Therapy progress

Every current and emerging approach — by mechanism, by what it does to the cardiolipin defect, by how it's delivered, and by how close it is to real use.

There is now an approved drug. On September 19, 2025 the FDA granted accelerated approval to elamipretide (brand name Forzinity, Stealth BioTherapeutics) — the first treatment ever approved for Barth syndrome, and the first approved mitochondria-targeted drug. It is approved for patients weighing at least 30 kg, with important caveats spelled out in its card below.

And the gene therapy has moved further than the literature shows. Unpublished

At the 2026 Barth Syndrome Foundation conference (July 19–26, Bonita Springs, FL), Bill Pu's lab presented an optimized vector called TAZ-02 — a rebuilt version of the 2020 proof-of-concept therapy with a myotropic capsid, a codon-optimized TAZΔ5 transgene, a muscle-specific promoter, and a 6.7-fold lower dose. Stated next steps: a pre-IND meeting with the FDA and large-animal toxicity testing. None of it is published — it exists as conference slides.

Every therapy here is trying to solve the same core problem: broken tafazzin (TAZ) means cardiolipin can't be remodeled, so mature cardiolipin falls, the toxic intermediate MLCL builds up, the mitochondrial energy machinery destabilizes, and ATP-hungry tissues (heart, skeletal muscle) suffer. What separates the approaches is where in that chain they intervene — which splits them into two families — plus a separate question that turns out to matter just as much: how you physically get the fix into the right cells.

● Compensatory strategies

Leave the broken cardiolipin as-is, but help the surviving mitochondria work better — prop up the machinery, or build more of it, to restore ATP. Treats the consequence, not the cause. Faster to develop; symptomatic. This is where today's only approved drug sits.

● Root-cause strategies

Fix the lipid or the gene itself — restore functional cardiolipin (or working tafazzin) so the machinery is built correctly. Treats the cause. More curative in concept; earlier and harder to develop.

The eight approaches side by side

Therapy	Family	How it works	Route	Cardiolipin	MLCL	ATP
Standard & supportive care	Care	Manages heart failure, neutropenia, nutrition	Varies	→	→	—
Elamipretide / Forzinity (SS-31) APPROVED	Compensatory	Binds cardiolipin; stabilizes cristae & supercomplexes	Injection (SC)	→	→	↑
MA-5 (Mitochonic Acid 5)	Compensatory	Binds mitofilin; boosts ATP-synthase assembly	Oral (pill)	→	→	↑
NR / NAD⁺ boosting	Compensatory	Raises NAD⁺ → stimulates mitophagy & biogenesis	Oral (supplement)	→	→	↑
ABHD18 inhibition	Root-cause	Blocks CL-degrading enzyme → preserves nascent CL, cuts MLCL	Oral / oligo (envisioned)	↑	↓ corrects	↑
Gene therapy (TAZ replacement)	Root-cause	Delivers a working TAZ gene → restores tafazzin	One-time infusion	✓ restores	✓ normalizes	↑
Gene correction / base editing	Root-cause	Rewrites the DNA typo in place → the native gene works	LNP or viral	✓ restores	✓ normalizes	↑
Enzyme replacement (recombinant TAZ protein)	Root-cause	Inject the tafazzin protein itself, tagged to enter cells	Repeat infusion	✓ in principle	✓ in principle	↑

↑ improves · → essentially unchanged · ✓ / ↓ fixes the lipid defect · — not a direct effect

The therapies in depth

Standard & supportive care

Care

The foundation of care today — not a cure, but what keeps children stable and well.

Who leads it: Not a lab but a clinical network. The main Barth-specific centers are Hilary Vernon (Johns Hopkins, Director of the Mitochondrial Medicine Center) and Barry Byrne (University of Florida). The Barth Syndrome Foundation's Scientific & Medical Advisory Board coordinates guidance across them.
Mechanism: Doesn't touch cardiolipin. Treats the downstream problems: heart-failure medications (ACE inhibitors, beta-blockers) for the cardiomyopathy, G-CSF and infection vigilance for neutropenia, nutritional and growth support, physical therapy, and — in the most severe cardiac cases — heart transplant.
On CL / MLCL: No change — it manages consequences, not the lipid.
On ATP: No direct effect; supports the tissues that are energy-starved.

