Can artificial intelligence lead to a drug that slows or even reverses human aging? Results of a new study have raised this question again. Researchers, in a recent analysis, found that a drug designed with the help of AI to treat idiopathic pulmonary fibrosis was associated in patients receiving the drug with a reduction in predicted biological age across six models called “aging clocks.”
This drug, now named Rentosertib and previously known as ISM001-055, is a TNIK protein inhibitor. The manufacturer, Insilico Medicine, developed it using generative AI technologies. The drug was evaluated in a phase 2a trial on 71 patients with idiopathic pulmonary fibrosis, with the primary aim of assessing safety and tolerability alongside an initial efficacy evaluation.
In the preliminary results of that trial, the drug demonstrated improved lung function at some doses. For example, in the group receiving 60 mg once daily, the mean change in forced vital capacity or FVC increased by about 98.4 milliliters, while this metric decreased by about 20.3 milliliters in the placebo group.
But researchers then reexamined serum samples collected in the same trial from another perspective: could changes in blood proteins indicate a change in patients’ biological age?
How do the six aging clocks work?
An “aging clock” is not a literal clock; rather it is a set of statistical and biological models that use molecular data, including amounts of proteins in the blood, to provide an index of the body’s biological state.
In the new study, researchers applied six proteome-based aging clocks to patients’ blood samples. These models attempt to estimate biological age or certain indices related to disease and mortality risk using patterns across thousands of proteins. However, the concept of “biological age” itself remains a matter of scientific debate, and the researchers emphasized the limitations of different aging clocks.
The study’s results showed that all six clocks recorded reductions in predicted biological age in the Rentosertib groups, whereas the placebo group showed much smaller changes or, in some cases, slight increases. Overall, 21 of 54 comparisons reached a pre-specified statistical threshold, with the strongest signal observed at week four.
In one dosing regimen-30 mg twice daily-9 comparisons showed significant reductions in biological age. However, these figures should not be interpreted as patients actually “becoming younger.”
Why we still cannot call this an anti‑aging drug?
The main limitation of the study is that proteomic clocks alone cannot determine whether the observed change was due to slowing of the aging process or a result of disease improvement and reduction of inflammation and fibrosis in patients’ bodies.
The researchers themselves emphasized in the paper that proteomic aging clocks alone are unable to disentangle aging-related effects from disease-specific effects. To probe this, they also analyzed biological pathways related to cellular aging and metabolism and observed signs of potential changes linked to anti‑aging processes.
On the other hand, the study was conducted on patients with pulmonary fibrosis, not healthy individuals; therefore its results cannot be directly generalized to the general population. The main trial also included only 71 participants, and the new analysis relies on samples from that same study.
According to Nature Biotechnology, the new findings indicate that aging metrics can be measured concurrently in clinical trials of drugs for age-related diseases; however, these results do not yet prove that Rentosertib is an anti‑aging drug for humans.
The significance of this research lies more in its approach: a drug whose primary goal was to treat a specific disease has, using molecular measurement tools and AI, now provided clues about its possible effects on aging‑related processes. To turn these clues into a real anti‑aging therapy will require larger studies, healthy cohorts, and trials that can separate the drug’s independent effect on aging from its effect on the disease.