Intermittent theta burst stimulation for poststroke non-spatial attention deficit: a protocol of prospective, double-blinded, single-centre, randomised controlled trial in China

Abstract

Introduction

Attention deficit is the most common cognitive impairment after stroke, which can significantly hinder the recovery of both other cognitive domains and motor functions. Increasing evidence suggests that the left dorsolateral prefrontal cortex (DLPFC) is related to non-spatial attention functions, which indicates that it may be a promising target of repetitive transcranial magnetic stimulation (rTMS) for treating poststroke non-spatial attention deficit. Theta burst stimulation (TBS) is a modified pattern of rTMS that delivers shorter stimulation times and exhibits superior therapeutic efficacy. This study aims to provide evidence regarding the efficacy of intermittent TBS (iTBS) over the left DLPFC to improve poststroke non-spatial attention deficits and elucidate the potential neurophysiological mechanisms.

Methods and analysis

In this single-centre, prospective, randomised, sham-controlled clinical trial, patients with non-spatial attention deficits (n=38) received 10 sessions of real iTBS (n=19) or sham iTBS (n=19) over the left DLPFC and a 30-min conventional attention training. Neuropsychological evaluations, electrophysiological examination and neuroimaging scan will be conducted at baseline, postintervention (second week) and 2-week follow-up (fourth week). The primary outcomes are the change in the Montreal Cognitive Assessment scores and the Digital Span Test scores from baseline to the end of the intervention (second week). The secondary outcomes comprise changes in magnetic resonance spectroscopy neuroimaging from baseline to the end of the intervention (second week) as well as attention test batteries (including tests of selective attention, sustained attention, divided attention and shifting attention) and ERP P300 from baseline to endpoint (fourth week).

Ethics and dissemination

This study has been approved by the Institutional Ethical Committee of Tongji Hospital (ID: TJ-IRB20230879). All participants will sign the informed consent. Findings will be published in peer-reviewed journals and conference presentations.

Trial registration number

ChiCTR2300068669.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study will be performed at a tertiary rehabilitation centre, which makes it easy to obtain an adequate sample size.

• The outcome measures include conventional neurobehavioural scales, electrophysiological and neuroimaging techniques.

• All participants will receive conventional attention training whether or not they received intermittent theta burst stimulation.

• Two-week follow-up results will be analysed.

• The limitation of this study is its single-centre design.

Introduction

Attention is the most commonly affected cognitive function after stroke, with a prevalence ranging from 46% to 92%.¹ It plays a crucial role in the rehabilitation of various cognitive domains, such as language, memory and executive control,² as well as in the recovery of everyday activities related to movement function.³ For instance, sustained attention within the first 2 months has been identified by Robertson et al as a predictive factor for motor and functional recovery within the next 2 years.⁴ Both selective and shifting attention play important roles in balancing function and daily living.⁵ Additionally, selective auditory attention is related to balance control and falls.⁶

Since brain damage selectively affects different domains of attention,⁷ it has been shown that individual training in the specific attention component may be more effective than training in other domains during recovery from attention deficits.² Therefore, it is necessary to distinguish the various attention components. Based on current theories,^{2 8} there are two types of attention: spatial attention, which leads to neglect when disturbed; and non-spatial attention, which involves the following attention components: selective attention, sustained attention (vigilance), divided attention and shifting attention (see table 1). The prevailing approach for managing poststroke non-spatial attention deficits primarily entails conventional cognitive rehabilitation techniques and computer-based tasks.¹ However, the efficacy of these methods is limited, and patient compliance is often poor. Thus, there is an urgent need to identify novel interventions that can enhance attention function.

Table 1.

