Our results indicate that the signal intensity and heterogeneity of hepatocellular carcinomas on MR images correlate with the degree of VEGF expression in hepatocellular carcinomas.

Angiogenesis is important for tumor growth, and is regulated by angiogenetic factors such as vascular endothelial growth factor (VEGF). In the present study, we investigated whether or not expression of VEGF receptors (VEGFRs) is related to the proliferation of tumor cells in hepatocellular carcinoma (HCC). We simultaneously stained proliferation marker Ki-67 antigen and either VEGFR1 (Flt-1) or VEGFR2 (Flk-1) on paraffin-embedded tissue sections from 50 cases of surgically resected human HCC. Based on the staining pattern of VEGFRs, we classified the cases into 4 categories; receptor double-negative, Flt-1 single-positive, Flk-1 single-positive, receptor double-positive. Interestingly, the Ki-67 index was significantly lower in receptor doublenegative cases in comparison to that in either Flt-1 singlepositive or Flk-1 single-positive cases (P=0.0491, P=0.0196, respectively). Moreover, the index was also significantly lower in receptor double-positive cases in comparison to either Flt-1 single-positive or Flk-1 single-positive cases (P=0.0026, P<0.0001, respectively). We further investigated 35 cases showing a Ki67 index >10% to determine the expression of VEGFRs on Ki-67 antigen-positive proliferating cells. Surprisingly, the histological grade of HCC and the expression pattern of VEGFRs showed a characteristic relation; the well-differentiated HCC cases were all distributed in the Flk-1-positive group (7/7), moderately differentiated HCC cases were distributed in either the Flt-1 or Flk-1 single-positive group (20/21), and poorly differentiated HCC cases were predominantly distributed in either the receptor doublenegative or double-positive group (6/7). These findings suggest that the expression pattern of VEGFRs influences the histological differentiation of HCC.

We summarize and discuss our previous research results on the correlation between findings on magnetic resonance (MR) imaging and angiographically assisted computed tomography (CT) and the intensity of vascular endothelial growth factor (VEGF) expression in hepatocellular carcinoma (HCC) and in the surrounding nontumorous liver. MR images (n = 22), CT during arterial portography (n = 20), and CT hepatic arteriography (n = 17) were retrospectively correlated quantitatively and qualitatively with VEGF expression in HCCs and in the surrounding liver assessed by western blotting. HCC-to-liver contrast-to-noise ratio correlated with VEGF expression index (VEGF(IND)) values of HCCs inversely on opposed-phase, T1-weighted, spoiled gradient recalled-echo (GRE) images, directly on T2-weighted, fast spin-echo images, and marginally and inversely on gadolinium-enhanced hepatic arterial-phase GRE images. On T2-weighted fast spin-echo images, standard deviation ratio of HCCs correlated directly with VEGF(IND) values of HCCs. By CT hepatic arteriography, the contrast-enhancement index of HCCs showed a moderate inverse correlation with VEGF(IND) values of HCCs, and the contrast-enhancement index of the liver showed marginal, moderate direct correlation with VEGF(IND) values in the liver. Heterogeneities of HCCs on images correlated directly with VEGF(IND) values of HCCs on opposed-phase T1-weighted GRE images, T2-weighted fast spin-echo images, hepatic arterial-phase GRE images, equilibrium-phase GRE images, and CT hepatic arteriogram. Our results may reflect that MR signal intensity, hepatic arterial vascularity, and heterogeneity of HCCs on CT or MR images are closely related to the intensity of VEGF expression in HCC as upregulated by hyper- or hypoxia in HCCs. Although the real effects of our results on radiologic practice are debatable at this moment, we believe that our results may help future radiologic practice in conjunction with biomolecular or genetic treatment for HCCs.

This study on the expression of VEGF might be useful to estimate the liver condition and the clinicopathological features of HCC.

We summarized and discussed our previous research results on correlation between magnetic resonance (MR) imaging findings and vascular endothelial growth factor (VEGF) expression in benign or borderline hepatocellular nodules in cirrhosis, hepatocellular carcinomas (HCCs), and in the surrounding liver. Magnetic resonance images were retrospectively correlated quantitatively and qualitatively with VEGF expression in hepatic nodules and in the surrounding liver. By immunohistochemistry, hepatic nodules with moderate to strong immunoreactivity for VEGF showed higher T1 signal intensity, and those with intense immunoreactivity for VEGF showed higher T2 signal intensity. By Western blotting, HCC-to-liver contrast-to-noise ratio correlated with VEGF indices (VEGFs) of hepatocellular carcinomas inversely on opposed-phase T1-weighted, directly on T2-weighted, and marginally and inversely on gadolinium-enhanced hepatic arterial-phase images. On T2-weighted images, standard-deviation ratio of hepatocellular carcinomas correlated directly with VEGFs of hepatocellular carcinomas. Heterogeneities of hepatocellular carcinomas on MR images correlated directly with VEGFs of HCCs on opposed-phase T1-weighted, T2-weighted, hepatic arterial-phase, and equilibrium-phase images. Our results may reflect that MR signal intensity, hepatic arterial vascularity, and heterogeneity of hepatocellular nodules on MR images are closely related to the intensity of VEGF expression as up-regulated by hyper- or hypoxia in the nodules. Gadolinium-enhanced MR imaging may be useful to monitor ischemic state of hepatocelluar nodules. Although real impacts of our results on radiologic practice have been still debatable, we believe that our results may help future radiologic practice in conjunction with biomolecular or genetic treatments for hepatocellular carcinomas.