SS-31 / elamipretide

SS-31 — elamipretide, one of the Szeto-Schiller peptides — is an aromatic-cationic tetrapeptide that concentrates on the inner mitochondrial membrane and binds cardiolipin, the phospholipid that holds mitochondrial cristae and the respiratory machinery together. Unlike most of this section, it is a genuine pharmaceutical with a completed clinical program and, since September 2025, an FDA-approved product — just not for anything involving a joint.

Why cartilage might care about a mitochondrial drug

Chondrocytes in osteoarthritic cartilage show measurably reduced electron-transport-chain activity, and mitochondrial dysfunction plausibly feeds the oxidative stress, catabolic signaling, and cell death that degrade cartilage matrix. More specifically: when cartilage takes a single hard impact, chondrocyte mitochondria depolarize and lose respiratory function within minutes to hours — before the cells die. That makes mitochondria a genuinely upstream target after joint injury, which is exactly where SS-31 acts: by binding cardiolipin it stops cytochrome c from turning peroxidase and destroying the cristae, preserving the cell's ability to make ATP through the crisis.

The impact-injury rescue

The cartilage evidence comes from one coherent line of work at Cornell, with the peptide's inventor as a co-author. In bovine cartilage explants given a hard single impact, injury multiplied chondrocyte death two-and-a-half-fold. SS-31 added to the media up to six hours after impact cut that induced death by roughly forty percent, leaving viability statistically indistinguishable from cartilage that was never injured — and the protection held for the week the explants stayed in culture. Treated cartilage also lost no more of its glycosaminoglycan matrix than uninjured tissue, where untreated injured explants lost about thirty percent more. Two sharp caveats live in the same data: delaying treatment to twelve hours lost the effect entirely, and most of the tissue was neonatal.

A follow-up that tracked thousands of individual cells showed what the drug is doing: impact triggers a wave of mitochondrial depolarization within minutes, tightly correlated with how hard the local tissue was strained. Pretreated explants showed depolarization and death sixty to ninety percent lower — comparable to unimpacted tissue — and the strain correlation vanished.

The one in-vivo joint test is a twelve-horse study: arthroscopic impact injuries to the talus, then a blinded intra-articular injection of SS-31 or saline one hour later. Treated joints released less mitochondrial DNA into synovial fluid over the following month and showed less cell death at every time point. Encouraging — but it measured biomarkers and cell death in six horses per arm, not cartilage structure, function, or whether arthritis developed.

What the human trials do and don't establish

Elamipretide has been through more randomized human testing than everything else in this section combined: a Barth syndrome crossover trial with a multi-year extension, a 218-patient mitochondrial myopathy trial, a 71-patient heart failure trial, and a 176-patient macular degeneration trial. Read precisely, that record establishes two things.

First, safety exposure. Hundreds of people have injected 40 mg daily, some for over three years, under trial monitoring — an evidence base peptides like BPC-157 simply do not have.

Second, and just as precisely: not one of those trials measured a joint. The blinded phases were also mostly negative on their own terms. The Barth crossover missed both primary endpoints, with improvements appearing only in the uncontrolled extension. The mitochondrial myopathy trial — the indication closest to the drug's mechanism — produced Class I evidence of no benefit on walking distance or fatigue. The heart failure trial showed nothing at either dose, and the macular degeneration trial missed its primaries while generating retinal-layer signals now being tested in phase 3. The FDA's September 2025 accelerated approval for Barth syndrome rests on a knee-extensor muscle-strength surrogate and requires a confirmatory trial; the word "knee" in that endpoint refers to a myopathy strength measure, not to cartilage. A sweep of every elamipretide registration on ClinicalTrials.gov finds no OA, cartilage, or joint trial of any kind as of August 2026.

There is also a translational mismatch worth stating plainly. The cartilage data show a rescue window of about six hours after a defined impact, delivered into the joint or the culture media. That is a model of acute post-injury treatment — closer to an emergency-department intervention than to ongoing subcutaneous use in a joint injured years ago, which no study of any kind has examined.

Safety

The pharmaceutical safety profile is consistent across trials: injection-site reactions dominate — erythema in twelve of twelve Barth patients on drug versus three of twelve on placebo, and in seventy to eighty percent of long-term extension patients — with occasional urticaria or drug eruption driving discontinuation. In the largest trial, adverse events were near-universal on drug and discontinuations for adverse events ran about seven percent versus two percent on placebo. Serious adverse events have not been attributed to treatment, and no deaths occurred in any trial.

All of that describes pharmaceutical-grade elamipretide under monitoring. The approved product is prescription-only for Barth syndrome; what peptide vendors sell as "SS-31" is a research chemical whose identity, purity, sterility, and actual dose are unverified, and no safety data exist for self-injected use in healthy or osteoarthritic adults at any dose or schedule. The only randomized human experience at a low dose — 4 mg daily for four weeks, in heart failure — was well tolerated and showed no efficacy signal on any endpoint.

What would change the tier

A human joint trial — any human joint trial. The pieces for one exist in a way that is rare for this section: a plausible upstream mechanism, a reproducible ex vivo rescue, an in-vivo delivery test, and even a candidate response biomarker in synovial-fluid mitochondrial DNA, developed explicitly to make such a trial measurable. Until someone runs it, SS-31 in a joint remains an elegant preclinical story attached to a drug whose human wins, so far, have come everywhere except where the syringe is usually aimed.