Correlations of 3T DCE-MRI Quantitative Parameters with Microvessel Density in a Human-Colorectal-Cancer Xenograft Mouse Model

Objective: To investigate the correlation between quantitative dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) parameters and microvascular density (MVD) in a human-colon-cancer xenograft mouse model using 3 Tesla MRI. Materials and Methods: A human-colon-cancer xenograft model was produced by subcutaneously inoculating $1{\times}10^{6}$ DLD-1 human-colon-cancer cells into the right hind limbs of 10 mice. The tumors were allowed to grow for two weeks and then assessed using MRI. DCE-MRI was performed by tail vein injection of 0.3 mmol/kg of gadolinium. A region of interest (ROI) was drawn at the midpoints along the z-axes of the tumors, and a Tofts model analysis was performed. The quantitative parameters ($K^{trans}$, $K_{ep}$ and $V_{e}$) from the whole transverse ROI and the hotspot ROI of the tumor were calculated. Immunohistochemical microvessel staining was performed and analyzed according to Weidner's criteria at the corresponding MRI sections. Additional Hematoxylin and Eosin staining was performed to evaluate tumor necrosis. The Mann-Whitney test and Spearman's rho correlation analysis were performed to prove the existence of a correlation between the quantitative parameters, necrosis, and MVD. Results: Whole transverse ROI of the tumor showed no significant relationship between the MVD values and quantitative DCE-MRI parameters. In the hotspot ROI, there was a difference in MVD between low and high group of $K^{trans}$ and $K_{ep}$ that had marginally statistical significance (ps = 0.06 and 0.07, respectively). Also, $K^{trans}$ and $K_{ep}$ were found to have an inverse relationship with MVD (r = -0.61, p = 0.06 in $K^{trans}$; r = -0.60, p = 0.07 in $K_{ep}$). Conclusion: Quantitative analysis of T1-weighted DCE-MRI using hotspot ROI may provide a better histologic match than whole transverse section ROI. Within the hotspots, $K^{trans}$ and $K_{ep}$ tend to have a reverse correlation with MVD in this colon cancer mouse model.