Effect Of RTMS On Swallowing Neural Network And Its Combination With NMES On Dysphagia After Stroke

Background:Dysphagia refers to the structural abnormalities of the swallowing organs such as the mouth,nose,throat,and esophagus,or the neurological diseases that innervate these organs,which results in the inability of food to enter the stomach safely and effectively from the mouth.According to this definition,if dysphagia is caused by structural abnormality of the swallowing organs,it is called structural dysphagia;if the nervous system that innervates the swallowing organs is damaged,it is called neurogenic dysphagia.Post-stroke dysphagia(PSD)is the most common neurogenic dysphagia,with an acute incidence of 29%-80%.Most patients with PSD can spontaneously recover their swallowing function within 2 weeks of onset,but about 11%-50%of patients still have long-term dysphagia.Patients with PSD will have clinical symptoms such as salivation,coughing on eating,laborious eating,choking,and repeated fever,which can lead to complications such as dehydration,malnutrition,and aspiration pneumonia.Therefore,how to improve dysphagia effectively and reduce complications is very important to patients.Traditional dysphagia therapy(TDT)includes adjustments of feeding appliances and position,oral sensorimotor therapy,direct feeding training,electromyographic biofeedback,balloon dilation and other targeted treatment techniques.The compensatory methods take effect immediately,but not last.Behavioral interventions have a cumulative effect,which can improve the swallowing-related muscles force and restore the normal swallowing biomechanism.However the effect shows slowly and relies on a lot of manpower and material resources,and some patients cannot cooperate with training actively due to disorders of consciousness and cognitive dysfunction,etc.So the treatment effect is poor.Balloon dilatation has not been widely used due to its narrow indications.Therefore,further exploration of new strategies for PSD treatment is of great significance to the patients with PSD.With the increasing study of the neural network of swallowing function,the focus of swallowing rehabilitation has shifted from improving the biomechanics of peripheral swallowing organs to the regulation of the underlying neural network.Neurostimulation technique is widely used in the rehabilitation of PSD due to its easy operation,safety and non-invasiveness,and low requirements for the patient's consciousness and cognitive level.It promotes neural plasticity through peripheral nerve or brain stimulation and helps patients to recover their impaired functions.Nowdays repetitive transcranial magnetic stimulation(rTMS)is the most widely used and possibly the most effective neurostimulation technique.It generates a pulsed magnetic field through a briefly changing current,which can penetrate the skull without attenuation.The magnetic field evokes subcortical micro-currents,which affect the metabolism and neuroelectric activity of the cortex,thereby regulate the neurons in the target cortex and the remote areas related to its function.In clinical application,there are many factors that affect the therapeutic effect of rTMS.In addition to the patient's own differences such as age,attention,hormone level,etc.,stimulation hemisphere,stimulation frequency,stimulation intensity,stimulation interval,number of pulses,presence or absence of pre-stimulation and the initial state of target muscles will affect the effect,so the application of rTMS in PSD presents a variety of stimulation modes.In recent years,there have been more and more combined applications of rTMS with NMES,tDCS or PES.However,because the physiological effects of rTMS are greatly affected by the stimulation parameters,it is still uncertain whether these combined stimulations can have a summative effect.On the other hand,rTMS is currently all applied to the cerebral cortex,requiring that the target skull to be stimulated is intact.However,some patients with hemorrhage in the cortex or subcortical basal ganglia have a skull defect or titanium mesh repair due to surgery,which limits the use of rTMS.Therefore,it is imperative to find new stimulation targets besides the cerebral cortex.Previous studies have confirmed that the cerebellum is involved in the physiological process of swallowing,so the cerebellum may become a new stimulation target for rTMS.However,it is still unclear about the effect of cerebellar rTMS stimulation on the swallowing neural network,thus limiting the application of rTMS in the cerebellum.This project intends to study the effect of cerebellar high-frequency rTMS stimulation on the swallowing neural network,evaluate the clinical efficacy of the combination of high-frequency rTMS on ipsilateral cerebral hemisphere and NMES on PSD,and analyze