Stage: Standard of care, in use everywhere

ConceptPreclinicalPhase 1Phase 2/3Approved

Everything else is layered on top of this. Even the newly approved drug — and any future disease-modifying therapy — sits alongside good cardiac and immune management, it doesn't replace it.

Elamipretide / Forzinity (SS-31)

Compensatory · APPROVED

The first-ever approved Barth drug — a cardiolipin-binding peptide.

Who leads it: Stealth BioTherapeutics (the sponsor). The peptide chemistry traces to Hazel Szeto and Peter Schiller — the “SS” in SS-31 is their initials. The pivotal TAZPOWER trial was designed and run by Hilary Vernon at Johns Hopkins.
Mechanism: A small peptide that concentrates in the inner mitochondrial membrane and physically binds the (malformed) cardiolipin already there, stabilizing membrane curvature and holding the respiratory-chain supercomplexes together. Better-organized machinery means more ATP.
On CL / MLCL: Unchanged. In tafazzin-knockdown hearts it improved respiration without shifting total cardiolipin or the MLCL:CL ratio. It makes the bad lipid work better; it doesn't fix it — a textbook compensatory drug.
On ATP: Improves — via supercomplex stabilization.
Route: Daily subcutaneous injection (40 mg). Most common side effect is injection-site reactions.
Approval: FDA accelerated approval, Sept 19, 2025 (brand Forzinity, Stealth BioTherapeutics), for patients ≥30 kg. Based on TAZPOWER (NCT03098797) plus its long open-label extension.

Stage: Approved (accelerated pathway) — available now for eligible patients

ConceptPreclinicalPhase 1Phase 2/3Approved

Three things sit together. Elamipretide exists and is prescribable, which is a first for this disease. But it cleared on accelerated approval against a surrogate endpoint — improvement in knee-extensor muscle strength, an intermediate measure rather than a hard clinical outcome — and the evidence path was uneven: the blinded 12-week portion of TAZPOWER did not cleanly meet its primary endpoint, and the benefit emerged in open-label follow-up. Confirmatory evidence is still owed. The label also excludes children under 30 kg; Stealth has said it intends to pursue expanded access for lighter children and generate more data.

MA-5 (Mitochonic Acid 5)

Compensatory

An oral mitochondria-homing molecule — same goal as elamipretide, different target and delivery.

Who leads it: Takaaki Abe and Takafumi Toyohara at Tohoku University Graduate School of Medicine, Japan; Yoshiyasu Tongu first-authored the 2025 Barth paper. Abe's group developed MA-5 originally.
Mechanism: Derived from a natural plant compound; homes to mitochondria, interacts with mitofilin (part of the cristae-shaping MICOS complex) and promotes ATP-synthase assembling into its efficient oligomers. Result: more ATP without burning more oxygen — which also lowers oxidative stress.
On CL / MLCL: Unchanged — like elamipretide, it's an efficiency booster, not a lipid fix.
On ATP: Improves. In Barth iPSC cells and fruit flies: better mitochondrial structure, higher ATP, less cell stress, and improved heart function plus climbing (muscle) in flies.
Route: Oral — a meaningful practical advantage over injected elamipretide.

Stage: Phase 1 safety done in healthy adults (Japan); Abe's group says it is preparing a Phase 2 trial

ConceptPreclinicalPhase 1Phase 2/3Approved

A real pill that has cleared an initial human-safety bar, and the Tohoku group has said publicly that it is preparing to start Phase 2 — which would make it the second Barth-specific drug to reach patients. The Barth evidence underneath it is still cells and flies, so a Phase 2 would be testing a hypothesis that has never been through a mammal.

NR / NAD⁺ boosting

Compensatory

Nicotinamide riboside — raise the cell's NAD⁺ to restart mitochondrial quality control.

