Off-label Non-vitamin K Antagonist Oral Anticoagulants Does Not Decrease the Stroke Risk of Atrial Fibrillation Patients in East China

Peng Zhou Huashan Hospital Fudan University Rongchen Liu Huashan Hospital Fudan University Yangjie Yu Huashan Hospital Fudan University Wei Chen Huashan Hospital Fudan University Yunzhi Ma Huashan Hospital Fudan University Hengyang Wang Huashan Hospital Fudan University Maieryemu Waresi Huashan Hospital Fudan University Meng Wang Huashan Hospital Fudan University Xinping Luo Huashan Hospital Fudan University Jian Li (  13816066763@163.com ) Huashan Hospital Fudan University


Introduction
Atrial brillation (AF) is the most common sustained arrhythmia worldwide [1,2], while anti-coagulation therapy remains the main strategy for AF including vitamin K antagonists (VKAs) and non-vitamin K antagonist oral anticoagulants (NOACs) [3]. Some studies have reported that the e cacy of NOACs are superior than VKAs among AF patients in East Asia [4].
Because of a greater risk for anti-coagulation related intracranial hemorrhage in east Asia, there is a compromised dosage of NOACs concerning the bleeding risk for anti-coagulation therapy [5]. Patients are usually given a so called "Asian dose" of NOACs such as low-dose rivaroxaban (10-15 mg once daily) or dabigatran (110-150 mg once daily) in China [6,7]. Several studies have demonstrated that Asian dose brings much bene t for AF patients in East Asian due to the lower bleeding risk [4,8]. However, the latest European Society of Cardiology (ESC) guideline indicates that reduced dose may be only favorable for patients with renal disease for rivaroxaban and apixaban [9]. The bene t of "Asian dose" was still controversial. Here, we set out this observational study to investigate whether the "Asian dose" of NOACs really exists for stroke prevention in AF patients in China.

Subjects And Methods
Patients and study design This was a retrospective study of prospectively obtained data from patients seen between January 2015 and December 2015 at a tertiary referral center in the East of China. Patients who diagnosed as Atrial brillation (AF) with an estimated glomerular ltration rate ≥ 30 ml/min/1.73 m 2 were recruited. The diagnosis of AF was based on the ESC guideline that at least 30 seconds electrocardiogram or 24 h-Holter before or during hospitalization. Patients who uses warfarin irregularly or with INR value < 2 were excluded. Finally, a total of 342 patients were recruited in this study. Then, according to the anticoagulation therapy, patients were divided into three groups including standard anti-coagulation (SAG) group, off-label (OFL) group and non-coagulation (NCG) group. Patients in SAG group received standard dosage of anti-coagulation therapy, including rivaroxaban 20 mg once daily or dabigatran 110 mg/150 mg twice daily, or warfarin with INR 2-3. Patients in OFL group received rivaoxaban < 20 mg daily or dabigatran 110 mg/150 mg once daily. Patients without anti-coagulation therapy were classi ed into NAG group. Patients were followed up every 6 months up to 5 years. All clinical events during followup including stroke and other thromboembolism events, acute coronary syndrome, major bleeding, allcause mortality and re-hospitalization were collected.

Statistical analyses
All analyses were carried out using SPSS 24.0. Descriptive statistics were used to evaluate patient related secondary objectives. For patient characteristics, two-sided Student's t-test, Chi square test, and Fisher's exact tests were utilized as appropriate. Comorbid conditions and responsible physician type were analyzed using Chi square test, and concomitant medication use was analyzed using Fisher's exact test.
All statistical tests were two-sided and a p-value of less than or equal to 0.05 was considered statistically signi cant.

Baseline characteristics
A total of 342 patients with AF were included into this study, 296 of which were regularly followed up while 46 were missed because of incorrect contact information. There were 148 patients in NAG group, 37 patients in OFL group and 111 patients in SAG group, respectively. The baseline characteristics of these three groups were summarized in Table 1. The CHA 2 DS 2 -VASc score of the NAG group, OFL group and SAG group were 3.6 ± 1.8, 3.7 ± 1.7, 3.5 ± 1.7, separately. Compared to OFL group (71 ± 10) and SAG group (69 ± 13), the age in NAG group was higher (73 ± 12). Unexpectedly, anti-platelet therapy was more common among patients in NAG group (49% in NAG to 35% in OFL and 10% in SAG). Figure 1 shows the anti-coagulation drugs used in OFL group (Fig. 1a) and SAG group (Fig. 1b). Most of the OFL group used rivaroxaban, while warfarin remained predominant in SAG group. We collected clinical events during our follow-up up to 5 years (Fig. 2). Compared to patients in SAG group, patients in OFL group had higher risk in stroke and thromboembolism events(P = 0.000). The risk of other events including major bleeding (P = 0.597) and re-hospitalization (P = 0.213) was similar in both groups, as well as the cardiovascular mortality (P = 0.640). Minor bleeding such as gingival bleeding and skin bleeding occurred more in SAG group. All adverse events were more common in NAG group except major bleeding.
As shown in Fig. 3, there was no signi cant difference on all-cause mortality between OFL and SAG (0.14 and 0.19, P = 0.601), while these two groups were dramatically lower than NAG group (0.52, P = 0.000).
As shown in Fig. 4, the total cardiovascular events and mortality rate were not signi cant different between OFL group and SAG group (P = 0.863), while both were better than NCG group (P = 0.000).

