Hemodynamic signal changes and swallowing improvement of repetitive transcranial magnetic stimulation on stroke patients with dysphagia: A randomized controlled study

Our study aims to measure the cortical correlates of swallowing execution in patients with dysphagia after repetitive transcranial magnetic stimulation (rTMS) therapy using functional near-infrared spectroscopy (fNIRS), and observe the change of pattern of brain activation in stroke patients with dysphagia after rTMS intervention. In addition, we tried to analyze the effect of rTMS on brain activation in dysphagia patients with different lesion sides. This study also concentrated on the effect of stimulating the affected mylohyoid cortical region by 5 Hz rTMS, providing clinical evidence for rTMS therapy of dysphagia in stroke patients.

This study was a sham-controlled, single-blind, randomized controlled study with a blinded observer. A total of 49 patients completed the study, which was randomized to the rTMS group (n = 23) and sham rTMS group (n = 26) by the random number table method. The rTMS group received 5 Hz rTMS stimulation to the affected mylohyoid cortical region of the brain and the sham rTMS group underwent rTMS using the same parameters as the rTMS group, except for the position of the coil. Each patient received 2 weeks of stimulation followed by conventional swallowing therapy. Standardized Swallowing Assessment (SSA), Fiberoptic Endoscopic Dysphagia Severity Scale (FEDSS), Penetration-Aspiration Scale (PAS), and functional oral intake status were assessed at two times: baseline (before treatment) and 2 weeks (after intervention). Meanwhile, we use the fNIRS system to measure the cerebral hemodynamic changes during the experimental procedure.

The rTMS group exhibited significant improvement in the SSA scale, FEDSS scale, and PAS scale after rTMS therapy (all P < 0.001). The sham rTMS group had the same analysis on the same scales (all P < 0.001). There was no significant difference observed in clinical assessments at 2 weeks after baseline between the rTMS group and sham rTMS group (all P > 0.05). However, there were statistically significant differences between the two groups in the rate of change in the FEDSS score (P = 0.018) and PAS score (P = 0.004), except for the SSA score (P = 0.067). As for the removal rate of the feeding tube, there was no significant difference between the rTMS group and sham rTMS group (P = 0.355), but there was a significant difference compared with the baseline characteristics in both groups (P _rTMS < 0.001, P _shamrTMS = 0.002). In fNIRS analysis, the block average result showed differences in brain areas RPFC (right prefrontal cortex) and RMC (right motor cortex) significantly between the rTMS group and sham rTMS group after intervention (P _channel30 = 0.046, P _channel16 = 0.006). In the subgroup analysis, rTMS group was divided into left-rTMS group and right-rTMS group and sham rTMS group was divided into sham left-rTMS group and sham right-rTMS group. The fNIRS results showed no significance in block average and block differential after intervention between the left-rTMS group and sham left-rTMS group, but differences were statistically significant between the right-rTMS group and sham right-rTMS group in block average: channel 30 (T = -2.34, P = 0.028) in LPFC (left prefrontal cortex) and 16 (T = 2.54, P = 0.018) in RMC. After intervention, there was no significance in left-rTMS group compared with baseline, but in right-rTMS group, channel 27 (T = 2.18, P = 0.039) in LPFC and 47 (T = 2.17, P = 0.039) in RPFC had significance in block differential. In the sham rTMS group, neither sham left-rTMS group and sham right-rTMS group had significant differences in block average and block differential in each brain area after intervention (P > 0.05).

The present study confirmed that a 5-Hz rTMS is feasible at the affected mylohyoid cortical region in post-stroke patients with dysphagia and rTMS therapy can alter cortical excitability. Based on previous studies, there is a dominant hemisphere in swallowing and the results of our fNIRS analysis seemed to show a better increase in cortical activation on the right side than on the left after rTMS of the affected mylohyoid cortical region. However, there was no difference between the left and right hemispheres in the subgroup analysis. Nevertheless, the present study provides a novel and feasible method of applying fNIRS to assessment in stroke patients with dysphagia.