evipedia.ai
← Back

PAI-1 Inhibitors for Health & Longevity

Evidence Review created on 09/01/2026 using AI4L / Opus 5

Also known as: Plasminogen Activator Inhibitor-1 Inhibitors, PAI-1 Antagonists, SERPINE1 Inhibitors, TM5614, RS5614, TM5509, TM5484, TM5441, TM5275, Tiplaxtinin, PAI-039, MDI-2517

Motivation

PAI-1 inhibitors (plasminogen activator inhibitor-1 inhibitors) are experimental oral medications built to block one protein. That protein is the body’s main brake on the system that dissolves blood clots, and it also drives scar tissue formation and pours out of worn-out cells that have stopped dividing but refuse to die. Its levels climb with age, with body fat, and with long-running inflammation.

Interest in switching it off grew out of an unusual natural experiment. In a genetically isolated farming community in Indiana, one extended family carries a rare mutation that halves production of the protein. Carriers turned out to live longer than their relatives, with longer chromosome end-caps and no diabetes. In parallel, a Japanese university group had spent two decades turning a computer-designed molecule into an oral drug that has now reached late-stage cancer trials.

This review examines what is known about blocking this protein in people: how the compounds work, which effects have actually been measured in completed trials, what remains laboratory and animal work only, the safety signals reported so far, and who paid for the evidence.

Benefits - Risks - Protocol - Conclusion

This section collects high-level overviews of PAI-1 (plasminogen activator inhibitor-1, the protein that blocks clot breakdown, or fibrinolysis) and of the drugs designed to inhibit it.

Five qualifying items were found, so the list is not padded. None comes from a priority expert platform: independent site searches of foundmyfitness.com, peterattiamd.com, hubermanlab.com, chriskresser.com, lifeextension.com and lifespan.io returned no article, episode or newsletter on this intervention. The single near-miss was a passing mention of PAI-1 as one of seven proteins inside an epigenetic ageing clock on FoundMyFitness, which does not discuss the drugs at all.

Grokipedia

  • Plasminogen activator inhibitor-1

    Grokipedia has no page for the drug class, so this is its dedicated page on the target protein, covering gene location, the vitronectin interaction, disease associations and the small-molecule inhibitors under development.

Examine

No Examine article exists for PAI-1 inhibitors. Examine.com covers dietary supplements and does not typically cover investigational prescription medications; its only search result is an outcome-database entry describing the PAI-1 biomarker, which is a database record rather than an article about the intervention.

ConsumerLab

No ConsumerLab article exists for PAI-1 inhibitors. ConsumerLab tests dietary supplements and does not typically cover investigational prescription medications; the site search returned only unrelated recall notices for herbal weight-loss products with similar-looking names.

Systematic Reviews

No systematic reviews or meta-analyses for PAI-1 Inhibitors were found on PubMed as of 1 September 2026.

Neither side of this intervention’s trade-off is represented at this evidence level: no systematic review or meta-analysis exists for the claimed effect (reduced senescence, fibrosis and clotting) or for the principal risk (bleeding from unopposed clot breakdown). Both are unrepresented.

Mechanism of Action

PAI-1 is a serpin — a serine protease inhibitor — built from the gene SERPINE1 (the gene carrying the instructions for making PAI-1). It is the fastest-acting brake on tissue-type plasminogen activator (tPA, the enzyme that starts clot dissolution) and urokinase-type plasminogen activator (uPA, its tissue-remodelling counterpart), so it suppresses production of plasmin, the clot-dissolving enzyme, and lets fibrin persist. Binding to vitronectin holds it in its active shape. Transforming growth factor beta (the master switch for scar formation), p53 (the cell’s damage-response gene), low oxygen and insulin all raise its output, and it is among the most abundant products of the senescence-associated secretory phenotype (the inflammatory mixture that worn-out cells release).

The TM-series compounds do not plug the enzyme’s cutting site. Crystallography places them in the vitronectin pocket, destabilising active PAI-1 and speeding its clearance, a two-step mechanism demonstrated for TM5484. TM5614’s half-maximal inhibitory concentration (IC50, the amount needed to block half the target’s activity) is 3.63 micromolar, against 9–12 micromolar for tiplaxtinin.

Two readings compete. One treats PAI-1 as an active driver of senescence, fibrosis and clotting; the other treats it as a downstream marker of fat mass and inflammation, in which case lowering it would not shift outcomes.