the effectiveness and safety of the application of rTMS in PSD.It provides a theoretical basis for understanding the neural network of swallowing function and further optimizing the treatment strategy of PSD.Purposes:1.To study the effect of cerebellar high frequency rTMS stimulation on the swallowing neural network.2.To explore the effect of combined stimulation of rTMS and NMES on the clinical efficacy,biomechanics and cortical excitability of PSD.3.Meta-analysis of the effectiveness and safety of the application of rTMS in PSD.Methods:Part 1:The study recruited 23 healthy volunteers.1.Determining the dominant cerebellar hemisphere for swallowing function and the cerebellar "hot-spot" of mylohyoid muscleSingle pulse TMS was used to act on both cerebellar hemispheres to induce MEP of mylohyoid muscle,and the hemisphere with the lower resting motor threshold(RMT)was determined to be the dominant cerebellar hemisphere for swallowing function.Marking the posterior occipital scalp,this area is the cerebellar "hot-spot" of mylohyoid muscle.2.Measuring the MEP amplitude of the mylohyoid muscle of the cerebral cortex on both sides before rTMS stimulationThe single pulse TMS was used to act on both cerebral hemispheres,and MEP of mylohyoid muscle was induced with the lowest stimulation intensity,and the MEP amplitude and stimulation intensity used were measured and recorded.3.Performing high frequency rTMS stimulation on the dominant cerebellumThe representative area of the mylohyoid muscle cortex on the dominant side of the cerebellar hemisphere was used as the stimulation target.10 Hz rTMS,100%RMT,250 pulses were used for stimulation,and the interval was 9s for I s.4.Re-measuring MEP amplitude of mylohyoid muscles of the cerebral cortex on both sides after rTMS stimulationAfter rTMS stimulation,the single pulse TMS of the same stimulation intensity was used to induce MEP of the mylohyoid muscles of the bilateral cerebral cortex again,and the amplitude was measured and recorded.Compared with the value before the stimulation,excitability changes of the cerebral cortex of the patient were evaluated.5.Evaluating the change of neural activity after cerebellar rTMS stimulation with resting-state functional magnetic resonance imaging(rs-fMRI)48 hours after elution,the same rTMS protocal was performed,and rs-fMRI were performed before and immediately after stimulation.After collecting and preprocessing rs-fMRI data,PerAF,DC,and ReHo were used to compare the changes of neural activity before and after rTMS stimulation,observing the effect of high-frequency cerebellar rTMS on the neural network of swallowing function.Part 2:The study included 45 patients with PSD lasting less than 6 months.They were divided into 3 groups by randomization:(1)5Hz ipsilateral cerebral rTMS with sham NMES(n=15),(2)ipsilateral cerebral sham TMS with NMES(n=15),(3)5Hz ipsilateral cerebral rTMS with NMES(n=15).Each of the active and sham rTMS lasted for 10 minutes,and each of the active and sham NMES lasted 30 minutes.The above treatment was once a day,5 times a week,totally 2 weeks.Before treatment and 2 weeks after treatment,the patients were subjected to videofluoroscopic swallowing study(VFSS),and the patients' penetration and aspiration were scored using the Rosenbek penetration-aspiration scale(PAS)to evaluate the clinical efficacy.The forward and upward movement speed of the hyoid bone was measured to evaluate the biomechanical changes.The MEP amplitude of the bilateral mylohyoid muscles was recorded to evaluate the cortical excitement.Effects of the combination of rTMS and NMES on PSD were explained from the perspectives of clinical manifestations,biomechanics and electrophysiology.Part 3:Two searchers independently searched the keywords published on PubMed,Embase,Medline database and Cochrane library as of January 2020 for deglutition disorder,swallowing disorders,dysphagia,esophageal dysphagia,stroke,cerebrovascular accident,cerebrovascular apoplexy,transcranial magnetic stimulations,TMS,transcranial neurostimulation,cortical stimulation,brain,rehabilitation.All English literature and the references that meet the inclusion criteria were adopted,and each of them is independently evaluated according to the Cochrane risk of bias assessment tool.The changes in related data before and after treatment were extracted from it,the combined effect size was calculated,and the RevMan 5.3 software was used for META analysis to analyze the effectiveness and safety of the application of rTMS in PSD.Results:Part 1:A total of 23 healthy volunteers were recruited in this experiment.During the study period,2 persons fell off(1 person could not tolerate the cerebellar rTMS,1 person could not cooperate with fMRI),and 2 persons were excluded from data preprocessing(1 person whose head moved more than 