Brain Stimulation 7 (2014) 864e870

Contents lists available at ScienceDirect

Brain Stimulation journal homepage: www.brainstimjrnl.com

Inter-subject Variability of LTD-like Plasticity in Human Motor Cortex: A Matter of Preceding Motor Activation Mitchell R. Goldsworthy a, *, Florian Müller-Dahlhaus b, c, Michael C. Ridding a,1, Ulf Ziemann b, c,1 a b c

Robinson Research Institute, School of Paediatrics and Reproductive Health, University of Adelaide, Adelaide 5005, Australia Department of Neurology and Stroke, and Hertie Institute for Clinical Brain Research, Eberhard-Karls-University Tübingen, Tübingen D-72076, Germany Department of Neurology, Goethe-University Frankfurt, Frankfurt/Main D-60590, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 June 2014 Received in revised form 7 August 2014 Accepted 7 August 2014 Available online 15 September 2014

Background: Continuous theta burst stimulation (cTBS) of the human primary motor cortex (M1) induces long-term depression (LTD)-like plastic changes in corticospinal excitability, but several studies have reported high inter-subject variability of this effect. Most studies use a tonic voluntary contraction of the target muscle before cTBS to set stimulation intensity; however, it is unclear how this might affect response variability. Objective: To examine the influence of pre-activation of the target hand muscle on inter-subject response variability to cTBS of the human M1. Methods: The response to cTBS was assessed by changes in motor evoked potential (MEP) amplitude in the right first dorsal interosseous (FDI) muscle. For Study 1, ten healthy subjects attended two sessions. They were instructed in one session to keep their FDI relaxed for the entire testing period (pre-relax), and in the other to perform a 2-min 10% of maximal voluntary tonic contraction 15 min before cTBS (preactive). For Study 2, data from our previous study were re-analyzed to extend the pre-relax condition to an additional 26 subjects (total n ¼ 36). Results: cTBS-induced highly consistent LTD-like MEP depression in the pre-relax condition, but not in the pre-active condition. Inter-subject response variability increased in the pre-active condition. Conclusions: cTBS induces consistent LTD-like plasticity with low inter-subject variability if pre-activation of the stimulated motor cortex is avoided. This affirms a translational potential of cTBS in clinical applications that aim at reducing cortical excitability. Ó 2014 Elsevier Inc. All rights reserved.

Keywords: rTMS Theta burst stimulation Transcranial magnetic stimulation Long-term depression Metaplasticity Voluntary contraction

Introduction The strength of synaptic transmission in the human cortex is adaptable, undergoing constant change in an activity-dependent manner. This plasticity is important for learning and memory, as well as recovery from neurological damage, and has been the focus of considerable research. Numerous non-invasive brain stimulation (NIBS) techniques [1] have been developed that are capable of

This work was supported by a grant from the National Health and Medical Research Council (NHMRC; grant 565302). M.R.G. was supported by a German Academic Exchange Service (DAAD; A/11/96158) Research Grant and Australian Postgraduate Award(APA), and is an Alzheimer’s Australia Dementia Research Foundation (AADRF) Postdoctoral Fellow. * Corresponding author. DX 650-517, Robinson Research Institute, School of Paediatrics and Reproductive Health, University of Adelaide, SA 5005, Australia. Tel.: þ61 8 8313 1323. E-mail address: [email protected] (M.R. Goldsworthy). 1 M.C.R. and U.Z. contributed equally to this work. 1935-861X/$ e see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.brs.2014.08.004

inducing short-term plasticity in the human cortex that can interact with and modify behavior [2,3]. This raises the possibility that such techniques may be useful as therapeutic tools for neurological and psychiatric disorders [4]. One NIBS technique that has gained much recent interest is theta burst stimulation (TBS), a patterned repetitive transcranial magnetic stimulation (rTMS) paradigm consisting of repeated bursts of highfrequency, sub-threshold magnetic stimuli. When applied to the human primary motor cortex (M1), TBS induces changes in corticospinal excitability that are either facilitatory when applied as intermittent TBS (iTBS), or inhibitory when applied as continuous TBS (cTBS) [2]. There is evidence that these after-effects are NMDA receptor-dependent [5,6] and occur at the level of the cortex [7,8], and are therefore likely to reflect long-term potentiation (LTP) and long-term depression (LTD)-like plastic changes in M1. The short application times and low stimulation intensity requirements of TBS make it more suitable for patient populations than other currently available options. However, a major limitation

M.R. Goldsworthy et al. / Brain Stimulation 7 (2014) 864e870

865

Figure 1. Experimental design of Study 1. Open rectangles designate blocks of 15 MEP recordings, measured using single-pulse TMS. 15 min prior to cTBS application in the preactive condition, subjects performed a 2-min contraction of their right FDI (black filled rectangle) at 10% of maximal voluntary contraction (MVC). RMT, determination of resting motor threshold.

