In this issue of the Korean Circulation Journal, Lee et al.5) report the impact of anticoagulation intensity of warfarin in Korean population. Patients were categorized into 5 groups by mean INR value (<1.6, 1.6–1.99, 2.0–2.59, 2.6–2.99, and ≥3.0). Obviously, mean INR under 1.6 group showed the highest thromboembolic events as expected. Also, mean INR ≥3.0 groups showed higher major bleeding events. Major bleeding, primary net clinical outcome, and secondary net-clinical outcome were the lowest in patients with mean INR 1.6–1.99 and 2.0–2.59. The dichotomous analysis showed that mean INR 1.6 and 2.6 could be appropriate cut-off to avoid both excessive thromboembolic and major bleeding events. Patients with mean INR 1.6–2.6 was significantly associated with the reduced risk of thromboembolic events, major bleeding, primary and secondary net clinical outcome, and mortality compared with patients with mean INR <1.6 or >2.6. Mean INR ranged from 2.0 to 3.0 did not show significant benefit compared with mean INR <2.0 or >3.0. In the aspect of TTR, TTR ≥70% was associated with less major bleeding risk than TTR <70% in all of TTR calculating criteria suggested in this study. Combined ideal mean INR (1.6–2.6) and satisfied TTR (≥70%), major bleeding, primary, and secondary net clinical outcome was significant lower in patients with ideal mean INR range and satisfied TTR compared to those with unsatisfied INR or TTR. Statistically insignificant, but numerically, the risk of thromboembolic events was lower in patients with ideal mean INR range and satisfied TTR. These data provide evidence to use warfarin more safely in a lower target INR range in Korean population.

There are several limitations to be mentioned. Firstly, this study was a single center study, thus, identification of clinical outcomes was limited. Second, this study was a retrospective analysis. The hypothesis about the optimal ranges of INR for Korean population could be suggested from this kind of retrospective observational studies, however, what optimal target INR is in a certain population can be warranted by prospective randomized clinical trials. In patients with mean INR <1.6, major bleeding incidence was numerically higher than INR 1.6–1.99 and 2.0–2.59 (7.0%/year vs. 3.7%/year vs. 4.3%/year, respectively). This finding suggested that physician might intentionally target lower INR in patients with high bleeding risk. Physicians' intention to achiever lower or higher INR target in a certain subset of patients cannot be adjusted in a retrospective study. Third, Japan suggested lower INR in elderly patients aged ≥70 years. Although median age of this study population was about 70 years, almost 30% of patients were younger than 70 years. Further studies are needed to determine whether age or body size-specified INR target can achieve better net clinical outcome in Asian patients. Lastly, the use of warfarin dramatically decreased in patients with non-valvular AF and rapidly replaced with NOACs, therefore, the clinical impact of this study might be attenuated in contemporary era. However, warfarin still had a certain role for special subgroups such as patients with moderate chronic kidney disease or end-stage renal disease, thus, optimal INR target for these patients with special considerations should be evaluated in future studies. In addition, in the use of NOACs, what optimal dose for Asian or Korean is controversial. There were pre-defined dose reduction criteria for all NOACs from pivotal randomized clinical trials, but these were not tested in a large-size Asian population. Many observational data from Asian population told us that lower doses of NOACs could be safely used without increase of thromboembolic event.6)7)8) Considering low TTR in Asian population even in randomized clinical trial,9) these previous observational data should be cautiously interpreted, because common comparator of these studies was warfarin who might have low TTR in real-world setting. Off-label underdosed NOAC was associated with increased risk of thromboembolic events and at least attenuated the net clinical benefit of NOACs compared with warfarin in previous studies.8)10)

In conclusion, warfarin plays an important role in OAC therapy for AF patients, and the current study provides for applying of a lower target INR (1.6–2.6) in Korean population treated with warfarin, with better clinical effectiveness and safety compared to a standard target INR (2.0–3.0). However, there is still limited evidence to support a lower INR range for warfarin or lower dose for NOAC in East Asian patients; therefore, further studies are needed to find both optimal target INR in warfarin therapy and optimal dose in NOAC therapy for Korean AF patients in various clinical scenario.