TM5614 is orally active and dosed once daily. Its human half-life, tissue distribution and metabolising enzymes are not published in peer-reviewed form. Selectivity is reported only as an absence of action on platelets and clotting factors; cross-reactivity with other serpins is unpublished.

Historical Context & Evolution

PAI-1 was identified in the early 1980s as the fast-acting inhibitor of tissue plasminogen activator, and the first therapeutic idea was strictly cardiological: high levels predicted recurrent heart attacks, so blocking the protein should reopen the clot-dissolving pathway. In the 1990s an Old Order Amish family in Berne, Indiana was found to carry a mutation abolishing PAI-1 outright. People with two copies bleed after trauma, surgery and childbirth but are otherwise healthy, which established that the protein can be reduced without threatening life.

Wyeth’s tiplaxtinin, the first orally available antagonist, cleared rat models of thrombosis (blood-clot formation) in 2008 but was not carried forward. The company never published its reasons, so this is a gap in the record rather than evidence against the target.

From the early 2000s a Tohoku University group screened roughly two million virtual compounds against the PAI-1 crystal structure, narrowed 96 hits to TM5275, then synthesised more than 1,400 derivatives to arrive at TM5614. The framing shifted in parallel: work in nitric-oxide-blocked mice and in mice lacking klotho (a gene whose loss causes rapid, premature ageing) recast PAI-1 as a mediator of senescence and vascular ageing rather than a clotting factor. The 2017 report that Amish carriers outlived their relatives, widely covered in the press, carried it into longevity medicine.

First-in-human dosing followed in 2019 in chronic myeloid leukaemia (a blood cancer). Whether the driver account or the marker-only account holds has not been settled by any completed trial.

Expected Benefits

High 🟩 🟩 🟩

No benefit reaches High: no outcome has been reproduced in more than one controlled human trial, because every completed study of a PAI-1 inhibitor is a single small open-label or phase II study and only one of them was placebo-controlled.

Medium 🟩 🟩

Deeper Molecular Remission in Chronic Myeloid Leukaemia

Adding TM5614 to a tyrosine kinase inhibitor (a targeted cancer medication) pushed more patients into deep molecular remission, the response depth that must be reached and held before treatment can stop. The proposed mechanism is eviction of dormant leukaemic stem cells from their bone-marrow niche. The evidence basis is a single-arm phase II trial in 33 patients, compared against historical rates rather than a concurrent placebo group. No drug-related bleeding or clotting abnormality was reported (Takahashi et al., 2023).

Magnitude: Cumulative incidence of deep molecular response at 12 months was 33.3% (95% confidence interval, the range within which the true value probably lies, 18.0–51.8%), against a historical 8% on the targeted medication alone.

Restored Tumour Response after Checkpoint-Inhibitor Failure

In melanoma that had already stopped responding to an anti-PD-1 antibody (a medication that releases an immune brake called programmed cell death protein 1), adding TM5614 produced measurable tumour shrinkage. The proposed mechanism is remodelling of the tumour’s surrounding tissue, lowering checkpoint-molecule expression and suppressing the scaffold cells and scavenger cells that shield tumours from immune attack. The evidence basis is one open-label, single-arm, multicentre phase II trial in 39 Japanese patients, with no control arm and a short eight-week endpoint (Fujimura et al., 2024).

Magnitude: Overall response rate at eight weeks was 25.9% (95% confidence interval 12.9–44.9%) among 27 evaluable patients refractory to anti-PD-1 therapy; the disease control rate was 62.0%.

Low 🟩

Longer Lifespan and a Healthier Metabolic Profile with Lifelong Lower PAI-1

Carrying one broken copy of SERPINE1 from birth is associated with longer life and better glucose handling. This is a lifelong genetic halving, not a drug effect: the basis is one uncontrolled family cohort plus a matching mouse model. No trial has tested adult-onset inhibition (Khan et al., 2017).

Magnitude: In the Berne Amish kindred, median survival was 85 years in carriers of one broken copy against 75 years in non-carrier relatives, with 10% longer white-blood-cell telomeres, 28% lower fasting insulin and diabetes in 0% against 7%; a mouse engineered with the same variant lived 17% longer (Khoddam et al., 2025).

Slower Progression of Viral Pneumonia ⚠️ Conflicted

The rationale is that blocking PAI-1 counters the clotting, inflammation and scarring that worsen viral pneumonia. Every secondary measure favoured the drug, but the placebo-controlled trial missed its primary endpoint and closed under-enrolled (Hirai et al., 2024). The net reading is a real direction of effect that remains unconfirmed.

