SS-31 (elamipretide): the phase 3 trial that missed

SS-31 is elamipretide, a synthetic tetrapeptide developed to bind cardiolipin on the inner mitochondrial membrane. It is the most thoroughly tested mitochondria-targeting peptide in the published literature, and it has already had a phase 3 readout. That readout was negative on both primary endpoints. This page reports it as printed, then reports the post hoc analysis that followed.

The phase 3 trial

MMPOWER-3 was a pivotal, phase 3, randomised, double-blind, placebo-controlled trial in adults with genetically confirmed primary mitochondrial myopathy. Eligible participants were randomised 1:1 to 24 weeks of subcutaneous elamipretide at 40 mg per day or placebo. The two primary efficacy endpoints were change from baseline to week 24 in distance walked on the six-minute walk test and total fatigue on the Primary Mitochondrial Myopathy Symptom Assessment.

218 participants were randomised, 109 to elamipretide and 109 to placebo. Mean age was 45.6 years, 64% were women and 94% were White. Most participants — 162, or 74% — had a mitochondrial DNA alteration, the remainder having nuclear DNA defects. The most frequent bothersome symptom at screening was tiredness during activities, at 28.9%.

The baselines were: mean 6MWT distance 336.7 ± 81.2 metres, mean PMMSA total fatigue score 10.6 ± 2.5, mean Neuro-QoL Fatigue Short-Form T score 54.7 ± 7.5.

The trial did not meet either primary endpoint. The difference in least squares mean from baseline to week 24 on the 6MWT was -3.2 metres (95% CI -18.7 to 12.3; p = 0.69) and on the PMMSA total fatigue score -0.07 (95% CI -0.10 to 0.26; p = 0.37). The authors classify the result as Class I evidence that elamipretide does not improve the 6MWT or fatigue at 24 weeks compared with placebo. The paper reports treatment as well tolerated, with most adverse events mild to moderate.

The post hoc analysis

A 2024 paper re-analysed MMPOWER-3 by genotype. Subjects with mtDNA pathogenic variants or single large-scale mtDNA deletions made up 74% of the population, and 70% of the mtDNA cohort had either single large-scale mtDNA deletions or MT-TL1 pathogenic variants. Most subjects in the nuclear-DNA cohort had pathogenic variants in genes required for mtDNA maintenance, the majority in POLG and TWNK.

In the mtDNA replisome post hoc cohort, the 6MWT change in the elamipretide group was 25.2 ± 8.7 metres against 2.0 ± 8.6 metres for placebo, p = 0.06 — a trend the authors describe as trending towards significant, not a result that met significance. Within that cohort, participants whose genotype presented as chronic progressive external ophthalmoplegia showed a 6MWT change of 37.3 ± 9.5 metres against -8.0 ± 10.7 metres for placebo, p = 0.0024.

Both of those are post hoc subgroup results inside a trial that missed its primary endpoints. They are the basis for the authors’ proposed follow-up phase 3 in mtDNA maintenance-related disorders, which is the honest description of where this molecule stands: not a failed drug so much as a drug whose phase 3 design did not stratify by genotype.

Exposure-response analyses in the nuclear-DNA cohort showed a weak positive correlation between plasma elamipretide concentration and 6MWT improvement. “Weak” and “correlation” are the abstract’s words.

What was still measurable in healthy muscle

A 2021 PLoS One randomised, double-blind, placebo-controlled trial measured mitochondrial capacity by non-invasive magnetic resonance and optical spectroscopy in the first dorsal interosseous muscle of 39 healthy older adults aged 60 to 85, 46% female, enrolled on the presence of poorly functioning mitochondria. A single dose, given as a two-hour infusion, elevated mitochondrial energetic capacity relative to placebo: ΔATPmax p = 0.055 and %ΔATPmax p = 0.045, immediately after the infusion. By day 7 there was no difference, which the authors attribute to the half-life of elamipretide in human blood.

Two things in that paragraph matter. Resting mitochondrial coupling showed no significant change. And despite the ATPmax rise, there was no significant effect of treatment on fatigue resistance in that muscle. So the physiological readout moved and the functional readout did not — the same pattern as the negative phase 3.

Mechanism, and how far it is carried by animals

A 2025 review describes elamipretide as targeting the inner mitochondrial membrane, with the cardiolipin binding properties corroborated by different investigative teams, and new findings on modulation of mitochondrial membrane electrostatic potentials and assembly of cardiolipin-dependent proteins. The review notes that initial descriptions involved reactive oxygen species scavenging and that the last ten years expanded that account.

The functional claims in that literature are animal. A 2022 review in Heart Failure Reviews reports that elamipretide has been shown to increase left ventricular ejection fraction in dog models of heart failure with reduced ejection fraction, and to prevent left ventricular remodelling in rats. The same review reports that in early-phase clinical trials administration has not resulted in severe adverse events and has shown improvements in cardiac haemodynamics at the highest doses, and that additional studies are necessary to describe long-term safety and efficacy.

What the data do not show

They do not show an improvement in the six-minute walk test in primary mitochondrial myopathy. MMPOWER-3 is the trial that tested exactly that, in 218 people, for 24 weeks, and it read -3.2 metres against placebo.

They do not show a fatigue benefit. The PMMSA total fatigue score difference was -0.07 with a confidence interval crossing zero.

They do not show a functional benefit in healthy older muscle. ATPmax moved; fatigue resistance did not.

They do not show a human animal-strength or exercise-capacity result, because the ejection fraction and remodelling findings cited here are in dogs and rats.

The genotype subgroup results are real and printed, but they are post hoc inside a trial that missed its primary endpoints.

Where this sits on the shelf

The compounds as supplied by the manufacturer: SS-31 · the 50 mg SS-31 presentation

What we supply

The SS-31 preparations above, as supplied by the manufacturer, tracked, from inside the EU.

Research use only. This page summarises published studies for research reference. It is not medical advice, not a protocol, and not a suggestion for human use. Nothing we supply is for human or veterinary use.