SS-31 is now an FDA-approved drug. Its entire clinical program was twelve patients.

Why this note exists

Most compounds in pepmg's index are unapproved everywhere. SS-31 changed category last year, and that is worth reporting accurately in both directions: the approval is genuine and it is narrow, and the evidence underneath it is smaller than the word "approved" will suggest to almost anyone who reads it on a product page.

In the index generated on August 12, 2026, 51 of 112 vendors carried an SS-31 listing — 86 listings in total, twenty-fifth by vendor coverage across the registry.[24] Ten of those 86 listings name elamipretide anywhere in the product title, and two name MTP-131; the other 74 sell it as SS-31 alone.[24] Where a size is stated, 10 mg is the most common vial (36 listings), then 50 mg (23).[24] The approved product's dosage is 40 mg subcutaneously once daily.[1] Those two numbers describe different things, and pepmg does not translate a label dose into a protocol for a research vial.

A note on what kind of evidence this is: every efficacy and safety figure in the next six sections is human data, carrying its design and participant count. There is one section of animal data and it is labeled as such. Where a result comes from an open-label or single-arm phase, this note says so in the same sentence as the number.

The approval

What FDA actually approved, in FDA's words

The indication is one sentence long, and the qualifiers are not decoration. Forzinity is indicated "to improve muscle strength in adult and pediatric patients with Barth syndrome weighing at least 30 kg."[1][2] Barth syndrome is an ultra-rare disorder of cardiolipin metabolism caused by a tafazzin mutation; FDA's opening remarks at the advisory committee described its manifestations as "fatigue, skeletal and cardiac myopathy, neutropenia, premature death," with treatment supportive and no therapies specifically approved for it at that time.[25]

Immediately under that sentence the label carries the conditional: "This indication is approved under accelerated approval based on an improvement in knee extensor muscle strength, an intermediate clinical endpoint. Continued approval for this indication may be contingent upon verification and description of clinical benefit in a confirmatory trial."[1] Accelerated approval is a pathway that lets FDA approve on a surrogate or intermediate endpoint reasonably likely to predict benefit, with the real question deferred.[4]

The path there was not smooth, and the approval letter records it plainly: the application is dated January 29, 2024, and the approval acknowledges an amendment dated August 15, 2025 "which constituted a complete response to our May 15, 2025, action letter."[2] In plain terms, FDA declined to approve it in May 2025, and approved a resubmission four months later.

Outside the United States it remains unapproved. Elamipretide holds an EMA orphan designation for Barth syndrome, granted May 20, 2021 — and EMA's own page states that orphan designation "does not mean the medicine is available or authorised for use."[22]

The pivotal trial

Twelve patients, and both primary endpoints missed

TAZPOWER was a randomized, double-blind, placebo-controlled crossover trial: 12 subjects randomized to 40 mg per day of elamipretide or placebo for 12 weeks, a 4-week washout, then 12 weeks on the opposite arm.[7][15] The primary endpoints were the six-minute walk test and a Barth syndrome symptom assessment. The publication's result sentence is four words long: "In part 1 neither primary endpoint was met."[7] The label says the same thing in its own register: "FORZINITY was not superior to placebo on these primary endpoints."[1]

Knee extensor strength — the endpoint the approval is built on — was a secondary outcome, and the label is unusually direct about how it behaved: "Increases in knee extensor muscle strength were not observed during the randomized trial but were observed during the extension period."[1] The numbers are in the label's Table 3, against a pre-dose baseline median of 124 newtons.[1]

Median change from the pre-dose baseline at the start of the randomized trial, as tabulated in the prescribing information.[1]

Read the three boxes left to right and the shape of the evidence is visible. Against a placebo, over twelve weeks, in twelve people, the drug and the placebo landed within nine newtons of each other with overlapping ranges. The large, sustained-looking numbers appear only after the blind comes off. The label's own pediatric section says the intermediate endpoint supporting use in children was "observed in an open-label extension study of FORZINITY that included seven pediatric patients aged 12 years and older."[1]

The extension is real data, and it is not a controlled comparison

This deserves stating fairly, because the extension results are not nothing. Ten patients continued into the open-label extension; eight reached week 168, and three of those received elamipretide for a total of 192 weeks.[1][8] Over that period the published report describes a cumulative 96.1 metres of improvement on the six-minute walk test at week 168 (P = .003), fatigue scores below baseline at every timepoint, and improving trends in three-dimensional left ventricular volumes.[8] Four years of daily injections in a progressive disease, with function moving the right way, is a serious observation.

It is also a single-arm observation. The extension's own primary endpoints were safety and tolerability, not efficacy.[8] There is no concurrent placebo group, no blinding, and eight patients remained — so the improvements cannot be separated from practice effects on a walking test, from the selection involved in being one of the people who chose to keep going, or from ordinary variation in a very small group. That is not a criticism of the investigators, who reported the design accurately. It is a description of what the design can answer.

The comparison group

Nineteen patients who were never in a trial

To give the extension something to be compared against, the sponsor built one retrospectively. The natural history comparison study was, in its authors' words, "a retrospective, non-interventional study" using a propensity score model, and the analysis "included 8 patients from the TAZPOWER OLE and 19 untreated NHCs."[9]

Its results were large. On the six-minute walk test the least-squares mean difference between groups was 79.7 m (P = 0.0004) at week 64 and 91.0 m (P = 0.0005) at week 76; on handheld dynamometry the differences were 40.8 newtons at 64 weeks (P = 0.0002) and 56.7 newtons at 76 weeks (P = 0.0005).[9] The authors' own conclusion uses the conditional throughout: the results "suggest that elamipretide may improve natural history" of the disease.[9]

Two structural facts belong beside those p-values. A propensity-matched comparison of 8 treated patients against 19 historical controls is not a randomized comparison, and it cannot control for the reasons a patient did or did not end up in a trial. And the study was not independent: two of its authors are identified as employees of Stealth BioTherapeutics, and the senior author declares research funding from the company.[9] The same disclosure pattern runs through the TAZPOWER papers.[7][8] Sponsor involvement does not make a result wrong; it is a fact a reader is entitled to have in front of them.

The vote

Ten to six, and both sides said why

On October 10, 2024 FDA's Cardiovascular and Renal Drugs Advisory Committee took up NDA 215244 and voted on a single question: "Based on available evidence, do you conclude that elamipretide is effective for the treatment of Barth syndrome?"[5][6] The result recorded in the final summary minutes was 10 yes, 6 no, 0 abstain.[5]

The minutes are worth quoting on both sides, because the disagreement was not about the data — everyone was looking at the same twelve patients. Reasons cited for the yes votes "included use of regulatory flexibility given the rare disease, anecdotal evidence from the open public hearing, and sizeable effect sizes that appear maintained over time," and other yes voters "noted that the decision was difficult, that there is significant uncertainty of the drug's effect, but it is not clear what can be done to resolve the uncertainty, and that it is not clear there is evidence from an adequate … controlled trial."[5]

Reasons cited for the no votes "included that there are too many uncertainties even with regulatory flexibility, that the randomized, placebo controlled trial (SPIBA-201 Part 1) did not show a treatment effect and Part 2 and SPIBA-001 are not adequate for establishing effectiveness, that one cannot reliably rely upon anecdotal data, and that more systematic evidence should and could be generated."[5]

That is a regulatory body deciding, explicitly and on the record, to accept uncertainty because the disease is rare, serious and untreated. It is a defensible decision about a fatal childhood condition. None of the reasoning transfers to a healthy adult buying a vial.

Evidence ledger

The bigger trials, and what they found

The Barth program is the small end of elamipretide's record. The compound has been through substantially larger, better-powered human trials in other diseases, and those are the ones a claim about mitochondrial function in general has to answer to.

Enrollment as recorded in each registry entry and publication.[16][20][17][15]

MMPOWER-3 is the most informative negative result in the file. It randomized 218 participants with genetically confirmed primary mitochondrial myopathy 1:1 to 40 mg/day subcutaneous elamipretide or placebo for 24 weeks. The difference in least-squares mean change on the six-minute walk test was −3.2 metres (95% CI −18.7 to 12.3; p = 0.69), and on the fatigue score −0.07 (95% CI −0.10 to 0.26; p = 0.37).[10] The authors' own classification is unambiguous: "This study provides Class I evidence that elamipretide does not improve the 6MWT or fatigue at 24 weeks compared with placebo in patients with primary mitochondrial myopathy."[10] The registry lists the trial as terminated.[16]

A later post hoc analysis of that trial reported that a subgroup with nuclear-DNA pathogenic variants improved on the walk test while the mitochondrial-DNA subgroup did not.[11] That is a hypothesis generated after the fact from a failed trial, and it is the hypothesis NuPOWER was built to test — a 48-week randomized phase 3 of 60 mg/day against placebo. NuPOWER enrolled 102 participants against a stated target of 130 and completed on December 4, 2024; as of a registry query on August 14, 2026 no results are posted and the entry was last updated in October 2025.[17] An unreported phase 3 is not a negative result. It is an absence, and it should be counted as one.

Across the whole registry, a ClinicalTrials.gov search on August 14, 2026 returned 21 registered studies of elamipretide — 20 interventional and one expanded-access protocol — spanning heart failure, acute coronary events, macular degeneration, Fuchs' dystrophy, Leber's hereditary optic neuropathy, Friedreich ataxia and mitochondrial myopathy.[21] This is a well-studied molecule. It has one approved indication.[3]

The aging claim rests on a single infusion

SS-31 is sold in the research market on a mitochondrial-decline story, so it matters what the human aging data actually is. There is one randomized trial, and it is a single-dose study. Thirty-nine healthy older adults aged 60 to 85 (46% female), enrolled specifically on the basis of poorly functioning mitochondria, received one two-hour infusion of elamipretide or placebo in a double-blind design.[14]

The result was a rise in mitochondrial energetic capacity relative to placebo immediately after the infusion (ΔATPmax P = 0.055; %ΔATPmax P = 0.045), with the authors reporting "no difference … on day 7 after treatment, which is consistent with the half-life of ELAM in human blood."[14] Resting mitochondrial coupling did not change significantly. And on the outcome closest to how a person would experience it: "Despite the increase in ATPmax there was no significant effect of treatment on fatigue resistance."[14] The work was funded by Stealth BioTherapeutics, which the authors disclose.[14]

