Quantitative Assessment of Invasion of High-Grade Gliomas Using Diffusion Tensor Magnetic Resonance Imaging.

Abstract

To determine heterogeneity of high-grade glioma (HGG) and its surrounding area and explore quantitative analysis of invasion of HGG using diffusion tensor imaging. This study included 14 patients with HGG and preoperative magnetic resonance imaging and diffusion tensor imaging examinations. Three regions of interest were placed. Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) values of these regions of interest were measured, and specimens from the 3 regions of interest were obtained under navigation guidance. Postoperative examinations of specimens were carried out. Correlations between ADC and FA values and tumor cell density were evaluated. Median survival was 36.7 months. As distance from the tumor increased, the number of tumor cells significantly decreased. Regarding levels of matrix metalloproteinase-9 and Ki-67, only the differences between tumor and distances of 1 cm and 2 cm away from the tumor were statistically significant. For analysis of the relationship between tumor cell density and ADC and FA values, the discriminant formulas were as follows: G1=-13.678+ 14984.791 (X)+ 14443.847 (Y) (tumor cell density ≥10%); G2=-11.649+ 14443.847 (X)+ 33.285 (Y) (tumor cell density <10%). We verified the heterogeneity of HGG and its surrounding area and found that patients with extensive resection may have longer survival. We also found a few formulas using FA and ADC values to predict tumor cell density.