Elamipretide took about twenty years to get from a university lab to a pharmacy shelf, and for most of that time it looked like it would not make it. Trials missed their targets. A regulator refused to even review the application. The company behind it nearly ran out of road more than once. Then, in September 2025, it was approved.
We sell the research-grade form of this molecule, so we have followed the story closely, partly out of professional interest and partly because customers ask us about it all the time. What follows is the history as we understand it, failures included. If you are planning research with elamipretide, knowing where it stumbled is at least as useful as knowing where it worked.
One thing up front. What we sell is a laboratory chemical for in-vitro research. It is not the approved medicine, and nothing here is advice about treatment.
Table of Contents
Table of Contents
Where elamipretide came from
The molecule was first made in the early 2000s by Hazel Szeto at Weill Cornell Medical College and Peter Schiller at the Clinical Research Institute of Montreal. They called it SS-31, after their initials and its place in a numbered series. We wrote about the rest of that series in our guide to the Szeto-Schiller peptide family.
The two were not looking for a mitochondrial drug. They were making analogues of opioid peptides. Along the way they found that certain small peptides with alternating aromatic and positively charged amino acids could cross cell membranes and collect in mitochondria. A 2004 paper in the Journal of Biological Chemistry described the effect and showed that these peptides protected cells and tissue from oxidative damage.
That paper is where the story really starts. It reported a compound that got into mitochondria without depending on the membrane's electrical charge, which meant it could reach sick mitochondria as well as healthy ones. Nobody had a good tool for that at the time.
From SS-31 to MTP-131 to Bendavia to elamipretide
In 2006 a company called Stealth Peptides was set up to develop elamipretide. It licensed the rights from Cornell and from the Montreal institute. The company later renamed itself Stealth BioTherapeutics.
The molecule picked up names as it went. Inside the company it was MTP-131. For early clinical work it carried the brand name Bendavia. When it was assigned an international non-proprietary name, that name was elamipretide. Ocular formulations were sometimes called Ocuvia.
If you search the literature you need all of these. A 2010 paper will say SS-31. A 2014 trial registration will say Bendavia or MTP-131. Anything after about 2016 will say elamipretide. They are one compound. We untangle the naming in more detail in our article on MTP-131.
The first big test: heart attacks
The early animal data were strongest in ischaemia-reperfusion injury. That is the damage done when blood supply to a tissue is cut off and then restored, as happens when a blocked coronary artery is reopened after a heart attack. Much of that damage comes from mitochondria, so it was a logical place to start.
The trial was called EMBRACE STEMI. Patients having a first heart attack were given elamipretide or placebo by infusion around the time their artery was reopened. The main measure was the size of the damaged area of heart muscle.
Results were published in 2016. Elamipretide was safe and well tolerated. It did not reduce infarct size. The primary endpoint was missed.
This was a blow, but not an unusual one. Cardioprotection is a graveyard for promising compounds. Dozens of things that shrink infarcts in animals have failed in people. Timing is brutally hard, patients differ, and a single short infusion may simply be too little. The trial did not show the mechanism was wrong. It showed that this way of testing it did not work.
Heart failure: a hint, then nothing
The company moved on to chronic heart failure, where failing mitochondria are part of the picture and treatment can be given over weeks, not minutes.
A small early study gave a single infusion to patients with heart failure and reduced ejection fraction. At the highest dose, heart volumes improved over a few hours. That was encouraging enough to run a longer study.
The follow-up gave elamipretide by daily injection under the skin for four weeks. It did not improve the primary measure of heart size compared with placebo.
Two cardiac programmes, two misses. By this point a less stubborn company might have stopped.
Primary mitochondrial myopathy and the MMPOWER trials
The next target made more obvious sense. Primary mitochondrial myopathies are genetic diseases in which the mitochondria themselves are faulty. Patients have muscle weakness, fatigue and poor exercise tolerance. There were no approved treatments.
Early studies, known as MMPOWER and MMPOWER-2, were small and short. They suggested patients could walk further in six minutes after a few days of treatment. On the strength of that, a Phase 3 trial called MMPOWER-3 enrolled more than two hundred patients for 24 weeks of daily injections.
The results came at the end of 2019. The trial missed both of its primary endpoints. Walking distance and fatigue scores were no better on drug than on placebo.
This was the lowest point. A Phase 3 failure in the disease that seemed the best fit. But there was a wrinkle. When the data were broken down by the genetic cause of each patient's disease, those with defects in nuclear genes, particularly genes involved in copying mitochondrial DNA, appeared to respond, while those with defects in mitochondrial DNA itself did not. Subgroup findings like this are notoriously unreliable. The company chose to test it properly in a new trial restricted to that group. That work was still going on at the time of writing.
