The effect of intermittent theta burst stimulation on central poststroke pain

Abstract

Background:

Central poststroke pain (CPSP) is neuropathic pain caused by a cerebrovascular lesion. Repetitive transcranial magnetic stimulation has been shown to effectively reduce CPSP. Intermittent theta burst stimulation (iTBS), an alternative stimulation method to conventional repetitive transcranial magnetic stimulation, has been shown to have beneficial effects for neuropathic pain. The objective of this study was to investigate the effects of iTBS on CPSP and demonstrate the effects using clinical assessment scales.

Methods:

Patients with CPSP who met the neuropathic pain criteria of the PainDetect questionnaire (PD-Q), Leeds assessment of neuropathic symptoms and signs (LANSS), and Douleur neuropathique 4 questions (DN4) were recruited. Twenty patients were randomly assigned to a real or sham iTBS and underwent 5 sessions of iTBS to the motor cortex. Before the initial stimulation and after the completion of 5 sessions, PD-Q, LANSS, DN4, and numeric rating scale were assessed.

Results:

PD-Q, LANSS, DN4, and numeric rating scale decreased significantly in the real iTBS group but not in the sham iTBS group. There were no adverse effects regarding the iTBS sessions.

Conclusion:

iTBS was effective in reducing CPSP. Considering the safety and convenience of application, iTBS can be suggested as an effective treatment for patients with CPSP.

1. Introduction

Central poststroke pain (CPSP) is neuropathic pain caused by a cerebrovascular lesion such as an infarct or a hemorrhage of the brain or brainstem.^[1,2] The prevalence of CPSP has been reported to range from 1% to 35%.^[3] It can appear during either the acute or chronic phase of stroke.^[4,5] Most patients with CPSP show symptoms after 1 to 3 months.^[2] The quality of life and recovery of stroke patients are affected by CPSP.^[6–8]

Pharmacological therapy has been shown to have limited effects in the treatment of CPSP, and even resistance to this treatment is often encountered; accordingly, neuromodulation therapy has been tried.^[3,7,9] High-frequency repetitive transcranial magnetic stimulation (rTMS) has proven to be a promising treatment tool in meta-analyses.^[3,7,10] Theta burst stimulation (TBS), an alternative stimulation method to conventional rTMS, has also been tried.^[11] TBS has advantages with respect to session duration and neuroplasticity effects.^[12,13] TBS comprises a group of repetitive bursts with intervals, and intermittent TBS (iTBS) is excitatory, whereas continuous TBS is inhibitory.^[14,15] iTBS was reported to be effective for the treatment of neuropathic pain.^[16,17] However, the effects of iTBS on CPSP alone have not yet been studied.

Multiple brain areas and neural circuits are related to the development of CPSP.^[7,8] Noninvasive brain stimulation to the M1 area has proven to be more effective than stimulation to other points such as the dorsolateral prefrontal cortex, anterior cingulate cortex, and posterior superior insula.^[7]

Various scales are used to assess and diagnose CPSP. The main function of these tools is to distinguish neuropathic pain from non-neuropathic pain.^[8] The PainDetect questionnaire (PD-Q), Leeds assessment of neuropathic symptoms and signs (LANSS), Douleur neuropathique 4 questions (DN4), and numeric rating scale (NRS) have been used in various research studies.^[4,8,18]

The objective of this study was to investigate the effects of iTBS on CPSP and demonstrate these effects using clinical assessment scales.

2. Methods

2.1. Trial design and study participants

This study was a randomized, double-blind, controlled trial. It was conducted in compliance with the principles of the Declaration of Helsinki and its contemporary amendments, as well as the ethical standards of the Guidelines for Good Clinical Practice. The protocol was approved by the Institutional Review Board of the hospital (EMC 2020-02-007). This study was registered as a clinical trial with the Clinical Research Information Service (KCT0004976) prior to the recruitment of the 1st participant. A separate analysis of neuropathic pain related to spinal cord injury under this Institutional Review Board protocol was reported elsewhere.^[19]

We recruited patients with neuropathic pain after a 1st-episode unilateral ischemic or hemorrhagic stroke who had been undergoing inpatient treatment at the Department of Rehabilitation Medicine of a university hospital. The diagnosis of neuropathic pain was based on the Neuropathic Pain Special Interest Group guidelines, combining pain and positive sensory signs in the neuroanatomically relevant area with lesion confirmation by brain magnetic resonance imaging or computed tomography.^[20] Patients who met the neuropathic pain criteria of PD-Q, LANSS, and DN4 were designated as having neuropathic pain (PD-Q ≥ 19, LANSS ≥ 12, and DN4 ≥ 4).^[18,21–23] The exclusion criteria were musculoskeletal pain in the area of neuropathic pain, increased intracranial pressure, headache, epileptic seizures, cognitive impairment that might interfere with understanding the questionnaire, metallic foreign body implant, pregnancy, or being under the age of 18. All subjects gave their written informed consent for inclusion before their participation.

