Application of the Modified High Bleeding Risk Criteria for Japanese Patients in an All-Comers Registry of Percutaneous Coronary Intervention　― From the CREDO-Kyoto Registry Cohort-3 ―

Abstract

Background: The prevalence of and expected bleeding event rate in patients with the Japanese version of high bleeding risk (J-HBR) criteria are currently unknown in real-world percutaneous coronary intervention (PCI) practice.

Methods and Results: We applied the J-HBR criteria in the multicenter CREDO-Kyoto registry cohort-3 that enrolled 13,258 consecutive patients who underwent first PCI. The J-HBR criteria included Japanese-specific major criteria such as heart failure, low body weight, peripheral artery disease and frailty in addition to the Academic Research Consortium (ARC)-HBR criteria. There were 8,496 patients with J-HBR, and 4,762 patients without J-HBR. The J-HBR criteria identified a greater proportion of patients with HBR than did ARC-HBR (64% and 48%, respectively). Cumulative incidence of the Bleeding Academic Research Consortium (BARC) type 3 or 5 bleeding was significantly higher in the J-HBR group than in the no-HBR group (14.0% vs. 4.1% at 1 year; 23.1% vs. 8.4% at 5 years, P<0.0001). Cumulative 5-year incidence of BARC 3/5 bleeding was 25.1% in patients with ARC-HBR, and 23.1% in patients with J-HBR. Cumulative incidence of myocardial infarction or ischemic stroke was also significantly higher in the J-HBR group than in the no-HBR group (6.9% vs. 3.6% at 1 year; 13.2% vs. 7.1% at 5 years, P<0.0001).

Conclusions: The J-HBR criteria successfully identified those patients with very high bleeding risk after PCI, who represented 64% of patients in this all-comers registry.

The Academic Research Consortium criteria for high bleeding risk (ARC-HBR) were proposed from literature review and by consensus of experts from the USA, Asia, and Europe to standardize the definition of high bleeding risk (HBR).¹ In the ARC-HBR initiative, HBR was arbitrarily defined as Bleeding Academic Research Consortium (BARC) type 3 or 5 bleeding ≥4% at 1 year or a risk of an intracranial hemorrhage ≥1% at 1 year after percutaneous coronary intervention (PCI). However, trials referenced in the ARC-HBR criteria were conducted outside Japan and several Japanese-specific factors associated with HBR were not included.¹ Patient characteristics that are not included in the ARC-HBR criteria, such as low body weight (BW), frailty, heart failure (HF), and peripheral artery disease (PAD), have been reported to be independently associated with bleeding complications, particularly in Japanese.^2–4 Therefore, the bleeding risk factors commonly seen in Japanese patients were added to the ARC-HBR criteria and the Japanese version of the HBR (J-HBR) criteria has been proposed by consensus of the Working Group of the Guidelines in the Japanese Circulation Society.⁵ However, the prevalence of and the expected bleeding event rate in patients who fulfil the J-HBR criteria are currently unknown.

Editorial p ????

Therefore, we sought to apply the HBR criteria modified for Japanese patients in a large Japanese PCI database to explore the prevalence of and the bleeding event rates in patients who fulfilled the J-HBR criteria in real-world clinical practice.

Methods

Study Population

The Coronary Revascularization Demonstrating Outcome Study in Kyoto (CREDO-Kyoto) PCI/coronary artery bypass grafting (CABG) registry cohort-3 is a physician-initiated, non-company-sponsored, multicenter registry enrolling consecutive patients who underwent their first coronary revascularization with PCI or isolated CABG without combined non-coronary surgery among 22 centers in Japan between January 2011 and December 2013 after approval of the new-generation drug-eluting stent in 2010 (Supplementary Appendix A). The relevant ethics committees in all of the participating centers approved the study protocol. Because of the retrospective enrollment, written informed consent from patients were waived; however, we excluded those patients who refused participation in the study when contacted for follow-up, which was concordant with the guidelines of the Japanese Ministry of Health, Labor and Welfare.

Between January 2011 and December 2013, 14,927 patients had their first coronary revascularization (PCI: 13,307 patients; isolated CABG: 1,620 patients) in 22 centers in Japan. After excluding 60 patients who refused study participation and the 1,620 CABG patients (11 CABG patients refused study participation), the current study population consisted of 13,258 patients who underwent their first PCI (Figure 1). The study population was divided into the 2 groups (J-HBR, and no-HBR) according to the definition of the J-HBR criteria (Supplementary Table 1 and for the web calculator, see https://www.seiken-j.or.jp/CREDO-Kyoto.risk.score_J-HBR).⁵ The J-HBR criteria included the Japanese-specific major criteria such as HF, low BW, PAD, and frailty as well as the ARC-HBR major or minor criteria. Patients were considered to have HBR if at least 1 major or 2 minor J-HBR criteria were met. In the present analysis, several major and minor J-HBR criteria were not captured. Therefore, those patients with at least 1 major criterion such as severe chronic kidney disease (CKD), thrombocytopenia, severe anemia, liver cirrhosis, prior hemorrhagic stroke, active malignancy, anticoagulation, HF, low BW, PAD, and frailty and those with ≥2 minor criteria such as age ≥75 years, mild anemia, prior ischemic stroke, prior bleeding or moderate CKD were classified as the J-HBR group. Patients with only 1 minor criterion or without any criterion were classified as the no-HBR group (Figure 1). Active malignancy was defined as diagnosis within previous 12 months or ongoing requirement for treatment including surgery, chemotherapy, or radiotherapy according to the definition in the ARC-HBR.¹ Low BW was defined as <55 kg for men and <50 kg for women.⁵ Frailty was regarded as present when inability to perform normal activities of daily living was documented in the hospital charts. Therefore, patients with frailty were considered to have severe frailty, and mild or moderate frailty was not included in this study. The information on bleeding diathesis, brain arteriovenous malformation, nondeferrable major surgery on dual antiplatelet therapy (DAPT), and recent major trauma or surgery, use of non-steroidal anti-inflammatory drugs (NSAIDs) or steroids were not captured in this trial, and these criteria were regarded as absent.

Figure 1.

Study patient flowchart. ARC-HBR, Academic Research Consortium for High Bleeding Risk; BW, body weight; CABG, coronary artery bypass grafting; CREDO-Kyoto, the Coronary Revascularization Demonstrating Outcome study in Kyoto; CKD, chronic kidney disease; HBR, high bleeding risk; HF, heart failure; J-HBR, Japanese version of high bleeding risk; PAD, peripheral artery disease; PCI, percutaneous coronary intervention.

