Brain Stimulation 14 (2021) 949e964 Contents lists available at ScienceDirect Brain Stimulation journal homepage: http://www.journals.elsevier.com/brain-stimulation Reproducibility of cortical response modulation induced by intermittent and continuous theta-burst stimulation of the human motor cortex Recep A. Ozdemir a, b, *, Pierre Boucher a, Peter J. Fried a, b, Davide Momi a, b, Ali Jannati a, b, c, Alvaro Pascual-Leone b, d, e, 1, Emiliano Santarnecchi a, b, f, 1, Mouhsin M. Shafi a, b, **, 1 a Berenson-Allen Center for Noninvasive Brain Stimulation, Department of Neurology, Beth Israel Deaconess Medical Center, Boston, MA, USA b Department of Neurology, Harvard Medical School, Boston, MA, USA c Neuromodulation Program and Division of Epilepsy and Clinical Neurophysiology, Department of Neurology, Boston Children's Hospital, Boston, MA, USA d Hinda and Arthur Marcus Institute for Aging Research and Deanne and Sidney Wolk Center for Memory Health, Hebrew Senior Life, Boston, MA, USA e Guttmann Brain Health Institute, Institut Guttmann de Neurorehabilitacio, Universitat Autonoma de Barcelona, Badalona, Spain f Department of Medicine, Surgery and Neuroscience, University of Siena, Italy a r t i c l e i n f o a b s t r a c t Article history: Background: Over the past decade, the number of experimental and clinical studies using theta-burst- Received 10 December 2020 stimulation (TBS) protocols of transcranial magnetic stimulation (TMS) to modulate brain activity has Received in revised form risen substantially. The use of TBS is motivated by the assumption that these protocols can reliably and 8 May 2021 lastingly modulate cortical excitability despite their short duration and low number of stimuli. However, Accepted 24 May 2021 Available online 12 June 2021 this assumption, and thus the experimental validity of studies using TBS, is challenged by recent work showing large inter- and intra-subject variability in response to TBS protocols. Objectives: To date, the reproducibility of TBS effects in humans has been exclusively assessed with Keywords: TMS-EEG motor evoked potentials (MEPs), which provide an indirect and limited measure of cortical excitability. Theta burst stimulation Here we combined TMS with electroencephalography (TMS-EEG) and report the first comprehensive Transcranial evoked potentials (TEPs) investigation of (1) direct TMS-evoked cortical responses to intermittent (iTBS) and continuous TBS Reliability (cTBS) of the human motor cortex, and (2) reproducibility of both iTBS- and cTBS-induced cortical response modulation against a robust sham control across repeat visits with commonly used cortical responsivity metrics. Results: We show that although single pulse TMS generates stable and reproducible cortical responses across visits, the modulatory effects of TBS vary substantially both between and within individuals. Overall, at the group level, most measures of the iTBS and cTBS-induced effects were not significantly different from sham-TBS. Most importantly, none of the significant TBS-induced effects observed in visit- 1 were reproduced in visit-2. Conclusions: Our findings suggest that the generally accepted mechanisms of TBS-induced neuro- modulation, i.e. through changes in cortical excitability, may not be accurate. Future research is needed to determine the mechanisms underlying the established therapeutic effects of TBS in neuropsychiatry and examine reproducibility of TBS-induced neuromodulation through oscillatory response dynamics. © 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). * Corresponding author. Berenson-Allen Center for Noninvasive Brain Stimulation, Department of Neurology, Beth Israel Deaconess Medical Center, Boston, MA, USA. ** Corresponding author. Department of Neurology, Harvard Medical School, Boston, MA, USA. E-mail addresses: rozdemir@bidmc.harvard.edu (R.A. Ozdemir), mshafi@bidmc.harvard.edu (M.M. Shafi). 1 Contributed equally. https://doi.org/10.1016/j.brs.2021.05.013 1935-861X/© 2021 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Introduction systematic research has been performed to evaluate reproducibility of TBS-induced effects in cortical responses using different TBS Repetitive Transcranial Magnetic Stimulation (rTMS) involves protocols with a robust sham-control. Furthermore, even the the application of sequences of TMS pulses at different patterns and impact of TBS on corticospinal excitability is called into question by frequencies to induce a lasting modulation of brain activity, pre- accumulating evidence showing high variability between in- sumably via Hebbian synaptic plasticity mechanisms [1,2]. A variety dividuals [28,29]. In particular, several studies [30e34] with large of different rTMS protocols have been developed, with the goal of sample sizes have failed to produce consistent and expected reliably potentiating or suppressing cortical activity in the stimu- response patterns, with only ~50% of subjects showing the ex- lated brain region [3,4]. These rTMS protocols have been increas- pected neuromodulation in response to TBS protocols. Conse- ingly used in neuroscience research to gain insights into causal quently, TBS protocols used in majority of these studies [30,31,33] relations between brain activity and behavior [5,6], to study brain did not show any significant neuromodulation at the group level. networks [7], explore their dynamics, and characterize mecha- Recent research has also reported a considerable degree of intra- nisms of brain plasticity [8]. Furthermore, a number of therapeutic subject variability in TBS effects, such that the MEP response to rTMS protocols have already been developed and cleared by the TBS can be dramatically different (e.g. suppression versus facilita- Food and Drug Administration for treatment of medication- tion) across day-to-day measurements for a given individual [35]. resistant depression [9], obsessive compulsive disorder [10], and Most importantly, relatively few studies included sham TBS pro- migraine [11]. Although randomized controlled clinical trials tocols, and thus the extent to which observed changes in excit- [9e11] have provided robust evidence on the therapeutic effects of ability can be attributed to the TMS stimulation rather than other rTMS, most uses of rTMS in basic neurophysiological research are aspects of the experimental design remains unclear. predicated on the premise that rTMS modulates neural excitability Here, we aimed to address these fundamental knowledge gaps in the stimulated region and/or its associated brain networks, and by first (1) investigating the effects of iTBS, cTBS and sham stimu- that these modulatory effects are reliable and predictable. lation of primary motor cortex (M1) on single-pulse TMS-evoked Theta-Burst Stimulation (TBS) is a particular form of rTMS cortical (EEG) and motor (MEP) responses in a single session, and inspired by invasive animal studies [12] reporting synaptic long- then (2) examining the reproducibility of TBS-induced neuro- term-potentiation (LTP) in hippocampal axons in response to modulation with identical repeat sessions performed at least one patterned electrical stimulation with high-frequency gamma bursts month apart. For corticospinal responses, we expected significant (50hz) repeated at the theta frequency (4e7 Hz). The first trans- facilitation of MEPs following iTBS, suppression following cTBS, and lation of this protocol into humans reported that when TBS was no significant change following sham TBS. Given the paucity of applied as continuous (cTBS) trains with a total of 600 gamma research on TEP responses to different TBS protocols, we did not pulses in 40 s over the hand region of the primary motor cortex, it predict directional differences between active TBS (iTBS vs cTBS) suppressed corticospinal excitability for up to 60 min [13], as protocols on specific TEP elements, but rather hypothesized that indexed by decreased motor-evoked potentials (MEPs) recorded both iTBS and cTBS would significantly modulate TEP responses to from the contralateral hand muscles. Conversely, if the same TMS, while sham-TBS would not induce any significant TEP mod- amount of gamma pulses were applied as 2 s of intermittent trains ulation. We also hypothesized that such TBS-induced MEP and TEP (iTBS) with 8 s of inter-stimulus intervals, corticospinal excitability modulations would be reliable across sessions. is facilitated for up to 20 min. Because of their low stimulation intensities, shorter durations, and reported durability of effects Methods [13], TBS protocols are often considered among the most effective, safe and practical approach to noninvasive neuromodulation [14]. Participants: Twenty-four right-handed healthy volunteers (16 Traditionally, TBS has been applied to motor cortex to modulate males; mean ± SD age ¼ 29.67 ± 10.60 years, range ¼ 18 to 49) corticospinal excitability and evaluate the mechanisms of plasticity participated in this study. None of the participants had self- [15,16]. Further, a growing body of behavioral and clinical research reported history of psychiatric or neurological diseases. In accor- has employed TBS to non-motor regions to modulate local cortical dance with the Declaration of Helsinki, experimental protocols and function and network-level connectivity in different non-motor voluntary participation procedures were explained to all partici- cortices [17,18], and claimed therapeutic effects with specific pants before they gave their written informed consent to the study. brain-behavior relationships through presumed mechanisms of All questionnaires and procedures were approved by the Institu- TBS in modulating cortical excitability [9,19e21]. tional Review Board of the Beth Israel Deaconess Medical Center, Despite its increasingly widespread experimental and