Efficacy of three‐dimensional arterial spin labeling and how it compares against that of contrast enhanced magnetic resonance imaging in preoperative grading of brain gliomas

Abstract Purpose To evaluate the efficacy of three‐dimensional arterial spin labeling (3D‐ASL) imaging in preoperative grading of brain gliomas, and compare the discrepancy between images obtained from 3D‐ASL and contrast enhanced magnetic resonance imaging (CE‐MRI) in grading of gliomas. Methods Fifty‐one patients with brain gliomas received plain MRI, CE‐MRI and 3D‐ASL scanning before surgery. In 3D‐ASL images, the maximum tumor blood flow (TBF) of tumor parenchyma was measured, relative TBF‐M and rTBF‐WM were calculated. The cases were categorized into “ASL dominant” and “CE dominant” to compare the discrepancy between 3D‐ASL and CE‐MRI results. Independent samples t test, Mann‐Whitney and U test and one‐way analysis of variance (ANOVA) were performed to test the differences of TBF, rTBF‐M and rTBF‐WM values among brain gliomas with different grades. Spearman rank correlation analysis was performed to assess the correlation between TBF, rTBF‐M, rTBF‐WM and glioma grades respectively. To compare the discrepancy between 3D‐ASL and CE‐MRI results. Results In high‐grade gliomas (HGG) group, TBF, rTBF‐M and rTBF‐WM values were higher than those in low‐grade gliomas (LGG) group ( p < .05). Multiple comparison showed TBF and rTBF‐WM values were different between grade I and IV gliomas, grade II and IV gliomas (both p < .05), the rTBF‐M value was different between grade I and IV gliomas ( p < .05). The values of all 3D‐ASL derived parameters were positively correlated with gliomas grading (all p < .001). TBF showed highest specificity (89.3%) and rTBF‐WM showed highest sensitivity (96.4%) when discriminating LGG and HGG using ROC curve. There were 29 CE dominant cases (23 cases were HGG), 9 ASL dominant cases (4 cases were HGG). Conclusion 3D‐ASL is of significance to preoperative grading of brain gliomas and might be more sensitive than CE‐MRI in detection of tumor perfusion.