Definitions

The primary bleeding outcome measure was major bleeding defined as Bleeding Academic Research Consortium (BARC) type 3 or 5 bleeding,⁶ and the primary ischemic outcome measure was defined as a composite of myocardial infarction (MI) or ischemic stroke. Death was regarded as cardiac in origin unless obvious non-cardiac causes could be identified. Death of unknown cause and any death during the index hospitalization for coronary revascularization were regarded as cardiac death. Cardiovascular death included cardiac death, and other vascular deaths related to stroke, renal disease, and vascular disease. MI and stent thrombosis were adjudicated according to the ARC definition.⁷ To keep consistency with the CREDO-Kyoto PCI/CABG registry cohort-2,⁸ MI was also adjudicated according to the Arterial Revascularization Therapy Study (ARTS) definition,⁹ in which only Q-wave MI was regarded as MI when it occurred within 7 days of the index procedure. Stroke was defined as ischemic or hemorrhagic with neurological symptoms lasting >24 h. To keep consistency with the CREDO-Kyoto PCI/CABG registry cohort-2,⁸ bleeding was also adjudicated according to the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO) classification.¹⁰ Target vessel revascularization was defined as either PCI- or CABG-related to the target vessel. Any coronary revascularization was defined as either PCI or CABG for any reason. Scheduled staged coronary revascularization procedures performed within 3 months of the initial procedure were not regarded as follow-up events, but included in the index procedure. Duration of DAPT was left to the discretion of each attending physician. Persistent discontinuation of DAPT was defined as withdrawal of either thienopyridines or aspirin for at least 2 months. Other definitions are shown in Supplementary Methods.

Data Collection for Baseline Characteristics and Follow-up Events

Clinical, angiographic, and procedural data were collected from hospital charts or hospital databases according to prespecified definitions by experienced clinical research coordinators from an independent clinical research organization (Supplementary Appendix B). Follow-up data were collected from hospital charts and/or obtained by contacting patients, their relatives or referring physicians by mail with questions regarding vital status, subsequent hospitalizations, and status of antiplatelet therapy between January 2018 and December 2019. Follow-up was regarded as completed if follow-up data beyond July 1, 2017 were obtained. The clinical event committee independently adjudicated events such as death, MI, stent thrombosis, stroke, and major bleeding (Supplementary Appendix C).

Statistical Analysis

Categorical variables are presented as number and percentage, and compared with the chi-square test. Continuous variables are expressed as mean±standard deviation, or median and interquartile range. Based on their distribution, continuous variables were compared between groups with Student’s t test or the Wilcoxon rank sum test.

Cumulative incidence was estimated by the Kaplan-Meier method and differences between the J-HBR and no-HBR groups were assessed with the log-rank test. To distinguish the periprocedural events at the index PCI from the long-term events, we performed a landmark analysis at 30 days. Those patients with the individual endpoint events before 1 month were excluded in the landmark analysis beyond 30 days. We used univariable Cox proportional hazard models to estimate the effects of the J-HBR group relative to the no-HBR group for clinical events. The effect of the J-HBR group relative to the no-HBR group was expressed as a hazard ratio (HR) with 95% confidence intervals (CI). We did not make any statistical adjustment models, because J-HBR should be considered as an indicator variable for future events, and we summarized the relevant confounders.

Furthermore, the cumulative incidences of the primary bleeding outcome measure were assessed according to the presence or absence of the individual J-HBR criterion, in isolation. The cumulative incidences of the primary bleeding endpoint were also assessed according to the number of J-HBR major and minor criteria. Patients with J-HBR major criteria were included in the ≥3 majors, 2 majors or 1 major groups according to the number of major criteria, regardless of the number of overlapped minor criteria. Patients with ≥2 J-HBR minor criteria without major criteria were included in the ‘≥2 minors without major’ group. The cumulative incidences of the primary bleeding endpoint were also assessed in J-HBR patients aged ≥75 years and <75 years to evaluate the effect of J-HBR on bleeding events according to age. Sensitivity, specificity, and Youden Index for bleeding events at 5 years were assessed for ARC-HBR criteria and J-HBR criteria, individually.

Statistical analyses were conducted by a physician (M.N.) and a statistician (T.M.) using JMP 14.0 and SAS 9.2 (SAS Institute Inc., Cary, NC, USA) software. All the statistical analyses were 2-tailed. P values <0.05 were considered statistically significant.

Results

Prevalence of J-HBR and ARC-HBR Criteria

There were 8,496 patients in the J-HBR group, including 6,407 patients with ARC-HBR, and 4,762 patients in the no-HBR group. The J-HBR criteria identified a greater proportion of patients with HBR than did the ARC-HBR (64% and 48%, respectively). Numbers and proportion of patients with the J-HBR and ARC-HBR major or minor criteria are shown in Figure 1, Figure 2, and Table 1.

Figure 2.

Prevalence of the criteria included in J-HBR. Abbreviations as in Figure 1.