of TBS is the high inter-subject response variability [9]. There are likely many factors that lead to this variability, one of which being the history of synaptic activity in the targeted cortical region [10]. In animal experiments, prior synaptic activity can interact with subsequent LTP/LTD induction [11,12]. Similarly, in humans, NIBSinduced plasticity in M1 is modulated when the motor regions are voluntarily engaged prior to stimulation, either by phasic [13] or tonic [14,15] contraction of the targeted hand muscle. Despite this, few studies using NIBS have controlled the activation history of the target muscle. Indeed, in many studies a tonic voluntary contraction of variable duration is employed prior to TBS to establish active motor threshold (AMT), which is then subsequently used to set stimulation intensity [2,8,9]. The present investigation examined the influence that preactivation of the target hand muscle has on the inter-subject variability of plastic responses to cTBS of the human M1. We also present a meta-analysis incorporating data collected as part of our previous study [16] to demonstrate, in a relatively large sample, the consistency of subject responses to cTBS in the absence of pre-activation. Material and methods Subjects A total of 36 healthy subjects (33 right-handed and three lefthanded; 15 females) aged 19e49 [24.0  5.6 years (mean  SD)] were included in this study. All subjects were screened for contraindications to TMS [17] and gave informed written consent prior to enrollment. All experiments were performed in accordance with the 2008 Declaration of Helsinki, and were approved by the University of Adelaide Human Research Ethics Committee and the ethics committee of the medical faculty of the Goethe-University of Frankfurt am Main. Experimental design This research consisted of two studies. Study 1 was performed in a subset of 10 subjects (five females; 22.8  4.4 years) to test the impact that pre-activation of the target hand muscle [i.e., the right first dorsal interosseous (FDI) muscle] has on the response to cTBS (Fig. 1). Subjects participated in two sessions separated by at least one week. In one session, subjects were instructed to keep their right FDI relaxed for the entire testing period (i.e., pre-relax condition), whereas in the other session, they performed a 2-min 10% of maximal voluntary tonic contraction (MVC) of their right FDI prior to cTBS (i.e., pre-active condition) (see Voluntary contraction section below). The interval between the contraction and cTBS was 15 min, consistent with the interval used in a recent study demonstrating

large inter-subject response variability to cTBS [9]. The order of sessions was randomized between subjects, with six subjects tested with pre-relax first, and four tested with pre-active first. Study 2 was a meta-analysis combining the pre-relax condition of Study 1 with data collected as part of a previous investigation in 26 subjects [16]. For all 36 subjects, the response to cTBS was tested in the absence of prior motor activity, and was measured using a standardized protocol (see Quantification of cTBS effects section). All experiments were performed in the afternoon to control for possible time-of-day effects on plasticity induction [18]. Stimulation and recording procedures Subjects were seated in a comfortable chair at the beginning of each session, and were asked to relax their right arm and hand. Surface electromyography was recorded from the right FDI using two Ag/AgCl electrodes arranged in a belly-tendon montage and sampled at 5 kHz (Cambridge Electrical Design 1401, Cambridge, UK). Signals were amplified (1000) and band-pass filtered between 20 and 1000 Hz (Cambridge Electrical Design 1902 amplifier, Cambridge, UK) or 2000 Hz (Counterpoint Mk2 electromyograph, Dantec, Denmark), before being stored on a computer for offline analysis. Background surface electromyography was monitored throughout the entire testing period to ensure complete relaxation of subjects’ right FDI. Single-pulse TMS with monophasic waveform was applied to the left M1 to elicit motor evoked potentials (MEPs) in the right FDI. Stimuli were applied using a figure-of-eight magnetic coil (external wing diameter, 90 mm) connected to a Magstim 200 magnetic stimulator (Magstim, Whitland, UK). The coil was positioned tangentially to the skull, with the handle pointing posterolaterally at a 45 angle to the sagittal plane (i.e., posterioreanterior current flow across M1). The hand representations of the left M1 were identified using a marginally supra-threshold stimulus intensity, and the optimal position for activating the right FDI was marked on the subject’s scalp using a felt marker. TMS intensity was adjusted to evoke baseline MEPs with peak-to-peak amplitudes around 1 mV; this intensity (SI1mV) was used for all subsequent MEP recordings. MEPs were recorded in blocks of 15 trials, with an intertrial interval of 7 s (10%). Trials contaminated with background muscle activation during the 100 ms prior to TMS were excluded from analysis (less than 5% of all trials). Continuous theta burst stimulation (cTBS) cTBS was applied with biphasic waveform (posterioreanterior/ anterioreposterior current flow) using either Magstim Super Rapid (Magstim) (Experiments 1 and 2) or MagPro X100 (MagVenture, Farum, Denmark) (Experiment 2 only) magnetic stimulators. The

