SS-31 in mitochondrial myopathy and heart failure research

SS-31's clinical history reflects this tension clearly, and honestly, and it's worth tracing in some detail — because the story isn't one of failure. It's one of a compound that turns out to be more specific in its utility than early enthusiasm suggested, and where the most important confirmation of the underlying science came not from the large trials everyone was watching but from a small, rare, genetically precise disease that most cardiologists had never heard of.

The early SS-31 research, in the laboratory of Hazel Szeto at Cornell, focused on ischemia-reperfusion injury — the particular pattern of damage that occurs when a tissue is deprived of blood flow and then has flow restored. The restoration is paradoxically damaging: the sudden return of oxygen generates a burst of reactive oxygen species that hits mitochondrial membranes at the moment they're most vulnerable. Cardiolipin, the structural lipid that maintains cristae geometry in the inner mitochondrial membrane, is particularly susceptible to this oxidative burst. In preclinical models, SS-31 administered at the time of reperfusion preserved mitochondrial membrane integrity, reduced infarct size in hearts, and protected kidneys from ischemic injury. The effect sizes were substantial. The mechanism was elegant. The obvious next move was toward cardiac disease, and specifically toward heart failure.

The company that took SS-31 into clinical development was Stealth BioTherapeutics — a name that referred to the peptide's stealth-like targeting mechanism, not to any particular secrecy. They ran a series of clinical programs that, collectively, tell the story of the gap between promising mechanism and clinical efficacy in heterogeneous patient populations.

The most visible program was the EMBRACE trial series for heart failure with preserved ejection fraction, or HFpEF. HFpEF is the form of heart failure in which the heart pumps adequately but cannot relax and fill properly — a diastolic problem rather than a systolic one. It represents roughly half of all heart failure cases, it disproportionately affects older women, it has far fewer proven treatments than its systolic counterpart, and mitochondrial dysfunction in cardiac muscle is a plausible contributor to the disease. The EMBRACE-HFpEF trial looked at whether SS-31 could improve exercise tolerance and symptoms in HFpEF patients. The results were mixed — some endpoints moved in promising directions, but the trial did not demonstrate the kind of clear efficacy signal that would support a broad indication. It was not a clean success.

This was disappointing to people who had followed the mechanism. But mixed results in HFpEF are not unique to SS-31. The disease is heterogeneous: different patients get there through different paths, with different degrees of mitochondrial involvement, different amounts of fibrosis, different metabolic profiles. A mechanism that is genuinely important in some patients with HFpEF may be diluted into statistical noise when the trial enrolls the full spectrum of HFpEF patients regardless of their mitochondrial status. This is a general problem in trial design for conditions with multiple pathophysiological subtypes, not a condemnation of the underlying science.

The MMPOWER program for primary mitochondrial myopathy followed a different arc. Mitochondrial myopathy — a group of conditions caused by mutations in mitochondrial or nuclear DNA that directly impair mitochondrial function in muscle tissue — represents a more mechanistically targeted population. These are patients for whom mitochondrial dysfunction is not a secondary feature of their disease but the primary pathology. Exercise intolerance, muscle weakness, fatigue at levels that significantly impair daily function. The MMPOWER-3 trial was the pivotal study. It, too, had a primary endpoint that did not achieve statistical significance in the full study population. But the signal in certain measures and subgroups was meaningful enough to sustain the program's trajectory.

What changed the trajectory more significantly was Barth syndrome.

Barth syndrome is an X-linked genetic disorder caused by mutations in the TAZ gene, which encodes tafazzin — an enzyme responsible for cardiolipin remodeling. In healthy mitochondria, cardiolipin is synthesized in an immature form and then remodeled: the enzyme exchanges its fatty acid groups to produce a mature cardiolipin species with a specific composition that optimizes its structural role in the inner membrane. In Barth syndrome, this remodeling fails. Immature cardiolipin accumulates. Mature cardiolipin is depleted. The inner mitochondrial membrane is structurally compromised in a way that is genetically fixed and unavoidable. The consequences are severe: dilated cardiomyopathy, skeletal myopathy, neutropenia, growth retardation, and historically a significant risk of early death from cardiac causes. The syndrome predominantly affects young males. The condition is rare — estimated at roughly 1 in 300,000 to 400,000 live births.

