In a 31-person trial, 47 percent of patients on dasatinib plus quercetin had less liver scarring, against 7 percent on placebo. Authors want a larger trial.
Liver scarring eased in 8 of 17 patients given dasatinib plus quercetin, two senolytics (drugs built to clear worn-out cells), against 1 of 14 on placebo (1). The two earlier randomized trials we checked that tested a senolytic for benefit missed their main goals, and no drug is approved as a senolytic. This trial met its main goal.
It was double-blind, ran 21 weeks and enrolled 31 people with MASH, a severe form of fatty liver disease once called NASH. It is reported in "Senolytics dasatinib and quercetin in metabolic dysfunction-associated steatohepatitis: a proof-of-principle randomized, controlled trial", published in Nature Metabolism on 1 October 2026 (1). The lead author is Mijra Koning and the senior author is Abraham Meijnikman. The trial is registered as NCT05506488.
From Laennec to Senolytics
The scarring this trial measured has a name that is two centuries old. In 1819 the French physician Rene Laennec published De l'auscultation mediate, the book built around the stethoscope he had invented three years earlier (7). In a footnote to that book he gave cirrhosis its name, from the Greek word kirrhos, meaning tawny. The liver lesion itself had been recognized in England more than a century earlier; what stuck was his name for it (8).
The cells these drugs target have a shorter history. In 1961 Leonard Hayflick and Paul Moorhead reported that ordinary human cells in a dish divide only a set number of times before stopping. Hayflick extended the finding in 1965 (9).
That arrested state was later named cellular senescence, now counted among the hallmarks of aging: the cells refuse to die and leak inflammatory signals. In 2015 a team led from Mayo Clinic by James Kirkland proposed a class of drugs they called senolytics, molecules that selectively kill senescent cells, and reported two of them, dasatinib and quercetin, which they also tested as a pair (10).
Kirkland and Tamar Tchkonia, a co-author of that paper, are co-authors of the new trial too. Both now work at Cedars-Sinai in Los Angeles. Both have a financial interest in patents on senolytic drugs held by Mayo Clinic, which the paper declares (1). The details are under Funding Transparency.
What the Dasatinib and Quercetin Trial Tested
MASH, short for metabolic dysfunction-associated steatohepatitis, is the stage of fatty liver disease where fat in the liver tips into inflammation and cell damage. The parent condition is metabolic dysfunction-associated steatotic liver disease, MASLD: fatty liver, under the newer names, and NAFLD and NASH under the older ones. It is the leading liver disease worldwide, with an estimated prevalence of almost 40 percent (1).
Over years, the damage lays down scar tissue called fibrosis, which doctors stage from 0, none, to 4, cirrhosis. Everyone in this trial enrolled at stage 2 or 3. Fibrosis can progress to cirrhosis, liver failure or liver cancer. The paper calls it the most important factor in predicting liver-related complications and death in people with fatty liver disease (1).
Dasatinib plus quercetin, called D+Q, is among the best-studied senolytic pairs. Dasatinib is a prescription leukaemia drug. Quercetin is a flavonoid, a plant compound, found in onions and apples and sold as a supplement.
In laboratory and animal studies, the pair kills senescent cells while largely sparing healthy ones (10). Senescent cells also hide from the immune system, which is one reason they pile up in older and injured livers (1). One more boundary: this trial tested a liver disease, not ageing itself.
In the trial, participants took the two senolytics three days a week for three weeks, then had four weeks off. That seven-week cycle was repeated three times, for a total of 21 weeks. This intermittent pattern, carried over from earlier senolytic studies, aims to clear senescent cells without continuous exposure (1).
Who Was in the Trial
Thirty-one people with fibrotic MASH, confirmed by biopsy, were randomized: 17 to D+Q and 14 to placebo. The median age was 58 in the D+Q group and 51 in the placebo group. Men numbered 13 of 17 on D+Q (76.5 percent) and 9 of 14 on placebo (64.3 percent), 22 of 31 overall, or 71 percent (1). Twenty-seven participants completed the 21 weeks.
