MR Imaging Studies On Glioma Microvascular Characteristics And Differential Diagnosis In Low Grade Gliomas

Part One：T1weighted dynamic contrast-enhanced MR imaging(DCE-MRI)inassessing microvessel characteristics and glioma gradingObjectiveTo assess the ability of Ktrans(volume transfer constant), Ve(volume fraction ofextravascular extracellular space) derived from dynamic contrast enhanced magneticresonance imaging(DCE-MRI) in evaluating microvessel characteristics and glioma grading.Materials and MethodsTwenty-six patients with different types of gliomas (12grade II,8grade III and6gradeIV) were enrolled in this study. All patients were preoperatively underwent routine MRimaging and DCE MRI imaging, the quantitative parameter values like Ktrans and Ve wereobtained from software. Analysis of variance(ANOVA) were used to compare the maximalvalues of Ktrans、Ve and mean MVD (microvessel density) and VD (vessel diameter) amongthose gliomas. Pearson correlation analyses were used to determine relationship betweenKtrans, Ve, and glioma grades, and it also determine between Ktrans, Ve, MVD and VD.Receiver operating characteristic (ROC) curve analyses were performed to determine thethreshold for Ktrans and Ve to distinguish different grades of gliomas.ResultsThe mean values of Ktrans and Ve in LGG (0.026/min-1and0.121, respectively) wereboth significentlt lower than that of HGG (0.117/min-1and0.505, respectively)(both P <0.01).In addition, there were significant differences for Ktrans and Ve values between grade II andgrade III, and between grade II and grade IV. However, no statistical difference was foundbetween grade III and IV. There were significant differences for mean MVD and VD betweenLGG and HGG.The cutoff value of Ktrans=0.054/min-1for differentiation between LGG and HGGprovided the best combination of sensitivity (94.1%) and specificity (93.3%). The area underthe curve (AUC) of Ktrans was0.941. The cutoff value of Ve=0.296provided the bestcombination of sensitivity (92.9%) and specificity (91.7%).The AUC of Ve was0.937. Inaddition, the cutoff value of Ktrans=0.045/min-1for differentiation between grade II andgrade III provided the best combination of sensitivity (87.5%) and specificity (86.7%). The area under the curve (AUC) of Ktrans was0.833. The cutoff value of Ve=0.296provided thebest combination of sensitivity (92.9%) and specificity (91.7%).The AUC of Ve was0.937.The cutoff value of Ktrans=0.064/min-1for differentiation between grade II and gradeIII provided the best combination of sensitivity(100%) and specificity (93.3%). The areaunder the curve (AUC) of Ktrans was0.993. The cutoff value of Ve=0.345provided the bestcombination of sensitivity (88.9%) and specificity (93.3%). The AUC of Ve was0.925. Ktransand Ve values were both strongly correlated with glioma grades (r=0.682and0.761,respectively, both P<0.01), Ktrans values were strongly correlated with Ve(r=0.696, P<0.01).Ktrans values were moderately correlated with MVD(r=0.440, P<0.05) and stronglycorrelated with VD(r=0.652, P<0.01). Ve values were strongly correlated with VD(r=0.625，p<0.01), However, Ve values were weakly correlated with MVD but without statisticalsignificance(r=0.625, P<0.01).ConclusionDCE-MR imaging can be used to distinguish not only LGG from HGG, but also grade II fromIII or IV via testing microvascular permeability, it provide a significant auxiliary method of imagingto clinical diagnosis of gliomas. There was good correlation between Ktrans values and MVDvalues, between Ktrans values and VD values, and between Ve values and VD values. Prat Two：dynamic susceptibility contrast (DSC) perfusion weighted imaging(PWI)and diffusion weighted imaging(DWI) in differentiation of grade II gliomasObjectiveTo assess the value of dynamic susceptibility contrast (DSC) MR imaging and MRdiffusion weighted imaging (DWI) in the differentiation of intracranial grade II gliomas(astrocytoma, oligodendroglioma and oligodendroastrocytoma).Materials and MethodsForty-six patients with pathologically proved low grade gliomas were retrospectivlyanalysed, those gliomas were divided into three groups:22astrocytomas (AC),16oligodendrogliomas (OD) and8oligodendroastrocytomas (OA). The mean maximum of relative Cerebral Blood Volume (rCBV) values and relative apparent diffusion coefficient(rADC) within solid part of tumor were measured resectively by special software, thenanalysis of variance (ANOVA) was performed to compare the mean rCBVmax and rADCamong3groups.Receiver operator characteristic (ROC) curve analyses were performed todetermine optimum of rCBVmax and rADC between AC and OD/OA, respectively. Theanalyses also permitted the calculation of the diagnostic accuracy, sensitivity and specificityassociated with optimal threshold. Pearson correlation analysis was used to show therelationships between the rCBVmax and rADC within those gliomas.ResultsThe mean rCBVmax values were1.43,3.34and2.77for AC, OD and OA, respectively.The mean rADC values were2.09,1.51and1.60for AC, OD and OA respectively. Therewere a significant difference in rCBVmax and rADC between AC and OD, between AC andOA, even between AC and OD/OA (all the P<0.01). However, no significant difference wasfound between OD and OA in both rADC and rCBVmax. The optimal threshold of rCBVmax=2.15for differentiation between AC and OD/OA provided the best diagnostic accurancy(88.7%) with combination of sensitivity (96.0%) and specificity (81.0%). The optimalthreshold of rADC=1.94for differentiation between AC and OD/OA provided the bestdiagnostic specificity (100%), However, the diagnostic accurancy (81.9%) and sensitivity(76.2%) was relatively lower than those of rCBVmax. Pearson correlation analysis found thata negative correlation between rADC and rCBVmax in AC (r=-0.602, P<0.01), but nocorrelation was found not only in OD or OA but also in OD/OA.ConclusionBoth dynamic susceptibility contrast (DSC) MR perfusion weighted imaging and MRdiffusion weighted imaging (DWI) may be helpful in differentiating OD and OA from AC, butthere were some difficulties between OD and OA. In differentiating AC from OD/OA, MRDWI has higher diagnostic Specificity, but the DSC-MR Perfusion has the higher diagnosticaccuracy and sensitivity. In addition, the correlation of the two parameters or not may relatedto the pathologic characters of those gliomas.