Ultrasonic Perfluorohexane-Loaded Monocyte Imaging: Toward a Minimally Invasive Technique for Selective Detection of Liver Inflammation in Fatty Liver Disease.

Objectives-To investigate the utility of ultrasonic (US) perfluorohexane (PFH)-loaded monocyte imaging for detection of liver inflammation in fatty liver disease. Methods-C57Bl6 mice were injected intraperitoneally with tumor necrosis factor a and assessed by US PFH-loaded monocyte imaging 3 hours later. Echogenic monocytes were injected intravenously, leading to a transient increase in liver tissue intensity on a US perfusion scan. The contrast wash-out time constant was hypothesized to reflect the degree of inflammation. Next, we evaluated US PFH-loaded monocyte imaging in Ldlr(-/-) mice fed a 1-week high-fat/high-cholesterol diet as model for early developing nonalcoholic steatohepatitis. Adjunct analyses included tissue markers of liver inflammation. Results-Tumor necrosis factor alpha-injected mice showed a reduced wash-out time constant (mean +/- SEM, 0.013 +/- 0.003; n=8) compared to controls (0.054 +/- 0.009; n=7; P=.0006), indicative of increased inflammatory adhesion molecule expression on the endothelium. The Ldlr(-/-) mice fed the high-fat/high-cholesterol diet showed liver inflammation, as reflected by increased (3- to 4-fold) infiltration of inflammatory cells and increased (3- to 4-fold) gene expression of tumor necrosis factor alpha, integrin alpha M, intracellular adhesion molecule, and vascular cell adhesion molecule. However, in these mice, no difference was detected in the wash-out time constant as assessed by US PFH-loaded monocyte imaging (high-fat/high-cholesterol, 0.050 +/- 0.017; n=5; chow, 0.048 +/- 0.006; n=6; P=.91). Conclusions-Our results indicate that US PFH-loaded monocyte imaging is able to detect vascularly expressed inflammatory adhesion molecules in the mouse liver on direct endothelial stimulation. However, in our mouse model of early developing nonalcoholic steatohepatitis, we did not detect inflammation by this method, which may suggest that the time-dependent relationship between parenchymal and endothelial inflammation remains a fundamental issue to be addressed.