Poster B37, Thursday, August 16, 3:05 – 4:50 pm, Room 2000AB

Leukoaraiosis Predicts Naming Outcomes after Treatment in Aphasia

Lisa Johnson1, Alexandra Basilakos1, Brielle C. Stark1, Sigfus Kristinsson1, Bonilha Leonardo2, Chris Rorden3, Julius Fridriksson1;1University of South Carolina Department of Communication Sciences and Disorders, 2Medical University of South Carolina, 3University of South Carolina Department of Psychology

Background Baseline aphasia severity and lesion volume has frequently been found to account for the greatest amount of variance in language recovery. In addition to these factors, it is likely that overall brain health, which is important for neuroplasticity and is characterized by measures of functional and structural integrity, may predict language recovery. In this study, we used contralesional leukoaraiosis (white matter hyperintensities attributed to, for example, artherosclerotic disease) as a surrogate measure of brain health. Based on prior work showing that the presence of leukoaraiosis negatively influences stroke outcomes, we predicted that those individuals with chronic aphasia demonstrating more advanced leukoaraosis in the contralesional hemisphere would make less or no improvement following speech-language treatment. This is the first study, to our knowledge, using leukoaraiosis as a potential predicting factor in aphasia recovery. Aims The purpose of this study was to determine whether leukoaraiosis can predict naming outcomes after treatment in individuals in the chronic stage of aphasia recovery. Methods Behavioral and neuroimaging data from 14 participants (3 F, age, M=63.14±7.9 years) in the chronic phase of stroke (months post-onset, M=73.2±61.2) were included. These data were obtained from a larger database as part of an ongoing study investigating factors that predict chronic aphasia recovery. Mean baseline aphasia quotient (AQ), a measure of aphasia severity obtained from the Western Aphasia Battery (WAB), was 46.2 (SD=14). All participants completed three weeks of daily aphasia therapy. The Philadelphia Naming Test (PNT) was administered at baseline and immediately following treatment. Proportional changes in PNT scores were used as the primary dependent variable. In addition, we obtained high-resolution T2 MRI scans for all participants. Leukoaraiosis severity was scored on T2 MRI scans using the modified Fazekas scale, where periventricular hyperintensities (PVH) and deep white matter hyperintensities (DWMH) are rated on a 4-point scale (0 being absent, 3 being severe). Fazekas scores, lesion volume, and baseline aphasia severity were separately correlated with proportional change in correct naming and subsequently compared using Fisher r-to-z transformation. Results Baseline WAB AQ, lesion volume, and PVH were not correlated with proportional change in correct naming (r=-0.26, p>0.05 for all); however, the correlation between DWMH and proportional change in correct naming resulted in a significant correlation (r=-0.85, p<0.0001). That is, higher DWMH ratings were significantly correlated with worse naming outcomes. Moreover, using Benjamini and Hochberg multiple comparison correction, the strength of the correlation between DWMH and naming outcomes was significantly greater than the relationship between naming outcomes and AQ (z=0.28, p=0.20), lesion volume (z=0.28, p=0.21) and PVH (z=0.37, p=0.37). Summary Given the small sample size, these results are preliminary. However, our results suggest that brain health, as measured by contralesional leukoaraiosis, provide an additional variable to consider for predicting recovery. Results from this study indicate general brain health should be considered when investigating neural correlates of disordered language, rather than lesion and baseline severity factors alone.

Topic Area: Language Disorders

Back