Magnitude: Total oxygenation-deterioration score over 14 days was 1.5 with TM5614 versus 4.0 with placebo (p = 0.22, meaning a difference this size could easily arise by chance), and oxygen was needed for 2.0 versus 3.5 days (p = 0.34).

Speculative 🟨

Lower Estimated Biological Age

Four months of TM5614 in 20 relatively healthy adults aged 50–75 lowered estimated epigenetic age by 2–3 years and shifted senescence markers. The basis is one uncontrolled open-label study, per a science-agency news report.

Reduced Senescent-Cell Burden and Arterial Stiffening

TM5441 blocked the senescence marker p16 and preserved telomere length in mouse arteries, while cutting blood pressure and aortic scarring. No human arterial-stiffness endpoint exists; the human senescence data are uncontrolled.

Antifibrotic Protection across Organs

TM-series compounds reduced scarring in mouse kidney, liver, bowel and lung models. The basis is animal work only; a trial in systemic sclerosis (a scarring autoimmune disease) has reported nothing.

Preservation of Muscle Fibre Size

A PAI-1 inhibitor prevented age-related shrinkage of muscle fibres in mice, and PAI-1 rises in ageing muscle where it impedes repair. The basis is animal work only; no human trial has measured function.

Preserved Bone Mass

A PAI-1 inhibitor stimulated bone formation and prevented bone loss in oestrogen-deficient mice (Jin et al., 2018). The basis is animal work only; no human bone-density endpoint has been measured.

Weight Loss and Reversal of Fatty Liver

In diet-induced obese mice, TM5441 triggered fat-cell breakdown and cleared liver fat faster than diet change alone. The basis is rodent work only; no human study has reported body weight or liver fat.

Preserved Cognitive Function

PAI-1 blocks the same enzymes that clear amyloid protein deposits from the brain, and rises in Alzheimer’s disease. The basis is mechanistic and observational only; no controlled study has measured cognition on an inhibitor.

Clot Prevention without Prolonged Bleeding

These compounds act on the fibrinolysis brake rather than on platelets, so they may prevent clots without bleeding. In monkeys the effect matched clopidogrel without lengthening bleeding time; no human thrombosis trial has begun.

Benefit-Modifying Factors

  • SERPINE1 4G/5G promoter variant (rs1799889): The 4G allele raises baseline PAI-1 production. People carrying two 4G copies start from a higher level and, in principle, have more to gain from inhibition, but no trial has stratified results by genotype.

  • Baseline PAI-1 level: Circulating PAI-1 rises with visceral fat, insulin resistance, fatty liver and inflammation. Someone already at the low end of the range has little headroom, and the developers themselves call for biomarker-guided patient selection in future trials.

  • Sex-based differences: Men and postmenopausal women generally run higher PAI-1 than premenopausal women, and oral oestrogen lowers it. Completed trials enrolled 58% men in leukaemia and 72% in pneumonia, so any sex-specific efficacy difference is undocumented.

  • Pre-existing health conditions: Benefit signals came from disease states with very high PAI-1 — leukaemia, checkpoint-resistant melanoma, viral pneumonia. Whether a metabolically healthy adult with a normal level derives any benefit has never been tested.

  • Age-related considerations: PAI-1 climbs steadily across the adult lifespan, so older adults have the largest theoretical margin for reduction. Trial populations had a median age near 56–59 years, and the only study in relatively healthy adults reached 75.

Potential Risks & Side Effects

High 🟥 🟥 🟥

No adverse effect reaches High: no harm has been documented across more than one controlled human trial, because only one placebo-controlled trial of a PAI-1 inhibitor has been completed and it enrolled 75 people.

Medium 🟥 🟥

Liver dysfunction was the main adverse event judged possibly attributable to TM5614 in the melanoma trial, appearing as raised liver enzymes. The mechanism is not established, and the signal is confounded because every affected patient was also receiving an anti-PD-1 antibody, which causes immune-mediated hepatitis in its own right. The evidence basis is one open-label single-arm phase II trial; the placebo-controlled pneumonia trial reported adverse-event rates similar to placebo, and the leukaemia trial recorded one non-severe case (Fujimura et al., 2024).

Magnitude: Hepatic dysfunction possibly related to the study drug occurred in 2 of 34 patients (5.9%); treatment-related events of grade 3 or higher (severe or worse) occurred in 3 of 39 treated patients (7.7%).

