Are longevity drugs hiding in plain sight?

There is a missed opportunity in clinical trials already under way. Most drug trials draw blood from participants at several visits and measure whether a treatment helped the disease it was designed for. Very few check whether the drug also changed how fast participants are aging. If aging clocks were measured in every trial, we could identify drugs that act across several diseases by slowing aging itself, using studies that are already happening. Two recent studies show what this could look like. Both reanalyzed stored blood samples from completed trials. The trials tested semaglutide in people with HIV and an experimental lung fibrosis drug from Insilico Medicine called rentosertib.

What an aging clock measures. Aging clocks estimate biological age from a blood sample using patterns that change with age. Aging clocks include epigenetic clocks, which read age-related marks on DNA, transcriptomic clocks, and proteomic clocks, which read the levels of thousands of blood proteins. Clocks differ in what they are trained to predict, which can be chronological age, disease states, mortality risk, or the current pace of aging. I have written about what biological age means and about which clocks respond to interventions in earlier articles.

A GLP-1 drug produced the first randomized trial evidence of slower epigenetic aging.

Why GLP-1 drugs are part of the longevity conversation. GLP-1 drugs such as semaglutide are often described as possible first longevity drugs. They treat multiple age-related diseases (heart disease, diabetes) and extend lifespan in mice.

The study in people with HIV. A study published in May in Nature Communications reanalyzed blood from a 32-week randomized, placebo-controlled trial of weekly semaglutide in adults with HIV-associated lipohypertrophy, a condition in which excess fat builds up around the abdomen. People with HIV tend to show faster epigenetic aging. Aging was not an outcome in the original trial, which was designed to measure fat around the organs. Researchers ran stored blood from 84 participants on 17 epigenetic clocks at the start of the trial and at 32 weeks.

The results across clocks. Compared with placebo, the semaglutide group measured about 4.9 years younger on PhenoAge and 1.4 to 3.1 years younger on versions of GrimAge, two clocks built to predict disease and mortality risk. DunedinPACE, which estimates the current speed of aging, showed a pace about 9% slower. Clocks for 11 organ systems moved in the same direction, with the largest changes in the inflammation, brain, and heart clocks. A clock estimating intrinsic capacity, a measure of physical and mental function, did not change. Because the analysis was added after the trial ended, the authors describe it as exploratory and call for trials designed from the start to test aging.

Rentosertib was designed with aging biology in mind.

What the drug is. Rentosertib is an experimental small-molecule drug (a chemically made compound taken as a pill) from Insilico Medicine, originally developed to treat idiopathic pulmonary fibrosis (IPF), a disease in which scar tissue builds up in the lungs. It works by blocking a protein called TNIK, which has been linked to fibrosis and to several biological processes associated with aging. Insilico used AI at both steps. After its PandaOmics platform flagged TNIK as a gene involved in six hallmarks of aging (a widely used list of biological processes that drive aging), its Chemistry42 generative chemistry platform designed the molecule itself. The discovery and early testing were published in Nature Biotechnology.

The aging connection was built in from the start. Where most geroscience studies (research on how aging drives disease) test repurposed drugs like rapamycin or metformin, Insilico folded aging biology into its target selection by prioritizing TNIK as a dual-purpose target relevant to both aging and IPF. IPF itself is an age-related disease, with onset averaging around age 65. Aging processes such as cellular senescence and inflammation play a large role in it.

NeuroAge’s platform similarly identifies drugs for dementia that target brain aging itself. This is a common theme in the new wave of longevity therapeutics companies.

The lung results. The Phase IIa trial, published in Nature Medicine in 2025, met its main safety goal and showed a dose-dependent trend toward better lung function. A Phase III trial began in China this year.

Six proteomic clocks pointed toward a younger biological age with rentosertib.

What the new paper found. The Nature Biotechnology study by Dr. Alex Zhavoronkov and colleagues reanalyzed blood samples from the Phase IIa trial. Forty-two of the trial’s 71 participants, with an average age of 67.1, provided samples at baseline and weeks 2, 4, and 12, in which 2,841 proteins were measured. Although the six clocks were built by different groups using different methods and trained on either chronological age or mortality risk, all six recorded a drop in predicted biological age in the treated arms compared with placebo. The largest effect appeared at week 4 in the 30 mg twice-daily group, roughly three to four years younger on several measures and up to six years on one clock.

The limits of the analysis. The trial enrolled only people with IPF, so no healthy participants took rentosertib and had their clocks measured. Giving the drug to people without lung disease and measuring the same clocks would show whether it lowers biological age outside a disease setting. The authors state that fully separating a broader slowing of aging from protein changes caused by the lung disease improving is not possible within an IPF group and would require testing in healthy volunteers. The clock signal also leveled off after week 4. By week 12 fewer of the comparisons with placebo were statistically significant, although predicted ages did not change significantly between weeks 4 and 12. The reason for this plateau is not known. It could reflect random variation or a compensating response by the body to ongoing TNIK blocking. More studies are needed to understand this result. This secondary analysis came from a trial that was not designed to show an anti-aging benefit. The analysis included a small sample of 42 people followed for 12 weeks. According to its authors, the study is the first head-to-head comparison of multiple proteomic clocks within a drug trial.

