European Primary Care Cardiovascular Society

BMJ criticises safety and use of dabigatran

News - Aug. 20, 2014

-Also read the view of dr. Michiel Coppens (internist, AMC Amsterdam, NL) about this –
In the British Medical Journal (BMJ) of July 23, critical articles were published about the presentation of data from the most important study (RE-LY) on the efficacy and safety of the new oral anticoagulans (NOAC) dabigatran [1-4]. The discussion concerns one of the most important arguments to use dabigatran for stroke prevention in non-valvular atrial fibrillation (NVAF), namely that time-consuming monitoring of anticoagulation activity or plasma levels of the agent are not needed, like with vitamine K antagonists such as warfarin. It is also suggested that the risk of major bleedings as published is an underestimate of the true risk.
The BMJ examined documents of the producer of dabigatran, Boehringer Ingelheim (BI), which came available during an American trial against the company, by requests based on the right of information. Based on these data, Deborah Cohen, investigations editor at BMJ, raises questions about the evidence on which the claim was based that monitoring anticoagulation activity is not needed [1,2].
The released internal documents reveal that BI had conducted extensive analyses that show that the risk of major bleedings could be reduced by 30-40% as compared with well-regulated warfarin, if dabigatran plasma levels were measured. Moreover, plasma levels were identified at which dose would need to be adjusted. These findings had not been communicated to physicians and regulating bodies, because the analyses did not give trustworthy predictions about the patient outcomes, according to the company.
Reilly and colleagues published a RE-LY subtrial at the end of last year, in which they showed that part of the patients can benefit from adjusting the dose of dabigatran. This publication does not mention the optimal plasma concentration range of the agent, which was mentioned in an earlier version of the article, as seen by BMJ [1].
In addition, Cohen states that there are indications that dabigatran might give a higher bleeding risk than is reported in the publication of the clinical trial, based on which it received regulatory approval. When the research data were first sent to the American Food and Drug Administration (FDA) for approval of dabigatran, the FDA asked for revision of the data due to too high a number of irregularities in the data set. The revision of the data yielded more events, in both treatment groups, but did not change the results importantly. There was, however, discussion about the updated results, predominantly about the number of (silent and clinical) myocardial infarctions. That RE-LY had an open label study design, may not have helped, although the FDA had requested for the clinical events to be adjudicated in a blinded fashion.
Cohen raises questions as to whether this protocol has always been followed, and whether major bleedings have always reached the statistics as such. In answer to BMJ’s question whether all missed events have now been identified, a spokesman of BI said that they “We are confident in RE-LY’s conclusions. While it is still possible that some events have not been identified, based on our reviews of RE-LY and the post-marketing analyses of our medicine, we are confident that any such events would be distributed evenly among treatment groups and would not affect RE-LY’s findings.”
The publications also points a critical finger towards the FDA. Because dabigatran was the first NOAC, it followed a ‘fast track’ procedure, to stimulate innovation. That is why only one large phase III trial was needed for approval, instead of two. Thus, an accelerated FDA review contributed to a less robust evaluation of the benefits and risks. Costs and benefits for society of such an accelerated process should be better weighted, according to Charlton and Redberg [3].
The FDA approved use of fixed-dose dabigatran without clinical monitoring at the time, despite the fact that one of the advisors of the FDA had remarked that plasma levels of the agent can vary considerably; very much for an agent that will be used without monitoring. The European Medicines Agency (EMA) followed  a year later with approval, but mentioned the necessity of education of the physician, monitoring renal function, publication of therapeutic medicine levels and the availability of a test to measure anticoagulation [3]. EMA expressed its concern earlier about the necessity to monitor plasma levels to reduce the risk of bleedings. In a report about the agent EMA mentions plasma concentrations that should be regarded as upper and lower borders to prevent clinical events. Moreover, it was organised that an accurate, validated test would become available: Hemoclot direct thrombin inhibitor assay. The necessity of monitoring was, however, not included when the agent was approved [1].
In response to the high number of bleedings in post-marketing reports from the USA and Japan, EMA renewed the conversation with BI about the necessity of monitoring and dosis adjustment of dabigatran. An advisory committee of EMA voted against monitoring because the desired plasma concentrations and the therapeutic range were not known. The final recommendation just emphasised the importance of monitoring renal function and patient characteristics before and during treatment, and in some patients to adjust the dose based on this. Cohen writes that not all information from the unpublished version of the article on plasma concentrations was presented to EMA. When confronted with the fact that the data in the article by Reilly had been brought into the public domain earlier, Reilly and colleagues point out that RE-LY is the only NOAC study that has published extensive data about plasma concentrations. That raises the question as to what extent other NOACs also show variable plasma levels. Independent and systematic research will have to shed more light on the balance between prevention of bleeding and death, and convenience [4].
The safety of dabigatran in NVAF can be improved considerably, according to Moore and colleagues. The producer, the FDA and EMA should reach consensus about a therapeutic range and should recommend initial dose adjustments based on plasma concentration measurements. The regulating bodies should recommend plasma measurements in all new patients, and the recommendation ‘does not in general require routine anticoagulant monitoring’ should be removed. If this is true for dabigatran, then possibly also for other NOACs [4].
BI wrote a press release, in response to ‘misleading statements in BMJ (...), as we are concerned that the reporting might put patients at risk of stopping their important stroke preventing medication.’ [5] BI refers to a big real-world analysis of the FDA of May 13 2014, in which the FDA confirms the positive efficacy-safety profile of dabigatran, based on a study in 134000 patients of 65 years and older, who were not monitored. They declare to have provided the regulatory bodies with the complete data set and analyses of the clinical evidence of efficacy and safety. About the accusation that BI would have withheld analyses, the press release present the following facts: ‘RE-LY showed that dabigatran improves stroke prevention versus standard of care. BI scientists performed preliminary, exploratory simulations with mathematical models to understand whether dose adjustments based on plasma concentrations might further improve the efficacy and safety profile of dabigatran. The initial hypothesis from this mathematical model could not be supported when applied to the actual clinical data from the RE-LY population. Therefore, they were not provided to regulators. The totality of scientific evidence does not support dosing decisions for Pradaxa® based solely on blood levels. The research shows that individual patient characteristics, such as age, kidney function and certain medications, are critical factors in contributing to the risk of bleeding.’
"BI made a robust effort to try and find ways to utilize plasma levels to further improve the efficacy and safety profile of Pradaxa® and it is wrong to suggest otherwise. The truth is, if these approaches could have been developed, we would have used them to the benefit of patients."
You may also want to read the view of dr. Michiel Coppens (vascular medicine, AMC Amsterdam The Netherlands) on this topic.

News • 20-8-2014

A more nuanced view from the clinic on the critical BMJ publications about dabigatran

Dr. Michiel Coppens (vascular medicine, Academic Medical Center, Amsterdam, NL) shares his views on the critical publications in BMJ about the safety of using dabigatran.


1. Cohen D. Dabigatran: how the drug company withheld important analyses. BMJ 2014;349:g4670
2. Cohen D. Concerns over data in key dabigatran trial. BMJ 2014;349:g4747
3. Charlton B, Redberg R. The trouble with dabigatran. BMJ 2014;349:g4681
4. Moore TJ, Cohen MR, Mattison DR. Dabigatran, bleeding, and the regulators. BMJ 2014;349:g4517
5. Persbericht Boehringer Ingelheim 23 juli 2014.

Share this page with your colleagues and friends: