Dasatinib + quercetin and ovarian function: what the preclinical evidence demonstrates, what is missing in humans, and which hypotheses can be falsified

1. Executive scientific abstract

The available evidence does not support the simple chain “dasatinib + quercetin (D+Q) eliminates senescent ovarian cells, restores ovarian function, and extends women's health or lifespan.” Each link contains a different object, and none currently closes the complete chain.

In adult female ovaries without a gonadotoxic exposure, the study closest to natural reproductive aging found no statistically significant differences in follicles, pregnancy, or litter size under D+Q; it did observe an isolated reduction in lipofuscin in a young cohort. This does not demonstrate equivalence, but neither does it provide functional restoration. In young females exposed to doxorubicin, D+Q reduced p16, p21, SA-β-gal, and Il6/Mcp1 without reversing follicles, fibrosis, γH2AX, hormones, cyclicity, or fertility, and without a concordant apoptosis signal. In contrast, several cisplatin or cyclophosphamide studies reported preservation of follicles, AMH, cycle, or litter when D+Q was administered before or from the same day as the toxicant. These are models of acute-injury prevention in young females, not reversal of natural ovarian aging.

Attribution to “senolysis” is also unidentified. The founding mechanism from Zhu et al. was discovered in preadipocytes, endothelial cells, and fibroblasts, with cell-type-dependent sensitivity. Granulosa cultures use 1 nM dasatinib and 10–20 µM quercetin, whereas dasatinib selectivity in preadipocytes appeared from 50 nM and the HUVEC combination contained 100 nM. Moreover, p21 is an anti-apoptotic defense in the irradiated preadipocyte, but it also participates in luteal exit and reversible quiescence in the ovary. Lowering p21 alone neither identifies a senescent cell nor demonstrates that it died.

One competing explanation already has direct support: quercetin alone, administered preventively in cyclophosphamide, attenuated primordial-follicle loss and PI3K–AKT–mTOR/rpS6–FOXO3A activation, without dasatinib and without detectable primordial-follicle apoptosis at 24 h. This fits restraint of follicular recruitment and protection of growing follicles, not selective elimination of a senescent cell. It does not prove that quercetin explains all D+Q studies because route, exposure, and model differ, and the complete ovarian factorial has never been performed.

Human translation remains incomplete, but the gap is more precise than “no women have been treated.” Hickson and Justice included only two women each and studied adipose tissue or lung. Farr et al. conducted a randomized phase 2 trial in 60 postmenopausal women: D+Q did not modify the primary bone-resorption endpoint at 20 weeks; P1NP increased transiently, and an exploratory subgroup with high p16 showed skeletal signals. The trial measured neither ovary, AMH, cycle, nor STRAW+10. Female human exposure to D+Q therefore exists, but the human chain qualified ovarian cell → free exposure → selective death → ovarian function → organ event does not.

Scientific delta of this report (L10-2-REPORT-EN-C1): the Farr 2024 randomized trial corrects the previous formulation of the human gap. What is absent is not women's exposure to D+Q, but ovarian mechanism and outcome; moreover, the overall primary skeletal endpoint was null, so an exploratory signal stratified by lymphocyte p16 cannot be transported to the ovary.

The link to female longevity remains open and cannot be inferred from AMH, follicles, cyclicity, litter, or age at menopause. A longer reproductive lifespan can associate with skeletal or metabolic benefits and, simultaneously, with greater exposure to hormone-dependent risks. The correct scientific unit is an organ event with mediation and competing risks, not a reproductive surrogate.

The operational conclusion is negative but actionable: D+Q is not mature as an ovarian senolytic or as a female-longevity strategy. The smallest next step is neither a human ovarian trial nor another chemotherapy model comparing the cocktail with vehicle. It is first to resolve, without new animals, whether Fang's systemic results and García's ovarian results are quantitatively compatible; only then, if that gate opens, to replicate the adipose percentage in females under Fang's exact schedule and split dasatinib, quercetin, and the combination. Any senolysis claim first requires identification of durable arrest separate from luteinization, atresia, and cycling, plus measurement of the free pharmacological species in the compartment.