Who leads it: No single lab owns “NR for Barth” — it's an inference from two lines of work. The mitophagy defect was established by Yuguang Shi's lab at the Barshop Institute, UT Health San Antonio (first author Jun Zhang). Riekelt Houtkooper (Amsterdam UMC) works on NAD⁺ and mitochondrial metabolism and holds a current BSF grant. NR itself is sold by supplement companies, not developed for Barth by anyone.
Mechanism: NR is a vitamin-B3 precursor to NAD⁺, a central metabolic coenzyme that falls in many failing and mitochondrial hearts. Raising NAD⁺ activates sirtuins → PGC-1α (the master switch for building new mitochondria) and revs up mitophagy (clearing damaged mitochondria). In Barth models, restoring this quality-control cycle improved cardiac function — though the Shi lab did that with rapamycin, not with NR.
On CL / MLCL: Unchanged. It works entirely downstream — it doesn't repair cardiolipin; it improves the turnover and replacement of the mitochondria that contain it.
On ATP: Improves — through more, healthier mitochondria (biogenesis) and better cleanup (mitophagy).
Route: Oral — sold as a nutritional supplement, with a good human safety record from trials in other conditions. Unusually accessible relative to experimental drugs.

Stage: Preclinical in Barth models; NR itself is trialed/available for other conditions

ConceptPreclinicalPhase 1Phase 2/3Approved

NR targets the mitophagy defect directly, and because it sells as a supplement the accessibility gap is unusually small. Against that: Barth-specific human evidence is lacking, “available supplement” is not “proven for Barth,” and using NR to get there is an inference from NAD⁺ biology rather than something demonstrated in a Barth mammal.

ABHD18 inhibition

Root-cause

The genetic-suppressor approach — fix the lipid by removing the enzyme that destroys it. No working TAZ required.

Who leads it: Jason Moffat (University of Toronto) with Vincent Blomen and Sebastian Nijman (Scenic Biotech, Netherlands), drawing on the Brummelkamp genetic-screening approach, plus Charles Boone (Toronto), Chad Myers (Minnesota) and Ian Scott (SickKids). Moffat currently holds a Barth Syndrome Foundation grant on TAZ genetic interactions.
Mechanism: ABHD18 is the enzyme that deacylates nascent cardiolipin into MLCL (and degrades it further). Block ABHD18 and you never generate the toxic MLCL in the first place — nascent cardiolipin is preserved instead. And nascent (unremodeled) CL is functional enough to support the ATP machinery, so you bypass the need for TAZ entirely.
On CL / MLCL: Corrects both: cardiolipin rises, MLCL falls, the diagnostic ratio moves toward normal, and respiratory-chain assembly is restored.
On ATP: Improves — at the source, by rebuilding the machinery correctly rather than propping it up. Rescued survival in Barth mice.
Route: Envisioned as an oral pill (ABHD18 is a serine hydrolase — an unusually druggable enzyme class), or as an siRNA/ASO that lowers the enzyme (this is where a muscle-homing conjugate would matter).

Stage: Preclinical proof-of-concept (cells + mice); target validated and patented; drug discovery funded, no molecule yet

ConceptPreclinicalPhase 1Phase 2/3Approved

Mechanistically the most elegant option in the pipeline — it corrects the actual lipid defect and rescued survival in Barth mice — but there is no drug molecule yet. What exists is a validated target, a patent, and funded discovery work in the Moffat and Blomen labs. Years, not months. The druggable enzyme class is the reason for optimism.

Gene therapy (TAZ replacement)

Root-cause

The most literal fix — give the body a working copy of the gene.

Who leads it: William (Bill) Pu, Boston Children's Hospital / Harvard — the TAZ-02 program. In parallel, Barry Byrne and Christina Pacak (University of Florida) ran the other major AAV-TAZ effort. Pu also sits on the BSF Scientific & Medical Advisory Board.
Mechanism: Deliver a functional TAZ gene (packaged in an AAV viral vector) so cells make real tafazzin again and remodel cardiolipin normally. The gene is tiny (under 1 kb) and fits an AAV easily — so, unlike Duchenne, Barth's challenge was never fitting the gene. It's delivery: reaching heart and skeletal muscle at once, and durability over a lifetime.
Evidence: Bill Pu's lab (Circulation Research 2020) showed an AAV9-TAZ vector prevented and reversed cardiomyopathy in TAZ-knockout mice. Two other groups — Byrne & Pacak at the University of Florida, and a separate Duke/Minnesota/WashU effort — did parallel AAV-TAZ work. The payload clearly works in mice; the 2020 vector's weakness was skeletal muscle and dose.
On CL / MLCL: Fully corrects — restores normal remodeling, so mature cardiolipin returns and MLCL normalizes. Confirmed in quadriceps for TAZ-02: cardiolipin back to wild-type levels, MLCL down, MLCL:CL ratio corrected.
On ATP: Improves at the source.
Route: Potentially a one-time infusion. Limits seen with the 2020 vector: needed roughly 70% of heart cells transduced for durable benefit; skeletal-muscle levels declined over time; no re-dosing once neutralizing antibodies form.

