Comparative Study Between DSC-MRI And DCE-MRI In Predicting Molecular Classification Of Gliomas

Background: Over the past 20 years,the genetic basis of cancer has been gradually elucidated.There is sufficient evidence to show that tumors with the same or similar tissue characteristics can have different molecular genetic backgrounds,and different histological tumors can have the same genetic background.According to the histological classification,the prognosis varies greatly among individuals,but molecular pathological classification based on the level of tumor genetics can more accurately judge clinical prognosis and guide clinical treatment.The World Health Organization(WHO)revised 2016 central nervous system tumor classification broke the diagnostic principle based entirely on the diagnosis under the microscope.The new classification adds molecular changes to the CNS tumor classification,which opens a new concept of morphology + molecular pathological diagnosis mode and integrated diagnosis.However,due to the genotype and protein expression pattern of solid tumors,there is heterogeneity in time and space distribution.Therefore,invasive biopsy cannot fully reflect the biological characteristics of the tumor,while imaging examination can comprehensively,dynamically,and non-invasively reflect the overall information of the tumor.Magnetic resonance perfusion imaging is a functional imaging method that uses magnetic resonance rapid imaging sequences and image post-processing techniques to reflect the degree of tissue vascularization and blood perfusion status,and to provide information on tissue and organ hemodynamics.Purpose: To evaluate the performance of dynamic susceptibility contrast(DSC)-and dynamic contrast-enhanced(DCE)-magnetic resonance imaging(MRI)histogram analyses in discriminating the states of molecular biomarkers and survival in glioma patients.Materials and Methods: Forty-three glioma patients who underwent DCE-and DSC-MRI were enrolled.Relevant molecular test results,including those on isocitrate dehydrogenase(IDH),O6-methylguanine-DNA methyltransferase(MGMT)and telomere reverse transcriptase(TERT),were collected.The mean relative cerebral blood volume of DSC-MRI and histogram parameters derived from DCE-MRI(volume transfer coefficient(Ktrans),fractional volume of the extravascular extracellular space(Ve),fractional blood plasma volume(Vp),rate constant between the extravascular extracellular space and blood plasma(Kep)and area under the curve(AUC))were calculated.Differences in each parameter between gliomas with different expression states(IDH,MGMT,and TERT)were evaluated.The diagnostic efficiency of each parameter was analyzed.The overall survival of all patients was assessed.Results: The 10 th percentile AUC(AUC = 0.830,sensitivity = 0.78,specificity = 0.80),the 90 th percentile Ve(AUC = 0.816,sensitivity = 0.84,specificity = 0.79),and the mean Kep(AUC= 0.818,sensitivity = 0.76,specificity = 0.78)provided the highest differential efficiency for IDH,MGMT,and TERT,respectively.KaplanMeier curves showed a significant difference between subjects with a 10 th percentile AUC higher or lower than 0.028(log-rank = 7.535;p = 0.006)for IDH and between subjects with different 90 th percentile Ve values(log-rank = 6.532;p=0.011)for MGMT.Conclusion: Histogram DCE-MRI demonstrates good diagnostic performance in identifying different molecular types and for the prognostic assessment of glioma.