Detecting acute reperfusion myocardial hemorrhage with CMR: a translational study

Abstract

Methods Canines (n=14), subjected to ischemia-reperfusion (I/R) injury (3 hrs of LAD occlusion followed by reperfusion), underwent CMR (1.5T) studies on day 5 post reperfusion. T2*-weighted (multi GRE; TE=3.4-18.4ms (6 echoes)), T2-weighted (T2-prep SSFP; prep times=0, 24 and 55 ms), T2-STIR (TE=64 ms) and Late Enhancement (LE) images covering the LV were acquired. Imaging resolution of all the scans was 1.3x1.3x8 mm3. Patients (n=14) underwent CMR (1.5T) on day 3 post angioplasty for STEMI after providing informed consent. T2*-weighted (TE=2.6-13.8ms (6 echoes)), T2-STIR (TE=61ms) and LE images covering the LV were acquired. Imaging resolution of all the scans was 1.4x1.4x10 mm3. T2* and T2 maps were constructed by fitting the multi-echo data to monoexponential decay. A thresholdbased signal analysis was used to identify hemorrhagic (Hemo+) and non-hemorrhagic (Hemo-) infarcts. T2STIR signal intensity (STIR-SI), T2* and T2 values, measured from Remote, Hemo- and Hemo+ regions, were compared. Statistical significance was set at p<0.05. Results Representative T2* and T2 maps, and T2-STIR and LE images (acquired from a canine on day 5 post I/R injury) are shown in Fig. 1A. Mean T2* of Hemo+ was lower than both Remote and Hemo- regions (-39%; p<0.001; Fig. 1B and 1C), while no differences were observed in T2* between Remote and Hemo- (p=0.27). Compared to Remote, mean T2 of both Hemo- and Hemo+ regions were elevated (26% and 17% respectively; p<0.001), with T2 of Hemo- greater than T2 of Hemo+ (p<0.001). A similar trend was observed in T2STIR images as well; STIR-SI of Hemo+ and Hemowere greater than Remote (59% and 31% respectively; p<0.001), while STIR-SI of Hemo- was greater than Hemo+ (p<0.001). Representative T2* map, T2-STIR and LE images, acquired from a patient (day 3 post angioplasty) are shown in Figure 2. Consistent with the animal studies, mean T2* of Hemo+ was lower than the mean T2* of both Remote and Hemo- (-46%, p<0.001; Fig.2), while no differences were observed between T2* of Remote and Hemo- (p=0.61). Mean STIR-SI of both Hemo- and Hemo+ were greater than that of Remote (78% and 33% respectively; p<0.001), with the mean STIR-SI of Hemogreater than that of Hemo+ (p<0.001).