clinical Boston, MA. use, however, the effects of TBS on cortical activity have not been Experimental Procedures: A T1-weighted (T1w) anatomical MRI clearly established. While the time course of TBS-induced neuro- scan was obtained from all participants at the beginning of the modulation has been extensively studied at the corticospinal level, study and used for neuronavigation during TMS visits. In each visit, MEPs are indirect readouts of cortical excitability, and thus pre- TMS-EEG and TMS-EMG data were collected synchronously. Details cisely how cortical circuits are being modulated remains unclear. of MRI scanning, TMS, EEG and EMG systems are provided in the Direct epidural recordings from the cervical region of the spinal supplementary methods. cord have revealed that while cTBS suppresses early I-waves [22], Each participant completed a total of six visits: two visits for iTBS preferentially potentiates later I-waves [23]. These results iTBS, two visits for cTBS and two visits for sham TBS conditions. The suggest that TBS-induced changes observed in MEPs may reflect order of first visits for each TBS condition was randomized between modulation of local and specific subpopulations of neuronal circuits subjects and the same order was kept for repeat visits. First visits of in the stimulated motor region. Given the limitations of MEPs, TMS TBS conditions were spaced at least 2 days apart to minimize carry in combination with EEG (TMS-EEG) can be used to characterize over effects, and repeat visits of each TBS condition were performed TBS-induced neuromodulation directly at the cortical level, and at least 1 month after the first set of visits. Visits for a given thus may provide a more comprehensive and thorough examina- participant were scheduled at the same time of the day as much as tion of TBS effects across the cortex [24]. To date, few studies have possible. reported TBS-induced changes in TMS evoked potentials (TEPs) The general layout of our experimental design for TBS protocols from a single session [25e27] with conflicting results. No is summarized in Fig. 1A. At the beginning of each visit, the motor 950 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 hotspot was determined over the hand region of left motor cortex Following determination of motor thresholds in each session, (L-M1) for eliciting MEPs in the right FDI muscle. The hotspot was 120 single pulses of TMS (spTMS) at 120% of RMT were delivered to defined as the region where single-pulse TMS elicited consistent the motor hotspot as a baseline measure of corticospinal (MEPs) MEPs in the FDI muscle. Following International Federation of and cortical (TEPs) excitability. The TBS protocol for that session Clinical Neurophysiology (IFCN) guidelines, resting motor was then applied, followed by 4 blocks of 60 spTMS at 120% of RMT threshold (RMT) was determined on the FDI hotspot as the mini- at 5 (T5), 10 (T10), 20 (T20), and 30 (T30) minutes after the TBS mum stimulation intensity eliciting at least five MEPs (50 mV) out protocol (Fig. 1A). This manuscript primarily focuses on TBS effects of ten pulses in the relaxed FDI using monophasic (posterior- on cortical excitability as assessed via TEPs; MEP effects are pre- anterior in the brain) current waveforms [36,37]. In compliance sented briefly here and are reported in detail in a separate manu- with the IFCN safety recommendations, participants were asked to script (Boucher et al., submitted). wear earplugs during hotspot and RMT trials to protect their TBS Procedures: TBS was applied to the motor hotspot at 80% of hearing, and to minimize external noise [38,39]. Active motor AMT. All TBS protocols were delivered as 3 pulse bursts at 50 Hz threshold was determined by again asking participants to flex their with 200 ms between bursts (600 pulses total). This pattern was right index finger to engage the FDI muscle to approximately 20% of delivered continuously in cTBS and in a 2 seceon, 8 seceoff pattern maximum voluntary contraction (MVC) and determined at the for iTBS. Participants were randomly assigned to receive either lowest intensity to produce MEPs of 200mv at least 5/10 times. sham using either the cTBS or iTBS pattern, which was maintained TMS was administered with a thin layer of foam placed under the across both sham visits. Sham cTBS and iTBS protocols were coil to minimize somatosensory contributions to TEPs. To minimize administered on the motor hot spot from the placebo side of the auditory evoked potentials related to the TMS coil click, auditory Cool-B65 A/P coil with a 3D printed 3 cm spacer additionally white noise masking was used throughout the TMS stimulation. attached to the placebo side (MagVenture A/S, Farum, Denmark). Fig. 1. Experimental protocol and TMS-EEG metrics. A: Overall layout of a representative TMS-EEG session. B: TMS evoked potentials from 63 channels (upper panel) with selected peaks (colored vertical lines). Lower panel in B shows topographical distribution of selected peaks (blue: n15, red: P30). C: Computation of GMFP (upper panel), LMFP in left motor cortex (middle panel) within pre-defined time-windows (Early responses: 15e75 ms and Late responses: 76e300 ms following TMS pulse), and selected TEP peaks (lower panel) extracted from C3 electrode before (red line) and after (black line) iTBS at T20 in a representative subject. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 951 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 2. ICA-based cleaning of TMS-EEG. Topography (upper panels), sorted single trial amplitudes (middle panels) and averaged times series (lower panels) representing pulse, eye, muscle, EKG artifacts and auditory evoked potential (AEPs) components from a representative subject. These components were removed from the data using fICA (see methods for details of EEG preprocessing and supplementary materials for details of AEP identification). Both active and sham-TBS protocols also included delivery of weak components with early TMS evoked high amplitude electrode and current pulses (between 2 and 4 mA and proportional to the in- EMG artifacts (1 ± 1 components were removed; range 0e3 out of tensity of actual TMS pulse) via surface electrodes (Ambu Neuroline 63). After the first round of fICA, the EEG data were interpolated for 715 12/Pouch) placed approximately 1 cm below the inion bump previously zero-padded time window around TMS pulse and synchronized with the TBS trains to produce scalp sensations (-2 mse14 ms) using linear interpolation, band pass filtered using a during both active and sham TBS conditions. This was done with forward-backward 4th order butterworth filter from 1 to 100 Hz, the intention of blinding participants as to what kind of stimulation notch filtered between 57 and 63 Hz, and referenced to global they were receiving when the direct somatosensory sensations of average. Subsequently, a second round of fICA was run to manually active TBS were not present during the sham stimulation. remove all remaining artifact components including eye move- ment/blink, muscle noise (EMG), single electrode noise, TMS EEG preprocessing evoked muscle, cardiac beats (EKG), as well as auditory evoked potentials. Fig. 2 shows representative topographies, single trial All EEG data pre-processing was performed offline using amplitudes and averaged time series of ICA components removed EEGLAB 18.1 [40], and customized scripts running in Matlab R2017b from the data in second round of fICA. (Math-Works Inc., USA). For TEP analyses, 120 single-pulse M1 Details for identifying and validating AEP components were block (Fig. 1A) before TBS were used as baseline (PreTBS) measures. provided in supplementary materials (see supplementary methods 60 single pulse TEPs following TBS from both T5 and T10 were and Supplementary Fig. 1). In both rounds of fICA, a semi- merged to create a single block of 120 trials and labeled as T5, while automated artifact detection algorithm incorporated into the TEPs from T20 and T30 were merged to create another single block open source TMS-EEG Signal Analyzer (TESA v0.1.0-beta [41]) and labeled as T20 to measure TBS effects at two separate time extension for EEGLAB was used to classify and visually inspect points (Fig. 1A, red blocks). All trials within each block were then components based on their frequency, activity power spectrum, segmented into 3000 ms epochs, each starting 1000 ms before amplitude, scalp topography, and time course (http://nigelrogasch. (pre-stimulus) and ending 2000 ms (post-stimulus) following TMS github.io/TESA/). Finally, the data were low pass filtered with a 4th pulse, respectively. Baseline correction was performed by sub- order Butterworth filter at 50 Hz, and interpolated for missing/ tracting the mean pre-stimulus (900 to 100) signal amplitude removed channels using spherical interpolation. from the rest of the epoch in each channel. Following baseline TMS-EEG and TMS-EMG metrics: Details for computing Global correction, data were visually inspected to identify noisy channels Mean Field Power (GMFP) and Local Mean Field Power (LMFP) as (2.5 ± 1.5 channels were deleted on average; range 0e4 out of 63). TMS-EEG metrics are provided in the supplementary methods. 