PT  - JOURNAL ARTICLE
AU  - S.C. Jung
AU  - J.a. Yeom
AU  - J.-H. Kim
AU  - I. Ryoo
AU  - S.C. Kim
AU  - H. Shin
AU  - A.L. Lee
AU  - T.J. Yun
AU  - C.-K. Park
AU  - C.-H. Sohn
AU  - S.-H. Park
AU  - S.H. Choi
TI  - Glioma: Application of Histogram Analysis of Pharmacokinetic Parameters from T1-Weighted Dynamic Contrast-Enhanced MR Imaging to Tumor Grading
AID  - 10.3174/ajnr.A3825
DP  - 2014 Jan 02
TA  - American Journal of Neuroradiology
4099  - http://www.ajnr.org/content/early/2014/01/02/ajnr.A3825.short
4100  - http://www.ajnr.org/content/early/2014/01/02/ajnr.A3825.full
AB  - BACKGROUND AND PURPOSE: The usefulness of pharmacokinetic parameters for glioma grading has been reported based on the perfusion data from parts of entire-tumor volumes. However, the perfusion values may not reflect the entire-tumor characteristics. Our aim was to investigate the feasibility of glioma grading by using histogram analyses of pharmacokinetic parameters including the volume transfer constant, extravascular extracellular space volume per unit volume of tissue, and blood plasma volume per unit volume of tissue from T1-weighted dynamic contrast-enhanced perfusion MR imaging. MATERIALS AND METHODS: Twenty-eight patients (14 men, 14 women; mean age, 49.75 years; age range, 25–72 years) with histopathologically confirmed gliomas (World Health Organization grade II, n = 7; grade III, n = 8; grade IV, n = 13) were examined before surgery or biopsy with conventional MR imaging and T1-weighted dynamic contrast-enhanced perfusion MR imaging at 3T. Volume transfer constant, extravascular extracellular space volume per unit volume of tissue, and blood plasma volume per unit volume of tissue were calculated from the entire-tumor volume. Histogram analyses from these parameters were correlated with glioma grades. The parameters with the best percentile from cumulative histograms were identified by analysis of the area under the curve of the receiver operating characteristic analysis and were compared by using multivariable stepwise logistic regression analysis for distinguishing high- from low-grade gliomas. RESULTS: All parametric values increased with increasing glioma grade. There were significant differences among the 3 grades in all parameters (P < .01). For the differentiation of high- and low-grade gliomas, the highest area under the curve values were found at the 98th percentile of the volume transfer constant (area under the curve, 0.912; cutoff value, 0.277), the 90th percentile of extravascular extracellular space volume per unit volume of tissue (area under the curve, 0.939; cutoff value, 19.70), and the 84th percentile of blood plasma volume per unit volume of tissue (area under the curve, 0.769; cutoff value, 11.71). The 98th percentile volume transfer constant value was the only variable that could be used to independently differentiate high- and low-grade gliomas in multivariable stepwise logistic regression analysis. CONCLUSIONS: Histogram analysis of pharmacokinetic parameters from whole-tumor volume data can be a useful method for glioma grading. The 98th percentile value of the volume transfer constant was the most significant measure. Abbreviations AUCarea under the curveDCEdynamic contrast-enhancedKtransvolume transfer constantROCreceiver operating characteristic analysisVeextravascular extracellular space volume per unit volume of tissueVpblood plasma volume per unit volume of tissue