Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Does Integrity of the Phonological and Semantic Brain Networks Predict Response to Phonological and Semantic Aphasia Treatment?
Poster Session A, Friday, September 12, 11:00 am - 12:30 pm, Field House
Sigfus Kristinsson1, Dirk den Ouden1, Christopher Rorden2, Roger Newman-Norlund2, Leonardo Bonilha3, Julius Fridriksson1; 1University of South CarolinaDepartment of Communication Sciences and Disorders, University of South Carolina, Columbia, South Carolina, United States, 2Department of Psychology, University of South Carolina, Columbia, South Carolina, United States, 3Department of Neurology, University of South Carolina, Columbia, South Carolina, United States
Introduction Treatment-induced language reorganization in aphasia is a function of lesion and integrity of the residual language network.1,2 However, the differential effect of treatment focused on restoring impaired functional networks or strengthening preserved networks remains unclear.3 Here, we examined the extent to which integrity of brain networks predominantly involved in phonological vs. semantic processing predicts response to phonologically- and semantically-focused aphasia treatment, respectively. Method A total of 93 participants with chronic aphasia (>12 months post-stroke) after left-hemisphere stroke underwent language assessments and neuroimaging prior to receiving 15 hours of phonological and semantic treatment in a cross-over design.4 Naming performance was assessed with the Philadelphia Naming Test (PNT)5 before and after each treatment. Lesions were manually demarcated on T2-weighted structural images4 and functional naming-specific activity was obtained by contrasting naming-related activity and activity elicited by abstract image viewing acquired with task-based functional MRI.6 Lesion and activation maps were parcellated using the Johns Hopkins University atlas7 to define the regional distribution of each network. With reference to prior literature8-13, the phonological network was defined as the inferior frontal gyrus (IFG) pars opercularis, supramarginal gyrus, anterior insula, superior temporal gyrus (STG), and posterior STG, and the semantic network was defined as the IFG pars triangularis, angular gyrus, superior and middle temporal gyri poles, and middle temporal gyrus. Proportional lesion load and functional activity were averaged across areas in both networks, resulting in four values reflecting residual neural integrity for each participant: lesion load and residual activity in the phonological and semantic networks. Linear mixed-effects models (LMMs) were used test the main effect of lesion load and functional activity in each network as well as their interaction with lesion volume on pre- and post-treatment PNT scores. All LMMs were fit with subject-specific intercepts and adjusted for lesion volume, aphasia severity, age, and time post-onset. Results A significant main effect of lesion load in the phonological network on naming was revealed for both treatment phases (phonological: unstandardized ß=43.5, p=.028; semantic: ß=39.0, p=.034), but no effect of lesion load in the semantic network. No main effect of activity in either network on treated naming was observed (all p>.200). After adjusting for lesion volume interactions, activity in the phonological (ß=-11.3, p=.034) and semantic (ß=-19.7, p=.008) network was significantly associated with post-treatment relative to pre-treatment naming after semantic treatment only, and activity in the semantic network only was significantly influenced by lesion volume (ß=0.01, p=.046). Discussion Our results demonstrate that integrity of the phonological network is associated with naming performance irrespective of treatment focus, whereas integrity of the semantic network is selectively associated with semantic treatment effect. Moreover, the association between activity in the semantic network and semantic treatment effect is partially dependent on lesion volume, with greater activity predicting better recovery in larger relative to smaller lesions. These findings suggest that language reorganization following different aphasia treatments depends on the integrity of dissociable functional networks as well as overall lesion size, emphasizing the need for further research on restorative vs. preservative treatment foci.
Topic Areas: Disorders: Acquired, Speech-Language Treatment