Domains of non-spatial attention

Domain of attention	Definition	Functional example
Selective attention	Ability to focus on specific stimuli while ignoring irrelevant stimuli	Reading while people talk in the background
Sustained attention (vigilance)	Ability to maintain attention over a prolonged period	Driving a car for long distances
Divided attention	Ability to multitask and divide attention between two or more tasks	Talking on the telephone while cooking
Shifting attention	Ability to shift attention flexibly between different tasks	Stop reading and chat with friends

Repetitive transcranial magnetic stimulation (rTMS) is a non-invasive and relatively safe technique for neural modulation, widely applied in neurological and psychiatric rehabilitation.⁹ Several studies have demonstrated the potential of transcranial magnetic stimulation (TMS) in improving attention deficits in healthy adults, and in patients with Alzheimer’s disease, depression or schizophrenia.^10–13 Intermittent theta burst stimulation (iTBS) is a modified version of rTMS proposed by Huang et al,¹⁴ 10 short sequences of 2 s duration are given every 10 s for 20 cycles, associated with excitatory after effects of cortical activity. It has presented advantages over other conventional rTMS strategies in its low intensity, short duration of application and long-lasting effects.^{15 16} In particular, the TBS protocol has been used to mimic the brain’s natural firing patterns to upregulate or downregulate the excitability of focal regions of the cortical surface with relatively high accuracy.^{14 17}

Despite the exact mechanism of poststroke attention deficit is still unclear, study has indicated stroke commonly results in cognitive impairment that is biased towards the left hemisphere.¹⁸ In terms of prognosis, a 2-year non-interventionist stroke longitudinal study concluded that patients with left hemisphere damage exhibited a significantly lower rate of improvement in attention deficits compared with those with right hemisphere damage.¹⁹ Besides, the prefrontal cortex plays a crucial role in the network responsible for non-spatial attentional control during cognitive processes.⁷ In particular, the left dorsolateral prefrontal cortex (DLPFC)^{20 21} is instrumental in managing deficits in selective and divided attention⁷ as well as sustained attention.²² Similar to language functions, attention also relies on specific brain regions as well as bilateral distributed brain networks and connections.²³ Studies have found that poststroke attention deficits in the internal capsule, caudate tail and thalamus are associated with disruption of the fronto-striato-thalamic circuit, which in turn affects the control of attention and executive function by the DLPFC and anterior cingulate cortex.¹⁸ Likewise, patients with basilar artery occlusion disease exhibit impairments in selective, sustained and shifting attention, which is attributed to the damaged connections between the thalamus and the parietal and frontal cortices.²⁴ The evidence above suggests that the left DLPFC may be an optimal stimulation site for neuromodulation in the treatment of attention deficits.

Recently, research on the effects of TMS on poststroke patients' attention has mainly focused on the spatial attention domain,²⁵ with limited reports on non-spatial attention deficits, which only address specific attention components.²⁶ In the studies conducted by Chu et al²⁷ and Tsai et al²⁸ on the effects of iTBS targeting the left DLPFC in the treatment of poststroke cognitive impairment, it was found that the attention function, as a major component of cognition, also be improved.

We hypothesise that stimulating the left DLPFC with excitatory iTBS can produce beneficial effects in augmenting poststroke attention recovery by regulating cortical local brain electrical activity and metabolism involved in attention processing. Besides, we will assess objective indicators using event-related potential (ERP) P300 and changes in neuroimaging using magnetic resonance spectroscopy (MRS). We hope to provide new evidence regarding the therapeutic efficacy of iTBS over the left DLPFC and elaborate on underlying neurophysiological mechanisms.

Methods and analysis

Study design

The study will be a prospective, randomised, sham-controlled, single-centre trial (registration number: ChiCTR2300068669) that will be conducted in the Department of Rehabilitation Medicine of Tongji Hospital in Wuhan, China. This protocol follows the Consolidated Standards of Reporting Trials Statement on randomised trials and it will be conducted according to the Standard Protocol Items: Recommendations for Interventional Trials guidelines. Hospitalised patients with poststroke attention deficits will be recruited to participate in the study. The flowchart overview and data collection of the study are presented in figure 1 and table 2. Thirty-eight eligible subjects will be randomly assigned with a ratio of 1:1 to a real or sham stimulation group. On the day of enrolment, after the end of iTBS intervention, and 2 weeks post-treatment completion, a battery of attention tests (including tests on selective attention, sustained attention, divided attention and shifting attention), Montreal Cognitive Assessment (MoCA), Digital Span Test (DST) and P300 are performed to evaluate attention function in poststroke attention deficit. Also, MRS data are collected before and after the end of the intervention. The expected overall study duration is approximately 15 months, from March 2023 to June 2024.

