Loading
What changed between revision 1 and revision 3, compared over the article's markdown source with its frontmatter — so a change to status, evidence or reversibility shows up as the edit it is.
Older · revision 1
Import· initial corpus import, not an edit
Initial import of content/gene-therapy-for-aging.md — the filesystem corpus, unchanged. Not an edit.
10.8 kB
Revision 1
Revision 3
lines 8–18 → 8–187 unchanged lines not shown
categories: ["longevity", "genetics"]categories: ["longevity", "genetics"]tags: ["gene therapy", "aav", "telomerase", "klotho", "follistatin", "aging", "self-experimentation"]tags: ["gene therapy", "aav", "telomerase", "klotho", "follistatin", "aging", "self-experimentation"]summary: "The use of delivered genes to modify aging biology systemically, demonstrated in mice with telomerase and other constructs and untested in any controlled human trial."summary: "The use of delivered genes to modify aging biology systemically, demonstrated in mice with telomerase and other constructs and untested in any controlled human trial."updated: "2026-07-28"updated: "2026-10-29"humanEvidence: "No controlled trial has tested any of these constructs against aging in people; the human record is unregulated self-administration reported by press release, plus early-phase follistatin trials in muscular dystrophy."humanEvidence: "No controlled trial has tested any of these constructs against aging in people; the human record is self-administration and single-arm offshore studies without published results, plus early-phase trials in muscle and eye disease."access: "Available through no approved route; pay-to-participate ventures in permissive jurisdictions have offered telomerase and follistatin constructs, one at a reported price of about a million dollars per participant."access: "Not approved for any aging indication; offshore ventures have charged for unapproved telomerase and follistatin constructs at reported prices from about $25,000 to $1 million per dose."reversibility: "context"reversibility: "context"issues: ["The BioViva, Libella, and Honduras episodes are described without citations.", "The klotho monkey result and the OSK mouse lifespan claim are stated without sources."]issues: ["The klotho monkey result is stated without a source.", "Minicircle's announced 2026 commercial klotho offering is not covered pending a primary source."]------ ```infobox```infoboxlines 32–53 → 32–5313 unchanged lines not shown
**Gene therapy for aging** is the delivery of genetic material to a large fraction of an adult's**Gene therapy for aging** is the delivery of genetic material to a large fraction of an adult'scells with the aim of changing the trajectory of aging itself rather than treating a single inheritedcells with the aim of changing the trajectory of aging itself rather than treating a single inheriteddisease. It differs from ordinary [[somatic-gene-therapy]] in ambition rather than in method: thedisease. The methods are those of ordinary [[somatic-gene-therapy]]; the ambition is larger. Thesame viral vectors and nucleic-acid payloads, aimed at systemic targets like telomere maintenance,same viral vectors and nucleic-acid payloads are aimed at systemic targets such as telomeremuscle mass, or epigenetic state, and given to people who are not sick by any current definition. Asmaintenance, muscle mass, or epigenetic state, and given to people who are not sick by any currentof 2026 the animal results are real and the human record consists almost entirely of unregulateddefinition. As of 2026 the animal results are real, and the human record consists almost entirely ofself-experimentation.unregulated self-experimentation. ## The animal evidence## The animal evidence The founding result is telomerase. A 2012 study delivered the mouse TERT gene in an adeno-associatedThe founding result came from telomerase. A 2012 study delivered the mouse TERT gene in anviral vector to adult mice and reported extended median lifespan, with a larger effect in animalsadeno-associated viral vector to adult mice and reported extended median lifespan, with a largertreated at one year of age than at two, and without an observed increase in cancereffect in animals treated at one year of age than at two, and without an observed increase in cancerincidence.[^bernardes2012] That last clause is the striking part, because telomerase reactivation isincidence.