SS-31

The trial did not meet either main goal. At week 24, the difference between elamipretide and placebo was -3.2 meters in the 6-minute walk (95% confidence interval -18.7 to 12.3; P=0.69) and -0.07 points in the total fatigue score (P=0.37), neither statistically significant. Secondary measures also did not show a benefit. In a later analysis that was not planned in advance (post hoc), people with nuclear-DNA changes walked 25.2 meters further on elamipretide (29 people per group; P=0.03), while people with mitochondrial-DNA changes did not differ from placebo; this does not change the negative result for the whole group. The registry notes the trial was ended after the double-blind part missed its main goals.

• The trial missed both primary endpoints, so no benefit was shown for primary mitochondrial myopathy.
• The nuclear-DNA result is a post hoc subgroup (29 people per group) in which 3 people were reclassified after a data-entry error, so it can only generate a hypothesis.
• The trial was ended after the double-blind part, so the open-label extension was cut short and lacks a placebo comparison.
• The study included adults and teenagers aged 16 to 80 with mild to moderate impairment at the start, and 94% were White, so results may not apply to other groups.
• The registry and the journal report different numbers of people who stopped because of adverse events, and the sources do not explain why.
• The registry reports group averages without between-group comparisons; those come only from the journal paper.

• People who stopped because of an adverse event during the 24-week double-blind period: ClinicalTrials.gov participant flow reports 3 on elamipretide, 1 on placebo; Neurology paper, safety section reports Adverse events led to discontinuation in 8 of 109 (7.3%) on elamipretide and 2 of 109 (1.8%) on placebo. The paper counts more discontinuations; the paper's disposition text also mentions treatment withdrawn for an adverse event in 2 and 1 people. The reason for the difference is not established from the sources reviewed.

Randomized part (12 people, both treatments for 12 weeks each): neither main goal was met. The average six-minute walk distance rose from 400.1 m to 443.1 m on elamipretide and from 412.6 m to 443.9 m on placebo. The fatigue score fell from 7.67 to 6.31 on elamipretide and from 7.43 to 6.24 on placebo; the FDA label states that elamipretide was not superior to placebo on these endpoints, and knee-extensor strength changes were not seen in the randomized part. Open-label extension (uncontrolled; 10 people, 8 at week 168): the journal abstract reports significant improvements in six-minute walk distance (+95.9 m at 36 weeks in the 2021 report; +96.1 m cumulatively at week 168 in the 2024 report) and the FDA label reports knee-extensor strength rising by a median of 63 newtons at week 168. Because the extension had no placebo group, these gains cannot be separated from other influences.

• Only 12 people took part, all males with Barth syndrome aged 12 or older, so the results cannot be generalized.
• Both primary endpoints in the randomized crossover part were not met.
• The gains described in the extension come from an open-label, uncontrolled period with 10 people (8 at week 168), without a placebo group and with everyone knowing the treatment.
• The sources describe the extension differently: 168 weeks (registry text and the 2024 abstract) versus 192 weeks (FDA label), and use different baselines for the walking-test change.
• The FDA approval is an accelerated approval based on an intermediate endpoint, knee-extensor strength; clinical benefit still has to be confirmed in a trial.
• The trial was run before approval, when elamipretide was investigational. Only the abstracts of the two journal reports and the FDA label were reviewed, and the trial was funded by the drug's developer.

• Length of the open-label extension: ClinicalTrials.gov record and the 2024 journal abstract reports 168 weeks; FDA label reports 192 weeks. The label describes a 192-week extension; the registry's design text and the 2024 abstract say 168 weeks, while the registry's outcome tables also list week-192 visits. The reason for the difference is not established from the sources reviewed.
• Six-minute walk improvement at week 168 of the extension: ClinicalTrials.gov (change from the first baseline, by sequence) reports +89.3 m and +100.2 m; 2024 journal abstract (cumulative change in the extension) reports +96.1 m. The figures are measured against different baselines and are reported by sequence group in the registry versus overall in the abstract. The registry defines baseline as the pre-dose visit of the first treatment period; the abstract reports improvement from the extension baseline.