E cacy of Continuous Theta Burst Stimulation on Brain Clearance Recovery in Post-stroke Cognitive Impairment

Guang-qing Xu (  guangchingx@163.com ) Guangdong Provincial People's Hospital https://orcid.org/0000-0003-2029-8177 Yi-qing Huang Shanghai East Hospital Ying-hua Jing Guangzhou First People's Hospital Qian Ding Guangzhou First People's Hospital Cheng Wu Guangzhou First People's Hospital Tuo Lin Guangzhou First People's Hospital Wan-qi Li Guangzhou First People's Hospital Zhong Pei Sun Yat-Sen University Chu-huai Wang Sun Yat-Sen University Yue Lan Guangzhou First People's Hospital


Background
Stroke induces substantial cognitive decline in many survivors [1] , and it is estimated that 70-90% of stroke patients incur at least some degree of post-stroke cognitive impairment [2] . Moreover, up to 50% of stroke patients demonstrate impairments in multiple cognitive domains [2] . This post-stroke cognitive impairment (PSCI) is also associated with progressive dementia, with up to 41% of patients meeting criteria for clinical dementia within the rst year post-stroke [3] . Although there is no standard de nition, PSCI typically refers to any cognitive de cit after a cerebrovascular event and can range from mild cognitive impairment to severe vascular dementia. Cognitive abilities determine functional performance [4,5] , so PSCI is a major cause of functional disability in activities of daily living (ADLs), impaired quality of life, post-stroke depression (PSD), progression to dementia, and reduced long-term survival [6] .
The treatment of cognitive impairment and prevention of further decline are essential aspects of stroke rehabilitation. A variety of interventions have been assessed, but there is only limited evidence to suggest bene cial effects of physical activity, cognitive training, and risk factor reduction; thus, more research is needed [7,8] . Indeed, improvement of cognition after stroke has been identi ed as a research priority by stroke survivors, caregivers, and health professionals [9] . Noninvasive brain stimulation (NIBS) techniques such as transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) are promising tools to enhance post-stroke recovery because they have the potential to induce long-term increases or decreases in regional cortical excitability [10] . In fact, the e cacy of NIBS in well documented for motor recovery [11,12] and language recovery [13] , highlighting the potential e cacy of NIBS for rehabilitation. However, further re nement of stimulation protocols is required for enhanced reliability and therapeutic e cacy. For instance, the potential of NIBS for post-stroke recovery of cognitive impairments outside the domain of language is still unclear, as are the methodological variables related to treatment e cacy. Kim [14] reported that 1 Hz repetitive transcranial magnetic stimulation (rTMS) over the contralesional posterior parietal cortex (PPC) resulted in a greater improvement of line bisection and letter cancellation tasks among stroke patients with visuospatial neglect. Similarly, Yang [15] found that continuous theta burst stimulation (cTBS) over the contralesional PPC improved symptoms of neglect, while Cao et al. reported that iTBS over the contralesional dorsolateral prefrontal cortex (DLPFC) also had signi cant e cacy in neglect patients [16] . Furthermore, 1 Hz rTMS over the right DLPFC resulted in a signi cant improvement of everyday memory function among PSCI patients [17] , and bene ts of TMS on executive function and other cognitive domains have been reported [18] . Although stimulation protocols and treatment regimens have varied substantially across these studies, most used a gure-8 coil and TMS to modulate cortical excitability in speci c brain regions. In addition to changing cortical excitability, our previous animal studies demonstrated that cTBS also enhances metabolite clearance via glymphatic pathways. Therefore, we speculated that enhanced brain glymphatic pathway function by cTBS treatment would accelerate the clearance of metabolic waste and improve cognitive function in patients with PSCI.
The objective of this study was to evaluate the effects of a 2-week daily cTBS regimen combined with computer-assisted cognitive training on global cognitive function, glymphatic pathway clearance e ciency, speci c cognitive domains, ADL, and quality of life in patients with PSCI. We speculated that cTBS would enhance the clearance e ciency of glymphatic pathways and improve cognitive function compared to sham cTBS.

