Abstract 17629: Myocardial Scar by Standard and Dual Energy Late Enhancement Computed Tomography: Comparison to Pathology in a Chronic Infarct Porcine Model
Background: Myocardial scar is a substrate for ventricular tachycardia (VT) and sudden cardiac death. Using a chronic myocardial infarction (MI) porcine model, we aimed to (1) quantitatively compare myocardial scar size using standard (LECT) and dual energy (DECT) late enhancement computed tomography to pathology, and (2) qualitatively compare anatomical location with pathology.
Methods: In 13 swine, we performed balloon occlusion of a left coronary vessel (either left anterior descending or circumflex artery) to create a ST elevation MI. After inducing this ischemia-reperfusion model, 8 pigs (75% male, 59±5 kg) underwent 128-slice dual source CT at 33±9 days and then sacrificed at 38±9 days. We had all 3 modalities available for analysis: LECT, DECT and pathology. We performed measurements on the systolic phase of the LECT and DECT acquisitions obtained at 10 minutes after contrast. We measured the area of scar and left ventricular (LV) myocardium at 8 mm slice thickness on co-registered LECT, DECT, and pathology images and calculated the percentage of scar. On a segment basis (n=136), we determined the presence of scar on all 3 modalities using the AHA 17-segment model.
Results: Mean time between CT and pathology was 5±2 days. Percent scar of the LV myocardium was 4.2±3.6% on pathology, 5.9±4.6% on LECT, and 6.6±4.2% on DECT. Both LECT and DECT had excellent correlation with each other (ρ=0.95, p<0.0001) and with pathology for percent scar (both ρ=0.98, p<0.0001). Compared to pathology, percent scar was larger on LECT (Δ 1.7%, p=0.008) and DECT (Δ 2.4%, p=0.008) but similar between LECT and DECT (p=0.55). On a segment basis, both LECT and DECT had excellent specificity with modest sensitivity for scar detection compared to pathology (LECT: sensitivity 53%, specificity 92%; DECT: sensitivity 55%, specificity 93%), with no difference in diagnostic accuracy for LECT and DECT (both McNemar’s test >0.56).
Conclusions: Following chronic MI, LECT and DECT have excellent correlation to pathology for scar quantification with slight overestimation of percent scar. Both have high specificity for scar detection compared to pathology. Late enhancement CT can localize scar and has the potential to be used pre-procedurally to guide VT ablation. Human studies are needed.
- © 2012 by American Heart Association, Inc.