60 Zero-padding between -2 ms and 14 ms time range were then single pulse MEPs and TEPs following TBS from both T5 and T10 applied to remove the early TMS pulse artifact from the EEG data. were merged to create a single block of 120 trials and labeled as T5, All zero padded epochs were then tagged based on voltage while MEPs and TEPs from T20 and T30 were merged to create (100 mV), kurtosis (3), and joint probability (Single channel- another single block and labeled as T20 to measure TBS effects at based threshold  3.5sd; All channel-based threshold  5sd) two separate time points (Fig. 1A, red blocks). metrics to identify excessively noisy epochs. Visual inspection was Both for GMFP and LMFP, we took the integral of the time series performed on the tagged epochs for the final decision for the and computed area under the curve (AUC) as the amount of global removal of noisy epochs (18 ± 6 epochs were deleted on average; (GMFP) and local (LMFP) cortical response measure in two separate range 2e39 out of 120). Next, an initial round of fast independent time windows. The first AUC window ranges from 15 to 75 ms and component analysis (fICA) was performed to identify and remove represents the sum of low amplitude and “early responses”, while 952 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 the second AUC window ranges from 76 to 300 ms and represents point across blocks (PreTBS, T5, T20) for each TBS-condition and the sum of large and “late responses” (Fig. 1C, second and third visit. First, we ran paired sample t-tests at each sample to deter- panels from the top). AUC values at each postTBS time point (T5 and mine significant time points between PreTBS and T5, and between T20) were normalized to baseline (PreTBS) by taking the percentage PreTBS and T20 separately. We then computed the length of adja- of “postTBS/PreTBS” for each subject. These AUC ratios were used to cent significant time points and sum of t-scores for significant time classify postTBS response of each subject into three categories: A points to determine (1) cluster size and (2) cluster magnitude in the “Facilitation” response was defined as a minimum of 10% increase, main analyses, respectively. Following main analyses, we per- while a “Suppression” response is defined as a minimum of 10% formed permutation t-tests [44] (n ¼ 1000) by randomly shuffling decrease in AUC values at postTBS time points (T5 and T20). This 10% 50% of subjects across compared blocks (i.e., 50% of subjects shifted change threshold in postTBS is chosen to ensure a clear TBS effect from PreTBS to T5 or vice versa) and determined cluster size and accounting possible random fluctuations around baseline levels magnitude of significant adjacent time points at each iteration. [42]. Therefore, subjects showing less than 10% increase or decrease Finally, we re-compute p-values of significant clusters in the main in their responses at postTBS were classified with a “No Change” analyses by calculating the probability of their size and magnitude response type. In addition to AUC ratios, time series analyses were in the permutation analyses. A cluster in the main analyses is also performed both for GMFP and LMFP responses to examine considered to survive permutation, and thus significant, only if significant differences between TBS blocks (Baseline vs T5 and both the size and magnitude of a given cluster is above 95% of all Baseline vs T20) at the highest temporal resolution. cluster sizes and magnitudes derived from permutation tests. TEP peaks for each EEG channel were extracted by using pre- Similarly, cluster-based permutation paired sample t-test statistics determined time windows. The first negative peak N15 with multiple corrections were performed separately for each TEP (14e24 ms following TMS), the first positive peak P30 (25e45 ms peak (N15, P30 and N100) to identify significant cluster of channels following TMS) and a second negative peak N100 (80e120 ms across comparisons (PreTBS vs T5 and PreTBS vs T20). A significant following TMS) were identified using these time windows, and the cluster in main analyses is defined as at least two neighboring voltage amplitudes within each time range were averaged to channels at T5 or T20 that significantly differ from baseline mea- compute TEP peaks (Fig. 1C, bottom panel). surements (PreTBS) and survived permutation (n ¼ 1000). MEPs were computed as the absolute amplitude difference be- ICCs were computed for all metrics to examine test-retest reli- tween minimum and maximum voltage peaks from 20 to 50 ms ability of TBS-induced effects. We first normalized all metrics following TMS for each trial. For a given block, individual MEPs (GMFP, LMFP, TEP and MEPs) to baseline assessments by taking the greater than 2.5 standard deviations (SDs) from the mean were ratio of each metric at T5 and T20 in reference to preTBS. We then considered outliers and rejected from further analyses. Similar to performed ICCs on the ratio scores across all visits and between GMFP and LMFP, mean MEP amplitudes for each post stimulation pairs of identical visits for T5 and T20 separately. ICCs < 0.25 were time-point were expressed as the percentage change from baseline. considered as “very low”, between 0.25 and 0.50 were considered as “low”, between 0.50 and 0.75 were considered as moderate, and Statistical analyses >0.75 were considered as “high” reproducibility, respectively [45]. Data from iTBS in one participant, from cTBS in another participant, To evaluate the stability of baseline measurements across visits, and from a sham sessions in two participants were excluded from preTBS GMFP, and LMFP were entered into a repeated-measure the final analyses due to insufficient TMS-EEG and TMS-EMG data analyses of variance (Rm-ANOVA) with Visit (V1, V2, V3, V4, V5, sets following the respective TBS protocols. V6) as a within-subject factor. To test overall reproducibility of cortical responses to spTMS, intra-class correlation coefficients Results (ICCs) were computed for all metrics (GMFP, LMFP AUC, and TEP peaks) to calculate absolute agreement across all visits [43]. For the Baseline TMS-evoked cortical responses are stable over time neuromodulatory effects of TBS, our goal was to examine if there is any “TBS-condition” main or “TBS-condition x Block” interaction We first examined how baseline measurements of TMS evoked effect in the first visit (Visit-1) and, if so, whether we can reproduce cortical responses varied across repeated sessions (Fig. 3). Rm- these effects in repeat data sets with an independent set of analyses ANOVA revealed no significant main effect of Visit for both GMFP (Visit-2). Thus, GMFP, and LMFP AUC, values from the 1st and 2nd (p > 0.05) and LMFP (p > 0.05) responses, indicating statistically visit of each condition were entered into separate Rm-ANOVAs stable spTMS evoked global and local responses across all visits with TBS-condition (iTBS, cTBS, Sham) and Block (PreTBS, T5, T20) (Fig. 3A, Supplementary Table 1). ICC analyses for GMFP and LMFP as within-subject factors. In addition, to assess the effects of TBS on in L-M1 showed significant correlations both for early (p < 0.00) more conventional measures of corticospinal excitability, similar and late (p < 0.00) responses with moderate reproducibility analyses were conducted for MEP values. We also performed pre- (0.50  r < 0.75) in late responses across visits (Fig. 3A, right liminary analyses to examine the relationships between the TBS- panels). For TEPs, the highest reproducibility was achieved for the induced changes in early cortical responses (early LMFP, N100 peaks, followed by P30 and N15 peaks (Fig. 3B). For N15 15e75 ms) and baseline measures of corticospinal (MEPs) and peaks, ICCs were not significant for one electrode (C3) on the cortical (early LMFP) excitability. For this aim, early LMFP responses stimulation site and four other (FT7, T8, TP8 and TP10) temporal following each TBS condition were grouped as facilitation (post-TBS electrodes (Fig. 3B, left panel). All electrodes for the P30 (Fig. 1B, LMFP > pre-TBS LMFP) and suppression (post-TBS LMFP < pre-TBS middle panel) and N100 (Fig. 3B, right panel) peaks were signifi- LMFP) at T5, and baseline MEP and early LMFP amplitudes were cantly correlated across visits. retrospectively compared between groups using independent t-test statistics with Bonferroni correction for multiple comparisons. Descriptive classification of global and local cortical responses to TBS Further, we used bivariate Pearson correlations to examine the relationships between changes in MEPs and early LMFP responses, Results of GMFP and LMFP from L-M1 responses to TBS are and between changes in MEPs and the early TEP peaks (N15, P30). summarized in Fig. 4 and Fig. 5, respectively. For GMFP, only Cluster-based permutation paired sample t-test statistics were 35e44% of participants showed excitatory iTBS response (Facilita- performed to compare GMFP and LMFP time series at each time tion: post TBS ratios >110%) with increased ratios across visits 953 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 3. Reproducibility of TMS-EEG metrics at baseline (Pre-TBS) measurements. A: Group average AUC values for GMFP (upper panels) and LMFP responses (Early responses: left panels and Late responses: right panels) at baseline (PreTBS) measurements of each visit. Error bars show one unit of standard error. B: Topographical distribution of intra-class correlation coefficients of each TEP peaks across all visits at the electrode level. ICCs are not significant for red colored electrodes, and significant for black colored electrodes. