Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and computed tomography (DCE-CT) provide independent measures of biomarkers related to tumor perfusion. We compared the reproducibilities and absolute values of DCE-MRI and DCE-CT biomarkers in the same tumors in an animal model, to investigate the physiologic validity of both approaches. DCE-MRI and DCE-CT were each performed sequentially on three consecutive days in each of twelve rats bearing C6 glioma xenografts. DCE-MRI yielded endothelial transfer constant (Ktrans), extracellular, extravascular space volume fraction (ve), and contrast agent reflux rate constant (kep); and DCE-CT, blood flow (BF), blood volume (BV), mean transit time (MTT), and permeability-surface area product (PS) using Tofts and deconvolution physiological models, with 6.6 and 0.4 seconds temporal resolutions, respectively. Variability in DCE-CT and DCE-MRI were evaluated by variance components analysis. Intra-rat coefficients of variation for DCE-CT parameters BF, BV, MTT and PS were 25%, 22%, 18% and 23%; and for DCE-MRI parameters Ktrans, kep and ve were 23%, 16% and 20%, respectively. Mean (±SD) BF, BV, MTT and PS were: 44.6 (±13.7) ml min−1 100 g−1, 5.7 (±1.5) ml 100 g−1, 10.8 (±2.3) seconds, and 14.6 (±4.7) ml min−1 100 g−1, respectively. Mean (±SD) Ktrans, kep and ve were: 0.21 (±0.05) min−1, 0.68 (±0.14) min−1, and 0.29 (±0.06), respectively. Permeability estimates from DCE-MRI (Ktrans) were 44% higher than from DCE-CT (PS), despite application of appropriate corrections. DCE-MRI and DCE-CT biomarkers of tumor perfusion have similar reproducibilities suggesting that they may have comparable utility, but their derived parameter values are not equivalent.
