Detection of Myocardial Ischemia Using Cardiovascular MRI Stress T1 Mapping: A Miniature-Swine Validation Study

Abstract

Purpose: To assess the efficacy of cardiac MRI stress T1 mapping in detecting ischemic and infarcted myocardium in a miniature-swine model, using pathologic findings as the reference standard.

Materials and methods: Ten adult male Chinese miniature swine, with coronary artery stenosis induced by an ameroid constrictor, and two healthy control swine were studied. Cardiac 3-T MRI rest and adenosine triphosphate stress T1 mapping and perfusion images, along with resting and late gadolinium enhancement images, were acquired at baseline and weekly up to 4 weeks after surgery or until humanely killed. A receiver operating characteristic analysis was used to analyze the performance of T1 mapping in the detection of myocardial ischemia.

Results: In the experimental group, both the infarcted myocardium (ΔT1 = 10 msec ± 2 [SD]; ΔT1 percentage = 0.7% ± 0.1) and ischemic myocardium (ΔT1 = 10 msec ± 2; ΔT1 percentage = 0.9% ± 0.2) exhibited reduced T1 reactivity compared with the remote myocardium (ΔT1 = 53 msec ± 7; ΔT1 percentage = 4.7% ± 0.6) and normal myocardium (ΔT1 = 56 msec ± 11; ΔT1 percentage = 4.9% ± 1.1). Receiver operating characteristic analysis demonstrated high diagnostic performance of ΔT1 in detecting ischemic myocardium, with an area under the curve (AUC) of 0.84 (P < .001). Rest T1 displayed high diagnostic performance in detecting infarcted myocardium (AUC = 0.95; P < .001). When rest T1 and ΔT1 were combined, the diagnostic performance for both ischemic and infarcted myocardium were improved (AUCs, 0.89 and 0.97, respectively; all P < .001). The collagen volume fraction correlated with ΔT1, ΔT1 percentage, and Δ extracellular volume percentage (r = -0.70, -0.70, and -0.50, respectively; P = .001, .001, and .03, respectively).

Conclusion: Using histopathologic validation in a swine model, noninvasive cardiac MRI stress T1 mapping demonstrated high performance in detecting ischemic and infarcted myocardium without the need for contrast agents.Keywords: Coronary Artery Disease, MRI, Myocardial Ischemia, Rest T1 Mapping, Stress T1 Mapping, Swine Model Supplemental material is available for this article. © RSNA, 2023See also commentary by Burrage and Ferreira in this issue.

Keywords: Coronary Artery Disease; MRI; Myocardial Ischemia; Rest T1 Mapping; Stress T1 Mapping; Swine Model.

Graphical abstract

Figure 1:

Diagram of MRI protocol sequence shows images from the survey to late gadolinium enhancement (LGE). AIF = arterial input function, ATP = adenosine triphosphate, ECV = extracellular volume, Gd = gadolinium.

Figure 2:

Myocardial infarction (purple circle) with corresponding rest and stress cine MRI, perfusion, T1 mapping, and late gadolinium enhancement (LGE) images and 2, 3, 5-triphenyltetrazolium (TTC) staining. A region is highlighted in the LGE image, with pale coloring in the corresponding area on the TTC image (arrows). ECV = extracellular volume, MBF = myocardial blood flow, MPR = myocardial perfusion reserve.

Figure 3:

Ischemic (pink circles) and remote (green circles) myocardium on corresponding rest and stress cine MR images, perfusion images, T1 mapping, and late gadolinium enhancement (LGE) images and 2, 3, 5-triphenyltetrazolium (TTC) stain. The defect is visible on the stress perfusion image (arrow), whereas there is no defect at rest perfusion (ie, reversible defect). ECV = extracellular volume, MBF = myocardial blood flow, MPR = myocardial perfusion reserve.

Figure 4:

Bar graphs show myocardial T1 and extracellular volume (ECV) at rest and during adenosine triphosphate (ATP) stress. Reactivity is shown for different types of myocardia (bottom row).* P < .05. The T1 and ECV values in infarcted myocardium are higher compared with values in ischemic myocardium at both rest and stress. There was no evidence of a difference in rest T1 value and ECV of the ischemic myocardium and remote myocardium compared with the normal myocardium in the control group (all P > .05). The ∆T1, ∆ECV, ∆T1 percentage, and ∆ECV percentage in ischemic myocardium are lower than those in the remote myocardium and normal myocardium.

Figure 5:

Graphs show the correlation between collagen volume fraction (CVF) and ∆T1 (left), ∆T1 percentage (middle), and change in extracellular volume fraction (∆ECV) percentage (right).

Figure 6:

Receiver operating characteristic analyses show the diagnostic performance of T1 and T1 reactivity in the detection of infarcted myocardium (A, C) and ischemic myocardium (B, D) at region of interest (ROI) and segment levels. AUC = area under the receiver operating characteristic curve.