“This is a post-peer-review, pre-copyedit version of an article published in The Cerebellum. The final authenticated version is available online at: http://dx.doi.org/10.1007/s12311-018-0971-0” 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Title: Both 50 and 30 Hz continuous theta burst transcranial magnetic stimulation depresses the cerebellum Authors: Nicholas D.J. Strzalkowski1, Aaron D. Chau1, Liu Shi Gan1, Zelma H.T. Kiss1* Affiliations: 1Department of Clinical Neurosciences, Hotchkiss Brain Institute, University of Calgary, Alberta, Canada *Corresponding author: Dr. Zelma H.T. Kiss Hotchkiss Brain Institute, Department of Clinical Neurosciences, Faculty of Medicine, University of Calgary, Calgary, AB, T2N4N1, Canada Email: zkiss@ucalgary.ca Fax: +1 403 210 9550 Running title Depressing cerebellar activity with cTBS Number of tables: 1 Number of figures: 4 Supplementary figure: 1 1 26 27 Abstract The cerebellum is implicated in the pathophysiology of numerous movement disorders, which 28 makes it an attractive target for noninvasive neurostimulation. Continuous theta burst stimulation 29 (cTBS) can induce long lasting plastic changes in human brain; however, the efficacy of different 30 simulation protocols has not been investigated at the cerebellum. Here we compare a traditional 50 Hz 31 and a modified 30 Hz cTBS protocols at modulating cerebellar activity in healthy subjects. Seventeen 32 healthy adults participated in two testing sessions where they received either 50 Hz (cTBS50) or 30 Hz 33 (cTBS30) cerebellar cTBS. Cerebellar brain inhibition (CBI), a measure of cerebello-thalamocortical 34 pathway strength, and motor evoked potentials (MEP) were measured in the dominant first dorsal 35 interosseous muscle before and after (up to ~40 minutes) cerebellar cTBS. Both cTBS protocols induced 36 cerebellar depression, indicated by significant reductions in CBI (P < 0.001). No differences were found 37 between protocols (cTBS50 and cTBS30) at any time point (P = 0.983). MEP amplitudes were not 38 significantly different following either cTBS protocol (P = 0.130). The findings show cerebellar excitability 39 to be equally depressed by 50 Hz and 30 Hz cTBS in heathy adults and supports future work to explore 40 the efficacy of difference cerebellar cTBS protocols in movement disorder patients where cerebellar 41 depression could provide therapeutic benefits. 42 Keywords 43 Cerebellum; Transcranial magnetic stimulation; Theta burst stimulation; TMS; cTBS 44 45 Introduction 46 Transcranial magnetic stimulation (TMS) is a commonly used research technique and a 47 promising clinical tool to noninvasively stimulate the brain. A TMS coil produces a magnetic field that 48 penetrates the scalp with minimal impedance and can induce electrical currents in underlying, 49 superficial tissue; i.e. cerebral and cerebellar cortices. These induced currents in the brain act to 2 50 depolarize neurons in stimulated regions [1]. TMS not only modulates the locally stimulated neurons but 51 can also modulate distant connected structures allowing network effects to be studied. The effects of 52 TMS on neuronal activity depend on the number and pattern of stimulation pulses delivered to the 53 brain. Single TMS pulses delivered to the primary motor cortex (M1) evoke motor evoked potentials 54 (MEPs) in contralateral muscles; which provides a measure of M1 excitability. Repetitive trains of TMS 55 pulses (rTMS) can induce excitatory and inhibitory neuroplasticity that outlast the stimulation duration 56 [2]. The noninvasive nature and therapeutic potential of rTMS has motivated research and therapeutic 57 applications in a variety of clinical populations including Parkinson’s disease, stroke, major depression, 58 and schizophrenia [3]. 59 The application of rTMS in short bursts of stimuli repeated at theta frequencies (4-7Hz) has 60 emerged as a popular technique due to its relatively short application time (~3 min or less) and lasting 61 neuroplastic aftereffects [3]. Theta burst stimulation (TBS) was first proposed by Huang and colleagues 62 (2005), who described three different TBS protocols that each consisted of three-pulse bursts delivered 63 at an inter-pulse interval of 50 Hz and an inter-burst interval of 5 Hz. Intermittent TBS (iTBS), a 2-s train 64 of TBS repeated every 10 s for 190 s was shown to increase MEP amplitudes. Conversely, continuous TBS 65 (cTBS), a 40-s train of uninterrupted TBS was shown to reduce MEP amplitudes. Intermediate TBS 66 (imTBS), a 5-s train of TBS repeated every 15s for 110s was found to have no effect on neuronal 67 excitability [2]. These original TBS protocols have been used in hundreds of studies [4], however the 68 parameters have remained largely unchanged from the original 50 Hz protocol [2]. An exception to the 69 50 Hz protocol is the limited use of a modified 30 Hz cTBS protocol first shown to induce behavioral 70 effects in the oculomotor system when applied to the frontal eye field region [5] and posterior parietal 71 cortex [6]. 30 Hz cTBS was originally investigated for its capability to deliver higher intensity stimulation 72 compared to traditional 50 Hz cTBS; which is advantageous in populations with high motor thresholds 73 [7]. When applied to the M1, 30 Hz cTBS evokes longer lasting aftereffects (MEP suppression) with less 3 74 inter-individual variability compared to 50 Hz cTBS [8]. These studies indicate that different cTBS 75 parameters can effectively modulate cerebral cortex excitability and supports additional studies to 76 explore the efficacy of different TBS protocols across brain regions. 