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) (Fig. 4A, upper panel), while 31e52% of participants did not show Block interaction for early responses in both visits (Supplementary changes with stable GMFPs (No change: post TBS ratios between 90 Table 1). As for late GMFP responses, we found a significant main and 110%) and 9e22% of participants showed inhibitory responses effect of Block both for visit-1 (F(2,38) ¼ 6.06, p ¼ 0.001, h2p ¼ 0.24, with decreased GMFPs (Suppression: post TBS ratios < 90%). GMFP power ¼ .892) and visit-2 (F(2,38) ¼ 4.521, p ¼ 0.010, h2p ¼ 0.20, responses to cTBS were also highly variable with 9e40% of partic- power ¼ .754), but no significant main effect of TBS-condition and ipants having inhibitory response with suppressed GMFP, while no significant TBS-condition by Block interaction (p > 0.05). Follow 23e59% and 33e46% of participants showed no change or facili- up pairwise comparisons revealed significant GMFP increase at T5 tation, respectively. Fig. 4B illustrates both the inter-and intra- compared to PreTBS block for both visits (Visit-1: F(1,19) ¼ 8.322, subject variability to TBS conditions across visits. While the ma- jority of group-averaged responses are around 100% of baseline, p ¼ 0.009, h2p ¼ 0.305, power ¼ .881; Visit-2: F(1,19) ¼ 10.706, individual responses show high variability across subjects, and p ¼ 0.004, h2p ¼ 0.360, power ¼ .911). Altogether, these results generally without any clear tendency towards facilitation versus suggest that the late GMFP response show nonspecific increases at suppression. Furthermore, when each subject's own responses are T5 across all TBS-conditions. connected across visits, substantial changes in the direction of in- ICCs revealed low reproducibility for only early responses at T5 dividual responses are observed such that individual responses in for Sham condition (r ¼ 0.38, p ¼ 0.02), while all other TBS- the first session (i.e. facilitation versus no-response versus sup- conditions showed very low to no reproducibility both for early pression) are not clearly associated with the response in the second and late responses, with ICCs ranging from r ¼ 0.17e0.20 (Fig. 4C). session. Similar to GMFPs, LMFPs from L-M1 (ipsilateral to the Cluster based permutation t-tests for the TEP time series at T5 stimulation) showed high variability in response classifications for (Fig. 4D) showed significant clusters between 141 and 169 ms for all TBS-conditions (Fig. 5A), and high inter- and intra-subject iTBS, and between 141-668 and 290e332 ms for cTBS in Visit-1. variability in TBS-induced modulation within and across visits However, none of these clusters were reproduced in Visit-2, sug- (Fig. 5B). gesting lack of reproducibility for the TBS effects on GMFP re- sponses in the temporal dimension (Fig. 4D). No significant TBS effects on TMS-evoked cortical and motor responses are not differences that survived permutations were found between pre- reproducible TBS and T20 (Fig. 4D). Similar to GMFPs, Rm-ANOVA for LMFP revealed no significant Rm-ANOVA for GMFP values revealed no main effects of TBS- main effects of TBS-condition (p > 0.05) or Block (p > 0.05), and no condition (p > 0.05) or Block (p > 0.05), and no TBS-condition by TBS-condition by Block interaction (p > 0.05) for early responses in 954 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 4. TBS effects on GMFP. A: Classification of subjects according to response types to TBS in ratios (Red: increased response following TBS/facilitation, Yellow: decreased response following TBS/suppression and Gray: no change in response following TBS/no response). B: Bars show group averaged response ratios normalized to baseline (PreTBS) at T5 (left panels) and T20 (right panels) both for early (upper panels) and late (lower panels) responses for each visit. Colored dots superimposed over group average bars represent individual responses and black lines connecting dots over group bars track individual response changes across visits. C: Scatter plots, regression lines and ICCs of GMFP ratios both early and late responses (Blue: reproducibility of responses at T5 and Red: at T20) across visits. D: Statistical comparison GMFP responses at the millisecond level for both visit-1 (upper panels) and visit-2 (lower panels). Colored lines represent group-averaged responses to TBS conditions while shaded regions show standard error of measurement (1 unit). Colored blocks at the bottom of each panel show significant cluster of time-points between post-TBS (Blue: T5 and Red: T20) and baseline (PreTBS) measurements (Empty blocks: Significant clusters for bivariate comparisons but did not survive permutation tests and Filled Blocks: Significant clusters survived permutation tests. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 955 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 5. TBS effects on LMFP (Left Motor Cortex). A: Classification of subjects according to response types to TBS in ratios B: Bars show group averaged response ratios normalized to baseline for each visit. Colored dots superimposed over group average bars represent individual responses and black lines connecting dots over group bars track individual response changes across visits. C: Scatter plots, regression lines and ICCs of LMFP ratios both early and late responses (Blue: reproducibility of responses at T5 and Red: at T20) across visits. D: Statistical comparison LMFP responses at the millisecond level for both visit-1 (upper panels) and visit-2 (lower panels). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 956 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 both Visit-1 and Visit-2. For late responses, we found a significant left sensorimotor cortices, while cTBS showed increased N100 main effect of Block both in visit-1 (F(2,38) ¼ 7.352, p ¼ 0.002, response over a series of left parietal and posterior parietal region h2p ¼ 0.28, power ¼ .920) and Visit-2 (F(2,38) ¼ 4.048, p ¼ 0.025, as well as a reduced (more positive) N100 response over right frontal and temporal regions at T5 (Fig. 7A, lower-left and middle h2p ¼ 0.18, power ¼ .689), but no significant main effect of TBS- panels). Importantly, however, none of these significant clusters condition and no significant “TBS-condition by Block interaction. was reproduced at Visit-2, but different patterns of significant Follow up pairwise comparisons revealed significant LMFP increase modulations were observed for N15 and P30 peaks after cTBS and at T5 compared to PreTBS block for both visits (Visit-1: Sham conditions, respectively (Fig. 7B, upper and middle panels). F(1,19) ¼ 9.112, p ¼ 0.016, h2p ¼ 0.342, power ¼ .843; Visit-2: Similarly, significant set of clusters in each TEP peak observed at T20 F(1,19) ¼ 6.762, p ¼ 0.024, h2p ¼ 0.305, power ¼ .718), suggesting for Visit-1 were generally not reproduced in Visit-2 for all TBS- that the late LMFP values show nonspecific increases at T5 across all conditions (Fig. 8A, and Fig. 8B). Sham TBS was the only condition TBS-conditions (Supplementary Table 1). with reproducible modulation for a subset of electrodes for P30 Intra-class correlations revealed very low reproducibility both peaks across visits (Fig. 8A and B middle-right panels). for early and late responses for iTBS and cTBS with ICCs ranging Next, we tested whether modulations induced by iTBS and cTBS from r ¼ 0.18 to 0.26 (Fig. 5C). Early responses at T5 for Sham TBS were significant compared to Sham TBS and, if so, whether these showed the highest ICCs (r ¼ 0.29, p ¼ 0.048). Although, cluster significant differences from Sham TBS were reproducible across based permutations for time series analyses revealed significant visits (Fig. 9). Both iTBS and cTBS had significantly larger (more differences between PreTBS and T5 for cTBS at multiple time win- negative) N15 amplitudes over left motor cortex when compared to dows in Visit-1 (Fig. 5D middle upper panel), none of these clusters sham TBS at T5 for Visit-1 (Fig. 9A, Visit-1 upper panels). Only were reproduced in Visit-2 (Fig. 5D middle lower panel). LMFPs significant cTBS differences from sham in left motor cortex were from R-M1 also showed significant effect of Block for both early partially reproduced in Visit-2, but within a substantially larger (F(2,38) ¼ 5.938, p ¼ 0.001, h2p ¼ 0.24, power ¼ .895) and late cluster extending from left parietal to left temporal cortex observed in visit 2 (Fig. 9A, Visit-2 upper panels). Significant differences (F(2,38) ¼ 3.590, p ¼ 0.031, h2p ¼ 0.16, power ¼ .714) responses in between active and sham TBS conditions observed for N100 Visit 2 only, but no main effect of TBS-condition and no TBS-condi- response in Visit-1 were also not reproduced in Visit-2 (Fig. 9A, tion by Block interaction effect (See Supplementary Table 1 for visit-1 and visit-2 lower panels). Similarly for T20, significant dif- statistics and Supplementary Fig. 4 for details). No significant main ferences between active and sham TBS conditions observed in effects or interactions were found for V1 (See Supplementary Visit-1 were not reproduced in Visit-2, except partially for larger Table 1 for statistics and Supplementary Fig. 5 for details). (more negative) N15 amplitudes over left motor cortex in cTBS TBS effects on MEPs are summarized in Fig. 6. Briefly, similar to condition (Fig. 9B, visit-1 and visit-2 upper panels). GMFP and LMFP, individual responses show high variability across We also tested the reproducibility of TBS effects on TEPs at the subjects, and within subjects across visits for all TBS conditions electrode level by computing ICCs for each channel across TEP (Fig. 6A.) Rm-ANOVA revealed no main effects of TBS-condition peaks and TBS conditions (Fig. 10). There was no clear pattern on (p > 0.05) and Block (p > 0.05), and no TBS-condition by Block the spatial distribution and direction of ICCs, with different elec- interaction (p > 0.05) in Visit-1 (Supplementary Table 1). In Visit-2, trodes both negatively and positively weakly correlated across we found a significant main effect of Block (F(2,38) ¼ 5.51, p ¼ 0.001, visits (r ¼ 0.5e0.5). h2p ¼ 0.22, power ¼ .823), but no significant main effect of TBS- condition (p > 0.05) and no significant TBS-condition by Block Role of baseline excitability on TBS effects and relationship between interaction (p > 0.05). Follow up pairwise comparisons revealed changes in MEPs and TEPs significant increase in MEP amplitudes at T20 compared to PreTBS block (F(1,19) ¼ 14.712, p ¼ 0.001, h2p ¼ 0.436, power ¼ .953), indi- Here, we performed preliminary analyses to better understand