866

M.R. Goldsworthy et al. / Brain Stimulation 7 (2014) 864e870

same stimulator was used when subjects were required to attend more than one session. We employed the standard cTBS pattern, which consisted of bursts of three stimuli at 50 Hz, repeated at a frequency of 5 Hz for 40 s (600 stimuli in total) [2]. Rather than the conventional 80% of AMT, we used the subjects’ resting motor threshold (RMT) to set cTBS intensity to avoid the requirement for pre-activation of the right FDI. 70% of RMT was used, since this is roughly equivalent to 80% of AMT [14,19]. RMT was assessed to the nearest 1% of maximal stimulator output (MSO) prior to cTBS application using the same biphasic stimulator and coil position, and was defined as the minimum stimulus intensity required to elicit an MEP from the relaxed right FDI muscle with peak-to-peak amplitude greater than 50 mV in at least five of ten consecutive trials. Voluntary contraction The force of the tonic contraction preceding cTBS in the preactivation condition of Experiment 1 was monitored using a strain gage, and visual feedback was displayed on an oscilloscope to ensure constant force production. Subjects performed an index finger abduction (where FDI is the prime mover) with their right hand for 2 min, using a constant force that was 10% of their maximal effort. This was determined at the beginning of the session for each subject by having them perform three MVCs. Quantification of cTBS effects For both studies, three blocks of MEPs were recorded at baseline to provide a stable measure of corticospinal excitability. For Study 1, an additional block was also recorded after the voluntary contraction (pre-active condition) or waiting period (pre-relax condition; see Fig. 1), immediately before cTBS (i.e., pre-cTBS time point), to determine the influence of FDI pre-activation on baseline corticospinal excitability. For quantification of the cTBS effect, MEPs were recorded at 0, 5, 10, 15, 20, 25, and 30 min post-cTBS (Fig. 1). For Study 2, MEPs were recorded using varying numbers of postcTBS recording blocks at varying times (up to 30 min) after cTBS, depending on the experiment from which data were collected. Multiple blocks of post-cTBS MEPs were recorded in 27 subjects. This included 10 subjects with MEPs recorded at 0, 10, 15, 20, 25, and 30 min post-cTBS (i.e., Experiment 2 of Ref. [16]); seven subjects with MEPs recorded 5, 10, 20, 25, and 30 min post-cTBS (i.e., Experiment 4 of Ref. [16]); as well as the 10 subjects from Study 1. An additional nine subjects were included, for whom a single block of MEPs were recorded either 0 min post-cTBS (n ¼ 4; Experiment 5 of Ref. [16]) or 5 min post-cTBS (n ¼ 5; Experiment 3 of Ref. [16]). Note that the numbers of subjects used here are not necessarily the same as those tested for the various experiments of Goldsworthy et al. [16], since several subjects in that study were tested in multiple experiments. For these subjects, data collected from the experiment with the greater number of post-cTBS time points were included for Study 2. Data analysis All statistical analyses were performed on normally distributed data with IBM SPSS Statistics 20 (IBM SPSS, Armonk, NY, USA). Mean peak-to-peak MEP amplitude was calculated for each block of 15 trials for each subject in both studies. Paired t tests were used to compare SI1mV and cTBS intensity between the pre-relax and preactive conditions in Study 1. To assess the stability of baseline MEP amplitudes between conditions and between the three baseline recording blocks, a two-way repeated-measures ANOVA (ANOVARM) with CONDITION (two levels e pre-relax and preactive) and TIME (three levels) as within-subject factors was used. Additionally, the influence of the contraction on baseline