Low 🟥

Bleeding from Unopposed Clot Breakdown ⚠️ Conflicted

Removing the fibrinolysis brake should promote bleeding. People born without any PAI-1 have a lifelong bleeding disorder (Heiman et al., 2014); yet the drugs spare platelets and clotting factors and showed no bleeding signal in trials excluding anticoagulant (blood-thinner) users. The net reading is a plausible risk not yet triggered.

Magnitude: No drug-related bleeding or coagulation abnormality occurred in the 33-patient leukaemia trial (Takahashi et al., 2023); one early onset of menstrual bleeding was reported in a patient also receiving low-molecular-weight heparin (an injected blood thinner) (Hirai et al., 2024).

Speculative 🟨

Impaired Wound Healing and Post-Surgical Repair

PAI-1 stabilises the early fibrin scaffold that wounds rebuild on, so sustained inhibition could slow healing after surgery. The developers list wound healing as a required monitoring item, but the basis is mechanistic only.

Context-Dependent Tumour Promotion

PAI-1 restrains cell migration as well as promoting it, and its cancer role shifts with tumour type. Blocking it could favour invasion in some settings; the basis is conflicting laboratory data only.

Disturbed Phosphate and Mineral Regulation

PAI-1 controls clearance of the bone hormone FGF23, and inhibition sharply lowers it in mice. Whether that helps or destabilises phosphate handling in people is untested; the basis is animal work only.

Unmasking of Latent Autoimmunity

TM5614 lowers immune-checkpoint molecules, which is how it is meant to help in cancer. The same action could release restraint on self-reactive immunity during long-term use; the basis is mechanistic reasoning only.

Risk-Modifying Factors

  • SERPINE1 genotype: Carrying one copy of a loss-of-function SERPINE1 variant already halves PAI-1, so adding a drug could push activity toward the bleeding range seen in people with two copies. No trial has genotyped participants.

  • Baseline biomarker levels: Low platelets, low haematocrit, prolonged clotting times or already-raised liver enzymes flag people most exposed to bleeding and hepatic harm. Trials excluded platelets below 50 × 10⁹/L and haematocrit below 30%.

  • Sex-based differences: Menstruating women are the group in whom fibrinolytic excess shows earliest; heavy menstrual bleeding is the commonest presentation of complete PAI-1 deficiency, and the single bleeding report in trials was menstrual.

  • Pre-existing health conditions: Active bleeding, prior bleeding into the brain, recent surgery, stomach ulcers, liver impairment and untreated hypertension all raise the consequence of tipping fibrinolysis. Cancer itself raises both clotting and bleeding risk.

  • Age-related considerations: Older adults carry more cerebral small-vessel disease (damage to the brain’s tiny vessels), use more antiplatelet (clot-preventing) drugs, and clear medications more slowly. No PAI-1 inhibitor has been dosed in a trial population selected for advanced age.

Key Interactions & Contraindications

  • Anticoagulants, or blood thinners (warfarin, apixaban, rivaroxaban, edoxaban, heparins): Additive bleeding risk by a separate route — the anticoagulant blocks clot formation while the inhibitor accelerates clot breakdown. Caution; every trial excluded full-dose anticoagulation, so combination safety is undocumented.

  • Antiplatelet medications, which stop platelets clumping (aspirin, clopidogrel, ticagrelor, prasugrel): Caution. Platelet inhibition plus enhanced clot breakdown compounds bleeding risk, in the gut and brain especially. Bruising is monitored; the mitigation is avoiding elective combination outside a trial.

  • Thrombolytic (clot-busting) agents (alteplase, tenecteplase): Absolute contraindication. Both act on the same axis, and PAI-1 inhibition directly amplifies administered tissue plasminogen activator, risking major bleeding. Any contraindication to a thrombolytic applies here too.

  • Over-the-counter analgesics (aspirin, ibuprofen, naproxen, diclofenac): Caution. These non-steroidal anti-inflammatory medications impair platelet function and erode the stomach lining, adding a bleeding source to enhanced clot breakdown. Paracetamol is the mitigating substitution for routine pain relief.

  • Immune checkpoint inhibitors, which release brakes on the immune system (nivolumab, pembrolizumab): Deliberate combination in the melanoma and lung-cancer trials, but overlapping liver toxicity. Liver enzymes are monitored at least every treatment cycle; both agents cause immune-mediated hepatitis.

  • Tyrosine kinase inhibitors (imatinib, dasatinib, nilotinib): Deliberate combination in leukaemia trials with no reported interaction. Several of these agents carry their own bleeding tendency, so platelet counts warrant monitoring at the existing treatment interval.