The random draw was not perfectly balanced. Type 2 diabetes was more common in the treated group: 13 of 17 (76.5 percent) against 5 of 14 (35.7 percent). So was oral antidiabetic medication, 64.7 percent against 21.4 percent. The treated group also started with more scarring: 11 of 17 had the more advanced F3 fibrosis against 4 of 14 on placebo (P = 0.07).
Both groups were heavy, with an average BMI over 31: MASH sits in metabolic health territory, driven by weight and blood sugar. In a trial this small, such imbalances can shift results in either direction, so the comparison is not a clean like-for-like.
Four participants left early, three of them on D+Q. Their reasons: side effects, mainly stomach complaints and fatigue; drinking more than 10 units of alcohol a day, which broke the trial's rules; tests tied to a cancer not disclosed at screening. The fourth, on placebo, declined the second biopsy. All four were counted as non-responders in the main analysis (1).
The Liver Scarring Result in Numbers
The primary endpoint, the main goal the trial set itself, was at least a one-stage improvement in fibrosis on a second biopsy, without the disease worsening. It was met by 8 of 17 patients on D+Q (47 percent) against 1 of 14 on placebo (7 percent) (1).
The P value was 0.021: a 2.1 percent chance of seeing a gap at least this large if there were genuinely no effect. By convention a P value below 0.05 earns the label statistically significant; the label says nothing about how large or lasting a benefit is.
Treated patients met the goal 6.59 times as often as patients on placebo. A risk ratio is a ratio of two rates, not a count of people. The absolute gap was 40 percentage points (1).
Every estimate comes with a confidence interval, the range of values compatible with this data. For a ratio, 1 is the no-difference point; for an absolute gap, it is 0.
The ratio's interval runs from 0.93 to 46.53, so on this yardstick no difference is still a value that fits. The gap's interval runs from 8.2 to 64.5 points and excludes 0. The ratio is the shakier of the two measures, because it rests on a single placebo responder (1).
A secondary result went the same way. MASH itself resolved in 9 of 17 treated patients (53 percent) against 1 of 14 on placebo (7 percent), also statistically significant (P = 0.02). Resolved here means the ballooning of liver cells, a swelling that marks injury, improved, with or without the inflammation improving (1).
A second scarring scale fell short. On the EPoS score, another way pathologists grade fibrosis, 8 of 17 on D+Q improved against 2 of 14 on placebo (P = 0.068), which is not statistically significant (1).
How Sure Should We Be
The trial has real strengths. It was a phase 2, double-blind, placebo-controlled trial. Three pathologists scored every biopsy without knowing the group or the timepoint. It ran at Amsterdam University Medical Center, from July 2023 to February 2026 (1).
The authors write that the biopsy result "should be considered hypothesis-generating rather than confirmatory, pending replication in a larger trial" (1). Hypothesis-generating means a lead worth testing properly, not yet proof. The main weakness is arithmetic: the result rests on a small number of responders, so a few patients either way can move the numbers.
Four patients had no follow-up biopsy. When the authors counted the missing data in the way least favourable to the drug, 8 responders against 2 in absolute numbers, the risk ratio fell from 6.59 to 3.29. The P value rose to 0.068, just above the usual line for statistical significance.
Three other ways of handling the missing data kept the result statistically significant, with risk ratios from 4.53 to 9.06 (1). The worst case, 3.29, is the one that did not.
The trial ran at one centre and lasted 21 weeks. It measured scarring on biopsy, not long-term outcomes, so it cannot say whether the improvement lasts or whether it lowers the risk of cirrhosis, liver failure or liver cancer.
The tests a patient would know found no detectable difference between the groups. That covers FIB-4 and the NAFLD Fibrosis Score, two blood-based scarring scores, the ProC3 and ELF blood tests, and the ultrasound stiffness measures. Liver enzymes ended lower on treatment, but not to a statistically significant degree. The authors say a trial this small may have been unable to detect changes in these markers (1).
So how sure should we be? Sure that the treated biopsies improved more often. Not sure the benefit is large, lasting, or real beyond this small group. That is what the larger trial is for.
