ROLE OF MYOCARDIAL HAEMORRHAGE IN PREDICTING ADVERSE LEFT VENTRICULAR REMODELING AFTER REPERFUSED ACUTE MYOCARDIAL INFARCTION. CAN WE OUTPERFORM MICROVASCULAR OBSTRUCTION?

Abstract

Abstract Category: MRIPresentation Number: 0911-07Authors: Claudia Raineri, Annalisa Turco, Sergio Leonardi, Mara Bonardi, Michela Cottini, Guido Tavazzi, Margherita Calcagnino, Gabriele Crimi, Maurizio Ferrario, Mario Previtali, Fondazione IRCCS Policlinico S.Matteo, Pavia, Italy Background: Microvascular obstruction (MVO) detected by contrast-enhanced cardiac magnetic resonance (CE-CMR) is a well-known predictor of adverse left ventricular remodeling (aLVR) after reperfused acute myocardial infarction (AMI), while the influence of myocardial haemorrhage (MH) on aLVR is still unclear. Aim of the present study was to test the hypothesis that MH can improve the ability to identify patients at risk of aLVR beyond MVO.Methods: Sixty-two AMI patients (88% males, 57±11 years, 64% anterior location) reperfused with primary percutaneous coronary intervention were referred for a CE-CMR assessment during index hospitalization and at 4 months. Left ventricular remodeling was defined as the difference between acute and 4-month LV end-systolic volume index. MVO and MH were semi-quantitatively estimated using a per-segment basis (presence or absence) score on a 16-segment model of the LV. The ability of MVO, MH and other potential predictors, including age, sex, MI location, TIMI flow pre- and post-reperfusion, acute LV ejection fraction, MI size (extent of delayed enhancement at CE-CMR), CK-MB peak, pain-to-balloon time, to predict aLVR was assessed by linear regression, with univariable and multivariable analysis.Results: MH and MVO were observed in 24 (39%) and 46 (74%) patients respectively. MH was always associated with MVO and the extent of MVO was significantly larger in the MH group (3.9±1.5 vs 2.3±1.2 segments, p<.0001). Compared to patients with MVO only, those with both MH and MVO had similar acute end-systolic LV volume index (MVO only:44.2±15.4 vs MVO+MH:48.8±16 ml/m2, p=.77 respectively), MI size (28±7 vs 27.5 ±12 %, p=.99), acute LV ejection fraction (46.7±9.1 vs 45.4±10 %, p=.96). At univariable analysis MVO (F=9.8, r=.36, p=.003), MH (F=10.2, r=.38, p=.002), peak CK-MB (F=6.3, r=.30, p=.015), MI size (F=5.3, r=.27, p=.025) were significant predictors of aLVR. In multivariable analysis only MVO retained the ability to predict aLVR (F=11.7, r=.41, p=.001).Conclusion: MH is always associated with MVO and can predict aLVR in univariable analysis but did not show independent predictive value beyond MVO.