Rentosertib: Up to -6 Years on Blood Age Clocks

A study involving 42 patients with IPF reports a decrease in predicted biological ages among those treated with rentosertib, a candidate designed by AI. Specialists acknowledge a signal but point out the small sample size and the lack of trials in healthy volunteers. The drug is already in phase III for IPF.

Methodological Limitations and an Inconclusive Trial

Rentosertib is currently being evaluated in phase III for idiopathic pulmonary fibrosis, the final step before potential approval. Eric Topol considers the drug promising while reminding that there is no definitive trial to draw conclusions. Vadim Gladyshev highlights the small sample size and the variable reliability of biological clocks, while describing this study as the first to clearly show a reduction in predicted biological age. However, he believes a trial in healthy individuals is necessary, as the results may only pertain to patients with pulmonary disease. Experts, more broadly, commend the encouraging results but point to the small cohort size and the absence of tests in healthy volunteers. According to the company, the observed correlations indicate potential without constituting proof.

Six AI Clocks Converge on a Lower Predicted Age

Six independent aging clocks, developed by teams at Harvard, Oxford, Beijing, and Insilico, were used to analyze blood samples from a trial involving 42 patients. These AI models, which estimate biological age based on blood proteins, all predicted a lower age for those treated compared to the placebo group. The most pronounced effect was observed by the fourth week, with a decrease of three to four years, one model even suggesting a reduction of up to six years. These results indicate that the protein profiles have evolved towards a younger interpreted age, without demonstrating a measurable rejuvenation of the patients.

Doses That Dissociate Pulmonary Function and Predicted Age

Younger blood values could be explained by an improvement in pulmonary function and a decrease in systemic stress. However, the dose associated with the best pulmonary effect, 60 mg once daily, differed from that linked to the greatest decrease in predicted age, 30 mg twice daily, suggesting an effect that is at least partially independent of respiratory improvement.

Comparison to Over 55,000 Profiles from the UK Biobank

The protein profiles of treated patients were compared to over 55,000 profiles from the UK Biobank, which documents typical protein changes with age. According to Insilico, rentosertib has precisely reversed these aging-related changes.

The TNIK Target, a Generative Design, and an Industrial Roadmap

Insilico has mobilized two AI systems: one identified relevant proteins from health data and literature, while the other generated suitable candidate molecules. The TNIK protein was selected as a target for aging and pulmonary fibrosis, and the transition from target to candidate took about 18 months. Rentosertib, designed for IPF, illustrates Insilico's ambition to accelerate drug development through generative AI. The company, founded in Hong Kong in 2014, has received investment from Eli Lilly to commercialize drugs co-developed with AI. Its founder, Alex Zhavoronkov, claims to have at least 28 AI-designed candidates by March 2026, many of which are in clinical trials.