Zu-shan Xu, F. Ren, Ping Wang, Jie Cao, Chunting Tan, Dedong Ma, Li Zhao, Jinghong Dai et al.
Phase 2a trial of rentosertib, a TNIK-targeting drug for pulmonary fibrosis discovered by generative AI, confirmed safety and showed signs of lung function improvement.
Idiopathic pulmonary fibrosis (IPF) is a progressive fatal disease with no current therapies that reverse disease progression. AI-based drug discovery has raised high expectations, but few candidates have reached clinical trials.
Using generative AI, the small molecule rentosertib was designed to inhibit TNIK (a novel target). A 12-week randomized, double-blind, placebo-controlled trial was conducted in 71 patients (30 mg QD, 30 mg BID, 60 mg QD, placebo). Primary endpoint was safety; secondary endpoints included lung function and symptom changes.
Safety was acceptable across all dose groups with low serious adverse event rates. In the 60 mg QD group, forced vital capacity (FVC) increased by a mean of +98.4 ml (vs. -20.3 ml in placebo), indicating an efficacy signal. This is the first evidence of an AI-generated drug showing disease modification potential in a randomized trial, marking a milestone in AI-driven drug development.