Quantitative Analysis of Acute Benign and Malignant Vertebral Body Fractures Using Dynamic Contrast-Enhanced MRI

Abstract

The objective of our study was to evaluate quantitative dynamic contrast-enhanced MRI (DCE-MRI) based on tracer kinetic modeling of perfusion in the differentiation of benign from malignant vertebral fractures. Patients with 26 osteoporotic vertebral fractures (18 women, eight men; mean age, 69 years) and patients with 20 malignant vertebral fractures (nine women, 11 men; mean age, 63.4 years) underwent dynamic contrast-enhanced MRI. T1-weighted, STIR, and T2-weighted sequences were acquired at 1.5 T. Dynamic contrast-enhanced image sets were acquired with a 2D saturation-recovery spoiled gradient-echo sequence. Regions of interest in parameter maps of mean transit time (MTT) and plasma flow in the fractured vertebral bodies were analyzed with a two-compartment tracer kinetic model. Plasma flow, plasma volume (PV), extraction flow, and interstitial volume were calculated. The forward volume transfer constant (K(trans)) and the extracellular volume (ECV) were derived. A two-tailed Fisher exact test, Mann-Whitney U test, and receiver operating characteristic analysis were performed. Forty-four vertebral fractures in 44 patients could be evaluated. In spots of increased plasma flow, interstitial volume (p = 0.0003), ECV (p = 0.002), and extraction flow (p = 0.03) for osteoporotic and malignant vertebral fractures were significantly different. The mean interstitial volume was 28.62 mL/100 mL for osteoporotic fractures and 11.73 mL/100 mL for malignant fractures, and the area under the curve (AUC) was 0.819 for a cutoff of 11.72 mL/100 mL or less indicating malignancy (sensitivity, 63.2%; specificity, 96.0%). The mean ECV was 52.68 mL/100 mL for osteoporotic fractures and 36.71 mL/100 mL for malignant fractures, and the AUC was 0.802 for a cutoff of 35.83 mL/100 mL or less indicating malignancy (sensitivity, 63.2%; specificity, 92.0%). The mean extraction flow was 15.19 mL/100 mL/min for osteoporotic fractures and 23.67 mL/100 mL/min for malignant fractures, and the AUC was 0.693 for a cutoff of 6.52 mL/100 mL/min or less indicating malignancy (sensitivity, 57.9%; specificity, 92.0%). K(trans), plasma flow, and PV in the spots of increased plasma flow and all quantitative perfusion parameters in the regions of increased MTT did not show any significant differences between benign and malignant fractures. In spots of high plasma flow, which can be determined with a deconvolution analysis, the quantitative perfusion parameters of interstitial volume, ECV, and extraction flow are significantly different between acute osteoporotic and malignant vertebral fractures and can aid in the distinction between the two entities.