With astonishing advancement in the field of medical science, overall life expectancy has risen and with increased life expectancy, prevalence of atrial fibrillation (AF) in elderly patients have also showed incremental tendency.1)2)3) AF, one of the most common co-morbidity in elderly cardiovascular patient and one of the highest growing disease entity in general public, is relatively common in general practice and can easily be neglected due to predominant clinical nature being asymptomatic.4)5) However, AF is a very potent predisposing factor of disabling cerebral infarction and along with rhythm and rate control strategy, anti-coagulation treatment to prevent catastrophic cerebrovascular event is one of the key elements to improving clinical outcome. The concept of anti-thrombotic therapy to prevent ischemic stroke in AF patient is one of the cornerstones of AF treatment.6)7)8)9) Aspirin and warfarin were introduced first as treatment choice and patient selection for anticoagulation became more feasible by accompanying risk stratification system utilizing CHA₂DS₂-VASc score.10) Despite the clinical benefits of anti-coagulation, many clinicians were hesitant to use warfarin due to complex nature of the medication itself and also the risk of bleeding. However, the introduction of non-vitamin K antagonist oral anticoagulant (NOAC) and the consequential results from many researches enabled patients to receive proper anti-coagulation therapy with high efficacy and safety.11)12)13) In spite of the results from many studies conducted to validate the clinical benefit of NOAC, general consensus on whether to treat very elderly patients (age above 85) with anticoagulants is still relatively subjective. Recently, many researchers delved into the idea of utilizing nationwide insurance cohort data to analyze the benefit of NOAC compared to warfarin in elderly patients.14)15)16)17) Furthermore, numerous reports including ELDERCARE AF, have demonstrated that regardless of patient frailty, anti-coagulation with NOAC had beneficial effect with respect to stroke and cerebrovascular accident (CVA).18) However, clinical studies designed specifically designating very elderly patients are relatively sparse. Therefore, our aim is to assess and evaluate the efficacy and safety of NOAC and warfarin in very elderly patients via Korean national insurance cohort registry.

Our study focused on very elderly AF patients who were in-need of anti-coagulation therapy from 2015 to 2019. Through nationwide insurance cohort data, clinical information of over 66,000 patients was collected for analysis. Amongst the entire insurance cohort database, patients with prior history of CVA, ICH and GI bleeding were excluded. A total of 10,625 patients who were either prescribed with aspirin or OAC were enrolled. Patients who were on OAC had the highest event free survival regarding stroke, TIA and CVA and within the OAC group, NOAC showed better efficacy regarding primary endpoint.

Over the course of several decades, many studies have looked into the prospect of finding an OAC with high efficacy and safety. With the discovery and introduction of NOAC back in the early 2010’s, new era of anticoagulation in AF patients emerged. Physicians treating AF patients had more evidence to prescribe NOAC with higher efficacy and safety compared to warfarin and moreover, patients clearly benefitted from this NOAC based anticoagulation approach since failure to successfully anticipate anticoagulation could ultimately result in catastrophic debilitating consequences.

Nevertheless, the issue of ischemic and bleeding risk in very elderly AF patients is still an ongoing agenda with relatively sparse clinical data. There have been studies conducted to investigate the clinical impact of NOAC in very elderly patients. Chao et al.16) successfully demonstrated that in very elderly patients, NOAC showed high efficacy and safety with respect to ischemic stroke and ICH. ELDERCARE-AF also investigated the clinical benefit of anticoagulation with NOAC in very elderly AF patients and was able to show that patients with NOAC showed higher event free survival with no significant safety deficit.19) However, despite successful elaboration of the importance of anticoagulation in elderly AF patients, lack of contemporary real world data preclude physicians to actively initiate anticoagulation in such patients.

To our knowledge, our study is the largest nationwide cohort database established during the high time of NOAC era and study protocol comprising relatively recent contemporary follow up period with safety profile encompassing entire bleeding spectrum. In our study, patients who were treated with NOAC showed higher event free survival regarding systemic thrombotic events compared to those who were on aspirin or no antithrombotic therapy. Even after adjusting for clinical and classical risk factors via Cox regression analysis, NOAC seemed to attribute patients with the highest clinical benefit with respect to primary outcome. Patients prescribed with NOAC did show higher bleeding tendency when entire bleeding spectrum was included for analysis. However, with matters pertaining to debilitating bleeding events such as ICH which could require additional surgical measures, NOAC did not show any clinical difference compared to aspirin. Furthermore, individual comparison of death rate resulting from CVA and bleeding showed that despite the overall incidence of bleeding event being higher in NOAC group, the rate of death was significantly lower in NOAC group compared to aspirin group for in-hospital, 6 months and 1year and beyond mortality.

The everlasting issue of proper anti-coagulation strategy is an ongoing debate stemming back from the 1950’s. The successful risk stratification and development of NOAC enabled AF patients with improved overall clinical outcome. Despite many endeavors to optimize successful anti-coagulation in AF patients, the apprehension of bleeding risk always lingered on the minds of physicians. The greatest clinical risk AF patients possess is debilitating CVA consequences which AF can ensue. It is evident that AF is one of the most commonly known predisposing factor for ischemic stroke. With improved social awareness and accessibility of clinics treating CVA, the overall survival has increased drastically. However, due to the diverse clinical nature of ischemic stroke, whether it being TIA without sequalae to malignant MCA infarction resulting permanent debilitation, early recognition and intervention is key to a successful management of AF. Recently, many researchers have shown great interest on the social economic burden CVA can have both individually and collectively. Due to the complex nature of CVA, it requires lengthy hospital stays along with costly healthcare management and inability to perform adequate ‘work performance’ causes economic and social revenue loss therefore appropriate anticoagulation with adequate medication providing both high efficacy and safety is essential in AF patients regardless of age.20)21) Our study was able to successfully reiterate previous notion on the importance of anti-coagulation in elderly patients regarding NOAC with large nationwide cohort data which comprised wide spectrum of ischemic and bleeding disease entities and in spite of the high bleeding event rate of NOAC compared to aspirin, the commensurate death rate was lower in NOAC. In spite of our results, this current study has some limitations. Firstly, due to the retrospective study design and limited access to NHIS data we were not able to fully anticipate factors which might have affected the decision to proceed with anticoagulation (i.e., malignancy, liver disease, transfusion and proton pump inhibitor use). Furthermore, the unique nature of the NHIS data made the research less feasible for acquisition of clinical parameter (i.e., systolic/diastolic blood pressure, heart rate, left ventricular ejection fraction, etc.) Therefore, despite our efforts to minimize biases via propensity score matching utilizing parameters comprising the CHA₂DS₂-VASc score, potential bias caused by clinical parameters still remains. Secondly, the study population being ‘very elderly patients’ renders possible criticism for survival selection bias. These group of patients have already “survived” to the point where they are eligible for another survival analysis. Therefore, relatively healthier patients could have been selected for analysis compared to other age groups. Thirdly, our study only comprised Koreans therefore cannot fully represent ‘very elderly’ population. Therefore, further multi-national prospective randomized control study designed specifically to investigate the efficacy and safety of NOAC in very elderly patients is warranted.

In conclusion, elderly patients who were prescribed with NOAC showed higher efficacy regarding CVA and lower death rate resulting from bleeding event which is an optimistic finding for both AF patients and the medical care provider.