SS-31 Was Approved on Eight Unblinded Patients, and the Version Sold Is a Different Salt
The trial programme is real, large and mostly failed. For once the dose people run sits below what was studied rather than above it.
What it is
SS-31 is one molecule with four names, and PubChem, the public chemistry database, holds one record covering all of them, CID 11764719, CAS 736992-21-5. The name is the Cornell laboratory code, after Hazel Szeto and Peter Schiller.
The mechanism story changed once. The 2004 founding paper from that laboratory calls this class antioxidants, and it is where the line about concentrating a thousandfold in the inner membrane of mitochondria comes from. (PMID 15178689) Ten years on Szeto restated it as binding to cardiolipin, a fat found almost nowhere except that membrane, where it holds the energy machinery in shape. (PMID 24117165) The vendor pages I read all trace back to that review.
What the trials found
Twenty-one records for elamipretide sit on the United States government's trial registry, 1,869 people across the twenty interventional ones. Go and read the results field on each. Most missed.
Mitochondrial myopathy first, an inherited fault in the cell's energy machinery. Karaa and colleagues, 2018: 36 adults, escalating infusions into a vein over five days, top dose group 64.5 metres further at day 5 against 20.4 on the dummy, just outside the usual threshold at p = 0.053. Two other analyses in the same paper did clear it: the gain grew with the dose, and a model adjusted for other things that could shift the answer gave 51.2 metres against 3.0. Neurology graded it Class I evidence that the drug improved walking, Class I being the journal's top grade for trial quality. (PMID 29500292) MMPOWER-2 moved under the skin, 40 mg a day for four weeks, thirty people each taking drug and dummy in turn: 19.8 metres further on the drug, not enough to clear the threshold. (PMID 32096613) The pivotal phase 3 then put 109 people on 40 mg a day under the skin for 24 weeks against 109 on placebo: 3.2 metres worse, fatigue unchanged, graded Class I in the other direction. (PMID 37268435)
Elsewhere the same, all of them missing their primary endpoints. 300 people given MTP-131 into a vein for one hour during angioplasty, the emergency procedure that reopens a blocked heart artery (PMID 26586786). 71 heart failure patients on 4 mg or 40 mg under the skin daily for 28 days (PMID 32068002). 176 with an eye disease on 40 mg under the skin daily for 48 weeks (PMID 39605874).
The only human study I found in people without a diagnosed disease is the infusion at the top of this entry, 39 adults aged 60 to 85 picked for poorly working mitochondria, sponsor funded. (PMID 34264994) The ageing case underneath it is a mouse case. Siegel and colleagues, 2013: mice of 5 and 27 months, one injection into the abdominal cavity at 3 mg per kilogram, and an hour later the age-related drops in energy production had gone back to young levels, with nothing in the young animals. (PMID 23692570) A route people do not use, and I found no human trial that reproduces it.
Then the trial that produced the approval. TAZPOWER, 2021: twelve male patients with Barth syndrome, 12 weeks each on 40 mg a day under the skin and on placebo, neither primary endpoint met. Ten continued into an open label extension where everybody knew they were on the drug; eight reached week 168. The strength numbers on the approved label come from those eight: a baseline median of 124 newtons on a handheld gauge the patient pushes against, rising by a median of 63 at week 168. In the randomised part the same patients moved 4 newtons, and the placebo arm minus 5. (PMID 33077895, PMID 38602181) A comparison group was built afterwards from the records of 19 untreated patients. (PMID 36056411) FDA, the United States drug regulator, granted accelerated approval on 19 September 2025, to improve muscle strength in Barth syndrome patients weighing at least 30 kg.
The one place the four names come apart
Everything above is one molecule. The product is not.
FDA's drug listing directory returned seven elamipretide records on 6 August 2026: one is Forzinity, the approved hydrochloride, and six are bulk powder listings, four of them the acetate, two labelled by their own filers as bulk ingredient for human prescription compounding. All six started marketing after the approval.