Figure 1.

Flow chart of the study procedure. 1H-MRS, proton magnetic resonance spectroscopy; DCT, Digital Cancellation Test; DST, Digital Span Test; iTBS, intermittent theta burst stimulation; MoCA, Montreal Cognitive Assessment; PASAT, Paced Auditory Serial Addition Test; SCWT, Stroop Colour-Word Test; TMT-B, Trail-Making Test B.

Table 2.

Schedule of enrolment, interventions and assessments throughout the trial

Study period visits	Pre-enrolment	Enrolment	Intervention	Postintervention assessment	Follow-up
Time point	1 week	0	0–2 week	2nd week	4th week
Eligibility screen	×
Informed consent	×
Allocation		×
MoCA	×			×	×
DST	×			×	×
SCWT		×		×	×
TMT-B		×		×	×
DCT		×		×	×
PASAT		×		×	×
P300		×		×	×
1H-MRS		×		×

DCT, Digital Cancellation Test; DST, Digital Span Test; 1H-MRS, proton magnetic resonance spectroscopy; MoCA, Montreal Cognitive Assessment; PASAT, Paced Auditory Serial Addition Test; SCWT, Stroop Colour-Word Test; TMT-B, Trail-Making Test B.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Participants

The inclusion and exclusion criteria of the present study correspond to the guidelines for using rTMS in clinics and research.⁹ The inclusion criteria are as follows: (1) first stroke meeting the diagnostic criteria established by the Fourth National Cerebrovascular Disease Academic Conference in 1995 confirmed by MRI or CT; (2) stabilised vital signs and neurological symptoms, course of stroke between 2 weeks and 6 months; (3) age 30–80 years old, right-handed; (4) mild or moderate cognitive impairment, defined by MoCA²⁹ score ≥10 and <26, with evidence of attention deficit initially screened by MoCA (attention function score ≤4) and digit span test (digits forwards/backwards <5); (5) patients can cooperate to complete the scale assessment without mental illness or communication barriers; and (6) sign the informed consent. The exclusion criteria are as follows: (1) patients with brain diseases other than cerebrovascular diseases, such as brain trauma, encephalitis, brain tumour, and so on; (2) cognitive impairment that occurred before stroke or caused by Alzheimer’s disease, Parkinson’s disease, dementia with Lewy bodies, and so on; (3) history of substance or alcohol abuse, premorbid seizures or neuropsychiatric diseases; (4) contraindications involving TMS and MRS (eg, skull defect or skin damage at the stimulation site, intracranial implants, cardiac pacemakers, implanted drug pumps, etc); (5) serious impairments of the heart, lung, liver, kidney and other organs, cannot tolerate training; (6) unable to cooperate with assessment and treatment due to severe visual, hearing or speech impairment; and (7) patients who had received TMS treatment or are currently participating in other clinical trials.

Randomization and blinding

Once the inclusion criteria are satisfied, the physician will explain the whole content of the trial to the participants and ask them to sign the informed consent. Then, the participants will be allocated to one real or sham iTBS group using an electronic random sequence generator (www.random.org) to receive a real or sham iTBS over the left DLPFC coupled with conventional attention training. The concealment of allocation is performed using sealed envelopes with numbers. The participants, assessors and physicians performing conventional attention training or the data collection/processing are blinded to the grouping. Only physicians who implement the iTBS intervention were aware of the allocation results due to the nature of the iTBS intervention.

