Whole-brain structural connectivity predicts cognitive deficits in pretreated glioma patients

Abstract Purpose Glioma patients frequently suffer from cognitive dysfunction potentially caused by tumor invasion or treatment effects. We hypothesized that cognitive functioning in pretreated glioma patients critically depends on the maintained structural connectivity of multiple brain networks. Methods The study included 121 pretreated glioma patients (median age, 52 years; median ECOG score 1; CNS WHO grade 3 or 4) who had biopsy or resection plus chemoradiation. Cognitive performance was assessed by ten tests in five main cognitive domains 14 (1-214) months after therapy initiation. Hybrid amino acid PET/MRI using the tracer O-(2-[ 18 F]fluoroethyl)-L-tyrosine, a network-based cortical parcellation, and advanced tractography methods were used to generate whole-brain fiber count-weighted connectivity matrices. The matrices were applied to cross-validated machine learning-based models to identify fiber connections and underlying networks predictive for cognitive performance in the evaluated domains. Results Compared to healthy controls (n = 121), the cognitive scores were significantly lower in nine cognitive tests. The models well predicted 7/10 scores (median correlation coefficient, 0.47; range, 0.39–0.57) from 24–230 (0.5–4.6%) of the possible connections/edges, 84% were between nodes of different networks. Critically involved cortical regions/nodes (≥ 10 adjacent predictive edges) included predominantly left-sided nodes of the visual, somatomotor, dorsal/ventral attention, and default mode networks. Highly critical nodes (≥ 15–20 edges) included the default-mode network’s left temporal and bilateral posterior cingulate cortex. Conclusions These results suggest that the cognitive performance of pretreated glioma patients is strongly related to structural connectivity between multiple brain networks and depends on the integrity of known network hubs also involved in other neurological disorders.