TAZ-02 — the rebuilt vector, component by component Conference data, unpublished

Component	Proof-of-concept (2020)	TAZ-02 (presented 2026)
Capsid	AAV9	A myotropic capsid designated C4
Transgene	TAZ, full length	TAZΔ5 — the predominant human isoform — codon-optimized (PCTZO)
Promoter	Desmin / CAG	Pr1, a strong striated-muscle-specific promoter
Genome	Single-stranded	Self-complementary (scITR), with an HBB2 intron — faster, stronger onset
Skeletal muscle	Low transduction, low efficacy	Efficient delivery and functional rescue
Dose	1–2 × 10¹³ vg/kg	3 × 10¹² vg/kg — a 6.7-fold reduction

TAZ-02 results: Tested in the cardiac-and-skeletal-muscle knockout (mKO) mouse Pu's lab published in 2025. Safety: maximum tolerated dose 3 × 10¹³ vg/kg against a minimum effective dose of 3 × 10¹² — a roughly ten-fold therapeutic window; histology across heart, quadriceps, liver, kidney, lung, spleen, brain, spinal cord and gut was normal-to-mild, and serum markers (troponin, CK, ALT, AST, BUN, creatinine, glucose) were mostly unchanged. Efficacy: normalized mitochondrial morphology in heart and quadriceps, and corrected the cardiolipin profile in skeletal muscle. Cardiolipin data credited to Seul Kee Byeon.
Next steps: As stated on the slide: a pre-IND meeting with the FDA, and large-animal toxicity testing. That is the standard runway into a first-in-human trial.

Presented at the 2026 Barth Syndrome Foundation International Scientific, Medical & Family Conference (July 19–26, 2026, Bonita Springs, FL). Conference slides only — not peer-reviewed, not indexed, and not citable as a paper.

Stage: Late preclinical — optimized vector characterized in mice, heading into pre-IND and large-animal toxicity testing. No human trial yet.

ConceptPreclinicalPhase 1Phase 2/3Approved

This is the furthest-along root-cause program, and it sits ahead of where the published record places it. The long-standing bottleneck — the capsid — has now been addressed in-house with C4 rather than by licensing an outside capsid. Two caveats apply: this is unpublished conference data, so it has not been through peer review and the dose and safety numbers could shift; and a pre-IND meeting is the start of the regulatory conversation, not the end of it — large-animal toxicity testing alone typically runs a year or more before a first patient is dosed.

Gene correction / base editing

Root-cause

Instead of adding a spare gene copy, rewrite the original typo so the native gene works.

Who leads it: Nobody, for Barth specifically — this is the one entry with no named program. The relevant precedent is the bespoke base editor built for an infant with CPS1 deficiency by Rebecca Ahrens-Nicklas (CHOP) and Kiran Musunuru (Penn), which BSF has flagged for families as the template. The underlying chemistry comes from David Liu's lab at the Broad Institute.
Mechanism: A base editor (a guided enzyme, delivered as mRNA plus a guide RNA, often in a lipid nanoparticle) changes a single mis-spelled DNA letter back to the correct one — repairing the patient's own TAZ gene in place. No foreign gene copy left behind.
On CL / MLCL: Corrects — the native gene is restored, so remodeling resumes.
On ATP: Improves at the source.
Route: LNP infusion (transient — the editor does its job and clears) or a viral vector. Mutation-specific: each family's exact TAZ variant needs its own guide.

Stage: Concept / early preclinical for Barth (the tools are advancing fast in other diseases)

ConceptPreclinicalPhase 1Phase 2/3Approved

Conceptually the cleanest cure — fix the original, leave nothing extra — and LNP delivery sidesteps the anti-viral-antibody problem. But reaching heart and skeletal muscle with an editor is still unsolved, and mutation-specificity makes it a per-family effort. This is the furthest-out of the root-cause options.

Enzyme replacement (recombinant TAZ protein)

Root-cause

Skip the gene entirely — manufacture the tafazzin protein and deliver it directly.

