SS-31 (elamipretide): the long road from failed trials to an approved drug

A drug for the power plant's walls

Most drugs act on receptors at the cell surface or on enzymes floating inside it. SS-31 travels much deeper. It slips through the cell membrane and concentrates about 1,000-fold in the innermost membrane of the mitochondria, the folded sheets where the cell makes most of its energy ^[1].

For two decades that unusual address made it a favourite of ageing researchers. In old mice, a single dose restored muscle energy production to youthful levels within an hour ^[4]. Yet in people, trial after trial came up short. Its first approval, in September 2025, came in a disease so rare that its developer estimates about 150 patients in the United States ^[12,13].

Origin: an opioid project that went sideways

The story began with painkillers. Szeto, a researcher at Weill Cornell Medical College in New York, was working with Peter Schiller on a synthetic opioid peptide with very high affinity for the mu-opioid receptor ^[2]. To their surprise, this highly polar, water-soluble peptide relieved pain in rodents after injection under the skin, which meant it was crossing the blood-brain barrier. Small, polar peptides were not supposed to do that ^[2].

Chasing that puzzle, the team found that their peptides entered cells and accumulated in mitochondria. They then designed analogues, including SS-31, that kept the mitochondrial targeting but had almost no affinity for opioid receptors ^[2]. The 2004 paper introducing these "cell-permeable peptide antioxidants" showed they reduced oxidative damage and cell death and protected isolated hearts after interrupted blood flow ^[1].

The peptides were licensed to a start-up, Stealth Peptides, of which Szeto was the scientific founder ^[2]. In clinical development SS-31 also carried the names MTP-131 and Bendavia ^[2].

How a four-letter peptide protects cardiolipin

Picture a factory whose production lines run along folded interior walls. In mitochondria those walls are the cristae, and their shape depends on cardiolipin, a phospholipid found in the inner membrane ^[3]. Cardiolipin also holds cytochrome c, a protein that ferries electrons along the energy chain ^[1,3].

Under stress, cytochrome c can act as a peroxidase instead, an enzyme that oxidises cardiolipin itself. The walls start to crumble and the mitochondria are damaged ^[3]. In 2013, Szeto's group used a fluorescent version of SS-31 to show that it binds cardiolipin with high affinity. The SS-31 and cardiolipin pair blocked cytochrome c's peroxidase activity, and in rats it protected cristae during kidney ischaemia ^[3].

Early papers framed SS-31 as an antioxidant, thanks to the dimethyltyrosine (Dmt) residue that can neutralise reactive oxygen species ^[1]. Later work shifted the emphasis to cardiolipin: less a sponge for damage, more a support beam for the membrane itself ^[2,3].

What the research shows

Ageing animals

In aged mice, one injection reversed age-related declines in resting and maximal mitochondrial ATP production and in the efficiency of energy coupling within an hour. It had no measurable effect in young 5-month-old mice ^[4]. Aged muscle became more resistant to fatigue, and 8 days of treatment improved whole-body endurance ^[4]. In a separate study, 8 weeks of SS-31 substantially reversed the stiffening of the ageing mouse heart known as diastolic dysfunction ^[5].

Heart attack and heart failure

The animal promise did not carry over to common heart disease. In EMBRACE STEMI, a phase 2a trial, SS-31 given intravenously during emergency treatment for a first anterior heart attack did not reduce infarct size ^[6]. In PROGRESS-HF, 71 patients with heart failure with reduced ejection fraction received placebo or one of two doses for 28 days. Neither dose changed heart volume or pumping function compared with placebo ^[7].

Primary mitochondrial myopathy

MMPOWER-3 was the pivotal test in inherited mitochondrial muscle disease. It randomised 218 adults to elamipretide or placebo for 24 weeks ^[8]. The trial missed both primary endpoints: the difference in six-minute walk distance was −3.2 metres (p = 0.69), and fatigue scores did not differ. The drug was well tolerated ^[8]. A follow-up phase 3 trial, NuPower, enrolled 102 people with nuclear-DNA forms of mitochondrial disease and finished in December 2024. No results were posted on ClinicalTrials.gov when we checked ^[15].

Barth syndrome

Barth syndrome is caused by defects in the TAZ gene, which leave cardiolipin abnormal ^[9]. It mainly affects males, typically starts with severe heart failure in infancy, and survivors often live with fatigue and poor stamina ^[12]. That direct link to cardiolipin made it a logical target.

The TAZPOWER trial randomised just 12 patients to elamipretide and placebo in a crossover design, 12 weeks each. Neither primary endpoint was met ^[9]. In the open-label extension, though, 6-minute walk distance improved by 95.9 metres at 36 weeks, with gains in symptom scores and knee-extensor strength ^[9]. Eight patients reached 168 weeks, with a cumulative walk improvement of 96.1 metres; injection-site reactions were the most common side effect ^[10].

The road to approval

Stealth filed for approval in January 2024. A 2024 FDA advisory committee concluded the drug was effective, but the FDA missed its April 2025 action date and then issued a rejection, citing observations from an inspection of a third-party manufacturing site ^[11]. The agency proposed a different route: accelerated approval based on knee-extensor muscle strength, which had improved by more than 45% in the phase 2 study ^[11].

The FDA approved Forzinity on stronger thigh muscles, a stand-in measure it judged "reasonably likely" to predict real benefit, not on proof that patients live or walk better ^[12].

The FDA granted accelerated approval on 19 September 2025 ^[12]. Supply began in late November 2025 ^[13]. In July 2026, Mighty Therapeutics, the parent of Stealth, dosed the first patient in 4TAZPower, a randomised, double-blind, placebo-controlled phase 4 trial lasting 72 weeks in patients aged 5 and older ^[14].

What we don't know yet

• Whether stronger muscles mean better lives. The approval rests on an intermediate endpoint, and continued approval may depend on the confirmatory trial ^[12,14].
• The size of the evidence base. The Barth data come from 12 randomised patients, and the clearest gains appeared in open-label phases without a placebo group ^[9,10].
• Why common diseases did not respond. Heart attack, heart failure and myopathy trials all missed their primary endpoints ^[6-8].
• Ageing in humans. The striking reversal of muscle and heart ageing has been shown in mice only ^[4,5].
• Long-term safety. The longest data cover 8 patients over about three years ^[10].

Regulatory status

• United States: FDA accelerated approval on 19 September 2025 as Forzinity (elamipretide) for Barth syndrome in patients weighing at least 30 kg. It received priority review, rare pediatric disease designation and a priority review voucher ^[12]. A study in children under 30 kg is planned to support a possible 2027 filing ^[14].
• Europe: We found no EU marketing authorisation; the company's commercial statements concern US patients ^[13].
• Other uses: Elamipretide remains investigational in dry age-related macular degeneration, where a phase 3 trial is expected to read out in late 2027, and in nuclear-DNA mitochondrial disease ^[13,14].
• Research material: Research-grade SS-31 is the same molecule, but it is not the approved medicine and is intended for laboratory research only.

In the lab: handling and storage

SS-31 is supplied as a lyophilised powder, usually as a salt. Standard practice is long-term storage at −20 °C, with 2–8 °C acceptable for short periods, protected from light and moisture. Because the sequence contains a non-standard residue (Dmt) and a D-amino acid, identity is best confirmed by mass spectrometry; the expected molecular weight of the free peptide is about 639.8. After reconstitution, many labs prepare single-use aliquots to avoid repeated freeze-thaw cycles. See our guide to reading a certificate of analysis and the SS-31 product page.

FAQ