corticospinal excitability was analyzed using two-way ANOVARM with CONDITION (two levels) and TIME (two levels e average baseline, pre-cTBS) as within-subject factors. For comparison of the cTBS effects between conditions, mean peakto-peak MEP amplitudes for all recording blocks were normalized to the mean MEP amplitude recorded at the pre-cTBS time point for each given individual. A two-way ANOVARM was run on these normalized data, with CONDITION (two levels) and TIME (10 levels e three baseline, and 0, 5, 10, 15, 20, 25, and 30 min post-cTBS) as within-subject factors. Separate one-way ANOVARM were performed on raw data for each condition with TIME (eight levels e pre-cTBS, and 0, 5, 10, 15, 20, 25, and 30 min post-cTBS) as the within-subject factor. Conditional on a significant main effect, Bonferroni-corrected post hoc paired t tests were performed to compare post-cTBS MEPamplitudes with pre-cTBS. Pearson correlation coefficient was used to assess the relationship of responses to cTBS (i.e., the grand average of all post-cTBS MEP amplitudes, normalized to pre-cTBS) between conditions, and was also used to test for associations between cTBS response and the subjects’ age, SI1mV, and RMT. The inter-subject response variability to cTBS was characterized for the pre-relax and pre-active conditions of Study 1 by classifying each subject’s change in MEP amplitudes from baseline. Similar to the limits used in previous studies to identify responders to another NIBS protocol [20,21], a subject’s response to cTBS was classified as ‘depression’ if the grand average of all post-cTBS MEP amplitudes (normalized to pre-cTBS) was 0.8, and ‘potentiation’ if 1.2. A response was classified as ‘no change’ if normalized post-cTBS MEPs were within these limits. Variability of responses was assessed by calculating the inter-subject coefficient of variation (CV) at each time point before and following cTBS. CVs were normalized to pre-cTBS, and a paired t test was used for comparison between conditions, with each post-cTBS time point representing a matched pair. In Study 2, the three baseline MEP recording blocks were averaged to give a baseline measure of corticospinal excitability for all 36 subjects. An initial one-way ANOVARM with within-subject factor TIME was performed on raw data in the subset of 27 subjects for whom multiple blocks of post-cTBS MEPs were recorded, using those time points consistent for all subjects (i.e., five levels e average baseline, and 10, 20, 25, and 30 min post-cTBS). Following this, the data for the remaining nine subjects were added, and Bonferroni-corrected post hoc paired t tests were performed to compare post-cTBS MEP amplitudes with the average baseline for all post-cTBS time points (i.e., 0, 5, 10, 15, 20, 25, and 30 min). The inter-subject CV was also calculated for all time points to assess inter-subject variability. Finally, a change in corticospinal excitability from baseline was assessed for each subject using either the grand average of all post-cTBS MEP amplitudes (for those 27 subjects in which multiple post-cTBS blocks were recorded), or MEP amplitudes recorded at a single time point 0 (n ¼ 4) or 5 (n ¼ 5) min post-cTBS. Post-cTBS MEP amplitude was normalized to the average baseline, and each subject’s response to cTBS was classified using the same method as that used in Study 1. All statistical analyses were two-tailed, and unless indicated otherwise, all data represent group mean  standard deviation. Where necessary, HuynheFeldt corrections were used to control for violations in sphericity. Statistical significance was accepted for P < 0.05. Results Study 1 e pre-activation reduces the cTBS response by increasing inter-subject variability Average baseline MEP amplitudes in the pre-relax and preactive conditions were 0.84  0.29 and 0.87  0.20 mV,

M.R. Goldsworthy et al. / Brain Stimulation 7 (2014) 864e870

867

change. Conversely, for the pre-active condition, five of 10 subjects showed no effect of cTBS, with three subjects responding with depression and two with potentiation (Fig. 3C). The increased variability of subject responses to cTBS was also reflected in the inter-subject CV, which was greater following cTBS in the pre-active than in the pre-relax condition (paired t(6) ¼ 3.50, P ¼ 0.013) (Fig. 3D). The variability of subject responses for each condition was not affected by the order of testing sessions (Supplementary Table 1). Study 2 e pre-relaxed cTBS, extended to a larger sample

Figure 2. The influence of target muscle pre-activation on the MEP response to cTBS. MEP amplitudes are expressed as a percentage of pre-cTBS. The gray column indicates a 2-min 10% of maximal voluntary tonic contraction of the right FDI for the pre-active condition. Whereas cTBS in the pre-relax condition (closed circles) induced a significant depression of MEP amplitudes, there was no change from baseline in the pre-active condition (open circles). *P < 0.05, when compared with pre-cTBS. #, significant after Bonferroni correction. Data are shown as group means  SEM.

