Outrunning Ageing

The tree · Worn-out cells

A drug clears aged cells from a person

First human Next rung within 5 to 20 years 1 fact still to check

Some cells stop dividing but do not die, and sit in tissue giving off signals that inflame their neighbours; they build up with age. Clearing them makes old mice healthier and longer-lived. In 2019 a Mayo Clinic group gave a short course of two drugs that kill such cells to nine people with kidney disease, took tissue samples before and after, and found fewer of the cells. That is a first human, on a measurement. No such drug is approved anywhere. The first company drug to reach a controlled trial on a clinical outcome, for knee osteoarthritis, failed in 2020; a small academic controlled trial of the same two drugs in scarred fatty liver read out positive on 2026-10-01 (31 people, a biopsy measure), the first result on a clinical-type measure in this class that went the other way.

What would count

A person given a drug meant to kill senescent (aged, non-dividing) cells afterwards has measurably fewer of them in a tissue sample or a validated blood marker, compared with before treatment, in a peer-reviewed report. For the clinic rung: a regulator approves such a drug for an age-related condition, or it is in routine use in a named clinical setting, so a doctor can give it outside a trial.

Who settles it: Peer-reviewed trial reports for the first-human rung; a regulator's approval record, or a named clinical setting's published practice, for the clinic rung.

Where it stands

First human. Reached 18 September 2019, on the record of Hickson and 32 others, 18 September 2019 (doi.org). Next rung within 5 to 20 years.

This is one of the 5 treatments shared with the sister site The Upgrade Tree, which keeps its own record of it at the same test and the same stage. The weekly reading checks that the two stages agree.

When it reaches the clinic: the project's estimate

This is the one thing on this page that is a judgment rather than a record. The project's estimate of the year it reaches the clinic is 2038, no earlier than 2031 and no later than 2060; the model draws from that range every run, with the likely year the most probable draw. Set 16 October 2026, a first pass awaiting an outside review.

RungNo earlier thanMost likelyNo later than
In clinical use 2031 2038 2060 feeds the year at the top
In wide use 2036 2046 2080

Why. A first human result on a measurement exists since 2019, so the remaining step is a controlled trial on a clinical outcome that a regulator accepts, and then the approval. The one such trial to read out so far failed (2020); several smaller trials in knee, lung and eye conditions are under way or planned, and the field has had fifteen years since the 2011 mouse result without an approval. The likely year assumes one of the current programmes reaches a positive phase 3 in the mid 2030s and approval follows within two to three years; the latest year allows for the whole first generation of these drugs failing and a second generation starting over. [Note 2026-10-17, from the weekly reading of that date: the sentence above that the one such trial to read out so far failed is out of date. A small academic controlled trial of the same two drugs in scarred fatty liver read out positive on 2026-10-01 (koning-2026). The years are unchanged; John may want to weigh it at the next review of the ranges.]

Changes to this record

  • 17 October 2026, description: human, on a measurement. No such drug is approved anywhere, and the one company drug to reach a controlled trial on a clinical outcome, for knee osteoarthritis, failed in 2020. became human, on a measurement. No such drug is approved anywhere. The first company drug to reach a controlled trial on a clinical outcome, for knee osteoarthritis, failed in 2020; a small academic controlled trial of the same two drugs in scarred fatty liver read out positive on 2026-10-01 (31 people, a biopsy measure), the first result on a clinical-type measure in this class that went the other way.. [reading 2026-10-17] The sentence 'the one company drug to reach a controlled trial on a clinical outcome ... failed in 2020' was out of date once the Amsterdam liver trial read out positive on 2026-10-01. The stage is unchanged: the 2019 kidney report stays the record of first human. Source: Koning and 12 others, 1 October 2026.
  • 17 October 2026, estimate.why: (nothing; a dated note added at the end of the reasoning; no year changed) became [Note 2026-10-17, from the weekly reading of that date: the sentence above that the one such trial to read out so far failed is out of date. A small academic controlled trial of the same two drugs in scarred fatty liver read out positive on 2026-10-01 (koning-2026). The years are unchanged; John may want to weigh it at the next review of the ranges.]. [reading 2026-10-17] A dated note that the estimate's premise that the one controlled trial so far failed no longer holds; no year changed. Source: Koning and 12 others, 1 October 2026.
  • 17 October 2026, notes: (nothing; text added at the end of the notes) became Added 2026-10-17, from the weekly reading of that date: in the Amsterdam liver trial (koning-2026), single-cell sequencing of liver biopsies showed fewer signs of aged cells in the treated group than the placebo group. That is a supporting measurement, not a replacement for the 2019 kidney report, and the stage stays where it was.. [reading 2026-10-17] Records the supporting senescence measurement in the new liver trial. Source: Koning and 12 others, 1 October 2026.

