As part of its ongoing investigation The BMJ has claimed the manufacturer of rivaroxaban (Xarelto, Janssen) withheld from regulators information about a faulty device used in the pivotal ROCKET AF trial that could have skewed data in favour of the drug.
In its investigation published on Thursday the journal also claims that patients involved in the trial may have been put at undue risk of harm because of the faulty device.
Results from the ROCKET AF trial were published in the New England Journal of Medicine in 2011. It compared rivaroxaban with warfarin for preventing strokes in patients with non-valvular atrial fibrillation and found that the drug was non-inferior to warfarin.
But in February 2016, The BMJ raised concerns about the validity of the trial results after the FDA recalled the device the study used to measure INR values.
The point of care device – INRatio, initially marketed by HemoSense and later by Alere – was used to measure international normalised ratio (INR) values in the 7133 participants in the warfarin arm of the study. However, the FDA recalled the device in December 2014 because INR results it generated could be “clinically significantly lower” than those found by a laboratory method.
Now The BMJ’s Associate Editor, Dr Deborah Cohen, who led the journal’s investigation, claims that Janssen executives knew and withheld information about the faulty device well before it was recalled.
According to Dr Cohen soon after the trial commenced in 2007 investigators wrote to Janssen executives questioning the accuracy and reliability of INRatio. The drug manufacturer responded to investigator concerns by initiating a safety program, dubbed the Covance recheck program, in 2008 to check the reliability of blood test readings against laboratory results.
The BMJ investigation revealed that study investigators submitted 149 samples to the recheck program (rivaroxaban 78; warfarin 71), and in legal documents obtained by The BMJ, Janssen stated that there were “16 instances where the value from the point of care device and lab were recorded as inconsistent.”
“Since the rivaroxaban arm used a sham device, this means that 23% of warfarin participants in the program had mismatching laboratory and point-of-care INR values,” Dr Cohen said.
What’s more the mean time in therapeutic range was 55% – lower than in any of the phase III trials of other direct oral anticoagulants, The BMJ reports.
The journal alleges that Janssen failed to share these data with the trial’s data safety monitoring board, Bayer (co-developer of rivaroxaban), and the FDA prior to the drug’s approval and “despite the safety of trial participants potentially being compromised,” the journal said in its investigation.
During a review by the European Medicines Agency, published in February 2016, its analysis of blood samples taken at weeks 12 and 24 revealed that in about 35% of cases, there were “discrepancies of potential clinical relevance”.
Further analyses found that the larger the difference between the INRatio and laboratory readings, the higher the rate of major bleeding.
But in a statement to the limbic Bayer maintains that both [Bayer and Janssen] have properly disclosed safety data to regulators and deny the central premise of BMJ’s report.
“Several recent reassessments of the trial data by Bayer and Janssen as well as by the ROCKET AF Executive Committee plus two reviews by the European Medicines Agency (EMA) – including one concluded in July 2016 that assessed the Covance data at the centre of BMJ’s story – confirm that the benefit-risk profile of Xarelto remains unchanged and favourable,” a spokesperson for Bayer said.
Chris Ward, Associate Professor and clinical haematologist at Royal North Shore Hospital in Sydney has been an investigator on the RE-LY trial for dabigatran (Pradaxa, Boehringer) and said he doesn’t buy into The BMJs argument that rivaroxaban is unsafe and that the registration trial is fatally flawed.
“Point of care devices for INR are all less accurate than lab tests – the ones used in trials, and mandated by the regulators, are older models and likely to be even less reliable … No one who has used these POC devices would be surprised at that level of discrepancy especially with INRs over 3 or over 4,” he told the limbic.
I don’t think this device is uniquely bad … I think they’re all pretty similar. In fact, not only was the device approved at the time of the study it was mandated by regulators to use that type of device … it’s the instance of the regulators to use machines which are inherently unreliable.”
According to Dr Ward, while there might indeed be a case for Janssen and Bayer to answer about conducting an extra INR monitoring process that wasn’t part of the trial submission, that doesn’t impact on the outcomes of the rivaroxaban arm in ROCKET AF.
“Even if the warfarin arm were managed suboptimally, we can compare their results with all the other SPAF [stroke prevention in atrial fibrillation] trials, and warfarin outcomes were fairly similar.”
Dr Ward said that there is now real-world data on large numbers of rivaroxaban users with very similar safety and efficacy results published in trials.
“If you’re not seeing a safety signal in other studies where exactly the same drug is being used then, for me, it’s unbalanced to say that the drug should never have been registered and for safety to be called into question.”
The EMA and FDA say they have not changed their recommendations on the use of rivaroxaban however the FDA review is ongoing.