77 The cerebellum is an attractive target for TBS because of its involvement in motor control [9,10] 78 and role in movement disorders [11]. The cerebellum exerts influence on motor cortex excitability 79 through the disynaptic cerebello-thalamocortical (CTC) pathway [12], and cerebellar TMS is thought to 80 act on Purkinje cells in the cerebellar cortex [13,14]. When active, Purkinje cells inhibit the dentate 81 nucleus, which leads to a disfacilitation of thalamocortical drive. Single TMS pulses to the lateral 82 cerebellum can reduce the size of subsequent M1 evoked MEPs, and this technique is called cerebellar 83 brain inhibition (CBI) [15]. 50 Hz cTBS applied to the cerebellum reduces CBI [9] and inhibits galvanic 84 vestibular reflexes [16]. TMS induced cerebellar inhibition has previously been shown to reduce 85 levodopa-induced dyskinesia in Parkinson’s disease (rTMS) [17], and improve cervical dystonia (cTBS) 86 [18]. Modulating cerebellum activity through TMS has clear research and clinical applications; however, 87 the effectiveness of different rTMS and TBS protocols in this location is unknown. 88 Despite the clear anatomical and functional differences between M1 and cerebellum, most 89 studies employ the same TMS protocols on both brain regions [3,4]. Here we built upon a previous study 90 where a 30 Hz cTBS protocol applied to the motor cortex was more effective at depressing M1 91 excitability than a 50 Hz protocol [8] and compared the efficacy of traditional 50 Hz and a modified 30 92 Hz cTBS protocol in modulating cerebellar activity. Maximizing the magnitude and duration of cerebellar 93 depression would increase the usefulness of TBS as both a clinical and research tool. 94 95 96 Methods Subjects 4 97 Seventeen subjects (9 male, 8 female, mean age 24, range 19-36) with no history of neurological 98 disorders participated in the study. Subjects were screened for eligibility (Rossi et al. 2011) and gave 99 written informed consent prior to data collection. The experimental procedures were approved by the 100 University of Calgary Research Ethics Board and complied with the Declaration of Helsinki. 101 Experimental overview 102 All subjects participated in two testing sessions, separated on average by 8 days (range 1-30). 103 Subjects were seated for all procedures, with their arms resting on a pillow across their lap. In the first 104 session, subjects were randomly assigned to receive either 50 Hz cTBS (cTBS50) or 30 Hz cTBS (cTBS30) to 105 the lateral cerebellum ipsilateral to their dominant hand. The alternative cTBS protocol was tested in the 106 second session. Cerebellar brain inhibition (CBI) and MEP amplitudes were recorded from the first dorsal 107 interosseous (FDI) of the dominant hand before (Pre) and after (Post1: 4-20mins, Post2: 25-40mins) 108 cerebellar cTBS. Neuronavigated TMS (Brainsight2. Rogue Research Inc, Montreal, Canada) was used to 109 ensure consistent coil orientation over the FDI motor cortex hotspot within and between experimental 110 sessions. Subjects wore ear plugs during CBI testing. MEPs were recorded in the FDI using surface 111 electromyography (EMG), with one electrode placed over the muscle belly and another over the 112 metacarpophalangeal joint of the index finger (1 x 1 cm2, Kendall H69P electrodes, Covidien, MA, USA). 113 Resting (RMT) and active (AMT) motor thresholds were first obtained using the built-in EMG system of 114 Brainsight2 (digitized at 3KHz, gain of 2500). A Bortec AMT-8 EMG system (gain of 1000, band pass 115 filtered between 10 and 1000Hz; Bortec Biomedical Ltd, Calgary, AB, Canada) with Clampex software 116 (digitized at 10kHz ; Clampex 10.6, Molecular Devices, San Jose, CA, USA) was then used to collect CBI 117 and MEP data pre and post cerebellar cTBS. 118 Cerebellar cTBS 119 A 70 mm double Airfilm coil and Super Rapid2 Plus 1 TMS stimulator (The Magstim Company Ltd., 120 Whiteland, UK) was used to deliver cTBS over the lateral cerebellum ipsilateral to the dominant hand, 3 5 121 cm lateral to the inion on the line joining the inion and the external auditory meatus. The coil was 122 positioned tangentially to the head, with the handle pointing upwards [9]. In both cTBS protocols, 600 123 pulses were delivered at 80% active motor threshold (AMT). AMT was determined in each test session 124 as the minimum stimulation intensity that could evoke an MEP ≥200µV in 5 of 10 trials as the subjects 125 held a 10% maximum contraction of their FDI [2]. cTBS50 involved three-pulse bursts at 50Hz with bursts 126 repeated at 5Hz. In contrast, cTBS30 involved three-pulse burst at 30Hz with burst repeated at 6Hz [8] 127 (Fig. 1). 128 CBI protocol 129 CBI was conducted with two TMS coils in a paired-pulse paradigm. Conditioning stimuli (CS) 130 were delivered to the lateral cerebellum with a 110 mm double-cone coil and Magstim BiStim2 131 stimulator (The Magstim Company Ltd., Whitland, UK); the most efficient coil design at eliciting CBI [19]. 