SS-31 (elamipretide): benefits, evidence and legal status

SS-31 (elamipretide) is a synthetic four-amino-acid peptide that binds cardiolipin in the inner mitochondrial membrane; the FDA granted it accelerated approval as Forzinity on September 19, 2025, for Barth syndrome only.

That approval is narrow. The label covers muscle strength in patients with Barth syndrome who weigh at least 30 kg, and it rests on a secondary endpoint seen in an open-label extension of a 12-patient trial whose randomized phase missed both primary endpoints (Forzinity label). For every other use, including the “SS-31” sold online for energy, ageing or recovery, elamipretide is unapproved, and its largest published trials outside Barth syndrome were negative.

Item	Status
US FDA	Approved only as Forzinity (NDA 215244), accelerated approval 2025-09-19, to improve muscle strength in Barth syndrome, patients ≥30 kg (Drugs@FDA); unapproved for any other use
Research-use “SS-31”	Named by FDA as an unapproved new drug in an August 24, 2026 warning letter to an online seller (FDA warning letter)
Compounding (503A/503B)	Not on FDA’s Category 2 list; no FDA statement specific to compounded elamipretide found (see the legal section)
WADA 2026	Not named on the List; see the legal section (WADA 2026 List)
EU	Orphan designation for Barth syndrome (2021); designation is not a marketing authorisation (EMA)
Strongest evidence	Approved label (Barth syndrome); Human RCTs (n=12 to 300), mostly negative on primary endpoints
Human trials	29 interventional trials registered on ClinicalTrials.gov, plus 1 expanded-access program, e.g. NCT03098797, NCT03323749
Last checked	2026-10-03

Identity	Value
Sequence	D-Arg-2,6-dimethyl-Tyr-Lys-Phe-NH2 (4 amino acids; D-arginine, the rest L)
Molecular formula	C32H49N9O5 (free base); hydrochloride salt C32H49N9O5·3HCl
Molecular weight	639.8 g/mol (free base); 749.2 (salt)
CAS	736992-21-5
PubChem CID	11764719
Other names	SS-31, MTP-131, Bendavia, Forzinity (brand)
Half-life	About 3 to 4 hours in plasma (FDA briefing)
Routes studied	Subcutaneous (approved), intravenous, oral, eye drops
Class	Mitochondrial cardiolipin binder (Szeto-Schiller peptide)

What is SS-31 (elamipretide)?

SS-31 is one of the Szeto-Schiller (SS) peptides, short peptides built from alternating aromatic and basic amino acids with a dimethyltyrosine; in lab work these peptides crossed cell membranes and concentrated about 1,000-fold in the inner mitochondrial membrane [Lab] (Zhao 2004). Stealth BioTherapeutics developed it under the codes MTP-131 and Bendavia before it received the name elamipretide (PubChem). It was tested in heart attack, heart failure, kidney disease, eye disease and mitochondrial myopathy before its approval in Barth syndrome, a rare X-linked disease caused by defects in the TAFAZZIN gene that disturb cardiolipin; FDA estimated about 130 affected people in the US (FDA briefing).

How does SS-31 work?

Cardiolipin is a fat (phospholipid) found almost only in the inner membrane of mitochondria, where it shapes the folds (cristae) that hold the energy-producing machinery.

• Cardiolipin binding: SS-31 bound cardiolipin with high affinity in lab tests [Lab] (Birk 2013); the label describes the drug as a cardiolipin binder that localizes to the inner mitochondrial membrane [Approved label] (Forzinity label).
• Cytochrome c peroxidase block: by binding cardiolipin, SS-31 stops cytochrome c from acting as a peroxidase that damages cardiolipin, while leaving its electron-carrying role intact [Lab] (Szeto 2014).
• Cristae protection after ischemia: in rats, pretreatment protected cristae during kidney ischemia, sped ATP recovery and reduced kidney dysfunction [Animal] (Birk 2013).
• Reactive oxygen species: SS peptides reduced mitochondrial oxidant production and cell death in cultured cells and improved contractile force in an isolated heart model [Lab] (Zhao 2004).
• Muscle energetics in people: a single infusion raised maximal ATP production in older adults’ hand muscle right after dosing, with no difference by day 7 [Human RCT, n=39] (Roshanravan 2021).

What the research shows

The table grades each common claim by the strongest evidence behind it, following the site’s editorial policy and methodology. A pattern runs through it: small or early trials reported signals, and the larger controlled trials did not confirm them.

