Clinical Research · global
Pulmonary fibrosis drug candidate rentosertib shows signs of reduced biological age: How should six protein clocks be interpreted?
Blood analysis from 42 patients extends research on an AI-designed drug into aging biology. But whether the models’ “younger” readings reflect antifibrotic effects, changes in aging mechanisms, or both remains unresolved.
Could aging signals in the blood also change during treatment for pulmonary fibrosis? A study published on September 7 in *Nature Biotechnology* applied six proteomic aging clocks to clinical trial samples for the drug candidate rentosertib and found signals of reduced predicted biological age in patients receiving treatment. This connects disease treatment with aging research, but is not sufficient to prove that patients actually experienced “age reversal.”
Rentosertib is a TNIK inhibitor developed by Insilico Medicine for investigation in idiopathic pulmonary fibrosis (IPF). According to the company, the team first used AI to help identify drug targets associated with fibrosis and aging, then used a generative chemistry platform to design small molecules. The specific question in this study was whether, following treatment with the drug, the protein composition of patients’ blood also shifted toward a state the models identified as younger.
The material analyzed came from a 12-week, randomized, double-blind, placebo-controlled Phase 2a trial. The original trial enrolled 71 people; this protein analysis covered 42 of those patients who consented to participate and had complete sampling data. Researchers examined serum samples collected before treatment and at weeks 2, 4, and 12, analyzing 2,841 proteins. This was therefore an exploratory extension of an existing trial, rather than a separate trial specifically designed to validate anti-aging efficacy.
Proteomic aging clocks are models that estimate biological age using relationships between protein concentrations and age or mortality risk. The six clocks use different methods and training objectives. Similar directional changes across multiple models help assess whether the signal extends beyond the preferences of a single algorithm; however, they read samples from the same group of patients and cannot be treated as six independent clinical validations.
In the study, all six clocks detected changes indicating reduced predicted biological age in the treatment groups. The regimen of 30 milligrams per dose, twice daily, produced the most consistent results, and signals across models were most concentrated at week 4. This does not mean that every dose at every time point reached statistical significance. By week 12, the number of statistically significant comparisons had decreased, making the persistence of these short-term changes a question that must be addressed next.
Insilico Medicine described the results in its announcement as showing “potential for biological age reversal,” while the paper explicitly leaves room for interpretation: proteomic clocks alone cannot distinguish whether the drug changed the aging process, improved fibrosis, or affected both. Disease activity itself can influence blood proteins; a lower score cannot be directly translated into longer lifespan, rejuvenated organs, or proven anti-aging benefits.
Another contribution of this work is its attempt to incorporate aging indicators into a disease trial. The company said the relevant protein data have been made publicly available under accession number OMIX008341 for further examination and research. The study’s authors also include the company’s founder and employees, which should be considered when interpreting the results. Before such indicators can be used to judge treatment efficacy, larger and longer-term validation is still needed, along with confirmation that changes in the models correspond to actual improvements in patients’ health.