On 8 December 2025, eighty days after the approval, FDA wrote to Darmerica, LLC of Davie, Florida naming substances it had distributed to compounders that may not lawfully be used. Between buserelin acetate and larazotide acetate sits "elamipretide (SS-31) acetate". On my reading of the statute, approving the hydrochloride does not make the acetate a component of an approved drug, and I found no FDA document explaining the salt point.
The same letter contains something worse. In that listing, the active ingredient named in the filing is something called peptide B27PD, a pattern FDA says runs across at least 23 of the firm's listings.
How it compares
| The usual problem | Does it apply here |
|---|---|
| A forged registry record | I could not find one. Elamipretide and SS-31 are not among the eight fabricated registrations documented on 31 July 2026 |
| Independent replication | Thin. Nineteen of 21 registry records have one sponsor, and the mechanism review is by the inventor |
| Evidence in healthy people | One infusion study in 39 older adults picked for poor mitochondrial function, in which the functional measure did not move |
Where the 40 mg came from
Vendor and protocol pages converge on injection under the skin between 0.1 and 40 mg a day, daily to three times a week, with cycling one page admits is not scientifically established. Another states correctly that the only published human trial dose in Barth syndrome is 40 mg a day under the skin.
That page is right. Forty milligrams a day under the skin is the dose in five published trials, in mitochondrial myopathy, Barth syndrome and the eye disease, and in a phase 3 still running, and it is the dose on the approved label, which restricts it to Barth syndrome patients weighing at least 30 kg and halves it to 20 mg daily in adults with severe kidney impairment not on dialysis.
The gap opens downward instead. I could find no human trial reporting a clinical result at 2, 5 or 10 mg a day. What the label has is a line in its pharmacokinetics section, the part on how much drug reaches the blood: exposure rises in proportion from 2 to 80 mg by daily injection under the skin. So somebody has given 2 mg a day and published only the blood levels. I looked for a study in any species testing three times weekly dosing and could not find one.
What could go wrong
The safety file is full and unglamorous. In the 12 patient Barth syndrome trial, injection site redness occurred in every patient on the drug against 25 percent on the dummy, pain in 75 against 42, hardening of the tissue in 67 against 17, itching in 67 against 17. Two thirds of the dummy injections produced a local reaction too, which is the number that keeps the first one honest. Injection site reactions ran to 80 percent over four weeks in the thirty people of MMPOWER-2. In the 176 patient eye trial, over 48 weeks, unwanted effects of any kind ran to 86 percent on the drug against 71 percent on the dummy, most commonly injection site itching, pain, bruising and redness.
Two warnings sit on the label: serious allergic reactions needing emergency treatment, and benzyl alcohol in the vial, which causes a life-threatening condition in newborns and is why the product is not approved for them. The drug was also not studied in patients aged 65 and over, the population most of the grey market interest comes from.
What we do not know
Whether the strength gain survives a proper trial. The confirmatory study plans 48 patients over 72 weeks against placebo, started 2 July 2026, due November 2029.
Whether the acetate and the hydrochloride behave the same way in a person. Every human trial used material supplied by the sponsor and I found no published comparison, nor any independent analysis of a grey market vial: no purity, content, sterility or bacterial contamination result. Nor could I find anything describing what peptide B27PD is.
Why the effect keeps vanishing. Up immediately after one infusion, gone by day 7, and something that reversible keeps being tested at 24 to 48 weeks.
My take
The approval is the most interesting thing here and it recommends nothing. A permission slip for one rare disease, resting on strength readings in eight men who all knew what they were taking. Closer to eight people telling you the same road is quicker than to a trial result. Whether that is a scandal or exactly what accelerated approval is for is a judgement.
The salt bothers me more than the eight patients do. It sits upstream of the vendor pages arguing about 5 mg against 20 mg.
If you have bought SS-31, I want to know what the certificate of analysis says the salt is, and whether the vendor mentioned Forzinity or the approval anywhere in its copy.
Frequently asked
Is SS-31 approved by the FDA?