Participants
Participants were recruited from the First A liated Hospital of Sun Yat-sen University and the First People's Hospital of Guangzhou. Inclusion criteria were rst stroke within the previous 12 months as con rmed by computed tomography (CT) scan or magnetic resonance imaging (MRI), evidence of strokerelated cognitive impairment as documented by neuropsychological evaluation, aged 40-80 years, primary school (grade 6) education or above, Mini-Mental State Examination (MMSE) score ≥ 15, Montreal Cognitive Assessment (MoCA) score less than 26, Hamilton Depression Rating Scale (HAMD) score < 17, not on drugs for cognitive improvement, and having a stable and reliable caregiver. Exclusion criteria were as follows: bilateral or subtentorial lesions; history of nervous system diseases such as stroke, brain tumor, and brain trauma; other causes of cognitive decline such as Alzheimer's disease, Parkinson's disease, frontotemporal dementia, dementia with Lewy bodies, VitB12 de ciency, or hypothyroidism; severe alcoholism and drug abuse; currently taking medications that may negatively affect cognitive functions such as anticholinergics, antipsychotics, and anticonvulsants; contraindications to MRI; contraindications to TMS treatment; psychiatric disorders such as schizophrenia, bipolar disorder, major depression and delirium; vision or hearing problems that could impede performance on cognitive tests; language de cits or aphasia. A researcher with a background in rehabilitation was responsible for screening and enrolling the participants.

Ethical considerations
The study protocol was reviewed and approved by the Ethics Committee for Clinical Research and Animal Trials of the First A liated Hospital of Sun Yat-sen University ([2018]111). The potential risks and bene ts of participation in this study were explained to each participant in advance, and all participants provided signed informed consent before participation. The study is registered at http://www.chictr.org.cn (ChiCTR1800017997).

Study Design
This was a single-blinded randomized controlled trial. After baseline evaluation, participants were randomized equally to the cTBS group or sham cTBS group. Group allocation was concealed in sealed sequentially numbered envelopes that were not opened until completion of baseline assessments. All randomization procedures were performed by an off-site researcher who was not involved in other aspects of the study. All assessments were performed by two researchers blinded to group allocation.

Intervention
Patients in the cTBS group received cTBS treatment and computer-assisted cognitive training for 2 weeks. Continuous theta-burst stimulation was applied using a Yiruide CCY-IA magnetic stimulator (Wuhan Yiruide Medical Equipment New Technology Co., Ltd., Wuhan, China) with a circular coil (Y125, 125 mm outer diameter). The cTBS protocol comprised 600 pulses delivered in 200 equally spaced bursts over 40 s (5 Hz), with each burst consisting of three pulses at 50 Hz [19] . The cTBS was applied over Cz according to the international 10-20 EEG System. The coil was held tangentially to the scalp, with the handle pointing in the posterior direction. Power was set to 80% of each patient's active motor threshold.
Participants received one treatment session per day, 5 days per week, for a total of 10 treatment sessions. Computer-assisted cognitive training was delivered using a computerized, multidomain, adaptive training program (Nanjing Wise Spirit Education Technology Co., Ltd., Nanjing, China). The training domains included processing speed, attention, perception, long-term memory, working memory, calculation, executive control, reasoning, and problem solving [20] . Participants completed 30 min of training per day, 5 days a week for 2 weeks. Patients in the sham cTBS group received the same protocol, except that the coil was held perpendicular to the scalp.

Measurements
Demographic information was obtained during a baseline interview. The outcome metrics detailed below were obtained before and after the intervention by researchers blinded to treatment group. The primary outcomes of this study were changes in Alzheimer's Disease Assessment Scale-cognitive subscale (ADAS-cog) score and analysis along the perivascular space (ALPS) index [21] . Safety was evaluated by recording adverse events and laboratory parameters.

ADAS-cog
The ADAS-cog score change from baseline was chosen as the primary endpoint of the trial [22] . The ADAScog has been widely used to measure cognitive changes in clinical trials of patients with Alzheimer's disease [23] and vascular cognitive impairment [24] . The ADAS-cog measures cognitive performance by combining ratings on 12 items addressing short-term memory, language, ability to orientate (re ects memory), construction/planning of simple designs, and performance. Although the ADAS-cog scale has not been validated in stroke, we considered it to be the most appropriate tool to assess cognition in stroke.
ALPS index: All images were acquired using a 3.0 T clinical scanner (Magnetom Verio, Siemens AG, Erlangen, Germany). T1-weighted images were acquired with the following parameters: 176 axial slices, The DTI data were preprocessed using PANDA software, a pipeline tool for analyzing brain diffusion images (http://www.nitrc.org/projects/panda/) [25] . A color-coded fractional anisotropy (FA) map and diffusivity map were also acquired by the software. Then, we split the diffusivity map and acquired diffusivity into separate x-axis, y-axis, and z-axis datasets. On the color-coded FA map of the plane at the level of the lateral ventricle body, we placed 5-mm diameter spherical regions of interest (ROIs) in the area of left hemisphere projection bers and association bers, and then calculated mean x-axis diffusivity values, abbreviated D x,p and D x,a . We also calculated the mean y-axis diffusivity in the ROI of projection bers and the mean z-axis diffusivity in the ROI of association bers, abbreviated as D y,p and D z,a . To evaluate the activity of the glymphatic system and the analysis along the perivascular space (ALPS), we calculated a ALPS index using the following formula:

