AI-Designed Drug Rentosertib Shows Signs of Slowing Aging

Insilico Medicine's AI-designed drug rentosertib reduced patients' predicted biological age across six aging clocks in a small Phase 2a trial, a first.

@sharedot

AI-Designed Drug Rentosertib Shows Signs of Slowing Aging

By @sharedot · · 7 pages

  • AI
  • Biotech
  • Longevity
  • Drug Discovery

Insilico Medicine's AI-designed drug rentosertib reduced patients' predicted biological age across six aging clocks in a small Phase 2a trial, a first.

What Happened

Insilico Medicine reported in Nature Biotechnology that rentosertib, an experimental drug for idiopathic pulmonary fibrosis (IPF), showed reductions in predicted biological age among trial participants.

Both the drug's molecular structure and its target, a protein called TNIK, were generated using AI methods. In a 12-week Phase 2a trial, patients on the drug showed reductions in predicted biological age compared with placebo when blood samples were reanalyzed using six AI-based "aging clocks."

Why It Is Surprising

The finding is a surprise because the trial was designed to test a lung disease treatment, not a longevity therapy.

The anti-aging signal emerged only from a reanalysis of samples, and patients showed reductions ranging from roughly 2.6 to 6.2 predicted years across the six clocks, with the strongest effect around the fourth week.

The Indian Express reports this is a milestone in the broader effort to use the same AI techniques behind chatbots and image generators to improve healthcare, pushing drug discovery into longevity research.

The Evidence

The trial initially enrolled 71 participants in China; researchers analyzed serum samples from 42 who consented, with an average age of 67.1.

All six proteomic aging clocks detected changes associated with reduced biological age in treated participants, while the placebo group showed little change or a slight increase.

AI Insider reports the 30-milligram twice-daily dose generated the broadest agreement, and protein analysis found changes in the trajectories of 326 proteins across rentosertib groups versus only two in the placebo group.

The Stakes

The result is a proof of concept for AI-driven drug discovery extending beyond specific diseases into longevity research. Insilico founder Alex Zhavoronkov described the approach as being like scanning a lock and generating a key that fits the lock, according to The Indian Express.

Harvard's Vadim Gladyshev called it the first study showing clearly that predicted biological age can be reduced, while cardiologist Eric Topol said the drug looks encouraging but there is not yet a definitive trial for a final judgment, The Indian Express reports.

What Comes Next

Scientists cautioned against reading too much into results from a small, ill patient population. The study did not test healthy patients, aging clocks remain a debated tool, and it could not separate the drug's anti-fibrotic activity from any independent anti-aging effect.

AI Insider reports researchers proposed dual-purpose trial designs integrating aging endpoints into disease studies, and that future work needs larger groups, longer observation, and non-IPF participants.

Rentosertib still requires further trials and regulatory review before any approval, according to NDTV Profit.

Sources

  1. ndtvprofit.com › AI-Generated Drug Shows Early Signs Of Slowing Aging: Report
  2. indianexpress.com › Early data indicates an AI-generated drug could slow aging
  3. theaiinsider.tech › AI-Designed Drug Shows Signs of Lowering Biological Age in Early Clinical Study

More on AI