Two Earlier Senolytic Trials That Missed
Two earlier randomized senolytic trials we checked, one in bone and one in knee joints, missed their primary endpoints. In 2024, a trial in 60 postmenopausal women tested intermittent D+Q against bone loss for 20 weeks. The main goal, the change in a bone resorption marker, did not differ between groups (P = 0.611).
A bone formation marker rose at 2 and 4 weeks, by about 16 percent, and was no different at 20 weeks. A subgroup result in women with a high senescent cell burden is exploratory (2).
In a trial reported in 2020, a single injection of a different senolytic, UBX0101, was tested in 183 people with painful knee osteoarthritis. It failed to beat placebo on its main goal, knee pain measured on WOMAC-A, a knee pain score, at 12 weeks. The developer did not take the drug further (3).
Three other small senolytic studies, all of D+Q, were pilots. One, in lung fibrosis, was open-label, with no placebo group (4). A 2023 trial in 12 people with lung fibrosis did randomize against placebo, but it was built to test feasibility and tolerability, not benefit (5). A pilot in diabetic kidney disease was open-label too (6).
Cross-trial comparisons like these are a rough guide, not a ranking: the diseases, drugs and endpoints all differ.
Two Medicines Are Already Approved for MASH
Two medicines are approved for MASH with fibrosis, and both were tested on biopsies too. The first, resmetirom, was approved in the United States on 14 March 2024, after the MAESTRO-NASH trial, a phase 3 trial, the large kind regulators ask for before approval.
In that trial's 966-patient primary analysis at 52 weeks, the disease resolved without fibrosis worsening in 25.9 and 29.9 percent in its two dose groups, against 9.7 percent on placebo. Fibrosis improved by at least one stage in 24.2 and 25.9 percent, against 14.2 percent on placebo (11, 12).
Semaglutide, a GLP-1 drug, gained a MASH approval in the United States in August 2025, after the ESSENCE trial. In the first 800 patients of that trial, read at 72 weeks, the disease resolved without fibrosis worsening in 62.9 percent on semaglutide, against 34.3 percent on placebo. Fibrosis improved without the disease worsening in 36.8 percent, against 22.4 percent on placebo (13, 14).
Each trial defined resolution in its own way, and the patients and endpoints differ, so these percentages are not one measure and cannot be ranked against the D+Q result. They set the bar any confirming senolytic trial would be compared with.
What Happened Inside the Liver
The researchers also ran single-nucleus RNA sequencing on the biopsies, reading gene activity cell by cell. An enrichment score measures whether a set of genes moves up or down together; a negative score means down. In the treated group, the SenMayo senescence gene set fell (score -1.4, P = 0.018) and a fibrosis gene set fell further (score -2.11, P less than 0.001). The placebo group did not fall on either (1).
What the trial did not do is count senescent cells directly. That count was a planned outcome, and the authors replaced it with the gene sequencing. In that data the usual marker genes, p16, p21 and p53, were only weakly active, so the team read whole sets of senescence genes instead.
The biopsies were small, so the 53 samples were pooled into 26 groups. The trial has therefore not shown that the senolytics cleared senescent cells from the liver; it found that their genetic footprints faded (1).
That pattern matches what senolytics are built to do: fewer senescence signals, fewer fibrosis signals, and fewer of the fibrogenic cell populations that lay down scar. It is mechanistic support, not proof, because it comes from the same small sample (1).
Dasatinib and Quercetin Side Effects in the Trial
More treated patients reported side effects: 14 of 17 (82 percent) against 6 of 14 (43 percent) on placebo, a statistically significant gap (P = 0.03). Together they reported 28 adverse events against 9 on placebo (P = 0.002). Headache or other neurological complaints were the most frequent, 13 of the 28 reports (46.4 percent). Infections, mostly self-limiting viral ones, appeared in both groups, and one participant stopped the treatment over gastrointestinal side effects (1).
