ESTRO 2020 Abstract book

S1004 ESTRO 2020

average template, the four cerebral lobes (frontal, parietal, occipital, temporal) were automatically contoured and a classification between white matter (WM), gray matter (GM) and cerebrospinal fluid (CSF) was performed for each voxel, allowing for volume assessments (Fig. 1). We then quantified the growth trajectory of different structural volumes (whole brain, CSF, GM, WM; as well as frontal, parietal, occipital, temporal GM/WM) and compared it to published results based on the National Institutes of Health (NIH) MRI data of brain development in the pediatric population (NeuroImage 2013;82:393-402).

Results For four patients we observed a reduction in the total volume of brain with increasing age, while the opposite trend was seen in the NIH population (Fig. 2). The largest discrepancies between the two groups were seen for the GM volume, with the patients presenting overall smaller GM volumes at all ages. When focusing on the four cerebral lobes, the frontal and parietal GM/WM had the largest reduction in volume compared to the NIH population. For five patients (all having had shunt insertion), the volume of CSF was markedly higher than in the NIH population but decreased with time.

Conclusion Our voxel-based method quantifying longitudinal MRI changes can be applied to pediatric brain tumor patients treated with RT to identify changes in substructures volumes. We observed that pediatric brain tumor patients have overall a reduced volume of GM. In further studies, we will relate at a voxel-level the observed MRI changes to the delivered dose.