2. Scientific question and relevance

The question is whether D+Q modifies ovarian function by selectively eliminating a senescent population and whether that effect can be causally connected to women's health and longevity. Answering it requires separating five questions:

1. Is the causal agent dasatinib, quercetin, or the combination?
2. Does the affected cell meet durable-arrest criteria distinguishable from luteinization, atresia, reversible quiescence, and proliferation?
3. Is there selective apoptosis, or do markers merely decrease?
4. Does an independent ovarian function follow the cellular change?
5. Does that trajectory mediate an organ event in women under competing risks?

The relevance is high because cellular senescence is a plausible mechanism of tissue injury, but the ovary is especially vulnerable to identity errors. p16, p21, SA-β-gal, lipofuscin, IL-6, AMH, follicle counts, and litter are not interchangeable. Neither are gonadotoxic injury in a young female and natural reproductive transition.

3. Scope, population, and life stage

The primary biological estimand is the adult female undergoing natural reproductive aging without chemotherapy. For human translation, the relevant population is women classified by STRAW+10: late reproductive (−3b/−3a), transition (−2/−1), and early postmenopause (+1c), analyzed separately. Chronological age alone does not substitute for stage.

Doxorubicin, cisplatin, and cyclophosphamide models are retained as injury contrasts. They inform prevention of or response to acute damage but are not pooled with natural aging. IVF granulosa cultures are human-cell evidence, not an intervention in a woman. Trials in adipose tissue, lung, or bone are extraovarian human evidence.

Prescription, dose recommendations, fisetin ranking, combination with NAD+ precursors, product proposals, user data, and any inference from fertility or menopause to longevity without an organ event are excluded.

4. Background knowledge and mechanistic map

4.1 Operational definitions

• D+Q: a pharmacological combination; it is not one mechanism until neither component alone is shown to suffice.
• Senolysis: selective death of a cell that already met durable-arrest criteria plus an orthogonal readout, with survival of non-senescent sister cells.
• Ovarian function: an endpoint independent of the senescence marker. The marker cannot be the success criterion.
• Exposure: free dasatinib and free quercetin aglycone in the compartment where the effect is claimed; total or conjugated concentration does not substitute for free species.
• Healthspan/longevity: years free of an organ event or survival, with mediation and competing risks; not AMH, follicles, cycle, pregnancy, or litter.

4.2 Favored chain that the literature would have to demonstrate

D or Q or D+Q
  → free species in the ovary
  → occupancy of a prespecified target
  → selective death of a qualified durable arrest
  → change in an independent ovarian function
  → endocrine or systemic trajectory
  → organ event in women
  → net effect under competing risks

The current literature breaks before the third link. No in-vivo ovarian measurement includes vehicle, dasatinib alone, quercetin alone, and D+Q together with free exposure, prior cell identity, selective death, and independent function.

4.3 Competing mechanisms

Mechanism A — marker reduction without functional senolysis. D+Q lowers p16/p21/SA-β-gal or cytokines without selective apoptosis or structural restoration. Gao is the clearest instance.

Mechanism B — protection against a gonadotoxic agent. Before or from toxicant initiation, the intervention reduces damage, apoptosis of growing follicles, or accelerated primordial recruitment. The outcome may be cytoprotection, not senescent-cell elimination.

Mechanism C — quercetin alone. Modulation of PI3K–AKT–mTOR/rpS6–FOXO3A preserves primordial quiescence in cyclophosphamide. The micromolar band in ovarian cultures resembles published quercetin pharmacology more than dasatinib pharmacology in preadipocytes.

