Benign and Malignant Musculoskeletal Tumors Evaluated with the Aid of Pa rametric Maximum Slope and Blood Volume Images

Abstract

Purpose: To assess the values of parametric maximum slope(MS) and blood volume(BV) images as indicators of tissue vascularization and perfusion in distinguishing benign from malignant musculoskeletal tumors. Material and Methods: Dynamic inversion recovery spin-echo echo-planar imaging (TR/TE/TI/NEX: 1500/24/500 ms/1; matrix 128 128; slice thickeness 5 mm, interleaved) at 1.5-second intervals(200 phases) was performed after intravenous bolus injection of Gd-DTPA. A total of 32 pathologically proven muscu-loskeletal masses(benign 9, malignant 23) were included in this study. MS was derived by fitting a time-intensity curve using a polynomial model. BV was determined by integration until maximum slope was reached. On a pixel-by-pixel basis, MS and BV images were generated and displayed by gray-scale. We selected a region of interest(ROI, more than 6 6 pixels) including the highest value of the tumors and calculated mean values of MS and BV. Wilcoxon’s signed rank test was used to compare benign tumors with malignant pre- and postchemotherapy tumors. Result: The mean values of ROIs selected on MS images were significantly different (p=.008) between benign (mean 3.33, range 0.01 -16.47 ) and prechemotherapy malignant(mean 8.54, range 1.61 -16.90) tumors, as were the mean values of ROIs on BV images(p=.005; benign tumors: mean 162.17, range 91.17 -283.7 ; prechemotherapy malignant tumors: mean 330.18, range 117.5 -845.1). MS and BV values of benign and malignant tumors overlapped but tended to be separated. Conclusion: BV and MS images may help distinguish benign from malignant musculoskeletal tumors.