Dose-Dependent Changes In Volume Of Cerebral Cortex And Subcortical Grey Matter Structures After Radiotherapy: A Need For Reconsidering RT Planning Strategies

Abstract

Cognitive decline after radiotherapy (RT) remains one of the biggest challenges in the treatment of brain tumors. Structural changes in the post-RT brain have been examined in order to explain and possibly prevent this debilitating consequence of therapy. The focus has largely been on white matter and hippocampus, with the cortical and other subcortical grey matter (GM) receiving limited attention. We hypothesize that changes in the GM are also involved in the observed post-RT cognitive decline. As a first step, we measured MRI-derived volumetric changes in GM after RT, and related these to the applied dose. A cohort of 28 grade II-IV glioma patients with optimal imaging available was selected from our patient records. GM was automatically segmented via the Computational Anatomy Toolbox (CAT12) in T1-weighted MR images before RT and one year after RT. Both scans were registered to standard space and smoothed to increase SNR. The PTV was censored from analysis to prevent errors due to tissue misclassification. Regional changes in GM volume were determined with voxel-based morphometry (VBM), and were related to the locally applied dose by means of linear regression, corrected for age and sex. Results were further corrected for multiple testing. Local GM volume changes per Gy are presented per GM region of interest (ROI), according to the Neuromorphometrics brain atlas. Dose-dependent changes in cortical and subcortical GM volume were observed in several areas of the brain. In total, 42 (29.2% of 142) of the studied ROIs showed significant volume loss with increasing dose, both in the cortical and subcortical GM. The table shows the areas in which more than 25% of voxels showed volume decrease, along with the effect size. No regions showed increase in GM volume after RT. Both the cortical and subcortical GM are susceptible to dose-dependent radiation-induced damage. This result may have implications for the treatment of brain tumors, as current sparing strategies do not take these areas into account. Currently, only the hippocampus is advised to be avoided to reduce cognitive symptoms after treatment. In this research we have shown that with increased dose other GM structures, both cortical and subcortical, are diminished. Future research should relate these GM morphometric changes to cognitive symptoms after RT, which could conclude that RT strategies should indeed include sparing of certain GM areas.Abstract 3701; TableRegion nameAffected voxels within area (%)Relative volume change (%/30 Gy)p-valueLeft Frontal Operculum61.016.10.002Left Posterior Orbital Gyrus48.713.80.012Left Anterior Insula47.212.90.004Left Inferior Frontal Gyrus35.16.10.005Left Thalamus Proper31.121.20.002Left Inferior Frontal Orbital Gyrus28.96.10.020Left Inferior Frontal Angular Gyrus25.75.00.008 Open table in a new tab