Quantification of angiogenesis by functional computed tomography in a Matrigel model in rats

Abstract

Rationale and objectives: The aim was to evaluate functional computed tomography (fCT) in the quantification of angiogenesis by comparing the tissue perfusion parameters measured by CT perfusion (CTP) software with histologic vascular parameters in a Matrigel model in rats. It was hypothesized that tissue perfusion parameters and histologic vascular parameters are related.

Materials and methods: In vivo angiogenesis assays were performed using Matrigel supplemented with escalating doses (0 ng [control group], 250 ng, and 1,000 ng) of recombinant rat vascular endothelial growth factor (VEGF164) subcutaneously injected into the backs of Sprague Dawley rats. On day 7, rats with Matrigel plug underwent fCT following a bolus injection of iodinated contrast medium. Using CTP software, fCT parameters were generated (blood flow [BF], blood volume [BV], mean transit time, and permeability-surface area product) and functional maps on the basis of a distributed parameter tracer kinetic model, the adiabatic approximation to the tissue homogeneity model. The animals were then sacrificed. Matrigel plug was sectioned into slices corresponding to the CT scan plane and stained with CD31 immunohistochemical stain. Histologic vascular parameters, including microvascular density (MVD), vessel number (VN), vascular area, and vascular perimeter, were measured. CTP and histologic parameters were correlated.

Results: The Matrigel plugs with the 1,000-ng VEGF group exhibited a higher MVD than the 250-ng VEGF and control groups (P < .05). VN differed significantly between the control versus the 250-ng VEGF groups and 250-ng versus 1,000-ng VEGF groups (P < .05), with the highest VN in the 250-ng VEGF group. BF, mean transit time, and permeability-surface area product each differed significantly to VEGF levels. Changes in BF and BV did not correspond with increases in MVD or VN; however, in the 250-ng VEGF group, there was a strong positive correlation (r = 0.9) between BV and VN, vascular area, and vascular perimeter, which was not seen in the control or 1,000-ng VEGF group. All fCT parameters significantly correlated with each other (P < .05), with strong correlations between BF and mean transit time (r = -0.7) and between BF and permeability-surface area product (r = 0.7) and a weak correlation between BF and BV (r = 0.3).

Conclusion: These results validate the VEGF-induced endothelial cell in a rat Matrigel model. In addition, histologic vascular parameter MVD does not correlate with fCT parameters measured by CTP software.