  • Medications that also lower PAI-1 — angiotensin-converting enzyme inhibitors (ramipril), angiotensin receptor blockers (losartan), statins (atorvastatin), fibrates (fenofibrate), metformin: These are blood-pressure, cholesterol and glucose drugs; the effect is additive PAI-1 suppression. Monitoring rather than avoidance is the stated mitigation.

  • Supplements with fibrinolytic or antiplatelet activity (nattokinase, lumbrokinase, serrapeptase, high-dose fish oil, high-dose vitamin E, garlic, ginkgo, curcumin, bromelain): Additive clot breakdown and platelet inhibition. Caution; timing separation does not help, so discontinuation before any procedure is the mitigation.

  • Other interventions: Elective surgery, dental extraction, colonoscopy with biopsy and cosmetic procedures all carry additive bleeding consequence. Caution; trial protocols and the developers’ own safety guidance call for wound-healing surveillance around any procedure.

Populations who should avoid PAI-1 Inhibitors:

  • Active clinically significant bleeding of any site
  • Known bleeding disorder, including complete PAI-1 deficiency and haemophilia
  • Platelet count below 50 × 10⁹/L or haematocrit below 30%
  • Any history of stroke, or a brain tumour, arteriovenous malformation (a tangle of abnormal blood vessels) or aneurysm (a ballooned artery wall)
  • Surgery or trauma involving the brain or spinal cord within 90 days
  • Severe uncontrolled hypertension (systolic blood pressure above 200 mmHg persisting beyond 12 hours)
  • Concurrent full-dose anticoagulation
  • Liver impairment of Child-Pugh Class B or C (moderate to severe), or liver enzymes above the upper limit of normal
  • Pregnancy and breastfeeding
  • Active venous thromboembolic disease (clots in the veins or lungs) under treatment

Risk Mitigation Strategies

  • Baseline clotting screen before the first dose: A complete blood count, prothrombin time with international normalised ratio, and activated partial thromboplastin time establish whether bleeding risk is already elevated, mitigating the bleeding harm that inhibition can unmask.

  • Liver enzyme surveillance on a fixed schedule: Alanine and aspartate aminotransferase plus bilirubin at baseline, every two weeks for eight weeks, then monthly. This detects the drug-related liver injury seen in 5.9% of the melanoma cohort before it becomes symptomatic.

  • Stepwise dose escalation rather than starting high: Trials began at 120–150 mg once daily and moved to 180 mg only after four to eight weeks of tolerance. Escalating slowly limits early hepatic and bleeding events at an unknown individual threshold.

  • Procedural washout: Trial protocols withhold the medication and all fibrinolytic supplements for at least seven days before elective surgery, dental extraction or biopsy, resuming only after wound closure. This mitigates surgical bleeding and impaired wound repair.

  • Avoidance of stacked fibrinolytic agents: Combination with anticoagulants, thrombolytics, nattokinase or serrapeptase is excluded in every protocol. Removing the additive load is the only reliable mitigation for bleeding, since no reversal agent for PAI-1 inhibition exists.

  • Menstrual bleeding tracking in women of reproductive age: Protocols log cycle timing and pad or cup counts from baseline. Heavy or early bleeding is the earliest clinical signal of excessive fibrinolysis and prompts dose reduction or discontinuation.

Therapeutic Protocol

  • No approved protocol exists: No PAI-1 inhibitor is licensed anywhere. Everything below describes the regimens used by investigators inside registered trials, not a treatment protocol available outside them.

  • Standard trial regimen: TM5614 at 120–150 mg orally once daily for the first four to eight weeks, escalated to 180 mg once daily thereafter. The 180 mg dose is given as six 30 mg tablets.

  • Short-course regimen: In acute inflammatory illness, 120 mg once daily for seven days followed by 180 mg once daily was used, or 180 mg once daily for up to seven days in the American protocol.

  • Competing approach — indirect PAI-1 lowering: Weight loss, aerobic training, metformin, blood-pressure drugs, statins and fibrates all lower circulating PAI-1; a meta-analysis of statin trials quantifies one. This route is available now; direct inhibition is not.

  • Who developed each approach: The TM series comes from Toshio Miyata’s group at Tohoku University with Renascience Inc.; the longevity framing comes from Douglas Vaughan’s group at Northwestern; MDI Therapeutics is developing the separate compound MDI-2517.

  • Best time of day: PAI-1 peaks in the early morning under circadian control, independently of the sleep-wake cycle (Scheer & Shea, 2014). Trials dosed in the evening, which positions drug exposure against that peak.

