Rivaroxaban for the prevention of outcomes in patients with atrial fibrillation in clinical practice: an indirect comparison of national and international registries

EMIR was a noninterventional and observational study that included patients ≥18 years old with AF, taking rivaroxaban according to clinical practice ≥6 months before being enrolled. Patients with a valvular prosthesis or rheumatic mitral stenosis were excluded from the study. Patients were recruited from 79 Spanish centers (hospitals and private clinics) and followed up during 2.5 years with four visits (baseline, 12 months, 24 months and study end) that should coincide with any of the patients' routine visits. In the EMIR study, MACE were defined as a combination of nonfatal myocardial infarction, revascularization and cardiovascular death (death for coronary events, progressive heart failure death and sudden cardiac death) [9].

ROCKET-AF was a double-blind clinical trial in which 14,264 patients with nonvalvular AF who were at increased risk for stroke were randomized to receive either rivaroxaban or dose-adjusted warfarin. The median duration of treatment exposure was 590 days, with a median follow-up period of 707 days [7].

In addition to the EMIR study, eight other Spanish observational studies (one with a prospective design and seven with a retrospective design) of patients with AF taking rivaroxaban for the prevention of stroke or systemic embolism were analyzed. Brun et al. performed a retrospective study of AF patients who had started rivaroxaban therapy between May 2013 and February 2017 in the cardiology department of two hospitals from Huesca. Mean duration of treatment was 30.5 months [10]. Cerdá et al. developed an observational and prospective study of all patients with AF, with or without prior stroke, who had attended the Oral Anticoagulant Treatment Unit of the Hemostasis and Thrombosis Department of the University Hospital Vall d’ Hebron from Barcelona and who had started treatment with any DOAC between January 2015 and September 2017. Mean follow-up was 1.7 years [11]. Gavin et al. designed a retrospective study that included all patients with AF who had started treatment with rivaroxaban for the prevention of thromboembolic events in the hematology unit of a tertiary hospital between February 2012 and June 2016 (mean follow-up of 16.5 months) [12]. Pérez Cabeza et al. analyzed, in a retrospective study, the drug persistence and outcomes after 2 years of follow-up in a cohort of patients with AF in whom rivaroxaban had been prescribed during the first quarter of 2014 by cardiologists or internists, in the sanitary area of Costa del Sol (Malaga, Spain) [13]. Muñiz et al. developed a retrospective study that included all patients with AF who had started anticoagulant treatment with rivaroxaban for the prevention of stroke or systemic embolism between July 2013 and May 2017 (mean follow-up of 26.6 months) in a regional hospital in Asturias (Hospital San Agustin, Aviles, Spain) [14]. Marti et al. retrospectively analyzed patients with AF that had started treatment with rivaroxaban for the prevention of stroke from July 2012 to December 2015 (mean duration of treatment of 29.7 months) in a healthcare area of 330,000 inhabitants of Valencia (Spain) [15]. Pimentel et al. performed a retrospective study of AF patients treated with rivaroxaban in the cardiology department of the tertiary hospital San Juan de la Cruz at Ubeda (Jaen, Spain), between November 2012 and January 2017 (median follow-up of 24 months) [16]. Caro et al. developed a retrospective study that included all AF patients who had started treatment with rivaroxaban for the prevention of stroke or systemic embolism between December 2012 and December 2015 (mean follow-up period of 22.7 months) from two health areas belonging to two tertiary hospitals (Hospital Vega Baja [(]Alicante, Spain], and University Hospital Clinico Virgen de la Arrixaca [El Palmar, Spain]) [17].

Furthermore, one international study of patients taking rivaroxaban for the prevention of thromboembolic events was included. Thus, the XANTUS study was the first international, real-world, prospective, observational study of patients treated with rivaroxaban for stroke prevention in AF. This study included patients with AF newly started on rivaroxaban who were followed up at ∼3-month intervals for 1 year, or for at least 30 days after permanent discontinuation (mean treatment duration of 329 days) [21].

