Clinical Outcomes of Anticoagulation Therapy With Direct Oral Anticoagulants or Warfarin in Patients With Atrial Fibrillation and Renal Impairment After Bioprosthetic Valve Replacement

Abstract

Background: Atrial fibrillation (AF) after bioprosthetic valve (BPV) replacement is common in older patients with multiple comorbidities and is associated with a heightened risk of thromboembolism. Anticoagulation therapy is often indicated, but renal impairment and other comorbidities elevate bleeding risk, making clinical decisions complex. This study compared clinical outcomes between warfarin and direct oral anticoagulants (DOACs) in this high-risk population.

Methods and Results: This subgroup analysis of the BPV-AF Registry included 612 patients treated with oral anticoagulants after BPV replacement, stratified by renal function: normal or mild impairment (creatinine clearance [CCr] ≥50 mL/min), mild-to-moderate impairment (30 mL/min ≤ CCr < 50 mL/min), and moderate-to-severe impairment (15 mL/min ≤ CCr < 30 mL/min). Baseline characteristics and outcomes were analyzed within each stratum. The composite outcome of stroke, systemic embolism, and cardiovascular events was numerically less frequent in the DOAC than warfarin group across all strata, although the differences were not statistically significant. Major bleeding also tended to be lower in the DOAC group.

Conclusions: In this study from a Japanese nationwide registry comparing outcomes of AF patients after BPV replacement with severe renal impairment between those treated with DOACs and those treated with warfarin, comparative conclusions between DOACs and warfarin cannot be drawn because of the small sample size. Nonetheless, both anticoagulants may be acceptable in clinical practice, highlighting the need for individualized decision-making based on patient risk.

Central Figure

Kaplan-Meier curves for the composite outcome of stroke, systemic embolism, and cardiovascular events in patients with moderate-to-severe renal impairment (15 mL/min ≤ creatinine clearance < 30 mL/min) treated with warfarin or direct oral anticoagulants (DOACs).

Most patients who undergo aortic valve replacement with bioprosthetic valves (BPVs) are elderly, with the number of patients with valvular heart disease requiring BPV replacement increasing with the aging society.^1–3 Atrial fibrillation (AF) becomes more prevalent with age and is a common comorbidity in the elderly population. AF and valvular heart disease are both influenced by aging and lifestyle-related risk factors, such as hypertension, diabetes, and dyslipidemia, which can lead to polypharmacy. AF in patients undergoing valve replacement surgery further increases the complexity of treatment decisions.⁴ In addition, older patients often experience a decline in renal function and frailty, which exacerbate drug metabolism, increase concerns regarding elevated blood concentrations of anticoagulants due to drug–drug interactions, and lead to a higher risk of bleeding.^5–8 Despite the increasing number of older patients with AF undergoing BPV replacement therapy, data on anticoagulant therapy, particularly for those with impaired renal function, are limited.^9,10

In clinical practice, inappropriate underdosing of direct oral anticoagulants (DOACs) in patients with atrial fibrillation (AF) has been reported to increase the incidence of ischemic stroke without reducing bleeding events,^11,12 suggesting the need to clarify actual outcomes in high-risk patients. In the BPV-AF Registry, no explicit age criterion was established for inclusion; however, the mean age of the enrolled patients was approximately 75 years. In this study, the term “elderly” refers broadly to older adults, generally aged 65 years and above, in accordance with conventional clinical usage. The aim of this study was to describe the current status of anticoagulant therapy in older patients with AF, BPV replacement, and impaired renal function using data from the multicenter prospective observational BPV-AF Registry.^13,14 As highlighted in the Japanese Circulation Society (JCS) / Japan Heart Rhythm Society 2024 Guidelines,¹⁵ individualized anticoagulant therapy is warranted in elderly patients with AF and comorbidities such as frailty and cognitive decline, considering the increased risk of both thromboembolic and bleeding events in this population.¹⁶

Methods

Study Population

The BPV-AF Registry was a multicenter, prospective, observational study conducted across 16 hospitals in Japan to evaluate antithrombotic therapy and associated clinical events in patients with AF following BPV replacement. Patients were enrolled between September 2018 and October 2019 and were followed for at least 1 year. BPV replacement included aortic, mitral, or both valves and was performed via either surgical valve replacement or transcatheter aortic valve implantation. AF was defined as paroxysmal, persistent, or permanent; patients with transient postoperative AF were excluded. The primary conclusion of the BPV-AF Registry was that the risks of major bleeding and stroke/systemic embolism were similar between warfarin and DOACs therapy.

Of the 894 participants in the BPV-AF Registry, the present subgroup analysis included 612 patients who were prescribed oral anticoagulants (OACs). These patients were stratified by renal function into 3 groups according to creatinine clearance (CCr): normal or mild impairment (CCr ≥50 mL/min), mild-to-moderate impairment (30 mL/min ≤ CCr < 50 mL/min), and moderate-to-severe impairment (15 mL/min ≤ CCr < 30 mL/min). Although baseline characteristics were compared across the 3 renal function groups, the present analysis specifically focused on the moderate-to-severe impairment group, which represents a particularly high-risk population. In these patients, impaired renal function is known to increase both thromboembolic and bleeding risks, and the use of DOACs remains less established. Therefore, clarifying the safety and efficacy of DOACs in this study has meaningful clinical implications.