The trial's table gives the detail the abstract compresses. Of the 28 reports on D+Q, 24 resolved, 3 recovered with minor after-effects and 1 was still ongoing. Seventeen were moderate and 1 severe; 8 were rated probably related to the treatment. The abstract calls all adverse events self-limiting; the table lists three reports that left minor after-effects and one still ongoing (1).
No cases of myelosuppression, a drop in blood cell production, or QTc prolongation, a change in heart tracing, occurred. Even so, platelets, the blood cells that help clotting, ended the trial lower on D+Q than on placebo. The adjusted gap was about 50 billion per litre of blood (P = 0.02) (1).
The one serious adverse event, the term for events that lead to hospital care, was a planned elective knee operation, considered unrelated to the study treatment. Separately, one participant was withdrawn after swallowing difficulty was traced to an oesophageal cancer that was already active at screening and had not been disclosed (1).
What Would Settle It
What would a confirming trial need? Our view: hundreds of patients rather than tens, so a handful of missing biopsies cannot flip the result. Years rather than weeks, to learn whether the scarring stays down. And hard outcomes: hospital admissions for liver failure, transplant and death, not only biopsy scores.
The authors themselves ask for a larger trial and for studies of dosing, patient selection, durability of response and long-term safety (1).
What You Can Do Now
Nothing in this trial is a reason to take either drug. Doctors checked blood counts and heart tracings, and platelet counts still ended lower than on placebo. We give no doses in this article for that reason.
If you have MASH, two questions are worth taking to a specialist: what stage is my scarring, and does either medicine approved in the United States fit me. And tell your liver doctor if you take quercetin or fisetin supplements.
Quercetin sold as a supplement is not this regimen. Quercetin alone was not tested here: it was given only together with dasatinib, so this result says nothing about quercetin by itself. And no supplement has produced this biopsy result.
Fisetin is another flavonoid sold as a senolytic. In old mice it improved artery function, but no completed human trial has yet found it improves artery function or any ageing outcome.
The everyday levers with human evidence behind them remain weight loss and exercise: the first approved MASH medicine is prescribed to be used along with diet and exercise (12).
Frequently Asked Questions
Do senolytics work?
Not proven yet. In a 31-person trial in Nature Metabolism, October 2026, liver scarring eased in 8 of 17 treated patients against 1 of 14 on placebo. Two earlier randomized senolytic trials we checked missed their main goals. The authors of the new trial call its result a lead to test, not proof.
What are senolytic drugs?
Senolytics are drugs designed to kill senescent cells, worn-out cells that stop dividing but refuse to die and leak inflammatory signals. Among the best-studied pairs is dasatinib plus quercetin. Several senolytics have cleared these cells in animals. Human evidence is still early and limited to small trials.
Are senolytics safe?
Long-term safety in people is not known. In a 31-person trial in Nature Metabolism, October 2026, 82 percent of treated patients had side effects against 43 percent on placebo over 21 weeks. Platelet counts ended lower on treatment, though the trial recorded no case of bone marrow suppression and none of QTc prolongation, a change in heart rhythm.
Can you buy senolytics?
Not as medicines. No drug is approved as a senolytic. Dasatinib is a prescription leukaemia drug, and quercetin supplements are not the tested regimen and vary in content. Buying either to copy a trial protocol would mean taking an unproven, unsupervised risk.
Should I take dasatinib and quercetin?
No. One small trial is not a reason to. That trial, in Nature Metabolism, October 2026, followed 31 people for 21 weeks under supervision, with blood counts and heart tracings checked. Its authors call the result a lead to test, not proof. Dasatinib is a prescription cancer drug with real risks.
Is quercetin hard on the liver?
In a 31-person trial in Nature Metabolism, October 2026, 17 people took quercetin only together with dasatinib for 21 weeks, so the trial tested the pair, not quercetin alone. No detectable difference in liver enzymes appeared between the groups. If you take quercetin, ask your liver doctor.
