SS-31 Became an Approved Drug. The Trials It Failed Are the More Useful Reading.
In September 2025 the FDA granted accelerated approval to elamipretide, making it the first drug cleared in the United States that acts directly on mitochondria. The same molecule has been sold in gray-market vials for a decade under the name SS-31, usually alongside claims about energy, recovery and biological age. The approval is being read in those circles as vindication, and it is easy to see why. A compound that longevity forums have been buying from Chinese suppliers now has a brand name, a label and a package insert. What the approval does not do is transfer to the use most people are buying it for. It covers one ultra-rare genetic disease, in patients above 30 kilograms, on the basis of a muscle strength measurement in twelve people. Two much larger trials of the same drug in two other diseases both missed their primary endpoints. The one randomized trial in ordinary older adults produced a result that is more interesting than it is encouraging. This is educational background on a prescription drug and an unapproved research chemical sharing a molecule. Nothing here is medical advice or a protocol.
A Drug Approved for Roughly One Person in a Million
Forzinity, from Stealth BioTherapeutics, was approved on 19 September 2025 to improve muscle strength in adults and children with Barth syndrome weighing at least 30 kilograms. Barth syndrome is an X-linked disorder caused by mutations in the TAFAZZIN gene, which means it appears almost exclusively in boys and men. It affects cardiac muscle, skeletal muscle and neutrophil counts, and it does so through a defect in exactly one thing: the remodeling of cardiolipin, a lipid found only in the inner mitochondrial membrane. The path to that approval was not smooth. An FDA advisory committee voted 10 to 6 in favor in October 2024. The agency then issued a complete response letter in May 2025, the company resubmitted in August, and the approval followed a month later. It is an accelerated approval, granted on an intermediate endpoint rather than a demonstrated clinical benefit, and the FDA is requiring a post-approval randomized, double-blind, placebo-controlled trial to confirm that the strength measurement actually corresponds to patients living better.
The Approval Came From the Smallest Trial
Participants randomized across the main elamipretide programs. The two largest trials, in primary mitochondrial myopathy and in dry age-related macular degeneration, both missed their primary endpoints. The approval rests on TAZPOWER, a crossover study in twelve people with Barth syndrome. That is not a criticism of the approval, because Barth syndrome affects perhaps one in a million births and a larger trial does not exist to run. It is a statement about how much evidence there is.
Why a Peptide Would Aim at Cardiolipin
Elamipretide is four amino acids long and carries a net positive charge. The inner mitochondrial membrane carries a strongly negative surface charge, largely because of cardiolipin, and the peptide concentrates there without needing a receptor or a transporter. Cardiolipin is structurally unusual, with four fatty acid tails rather than the usual two, and it does mechanical work: it bends the membrane into cristae, and it holds the electron transport chain complexes together into the assemblies that make oxidative phosphorylation efficient. When cardiolipin is damaged or abnormally remodeled, cristae flatten, complexes come apart, electrons leak, and the mitochondrion produces more reactive oxygen species while producing less ATP. That damage accumulates with age and it is the specific defect in Barth syndrome. So the logic is clean, which is part of why the compound has attracted the attention it has. Stabilize the lipid, restore the architecture, recover the output. The trials are where that logic meets resistance.
The Large Trial Came Back Flat
MMPOWER-3 is the study that should have settled the question. It screened 296 people with genetically confirmed primary mitochondrial myopathy, randomized 218 of them one to one between October 2017 and December 2019, and ran 40 mg of daily subcutaneous elamipretide against placebo for 24 weeks. The two primary endpoints were distance walked in six minutes and a total fatigue score. On the walk test, the difference between arms was 3.2 meters in the wrong direction, with a confidence interval running from 18.7 meters below to 12.3 meters above and a p value of 0.69. Fatigue produced nothing either, at p equals 0.37. Secondary endpoints did not rescue it. This is a well-run trial in a population selected for having exactly the problem the drug was designed to fix, with a dose and duration the sponsor chose, and it found nothing. Any honest account of elamipretide has to start there rather than end there.
