Z Gastroenterol 2012; 50 - K253
DOI: 10.1055/s-0032-1324188

Non-invasive quantification of hepatic fibrosis in toxic and cholestatic mouse models by magnetic resonance relaxometry (MRR)

K Hochrath 1, A Müller 2, M Krawczyk 1, A Bücker 2, F Lammert 1
  • 1Saarland University Medical Center, Department of Medicine II, Homburg, Germany
  • 2Saarland University Medical Center, Department of Diagnostic and Interventional Radiology, Homburg, Germany

Aims: To date studies in mouse models of hepatic fibrogenesis have been hampered by the need for tissue specimens to quantify liver fibrosis. Hence, the development of non-invasive methods is crucial to monitor progression, resolution and potential therapies for hepatic fibrosis and to reduce the number of experimental animals. Our aim now was to develop a non-invasive magnetic resonance-based quantification of liver in mice with various aetiologies of hepatic fibrogenesis.

Methods: We analysed 13 BALB/cJ mice after toxic induced liver fibrosis (1.4mg CCl4/kg/wk i.p. for 6 wks) and 23 Abcb4 knockout (Abcb4-/- ) mice with secondary biliary fibrosis as well as 27 wild-type fibrosis-free controls. To assess hepatic fibrosis, we ascertained relative collagen areas after Sirius red staining and collagen contents by hydroxyproline. Magnetic resonance relaxometry (MRR) was performed on a horizontal bore 9.4 Tesla animal scanner (Bruker Biospec 9.4/200) with a circular polarized receive/transmit coil. The relaxation times T1, T2 and T2* as well as signal intensity were acquired with turbo spin echo and multiple gradient echo techniques.

Results: Compared to untreated wild-type mice, animals treated with CCl4 and Abcb4-/- mice demonstrate significantly (p<0.01) increased hepatic collagen contents. The CCl4-challenged mice display a significant (p<0.05) increase in T2* (7.0, range 5.0–11.4 vs. 8.0, range 7.1–8.7ms) relaxation times as compared to controls. Furthermore, T2* times correlate significantly (p<0.05) with relative collagen area in CCl4-treated animals. Nevertheless, the T1 relaxation times do not differ between CCl4-treated and controls. In contrast to the above results, Abcb4-/- mice display a significant (p<0.001) rise in T1- (923.8, range 717.9–1248.8 vs. 1073.7, range 681.4–1507.8ms) but no differences in T2- and T2* relaxation times in comparison to controls.

Conclusions: Our study demonstrates the feasibility of MRR as a non-invasive method to discriminate between fibrotic and non-fibrotic liver tissue in mice independent of the aetiologies of hepatic injury. Additional studies are currently being performed to develop algorithms allowing refined differentiation of fibrosis stages and assessment of fibrosis resolution.