respectively. Baseline MEPs did not differ between conditions (F1,9 ¼ 0.12, P ¼ 0.74), nor did they differ between baseline recording blocks (F2,18 ¼ 1.26, P ¼ 0.31). Neither SI1mV nor cTBS intensity differed between the pre-relax and pre-active conditions (SI1mV: 49.3  10.2 and 49.9  10.6% MSO, respectively; paired t(9) ¼ 0.90, P ¼ 0.39; and cTBS intensity: 35.9  6.4 and 36.4  6.5% MSO, respectively; paired t(9) ¼ 1.25, P ¼ 0.24). There was no difference between MEP amplitudes recorded at baseline and precTBS (F1,9 ¼ 0.07, P ¼ 0.80), nor was there a difference between conditions (F1,9 ¼ 0.01, P ¼ 0.91) or CONDITION*TIME interaction (F1,9 ¼ 0.65, P ¼ 0.44), suggesting that contraction alone had no effect on MEP amplitudes. A two-way ANOVARM of MEP data normalized to pre-cTBS revealed significant main effects of CONDITION (F1,9 ¼ 11.0, P ¼ 0.009) and TIME (F5.0,44.7 ¼ 3.51, P ¼ 0.001), with a trend for a CONDITION*TIME interaction (F6.7,60.1 ¼ 2.09, P ¼ 0.061). This was due to a reduction in post-cTBS MEP amplitudes compared to pre-cTBS in the pre-relax (F3.7,33.6 ¼ 3.17, P ¼ 0.028), but not in the pre-active (F7,63 ¼ 0.99, P ¼ 0.45) condition (Fig. 2). Bonferronicorrected post hoc comparisons revealed that, relative to pre-cTBS, there was a significant depression of MEP amplitudes recorded at 25 min following cTBS in the pre-relax condition (P ¼ 0.003). Additionally, comparison of grand-averaged post-cTBS MEP amplitudes (normalized to pre-cTBS) revealed a significant difference between conditions, with MEP depression observed for the pre-relax (67.1  21.7% of pre-cTBS) but not pre-active (97.4  27.4% of pre-cTBS) condition (paired t(9) ¼ 3.90, P ¼ 0.004). Correlation analysis indicated no significant relationship of the responses to cTBS between conditions (r ¼ 0.52, P ¼ 0.13) (Supplementary Fig. 1), nor was there a relationship with the age of study participants (pre-relax: r ¼ 0.40, P ¼ 0.26; pre-active: r ¼ 0.28, P ¼ 0.43), SI1mV (pre-relax: r ¼ 0.37, P ¼ 0.30; preactive: r ¼ 0.37, P ¼ 0.29), or RMT (pre-relax: r ¼ 0.46, P ¼ 0.18; pre-active: r ¼ 0.34, P ¼ 0.33). Whereas MEP amplitudes were consistently depressed in most subjects following cTBS in the pre-relax condition (Fig. 3A), more variable responses were observed when the target muscle was activated prior to cTBS (Fig. 3B). Of the 10 subjects tested, seven showed the expected depressive response to cTBS in the pre-relax condition, with the remaining three subjects showing no overall

The one-way ANOVARM in the subset of 27 subjects with multiple post-cTBS time points showed that in the absence of preactivation, there was a highly significant main effect of TIME (F4,104 ¼ 12.3, P < 0.0001). When taking into account data from all 36 subjects, the sample size for each post-cTBS time point ranged from n ¼ 20 at 15 min post-cTBS to n ¼ 27 at 10, 20, 25, and 30 min post-cTBS (Fig. 4A). Bonferroni-corrected post hoc comparisons revealed that, relative to the average baseline, there was a significant depression of MEP amplitudes recorded at all time points following cTBS (P < 0.001 for all) (Fig. 4B). Similar to Study 1, intersubject variability increased only slightly following cTBS (Fig. 4C). Of the 36 subjects tested, 26 (72%) responded to cTBS with MEP depression, while the remaining 10 subjects (28%) showed no overall change and none showed MEP potentiation (Fig. 4D). Discussion The present studies show that pre-activation of the target hand muscle increases the inter-subject response variability to cTBS of the human M1, reducing the group-level LTD-like plasticity effect. We demonstrate in a relatively large sample that, in the absence of pre-activation, a highly consistent LTD-like MEP depression is induced. Influence of voluntary motor activity on cTBS-induced M1 plasticity There is a growing body of evidence demonstrating an interaction between voluntary motor activity and cTBS-induced M1 plasticity. Activation of the target hand muscle during [22] or after [16,22] cTBS can abolish or reverse its effects on MEPs, and a similar reversal may also occur when target muscle activity precedes cTBS [13e15]. Gentner et al. [14] showed that a prior tonic voluntary contraction of the target hand muscle for 5 min at w25% of maximal voluntary effort reversed the MEP facilitation induced by a short, 300 pulse train of cTBS. Using a phasic movement task instead of tonic contraction, Iezzi et al. [13] found that prior activity reversed the polarity of after-effects induced by both iTBS and cTBS, converting iTBS-induced MEP facilitation to depression and cTBSinduced MEP depression to facilitation. The influence of target muscle activity on subsequent cTBSinduced plasticity in Study 1 is largely consistent with these previous findings, and may reflect metaplasticity in M1. Metaplasticity has been described extensively in vitro in animal models [11,12] and in the human cortex using various NIBS techniques [23], with preactivation of synaptic connections causing changes in Ca2þ signaling, modulating the threshold for subsequent LTP/LTD induction. This could be mediated by changes in Ca2þ dynamics through L-type voltage-gated Ca2þ channels, which have been implicated in the polarity-reversal of cTBS-induced effects by prior motor activity [24]. Typically, priming activity that increases postsynaptic neuronal firing will increase the threshold for subsequent LTP induction and lower the threshold for LTD induction to maintain homeostasis of synaptic excitability [25]. Our findings

868

M.R. Goldsworthy et al. / Brain Stimulation 7 (2014) 864e870

Figure 3. Influence of target muscle pre-activation on the inter-subject response variability to cTBS. (A, B) Single subject data showing the MEP response to cTBS (expressed as a percentage of the pre-cTBS MEP amplitude) for the (A) pre-relax and (B) pre-active conditions. (C) The percentage of subjects that responded to cTBS with no change, depression, or potentiation of MEP amplitudes for each condition. (D) The inter-subject coefficient of variation (CV) (expressed as a percentage of the pre-cTBS CV) following cTBS in the pre-relax (closed circles) and pre-active (open circles) conditions. The gray column indicates a 2-min 10% of maximal voluntary tonic contraction of the right FDI for the pre-active condition.