Mechanism D — dasatinib or other pleiotropy. The reduction in primordial follicles observed with D+Q without doxorubicin could depend on dasatinib, the interaction, or another toxicity. No separated arm exists.

Mechanism E — systemic effect. Changes in adipose tissue, energy, immunity, endothelium, or an endocrine axis can modify reproductive readouts without local ovarian senolysis. Fang shows an adverse systemic direction in females; García did not test the same regimen.

5. Evidence method

The complete scoping, evidence-map, verification, mechanistic-synthesis, compute-decision, hypothesis, adversarial-review, and experiment-design artifacts were integrated. Decisive claims were preserved by layer: human, animal, in vitro/ex vivo, computational, and inferred.

The prior ovarian search identified 16 PubMed records for D+Q and ovary/follicle; available primary studies were audited in PubMed/PMC. This report also verified two critical updates: the recent Su 2026 primary study (PMID 42011821) and the phase 2 trial in 60 postmenopausal women by Farr 2024 (PMID 38956196). Abstracts were checked against methods and figures when available. A value of p>0.05 was never interpreted as equivalence. Sections, fields, ovaries, and mating rounds were not promoted to numbers of females.

No computation was run. The distribution of EFNB1, EFNB3, and PI3KCD in atlas GSE202601 cannot distinguish ligand from occupied kinase or cell death from RNA abundance. GSE264327 lacks single-drug arms. No expression analysis could change the hypothesis ranking.

6. Evidence map

7. Contradictory evidence and null findings

7.1 The marker does not predict structure or function

Gao demonstrates the strongest dissociation: p16, p21, SA-β-gal, Cdkn2a, Cdkn1a, Il6, and Mcp1 decrease; Tgfβ1, fibrosis, γH2AX, follicles, hormones, cyclicity, and fertility do not recover. TUNEL shows no evident separation, and Bax/caspase 3 do not differ. Even if an early apoptotic peak was missed, the proposed death did not restore the organ.

García adds a null close to natural aging: follicles, pregnancy, and litter do not separate. Hense 2022 and the MASLD abstract again separate marker burden or pregnancy from the reserve. These observations are not equivalence results, but they preclude treating any decrease in p16/p21/lipofuscin as ovarian restoration.

7.2 Apoptosis changes direction across models

• Gao: TUNEL does not separate groups.
• Du: D+Q reduces TUNEL after cisplatin, consistent with protection.
• Su: D+Q increases TUNEL, interpreted by the authors as senescent-cell elimination.

Without a qualified cell population before intervention, all three directions remain compatible with injury, protection, sampling time, or nonspecific death. None demonstrates selective senolysis.

7.3 Toxicant and schedule dominate

D+Q does not restore follicles under doxorubicin but is associated with preservation under cisplatin or cyclophosphamide. γH2AX is not reversed in Gao and decreases in Su. Treatment is preventive in Du, Xu, and Su. Delayed intervention after established chemical injury, represented only weakly by the Ávila abstract, does not provide a positive rescue. The best-positioned explanation is that toxicant identity and exposure timing condition the result more strongly than the “senolytic” label.

7.4 Female systemic evidence is not uniformly favorable

Fang observed an adverse phenotype in females: more adipose tissue and cytokines, lower energy expenditure, and worse novel-object recognition. However, gonadal-adipose p16, glucose, and insulin did not worsen, and the water maze was null. The combination did not yield a coherent “less senescence = better healthspan” signal. It also cannot be merged with García: Fang administered one monthly dose in DMSO/canola from 4 to 13 months; García used three consecutive days per month in PHOSAL/PEG/ethanol from 6 to 10 months.

7.5 The strongest female human trial has a null primary endpoint

In Farr, D+Q did not reduce CTx in the full set of 60 women at 20 weeks. The P1NP increase was transient, and the high-p16 subgroup was exploratory. Lymphocyte p16 is not a qualified ovarian cell. The study shows that D+Q can be investigated in women and that effects may depend on context, but it validates neither the ovary nor a global longevity benefit.