Table 1. Baseline Characteristics and Medications

	Entire cohort (n=13,258)	J-HBR (n=8,496)	No-HBR (n=4,762)	P value
Clinical characteristics
Age (years)	69.4±11.3	72.8±10.6	63.4±9.8	<0.0001
≥75 years	4,681 (35%)	4,182 (49%)	499 (10%)	<0.0001
Male	9,672 (73%)	5,741 (68%)	3,931 (83%)	<0.0001
Body mass index (kg/m2)	23.8±3.6 (13,074)	23.1±3.7 (8,347)	25.1±3.1 (4,727)	<0.0001
<25.0	8,901 (67%)	6,287 (74%)	2,614 (55%)	<0.0001
Low body weight (<55 kg for men, <50 kg for women)	3,017 (23%)	3,017 (36%)	0 (0%)	<0.0001
Acute MI	5,316 (40%)	3,376 (40%)	1,940 (41%)	0.26
Hypertension	10,900 (82%)	7,065 (83%)	3,835 (81%)	0.0002
Diabetes mellitus	5,039 (38%)	3,345 (39%)	1,694 (36%)	<0.0001
On insulin therapy	1,091 (8.2%)	837 (9.9%)	254 (5.3%)	<0.0001
Current smoking	3,672 (28%)	1937 (23%)	1,735 (36%)	<0.0001
Heart failure	3,117 (24%)	3,117 (37%)	0 (0%)	<0.0001
Multivessel disease	7,601 (57%)	5,165 (61%)	2,436 (51%)	<0.0001
LVEF (%)	58.6±13.0 (11,531)	56.9±14.0 (7,324)	61.6±10.5 (4,207)	<0.0001
≤40%	1,137/11,531 (9.9%)	1,010/7,324 (14%)	127/4,207 (3.0%)	<0.0001
Mitral regurgitation grade ≥3/4	894/11,700 (7.6%)	767/7,430 (10%)	127/4,270 (3.0%)	<0.0001
Prior MI	1,460 (11%)	1,038 (12%)	422 (8.9%)	<0.0001
Prior stroke	1,699 (13%)	1,557 (18%)	142 (3.0%)	<0.0001
Prior hemorrhagic stroke	304 (2.3%)	304 (3.6%)	0 (0%)	<0.0001
Prior ischemic stroke	1,466 (11%)	1,324 (16%)	142 (3.0%)	<0.0001
PAD	1,213 (9.2%)	1,213 (14%)	0 (0%)	<0.0001
Atrial fibrillation	1,286 (9.7%)	1,184 (14%)	102 (2.1%)	<0.0001
COPD	519 (3.9%)	404 (4.8%)	115 (2.4%)	<0.0001
Liver cirrhosis	335 (2.5%)	335 (3.9%)	0 (0%)	<0.0001
Malignancy	1,677 (13%)	1,351 (16%)	336 (6.9%)	<0.0001
Active malignancy	270 (2.0%)	270 (3.2%)	0 (0%)	<0.0001
Severe frailty†	550 (4.1%)	550 (6.5%)	0 (0%)	<0.0001
Moderate CKD (eGFR 30–59 mL/min/1.73 m2)	4,175 (31%)	3,455 (41%)	720 (15%)	<0.0001
Severe CKD	1,197 (9.0%)	1,197 (14%)	0 (0%)	<0.0001
eGFR <30 mL/min/1.73 m2, not on dialysis	586 (4.4%)	586 (6.9%)	0 (0%)	<0.0001
Dialysis	611 (4.6%)	611 (7.2%)	0 (0%)	<0.0001
Severe anemia (Hb <11 g/dL)	1,632 (12%)	1,632 (19%)	0 (0%)	<0.0001
Mild anemia (Hb 11–12.9 g/dL for men and 11–12.9 g/dL for women)	2,697 (20%)	2,422 (29%)	275 (5.8%)	<0.0001
Thrombocytopenia (Platelet <100×109/L)	262 (2.0%)	262 (3.1%)	0 (0%)	<0.0001
Prior bleeding	446 (3.4%)	400 (4.7%)	46 (1.0%)	<0.0001
Procedural characteristics
No. of target lesions	1.48±0.78	1.48±0.79	1.47±0.78	0.27
Target of proximal LAD	7,999 (60%)	5,075 (60%)	2,924 (61%)	0.06
Target of unprotected LMCA	591 (4.5%)	448 (5.3%)	143 (3.0%)	<0.0001
Target of chronic total occlusion	1,331 (10%)	867 (10%)	464 (9.7%)	0.4
Target of bifurcation	5,242 (40%)	3,304 (39%)	1,938 (41%)	0.04
Side-branch stenting	524 (4.0%)	334 (3.9%)	190 (4.0%)	0.87
Total number of stents	1 (1–2) 1.91±1.33 (12,691)	1 (1–2) 1.95±1.35 (8,065)	1 (1–2) 1.85±1.3 (4,626)	<0.0001
Total stent length (mm)	28 (18–52) 41.3±33.1 (12,689)	30 (18–54) 42.2±33.5 (8,063)	28 (18–50) 39.8±32.4 (4,626)	<0.0001
Total stent length >28 mm	6,223/12,689 (49%)	4,116/8,063 (51%)	2,107/4,626 (46%)	<0.0001
Minimum stent size (mm)	2.86±0.44 (12,690)	2.84±0.43 (8,064)	2.91±0.46 (4,626)	<0.0001
Minimum stent size <3.0 mm	6,671/12,690 (53%)	4,433/8,064 (55%)	2,238/4,626 (48%)	<0.0001
Stent use	12,691 (96%)	8,065 (95%)	4,626 (97%)	<0.0001
DES use	10,334 (78%)	6,537 (77%)	3,797 (80%)	0.0001
New-generation DES use	10,183 (77%)	6,429 (76%)	3,754 (79%)	<0.0001
IVUS or OCT use	9,738 (73%)	6,160 (73%)	3,578 (75%)	0.001
IVUS use	9,629 (73%)	6,088 (72%)	3,541 (74%)	0.0008
OCT use	288 (2.2%)	171 (2.0%)	117 (2.5%)	0.09
Transradial approach	4,973/13,243 (38%)	2,886/8,487 (34%)	2,087/4,756 (44%)	<0.0001
Transfemoral approach	6,987/13,243 (53%)	4,675/8,487 (55%)	2,312/4,756 (49%)
Transbrachial approach	1,283/13,243 (9.7%)	926/8,487 (11%)	357/4,756 (7.5%)
Staged PCI	2,650 (20%)	1,628 (19%)	1,022 (21%)	0.002
Baseline medications
Antiplatelet therapy
Thienopyridine	13,043 (98.4%)	8,313 (97.9%)	4,730 (99.3%)	<0.0001
Ticlopidine	358/13,043 (2.7%)	255/8,313 (3.1%)	103/4,730 (2.2%)	<0.0001
Clopidogrel	12,571/13,043 (96.4%)	8,016/8,313 (96.4%)	4,555/4,730 (96.3%)
Others	114/13,043 (0.9%)	42/8,313 (0.5%)	72/4,730 (1.5%)
Aspirin	13,110 (98.9%)	8,363 (98.4%)	4,747 (99.7%)	<0.0001
Cilostazole	357 (2.7%)	289 (3.4%)	68 (1.4%)	<0.0001
Other medications
Statins	10,268 (77%)	6,130 (72%)	4,138 (87%)	<0.0001
High-intensity statins‡	211 (1.6%)	112 (1.3%)	99 (2.1%)	0.001
β-blockers	5,204 (39%)	3,531 (42%)	1,673 (35%)	<0.0001
ACEI/ARB	8,527 (64%)	5,414 (64%)	3,113 (65%)	0.06
Nitrates	2,563 (19%)	1,615 (19%)	948 (20%)	0.21
Calcium-channel blockers	5,220 (39%)	3,372 (40%)	1,848 (39%)	0.32
Nicorandil	2,220 (17%)	1,513 (18%)	707 (15%)	<0.0001
Oral anticoagulants	1,404 (11%)	1,404 (17%)	0 (0%)	<0.0001
Warfarin	1,223 (9.2%)	1,223 (14%)	0 (0%)	<0.0001
DOAC	183 (1.4%)	183 (2.2%)	0 (0%)	<0.0001
PPI or H2-blockers	10,234 (77%)	6,635 (78%)	3,599 (76%)	0.001
PPI	8,648 (65%)	5,619 (66%)	3,029 (64%)	0.003
H2-blockers	1,659 (13%)	1,068 (13%)	591 (12%)	0.79

Continuous variables are expressed as mean±standard deviation, or median (interquartile range). Categorical variables are expressed as number (percentage). Values were missing for body mass index in 184 patients, for LVEF in 1,727 patients, for mitral regurgitation in 1,558 patients, for eGFR in 44 patients, for hemoglobin level in 40 patients, and for platelet count in 45 patients; however the numbers of missing values for body mass index, eGFR, hemoglobin level, and platelet count were negligible. The missing values for these variables were imputed as “normal” in the binary classification, because data would have been available if abnormalities were suspected. On the other hand, the missing values for LVEF, and mitral regurgitation were not imputed in the categorical classification, because the numbers of missing values were substantial. Values were missing for stent length in 2 patients and for stent size in 1 patient. Patients with no-stent and missing values for stent length and stent size were classified as follows: total stent length ≤28 mm and minimal stent size ≥3.0 mm. ^†Severe frailty was regarded as present when the inability to perform usual activities of daily living was documented in the hospital chart. ^‡High-intensity statin therapy was defined as statin doses ≥atorvastatin 20 mg, pitavastatin 4 mg, or rosuvastatin 10 mg.²⁹ ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin II receptor blocker; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; DES, drug-eluting stents; DOAC, direct oral anticoagulants; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; HBR, high bleeding risk; IVUS, intravascular ultrasound; J-HBR, Japanese version of high bleeding risk; LAD, left anterior descending coronary artery; LMCA, left main coronary artery; LVEF, left ventricular ejection fraction; MI, myocardial infarction; OCT, optical coherence tomography; PAD, peripheral artery disease; PCI, percutaneous coronary intervention; PPI, proton pump inhibitor.

