P1950How to improve early prediction of heart failure development after STEMI in clinical arena-is there a place for myocardial deformation imaging?

Abstract

Abstract Heart failure (HF) development after myocardial infarction with ST segment elevation (STEMI) in the modern era varies greatly (between 4% and 28%), with the highest incidence in the first year. Since, HF still carries substantial morbidity and mortality, accurate and early identification of high-risk patients for HF development after pPCI allows for targeted use of intensive therapy. Aim The current study is a sub-study of PREDICT-VT study (NCT03263949). Its aim is to define multi-parametric model for early HF prediction in STEMI patients treated by pPCI, based on three data sets: clinical data, conventional echocardiographic data and data from myocardial deformation analysis obtained by early speckle tracking echocardiography. Methods In 264 consecutive pts enrolled in PREDICT-VT study early echocardiography (5±2 days after pPCI) was done and included LA and multilayer LV deformation analysis with longitudinal (L), radial (R) and circumferential (C) strain (S; %) and strain rate (SR, 1/sec). LV index of post systolic shortening for longitudinal strain (PSS LS) and for circumferential strain (PSS CS) were calculated for the purpose of this study as average of PSS over 18 LV segments. Results From 195 patients who completed 1-year follow-up, 17 (8.7%) develop HF NYHA class 3 or 4, 60 (30.8%) NYHA class 2 and remaining 118 patients (60.5%) were in NYHA class I. Significant univariate predictors of NYHA were: from clinical parameters – female gender (β=0.169, p=0018), Killip class on admission (β=0.253, p<0.001) and previous atrial fibrillation (β=0.205, p=0.004); from conventional echocardiographic parameters – LV WMSI (β=0.223, p=0.0072), LVEF (β=−0.256, p<0.001), LAVI (β=0.174, p=0.020) and TAPSE (β=−0.263, p=0.001); from parameters of longitudinal LV deformation – LS at the epicardial level (β=0.242, p=0.001) and PSS LS (β=0.360, p<0.001); from parameters of LV circumferential deformation – CS at epicardial level (β=0.225, p=0.001) and PSS CS at epicardial level (β=0.124, p=0.004); from left atrial mechanics – LA strain (β=−0.199, p=0.007). In multivariable stepwise regression model 5 variables were further identified as independent predictors that significantly increased model power to predict HF development (R square from 0.134 to 0.270, p<0.001). They are: PSS LS (β=0.255, p=0.002), previous atrial fibrilation (β=0.205, p=0.008), TAPSE (β=−0.176, p=0.031), female gender (β=0.165, p=0.032) and PSS CS (β=0.155, p=0.047). Conclusion Independently from and above classical clinical and echocardiographic parameters, amount of left ventricular post-systolic deformation in longitudinal and circumferential direction, expressed as LV indexes of post-systolic shortening, significantly improved early prediction of HF development after pPCI.