MR Assessment of Acute Pathologic Process after Myocardial Infarction in a Permanent Ligation Mouse Model: Role of Magnetic Nanoparticle-Contrasted MRI

Figure 1. MRI images were acquired at 1 day (LGE: a, g), 2 days (MEMRI: b, h), and 3 days (MNP-MRI: c, i) after induction of myocardial infarction (MI). Histological examination of the mouse heart was performed at 3 days after MI (d, j). Representative cardiac MRI images and corresponding histological images are shown. Images in the upper row were acquired from anterolateral LV wall infarction (a–f) and bottom row images were obtained from entire LV free wall infarction (g–l). The infarcted myocardium appears hyperintense in LGE and hypointense in MEMRI images, and the corresponding infarct region is evident in a immunohistochemistry stain for myoglobin (d, j); myoglobin-positive areas (dark blue) are viable myocardium. MNP- (red) and CD68-positive (green) regions are consistent with hypointense signals in MNP-MRI representing inflammatory areas after myocardial infarction (e, k). Colocalization of MNPs with macrophages is observed in the peri-infarct zone (f, l).

Figure 2. Area-based approach for contrast-enhanced size measurement. Enhanced area (red color) was demarcated using Otsu's threshold method (b, e, h). The epicardium and endocardium are traced in yellow. The dotted lines in the histogram correspond to the estimated threshold values and green bars represent enhanced pixels in MRI images (c, f, i). The sizes of enhanced area for LGE, MEMRI, and MNP-MRI was 38.7%, 46.3%, and 50.4%, respectively. The horizontal axes represent normalized signal intensity in the MR image.

Figure 3. Sector-based analysis of contrasted CMR images. LV myocardium was registered to 360 radial segments for sector analysis (b, f, j). Enhanced sectors were determined by Otsu's thresholding (c, g, k). The green region denotes the enhanced area both in the histogram and in radial profile. Dashed lines both in the histogram and in radial profile are threshold values according to Otsu's method. The enhanced sizes as determined by LGE, MEMRI, and MNP-MRI were 30.8%, 23.9%, and 49.2%, respectively.

Figure 4. Comparison of contrasted sizes of MI mice by different MRI methods. The sizes were evaluated by area-based (a) and sector-based approaches (b). Enhancement size of MNP-MRI was significantly larger than that of LGE and MEMRI for both approaches. Difference in the enhancement size between LGE and MEMRI was not significant for both approaches. ∗∗P < 0.01, ∗∗∗P < 0.001, ∗∗∗∗P < 0.0001, ns = nonsignificant. (area-based: P = 0.34 for LGE versus MEMRI, P < 0.0001 for LGE versus MNP-MRI and P < 0.0001 for MEMRI versus MNP-MRI; sector-based: P = 0.99 for LGE versus MEMRI, P = 0.0015 for LGE versus MNP-MRI and P = 0.0001 for MEMRI versus MNP-MRI).

Figure 5. Correlation between infarct lesion size based on LGE or MEMRI and inflammatory lesion size based on MNP-MRI. Pearson's correlation coefficients (r) and P values are described in the figure. Significant positive correlations between the infarct and inflammatory lesion size were observed by both area-based (a, c) and sector-based approaches (b, d). The solid line denotes the linear regression fit, and the dotted lines represent the 95% confidence limits. Each data point represents the individual enhanced size for each slice (4 slices per mouse).

Figure 6. Scatterplot of infarction lesion size based on MRI (sector-based method) and histology (length-based method). There is a significant correlation of infarct size between MRI and histology results. Spearman's rank correlation coefficients (r) and P values are r = 0.9058 (P = 0.0129) for LGE (◯), and r = 0.9301 (P = 0.0071) for MEMRI (△). The dotted line represents the identity line. The infarct size was measured at the mid-ventricular level in 6 mice.