[^bernardes2012] The lack of an observed cancer increase is the surprising part, becausea near-universal feature of tumours, and the finding has not been reproduced widely enough to settletelomerase reactivation is a near-universal feature of tumours, and the finding has not beenthe safety question. Its interpretation depends on the broader argument covered underreproduced widely enough to settle the safety question. How much it means depends on the broader[[telomeres-and-telomerase]], where telomere attrition is a weaker driver of human aging than ofargument covered under [[telomeres-and-telomerase]]: in several respects telomere attrition is a weaker driver ofmouse aging in several respects.human aging than of mouse aging. Two other targets recur. [[klotho|Klotho]], a protein whose loss in mice produces a syndrome resemblingTwo other targets recur. [[klotho|Klotho]], a protein whose loss in mice produces a syndrome resemblingaccelerated aging and whose overexpression extends mouse lifespan,[^kurosu2005] has been pursued bothaccelerated aging and whose overexpression extends mouse lifespan,[^kurosu2005] has been pursued bothlines 55–69 → 55–731 unchanged line not shown
after peripheral klotho administration, in small numbers and without long-term follow-up. Follistatinafter peripheral klotho administration, in small numbers and without long-term follow-up. Follistatinblocks myostatin signalling and reliably increases muscle mass when delivered by AAV to mice andblocks myostatin signalling and reliably increases muscle mass when delivered by AAV to mice andmonkeys; early-phase trials in muscular dystrophy have reported modest and inconsistent functionalmonkeys; early-phase trials in muscular dystrophy have reported modest and inconsistent functionalchange,[^mendell2015] and the relationship between added muscle mass and added strength is notchange,[^mendell2015] and added muscle mass does not translate simply into added strength, thestraightforward — the issue treated in [[myostatin-inhibition]].issue treated in [[myostatin-inhibition]]. The newest line delivers reprogramming factors rather than a single protein. Vectors carrying OSK, theThe newest line of work delivers reprogramming factors instead of a single protein. Vectors carryingthree-factor subset of the [[yamanaka-factors]] that omits the oncogene MYC, have been reported toOSK, the three-factor subset of the [[yamanaka-factors]] that omits the oncogene MYC, have beenextend remaining lifespan in mice treated in old age, in studies with small cohorts that have not beenreported to extend remaining lifespan in mice treated in old age: one study gave an inducibleindependently replicated. This work sits at the intersection of gene therapy andAAV-OSK system to male mice aged 124 weeks and reported a 109% increase in median remaining[[epigenetic-reprogramming]]; the dosing and duration questions are those oflifespan, a result that has not been independently replicated.[^macip2024] In 2026 an OSK construct[[partial-reprogramming]].entered a registered human trial aimed at a disease rather than at aging: Life Biosciences' ER-100,an AAV vector carrying doxycycline-controlled OSK injected into one eye, began an open-label phase 1safety study in glaucoma and ischaemic optic neuropathy.[^registry2026] This work sits at theintersection of gene therapy and [[epigenetic-reprogramming]]; the dosing and duration questions arethose of [[partial-reprogramming]]. > [!caution] Mice are not the hard case> [!caution] Mice are not the hard case> A mouse lives under thirty months, weighs thirty grams, and can receive a vector dose per kilogram> A mouse lives under thirty months, weighs thirty grams, and can receive a vector dose per kilogramlines 73–84 → 77–883 unchanged lines not shown
## Why aging is a difficult target## Why aging is a difficult target Gene therapy works best where a single defective gene in a reachable tissue causes a defined disease.Gene therapy works best where a single defective gene in a reachable tissue causes a defined disease.Aging has none of those properties.Aging meets none of those conditions. *No single target.* The processes catalogued as [[hallmarks-of-aging]] are numerous and partly*No single target.* The processes catalogued as [[hallmarks-of-aging]] are numerous and partlyindependent. A construct that addresses telomere length leaves the others untouched,independent. A construct that addresses telomere length leaves the others untouched, so proposalswhich is why proposals in this space tend toward multi-gene combinations, each adding immunologicalin this space tend toward multi-gene combinations, and each added gene brings immunological andand manufacturing complexity.manufacturing complexity. *Delivery.