The Application Study Of Diffusional Kurtosis Imaging In Grading And Evaluating The Microstructure Of Gliomas

Objective:To evaluate the application value of diffusional kurtosis imaging(DKI)in grading gliomas and relation between DKI metrics and the histological parameters;and then to determine whether DKI can assess the microstructural changes of gliomas.Methods:Thirty-seven patients(14 women and 23 men;age range,23-66 years;mean age,47 years)with cerebral gliomas underwent conventional MRI and DKI at 3.0Tesla MR imager.The parametric maps of fractional anisotropy(FA),mean diffusivity(MD)and mean kurtosis(MK)were obtained using the post-processing software.The regions of interest(ROIs)were manually drawn around the tumor center,tumor periphery and contralateral nomal-appearing white matter(NAWMc)by an experienced neuroradiologist.The parameter values for the tumor center and periphery were normalized to the corresponding values in the NAWMc of each patient.Normalized parameters(nFA,nMD and nMK)were compared between low grade gliomas(LGGs)and high grade gliomas(HGGs)using the independent-samples t test.Receiver operating characteristic(ROC)curve analyses were applied to the normalized values for the discrimination of LGGs and HGGs.For comparisons normalized metrics between the tumor center and periphery,a two-sided paired Student’s t test was used for statistical analysis.During surgery,the neurosurgeon obtained tissue from the tumor center and peritumoural region determined preoperatively via conventional MR images.Immunohistochemical staining with Ki-67,CD34,and microtubule-associated protein 2(MAP2)antibodies was performed on tumor center and periphery samples.The Pearson correlation analyses were performed to evaluate correlations between normalized DKI metrics(nFA.nMD,and nMK)and the corresponding histological parameters(tumor cell density,total vascular area[TVA],and Ki-67 labeling index[LI]).Results:1.Relative to the LGGs,the nFA and nMK values in the tumor center were significantly higher in the HGGs(P=0.02 and P<0.001,respectively);in contrast,the nMD was significantly lower in the HGGs(P = 0.001).2.The nMK exhibited the maximal area under the curve(AUC)value(0.889),sensitivity(80%)and specificity(100%)for differentiating between LGGs and HGGs,and the difference was significant(all P<0.05).3.Regarding the nFA and nMD values,a significant difference was found between the tumor center and periphery among all of the patients(P<0.001 and P = 0.006,respectively).No significant difference,however,was found for nMK(P = 0.28)values,which were significantly increased in the tumor periphery for the LGGs(P<0.001)and significantly decreased for the HGGs(P = 0.002)compared with the values in the tumor center.4.For the samples obtained from the tumor center,the nMK values were found to be correlated with tumor cell density(r = 0.596,P= 0.006),TVA(r = 0.764,P<0.001),and Ki-67 LI(r = 0.766,P<0.001).The nFA were also found to be correlated with tumor cell density(r = 0.534,P = 0.01).A significant negative correlation was found between the nMD and tumor cell density(r =-0.467.,P=0.03)and nMD and Ki-67 LI(r=-0.630,P = 0.003).Conclusion:1.DKI,especially concerning the MK parameter,demonstrated significant difference between LGGs and HGGs.MK showed high sensitivity and specificity in pathological grade of gliomas,comparing with conventional parameters(FA and MD).2.DKI,especially the MK parameter was closely related to the histological features of gliomas,such as tumor cell proliferation,vascularity and tumor cell density,and can act as a noninvasive technology to evaluate the microstructure change of gliomas.