Additionally, in order to put into context the results of rivaroxaban compared with those of other DOACs, three prospective registries of patients with AF taking any DOAC were also included for analysis (one developed in Spain and two international studies performed in different countries across Europe). The FANTASIIA registry included consecutive AF outpatients on anticoagulant therapy, either with vitamin K antagonists or DOACs, from 50 Spanish centers between 1 June 2013 and 15 October 2014 (median follow-up of 32.4 months). For this analysis, the 533 patients receiving DOACs were included [18–20]. GLORIA-AF was a large, prospective, global registry program run in three phases, assessing the long-term safety and effectiveness of dabigatran in patients with newly diagnosed AF in clinical practice. For this analysis, patients from phase II, enrolled between 2011 and 2014 with a mean follow-up of 18.0 months, were included [22–24]. ETNA-AF-Europe was a prospective, multicenter, postauthorization, observational study enrolling patients treated with edoxaban in ten European countries. Data from 852 sites that completed the 1-year follow-up (mean and median follow-up of 352 and 366 days, respectively) were analyzed [25].

With the exception of the ROCKET-AF trial, all other studies had an observational design. All studies were approved by the appropriate clinical research ethics committees. In each study, data were collected from the medical history and physician interview. Biodemographic data, comorbidities and thromboembolic and bleeding risk were recorded if available. Information regarding thromboembolic and bleeding events, as well as death, was recorded from all studies if available and expressed as annual rates (events per 100 patient-years).

The EMIR study showed in a wide sample of real-life patients with AF that rivaroxaban was prescribed in elderly patients with a high thromboembolic risk. Despite that, rates of thromboembolic, cardiovascular and major bleeding events remained low. These numbers were in the range of or were even better than those reported in other national or international registries.

The ROCKET-AF trial was the study that included those patients with the highest thromboembolic risk and the greatest number of comorbidities when compared with the other pivotal clinical trials [7,26–28]. As a result, it was necessary to ascertain whether the results of the ROCKET-AF trial could be applied to real-life patients [8,29]. On the other hand, the clinical profile of patients for which rivaroxaban is prescribed has changed over time. Thus, as experience with rivaroxaban has increased, its prescription has been extended to the whole AF population, including those patients with a more challenging clinical profile [16,30]. In this context, the present study provided relevant and current information. In the EMIR study, patients were slightly older than in the ROCKET-AF trial, and despite the fact that they had a high thromboembolic risk and a great number of comorbidities, these were lower than in the ROCKET-AF trial [7,9]. Compared with other Spanish registries of patients taking rivaroxaban, the number of patients included in the EMIR study was markedly higher. Interestingly, overall, these studies showed that in Spain, rivaroxaban had been prescribed in elderly patients, with a high thromboembolic risk and a great number of conditions, in line with the data reported in the EMIR study [9–17]. Additionally, the XANTUS study exhibited a clinical profile similar to that of the EMIR study [21]. In summary, all these data are consistent and indicate that rivaroxaban is prescribed for the whole spectrum of the AF population, including patients with more complex disease.

In the EMIR study [9], the rate of stroke was low (0.57 events per 100 patient-years). This was much lower than that reported in the ROCKET-AF trial [7] but similar to that of the XANTUS study and other Spanish registries [10–17,21]. With regard to major bleeding, the EMIR study showed lower rates than those observed in the ROCKET-AF trial and some national and international registries of patients treated with rivaroxaban, such as the XANTUS study [7,9–17,21]. In addition, rates of stroke and bleeding events in the EMIR study were similar to those reported with other DOACs [7,18–20,22–25]. Although disparities in the clinical profiles of patients among the registries could explain, at least in part, some differences in the numbers, it should be emphasized that the EMIR study is a contemporary registry, and consequently, the more experience with the use of rivaroxaban and in the management of AF could have had some impact on the results. In fact, in the EMIR study, around 17% of patients were taking an inappropriate dose of rivaroxaban (vs 19% in the XANTUS study); underdosing was observed in 9.7% and 8.4% of patients, respectively [9,21]. In the Spanish observational studies, underdosing with rivaroxaban ranged from 3.0% to 31.4% [10–18]. These figures could be higher with other DOACs, likely related to the simplicity of dose adjustment with rivaroxaban, only depending on renal function [9,31,32]. Thus, in the FANTASIIA study, 32% of patients did not receive the proper dose [18]. In another study, underdosing was observed in one out of five patients taking DOACs [32]. This is very relevant, as underdosing is associated with an increased risk of stroke, without reducing the risk of bleeding [33,34]. As a result, prescribing the appropriate dose of DOACs according to the summary of product characteristics should be encouraged in order to provide the highest benefit [2].