Within the moderate-to-severe impairment group, we compared the incidence of thrombotic and bleeding events between patients treated with warfarin and those treated with DOACs. Stroke risk was assessed using 3 scoring systems: the widely used CHADS₂ and CHA₂DS₂-VASc scores, and the more recently developed HELT-E₂S₂ score. The HELT-E₂S₂ score, developed from the Japanese Risk of Stroke in Atrial Fibrillation Patients (J-RISK) study,^17,18 a pooled analysis of 5 major Japanese AF registries, has shown superior predictive accuracy for ischemic stroke in Japanese patients,^15,17,19,20 making it particularly relevant for our elderly cohort with impaired renal function. The HELT-E₂S₂ score assigns 1 point each for hypertension (H), elderly age (E; age 75–84 years), low body mass index (L; BMI <18.5 kg/m²), and AF type (T; persistent or permanent), and 2 points each for extreme elderly (EE; age ≥85 years) and prior stroke (S; ischemic stroke or transient ischemic attack). In addition, data on medication regimens, the number of prescribed drugs, medical history, and comorbidities were analyzed. The number of medications and the presence of comorbid conditions were used as surrogate markers for frailty and polypharmacy, which are commonly observed in elderly patients with renal dysfunction and have been associated with adverse outcomes in large observational studies such as the REPOSI registry.⁶

This study was conducted in accordance with the Declaration of Helsinki, the Ethical Guidelines for Medical and Health Research Involving Human Subjects issued by the Ministry of Health, Labor and Welfare of Japan, and all other applicable regulatory requirements. The study protocol and informed consent documents were approved by the Ethics Committee of the National Cerebral and Cardiovascular Center (M30-068; September 26, 2018) and each participating hospital. This study was registered with the University Hospital Medical Information Network (UMIN) Clinical Trials Registry (ID: UMIN000034485).

Outcomes

The primary outcome was a composite of stroke, systemic embolism, and cardiovascular events, including heart failure requiring hospitalization or cardiac death. For the efficacy outcome, the incidence of stroke or systemic embolism was evaluated. For the safety outcome, the occurrence of major bleeding was analyzed based on the International Society on Thrombosis and Haemostasis criteria.²¹

Statistical Analyses

Continuous variables are presented as the mean±SD, whereas categorical variables are presented as numbers with percentages. The significance of differences in baseline characteristics among the renal function groups was determined using analysis of variance for continuous variables and the Chi-squared test or Fisher’s exact test for categorical variables. Patients receiving warfarin- and DOAC-based treatments were compared using t-tests for continuous variables. The relationship between HELT-E₂S₂ scores and renal function was analyzed using a generalized linear regression model. For composite and primary outcomes, event rates (%/year) and 95% confidence intervals (CIs) were calculated using a Poisson distribution. The cumulative incidence rates of the composite and primary outcomes were calculated using the Kaplan-Meier method and compared using log-rank tests. Univariate Cox proportional hazards regression models were used to calculate hazard ratios (HRs) and 95% CIs for composite and primary outcomes.

In the multivariable model, the HELT-E₂S₂ score was used to adjust for baseline thromboembolic risk. Additional covariates, namely antiplatelet use, malignancy, and dementia, were selected based on their established clinical relevance in older patients with impaired renal function and their potential to independently affect both bleeding and thromboembolic outcomes. In particular, malignancy was included given that it has been reported as one of the major causes of death, in addition to cardiovascular death, in elderly AF patients.¹⁶ The number of adjusting variables was intentionally limited to avoid model overfitting.

All reported P values are 2-tailed, and P<0.05 were considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Baseline Patient Characteristics

Baseline characteristics of patients according to renal function group are presented in Table 1. As renal function declined, age increased significantly, whereas body weight and BMI decreased significantly. CHADS₂, CHA₂DS₂-VASc, and HAS-BLED scores increased in parallel with worsening renal function, indicating elevated risks of thromboembolism and bleeding. HELT-E₂S₂ scores also increased with increasing renal impairment. Paroxysmal AF became more prevalent with worsening renal impairment. Warfarin was more frequently prescribed than DOACs in all renal function groups, although the difference was not statistically significant (P=0.448). Antiarrhythmic drugs were more commonly prescribed in patients with normal or mildly impaired renal function, whereas antihypertensive drugs were significantly more common in the moderate-to-severe renal impairment group. There were no significant differences in the mean number of prescribed medications or in the proportion of patients receiving ≥5 medications across the groups. No significant differences were observed in the medical history or comorbidities, except for a significantly higher prevalence of heart failure in the moderate-to-severe renal impairment group.

Table 1.

Baseline Clinical Characteristics for All Patients and According to Renal Function Impairment

