⚠️ CONFERENCE ABSTRACT ONLY — Auto-extracted by Claude on 2026-06-30. This page is based on (a) Crossref metadata for DOI 10.1016/j.joca.2025.02.667 (title, authors, volume/page confirmed) and (b) ClinicalTrials.gov NCT04210986 posted results (design, dosing, endpoints, outcome data). The abstract body text was NOT accessible to automated retrieval (OARSI journal page is Cloudflare-blocked). The full paper had not been published as of mid-2026. Do NOT cite specific p-values as abstract-derived — they come from the ClinicalTrials.gov results posting. Verify all claims against the published paper when it appears.

RESULTS: FROM A RANDOMIZED CLINICAL TRIAL EVALUATING THE SENOLYTIC FISETIN FOR TREATING KNEE OSTEOARTHRITIS

TL;DR

Abstract-level evidence only. A Phase I/II randomized, double-blind, placebo-controlled trial (n=74) of intermittent oral fisetin versus placebo in adults with symptomatic knee osteoarthritis found no statistically significant benefit on any measured efficacy endpoint — patient-reported pain (NRS), function (WOMAC), serum biomarkers (CRP, COMP), physical performance tests, or quantitative MRI cartilage quality. Safety was comparable between arms. This is a negative senolytic trial in humans and as of mid-2026 represents the most substantive Phase I/II dataset bearing on whether fisetin’s established preclinical senolytic activity translates to clinical OA benefit. The full paper was not yet published as of mid-2026. Results summarised here are drawn from the ClinicalTrials.gov posted outcome data, not the abstract text itself.

Design

Source note: The following design details are drawn from ClinicalTrials.gov registration NCT04210986 (“Senolytic Drugs Attenuate Osteoarthritis-Related Articular Cartilage Degeneration: A Clinical Trial”), which matches the abstract’s declared RCT. The abstract body was inaccessible; design details confirmed against the registry.

Parameter	Detail
Phase	I/II
Design	Randomized, double-blind, placebo-controlled
Site	The Steadman Clinic / Steadman Philippon Research Institute, Vail CO
Enrolment period	January 2020 – February 2023 (completed)
Total n	74 (fisetin n=34; placebo n=40)
Age range	40–80 years
OA severity	Kellgren-Lawrence grade II–IV (radiographically confirmed)
Pain criterion	NRS pain score 4–10 (ambulatory)
Key exclusions	Pregnancy; recent knee surgery; concurrent warfarin or losartan; other senolytic agents; BMI >40; diabetes (HbA1c >6.5%); significant hepatic or renal disease
Trial registration	NCT04210986

Intervention

Fisetin 100 mg oral capsules at ~20 mg/kg/day for 2 consecutive days, followed by 28 days off; a second 2-day course given on days 31–32 of the study. Placebo capsules followed the identical schedule.

This “hit-and-run” dosing schedule mirrors the intermittent senolytic regimen conceptualized in the foundational mouse work ¹ — short exposure sufficient to eliminate senescent cells, relying on their non-renewal rather than continuous drug exposure for sustained effect. The same logic underlies dasatinib + quercetin intermittent dosing used in human senolytic trials.

Primary outcome

Safety: number of participants experiencing one or more treatment-emergent adverse event over the 12-month observation period.

Secondary outcomes (NCT04210986 registered)

• Serum biomarkers: CRP (C-reactive protein), COMP (cartilage oligomeric matrix protein)
• Physical performance: 6-minute walk test, timed-up-and-go, 4-meter walk speed, stair-climbing test, knee adduction moment
• Muscle strength: isokinetic dynamometry
• Patient-reported outcomes: Numeric Rating Scale (NRS) pain, WOMAC (Western Ontario and McMaster Universities Osteoarthritis Index) function subscale
• Cartilage quality: quantitative MRI (T2 relaxation times)
• Conversion to alternative therapies (surgical or injection-based)

Key results

Data source: Results below are from NCT04210986 posted ClinicalTrials.gov data, NOT the abstract body (which was inaccessible). Treat p-values as ClinicalTrials.gov registry-reported, not abstract-confirmed. The full paper will carry the authoritative statistics.

Safety (primary endpoint): No significant difference in adverse event rate between fisetin and placebo (p>0.9 per ClinicalTrials.gov posted results). Fisetin was well-tolerated.

Serum biomarkers: No statistically significant differences in CRP or COMP between groups at any timepoint.

Physical performance: No significant differences in 6-minute walk distance, timed-up-and-go time, 4-meter walk speed, stair-climbing performance, or knee adduction moment.

Patient-reported outcomes: No significant differences in NRS pain scores or WOMAC functional scores.

Cartilage quality (MRI T2): No significant difference in T2 relaxation time changes between groups.

Treatment conversion: 4 participants total converted to alternative therapy (1 fisetin, 3 placebo; p=0.474, NS).

Summary: null result across all efficacy endpoints. No significant benefit of fisetin over placebo was detected on any measured dimension of OA — biomarker, functional, patient-reported, or imaging.