  • Half-life and dosing interval: No peer-reviewed human half-life has been published. Once-daily dosing across every trial implies an exposure profile compatible with a 24-hour interval, but this is inference, not measured drug-level data.

  • Single versus split dosing: All TM5614 trials used a single daily dose. The second-generation compound MDI-2517 is being tested both once and twice daily in its phase I programme, so the question is not settled for the class.

  • Genetic polymorphisms: No pharmacogenetic dosing rule exists. The SERPINE1 4G/5G variant sets baseline target abundance and would be the obvious candidate, but no trial has genotyped participants or adjusted dose accordingly.

  • Sex-based differences: No sex-specific dosing has been studied. Trial populations ran from 58% male in leukaemia to 72% in pneumonia, and premenopausal women start from lower baseline PAI-1, so identical dosing may produce deeper relative inhibition in them.

  • Age-related considerations: No dose adjustment for age has been established. Trials enrolled adults with a median age near 56–59 years and applied no geriatric modification; renal and hepatic clearance data that would guide one are unpublished.

  • Baseline biomarker levels: Response is expected to track baseline PAI-1 activity, and the developers explicitly call for biomarker-guided selection. No threshold has been validated, so this remains a stated intention rather than a protocol.

  • Pre-existing conditions: Regimens were identical across leukaemia, melanoma and pneumonia, so disease state did not alter dosing. Liver impairment and full anticoagulation were handled by exclusion rather than dose modification.

Discontinuation & Cycling

  • Intended duration: Not settled. Cancer trials dose for 8–52 weeks, the pneumonia trials for 7–14 days, and the lung-fibrosis trial for 48 weeks. Any longevity application would imply indefinite use that nobody has studied.

  • Withdrawal effects: None reported. The pharmacology gives no reason to expect a withdrawal syndrome; PAI-1 production is not suppressed at the gene level, so levels should simply return to baseline.

  • Rebound consideration: No rebound above baseline has been described, but no trial has measured PAI-1 activity after stopping. Whether an unmeasured rebound occurs is genuinely unknown rather than excluded.

  • Tapering protocol: No taper has been used or is thought necessary. Trials stopped the medication outright at the end of the treatment period without dose reduction and without reported consequence.

  • Cycling: Not studied. No tolerance or loss of effect has been reported over 52 weeks of continuous dosing in leukaemia, so no rationale for scheduled breaks currently exists.

  • Stopping rules used in trials: Discontinuation was triggered by disease progression or by adverse events, principally liver dysfunction. Those remain the only validated reasons to stop.

Sourcing and Quality

  • Not commercially available: No PAI-1 inhibitor is approved or sold as a medicine anywhere. Legitimate access exists only through enrolment in a registered trial, almost all of which are running at Japanese sites.

  • Developers and manufacturers: TM5614 is manufactured to good-manufacturing-practice standards for Renascience Inc. and Tohoku University; MDI-2517 is manufactured for MDI Therapeutics. Neither supplies material outside its own clinical programme.

  • Research-chemical vendors: Compounds sold online as “TM5614” or “tiplaxtinin” are research reagents intended for laboratory use. They carry no identity, purity or sterility guarantee and are not manufactured to pharmaceutical standards.

  • No compounding route: No compounding pharmacy can legitimately prepare these compounds, because there is no approved active ingredient monograph and no pharmaceutical-grade starting material in the supply chain.

  • What to look for if evaluating any supplied material: A certificate of analysis from an independent laboratory naming the assay method, plus identity confirmation by mass spectrometry. Absence of either means the contents are unverified.

Practical Considerations

  • Time to effect: Molecular response in leukaemia was assessed at 12 months, tumour response in melanoma at 8 weeks, and oxygen requirement in pneumonia within 3–5 days. Nothing gives a timeline for a longevity endpoint.

  • Common pitfall — treating the genetic finding as a drug result: The Amish lifespan observation describes lifelong halved production from conception. No trial has shown that starting an inhibitor in mid-life reproduces any part of that outcome.

  • Common pitfall — stacking fibrinolytic supplements: People drawn to this target often already take nattokinase, serrapeptase or high-dose fish oil. Combining them adds bleeding risk without any evidence of additive benefit.

  • Regulatory status: Investigational everywhere. TM5614 received orphan drug designation, a status easing review for rare diseases, from Japan’s Ministry of Health, Labour and Welfare in August 2024. There is no United States Food and Drug Administration approval and no off-label route.

  • Cost and accessibility: Access is the binding constraint rather than price. Most trials recruit only in Japan and only in specific disease populations, so healthy adults seeking this intervention have no legitimate route at all.