	All (n=612)	Renal function impairmentA			P value
		Moderate to severe (n=170)	Mild to moderate (n=259)	None/mild (n=183)
Male sex	276 (45.1)	53 (31.2)	103 (39.8)	120 (65.6)	<0.001
Age (years)	80.4±6.9	84.2±5.7	81.2±6.2	75.8±6.2	<0.001
Weight (kg)	54.1±11.3	47.2±8.8	53.1±9.9	61.9±10.4	<0.001
Weight <50 kg	230 (37.6)	105 (61.8)	107 (41.3)	18 (9.8)	<0.001
BMI (kg/m2)	22.3±3.7	20.8±3.9	22.2±3.4	23.9±3.4	<0.001
BMI <18.5 kg/m2	81 (13.5)	40 (24.0)	35 (13.7)	6 (3.4)	<0.001
CHADS2 score	2.5±1.2	2.8±1.1	2.5±1.2	2.3±1.3	<0.001
CHADS2 score ≥2.0	457 (79.8)	136 (89.5)	197 (79.8)	124 (71.3)	<0.001
CHA2DS2-VASc score	4.2±1.5	4.7±1.4	4.3±1.4	3.7±1.5	<0.001
CHA2DS2-VASc score ≥3.0	513 (88.8)	149 (96.8)	227 (90.8)	137 (78.7)	<0.001
HAS-BLED score	2.4±1.0	2.6±1.1	2.3±1.0	2.2±1.0	<0.001
HAS-BLED score ≥3.0	232 (40.6)	78 (51.7)	95 (38.6)	59 (33.9)	0.004
HELT-E2S2 score	2.7±1.0	3.3±1.0	2.8±1.0	2.2±0.9	<0.001
HELT-E2S2 score ≥3.0	362 (59.2)	133 (78.2)	156 (60.2)	73 (39.9)	<0.001
eGFR (mL/min/1.73 m2)	48.3±15.9	32.5±8.0	48.0±10.8	63.6±12.3	<0.001
CCr (mL/min)	42.2±16.8	23.8±4.0	39.6±5.7	63±10.9	<0.001
PT-INRB	1.9±0.5	1.9±0.5	2.0±0.5	1.9±0.5	0.187
Paroxysmal AF	201 (32.8)	56 (32.9)	97 (37.5)	48 (26.2)	0.047
Anticoagulant/antiplatelet therapy					0.448
Warfarin	378 (61.8)	102 (60.0)	156 (60.2)	120 (65.6)
DOAC	234 (38.2)	68 (40.0)	103 (39.8)	63 (34.4)
Antiplatelet drug	159 (26.0)	46 (27.1)	61 (23.6)	52 (28.4)
Antiarrhythmic drugs	199 (33.3)	51 (30.7)	75 (29.3)	73 (41.5)	0.022
Antihypertensive drugs	416 (69.6)	128 (77.1)	180 (70.3)	108 (61.4)	0.006
Lipid-lowering drugs	144 (24.1)	38 (22.9)	69 (27.0)	37 (21.0)	0.336
Diabetes drugs	78 (13.0)	14 (8.4)	35 (13.7)	29 (16.5)	0.081
Drugs for peptic ulcers	365 (61)	107 (64.5)	162 (63.3)	96 (54.6)	0.107
P-Glycoprotein inhibitors	28 (4.7)	8 (4.8)	10 (3.9)	10 (5.7)	0.689
No. medications	3.2±1.2	3.3±1.1	3.3±1.2	3.2±1.2	0.652
≥5 medications	83 (13.6)	20 (11.8)	37 (14.3)	26 (14.2)	0.535
Previous history of CVD
Ischemic stroke	91 (14.9)	28 (16.5)	35 (13.5)	28 (15.3)	0.688
Hemorrhagic stroke	15 (2.5)	4 (2.4)	8 (3.1)	3 (1.6)	0.641
Intracranial hemorrhage	21 (3.4)	5 (2.9)	13 (5.0)	3 (1.6)	0.145
Systemic embolism	8 (1.3)	2 (1.2)	3 (1.2)	3 (1.6)	0.906
Major bleeding	29 (4.7)	11 (6.5)	13 (5.0)	5 (2.7)	0.246
Comorbidities
Hypertension	457 (74.7)	136 (80.0)	186 (71.8)	135 (73.8)	0.154
Heart failure	323 (52.8)	116 (68.2)	136 (52.5)	71 (38.8)	<0.001
Dyslipidemia	297 (48.5)	87 (51.2)	127 (49.0)	83 (45.4)	0.538
Diabetes	132 (21.6)	31 (18.2)	54 (20.9)	47 (25.7)	0.220
Chronic respiratory disease	61 (10.0)	15 (8.8)	26 (10.0)	20 (10.9)	0.803
Malignant tumor	51 (8.3)	16 (9.4)	20 (7.7)	15 (8.2)	0.823
Myocardial infarction	23 (3.8)	6 (3.5)	12 (4.6)	5 (2.7)	0.575
Peripheral arterial disease	23 (3.8)	7 (4.1)	11 (4.3)	5 (2.7)	0.682
Thrombosis and embolism	21 (3.4)	7 (4.1)	6 (2.3)	8 (4.4)	0.427
Liver dysfunction	13 (2.1)	5 (2.9)	4 (1.5)	4 (2.2)	0.601
Dementia	31 (5.1)	13 (7.7)	11 (4.3)	7 (3.8)	0.192
LVEF <40%	39 (6.7)	13 (8.1)	20 (7.9)	6 (3.6)	0.153

Unless indicated otherwise, data are presented as n (%) or the mean±SD. ^APatients were divided into 3 groups according to creatinine clearance (CCr): normal or mild impairment of renal function (CCr ≥50 mL/min); mild-to-moderate impairment (30 mL/min ≤ CCr < 50 mL/min); and moderate-to-severe impairment (15 mL/min ≤ CCr < 30 mL/min). ^BInternational normalized ratio of prothrombin time (PT-INR) values are shown only for patients receiving warfarin. AF, atrial fibrillation; BMI, body mass index; CVD, cardiovascular disease; DOAC, direct oral anticoagulant; eGFR, estimated glomerular filtration rate; LVEF, left ventricular ejection fraction.

Clinical Characteristics and OACs in the Moderate-to-Severe Renal Impairment Group

Patients in the moderate-to-severe renal impairment group (15 mL/min ≤ CCr < 30 mL/min) were divided into 2 groups according to treatment: warfarin (n=102) and DOACs (n=68). Clinical characteristics were compared between these 2 groups (Table 2). The mean age of patients with moderate-to-severe renal impairment was 84.2 years, and patients in the DOACs group were significantly older than those in the warfarin group. Body weight was also significantly lower in the DOACs group, although BMI did not differ between the 2 groups. Although CHADS₂ scores did not differ significantly between the 2 groups, they tended to be higher in the DOACs group. The CHA₂DS₂-VASc scores were significantly higher in the DOACs group, suggesting that DOACs were selected for patients with a higher thromboembolic risk. HAS-BLED scores did not differ significantly between the 2 groups. The mean HELT-E₂S₂ score was similar between the 2 groups, but a score of ≥3 was significantly more frequent in the DOACs group. Renal function was significantly lower in the warfarin group, whereas paroxysmal AF was more frequent in the DOACs group. The number of prescribed medications and the proportion of patients with polypharmacy were similar between the 2 groups. However, impaired cardiac function was significantly more frequent in the warfarin group.

Table 2.

