Search Abstracts | Symposia | Slide Sessions | Poster Sessions
The left hemisphere isn't so dominant after all: Prevalence of aphasia in acute right-hemisphere stroke
Poster Session C, Friday, October 25, 4:30 - 6:00 pm, Great Hall 3 and 4
Jeremy Yeaton1, Sigfus Kristinsson2, William Matchin2, Roger Newman-Norlund2, John Absher2, Julius Fridriksson2, Gregory Hickok1; 1University of California, Irvine, 2University of South Carolina
Background. It is the working assumption in the field that language is a left-lateralized system. Although there is some evidence for left-lateralization of some language functions, there is also evidence for bilateral organization of receptive language processing (Rogalsky et al., 2022). This left-hemispheric dominance assumption has led to a sampling bias in aphasia research: most large-scale aphasia studies focus on participants who a) have left-hemisphere lesions, and b) present with aphasia. This excludes two critical groups: 1) participants with left hemisphere lesions and no aphasia, and 2) participants with right hemisphere lesions who show aphasia. Revisiting these assumptions with a large sample could provide key insights. Methods. To provide some preliminary insights into this question, we analyzed data from a uniquely large, unselected sample of acute stroke patients (n=888). From a retrospective analysis of patient records, three coarse language measures were identified: 1) the language subscore of the NIH Stroke Scale (NIHSS), 2) a binary bedside assessment of presence of aphasia by an attending physician or Speech-Language Pathologist, and 3) proportion of correctly repeated items from WAB-R repetition subtest. Each participant’s lesion was traced from their MRI and automatically converted into percent of damage to ROIs under the JHU parcellation. We performed univariate statistics testing the reliability of damage to each ROI in predicting each of our language scores across the whole group, as well as filtering out participants who did not have contralateral damage when examining LH and RH ROIs separately. We used total lesion volume as well as stroke severity (total NIHSS score) as covariates. Results. When considering all gray matter ROIs across all participants, we found that damage to bilateral inferior frontal ROIs, as well as RH temporal ROIs were significant predictors of a bedside classification of aphasia. NIHSS Language score was predicted only by damage to LH ROIs: superior & inferior frontal gyri, SMG, STG, and insula. Repetition score was predicted by damage to bilateral superior and middle frontal gyri, right STG, MTG, and postcentral gyrus, and left fusiform gyrus. When considering only LH ROIs, we found that damage to IFG, STG, insula, and fusiform gyrus were predictive of NIHSS language score, while pre- and post-central gyri, as well as middle occipital gyrus were predictive of repetition score. When considering only RH ROIs, damage to inferior frontal and insular ROIs were predictive of bedside assessment of aphasia, and damage to STG was predictive of repetition score. Conclusions. Overall, damage to both LH and RH ROIs in the canonical language network were predictive of language scores in acute stroke. Critically, most of these ROIs are not implicated in motor-speech impairments, which may be conflated with language impairments in clinical scales. Although the measures used in this work are coarse, the substantial size and unselected nature of the sample lends credence to these findings, pointing to a need for more investigation into the role of the right hemisphere in aphasia, as well as for a reconsideration of the assumption of left hemisphere dominance in the neurobiology of language.
Topic Areas: Disorders: Acquired,