Baseline Characteristics

The study population reflected a typical Japanese PCI population including large proportions of patients with advanced age (mean 69.4 years), male sex (73%), and diabetes (38%); 40% of patients had acute MI presentation at index PCI (Table 1).

Patient in the J-HBR group were much older and had smaller body mass index than those in the no-HBR group. Patients in the J-HBR group more often had hypertension, diabetes, multivessel disease, prior MI, prior stroke, atrial fibrillation and chronic obstructive pulmonary disease than those in the no-HBR group, although there were fewer males and current smokers (Table 1).

In terms of procedural characteristics, the total number of stents was larger, the total stent length was longer and the minimum stent size was smaller in the J-HBR group than in the no-HBR group. The transradial approach was less often selected in the J-HBR group than in the no-HBR group. Regarding medications, patients in the J-HBR group less often received statins and more often received proton pump inhibitors or H₂-blockers than those in the no-HBR group (Table 1).

Long-Term Clinical Outcomes

Median follow-up duration of the survivors was 6.0 (interquartile range: 5.1–6.8) years. Cumulative incidence of persistent discontinuation of DAPT was significantly, but only slightly, higher in the J-HBR group than in the no-HBR group, indicating that DAPT duration was significantly shorter in the J-HBR group than in the no-HBR group (Supplementary Figure 1). Cumulative incidence of the primary bleeding outcome measure (BARC type 3 or 5 bleeding) was significantly higher in the J-HBR group than in the no-HBR group (14.0% vs. 4.1% at 1 year; 23.1% vs. 8.4% at 5 years, P<0.0001) (Table 2, Figure 3A). Patients with J-HBR had a major bleeding risk higher than ≥4% at 1 year by the definition of ARC-HBR.¹ The cumulative incidence of intracranial bleeding was also significantly higher in the J-HBR group than in the no-HBR group (1.3% vs. 0.3% at 1 year; 3.7% vs. 1.0% at 5 years, P<0.0001) (Table 2). The cumulative incidence of the primary ischemic outcome measure was also significantly higher in the J-HBR group than in the no-HBR group (6.9% vs. 3.6% at 1 year, 13.2% vs. 7.1% at 5 years, P<0.0001) (Table 2, Figure 3B).

Table 2. Clinical Outcomes

	No. of patients with event (cumulative 5 years incidence)		Univariable
	J-HBR (n=8,496)	No-HBR (n=4,762)	HR (95% CI)	P value
Major bleeding
BARC type 3 or 5	1,916 (23.1%)	437 (8.4%)	2.95 (2.66–3.27)	<0.0001
Gastrointestinal	546 (7.5%)	146 (3.0%)	2.58 (2.15–3.1)	<0.0001
Access site	340 (4.2%)	68 (1.4%)	2.94 (2.26–3.81)	<0.0001
Intracranial	269 (3.7%)	54 (1.0%)	3.65 (2.72–4.89)	<0.0001
Other	761 (9.9%)	169 (3.2%)	3.04 (2.57–3.59)	<0.0001
BARC type 5	107 (1.4%)	15 (0.3%)	4.68 (2.73–8.04)	<0.0001
GUSTO moderate/severe bleeding	1,699 (20.5%)	301 (5.5%)	3.82 (3.38–4.32)	<0.0001
GUSTO severe bleeding	795 (9.7%)	158 (2.8%)	3.33 (2.81–3.95)	<0.0001
Hemorrhagic stroke	200 (2.4%)	36 (0.6%)	3.79 (2.66–5.41)	<0.0001
MI†/ ischemic stroke	1,112 (13.2%)	392 (7.1%)	1.87 (1.66–2.09)	<0.0001
MI
ARC definition	640 (7.6%)	264 (5.0%)	1.55 (1.34–1.78)	<0.0001
ARTS definition	503 (6.0%)	208 (3.8%)	1.59 (1.35–1.87)	<0.0001
Ischemic stroke	531 (6.4%)	139 (2.3%)	2.57 (2.13–3.1)	<0.0001
All-cause death	2,390 (24.9%)	286 (4.4%)	5.64 (4.99–6.37)	<0.0001
Cardiac death	1,121 (12.7%)	85 (1.3%)	8.62 (6.91–10.7)	<0.0001
Cardiovascular death	1,319 (14.7%)	101 (1.6%)	8.61 (7.03–10.5)	<0.0001
Non-cardiovascular death	1,071 (12.0%)	185 (2.9%)	4.01 (3.43–4.69)	<0.0001
Definite stent thrombosis	77 (0.9%)	38 (0.7%)	1.28 (0.87–1.89)	0.22
TVR	1,543 (19.7%)	883 (17.4%)	1.13 (1.04–1.23)	0.004
Any coronary revascularization	2,216 (28.3%)	1,299 (25.8%)	1.11 (1.04–1.19)	0.002

Number of patients with event was counted until the end of follow-up. Cumulative 5-year incidence was estimated by the Kaplan-Meier method. HR with 95% CI of the J-HBR group relative to the no-HBR group for the outcome measures was estimated throughout the entire follow-up period by univariate Cox proportional hazard models. ^†MI as a component of the composite outcome measure was adjudicated according to the ARC definition.⁷ ARC, Academic Research Consortium; ARTS, Arterial Revascularization Therapy Study; BARC, Bleeding Academic Research Consortium; CI, confidence interval; GUSTO, Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries; HBR, high bleeding risk; HR, hazard ratio; J-HBR, Japanese version of high bleeding risk; MI, myocardial infarction; TVR, target vessel revascularization.

Figure 3.

Kaplan-Meier curves for (A) the primary bleeding and (B) ischemic outcome measures: J-HBR vs. No-HBR. The primary bleeding outcome measure was major bleeding defined as BARC type 3 or 5 bleeding, while the primary ischemic outcome measure was defined as a composite of myocardial infarction or ischemic stroke. BARC, Bleeding Academic Research Consortium; HBR, high bleeding risk; J-HBR, Japanese version of high bleeding risk; PCI, percutaneous coronary intervention.

