Researchers from Peking Union Medical College Hospital and artificial intelligence (AI) drug developer Insilico Medicine published phase 2a clinical trial data for a small-molecule treatment for idiopathic pulmonary fibrosis (IPF) in Nature Medicine magazine on Tuesday. Data demonstrates preliminary safety and efficacy of the drug candidate, making it the fastest-progressing AI-discovered drug worldwide to date, according to a report by The Paper.
IPF is a chronic, progressive lung disease characterized by irreversible decline in lung function, affecting an estimated 5 million people worldwide. The median survival is only three to four years. Existing treatments can slow the disease's progression but cannot halt or reverse it, according to The Paper. The molecule tested in this trial, Rentosertib, targets TNIK — a novel site identified with the help of AI, as research group say in the Nature Medicine.
In the development of new drugs, a "target" is like a lock that a drug — the key — must fit. Traditionally, identifying the correct target is time-consuming and depends heavily on trials and errors. In the case of Rentosertib, AI helped researchers pinpoint a previously overlooked target, significantly enhancing discovery efficiency, Chen Jing, a vice president of the Technology and Strategy Research Institute, told the Global Times on Wednesday.
"AI has already made major breakthroughs in drug-related fields," Chen said. "Its success in identifying a novel target represents a natural extension of that capability into basic research. The core competency is present — with ongoing data accumulation and training, large-scale application and major results are only a matter of time."
While AI has long been expected to accelerate drug development by efficiently screening compounds, simulating mechanisms and supporting clinical trial design, no AI-discovered drugs had advanced beyond phase 1 trials until now. Rentosertib's progression into phase 2a marks a significant milestone, demonstrating AI's tangible impact on real-world pharmaceutical innovation, per the report.
Clinical observations showed that Rentosertib had demonstrated manageable safety and tolerability, laying the groundwork for larger and longer-term trials. The experience with Rentosertib highlights the transformative potential of AI in drug development and offers a valuable reference for more efficient and innovative approaches, said Dr. Alex Zhavoronkov, founder and CEO of Insilico Medicine, the report noted.
"AI is making steady progress in fields such as medical imaging, clinical decision support, and personalized health monitoring," Chen said. "But in drug development, it still faces major challenges — from limited access to high-quality data and poor model interpretability to high computational costs and a shortage of interdisciplinary talent. Intellectual property and regulatory frameworks also need to evolve. If these issues are properly addressed, AI has the potential to truly advance inclusive and accessible healthcare forward."
AI applications in China's healthcare sector are expected to expand in scale and diversify across more use cases in the future, expert said. China's medical AI market is projected to reach $13.4 billion by 2026, with an annual growth rate exceeding 40 percent, according to an IDC report, quoted by the Xinhua News Agency.
京公網安備 11010202009201號