Assessment of Myocardial Microstructural Dynamics by In Vivo Diffusion Tensor Cardiac Magnetic Resonance

Figures

• 

  FIGURE 1. LV Myocardial Helical and Sheetlet Microstructures (A) Multi-slice tractogram obtained from ex vivo diffusion tensor cardiac magnetic resonance depicts the primary diffusion eigenvector (E1) direction color-coded according to the helix angle (HA). (B) Schematic diagram of the helical structure of cardiomyocytes with zoom. Sheetlet angle (SA) rotation during cardiac contraction is depicted during diastole (C to F) and systole (G to J) with myocardial sheetlet microstructures shown in relaxed (F) and contracted (J) states. In mid-myocardial zooms of histology sections cut perpendicular to the local cardiomyocytes acquired from relaxed (C) and contracted (G) ex vivo heart samples, a yellow ellipse surrounds a single sheetlet composed of closely packed cardiomyocytes bounded by the pale cracks of shear layers. Additionally, the sheetlet angle is defined as the angle between the sheetlet and the local epicardial left ventricular (LV) wall. Here, SA varied from a low value (SA ~15°) in diastole (D) to a high value (SA ~60°) in systole (H). See Online Video 1. Blue = left-handed epicardial helixes; yellow = mid-myocardium circumferential alignments; red = right-handed endocardial…

• 

  FIGURE 2. E2A Changes Throughout the Cardiac Cycle (A) In vivo secondary diffusion tensor eigenvector angle (E2A) maps are depicted at multiple points of the cardiac cycle, together with a plot of E2A throughout the entire cardiac cycle for all in vivo experiments. As the measured planes swivel from diastole to systole, E2A increases and their color changes from blue to red. (B) In situ and (C) ex vivo E2A maps depict relaxed and contracted hearts after injection of potassium chloride (KCl) and barium chloride (BaCl2). (D) Long-axis histological cuts with mesocardial layer details show sheetlets in relaxed and contracted heart tissue samples, with their corresponding angular histograms, demonstrating the sheetlet and cleavage plane reorientation. See Online Video 2. DT-CMR = diffusion tensor cardiac magnetic resonance.

• 

  FIGURE 3. E2A Increase and Wall Thickening Color-coded plots of E2A change (ΔE2A) versus wall thickness changes (ΔWT) are shown for both in vivo (A) and in situ (B) with BaCl2 arrest acquisitions. The strong correlation between ΔE2A and ΔWT supported the thesis that the reorientation of the sheetlets contributes significantly to systolic wall thickening. Abbreviations as in Figures 1 and 2.

• 

  FIGURE 4. E1A Distribution Through Depth of LV Wall (A) Typical in vivo primary diffusion eigenvector angle (E1A) maps derived from DT-CMR at 6 time points in the cardiac cycle, and mean E1A against wall depth plots at systole and diastole for all in vivo experiments; inset illustrates E1 and E1A changes with LV wall depth. (B) Typical in situ E1A maps from DT-CMR after arrest by injection of KCl and BaCl2. See Online Video 2. (C) Typical ex vivo E1A maps at 3 mid-ventricular slices and E1A against wall depth plots in relaxed and contracted hearts. (D) Typical mid-layer wall-parallel histology sections showing circumferentially aligned cardiomyocytes that appear approximately horizontal on the images and expanded views, and their respective radial histograms measuring HA ~0°, with line plots of HA against wall depth derived from multiple wall parallel sections of all relaxed and contracted hearts. See Online Videos 3 and 4. endo = endocardium; epi = epicardium; meso = mesocardium; other abbreviations as in Figures 1 and 2.

• 

  FIGURE 5. Plots of Median E2A (A) Variations in diastolic and systolic secondary diffusion tensor eigenvector angle (E2A) are seen among patients with hypertrophic cardiomyopathy (HCM) or dilated cardiomyopathy (DCM) as well as control subjects. (B) These 3 groups are clearly differentiated when E2A is measured against left ventricular ejection fraction (LVEF).

• 

  FIGURE 6. E1A and E2A Maps The E1A maps show similar E1A distributions in all populations and contractile states whereas E2A maps are color coded according to absolute E2 angle. E2A changes are from blue in diastole to red in systole in healthy control subjects. The HCM example takes a healthy systolic conformation (wall perpendicular sheetlets) but an incomplete diastolic conformation (mix of wall parallel and wall perpendicular sheetlets). The DCM example takes a healthy diastolic conformation (wall parallel sheetlets) but an incomplete systolic conformation (mix of wall parallel and wall perpendicular sheetlets). Abbreviations as in Figures 2, 4, and 5.

• 

  FIGURE 7. E1A and E2A Histograms at Diastole and Systole Whereas E1A histograms overlap for all groups in both diastole and systole, with a slightly broader E1A distribution in systole, E2A distribution in diastole is predominantly low in control subjects, swine (preclinical), and DCM, with a relatively wide even distribution in HCM. In systole, E2A is predominantly high in control subjects, swine, and HCM, with a relatively wide even distribution in DCM. Abbreviations as in Figures 2, 4, and 5.

• 

  CENTRAL ILLUSTRATION. Assessment of Myocardial Microstructural Dynamics by In Vivo Diffusion Tensor Cardiac Magnetic Resonance Myocardial microstructure dynamics can be characterized by in vivo diffusion tensor cardiac magnetic resonance (DT-CMR). E2 angle (E2A) is a DT-CMR index of myocardial sheetlet rotation; increase in E2A from diastole to systole was consistently observed in healthy swine in vivo, in situ, and ex vivo, and correlated well with histology. E2A changes in healthy volunteers match closely those in healthy swine. Varying E2A patterns emerged in patients with hypertrophic cardiomyopathy (HCM) and dilated cardiomyopathy (DCM). E2A mobility was highest for control subjects and reduced for HCM and DCM. Sheetlet function was abnormal in DCM with altered systolic conformation and reduced mobility, contrasting with HCM, which showed reduced mobility with altered diastolic conformation. These novel insights significantly improve understanding of contractile dysfunction at a level of noninvasive interrogation not previously available in humans. SA = sheetlet angle.