Landmark Analysis Within and Beyond 30 Days

In the 30-day landmark analysis, the cumulative incidence of the primary bleeding outcome measure was significantly higher in the J-HBR group than in the no-HBR group both at 30 days (8.6% vs. 2.3%, P<0.0001), and beyond 30 days (5.9% vs. 1.9% at 1 year; 15.8% vs. 6.2% at 5 years, P<0.0001) (Table 3, Figure 4A). Patients with J-HBR had major bleeding risk ≥4% at 1 year even after excluding bleeding events within 30 days.

Table 3. Clinical Outcomes Within and Beyond 30 Days

	No. of patients with events (cumulative incidence)		Univariable
	J-HBR	No-HBR	HR (95% CI)	P value
Within 30 days	(n=8,496)	(n=4,762)
Major bleeding
BARC type 3 or 5	728 (8.6%)	109 (2.3%)	3.87 (3.16–4.73)	<0.0001
Access site bleeding	299 (3.6%)	60 (1.3%)	2.85 (2.16–3.77)	<0.0001
Gastrointestinal bleeding	126 (1.6%)	6 (0.1%)	12.4 (5.45–28.0)	<0.0001
Intracranial bleeding	25 (0.3%)	2 (0.04%)	7.36 (1.74–31.1)	0.001
Other	278 (3.4%)	41 (0.9%)	3.94 (2.84–5.48)	<0.0001
BARC type 5	15 (0.2%)	4 (0.08%)	2.12 (0.7–6.39)	0.17
GUSTO moderate/severe bleeding	574 (6.8%)	35 (0.7%)	9.51 (6.76–13.4)	<0.0001
GUSTO severe bleeding	212 (2.5%)	20 (0.4%)	6.05 (3.82–9.56)	<0.0001
Hemorrhagic stroke	15 (0.2%)	1 (0.02%)	8.56 (1.13–64.8)	0.01
MI†/ ischemic stroke	366 (4.4%)	123 (2.6%)	1.7 (1.39–2.08)	<0.0001
MI
ARC definition	275 (3.3%)	110 (2.3%)	1.42 (1.14–1.77)	0.002
ARTS definition	125 (1.5%)	51 (1.1%)	1.4 (1.01–1.93)	0.04
Ischemic stroke	96 (1.2%)	14 (0.3%)	3.92 (2.24–6.87)	<0.0001
All-cause death	310 (3.7%)	9 (0.2%)	19.6 (10.1–38.1)	<0.0001
Cardiac death	300 (3.5%)	9 (0.2%)	19.0 (9.79–36.9)	<0.0001
Cardiovascular death	301 (3.6%)	9 (0.2%)	19.1 (9.82–37.0)	<0.0001
Non-cardiovascular death	9 (0.1%)	0 (0%)	NA	0.02
Definite stent thrombosis	40 (0.5%)	16 (0.3%)	1.42 (0.8–2.54)	0.23
TVR	175 (2.1%)	56 (1.2%)	1.79 (1.33–2.42)	0.0001
Any coronary revascularization	214 (2.6%)	67 (1.4%)	1.84 (1.4–2.42)	<0.0001
Beyond 30 days
Major bleeding
BARC type 3 or 5	1,188/7,523 (15.8%)	328/4,632 (6.2%)	2.63 (2.32–2.97)	<0.0001
Access site bleeding	41/7,523 (0.6%)	8/4,632 (0.1%)	3.58 (1.68–7.65)	0.0004
Gastrointestinal bleeding	420/7,523 (6.0%)	140/4,632 (2.9%)	2.15 (1.77–2.6)	<0.0001
Intracranial bleeding	244/7,523 (3.4%)	52/4,632 (0.9%)	3.5 (2.59–4.72)	<0.0001
Other	483/7,523 (6.7%)	128/4,632 (2.4%)	2.74 (2.26–3.33)	<0.0001
BARC type 5	92/8,121 (1.2%)	11/4,735 (0.2%)	5.64 (3.02–10.5)	<0.0001
GUSTO moderate/severe bleeding	1,125/7,676 (14.7%)	266/4,706 (4.8%)	3.05 (2.67–3.48)	<0.0001
GUSTO severe bleeding	583/7,977 (7.4%)	138/4,721 (2.4%)	2.92 (2.43–3.52)	<0.0001
Hemorrhagic stroke	185/8,110 (2.2%)	35/4,735 (0.6%)	3.65 (2.54–5.24)	<0.0001
MI†/ ischemic stroke	746/7,784 (9.3%)	269/4,614 (4.6%)	1.95 (1.69–2.24)	<0.0001
MI
ARC definition	365/7,862 (4.5%)	154/4,627 (2.7%)	1.64 (1.36–1.98)	<0.0001
ARTS definition	378/8,010 (4.5%)	157/4,686 (2.7%)	1.66 (1.37–2.0)	<0.0001
Ischemic stroke	435/8,038 (5.3%)	125/4,721 (2.0%)	2.41 (1.97–2.94)	<0.0001
All-cause death	2,080/8,119 (22.1%)	277/4,735 (4.2%)	5.17 (4.56–5.87)	<0.0001
Cardiac death	821/8,119 (9.5%)	76/4,735 (1.1%)	7.38 (5.83–9.33)	<0.0001
Cardiovascular death	1,018/8,119 (11.6%)	92/4,735 (1.4%)	7.57 (6.11–9.37)	<0.0001
Non-cardiovascular death	1,062/8,119 (11.9%)	185/4,735 (2.9%)	3.98 (3.41–4.65)	<0.0001
Definite stent thrombosis	37/8,085 (0.5%)	22/4,719 (0.3%)	1.16 (0.69–1.97)	0.57
TVR	1,368/7,957 (17.9%)	827/4,680 (16.4%)	1.08 (0.99–1.18)	0.07
Any coronary revascularization	2,002/7,924 (26.4%)	1,232/4,669 (24.7%)	1.07 (0.99–1.15)	0.053

In the landmark analysis beyond 30 days, the number of patients with event was counted until the end of follow-up, while the cumulative incidence was estimated by the Kaplan-Meier method between 30 days and 5 years. HR with 95% CI of the J-HBR group relative to the no-HBR group for the outcome measures was estimated throughout the entire follow-up period by univariate Cox proportional hazard models. ^†Myocardial infarction as a component of the composite outcome measure was adjudicated according to the ARC definition.⁷ Abbreviations as in Table 2.

Figure 4.

Landmark analysis within and beyond 30 days. Kaplan-Meier curves (A) for the primary bleeding outcome measure, and (B) for the primary ischemic outcome measure. The primary bleeding outcome measure was major bleeding defined as BARC type 3 or 5 bleeding, while the primary ischemic outcome measure was defined as a composite of myocardial infarction or ischemic stroke. Abbreviations as in Figure 3.

The cumulative incidence of the primary ischemic outcome measure was also significantly higher in the J-HBR group than in the no-HBR group both at 30 days (4.4% vs. 2.6%, P<0.0001), and beyond 30 days (2.6% vs. 1.1% at 1 year; 9.3% vs. 4.6% at 5 years, P<0.0001) (Figure 4B).

