Using 64-slice CT Perfusion Imaging Permeability Surface to Evaluate the Histopathologic Grade of Intracranial Gliomas

ZHANG Jia-wen,FENG Xiao-yuan,LIU Bin
2008-01-01
Abstract:Purpose: To detect the role and value of permeability surface(PS) using 64-slice CT perfusion imaging in preoperative grading gliomas. Materials and Methods:Thirty one glioma patients (11 low grade gliomas and 20 high grade gliomas) underwent 64-slice CT cerebral perfusion examination before operation. Low grade and high grade groups were categorized corresponding to WHO grade Ⅰ or II gliomas and WHO grade III or IV gliomas, respectively, as determined by histopathological examination. The patients’ raw data was transferred to workstation and the permeability surface (PS) maps were analyzed. The absolute value of PS was obtained from regions of maximal abnormality in tumor parenchyma on PS color perfusion maps.The receiver operating characteristic (ROC) curve was used to assess their values on distinguishing the low grade and high grade gliomas.Results:PS value of low grade gliomas and high grade gliomas was (4.16±0.76)ml/(100g·min), (11.68±6.09)ml/(100g·min),respectively. PS provided significant P-value in differentiating glioma grade( P< 0.001) using the Mann-Whitney U test. PS value of gliomas had strong correlation with the grade of glioma, and the r value was 0.701(P<0.001). The area under Receiver Operating Characteristic( ROC) curve was 0.923 for PS. ROC curves revealed better specificity and sensitivity in PS for glioma grading. With a PS cutoff value of 5.13 ml/(100g·min),sensitivity was 90% and specificity was 90.9% in differentiating high- from low-grade gliomas. One low-grade gliomas was incorrectly categorized which did not enhance. Two high-grade gliomas were incorrectly classified by the PS cutoff value. All of 4 oligodendrogliomas(one high-grade, one mixed high-grade and two low-grade tumors ) were correctly categorized. Conclusion:PS value using 64-slice CT perfusion imaging provides useful information for the grade of glioma and might have impact on clinical management of gliomas.
What problem does this paper attempt to address?