You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster B66, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

Areas predicting tDCS effects in primary progressive aphasia (PPA)

Kyrana Tsapkini1, Kim Webster1, Bronte Ficek1, Chiadi Onyike2, Brenda Rapp3, Argye Hillis1, Constantine Frangakis4;1Department of Neurology, Johns Hopkins University, Baltimore, MD, 2Department of Psychiatry and Behavioral Sciences, Johns Hopkins University, Baltimore, MD, 3Department of Cognitive Science, Johns Hopkins University, MD, 4Department of Biostatistics, Johns Hopkins School of Public Health, Baltiomore, MD

Introduction: Recently there has been a great interest and several positive outcomes using tDCS in patients with neurodegenerative diseases and in particular in PPA (Cotelli et al., 2014; Tsapkini et al., 2014). We have present the results of a large cohort of PPA participants 28 participants from all 3 variants of PPA. We tested the additive effects of 15 sessions of anodal tDCS over the left inferior frontal gyrus (a main language production area) coupled with written language production/spelling therapy, in a sham-controlled, double-blind, within-subjects crossover design. Here we present preliminary data from our study of 28 participants evaluating whether the volumes of language network areas predict therapy outcomes. Methods: Language therapy on written language production (written naming/spelling) was delivered for 15 daily sessions (45 min) by a speech-language pathology clinician 5 days/week. Electrodes were positioned on the F7 area of the participant’s scalp to simultaneously deliver either anodal tDCS or sham for the initial 20 min. Outcomes were assessed at multiple time points using graphemic scores, i.e., percent letters correct on trained and untrained words. Each outcome score was converted to a gain score by subtracting the outcome from the baseline score from the same period. For each follow-up time, we estimated the tDCS vs. Sham effect δ(T vs. S) based on the Generalized Estimating Equation approach (Liang & Zeger, 1986: Biometrika). Volumetric data for all language network areas and their right homologues as well as learning and memory areas were analyzed. The ROIs were: inferior frontal gyrus (IFG), supramarginal gyrus (SMG), middle temporal gyrus (MTG), superior temporal gyrus (STG), fusiform gyrus (FuG), anterior temporal pole (ATP), angular gyrus (AG) and hippocampus. MPRAGE (T1-weighted) scans were used and data were analyzed using software (Oishi et al., 2009: Neuroimage). We estimated the degree to which the additional tDCS effect over sham in therapy outcomes was associated with ROI volumes by a regression evaluating the interaction of treatment with volumes using the L1 (Lasso) penalty (Tibshirani, 1996: J R Stat Soc Ser B Method). We obtained significance levels by comparing the estimates with the distribution calculated by permuting the order of treatment assignment (tDCS then sham, or reverse) across patients. Results: Overall we found more beneficial effects of tDCS over sham in 28 participants with differential effects for each variant. Only the volumes of the left hippocampus and the left fusiform predicted performance on trained and untrained items in writing/spelling. Most importantly, a larger left hippocampus significantly predicted better performance in trained and untrained items. Additionally, tDCS helps therapy gains to maintain and generalize to untrained items only for those with larger left hippocampal and fusiform volumes. This finding shows that maintenance and generalization of therapy gains depend both on the hippocampus (learning substrate) and cortical regions specialized for the task (written naming).

Topic Area: Language Therapy

Back to Poster Schedule