Effects of the Individual J-HBR Criteria on the Primary Bleeding Endpoint

J-HBR major criteria such as active malignancy, severe CKD, severe anemia, liver cirrhosis, oral anticoagulation, prior hemorrhagic stroke, HF, PAD and low BW, in isolation, with the exception of thrombocytopenia and severe frailty, were associated with major bleeding risk ≥4% at 1 year. Furthermore, J-HBR minor criteria such as prior bleeding, age ≥75 years, mild anemia, and moderate CKD, in isolation, with the exception of prior ischemic stroke, was also associated with major bleeding risk ≥4% at 1 year (Supplementary Figure 2A).

Overall population with each J-HBR major or minor criterion were also associated with major bleeding risk ≥4% at 1 year, and with bleeding rate higher than those without that particular J-HBR major or minor criterion (Supplementary Figure 2B).

Effects of the Number of J-HBR Major and Minor Criteria on the Primary Bleeding Endpoint

Cumulative 5-year incidence of the primary bleeding outcome measure increased incrementally as the number of J-HBR major criteria increased (≥3 majors: 37.4%; 2 majors: 28.6%; 1 major: 18.7%; ≥2 minors: 12.2%; and no-HBR: 8.4%, P<0.0001) (Figure 5).

Figure 5.

Kaplan-Meier curves for the primary bleeding outcome measure according to the number of J-HBR major or minor criteria. The primary bleeding outcome measure was major bleeding defined as BARC type 3 or 5 bleeding. Abbreviations as in Figure 3.

Effects of the J-HBR Criteria According to Age on the Primary Bleeding Endpoint

Cumulative incidence of the primary bleeding outcome measure was significantly higher in both the J-HBR group with age ≥75 and age <75 than in the no-HBR group (15.3%, 12.8% vs. 4.1% at 1 year; 25.0%, 21.4% vs. 8.4% at 5 years, P<0.0001), indicating that the bleeding event rate was substantial even in patients with J-HBR and age <75 years (Supplementary Figure 3).

Clinical Outcomes in Patients With ARC-HBR Criteria and J-HBR Criteria

Cumulative 5-year incidence of the primary bleeding outcome measure was 25.1% in patients with ARC-HBR, and 23.1% in patients with J-HBR, whereas the cumulative 5-year incidence of the primary ischemic outcome measure was 15.1% in patients with ARC-HBR, and 13.2% in patients with J-HBR (Supplementary Table 2, Supplementary Figure 4). The sensitivity for the estimation of BARC 3/5 bleeding was 66.2% by the ARC-HBR criteria, and 82.3% by the J-HBR criteria, and the specificity was 62.1% by the ARC-HBR criteria, and 45.3% by the J-HBR criteria (Supplementary Table 3).

Discussion

The main findings of this study were as follows: (1) the J-HBR criteria identified a greater proportion of patients with HBR than the ARC-HBR (64% and 48%, respectively); and (2) cumulative 5-year incidence of the primary bleeding outcome measure remained high in patients with J-HBR, which was numerically similar to that in patients with ARC-HBR (25.1% and 23.1%, respectively).

In the CREDO-Kyoto registry cohort-2, 43% of patients had ARC-HBR with major bleeding risk higher than 4% at 1 year.¹¹ In the PENDULUM (PrasugrEl moNotherapy after DrUg eLUting stent deployMent as a Management Of patients who are uNsuitable for lOng-term dual antiplatelet therapy) registry enrolling Japanese patients undergoing PCI between December 2015 and June 2017, approximately half of the patients met the ARC-HBR criteria and had a higher incidence of major bleeding events.¹² In line with those reports, 48% of patients had ARC-HBR and were associated with a higher risk for major bleeding events in this registry. Furthermore, the presence of each ARC-HBR major or even minor criterion, in isolation, with the exception of thrombocytopenia and prior ischemic stroke, was associated with major bleeding risk higher than 4% at 1 year. Therefore, the ARC-HBR criteria are considered to be applicable to Japanese patients undergoing PCI. We could not capture information on bleeding diathesis, brain arteriovenous malformation, planned surgery on DAPT, recent major trauma or surgery, use of NSAIDs or steroids in this study. In the PENDULUM registry, prevalence of the use of NSAIDs or steroids was 13% in the HBR group. The 1-year incidence of major bleeding in those with NSAIDs or steroids was 4.9%, which was the highest incidence among the ARC-HBR minor criteria.¹² In a large-scale single-center registry, the prevalence of planned surgery was 8.2% in the HBR group. The 1-year incidence of major bleeding in patients with planned surgery was 8.9% overall and 4.8% in isolation.¹³ When we consider the missing information of several criteria, the prevalence of ARC-HBR patients could have been underestimated in this study.

In addition to the ARC-HBR criteria, HF, PAD, low BW, and frailty were included in the J-HBR as major criteria. HF was included as a predictor of the CRUSADE (Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the ACC/AHA guidelines) bleeding score that evaluated in-hospital major bleeding in patients with non-ST-segment elevation MI.¹⁴ PAD was also included as a predictor for bleeding events in the CRUSADE bleeding score.¹⁴ In the DAPT (Dual Antiplatelet Therapy) study, PAD was associated with the independent risk factors for moderate or severe bleeding.¹⁵ HF and PAD were also included as risk factors for major bleeding in the CREDO-Kyoto risk score.² In line with previous reports, HF and PAD, in isolation, were associated with major bleeding risk higher than 4% (≈10%) at 1 year in this study. Regarding low BW, previous studies have shown that underweight patients have an increased event rate for bleeding among those undergoing PCI.^16,17 Low body mass index was also included as a component of the PARIS (Patterns of Non-Adherence to Anti-Platelet Regimen in Stented Patients) bleeding score.¹⁸ In the PENDULUM registry, low BW was one of the independent predictors of major bleeding events.¹² In line with previous reports, low BW, in isolation, was associated with major bleeding risk higher than 4% at 1 year in this study. Frailty is also an important consideration for bleeding risk as a result of more frequent falls, or the inability to ambulate without assistance. Frailty was not included as a criterion of ARC-HBR because of the paucity of data and the lack of consensus on how frailty is best assessed.^1,19 In the ACTION (Acute Coronary Treatment and Intervention Outcomes Network) registry, frail patients are reported to have had a higher risk of major bleeding than non-frail patients when catheterization was performed.²⁰ In the present study, there were only 9 patients with severe frailty, in isolation, indicating that the vast majority of frail patients also had other HBR criteria. Indeed, overall the patients with severe frailty had much higher bleeding risk than those without.

