Renal Function and Clinical Outcomes Among Elderly Patients With Nonvalvular Atrial Fibrillation From ANAFIE

Background Advancing age, decreasing renal function, and atrial fibrillation are strongly associated. Real-world evidence of direct oral anticoagulant (DOAC) use among elderly patients ≥75 years of age with nonvalvular atrial fibrillation and renal dysfunction is limited. Objectives This study sought to assess 2-year outcomes and anticoagulant treatment, stratified by renal function. Methods Enrolled patients were divided into 4 subgroups by creatinine clearance (CrCl) to determine the impact of renal dysfunction on clinical outcomes. Results Of 32,275 patients, 26,202 with CrCl data were analyzed (median follow-up 2.00 [IQR: 1.92-2.00] years); 1.3% of patients had CrCl <15 mL/min, 10.7% had CrCl 15 to <30 mL/min, 33.4% had CrCl 30 to <50 mL/min, 35.8% had CrCl ≥50 mL/min, and 18.9% had unknown CrCl. Cumulative incidences of stroke/systemic embolic events, major bleeding, major plus clinically relevant nonmajor bleeding, cardiovascular death, all-cause death, and net clinical outcomes increased with decreasing CrCl. In multivariable Cox regression analysis, lower CrCl emerged as an independent risk factor for these clinical outcomes, except for major bleeding, compared with CrCl ≥50 mL/min. The effectiveness and safety of DOACs over warfarin were similar or better across 3 CrCl subgroups with CrCl 15 mL/min or more. DOAC use was associated with a lower risk of stroke/systemic embolic events, major bleeding, cardiovascular death, all-cause death, and net clinical outcome compared with warfarin in patients with CrCl 30 to <50 mL/min. Conclusions Incidences of major clinical outcomes increased with decreasing renal function in elderly nonvalvular atrial fibrillation patients. DOACs were effective and safe even in patients with renal dysfunction (CrCl 15-<50 mL/min). (Prospective Observational Study in Late-Stage Elderly Patients with Non-Valvular Atrial Fibrillation: All Nippon AF In Elderly Registry [ANAFIE Registry]; UMIN000024006)

A dvancing age, decreasing renal function, and atrial fibrillation (AF) are strongly associated. 1 About onethird of patients with AF have chronic kidney disease (CKD). A close bidirectional relationship between these 2 conditions has been described, whereby CKD increases the risk of AF and AF increases the risk of CKD. [1][2][3][4] The age-related decline in renal function raises several issues for the medical care of patients with AF. Renal dysfunction is associated with poor outcomes, including increased risk of stroke/systemic embolic events (SEEs), major bleeding, and increased mortality. [5][6][7][8] Additionally, reduced renal function has a considerable effect on the pharmacokinetics, pharmacodynamics, and safety of drugs that have high renal excretion rates. 9 Current clinical practice guidelines recommend the use of direct oral anticoagulants (DOACs), rather than warfarin, as the firstline therapy for stroke prevention in patients with nonvalvular atrial fibrillation (NVAF) and creatinine clearance (CrCl) $30 mL/min. [10][11][12] However, DOACs such as dabigatran, rivaroxaban, apixaban, and edoxaban are excreted by the kidney to varying extents, ranging from 27% to 80%, and thus are more dependent on renal excretion compared with warfarin, which is primarily metabolized by the liver. 13 There is a misconception that administering DOACs to elderly patients with renal dysfunction is challenging, and the practice of prescribing warfarin for these patients is deeply rooted in clinical settings.
However, delayed warfarin metabolism is associated with an increased risk of bleeding, and increasing age is correlated with increased bleeding complications among patients taking warfarin. 14,15 Compared with warfarin, DOACs appear to be equally safe and effective across all levels of CKD. 8 At the same time, some studies have reported that DOACs were associated with a lower risk of all-cause mortality, major bleeding, 16 and stroke/SEEs. 17 In clinical trials, the relative efficacy and safety of DOACs compared with warfarin were consistent, regardless of the renal function level of patients. [18][19][20] Table 2).
The risk of cardiac events, ischemic heart disease, myocardial infarction, heart failure requiring hospitalization, cardiac death and sudden death, and cardiovascular events was significantly higher in the 3 CrCl <50 mL/min subgroups than in the CrCl $50 mL/min subgroup. In general, the risk  of fractures and falls was significantly higher in the subgroups with lower CrCl than with CrCl $50 mL/min (Supplemental Table 1).
We conducted additional analyses by renal function, applying the Fine-Gray model and using all-cause death as a competing risk. The results of this analysis revealed that after incorporating the competing risk into the statistical analysis, the association between renal function and events overall did not change (Supplemental Table 2).    Table 4).
Supplemental    showed that the percentage of NVAF patients with CrCl <50 mL/min was 37%. 5

RELATIONSHIP BETWEEN CrCl AND EVENT RATES.
Considering that DOACs are not indicated for patients with CrCl <15 mL/min, in the present study, we registry also indicated that lower CrCl is not an independent predictor for major bleeding. 7 In the ANAFIE registry, bleeding events were assessed using the International Society on Thrombosis and Haemostasis criteria, which may be the reason for the difference between this and previous findings. 5,7 Possibly, hemoglobin levels were not adequately evaluated before and after the onset of hemorrhagic events. Alternatively, blood transfusions may have been given less frequently in elderly patients in actual clinical practice than in clinical trials. 26 These points could explain this study's low incidence of major bleeding. Additionally, all-cause mortality was very high in the CrCl <15 mL/min group and likely contributed to the lower incidence of major bleeding.
There was a significant difference in the cumulative incidences of major bleeding among the 4 CrCl subgroups. However, in the multivariable analysis, renal dysfunction was not identified as an independent risk factor for major bleeding in this study. This result indicates that multiple comorbidities rather than reduced renal function may further contribute to major bleeding in elderly patients. Nevertheless, renal dysfunction was an independent risk factor for bleeding events, including major plus clinically rele-