The last readings

  • 17 October 2026 Left for the author

    Nothing dated in the window meets or moves the test; the leaf stays at first-human on the 2019 kidney-disease measurement. Two things the record does not know, both dated before the window: the Nature Metabolism phase 2 placebo-controlled trial of dasatinib plus quercetin in 31 people with scarred fatty liver (published 2026-10-01 by the search summary; 47 per cent against 7 per cent on the main biopsy measure), and a Cell Metabolism paper from Cedars-Sinai (press release 2026-09-10, re-reported by ScienceDaily on 2026-10-13) reporting that a soluble form of the protein PD-L2 rises in blood with age and fell in people after senolytic drugs, which the authors offer as a possible blood marker of aged cells but say is not validated. The searches found no 2026 peer-reviewed report of fewer aged cells in a tissue sample after a senolytic drug beyond those the record holds. Left for the author, resolved on 17 October 2026: The change I would make: the description says the one drug to reach a controlled trial on a clinical outcome failed in 2020, and the estimate's reasoning says the one such trial to read out so far failed; the Nature Metabolism liver trial is a controlled clinical-outcome readout that went the other way, so both sentences are now wrong as written and should say that a small academic trial in liver scarring read out positive in 2026-10 (the candidate is logged under the controlled-trial leaf of this branch and need enter the publications log once). The stage does not change. The PD-L2 blood marker is worth a line in the notes only if the author wants to track candidate blood markers; it is not validated and the paper is mouse-and-samples work. Both are the author's call because they are dated before the window, so the 2026-10-16 entry missed them, and because I could not open either paper's page to confirm date, authors or venue.source 1, source 2, source 3, source 4, source 5

What the story so far says about it

Promised, in public

  • 27 October 2016Unity Biotechnology: Drugs that kill worn-out cells preventing, halting or reversing diseases of ageing in people, with first trials in the near future; no year. Verdict: cannot be settled. A promise moves nothing here; the stage above is what has been shown.

On the record

Notes

The animal record is the 2011 Mayo Clinic report in which clearing such cells from mice delayed several age-related conditions. An earlier 2019 pilot in people with lung scarring reported better walking and standing but did not measure the cells, so it is not the record here. The Upgrade Tree keeps the same leaf at first human; the weekly reading checks the two stages agree. Added 2026-10-17, from the weekly reading of that date: in the Amsterdam liver trial (koning-2026), single-cell sequencing of liver biopsies showed fewer signs of aged cells in the treated group than the placebo group. That is a supporting measurement, not a replacement for the 2019 kidney report, and the stage stays where it was.

Not yet verified

These facts in the record were written from memory or from a secondary source and have not yet been checked against the paper, the regulator's record or the registry. Nothing here is presented as settled until it is cleared, by checking it and either deleting the line or correcting the record.

  • A paper in Cell Metabolism, announced by Cedars-Sinai on 2026-09-10, reports that blocking the protein PD-L2 reduced aged cells in old mice and that PD-L2 rises in human aged cells; the weekly reading of 2026-10-17 also saw a report that a soluble form of it fell in people after senolytic drugs and might serve as a blood marker. The university's release was read, but the paper's own page was not opened, so the paper is not logged and the blood-marker claim is unchecked.