The Genotype Split Is Either a Clue or an Artifact
A post hoc analysis published afterward divided the MMPOWER-3 population by the kind of mutation causing their disease. Participants whose defect sat in nuclear DNA, 59 of them, walked 25.2 meters further on elamipretide against 0.3 meters on placebo, at p equals 0.03. Participants whose defect sat in mitochondrial DNA, 146 of them, gained 14.9 meters against 24.1 meters on placebo, which is to say placebo did better. A smaller group with replisome variants and chronic progressive external ophthalmoplegia, 32 people in total, showed a 37.3 meter gain against a loss of 8.0 meters, at p equals 0.0024. There is a plausible mechanism behind that split, since nuclear and mitochondrial DNA defects damage the machinery in different ways. There is also a plausible statistical explanation, which is that if you cut a null result into enough pieces, some of the pieces will look positive. Post hoc subgroups are hypotheses. They are how you design the next trial, not how you conclude the last one.
The Genotype Split Inside a Flat Result
Change in six-minute walk distance at week 24 in MMPOWER-3, broken out by the kind of genetic defect causing the disease. Across all 218 participants the difference between arms was 3.2 meters in favor of placebo. Split by genotype, participants with nuclear DNA variants gained 25.2 meters against 0.3 meters on placebo. Participants with mitochondrial DNA variants did slightly worse than placebo. These are post hoc subgroups, which means the split was found after the result was known.
- Placebo (m)
- Elamipretide (m)
How Twelve People Produced an Approval
TAZPOWER enrolled twelve patients with genetically confirmed Barth syndrome, all at least twelve years old and above 30 kilograms. Each received 40 mg daily for twelve weeks, then a four week washout, then twelve weeks of the opposite assignment. Both primary endpoints, the six-minute walk and a Barth-specific fatigue score, failed to separate from placebo. Knee extensor strength did separate, along with patient global impression and some cardiac measures, and knee extensor strength is what the label was eventually built on. What happened next is the part worth understanding. Ten patients continued into an open-label extension, eight of them reaching week 168, and over that period the walk test improved by a cumulative 96.1 meters at p equals 0.003, with fatigue scores below baseline at every timepoint and left ventricular volumes trending better. Nobody in that period was blinded and there was no placebo group. In a slowly progressive disease with a natural history of decline, a sustained gain like that is genuinely difficult to explain away, and the advisory committee clearly did not dismiss it. It also cannot carry the weight a controlled comparison carries, which is precisely why the FDA attached a confirmatory trial requirement.
Where the Barth Syndrome Improvement Shows Up
Cumulative change in six-minute walk distance during the TAZPOWER open-label extension, measured from the extension baseline. Ten patients entered and eight reached week 168. The gain is large and it is sustained. It is also uncontrolled, because everyone in this period knew they were receiving the drug and there was no placebo group left to compare against. During the randomized, blinded crossover that preceded it, neither of the two primary endpoints reached significance.
The Eye Trial Failed the Same Way, and Then Got Interesting
ReCLAIM-2 tested the same 40 mg daily dose for 48 weeks in 176 people with dry age-related macular degeneration and geographic atrophy, randomized roughly two to one. Both primary endpoints missed. Low-luminance visual acuity did not separate from placebo on mean change, and the atrophic lesion grew at a statistically indistinguishable rate. The secondary results are where people who follow this compound spend their attention. Attenuation of the ellipsoid zone, the mitochondria-dense photoreceptor layer, progressed 43 percent more slowly on elamipretide at a nominal p of 0.0034. Partial attenuation slowed by 47 percent. And 14.6 percent of treated participants gained at least ten letters of low-luminance acuity, against 2.1 percent on placebo. The word nominal is doing real work in those sentences: these p values were not corrected for the number of comparisons, and in a trial that missed both primaries, secondary findings are exploratory by definition. Notice the pattern across three programs, though. A mitochondrial structural measure moves, and the functional endpoint that patients would feel does not follow it.
The Eye Trial: A Secondary Endpoint Worth Noticing
Proportion of ReCLAIM-2 participants gaining at least ten letters of low-luminance visual acuity over 48 weeks. The trial missed both of its primary endpoints, which were mean change in low-luminance acuity and growth of the atrophic area. This responder analysis and the ellipsoid zone measurements were secondary, and the p values attached to them are nominal, meaning they were not adjusted for the number of comparisons made.