Interaction with Foundational Habits

  • Sleep: Indirect and bidirectional. PAI-1 follows a strong circadian rhythm peaking in the early morning independently of the sleep-wake cycle, and obstructive sleep apnoea (repeated breathing pauses during sleep) raises it substantially, per a review of PAI-1 in sleep apnoea. Treating apnoea lowers the target the medication aims at.

  • Nutrition: Indirect and potentiating. Visceral fat is a major source of PAI-1, so energy restriction and weight loss lower it directly; bariatric surgery produces a large reduction, per a meta-analysis of surgical cohorts. Practically, dietary fat loss reduces what the medication has to work against.

  • Exercise: Indirect and potentiating. Aerobic training lowers resting PAI-1 and raises tissue plasminogen activator, shifting the fibrinolytic balance in the same direction as the medication. There is no evidence that PAI-1 inhibition blunts training adaptation, but contact sports raise bruising risk.

  • Stress management: Indirect. Acute psychological stress and depressive illness both raise circulating PAI-1, and elevated levels are consistently reported in major depression. Stress reduction therefore lowers the target, though no study has tested whether it changes response to an inhibitor.

Monitoring Protocol & Defining Success

Because no PAI-1 inhibitor is approved, no validated monitoring schedule exists; what follows reflects the safety measurements built into the registered trials plus the functional-medicine practice of tracking the target itself. Before the first dose, trial protocols establish a clotting and liver baseline together with the metabolic markers that the intervention is proposed to influence, and record PAI-1 activity or antigen from a morning fasting draw so that later values are comparable. Ongoing monitoring in those protocols runs at one week, four weeks, eight weeks, then every three months while treatment continues, tightening to weekly if any bleeding symptom or liver enzyme rise appears. Success in this context is defined as a measurable fall in PAI-1 activity accompanied by stable liver enzymes, stable blood counts and no bleeding, rather than by any clinical endpoint that has been validated for this drug class.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
PAI-1 activity Below 10 IU/mL; ideally the lower half of the laboratory range Direct target engagement The draw is fasting and between 08:00 and 10:00 every time; levels peak in the early morning, so timing drift ruins comparability
PAI-1 antigen Below 20 ng/mL Total protein pool, active plus latent Paired with activity; antigen can stay flat while activity falls, which is the expected drug signature
Platelet count and haemoglobin Platelets 150–400 × 10⁹/L; haemoglobin mid-reference for sex Detects bleeding before it is visible Part of a complete blood count (CBC, a standard panel of red cell, white cell and platelet measurements); trials excluded platelets below 50 × 10⁹/L
Prothrombin time with INR, and activated partial thromboplastin time Within laboratory reference range Baseline clotting competence INR is the international normalised ratio, a standardised measure of how long blood takes to clot. Both tests are insensitive to fibrinolytic drugs and usually stay normal; they screen for a coexisting clotting defect
D-dimer Below 0.5 mg/L fibrinogen-equivalent units Rises when clot breakdown accelerates A rise on treatment is expected pharmacology, not necessarily harm; interpret alongside symptoms, not alone
Fibrinogen 200–300 mg/dL Falling values suggest excessive fibrinolysis Conventional reference range extends to 400 mg/dL. Also an acute-phase protein, so infection or injury confounds it
Alanine and aspartate aminotransferase, bilirubin Alanine aminotransferase below 25 U/L in men and 20 U/L in women Detects the one adverse event attributed to this drug class Conventional laboratory cut-offs run to about 40 U/L, so a value called normal can still sit well above the functional target. Fasting draw preferred; any value above twice the upper limit of normal is rechecked within one week
Fasting insulin and HOMA-IR Insulin below 5 µIU/mL; HOMA-IR below 1.5 The metabolic axis linked to low PAI-1 in carriers Conventional reference ranges run to roughly 25 µIU/mL for insulin and 2.5 for HOMA-IR. HOMA-IR is the homeostatic model assessment of insulin resistance, a calculation combining fasting glucose and insulin. Requires a 10–12 hour fast; best paired with fasting glucose from the same draw
High-sensitivity C-reactive protein Below 1.0 mg/L Inflammation is the main upstream driver of PAI-1 Conventional cardiovascular risk cut-off is 3.0 mg/L, three times the functional target. Repeated after any infection resolves; a single high value reflects the illness, not the trend
eGFR Above 90 mL/min/1.73 m² Clearance route is unpublished, so kidney function is tracked as a precaution Conventional practice treats anything above 60 mL/min/1.73 m² as normal. eGFR is the estimated glomerular filtration rate, a calculated measure of kidney filtering capacity. Trials in healthy volunteers excluded values below 90 mL/min/1.73 m²

