Concerns over data in key dabigatran trialBMJ 2014; 349 doi: https://doi.org/10.1136/bmj.g4747 (Published 23 July 2014) Cite this as: BMJ 2014;349:g4747
All rapid responses
Concerns expressed by the BMJ investigation about the safety and efficacy of dabigatran in patients with atrial fibrillation are legitimate and well-founded.(1)
Boehringer Ingelheim (BI) posted a response the same day these articles were published (July 23, 2014): "Boehringer Ingelheim (BI) wants to set the record straight following misleading statements that the British Medical Journal (BMJ) published today regarding Pradaxa® (dabigatran etexilate mesylate). We are concerned that this publication may alarm patients and prompt them to stop taking PRADAXA, thereby increasing their risk of stroke."(2) Approximately one month later (August 25, 2014), BI also announced their participation at the Congress of the European Society of Cardiology (30th Aug to – 3rd Sept, 2014. Barcelona, Spain) presenting new data of dabigatran pharmacological effects, a media briefing called "Setting the record straight, Pradaxa® protection today and tomorrow" was scheduled during this event.(3)
BI insist that RE-LY demonstrated that dabigatran etexilate 150 mg (no need for coagulation monitoring and without dose adjustment) bid was a superior to warfarin (monitored and dose-adjusted) in terms of their efficacy & safety outcomes in patients with atrial fibrillation.(4) Despite the fact that two corrections have been published by the NEJM, Connolly and colleagues wrote in a BMJ letter “The fact that the results of RE-LY do not change despite multiple and repeated checks of the data, attest to the robustness of the RE-LY findings.”(5)
BI responses are characterized by scarce numbers among lots of words, adding more evidence that the company still withholding key data of dabigatran studies. In addition, ClinicalTrials.gov records of the RE-LY trial (NCT00262600) and their extension observational study RELY-ABLE (NCT00808067) adds more evidence supporting the findings by D. Cohen published in the dabigatran investigation articles.(1,6,7)
1. The RE-LY study ranks 1st by the number of modifications in the ClinicalTrials.gov record among the top 10 largest US phase 3 trials completed before Jan 1st, 2011:
49 changes (1st: 8/12/2005; 49th: 8/5/2014).(8) The observational extension of
RE-LY (RELY-ABLE) was modified 62 times (1st: 16/12/2008; 62th 8/6/2014).
2. Number of changes after the last patient follow-up:
- RE-LY = 16 (15/3/2009 - 8/5/2014)
- RELY-ABLE = 11 (9/1/2013 - 8/6/2014)
3. The name of RE-LY was modified to “Premenstrual Syndrome (PMS) Assessing the Long-Term Efficacy and Safety of Nevirapine Therapy (Combined With Other ARV Drugs) in HIV-1 Positive Patients in the Daily Clinical Practice in Poland” (changed: Oct 6, 2008 - renamed: Oct 12, 2008).
4. Location of patient recruitment centers:
- Number of times locations were modified in the RE-LY ClinicalTrials.gov registry = 35
- Joué-lès-Tours in France (population about 40000 people) ranks as the 1st by the number of patient recruitment centers (n=18) among RE-LY locations distributed across 43 countries (according to the ClinicalTrials.gov registry) or 44 countries (Connolly and colleagues, NEJM 2009). A close city to Joué-lès-Tours (Angers - population about 300000 people) ranks 3rd by the number of locations (n=11) and only one patient recruitment center was located in Paris. France also ranks 1st in Europe by number of RE-LY locations (ClinicalTrials.gov) but the country is absent in the EudraCT registry.
- Israel is not a location of patient recruitment in the RE-LY registry but appears in the registry of their extension study RELY-ABLE (26 patient recruitment centers) and the NEJM 2009 publication (24 investigators recruited at least 12 patients). There are many other inconsistencies in the ClinicalTrials.gov locations of RE-LY and RELY-ABLE (Figure 1.).
RE-LY primary outcome in EudraCT and ClinicalTrials.gov are not the same.
Results were modified in the ClinicalTrials.gov registry of RE-LY and RELY-ABLE.
Results of RE-LY published by peer-review medical journals are not the same safety and efficacy outcomes disclosed to ClinicalTrials.gov.
Primary and secondary outcomes of the RE-LY study were modified (e.g., silent myocardial infarction was not a predefined outcome).
6. Limitations in the external validity of the sample of patients (n = 18113) recruited by the RE-LY study:
- Ethnic backgrounds not sufficiently represented: black ethnicity 1% (n = 176); hispanic or latinos 7% (n = 1244)
- Gender (female) 36% (n = 6598)
- External validity limitations are also evident in the RE-LY genetic subanalysis
7. Patient flow required by the CONSORT statement is missing in RE-LY papers
8. Criteria of patient to select patients in the multiple RE-LY subanalysis are not described in the methods of these articles.
1. The BMJ dabigatran investigation. Publication date: July 23, 2014
URL: http://www.bmj.com/investigation/dabigatran (access date: October 15, 2014).
2. Benefits and Safety of Pradaxa® (dabigatran etexilate mesylate) Repeatedly Confirmed. Publication date: July 23, 2014
URL:http://us.boehringer-ingelheim.com/news_events/press_releases/press_rele... (access date: October 15, 2014).
3. New Pradaxa® data to be presented at the ESC Congress 2014
URL:http://www.boehringer-ingelheim.com/news/news_releases/press_releases/20... (access date: October 15, 2014).
4. Dabigatran and the monitoring of plasma levels in patients http://www.bmj.com/content/349/bmj.g4670/rr/762688 (access date: October 15, 2014).
5. Efficacy, safety and optimal dosing of dabigatran – a commentary on the BMJ articles on dabigatran. URL: http://www.bmj.com/content/349/bmj.g4517/rr/768447 (access date: October 16, 2014).
6. RE-LY: "Randomized Evaluation of Long Term Anticoagulant Therapy With Dabigatran Etexilate."
- History of changes: http://clinicaltrials.gov/archive/NCT00262600
7. RELY-ABLE: "Long Term Multi-center Extension of Dabigatran Treatment in Patients With Atrial Fibrillation Who Completed RE-LY Trial."
- History of changes: http://clinicaltrials.gov/archive/NCT00808067
8. Ramirez, Jorge H (2014): Top 10 US phase 3 trials by sample size: how many times changes were made?. figshare. http://dx.doi.org/10.6084/m9.figshare.1122535
Competing interests: I have requested the retraction of four publications of the RE-LY study to three different peer-review medical journals, more information is available in the reference below. Ramirez, Jorge H (2014): Requested (Jul 29, 2014) & Retracted by the author (Aug 23, 2014): "Conelly S, et al. Dabigatran versus Warfarin in Patients with Atrial Fibrillation. N Engl J Med 2009; 361:1139-1151"] - Question Thread Open. figshare. http://dx.doi.org/10.6084/m9.figshare.1144305