Barth syndrome: the indication that fit
Barth syndrome is very rare. It affects boys and men almost exclusively, and only a few hundred are known worldwide. It is caused by mutations in a gene called TAZ, which codes for an enzyme that remodels cardiolipin. Without it, cardiolipin is abnormal, mitochondria work poorly, and patients develop heart muscle disease, skeletal muscle weakness and severe fatigue. Many die young.
By the mid 2010s, research had shown that the molecule binds to cardiolipin. A disease defined by broken cardiolipin was about as direct a match as could be imagined.
The trial, TAZPOWER, was tiny because the disease is tiny. Twelve patients. Each received elamipretide and placebo in turn for twelve weeks, in random order. At the end of that blinded phase there was no significant difference.
Then the patients carried on taking elamipretide in an open-label extension. Over the following months, and then years, their walking distance and muscle strength improved steadily, and measures of heart function improved too. The company also compared these patients with historical records of untreated patients, who do not get better on their own.
The long fight with the regulator
This is where it got difficult. The US Food and Drug Administration generally wants to see a benefit in the blinded, placebo-controlled part of a trial. TAZPOWER had not shown one. The improvement came later, in the unblinded part, where patients and doctors knew what was being given.
The company argued that twelve weeks had simply been too short for muscle to rebuild, and that a larger or longer placebo-controlled trial was not feasible in a disease with so few patients. The agency was not persuaded at first. In 2021 it declined to accept the application for review.
Patients and families pushed hard. The company went back with more follow-up data. An application was eventually accepted, and in late 2024 an advisory committee of outside experts voted narrowly that elamipretide was effective. Even then the agency did not approve it straight away. In 2025 it issued a letter declining approval in the form requested but pointing toward the accelerated approval route, which allows a drug to be approved on an intermediate measure likely to predict benefit, with confirmation required afterwards.
In September 2025, elamipretide received accelerated approval for Barth syndrome under the brand name Forzinity. It was the first treatment ever approved for the disease and the first approved drug designed to act on mitochondria. The FDA trial snapshot sets out the basis for the decision.
The eye programme
Running alongside all of this was work on dry age-related macular degeneration. The cells of the retina are among the most energy-hungry in the body, and mitochondrial decline is thought to contribute to their loss.
A Phase 2 trial called ReCLAIM-2 gave daily injections under the skin for 48 weeks to patients with an advanced form of the disease. It missed its two primary endpoints. However, a structural measure of photoreceptor health did favour elamipretide, and more treated patients gained vision in low light. Larger Phase 3 trials were started to follow that up.
The pattern will be familiar by now. A miss on the headline measure, a signal underneath it, and a decision to keep going.
What researchers can take from the elamipretide story
We think there are several practical lessons here for anyone working with elamipretide at the bench.
Time matters
Short exposures have repeatedly shown less than long ones. A few hours or a few days often produced little. Months produced more. If mitochondrial structure needs to be rebuilt and tissue then needs to remodel, that is not surprising. When you design a study, think hard about duration.
The model matters
The clearest results came where cardiolipin is plainly involved. The murkiest came in conditions with many contributing causes. If your model has a clear cardiolipin or inner membrane defect, you are on firmer ground.
Endpoints matter
Several trials missed a functional endpoint while showing a change in a structural or biochemical one. In lab work you have the freedom to measure both. Use it. Do not rely on a single readout.
Safety has been consistent
Across every trial, the main side effect was irritation at the injection site. That is clinical information and not directly relevant to in-vitro work, but it does tell you that the molecule is not broadly toxic to cells at the exposures studied.
There is a lot more detail on mechanism in our article on MTP131 and cardiolipin, and on study design in our piece on Szeto Schiller 31 research models.
Research-grade elamipretide is not Forzinity
We need to be clear about this, because approval has caused some confusion.
The approved product is a sterile pharmaceutical made under drug manufacturing regulations, prescribed by specialists and dispensed to named patients. It has gone through every check that implies.
What we supply is elamipretide as a research chemical. It is the same molecule, made by peptide synthesis, purified, tested by HPLC and mass spectrometry and freeze-dried into vials. It is sold to laboratories for in-vitro research. It is not sterile-filled for injection, it is not a medicine, and it is not for human or animal use. We will not discuss dosing, and we will not sell to anyone who tells us they intend personal use.
If you or someone in your family has Barth syndrome, the right route is a specialist clinic and the approved product, not a research supplier.
Ordering elamipretide for laboratory research
For labs, the practical details are these. We stock elamipretide in four purity grades and in vials from 5 mg to 500 mg.
- Standard Grade, 5 mg, 98 percent or better, for method development.
- High Purity Grade, 10 mg, for routine cell work.
- Ultra-Pure Research Grade, 10 mg, 99.5 percent or better, HPLC and MS verified.
- Bulk Research Grade, 100 mg, for higher volume users.
Our minimum order is 50 units, and you can mix grades and sizes. Go to the SS-31 peptide page, press Order next to the size you want, and fill in the quantity and delivery address. We reply by email with the total, usually the same day. Payment is by Apple Pay, Zelle, Chime, Cash App, USDT or Bitcoin, and every order ships with the Certificate of Analysis for its lot.
If you need help choosing, the SS31 buying guide covers grades and quantities, and you can always ask through the contact page.
The elamipretide timeline at a glance
It is easy to lose the thread across two decades, so here is the sequence in one place. Dates are approximate where the public record is.
- Early 2000s. Elamipretide is made and characterised in the Szeto and Schiller laboratories as SS-31.
- 2004. First major paper describing mitochondrial uptake and protection.
- 2006. A company is formed to develop the Szeto-Schiller compounds.
- Around 2010 to 2012. First human studies, under the code MTP-131 and the brand name Bendavia.
- 2013. Papers show binding to cardiolipin and rapid effects in aged mouse muscle.
- 2016. The heart attack trial reports. Safe, but the main endpoint is missed. The generic name elamipretide comes into general use.
- 2017 to 2020. Heart failure studies. An early signal, then a negative four-week trial.
- 2019. The Phase 3 trial in mitochondrial myopathy misses its endpoints.
- 2020 onward. Long-term follow-up in Barth syndrome shows steady improvement.
- 2021. The regulator declines to review the first application.
- 2022. The Phase 2 eye trial misses its primary endpoints but shows a structural signal.
- 2024. An advisory committee votes narrowly in favour.
- 2025. A further setback in the spring, then accelerated approval of elamipretide for Barth syndrome in September.
Seen like that, the pattern is plain. The science moved quickly in the first ten years. The clinic took the next ten to catch up, and only did so in one small corner.
The part patients played
No account of elamipretide is complete without the families. Barth syndrome is so rare that most doctors never see a case. The patient community is small, well organised and extremely determined.
When the first application was turned away, parents and patients did not accept it quietly. They gave testimony. They described boys who could not climb stairs before treatment and could afterwards. They pointed out, reasonably, that a regulator demanding a large placebo-controlled trial in a disease with a few hundred known patients was asking for something that could not be done.
That pressure did not change the data. It did change how much attention the data received, and it kept the application alive through several rounds. Whatever view one takes of the evidence, it is hard to imagine the approval happening without them.
There is a wider point in this for anyone working on rare conditions. The standard tools of drug evaluation were built for common diseases. Applying them to very small populations forces awkward choices, and the story of elamipretide is now one of the reference cases for how those choices get made.
What is still being tested
Approval for one indication is not the end of the research. Several questions remain open, and they are worth knowing about because they shape what laboratory work is most useful.
Does the benefit in Barth syndrome hold up in the confirmatory study that accelerated approval requires? That is the condition attached to the decision.
Does the subgroup finding in mitochondrial myopathy stand when tested on its own? A trial limited to patients with defects in nuclear genes for mitochondrial DNA maintenance was set up to find out.
Do the larger eye trials confirm the structural signal seen in Phase 2? Dry macular degeneration affects very large numbers of people, so the answer matters well beyond the mitochondrial field.
And underneath all of these: why does elamipretide appear to work in some settings and not in others that look similar? That is a mechanistic question, and it is one that bench researchers are better placed to answer than clinical trialists. Work on membrane structure, on which proteins the molecule sits beside, and on how different tissues handle cardiolipin all feed into it.
If you are choosing a project, that last question is where we would look. There is plenty of room, the tools exist, and elamipretide is easy to obtain.
A practical note for labs
None of the clinical history changes how research-grade elamipretide behaves at the bench. It is the same stable, water-soluble tetrapeptide it was in 2004. What the history does change is the questions worth asking of it, and the standard of evidence a reader will expect. Plan for longer exposures, more than one readout, and a lot-specific certificate on file.
Questions people ask about elamipretide
Is elamipretide the same thing as SS-31?
Yes. SS-31 was the original laboratory name. elamipretide is the official generic name assigned later.
Is elamipretide FDA approved?
A pharmaceutical form received accelerated approval in the United States in September 2025 for Barth syndrome. That approval covers that product for that condition. It does not cover research chemicals.
Why did so many elamipretide trials fail?
Different reasons in different trials. Short treatment periods, mixed patient populations and endpoints that may have been too blunt all played a part. Elamipretide also may simply not help in some conditions.
Where can I read the trial records?
Registered studies are listed on ClinicalTrials.gov, and published papers can be found by searching PubMed.
Do you sell elamipretide for personal use?
No. Research use only, sold to laboratories and institutions.
A last word
It would be easy to tell this as a story of triumph. It is closer to a story of persistence. Most of the trials failed on their main measure. One rare disease, with a direct link to the molecule's target, eventually came good after years of follow-up and a great deal of argument.
For researchers that is not a discouraging picture. It says the biology is real, that it is specific, and that there is a lot still to work out about where and how it applies. That is what research is for. If you are setting up work in this area and need material, the sizes and prices are on the SS-31 peptide page. A general overview of the molecule is also available on Wikipedia.