Interventions

The iTBS treatment will be administered using a CCY-I-type stimulator (Wuhan Yiruide Medical Equipment New Technology Co, Ltd, China) equipped with a figure 8-shaped coil. Before the intervention, a single-pulse TMS will be applied over the motor cortex of the healthy hemisphere to determine the resting motor threshold (RMT). The RMT is defined as the minimum stimulus intensity that produces a motor-evoked potential >50 μV in more than 5 out of 10 trials after a recording electrode was attached to the contralesional first dorsal interosseous muscle of the non-paretic hand as the target muscle.⁹ A TMS-navigation system (Localite TMS navigator, Germany) will be used to locate the therapeutic target at the left DLPFC, with an average functional coordinate position of (x, y, z) = −46, 25, 44 (Talairach coordinates (x, y, z) = −47, 24, 48 Montreal Neurological Institute coordinates).³⁰ Each participant receives real or sham iTBS over the left DLPFC at 100% RMT.²⁸ A standard iTBS protocol consists of bursts of three pulses at 50 Hz repeated at 5 Hz (2 s on, 8 s off) for a total of 200 s and 600 pulses.^{14 31} In the real stimulation group, the stimulation coil will be oriented vertically to the target, while the stimulation face of the coil will be rotated 90° perpendicular to the target area to generate minimum stimulation during sham stimulation. The stimulation parameters and sites of the two groups are consistent. Immediately following real or sham iTBS, all participants receive a 30-min conventional attention training that includes training on selective attention, sustained attention, divided attention and shifting attention. Based on the previous studies, each patient will receive 10 sessions of iTBS of the left DLPFC within 2 weeks.^{28 32}

Tolerability and safety

Participants will be monitored during the treatment to identify any negative experiences, such as pain on the stimulation site, nausea, headache and so on. Particularly, seizures that are the most severe rTMS-related side effects, with a crude risk of approximately 0.02%,³³ are only expected to occur during or immediately after iTBS treatment. The investigator will record and report all adverse events during the treatment and within 1 week after the end of treatment. If adverse events occur in the subjects, the following procedures should be followed: if necessary, symptomatic treatment can be initiated first. The investigator will conduct an initial assessment of the severity of the adverse event and its relevance to TBS stimulation, and provide further treatment advice. In the case of a serious adverse event such as a seizure, prompt treatment will be administered and consideration will be given to terminating the trial. Furthermore, any serious event will be immediately reported to the Medical Ethics Committee of Tongji Hospital Affiliated to Tongji Medical College of Huazhong University of Science and Technology.

Outcome measurements

The primary outcomes are the change in the MoCA scores and the DST scores before and after a 2-week treatment, which are used to evaluate improvements in cognitive function and attention function. As secondary outcomes, attention test batteries will be used to assess various aspects of attention function, including Stroop Colour-Word Test for selective attention, Digital Cancellation Test (DCT) for sustained attention, Trail-Making Test Part B (TMT-B) for shifting attention and Paced Auditory Serial Addition Test (PASAT) for divided attention; in addition, ERP P300 and MRS will be examined. These assessments are conducted at baseline, postintervention (second week, and 2-week follow-up (fourth week). The neuroimaging test will be evaluated at baseline and the end of the intervention (second week).

Montreal Cognitive Assessment

The MoCA scale is used to screen the overall cognitive function and attention deficit of patients before and after the treatment. The cognitive domains assessed by the MoCA scale include visual-spatial and executive functions, naming, memory, attention, language, abstraction, delayed recall and orientation, with a total score of 30 points.²⁹ A score of ≥26 indicates normal cognitive function, while cognitive impairment is classified as mild for scores between 18 and 26, moderate for scores between 10 and 17 and severe for scores <10 (if the subject has an education level of ≤12 years (high school level), add 1 point to the score).³⁴ Among them, the total score for the attention function is 6 points, including three items: digits forward and backward (1 point each), a sustained attention task (target detection using tapping; 1 point) and a serial subtraction task (3 points). Attention function subtest score ≤4 points, suggesting attention deficit.³⁵

Digital Span Test

The scale is used to assess attention and working memory,³⁵ which comprises two sections: digits forward and digits backward. In the forward section, participants are required to repeat a sequence of 10 numbers in the same order, while in the backward section, they must repeat a sequence of 9 numbers in reverse order. Scores are calculated based on the number of digits accurately recalled, with the maximum number of digits completed representing the test score. All participants are allowed two attempts for each item, with a score of each section <5 points, indicating attention deficit.

Stroop Colour-Word Test

The Stroop test is commonly used to assess selective attention.³⁶ The test involves three cards (Card A, Card B and Card C). Card A has black Chinese characters printed in red, yellow, blue and green representing four different colours. Card B has four coloured dots with a diameter of 6 mm (red, green, yellow and blue). Card C has coloured Chinese characters (red, yellow, green and blue), but the colour of the font does not match their pronunciation. The patients are required to accurately and quickly read the pronunciation of the black Chinese characters on card A, the colour of each coloured dot on card B and the colour of each coloured Chinese character on card C, rather than their pronunciation. The time taken by participants to read each card will be recorded, along with the number of mistakes made on Card C.

