The early TAZPOWER report contains two kinds of evidence in one publication. Its randomized comparison and subsequent open-label extension should be described separately. Primary report
What the report found
Twelve participants with Barth syndrome entered a double-blind, placebo-controlled crossover trial of elamipretide. Each treatment period lasted 12 weeks, separated by a washout. Neither primary endpoint, the six-minute walk test nor the Barth syndrome symptom assessment, was met in that randomized phase. Ten participants entered the open-label extension, with eight reaching 36 weeks. Improvements in walking distance and symptom scores were reported during the extension. The disclosure identifies one author as a Stealth BioTherapeutics employee. Primary report
Our appraisal: do not move results between phases
A later improvement does not change which endpoint the earlier randomized comparison met. Conversely, a randomized phase that did not meet its endpoints does not make all later observations meaningless. The task is to assign each observation the evidential weight supported by its design.
For example, an uncontrolled change over time can generate a hypothesis worth testing. It cannot, on its own, separate the treatment effect from every other explanation for change. That distinction should remain visible even when a within-group comparison has a small P value.
A practical reading order
First record the phase and comparator for each estimate. Next record how many participants contributed data at that time point. Then inspect the analysis plan, missing-data handling, and uncertainty. Finally, read the clinical interpretation without allowing it to erase the design labels.
For a crossover study, our full-text checklist also includes period effects, sequence, washout rationale, and the possibility of carryover. For an extension, it includes selection into follow-up and whether comparisons are to baseline or to a concurrent control. These are questions to investigate, not accusations about the report.
Frequently asked questions
Did the randomized phase meet its primary endpoints?
No. The report states that neither was met.
Were the extension improvements placebo-controlled?
The extension was open-label; it was not the same randomized comparison.
Were twelve people still represented at 36 extension weeks?
No. Eight reached that time point in this report.
Is an extension an independent replication?
Following members of the original cohort does not create an independent new cohort.
Does this establish general anti-aging benefits?
No. Such a claim is outside the population and outcomes summarized here.
Is this a current regulatory-status page?
No. It appraises a historical study and does not substitute for current product labeling.
How much of the source was reviewed?
The primary indexed abstract and disclosure. Full methods and later publication history require separate release review.