suggest that pre-activation of M1 by voluntary contraction initiates non-homeostatic metaplasticity, increasing the threshold for subsequent LTD induction by cTBS. Although similar to the non-homeostatic metaplasticity described previously [13], the influence of prior target muscle activity in Study 1 was not polarity-reversing. Whereas Iezzi et al. showed the inhibitory response to cTBS was converted to facilitation, we observed no overall group-level change in MEP amplitudes following cTBS primed with voluntary motor activity, with highly variable responses ranging from MEP depression to facilitation in some subjects (see Fig. 3B and C). This could be explained by differences in the type of motor task employed by Iezzi et al. (i.e., phasic, compared with tonic in this study), and the interval between motor activity and cTBS (i.e., w2 min, compared with 15 min in this study). It is possible that a prior activation with higher intensity, longer duration, and/or shorter interval before cTBS may have resulted in more subjects’ response profiles converting to facilitation; however, this was not tested systematically in this study. Also, subjects were required to perform three MVCs at the beginning of the pre-active condition to set the intensity for the 2-min sub-maximal tonic contraction, and as a result, we cannot exclude the possibility that this additional motor activity contributed to the increased response variability to cTBS. Therefore, further studies are required to determine the impact of these variables on subsequent plasticity induction. Another factor that may have influenced the response variability to cTBS preceded by motor activity was the strength of the cTBS response in the absence of prior activity. Although subjects’ responses in the pre-relax condition of Study 1 were more consistent than in the pre-active condition, a degree of variability was observed with some subjects responding with strong MEP depression and others with no change (see Fig. 3A and C). There was a tendency for a linear positive relationship between subjects’

responses to cTBS in the two conditions, with subjects showing a weak response to cTBS in the pre-relax condition being more likely to show reversal to MEP potentiation when cTBS was preceded with prior activity (see Supplementary Fig. 1). However, this association did not reach statistical significance, possibly due to the small sample size used in this study. Consistency of the cTBS response in the absence of target muscle pre-activation An important finding of the present studies was that cTBS induced a strong LTD-like depression of corticospinal excitability, but only in the absence of prior voluntary motor activity. Furthermore, in Study 2 we were able to show in a relatively large sample (n ¼ 36) that, when pre-activation of M1 was avoided, the LTD-like response to cTBS was highly consistent between subjects, with 72% showing a strong depression of MEP amplitudes and 0% showing MEP potentiation. While the criteria used to classify response to cTBS was to some extent arbitrary, similar criteria have been used in previous studies for identifying responders to other plasticity-inducing NIBS protocols [20,21]. Indeed, other studies of this type have tended to use less stringent criteria, with some using any change in cortical excitability from baseline values, however small, as evidence of MEP potentiation/depression [9], and others using a 10% difference from baseline as indicative of clinically meaningful change [26]. Therefore, the 20% change required for a response to be classified as either depression or potentiation in this study is quite conservative. Although the initial report by Huang et al. [2] showed cTBS induced a strong depression of MEP amplitudes lasting up to 1 h in duration, several more recent studies have shown either a modest or no effect of cTBS applied to the human M1 [9,15,27e32]. This lack of consistency between studies is likely a combination of the high

M.R. Goldsworthy et al. / Brain Stimulation 7 (2014) 864e870

869

hand muscle prior to cTBS, and although an interval of 15 min separated motor activity and subsequent cTBS application, it is unclear to what extent this prior activity influenced the cTBS response. Here, we show that even a short, low-intensity activation of the target hand muscle 15 min before cTBS application can increase inter-subject response variability, resulting in a nonsignificant group-level response. Considering the majority of studies investigating the effects of cTBS applied to the human M1 require pre-activation of the target muscle during AMT assessment to set stimulation intensity, our findings may help to explain the lack of consistency in the cTBS response, both between subjects and between studies investigating small samples.

