Structural connectome-based predictive modeling of cognitive deficits in treated glioma patients

In glioma patients, tumor growth and subsequent treatments are associated with various types of brain lesions. We hypothesized that cognitive functioning in these patients critically depends on the maintained structural connectivity of multiple brain networks.

The study included 121 glioma patients (median age, 52 years; median Eastern Cooperative Oncology Group performance score 1; CNS-WHO Grade 3 or 4) after multimodal therapy. Cognitive performance was assessed by 10 tests in 5 cognitive domains at a median of 14 months after treatment initiation. Hybrid amino acid PET/MRI using the tracer O-(2-[18F]fluoroethyl)-L-tyrosine, a network-based cortical parcellation, and advanced tractography were used to generate whole-brain fiber count-weighted connectivity matrices. The matrices were applied to a cross-validated machine-learning model to identify predictive fiber connections (edges), critical cortical regions (nodes), and the networks underlying cognitive performance (Fig. 1).

Structural connectome-based predictive modeling of cognitive deficits in treated glioma patients
Figure 1: Overview of methods for determining the whole-brain structural connectome and generating and validating a connectome-based predictive model (CPM) for cognitive functioning in glioma patients following multimodality treatment. SS3T-CSD, single-shell 3-tissue constrained spherical deconvolution; WM, white matter; FOD, fiber orientation distribution function; DWI, diffusion-weighted magnetic resonance imaging; GM, gray matter; CSF, cerebrospinal fluid; ACT, Anatomically Constrained Tractography; Pts, patients; Pat, patient.

Compared to healthy controls (n = 121), patients’ cognitive scores were significantly lower in 9 cognitive tests. The models predicted the scores of 7/10 tests (median correlation coefficient, 0.47; range, 0.39–0.57) from 0.6% to 5.4% of the matrix entries; 84% of the predictive edges were between nodes of different networks. Critically involved cortical regions (≥10 adjacent edges) included predominantly left-sided nodes of the visual, somatomotor, dorsal/ventral attention, and default mode networks. Highly critical nodes (≥15 edges) included the default mode network’s left temporal and bilateral posterior cingulate cortex (Fig. 2).

Structural connectome-based predictive modeling of cognitive deficits in treated glioma patients
Figure 2: Left: Binary connectivity matrices labeling cross-validated predictive edges; membership of nodes to networks marked in gray. Middle/Right: Anatomical representation of the critical nodes by visualization of node degree and connecting edges. Results are shown for 3 representative cognitive tests. (A) Corsi Block Tapping (visual working memory), (B) Trail-making Test B (processing speed), (C) Word List, immediate recall (verbal semantic memory). (D) Heatmap of the networks and nodes and their degrees of adjacent cross-validated predictive edges concerning the raised cognitive scores. Bw, Backward; L, left; R, right; Vis, visual; SM, somatomotor; dAtt, dorsal attention; vAtt, ventral attention; Limb, limbic; Ctrl, frontal control; DMN, default mode network; Temp, temporal; TMT-A (A), Trail-Making Test A (attention); TMT-B (E), Trail-Making Test B (executive function); SupM (L), Imagined Shopping Tour (language); DSf (VM), Digit Span Forward (verbal working memory); DSb (VM), Digit Span Backward (verbal working memory); CBTf (vM), Corsi Block Tapping Forward (visuospatial working memory); CBTb (vM), Corsi Block Tapping Backward (visuospatial working memory); WLi (eM), Word List, immediate recall (verbal episodic memory); WLd (eM), Word List, delayed recall (verbal episodic memory); PCC, posterior cingulate cortex.

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.


Friedrich, M., Filss, C. P., Lohmann, P., Mottaghy, F. M., Stoffels, G., Weiss Lucas, C., Ruge, M. I., Shah, N. J., Caspers, S., Langen, K.-J., Fink, G. R., Galldiks, N. & Kocher, M. (2024). Structural connectome-based predictive modeling of cognitive deficits in treated glioma patients. Neuro-Oncology Advances, 6(1), vdad15. doi: 10.1093/noajnl/vdad151

Correspondence to:

Michel Friedrich

Letzte Änderung: 12.01.2024