Clinical Characteristics at Baseline of Patients With Moderate-to-Severe Impairment of Renal Function^A Overall and According to Treatment Group

	All (n=170)	Warfarin (n=102)	DOAC (n=68)	P value
Male sex	53 (31.2)	38 (37.3)	15 (22.1)	0.036
Age (years)	84.2±5.7	83.0±5.8	86.1±5.2	0.001
Weight (kg)	47.2±8.8	48.3±9.2	45.5±7.9	0.038
Weight <50 kg	105 (61.8)	59 (57.8)	46 (67.7)	0.198
BMI (kg/m2)	20.8±3.9	20.9±3.1	20.7±4.9	0.755
BMI <18.5 kg/m2	40 (24.0)	21 (21.2)	19 (27.9)	0.317
CHADS2 score	2.8±1.1	2.6±1.1	3.0±1.2	0.058
CHADS2 score ≥2.0	136 (89.5)	78 (85.7)	58 (95.1)	0.065
CHA2DS2-VASc score	4.7±1.4	4.5±1.3	5.1±1.4	0.003
CHA2DS2-VASc score ≥3.0	149 (96.8)	87 (94.6)	62 (100.0)	0.082
HAS-BLED score	2.6±1.1	2.6±1.1	2.7±1.1	0.736
HAS-BLED score ≥3.0	78 (51.7)	46 (51.1)	32 (52.5)	0.871
HELT-E2S2 score	3.3±1.0	3.2±1.0	3.4±1.0	0.124
HELT-E2S2 score ≥3.0	133 (78.2)	74 (72.6)	59 (86.8)	0.028
eGFR (mL/min/1.73 m2)	32.5±8.0	29.9±7.3	36.2±7.7	<0.001
CCr (mL/min)	23.8±4.0	22.9±4.3	25±3.3	<0.001
PT-INRB	–	1.9±0.5	–	–
<1.6	–	22 (22.5)	–	–
1.6–2.6	–	69 (70.4)	–	–
>2.6	–	7 (7.1)	–	–
DOAC type and doseC
Edoxaban (mg)	23 (33.8)	–	23 (33.8)	–
60 mg	0 (0)	–	0 (0)	–
30 mg	22 (32.4)	–	22 (32.4)	–
15 mg	1 (1.5)	–	1 (1.5)	–
Apixaban (mg)	29 (42.6)	–	29 (42.6)	–
10 mg	1 (1.5)	–	1 (1.5)	–
5 mg	28 (41.2)	–	28 (41.2)	–
Rivaroxaban (mg)	14 (20.6)	–	14 (20.6)	–
15 mg	0 (0)	–	0 (0)	–
10 mg	14 (20.6)	–	14 (20.6)	–
Dabigatran (mg)	2 (2.9)	–	2 (2.9)	–
300 mg	0 (0)	–	0 (0)	–
220 mg	2 (2.9)	–	2 (2.9)	–
Paroxysmal AF	56 (32.9)	27 (26.5)	29 (42.7)	0.028
Antiplatelet drugs	46 (27.1)	46 (27.1)	0 (0)	–
Antiarrhythmic drugs	51 (30.7)	33 (33.3)	18 (26.9)	0.376
Antihypertensive drugs	128 (77.1)	76 (76.8)	52 (77.6)	0.899
Lipid-lowering drugs	38 (22.9)	20 (20.2)	18 (26.9)	0.316
Diabetes drugs	14 (8.4)	9 (9.1)	5 (7.5)	0.711
Drugs for peptic ulcers	107 (64.5)	67 (67.7)	40 (59.7)	0.292
P-Glycoprotein inhibitors	8 (4.8)	5 (5.1)	3 (4.5)	1.000
No. medications	3.3±1.1	3.3±1.0	3.3±1.1	0.930
≥5 medications	20 (11.8)	11 (10.8)	9 (13.2)	0.627
Previous history of CVD
Ischemic stroke	28 (16.5)	13 (12.8)	15 (22.1)	0.109
Hemorrhagic stroke	4 (2.4)	1 (1.0)	3 (4.4)	0.303
Intracranial hemorrhage	5 (2.9)	3 (2.9)	2 (2.9)	1.000
Systemic embolism	2 (1.2)	1 (1.0)	1 (1.5)	1.000
Major bleeding	11 (6.5)	10 (9.8)	1 (1.5)	0.052
Comorbidities
Hypertension	136 (80.0)	80 (78.4)	56 (82.4)	0.531
Heart failure	116 (68.2)	65 (63.7)	51 (75.0)	0.122
Dyslipidemia	87 (51.2)	52 (51.0)	35 (51.5)	0.950
Diabetes	31 (18.2)	23 (22.6)	8 (11.8)	0.074
Chronic respiratory disease	15 (8.8)	6 (5.9)	9 (13.2)	0.098
Malignant tumor	16 (9.4)	8 (7.8)	8 (11.8)	0.391
Myocardial infarction	6 (3.5)	3 (2.9)	3 (4.4)	0.684
Peripheral arterial disease	7 (4.1)	4 (3.9)	3 (4.4)	1.000
Thrombosis and embolism	7 (4.1)	5 (4.9)	2 (2.9)	0.704
Liver dysfunction	5 (2.9)	2 (2.0)	3 (4.4)	0.390
Dementia	13 (7.7)	7 (6.9)	6 (8.8)	0.637
LVEF <40%	13 (8.1)	12 (12.2)	1 (1.6)	0.017
TTRB (%)
Mean±SD	–	68.3±35.2	–	–
Median [IQR]	–	81.7 [35.5–100]	–	–

Unless indicated otherwise, data are presented as n (%) or the mean±SD. ^AModerate-to-severe impairment of renal function was defined as CCr ≥15 and <30 mL/min. ^BPT-INR and time in therapeutic range (TTR) values are applicable only to patients in the warfarin group. ^CDOAC type and dosage data are shown only for those using DOACs; identical values are presented in the All and DOAC columns. IQR, interquartile range. Other abbreviations as in Table 1.