Bleeding events are reported to be associated with subsequent mortality risk at least comparable to that of MI.²¹ The current guidelines recommend firstly evaluating HBR when considering the intensity of antithrombotic therapy, DAPT duration in particular, in patients undergoing PCI.^5,22 The rates of major bleeding in HBR patients were extremely high in clinical trials and in the present real-world registry.^11,23 Therefore, de-escalation of the antithrombotic regimen would be mandatory in HBR patients. Indeed, several recent randomized trials evaluating short DAPT showed it significantly reduced the bleeding risk without increasing the ischemic events.^24–26 In the STOPDAPT (ShorT and OPtimal duration of Dual AntiPlatelet Therapy)-2 trial comparing 1-month DAPT followed by clopidogrel monotherapy with 12-month DAPT, clopidogrel monotherapy was associated with a lower incidence of major bleeding than clopidogrel plus aspirin without increasing the cardiovascular events.²⁴ The benefit of 1-month DAPT in reducing major bleeding was numerically greater in HBR patients.²⁷ In the TWILIGHT (Ticagrelor with Aspirin or Alone in High-Risk Patients after Coronary Intervention) trial, patients with a higher risk of thrombotic or bleeding events were randomly assigned to a group that discontinued aspirin after 3-month DAPT with ticagrelor plus aspirin and a group that continued DAPT. The 3-month DAPT group showed a significantly lower incidence of bleeding events and had a comparable incidence of ischemic events.²⁵ In the TICO (Ticagrelor Monotherapy After 3 Months in the Patients Treated With New Generation Sirolimus-eluting Stent for Acute Coronary Syndrome) trial, ticagrelor monotherapy after 3 months of DAPT, compared with ticagrelor-based 12-month DAPT, resulted in a significant reduction in the composite outcome of major bleeding and cardiovascular events at 1 year in patients with acute coronary syndrome.²⁶ The ARC-HBR initiative selected a series of bleeding risk factors as the ARC-HBR criteria by consensus of experts, but did not cover all the bleeding risk factors. We proposed the J-HBR criteria included 4 additional bleeding risk factors that are commonly found in Japanese patients. The J-HBR criteria identified a greater proportion of patients with HBR than did the ARC-HBR, while the rate of major bleeding remained high in patients with J-HBR, which was numerically similar to that in patients with ARC-HBR. The J-HBR criteria might be clinically more useful than the ARC-HBR criteria because the former captured HBR patients more widely than the latter, despite the cost of somewhat less specificity.

Study Limitations

Some limitations to our study should be considered. First, because the CREDO-Kyoto PCI/CABG registry cohort-3 was not designed to investigate the performance of the ARC-HBC or J-HBR criteria, data were not available for some HBR criteria. Therefore, the prevalence of HBR patients was underestimated in this study. Second, we cannot deny ascertainment bias for severe frailty, because the prevalence of severe frailty in the present study was apparently lower than in previous studies.²⁸ Furthermore, due to the retrospective study design, we could not assess other important factors such as moderate frailty and cognitive impairment. Third, we cannot deny potential selection bias. In the decision making for PCI, patients with HBR might more often be averted from proceeding to PCI than those with no-HBR. Survival bias might also dilute the results of major bleeding. However, such excluded patients could be those with much severe conditions and the effect of J-HBR could be bigger if they were included; therefore, the findings can be considered as robust. Fourth, any type of liver cirrhosis was considered to be a major criterion in the current study, although the ARC-HBR definition considers only liver cirrhosis with portal hypertension as a major criterion. All previous bleeding history was regarded as a minor criterion, because we did not have information on the timing, requirement of hospitalization or transfusion, and recurrence with a previous history of spontaneous bleeding. Furthermore, any type of ischemic stroke was considered a minor criterion in the current study, although the ARC-HBR definition considers moderate or severe ischemic stroke within the past 6 months as a major criterion. The original definitions of some ARC-HBR criteria were not directly applicable to the dataset in the present study, and approximation was necessary when they were retrospectively implemented. Finally, the low rates of using the transradial approach, and of patients treated with direct oral anticoagulants could have influenced the results of this study.

Conclusions

The J-HBR criteria successfully identified those patients with very high bleeding risk after PCI, who represented 64% of patients in this all-comers registry.

Acknowledgment

We thank the clinical research coordinators of the Research Institute for Production Development.

Sources of Funding

This study was supported by an educational grant from the Research Institute for Production Development (Kyoto, Japan).

Disclosures

Dr. Natsuaki reports modest honoraria from Abbott Vascular, Amgen, Boehringer-Ingelheim, Boston Scientific, Daiichi-sankyo, Medtronic, Sanofi, Takeda and Terumo. Dr. Morimoto reports modest honoraria from Bayer and Kowa, and modest expert witness from Boston Scientific and Sanofi. Dr. Ehara reports modest honoraria from Abbott Vascular, Bayer, Boston Scientific, Daiichi-sankyo, Edwards Scientific, Medtronic, Pfizer and Terumo. Dr. Furukawa reports modest honoraria from Bayer, Kowa, and Sanofi. Dr. Nakagawa reports modest research grant from Abbott Vascular and Boston Scientific, and modest honoraria from Abbott Vascular, Bayer, and Boston Scientific. Dr. Kimura reports significant honoraria from Abbott Vascular, and modest honoraria from Astellas, AstraZeneca, Bayer, Boston Scientific, Kowa, and Sanofi.

Dr. Kimura is a member of Circulation Journal ’s Editorial Team.

Data Availability

The deidentified participant data will not be shared.

IRB Information

The research protocol of the CREDO-Kyoto Cohort-3 was approved by the Institutional Review Board of Kyoto University (E2400) and by the local ethics committees of all participating medical centers.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-20-0836