For SS-31, Barth syndrome was not a pivot to a consolation prize. It was the most precise possible test of the cardiolipin-targeting hypothesis. Here was a population in which cardiolipin dysfunction was not incidental, not secondary, not one feature among many — it was the central genetically determined defect. If SS-31 worked by stabilizing cardiolipin and protecting inner membrane architecture, Barth syndrome patients were the group most likely to respond to it, and the most meaningful population in which to validate the mechanism.

The TAZPOWER trial was a double-blind, placebo-controlled crossover study in Barth syndrome patients. It measured functional outcomes including the six-minute walk test — a measure of exercise tolerance — and echocardiographic parameters. The results were positive. Patients on elamipretide showed improvements in exercise tolerance and some cardiac function measures compared to placebo. An open-label extension study followed patients longer and reinforced the functional benefit pattern. Based on this data, the FDA granted accelerated approval for elamipretide — marketed as Stegazo — for the treatment of Barth syndrome in 2023, under the accelerated approval pathway, which requires post-marketing confirmatory trials to maintain the approval. It was the first FDA-approved treatment specifically for Barth syndrome, and the first FDA-approved cardiolipin-targeted therapy of any kind.

The accelerated approval designation is worth understanding clearly. It is not provisional in the sense of being uncertain — it reflects the FDA's judgment that the available data supports efficacy on the basis of a surrogate or intermediate endpoint likely to predict clinical benefit. Post-marketing confirmatory studies are required. The approval is conditional on those studies confirming benefit, and the process is ongoing. But the FDA did find the data sufficient to justify approval in a disease with no prior approved therapies, in a patient population with significant unmet need. That is meaningful.

What the Barth syndrome approval validates is not merely that SS-31 works for one rare disease. It validates the targeting logic at the center of the compound's design. A regulatory agency has formally agreed that a peptide designed to bind cardiolipin and protect inner mitochondrial membrane architecture can produce measurable clinical benefit in a disease defined by cardiolipin remodeling failure. This is the mechanistic claim SS-31 has been carrying since Szeto's laboratory work, now endorsed by the most stringent evidence standard in clinical medicine.

The question that follows — how far does this logic extend? — is the one now being explored across several parallel research programs. In Leber's hereditary optic neuropathy, an inherited mitochondrial disease affecting retinal ganglion cells, mitochondrial dysfunction is the direct cause of vision loss. The logic of inner membrane protection applies; trials are ongoing. In Friedreich's ataxia, a progressive neuromuscular disease caused by frataxin deficiency that impairs iron-sulfur cluster assembly and leads to mitochondrial dysfunction in cerebellar and cardiac tissue, there is mechanistic rationale for cardiolipin protection and research is active. In primary mitochondrial myopathies beyond the MMPOWER populations already studied, the ongoing collection of evidence continues to refine who the most likely responders are.

There is a theme in the data that matters for interpreting all of this: SS-31 appears to be more useful when mitochondrial dysfunction is mechanistically primary and when the inner membrane damage pathway is directly implicated, and less reliable when mitochondrial dysfunction is a secondary downstream feature of a more heterogeneous condition. HFpEF is heterogeneous. Primary mitochondrial myopathies are not. Barth syndrome is maximally specific. The pattern is consistent with the targeted nature of the mechanism — and it suggests that the path forward for elamipretide is not in broad cardiometabolic conditions but in the narrower set of diseases where cardiolipin damage is not a contributing factor among many but the central pathology being addressed.

The commercial landscape around ultra-rare disease approvals is complicated in its own right. Stegazo's approval for Barth syndrome represents a genuine clinical advance for a small population with previously no approved options. It also enters a market where rare disease drug prices generate significant access and ethical questions. These are conversations that sit outside the mechanism science, but they're worth noting honestly: the approval creates access for patients who had none, and does so through a price structure that raises questions about sustainability and equity in rare disease pharmacology that are not unique to this drug.

The arc from Szeto's laboratory through ischemia-reperfusion models, through the large heart failure and myopathy trials, to the precise genetic confirmation in Barth syndrome is not a story of a drug that mostly failed and finally found a niche. It's a story of a mechanism that turned out to be real and specific — specific enough that its utility tracks closely with how central cardiolipin biology is to a given disease. The broader investigational programs now operating in optic neuropathy, ataxia, and primary mitochondrial disease are probing that specificity further. The results will take years to accumulate. But the evidentiary foundation has been established in a form that matters: one FDA approval, one genetically validated mechanism, one disease where the lipid this peptide was designed to protect is the lipid whose dysfunction is the disease.