The risk of MACE was specifically analyzed in the EMIR study [9]. Remarkably, despite anticoagulation, AF patients have a high risk for myocardial infarction or cardiovascular death [2,3]. In fact, the European guidelines recommend a comprehensive approach in the management of patients with AF that includes the prevention of not only thromboembolic events but also cardiovascular disease [2]. Unfortunately, this has not been fully analyzed in real-life studies. For example, only Gavin et al. and the XANTUS study specifically investigated this point, but not other registries, including those with other DOACs [12,21]. On the other hand, it is difficult to compare the rates of MACE among these studies, as the definition of MACE differed among them. In any case, the rate of MACE of the EMIR study was markedly lower than that of the ROCKET-AF trial and similar to those observed in the real-life studies with rivaroxaban [7,9,12,21]. As no head-to-head clinical trials between DOACs are available, observational studies can provide relevant information on the comparative effectiveness and safety of DOACs. However, these studies cannot be considered individually, since differences in the data can be the result of disparities in the clinical profile and methods [11,35]. By contrast, meta-analysis can show more accurate information. In this context, previous meta-analyses have reported that the use of rivaroxaban could be associated with a lower risk of ischemic events in the AF population [36–38].

This narrative review has some limitations. As no control groups are available in observational studies, no definite conclusions can be obtained from these studies (this occurred in national and international registries, including XANTUS, GLORIA-AF and ETNA-AF). However, these studies best reflect clinical practice, as no specific diagnostic or therapeutic intervention is performed for being included. On the other hand, data were taken from different studies, with particular inclusion criteria, follow-up and outcomes definitions. Therefore, only an indirect comparison can be performed between the studies and no more than hypothesis can be suggested. However, the high number of patients included throughout Spain and the robustness of the international registries may reduce these potential biases.

National and international registries confirm that in clinical practice, despite the fact that rivaroxaban is prescribed in elderly patients with a high thromboembolic risk and many comorbidities, rates of outcomes such as stroke, MACE, myocardial infarction, death and major bleeding are low. Therefore, rivaroxaban should be considered as a first-line therapy in the comprehensive management of patients with AF, to reduce the risk of not only thromboembolic complications but also cardiovascular disease. Using the appropriate dose of DOACs is highly recommended to obtain the maximum benefit.

Although the information provided by clinical trials is important, real-life studies are necessary to determine whether the results can be extended to the whole population [8]. Additionally, as observational studies cannot control all potential bias, it is important to compare data between different real-world evidence studies. In this context, as the information provided was consistent between different studies, we can conclude that rivaroxaban represents a good choice for anticoagulation in AF patients, even in elderly patients with many comorbidities.

F Marín has received consultancy/lecturing fees from Bayer, Boehringer Ingelheim, Pfizer, Bristol Myers Squibb, Daiichi Sankyo and AFNET. M Sanmartín Fernández has received speaker and advisory fees from the following companies in the past 3 years: Bayer, Boehringer Ingelheim, BMS and Pfizer. I Lekuona has received honoraria for presentations from Bayer, Boehringer-Ingelheim, Daiichi Sankyo and Pfizer-BMS. F Arribas has received personal fees for educational activities or participation on boards from Daiichi Sankyo, Bayer, Boehringer Ingelheim and Bristol Myers Squibb. G Barón-Esquivas has received honoraria as advisor from Bayer, Daiichi-Sankyo, BMS-Pfizer and Rovi and honoraria as speaker from Boehringer-Ingelheim, Bayer, Daiichi-Sankyo, BMS and Pfizer. V Barrios has received consultancy/lecture fees from Bayer, BMS/Pfizer, Boehringer Ingelheim and Daiichi Sankyo. J Cosin-Sales has received consultancy/lecture fees from Bayer, BMS/Pfizer, Boehringer Ingelheim and Daiichi Sankyo. R Freixa-Pamias has received honoraria for presentations from Bayer, Boehringer-Ingelheim, Daiichi Sankyo and Pfizer-BMS. J Masjuan has received consultancy/lecturing fees from Bayer, Boehringer Ingelheim, Pfizer, Bristol Myers Squibb and Daiichi Sankyo. AI Pérez-Cabeza has received personal fees for educational activities or participation on boards from Daiichi Sankyo, Bayer, Boehringer Ingelheim and Bristol Myers Squibb. VR Schilling has received honoraria for presentations from Bayer, Boehringer-Ingelheim, Daiichi Sankyo and Pfizer-BMS. JMV Rodríguez has received lecturing fees from Bayer, Pfizer and Daiichi Sankyo. CR Priu is an employee of Bayer Hispania SL. MA Sánchez has received funding for consulting and conference services from Bayer, Daiichi-Sankyo and Pfizer. 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.

Writing and editorial assistance was provided by Content Ed Net (Madrid, Spain) with funding from Bayer Hispania.