The participants were randomly assigned to either the real or sham iTBS group. A permuted block randomization with a block size of 4 was applied.^[16] A randomization sequence of blocks was generated by a computer and placed in sequentially numbered opaque and sealed envelopes. Each participant received assigned stimulation daily for 5 consecutive days. If the participants were on medications or physical treatment for CPSP, the treatment was not changed during the study period. The outcome measures were assessed before the initial stimulation and immediately after the completion of 5 iTBS sessions.

2.2. Interventions

This study used a Magstim Rapid2 System (Magstim) and a double 70 mm air film coil. Motor evoked potential was recorded at the 1st dorsal interosseous muscle contralateral to the stimulated hemisphere using a Nicolet EDX (Natus Medical Inc.) connected to the Magstim Rapid2 System. The motor threshold was identified by single-pulse stimulation of the ipsilesional M1.^[24] The active motor threshold (aMT) was defined as the lowest intensity needed to induce motor evoked potentials of more than 200 μV peak-to-peak amplitude in at least 5 of 10 stimuli performed while contracting the 1st dorsal interosseous muscle isometrically.^[25,26] The aMT was expressed as a percentage of the maximal stimulator intensity.

The real iTBS involved 3 pulses (60 milliseconds) of stimulation delivered at 50 Hz, and a 5 Hz TBS train repeated every 10 seconds for 2 seconds, for a total of 200 seconds (total 600 pulses) at an intensity of 80% aMT.^[14,27] The hand area was specified as the hot spot of the primary motor cortex for participants with CPSP in the upper extremities.^[16] The leg area, designated as the intersection of the motor cortex and interhemispheric line, was stimulated for participants with CPSP in the lower extremities.^[28–30] The sham iTBS was conducted by tilting the coil 90° to the scalp.^[31]

2.3. Outcome measurements

Patients were evaluated using PD-Q, LANSS, DN4, and NRS. The PD-Q includes 3 categories of questions.^[21] The 1st category includes 7 questions regarding the gradation of pain, and each question is scored from 0 to 5. The 2nd category deals with the pain course pattern, while the last category is a question about the radiation of pain. The maximum PD-Q score is 38, while the minimum is−1. A PD-Q score of 19 or more suggests neuropathic pain. A PD-Q score of 12 or lower indicates that neuropathic pain is unlikely to be present. The LANSS pain scale consists of 7 binary items with different score allocations, ranging from 0 to 24.^[22] A LANSS score of 12 or higher indicates the presence of neuropathic pain. The DN4 consists of 10 dichotomous items, of which 7 are associated with pain quality and 3 are related to clinical examination.^[23] The DN4 ranges from 0 to 10, and a DN4 score of 4 or higher implies the existence of neuropathic pain. The NRS is composed of an 11-point scale ranging from 0 to 10.^[28] The evaluation was conducted by one of the investigators, who was blind to the group allocation.

2.4. Sample size

The sample size was calculated based on the standardized effect sizes obtained from previously published data.^[16] Power analysis was performed using the G^*Power software version 3.1 (University of Dusseldorf).^[32] We expected that only a small number of patients would meet the criteria, and a priori analysis using the independent t test indicated that a sample size of 14 patients would provide a statistical power of 80% (β = 0.20 and α = 0.05). The asymptotic relative efficiency of the Mann–Whitney U test versus the independent t test was 0.955; taking into account asymptotic relative efficiency and the 15% dropout rate, a total of 18 patients were needed.^[33]

2.5. Statistical analysis

The statistical analysis was conducted using SPSS version 14.0 (SPSS Inc.). Baseline clinical characteristics and follow-up assessment data for the 2 groups were compared using the Mann–Whitney U test and Fisher exact test. Changes within each group were analyzed using the Wilcoxon signed-rank test. A P value of <.05 was considered statistically significant.

3. Results

Among 149 screened poststroke patients with suspected neuropathic pain, only 20 met the neuropathic pain criteria of PD-Q, LANSS, and DN4. These patients were randomly assigned to the real iTBS group or the sham iTBS group (Fig. 1). The participants completed the study without any adverse effects during or after the stimulation.

Figure 1.

Study flowchart. iTBS = intermittent theta burst stimulation.