132 The double-cone coil was centered over the lateral cerebellum ipsilateral to the dominant hand, 3 cm 133 lateral to the inion on the line joining the inion and the external auditory meatus [15]. A figure-of-eight 134 coil (D702, MagStim BiStim2 stimulator, Magstim Company, UK) was used to deliver test stimuli (TS) to 135 the FDI M1 motor hotspot. The TS coil was placed tangential to the scalp, orientated to have the lowest 136 FDI MEP thresholds; with the handle oriented posterior and approximately 45˚ lateral from the midline 137 (Fig. 2A). FDI MEPs were recorded in response to the TS alone, and with the CS delivered before the TS 138 at three inter-stimulus intervals (ISI: 3, 5, 7 ms). CBI is expressed as the ratio of conditioned (3, 5, 7 ms 139 ISI) to unconditioned (TS alone) MEP peak-to-peak amplitudes (Fig. 2B). Under normal conditions (i.e. 140 pre cTBS), the cerebellar CS acts to inhibit the M1, and conditioned MEPs are expected to be smaller 141 compared to unconditioned MEPs (smaller ratio of conditioned to unconditioned MEP amplitudes) [20] 142 CS intensity was set to 100% of 50µV RMT; and three TS intensities were tested: 90%, 100%, and 110% 143 of 500µV RMT. Both 50µV and 500µV RMTs (RMT50 and RMT500) were measured at the FDI hotspot using 144 the figure-of-eight coil. At each TS intensity-ISI combination (including TS-alone), 10 MEPs were 6 145 recorded Pre, and 20 MEPs were recorded Post cerebellar cTBS. MEPs were recorded in blocks of eight 146 trials at a single TS intensity. Within a block, two trials at each ISI (including TS-alone) were randomly 147 delivered every 6 seconds. Blocks were randomly delivered in sets of three (one of each TS intensity), 148 and five sets were delivered Pre, Post1, and Post2 (Fig. 1). This method allows for CBI and MEPs at each 149 TS intensity to be investigated at similar time points post cTBS. 150 Data analysis 151 CBI and MEP data were analyzed using a custom MATLAB script (MATLAB R2017, The 152 MathWorks, Inc., Natick, United States). The effect of cTBS protocol on cerebellar function was 153 quantified by comparing CBI induced changes in MEP amplitudes pre and post cTBS. Trials that did not 154 evoke an MEP were excluded from MEP averages (8 trials, only occurred using TS of 90% in 3 subjects). 155 Data were excluded from analysis if >2 trials were removed within the same TS intensity-ISI combination 156 either Pre, Post1, or Post2. The remaining 8 to 10 MEPs within the Pre, Post1, and Post2 sets were 157 averaged at each TS intensity-ISI combination (Fig. 1). To investigate the time course of cTBS 158 aftereffects, block averages (average of 2 MEPs within each block at each ISI) were calculated post cTBS. 159 The TS intensity-ISI combination (e.g. TS 90% and ISI 5 ms) that evoked the most CBI Pre was determined 160 for each subject. This optimum TS intensity-ISI combination was used to compile Pre and Post data for 161 each subject. In addition to measurements of CBI, the peak-to-peak MEP amplitudes at each TS intensity 162 were averaged for the TS-alone trials for each subject. MEP amplitude was compared pre and post cTBS. 163 Statistical analysis was performed using SPSS (IBM SPSS Statistics, Version 24.0. Armonk, NY, 164 USA). A linear mixed model analysis was used to compare CBI data across time (Pre, Post1, Post2) and 165 cTBS condition (cTBS50, cTBS30). The best-fitting covariance structure for the residuals was an 166 autoregressive structure. Diagonal, scaled identity, compound symmetry, and unstructured covariance 167 structures were also tested but none showed an improved fit of the model. Significant main effects were 168 followed up with pairwise comparisons with Bonferroni adjustments for multiple comparisons. A three7 169 way repeated measures analysis of variance (ANOVA) with factors time (Pre, Post1, Post2), TS intensity 170 (90%, 100%, 110%), and cTBS protocol (cTBS50, cTBS30) was conducted on the unconditioned MEP data. A 171 Greenhouse-Geisser correction was applied to comparison of intensity in those cases which violated 172 Mauchley’s test of sphericity. Figures were made using Prism5 (GraphPad Prism version 5.0c for Mac OS 173 X, San Diego CA). 174 175 176 Results The TMS intensities used for the TS, CS, and cTBS (500 and 50 µV RMT, and 200 µV AMT) were 177 similar (within 1% of stimulator output) between days (cTBS protocols) (Table 1). One subject was 178 excluded from analysis because they had Pre CBI values >0.9 on both days, indicating that they were a 179 non-responder to CBI. Four other subjects had Pre CBI >0.9 on one of the testing days (50 Hz or 30 Hz) 180 and their data for the corresponding stimulation protocol were excluded on these occasions. With non- 181 responders removed, data from 14 subjects were analyzed for each cTBS protocol. TS intensity set to 182 90% RMT500 was found to be the best at evoking CBI in 9 of 14 subject Pre cTBS50 and in 6 of 14 subjects 183 Pre cTBS30; making it slightly better than 100% or 110% RMT500 at evoking CBI. Both 5ms and 7ms ISI 184 durations were equally effective in evoking CBI. 185 Cerebellum excitability was reduced by both cTBS50 and cTBS30; however, a statistical difference 186 between protocols was not observed. The linear mixed model analysis with repeated factors of: time 187 (Pre, Post1, Post2) and cTBS protocol (cTBS50, cTBS30), showed a significant main effect of CBI for time 188 (F2, 38.597 = 35.903, P < 0.001), but not for cTBS protocol (cTBS50, cTBS30) (F1, 36.120 = 0.906, P = 0.348) or an 189 interaction effect (F2, 38.597 = 0.525, P = 0.596) (Fig.3). Pairwise comparisons across time points with 190 Bonferroni adjustment found CBI Post1 (P < 0.001) and Post2 (P = 0.002) to be significantly reduced 191 (larger conditioned/unconditioned MEP ratio) compared to Pre. Post1 was not found to be significantly 192 different from Post2 (P = 0.060). 