The Effect of Hypertrophy in CardioCEST Magnetization Transfer Contrast
Scott William Thalman1, Zhengshi Yang2, Ashley Pumphrey2, and Moriel Vandsburger1,2,3

1Department of Biomedical Engineering, University of Kentucky, Lexington, KY, United States, 2Saha Cardiovascular Research Center, University of Kentucky, Lexington, KY, United States, 3Department of Physiology, University of Kentucky, Lexington, KY, United States

Synopsis

Remodeling of the myocardium via hypertrophy and fibrosis increases the risk of adverse cardiac events. Quantitative magnetization transfer weighted imaging has shown promise as a method to identify fibrosis, however the effects of hypertrophy on such measures remains unknown. Using a murine model of chronic Angiotensin-II stimulation characterized by robust hypertrophy with little fibrosis, we demonstrate that despite large increases in myocardial mean cross-sectional area (AngII= 4825μm2±717 vs Saline= 2372μm2±158), no change occurs in the cardioCEST derived measure of magnetization transfer ratio (MTR) (AngII= 30.8%±7.3 vs Saline= 27.2%±8.6). Thus, increases in MTR due to fibrosis are unbiased by concomitant hypertrophy.

Target Audience

Individuals interested in endogenous imaging of cardiac remodeling.

Purpose

To demonstrate the effects of myocardial hypertrophy on MT-weighted cardioCEST MRI.

Introduction

Cardiovascular disease is characterized by both increased interstitial fibrosis and extracellular volume, and by hypertrophic remodeling of individual cardiomyocytes. These structural changes are associated with increased risk of heart failure, arrhythmia, and sudden cardiac death. Techniques to characterize myocardial tissue remodeling with gadolinium are limited in populations with reduced renal function who are at increased risk for adverse cardiac events. The endogenous contrast mechanism of magnetization transfer (MT) is influenced by changes in tissue structure, and recently MT-weighted CMR approaches have shown promise in identifying cardiac fibrosis(1,2,3). However, the impact of increased intracellular macromolecule concentration concomitant with hypertrophy on MT contrast remains unknown. In this study we use a murine model of chronic Angiotensin-II (AngII) stimulation to demonstrate that hypertrophy has little effect on MT contrast generated using cardiac chemical exchange saturation transfer (cardioCEST)(4).

Methods

Pulse Sequence: CEST encoding used a saturation pulse train of 88 spatially non-selective Gaussian pulses (flip angle= 270°, B1average= 5.25μT, bandwidth= 200Hz, duration= 8.8ms, saturation frequency offsets= 6 and 15ppm). Four averages of one phase encoding step for each cardiac phase were acquired following each saturation train. A constant repetition time (TR) cine gradient echo sequence (TR/TE= 10.2/3.5 ms, flip angle= 10°) was used to encode CEST contrast into the steady state longitudinal magnetization. Data acquisition was prospectively triggered using combined ECG and respiratory waveforms. Dummy pulses were used to maintain steady state magnetization between heart beats in cases of heart rate variability and during respiratory motion. Images were acquired in 1 mid-ventricular short-axis slice with parameters: FOV= 25.6x25.6cm, Matrix= 192x192, slice thickness= 1mm. Animal Model: Adult male C57Bl/6 mice (n=12) received either constant infusion of AngII (1000ng/kg/min, BACHEM, n= 7) or saline (n= 5) via mini osmotic pump (Alzet, Cupertino, CA USA). MRI was performed prior to and 10 days after pump implantation, with anesthesia maintained using 1.25% isoflurane in oxygen and body temperature maintained using circulating thermostated water. Imaging: All imaging was performed on a 7T Bruker ClinScan (Bruker Biospin, Ettlingen, Germany) using a cylindrical volume coil for excitation and a 4-channel phased array surface coil for reception. A reference image (saturation offset= 333ppm, saturation flip angle= 1°) was acquired to normalize signal for the receiver coil profile (See Fig-1). Image Analysis: Following a manual registration protocol to account for slight changes in bed position between scans, the magnetization transfer ratio (MTR) was calculated on a pixel-wise basis as MTR(ω)= [(SRef-S(ω))/SRef]*100. Myocardial borders were traced manually and the MTR was averaged over the entire myocardium. Regions of interest were then drawn surrounding the left anterior descending artery (LADA), the posterior right ventricular insertion (RVI) point, the septum (Sept) and the lateral wall (LW). Histology: Immediately following post-treatment scanning, all mice were euthanized and hearts were isolated, fixed in formalin, sliced transversely at the imaging location and stained with picrosirius red. Sections from four AngII-treated mice and two saline-treated mice were also stained with wheat germ agglutinin (WGA). Images were digitized using a Nikon A1R confocal microscope at 10x magnification. A thresholding method as described by Beliveau et al(5) was used to quantify the collagen volume fraction in the LAD, RVI, septum and lateral wall ROIs. Using the same ROIs, five mid-wall circumferentially oriented cardiomyocytes were selected in the WGA-stained sections to measure mean cross-sectional area (MCA) using ImageJ (NIH, Bethesda, Maryland USA).

Results

Angiotensin-II treatment provoked a perivascular pattern of fibrosis, however the collagen volume fraction was not significantly increased (AngII= 2.86% ± 0.9, Sal= 1.63% ± 0.3). Angiotensin-II treatment provoked a significant increase in MCA when compared to saline infusion (AngII= 4825μm2 ± 717, Sal= 2372μm2 ± 158, p= 0.01) (See Fig-2). Parametric maps (See Fig-1) reveal no significant changes in MTR at 6ppm in AngII-treated mice (30.8% ± 7.3) relative to saline-treated mice (27.2% ± 8.6) or pretreatment scans (25.8% ± 4.2) (See Fig-3). Linear regression revealed no correlation between regional MCA and MTR values (p= 0.943) (See Fig-4).

Discussion

Increased extracellular volume that is concomitant with myocardial fibrosis leads to a reduction in MT when compared to healthy myocardium. We examined whether cardiomyocyte hypertrophy, which is characterized by increased production of intracellular proteins, would impart an opposing MT effect. By using a mouse model of AngII induced hypertrophy that is largely absent of interstitial fibrosis, we demonstrated using a cardioCEST approach that cardiomyocyte hypertrophy does not significantly alter MT contrast.

Acknowledgements

Dr. Debra Rateri, for help with the AngII model; Dr. Ja Brandon for help with microscopy; Dr. Wendy Katz for histological sample preparation.

References

(1) Stromp et al. JCMR 2015; 17(1):90. (2) Weber et al. MRM. 2009; 62(3): 699-705. (3) Weiss et al. Radiology 2003; 226(3):723-30 (4) Pumphrey et al. NMR Biomed. 2015 In Press. (5) Beliveau et al. Comput Biol Med 2015; 65:103-13.

Figures

Figure 1: Representative cardioCEST images and MTR maps demonstrating no change in MTR in AngII-treated compared to saline-treated mice.

Wheat germ agglutinin staining for cell membranes demonstrates hypertrophy of cardiomyocytes in AngII-treated mice in all regions of the heart. White arrows indicate cell membranes.

Figure 3: Mean MTR values averaged over the entire myocardium. AngII-treatment evoked only a slight, nonsignificant increase in MTR despite profound global hypertrophy. Note that MTR values are reduced at 15ppm vs 6ppm as the saturation frequency moves further away from the broad excitation bandwidth of bound protons and any elevation in MTR is abolished.

Figure 4: Mean cross-sectional area values demonstrate hypertrophy in all regions of the AngII-treated mice, however, when plotted against the corresponding regional MTR data there is no correlation between hypertrophy and MTR.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2535