulatory-acceptable ways.
“With Unlearn’s guidance, our partners now have a framework that enables them to confidently implement TwinRCTs in alignment with the EMA’s rigorous standards.”
In April this year, the company raised Series B funds worth $50m led by Insight Partners to progress the usage of TwinRCTs in trials.