Dipg-37. Predicting Outcome in Childhood Diffuse Midline Gliomas Using Magnetic Resonance Imaging Based Texture Analysis

Abstract

BACKGROUND Diffuse midline gliomas (DMG) are aggressive brain tumours with 10% overall survival (OS) at 18 months. Predicting OS will help refine treatment strategy in this patient group. MRI based texture analysis (MRTA) is a novel technique that provides objective information about spatial arrangement of MRI signal intensity and has potential as an imaging biomarker. OBJECTIVES: To investigate MRTA in predicting OS in childhood DMG. METHODS Retrospective study of patients diagnosed with DMG, based on radiological features, treated at our institution 2007–2017. MRIs were accomplished at diagnosis and 6 weeks after radiotherapy (54Gy in 30 fractions). MRTA, performed using TexRAD software, on T2W sequence and Apparent Diffusion Coefficient (ADC) maps encapsulated tumour in the largest single axial plane. MRTA comprised filtration-histogram technique using statistical and histogram metrics for quantification of texture. Kaplan-Meier analysis determined association of MRI texture parameters with OS. RESULTS 32 children 2–14 years (median 7 years) were included. MRTA was undertaken on T2W (n=32) and ADC (n=22). MRTA on T2W was better at prognosticating than on ADC maps. Children with homogenous tumour texture, at medium scale on baseline T2W MRI, had worse prognosis (mean p=0.0098, SD p=0.0115, entropy p=0.0422, mean of positive pixels (MPP) p=0.0051, kurtosis p=0.0374). MPP was the most significant texture parameter. Median survival in this group as identified by MRTA (medium texture, MPP) was 7.5 months versus 17.5 months. CONCLUSIONS DMG with more homogeneous texture on diagnostic MRI is associated with worse prognosis. MPP texture parameter is the most predictive of OS in childhood DMG.