Clinical Outcomes

The incidence of the composite outcome and its components (stroke or systemic embolism, and major bleeding during follow-up) individually in patients with moderate-to-severe renal impairment is presented in Table 3 and the Supplementary Table. The incidence of the composite outcome in the warfarin and DOACs groups was 14.71%/year (95% CI 9.3–23.3) and 15.36%/year (95% CI 8.7–27.0), respectively. The incidence of stroke or systemic embolism in the warfarin and DOACs groups was 3.18%/year (95% CI 1.2–8.5) and 3.50%/year (95% CI 0.6–9.9), respectively, whereas the incidence of major bleeding was 4.78%/year (95% CI 2.2–10.6) and 1.20%/year (95% CI 0.2–8.7), respectively. No significant differences were observed between the 2 groups for the composite outcome, stroke or systemic embolism, or major bleeding (log-rank test P=0.882, P=0.760, and P=0.161, respectively; Central Figure, Figures 1,2). Unadjusted HRs for the DOACs group compared with warfarin group were 1.06 (95% CI 0.51–2.20; P=0.881) for the composite outcome, 0.77 (95% CI 0.14–4.20; P=0.761) for stroke or systemic embolism, and 0.25 (95% CI 0.03–2.05; P=0.195) for major bleeding. Adjusted HRs for the composite outcome, stroke or systemic embolism, and major bleeding were 0.92 (95% CI 0.44–1.93; P=0.821), 0.72 (95% CI 0.13–3.97; P=0.702), and 0.25 (95% CI 0.03–2.13; P=0.206), respectively (Table 4).

Table 3.

Summary of Primary and Secondary Outcomes in Patients With Moderate-to-Severe Impairment of Renal Function Overall and According to Treatment Group

	All (n=170)		Warfarin (n=102)		DOAC (n=68)
	No. patients (%)	Event rateA (%/years)	No. patients (%)	Event rateA (%/years)	No. patients (%)	Event rateA (%/years)
Composite outcomeB	30 (17.7)	14.96 (10.5–21.4)	18 (17.7)	14.71 (9.3–23.3)	12 (17.7)	15.36 (8.7–27.0)
Other outcomes
Stroke or systemic embolism (efficacy)	6 (3.5)	2.90 (1.3–6.5)	4 (3.9)	3.18 (1.2–8.5)	2 (2.9)	3.50 (0.6–9.9)
Major bleeding (safety)	7 (4.1)	3.38 (1.6–7.1)	6 (5.9)	4.78 (2.2–10.6)	1 (1.5)	1.20 (0.2–8.7)

^AValues in parentheses are 95% confidence intervals. ^BThe composite outcome included stroke, systemic embolism, cardiovascular events (including heart failure requiring hospitalization) or cardiac death. DOAC, direct oral anticoagulant.

Figure 1.

Kaplan-Meier curves for stroke or systemic embolism in patients with moderate-to-severe renal impairment (15 mL/min≤creatinine clearance<30 mL/min) treated with warfarin or direct oral anticoagulants (DOACs).

Figure 2.

Kaplan-Meier curves for major bleeding in patients with moderate-to-severe renal impairment (15 mL/min ≤ creatinine clearance < 30 mL/min) treated with warfarin or direct oral anticoagulants (DOACs).

Table 4.

Cox Proportional Hazards Regression Models for Outcome Events in Patients With Moderate-to-Severe Impairment of Renal Function Overall and According to Treatment Group

	Univariate model		Multivariate modelA
	HR (95% Cl)	P value	HR (95% CI)	P value
Composite outcomeB
Warfarin	Ref.		Ref.
DOAC	1.06 (0.51–2.20)	0.881	0.92 (0.44–1.93)	0.821
Stroke/systematic embolism (efficacy)
Warfarin	Ref.		Ref.
DOAC	0.77 (0.14–4.20)	0.761	0.72 (0.13–3.97)	0.702
Major bleeding (safety)
Warfarin	Ref.		Ref.
DOAC	0.25 (0.03–2.05)	0.195	0.25 (0.03–2.13)	0.206

^AAdjusted for HELT-E₂S₂ score, antiplatelet use, malignant tumor, and dementia. ^BThe composite outcome included stroke, systemic embolism, cardiovascular events (including heart failure requiring hospitalization), or cardiac death. CI, confidence interval; DOAC, direct oral anticoagulant; HR, hazard ratio.

Discussion

Compared with the general older population, patients in the BPV-AF Registry had more complex clinical backgrounds, including a high prevalence of comorbidities such as renal dysfunction, heart failure, and frailty, along with polypharmacy, which complicates decisions regarding anticoagulant therapy. This study focused on renal function, comorbidities, and the number of prescribed medications in these patients, specifically comparing the efficacy and safety of DOACs to that of warfarin in patients with moderate-to-severe renal impairment. Although previous studies evaluated anticoagulant therapy in patients with BPV, the present study specifically investigated a clinically vulnerable subgroup, namely older patients with AF and moderate-to-severe renal impairment (15 mL/min ≤ CCr < 30 mL/min), in whom DOACs may be used cautiously in real-world clinical practice. By focusing on this specific population, our study provides additional insights into anticoagulant use in a setting where evidence remains scarce.

The primary findings of the present study are as follows. First, as renal function declined, CHADS₂, CHA₂DS₂-VASc, and HELT-E₂S₂ scores increased, along with the HAS-BLED score, suggesting elevated thromboembolic and bleeding risks in patients with impaired renal function. Second, despite these elevated risk scores, the number of comorbidities and the extent of polypharmacy did not differ significantly across renal function groups. Finally, in patients with moderate-to-severe renal impairment, those with higher embolic risk (as assessed by CHADS₂, CHA₂DS₂-VASc, and HELT-E₂S₂ scores) were more frequently prescribed DOACs. That these findings were observed in AF patients after BPV replacement suggests that clinicians tended to select DOACs for patients perceived to be at higher thromboembolic risk. Although no significant differences in clinical outcomes were observed between the warfarin and DOAC groups, the trend may inform real-world treatment strategies in this high-risk population.