The baseline clinical characteristics of the subjects are presented in Table 1. Age, duration after stroke onset, aMT, and initial scores of PD-Q, LANSS, DN4, and NRS were compared using the Mann–Whitney U test. Sex, type of stroke, and side and location of pain were compared using Fisher exact test. There were no statistically significant differences between the 2 groups.

Table 1.

Baseline characteristics of the subjects.

	Real iTBS group (n = 10)	Sham iTBS group (n = 10)	P value*
Age, yr	62.80 ± 15.08	61.90 ± 18.74	.971
Sex	Male 7 female 3	Male 5 female 5	.650
Type of stroke	Ischemic 5 hemorrhagic 5	Ischemic 6 hemorrhagic 4	>.999
Pain side	Right 5 left 5	Right 2 left 8	.350
Pain location	UE 8 LE 2	UE 4 LE 6	.170
Duration, d	46.90 ± 14.81	48.10 ± 14.64	.796
aMT	41.80 ± 9.95	42.20 ± 4.85	.280
PD-Q	22.40 ± 1.84	22.70 ± 1.83	.739
LANSS	14.30 ± 2.21	15.40 ± 1.96	.247
DN4	5.70 ± 0.95	6.10 ± 1.20	.436
NRS	6.20 ± 1.62	6.00 ± 1.89	.631

Values are presented as mean ± standard deviation or number.

aMT = active motor threshold, DN4 = Douleur neuropathique 4 questions, iTBS = intermittent theta burst stimulation, LANSS = Leeds assessment of neuropathic symptoms and signs, LE = lower extremity, NRS = numeric rating scale, PD-Q = PainDetect questionnaire, UE = upper extremity.

^*P values are based on the Mann–Whitney U test or Fisher exact test, as appropriate.

In the real iTBS group, the Wilcoxon signed-rank test found significant differences between preintervention and postintervention values for all evaluation tools, and all scores decreased, indicating pain improvement (PD-Q P = .005, LANSS P = .016, DN4 P = .004, and NRS P = .004). In the sham iTBS group, the Wilcoxon signed-rank test revealed nonsignificant changes for all evaluation tools (PD-Q P = .235, LANSS P = .180, DN4 P = .564, and NRS P = .655). The Mann–Whitney U test found that the 2 groups were statistically different in terms of postintervention scores for all evaluation tools (Table 2).

Table 2.

Comparison of changes in neuropathic pain assessment.

	Real iTBS group (n = 10)			Sham iTBS group (n = 10)			P value
	Pre	Post	Decrease	Pre	Post	Decrease
PD-Q	22.40 ± 1.84	13.50 ± 3.84	8.90 ± 4.25*	22.70 ± 1.83	21.20 ± 3.71	1.50 ± 3.50	<.001**
LANSS	14.30 ± 2.21	10.70 ± 3.37	3.60 ± 3.13*	15.40 ± 1.96	14.80 ± 2.57	0.60 ± 1.35	.003**
DN4	5.70 ± 0.95	4.10 ± 1.10	1.60 ± 0.52*	6.10 ± 1.20	6.20 ± 1.23	−0.10 ± 0.57	.002**
NRS	6.20 ± 1.62	3.60 ± 1.58	2.60 ± 1.51*	6.00 ± 1.89	5.90 ± 2.28	0.10 ± 0.74	.029**

Values are presented as mean ± standard deviation.

DN4 = Douleur neuropathique 4 questions, iTBS = intermittent theta burst stimulation, LANSS = Leeds assessment of neuropathic symptoms and signs, NRS = numeric rating scale, PD-Q = PainDetect questionnaire.

^*P < .05 by the Wilcoxon signed-rank test within each group.

^**P < .05 by the Mann–Whitney U test between 2 groups.

4. Discussion

In this study, the CPSP-reducing effects of iTBS were demonstrated using PD-Q, LANSS, DN4, and NRS.

In this study, CPSP was diagnosed using 3 assessment scales, and the inclusion rate was 13.4%. The prevalence of CPSP has been found to vary widely because clear diagnostic criteria have not been established.^[34,35] In recent studies in which clinical scales were used, the prevalence of CPSP was reported to be between 5% and 12.3%.^[18,36–40] Meanwhile, the pooled prevalence of CPSP in a meta-analysis was 11%.^[41] Our study showed a prevalence rate similar to those established in the aforementioned studies.

