Molecular imaging by small-animal PET is an important noninvasive means to phenotype transgenic mouse models in vivo. When investigating pathologies of the left ventricular (LV) myocardium, the serial assessment of LV volumes is important. By this, the presence of LV dilation as a sign of developing heart failure can be detected. Whereas PET is usually used to derive biochemical and molecular information, functional parameters such as ventricular volumes are generally measured using echocardiography or MRI. In this study, a novel method to monitor LV dilation in mice with PET is presented and evaluated using cardiac MRI.
Methods: A semiautomatic 3-dimensional algorithm was used to delineate the LV myocardial wall on static PET images depicting myocardial glucose metabolism ((18)F-FDG PET) for 20 mice: 10 wild-type and 10 genetically modified littermates designed to develop a dilative cardiomyopathy phenotype (cardiomyocyte-specific knockout of survivin). The volume enclosed by the 3-dimensional midmyocardial contour was calculated as a measure for LV volume for each mouse. Data were compared with ventricular volumes measured by MRI in the same animals.
Results: LV volumes obtained by PET and MRI correlated well (R = 0.89) for hearts with small and large left ventricles. In accordance with the hypothesis, the LV volumes were increased significantly for transgenic mice examined at an older age compared with those examined at a younger age (MRI: 160.5 +/- 25.7 microL vs. 114.7 +/- 15.2 microL [P = 0.012]; PET: 129.3 +/- 15.3 microL vs. 73.8 +/- 15.0 microL [P < 0.001], all values shown as mean +/- SD; for MRI, mean of end-diastolic and end-systolic volumes are given), whereas they did not for their wild-type littermates (MRI: 106.2 +/- 12.3 microL vs. 94.7 +/- 14.6 microL [P = 0.214]; PET: 82.6 +/- 20.9 microL vs. 65.0 +/- 16.9 microL [P = 0.185]).
Conclusion: Evaluation and quantitation of LV dilation in both control and cardiomyopathic mice can be reliably and serially performed using small-animal PET and (18)F-FDG, yielding useful functional information in addition to metabolic data.