Qualitative markers worth tracking alongside the laboratory panel:

  • Easy bruising, gum bleeding when brushing, or nosebleeds lasting beyond ten minutes
  • Menstrual volume and cycle timing in women of reproductive age
  • Time for small cuts and dental sites to stop bleeding and to close
  • Energy levels and exercise tolerance
  • Cognitive clarity and sleep quality, both of which shift with the inflammatory state that drives PAI-1

Emerging Research

  • Phase III trial in chronic myeloid leukaemia: A placebo-controlled study of 60 participants testing whether adding TM5614 to a targeted cancer medication produces and sustains deep molecular remission for two years. Registered in Japan’s clinical trial registry as jRCT2031220084, with no ClinicalTrials.gov identifier, per the developer’s trial overview.

  • Phase III trial in melanoma: A placebo-controlled study of 124 participants with overall survival as its primary endpoint, started February 2025 and due to report in 2029, per the published trial protocol. Registered as jRCT2021240049; no ClinicalTrials.gov identifier exists. This is the first survival endpoint for the class.

  • Placebo-controlled trial in systemic sclerosis lung disease: Fifty participants, 48 weeks of dosing, testing the antifibrotic claim in humans for the first time. Registered as jRCT2021230022 with no ClinicalTrials.gov identifier, per the developer’s trial overview. A negative result would undercut the whole antifibrotic rationale.

  • Suspended American trial in severe pneumonia: The only registered United States trial of TM5614, a phase 1/2 study of 180 mg daily for seven days (NCT04634799). Suspended after enrolling 9 participants, pending new drug supply and Japanese results.

  • First healthy-volunteer drug-level study of the class: A completed phase 1 single- and multiple-ascending-dose study of MDI-2517 in 32 healthy adults, measuring half-life, renal excretion, metabolites and target engagement (NCT07220889).

  • Second-generation compound in autoimmune fibrosis: MDI-2517 reduced disease severity in a preclinical systemic sclerosis model, establishing a competitor programme outside the Tohoku group (Su et al., 2026). Independent replication of the antifibrotic effect matters because the existing evidence has one dominant source.

  • Completed ageing trial in relatively healthy older adults: Four months of TM5614 in 20 adults aged 50–75 lowered estimated epigenetic age by 2–3 years, with no control arm and no peer-reviewed publication yet (per a science-agency news report). An international successor trial of 100 or more participants is planned but unregistered.

  • Evidence that could weaken the case — the target may be a marker: Circulating PAI-1 tracks visceral fat and inflammation so tightly that its associations with diabetes may be confounded, as Yarmolinsky et al., 2016 show in a meta-analysis of observational studies. If so, lowering it pharmacologically would change little.

  • Evidence that could weaken the case — the one placebo-controlled trial missed: The randomised pneumonia trial failed its primary endpoint (Hirai et al., 2024). Every positive human result to date comes from single-arm designs compared with historical controls, which systematically overstate effect.

  • The open question for this audience: Whether lowering PAI-1 changes an ageing measure in healthy older adults rests on that single uncontrolled study, a genetic cohort (Khan et al., 2017) and a mouse engineered with the same variant (Khoddam et al., 2025), not a controlled trial.

Conclusion

PAI-1 inhibitors are oral medications that block the body’s main brake on clot dissolution, a protein that also drives scarring and floods out of worn-out cells. The interest in them for longevity rests on a family that makes half the normal amount from birth and lives longer with better blood sugar control, and on mouse work in which blocking the protein delays several features of ageing.

What has been measured in people is narrower. Two small cancer studies without a comparison group showed real responses, one in a blood cancer and one in melanoma that had stopped responding to immune treatment. The one study with a placebo group, in viral pneumonia, pointed the right way on every measure but was too small to settle the question and closed short of its planned size. In relatively healthy older adults, one small study without a comparison group reported a modest drop in a laboratory estimate of biological age.

Safety so far looks mild: liver enzyme rises in a small minority, and no bleeding despite the obvious mechanism, though the trials excluded anyone on blood thinners. People born with none of the protein bleed for life, which sets the outer limit of what full blockade would do.

The evidence base is concentrated. Almost all of it comes from the university group and company that own the compound, or from their long-standing collaborators, and no professional body has taken a position either way. That concentration is itself part of the uncertainty.

Top - Benefits - Risks - Protocol