Who leads it: Michael T. Chin, Molecular Cardiology Research Institute, Tufts Medical Center — the lineage runs from Ana Dinca's 2017 UW thesis (Chin was at the University of Washington then) to the 2024 penetratin paper. Related peptide-engineering work: Nathan Alder (UConn) on mitochondria-targeted peptides and Steven Glynn (Stony Brook) on tafazzin's structure — both BSF-funded.
Mechanism: Make purified human tafazzin in a bioreactor and infuse it, decorated with peptide tags that carry it into cells. Reading the construct name (HIS6-EEP-hTAFAZZINv2-CPP) left to right: HIS6 is a purification handle; EEP is an endosomal-escape peptide (springs the cargo out of the uptake bubble before it's digested); hTAFAZZIN v2 is the enzymatically active human isoform lacking exon 5; CPP is a cell-penetrating peptide (e.g. penetratin/TAT) that crosses the cell membrane.
On CL / MLCL: Corrects in principle — if the enzyme reaches the inner mitochondrial membrane in working form, it remodels cardiolipin like the real thing.
On ATP: Improves in principle.
Route: Repeat infusions (like enzyme replacement for Fabry, Gaucher, Pompe). Redoseable and titratable — no permanent genetic change, no antibody wall against re-dosing.

Stage: Early academic / preclinical (Chin lab lineage; Dinca UW thesis 2017)

ConceptPreclinicalPhase 1Phase 2/3Approved

The redoseable, no-permanent-change profile is attractive, but the delivery problem is the hardest of any approach in the pipeline. Approved enzyme-replacement drugs target the lysosome, which has a natural cell-surface delivery highway; mitochondria have no such receptor, and tafazzin must clear four membranes to reach its post. A related twist: when the Chin lab moved to delivering the gene with a penetratin tag, the plain gene fixed the MLCL:CL ratio but didn't improve heart function — only the penetratin version did, for reasons still unknown.

The other axis: delivery

Across this whole pipeline, the payload is usually the easy part; getting it into heart and skeletal muscle is the hard part. The four below aren't therapies — they're the vehicles the therapies ride in, and a better vehicle can rescue a payload that already works.

Delivery vehicle	What it carries	Barth relevance	Stage
AAV capsid	A gene (e.g. TAZ)	Natural AAV9 works but under-doses muscle. Engineered muscle-tropic capsids aim far more vector at heart and skeletal muscle and detarget liver: POLARIS-101 / AAV-SLB101 (Solid Biosciences), MyoAAV (Sharif Tabebordbar & Amy Wagers, Broad/Harvard), AAVMYO (Dirk Grimm, Heidelberg). Pu's lab went its own way — TAZ-02 uses an in-house myotropic capsid called C4, not a licensed one.	C4 characterized in Barth mice; POLARIS-101 clinical for other targets
LNP (lipid nanoparticle)	mRNA / base editor + guide RNA	Transient, non-viral, re-doseable, no anti-viral immunity — the vehicle for base editing. Today LNPs go mainly to liver; heart and muscle targeting is the open problem.	Proven in vaccines; muscle targeting unsolved
Bicycle / TfR1 conjugate & AOCs	siRNA / ASO (e.g. anti-ABHD18)	A small peptide (Bicycle Therapeutics) or antibody (AOC — Dyne Therapeutics, Avidity Biosciences) binds transferrin receptor 1 on muscle to haul an oligonucleotide inside — finally making siRNA/ASO viable in heart and skeletal muscle. The natural route for an anti-ABHD18 knockdown or a splice-correcting ASO.	Preclinical for Barth; clinical in other programs
CPP + EEP peptides	A protein (e.g. tafazzin)	Cell-penetrating plus endosomal-escape peptides drag a protein across the cell membrane — the delivery half of enzyme replacement. Still can't solve the mitochondrial-targeting step. Chin lab, Tufts.	Early academic

Who's working on what

The labs and companies behind each approach, collapsed into one scannable list.