8. Multiscale mechanistic synthesis

8.1 Protein and exposure

Dasatinib inhibits multiple tyrosine kinases. Zhu linked its action to ephrin-dependent survival in preadipocytes, but the decisive siRNAs targeted the ligands EFNB1/EFNB3, PI3KCD, p21, BCL-xL, and PAI-2. Expression of a ligand does not identify the occupied kinase. Neither target occupancy nor free concentration contemporaneous with the effect has been measured in the ovary.

Quercetin is pleiotropic. In ovarian cultures it appears at 10–20 µM, a band compatible with multiple cellular effects. Total human concentration or hydrolyzed conjugate is not equivalent to free ovarian aglycone. CD38 and COMT remain unproven hypotheses in this context.

8.2 Cell identity

p21 is not a specific marker. Its loss can remove an anti-apoptotic defense in preadipocytes; its presence can accompany luteal cell-cycle exit or reversible granulosa quiescence. CDKN2A also does not necessarily identify p16INK4a rather than p14ARF in RNA data. SA-β-gal can accompany injury or lysosomal activity. The cell that dies under D+Q cannot be used retrospectively to define what senescence was.

The required gate is Q0-STATE: durable arrest plus an orthogonal readout, explicitly failing in luteal, atretic, and cycling cells from the same ovary. Without this gate, “ovarian senolysis” is circular.

8.3 Follicle and tissue

In cyclophosphamide, quercetin alone reduces phosphorylation of PI3K/AKT/mTOR/rpS6/FOXO3A, preserves nuclear FOXO3A localization, and attenuates primordial recruitment. Apoptosis localizes to growing follicles, not the primordial follicle. This mechanism explains reserve preservation through restraint of recruitment and early cytoprotection.

In doxorubicin, lowering Il6/Mcp1 does not move Tgfβ1 or fibrosis. Matrix is partly decoupled from the measured SASP branch. In Gao's no-toxicant arm, D+Q reduces primordial follicles, the opposite of simple preservation. In García, reserve does not change statistically significantly. D+Q therefore has no universal follicular direction.

8.4 Female system

The system may change even when the ovary was not measured. Fang supports a possible adipose-energy chain: ACC up, brown-fat thermogenesis and VO2 down, adipose percentage up, and cytokines afterward. An endocrine alternative—altered cycle or estradiol—was not measured. A thyroid alternative proposed in the discussion was not measured either. The ovary is neither exonerated nor implicated.

8.5 Woman and organ event

Human data span adipose tissue, lung, physical function, and bone. They do not span ovarian cell identity, free ovarian exposure, or reproductive function under D+Q. Farr demonstrates that an intermediate endpoint in another organ can have a null primary result and time- or subgroup-dependent secondary findings. This reinforces the need not to merge organs or turn a marker into healthspan.

9. Computational layer

Decision: skip. GSE202601 permits counting EFNB1, EFNB3, and PI3KCD, but:

• ligand transcript is not occupied kinase;
• non-detection in snRNA is not absence;
• RNA identity demonstrates neither durable arrest nor death;
• the atlas is not stratified by STRAW+10.

GSE264327 contains control, cyclophosphamide, and cyclophosphamide+D+Q in ovary and oocyte, without D alone or Q alone and with n=3 per deposited condition. Reanalysis cannot distinguish the pharmacological hypotheses. BioNeMo offers no valid inference about exposure, senolysis, or function for this question.

10. Primary hypothesis

L10-2-AR-H1 v1 — García's ovary and Fang's systemic harm do not represent the same exposure

Falsifiable statement. Once extracted with mean, error, and n, the interval of body-weight change under D+Q in García's experiment 2 contains Fang's female point effect; if it does not contain it, García's regimen is excluded as the carrier of Fang's systemic phenotype.

