Early echocardiographic markers of adverse left ventricular remodeling in STEMI patients: mechanical dispersion and global longitudinal strain

Abstract

Despite advances in contemporary treatment strategies, patients with ST-elevation myocardial infarction (STEMI) continue to face a substantial risk of adverse cardiovascular outcomes. Early identification of patients with high risk of adverse cardiac remodeling (ACR) remains a major clinical challenge as post-infarction ACR drives heart failure progression and increases long-term mortality.

Aim

to assess left ventricular (LV) mechanical dispersion (MD) and global longitudinal strain (GLS) as predictive markers of ACR in patients with STEMI after successful revascularization after primary percutaneous coronary intervention (PCI) over a 12-months period.

Materials and Methods

Patients with acute STEMI who had no prior MI, no arrhythmias and no severe comorbidities were included. A total of 119 STEMI patients were enrolled. Clinical, hemodynamic, ECG and echocardiographic including Doppler and speckle-tracking markers were assessed in the acute phase and at 12 months post-event. Blood biomarker levels were measured at baseline. ACR was defined as an increase in LV end-systolic and/or end-diastolic volumes >10% at 12 months.

Results

Among 119 patients after STEMI with successful revascularization, TIMI III the majority were men (88.2%) with a mean age of 59 years. Patients were stratified into four groups according to MD levels. Univariate regression analysis identified the GRACE score, NT-proBNP, Troponin I levels, LV ejection fraction (LVEF), LV GLS, and LV MD as predictive markers of ACR in 1-year follow-up period. Multivariate regression analysis showed that GLS >–8%, high LV MD quartile, and NT-proBNP >953 pg/mL remained independent predictors of ACR. ROC curve analysis demonstrated that all three predictive models based on LV MD (AUC=0.84, p<0.001; sensitivity 80.8%, specificity 87.5%), LV GLS (AUC=0.62; p<0.001; sensitivity 82.0%, specificity 66.7%), and NT-proBNP (AUC=0.60, p<0.001; sensitivity 78.2%, specificity 73.5%) significantly differed from the baseline model allowing identification of high-risk post-infarction patients.

Adding LV MD to the baseline predictive model (NT-proBNP >953 pg/mL) significantly improved model discrimination (AUC 0.84 [0.72–0.90], p=0.001; NRI Median 95% CI =0.54, p=0.018; IDI Median 95% CI =0.53, p=0.042), whereas adding LV GLS did not enrich the predictive model either alone (AUC 0.62 [0.58–0.69], p=0.49; NRI Median 95% CI =0.30, p=0.25; IDI Median 95% CI=0.04, p=0.62) or when combined with all parameters (AUC 0.83 [0.73–0.91], p=0.001; NRI Median 95% CI =0.52, p=0.02; IDI Median 95% CI=0.052, p=0.044). Thus, LV MD proved to be an independent prognostic factor for ACR significantly improving predictive accuracy when applied.

Conclusion

Our study demonstrated that LV GLS and LV MD were the most independent markers of ACR. Measurement of the markers in the acute phase of STEMI may be useful for identifying patients at high risk of ACR during 12 months of follow-up.