Sources
- Koning M, Kovynev A, et al. Senolytics dasatinib and quercetin in metabolic dysfunction-associated steatohepatitis: a proof-of-principle randomized, controlled trial. Nature Metabolism. Published 1 October 2026. PMID: 42823536. DOI: 10.1038/s42255-026-01643-4. Open access. NCT05506488
- Farr JN, et al. Effects of intermittent senolytic therapy on bone metabolism in postmenopausal women: a phase 2 randomized controlled trial. Nature Medicine. 2024. PMID: 38956196. DOI: 10.1038/s41591-024-03096-2
- Unity Biotechnology. UNITY Biotechnology announces 12-week data from UBX0101 phase 2 clinical study in patients with painful osteoarthritis of the knee. Press release, 17 August 2020. GlobeNewswire. Registry: NCT04129944
- Justice JN, et al. Senolytics in idiopathic pulmonary fibrosis: results from a first-in-human, open-label, pilot study. EBioMedicine. 2019. PMID: 30616998
- Nambiar A, et al. Senolytics dasatinib and quercetin in idiopathic pulmonary fibrosis: results of a phase I, single-blind, single-center, randomized, placebo-controlled pilot trial on feasibility and tolerability. EBioMedicine. 2023. PMID: 36857968. NCT02874989
- Hickson LJ, et al. Senolytics decrease senescent cells in humans: preliminary report from a clinical trial of dasatinib plus quercetin in individuals with diabetic kidney disease. EBioMedicine. 2019. PMID: 31542391
- Roguin A. Rene Theophile Hyacinthe Laennec (1781-1826): the man behind the stethoscope. Clin Med Res. 2006. PMID: 17048358
- Duffin JM. Why does cirrhosis belong to Laennec? CMAJ. 1987. PMID: 3304599
- Hayflick L, Moorhead PS. The serial cultivation of human diploid cell strains. Exp Cell Res. 1961. PMID: 13905658. And Hayflick L. The limited in vitro lifetime of human diploid cell strains. Exp Cell Res. 1965. PMID: 14315085
- Zhu Y, et al. The Achilles' heel of senescent cells: from transcriptome to senolytic drugs. Aging Cell. 2015. PMID: 25754370
- Harrison SA, et al. A phase 3, randomized, controlled trial of resmetirom in NASH with liver fibrosis. N Engl J Med. 2024. PMID: 38324483. NCT03900429
- US Food and Drug Administration. FDA approves first treatment for patients with liver scarring due to fatty liver disease. Press announcement, 14 March 2024. fda.gov
- Sanyal AJ, et al. Phase 3 trial of semaglutide in metabolic dysfunction-associated steatohepatitis. N Engl J Med. 2025. PMID: 40305708. NCT04822181
- US Food and Drug Administration. FDA approves treatment for serious liver disease known as MASH. Drugs news page, content current as of 15 August 2025. fda.gov
Funding Transparency
LSD is editorially independent. We receive no funding from pharmaceutical, supplement or longevity companies. These are the funding ties behind each study above, taken from each paper's own funding and conflict statements where we could read them.
- Source 1 (Koning, 2026): Supported by a ZonMw GGG grant 2021 (10140262110036). The senior author was supported by a VENI 2023 grant (09150162310148), the Dutch Digestive Health fund 2024 (WO2423) and the Dutch Diabetes Research Foundation 2025 (2025.81.019). Another author is supported by NIH P30 DK120515, and two others by NIH grants R37AG013925 and R33AG061456, the Connor Fund, Robert J. and Theresa W. Ryan and the Noaber Foundation. One author, Bernd Schnabl (B.S.), has consulted for Boehringer Ingelheim Pharma and Mabwell Therapeutics. His institution, UC San Diego, received research support from Axial Biotherapeutics, Apollo Therapeutics, ChromoLogic, Intercept Pharmaceuticals and Prodigy Biotech, and has filed several patents with him as inventor. He founded Nterica Bio. Two other authors, Tamar Tchkonia (T.T.) and James L. Kirkland (J.L.K.), have a financial interest related to this research, including patents and pending patents covering senolytic drugs and their uses held by Mayo Clinic.
