Randomized trial of transcranial direct current stimulation for poststroke dysphagia.

ABSTRACT

OBJECTIVE: We evaluated whether transcranial direct current stimulation (tDCS) is able to enhance dysphagia rehabilitation following stroke. Besides relating clinical effects with neuroplastic changes in cortical swallowing processing, we aimed to identify factors influencing treatment success. METHODS: In this double-blind, randomized study, 60 acute dysphagic stroke patients received contralesional anodal (1mA, 20 minutes) or sham tDCS on 4 consecutive days. Swallowing function was thoroughly assessed before and after the intervention using the validated Fiberoptic Endoscopic Dysphagia Severity Scale (FEDSS) and clinical assessment. In 10 patients, swallowing-related brain activation was recorded applying magnetoencephalography before and after the intervention. Voxel-based statistical lesion pattern analysis was also performed. RESULTS: Study groups did not differ according to demographic data, stroke characteristics, or baseline dysphagia severity. Patients treated with tDCS showed greater improvement in FEDSS than the sham group (1.3 vs 0.4 points, mean difference = 0.9, 95% confidence interval [CI] = 0.4-1.4, p < 0.0005). Functional recovery was accompanied by a significant increase of activation (p < 0.05) in the contralesional swallowing network after real but not sham tDCS. Regarding predictors of treatment success, for every hour earlier that treatment was initiated, there was greater improvement on the FEDSS (adjusted odds ratio = 0.99, 95% CI = 0.98-1.00, p < 0.05) in multivariate analysis. Stroke location in the right insula and operculum was indicative of worse response to tDCS (p < 0.05). INTERPRETATION: Application of tDCS over the contralesional swallowing motor cortex supports swallowing network reorganization, thereby leading to faster rehabilitation of acute poststroke dysphagia. Early treatment initiation seems beneficial. tDCS may be less effective in right-hemispheric insulo-opercular stroke. Ann Neurol 2018;83328-340.