References

• 1.   Urban P, Mehran R, Colleran R, Angiolillo DJ, Byrne RA, Capodanno D, et al. Defining high bleeding risk in patients undergoing percutaneous coronary intervention. Circulation 2019; 140: 240–261.
• 2.   Natsuaki M, Morimoto T, Yamaji K, Watanabe H, Yoshikawa Y, Shiomi H, et al. Prediction of thrombotic and bleeding events after percutaneous coronary intervention: CREDO-Kyoto thrombotic and bleeding risk scores. J Am Heart Assoc 2018; 7: e008708.
• 3.   Numasawa Y, Inohara T, Ishii H, Kuno T, Kodaira M, Kohsaka S, et al. Comparison of outcomes of women versus men with non-ST-elevation acute coronary syndromes undergoing percutaneous coronary intervention (from the Japanese nationwide registry). Am J Cardiol 2017; 119: 826–831.
• 4.   Nakamura M, Kitazono T, Kozuma K, Sekine T, Nakamura S, Shiosakai K, et al. Prasugrel for Japanese patients with ischemic heart disease in long-term clinical practice (PRASFIT-Practice II): 1-year follow-up results of a postmarketing observational study. Circ J 2019; 84: 101–108.
• 5.   Nakamura M, Kimura K, Kimura T, Ishihara M, Otsuka F, Kozuma K, et al. JCS 2020 guideline focused update on antithrombotic therapy in patients with coronary artery disease. Circ J 2020; 84: 831–865.
• 6.   Mehran R, Rao SV, Bhatt DL, Gibson CM, Caixeta A, Eikelboom J, et al. Standardized bleeding definitions for cardiovascular clinical trials: A consensus report from the Bleeding Academic Research Consortium. Circulation 2011; 123: 2736–2747.
• 7.   Cutlip DE, Windecker S, Mehran R, Boam A, Cohen DJ, van Es GA, et al; Academic Research Consortium. Clinical end points in coronary stent trials: A case for standardized definitions. Circulation 2007; 115: 2344–2351.
• 8.   Kimura T, Morimoto T, Furukawa Y, Nakagawa Y, Kadota K, Iwabuchi M, et al. Long-term safety and efficacy of sirolimus-eluting stents versus bare-metal stents in real world clinical practice in Japan. Cardiovasc Interv Ther 2011; 26: 234–245.
• 9.   Serruys PW, Unger F, Sousa JE, Jatene A, Bonnier HJ, Schonberger JP, et al; Arterial Revascularization Therapies Study Group. Comparison of coronary-artery bypass surgery and stenting for the treatment of multivessel disease. N Engl J Med 2001; 344: 1117–1124.
• 10.   GUSTO investigators. An international randomized trial comparing four thrombolytic strategies for acute myocardial infarction. N Engl J Med 1993; 329: 673–682.
• 11.   Natsuaki M, Morimoto T, Shiomi H, Yamaji K, Watanabe H, Shizuta S, et al. Application of the academic research consortium high bleeding risk criteria in an all-comers registry of percutaneous coronary intervention. Circ Cardiovasc Interv 2019; 12: e008307.
• 12.   Nakamura M, Kadota K, Nakao K, Nakagawa Y, Shite J, Yokoi H, et al. High bleeding risk and clinical outcomes in East Asian patients undergoing percutaneous coronary intervention: The PENDULUM Registry. EuroIntervention, doi:10.4244/EIJ-D-20-00345.
• 13.   Cao D, Mehran R, Dangas G, Baber U, Sartori S, Chandiramani R, et al. Validation of the academic research consortium high bleeding risk definition in contemporary PCI patients. J Am Coll Cardiol 2020; 75: 2711–2722.
• 14.   Subherwal S, Bach RG, Chen AY, Gage BF, Rao SV, Newby LK, et al. Baseline risk of major bleeding in non-ST-segment-elevation myocardial infarction: The CRUSADE (Can Rapid risk stratification of Unstable angina patients Suppress ADverse outcomes with Early implementation of the ACC/AHA guidelines) bleeding score. Circulation 2009; 119: 1873–1882.
• 15.   Yeh RW, Secemsky EA, Kereiakes DJ, Normand SL, Gershlick AH, Cohen DJ, et al. Development and validation of a prediction rule for benefit and harm of dual antiplatelet therapy beyond 1 year after percutaneous coronary intervention. JAMA 2016; 315: 1735–1749.
• 16.   Lamelas P, Schwalm JD, Quazi I, Mehta S, Devereaux PJ, Jolly S, et al. Effect of body mass index on clinical events after acute coronary syndromes. Am J Cardiol 2017; 120: 1453–1459.
• 17.   Rocca B, Fox KAA, Ajjan RA, Andreotti F, Baigent C, Collet JP, et al. Antithrombotic therapy and body mass: An expert position paper of the ESC Working Group on Thrombosis. Eur Heart J 2018; 39: 1672–1691.
• 18.   Baber U, Mehran R, Giustino G, Cohen DJ, Henry TD, Sartori S, et al. Coronary thrombosis and major bleeding after PCI with drug-eluting stents: Risk scores From PARIS. J Am Coll Cardiol 2016; 67: 2224–2234.
• 19.   Aguayo GA, Donneau AF, Vaillant MT, Schritz A, Franco OH, Stranges S, et al. Agreement between 35 published frailty scores in the general population. Am J Epidemiol 2017; 186: 420–434.
• 20.   Dodson JA, Hochman JS, Roe MT, Chen AY, Chaudhry SI, Katz S, et al. The association of frailty with in-hospital bleeding among older adults with acute myocardial infarction: Insights from the ACTION registry. JACC Cardiovasc Interv 2018; 11: 2287–2296.
• 21.   Genereux P, Giustino G, Witzenbichler B, Weisz G, Stuckey TD, Rinaldi MJ, et al. Incidence, predictors, and impact of post-discharge bleeding after percutaneous coronary intervention. J Am Coll Cardiol 2015; 66: 1036–1045.
• 22.   Valgimigli M, Bueno H, Byrne RA, Collet JP, Costa F, Jeppsson A, et al. 2017 ESC focused update on dual antiplatelet therapy in coronary artery disease developed in collaboration with EACTS: The task force for dual antiplatelet therapy in coronary artery disease of the European Society of Cardiology (ESC) and of the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2018; 39: 213–260.
• 23.   Urban P, Meredith IT, Abizaid A, Pocock SJ, Carrie D, Naber C, et al. Polymer-free drug-coated coronary stents in patients at high bleeding risk. N Engl J Med 2015; 373: 2038–2047.
• 24.   Watanabe H, Domei T, Morimoto T, Natsuaki M, Shiomi H, Toyota T, et al. Effect of 1-month dual antiplatelet therapy followed by clopidogrel vs. 12-month dual antiplatelet therapy on cardiovascular and bleeding events in patients receiving PCI: The STOPDAPT-2 randomized clinical trial. JAMA 2019; 321: 2414–2427.
• 25.   Mehran R, Baber U, Sharma SK, Cohen DJ, Angiolillo DJ, Briguori C, et al. Ticagrelor with or without aspirin in high-risk patients after PCI. N Engl J Med 2019; 381: 2032–2042.
• 26.   Kim BK, Hong SJ, Cho YH, Yun KH, Kim YH, Suh Y, et al. Effect of ticagrelor monotherapy vs. ticagrelor with aspirin on major bleeding and cardiovascular events in patients with acute coronary syndrome: The TICO randomized clinical trial. JAMA 2020; 323: 2407–2416.
• 27.   Watanabe H, Domei T, Morimoto T, Natsuaki M, Shiomi H, Toyota T, et al. Very short dual antiplatelet therapy after drug-eluting stent implantation in patients with high bleeding risk: Insight from the STOPDAPT-2 trial. Circulation 2019; 140: 1957–1959.
• 28.   Singh M, Rihal CS, Lennon RJ, Spertus JA, Nair KS, Roger VL. Influence of frailty and health status on outcomes in patients with coronary disease undergoing percutaneous revascularization. Circ Cardiovasc Qual Outcomes 2011; 4: 496–502.
• 29.   Taguchi I, Iimuro S, Iwata H, Takashima H, Abe M, Amiya E, et al. High-dose versus low-dose pitavastatin in Japanese patients with stable coronary artery disease (REAL-CAD): A randomized superiority trial. Circulation 2018; 137: 1997–2009.