Abstract BACKGROUND Diffuse midline glioma (DMG) harboring the H3K27M mutation is an exceedingly aggressive and treatment-resistant pediatric brain tumor. Monitoring patterns of tumor cell infiltration, understanding the immune microenvironment and the identification of novel imaging biomarkers are essential for developing more effective therapies. This study combines advanced MRI with histopathological analysis, novel imaging techniques and immuno-profiling to characterize DMG in both preclinical models and selected DMG patients. METHODS Using a preclinical NSG/NXG model, we implanted DIPG-17 and DIPG-007 cells into the pons, cerebellum, thalamus or cortex to evaluate tumor proliferation, immune infiltration, and therapeutic responses. Imaging modalities comprised advanced MRI sequences, including chemical exchange saturation transfer (APTw-CEST) and spin-echo diffusion tensor imaging (SE-DTI), in conjunction with complementary histological and light sheet microscopy assessments. Selected H3K27M patients were monitored concomitantly by multiparametric and CEST MRI. RESULTS MRI data indicate unique imaging signatures associated with different primary tumor locations. We found strong infiltration along white matter tracts and distinct immune profiles. SE-DTI tractography was utilized to reconstruct white matter tracts to further identify DMG infiltration pathways. Myeloid cell distribution and polarization in the tumor microenvironment was assessed by IHC to further understand their distribution in the tumor core and adjacent infiltrative zone. Additionally, we assessed the biodistribution of fluorescently labeled cyclodextrin (CDNP)-nanoparticle therapy upon intravenous administration in the tumor microenvironment by light sheet microscopy. DISCUSSION This study advances the characterization of DMG by combining advanced imaging techniques with immune profiling, yielding insights into tumor evolution and treatment responses. Our findings could guide individualized imaging-based treatment approaches for DMG patients and improve non-invasive evaluation of therapeutic response in clinical trials.
Sinn et al. (2025) studied this question.