cating nonspecific MEP increases at T20 across all TBS-conditions whether (1) baseline corticospinal and cortical excitability indexed (Supplementary Table 1). ICCs revealed negative correlations by MEPs and early LMFPs, respectively, differ in subjects with (Fig. 6B, upper panel) with very low reproducibility for iTBS both at facilitatory versus suppressive cortical responses to TBS, and, if so, T5 (r ¼ 0.24, p ¼ 0.84) and T20 (r ¼ 0.14, p ¼ 0.73). For cTBS, (2) whether TBS-induced modulation of corticospinal and cortical however ICCs were positive across sessions (Fig. 6B, middle panel) responses are related (Fig. 11B and Fig. 11C). We found that in- with low reproducibility at T5 (r ¼ 0.39, p ¼ 0.02), and very low dividuals with facilitatory LMFP responses (n ¼ 14) following iTBS reproducibility at T20 (r ¼ 0.11, p ¼ 0.30). While MEPs from at T5 had significantly lower (t(1,22) ¼ -5.80, p ¼ 0.000) MEPs at shamTBS were positively correlated at T5 (Fig. 6B, lower panel) with baseline than those with a suppressive response (n ¼ 9) in visit-1 low reproducibility (r ¼ 0.28, p ¼ 0.08), no correlation pattern was (Fig. 11A, left panel) suggesting that the facilitatory effects of iTBS observed at T20 (r ¼ 0.02, p ¼ 0.64). MEP effects are explored with on cortical reactivity are seen primarily in those with lower corti- substantial detail in Boucher et al. (submitted). cospinal excitability at baseline. Contrary to iTBS, cTBS further suppressed cortical reactivity (n ¼ 12) in individuals with low baseline corticospinal excitability (t(1,22) ¼ 2.44, p ¼ 0.035) in Visit- TBS effects on TEP components are not reproducible 1. However, although a similar pattern was observed for iTBS, baseline MEP comparisons were not significant (p > 0.05) in visit-2 Group averaged topographies for each TEP peak across TBS for both iTBS and cTBS indicating poor reproducibility. No signifi- conditions, blocks and visits are shown in Supplementary Fig. S6. cant baseline MEP differences were found (p > 0.05) between Cluster based permutation t-test statistics revealed significantly facilitated versus suppressed LMFP responses following sham for reduced N15 amplitude (p < 0.05) for electrodes over left sensori- both visits. Furthermore, we found no statistical difference in motor cortices at T5 for Sham TBS (Fig. 7A, upper-right panel). As for baseline LMFP comparisons (p > 0.05) between facilitation P30 peaks, a series of mid-frontal electrodes showed increased P30 (LMFPþ) and suppression (LMFP-) cortical responses to TBS. response while left parietal electrodes showed decreased P30 Next, we examined relationships between changes in MEP and response only for cTBS at T5 (Fig. 7A, middle panel). iTBS showed LMFP responses following each TBS condition and visit (Fig. 11B). increased (more negative) N100 response in four electrodes over We found significant positive correlations between MEP and early 957 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 6. TBS effects on MEPs. A: Bars show group averaged response ratios normalized to baseline (PreTBS) at T5 (left panels) and T20 (right panels) for both visits. Colored dots superimposed over group average bars represent individual responses and black lines connecting dots over group bars track individual response changes across visits. B: Scatter plots, regression lines and ICCs of MEP ratios at T5 (Blue) and T20 (Red). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) LMFP ratios following iTBS both for visit-1 (r ¼ 0.44, p ¼ 0.005) and aims. The initial study establishing these protocols reported visit-2 (r ¼ 0.40, p ¼ 0.019). Although cTBS was associated with increased corticospinal excitability (assessed with TMS-elicited negative correlations between MEP and LMFP ratios, the magnitude MEPs) lasting up to 20 min following iTBS, and decreased cortico- of correlations were low (visit-1: r ¼ 0.17, visit-2: r ¼ 0.03) and spinal excitability lasting up to 60 min following cTBS [13,46]. Since not significant (p > 0.05) for both visits. There was no significant then, both iTBS and cTBS have been conceptualized as “excitatory” correlation between MEP and LMFP changes with sham stimula- and “inhibitory” rTMS protocols, respectively. Today these TBS tion. Finally, we examined changes in MEPs and early TEP compo- protocols have been widely used to induce lasting brain plasticity nents (Fig. 11C) to further understand specific contributions of each and modulate human behavior and cognition. Numerous studies, TEP component to the relationship between corticospinal and local for example, have applied TBS outside the motor cortex to disrupt cortical neuromodulation observed in Fig. 11B. We found that [19,20,47] or enhance normal behaviors [7,48], as well as to recover changes in MEP responses were positively and consistently corre- impaired behaviors [49,50] and treat neuropsychiatric conditions lated with changes in P30 component over the stimulated region [9]. Observed changes in behavior, cognition or neuropsychiatric following iTBS in both visits (Fig. 11C left lower panels). On the conditions are attributed to TBS-induced brain plasticity over the other hand, cTBS resulted in weak negative correlations between stimulated networks. The vast majority of these studies have MEP and TEP changes over the stimulated cortical region only in inferred mechanisms based on the presumed “excitatory” or visit-1 but not in visit-2. Taken together, these results suggest that “inhibitory” effects of the particular TBS without independent the P30 component reliably reflects relationships between corti- neurophysiologic measurements of the direct cortical responses to cospinal and local cortical response modulations following iTBS. TBS. However, a fundamental limitation inherent to the use of TBS is that the presumed effects on cortical excitability have not been Discussion adequately established. More specifically, the neuromodulatory effects of TBS have typically been assessed with MEPs, which pro- TBS is an rTMS protocol that is widely used with the goal of vide at best an indirect and limited measure of local cortical modulating cortical activity for both scientific and therapeutic excitability in the stimulated primary motor region, and cannot be 958 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 7. TEP responses to different TBS at T5. Topoplots show statistical comparison (t values) of TEP peaks between baseline (PreTBS) and postTBS measurements at T5 for each TBS protocol and TEP peak both in visit-1 (A) and visit-2 (B). Significant cluster of electrodes that survived permutation analyses are shown in red. Negative t values for N15 and N100 peaks represents increased amplitude (more negative peak) while positive t values represent decreased amplitudes (less negative peak). Negative t values for P30 peak represents decreased amplitude (less positive peak) and positive t values represent increased amplitude (more positive peak). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) used to probe overall neurophysiologic changes across brain re- reproducible cortical responses across sessions [44,51], the neuro- gions. Furthermore, the reliability of the TBS-induced effects on modulatory effects of TBS on TEPs were substantially variable be- cortical excitability have not been studied. tween individuals at the cortical level. Thus, both iTBS and cTBS- Here, we examined the impact and reproducibility of TBS- induced changes for global (GMFP) and local (LMFP) cortical re- induced neuromodulation directly at the cortical level by TMS- sponses were not significantly different from sham control at the EEG with a robust sham-controlled test-retest design. We found group level. When assessing the effects of TBS on individual TEP that although spTMS of the motor cortex generates stable and peaks, we found that the modulation of the N15 peaks over the Fig. 8. TEP responses to different TBS at T20. 959 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 9. Statistical comparison of active TBS effects with sham control. TBS effects for each TBS condition is calculated by subtracting PostTBS (T5 and T20) TEPs from Baseline (PreTBS) TEPs. These differences were then statistically compared between active TBS (iTBS vs Sham and cTBS vs Sham) and sham control both at T5 (A) and T20 (B) separately for both visits (Visit-1 and Visit-2). Significant cluster of electrodes that survived permutation analyses are shown in red. Negative t values for N15 and N100 peaks represents increased amplitude in active TBS as compared to sham control (more negative peak in active TBS), while positive t values represent decreased amplitudes amplitude in active TBS as compared to sham control (less negative peak in active TBS). Negative t values for P30 peak represents decreased amplitude (less positive peak in active TBS as compared to sham) and positive t values represent increased amplitude (more positive peak in active TBS as compared to sham). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) stimulated motor cortex following iTBS and cTBS, and the modu- decreased cortical excitability in Visit-2. Overall, there results lation of the N100 peaks within a large cluster of electrodes over indicate very poor reproducibility for TBS-effects on cortical frontal, premotor, and parietal regions following iTBS, were excitability in healthy individuals, and suggest that brain-behavior significantly different from Sham TBS in Visit-1. However, none of relationships established in behavioral research paradigms through the significant sham-controlled active TBS effects observed in Visit- the suggested “excitatory” or “inhibitory” mechanisms of TBS- 1 were reproduced in Visit-2. Importantly, we noticed that all TBS induced neuromodulation may need to be reevaluated. In partic- conditions also suffered from high intra-subject variability across ular, our findings strongly suggest that future studies applying TBS visits, such that a given person with increased cortical excitability with the aim of modulating human behavior and cognition need to in Visit-1 may show an opposite response pattern to TBS with employ techniques such as TMS-EEG to characterize the TBS- Fig. 10. Scalp distribution of ICCs for each TEP peak across identical TBS sessions. Higher values (>0) represent positive relationships while lower values (<0) represent negative correlations between electrode pairs across visits. 