The effects of equilibrium intercompartmental water exchange kinetics on MRI estimation of tissue concentration of contrast agents

Magnetic resonance imaging has become the appropriate imaging modality to characterize and evaluate a therapeutic response in pathologies such as cerebral tumors.;This work addresses a new approach to evaluate the effect of intercompartmental water exchange kinetics across the transendothelial walls and cell membranes on the tissue water relaxation rate R1 utilizing the Bloch-McConnell formalism to study a Look-Locker pulse sequence. A multiexponential signal from a three-site two exchange model generated with reasonable values (i.e., relaxation rates, exchange rates, compartmental size of each compartment, and pulse sequence acquisition parameters) was fitted by a monoexponential relaxation rate. The relationship between Look-Locker estimates of R 1 versus the relaxation rate of the extracellular space R1e (i.e., extracellular contrast agent concentration) was nearly linear across the experimental range and, further, was modeled to be an approximately linear across a wider range of R1e. The modeling result was confirmed in rats implanted with 9L cerebral tumor by comparing ΔR1 at 25 minutes after administration of gadolinium-labeled albumin to the autoradiographically estimated concentration 30 minutes after administration of this contrast agent's radiotracer analog, radioiodinated serum albumin in the same animal. A generalized estimating equation analysis of 15 clusters of data of AR1 maps of gadolinium-labeled albumin to the radioiodinated serum albumin maps 9L cerebral tumors animals was well correlated (r = 0.727, p < 0.0001).;The second work compares magnetic resonance imaging estimates of forward transfer rate K1 for albumin to the estimates K1 determined by quantitative autoradiographically for radioiodinated serum albumin. A Look-Locker estimate of ΔR1 was used as an index of the tissue concentration of the Gd-albumin. K1 was estimated from quantitative autoradiographic data using regions of interest (ROI) closely corresponding to those used to estimate of magnetic resonance imaging K1. The two estimates of K1 were well correlated (r = 0.55, p < 0.040) and generalized estimating equation analysis used for 15 cluster of longitudinal data showed correlation (r = 0.53, p < 0.0001). I conclude that an in vivo K1 derived from magnetic resonance imaging is well correlated with the gold standard for quantifying vascular permeability, quantitative autoradiographic.