8 193 To gain a more detailed look at how CBI measures change following cTBS over time, a linear 194 mixed model analysis was used to investigate main and interaction effects of time and cTBS protocol on 195 CBI with the Post cTBS data separated by block of time (1-10, ~4 min each). Similar to the grouped Post1 196 and Post2 analysis, a significant main effect for time (F10, 49.227 = 9.962, P < 0.001) was obtained, but not 197 for cTBS protocol (F1, 89.593 < 0.001, P = 0.983) nor for an interaction effect (F10, 49.070 = 1.080, P = 0.3.95) 198 (Fig.4). Pairwise comparisons across time points with a Bonferroni adjustment found CBI Pre to be 199 significantly greater than Post cTBS block 1 (P < 0.001), block 2 (P = 0.016), and block 9 (P = 0.007). 200 Unconditioned MEP amplitudes did not change between time (Pre, Post1, Post2) (F2, 13 = 2.397, P 201 = 0.130) or cTBS conditions (cTBS50, cTBS30) (F1, 14 = 0.285, P =0.602); however MEP amplitude increased 202 with TS intensity (F2, 13 = 11.979, P =0.001). Pairwise comparisons across TS intensity with a Bonferroni 203 adjustment found MEP amplitudes at 110% to be significantly larger than 100% (P = 0.022) and 90% (P = 204 0.001), and MEP amplitudes at 100% to be significantly larger than at 90% (P = 0.001). 205 206 207 Discussion In summary, both 50 Hz and 30 Hz cTBS can depress cerebellar activity, as evidenced by reduced 208 CBI. Both protocols were equally effective at reducing CBI, a finding that is in contrast to the results of 209 Goldsworthy et al. 2012 who found a 30 Hz cTBS protocol to evoke greater, less variable and longer 210 lasting depression than a 50 Hz protocol when applied to M1. Therefore, the cerebellum responds 211 differently than M1 to TMS stimulation and further research is needed to investigate if a more effective 212 cTBS protocol exists for cerebellum. The 50 Hz and 30 Hz cTBS protocols were selected because these 213 are the two most common patterns currently tested in humans. This is the first study to compare cTBS 214 protocols at the cerebellum and indicates that additional studies are needed to explore cTBS 215 mechanisms and additional stimulation patterns. 216 9 217 Comparison between cTBS50 and cTBS30 218 Theta burst stimulation originated from in vitro animal studies that demonstrated high- 219 frequency theta burst electrical stimulations induced long term potentiation (LTP) and long term 220 depression (LTD) [21]. This stimulation pattern was adopted in the development of TBS at the M1, 221 whose influence on neuronal excitability resembles LTP and LTD [2,22]. N-methyl-D-aspartate (NMDA) 222 receptors and calcium (Ca2+) channels are the most likely candidates for mediating TBS aftereffects 223 [3,22]. An NMDA receptor antagonist, memantine, has been shown to block the inhibitory effects of 224 cTBS and the facilitatory effects of iTBS at the motor cortex [23]. The cTBS stimulation pattern is 225 speculated to increase intracellular Ca2+ concentrations, which then evokes a cascade of inhibitory 226 factors that reduce the number and responsiveness of postsynaptic receptors [22]. Similar studies 227 investigating the mechanisms of TBS at the cerebellum have not been reported; however high- 228 frequency electrical stimulation in rat cerebellum slices have shown LTD between Purkinje cell and the 229 deep cerebellar nuclei [24]. Future studies are needed to fully understand TBS mechanisms at the 230 cerebellum but similar LTD-like responses reported at the M1 likely contribute to the observed effects at 231 the cerebellum. 232 In the present study, cerebellar cTBS inhibited the CTC pathway for approximately 30 minutes; 233 however significant difference between the cTBS protocols tested were not observed. This is in contrast 234 to a previous report which found 30 Hz cTBS to evoke longer lasting aftereffects (MEP suppression) with 235 less inter-individual variability compared to the 50Hz cTBS at the M1 [8]. Following M1 50 Hz cTBS, MEP 236 suppression occurred for up to 10 minutes, while MEP suppression persisted to 30 minutes following 30 237 Hz cTBS. The underlying neurological mechanisms for why 30 Hz cTBS may be more effective than 50 Hz 238 cTBS at the M1, and why in the present study cerebellar cTBS appears frequency independent are not 239 clear. 10 240 The cTBS50 and cTBS30 protocols differ in two aspects, the inter-pulse frequency (50 vs 30 Hz) 241 and the inter-burst frequency (5 vs 6 Hz). These differences are noteworthy because the temporal 242 pattern of stimulation is known to influence the aftereffects [2]. Assuming LTP and LTD like mechanisms 243 are responsible for the cTBS aftereffects, concentration and rate of change of intracellular Ca2+ play a 244 major role to the extent of these aftereffects [22]. It is possible that the higher inter-burst frequency of 245 the 30Hz cTBS protocol has a more pronounced effect on LTD Ca2+ mechanisms than the 50Hz cTBS 246 protocol; which can help explain the greater depressive effects of the 30 Hz protocol seen at the M1 [8]. 247 Higher inter-burst and/or inter-pulse intervals may be needed to see significant differences at the 248 cerebellum. 