Real-world clinical evidence on rivaroxaban, dabigatran, and apixaban compared with -vitamin K antagonists in patients with non-valvular atrial fibrillation: a systematic literature review

KEYWORDS:

• Anticoagulation
• direct oral anticoagulants
• non-valvular atrial fibrillation
• real-world evidence
• stroke prevention

We read with interest the recent review article by Briere JB et al. published in Expert Review of Pharmacoeconomics & Outcomes Research (31 August 2018 [Epub ahead of print]). We welcome the useful synthesis of the increasing literature on real-world clinical evidence, especially given the debate and discussion on how ‘real-world’ evidence compares to clinical trial cohorts, which have been described as the ‘unreal world’ given the specific inclusion and exclusion criteria of trial cohorts[1]. This is particularly relevant given the central role of stroke prevention in the management of atrial fibrillation[2].

Nevertheless, we have observed some major errors in the reporting of results from the reviewed literature that in turn led to mistakes in some of the conclusions. In particular, we found four outcomes reported in Table 4 that differed from what was actually in the cited publications, some of which were abstracts coauthored by us [3,4].

Table 4 reports that Deitelzweig et al. 2016 [3] (ref 66) found the risk of major bleeding to be significantly lower with VKAs compared to apixaban; however, that publication [3] actually reported the opposite: patients treated with warfarin had significantly higher risk of major bleeding. In addition, Table 4 reports that the risk of any bleeding was significantly lower with VKAs compared to apixaban in Lin et al. 2015 [4] (ref 58); yet, the publication [4] states that warfarin was associated with a significantly higher risk of bleeding compared to apixaban. The study by Yao et al. 2016 [5] (ref 18) presents data on the composite outcome of stroke/systemic embolism, which is defined to include both ischemic and hemorrhagic stroke. This outcome differs from the outcome presented in Table 4 as a composite of ischemic stroke/systemic embolism. Lastly, the table in the manuscript shows that only one study reports non-persistence and within that study, no significant differences between apixaban and warfarin treatments were observed. However, the study by Lamberts et al. 2016 [6] (ref 65, which was one of the apixaban studies identified by the authors) states that apixaban has a significantly higher persistence versus warfarin.

Based on the observed inconsistencies noted above, we respectfully suggest that the following statements in the Results and Conclusion sections require correction:

On p. 12: ‘Apixaban was strongly linked to a lower risk of GI bleeding and major bleeding compared with VKAs; however, one study reported VKAs were associated with a lower risk of major bleeding events compared with apixaban. Two studies evaluated the risk of any bleeding; one found apixaban to be associated with lower risk, while the other found VKAs to be associated with lower risk. While several studies evaluated the difference between apixaban and VKAs on ischemic stroke and non-persistence, significant differences between the two treatments were not reported.’ Instead this should state that apixaban was associated with a lower risk of GI bleeding and major bleeding compared with VKAs, as consistently shown across studies. Two studies evaluated the risk of any bleeding, both of which suggest that apixaban is associated with a lower risk of any bleeding compared to VKA. While several studies evaluated the difference between apixaban and VKAs on ischemic stroke, significant differences between treatments were not observed. One study reported non-persistence and the study found that apixaban had a significantly higher persistence compared to VKA.

On p. 13: ‘Results of any bleeding events were mixed; overall, dabigatran was associated with a lower risk of any bleeding compared with VKAs, whereas apixaban was comparable with VKAs.’ Instead this should state that dabigatran and apixaban were associated with a lower risk of any bleeding compared with VKAs.

On p. 14: ‘Studies evaluating non-persistence rates for DOACs versus VKAs tended to favor DOACs, in particular rivaroxaban, for treatment continuation.’ Instead this should state that studies evaluating non-persistence rates for DOACs versus VKAs tended to favor DOACs, in particular rivaroxaban and apixaban, for treatment continuation.

Finally, Table 5 should also be revised as it is based partly on the errors in Table 4.

We appreciate your corrections, as the medical community should have access to the most updated and accurate real-world clinical evidence.

Declaration of interest

S Deitelzweig is a consultant for Bristol-Myers Squibb Company/Pfizer Inc., Daiichi-Sankyo, Portola, and Boehringer Ingelheim; he also has been on the speakers’ bureau for Bristol-Myers Squibb Company/Pfizer Inc., and Boehringer Ingelheim. GY Lip is a consultant for Bayer/Janssen, BMS/Pfizer, Medtronic, Boehringer Ingelheim, Novartis, Verseon and Daiichi-Sankyo. Speaker for Bayer, BMS/Pfizer, Medtronic, Boehringer Ingelheim, and Daiichi-Sankyo. No fees are directly received personally. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgments

Allie Cichewicz (Evidera) for providing assistance regarding data collection, Evelien Bergrath (Evidera) for providing assistance regarding data collection and development of the letter, and Amy Earley (Evidera) for providing assistance regarding development of the letter.

Additional information

Funding

The letter to the editor was sponsored by Pfizer Inc. and Bristol-Myers Squibb Company.