That is a clean, interesting mechanistic finding — a measurable change in mitochondrial capacity in aging human muscle, reversing within a week, with no functional consequence detected. It is not a trial of taking SS-31 for months. As of August 14, 2026, the only registered study of elamipretide in healthy older adults is an open-label, single-arm phase 2a pilot at 30 participants over four weeks, sponsored by an academic investigator, whose primary outcome is safety and tolerability.[19] There is no randomized trial of chronic elamipretide in healthy aging to report on, because none has been registered.[21]

The safety record, including what has not been done

The label's safety database is the same twelve people: "In the FORZINITY clinical development program, 12 male patients aged 12 to 35 years with genetically-confirmed Barth syndrome received treatment with daily subcutaneous injections of 40 mg FORZINITY. Eleven of these 12 patients were Caucasian."[1]

What that database shows is dominated by the injection site. Any local administration reaction occurred in 12 of 12 on elamipretide against 8 of 12 on placebo; injection site erythema 100% vs 25%, pain 75% vs 42%, induration 67% vs 17%, pruritus 67% vs 17%, with bruising and urticaria each 25% vs none.[1] The label also records that absolute eosinophil counts rose frequently where dosing ran 30 days or more, peaking around 90 days (mean increase roughly 0.5 to 0.6 × 10³/µL) and returning to baseline after 6 to 12 months of continued dosing.[1] The larger trials in other diseases reported the drug as well tolerated, and MMPOWER-3's authors said so explicitly even as it missed its endpoints.[10]

The more important line is in section 13.1, and it is a sentence about animal studies that have not happened: "Carcinogenicity studies have not been conducted with elamipretide."[1] Genotoxicity assays were negative — a bacterial reverse mutation assay, a chromosomal aberration assay in Chinese hamster ovary cells, and an in vivo rat bone marrow micronucleus assay — and subcutaneous doses in rats up to 20 mg/kg/day, described as approximately 5 times the clinical exposure at the maximum recommended human dose, did not affect fertility or reproductive performance.[1] All of that is animal and in vitro data, and none of it is a carcinogenicity study.

FDA required those studies as postmarketing obligations at approval: a 26-week carcinogenicity study in TgRasH2 mice and a 2-year carcinogenicity study in rats, the latter with study completion listed as October 2029 and the final report due January 2030.[2] For a compound people inject daily for years, "the long-term animal carcinogenicity work is scheduled, not finished" is a material fact rather than a technicality.

The trial that decides this started six weeks ago

Accelerated approval carries a bill, and the approval letter itemizes it. Under section 506(c) of the FDCA and 21 CFR 314.510, Stealth is required to "conduct a randomized, double-blind, placebo-controlled trial in patients 5 years and older with Barth syndrome to verify and describe the clinical benefit of Forzinity predicted by improvements in knee extensor muscle strength assessed by handheld dynamometry."[2] The letter adds the standard consequence: if the trial "fails to verify clinical benefit or is not conducted with due diligence … we may withdraw this approval."[2]

The registry entry matches the letter. 4TAZPower is a phase 3b/4 randomized, quadruple-masked, placebo-controlled trial over 72 weeks with an estimated enrollment of 48, an actual start date of July 2, 2026 and an estimated primary completion of September 30, 2029; its stated objective is "to confirm the efficacy of elamipretide which is approved in the United States (FORZINITY™) under the accelerated approval based on an improvement in knee extensor muscle strength, an intermediate clinical endpoint."[18] The sponsor's timetable in the approval letter lists study completion 09/2029 and final report submission 03/2030.[2]

So the honest summary of where this stands is: the question of whether elamipretide produces clinical benefit in Barth syndrome is open, FDA has said as much in writing, and the answer is roughly three years away.

Three things this note is not saying

First, it is not saying the approval was wrong. Barth syndrome is fatal, ultra-rare, and had no approved treatment; a twelve-person trial is not a scandal in a disease with that few patients, and the committee majority said openly that they were exercising regulatory flexibility rather than pretending the evidence was strong.[5] Reasonable people voted both ways in the same room.

Second, it is not saying elamipretide does nothing. It has a plausible, well-characterized mechanism at cardiolipin, a measurable acute effect on mitochondrial capacity in aging human muscle, and years of tolerability data at 40 mg daily in the patients who stayed on it.[14][8]

Third, it is not saying the safety record is bad. Across the larger trials the adverse-event profile was mostly injection-site reactions.[1][10] But a twelve-person label safety database, in one sex, in one disease, with the carcinogenicity studies still pending, cannot characterize what happens to a healthy adult injecting it for years. That is a statement about the size and shape of the record, not a claim of known harm.

What it is saying is narrower and more useful: "FDA-approved" is now a true statement about SS-31, and it means an intermediate endpoint, in twelve patients, in a disease that almost nobody buying the peptide has — with the verification trial still recruiting.

Questions people are asking