* [[aav-vectors|AAV]] reaches the liver efficiently and most other tissues poorly, carries*Delivery.* [[aav-vectors|AAV]] reaches the liver efficiently and most other tissues poorly, carriesa payload limit of about five kilobases, and provokes an immune response that prevents redosing witha payload limit of about five kilobases, and provokes an immune response that prevents redosing withlines 89–125 → 93–1424 unchanged lines not shown
*Permanence.* An AAV genome persists episomally in non-dividing cells for years and cannot be*Permanence.* An AAV genome persists episomally in non-dividing cells for years and cannot bewithdrawn. For a transgene that promotes growth, blocks a tumour suppressor, or reactivateswithdrawn. For a transgene that promotes growth, blocks a tumour suppressor, or reactivatestelomerase, the absence of an off switch is a serious safety problem. Regulatable systems andtelomerase, the lack of an off switch is a serious safety problem. Regulatable systems and transienttransient mRNA delivery are the responses, at the cost of requiring repeated administration through amRNA delivery answer it, but they require repeated administration, and immunity tends to close thatroute that immunity tends to close.route. *Risk-benefit.* High-dose systemic AAV has caused severe hepatotoxicity and deaths in trials for*Risk-benefit.* High-dose systemic AAV has caused severe hepatotoxicity and deaths in trials forserious childhood diseases. That risk is defensible when the alternative is a fatal disorder. Appliedserious childhood diseases. That risk is defensible when the alternative is a fatal disorder. It isto a healthy sixty-year-old seeking a probabilistic benefit decades hence, the same risk is not, andnot defensible for a healthy sixty-year-old seeking a probabilistic benefit decades later. Nono regulator has an approval pathway for the indication — the same structural obstacle described underregulator has an approval pathway for the indication either, the same structural obstacle described[[geroscience-hypothesis]] and in [[metformin|the TAME proposal]].under [[geroscience-hypothesis]] and in [[metformin|the TAME proposal]]. ## Unregulated human use## Unregulated human use Because no legitimate trial exists, the human record has been written by people outside the system.Because no controlled trial for aging exists, most human use has happened outside conventional oversight. In 2015 the chief executive of the company BioViva, Elizabeth Parrish, received AAV constructsIn September 2015 the chief executive of the company BioViva, Elizabeth Parrish, received two genecarrying telomerase and a myostatin inhibitor outside the United States and reported subsequenttherapies outside the United States, reportedly in Colombia: follistatin, a myostatin inhibitor,lengthening of leukocyte telomeres measured by a commercial assay. The result was never peerinjected into muscle, and an intravenous viral vector carrying telomerase.[^regalado2015] BioVivareviewed, the assay is noisy, and one person is not a study.later announced, by press release, lengthening of her leukocyte telomeres measured by a commercialassay. The result was not published in a journal, the assay is noisy, and a single person cannotconstitute a study. In 2019 Libella Gene Therapeutics announced a pay-to-participate telomerase gene therapy offered inIn 2019 Libella Gene Therapeutics offered an AAV telomerase gene therapy at a clinic in Zipaquirá,Colombia at around a million dollars per participant, a structure that inverts the ethics of clinicalColombia, at $1 million per dose.[^swain2019] The pay-to-participate structure inverts the ethics ofresearch by charging subjects for an unproven intervention. No results have been published.clinical research by charging subjects for an unproven intervention. Its registry entries, eachfor five participants, were last updated in 2019 and report no results.