Multi parametric MRI evaluation of skeletal muscle growth and myopathies in mice
Kerryanne V. Winters1,2, Olivier Reynaud1,2, Dmitry S. Novikov1,2, Els Fieremans1,2, and Sungheon G. Kim1,2

1Center of Biomedical Imaging, Department of Radiology, NYU School of Medicine, New York, NY, United States, 2Center for Advanced Imaging Innovation and Research, NYU Langone Medical Center, New York, NY, United States

Synopsis

The random permeable barrier membrane (RPBM) model for diffusion tensor imaging (DTI) provides a non-invasive modality potentially useful for early and accurate diagnosis for the wide range of myopathies. We have utilized the DTI-RPBM method to assess myofiber changes in the Surface-to-Volume ratio S/V and sarcolemma permeability κ as markers in growing and wasting skeletal muscle. Preliminary results show that S/V and κ decrease in both wild-type and mdx mice, with a more pronounced change between weeks 3 and 4 in mdx mice. The conventional IDEAL-Dixon and T2 mapping measures were not sensitive enough to observe the same change.

Purpose

The most commonly used imaging method to assess for a the range of myopathies in vivo is the three-point Dixon MRI, which measures the fat fraction based on the fat and water composition in each voxel 1. This method relies on the fatty infiltration process that occurs during the later stages of disease, when significant wasting of the skeletal muscle is already present. The fat fraction also does not provide any information about changes in the lean tissue. Time dependent diffusion tensor imaging (DTI(t)) serves as a strong candidate for earlier diagnosis of myopathies as it can measure changes in the structural parameters of tissues at the cellular level 2-5. In our previous simulation and in vivo studies, we utilized the random permeable model (RPBM) with diffusion MRI data to quantify muscle fiber size and sarcolemma permeability 6. In this study, we have furthered our work by investigating the microstructural changes in the skeletal muscle of the murine model for Duchenne Muscular Dystrophy (DMD) longitudinally. The main purpose of this study is to examine the proposed DTI model and compare with conventional IDEAL-Dixon and T2 measures in the skeletal muscle of both normal and diseased mice.

Methods

To study the skeletal muscle properties during normal muscle growth and myopathies, we imaged C57/bl (n=22, males) and mdx mice (n=4, males) at 3, 4 and 6 weeks old. All images were acquired using a 7T Bruker Scanner with a Paravision 5 console and a volume transmit and receive coil. The MRI protocol included a T2-weighed rapid acquisition with relaxation enhancement (RARE) sequence (TR/TE=2s/35.4ms, RES=0.156x1.56 mm3, 10 slices) and a T1-weighed 3D FLASH sequence (TR/TE= 40ms/3.6ms, flip angle=10) to locate the muscle groups of the lower legs. A diffusion weighed (DW) stimulated-echo (STEAM) pulse sequence with 3D echo planar imaging (EPI) readout was used to acquire images with diffusion gradients in twenty non-collinear directions and one image without diffusion weighting. The DW-STEAM-EPI was run repeatedly for seven diffusion times t ranging from 20-700 ms with TR/TE=6s/27ms, FOV 2.20x2.20x1.20 cm and image matrix 64x64x8. The b value was held constant near 1000 s/mm2 by varying the diffusion weighting gradient strength as the diffusion times increased. For fat quantification, IDEAL-Dixon imaging was conducted to acquire images at six echo times with TR/TE=5 ms/7.1 ms, FOV 2.12x2.0x1.0 cm and image matrix 128x96x10. T2 relaxation times were measured using a spin-echo sequence of 32 echoes with TR=2.4 s and TE=7.1 ms. All together, an MRI session per mouse was about 2 hours. Data analysis was performed with a region of interest (ROI) drawn over the lower hindlimb muscle and repeated for each sequence. The average of second and third eigenvalues at each diffusion time t was assumed as the measure of diffusion D(t) perpendicular to the muscle fibers, to which the DTI-RPBM model was fitted. The DTI-RPBM model fitting provided estimates of surface-to-volume ratio S/V, membrane permeability κ, and unrestricted diffusion coefficient D0. Water and fat fractions were determined by the IDEAL data analysis method and T2 was calculated using a monoexponential fit to the multi-spin echo data above noise level.

