Evaluating disease response using volumetric analysis vs 2D cross sectional analysis
Mentor Name: Carl Koschmann
Our research project focuses on investigating perfusion MRI as a predictive tool for assessing the efficacy of ONC201 treatment in children and young adults diagnosed with diffuse midline glioma (DMG) harboring the H3K27M mutation. Despite the promising outcomes observed with ONC201, a significant proportion of patients (60%) do not exhibit clinical or radiographic responses. This project aims to address this gap by exploring baseline normalized relative cerebral blood volume (nrCBV) as a potential predictive biomarker for treatment responsiveness. The comprehensive study involves a multimodal analysis, incorporating various MRI sequences and a detailed longitudinal assessment of approximately 60 patients treated with ONC201 at the University of Michigan. Our approach encompasses routine longitudinal, multi-sequence MRI imaging, which will be subjected to consensus review by two expert neuroradiologists utilizing the Response Assessment in Pediatric Neuro-Oncology (RAPNO) grading. Volumetric segmentation will be performed using semi-automated techniques based on thresholding, providing a nuanced understanding of the tumor's spatio-temporal profile during treatment. Parametric maps generated during volumetric segmentations will be analyzed alongside post-processed imaging, employing Cox regression models to assess correlations with progression-free survival (PFS) and overall survival (OS). The anticipated outcomes include the establishment of a robust clinical radiographic dataset and an assessment of the predictive value of nrCBV at baseline and its alteration during ONC201 treatment. Even negative results from this specific analysis in the ONC201 treated population will contribute meaningfully to our understanding of treatment response. This research project represents a significant step towards personalized treatment strategies for H3K27M mutant-DMG patients, aiming to improve outcomes and enhance the quality of life for individuals facing this challenging, fatal diagnosis. In this retrospective analysis we plan to harness routinely performed longitudinal, multi-sequence, T2 weighted,T1 weighted, without and with intravenous contrast, fluid attenuated inversion recovery (FLAIR), Diffusion weighted imaging (DWI) and Perfusion weighted imaging (PWI) MR imaging from approximately 60 patients treated with ONC201 at the University of Michigan. Imaging data will be reviewed by two expert neuroradiologists completing consenus RAPNO grading, qualitative assessment of perfusion and diffusion Imaging.
