The Clinical Study On Brain Tumors And Perienhancement Area With Multivoxel ～1H-MRS And DWI

Objective To investigate the predictive value of multivoxel proton magnetic resonance spectroscopy (1H-MRS) and diffusion-weighted imaging (DWI) in discussing the metabolic changes and the apparent diffusion coefficient characteristic in tumor itself and perienhancement area of brain tumor. To evaluate the role of MRS and DWI on diagnosis and differentiation of brain tumors, delimiting the tumor border and deciding gliomas degrees.Methods Twenty normal persons underwent DWI. Thirty-five intracranial tumor patients were recruited in this study with 16 astrocytomas [gradeⅠ~Ⅱ(9),Ⅲ~Ⅳ(7)], 12 metastases and 7 meningiomas. All the patients had underwent MR routing scanning and DWI,1H-MRS. DWI was performed on a 1.5T MR scanner using echo planer imaging-spin echo sequence with TE of 101ms and TR of 6265ms. The ADC values were calculated within various brain tumors, various areas which were parenchyma of the tumors, perienhancement regions, 10mm to the tumors and corresponding contra lateral normal appearing white matter. The difference of the mean ADCs were made by analysis of variance. The relative ADCs were obtained from the ADCs of ROI contrasting to the ones of the contra lateral normals. Multivoxel proton MRS was performed orderly using Point-resolved spectroscopy (PRESS) sequence with TE of 135ms and TR of 1500ms. Spectra of three vowels were taken from parenchyma of the tumors, perienhancement region (with T2 hyperintensement areas) and corresponding contra lateral normal appearing white matter, and these regions were evaluated in every patient. Fitted areas in the spectrum for N-acetylaspartate (NAA), choline (Cho), creatine (Cr), lipid (lip) /lactate (lac) and myoInositol (mI) metabolite peaks were measured and NAA/Cho, NAA/Cr, Cho/Cho (normal), Cho/nCr, Cho/Cr ratios were calculated for each voxel (56.2 mm3 in size). One way, ANOVA (SPSS 11. 5 for windows, Chicago, Ill.) was used for statistical analysis in metabolic ratio's difference and the ADC values variety among the brain tumors.Results In DWI from different areas of normal brain, the mean ADCs were different and it was the lowest in cerebellum. The mean ADCs from parenchyma of gliomas, meningiomas and metastases were (1.26±0.19)×10-3mm2/s,(0.80±0.15)×10-3mm2/s and (1.21±0.23)×10-3mm2/s, respectively, and significant difference were obtained between meningiomas and gliomas or metastases. The mean ADC values of parenchyma of the tumors for high and low grade gliomas were (1.37±0.17)×10-3mm2/s and (1.12±0.21)×10-3mm2/s, there was significant difference between them. The values of perienhancement for glial tumors, metastases and meningiomas were (1.37±0.23)×10-3mm2/s, (1.59±0.17)×10-3mm2/s and (1.65±0.18)×10-3mm2/s, and they were statistically significant. In the areas 10mm to the tumors, there was no significant among them. The changes of rADC values were similar to the ADC values. In multivoxel of 1H-MRS from the parenchyma of the lesion, NAA were lower and Cho were higher than the contra lateral normal brain, the ratios of NAA/Cho, NAA/Cr and Cho/Cr changed significantly by comparing with that of normal control brain tissues (P<0.05), but there was no significant difference among gliomas and metastases (P>0.05). Lac occurred in 17 cases and lip presented to 7 cases. The lip peak was insolated and towering in 2 gradeⅣgliomas and 3 metastases. The lac peak appeared as inverse double peaks in 4 high grade gliomas and 5 metastases. 5 meningiomas and 1 glioma showed Ala peak while 4 malignant gliomas and 2 metastases with necrosis had high mI peak. On the contrary, in voxel B of MRS fromperienhancement region, The mean values of perienhancement for glial tumors, metastases and meningiomas were 0.83±0.21, 1.18±0.20 and 1.27±0.27 for NAA/Cho; 1.04±0.11, 1.45±0.22 and 1.42±0.47 for NAA/Cr; 0.94±0.14, 1.24±0.20 and 1.24±0.40 for NAA/nCr; 1.28±0.38, 0.87±0.14 and 0.83±0.22 for Cho/NAA; and 1.16±0.20, 1.08±0.20 and 1.00±0.25 for Cho/nCr, respectively. NAA/Cho, NAA/Cr, NAA/nCr and Cho/NAA ratios revealed strong correlations between metabolite concentrations and tumor types, allowing the differentiation of glial tumors from both metastases and meningiomas (P<0.05). With the malignant degree increasing, NAA decreasing while Cho increasing and abnormal Lac, lip appeared in gliomas. And the metabolic ratio (NAA/Cho, NAA/nCr, Cho/Cr and Cho/nCho) between malignant and benign gliomas was significantly difference.Conclusion Evaluation of brain tumors and specifically perienhancement region with multivoxel proton MRS may be useful in distinguishing glial tumors from metastases and meningiomas. It is useful in determining the tumor border with ADC values. These two technologies with enhanced MRI will improve the diagnostic yield of traditional MRI evaluation in patients with brain tumors; they are useful in the diagnosis and differentiation of tumors, deciding the border of tumors, histological classification of gliomas and mapping out the optimal surgery scheme.