249 250 Cerebellar Brain Inhibition 251 CBI is commonly used to measure the magnitude of inhibition that the cerebellum exerts over 252 the M1 [14]. The cerebellum conditioning stimulus is thought to activate inhibitory Purkinje cells in the 253 cerebellar cortex, which suppresses the excitatory output of the deep cerebellar nuclei and thalamus 254 (Molnar et al. 2004). The resulting decrease in excitatory thalamocortical drive, suppresses M1 255 excitability evidenced by reduced MEP amplitudes. Larger reductions in conditioned MEPs suggests 256 greater activity in the CTC pathway. CBI is reduced in Parkinson’s disease [25,26] and dystonia patients 257 [27] compared to healthy controls. Improving CTC pathway integrity may have therapeutic benefits for 258 movement disorder patients, and CBI measurements may prove a valuable assessment tool. 259 In the present study, CBI was used as the dependent variable, providing a measure of the 260 excitability of the CTC pathway. For this purpose, we implemented different TS intensities and ISIs to 261 maximize CBI for each participant on each day. We found average Pre cTBS CBI to be 0.73, which is 262 similar to previous studies (~0.65 Popa et al. 2010; ~0.80, Koch et al 2008; ~0.70, Carrillo et al. 2013). 263 CBI in the present study was found to be subtle and no combination of single TS intensity-ISI 11 264 demonstrated convincing CBI in this subject group. Because our primary aim was to compare cTBS 265 protocols on CBI, subject specific TS Intensity-ISI combinations were employed in the present study to 266 assist in the detection of cTBS (50 Hz vs 30 Hz) effects on CTC pathway integrity. This approach may have 267 exaggerated the inhibitory effect compared to previous studies that employed the same CBI protocol 268 across subjects. It is unclear why greater CBI was not found at a single TS intensity-ISI when averaged 269 across subjects as reported by others [9]; however we suspect that if larger CS amplitudes were applied 270 [28] we would have evoked great CBI. 271 272 No change in MEP amplitude 273 There are conflicting reports regarding the influence of cerebellar TBS on MEP amplitude. 274 Cerebellar cTBS is thought to depress the membrane excitability of Purkinje cells, and therefore increase 275 the output of the dentate nucleus which then leads to cortical excitation [13]. In agreement with this 276 model, MEP amplitudes have been shown to increase following inhibitory cerebellar 1 Hz rTMS [29]. It is 277 therefore surprising that unconditioned MEP amplitudes at 90%, 100%, and 110% RMT500 were not 278 significantly modified by cTBS in the present or in a previous study [9]. Furthermore, additional studies 279 have reported MEP amplitudes to decrease following cerebellar cTBS [13,26,30,31]. cTBS is delivered at 280 a relatively low stimulation intensity (80% AMT), which may help explain the heterogeneous M1 281 excitability aftereffects reported in the literature. cTBS likely only induces changes in Purkinje cells with 282 the lowest thresholds and their downstream M1 neurons [13]. In the present study, the absence of a 283 change in unconditioned MEP amplitude suggests that the excitability of these downstream M1 neurons 284 were not meaningfully modulated by cTBS. In contrast, the relatively high amplitude CS delivered during 285 CBI leads to a measurable decrease in M1 excitability due to the evoked Purkinje cell volley; which is 286 reduced following cerebellar cTBS. Cerebellar TMS, be it cTBS or CBI will influence a subset of 287 associated M1 neurons. The observation in the present study that TS at 90% RMT500 produced the best 12 288 CBI in the majority of subjects suggests that lower threshold neurons in the M1 are predominantly 289 influenced by cerebellum TMS. It is clear that the mechanisms of cerebellar cTBS are poorly understood, 290 and multiple pathways may be contributing to the aftereffects [13]. Future work is needed to better 291 understand how TMS intensity dictates the modulation of neuronal excitability. 292 293 Clinical significance 294 We are only in the infancy of TBS research; however there is already interest and demand for 295 clinical TBS applications [3]. The clinical goal of TMS as a therapeutic intervention is to improve brain 296 functions by modulating neuronal excitability. TMS interventions should therefore be quick and 297 noninvasive; requirements met by TBS. 298 The cerebellum is an attractive target for TBS investigations because of its accessibility and 299 involvement in many movement disorders. The cerebellum is thought to be hyperactive in Parkinson’s 300 disease [11,32] and dystonia [18], and normalizing CTC pathway activity may improve clinical symptoms 301 [17,25,33]. Parkinson’s patients have deficient CTC inhibitory interactions that are not restored by 302 standard dopaminergic medication [26]. Cerebellar cTBS and rTMS have been shown to have some 303 clinical benefits, including reductions in levodopa induced dyskinesia [17] and resting tremor [34] in 304 Parkinson’s disease. Previous studies have only investigated the efficacy of rTMS or 50Hz cTBS, and the 305 efficacy of alternative parameters is unknown. The present data indicate that 30 Hz cerebellar cTBS can 306 also effectively modulate activity in healthy adults. Future studies are needed to explore the efficacy of 307 30 Hz and additional cTBS protocols in clinical populations at the cerebellum and other brain sites. 