[^registry2026] Since 2023 a company has offered plasmid-based follistatin gene therapy in a special jurisdiction inMinicircle, a Delaware-registered startup, doses volunteers in Próspera, a special economic zone onHonduras, with participants including a prominent longevity entrepreneur. Plasmid delivery is lessthe Honduran island of Roatán with its own regulatory framework, whose legal standing Honduras haspersistent than AAV, which reduces some risks, and the venture has produced self-collected data rathercontested since its Congress voted in 2022 to repeal the law behind such zones. The company recruitedthan controlled evidence.volunteers in 2022 for a plasmid follistatin gene therapy, at one point asking them to buy an NFT totake part.[^clarke2023] That study enrolled 43 people between August 2022 and August 2023, wasregistered only in May 2024, after it had finished, and has posted no results.[^registry2026] InSeptember 2023 the longevity entrepreneur [[bryan-johnson|Bryan Johnson]] received the therapy, thenlisted at about $25,000 per treatment, on Roatán; Minicircle had published no trial data.[^fortune2023]Plasmid delivery is less persistent than AAV, which reduces some risks. In 2025 the company began twofurther single-arm studies, of a klotho plasmid and of klotho combined with follistatin, eachenrolling 14 people; as of October 2026 neither had posted results. A second company, UnlimitedBiotechnology, began an open-label phase 1/2a at a Roatán clinic in June 2026, combining an AAV9follistatin vector with a VEGF plasmid for age-related muscle decline.[^registry2026] These ventures share a pattern: jurisdiction shopping, absent control groups, surrogate endpointsThese ventures share a pattern: jurisdiction shopping, no control groups, surrogate endpoints chosenchosen after the fact, and outcomes announced by press release. They generate no usable knowledge andafter the fact, and outcomes announced by press release. The later ventures register onraise the regulatory temperature for the researchers doing the work properly.ClinicalTrials.gov, which makes their designs visible but adds no control group. The ventures generatelittle usable knowledge and raise regulatory scrutiny for researchers doing the work properly. ## What a legitimate path would look like## What a legitimate path would look like lines 128–136 → 145–1532 unchanged lines not shown
Alzheimer's disease have entered early-phase trials, and one-time editing of PCSK9 to lower lifelongAlzheimer's disease have entered early-phase trials, and one-time editing of PCSK9 to lower lifelongcholesterol, using [[base-editing]] delivered in lipid nanoparticles and first shown to work durablycholesterol, using [[base-editing]] delivered in lipid nanoparticles and first shown to work durablyin non-human primates,[^musunuru2021] is the closest existing test of a durable genetic interventionin non-human primates,[^musunuru2021] is the closest existing test of a durable genetic interventiongiven for prevention rather than rescue. Neither is marketed as an aging therapy, and both wouldgiven for prevention rather than rescue. Neither is marketed as an aging therapy, but both would setestablish precisely the precedents an aging therapy needs: acceptable risk in relatively healthy adults,the precedents an aging therapy needs: acceptable risk in relatively healthy adults, a durable effectdurable effect from a single administration, and a regulatory framework for prevention.from a single administration, and a regulatory framework for prevention. Progress also depends on measurement. Without an accepted [[aging-biomarkers|biomarker of aging]] aProgress also depends on measurement. Without an accepted [[aging-biomarkers|biomarker of aging]] atrial cannot read out in less than a decade, and neither an [[epigenetic-clock]] shift nor antrial cannot read out in less than a decade, and neither an [[epigenetic-clock]] shift nor anlines 138–152 → 155–1691 unchanged line not shown
## Outlook## Outlook The near-term question is not whether aging can be treated genetically but whether any of theThe near-term question is whether any of the candidate constructs are worth the delivery risk, acandidate constructs are worth the delivery risk. Telomerase carries an unquantified cancer liability;narrower question than whether aging can be treated genetically at all. Telomerase carries anfollistatin adds muscle of uncertain functional value; reprogramming factors carry dedifferentiationunquantified cancer liability; follistatin adds muscle of uncertain functional value; reprogrammingand teratoma risk that has been demonstrated in mice. Combination approaches multiply these problemsfactors carry dedifferentiation and teratoma risk that has been demonstrated in mice. Combining themrather than averaging them.multiplies these