Future directions As mentioned above, one key point that remains to be answered is the effects of varying levels of pre-activation on the plastic response to cTBS. The tonic voluntary contraction employed in Study 1 was highly standardized for force (i.e., 10% of individual maximal voluntary effort), duration (i.e., 2 min), and delay to cTBS (i.e., 15 min), and although typical of the types of contractions used for AMT assessment prior to cTBS application in most studies [2,8,9,28,29,31], these factors often vary between studies and between subjects of the same study. Therefore, additional studies are required to further examine the influence that these variations in contraction force, duration, and delay to cTBS, have on the variability of the cTBS response. Although the present studies have shown that prior motor activity influences the response to cTBS of the human M1, it is unclear whether this may also be the case for other NIBS techniques. In addition to the polarity-reversal of cTBS-induced MEP depression, Iezzi et al. [13] showed iTBS-induced MEP facilitation to be similarly reversed when phasic contractions preceded stimulation. Considering that, as with cTBS, iTBS is typically applied after target muscle pre-activation during AMT assessment, it will be important for future studies to determine the influence of this pre-activity on the variability of the iTBS response. Finally, several recent studies have shown reduced plastic responses to cTBS in patients with various hypokinetic [33,34] and hyperkinetic [35] movement disorders. In light of the results of the present studies, the interaction between abnormal voluntary/ involuntary activity in the stimulated cortex and the subsequent cTBS response may be a contributing factor to this reduced plasticity that needs to be explored further. Figure 4. Response to cTBS without target muscle pre-activation (Study 2). (A) The number of subjects for which data was included for each time point at baseline and following cTBS. (B) The time course of the cTBS response, with MEP amplitudes (shown as group means  SEM) depressed at each time point post-cTBS (#, P < 0.001, Bonferroni-corrected). (C) The inter-subject coefficient of variation (CV) for each time point before and following cTBS. (D) The percentage of subjects (total n ¼ 36) that responded to cTBS with either no change or depression of MEP amplitudes (no subject showed MEP potentiation).

variability between subject responses, together with small sample sizes. In a large sample of 52 subjects, Hamada et al. [9] were unable to show a significant group response to cTBS, with less than half of all subjects showing a change in corticospinal excitability in the expected direction (i.e., MEP depression). The large variability of subject response profiles observed by Hamada et al. was attributed, in part, to inter-individual differences in the interneuronal networks activated by single-pulse TMS, which was determined before cTBS by measuring differences in the latencies of MEPs elicited by TMS with different induced current directions. This procedure required extended periods of voluntary contraction of the target

Conclusions The capacity for NIBS techniques like cTBS to induce consistent plastic changes in human cortical excitability is critical for their application as therapeutic tools in neuropsychiatric disorders. The results of the present studies indicated that pre-activation of the target muscle introduces inter-subject variability to the cTBS response, reducing the LTD-like plastic effect. Conversely, in the absence of prior motor activity, cTBS induced a highly consistent LTD-like response. These findings highlight the importance of controlling baseline activity, and suggest that pre-activation of the stimulated cortex should be minimized when cTBS is used for induction of LTD-like plasticity.

Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.brs.2014.08.004.

870

M.R. Goldsworthy et al. / Brain Stimulation 7 (2014) 864e870

References [1] Ziemann U, Paulus W, Nitsche MA, et al. Consensus: motor cortex plasticity protocols. Brain Stimul 2008;1(3):164e82. [2] Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta burst stimulation of the human motor cortex. Neuron 2005;45(2):201e6. [3] Jung P, Ziemann U. Homeostatic and nonhomeostatic modulation of learning in human motor cortex. J Neurosci 2009;29(17):5597e604. [4] Ridding MC, Rothwell JC. Is there a future for therapeutic use of transcranial magnetic stimulation? Nat Rev Neurosci 2007;8(7):559e67. [5] Huang YZ, Chen RS, Rothwell JC, Wen HY. The after-effect of human theta burst stimulation is NMDA receptor dependent. Clin Neurophysiol 2007;118(5):1028e32. [6] Teo JT, Swayne OB, Rothwell JC. Further evidence for NMDA-dependence of the after-effects of human theta burst stimulation. Clin Neurophysiol 2007;118(7):1649e51. [7] Di Lazzaro V, Pilato F, Dileone M, et al. The physiological basis of the effects of intermittent theta burst stimulation of the human motor cortex. J Physiol 2008;586(16):3871e9. [8] Di Lazzaro V, Pilato F, Saturno E, et al. Theta-burst repetitive transcranial magnetic stimulation suppresses specific excitatory circuits in the human motor cortex. J Physiol 2005;565(Pt 3):945e50. [9] Hamada M, Murase N, Hasan A, Balaratnam M, Rothwell JC. The role of interneuron networks in driving human motor cortical plasticity. Cereb Cortex 2013;23(7):1593e605. [10] Ridding MC, Ziemann U. Determinants of the induction of cortical plasticity by non-invasive brain stimulation in healthy subjects. J Physiol 2010;588(Pt 13):2291e304. [11] Abraham WC. Metaplasticity: tuning synapses and networks for plasticity. Nat Rev Neurosci 2008;9(5):387. [12] Hulme SR, Jones OD, Abraham WC. Emerging roles of metaplasticity in behaviour and disease. Trends Neurosci 2013;36(6):353e62. [13] Iezzi E, Conte A, Suppa A, et al. Phasic voluntary movements reverse the aftereffects of subsequent theta-burst stimulation in humans. J Neurophysiol 2008;100(4):2070e6. [14] Gentner R, Wankerl K, Reinsberger C, Zeller D, Classen J. Depression of human corticospinal excitability induced by magnetic theta-burst stimulation: evidence of rapid polarity-reversing metaplasticity. Cereb Cortex 2008;18(9):2046e53. [15] Goldsworthy MR, Pitcher JB, Ridding MC. The application of spaced theta burst protocols induces long-lasting neuroplastic changes in the human motor cortex. Eur J Neurosci 2012;35(1):125e34. [16] Goldsworthy MR, Müller-Dahlhaus F, Ridding MC, Ziemann U. Resistant against de-depression: LTD-like plasticity in the human motor cortex induced by spaced cTBS. Cereb Cortex 2014. Epub 31.01.14 http://dx.doi.org/10.1093/ cercor/bht353. [17] Rossi S, Hallett M, Rossini PM, Pascual-Leone A. Safety, ethical considerations, and application guidelines for the use of transcranial magnetic stimulation in clinical practice and research. Clin Neurophysiol 2009;120(12):2008e39. [18] Sale MV, Ridding MC, Nordstrom MA. Factors influencing the magnitude and reproducibility of corticomotor excitability changes induced by paired associative stimulation. Exp Brain Res 2007;181(4):615e26.

