Occurrence and predictors of left ventricular systolic dysfunction at hospital discharge and in long−term follow−up after acute myocardial infarction treated with primary percutaneous coronary intervention

Abstract

Background: Post-ST-segment elevation myocardial infarction (STEMI) left ventricular systolic dysfunction (LVSD) has been identified as an important marker of poor prognosis. Aim: To assess the prevalence and course of LVSD at hospital discharge and in long-term follow-up in STEMI patients treated with primary percutaneous coronary intervention (pPCI). Methods: We enrolled 205 patients (157 male, 48 female) with a first STEMI. Echocardiography was performed before hospital discharge and 12 months after STEMI. Left ventricular systolic function (LVSF) parameters were assessed: left ventricular ejection fraction (LVEF), wall motion score index (WMSI), and average peak systolic mitral annular velocity (S’) by tissue Doppler echocardiography (TDE). B-type natriuretic peptide plasma concentration was measured at admission (BNPadmission) and at discharge (BNPdischarge). Results: We found moderate LVSD, both at hospital discharge and after 12 months. Significant global LVSD (LVEF £ 40%) was observed in 34% of patients at discharge, and 21% after 12 months (p < 0.001). Significant regional LVSD (WMSI ≥ 1.7) after 12 months was less frequent than at discharge (21% vs 33%; p < 0.001). More patients had significant longitudinal LVSD (S’ £ 6.0 cm/s) after 12 months compared to discharge (28% vs 23%; p < 0.001). Severe global LVSD (LVEF £ 30%) was rare. Univariate logistic regression analysis revealed the predictors of significant global LVSD at 12 months after STEMI to be: anterior location of STEMI; pre-discharge echocardiographic parameters of LVSF and left ventricle size and mass; prepPCI angiographic indices; ratio of the difference of BNPdischarge and BNPadmission to BNPadmission expressed as % (BNPdelta %); time from onset of pain to balloon, and the use of abciximab. Multivariate logistic regression analysis found independent predictors of significant global LVSD at 12 months to be: BNPdelta % and LVEF at discharge with optimal cut-off values of 728.2% for BNPdelta % and 37% for LVEF. Conclusions: Patients with a first STEMI treated with pPCI present moderate LVSD, both at hospital discharge and after 12 months. In long-term follow-up, we found an improvement in global LVSF, and, albeit a smaller, improvement in regional LVSF. No improvement in longitudinal LVSF was observed. The increase of BNP during hospitalisation, and LVEF at discharge, are independent predictors of significant global LVSD at 12 months after a first STEMI treated with pPCI. Pre-discharge peak systolic mitral annular velocity obtained by TDE may be useful in predicting LVEF in long-term follow-up in this group of patients.