Diffusion Kurtosis Imaging For Noninvasively Predicting Molecular Subtypes And Survival In Gliomas

Objective:To investigate the application value of diffusion kurtosis imaging(DKI)in predicting the molecular classification and prognosis in WHO grade 2 and 3gliomas,which consists of isocitrate dehydrogenase(IDH)and telomerase reverse transcriptase(TERT)gene status.Methods:The clinical information and imaging features of 33 patients with WHO 2 or 3gliomas confirmed by histopathology in the First Hospital of Shanxi Medical University were retrospectively analyzed and recorded in detail.All patients underwent gene sequencing to obtain complete IDH,TERT molecular classification results and had magnetic resonance imaging(MRI)to obtain complete preoperative DKI and conventional sequences(including conventional MRI scan,enhanced MRI).According to the gene sequencing results,all patients were divided into IDH-mutant type group(16cases)and IDH-wild type group(17 cases).The IDH-wild type group was divided into TERT-mutant type group(5 cases)and TERT-wild type group(12 cases).According to the overall survival(OS),the IDH-wild type group was divided into OS< 2-year group(8cases)and ≥ 2-year group(8 cases).The post-processing tool was used to process the DKI image,then the parameters of mean kurtosis(MK),axial kurtosis(Ka),radial kurtosis(Kr)and fractional anisotropy of kurtosis(FAK)pseudo color image were obtained.By drawing regions of interest(ROI)in the tumor parenchyma area and the contralateral normal white matter,doctors measured MK,Ka,Kr and FAK values.The ratio of tumor area to contralateral normal white matter was used as the corrected parameter values:relative mean kurtosis(rMK),relative axial kurtosis(rKa),relative radial kurtosis(rKr)and relative fractional anisotropy of kurtosis(rFAK).To make the comparisons of the clinical data and imaging parameters in different IDH,TERT and prognosis groups,chi-square test,two independent sample t test and rank sum test were adopted.Selecting statistically different parameters,the receiver operating characteristic curve(ROC)was drawn,then area under the curve(AUC),sensitivity,specificity and thresholds were calculated.Results:1.In all patients,the age of the IDH-wild type group was higher than IDH-mutant type group(P=0.003),and gender was no statistical difference between two groups(P>0.05).Among the parameters,the rMK,rKa and rKr values of IDH-wild type group were significantly higher than those of IDH-mutant type group(P=0.001,P=0.002,P=0.008);ROC analysis of the above parameters showed that AUC,sensitivity,specificity and threshold of rMK value were 0.820,75.00%,82.40% and 0.499,respectively.AUC,sensitivity,specificity and threshold of rKa value were 0.805,75.00%,98.20% and 0.660.The AUC,sensitivity,specificity and threshold of rKr value were0.768,81.20%,70.60% and 0.460,respectively.2.In IDH-wild type group,there was no significant difference in age and gender between TERT-mutant type group and TERT-wild type group(P>0.05).ROC analysis was performed on significant rMK,rKa and rKr values.The AUC,sensitivity,specificity and threshold of rMK value were 0.833,100%,66.70% and 0.643,respectively.The AUC,sensitivity,specificity and threshold of rKa value were 0.867,60.00%,100% and1.095,respectively.The AUC,sensitivity,specificity and threshold of rKr value were0.833,80.00 %,91.70 % and 0.681.3.In IDH-wild type group,there was no statistical difference in age and gender between the two groups(P>0.05).ROC analysis was performed on the significant rFAK value,and the AUC,sensitivity,specificity and threshold were 0.859,100%,62.50 % and0.486,respectively.Conclusion:In this study,we analyzed the value of DKI parameters in noninvasive prediction of IDH,TERT classification and prognosis of glioma patients.It is confirmed that rMK,rKa and rKr values can predict IDH status of gliomas to varying degrees.Among them,rMK value has higher diagnostic efficiency and rKa value has higher specificity.rMK,rKa and rKr values can predict the TERT status of gliomas,and rMK is highly sensitive.The rFAK value has reference value for the prognosis of glioma patients.Therefore,DKI helps to stratify glioma patients to develop personalized treatment and evaluate prognosis.