Loading...
Presurgical structural connectivity predicts postsurgical cognitive impairment in glioma patients
Smolders,Lars De Baene,Wouter Gehring,Karin van der Hofstad,Remco Florack,Luc Rutten,Geert-Jan
Smolders,Lars
De Baene,Wouter
Gehring,Karin
van der Hofstad,Remco
Florack,Luc
Rutten,Geert-Jan
Abstract
Glioma patients frequently suffer from cognitive impairments after surgery, but predicting these impairments preoperatively at an individual level remains challenging. Cognitive functions are increasingly studied from a network perspective, where an important role is played by the Default Mode Network (DMN) and Frontoparietal Network (FPN). Hypothesizing that postsurgical cognitive impairments arise from structural network vulnerabilities, we trained models using presurgical structural connectivity of DMN and FPN regions to predict postsurgical cognitive impairment. We obtained individualized structural connectomes in 63 glioma patients (grades II–IV) who underwent diffusion-weighted MRI before surgery (T0) and neuropsychological screening 3 months after surgery (T3) and, for a small majority, adjuvant treatment. Random forest classifiers were trained on a combination of baseline (sociodemographic and clinical), tumour location and structural network variables available before surgery to predict postsurgical cognitive impairment in individual patients. Classifier performance was measured as area under curve of the receiver operating characteristic (AUC-ROC), testing statistical significance via permutation testing. Predictor importance was calculated post-hoc using Shapley additive explanations for trees. Postsurgical impairment was predicted by baseline variables available at T0 (AUC = 0.69, P = 0.011), presurgical DMN degrees (AUC = 0.73, P = 0.001), presurgical FPN degrees (AUC = 0.73, P = 0.001) and combinations of network and baseline variables (AUC = 0.75, P < 0.001; AUC = 0.76, P < 0.001 for DMN and FPN, respectively), but not by tumour location only (AUC = 0.62, P = 0.068). The combination of baseline variables, DMN degrees and FPN degrees (AUC = 0.76, P < 0.001) did not improve results. Importantly, models including network variables performed better than models using baseline or tumour location variables only. The most important predictors of postsurgical cognitive impairment were older age and low connectivity of the left lateral superior frontal gyrus (DMN), right pars opercularis (FPN) and bilateral middle frontal gyrus (DMN). This study represents a step towards preoperative prediction of postsurgical cognitive impairments in individual glioma patients. Our results underscore the importance of the DMN and FPN for cognition and suggest a biomarker for cognitive resilience to damage from treatment. The success of our model illustrates the utility of individual structural connectomes for studying cognitive impairment. Future expansions, e.g. incorporating resting-state fMRI, could improve our model. Ultimately, a sufficiently accurate model could be applied in neurosurgical planning by assessing a patient’s risk of postsurgical impairment from presurgical information only, improving counselling of glioma patients regarding surgical expectations.
Description
Date
2025-10
Journal Title
Journal ISSN
Volume Title
Publisher
Files
Loading...
fcaf346.pdf
Adobe PDF, 1.06 MB
Research Projects
Organizational Units
Journal Issue
Keywords
machine learning, tractography, network hubs, network neuroscience
Citation
Smolders, L, De Baene, W, Gehring, K, van der Hofstad, R, Florack, L & Rutten, G-J 2025, 'Presurgical structural connectivity predicts postsurgical cognitive impairment in glioma patients', Brain communications, vol. 7, no. 5, fcaf346. https://doi.org/10.1093/braincomms/fcaf346
License
info:eu-repo/semantics/openAccess