Mouse Hepatic Tumor Vascular Imaging by Experimental Selective Angiography

Abstract

Purpose: Human hepatocellular carcinoma (HCC) has unique vascular features, which require selective imaging of hepatic arterial perfusion and portal venous perfusion with vascular catheterization for sufficient evaluation. Unlike in humans, vessels in mice are too small to catheterize, and the importance of separately imaging the feeding vessels of tumors is frequently overlooked in hepatic tumor models. The purpose of this study was to perform selective latex angiography in several mouse liver tumor models and assess their suitability.

Materials and methods: In several ectopic (Lewis lung carcinoma, B16/F10 melanoma cell lines) and spontaneous liver tumor (Albumin-Cre/MST1fl/fl/MST2fl/fl, Albumin-Cre/WW45fl/fl, and H-ras12V genetically modified mouse) models, the heart left ventricle and/or main portal vein of mice was punctured, and latex dye was infused to achieve selective latex arteriography and/or portography.

Results: H-ras12V transgenic mice (a HCC and hepatic adenoma model) developed multiple liver nodules that displayed three different perfusion patterns (portal venous or hepatic artery perfusion predominant, mixed perfusion), indicating intra-tumoral vascular heterogeneity. Selective latex angiography revealed that the Lewis lung carcinoma implant model and the Albumin-Cre/WW45fl/fl model reproduced conventional angiography findings of human HCC. Specifically, these mice developed tumors with abundant feeding arteries but no portal venous perfusion.

Conclusion: Different hepatic tumor models showed different tumor vessel characteristics that influence the suitability of the model and that should be considered when designing translational experiments. Selective latex angiography applied to certain mouse tumor models (both ectopic and spontaneous) closely simulated typical characteristics of human HCC vascular imaging.

Fig 1. Selective angiography using blue latex dye.

(A) Systemic arteriography and (B) direct portography in a wild-type mouse. The stomach, duodenum, and jejunum were removed to allow a clear view of the vessels. (C-E) Magnified images of the liver after selective angiography and subsequent tissue clearing procedures (upper panels) and histological assessment (lower panels, hematoxylin-eosin staining) were performed. (C) Arteriography, (D) portography, and (E) hepatic venography specimens are shown. Blue latex dye particles are seen only in the target vascular compartment (yellow arrows), which seems to have been slightly dilated during injection (single arrow, hepatic artery; double arrow, portal vein; thick arrow, central vein). Blue latex dye particles are not detectable in non-target vascular compartments (white arrows).

Fig 2. Liver dual angiography performed in a H-ras12V over-expressing transgenic mouse that developed multiple spontaneous tumors.

Selective latex arteriography (blue latex) and portography (green acrylic paint) were sequentially conducted. Gross images of (A) the whole liver superior aspect (left) and inferior aspect (right), and sectioned slices of (B) the medial lobe, (C) left lateral lobe, (D) right lobe, and (E) caudate lobe. Nodules were perfused by arterial supply (red arrow), portal venous supply (blue arrow), or both (double arrows).

Fig 3. Conventional angiography of human hepatocellular carcinoma and latex angiography in mouse Lewis lung carcinoma model.

(A) Representative images of conventional angiography of a typical case of human hepatocellular carcinoma. From left, superior mesenteric arterial angiography, indirect portography, celiac trunk arterial angiography (early phase), and celiac trunk arterial angiography (late phase) images are shown. The branches of the tumor-feeding artery supplying the medial portion of the mass (white arrow) and displaced portal veins (black arrow) are clearly visible. Note that the main mass (white arrowheads) and the multiple small satellite nodules are extremely hypervascular in nature. (B-E) Selective latex angiography of a liver Lewis lung carcinoma ectopic tumor model. (B-C) Selective latex arteriography and the subsequent tissue clearance procedure were performed in a liver Lewis lung carcinoma-bearing mouse. (B) Gross image of the whole liver and (C) sectioned slices demonstrate the hypervascular mass (arrows) supplied by the hepatic artery. The enlarged, slightly tortuous tumor-feeding artery (red arrow) is clearly visible. (D-E) Selective latex portography and the subsequent tissue clearance procedure were performed in a liver Lewis lung carcinoma-bearing mouse. (D) Gross image of the whole liver and (E) sectioned slices show that portal venules do not enter the mass (arrows).

Fig 4. Selective latex arteriography of an Albumin-Cre/MST1 ^fl/fl /MST2 ^fl/fl mouse that developed multiple spontaneous hepatocellular carcinomas.

(A) Histologic examination (hematoxylin and eosin staining, ×100). (B) The whole liver and (C, D) magnified images from a surgical microscope display a hypervascular mass supplied by feeding arteries (black arrow) and enlarged prominent arteries (white arrow) in the tumor-free background liver. The dotted box in (B) indicates the region magnified in (C, D).

Fig 5. Latex angiography of Albumin-Cre/WW45 ^fl/fl mice that developed spontaneous hepatocelluar carcinomas.

Representative images of (A) histologic examination (hematoxylin and eosin staining, ×100), (B) selective latex arteriography and (C) portography. The masses appeared as extremely hypervascular lesions (red arrows) on selective latex arteriography, but were devoid of portal venous perfusion (black/white arrows).