308 309 Limitations and recommendations 310 TMS is relatively quick and simple to deliver, however responses are often variable between 311 people and factors contributing to response amplitude and duration are poorly understood [35,36]. At 13 312 M1, intrinsic factors in the recruitment of TMS indirect waves (I-waves) can partly explain inter- 313 individual variability in TBS aftereffects [37]. It is unclear if a similar measurement (direct-indirect wave 314 latency) can be used to predict cerebellar TBS responders and non-responders. A test to pre-screen 315 individual susceptibility to TMS induced neuroplasticity would have great research and clinical value. The 316 duration of cTBS effects and when to appropriately retest subjects is an outstanding question. Two 317 weeks of bilateral cTBS was found to have a modest clinical improvement in cervical dystonia patients 318 when measured ~2 days but not two weeks post cTBS [18]. This suggest a cumulative influence of 319 repeated cTBS with effects that can last days. The research and therapeutic value of cTBS would be 320 increased by longer after-effects; however cTBS depression is only reported out to 30-to-60 minutes in 321 motor cortex [2,9] and cerebellar [26] after single applications; which is in agreement with the present 322 findings. 323 CBI is an indirect measure of cerebellum excitability and may not fully capture the cTBS 324 aftereffects on cerebellum or M1 excitability. In the present study we optimized CBI TS intensity and ISI 325 to account for inter-individual variability and found no differences in between cTBS50 and cTBS30 at 326 decreasing CBI. A more sensitive measure of cerebellum activity (i.e. fMRI) may have been able to detect 327 differences between cTBS protocols; however, the functional significance of a subtle difference, if 328 present, would be questionable. A consequence of optimizing the CBI parameters in the present study is 329 a reduced number of MEPs per block. The data presented in Figure 4 is an average of two MEPs per 4- 330 minute window and may only provide a rough approximation of CBI changes over these time windows. 331 One subject in the present study did not show CBI with any TS-ISI combination at any testing set, 332 and CBI was absent in four other subjects on one of the two testing days. It is unclear what factors 333 contribute to a subject’s susceptibility to demonstrating CBI. In some cases, the CS was likely insufficient 334 to evoke a measurable inhibitory volley from the cerebellar cortex. A recent study found a CS of 60% 335 maximum stimulator output evoked reliable CBI [28]. Whereas in our study, CS was normalized to 100% 14 336 of a 50µV RMT that resulted in CS amplitudes ranging from 38 to 54% of maximum stimulator output. It 337 is likely that with higher CS amplitudes, larger CBI would have been observed across subjects and TS-ISI 338 combinations. 339 In the present study a TS intensity at 90% RMT500 generally showed the best CBI pre cTBS. We 340 suggest that this may be due to the low threshold afferent pathways activated by the cerebellum CS. 341 Using even lower TS intensities, and higher CS intensity may improve the magnitude, and usefulness of 342 CBI as a measure of CTC inhibitory efficacy. Similarly, a higher cTBS intensity would likely evoke greater 343 cerebellar inhibition. The 80% AMT intensity used in the present study was well tolerated by all subjects 344 and was based on previous work at the M1 [8]. However, the cerebellar cortex is deeper than the M1 345 and may be better targeted with higher intensity stimulation. 346 347 Conclusions 348 Both 50 Hz and 30 Hz cTBS can equally suppress CTC pathway activity. Suppressive effects were 349 most pronounced in the first 15 minutes, but reduced cerebellum activity may persist up to 30 minutes. 350 These findings support further investigations to explore how additional changes in cTBS stimulation 351 parameters (inter-pulse, and inter-burst intervals) impact neuroplasticity induction in the cerebellum, in 352 healthy and diseased populations. Optimizing TBS protocols at the cerebellum is a critical step prior to 353 the development of clinical applications. 354 355 Acknowledgment 356 The authors would like to thank Dr. Oury Monchi for use of his laboratory space, and Rachel 357 Sondergaard for assistance with data collection. 358 359 Funding 15 360 This work was supported in part by Natural Sciences and Engineering Research Council of Canada grant # 361 2017-04126 (ZHTK), an Eyes High University of Calgary fellowship, Parkinson Alberta fellowship, and 362 Parkinson’s Foundation Fellowship Grant No. PF-FBS-1776 (NDJS), Branch Out Neurological Foundation 363 (ADC), as well as the Hotchkiss Brain Institute N3 Network. 364 365 Conflict of interest 366 The authors declare that there are no conflicts of interest, financial or otherwise. 367 368 Author contributions 369 N.D.J.S, L.G and Z.H.T.K conceived and designed the research; N.D.J.S, A.D.C and L.G performed the 370 experiments; N.D.J.S and A.D.C analyzed data and prepared figures; N.D.J.S drafted manuscript; N.D.J.S, 371 A.D.C, L.G and Z.H.T.K edited and revised manuscript; N.D.J.S, A.D.C, L.G and Z.H.T.K approved final 372 version of manuscript. 16 373 References 374 375 1. Chung SW, Hill AT, Rogasch NC, Hoy KE, Fitzgerald PB. Use of theta-burst stimulation in changing excitability of motor cortex: A systematic review and meta-analysis. Elsevier Ltd; 2016;:1–22. 376 377 2. Huang Y-Z, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta Burst Stimulation of the Human Motor Cortex. Neuron. 2005;45:201–6. 378 379 3. Suppa A, Huang YZ, Funke K, Ridding MC, Cheeran B, Di Lazzaro V, et al. Ten Years of Theta Burst Stimulation in Humans: Established Knowledge, Unknowns and Prospects. Elsevier Inc; 2016;:1–13. 