Subgroup analysis for anti-platelet therapy
There were 73 patients using anti-platelet drugs such as aspirin and clopidogrel in NCG group, 75 patients without any anti-thrombotic therapy. We compared the anti-platelet subgroup with OFL group (Table 2).
Off-label anti-coagulation therapy was not superior to prevent stroke or SE compared to anti-platelet therapy. The bleeding risk was higher in OFL group as well. Whereas, off-label anti-coagulation could decrease coronary events and all-cause mortality apparently.

Discussion
AF causes severe economic burden and medical cost [10]. In east Asia, NOACs have been proved to reduce the burden of stroke and death compared to VKA [11]. Previous study have indicated that nonwhites as Asians with AF were at greater risk for anti-coagulation related intracranial hemorrhage (ICH) [5]. Therefore, some compromised dosage of NOACs are applied for the treatment of atrial brillation in east Asia, which is so called "Asian dose".
Evidence has showed the bene t of low dose rivaroxaban in east Asia [8,12]. Low dose of NOACs have been proposed to decrease the risk of hemorrhage compared to standard dose of NOACs. However, Cheng. et al pointed out that off-label rivaroxaban was associated with an increased risk of ischemic stroke with 2.75 times [6]. The standard dose NOACs was associated with a relatively mild stroke in the patients compared to under-dosed NOACs [13]. The previous studies were similar to our study which showed a poor clinical bene t of off-label anti-coagulation. The thrombosis risk was signi cantly higher in OFL group compared to SAG group. However, the thrombosis risk in OFL group was even higher than NAG group. This might be partly because of the survival period was shorter in NAG group. In addition, our results indicated that Asian dose of anti-coagulation therapy indeed did not lead to a less bleeding risk compared to the standard anti-coagulation therapy. In our study, standard dose of NOACs is only related to slightly increased risk of gingival bleeding and skin bleeding.
Here, our results show that the CHA 2 DS 2 -VASc score among the groups were close. The proportion of previous stroke and SE history were higher in OFL and SAG group while the age was higher in NAG group.
This outcome is highly reasonable in clinic as the anti-thrombotic therapy are more active for stroke AF patients. The HAS-BLED score of NAG group was higher compared to other groups. That might be the reason why anti-coagulation therapy was not applied for these patients. For elder patients, anticoagulation needs more prudent due to the higher bleeding risk as well. In addition, rivaroxaban is used more commonly bene t from its various off-label dosage (10/15/20 mg per tablet).
Collectly, our study demonstrate that off-label anti-coagulation therapy can increase the survival rate of AF patients without decreasing the risk of stroke. We supposed that the major reason may be the reduction of coronary events, which are similar to a previous study in China [8]. Low dose rivaroxaban could reduce thrombus formation and platelet-dependent thrombin generation in patients with coronary heart disease, which are the potential ischemic bene ts [14]. However, off-label NOACs brought higher bleeding risk without any bene ts on thromboembolic events risk.
There were several limitations in our study. The sample size was relatively small, which may due to the loss of follow-up. In addition, the study was an observational study. More evidence such as randomized clinical trail is needed.

Conclusion
Asian dose of anti-coagulation therapy can not decrease the stroke risk of AF patients in east China. For the prevention of stroke and SE, standard dose of NOACs should be more appropriate for Asian AF patients. More experimental studies are needed to be carry out in the future.

Declarations
Ethics approval and consent to participate All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. All patients have signed informed consent previously.

Consent for publication
Written informed consent for publication was obtained from all participants.

Availability of data and materials
The datasets used or analysed during the current study are available from the corresponding author on reasonable request. Figure 1 The anti-coagulation drugs used in OFL group (a) and SAG group (b)