Effectiveness and avoidance of side effects of ARNI, SGLT2 onhibitor, MRA, and beta blockers in acute heart failure patients

2022 Fiscal Year Final Research Report

• Project/Area Number: 21K21277
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0909:Sports sciences, physical education, health sciences, and related fields
• Research Institution: National Cardiovascular Center Research Institute
• Principal Investigator: Kiyoshige Eri 国立研究開発法人国立循環器病研究センター, 研究所, リサーチフェロー (30911432)
• Project Period (FY): 2021-08-30 – 2023-03-31
• Keywords: AI / 循環器 / 疫学 / 予測モデル

Outline of Final Research Achievements

Decision of optimal medication for Heart Failure (HF) patients were difficult; however, the combined treatment called “Fantastic four”: an ARNI, a beta-blocker, an MRA, and an SGLT2 inhibitor, could avoid risks of mortality and worsening. The reported Fantastic four papers used simulated data, thus we used observational data and conducted analyze the association between decrease major adverse cardiovascular events (MACE) and Fantastic four medication and exploring how to avoid MACE. Firstly, we develop high accurate MACE prediction model by using XGBoost (AUC=0.91), of which feature importances were blood urea nitroge, systolic blood pressure, history of chronic kidney disease and heart failure, and troponin T level. The obtained results could suggest the importance of primary prevention for heart failure and of well-organized patients’ blood pressure, and those interpretations were meaningful in clinical practice.

Free Research Field

疫学・統計

Academic Significance and Societal Importance of the Research Achievements

心不全患者における重篤な心疾患イベント発症（MACE）を従来のモデルより高精度に予測でき、加えMACE予測に有用な変数は先行研究を支持するものであり、MACEに効果的な4剤併用(ARNI、SGLT2阻害薬、MRA、β遮断薬)よりも影響があると明らかになった。AIは従来の統計モデルよりも自由度が高いため、実臨床での日常業務で収集したデータをそのまま使用する事が可能である。本研究は、実臨床にある膨大なデータをAIにて分析可能にし、心不全患者の寿命の延伸とQOL向上に有用であることを示唆した。