Accuracy of infarct measurements by inversion recovery delayed-enhancement MRI during the hyper-acute phase of myocardial infarction in rats

Introduction

Quantification of infarct size is important in the assessment of myocardial infarction (MI) and is commonly performed by ex-vivo staining in experimental studies [1]. Delayed-enhancement (DE) is an accurate method for assessing infarct in-vivo in humans, and is increasingly used in experimental studies of chronic MI where its accuracy has been well validated [2]. However, the feasibility and accuracy of DE-MRI in measuring infarct during the hyperacute MI phase (i.e. hours) is less well-defined. Additionally, while inversion recovery (IR) is the MR sequence of choice clinically, its role has not been fully explored in small animal models.

Purpose

To compare infarct measurements by IR-based DE-MRI against tetrazolium staining during the hyperacute phase of MI in rats.

Methods

Male Wistar rats (n = 13) were anaesthetised and underwent 30 minutes of myocardial ischaemia and 2 hours of reperfusion. Short-axis DE images of the infarcted heart were acquired using a 9.4 T system (Varian) with Gd-DTPA (0.6 mmol/kg, i.v.) and a dual ECG/respiratory-gated IR sequence [3] (α = 90°, TE = 1.5 ms, TR = 3.6 ms, TI~400 ms, FOV = 40 × 40 mm², 192 × 192, ΔZ = 1 mm, 10-15 slices). Hearts were extracted following imaging for TTC staining and planimetry. Data were randomised and segmented by thresholding with manual corrections (ImageJ). Infarct size was expressed as a ratio to the LV (In/LV). Linear regression and Bland-Altman analysis were performed (SPSS) to compare the two methods.

Results

Figure 1 shows representative DE-MRI and TTC images. DE-MRI measurement of absolute LV and infarct sizes was systemically larger than TTC planimetry, which is most likely due to tissue shrinkage during TTC preparation. Analysis of normalised infarct size (In/LV) demonstrated a strong correlation between DE-MRI and TTC planimetry (R² = 0.893, F = 108.7, p < 0.0001; Figure 2a). Bland Altman analysis revealed negligible bias (0.4%; 95% limits of agreement = -6.6 to 7.4%; Figure 2b).

Figure 1

Figure 2

Correlation of In/LV by DE-MRI and TTC planimetry, illustrated here as linear regression (2a) and Bland-Altmann plots (2b). Dotted lines represent 95% confidence intervals/limits of agreement. In/LV measured by DE-MRI and TTC showed in non-statistically significant mean bias of +0.44 ± 3.72% (p = 0.67).

Conclusion

The application of IR-based DE-MRI in small animals has mainly been in chronic MI at lower fields (1.5-3 T) [4, 5]. By employing a customised IR sequence at 9.4 T, we achieved excellent image contrast between normal and infarcted myocardium, and demonstrated that this technique can provide accurate infarct measurements during the hyper-acute phase of MI as validated by TTC planimetry. Therefore, IR-based DE-MRI may serve as a valuable tool in monitoring infarct progression and therapeutic outcome in small animal models.