Date Published: March 18, 2019
Publisher: Public Library of Science
Author(s): Yi-Hsin Chan, Kuang-Tso Lee, Yi-Wei Kao, Chien-Ying Huang, Yung-Lung Chen, Samuel Chi-Ling Hang, Pao-Hsien Chu, Andreas Charidimou.
Previous studies indicated low-intensity warfarin (INR target of 1.5–2.5) achieved reduced hemorrhage without increasing thromboembolism for Asians with non-valvular atrial fibrillation (NVAF). Whether non-vitamin K antagonist oral anticoagulant (NOAC) is superior to warfarin with good time in the therapeutic range (TTR) based on lower INR target among Asians with NVAF remains unknown.
In this retrospective study collected from Taiwan Chang Gung Memorial Hospital Database, there were 5,197, 3,396, and 9,898 consecutive patients taking warfarin, NOAC, and no-treatment, respectively, from January 1, 2000 to December 31, 2015. Propensity-score weighting was used across the study groups. Patients were followed until the first occurrence of study outcome or end date of study.
Among those patients taking warfarin, the mean”artificial” TTR (aTTR) based on a lower INR target of 1.5–2.5 was 44.4±33.3%. Total 79.2% (n = 2,690) patients took low-dose NOACs. Patients with aTTR in the range from <30%(34.0%), 30–50%(17.6%), 50–70%(23.5%) to >70%(24.9%) showed decremental risks of efficacy and composite outcome compared with no-treatment. The risk of major bleeding didn’t increase among patients with top aTTR>70% compared to no-treatment. The NOAC group showed a comparable risk of composite outcome to the warfarin subgroup with aTTR of >70% (P = 0.485). The NOAC group had a lower risk of composite outcome than warfarin subgroup with TTR of>70% based on the INR target of 2.0–3.0 (P = 0.004).
NOACs showed a comparable risk of efficacy, safety, and composite outcome to well-managed warfarin based on a lower INR target of 1.5–2.5 in Asians with NVAF taking oral anticoagulants.
Warfarin is commonly used for prevention of thromboembolic events in patients with non-valvular atrial fibrillation (NVAF). Previous meta-analysis indicated that warfarin reduced the risk of thromboembolic events by 65% and all-cause mortality by 22% as well when compared with no treatment . However, the benefit of warfarin was largely compromised by its inconvenience to use and increased risk of major bleeding. The risks of bleeding and thromboembolism depend on the intensity of anticoagulation as measured by the International Normalized Ratio (INR) when taking warfarin. Both the European Society of Cardiology (ESC) and the American Heart Association (AHA) recommend a target of INR range of 2.0 to 3.0 for prevention of thromboembolism in patients with NVAF [2, 3], where the lowest risk of thromboembolism and bleeding cab be only achieved in such a narrow therapeutic range. However, several studies indicated that Asians are more sensitive to warfarin and vulnerable to warfarin related bleeding than Non-Asians [4, 5]. The meta-analysis indicated that low-intensity warfarin therapy (INR target of 1.5–2.5) can achieve reduced hemorrhage without increasing thromboembolism for Asian patients with NVAF taking warfarin [6–9]. Recently, non-vitamin K antagonist oral anticoagulants (NOACs) have been demonstrated to be effective and safe for prevention of thromboembolism in patients with NVAF . It is noted that NOACs were more effective and safer in Asians than in non-Asians, which was majorly contributed from the tendency of poorer TTRs (time in therapeutic range) with the INR target of 2.0–3.0 among Asians taking warfarin . However, the potential benefit of NOACs over warfarin with high TTR based on a lower INR target of 1.5–2.5 remained questionable among Asians with NVAF. This study aimed at elucidating the efficacy and safety of NOACs compared to warfarin with a variety of TTR with a lower INR target of 1.5–2.5 specifically focused on Asians with NVAF taking oral anticoagulant.
A total of 5,197, 3,396, and 9,898 consecutive patients taking warfarin, NOAC, and no-treatment, respectively, from January 1, 2000 to December 31, 2015, were enrolled. In general, NOAC patient group was older, had higher CHA2DS2-VASc and HAS-BLED scores, and had a higher proportion of comorbidities and multiple medications than the warfarin and no-treatment groups before propensity score weighting (Table 1). After propensity score weighting, the three study groups were well-balanced in most characteristics (Table 2).
In the present large in-hospital cohort study, we investigated the aTTR-specific risk of thromboembolic and major bleeding events in Asians with NVAF taking moderate-intensity anticoagulation therapy with the INR target of 1.5–2.5. The adjusted risk of efficacy, safety, and composite outcomes in patients taking NOACs were comparable to those taking warfarin with aTTR of > 70%. Our study indicated that NOACs may be an alternative choice to warfarin with aTTR of > 70% based on a lower INR target of 1.5–2.5 in Asian patients with NVAF.
The present study had several limitations. This study is multiple-center, retrospective, and observational study, which limited the level of evidence presented. The selection of anticoagulants was not in a randomized determination, which is evidenced by significant difference in baseline characteristics between patients on no treatment, warfarin, and NOACs. Although an extensive number of variables was included in our propensity score model and achieved a close balance for most factors, residual confounding by unmeasured factors including blood pressure control, use of herbal and over the counter medications, and vitamin K dietary intake cannot be excluded due to its retrospective design. The amount of patients without any treatment is quite staggering in the study (n = 9,898). This is the largest group by quite some margin. Furthermore, the mean CHA2DS2-VASc score of the no-treatment group was 3.36 in the present study (Table 1), indicating that the majority of the patients in the no-treatment group was still indicated for oral anticoagulation therapy according to the current guideline. Our data was compatible to previous large national cohort study, which also reported that there were 157,829 patients (85%) indicated for oral anticoagulant (CHA2DS2-VASc score of ≥ 2) among the total 186,570 AF patients without any oral anticoagulant exposure in Taiwan’s real-world practice .Why such a high prevalence of patients indicated for therapy but without receiving oral anticoagulant among the total AF population especially in the Asia is unclear. One important issue is that the risk of major bleeding regarding the use of oral anticoagulant in AF patients is higher in those of Asian ethnicity . Therefore, physicians in Asia may avoid prescribing any oral anticoagulants in those patients perceived to be at an increased risk of hemorrhage. Although the sensitivity analysis with restriction of the patient enrollment started from January 1, 2010 to December 31, 2015 has reduced the patient number without any treatment down to 4,878, and the result of sensitivity analysis was still compatible to those of the main analysis, this augments regarding selection bias still cannot be excluded. In addition, we observed a high prevalence of low-dose NOAC prescriptions in the present cohort. The lack of body weights and detailed renal function makes it difficult to determine if those patients taking low-dose NOACs were correctly prescribed an “adjusted” low-dose or “off-label” under-dose NOACs. Nevertheless, the present study may be highly prone to selection bias and confounded by indication as physicians may have avoided prescribing full-dose NOAC or warfarin with a target INR 2.0–3.0 in those patients perceived to be at an increased risk of bleeding. In the present study, only primary discharge diagnoses were adopted in order to improve the accuracy of clinical outcome, and some outcomes including a mild form of stroke or bleeding without hospitalization may be missed.
The efficacy, safety, and composite outcome of NOACs were comparable to warfarin with aTTR of more than 70% based on a lower INR target of 1.5–2.5 in Asians/Chinese with NVAF taking oral anticoagulants. NOAC may be an effective, safe, and convenient alternative to the well-managed warfarin among those patient groups. The results are hypothesis generating and may form a basis of a randomized control trial in the future.