Microvessel density and angiogenesis in primary hepatic malignancies: hock for Differential expression of CD31 and VEGFR-2 in hepatocellular carcinoma and intrahepatic cholangiocarcinoma

Background: Microvessel density is an indicator of tumor-driven neoangiogenesis. Hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) have distinct vascular patterns, which are also reflected in their imaging characteristics. Since a significant proportion of HCC are treated without biopsy confirmation, it is essential to discriminate HCC and ICC radiologically. The aim of our study was therefore to compare microvessel density and expression of VEGFR-2 in HCC and ICC, since these data may ultimately help us to better understand their imaging characteristics. Whereas CD31 documents vessel density, VEGFR-2 expression is an indicator of tumor-related neoangiogenesis. Methods: CD31 and VEGFR-2 expressing microvessels were quantified on tissue microarrays of 95 resection specimens of HCC and 47 cases of ICC. Microvessel density was evaluated by counting immuno-reactive vascular structures both within the tumor and adjacent liver control tissue, respectively. Further 16 cases of ICC were immunostained for CD31 and VEGFR-2 on full sections. Results: The frequency of VEGFR-2 (46.2/HPF; range 0-150) and CD31 (61.2/HPF; range 2.6-140) expressing vascular structures was significantly increased in HCC compared to adjacent liver parenchyma (VEGFR-2 33.3/ HPF, range 0-87, CD31 21.4/HPF, range 0-78, both p < 0,001). ICC revealed significantly less VEGFR2-positive microvessels (15.4/HPF; range 2-77) compared to matched control tissue (42.3/HPF; range 4.6-109), whereas microvessel density with CD31 was comparable between ICC and adjacent liver (32.1/HPF; range 5.3-78 versus 28.0/HPF; range 5.3-57; p = 0.89). In ICC, the tumor-to-normal microvessel density ratio was 0.38 for VEGFR-2 and 1.24 for CD31. These ratios were nearly identical (VEGFR: 0.38; CD31: 0,97) for the 16 cases of ICC studied on whole sections, confirming the validity of the TMA approach. In contrast, ratios of VEGFR-2 and CD31 in HCC vs. adjacent liver were significantly higher (VEGFR: 2.23; CD31: 6.57). Expression of VEGFR-2 by tumor cells was not observed in any of the cases. Conclusions: HCC and ICC differ significantly in their microvessel density, confirming the hypovascular nature of ICC as compared to the hypervascularity of HCC. Of note, inverse tumor-to-normal ratios of microvascular VEGFR-2 expression between the two neoplasms indicate distinct features of neoangiogenesis. Whether these differences can be exploited for improvements in imaging of hepatic tumors and may play a role for antiangiogenic treatment strategies requires further studies.