Claim	Evidence level	Best study and result
Improves muscle strength in Barth syndrome	Approved label (accelerated)	Forzinity label: no gain in the 12-week randomized phase; median +63 N at week 168 open-label (n=8)
Improves walking and fatigue in Barth syndrome	Human RCT (n=12), negative	Reid Thompson 2021: neither primary endpoint met
Long-term benefit in Barth syndrome	Human, open-label (n=8–10)	Thompson 2024: 6-minute walk +96.1 m at week 168, uncontrolled
Improves primary mitochondrial myopathy	Human RCT (n=218), negative	Karaa 2023: walk difference −3.2 m, p=0.69
Limits heart-attack damage	Human RCT (n=300), negative	Gibson 2016: no reduction in infarct size
Improves heart failure	Human RCT (n=71), negative	Butler 2020: no change in heart volume at 4 weeks
Slows dry macular degeneration	Human RCT (n=176), primary endpoints not met	Ehlers 2025: 43% slower ellipsoid-zone loss, nominal
Protects kidneys during stenting	Human RCT pilot (n=14)	Saad 2017: less hypoxia, higher renal blood flow
Raises muscle energy in older adults	Human RCT (n=39)	Roshanravan 2021: brief ATP rise, no fatigue benefit
Protects mitochondria under stress	Lab and Animal	Birk 2013, rats: cristae preserved after kidney ischemia

Barth syndrome

TAZPOWER randomized 12 patients to 12 weeks of elamipretide and 12 weeks of placebo; neither the 6-minute walk nor the fatigue score improved [Human RCT, n=12] (Reid Thompson 2021). FDA put the p-values at 0.97 and 0.89 and refused to file the first application in 2021 because it lacked a single adequate, well-controlled trial (FDA briefing). In the open-label extension, walking distance rose by 96.1 m at week 168 in the 8 patients who stayed [Human, open-label, n=8] (Thompson 2024), and a comparison of those 8 patients with 19 untreated patients from a natural-history study favored the drug [Human, external comparison, not randomized] (Hornby 2022). The approval rests on knee-extensor strength, which FDA treats as reasonably likely, not proven, to predict benefit (FDA press release).

Primary mitochondrial myopathy

A 5-day intravenous trial reported a dose-related gain in walking distance (64.5 m vs 20.4 m, p=0.053) [Human RCT, n=36] (Karaa 2018), and a 4-week crossover found a 19.8 m difference that was not significant [Human RCT, n=30] (Karaa 2020). The 24-week phase 3 MMPOWER-3 found no effect on walking or fatigue [Human RCT, n=218] (Karaa 2023). A post hoc analysis pointed to a possible signal in patients with mtDNA-maintenance (replisome) variants, which led to the NuPower trial [Human, post hoc] (Karaa 2024).

Heart and kidney

A single high-dose infusion reduced left-ventricular volumes in heart failure (−18 mL end-diastolic, P=0.009) [Human RCT, n=36] (Daubert 2017), but four weeks of daily injections in PROGRESS-HF did not change end-systolic volume or ejection fraction [Human RCT, n=71] (Butler 2020). In EMBRACE STEMI, a 1-hour infusion during angioplasty for heart attack did not reduce infarct size (Gibson 2016). In a kidney-artery stenting pilot, hypoxia rose 47% with placebo and fell 6% with elamipretide [Human RCT, n=14] (Saad 2017).

Eye disease

In ReCLAIM-2, 48 weeks of daily injections missed both primary endpoints in dry age-related macular degeneration but showed nominally slower loss of the photoreceptor layer [Human RCT, n=176] (Ehlers 2025); that measure is the primary endpoint of the phase 3 ReNEW trial. Eye drops in Leber hereditary optic neuropathy did not improve visual acuity [Human RCT, n=12] (Karanjia 2024).

Human studies and trials

A ClinicalTrials.gov search on October 3, 2026 (elamipretide, SS-31, MTP-131, Bendavia) returned 29 interventional elamipretide trials and one expanded-access program (NCT04689360). The main ones:

Trial	Design and n	Status and result
NCT03098797 TAZPOWER, Barth	Phase 2/3 crossover + extension, n=12	Completed 2021; primary endpoints missed; basis of approval
NCT07531251 4TAZPower, Barth	Phase 3b/4, 72 weeks, est. n=48	Recruiting since July 2026; confirmatory trial
NCT03323749 MMPOWER-3, myopathy	Phase 3, n=218	Terminated; primary endpoints not met
NCT02367014 MMPOWER	Phase 1/2, IV, n=36	Completed; borderline walk gain
NCT02805790 MMPOWER-2	Phase 2 crossover, n=30	Completed; walk difference not significant
NCT02976038 myopathy extension	Phase 2 open-label, n=28	Terminated after MMPOWER-3 failed
NCT05162768 NuPower	Phase 3, 48 weeks, n=102	Completed Dec 2024; no results posted
NCT01572909 EMBRACE STEMI	Phase 2a, IV, n=300	Completed; no infarct-size reduction
NCT02788747 PROGRESS-HF	Phase 2, n=71	Completed; no change in heart volume
NCT02914665 acute heart failure	Phase 2, n=308	Completed 2017; no results posted
NCT01755858 EVOLVE, renal stenosis	Phase 1/2, n=16	Terminated; pilot data published
NCT03891875 ReCLAIM-2, dry AMD	Phase 2, n=176	Completed; primary endpoints missed
NCT06373731 ReNEW, dry AMD	Phase 3, n=313	Active, not recruiting; est. completion 2027
NCT02693119 LHON eye drops	Phase 2, n=12	Completed; acuity endpoint missed
NCT05168774 Friedreich ataxia	Phase 1/2, n=20	Completed 2024; results posted, no publication found
NCT02245620 MOTION, older adults	Phase 2, n=41	Completed; brief ATP rise
NCT07275424 SHAPE, healthy ageing	Phase 2a open-label, est. n=30	Recruiting

The rest are phase 1 safety and pharmacokinetic studies in healthy volunteers, including NCT01115920 (intravenous, n=40) and two oral studies of 30 people each, plus small trials in heart failure, Fuchs corneal dystrophy and macular disease.