It was approved in the United States on 19 September 2025 under the brand name Forzinity, for one thing: to improve muscle strength in adults and children with Barth syndrome, a rare inherited heart and muscle disease, who weigh at least 30 kg. That is the whole approval. It is an accelerated approval, which means it stays provisional until a confirmatory trial reports, and that trial is not due to finish until November 2029.
Is SS-31 the same thing as elamipretide?
They are one molecule with four names, SS-31, elamipretide, MTP-131 and Bendavia, and the public chemistry database holds a single record covering all of them. The one place they come apart is the salt, meaning the partner the active molecule is paired with to make a stable powder: the approved medicine is elamipretide hydrochloride, while most bulk powder sold to compounding pharmacies is elamipretide acetate.
What dose of SS-31 was used in the trials?
Forty milligrams a day injected under the skin is the dose in five published human trials, in mitochondrial myopathy, Barth syndrome and an eye disease, and in a phase 3 still running, and it is the dose on the approved label. Every one of those trials enrolled people with a diagnosed disease rather than healthy adults. The label halves the dose to 20 mg once daily for adults with severe kidney impairment who are not on dialysis. The lower figures that circulate, 2 to 10 mg a day, appear in the label only as a range over which blood levels were measured, and I could find no human trial reporting a clinical result at those doses.
Does SS-31 do anything for healthy people?
The only study I found in healthy older adults gave a single two hour infusion into a vein to 39 people aged 60 to 85 who were selected for having poorly functioning mitochondria. A laboratory measure of energy production in a hand muscle rose immediately afterwards, the difference was gone by day 7, and how tired the muscle got did not change. The study was funded by the drug's sponsor.
Have all the SS-31 trials reported their results?
Not all of them. Two finished trials have posted nothing I could find. NuPOWER, a phase 3 that gave 102 people with mitochondrial disease 60 mg a day under the skin for 48 weeks, completed in December 2024. And a phase 2 that gave 20 mg into a vein daily for seven days to 308 patients hospitalised with heart failure, completed in November 2017. Neither has results on the United States registry, and I could find no publication for either.
What are the side effects of SS-31?
Injection site reactions are close to universal. In the 12 patient Barth syndrome trial at 40 mg a day under the skin, redness occurred in every patient on the drug against a quarter of those on the dummy injection, with pain in 75 percent, hardening of the tissue in 67 percent and itching in 67 percent. The approved label carries two warnings, serious allergic reactions and benzyl alcohol in the vial, which is why the product is not approved for newborns. The label also states the drug was not studied in patients aged 65 and over.
Is the SS-31 sold online the same as the approved drug?
The approved product is elamipretide hydrochloride. Four of the six bulk powder listings in FDA's directory are a different salt, elamipretide acetate, and FDA, the United States drug regulator, named elamipretide (SS-31) acetate in a warning letter of 8 December 2025 as ineligible for use in pharmacy compounding. That letter also records a listing for elamipretide acetate that named a different substance, peptide B27PD, as its active ingredient. I could find no published comparison of the two salts in people and no independent laboratory analysis of a grey market vial.
This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.
Get the next one in your inbox
Free. One email a week at most. Unsubscribe in one click.
Double opt-in · No spam · Unsubscribe anytime
More in Mitochondria and cellular energy
- AICAR: The Human Trials Were Run, and the Big One Was Stopped for FutilityRoughly 7,100 people were randomised and about 3,550 got the drug, all in heart surgery, and three separate infusion studies failed to move the enzyme it is sold for.
- BAM-15 Does What DNP Did, and the Margin Is Missing From the RecordMouse dosing stopped at 200 mg per kilogram because the compound would not dissolve any further, so what the record holds is a formulation limit rather than a safety ceiling.
- Humanin: The Lifespan Paper Ran 100 Mice and Found No Lifespan EffectThe word lifespan in that title is carried by worms, and most of the rest of the record belongs to a mutant version of the peptide rather than the one sold in vials.