Quantitative Evaluation of Diffusion and Dynamic Contrast-Enhanced Magnetic Resonance Imaging for Differentiation Between Primary Central Nervous System Lymphoma and Glioblastoma.

Abstract

This study aimed to evaluate the utility of diffusion and permeability parameters derived from diffusion-weighted imaging and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) for differentiating primary central nervous system lymphoma (PCNSL) and glioblastoma multiforme (GBM) and to assess the correlation among these parameters. Forty-two patients with GBM and 18 patients with PCNSL underwent conventional 3.0-T MRI, diffusion-weighted imaging, and DCE-MRI before surgery. Normalized apparent diffusion coefficient ratio (rADC) and DCE-MRI-derived parameters (the volume transfer constant [K], the flux rate constant, the volume fraction of extravascular extracellular space [Ve], and the fractional plasma volume) were measured within the entire enhancing tumor and compared between the 2 groups. The diagnostic ability of each parameter and their optimal combination for differentiating between PCNSL and GBM, and the correlation among these parameters, were statistically analyzed. The PCNSLs demonstrated significantly lower rADC (P = 0.000), higher K (P = 0.000), and higher Ve (P = 0.001) than GBMs. With the combination of rADC and K, the diagnostic ability for discriminating between PCNSL and GBM was significantly improved (area under the receiver operating characteristic curve [AUC] = 0.930) as compared with rADC (AUC = 0.858) and K (AUC = 0.852) alone (P < 0.001 for both). The rADC did not correlate with K or Ve derived from DCE-MRI. Apparent diffusion coefficient ratio, K, and Ve are useful parameters for differentiating between PCNSL and GBM. The combination of rADC and K helps to improve the diagnostic accuracy. The rADC may not show correlation with K or Ve.