Background: Contrast-enhanced MRI (ceMRI) has impaired accuracy in the prediction of functional recovery after revascularisation in cases of intermediate myocardial viability.
Objective: To evaluate the predictive value of layer-specific myocardial deformation analysis for improvement in ischaemic dysfunction after revascularisation.
Methods: In 132 patients with ischaemic left ventricular dysfunction undergoing revascularisation, myocardial viability was assessed by pixel-tracking-derived myocardial deformation imaging and ceMRI. Peak systolic circumferential strain was determined for total wall thickness and for three myocardial layers (endocardial, mid-myocardial and epicardial) in a 16-segment model. Analysis to predict recovery of function at 8±2 months after revascularisation was performed considering all dysfunctional segments or only segments with intermediate viability by ceMRI (hyperenhancement 25-75%, N=735 segments).
Results: Segments with functional recovery (N=568) had higher circumferential strain in all myocardial layers and a smaller degree of hyperenhancement than segments without functional recovery (N=433). Analysis of all dysfunctional segments showed that the predictive accuracy for functional recovery was high for endocardial strain, total wall thickness strain and hyperenhancement by ceMRI (area under the curve (AUC) 0.883, 0.782 and 0.834, respectively). Considering only segments with intermediate viability by ceMRI, endocardial circumferential strain allowed prediction of functional recovery with higher accuracy (specificity 75%, sensitivity 78%, AUC=0.811, 95% CI 0.776 to 0.851) than hyperenhancement analysis (specificity 59%, sensitivity 72%, AUC=0.705, 95% CI 0.659 to 0.747, p<0.05).
Conclusion: Analysis of layer-specific myocardial function using deformation imaging allows accurate identification of reversible myocardial dysfunction. In segments with intermediate viability analysis of layer-specific deformation may have special advantages for prediction of functional recovery.
Trial registration: ClinicalTrials.gov NCT00476320.