Data analysis
According to pilot experiments, an estimated sample size of 18 PSCI patients (nine per group) would be required for detecting a signi cant group difference in ADAS-cog score change with 80% power at a twosided α = 0.05 by independent samples t-test assuming a mean (± standard deviation) score change of 8.0 ± 3.0 in the cTBS group, 4.0 ± 2.0 in the sham cTBS group, and an estimated drop-out rate of 10%. Therefore, we recruited candidates until 20 participants were identi ed meeting all inclusion criteria, with no reasons for exclusion, and agreeing to participate.
Data entry was performed independently by two researchers to ensure accuracy. Measurement data are expressed as mean ± standard deviation. SPSS version 24.0 (IBM, Armonk, NY) was used to conduct all data analyses. A P ≤ 0.05 (two-tailed) was considered statistically signi cant for all tests.
Comparison of baseline values. Baseline measurement data such as age, years of education, and course of disease were compared by independent samples t-test if normally distributed or by the non-parametric rank sum test if not normally distributed according to X. Baseline count data, such as sex, stroke type, and hemiplegia side were compared by Fisher's exact test.
Outcomes analysis. Group differences in outcomes were compared by independent sample t test or nonparametric rank sum test as indicated.

Results
Patients were recruited from May 2018 to March 2019. Twenty out of 285 PSCI patients screened met eligibility criteria and were randomly allocated to the cTBS or sham group (Figure 1).

Demographics and baseline assessments
Participant characteristics are summarized in Table 1. There were no signi cant between-group differences in demographic and outcome variables at baseline (P>0.05).

Compliance and adverse events
All participants completed the treatments and measurements. No adverse events were reported during the treatment period.

Effect of cTBS on the glymphatic system
We used the ALPS index to evaluate glymphatic system function (Figure 3). At baseline, ALPS index was similar between cTBS and sham cTBS groups (1.166 ± 0.145 vs. 1.166 ± 0.125, t = 0.015, P = 0.988), while the change from baseline after the intervention was signi cantly greater in the cTBS group compared to the sham cTBS group (0.182 ± 0.097 vs. 0.093 ± 0.080, t = 2.236, P = 0.038).

