Detection and Quantification of Myocardial Reperfusion Hemorrhage Using T2*-Weighted CMR

ObjectivesThe purpose of this study was to validate T2*-weighted cardiac magnetic resonance (T2*-CMR) for the detection and quantification of reperfusion hemorrhage in vivo against an ex vivo gold standard, and to investigate the relationship of hemorrhage to microvascular obstruction, infarct size, and left ventricular (LV) functional parameters.BackgroundHemorrhage can contribute to reperfusion injury in myocardial infarction and may have significant implications for patient management. There is currently no validated imaging method to assess reperfusion hemorrhage in vivo. T2*-CMR appears suitable because it can create image contrast on the basis of magnetic field effects of hemoglobin degradation products.MethodsIn 14 mongrel dogs, myocardial infarction was experimentally induced. On day 3 post-reperfusion, an in vivo CMR study was performed including a T2*-weighted gradient-echo imaging sequence for hemorrhage, standard sequences for LV function, and post-contrast sequences for microvascular obstruction and myocardial necrosis. Ex vivo, thioflavin S imaging and triphenyl-tetrazoliumchloride (TTC) staining were performed to assess microvascular obstruction, hemorrhage, and myocardial necrosis. Images were analyzed by blinded observers, and comparative statistics were performed.ResultsHemorrhage occurred only in the dogs with the largest infarctions and the greatest extent of microvascular obstruction, and it was associated with more compromised LV functional parameters. Of 40 hemorrhagic segments on TTC staining, 37 (92.5%) were positive for hemorrhage on T2*-CMR (kappa = 0.96, p < 0.01 for in vivo/ex vivo segmental agreement). The amount of hemorrhage in 13 affected tissue slices as determined by T2*-CMR in vivo correlated strongly with ex vivo results (20.3 ± 2.3% vs. 17.9 ± 1.6% per slice; Pearson r = 0.91; r2 = 0.83, p < 0.01 for both). Hemorrhage size was not different between in vivo T2*-CMR and ex vivo TTC (mean difference 2.39 ± 1.43%; p = 0.19).ConclusionsT2*-CMR accurately quantified myocardial reperfusion hemorrhage in vivo. Hemorrhage was associated with more severe infarct-related injury.