960 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 Fig. 11. Corticospinal excitability and TBS effects on cortical responses: A: Comparison of baseline MEP amplitudes as a function of early LMFP response type to TBS conditions across visits. Blue bars show baseline MEP means for individuals with facilitation response (þLMFP), while red bars show baseline MEPs means for suppressive response to (-LMFP) each TBS protocol in left-M1 at T5. Error bars represent standard error of measurements (1 unit) and horizontal lines with asterisks over the bars show significant differences (p < 0.05) in baseline MEPs between facilitation and suppression responses to TBS. B: Scatter plots, regression lines and ICCs for the relationship between pre-post TBS changes in MEPs and early LMFP both for visit-1 (blue) and visit-2 (red). Changes in MEP and early LMFP amplitudes are expressed as response ratios normalized to baseline (PreTBS) at T5. C: Scalp distribution of bivariate Pearson correlations between changes in MEP and TEP responses for N15 (upper panels) and P30 peaks (lower panels) across visits and TBS conditions. Higher values (>0) represent positive relationships while lower values (<0) represent negative correlations at the electrode level. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) induced changes in cortical activity to understand the underlying critical to note that we could also reach similar conclusions in our brain-behavior relationships. study by ignoring results from the sham TBS, as both iTBS and cTBS One critical finding of this study is the importance of repeat produced significant differences in all of our metrics (GMFP, LMFP, (test-retest) sessions to draw valid conclusions on the effects of TEP peaks and MEPs at T20) compared to baseline values. However, TBS-induced neuromodulation. To date, direct cortical responses we noticed that most of the TBS effects on cortical responses (GMFP, following cTBS of the primary motor cortex have only been LMFP and P30 at T5, p30 and N100 at T20) and MEPs were not assessed in two previous reports, using either a single electrode significant when compared to sham effects, supporting the recent [26] or a sub-group of electrodes [52] over the stimulated cortex. evidence showing no significant differences between active TBS With regard to TBS effects in other brain regions, Chung and col- and primed sham conditions [35]. leagues employed iTBS over the prefrontal cortex in a series of TMS of M1 generates unique response dynamics compared to recent studies [27,53,54], and reported significant increases in TMS of non-motor regions, as M1 stimulation activates cortico- selected TEP components. Although our results from Visit-1 are spinal pathways in parallel to cortico-cortical and thalamo-cortical partially in line with these studies, our reliability analyses showed tracts. The early local cortical activation following TMS primarily that none of these results were observed in Visit-2, demonstrating propagates through corticospinal tracts, elicits a motor response in poor reproducibility across sessions, and thus suggesting the crit- the targeted muscle, and a following sensory return to somato- ical importance of validating TBS-induced neuromodulation with sensory cortex suggesting dynamic relationships among local repeat sessions. cortical and corticospinal excitability and evoked sensory re- A large body of prior research [13,55e59] also reported signifi- sponses in the brain. Accordingly, many recent studies [60,61] re- cant modulation of corticospinal excitability following TBS pro- ported that MEPs are related with early TEP components (P25/30) tocols using MEPs as the outcome measure. A recent meta-analysis and delayed oscillatory responses following spTMS [62], and that [15] and a large-scale analysis [16] pooling data from several pre- neuromodulatory TBS interventions induces correlated changes in vious reports also supported these prior findings by showing sig- MEPs and early TEP peaks following iTBS [25,57] and cTBS [26,57]. nificant modulation of corticospinal excitability, with increased A recent study [60] suggested modulation of pre-stimulus oscilla- MEPs up to 30-min following iTBS and decreased MEPs at 5e10 or tory power in the alpha band following transcranial direct current 60-min following cTBS. Our results for MEPs and TEPs did not stimulation as a possible mechanism for correlated changes in corroborate those findings. One major factor may be the lack of MEPs and TEPs. Our preliminary results from visit-1 showing pos- sham control in the majority of the previous studies included in itive correlations between changes in MEPs and P30 component these reports. For example, out of 87 studies re-examined in Chung following iTBS and negative correlations following cTBS localized to et al.‘s meta-analysis [15] only 3 studies performed a sham control the stimulated cortex are in line with these findings; however, only in their TBS protocols, and out of 22 studies included in Corp et al.‘s the correlation for iTBS was reliable across sessions. As iTBS and large-scale analysis [16] only 1 study included a sham control. It is cTBS were previously shown to modulate different components of 961 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 descending corticospinal volleys [22,23], further studies with variety of other EEG metrics including connectivity, perturbational simultaneous epidural cervical and scalp EEG recordings would be complexity index, and time-frequency responses. Future research is highly useful to establish underlying mechanisms of correlated needed to address TBS-induced neuromodulation and its repro- changes in MEPs and TEPs following iTBS. Taken together, these ducibility with these measures. Importantly, our study is not findings suggest that interventions that increase corticospinal designed to systematically examine methodological and subject- excitability reliably increase cortical excitability as well, but that related factors and their effects on the observed high inter and the effects of other interventions may not be as consistent. Addi- intra-subject response variability to TBS. Although a number of tionally, we showed that neuromodulatory effects of iTBS on methodological (i.e, stimulation duration, intensity, coil orienta- cortical responses might be influenced by baseline corticospinal tion, time of the day), inter-subject (genetics, age, variability in excitability; however, these effects have poor reproducibility. neural circuits activated by TMS) and intra-subject (brain state, Contrary to our initial prediction, we observed significant de- prior sleep, baseline cortico-spinal excitability) factors have been viations from baseline measurements following sham TBS (See N15 reported to contribute to the response variability to TBS protocols and P30 peaks in Figs. 6 and 7). One methodological reason for this [14], no systematic evaluation have been performed to examine could be the measurement of TBS neuromodulation through their causal effects on reproducibility of TBS responses. Further- repeated blocks of spTMS over time. A recent study by Pellicciari more, significant but inconsistent cortical responses following and colleagues [63] showed that spTMS, without having any active- and sham-TBS could potentially be attributed to presence of changes in stimulation parameters, can induce cumulative in- TMS-induced sensory potentials, such as somatosensory and creases in corticospinal excitability across multiple stimulation auditory evoked potentials, in TEPs. As these non-transcranially blocks within the session. Our results from MEPs in both visits evoked potentials have large amplitudes with stereotyped confirm these findings as MEP amplitudes increased following spatial-temporal dynamics, possible spurious fluctuations across shamTBS both at T5 and T20 compared to baseline (Fig. 5A, Lower blocks or visits may confound our ability to detect TBS-induced panels). Accordingly, significant sham effects on P30 peaks we neuromodulation and its’ reproducibility. To eliminate potential observed at T20 in both visits (Fig. 6A and B) are in line with effects of sensory evoked responses in our analyses we followed a increased MEP amplitudes reported in Pellicciari et al. [63] and may structured ICA approach to carefully identify and remove these extend their findings to cumulative effects of spTMS to direct non-transcranial responses from TEPs (Fig. 2). We then showed cortical responses. Another important aspect to consider for sig- reproducibility of baseline TEPs (Fig. 3) across sessions, suggesting nificant sham responses would be the possible placebo effects of that even if any residual sensory-evoked potentials remain in our TBS. A growing body of evidence suggests that the magnitude of post-processed TEPs, they are stable across sessions. However, to sham effects for therapeutic applications of rTMS in different clearly understand and eliminate any possible confounding effects clinical populations has considerably increased over the years, of TBS on sensory-evoked potentials, future studies could integrate possibly due to growing awareness on the potential of TMS in single pulse sham-TMS designs into their TBS