239. Histological Measurement of Myocardial Infarct Area Does Not Provide a Sensitive Assessment of Infarct Size in a Mouse Chronic Infarct Model

Abstract

Assessment of severity of a myocardial infarction (MI) is an important measure of success after experimental delivery of potentially therapeutic genes or cells in rodent MI models. In experiments involving an acute MI setting, histological measurement of the area of the infarcted region in tissue sections of the left ventricle (LV) is a standard assessment of |[ldquo]|infarct size.|[rdquo]| This approach is also used by numerous investigators in a chronic infarct setting, taking infarct area measurements several weeks post-MI. However, we hypothesized that wall thinning due to remodeling of a transmural infarct would invalidate the correlation between a severe infarct and a large infarct scar area measurement, reducing the legitimacy of this approach in a chronic setting. To address this issue, we assessed left ventricular ejection fraction (LVEF; cardiac systolic function) in 8 infarcted mouse hearts and ranked them according to increasing severity. Tissue sections were collected approximately every 300 uM from apex to base and were stained with Masson's trichrome to visualize the infarct scar. The scar was measured in each section by a blinded investigator using several strategies, the numbers obtained were combined to express infarct size as percentage of total LV volume, and the values were compared to the functional measurements for each heart. Our measurement strategies consisted of calculation of the ratio of (1) summed infarct areas to summed LV myocardial areas (Area Measurement: AM), (2) summed epicardial and endocardial infarct arc lengths to summed epicardial and endocardial LV circumferences (Length Measurement: LM), and (3) summed midline infarct arc lengths to summed midline LV circumferences (Midline Length Measurement: MLM). Linear correlation analyses showed that infarct size from all three measurement approaches correlated significantly with LVEF (r=0.87, p<0.01 for each method) and wall motion score index (r=0.97-0.98, p<0.01 for each). However, the infarct size calculated using AM was significantly smaller than those using the other measurements (average of values obtained for all hearts studied: AM 10.83|[plusmn]|6.53% vs. LM 26.72|[plusmn]|19.04%, MLM 26.08|[plusmn]|20.94%, p<0.01), and the range of values obtained was compressed 0.4-fold, substantially compromising the accuracy of this approach. We conclude that while wall thinning did not invalidate the use of area-based infarct size measurements, it underestimated the extent of the infarct and substantially reduced the sensitivity of this measure compared to measurements based on infarct arc lengths. We further conclude that the estimation of myocardial and infarct midlines provided comparable accuracy to the more cumbersome tracing of epicardial and endocardial lengths and circumferences. These findings demonstrate that the ratio of summed infarct midlines to summed LV midlines is significantly better than infarct area measurements as an assessment of infarct size in cell or gene therapy studies in chronic rodent MI models.