[19] Chen R, Tam A, Butefisch C, et al. Intracortical inhibition and facilitation in different representations of the human motor cortex. J Neurophysiol 1998;80(6):2870e81. [20] Heidegger T, Krakow K, Ziemann U. Effects of antiepileptic drugs on associative LTP-like plasticity in human motor cortex. Eur J Neurosci 2010;32(7):1215e22. [21] Korchounov A, Ziemann U. Neuromodulatory neurotransmitters influence LTP-like plasticity in human cortex: a pharmaco-TMS study. Neuropsychopharmacology 2011;36(9):1894e902. [22] Huang YZ, Rothwell JC, Edwards MJ, Chen RS. Effect of physiological activity on an NMDA-dependent form of cortical plasticity in human. Cereb Cortex 2008;18(3):563e70. [23] Müller-Dahlhaus F, Ziemann U. Metaplasticity in human cortex. Neuroscientist 2014. Epub 11.03.14, http://dx.doi.org/10.1177/1073858414526645. [24] Wankerl K, Weise D, Gentner R, Rumpf JJ, Classen J. L-type voltage-gated Ca2þ channels: a single molecular switch for long-term potentiation/long-term depression-like plasticity and activity-dependent metaplasticity in humans. J Neurosci 2010;30(18):6197e204. [25] Bienenstock EL, Cooper LN, Munro PW. Theory for the development of neuron selectivity: orientation specificity and binocular interaction in visual cortex. J Neurosci 1982;2(1):32e48. [26] Hinder MR, Goss EL, Fujiyama H, et al. Inter- and intra-individual variability following intermittent theta burst stimulation: implications for rehabilitation and recovery. Brain Stimul 2014;7(3):365e71. [27] Di Lazzaro V, Dileone M, Pilato F, et al. Modulation of motor cortex neuronal networks by rTMS: comparison of local and remote effects of six different protocols of stimulation. J Neurophysiol 2011;105(5):2150e6. [28] McAllister SM, Rothwell JC, Ridding MC. Cortical oscillatory activity and the induction of plasticity in the human motor cortex. Eur J Neurosci 2011;33(10):1916e24. [29] Todd G, Flavel SC, Ridding MC. Priming theta-burst repetitive transcranial magnetic stimulation with low- and high-frequency stimulation. Exp Brain Res 2009;195(2):307e15. [30] Vallence AM, Kurylowicz L, Ridding MC. A comparison of neuroplastic responses to non-invasive brain stimulation protocols and motor learning in healthy adults. Neurosci Lett 2013;549:151e6. [31] Zafar N, Paulus W, Sommer M. Comparative assessment of best conven tional with best theta burst repetitive transcranial magnetic stimulation protocols on human motor cortex excitability. Clin Neurophysiol 2008;119(6):1393e9. [32] Goldsworthy MR, Pitcher JB, Ridding MC. A comparison of two different continuous theta burst stimulation paradigms applied to the human primary motor cortex. Clin Neurophysiol 2012;123(11):2256e63. [33] Eggers C, Fink GR, Nowak DA. Theta burst stimulation over the primary motor cortex does not induce cortical plasticity in Parkinson’s disease. J Neurology 2010;257(10):1669e74. [34] Kishore A, Joseph T, Velayudhan B, Popa T, Meunier S. Early, severe and bilateral loss of LTP and LTD-like plassticity in motor cortex (M1) in de novo Parkinson’s disease. Clin Neurophysiol 2012;123(4):822e8. [35] Suppa A, Marsili L, Di Stasio F, et al. Cortical and brainstem plasticity in Tourette syndrome and obsessive-compulsive disorder. Mov Disord 2014. Epub 03.07.14, http://dx.doi.org/10.1002/mds.25960.