Therapy	People	Where
Supportive care	Hilary Vernon; Barry Byrne	Johns Hopkins (Mitochondrial Medicine Center); University of Florida
Elamipretide	Stealth BioTherapeutics (sponsor); Hilary Vernon (TAZPOWER); Hazel Szeto & Peter Schiller (original chemistry)	Company; Johns Hopkins
MA-5	Takaaki Abe; Takafumi Toyohara; Yoshiyasu Tongu	Tohoku University, Japan
NR / NAD⁺	Yuguang Shi & Jun Zhang (mitophagy defect); Riekelt Houtkooper (NAD⁺ metabolism)	Barshop Institute, UT Health San Antonio; Amsterdam UMC
ABHD18	Jason Moffat; Vincent Blomen; Sebastian Nijman; with Boone, Myers, Scott	University of Toronto; Scenic Biotech (NL); Minnesota; SickKids
Gene therapy	William (Bill) Pu — TAZ-02; Seul Kee Byeon (cardiolipin data); Barry Byrne & Christina Pacak (parallel program)	Boston Children's / Harvard; University of Florida
Base editing	Nobody, for Barth. Precedent: Rebecca Ahrens-Nicklas & Kiran Musunuru (CPS1); David Liu (base-editing chemistry)	CHOP; Penn; Broad Institute
Enzyme replacement	Michael T. Chin; Nathan Alder (mito-targeted peptides); Steven Glynn (tafazzin structure)	Tufts Medical Center; UConn; Stony Brook
Muscle-tropic capsids	Solid Biosciences (POLARIS-101); Sharif Tabebordbar & Amy Wagers (MyoAAV); Dirk Grimm (AAVMYO)	Company; Broad/Harvard; Heidelberg
Oligo delivery to muscle	Bicycle Therapeutics; Dyne Therapeutics; Avidity Biosciences	Companies
Underlying biology	Steven Claypool; Michael Schlame; Vishal Gohil; Colin Phoon; Mauro Corrado	Johns Hopkins; NYU; Texas A&M; NYU; Cologne

Where this leaves things

File the therapies by where they act. Compensatory approaches (elamipretide, MA-5, NR) leave the cardiolipin defect in place and raise ATP by different tricks — stabilizing the machinery or renewing the mitochondrial population. They're closer to use and lower-risk, but symptomatic, and this is the family that produced the first approved drug. Root-cause approaches (ABHD18 inhibition, gene therapy, base editing, enzyme replacement) actually correct the lipid or the gene, so cardiolipin and MLCL move back toward normal — more curative in concept, but earlier and harder, and each is gated by a delivery problem more than a payload problem. Gene therapy is the first of them to have visibly solved its delivery problem.

Whether a therapy moves the MLCL:CL ratio sorts it into a family — unchanged means compensatory, corrected means root-cause — but that test alone is not enough to judge one. The Chin-lab penetratin result showed why: plain gene delivery corrected the ratio and still did not improve heart function, so fixing the lipid is necessary but not always sufficient. Assessing a therapy takes two questions: does it move MLCL:CL, and does it move a functional endpoint — ejection fraction, strain, or muscle strength? The strongest candidates do both.

Where that lands today: supportive care is the foundation; elamipretide (Forzinity) is approved for patients ≥30 kg, on an accelerated pathway with an uneven evidence base behind it; MA-5 may be next to reach patients, on thin Barth-specific evidence; and among root-cause approaches gene therapy has pulled clearly ahead, with an optimized vector characterized in mice and pre-IND work stated as the next step. ABHD18 remains the most mechanistically elegant approach and still has no molecule. Nothing in the root-cause column is in a human yet.

This is an educational synthesis of the published science, not medical advice, and preclinical results in cells, flies, and mice do not reliably predict what helps a patient. Even the approved drug carries real nuance — an accelerated approval on a surrogate endpoint, a ≥30 kg weight limit, and daily injections. Several items here, including over-the-counter NR, are not proven for Barth. Any decision about care belongs with a metabolic specialist and cardiologist who know the specific mutation, weight, and clinical picture.

References

The capsid swap is real, but it is not in the literature. Pu's published gene therapy used AAV9 — the 2020 Circulation Research paper below — and the lab's most recent Barth publication (2025) is a new mouse model, not a therapy. The optimized TAZ-02 vector was presented at the 2026 BSF conference and has not been published, so a literature search alone will place the program where it stood in 2020. It should be cited as conference data, not as peer-reviewed work.