protocols. In the public and media, adoption of more realistic sham protocols, and current study, a sham control for spTMS was not feasible for our sophisticated experimental set-ups with neuronavigated TMS-EEG post TBS measurements, since we could only administer either designs [64]. An alternative and highly plausible explanation would active spTMS or sham spTMS at any one time point (e.g. T5) be mutually interacting effects of repeated blocks of spTMS and following TBS. Given the effects of TBS are not stable over time; it placebo effects of TBS. Given the lack of experimental evidence, would not be valid to compare the effects of iTBS on real spTMS at understanding possible neurophysiological and environmental T5 with the effects on sham spTMS at T15. When advances in TMS mechanisms of significant sham responses in TMS research clearly technology permit interleaved sham and real spTMS, sham spTMS warrants future systematic investigations. stimulation should be included in future studies to parcel out the Certain limitations of the current study should also be simultaneous effects of TBS on the transcranial and non- acknowledged. First, our results are limited to stimulation of the transcranial sensory-evoked components. Finally, as in most of primary motor cortex, and thus do not provide direct evidence for the previous literature, the experimenters were not blinded to the reproducibility of TBS-induced neuromodulation over non-motor applied stimulation protocol. Although currently it is not a standard cortices. Understanding direct cortical responses to TBS and, more experimental design in basic neurophysiological research, future importantly, examining reproducibility of such cortical responses studies could clearly benefit from double blind methods to mini- outside the motor cortex is essential for validating previously mize the risk of experimenter bias and ensure the robustness of suggested brain-behavior relationships through the mechanism of replication designs. TBS-induced neuromodulation. Secondly, our results do not generalize to other widely used rTMS protocols. In general, low Conclusions and future directions frequency rTMS (1 Hz) and high frequency rTMS (10 Hz, 20 Hz) have been used as “inhibitory” and “excitatory” protocols respec- After almost 30 years of experimental research, the large tively to modulate brain activity [65,66]. However, the effects of response variability to rTMS protocols still stands as a major these rTMS protocols on cortical excitability as assessed via EEG, obstacle for the clinical utility and experimental validity of rTMS- and the reproducibility of rTMS-induced neuromodulatory effects induced neuromodulation. This study reports the first compre- against a robust sham control have not been clearly established. hensive examination of (1) the direct cortical EEG responses to iTBS Thirdly, our results are limited to analyses of TEPs, and thus do not and cTBS of human motor cortex and (2) the reproducibility of both generalize to other measures of cortical responsivity and modula- iTBS- and cTBS-induced modulation of cortical excitability against a tion. A series of recent studies [27,53,67], for example, reported sham control across repeat sessions. We find that the neuro- significant TBS-induced modulations in TMS-evoked oscillatory modulatory effects of TBS are substantially variable both between responses in the frequency domain, but did not evaluate the and within individuals, with no significant group effects over sham reproducibility of these oscillatory responses. In this study, we control, and with very poor reproducibility across sessions. Our limited our analyses to TEPs, GMFPs and LMFPs in the temporal results suggest three key messages for future studies. First, a sham domain, as they are the most commonly used EEG metrics for control has to be an integral part of any experimental TBS design to measuring cortical response modulation in the current TMS-EEG rule out cumulative effects of spTMS, as well as to control possible literature. Clearly, cortical responsivity can be measured with placebo effects of rTMS interventions. Secondly, studies applying 962 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 TBS with the goal of establishing brain-behavior relationships References through the mechanisms of TBS neuromodulation should validate expected TBS effects by measuring direct cortical (EEG) responses [1] Ziemann U, Muellbacher W, Hallett M, Cohen LG. Modulation of practice- dependent plasticity in human motor cortex. Brain 2001;124:1171e81. to TBS. Finally, significant neuromodulatory and behavioral effects [2] Klomjai W, Katz R, Lackmy-Valle e A. Basic principles of transcranial magnetic of TBS over sham control should be confirmed with either repeat stimulation (TMS) and repetitive TMS (rTMS). Ann Phys Rehab Med 2015;58: sessions or independent data sets to establish the reproducibility 208e13. [3] Lefaucheur J-P, et al. Evidence-based guidelines on the therapeutic use of and robustness of observed effects. Further randomized controlled repetitive transcranial magnetic stimulation (rTMS). Clin Neurophysiol trials with double blind methods are urgently needed to minimize 2014;125:2150e206. the risk of bias and unravel the underlying reasons for such sub- [4] Di Lazzaro V, et al. Modulation of motor cortex neuronal networks by rTMS: comparison of local and remote effects of six different protocols of stimula- stantial response variability to rTMS by experimentally evaluating tion. J Neurophysiol 2011;105:2150e6. the causal role of possible methodological, inter- and intra-subject [5] Hilgetag CC, The oret H, Pascual-Leone A. Enhanced visual spatial attention factors on rTMS-induced neuromodulation. ipsilateral to rTMS-induced’virtual lesions' of human parietal cortex. Nat Neurosci 2001;4:953e7. [6] Tavor I, et al. Task-free MRI predicts individual differences in brain activity CRediT authorship contribution statement during task performance. Science 2016;352:216e20. [7] Wang JX, et al. Targeted enhancement of cortical-hippocampal brain networks and associative memory. Science 2014;345:1054e7. Recep A. Ozdemir: Methodology, conceptualized the frame- [8] Ziemann U, et al. Consensus: motor cortex plasticity protocols. Brain Stimul work, Data curation, preprocessed the TMS-EEG and EMG data. 2008;1:164e82. Pierre Boucher: Data curation, preprocessed the TMS-EEG and [9] Blumberger DM, et al. Effectiveness of theta burst versus high-frequency re- petitive transcranial magnetic stimulation in patients with depression EMG data. Peter J. Fried: Methodology. Davide Momi: Data cura- (THREE-D): a randomised non-inferiority trial. Lancet 2018;391:1683e92. tion. Ali Jannati: performed statistical analyses for reproducibility [10] Sachdev PS, Loo CK, Mitchell PB, McFARQUHAR TF, Malhi GS. Repetitive of GMFP and LMFPs, oversaw study conduction and edited the first transcranial magnetic stimulation for the treatment of obsessive compulsive disorder: a double-blind controlled investigation. Psychol Med 2007;37:1645. draft. Alvaro Pascual-Leone: Methodology, oversaw study con- [11] Brighina F, et al. rTMS of the prefrontal cortex in the treatment of chronic duction and edited the first draft. Emiliano Santarnecchi: Meth- migraine: a pilot study. J Neurol Sci 2004;227:67e71. odology. Mouhsin M. Shafi: Methodology, conceptualized the [12] Larson J, Lynch G. Theta pattern stimulation and the induction of LTP: the sequence in which synapses are stimulated determines the degree to which framework, oversaw study conduction and edited the first draft, All they potentiate. Brain Res 1989;489:49e58. authors critically reviewed the manuscript for content and [13] Huang Y-Z, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. Theta burst stim- approved the final version for publication. ulation of the human motor cortex. Neuron 2005;45:201e6. [14] Suppa A, et al. Ten years of theta burst stimulation in humans: established knowledge, unknowns and prospects. Brain Stimul 2016;9:323e35. Declaration of competing interest [15] 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. Neurosci Biobehav Rev 2016;63:43e64. Dr. A. Pascual-Leone is a co-founder of Linus Health and TI So- [16] Corp DT, et al. Large-scale analysis of interindividual variability in theta-burst lutions AG; serves on the scientific advisory boards for Starlab stimulation data: results from the ‘Big TMS Data Collaboration’. Brain Stimul Neuroscience, Neuroelectrics, Magstim Inc., Nexstim, Cognito, and 2020;13(5):1476e88. https://doi.org/10.1016/j.brs.2020.07.018. [17] Lowe CJ, Manocchio F, Safati AB, Hall PA. The effects of theta burst stimulation MedRhythms; and is listed as an inventor on several issued and (TBS) targeting the prefrontal cortex on executive functioning: a systematic pending patents on the real-time integration of noninvasive brain review and meta-analysis. Neuropsychologia 2018;111:344e59. stimulation with electroencephalography and magnetic resonance [18] Demirtas-Tatlidede A, Vahabzadeh-Hagh AM, Pascual-Leone A. Can noninva- sive brain stimulation enhance cognition in neuropsychiatric disorders? imaging. Neuropharmacology 2013;64:566e78. [19] Maier MJ, et al. Forgiveness and cognitive controleProvoking revenge via theta-burst-stimulation of the DLPFC. Neuroimage 2018;183:769e75. Acknowledgment [20] Verbruggen F, Aron AR, Stevens MA, Chambers CD. Theta burst stimulation dissociates attention and action updating in human inferior frontal cortex. We are grateful to all research assistants who helped to run these Proc Natl Acad Sci Unit States Am 2010;107:13966e71. [21] Stefan K, Gentner R, Zeller D, Dang S, Classen J. Theta-burst stimulation: study visits. We are grateful to the gracious funding from the MIT- remote physiological and local behavioral after-effects. Neuroimage 