Proposed mechanism. Duty cycle, vehicle, age at initiation, and duration alter effective exposure and can change weight, adipose tissue, and endocrine axis even with the same nominal 5+50 ratio.

Evidence for. The methods differ; Fang reports increased female weight/adiposity, and García states that no weight difference was observed. García's assertion that it was “the same scheme” does not match the methods.

Evidence against. Means and intervals were not extracted from García's supplement. “No difference” can conceal an effect of the same sign and a magnitude compatible with Fang.

Predictions.

1. If García's 95% CI contains Fang's point effect, the hypothesis dies as a numerical divergence.
2. If it excludes that effect, García's ovarian results cannot be used as the ovary of Fang's phenotype.
3. García's experiment 1 and fisetin do not vote.

Kill criteria. Compatible intervals; absence of mean/error/n that makes the contrast non-estimable; or a divergence claim based only on p>0.05.

Status and maturity. Proposed. H1 for non-identity of methods; H0 for quantitative divergence. Confidence: 0.86 in methodological non-identity and 0.42 that the supplement will exclude Fang's effect.

11. Competing hypothesis

L10-2-AR-H2 v1 — if Fang's adipose effect replicates, the ovarian axis mediates only if cycle or estradiol accompanies it

Falsifiable statement. Under Fang's exact regimen in C57BL/6 females without a gonadotoxic agent, perigonadal adipose percentage increases versus vehicle. In the same experimental unit, cyclicity or phase-fixed estradiol either separates—keeping endocrine mediation plausible—or remains within a prespecified margin while adipose tissue increases—favoring extraovarian pleiotropy.

Competing mechanisms.

• Direct adipose/energy pleiotropy of one or both drugs.
• Alteration of the reproductive axis that modifies adiposity without necessarily changing follicle counts.
• Inflammation secondary to adipose expansion.
• Systemic senolysis, less favored because gonadal p16 did not decrease in females and fisetin did not reproduce the phenotype.

Predictions.

1. The systemic coprimary is perigonadal adipose percentage of body weight, not a post hoc selected mass, SASP, NOR, glucose, or p16.
2. If the percentage increases and cycle/estradiol also changes, the axis remains a candidate mediator; thyroperoxidase is not named without measurement.
3. If the percentage increases with the axis in equivalence, an extraovarian effect is favored.
4. If quercetin alone reproduces D+Q, the pair is unnecessary for that endpoint; the same applies to dasatinib alone.

Adversarial evidence. Fang's ANOVA was followed by multiple t tests; n varies by panel; NOR may depend on reduced distance; SASP may be a consequence of adiposity. The study measured neither ovary, cycle, estradiol, nor TSH. A future non-significant result is not equivalence.

Kill criteria. A replication 95% CI excludes Fang's published effect in the same estimand; the depot does not replicate in two runs; or the call is adjudicated using a surrogate endpoint.

Status and maturity. Proposed. H1 for the lipid chain observed within Fang; H0 for endocrine mediation. Confidence: 0.57 that SASP is not independent of adipose tissue and 0.24 in endocrine mediation.

Preserved pharmacological rival

L10-2-HG-H2 v1 remains weakened but alive: quercetin alone separates from cyclophosphamide on primordial follicles, but equivalence with control was not demonstrated (P=0.275 does not establish it), D+Q was not split in the same experiment, and the 24 h western is not linked to each female in the day 7 count. Its maturity is H1 for that concrete separation and H0 for explaining the pair or natural aging.

12. Translational hypothesis

L10-2-AR-HT1 v1 — no reproductive or extraovarian surrogate yet qualifies as mediator of an event

Falsifiable statement. In women classified by STRAW+10, only a parameter that has passed cell identity, free exposure, selective death, and independent ovarian function can be tested as mediator of a single-organ event. It must improve out-of-sample prediction over stage and a systemic panel under competing risks. AMH, follicles, cycle, age at menopause, adipose p16/p21, and Fang's direction are excluded as mediators.