In addition to CPSP, stroke patients may experience various types of pain, including musculoskeletal pain, spasticity-associated pain, and headaches.^[42] Transcranial magnetic stimulation is known to be effective for pain other than neuropathic pain.^[43–45] Three assessment scales were therefore used to demonstrate the effects of iTBS on CPSP.^[18]

The initial PD-Q scores of both groups were >19, the cutoff point for the presence of neuropathic pain.^[21] The decrease in PD-Q scores after stimulation was significant in the real iTBS group only, but the postintervention PD-Q scores of this group were still above 12. This means that although the possibility of neuropathic pain being present decreased, the possibility was still not in the range of unlikeliness.

The LANSS, DN4, and NRS scores decreased significantly in the real iTBS group only. The postintervention LANSS scores of the real iTBS group fell below 12, the cutoff point, while their postintervention DN4 scores were still above 4, the cutoff point.^[22,23] The NRS is not specific to neuropathic pain, but it is frequently employed to assess neuropathic pain intensity.^[20,46] The decreases in all of these measures imply that iTBS was effective in reducing neuropathic pain, although the effect was not enough to eliminate it.

Regarding CPSP, various mechanisms were suggested, such as the functional imbalance between the inhibitory and facilitatory systems in the central pain pathways, neuroinflammation, and changes in the spinothalamic tract function.^[4,47] Accordingly, there is no gold standard treatment for CPSP, but pharmacologic treatment and rTMS have proven to be moderately effective.^[2] From recent meta-analyses regarding rTMS for CPSP, high-frequency rTMS applied to the ipsilesional M1 was found to be effective in reducing CPSP.^[3,10] Most studies have used the visual analog scale or NRS as pain outcome indicators. High-frequency rTMS is known to increase excitability. In stroke patients, decreased excitability of the affected M1 leads to interhemispheric inhibition, and this phenomenon causes increased excitability of the contralateral M1; thus, high-frequency rTMS to the ipsilesional M1 increases the excitability of this area and inhibits the augmented excitability of the contralateral M1.^[3,10] Neuropathic pain can be alleviated by this proposed mechanism. From this perspective, iTBS was applied to the ipsilesional M1 as a substitute for high-frequency rTMS because TBS has an advantage over rTMS in terms of lower seizure risk and shorter stimulation session durations.^[13,17,48] Stimulation over the hand area was reported to be ineffective for patients with neuropathic pain in the lower extremities; thus, the leg area located at the intersection of the motor cortex and interhemispheric line was stimulated in this study.^[28–30]

The effects of iTBS on neuropathic pain have been demonstrated in several studies, but the causes of neuropathic pain varied because the number of patients who were ultimately diagnosed as having neuropathic pain was small.^[16,17] The present study demonstrated that iTBS applied to the motor cortex was more effective in reducing CPSP than a sham stimulation. These findings were demonstrated with scales frequently used in CPSP studies.^[4,8,18]

High-frequency rTMS can be an option for patients with CPSP that is resistant to pharmacologic treatment.^[7] iTBS can be another optional treatment for patients with CPSP who are unable to tolerate a long single stimulation time of rTMS. Effects of iTBS have been reported to be longer than those of high-frequency rTMS in some studies.^[14,49] Therefore, iTBS may be an effective and efficient treatment for CPSP.

This study has several limitations. First, the number of participants was small. The strictness of the combined criteria for assessment tools made the inclusion rate of screened patients low. Conversely, it may better demonstrate the effects of iTBS. Further studies with large numbers of patients are needed. Second, the long-term effects of iTBS were not evaluated. The outcome measures were only evaluated before and immediately after the completion of 5 iTBS sessions, and no subsequent evaluations were conducted. Since CPSP is a chronic condition that often lasts a lifetime, additional research that includes follow-up assessments is necessary.^[50,51] Third, aMT was not measured after stimulation completion. Prospective studies with longer study durations and electrophysiologic follow-up are required.

5. Conclusions

In conclusion, iTBS was effective in reducing CPSP. The decreases in the scores of clinical assessment scales frequently used for CPSP studies support this. Considering the safety and convenience of application, iTBS could be suggested as an effective treatment for patients with CPSP.

Author contributions

Conceptualization: Sangpil Son, Sol Lee, JaeIn You, Jong Keun Kim, Jong Youb Lim.

Data curation: Sangpil Son, Sol Lee, JaeIn You, Jong Keun Kim.

Formal analysis: Sangpil Son, Jong Keun Kim, Jong Youb Lim.

Investigation: Sangpil Son, Sol Lee, JaeIn You, Jong Keun Kim, Jong Youb Lim.

Supervision: Jong Youb Lim.

Writing – original draft: Sangpil Son.

Writing – review & editing: Jong Youb Lim.