Correlation Between Neurocognitive Outcomes and Neuroaxonal Connectome Alterations After Whole Brain Radiotherapy: A Proof-of-Concept Study

TL;DR

This study shows that whole brain radiotherapy affects brain connectivity and is linked to cognitive decline in patients with brain metastases.

Contribution

The study introduces a novel method of using connectome analysis to assess the impact of whole brain radiotherapy on neurocognitive outcomes.

Findings

Baseline anomalies in brain networks increased after whole brain radiotherapy.

Neurocognitive decline correlated with changes in functional connectivity anomalies.

The multiple demand network showed the highest baseline anomaly frequency.

Abstract

Brain connectivity and its modulation by tumor-related treatment has garnered recent interest with developments in the field of “Connectomics”, which involves integrating data from multiple advanced imaging modalities and mapping the complex brain networks, providing insights into brain disorders. In the current study, we aimed to evaluate the effect of whole brain radiotherapy (WBRT) in patients with brain metastases on human brain networks and neurocognitive outcomes. In a prospective registry, data was integrated from diffusion tensor imaging and functional magnetic resonance imaging, and individualized brain maps of 15 neuronal networks were created for each patient at different time points. We observed that the baseline anomalies due to brain metastases increased post-WBRT across multiple networks. The anomaly changes also correlated with neurocognitive decline, suggesting the importance of functional connectivity in cognitive processes. The long-term outcomes might suggest connectome-tailored radiation dose constraints to preserve cognition.

Background/Objectives: Connectomics is an evolving branch of neuroscience that determines structural and functional connectivity in the brain. The objective of this prospective imaging study is to evaluate the effect of whole brain radiotherapy (WBRT) on the connectome. Methods: A combination of diffusion tensor imaging (DTI) and functional magnetic resonance imaging (fMRI) was used to study the structural and functional connectivity of the brain, and a machine learning algorithm trained to analyze subject-specific data was applied to create individualized brain maps with 15 neuronal networks for each patient. These brain maps were compared to normal brains from the human connectome project, producing an anomaly matrix. Connectome analysis and multi-dimensional neurocognitive testing on a web-based platform were performed at baseline and 3 months post-WBRT. The change in anomaly frequency was co-related with neurocognitive outcomes. Results: At baseline, connectome analysis revealed that the multiple demand network had the most anomalies (46%). Pre- and post-WBRT comparison revealed increases in proportional anomaly frequency across multiple networks. Pearson correlation showed correlation between neurocognitive domain decline and anomaly changes: learning and memory domain with subcortical network [Verbal recall (Pearson coefficient −0.94; p < 0.01), verbal revision (Pearson coefficient −0.89; p = 0.01), and verbal recognition (Pearson coefficient −0.94; p < 0.01)]. Conclusions: This proof-of-concept study integrated data from DTI and fMRI in the form of connectome and revealed significant changes in brain connectivity, with WBRT that also correlated with neurocognitive outcomes. Further studies in a larger cohort are underway, and correlations with white matter changes and tumor locations/numbers will be performed.

Figures