380 381 4. Wischnewski M, Schutter DJLG. Efficacy and Time Course of Theta Burst Stimulation in Healthy Humans. Brain Stimulation. Elsevier Inc; 2015;8:685–92. 382 383 384 5. Nyffeler T, Wurtz P, Lüscher H-R, Hess CW, Senn W, Pflugshaupt T, et al. Repetitive TMS over the human oculomotor cortex: Comparison of 1-Hz and theta burst stimulation. Neurosci. Lett. 2006;409:57–60. 385 386 387 6. Nyffeler T, Cazzoli D, Wurtz P, Lüthi M, Wartburg von R, Chaves S, et al. Neglect-like visual exploration behaviour after theta burst transcranial magnetic stimulation of the right posterior parietal cortex. Eur J Neurosci. 2008;27:1809–13. 388 389 390 7. Wu SW, Shahana N, Huddleston DA, Gilbert DL. Effects of 30Hz Theta Burst Transcranial Magnetic Stimulation on the primary motor cortex. Journal of Neuroscience Methods. Elsevier B.V; 2012;208:161– 4. 391 392 393 8. Goldsworthy MR, Pitcher JB, Ridding MC. A comparison of two different continuous theta burst stimulation paradigms applied to the human primary motor cortex. Clinical Neurophysiology. International Federation of Clinical Neurophysiology; 2012;123:2256–63. 394 395 9. Popa T, Russo M, Meunier S. Long-lasting inhibition of cerebellar output. Brain Stimulation. Elsevier Inc; 2010;3:161–9. 396 397 398 10. Manto M, Bower JM, Conforto AB, Delgado-García JM, da Guarda SNF, Gerwig M, et al. Consensus Paper: Roles of the Cerebellum in Motor Control—The Diversity of Ideas on Cerebellar Involvement in Movement. Cerebellum. 2nd ed. 2011;11:457–87. 399 11. Wu T, Hallett M. The cerebellum in Parkinson’s disease. Brain. 2013;136:696–709. 400 401 12. Bostan AC, Dum RP, Strick PL. Cerebellar networks with the cerebral cortex and basal ganglia. Trends in Cognitive Sciences. Elsevier Ltd; 2013;17:241–54. 402 403 404 13. Koch G, Mori F, Marconi B, Codecà C, Pecchioli C, Salerno S, et al. Changes in intracortical circuits of the human motor cortex following theta burst stimulation of the lateral cerebellum. Clinical Neurophysiology. 2008;119:2559–69. 405 406 14. Celnik P. Understanding and Modulating Motor Learning with Cerebellar Stimulation. Cerebellum. 2014;14:171–4. 17 407 408 15. Ugawa Y, Uesaka Y, Terao Y, Hanajima R, Kanazawa I. Magnetic stimulation over the cerebellum in humans. Ann. Neurol. Wiley Subscription Services, Inc., A Wiley Company; 1995;37:703–13. 409 410 16. Lam CK, Staines WR, Tokuno CD, Bent LR. The medium latency muscle response to a vestibular perturbation is increased after depression of the cerebellar vermis. Brain Behav. 2017;7:e00782–9. 411 412 17. Koch G, Brusa L, Carrillo F, Gerfo Lo E, Torriero S, Oliveri M, et al. Cerebellar magnetic stimulation decreases levodopa-induced dyskinesias in Parkinson disease. Neurology. 2009;73:113–9. 413 414 415 18. Koch G, Porcacchia P, Ponzo V, Carrillo F, Cáceres-Redondo MT, Brusa L, et al. Effects of Two Weeks of Cerebellar Theta Burst Stimulation in Cervical Dystonia Patients. Brain Stimulation. Elsevier Ltd; 2014;7:564–72. 416 417 19. Hardwick RM, Lesage E, Miall RC. Cerebellar Transcranial Magnetic Stimulation: The Role of Coil Geometry and Tissue Depth. Brain Stimulation. Elsevier Ltd; 2014;7:643–9. 418 419 20. Fernandez L, Major BP, Teo W-P, Byrne LK, Enticott PG. Assessing cerebellar brain inhibition (CBI) via transcranial magnetic stimulation (TMS): A systematic review. Neurosci Biobehav Rev. 2018;86:176–206. 420 421 21. Hess G, Aizenman CD, Donoghue JP. Conditions for the induction of long-term potentiation in layer II/III horizontal connections of the rat motor cortex. J. Neurophysiol. 1996;75:1765–78. 422 423 22. Huang Y-Z, Rothwell JC, Chen R-S, Lu C-S, Chuang W-L. The theoretical model of theta burst form of repetitive transcranial magnetic stimulation. Clinical Neurophysiology. 2011;122:1011–8. 424 425 23. Huang Y-Z, Chen R-S, Rothwell JC, Wen H-Y. The after-effect of human theta burst stimulation is NMDA receptor dependent. Clinical Neurophysiology. 2007;118:1028–32. 426 427 24. Aizenman CD, Manis PB, Linden DJ. Polarity of long-term synaptic gain change is related to postsynaptic spike firing at a cerebellar inhibitory synapse. Neuron. 1998;21:827–35. 428 429 25. Ni Z, Pinto AD, Lang AE, Chen R. Involvement of the cerebellothalamocortical pathway in Parkinson disease. Ann. Neurol. 2010;68:816–24. 430 431 432 26. Carrillo F, Palomar FJ, Conde V, Diaz-Corrales FJ, Porcacchia P, Fernández-del-Olmo M, et al. Study of Cerebello-Thalamocortical Pathway by Transcranial Magnetic Stimulation in Parkinson's Disease. Brain Stimulation. Elsevier Ltd; 2013;6:582–9. 433 434 435 27. Brighina F, Romano M, Giglia G, Saia V, Puma A, Giglia F, et al. Effects of cerebellar TMS on motor cortex of patients with focal dystonia: a preliminary report. Experimental Brain Research. 2008;192:651– 6. 436 437 438 28. Fernandez L, Major BP, Teo W-P, Byrne LK, Enticott PG. The Impact of Stimulation Intensity and Coil Type on Reliability and Tolerability of Cerebellar Brain Inhibition (CBI) via Dual-Coil TMS. The Cerebellum; 2018;:1–10. 439 440 441 29. Oliveri M, Koch G, Torriero S, Caltagirone C. Increased facilitation of the primary motor cortex following 1Hz repetitive transcranial magnetic stimulation of the contralateral cerebellum in normal humans. Neurosci. Lett. 2005;376:188–93. 18 442 443 444 30. Li Voti P, Conte A, Rocchi L, Bologna M, Khan N, Leodori G, et al. Cerebellar continuous theta-burst stimulation affects motor learning of voluntary arm movements in humans. Eur J Neurosci. 