Left ventricular flow kinetics and myocardial deformation following acute infarction: Additional predictive value of cardiac magnetic resonance four-dimensional flow for left ventricular remodeling post-ST-elevation myocardial infarction

Background The exact mechanism underlying myocardial maladaptive changes post ST-elevation myocardial infarction (STEMI) remains unclear. We sought to assess the impact of the tissue=flow interaction on the development of adverse cardiac remodeling 12 months(M) after acute STEMI. Materials and methods Forty-nine first-STEMI patients (M:F = 26:13; mean age = 58 ± 10) prospectively underwent 3T cardiovascular magnetic resonance (CMR) acutely, at 3 months (3M) and 12M post-STEMI. The CMR protocol included left ventricular (LV) cine-images for LV end-diastolic (LVEDV) and end-systolic volumes, stroke volume (SV), and ejection fraction (LVEF); four-dimensional (4D)-flow and late gadolinium enhancement imaging. The 3M outcome measures included 4D-flow derived LV flow kinetic energy indexed to EDV (KEiEDV) and functional flow components [LV-KEiEDV, minimal- KEiEDV, diastolic- KEiEDV, and residual volume (RV), retained inflow, delayed ejection, direct flow (DF)]; global radial, circumferential, and longitudinal strain (GRS, GCS, GLS) by feature tracking (FT); infarct size (IS). Adverse LV remodeling (LVremod) was defined by a ≥20% increase in LVEDVi at 12M from baseline, in opposition to the non-remodeling group (LVnon-remod). Association between SV, FT-strain, KE, and 4D flow parameters was assessed, as well as predictors of adverse remodeling at 12M post-STEMI. Results There were 23 LVremod patients. At 3M post-STEMI, LVremod patients had significantly reduced LVEF, increased IS, abnormal FT-strain, systolic KEiEDV, DF, and RV compared to LVnon-remod patients. There was no significant difference in SV between the two groups. FT-strain parameters significantly correlated with DF (GRS: r = 0.62; GCS: r = −0.67; GLS: r = −0.58, all p < 0.001); RV (GRS: r = −0.56; GCS: r = 0.51; GLS: r = 0.53, all p < 0.001); peak-A-wave KEiEDV (GRS: r = 0.38, p = 0.008; GCS: r = −0.30, p = 0.038; GLS: r = −0.29, p = 0.04); systolic KEiEDV (GRS: r = 0.31, p = 0.033, GLS: r = −0.35, p = 0.012). DF outperformed conventional LV function parameters (SV and LVEF) in the LVremod prediction. DF and IS were the only independent predictors of 12M adverse remodeling after adjustment for LVEF, SV, FT-strain, and KEiEDV parameters. Conclusions Our study suggests a potential early interaction between FT-strain and 4D-flow parameters post-STEMI leading to the development of adverse remodeling. Within the limitations of our sample size, DF and IS were independent predictors of LV remodeling after adjustment for LVEF, SV, FT-strain, and KE parameters. These findings suggest that these parameters may contribute to further risk stratification at 3M for the development of adverse remodeling at 12M post-STEMI, above conventional LV function parameters. Larger studies are needed to confirm these results.