Discrimination of Hepatic Inflammation and Fibrosis with Magnetic Resonance Elastography
Meng Yin1, Kevin J. Glaser1, Harmeet Malhi2, Amy Mauer2, Anuradha Krishnan2, Taofic Mounajjed3, Jason Bakeberg4, Christopher Ward4, Ruisi Wang2, Douglas Simonnetto2, Shennen Mao5, Jaime Glorioso5, Faysal Elgilani6, Vijay Shah2, Scott Nyberg6, Armando Manduca1, and Richard L. Ehman1

1Radiology, Mayo Clinic, Rochester, MN, United States, 2Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN, United States, 3Pathology, Mayo Clinic, Rochester, MN, United States, 4Nephrology and Hypertension Research, Mayo Clinic, Rochester, MN, United States, 5Surgery, Mayo Clinic, Rochester, MN, United States, 6Transplant Center, Mayo Clinic, Rochester, MN, United States

Synopsis

To investigate the utility of MRE-derived mechanical properties in discriminating hepatic inflammation and fibrosis in early-stage of chronic liver diseases, we performed multifrequency 3D MRE on five different in vivo animal models with chronic liver diseases. Liver stiffness and phase angle derived from complex shear modulus were selected for evaluation. Results demonstrated distinct and potentially characteristic changes in these mechanical properties with hepatic inflammation, fibrosis and increased portal pressure. The findings offer preliminary evidence of the potential to extend MRE to distinguish and independently assess necroinflammatory and fibrotic processes in the early phase of chronic liver diseases.

Purpose

To investigate MR Elastography (MRE)–assessed mechanical properties of shear stiffness and phase angle to distinguish hepatic inflammation and fibrosis in several in vivo animal models with progressively developed chronic liver diseases that involve inflammation, fibrosis, congestion and portal hypertension.

Methods

All activities related to animal subjects were reviewed and approved by our institutional animal care and use committee. A total of 188 animals, including 183 mice and 5 pigs that represent five different liver diseases with a varying combination and extent of hepatic inflammation, fibrosis, congestion and portal hypertension. 1) 55 wild type male mice with fast food diet (FFD) induced nonalcoholic fatty liver disease or steatohepatitis (NAFLD/NASH) and 40 gender and age-matched control mice; 2) 19 knockout mice with autosomal recessive polycystic kidney disease (ARPKD) and 18 gender and age-matched control mice; 3) 12 wild type mice with toxic chemical carbon tetrachloride (CCl4) induced hepatic fibrosis and 12 gender and age-matched control mice; 4) 16 wild type mice with partial inferior vena cava ligation (pIVCL) surgery induced venous congestion and 11 gender and age-matched sham mice; 5) 5 inborn fumarylacetoacetate hydrolase (FAH)-deficient pigs. An in vivo multifrequency 3D MRE examination was performed on each animal monthly (models #1&5) or one time (models #2, 3& 4) before euthanization1. The complex shear modulus was calculated with outputs of liver stiffness and phase angle at multiple mechanical frequencies (80-200 Hz in mice, 40-80 Hz in pigs).

Results

Results validated the well-established fact that both shear storage modulus and shear stiffness of the liver increased progressively with the severity of chronic liver diseases (p<0.05). However, it was not able to detect very early onset of inflammation at high frequencies. Results also demonstrated that phase angle increased significantly at the early onset of inflammation (p<0.05), even before histologically detectable macrophage transformation or migration. Moreover, it has better discrimination ability at lower frequencies. With the progressively developed fibrosis with coexisting/diminishing inflammation, phase angle decreased and lost its sensitivity to inflammation in models #1, 2&3 (p>0.1). In the congestive liver model without significant fibrosis, phase angle was inversely proportional to portal pressure (r=-0.6129, p<0.001, model #4), but increased significantly in response to cirrhosis induced spontaneous portal hypertension (model #5).

As shown in Figure-1 of the NAFLD/NASH model, monthly MRE results demonstrated that liver stiffness successfully detected the onset of inflammation (16 weeks, p=0.04) and the onset of fibrosis (24 weeks, p=0.04). Both loss modulus and phase angle increased significantly as early as 4 weeks (80Hz, p=0.005) before histologically detectable inflammation and stayed significantly high from 12 until 40 weeks. From 44 to 48 weeks, there were no significant change observed in phase angle (p>0.1).