The Only Trial Run in People Who Were Just Old
For readers of this site the most relevant study is also the smallest of the controlled ones. A team at the University of Washington randomized 39 adults aged 60 to 85, screened and selected for having poorly functioning mitochondria, to a single two-hour intravenous infusion of elamipretide at 0.25 mg per kilogram per hour or to placebo. Mitochondrial capacity in the first dorsal interosseous muscle of the hand was measured non-invasively. The primary endpoint was maximum ATP production, which rose 27 percent on the drug against 12 percent on placebo, p equals 0.045 on relative change and 0.055 on absolute change. That is a real effect from one infusion, and it is the strongest human evidence that this molecule does something in a body that is merely aging rather than diseased. Read the rest of the result, though. Coupling efficiency at rest did not change, meaning the extra ATP came from pushing more electrons through rather than from wasting fewer. Fatigue resistance across repeated contractions did not change in the pre-specified analysis. And by day seven everything had returned to baseline, consistent with a blood half-life of about sixteen hours. The mitochondria worked better. The muscle did not do more.
One Infusion, in People Who Were Only Old
Percent change in maximum mitochondrial ATP production in the hand muscle of 39 adults aged 60 to 85, selected for poor baseline mitochondrial function, after a single two-hour intravenous infusion. This is the primary endpoint and it was met, narrowly. Two things did not move: mitochondrial coupling efficiency at rest, and fatigue resistance during repeated contractions. The effect was gone by day seven.
Mice Got Better Without Getting Younger
An eight-week study in C57BL/6J mice published in Aging Cell in 2025 gave 23 to 24 month old animals elamipretide and measured both function and molecular age. The functional results were good. Cardiac strain and diastolic function partially recovered, skeletal muscle fatigue resistance rose, and frailty accumulation slowed. Then the authors ran epigenetic clocks and transcriptomic age estimates on the tissue and found no statistically significant shift. Pathway analysis did show fatty acid metabolism and oxidative phosphorylation genes turned up and inflammatory genes turned down, so something changed. The clocks that longevity research treats as the readout of biological age simply did not register it. That finding deserves more attention than it has received, and not only regarding this drug. It means an intervention can make an old animal work measurably better without moving the marker that gets sold as the measure of aging, which is an argument about what those clocks are actually tracking. It also cuts the other way. A supplement that shifts a methylation clock has not thereby been shown to make anybody stronger.
What the Approval Does Not Transfer To
Here is the practical summary. Elamipretide has one approved use, in one ultra-rare genetic disease, granted on an intermediate endpoint with a confirmatory trial still owed. In the two larger diseases where it was tested against placebo it did not beat placebo on the endpoints that were specified in advance. In healthy older adults a single infusion raised mitochondrial ATP output for less than a week without changing what the muscle could do. The vials sold online as SS-31 are not Forzinity. They come from suppliers with no obligation to demonstrate identity, purity, sterility or content, they are labeled for research use in a way that exists to sidestep drug regulation, and the peptide's own trial history offers no established subcutaneous protocol for the use people buy it for. There is a version of this compound's story where the nuclear DNA subgroup signal replicates, the confirmatory trial reads out well, and mitochondrial-targeted peptides become a real category. That version requires trials that have not been run yet. Anyone reasoning from a rare-disease approval to their own energy levels is skipping every step that would make the reasoning work.
Frequently Asked Questions
Have more questions? Contact us or explore our full peptide intelligence platform.
Related Articles
Does Telomere Length Determine Lifespan? Short Telomeres Track With Earlier Death. The Genetics Say Longer Ones Raise Cancer Risk.
Does telomere length determine lifespan? Short telomeres track with earlier death, yet genetic studies show longer ones raise the risk of several cancers.
Low-Dose Aspirin Added 8 Percent to Male Mouse Lifespan. In 19,114 Healthy Older Adults, It Added No Healthy Years and Deaths Ran Higher.
Does low-dose aspirin help longevity? It extended male mouse lifespan 8%, but the ASPREE trial in 19,114 healthy older adults found no gain and more bleeding.
Is NAD+ a Peptide? No. It Is a Coenzyme Built From Two Nucleotides, and the Peptide Label Comes From the Clinic, Not the Chemistry
Is NAD+ a peptide? No. NAD+ is a coenzyme built from two nucleotides. What NAD+ is used for, what IV NAD+ studies found, and how it differs from SS-31.