Poster B42, Thursday, August 16, 3:05 – 4:50 pm, Room 2000AB

Neurofunctional correlates of overall language function in aphasia

Jillian M. Lucanie1, Sarah M. Schneck1, Dana K. Eriksson2, Melodie Yen1, Deborah F. Levy1, Ian Quillen1, L. Taylor Davis1, Wayneho Kam1, Howard S. Kirshner1, Michael de Riesthal1, Stephen M. Wilson1;1Vanderbilt University Medical Center, 2University of Arizona

Aphasia is caused by damage to left hemisphere language regions, but language function in individuals with aphasia is the product not of the damage itself, but of the network of surviving regions. We used functional MRI in a diverse group of individuals with post-stroke aphasia to address two basic questions: (1) Which brain regions are involved in language processing in aphasia? (2) Are there any brain regions where functional activity is predictive of better overall language function? Critically, we addressed these questions using an adaptive semantic matching paradigm, which we have previously shown to be feasible, reliable and valid for identifying language regions in people with and without language impairments. Thirty-four individuals with subacute or chronic post-stroke aphasia took part in the study (age: 55.6 ± 15.0 years (range 24–80 years); 20 male, 14 female; 32 right-handed, 2 left-handed; time post onset 18 days–11 years). Aphasia was characterized with our Quick Aphasia Battery (QAB), and varied widely in severity: QAB overall score (out of 10) = 6.4 ± 2.5, range = 0.9–9.7. Random effects analyses were thresholded at voxelwise p < 0.005, then corrected for multiple comparisons using Gaussian random field theory. We found that as a group, people with aphasia recruited strongly left-lateralized inferior frontal and posterior temporal regions to perform the adaptive semantic matching task (p < 0.001), very similar to the previously reported pattern in neurologically normal participants. This suggests that surviving left hemisphere language regions continue to be responsible for language processing in post-stroke aphasia. We then correlated task-induced signal change with overall language function (QAB overall score). Two regions showed positive correlations between activation and overall language function. One was the left posterior superior temporal sulcus (pSTS) (2.46 cc, p = 0.046), which was one of the regions involved in the group average map described above. This finding was not surprising, since this is a critical language area where structural damage or dysfunction is known to impact language processing. The other region where activity correlated positively with language function was the homotopic area in the right pSTS (3.18 cc, p = 0.013). Comparison to data from neurologically normal participants showed that this correlation did not reflect reorganization of language to the right pSTS, since signal change in patients with better overall language did not exceed the normal range. Rather, the correlation may reflect either variable premorbid capacity of the right hemisphere to partially support language processing, or it may be a downstream consequence of the damage or dysfunction of the left pSTS. In sum, our findings suggest that most individuals with aphasia largely continue to process language using surviving left frontal and posterior temporal regions, of which only the latter are predictive of overall language function. There is little evidence for wholesale reorganization to the right hemisphere, but some evidence that the pre-existing linguistic capacity of the right hemisphere may be relevant for predicting outcome when left hemisphere language regions are damaged.

Topic Area: Language Disorders

Back