As shown in Figure-2 of the ARPKD model, fibrosis extent increased progressively with age; while the portal inflammation extent increased significantly and stayed consistently mild in the ARPKD mice of 3 and 6 months old, but not detectable in the histologic analysis of 1 month old mice. Liver stiffness successfully distinguished ARPKD mice for all age groups (80Hz only, not for higher frequencies). Phase angle showed significant augmentation in 1 month old mice (80Hz, p<0.0001), but no change in older mice (p>0.1).

As shown in Figure-3 of the CCl4 model, peak necroinflammation was observed at week 1, but decreased at week 2 and 4 then diminished at week 6; the onset of fibrosis was at week 2 and increased thereafter. Liver stiffness successfully detected the fibrosis at week 4 and after (200Hz, p<0.03); while phase angle discriminated early inflammation as early as week 1 (200Hz, p=0.03).

As shown in Figure-4 of the pIVCL model, portal pressure increased significantly for all time groups with insignificant inflammation and fibrosis developed in the congestive liver. Liver stiffness positively correlated with the portal pressure (best fit at 200Hz, r=0.8738, p<0.0001), while phase angle negatively correlated with portal pressure (best fit at 200Hz, r=-0.6129, p<0.001).

As shown in Figure-5 of the FAH model, both liver stiffness and phase angle increased and reached their maximum value through the animal’s lifespan at the time of spontaneous portal hypertension with gastrointestinal bleeding. In pigs without GI bleeding, at one month of age with onset of abnormality, the liver stiffness maintained its lowest value, while the phase angle reached its maximum values.

Discussion and Conclusion

Results in five animal models indicated that the phase angle, defined by the complex shear modulus, is useful for discriminating between inflammation and fibrosis alone in the early stage of liver diseases. It may also distinguish venous congestion and portal hypertension associated with cirrhosis in the advanced stage of liver diseases.

Acknowledgements

This work has been supported by NIH grants EB017197 and EB001981.

References

1. Yin, M., et al. Advanced assessment of liver diseases with magnetic resonance elastography in animal models. in International Society for Magnetic Resonance in Medicine. 2015. Toronto, Canada.

Figures

Figure 1 NAFLD/NASH Mouse Model.

Chart (a) shows that steatosis extent increased gradually from 1 to 12 weeks, then maintained around 30-35%. At 24 weeks, mild fibrosis started to develop. At 48 weeks, moderate fibrosis was developed. All MRE measurements were normalized by their control groups. Liver stiffness began to increase at 16 weeks, as shown in chart (b). Loss modulus and phase angle detected abnormality and increased significantly as early as 4 weeks, as shown in chart (c) and (d).


Figure 2 ARPKD Mouse Model.

ARPKD mice had mild fibrosis but no observable portal inflammation after wean, then developed mild inflammation and moderate fibrosis after 3 months old, as shown in charts (a) and (b). Corresponding histologic pictures were illustrated on the left respectively. ARPKD mice had progressively increased liver stiffness with the fibrosis extent, as shown in chart (c). The phase angle only elevated significantly in the 1-month-old ARPKD mice, as shown in chart (d).


Figure 3 CCl4 Mouse Model.

For the mice with CCl4 administration, peak inflammation and zone 3 necrosis were observed during week 1, fibrosis extent progressively increased with the time duration after week 2, as shown in chart (a). Liver stiffness elevated significantly in the mice with CCl4 injection more than 4 weeks, as shown in chart (b). The early liver injury was distinguished by significantly increased phase angle only, as shown in chart (c).


Figure 4 pIVCL Mouse Model.

Chart (a) demonstrates that portal pressure increased significantly in pIVCL mice compared with the time matched SHAM groups (p<0.001). However, no significant fibrosis (i.e., hydroxyproline concentration) was developed in these pIVCL mice for each time group (p>0.07). Chart (b) and (c) illustrates that MRE-assessed liver stiffness (200Hz), and phase angle (200 Hz) significantly correlates with portal pressure measurements.


Figure 5 FAH Pig Model.

Chart (a) shows monthly MRE data from a male pig through his life span. Both shear stiffness and phase angle increased to their maximum values at the time of spontaneous GI bleeding. Chart (b) shows monthly MRE data from a female pig through her life span. At the onset of abnormality, phase angle increased to its maximum value, while shear stiffness did not change compared with the baseline values.




Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0361