- Source 2 (Farr, 2024): Funded by US National Institutes of Health grants and run at Mayo Clinic. The grants: R21 AG065868, P01 AG062413, R01 AG076515, R01 DK128552, R01 AG055529, R37 AG13925 and R33 AG61456. Its competing interests statement reads: "N.K.L., T.T. and J.L.K. have a financial interest related to this research, including patents and pending patents covering senolytic drugs and their uses that are held by Mayo Clinic."
- Source 3 (UBX0101): A press release from Unity Biotechnology, the sponsor, plus the trial's registry entry. We found no full peer-reviewed paper; a 2021 conference abstract (Osteoarthritis and Cartilage 29:S52) reports the topline.
- Source 4 (Justice, 2019): Funded by NIH grants (including R37 AG13925), the Connor Group, Robert J. and Theresa W. Ryan, and the Noaber, Ted Nash Long Life and Glenn Foundations. Its declaration of interests names the same pair as source 1, and a third author: "JLK, TT, and NKL have a financial interest related to this research. Patents on senolytic drugs are held by Mayo Clinic."
- Source 5 (Nambiar, 2023): A pilot run by many of the same investigators as source 4. Its registry entry names Wake Forest University Health Sciences as sponsor, with Mayo Clinic and UT Health San Antonio as collaborators. Funded by NIH grants (R33AG61456, R37 AG013925 and others), Robert and Arlene Kogod, the Connor Fund, Robert J. and Theresa W. Ryan, and the Noaber and Ted Nash Long Life Foundations. Its declaration of interests reads: "TT and JK have a financial interest related to this research including patents and pending patents on senolytic drugs, and their uses are held by Mayo Clinic."
- Source 6 (Hickson, 2019): Funded by NIH grants (DK109134, DK118120, DK120292, AG013925, AG062413 and AG061456, among others), Robert and Arlene Kogod, Satellite Healthcare, the Connor Group, Robert J. and Theresa W. Ryan, and the Glenn, Ted Nash Long Life and Noaber Foundations. The paper says the sponsors had no role in the data or the writing. Its declaration of competing interest reads: "J.L.K., T.T., Y.Z., and N.K.L. have a financial interest related to this research. Patents on senolytic drugs are held by Mayo Clinic."
- Sources 7, 8 and 9: Historical papers on Laennec and Hayflick. We found no commercial tie relevant to the facts we cite.
- Source 10 (Zhu, 2015): A Mayo Clinic-led collaboration with the Scripps Research Institute and others. Funded by NIH grants (AG013925, AG041122, AG031736, AG044396, DK050456, HL111121 and AG043376), the Glenn, Ted Nash Long Life and Noaber Foundations, and CTSA grant UL1-TR000157. Its conflict statement reads: "Mayo Clinic, JK, TT, YZ, TP, NG, AP have a financial interest related to this research." Two authors are co-founders of, and hold equity in, Aldabra Bioscience.
- Source 11 (MAESTRO-NASH): Funded by Madrigal Pharmaceuticals, the maker of resmetirom.
- Source 13 (ESSENCE): Funded by Novo Nordisk, the maker of semaglutide.
- Sources 12 and 14: Announcements from the US medicines regulator.
Related Reading
- Cellular Senescence Leaves Cells That Will Not Die, the biology of the cells these drugs are built to kill
- Worn-Out Cells Hide From the Immune System as We Age, how senescent cells escape clearance
- Fisetin Improved Artery Function in Old Mice, Not People, the senolytic supplement with a mouse result
- The 12 Hallmarks of Aging, and Exercise Hits Seven, where senescence sits among the causes of aging
- GLP-1 Weight Loss Is Up to 40 Percent Muscle and Bone, what GLP-1 drugs have done in people
In a small, 21-week trial, biopsies improved in 8 of 17 treated patients against 1 of 14 on placebo. Whether the benefit holds in a larger trial, lasts, or prevents the outcomes that matter is not yet known.
Written with the help of AI tools, shaped and verified by humans. Every result in this article was checked against the study's own paper, abstract or registry entry, or the sponsor's or regulator's own announcement where no paper exists.
This article is for information only and is not medical advice. Do not start or stop any medicine on the basis of an article.