Prediction of 5-year survival in patients with heart failure and implanted cardiac resynchronization therapy devices

Abstract

Aim. Based on clinical parameters and diagnostic investigations, to create a complex model of personalized selection of patients with heart failure (HF) for cardiac resynchronization therapy (CRT). To establish the diagnostic value of the created model in predicting 5-year survival. Material and methods. The study included 141 patients with HF (men, 77,3%; women, 22,7%). The mean age of patients at the time of implantation was 60,0 [53,0; 66,0] years. All patients had New York Heart Association (NYHA) class II-IV HF, left ventricular ejection fraction (LVEF) ≤35%, and QRS ≥130 ms. Patients were randomly divided into training (n=95) and test (n=36) samples, which were comparable in main clinical and functional characteristics. Results. The index included parameters that had a significant relationship with 5-year survival according to the Cox regression: male sex, prior myocardial infarction, hypertension, QRS <150 ms, no left bundle branch block, PR ≥200 ms with sinus rhythm/absence of radiofrequency ablation in atrial fibrillation, NYHA class III, IV HF, LVEF <30%, left ventricular end-diastolic volume ≥235,0 ml, NT-proBNP ≥2692,0 ng/ml. All variables were scored based on the β-coefficients. In the training sample, a value ≥45 points demonstrated a sensitivity of 82,4% and a specificity of 67,2% in predicting 5-year survival (AUC, 0,873; p<0,001). The index use on the test sample showed comparable results (AUC, 0,718; p=0,020; sensitivity — 71,4%, specificity — 62,5%). Also, in the training sample, the index ≥45 points was associated with1-year survival (sensitivity — 84,6%, specificity — 58,1%, AUC, 0,811; p<0,001). Conclusion. An index of personalized selection for CRT has been created, which makes it possible to accurately predict the 5-year survival rate, as well as the 1-year survival rate, regardless of the current selection criteria.