2013;39:124–31. 445 446 447 31. Bologna M, Di Biasio F, Conte A, Iezzi E, Modugno N, Berardelli A. Effects of cerebellar continuous theta burst stimulation on resting tremor in Parkinson's disease. Parkinsonism and Related Disorders. Elsevier Ltd; 2015;21:1061–6. 448 449 32. Helmich RC, Janssen MJR, Oyen WJG, Bloem BR, Toni I. Pallidal dysfunction drives a cerebellothalamic circuit into Parkinson tremor. Ann. Neurol. 2011;69:269–81. 450 451 33. Molnar GF, Sailer A, Gunraj CA, Lang AE, Lozano AM, Chen R. Thalamic deep brain stimulation activates the cerebellothalamocortical pathway. Neurology. 2004;63:907–9. 452 453 34. Lefaivre SC, Brown MJN, Almeida QJ. Cerebellar involvement in Parkinson’s disease resting tremor. Cerebellum & Ataxias. Cerebellum & Ataxias; 2016;:1–7. 454 455 35. Ridding MC, Ziemann U. Determinants of the induction of cortical plasticity by non-invasive brain stimulation in healthy subjects. J. Physiol. (Lond.). 2010;588:2291–304. 456 457 458 36. Pell GS, Roth Y, Zangen A. Modulation of cortical excitability induced by repetitive transcranial magnetic stimulation: Influence of timing and geometrical parameters and underlying mechanisms. Prog. Neurobiol. Elsevier Ltd; 2011;93:59–98. 459 460 37. Hamada M, Murase N, Hasan A, Balaratnam M, Rothwell JC. The Role of Interneuron Networks in Driving Human Motor Cortical Plasticity. Cerebral Cortex. 2013;23:1593–605. 461 19 462 Table Session cTBS50 cTBS30 500µV RMT Mean (SD) Range 55.4 (9.9) 29-67 54.4 (8.8) 31-67 50µV RMT Mean (SD) Range 42.2 (5.5) 28-51 41.2 (5.4) 27-49 463 200µV AMT 464 Mean (SD) Range 49.5 (5.2) 37-57 465 48.9 (5.2) 37-57 466 467 468 469 470 471 472 473 474 Table caption Table 1: Resting (RMT) and active (AMT) motor thresholds used to select stimulator outputs for the CBI and cTBS protocols. RMT was determined with a figure-of-eight coil and AMT with an Airfilm coil. Test pulses were delivered at 90%, 100%, and 110% of the 500µV RMT, conditioning pulses were delivered at 100% of 50µV RMT, and cTBS was delivered at 80% of 200 µV AMT (10% maximum contraction). Values indicate percentages of maximum stimulator output. 20 475 476 477 478 479 480 481 Figures Figure 1 21 482 483 484 485 486 Figure 2 22 487 488 Figure 3 489 490 23 491 492 493 494 Figure 4 495 496 24 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 Figure Captions Fig. 1. Experimental design. Cerebellar brain inhibition (CBI) was measured before (Pre) and after (Post 1 and Post 2) continuous theta burst stimulation (cTBS) was applied to the lateral cerebellum using either cTBS50 or cTBS30 stimulation protocols. The inhibitory effect of cTBS50 and cTBS30 were measured on separate days in the same subjects. CBI was measured in five sets of three blocks, in which three test stimulation intensities (TS intensity: 90%, 100%, 110% of resting motor threshold) were randomly tested Pre, Post 1, and Post 2. Four different inter-stimulus interval trials (ISI: TS alone, 3ms, 5ms, 7ms) were randomly tested twice per block. The TS intensity-ISI combination that yielded the greatest CBI Pre was determined for each subject and used in Pre-Post comparisons. Fig. 2. Cerebellar brain inhibition (CBI). A) Subject set up. A double cone coil (① light grey) delivered a conditioning stimulus (CS) to the cerebellum 3 cm lateral to the inion, ipsilateral to the dominant hand. A figure-of-eight coil (② dark grey) delivered a test stimulus (TS) to the contralateral first dorsal interosseous (FDI) primary motor cortex hotspot. Motor evoked potentials (MEP) were recorded using surface electromyography (EMG) from the dominant FDI. B) Example of ten MEPs (grey) and their average (black). CBI is the ratio between conditioned MEPs (A) where the TS is preceded by a CS shown here at a 5ms inter-stimulus interval (ISI), and the unconditioned MEPs (B) where the TS delivered alone. Fig. 3. Cerebellar brain inhibition (CBI), before (Pre) and ~3-20 minutes after (Post1) and ~23-40 minutes after (Post2) A) 50 Hz cTBS and B) 30Hz cTBS applied to the lateral cerebellum. Lower values indicate greater cerebellar inhibition, while values close to 100% indicate no difference between unconditioned and conditioned motor evoked potentials (MEP). Circles and lines present Individual subject data, and group averages are shown by grey bars (mean, SD). CBI was significantly reduced (higher values) Post1 (P <0.001) and Post2 (P = 0.002) compared to Pre following both 50 Hz (A) and 30 Hz (B) protocols. No differences were found between Post1 and Post2 or between cTBS protocols at any time point (P >0.05). * indicates significant difference (P < 0.05). Fig .4. Cerebellar brain inhibition (CBI: mean, SE) before (Pre) and after (~4-40 minutes, blocks 1-10) cTBS50 (grey) and cTBS30 (black). CBI was significantly reduced (shown as larger %) at blocks 1, 2, and 9 compared to Pre, which indicates depression of the cerebello-thalamocortical (CTC) pathway. No differences were found at any time point between cTBS50 and cTBS30. * indicates significantly higher CBI relative to Pre, P < 0.05. 25 534 535 536 537 538 539 540 541 Supplementary Figure Fig A: Cerebellar brain inhibition (CBI) across each conditioning stimulus intensity: top 90%, middle 100%, and bottom 110%, and continuous theta burst stimulation (cTBS) protocol: left 50 Hz and right 30 Hz. CBI was not observed at any conditioning stimulus inter-stimulus interval (3ms, 5ms, 7ms) combination. Error bars are standard deviations. 26