3mm)or 3°,1 person's registration effect is poor).There were 19 cases included finally,including 13 women and 6 men with an average age of 25.53±4.29.6 people had the dominant hemisphere on the left cerebellum,and 13 people had the dominant hemisphere on the right cerebellum.1.Changes of the swallowing cortex excitability on the bilateral cerebral hemispheres after rTMS stimulationIn order to study the effect of cerebellar rTMS on the swallowing cortex excitability of the bilateral cerebral hemispheres,this study observed the changes in the MEP amplitude of the representative area of the mylohyoid muscle cortex.It was found that the MEP amplitude of the swallowing cortex of the bilateral cerebral hemispheres increased significantly after stimulation(P<0.05),indicating that high-frequency rTMS of one side of the cerebellum can induce increased excitability of the bilateral swallowing cortex.2.Changes in neural activity of brain areas after rTMS stimulationIn order to observe changes of the whole brain neural activity after high-frequency rTMS cerebellar stimulation,this study performed fMRI,analyzing the acquired data with the analysis methods of PerAF,DC,and ReHo.The fMRI analysis based on the PerAF method found that after cerebellar stimulation,the neural activity of anterior central gyrus,parietal lobes,marginal lobes,bilateral anterior cuneiform lobes,bilateral middle frontal gyrus,bilateral posterior central gyrus,bilateral medial and lateral cusps was found changed after cerebellar rTMS.PerAF values of belt back,bilateral posterior cingulate gyrus,right dorsolateral superior frontal gyrus,corpus callosum,right motor auxiliary area and left thalamus were higher than before,the difference was statistically significant(P<0.05),which indicated unilateral cerebellar rTMS can increase the neural activity of the bilateral central anterior gyrus,which was consistent with the results of MEP.The fMRI analysis based on the DC method found that the DC values of the pontine and medulla oblongata after cerebellar rTMS were higher than before,and the difference was statistically significant(P<0.05),indicating that high-frequency cerebellar rTMS stimulation can increase the pons and bulbar neural activity.The fMRI analysis based on the ReHo method found that after unilateral cerebellar stimulation,the ReHo value of ipsilateral supplementary motor cortex decreased,and the difference was statistically significant(P<0.05),which indicated that unilateral cerebellar rTMS may inhibit the neural activity of ipsilateral supplementary motor cortex.3.Correlation analysis between changes of neural activity in pons and medulla oblongata and changes in cerebral swallowing cortex excitabilityIn order to explore the relationship between changes in brainstem neural activity and cerebral swallowing cortex excitability,we conducted the analysis based on the Pearson correlation coefficient.The results showed that the increase of DC value of the pons and medulla oblongata and the MEP amplitude of the contralateral cerebral swallowing cortex is positively correlated(P<0.05),while no correlation with the increase in MEP amplitude in the ipsilateral cerebral swallowing cortex(P>0.05),which verifies the excitatory cortex-brainstem effect after cerebellar rTMS stimulation.Part 2:45 patients were enrolled in this study.During the study period,1 person in the rTMS group fell off due to personal reasons,and 1 person in the NMES group fell off due to lung infection.43 patients finally completed the study.1.Observation of clinical efficacy of PSD after rTMS and NMES treatmentAfter 2 weeks rTMS and NMES treatment,the PAS scores of the 3 groups were lower than before treatment(P<0.001).Comparing between the groups,the PAS scores of the 3 groups after treatment were statistically different(P<0.05),among which,the PAS score of rTMS+NMES was the lowest,indicating the most obvious improvement in leakage and aspiration,followed by the NMES,and rTMS had the least improvement.2.Biomechanical changes of patients swallowing muscles after rTMS and NMES treatmentAfter 2 weeks treatment,the 3 groups of patients' hyoid bone upward and foward movement speed increased(P<0.05).Comparing between the 3 groups,the speed of rTMS+NMES was faster than rTMS alone and NMES alone(P<0.05),while no difference was found between NMES and rTMS(P>0.05).3.MEP changes of the affected swallowing cortex after rTMS and NMES treatment After 2 weeks treatment,the MEP amplitudes of the affected cortex of the three groups were increased(P<0.05);Comparing between the 3 groups,the MEP amplitude of rTMS+NMES was higher than that of rTMS alone and NMES alone(P<0.05),while no difference was found between rTMS and NMES(P>0.05).4.MEP changes of the unaffected swallowing cortex after rTMS and NMES treatmentAfter 2 weeks treatment,the MEP