Predictions.

1. An AMH or age-at-menopause trial does not test this hypothesis.
2. High lymphocyte p16 in Farr's exploratory analysis does not transport ovarian identity.
3. Opposite adipose signs in Fang and Hickson must be preserved as different populations and organs.
4. A skeletal or physical signal can exist without the ovary and therefore does not demonstrate ovarian mediation.

Kill. A qualified ovarian parameter predicts and mediates an event better than stage and the systemic panel in external validation; or the effect is explained by direct drug action in the organ, which kills ovarian mediation while preserving pleiotropy.

Status and maturity. Proposed and parked. H0. HUMAN_QA_REQUIRED. It is not a biomarker.

13. Falsifiable predictions and kill criteria

The original senolytic hypothesis remains H0 because it has not been tested with identity, exposure, and a factorial. It is not considered refuted for a cell that has not yet been identified; it is outside what existing studies demonstrate.

14. Discriminating experiment

Step 1 — D0-WEIGHT-EXTRACT: zero animals

Extract separately:

• Fang, females, vehicle versus D+Q: body weight and perigonadal percentage.
• García experiment 2, females, vehicle versus D+Q: body weight.
• García experiment 1: archive without a vote.

Priority: table/source data; if unavailable, two independent digitizations. A difference >5% of the panel's axis span makes the panel non-estimable. With SEM, the error of the difference is √(SEM₁²+SEM₂²). If Fang only provides n=8–20, conservatively use n=8, df=14, t=2.145. Compatible means García's 95% CI contains Fang's point effect.

• Compatible: AR-H1 dies as divergence. The animal cohort does not open.
• Excludes: a Fang replication may open.
• Non-estimable: one retry with a higher-resolution source; animals do not open immediately.

Step 2 — D0-FANG-WAT-A: conditional minimal replication

Only if step 1 excludes or remains non-estimable after retry and an independent replication is authorized:

• C57BL/6 females; substrain/provider fixed in advance.
• No gonadotoxic agent.
• Exact Fang schedule: one oral administration per month, 2% DMSO in canola vehicle, start at 4 months, and readout seven days after the last dose at 13 months.
• Two arms: vehicle and D+Q.
• Confirmatory unit: cage; floor of four cages per arm, three females per cage. This is not a power calculation.
• Systemic coprimary: perigonadal percentage of body weight.
• Axis coprimary: proportion of days in estrus over 14 smears before the last dose.
• Weight, wet mass, VO2, and follicles are archived; they do not replace coprimary endpoints.

A p>0.05 result is inconclusive. AR-H2 dies only if the replication 95% CI excludes Fang's digitized effect in the same estimand.

Step 3 — D0-FANG-WAT-B: split the pair

Only if the percentage increases in A: add dasatinib alone and quercetin alone under the same schedule and unit. This factorial identifies which component carries the depot effect. It does not identify senolysis.

Step 4 — ovarian senolysis, still closed

Only after Q0-STATE and Q0-EXPOSURE:

• vehicle, D, Q, and D+Q;
• death of the qualified state versus sister, luteal, and cycling cells;
• independent function;
• female/donor as n;
• natural aging separated from gonadotoxic injury.

15. Biomarkers and stratification

There is no validated biomarker. The following are research readouts:

• Free dasatinib and free quercetin aglycone in the compartment: exposure gate.
• Durable arrest plus an orthogonal readout: identity gate.
• Selective death versus cellular controls: senolysis gate.
• Independent ovarian function: endpoint, not marker.
• STRAW+10: minimum human stratification.

p16/p21, SA-β-gal, lipofuscin, AMH, follicles, menstrual regularity, and age at menopause cannot be presented as one biomarker of senolysis, health, or longevity. Lymphocyte p16 used by Farr is exploratory stratification of skeletal response, not an ovarian measure.

