BMJ investigation reveals AstraZeneca’s billion-dollar heart drug built on flawed data
A damning BMJ investigation has blown open a scandal at the heart of modern cardiology, revealing how AstraZeneca’s billion-dollar blockbuster ticagrelor may have reached patients through manipulated data, missing evidence, and phantom investigators in clinical trials that regulators failed to properly scrutinise.
A comprehensive investigation by The BMJ has uncovered substantial data integrity concerns surrounding ticagrelor (Brilinta/Brilique), raising fundamental questions about the evidence base supporting one of cardiology’s most widely prescribed antiplatelet medications. The findings, published on 20 June 2025, extend previous concerns about the landmark PLATO trial and focus on two pivotal pharmacodynamic studies that convinced clinicians of ticagrelor’s therapeutic superiority.
Critical flaws in foundational studies
The investigation examined the ONSET/OFFSET and RESPOND trials, published in Circulation, which demonstrated ticagrelor’s ability to provide faster and more consistent platelet inhibition compared to clopidogrel. These studies were instrumental in establishing clinical confidence in ticagrelor following percutaneous coronary intervention procedures.
However, The BMJ’s analysis revealed that primary endpoint results for both trials were inaccurately reported in the published papers. Most significantly, the RESPOND study failed to achieve statistical significance according to its original protocol specification (P=0.157), yet was published as statistically significant (P=0.005) following an undeclared change in the primary endpoint definition.
The investigation also identified more than 60 missing platelet machine readings from the 282 total measurements that should have been present in FDA datasets. These absent readings represented approximately 25% of total measurements from the principal investigator’s site, with the missing data showing significantly higher platelet activity levels than those included in the final analyses.
Investigator accountability concerns
A particularly troubling finding involved the study authorship and investigator participation. The BMJ discovered that Tonny Nielsen, listed as a co-author of the RESPOND study and identified as a principal investigator in Denmark, told the journal he “did not participate in the RESPOND study” – a statement corroborated by two colleagues. Conversely, Alberto Yataco operated an active Baltimore study site, enrolling 12 patients, yet never received authorship credit despite his substantive contribution.
Most investigators contacted by The BMJ, including principal investigator Paul Gurbel, were either unreachable or declined interviews. This lack of investigator engagement represents an unusual pattern for clinical research accountability.
Technical execution challenges
The studies’ technical execution raised additional concerns among platelet function experts. Alan Michelson, director emeritus of the Center for Platelet Research Studies at Boston Children’s Hospital, expressed surprise at the extensive blood sampling requirements – up to 429 mL across multiple visits for ONSET/OFFSET participants.
The multicentre design, while typically enhancing study credibility, introduced complexities specific to platelet function testing. Light transmittance aggregometry requires execution within hours of blood collection and is “prone to artefacts of various kinds,” according to Michelson. The investigation could not confirm that all sites received appropriate training for standardised testing procedures.
Data analysis irregularities
The BMJ’s reanalysis of FDA-obtained patient-level data revealed systematic problems in data handling. At three of eight active ONSET/OFFSET sites, investigators apparently failed to conduct baseline platelet aggregation tests – the study’s primary endpoint measurement. Despite missing these crucial baseline values, blood continued to be drawn from affected patients an average of 11 additional times.
The analysis also identified patients with implausibly low baseline platelet aggregation levels (under 50%) and paradoxical increases in platelet aggregation following antiplatelet treatment. These aberrant datapoints underwent unpublished adjustments that obscured their impact on primary endpoint calculations.
Regulatory oversight questioned
The findings illuminate a concerning disconnect between pharmacodynamic evidence and clinical outcomes that FDA reviewer Thomas Marciniak had previously noted. Despite ticagrelor’s demonstrated superior platelet inhibition, early clinical benefits were not observed in the PLATO trial’s percutaneous coronary intervention population.
Victor Serebruany, adjunct faculty member at Johns Hopkins University and prominent ticagrelor critic, told The BMJ that “there are episodes of skyrocketing rebound and profound platelet inhibition after ticagrelor making patients prone to thrombosis or bleeding. If doctors had known what happened in these trials, they would never have started using ticagrelor.”
Clinical practice implications
As generic versions of ticagrelor prepare for market entry in 2025, these revelations carry significant implications for clinical practice. The drug generates over one billion dollars in annual sales and features prominently in international cardiovascular treatment guidelines.
Dan Atar, editor-in-chief of Cardiology and head of research at Oslo University Hospital, emphasised the importance of transparency in data adjustments: “Whether the adjustment was reasonable or not can be debated. But either way it is something that absolutely should have been reported by the authors. Without that transparency, one cannot even evaluate its appropriateness.”
The investigation underscores broader concerns about pharmaceutical trial integrity and regulatory oversight, particularly for medications with substantial clinical and economic impact. The accumulated evidence suggests fundamental reassessment of ticagrelor’s risk-benefit profile may be warranted.
Reference
Doshi, P. (2025). Ticagrelor doubts: inaccuracies uncovered in key studies for AstraZeneca’s billion dollar drug. BMJ, 389, r1201. https://doi.org/10.1136/bmj.r1201