problems instead of averaging them out. If the field does mature, its main constraint may be economic rather than biological. One-time geneIf the field does mature, its main constraint may be economic rather than biological. One-time genetherapies for rare disease already carry prices in the millions, and a therapy intended for everyonetherapies for rare disease already carry prices in the millions, and a therapy intended for everyoneabove sixty would face a manufacturing and distribution problem of a different order — the concernabove sixty would face a manufacturing and distribution problem of a different order, the concernargued out under [[access-and-inequality]].argued out under [[access-and-inequality]]. ## See also## See alsolines 166–169 → 183–19213 unchanged lines not shown
[^kurosu2005]: `paper` Kurosu, H. et al. "Suppression of aging in mice by the hormone Klotho." *Science*, 2005. {A transgenic overexpression model rather than a delivered therapy: the mice carried the extra klotho from conception.}[^kurosu2005]: `paper` Kurosu, H. et al. "Suppression of aging in mice by the hormone Klotho." *Science*, 2005. {A transgenic overexpression model rather than a delivered therapy: the mice carried the extra klotho from conception.}[^mendell2015]: `paper` Mendell, J.R. et al. "A phase 1/2a follistatin gene therapy trial for Becker muscular dystrophy." *Molecular Therapy*, 2015. {An open-label early-phase trial in a small number of men with Becker muscular dystrophy, with no placebo group.}[^mendell2015]: `paper` Mendell, J.R. et al. "A phase 1/2a follistatin gene therapy trial for Becker muscular dystrophy." *Molecular Therapy*, 2015. {An open-label early-phase trial in a small number of men with Becker muscular dystrophy, with no placebo group.}[^musunuru2021]: `paper` Musunuru, K. et al. "In vivo CRISPR base editing of PCSK9 durably lowers cholesterol in primates." *Nature*, 2021.[^musunuru2021]: `paper` Musunuru, K. et al. "In vivo CRISPR base editing of PCSK9 durably lowers cholesterol in primates." *Nature*, 2021.[^macip2024]: `paper` Macip, C. C. et al. "Gene Therapy-Mediated Partial Reprogramming Extends Lifespan and Reverses Age-Related Changes in Aged Mice." *Cellular Reprogramming*, 2024. {Male mice only, treated from 124 weeks of age; the 109% figure is for median remaining lifespan, not total lifespan.}[^regalado2015]: `news` Regalado, A. "A Tale of Do-It-Yourself Gene Therapy." *MIT Technology Review*, 14 October 2015. https://www.technologyreview.com/2015/10/14/165802/a-tale-of-do-it-yourself-gene-therapy/ {The magazine could not confirm many details: Parrish would not name the country or the doctor, and Colombia comes from people involved with the company.}[^swain2019]: `news` Swain, F. "Buyer beware of this $1 million gene therapy for aging." *MIT Technology Review*, 6 December 2019. https://www.technologyreview.com/2019/12/06/131657/buyer-beware-of-this-1-million-gene-therapy-for-aging/[^clarke2023]: `news` Clarke, L. "This biohacking company is using a crypto city to test controversial gene therapies." *MIT Technology Review*, 13 February 2023. https://www.technologyreview.com/2023/02/13/1068330/minicircle-prospera-honduras-biohacking-follistatin-gene-therapy/[^fortune2023]: `news` Royle, O. R. "Bryan Johnson is being used as a human guinea pig in unproven gene therapy injections worth $25,000 a pop in an island off Honduras." *Fortune*, 21 December 2023. https://fortune.com/2023/12/21/bryan-johnson-human-guinea-pig-unproven-tests-reprogram-human-genes-follistatin-roatan-island-honduras {Reports that Minicircle did not charge Johnson; his claimed results come from his own posts and from Bloomberg, not from a study.}[^registry2026]: `trial` ClinicalTrials.gov records NCT04133649 (Libella, AAV-hTERT), NCT06411366 (Minicircle, follistatin plasmid), NCT07216781 and NCT07285629 (Minicircle, klotho and klotho-follistatin plasmids), NCT07443826 (Unlimited Biotechnology, AAV9-follistatin with VEGF plasmid), and NCT07290244 (Life Biosciences, ER-100), checked October 2026. {All are single-arm or non-randomised and unmasked; none had posted results at the time checked.} removed, struck through added, underlinedLine numbers count the serialised markdown of each revision, frontmatter included.
109 lines are the same in both revisions, and 44 of them are not shown. Neither revision here is necessarily the current article — read Gene therapy for aging for the text this wiki stands behind today.