2008;40: Harvard Broad institute (6600024e5500000895) directly support- 265e74. ing the study and our line of research on brain plasticity biomarkers. [22] Di Lazzaro V, et al. Theta-burst repetitive transcranial magnetic stimulation suppresses specific excitatory circuits in the human motor cortex. J Physiol Dr. Santarnecchi is supported by the Beth Israel Deaconess Medical 2005;565:945e50. Center (BIDMC) via the Chief Academic Officer (CAO) Award 2017, the [23] Di Lazzaro V, et al. The physiological basis of the effects of intermittent theta Defence Advanced Research Projects Agency (DARPA) via burst stimulation of the human motor cortex. J Physiol 2008;586:3871e9. [24] Hill AT, Rogasch NC, Fitzgerald PB, Hoy KE. TMS-EEG: a window into the HR001117S0030, the NIH (P01 AG031720-06A1, R01 MH117063-01, neurophysiological effects of transcranial electrical stimulation in non-motor R01 AG060981-01) and ADDF (ADDF-FTD GA201902e2017902). Dr. brain regions. Neurosci Biobehav Rev 2016;64:175e84. Pascual-Leone is supported by grants from the National Institutes of [25] Gedankien T, Fried PJ, Pascual-Leone A, Shafi MM. Intermittent theta-burst Health (R24AG06142, and P01 AG031720), the National Science stimulation induces correlated changes in cortical and corticospinal excit- ability in healthy older subjects. Clin Neurophysiol 2017;128:2419e27. Foundation, and the Barcelona Brain Health Initiative funded pri- [26] Vernet M, et al. Insights on the neural basis of motor plasticity induced by marily by La Caixa. Dr. Shafi is supported by the Football Players theta burst stimulation from TMSeEEG. Eur J Neurosci 2013;37:598e606. Health Study (FPHS) at Harvard University, and the NIH (R01 [27] Chung SW, et al. Demonstration of short-term plasticity in the dorsolateral prefrontal cortex with theta burst stimulation: a TMS-EEG study. Clin Neu- MH115949, R01AG060987, P01 AG031720-06A1). The content of this rophysiol 2017;128:1117e26. paper is solely the responsibility of the authors and does not [28] Ziemann U, Siebner HR. Inter-subject and inter-session variability of plasticity necessarily represent the official views of Harvard University and its induction by non-invasive brain stimulation: boon or bane? Brain Stimul 2015;8:662e3. affiliated academic health care centres, or the National Institutes of [29] Huang Y-Z, et al. Plasticity induced by non-invasive transcranial brain stim- Health. ulation: a position paper. Clin Neurophysiol 2017;128:2318e29. [30] Hamada M, Murase N, Hasan A, Balaratnam M, Rothwell JC. The role of interneuron networks in driving human motor cortical plasticity. Cerebr Appendix A. Supplementary data Cortex 2013;23:1593e605. [31] Lopez-Alonso V, Cheeran B, Río-Rodríguez D, Ferna ndez-del-Olmo M. Inter- individual variability in response to non-invasive brain stimulation para- Supplementary data to this article can be found online at digms. Brain Stimul 2014;7:372e80. https://doi.org/10.1016/j.brs.2021.05.013. 963 R.A. Ozdemir, P. Boucher, P.J. Fried et al. Brain Stimulation 14 (2021) 949e964 [32] Lowe CJ, Hall PA. Reproducibility and sources of interindividual variability in [50] Koch G, et al. Theta-burst stimulation of the left hemisphere accelerates re- the responsiveness to prefrontal continuous theta burst stimulation (cTBS). covery of hemispatial neglect. Neurology 2012;78:24e30. Neurosci Lett 2018;687:280e4. [51] Kerwin LJ, Keller CJ, Wu W, Narayan M, Etkin A. Test-retest reliability of [33] Schilberg L, Schuhmann T, Sack AT. Interindividual variability and intra- transcranial magnetic stimulation EEG evoked potentials. Brain Stimul individual reliability of intermittent theta burst stimulation-induced neuro- 2018;11:536e44. plasticity mechanisms in the healthy brain. J Cognit Neurosci 2017;29: [52] Rocchi L, et al. Variability and predictors of response to continuous theta burst 1022e32. stimulation: a TMS-EEG study. Front Neurosci 2018;12. [34] Vallence A-M, et al. Inter-and intra-subject variability of motor cortex plas- [53] Chung SW, et al. The effects of individualised intermittent theta burst stim- ticity following continuous theta-burst stimulation. Neuroscience 2015;304: ulation in the prefrontal cortex: a TMS-EEG study. Hum Brain Mapp 2019;40: 266e78. 608e27. [35] Perello n-Alfonso R, et al. Similar effect of intermittent theta burst and sham [54] Chung SW, Rogasch NC, Hoy KE, Fitzgerald PB. The effect of single and stimulation on corticospinal excitability: a 5-day repeated sessions study. Eur repeated prefrontal intermittent theta burst stimulation on cortical reactivity J Neurosci 2018;48:1990e2000. and working memory. Brain Stimul 2018;11:566e74. [36] Rothwell JC, et al. Magnetic stimulation: motor evoked potentials. Electro- [55] Huang Y-Z, Rothwell JC, Edwards MJ, Chen R-S. Effect of physiological activity encephalogr Clin Neurophysiol Suppl 1999;52:97e103. on an NMDA-dependent form of cortical plasticity in human. Cerebr Cortex [37] Rossini PM, et al. Non-invasive electrical and magnetic stimulation of the 2008;18:563e70. brain, spinal cord, roots and peripheral nerves: basic principles and proced- [56] Gentner R, Wankerl K, Reinsberger C, Zeller D, Classen J. Depression of human ures for routine clinical and research application. An updated report from an corticospinal excitability induced by magnetic theta-burst stimulation: evi- IFCN Committee. Clin Neurophysiol 2015;126:1071e107. dence of rapid polarity-reversing metaplasticity. Cerebr Cortex 2008;18: [38] Rossi S, Hallett M, Rossini PM, Pascual-Leone A, Group, S. of T. C. Safety, ethical 2046e53. considerations, and application guidelines for the use of transcranial magnetic [57] Vernet M, et al. Reproducibility of the effects of theta burst stimulation on stimulation in clinical practice and research. Clin Neurophysiol 2009;120: motor cortical plasticity in healthy participants. Clin Neurophysiol 2014;125: 2008e39. 320e6. [39] Rossi S, et al. Safety and recommendations for TMS use in healthy subjects and [58] Jannati A, Block G, Oberman LM, Rotenberg A, Pascual-Leone A. Interindi- patient populations, with updates on training, ethical and regulatory issues: vidual variability in response to continuous theta-burst stimulation in healthy expert Guidelines. Clin Neurophysiol 2021;132(1):269e306. https://doi.org/ adults. Clin Neurophysiol 2017;128:2268e78. 10.1016/j.clinph.2020.10.003. [59] Fried PJ, et al. Humans with type-2 diabetes show abnormal long-term [40] Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single- potentiation-like cortical plasticity associated with verbal learning deficits. trial EEG dynamics including independent component analysis. J Neurosci J Alzheim Dis 2017;55:89e100. Methods 2004;134:9e21. [60] Ahn S, Fro €hlich F. Pinging the brain with transcranial magnetic stimulation [41] Rogasch NC, et al. Analysing concurrent transcranial magnetic stimulation and reveals cortical reactivity in time and space. Brain Stimul 2021;14(2):304e15. electroencephalographic data: a review and introduction to the open-source https://doi.org/10.1016/j.brs.2021.01.018. TESA software. Neuroimage 2017;147:934e51. [61] M€aki H, Ilmoniemi RJ. The relationship between peripheral and early cortical [42] Nettekoven C, et al. Inter-individual variability in cortical excitability and activation induced by transcranial magnetic stimulation. Neurosci Lett motor network connectivity following multiple blocks of rTMS. Neuroimage 2010;478:24e8. 2015;118:209e18. [62] Fecchio M, et al. The spectral features of EEG responses to transcranial mag- [43] McGraw KO, Wong SP. Forming inferences about some intraclass correlation netic stimulation of the primary motor cortex depend on the amplitude of the coefficients. Psychol Methods 1996;1:30. motor evoked potentials. PloS One 2017;12:e0184910. [44] Ozdemir RA, et al. Individualized perturbation of the human connectome [63] Pellicciari MC, Brignani D, Miniussi C. Excitability modulation of the motor reveals reproducible biomarkers of network dynamics relevant to cognition. system induced by transcranial direct current stimulation: a multimodal Proc Natl Acad Sci Unit States Am 2020. https://doi.org/10.1073/ approach. Neuroimage 2013;83:569e80. pnas.1911240117. [64] Burke MJ, Kaptchuk TJ, Pascual-Leone A. Challenges of differential placebo [45] Portney LG, Watkins MP. Foundations of clinical research: applications to effects in contemporary medicine: the example of brain stimulation. Ann practice, vol. 892. Upper Saddle River, NJ: Pearson/Prentice Hall; 2009. Neurol 2019;85:12. [46] Wischnewski M, Schutter DJ. Efficacy and time course of theta burst stimu- [65] Fitzgerald PB, Fountain S, Daskalakis ZJ. A comprehensive review of the effects lation in healthy humans. Brain Stimul 2015;8:685e92. of rTMS on motor cortical excitability and inhibition. Clin Neurophysiol [47] Rounis E, Maniscalco B, Rothwell JC, Passingham RE, Lau H. Theta-burst 2006;117:2584e96. transcranial magnetic stimulation to the prefrontal cortex impairs meta- [66] Fried PJ, Elkin-Frankston S, Rushmore RJ, Hilgetag CC, Valero-Cabre A. Char- cognitive visual awareness. Cognit Neurosci 2010;1:165e75. acterization of visual percepts evoked by noninvasive stimulation of the hu- [48] Debarnot U, et al. Intermittent theta burst stimulation over left BA10 en- man posterior parietal cortex. PloS One 2011;6:e27204. hances virtual reality-based prospective memory in healthy aged subjects. [67] Chung SW, et al. Impact of different intensities of intermittent theta burst Neurobiol Aging 2015;36:2360e9. stimulation on the cortical properties during TMS-EEG and working memory [49] Talelli P, Greenwood RJ, Rothwell JC. Exploring Theta Burst Stimulation as an performance. Hum Brain Mapp 2018;39:783e802. intervention to improve motor recovery in chronic stroke. Clin Neurophysiol 2007;118:333e42. 964