Purpose: Post-stroke aphasia is typically associated with ischemic damage to cortical areas or with loss of connectivity among spared brain regions. It remains unclear whether the participation of spared brain regions as networks hubs affects the severity of aphasia.
Methods: We evaluated language performance and magnetic resonance imaging from 44 participants with chronic aphasia post-stroke. The individual structural brain connectomes were constructed from diffusion tensor. Hub regions were defined in accordance with the rich club classification and studied in relation with language performance.
Results: Number of remaining left hemisphere rich club nodes was associated with aphasia, including comprehension, repetition and naming sub-scores. Importantly, among participants with relative preservation of regions of interest for language, aphasia severity was lessened if the region was not only spared, but also participated in the remaining network as a rich club node: Brodmann area (BA) 44/45 - repetition (p = 0.009), BA 39 - repetition (p = 0.045) and naming (p < 0.01), BA 37 - fluency (p < 0.001), comprehension (p = 0.025), repetition (p < 0.001) and naming (p < 0.001).
Conclusions: Disruption of language network structural hubs is directly associated with aphasia severity after stroke.
Keywords: Connectome; aphasia; diffusion tensor imaging; hubs; language performance; magnetic resonance imaging; rich club; stroke; structural networks.