Trends Based on Renal Function

In patients with AF after BPV replacement, CHADS₂ and CHA₂DS₂-VASc scores increased as renal function declined, similar to the trend observed for HELT-E₂S₂ scores. The HAS-BLED score also increased with worsening renal function, suggesting heightened risks of both thromboembolism and bleeding. Unlike CHADS₂, the HELT-E₂S₂ score does not include heart failure or diabetes, instead incorporating factors such as low BMI (<18.5 kg/m²) and AF type. Impaired renal function increases the risk of AF development due to the cardiovascular burden caused by anemia and electrolyte (e.g., potassium and magnesium) imbalances.^22,23 In addition, renal impairment contributes to coagulation activation through mechanisms such as chronic inflammation, oxidative stress, vascular stiffness, and endothelial dysfunction, similar to conditions seen with low body weight, endothelial dysfunction, platelet hyperactivity, and systemic inflammation.²⁴

Although renal impairment is frequently associated with multiple comorbidities in older populations,^25,26 we found no significant differences in the number of comorbidities according to renal function in the present study. This may reflect comparable distributions of common diseases such as hypertension, diabetes, and dementia across the renal function groups. Moreover, lifestyle factors such as smoking and alcohol use may have influenced the results. Despite the expectation that worsening renal function increases polypharmacy, the mean number of prescribed medications per patient was 3.0±1.0 across all renal function groups, with no significant differences. In contrast, the All Nippon AF in the Elderly (ANAFIE) registry reported a mean of 6.6±3.2 medications per patient, with approximately 60% of patients taking ≥5 medications.⁵ The lower rate of polypharmacy in the present study, as well as the absence of significant variation according to renal function, may be due to careful drug selection, prescription constraints for renally excreted medications, and close monitoring in this post-BPV replacement population.

The findings of the present study underscore the importance of individualized anticoagulant selection in patients with AF after BPV replacement, particularly those with moderate-to-severe renal impairment, where balancing efficacy and safety becomes more complex. Understanding these associations is essential for guiding appropriate anticoagulant therapy in this high-risk post-BPV population.

Comparison of DOACs and Warfarin in Patients With Moderate-to-Severe Renal Impairment

In patients with AF after BPV replacement, particularly those with moderate-to-severe renal impairment, DOACs were more frequently prescribed to those at higher thromboembolic risk (Table 2), whereas the incidence of thromboembolism and major bleeding did not differ significantly between the DOAC and warfarin groups. These findings suggest that even in the high-risk post-BPV setting, DOACs may be a viable alternative to warfarin, although careful patient selection remains essential. These findings are consistent with those of the ANAFIE registry subanalysis, a large-scale registry study of over 30,000 older Japanese patients with non-valvular AF,^27,28 further supporting our interpretation. In the ANAFIE registry subanalysis,²⁸ patients with impaired renal function had annual rates of stroke or systemic embolism of 3.6% with warfarin and 4.0% with DOACs (HR 0.89; P=0.541), whereas major bleeding occurred in 2.4% and 3.5% of patients, respectively (HR 0.67; P=0.065). DOACs were more frequently prescribed to higher-risk patients, as observed both in the ANAFIE registry²⁸ and in the present study, and the limited sample size reduces the statistical power to detect differences in bleeding outcomes. Therefore, the present study does not allow for valid statistical comparisons between DOACs and warfarin. Rather, it should be interpreted as a descriptive analysis of outcomes in a high-risk post-BPV replacement cohort with moderate-to-severe renal impairment.

Clinical Implications

Managing patients with both high thromboembolic and bleeding risks, such as those with AF after BPV replacement and moderate-to-severe renal impairment, remains a substantial clinical challenge. This study highlights the complexity of anticoagulant selection in this population, underscoring the importance of individualized risk assessment. Because renal function can progressively deteriorate over time, ongoing monitoring of renal parameters, periodic reassessment of bleeding risk, and flexible adjustment of anticoagulant therapy are essential to maintain optimal safety and efficacy in long-term management.

Study Limitations

This study has several limitations. First, the impact of impaired renal function on the pharmacokinetics of DOACs is significant. Because DOACs are partially excreted by the kidneys, their plasma concentrations can increase in patients with renal dysfunction, potentially increasing bleeding risk. Conversely, warfarin is primarily metabolized by the liver and is less affected by renal function, which may reduce the relative safety advantages of DOACs in this population. Second, the overall sample size for the group with moderate-to-severe renal impairment the present study was small (n=170), resulting in insufficient statistical power to detect meaningful differences, particularly for rare outcomes such as major bleeding. Third, patients in the DOAC group were significantly older and had higher CHA₂DS₂-VASc and HELT-E₂S₂ scores than those in the warfarin group, indicating higher baseline risk of both thromboembolism and bleeding. These differences may have offset the potential safety benefits of DOACs. Fourth, the time in therapeutic range for warfarin was well maintained, potentially improving warfarin safety and narrowing the risk gap between the DOACs and warfarin groups. Fifth, unmeasured confounders, such as nutritional status, albumin levels, and lifestyle factors, may have influenced outcomes. Indeed, poor nutritional status has recently been reported to be associated with increased bleeding risk in anticoagulated patients.⁸ These residual confounding factors could not be fully adjusted for. Large-scale cohort studies and subgroup analyses stratified by renal function are needed to confirm our findings.

Conclusions

This study reported clinical outcomes in patients with AF after BPV replacement and moderate-to-severe renal impairment treated with DOACs or warfarin. Although the observed event rates were comparable between the DOACs and warfarin groups, the limited sample size precludes valid statistical comparisons. These findings provide descriptive insights into anticoagulant therapy in this high-risk population and underscore the need for individualized treatment strategies and further large-scale studies.

Acknowledgments

The authors thank the staff and participants of the BPV-AF Registry for their significant contributions to this study. The authors also thank the National Cerebral and Cardiovascular Center for its collaboration and support. The authors also thank Daisuke Chiba, Kumiko Sugio, and Minako Oshima of Daiichi Sankyo Co., Ltd., for their assistance with the preparation of the manuscript. Their contributions were made in accordance with the Good Publication Practice (GPP3) guideline. The authors used an AI-based language model (Chat GPT, Open AI) to improve the clarity and language of the English text during the preparation of this manuscript.

Sources of Funding

This study was supported by Daiichi Sankyo Co., Ltd. (Tokyo, Japan) in collaboration with the National Cerebral and Cardiovascular Center.