Evidence-weight caveats

Several issues limit confidence in interpretation. Readers should weigh these before citing this abstract as settling the question of fisetin in OA:

1. Abstract-level evidence only. The abstract body was not accessible for automated review. Full methods, per-timepoint statistics, subgroup analyses, and any pharmacodynamic data await the full publication. Do not treat this page as equivalent to a full-paper extraction. gap/needs-full-paper-publication

2. Primary endpoint was safety, not efficacy. Phase I/II studies are typically powered for safety signals, not for efficacy. Whether the trial had adequate power to detect a clinically meaningful improvement in, e.g., WOMAC score is not stated in the registry data and cannot be confirmed from the abstract. gap/needs-replication

3. No pharmacodynamic confirmation. No endpoint in the registered outcomes measures whether fisetin actually reduced senescent cell burden in joint tissue (chondrocytes, synoviocytes). The trial cannot distinguish “fisetin failed because OA senescent cells weren’t cleared” from “fisetin cleared senescent cells but cell clearance doesn’t ameliorate established OA.” This is the key mechanistic gap. gap/no-mechanism

4. Dosing adequacy is uncertain. Two 2-day courses over 32 days — totalling 4 days of fisetin exposure — may be insufficient. Fisetin has a very short plasma half-life (~0.09 h rapid distribution; ~3.1 h terminal) ¹. Whether this schedule achieves senolytic concentrations in articular cartilage, a poorly vascularised tissue, is unknown.

5. Late-stage disease. Kellgren-Lawrence grade II–IV OA includes patients with structural damage. If senescent cells primarily accelerate cartilage loss in early OA, senolytics may not reverse established structural damage. The appropriate disease stage for senolytic intervention in OA is unsettled. gap/dose-response-unclear

6. Preregistered study — positive. NCT04210986 was registered prospectively, reducing selective reporting risk for primary safety data. However, without the full paper’s statistical analysis plan, secondary-endpoint interpretation is uncertain.

Significance in the fisetin evidence landscape

This trial is the first completed Phase I/II RCT of a senolytic compound for OA in humans to report results (as of mid-2026). Its null efficacy finding stands in contrast to:

• Preclinical evidence that intra-articular senescent cells drive OA progression and that their clearance (by navitoclax, D+Q, or p16 genetic ablation) is chondroprotective in mouse models ²
• The foundational fisetin senolytic characterization showing p16+ cell clearance in fat, spleen, liver, kidney, and blood of aged mice, with ex vivo human adipose tissue confirmation (n=3) ¹
• The Murray 2025 in-vivo mouse study showing intermittent fisetin improves frailty and grip strength in aged mice (functional endpoints, not OA) ³

The Tashman 2025 null result does not falsify fisetin’s senolytic mechanism per se but does constitute negative evidence for its clinical utility in symptomatic knee OA at this dose and schedule in this patient population. The framing is consistent with the general caution that mouse senolytic-to-human translation has been slower than preclinical data suggested.

Extrapolation to humans

Dimension	Status	Notes
Pathway conserved in humans?	yes	Cellular senescence and SASP are documented in human OA chondrocytes and synoviocytes
Phenotype conserved in humans?	yes	OA is a human disease; mouse models only approximate it
Replicated in humans?	no	This IS the human trial; efficacy not demonstrated; no independent replication exists

Limitations

• Full text unavailable at time of extraction. This page will require update when the full paper is published. gap/needs-full-paper-publication
• No pharmacodynamic data (senescent cell burden reduction) in any published endpoint.
• Phase I/II; likely underpowered for efficacy detection.
• Short follow-up context: the 12-month primary safety observation period may be too brief to detect disease-modifying structural effects (OA trials typically require ≥2 years for DMOAD endpoints).
• Single-site study (Steadman Philippon Research Institute); potential centre-effect on patient selection (elite-sports medicine patient population may differ from community OA populations).
• Asymmetric arm sizes (n=34 vs n=40) are unusual for a blinded RCT; randomisation ratio should be clarified in the full paper.
• No patient-level imaging biomarker data reported at abstract level.

Cross-references

• fisetin — compound page; mechanism, PK, clinical trial landscape
• cellular-senescence — senescent chondrocytes and synoviocytes as proposed OA drivers
• osteoarthritis — phenotype page; senescence framing of OA pathology
• senolytics — intervention class page; fisetin in context of other senolytic compounds
• yousefzadeh-2018-fisetin-senolytic — foundational fisetin senolytic paper (preclinical basis for this trial)
• murray-2025-fisetin-muscle-function — parallel fisetin in-vivo mouse study (frailty/grip; also null for direct senescent-cell count in muscle)

Footnotes

Footnotes

1. yousefzadeh-2018-fisetin-senolytic · n=8–9/group (lifespan arm); n=6–7 (acute senolysis arm); n=3 (human ex vivo) · in-vivo + in-vitro · p<0.05 (lifespan) · model: aged C57BL/6 ± FVB/n hybrid mice + human adipose ex vivo; establishes fisetin as senolytic and the “hit-and-run” mechanistic rationale ↩ ↩² ↩³

2. Mouse OA senolytic studies (navitoclax, D+Q, genetic p16 clearance) have shown chondroprotection in destabilization-of-the-medial-meniscus (DMM) and collagenase-injection models. Human OA differs in chronicity, joint mechanics, and disease stage at diagnosis. No study page currently exists for the preclinical OA-senolytic mouse literature; flag for future seeding. gap/needs-human-replication stub ↩

3. murray-2025-fisetin-muscle-function · n=39 (old fisetin) / n=31 (old veh) · in-vivo · p=0.0264 (frailty), p=0.0038 (grip) · model: aged male+female C57BL/6 mice (27–29 months); fisetin improves frailty and grip in mice; no effect on muscle senescent-cell counts (transcriptomic readout only) ↩