Doses used in studies

These amounts are reported for scientific context only. They are not recommendations.

Route and amount	Population	Study
Subcutaneous, 40 mg once daily (20 mg in severe renal impairment)	Barth syndrome, ≥30 kg (approved label)	Forzinity label
Subcutaneous, 40 mg daily, 24 weeks	Human, mitochondrial myopathy	Karaa 2023
Subcutaneous, 4 mg or 40 mg daily, 28 days	Human, heart failure	Butler 2020
Intravenous, 0.01–0.25 mg/kg/h for 2 hours	Human, mitochondrial myopathy	Karaa 2018
Intravenous, 0.05 mg/kg/h for 1 hour	Human, heart attack	Gibson 2016
Eye drops, 1% solution	Human, LHON	Karanjia 2024

No established human dose exists outside approved labels. The only labeled dose applies to Barth syndrome under a prescriber’s care.

Side effects and safety

The safety record comes mostly from controlled trials of a pharmaceutical-grade product, and it is reasonably consistent: injection-site reactions are near universal, and allergic reactions are the main serious risk.

• Injection-site reactions [Approved label, n=12]: redness in 12 of 12 patients on elamipretide vs 3 of 12 on placebo; pain 75% vs 42%; hardening and itching 67% each vs 17% (Forzinity label). Injection-site reactions occurred in 80% of patients in a myopathy trial [Human RCT, n=30] (Karaa 2020).
• Hypersensitivity [Approved label]: serious allergic reactions needing emergency treatment have been reported, with rash, eczema-like skin reactions or cough, from minutes to months after starting. Two Barth patients stopped treatment because of hives (FDA briefing).
• Eosinophilia [Approved label]: eosinophil counts rose often with 30 days or more of treatment, peaking near day 90 and returning to baseline, without symptoms.
• Kidney function: elamipretide leaves the body in urine; exposure rose 125% in severe renal impairment, and the label halves the dose there.
• Unknowns: carcinogenicity studies have not been done, and FDA requires mouse and rat studies after approval (approval letter). No one aged 65 or older with Barth syndrome was studied.
• Benzyl alcohol: Forzinity contains benzyl alcohol and is not approved for newborns because of the risk of fatal “gasping syndrome”.

Products sold online as “SS-31” are a separate problem. FDA’s August 2026 warning letter treated one seller’s SS-31 as an unapproved new drug (FDA warning letter); such products carry none of the label’s quality controls, and their content and purity are unverified. Per the label, people with a serious allergy to elamipretide or its ingredients must not receive it.

Legal and regulatory status

Jurisdiction	Status (checked 2026-10-03)
US FDA	Forzinity approved 2025-09-19 under accelerated approval, Barth syndrome only (Drugs@FDA)
US, online “SS-31”	Unapproved new drug per FDA warning letter, August 2026
US compounding	Not on the Category 2 list; FDA’s 503A rules allow bulk substances that are components of FDA-approved drugs (FDA 503A page), but FDA has published nothing specific on elamipretide
WADA	Not named on the 2026 or 2027 List
EU	Orphan designations for Barth syndrome (2021) and mitochondrial DNA depletion myopathy (2022); not authorised

Dated events: FDA refused to file the first application in 2021; Stealth resubmitted it on January 29, 2024; an FDA advisory committee met on October 10, 2024 and voted on whether elamipretide is effective in Barth syndrome (FDA questions); FDA issued an action letter on May 15, 2025 and approved the drug on September 19, 2025 (approval letter). Continued approval may depend on a confirmatory trial verifying clinical benefit; FDA’s schedule has that trial starting in March 2026 and reporting in March 2030.

For sport, elamipretide is not named on the WADA 2027 List. The S0 class covers substances with “no current approval by any governmental regulatory health authority”, and the US approval means S0 no longer obviously fits. No WADA statement on elamipretide was found, so its status in sport is unsettled rather than cleared; see the site’s overview of peptides prohibited in sport. In the EU, the orphan designation gives development support only; EMA states it “does not mean the medicine is available or authorised for use” (EMA).

Related peptides

MOTS-c, a peptide encoded in mitochondrial DNA, is another peptide marketed for “mitochondrial” effects; unlike elamipretide it is named on the WADA List, under S4.4.1 metabolic modulators (WADA 2026 List). For a peptide sold online with even thinner human evidence, see BPC-157.