Disclosures

Y.F. has received remuneration from Daiichi Sankyo Co., Ltd., Bayer Yakuhin Ltd., Ono Pharmaceutical Co., Ltd., and Novartis Pharma K.K. H.T. has received consultancy fees from AstraZeneca PLC and Ono Pharmaceutical Co., Ltd. K.A. has received remuneration from Japan Lifeline Co., Ltd., Terumo Corporation, and Medtronic Japan Co. Ltd. M. Izumo has received consultancy fees from Abbott Medical Japan LLC and remuneration from Edwards Lifesciences Corporation. T. Kimura has received research funding from Abbott Vascular and Boston Scientific Japan. K.K. Y. Sakata has received remuneration from Daiichi Sankyo Co., Ltd. and Nippon Boehringer Ingelheim Co., Ltd., and scholarship funding from Nippon Boehringer Ingelheim Co., Ltd., Bayer Yakuhin Ltd., and Daiichi Sankyo Co., Ltd. T. Kimura, M.F. is an employees of Daiichi Sankyo Co., Ltd. K.N. has received research funding from Philips Japan Ltd., Terumo Co., Ltd., TEPCO Power Grid Inc., and Asahi Kasei Pharma Co. H. Kanazawa has received support from an endowed department sponsored by Abbott Medical Japan Co., Ltd., Medtronic Japan Co., Ltd., Japan Lifeline Co., Ltd., Fukuda Denshi Co., Ltd., Boston Scientific Japan K.K., Biotronic Japan, Inc., Nipro Corporation, and Fides-one, Inc. K.T. has received remuneration from Amgen K.K., Bayer Yakuhin Ltd., Daiichi Sankyo Co., Ltd., Kowa Pharmaceutical Co. Ltd., Novartis Pharma K.K., Otsuka Pharmaceutical Co., Ltd., and Pfizer Japan Inc; research funding from AMI Co., Ltd., Bayer Yakuhin Ltd., Bristol-Myers Squibb K.K., EA Pharma Co., Ltd., and Mochida Pharmaceutical Co., Ltd.; scholarship funding from AMI Co., Ltd., Bayer Yakuhin Ltd., Nippon Boehringer Ingelheim Co., Ltd., Chugai Pharmaceutical Co., Ltd., Daiichi Sankyo Co., Ltd., Edwards Lifesciences Corporation, Johnson & Johnson K.K., Ono Pharmaceutical Co., Ltd., Otsuka Pharmaceutical Co., Ltd., and Takeda Pharmaceutical Co., Ltd.; and is affiliated with the endowed department sponsored by Abbott Japan Co., Ltd., Boston Scientific Japan K.K., Fides-one Inc., GM Medical Co., Ltd., ITI Co., Ltd., Kaneka Medix Co., Ltd., Nipro Corporation, Terumo Co., Ltd., Abbott Medical Co., Ltd., Cardinal Health Japan LLC., Fukuda Denshi Co., Ltd., Japan Lifeline Co., Ltd., Medical Appliance Co., Ltd., and Medtronic Japan Co., Ltd. C.I. has received remuneration from Daiichi Sankyo Co., Ltd., and research funding from Daiichi Sankyo Co., Ltd. C.I., T. Kimura, K.T., and Y. Sakata are members of Circulation Journal’s Editorial Team. K.T. is a member of Circulation Reports’ Editorial Team. The remaining authors have no disclosures to report.

IRB Information

The study protocol and informed consent documents were reviewed and approved by the Ethics Committee of the National Cerebral and Cardiovascular Center (Approval no. M30-068; September 26, 2018).

Data Availability

The datasets analyzed during the present study are not publicly available due to institutional and ethical restrictions, but are available from the corresponding author upon reasonable request.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circrep.CR-24-0156

