Regional contrast agent quantification in a mouse model of myocardial infarction using 3D cardiac T1 mapping

Abstract

Background: Quantitative relaxation time measurements by cardiovascular magnetic resonance (CMR) are of paramount importance in contrast-enhanced studies of experimental myocardial infarction. First, compared to qualitative measurements based on signal intensity changes, they are less sensitive to specific parameter choices, thereby allowing for better comparison between different studies or during longitudinal studies. Secondly, T1 measurements may allow for quantification of local contrast agent concentrations. In this study, a recently developed 3D T1 mapping technique was applied in a mouse model of myocardial infarction to measure differences in myocardial T1 before and after injection of a liposomal contrast agent. This was then used to assess the concentration of accumulated contrast agent.

Materials and methods: Myocardial ischemia/reperfusion injury was induced in 8 mice by transient ligation of the LAD coronary artery. Baseline quantitative T1 maps were made at day 1 after surgery, followed by injection of a Gd-based liposomal contrast agent. Five mice served as control group, which followed the same protocol without initial surgery. Twenty-four hours post-injection, a second T1 measurement was performed. Local ΔR1 values were compared with regional wall thickening determined by functional cine CMR and correlated to ex vivo Gd concentrations determined by ICP-MS.

Results: Compared to control values, pre-contrast T1 of infarcted myocardium was slightly elevated, whereas T1 of remote myocardium did not significantly differ. Twenty-four hours post-contrast injection, high ΔR1 values were found in regions with low wall thickening values. However, compared to remote tissue (wall thickening > 45%), ΔR1 was only significantly higher in severe infarcted tissue (wall thickening < 15%). A substantial correlation (r = 0.81) was found between CMR-based ΔR1 values and Gd concentrations from ex vivo ICP-MS measurements. Furthermore, regression analysis revealed that the effective relaxivity of the liposomal contrast agent was only about half the value determined in vitro.

Conclusions: 3D cardiac T1 mapping by CMR can be used to monitor the accumulation of contrast agents in contrast-enhanced studies of murine myocardial infarction. The contrast agent relaxivity was decreased under in vivo conditions compared to in vitro measurements, which needs consideration when quantifying local contrast agent concentrations.

Figure 1

3D black-blood retrospectively triggered CINE images used for determination of wall thickening. 3D black-blood retrospectively triggered CINE images used for determination of wall thickening. Images from left to right represent different slices through the heart towards the apex (movie provided in Additional File 1).

Figure 2

Pre-contrast myocardial T₁ (A), post-contrast myocardial T₁ (B), and post-contrast T₁-weighted images (C) in a mouse with myocardial infarction. Pre-contrast myocardial T₁ (A), post-contrast myocardial T₁ (B), and post-contrast T₁-weighted images (C) in a mouse with myocardial infarction. Arrows mark infarct regions showing contrast enhancement in post-contrast T₁-weighted imaging. Images represent the same slices as in Figure 1.

Figure 3

Relation between wall thickening and contrast agent accumulation. (A) Bull's eye plots of ΔR₁ and wall thickening (SWT) values corresponding to the animal in Figure 1. Regions with SWT < 15% are considered infarcted, whereas remote regions have SWT > 45%. (B) Relation between wall thickening and their corresponding ΔR₁ values (mean value of all animals, * p < 0.01).

Figure 4

Statistical analysis. Statistical analysis of pre-contrast T₁, post-contrast T₁ and ΔR₁ for infarcted, remote and control tissue (* p < 0.01).

Figure 5

Correlation between regional ΔR₁ and Gd concentration. Correlation between regional ΔR₁ values determined by in vivo CMR and Gd concentration determined by ex vivo ICP-MS. The analysis was done either on all individual heart segments (A) or after averaging all values in each heart slice (B).