Introduction: Diffusion tensor magnetic resonance (MR) imaging (DTI) can be used to characterize the microstructures of ordered biological tissues. This study was designed to assess histological features of gliomas and surrounding brain tissues in rats using DTI.
Methods: Three types of tumors, a 9L gliosarcoma (n = 8), a F98 glioma (n = 5), and a human glioblastoma xenograft (GBM22; n = 8) were incubated in rat brains and underwent conventional MRI and DTI scanning using a 4.7-T animal MRI system. Fractional anisotropy (FA), isotropic apparent diffusion coefficient, parallel diffusivity (λ//), and perpendicular diffusivity (λ⊥), as well as histological features within several regions of interest were analyzed.
Results: All tumor masses consisted of low-FA central zones (tumor center) and high-FA peripheral regions (tumor rim). Histological examination revealed the existence of highly coherent tumor organizations (circular for 9L and F98 or radial for GBM22) in the tumor rims. There were higher apparent diffusion coefficient, λ⊥, and λ// in the peritumoral edema compared to the contralateral gray matter. There were significantly lower FA and higher λ⊥ in the ipsilateral white matter than in the contralateral white matter for the GBM22 tumor, whereas there were no differences for the 9L and F98 tumors. Histologic examination showed GBM22 tumor infiltration into the ipsilateral damaged white matter.
Conclusions: Quantitative analysis of DTI indices provides useful information for assessing tumor microstructure and tumor cell invasion into the adjacent gray matter and white matter.
From the *Division of MR Research, Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD; and †FM Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD.
Received for publication June 7, 2012; accepted July 5, 2012.
Reprints: Silun Wang, MD, PhD, MPhill, Department of Radiology, Johns Hopkins University School of Medicine, 600 N Wolfe St, 344 Park Bldg, Baltimore, MD 21205 (e-mail: firstname.lastname@example.org).
This work was supported in part by grants from NIH/NIBIB (EB009112, EB009731, and EB015555).
The authors have no conflicts of interest to declare.