The direct aging evidence for TNIK centers on one hallmark.

Where the six-hallmark figure comes from. The six-hallmark figure comes from Insilico’s own literature-mining analysis, published in 2022 in Aging. The main text of that paper links TNIK to extracellular matrix stiffness (stiffening of the scaffolding around cells), with the full list of hallmark assignments in a supplementary table. The direct experimental evidence for TNIK in aging centers on one hallmark, cellular senescence, a state in which damaged cells stop dividing and release inflammatory signals.

What TNIK does biologically. TNIK is a kinase, an enzyme that switches other proteins on or off. In fibroblasts, it regulates TGF-β signaling and pushes these cells to multiply and turn into myofibroblasts, the scar-forming cells in fibrosis.

The direct aging experiment. The strongest aging-specific data is an Insilico study in Aging and Disease by Dr. Qiuqiong Tang and colleagues. The team used its automated robotics lab and found that blocking TNIK, either with the drug or by silencing the gene, reduced cellular senescence across several senescence models. The drug lowered the inflammatory factors that senescent cells release. Gene expression analysis showed reduced aging signatures and less fibronectin, a matrix protein, through TGF-β signaling. This puts it in the senomorphic category, meaning it dampens the harmful output of senescent cells without removing them. This work was done in cultured cells.

What’s missing. I found no animal lifespan or healthspan studies of TNIK inhibition, and no independent replication outside Insilico. Insilico itself says these findings don’t establish rentosertib as an anti-aging therapy but make the case for studying TNIK where fibrosis, inflammation, senescence, and age-related disease overlap.

TGF-β signaling connects fibrosis and aging.

A signal for growth and repair. TGF-β (transforming growth factor beta) is a signaling protein that cells use to control growth, repair, and immune activity. Its role in aging is mixed. In youth it keeps tissues in balance but with age its signaling tends to rise in many tissues and push toward fibrosis, senescence, and poor regeneration.

It depends on context. In healthy tissue, TGF-β stops cells from dividing too much, calms the immune response, and directs wound healing. It is needed for regulatory T cells, which keep immune responses in check, and for microglia, the brain’s immune cells, to stay in their resting state. In early cancer it suppresses tumors, but later it can help them spread. Aging tends to shift the balance toward its harmful effects.

Why drugs target the pathway indirectly. Blocking TGF-β throughout the body has caused problems in trials and animal studies, including heart valve damage with some receptor inhibitors and skin lesions with antibody treatments. Because of its role in immune balance and tumor suppression, complete blockade is risky. This has led developers to aim at points that are more specific to disease, such as integrins (surface proteins that activate TGF-β locally) or downstream kinases like TNIK. Pirfenidone, an approved IPF drug, lowers TGF-β activity among other effects.

Balance is the goal. TGF-β is part of normal tissue maintenance, but its long-term overactivity with age contributes to fibrosis, senescence, stem cell decline, and possibly brain inflammation. The challenge is to reduce its harmful effects without removing its protective ones.

Adding aging clocks to every trial would require several changes in how trials are designed and reported.

Storing blood at each visit. Many trials already draw and store blood. Both studies described here used samples collected for other purposes, which means completed trials with stored samples could be analyzed for aging effects now. One blood draw can supply both epigenetic and proteomic clocks.

Separating the treated disease from aging. A drug that improves lung disease or reduces abdominal fat would be expected to shift some blood markers for that reason alone. Organ-specific clocks can help with this. When clocks for organs outside the treated disease also change, as the brain and heart clocks did in the semaglutide study, the result is consistent with a broader effect on aging.

Reporting clock results for every trial. Publishing clock results from every trial would capture drugs that miss their main goal but still slow aging clocks. Null results matter too. In a trial of rapamycin combined with exercise, none of four clocks reached statistical significance, which is informative for anyone weighing that drug.

Linking clock changes to health outcomes. Regulators have not yet accepted any aging clock as an endpoint that can support drug approval. Following trial participants whose clocks changed would build the necessary evidence for that over time.

Some drugs already in testing may be slowing aging without anyone measuring it.

Where this could lead. Many drugs in development target inflammation, fibrosis, or metabolism, each of which overlaps with aging biology. Both the semaglutide and rentosertib analyses found signals the original trials were never designed to look for, using blood that had been stored in freezers. If clocks become a routine part of trial design, the next candidate longevity drug may come from a trial for lung, kidney, or liver disease, identified years earlier than it would be otherwise.