Poster Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions

Lesion Mapping of Language Deficits in Acute Stroke Patients: An Analysis of NIHSS Best Language Scores and Clinical Neuroimaging

Poster Session C, Saturday, September 13, 11:00 am - 12:30 pm, Field House

Sigfus Kristinsson¹, Nilofar Sherzad², Roger Newman-Norlund³, John Absher⁴, Leonardo Bonilha⁵, Christopher Rorden³, Julius Fridriksson¹; ¹Department of Communication Sciences and Disorders, University of South Carolina, Columbia, South Carolina, United States, ²Arnold School of Public Health, University of South Carolina, Columbia, South Carolina, United States, ³Department of Psychology, University of South Carolina, Columbia, South Carolina, United States, ⁴School of Medicine, University of South Carolina, Greenville, South Carolina, United States, ⁵Department of Neurology, University of South Carolina, Columbia, South Carolina, United States

Introduction Lesion characteristics and initial severity of language deficits are critical prognostic indicators for long-term recovery in post-stroke aphasia.1,2 However, lesion anatomy associated with acute aphasia remains poorly understood.3-6 One approach to examine lesion-symptom relationships in acute stroke without jeopardizing patient care is to leverage existing clinical tools and patients’ medical records in retrospective analyses. Here, we combined clinical neuroimaging and language measures from the National Institute of Health Stroke Scale (NIHSS)7 to investigate lesion anatomy associated with acute aphasia severity in a large, non-selective cohort of stroke cases admitted to the Prisma Health-Upstate hospital in Greenville, South Carolina, between January 2019 and December 2020. Method Eligible participants were drawn from the Stroke Outcome Optimization Project8 dataset, which includes basic demographic information, NIHSS scores, and neuroimaging data. A total of 478 participants with NIHSS Best Language scores and neuroimaging were included in data analyses. The Best Language subscale is a widely used clinical assessment that relies on visual stimuli to assess acute language impairment. Aphasia severity is assessed on a 4-point scale ranging from ‘0’ (no aphasia) to ‘3’ (global aphasia). Since only 6 individuals presented with global aphasia, ratings of ‘2’ and ‘3’ were considered indicative of severe aphasia in statistical analyses. Clinical MRI scans were completed within 48 hours of stroke onset for most participants and within 30 days for all participants. Stroke lesions were manually demarcated on the structural images (T1-/T2-weighted, FLAIR) available for each participant using MRIcroGL.8 We performed voxelwise and region-of-interests whole-brain analyses to identify lesion correlates of acute aphasia severity across three group contrasts: no (‘0’) vs. mild-to-moderate (‘1’) aphasia, no (‘0’) vs. severe (‘2’/’3’) aphasia, and mild-to-moderate (‘1) vs. severe (‘2’/’3’) aphasia. Lesion-symptom mapping was performed using the NiiStat software and analyses included voxels or regions damaged in at least 10 participants. Overall lesion volume was treated as a covariate in all analyses and permutation thresholding was applied to correct for multiple comparisons (N=5,000, corrected p<.05). Results Participants with severe aphasia (N=102) were older (68.014.4y) and had larger lesion volume (49.951.5cc) than participants with mild-to-moderate aphasia (N=149; 66.113.9y; 24.639.9cc) or no aphasia (N=227; 64.114.4y; 16.227.4) (p<.05). Our primary analyses revealed that mild-to-moderate aphasia was associated with dorsal, superior temporal, and deep white matter lesion in the right hemisphere (RH) relative to no aphasia, whereas severe aphasia was associated with lesion to frontal, temporal, and basal ganglia anatomy, and fronto-temporal white matter connections exclusively in the left hemisphere (LH). Comparison between mild-to-moderate vs. severe aphasia mirrored these findings, with the strongest effects observed in the RH frontal lobe (mild-to-moderate>severe, Z>3.0) and LH frontal, superior temporal, and fronto-temporal connections (severe>mild-to-moderate, Z>4.0). Discussion As expected, our results revealed that damage to the canonical perisylvian language network in the LH is associated with severe acute aphasia. However, somewhat surprisingly, we found that lesion affecting homolog regions in the RH was robustly associated with milder acute language deficits. These findings are clinically relevant and inform our understanding of the neural mechanisms underlying language impairment in post-stroke aphasia.

Topic Areas: Disorders: Acquired, Speech-Language Treatment