Digital Cancellation Test

The DCT is a paper-and-pencil Continuous Performance Test that measures sustained attention and is widely used in China.³⁷ The test comprises five sections and is composed entirely of Arabic numerals. To minimise the influence of executive function, this study selects the first two parts with low difficulty. In detail, participants are asked to identify and cross out digit ‘3’ (the total number is 45) and the digit prior to ‘3’ (the total number is 45). Sustained attention index = (total number of words read/review time) × [(number of correctly crossed out words − number of wrongly crossed out words)/number of words that should be crossed out].³⁸

Trail-Making Test Part B

The TMT-B is a classic measure of attention shifting, involving the alternation of 24 numbers and letters.³⁹ To accommodate Chinese patients, we have made some modifications, using numbers 1–12 (each appearing twice, respectively, in the circle and square graphics) randomly distributed in different positions on the paper. The subjects are instructed to connect the numbers in ascending order, alternating between circles and squares. Performance is measured based on the time required and the number of errors made.

Paced Auditory Serial Addition Test

The PASAT is a commonly employed tool for evaluating divided attention.⁴⁰ While there are many versions of the test, the present study specifically utilised the 2 s trial of the PASAT. The subjects are asked to listen carefully to a group of numbers played in the recording and to say the sum of two adjacent numbers heard as soon as possible. The time interval between each number was 2 s. The sequence consisted of 30 numbers, randomly arranged from 1 to 9. Participants could earn 1 point for each correct answer, with a maximum possible score of 30 points.

Event-related potentials

The P300 is a late component of the ERPs, first described by Sutton et al in 1965, and widely believed to be associated with context updating, allocation of attentional resources and working memory processes.⁴¹ TBS modulates neural activity in the DLPFC area, thereby influencing the processing of external stimuli by the subjects, as reflected through the P300 ERP. In general, P300 latency is negatively correlated with MoCA scores in visuospatial, executive function, and attention domains, whereas there is a positive correlation between P300 amplitude and these scores.⁴² Participants will be assessed and averaged on a Keypoint9033A07 multifunctional ultrasonic/electroacoustic spectrometer (Natus Medical Instrument Co). The recording electrodes are placed at the Pz, following the 10/20 system of the International Electroencephalograph Society. The forehead (FPz) is used for grounding, and the reference electrodes are connected to the two earlobes. P300 waves are recorded in the Oddball paradigm, and the impedance of the electrodes is kept below 50 kΩ. All patients should be in a sitting position, with eyes closed, relaxed but awake, listening to the sound and counting the number of occurrences of the target stimulation during the recording. Sound stimulation will be given through earphones, including two tones (75% non-target stimulation and 25% target stimulation appearing alternately and irregularly, with an interval of 1.5 s and a duration of 60 ms). The frequency of target stimulation is 2000 Hz, while the non-target stimulation is 1000 Hz, repeated 50 times. The intensity of the two tones is 100 dB. The latency and amplitude of the P300 potential are recorded before and after the treatment.