amplitude of the contralesional hemisphere in the rTMS group was slightly lower than that before,but the difference was not statistically significant(P>0.05);the MEP amplitude of the contralesional hemisphere increased in the NMES group and the rTMS+NMES group(P<0.05).In comparison between the groups,the increase in the NMES group and rTMS+NMES group was more significant than that in the rTMS group,but there was no statistically significant difference between the two groups(P>0.05).5.Correlation analysis between the clinical efficacy of PSD after rTMS and NMES treatment and the biomechanical changes of swallowing musclesAfter treatment,the difference in the forward and upward movement speed of the hyoid bone in the 3 groups was closely related to the difference in PAS scores(rTMS group pearson coefficient 0.850,0.740,P<0.05;NMES group pearson coefficient 0.744,0.777,P<0.05;The pearson coefficients of rTMS+NNMES group were 0.534 and 0.534,P<0.05).Part 3:The study included 6 published randomized controlled studies that met the inclusion criteria,involving 247 stroke patients(127 in the experimental group and 120 in the control group).1.Clinical efficacy of rTMS on PSD patientsCompared with the sham stimulation,the swallowing function of the patients with rTMS stimulation was improved,and the difference was statistically significant(SMD=1.19,95%CI:0.68 to 1.69;Z=4.69,P<0.00001;I2=68%).In order to find the source of heterogeneity,sensitivity analysis was performed.After excluding each study one by one,no significant change in I2 was found.2.Effect of rTMS stimulation hemisphere on the clinical efficacy of PSD patientsSubgroup analysis based on stimulation hemispheres showed that no matter stimulating the affected or unaffected cerebral hemisphere,or the combined stimulation of the bilateral cerebral hemispheres are effective(SMD=1.13,95%CI:0.06 to 2.2;Z=2.07,P=0.04;SMD=0.9,95%CI:0.42 to 1.38;Z=3.67,P=0.0002;SMD=1.78,95%CI:0.74 to 2.83;Z=3.34,P<0.0008).Especially,the combined stimulation effect of bilateral cerebral hemispheres seems to be better.3.Effect of rTMS stimulation frequency on the clinical efficacy of PSD patientsThe subgroup analysis based on the stimulation frequency showed that the low frequency subgroup(1Hz)and the high frequency subgroup(3Hz,5Hz,10Hz)were effective(SMD=1.02,95%CI:0.47 to 1.58;Z=3.6,P=0.0003;SMD=1.21,95%CI:0.46 to 1.96;Z=3.15,P=0.002).Although the overall treatment effect of the high frequency subgroup seems to be better than that of the low frequency subgroup,the improvement of swallowing function in the 10Hz subgroup alone was not statistically significant compared with the sham stimulation(SMD=0.6,95%CI:-0.02 to 1.22;Z=1.91,P=0.06).4.Observation of long-term effect of rTMS on PSDThe effect of rTMS can last for at least 1 month after the end of the treatment,and the effect scale is further improved compared with the end of the treatment(SMD=1.39,95%CI:0.86 to 1.92;Z=5.13,P<0.00001).Conclusions:Part 1:rTMS stimulation of the swallowing cortex in the dominant cerebellar hemisphere can increase the excitability of the swallowing cortex on both sides of the cerebral cortex,the neural activity of anterior central gyrus,corpus callosum,frontal lobe,parietal lobe,insula,cingulate gyrus,cuneiform lobe,and precuneus lobe,posterior central gyrus,medulla oblongata,and pons increased,and the increase of DC values in the medulla oblongata and pons was positively correlated with the excitability increase of the contralateral cerebral swallowing cortex,suggesting that contralateral excitatory cortex-brainstem effect appeared after high frequency rTMS stimulation of one side of the cerebellum.The excitability increase of the ipsilateral M1 cortex and the supplementary motor cortex were presumably due to the excitatory effect transmitted by the contralateral M1 cortex through the corpus callosum.Part 2:5Hz rTMS ipsilesional stimulation combined with NMES promoted the recovery of swallowing function in patients with PSD,and the effect was better than that of NMES alone and 5Hz rTMS ipsilesional stimulation.The peripheral mechanism may be related to the increase of hyoid bone upward and forward movement speed to improve the pharyngeal leakage and aspiration performance.The central mechanism may be that NMES enhances the excitability increase of the affected hemisphere induced by rTMS.The therapeutic effect of NMES was better than that of 5Hz rTMS stimulation alone,which may be related to the peripheral stimulation that strengthened the biomechanics of the mylohyoid muscle and excited the bilateral swallowing cortex.Part 3:rTMS treatment can improve the swallowing function of PSD,and the treatment effect existed for a long time.The choice of bilateral hemisphere combined stimulation or high frequency stimulation may be better than other parameters.