16. Individual variability

Expected heterogeneity includes reproductive stage, baseline reserve, reproductive history, toxicant and time from injury, tissue composition, luteal state, adiposity, renal/hepatic function, free exposure, vehicle, microbiota, and dosing schedule. Human studies must distinguish:

• natural transition from iatrogenic ovarian insufficiency;
• follicle-bearing ovary from late postmenopause;
• local action from systemic disease;
• organ-specific response from a global index.

Women's participation does not equal estimation of stage modification. The Hickson/Justice pilots and Farr trial inform safety and extraovarian pharmacodynamics in specific populations; they do not estimate ovarian transition.

17. Pharma relevance and maturity

Pharma relevance is currently negative derisking, not readiness:

1. No nominated and occupied ovarian target exists.
2. D, Q, and D+Q remain three exposures.
3. Preclinical benefit is concentrated in prevention of gonadotoxicity; reversal of natural aging has not been demonstrated.
4. Fang's female systemic direction is adverse and must be replicated before any healthspan interpretation.
5. Farr shows that a female human trial is feasible, but the overall primary endpoint was null and the subgroup was exploratory.
6. No female-longevity indication or validated stratum exists.

Maturity:

• D+Q ovarian senolysis: H0.
• Preventive quercetin in cyclophosphamide: H1 for the specific model.
• Marker–function dissociation in doxorubicin and the natural window: H1 from the literature, without formal equivalence.
• Female human extraovarian exposure: H3/H4 by organ design, without ovarian translation.
• Translational event hypothesis: H0, HUMAN_QA_REQUIRED.

18. Limitations

1. No systematic review with a PRISMA protocol was conducted; the search was directed to the estimand.
2. Several studies have small n, ambiguous units, or counts per section.
3. Ávila 2024 and Hense 2026 remain abstract-layer evidence; n and dose were not invented.
4. Free ovarian exposure and target occupancy are unavailable.
5. No in-vivo ovarian study splits D, Q, and D+Q.
6. None satisfies the required cell-identity gate.
7. Schedules, vehicles, strains, and toxicants are heterogeneous.
8. p>0.05 does not establish equivalence; several “nulls” may be imprecise.
9. Farr did not measure the ovary, and its high-p16 analyses were exploratory.
10. D0-WEIGHT-EXTRACT was not executed; the smallest experiment remains a design, not a result.
11. Animal data authorize neither doses nor intervention in people.
12. The longevity bridge has no measured event, mediation, or competing risks.

19. Conclusions

D+Q can modify ovarian markers and, in some preventive chemotherapy models, preserve follicles or reproductive functions. That does not demonstrate ovarian senolysis, reversal of natural aging, or a longevity benefit. The evidence closest to aging without a toxicant is functionally null or imprecise; evidence under doxorubicin separates marker from structure; and evidence under cyclophosphamide admits a quercetin-alone explanation through restraint of recruitment.

The new female human evidence does not change that verdict. Farr demonstrates randomized exposure in postmenopausal women and a null overall primary skeletal endpoint, not a treated ovary. The human gap is identity, exposure, selectivity, function, and mediation.

The scientific priority is neither to expand enthusiasm nor to repeat cocktail versus vehicle. It is to separate exposure, drugs, cells, and organs. The first discriminant costs zero animals: determine whether García's and Fang's published weights are compatible. Only an incompatibility or persistent non-estimability justifies replicating Fang's adipose percentage, and only a positive replication justifies splitting D from Q. Ovarian senolysis remains closed until a cell is qualified before drug exposure.

