Reproducibility of quantitative measurement of infarct size in mice with high resolution PET/CT imaging.

Aim: Positron emission tomography (PET) with 18-F fluorodeoxyglucose (FDG) is increasingly used for noninvasive evaluation of myocardial pathophysiology in small animals. Mice models are of importance due to the availability of a multitude of disease and knock-out models. However, the dimensions of mouse heart are less than in rats and pose a challenge for imaging acquisition and analysis. This study assessed the reproducibility of 18F-FDG PET for quantification of infarct size in mice using a high-resolution animal PET/computed tomography (CT) system. Material and Methods: Forty outbred wild-type mice were used for this study. Twenty mice were studied after induction of myocardial infarction (MI group), obtained through ligation of the descendant anterior branch of the left coronary artery and 20 mice under control conditions (C group). Imaging was performed 45 min after intravenous administration of 18F-FDG (7.5 MBq) using a highresolution dedicated PET/CT (eXplore Vista, GE Healthcare) with a PET spatial resolution of 1.6 mm FWHM and a CT spatial resolution of 0,2 mm. PET data were reconstructed using a 2D-OSEM including random and scatter corrections, and attenuation correction using CT data. Volumetric analysis was used to measure infarct size expressed as area/cm2 and % area map. Intraobserver and interobserver post-processing reproducibility was evaluated by the repeatability coefficient (RC). The intraclass coefficient correlation (ICC) was used to assess intra- and interobserver interpretative reproducibility of analysis. The performers were blinded to mice data and to each other's results. Results: Three mice of the MI group died during surgery or in the early postoperative phase; complete data from 37 animals were available for analysis. Measures of infarct size in the mice of the MI group were 0.35±0.22 area/cm2 and 36±13 % area map. In all 37 mice, intraobserver RC was 0.1 (area/cm2) and 3 (% area map) and interobserver RC was 0.11 (area/cm2) and 3.6 (% area map). In MI group, intraobserver RC was 0.12 (area/cm2) and 4.3 (% area map) and interobserver RC was 0.16 (area/cm2) and 5 (% area map). In C group, intra-observer RC was 0.01 (area/cm2) and 0.71 (% area map) and interobserver RC was 0.01 (area/cm2) and 1.5 (% area map). In all 37 mice, the intraobserver and interobserver ICC were 0.96 (F value 49) and 0.81 (F value 9), respectively (p<0.01). In MI group, the intraobserver and interobserver ICC were 0.95 (F value 42) and 0.82 (F value 10), respectively (p<0.01). In C group, the intraobserver and interobserver ICC were 0.89 (F value 18) and 0.54 (F value 4), respectively (p<0.01). Conclusions: The quantitative measurement of infarct size in mice with a high-resolution PET/CT has excellent intraobserver and interobserver interpretative reproducibilities. This noninvasive methodology will allow longitudinal studies of cardiac biochemical parameters and facilitate studies to assess the effect of intervention after acute MI.