Effects of cTBS on other outcomes
There were no signi cant differences in MoCA, DST, Stroop Test, TMT, AVLT-H, VCFT, CDT, MBI, and SS-QoL between groups (P>0.05) ( Table 2) Discussion A brief (2-week, 10-session) cTBS regimen enhanced clearance e ciency along the perivascular space and improved cognitive outcomes in PSCI patients compared to sham cTBS. This is the rst demonstration that cTBS can promote metabolite clearance by the glymphatic system in order to improve cognitive function in PSCI.
Post-stroke cognitive impairment arises from multiple pathogenic processes. Cerebrovascular events can directly damage neurons in brain regions critical for higher cognition, such as the frontal, temporal, and parietal lobes, through mutually reinforcing episodes of metabolic failure, excitotoxicity, oxidative stress, osmotic stress, and in ammation. In addition, subcortical infarction or small vessel disease can damage nerve ber pathways and disrupt functional connectivity among neural networks [26] . Post-stroke cognitive impairment patients with previously normal cognition and lacking comorbid neurological diseases may show full or partial recovery without further deterioration of cognitive status [27] . However, in PSCI patients with preexisting pathological changes, such as clinically signi cant Aβ deposition, the cerebrovascular event may damage the cognitive reserve that normally serves to compensate for progressive degenerative changes, limiting recovery and resulting in irreversible cognitive de cits [26,28,29] . Further, there is a strong relationship among abnormal Aβ deposition, cerebral hypoperfusion, and impaired glymphatic pathways [30] . Koistinaho et al found that patients with PSCI exhibited lower perfusion in the hemisphere with greater Aβ, suggesting that persistent hypoperfusion accelerates Aβ deposition [31] . Further, the clearance e ciency of glymphatic pathways was signi cantly reduced by vessel ligation or occlusion [32,33] . Therefore, hypoperfusion may impair removal of metabolic waste by glymphatic pathways, resulting in enhanced Aβ deposition. A clinical PET study using the Aβ imaging agent 11 C-PIB also found increased Aβ deposition in patients with PSCI compared to age-matched control stroke patients, as well as a signi cant negative correlation between the level of cognitive dysfunction and Aβ deposition [34] . Therefore, hypoperfusion appears to impair cognitive function by exacerbating Aβ-related neurodegeneration Moreover, when the changes secondary to strokes, such as hypoperfusion, interact with the previously existing abnormal deposition of Aβ, it not only hinders the recovery of cognitive function, but also accelerates the decline of cognitive function [27] . In summary, we speculate that reduced clearance of metabolic waste due to impaired glymphatic pathways is a major contributor to the development of PSCI and an impediment to the recovery of cognitive function. Alternatively, previous animal studies by our research team have con rmed that cTBS can improve the clearance e ciency of glymphatic pathways in mice [35] . Therefore, we speculate that cTBS may improve cognitive function in PSCI patients by enhancing glymphatic clearance.
The outcomes of TMS for PSCI have been inconsistent across studies, possibly due to differences in stimulus protocols, outcome measures, and patient heterogeneity. This is the rst study in which cTBS was delivered by a circular coil at the apex of the skull (Cz) to in uence widely distributed glymphatic pathways. This cTBS delivery method has many advantages, including brief stimulation time (only 40 seconds), which may enhance treatment acceptance, and easy positioning of the coil for greater reproducibility. In addition, TBS uses lower stimulation intensity than traditional rTMS and so is less likely to induce epilepsy or other adverse events. Therefore, we believe that this TMS stimulation protocol can be applied safely and effectively in clinical practice for treatment of PSCI.
The primary outcomes of this study were ADAS-cog score and ALPS index. The change in ADAS-cog score from baseline was signi cantly larger in the cTBS group than the sham cTBS group, indicating that cTBS can improve global cognitive function in patients with PSCI. In addition, cTBS increased the ALPS index from baseline, indicating improved clearance of metabolites via glymphatic pathways, consistent with our previous animal studies [35] . A clinical trial by our research team found that global cognitive function was positively correlated with the clearance e ciency of glymphatic pathways (in press).
Among the main targets of glymphatic clearance is Aβ, and cognitive function in PSCI patients is negatively correlated with Aβ deposition [34] . Therefore, we speculate that cTBS may improve the cognitive function of PSCI patients by accelerated the clearance of metabolic waste, such as Aβ via the glymphatic system.
In contrast to global cognition and glymphatic function, there were no signi cant differences in secondary outcomes between cTBS and sham groups. However, it is possible that the intervention was too brief (2 weeks) to induce measurable improvements in speci c cognitive domains, ADL, and quality of life. Conversely, the sample size may be insu cient to detect more modest changes in these outcomes. Longer, larger-scale studies are warranted to comprehensively assess the bene ts of this cTBS paradigm on the functional recovery of stroke patients.

Conclusions
Continuous theta burst stimulation can improve cognitive outcome and enhance ALPS. Enhanced waste clearance via the glymphatic system induced by cTBS may contribute to improved cognitive function following stroke.

Declarations
Author Contribution GX, YH, YJ, CW, TL and WL contributed to the conception and design of the study and interpretation of the data. YH, YJ, CW and WL performed the research. YH, QD and YJ conducted the statistical analyses. ZP and CW provided additional statistical expertise related to the analysis. YH, QD and GX drafted the manuscript. YL and GX were the principal investigator of the study and were responsible for the study conception and interpretation of data. YL and GX had nal responsibility for the decision to submit for publication. All authors provided nal approval for the version of the manuscript submitted for publication and agreed to be accountable for the work.

Availability of data and materials
The datasets generated during the present study are available from the corresponding author upon request. Hypertension (n) 9 9 Myocardial ischemia (n) 0 0 Cerebral arteriosclerosis (n) 6 5 Coronary artery disease (n) 1 2 MMSE: Mini-Mental State Examination   ADAS-cog scores: mean differences from baseline. Differences between the cTBS and sham groups were signi cant at day 14 (P = 0.006).