References

• 1.   Committee for Scientific Affairs, The Japanese Association for Thoracic Surgery; Shimizu H, Endo S, Natsugoe S, Doki Y, Hirata Y, Kobayashi J, et al. Thoracic and cardiovascular surgery in Japan in 2016: Annual report by The Japanese Association for Thoracic Surgery. Gen Thorac Cardiovasc Surg 2019; 67: 377–411, doi:10.1007/s11748-019-01068-9.
• 2.   Hollenberg SM. Valvular heart disease in adults: Etiologies, classification, and diagnosis. FP Essent 2017; 457: 11–16.
• 3.   Ray S. Changing epidemiology and natural history of valvular heart disease. Clin Med (Lond) 2010; 10: 168–171, doi:10.7861/clinmedicine.10-2-168.
• 4.   Parikh K, Dizon J, Biviano A. Revisiting atrial fibrillation in the transcatheter aortic valve replacement era. Interv Cardiol Clin 2018; 7: 459–469, doi:10.1016/j.iccl.2018.06.001.
• 5.   Koretsune Y, Yamashita T, Akao M, Hirotsugu A, Ikeda T, Okumura K, et al. Baseline demographics and clinical characteristics in the All Nippon AF in the Elderly (ANAFIE) registry. Circ J 2019; 83: 1538–1545, doi:10.1253/circj.CJ-19-0094.
• 6.   Mannucci PM, Nobili A, Pasina L. Polypharmacy in older people: Lessons from 10 years of experience with the REPOSI Collaborators. Intern Emerg Med 2018; 13: 1191–1200, doi:10.1007/s11739-018-1941-8.
• 7.   Yamashita T, Akao M, Atarashi H, Ikeda T, Koretsune Y, Okumura K, et al. Effect of polypharmacy on clinical outcomes in elderly patients with non-valvular atrial fibrillation: A sub-analysis of the ANAFIE registry. Circ J 2022; 87: 6–16, doi:10.1253/circj.CJ-22-0170.
• 8.   Fujita T, Sugihara M, Mine K, Kato Y, Hirata T, Arimura T, et al. Controlled nutritional status scores and bleeding events in lower extremity atrial disease: 3-year clinical outcome study. Circ Rep 2025; 7: 677–684, doi:10.1253/circrep.CR-25-0076.
• 9.   Masunaga N, Ishii M, Oka K, Okamoto K, Yoshida Y, Minami K, et al. 10-year trends of antithrombotic therapy status and clinical outcomes in patients with atrial fibrillation and renal dysfunction: The Fushimi AF registry. Circ J 2025; 89: 174–183, doi:10.1253/circj.CJ-24-0614.
• 10.   Choi JI, Kiatchoosakun S, Jiampo P, Tse H, Soo Y, Wang C, et al. Prescribing patterns and outcomes of edoxaban in atrial fibrillation patients from Asia: One- year data from the Global ETNA-AF program. Circ Rep 2024; 6: 86–93, doi:10.1253/circrep.CR-23-0098.
• 11.   Kitahara H, Yamazaki T, Hiraga T, Suzuki S, Ohno Y, Harada J, et al. Impact of under dosing of direct oral anticoagulants on clinical outcomes in patients with atrial fibrillation undergoing percutaneous coronary intervention. Circ J 2025; 89: 195–203, doi:10.1253/circj.CJ-24-0418.
• 12.   Akao M, Inoue H, Yamashita T, Atarashi H, Ikeda T, Koretsune Y, et al. Relationship between direct oral anticoagulant doses and clinical outcomes in elderly patients with non-valvular atrial fibrillation: ANAFIE Registry sub-analysis. Circ J 2023; 87: 1550–1560, doi:10.1253/circj.CJ-23-0143.
• 13.   Furukawa Y, Miyake M, Fujita T, Koyama T, Takegami M, Kimura T, et al. Rationale, design, and baseline characteristics of the Bioprosthetic Valves with Atrial Fibrillation (BPV-AF) study. Cardiovasc Drugs Ther 2020; 34: 689–696, doi:10.1007/s10557-020-07038-1.
• 14.   Izumi C, Miyake M, Fujita T, Koyama T, Tanaka H, Ando K, et al. Antithrombotic therapy for patients with atrial fibrillation and bioprosthetic valves: Real-world data from the multicenter, prospective, observational BPV-AF Registry. Circ J 2022; 86: 440–448, doi:10.1253/circj.CJ-21-0564.
• 15.   Iwasaki Y, Noda T, Aoki M, Fujino T, Hirano T, Inoue K, et al. JCS/JHRS 2024 Guideline focused update on management of cardiac arrhythmias. Circ J 2025; 89: 1012–1073, doi:10.1253/circj.CJ-24-0073.
• 16.   Yamashita T, Akao M, Atarashi H, Ikeda T, Koretsune Y, Okumura K, et al. Causes of death in elderly patients with non-valvular atrial fibrillation: Results from the ANAFIE Registry. Circ J 2023; 87: 957–963, doi:10.1253/circj.CJ-22-0614.
• 17.   Okumura K, Tomita H, Nakai M, Kodani E, Akao M, Suzuki S, et al. A novel risk stratification system for ischemic stroke in Japanese patients with non-valvular atrial fibrillation. Circ J 2021; 85: 1254–1262, doi:10.1253/circj.CJ-20-1075.
• 18.   Okumura K, Tomita H, Nakai M, Kodani E, Akao M, Suzuki S, et al. Risk factors associated with ischemic stroke in Japanese patients with nonvalvular atrial fibrillation. JAMA Netw Open 2020; 3: e202881, doi:10.1001/jamanetworkopen.2020.2881.
• 19.   Nakano M, Kondo Y, Shiko Y, Nakano M, Kajiyama T, Ito R, et al. Validity of HELT-E₂S₂ score in predicting ischemic stroke in Japanese patients with non-valvular atrial fibrillation with cardiac implantable electronic device. Circ J 2025; 89: 765–773, doi:10.1253/circj.CJ-24-0715.
• 20.   Yamauchi T, Okumura Y, Nagashima K, Watanabe R, Saito Y, Yokoyama K, et al. External validation of the HELT-E₂S₂ score in Japanese patients with nonvalvular atrial fibrillation: A pooled analysis of the RAFFINE and SAKURA registries. Circ J 2023; 87: 1777–1787, doi:10.1253/circj.CJ-23-0318.
• 21.   Schulman S, Kearon C; Subcommittee on Control of Anticoagulation of the Scientific Standardization Committee of International Society on Thrombosis Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medical products in non-surgical patients. J Thromb Haemost 2005; 3: 692–694, doi:10.1111/j.1538-7836.2005.01204.x.
• 22.   Olesen JB, Lip GYH, Kamper AL, Hommel K, Køber L, Lane DA, et al. Stroke and bleeding in atrial fibrillation with chronic kidney disease. N Engl J Med 2012; 367: 625–635, doi:10.1056/NEJMoa1105594.
• 23.   Lau YC, Proietti M, Guiducci E, Blann AD, Lip GYH. Atrial fibrillation and thromboembolism in patients with chronic kidney disease. J Am Coll Cardiol 2016; 68: 1452–1464, doi:10.1016/j.jacc.2016.06.057.
• 24.   Hamatani Y, Ogawa H, Uozumi R, Iguchi M, Yamashita Y, Esato M, et al. Low body weight is associated with the incidence of stroke in atrial fibrillation patients: Insight from the Fushimi AF registry. Circ J 2015; 79: 1009–1017, doi:10.1253/circj.CJ-14-1245.
• 25.   Japanese Society of Nephrology. Management of CKD in the elderly. 2020. https://jsn.or.jp/guideline/ckd/elderly.php (accessed March 18, 2025).
• 26.   Japanese Society of Nephrology. Clinical practice guideline for prevention of CKD progression and transition to renal replacement therapy in patients with CKD stage G3b–5. 2017. https://jsn.or.jp/guideline/ckd/CKD2017.php (accessed March 18, 2025).
• 27.   Yamashita T, Suzuki S, Inoue H, Akao M, Atarashi H, Ikeda T, et al. Two-year outcomes of more than 30 000 elderly patients with atrial fibrillation: Results from the All Nippon AF In the Elderly (ANAFIE) Registry. Eur Heart J Qual Care Clin Outcomes 2022; 2: 202–213, doi:10.1093/ehjqcco/qcab025.
• 28.   Shimizu W, Yamashita T, Akao M, Atarashi H, Ikeda T, Koretsune Y, et al. Renal function and clinical outcomes among elderly patients with nonvalvular atrial fibrillation from ANAFIE. JACC Asia 2023; 3: 475–487, doi:10.1016/j.jacasi.2023.02.002.