20. References

1. Zhu Y, Tchkonia T, Pirtskhalava T, et al. The Achilles' heel of senescent cells: from transcriptome to senolytic drugs. Aging Cell. 2015;14:644–658. DOI: 10.1111/acel.12344. PMID: 25754370.
2. Gao Y, et al. Increased cellular senescence in doxorubicin-induced murine ovarian injury: effect of senolytics. GeroScience. 2023;45:1775–1790. DOI: 10.1007/s11357-023-00728-2. PMID: 36648735.
3. García DN, et al. Study of D+Q/fisetin, ovarian reserve, and fertility in C57BL/6 females. GeroScience. 2024. DOI: 10.1007/s11357-024-01089-0. PMID: 38358579.
4. Du X, et al. Study of metformin and D+Q in cisplatin-induced ovarian insufficiency. Oxidative Medicine and Cellular Longevity. 2022;2022:9144644. DOI: 10.1155/2022/9144644. PMID: 35693700.
5. Xu M, et al. The Role of Cellular Senescence in Cyclophosphamide-Induced Primary Ovarian Insufficiency. International Journal of Molecular Sciences. 2023;24:17193. DOI: 10.3390/ijms242417193. PMID: 38139022.
6. Su et al. Senolytics alleviate cyclophosphamide-induced premature ovarian insufficiency by eliminating senescent cells. European Journal of Histochemistry. 2026. DOI: 10.4081/ejh.2026.4537. PMID: 42011821.
7. Guan et al. Study of D+Q nanoparticles in cyclophosphamide-induced ovarian injury and postovulatory aging. Cellular and Molecular Life Sciences. 2025. DOI: 10.1007/s00018-025-05697-7. PMID: 40249520.
8. Li J, et al. Quercetin prevents primordial follicle loss via suppression of PI3K/Akt/Foxo3a pathway activation in cyclophosphamide-treated mice. Reproductive Biology and Endocrinology. 2021. DOI: 10.1186/s12958-021-00743-y. PMID: 33892727.
9. Hense JD, et al. Study of ob/ob obesity, ovarian senescence, and D+Q. GeroScience. 2022. DOI: 10.1007/s11357-022-00573-9. PMID: 35460445.
10. Ávila et al. Study of D+Q/fisetin after VCD-induced chemical estropause. Life Sciences. 2024;123073. DOI: 10.1016/j.lfs.2024.123073. PMID: 39307182.
11. Hense et al. Study of Western diet/MASLD, D+Q, and ovarian function. Biology of Reproduction. 2026. DOI: 10.1093/biolre/ioag060. PMID: 41863601.
12. Fang Y, et al. Comparative study of D+Q and fisetin in male and female C57BL/6 mice. GeroScience. 2023. DOI: 10.1007/s11357-023-00843-0. PMID: 37296266.
13. 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. DOI: 10.1016/j.ebiom.2019.08.069. PMID: 31542391.
14. Justice JN, et al. Senolytics in idiopathic pulmonary fibrosis: Results from a first-in-human, open-label, pilot study. EBioMedicine. 2019. DOI: 10.1016/j.ebiom.2018.12.052. PMID: 30616998.
15. Farr JN, Atkinson EJ, Achenbach SJ, et al. Effects of intermittent senolytic therapy on bone metabolism in postmenopausal women: a phase 2 randomized controlled trial. Nature Medicine. 2024;30:2605–2612. DOI: 10.1038/s41591-024-03096-2. PMID: 38956196.
16. Xu M, Pirtskhalava T, Farr JN, et al. Senolytics improve physical function and increase lifespan in old age. Nature Medicine. 2018. DOI: 10.1038/s41591-018-0092-9. PMID: 29988130.
17. Jin et al. Single-cell/nuclear atlas of human ovarian aging. Nature Aging. 2025. DOI: 10.1038/s43587-024-00762-5. PMID: 39578560.

Claims added in REPORT_EN: L10-2-REPORT-EN-C1.

Hypotheses touched: L10-2-AR-H1 v1, L10-2-AR-H2 v1, L10-2-AR-HT1 v1, L10-2-HG-H2 v1, L10-2-SCOPE-H1 v1.

This is a scientific report and does not constitute medical advice, a prescription, or authorization to use dasatinib, quercetin, or their combination.