Results and Discussion

Figure 1 shows an example of the lower leg muscle images acquired during an MRI session: T2 weighted, T1 weighted, fractional anisotropy (FA), water/fat images and T2 map. Figure 2 shows the eigenvalues from the mouse hindlimb and illustrates the relationship between diffusion eigenvalues perpendicular to the myofibers and diffusion time, and also the asymptotically linear dependence of D(t) perpendicular to the muscle fiber on 1/t1/2. Figure 3 shows longitudinal changes of DTI-RPBM parameters in which both S/V and κ of the mdx mice decrease substantially between week 3 and week 4 while D0 does not show any noticeable change. This pattern of change was not observed in T2 or fat fraction (Figure 4) although the fat fraction of mdx mice remained higher than that of the control mice. This observation suggests that the changes in the muscle fibers may not be fully reflected in the measurement of fat infiltration.

Conclusions

In this study, we investigated the complimentary roles of DTI and other imaging methods for a comprehensive assessment of myofiber development in the mouse hindlimb. Together, these methods can be used to investigate the effect that established and potential therapies would have on myopathies.

Acknowledgements

This work was supported by NIH R21 NS081230

References

[1] Wren TA, Bluml S, Tseng-Ong L, and Gilsanz V. Three-point technique of fat quantification of muscle tissue as a marker of disease progression in Duchenne muscular dystrophy: preliminary study. AJR Am J Roentgenol. 2008;190:W8–W12.

[2] Kim, S., Chi-Fishman, G., Barnett, A. S. and Pierpaoli, C. Dependence on diffusion time of apparent diffusion tensor of ex vivo calf tongue and heart. Magn Reson Med. 2005;54: 1387–1396. doi: 10.1002/mrm.20676

[3] Novikov, DS, Fieremans E, Jensen JH, and Helpren JA. Random walks with barriers. Nature Physics. 2011;7:508-14

[4] Novikov DS, Jensen JH, Helpern JA, and Fieremans E. Revealing mesoscopic structural universality with diffusion. PNAS. 2014;111 (14) 5088-5093

[5] Reynaud O, Winters KV, Hoang DM, Wadghiri YZ, Novikov DS and Kim SG. Surface-to-volume ratio mapping of tumor microstructure using diffusion weighted imaging. Magn Reson Med. 2015;doi:10.1002/mrm.25865

[6] Winters KV, Reynaud O, Novikov DS, Fieremans E and Kim SG. Multi parametric MRI evaluation of muscle development. ISMRM. 2015;p.4948

Figures

Figure 1: Anatomical T2-w RARE and T1-w images to visualize the lower leg muscles of a wild-type mouse (A, B) and the corresponding FA color map (B): red, left-right; blue, caudal-rostral; green, dorsal-ventral. The water and fat images are (D) and (E) respectively, with a T2 map (F).

Figure 2: The eigenvalue plots show the diffusion time dependence of the diffusion tensor and average in the hindlimb (A). The blue line is a DTI-RPBM model fit to the transverse D(t)=(λ2+λ3)/2. (B) D(t) replotted as a function of t-1/2 becomes asymptotically linear at small t-1/2, i.e. at large t.

Figure 3: DTI-RPBM was used to measure D0 (A), S/V (B) and κ (C) in control (green) and mdx (blue) mice with age. With normal development, WT mice had a slight decrease in S/V. The mdx mice showed a similar trend, but experienced a greater change between weeks 3-4.

Figure 4: T2 relaxation times (A) were stable throughout the developmental stages, while fat fractions (B) were shown to increase with age. Both the wild-type (green) and mdx (blue) mice showed a slight increase over the 3 weeks, however, the mdx mice were overall at a high fat composition.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0757