Mathematical model for histogram analysis of dynamic contrast-enhanced MRI: A method to evaluate the drug treatment response in rheumatoid arthritis

Purpose: To evaluate the effectiveness of a mathematical model for histogram analysis of DCE-MRI in distinguishing responders from non-responders during RA drug treatment. Method: Twenty-three consecutive RA patients with clinically active inflammation prospectively underwent DCEMRI at baseline and after treatment. Manual segmentation of the enhanced synovium was performed on the last phase of DCE-MRI. The voxel-based contrast enhancement was calculated in each phase to obtain 75th percentile values. Kinetic curves made from the 75th percentile values were fitted to mathematical model as follows, delta S(t) = A(1-e- alpha t)e-beta t, where A is the upper limit of signal intensity (%), alpha (sec-1) is the rate of signal increase, and beta (sec-1) is the rate of signal decrease during washout. AUC30 was calculated by integration of 30 s. SER was calculated as the signal intensity at the initial time point (t = 60) relative to the delayed time point (t = 300). The volumes of enhanced synovium (sum of the number of voxels) were also calculated. Results: After treatment, alpha, A alpha, AUC30 and SER were significantly lower in the responder group than in the nonresponder group (p = 0.033, 0.024, 0.015, and 0.007). The p value of SER was lowest. A alpha, AUC30, and the volume of enhanced synovium had significantly larger changes from baseline to after treatment in the responder group than in the non-responder group (p = 0.045, 0.017, and 0.008). The volume of enhanced synovium had the lowest p value. Conclusions: SER after treatment and change in the volume of enhanced synovium might be effective for distinguishing responders from non-responders.