Multiparametric characterization of grade 2 glioma subtypes using magnetic resonance spectroscopic, perfusion, and diffusion imaging

Abstract

Background and purpose: The purpose of this study was to derive quantitative parameters from magnetic resonance (MR) spectroscopic, perfusion, and diffusion imaging of grade 2 gliomas according to the World Health Organization and to investigate how these multiple imaging modalities can contribute to evaluating their histologic subtypes and spatial characteristics.

Materials and methods: MR spectroscopic, perfusion, and diffusion images from 56 patients with newly diagnosed grade 2 glioma (24 oligodendrogliomas, 18 astrocytomas, and 14 oligoastrocytomas) were retrospectively studied. Metabolite intensities, relative cerebral blood volume (rCBV), and apparent diffusion coefficient (ADC) were statistically evaluated.

Results: The 75th percentile rCBV and median ADC were significantly different between oligodendrogliomas and astrocytomas (P < .0001) and between oligodendrogliomas and oligoastrocytomas (P < .001). Logistic regression analysis identified both 75th percentile rCBV and median ADC as significant variables in the differentiation of oligodendrogliomas from astrocytomas and oligoastrocytomas. Group differences in metabolite intensities were not significant, but there was a much larger variation in the volumes and maximum values of metabolic abnormalities for patients with oligodendroglioma compared with the other tumor subtypes.

Conclusions: Perfusion and diffusion imaging provide quantitative MR parameters that can help to differentiate grade 2 oligodendrogliomas from grade 2 astrocytomas and oligoastrocytomas. The large variations in the magnitude and spatial extent of the metabolic lesions between patients and the fact that their values are not correlated with the other imaging parameters indicate that MR spectroscopic imaging may provide complementary information that is helpful in targeting therapy, evaluating residual disease, and assessing response to therapy.

Figure 1

MR images and spectra from a patient with oligodendroglioma. (A) Axial T2-weighted FSE image with contours of T2ALL (black) and NAWM (white). (B) T1-weighted postcontrast image showing no contrast enhancement. (C) rCBV map showing a large focus of increased blood volume (arrows) in the tumor. (D) ADCmap showing mildly increased magnitude of diffusion (arrows) in the tumor. (E and F) T2-weighted FSE image overlaid with arrays of spectra showing voxels with elevated Cho and reduced NAA peaks in the tumor.

Figure 2

MR images and spectra from a patient with astrocytoma. (A) Axial T2-weighted FSE image. (B) T1-weighted postcontrast image showing no contrast enhancement. (C) rCBV map showing minimally increased blood volume (arrows) in the tumor. (D) ADC map showing a large focus of increased magnitude of diffusion (arrows) in the tumor. (E and F) T2-weighted FSE image overlaid with arrays of spectra showing voxels with elevated Cho and reduced NAA peaks in the tumor.

Figure 3

MR images and spectra from a patient with oligoastrocytoma. (A) Axial T2-weighted FSE image. (B) T1-weighted postcontrast image showing no contrast enhancement. (C) rCBV map showing mildly increased blood volume (arrows) in the tumor. (D) ADC map showing mildly increased magnitude of diffusion (arrows) in the tumor. (E and F) T2-weighted FSE image overlaid with arrays of spectra showing voxels with elevated Cho and reduced NAA peaks in the tumor.

Figure 4

T2-weighted FSE images (left) and spectra (right) from two patients who had oligodendrogliomas but with a different number of voxels with elevated CNI. Voxels highlighted in light gray have CNI > 2 and voxels highlighted in dark gray have CNI > 3. Patient A had 160 voxels with CNI > 2, 129 voxels with CNI > 3, and the max CNI = 16.8. Patient B had 10 voxels with CNI > 2, no voxels with CNI > 3, and the max CNI = 2.9.