Proton magnetic resonance spectroscopy

Based on nuclear magnetic resonance and chemical shift phenomena, proton magnetic resonance spectroscopy imaging (1H-MRS) is currently the only non-invasive detection method capable of performing quantitative analysis of tissue metabolism in vivo.⁴³ All patients will undergo MRI and ¹H MRS examination on a 3.0 T MR scanner (Discovery LS MR 750; GE Healthcare, Chicago, Illinois, USA) with a 32-channel head coil. The sequences of imaging with a slice thickness of 5 mm include a T1-weighted sequence (repetition time/echo time, TR/TE=2000/20 ms), a T2-weighted sequence (TR/TE=2600/80 ms), a fluid-attenuated inversion recovery sequence (TR/TE=8000/160 ms) and diffusion-weighted imaging. Based on the PROBE-SV 35 (3D BRAVO), single-voxel MRS is performed with a point-resolved spectroscopic sequence with TR/TE=1500/35 ms. Region of interest, a 1.0 cm³ (1.0×1.0×1.0 cm) voxel, is prescribed on the left DLPFC. To ensure accurate examination results, structures such as the skull, fat, air cavities and cerebrospinal fluid are deliberately excluded from the analysis. The metabolite identification and ratios will be automatically completed by the equipment’s own software. In this study, the chemical shifts of the metabolites are as follows: N-acetyl aspartic acid (NAA) 2.0 ppm, choline (Cho) 3.2 ppm, myoinositol (mI) 3.5 ppm, creatine (Cr) 3.0 ppm. Record the relative concentrations of NAA, Cho and mI and calculate the NAA/Cr, Cho/Cr and mI/Cr ratios. Among the main compounds detected, NAA is considered a sensitive indicator of neuronal activity, which is closely related to cognitive functions such as attention, executive function and working memory.⁴⁴ Cho participates in cell membrane and phospholipid metabolism, and mI, which is closely related to the content and activity of glial cells. Following brain tissue damage, Cho and mI expression levels can significantly increase due to the proliferation and repair of glial cells.⁴⁵ The Cr peak is generally stable and frequently serves as a reference for the signal strength of other metabolites. All data collection will be carried out by two experienced radiologists who are blinded to the clinical information.

Sample size

The sample size is calculated using the software G-Power 3.1 (t-test model, two-tailed).⁴⁶ According to a previous study,⁴⁷ we set the value of effect size d = 1.0 based on the MoCA scale. We expect that the target effect size has 80% of the power with a type I error of 5% (α=0.05). Thus, the sample size of each group is at least 17. Allowing for a dropout rate of 10%, we aim to recruit totally 38 subjects who underwent stroke.

Data monitoring and management

The responsibilities of the data monitoring committee (DMC) include safety oversight, test data oversight and recommendations for test design adjustments in the absence of conflicts of interest. The Rehabilitation Department of Tongji Hospital will be responsible for quality assurance of informed consent, recruitment of qualified participants, implementation of interventions and data management. The designated person will be responsible for collecting the case report forms, transferring data and conducting data analysis. The investigator and director are responsible for retaining all records, whereas the anonymised case report form data will be kept by the data centre. To enhance the security and confidentiality of data, electronic information will be stored on computers that need a password, while all paper documents will be kept in locked filing cabinets.

Statistical analysis

Statistical analyses will be performed with the SPSS V.25.0 programme and the level of significance was set at p≤0.05. All analyses will follow the intention-to-treat principle, and missing data will be allocated according to the last observation carried forward and/or mixed data. The primary strategy for handling missing data will be sensitivity analysis and weighted estimating equations. The normal distribution will be checked using the Shapiro-Wilk test. For normally distributed quantitative data, the mean±standard deviation (${\displaystyle \overline{x}\pm s}$) will be used, and the within-group comparisons and between-group comparisons will be, respectively, conducted using paired-sample t-test and independent-sample t-test. For non-normal quantitative data, the median will be used as a measure of central tendency. Within-group comparisons and between-group comparisons will be, respectively, conducted using the Wilcoxon rank-sum test and the Mann-Whitney U test. Categorical variables will be transformed into numeric variables. Besides, the Spearman correlation will be used to measure the association between primary and secondary outcomes. Furthermore, to examine the effect of stroke characteristics (ie, demographics, disease onset, lesion location) on the variable object of our study, we will carry out analyses on raw data while controlling for these variables as covariates. Follow-up data will be further analysed using repeated measures analysis of variance or mixed-effects model. The Bonferroni correction method and false discovery rate <5% will be considered for multiple comparisons.

Ethics and dissemination

Ethical approval has been obtained for this study protocol by the Institutional Ethical Committee of Tongji Hospital (ID: TJ-IRB20230879). All participants or their legally authorised representatives will sign a written informed consent ahead of participation. Patients are allowed to withdraw from the study at any point without providing a reason, and their decision will not impact their care in any way. The DMC will monitor the conduct and safety of the study to ensure the safety of participants and study’s data. Results will be published in peer-reviewed journals and conference presentations.

Ethics statements

Patient consent for publication

Not applicable.