Structural network hubs as potential organs at risk in glioma patients after radiotherapy.

Abstract

Background: Cognitive sequelae are a concern in glioma patients post-radiotherapy. As there is uncertainty regarding which brain regions to spare during radiotherapy to preserve cognition, we explored structural brain network hubs as potential organs at risk.

Methods: We conducted a cross-sectional study, involving 39 irradiated adult WHO grade 2 and 3 gliomas along with 50 healthy controls. Cognitive domains (language, memory, attention, motor-, executive functioning) were assessed ≥1-year post-radiotherapy. Using multi-shell diffusion-weighted imaging, weighted structural graphs were constructed, and graph measures calculated to define hubs. The association between mean RTdose in each region and 1) nodal strength and 2) cognitive domains were tested with a linear regression model and Spearman's rho correlations, respectively.

Results: Lower nodal strength was significantly associated with increasing RTdose in nine brain regions, significantly (Mc Nemar's test, p<.001) impacting hubs more often than non-hubs (58% vs. 7%). Executive performance (r(37)≥-.474, p_FDR≤.045) and attention (r(37)≥-.471, p_FDR≤.045) were significantly correlated with RTdoses to the left pre- and postcentral gyrus and right posterior cingulate cortex, while poorer language outcomes were observed in patients receiving higher doses to the left insula, superior frontal, and precentral gyrus (r(37)≥-.460, pFDR.≤.045). These correlations were more prevalent in hubs than non-hubs (p=.033), and higher than those between memory and left (r(37)=-.359) and right (r(37)=.059) hippocampal dose.

Conclusion: Higher RTdoses to specific brain regions, particularly left-sided hubs, were associated with reduced nodal strength (i.e., lower network centrality) and poorer cognitive performance. While baseline cognitive testing is unavailable and cognitive functioning is influenced by multiple factors, this study highlights the potential value of network- or hub-sparing RT dose planning. Future longitudinal studies are needed to validate these findings before clinical implementation.