Approved therapy — elamipretide (Forzinity)

Long-term efficacy and safety of elamipretide in Barth syndrome: 168-week open-label extension of TAZPOWER — Genetics in Medicine 2024; the open-label data the approval leaned on
SS-31 ameliorates cardiac mitochondrial morphology and defective mitophagy in a murine model of Barth syndrome — Scientific Reports 2024 (mechanism)
FDA elamipretide updates and regulatory timeline — Barth Syndrome Foundation
FDA grants accelerated approval to elamipretide, first treatment for Barth syndrome — Pharmacy Times, Sept 2025 (approval details, ≥30 kg, surrogate endpoint)
FDA approves first mitochondrial disease therapy: Stealth's elamipretide for Barth syndrome — UMDF

Gene therapy — TAZ replacement

AAV Gene Therapy Prevents and Reverses Heart Failure in a Murine Knockout Model of Barth Syndrome — Wang S, … Pu WT (senior author); Circulation Research 2020;126(8):1024–1039. The definitive Pu paper — uses AAV9.
AAV-Mediated TAZ Gene Replacement Restores Mitochondrial and Cardioskeletal Function in Barth Syndrome — Suzuki-Hatano, … Byrne BJ, Pacak CA; Human Gene Therapy 2019 (University of Florida group)
Cell-Penetrating Peptide Enhances Tafazzin Gene Therapy in Mouse Model of Barth Syndrome — Raghav, … Chin MT; Int. J. Mol. Sci. 2024 (the “ratio fixed but function not” result)
Yazawa E, … Pu WT. A murine model of Barth syndrome with cardiac and skeletal muscle selective inactivation of tafazzin — Dis. Model. Mech. 2025 (a model, not a therapy)
TAZ-02 — no citation exists. Presented at the 2026 Barth Syndrome Foundation International Scientific, Medical & Family Conference, July 19–26 2026, Bonita Springs FL. Conference slides only; not peer-reviewed, not citable as a paper.
Directed evolution of a family of AAV capsid variants enabling potent muscle-directed gene delivery across species — Tabebordbar, … Wagers; Cell 2021 (MyoAAV). Companion: AAVMYO, Grimm lab, Science Advances.

Root-cause — ABHD18 inhibition

Genetic suppression features ABHD18 as a Barth syndrome therapeutic target — Nature 2025; corresponding authors Jason Moffat (University of Toronto), Vincent Blomen and Sebastian Nijman (Scenic Biotech)
Rebalancing cardiolipin biosynthesis to treat cardiomyopathy — Nature Reviews Drug Discovery highlight (plain-language context)
BSF research summary of the ABHD18 paper — Barth Syndrome Foundation

Compensatory — MA-5 and NR / NAD⁺

Mitochondria-Homing Drug Mitochonic Acid 5 Improves Barth Syndrome Myopathy in a Human iPSC Model and Drosophila Model — Tongu et al.; Abe & Toyohara labs, Tohoku University; FASEB J. 2025
First oral drug shows promise for Barth syndrome — Tohoku University press release; the source for “preparing to start Phase 2”
Restoration of mitophagy ameliorates cardiomyopathy in Barth syndrome — Jun Zhang et al., Yuguang Shi lab, Barshop Institute, UT Health San Antonio; Autophagy 2022. Establishes the mitophagy defect — the rescue used rapamycin, not NR.

Base editing — the precedent, not a Barth program

World's first patient treated with personalized CRISPR gene editing therapy — Rebecca Ahrens-Nicklas (CHOP) and Kiran Musunuru (Penn), CPS1 deficiency, 2025
BSF's summary of the CPS1 case for Barth families — why the foundation is watching it

Delivery platforms & enzyme replacement

A transferrin-receptor-binding bicyclic peptide enhances oligonucleotide delivery to heart and skeletal muscle — Nucleic Acids Research 2025 (Bicycle/TfR1 — the route for an anti-ABHD18 oligo)
Dinca A. Development of Enzyme Replacement Therapy in Mammalian Models of Barth Syndrome — PhD thesis, University of Washington 2017 (the recombinant tafazzin + CPP lineage)

Emerging

Allogenic mitochondria transfer improves cardiac function in iPSC-derived cardiomyocytes of a patient with Barth syndrome — Experimental & Molecular Medicine 2025

Synthesized from the August 2026 Barth therapy comparative review. Preclinical items are lab-stage, not available treatments. Spot something out of date? Suggest an update.