The British Medical Journal has retracted a phase 3 randomized clinical trial that claimed to demonstrate prevention of heart failure following acute myocardial infarction through intracoronary infusion of mesenchymal stem cells. The retracted article, authored by Attar and colleagues and originally published on October 29, 2025 under the designation PREVENT-TAHA8, has been withdrawn following a review process that identified substantial concerns regarding data integrity, trial conduct, and reporting reliability.

The retraction represents a consequential development in the field of cardiac regenerative medicine, where mesenchymal stem cell (MSC) therapy has attracted considerable investigative interest over the past two decades. The withdrawal of a phase 3 trial on these grounds raises fundamental questions about the evidentiary basis supporting MSC-based interventions in post-infarction heart failure, and reinforces ongoing concerns about research integrity standards in cardiovascular clinical trials.

Grounds for Retraction

The BMJ editors identified multiple categories of concern that, taken together, rendered the trial unreliable. These concerns fell broadly into three areas: enrollment irregularities, participant count discrepancies, and data pattern anomalies.

With respect to enrollment, the journal identified apparent recruitment of participants outside the study’s stated inclusion criteria. Specifically, the record indicated enrollment of individuals older than 65 years, a population that the trial’s protocol appears to have excluded. Recruitment outside defined inclusion criteria represents a protocol deviation of considerable consequence. It introduces uncontrolled heterogeneity into the study population, potentially confounds outcome analyses, and undermines the internal validity of any reported treatment effect. In a phase 3 trial, where the objective is to confirm efficacy and safety in a well-characterized population, such deviations are not minor administrative errors. They affect the interpretation of every endpoint.

The second category involved discrepancies in the reported number of enrolled participants. A mismatch between enrollment figures at different stages of a trial, whether between the screening log, the randomization record, or the intention-to-treat population, signals either administrative failure or deliberate manipulation of the dataset. Without reconciliation, it is not possible to determine the true denominator against which outcomes were measured, which invalidates statistical inference for both primary and secondary endpoints.

The third and arguably most technically damning category involved anomalous patterns within the data itself. The journal’s review identified what it characterized as improbable values and repeated numbers within the dataset. In clinical trial data, repeated identical values across independent observations, particularly for continuous physiological measurements such as ejection fraction, biomarker concentrations, or functional capacity scores, are statistically implausible under natural biological variation. Such patterns are a recognized signal of data fabrication or inappropriate manipulation. The review additionally identified mismatched and missing secondary outcomes, further eroding confidence in the completeness and accuracy of the reported results.

Timeline of the Retraction Process

The journal was first alerted to potential problems following the article’s October 2025 publication, consistent with the post-publication review mechanisms that high-impact journals increasingly rely upon. The BMJ published an expression of concern while the investigation was ongoing, a standard intermediate step that signals active scrutiny without prejudging the outcome. The authors cooperated with the review process and provided supplementary documentation. They also responded publicly through PubPeer, the post-publication peer review platform where researchers can submit and discuss concerns about published work.

Notwithstanding the authors’ cooperation and responses, the journal determined that the concerns were substantiated and that retraction was warranted. The formal retraction was published under BMJ reference bmj.s579.

Clinical Context and Significance of the Retraction

MSC therapy for ischemic cardiomyopathy has been an active area of investigation since early pilot trials first reported feasibility and preliminary signals of benefit. The biological rationale involves the proposed capacity of MSCs to modulate inflammatory responses, reduce myocardial fibrosis, and potentially support cardiomyocyte survival in the peri-infarct zone. Delivery by intracoronary infusion during or shortly after percutaneous coronary intervention represents one of several administration strategies that have been evaluated.

However, the evidence base for MSC-based cardiac therapies has been repeatedly contested. Several trials have been retracted or subjected to post-publication scrutiny on grounds of data irregularities, including work attributed to Piero Anversa and colleagues at Harvard-affiliated institutions, which resulted in multiple retractions and institutional investigations. The current retraction of PREVENT-TAHA8 extends this pattern and raises renewed concern about whether phase 3 data in this specific therapeutic area can be accepted at face value without independent verification.

From a clinical standpoint, heart failure following acute myocardial infarction carries a prognosis that remains poor despite advances in reperfusion therapy, neurohormonal blockade, and device-based interventions. Left ventricular ejection fraction (LVEF) at 90 days post-infarction and major adverse cardiovascular events (MACE) at 12 months are among the endpoints typically examined in this population. The retracted trial reportedly assessed a clinical strategy with substantial therapeutic implications. If the trial had been reliable and its reported outcomes accurate, it would have provided phase 3 evidence to support a novel cellular intervention in a population with limited alternatives. Its retraction eliminates that evidentiary contribution and leaves the field without the data it purported to provide.

Implications for Systematic Reviews and Clinical Guidelines

The retraction of a phase 3 trial carries downstream consequences that extend beyond the original publication. Any meta-analysis or systematic review that incorporated data from PREVENT-TAHA8 must now be reconsidered. Pooled estimates of treatment effect that included this trial’s results may have been biased by the inclusion of unreliable data, potentially influencing effect size estimates in either direction depending on the magnitude and direction of any data manipulation.

Clinical guidelines or protocol committees that referenced PREVENT-TAHA8 in support of recommendations or investigational frameworks for MSC therapy in post-infarction heart failure face a similar obligation to reassess. In the current environment of evidence-based guideline development, the removal of a phase 3 trial from the evidentiary record is not a trivial adjustment. It may alter the classification of a given intervention from one with robust trial support to one for which the available evidence is considered insufficient or of low quality.

Broader Observations on Research Integrity in Regenerative Cardiology

The PREVENT-TAHA8 retraction provides an opportunity to examine the structural conditions that permit trial irregularities to persist through peer review and into publication. Several features of the irregularities identified, including participant recruitment outside criteria, enrollment count discrepancies, and improbable data patterns, are detectable through systematic data auditing. Yet peer review, as traditionally conducted, does not routinely include access to raw trial data. Reviewers evaluate the manuscript as submitted, without the ability to interrogate underlying datasets for anomalies.

Statistical methods for detecting data fabrication, including digit frequency analysis, Benford’s law applications, and intra-cluster correlation screening, are increasingly available and have been applied productively in post-publication audits. Several journals have begun requiring data sharing as a condition of publication, a policy that would facilitate the kind of scrutiny that appears to have been necessary in the present case. The BMJ itself maintains data availability requirements, and the post-publication detection of these irregularities underscores the value of those mechanisms when enforced.

For the Pacific Islander and Native Hawaiian populations served by institutions contributing to Hawaii Medical Journal, the relevance of this retraction is not abstract. Acute myocardial infarction and its sequelae, including post-infarction heart failure, affect these communities at rates influenced by the elevated prevalence of type 2 diabetes mellitus, hypertension, and obesity-related cardiovascular risk factors in the region. Any phase 3 trial offering a novel therapeutic strategy for this indication warrants scrutiny from the perspective of regional applicability, including whether Pacific Islander participants were represented in trial enrollment, and whether subgroup analyses were conducted or reported for this population.

PREVENT-TAHA8 no longer provides any such evidence. Its retraction removes it entirely from the literature that informs therapeutic decision-making, and the field must proceed without the conclusions it claimed to establish.

Clinicians and researchers who previously encountered or cited this work are advised to consult the retraction notice published in the BMJ and to review any derivative materials, including review articles, guidelines documents, or institutional protocols, that may have relied upon its findings.