We use cookies to ensure that we give you the best experience on our website Learn more

Home

Saved research

Submission

Prediction of Language Outcome after Chronic Infarct Associated Aphasia Using Multiple Compartment Lesion Analysis

Submitted

21 Views
7 Downloads
0 Saves

Presentation

thumbnail

Abstract

Purpose Infarct extent and location are related to aphasia symptoms and prognosis.1 Prior studies on voxel-based lesion-symptom mapping (VLSM) in stroke have largely considered only lesions on T1W images to represent overt damage; though a lack of consensus exists about what should be considered lesioned tissue, and some evidence suggests white matter (WM) disease on FLAIR may also predict outcomes.2-4 Our aim was to determine which lesions contribute to disease severity and treatment outcome Materials and Methods 79 participants (mean age 58.6, 26 female) with chronic stroke-induced aphasia were enrolled from 3 sites. The western-aphasia battery aphasia quotient (WABAQ) was used as a metric of global aphasia severity. Two neuroradiologists used Freesurfer to manually segment co-registered T1 MPRAGE and FLAIR images into core infarct, periinfarct gliosis, periventricular hyperintensities, right and left sided lacunes and WM hyperintensities (Figure 1), building on prior work in which only the infarct was segmented.4 Partial Spearman Rank-Order correlations controlling for intracranial volume and age were examined between WABAQ and volumetric estimates of each lesion class, while lacunes and WM lesions were tested using lesion counts. VLSM was used to analyze the spatial relationship between each lesion class and WABAQ. For VLSM, lesion masks were normalized into a standard (MNI) space, using enantiomorphic lesion replacement to minimize distortion caused by warping lesioned brains. Lesion size (core infarct + periinfarct gliosis) was regressed from both the voxel-level and behavioral data. At each voxel, a t-test was conducted on WABAQ scores between individuals with and without lesions in that location, and maps were cluster-corrected to a whole brain Family-Wise error rate of 0.01 using permutation testing (1000 permutations).5 Results In the volumetric analysis of lesions vs WABAQ, core infarct (ρ=-0.33, p<0.003) and peri-infarct gliosis (ρ=-0.55,p<0.001) showed significant negative associations, while periventricular WMH was marginally significant (ρ=-0.21,p=0.06). The effects of the peri-infarct gliosis remained significant after controlling for core infarct volumes (ρ=-0.488,p<0.001). Left hemisphere WM showed a positive association (ρ=0.3,p=0.006), which appeared driven by outliers and was no longer significant after removing subjects without any such damage (ρ=0.27,p=0.07). A preliminary power analysis suggested core infarct, periinfarct gliosis, and periventricular white matter disease classes were viable for VLSM analyses. VLSM maps showed effects of the core infarct in the superior temporal gyrus (STG), insula, and inferior frontal gyrus (IFG) and peri-infarct gliosis along the inferior longitudinal fasciculus (ILF) and arcuate fasciculus (AF) (Figure 3). No effects of periventricular WMH survived correction. Conclusion Results show that, no matter how severe the underlying burden of WM disease and lacunae, only core infarct and peri-infarct gliosis predict global language outcomes after an ischemic stroke, with the worst language outcomes having lesions in the STG, insula, IFG, ILF and AF. References 1. Sul, B., et al., Association of Lesion Location With Long-Term Recovery in Post-stroke Aphasia and Language Deficits. Frontiers in Neurology, 2019. 10 (776). 2. Wilmskoetter, J., et al., Long-range fibre damage in small vessel brain disease affects aphasia severity. Brain, 2019. 142(10): p. 3190-3201. 3. Yang, M., et al., Effect of White Matter Hyperintensity on the Functional Outcome of Ischemic Stroke Patients after Inpatient Stroke Rehabilitation. Brain Neurorehabil, 2019. 12(2). 4. Lukic, S., et al., Right Hemisphere Grey Matter Volume and Language Functions in Stroke Aphasia Neural Plasticity, 2017. https://doi.org/10.1155/2017/5601509. 5. Bates, E., et al., Voxel-based lesion–symptom mapping. Nature Neuroscience, 2003. 6(5): p. 448-450.

Datasets

No datasets are available for this submission.

Company

Legal

Follow us

© Copyright 2019 Morressier GmbH. All rights reserved.

© Copyright 2019 Morressier GmbH.
All rights reserved.