Impact of BNP level in patients with heart failure on major bleeding events after percutaneous coronary intervention

Aims The Academic Research Consortium for High Bleeding Risk (ARC-HBR) presents a bleeding risk assessment in antithrombotic therapy for patients post percutaneous coronary intervention (PCI). In Japanese patients, heart failure (HF), peripheral vascular disease, and frailty are established as bleeding risk factors in addition to ARC-HBR. However, it is unknown whether left ventricular function or severity of HF is associated with HBR. The aim of this study was to investigate the association between the severity of HF measured by BNP and future bleeding events after PCI. Methods Clinical Deep Data Accumulation System (CLIDAS), a multicenter database with 7 tertiary medical hospitals in JAPAN, was developed to collect data directly for patient characteristics, medications, laboratory test, physiological test, cardiac catheterization and PCI treatment in electronic medical records using Standardized Structured Medical Information eXchange Extended Storage (SS-MIX). This retrospective analysis using CLIDAS database included 7160 patients who underwent PCI during April 2014 and March 2020 in the participating hospitals and also who have completed 3-year follow-up were divided into two groups: No HF (n=6645) and HF (n=515). HF patients were furthermore divided based on high BNP (≥100 pg/ml) group (n=384) and low BNP (<100 pg/ml) group (n=131). Primary outcome was defined as bleeding events according to the moderate and severe bleeding in the GUSTO classification. In addition, secondary endpoint was major adverse cardiovascular events (MACE) defined as a composite of cardiac death, myocardial infraction and stroke. Results Multivariable Cox regression adjusted for age, sex, BMI, acute coronary syndrome, hypertension, diabetes, dyslipidemia, chronic kidney disease, hemodialysis, previous PCI, previous coronary artery bypass grafting, prior myocardial infraction, prior stroke, prior atrial fibrillation, prior PVD, left main trunk disease, multivessel disease, and anticoagulants use showed that HF with high BNP was significantly associated with bleeding events (hazard ratio [HR], 1.66; 95% confidence interval [CI], 1.10–2.50), MACE (HR, 2.16; 95% CI, 1.60–2.90), and all-cause death (HR, 1.74; 95% CI, 1.30–2.33), but not HF with low BNP. Conclusions The CLIDAS real-world database revealed that HF with high BNP was associated with future bleeding events, suggesting that bleeding risk might be altered depending on severity of HF. Funding Acknowledgement Type of funding sources: None.