Development of Non-Invasive Brain Stimulation Techniques

Objectives

Noninvasive brain stimulation (NIBS) using magnetic and electrical means, through the use of transcranial magnetic stimulation (TMS) and transcranial direct and alternating current stimulation (tDCS and tACS respectively), has proven to be a versatile tool in the investigation of cortical organization and function, and has shown great potential as a means to diagnose and treat psychiatric and neurological illness. Unfortunately, this research has been slowed due to two fundamental problems that directly impact the usefulness of NIBS. First, the space of parameters used to produce NIBS is immensely large and has barely been explored. Such parameters include intensity of stimulation, pulse wave form and duration, number and timing of pulses (which in turn can involve trains of pulses: their frequency, duration, number, and intertrain intervals), the shape of the magnetic or electric field produced (for TMS, meaning type of coil used, its position, and its orientation (with three degrees of freedom for a figure eight coil), and for tDCS and tACS, the numbers, sizes and placements of electrodes), and, for creating long-lasting effects, the number and timing of NIBS sessions. Differences- often even very small differences- in the particular values of any of these parameters have been shown to have significant impact on the size and duration of NIBS effects, and yet little is known. Parametric exploration has only been addressed in a limited way in the field, and primarily only in motor cortex. Second, the interaction of any set of NIBS parameters with any individual brain is poorly understood, and this has led to unpredictable efficacy and a large amount of inter-individual variability in NIBS studies. The human brain is always in a complex, dynamic state of change, and it has become increasingly clear that the state of a cortical region and its associated network connections when NIBS is applied strongly influences its effects There are now a great many demonstrations of how the state of the cortical region being stimulated plays a large role in determining what specific NIBS effect occurs. Not knowing the particular state of a brain when stimulating it has led to wide variability and unpredictability in NIBS effects. Thus while the variable state of the target region in each subject is likely an important source of the large amount of inter-individual variability found in TMS and tDCS studies, little research has been performed as yet. This poses a huge challenge to the effectiveness of NIBS therapeutic interventions on an individual and precision medicine basis, to the design and implementation of brain stimulation studies and to the test-retest reliability of the neurophysiological and behavioral measures used (and their optimization).

The purpose of this technical development protocol is to address these two fundamental deficiencies in NIBS research. To do so, we would like to 1) explore, within safe ranges as established by international consensus, the effects of different sets of NIBS stimulation parameters on behavioral and cortical function, as measured by behavioral performance, electroencephalography (EEG), electromyography (EMG), eye movements, and MRI. and, 2) to develop the methodologies and analytic techniques behind those behavioral and physiological measures in order to increase their usefulness in interpreting NIBS effects. In addition, this protocol will be used to train new fellows coming to the Non-invasive Neuromodulation Unit (NNU) in the use of TMS techniques. The NNU is exceptionally well-situated, both in terms of expertise and resources, to perform this protocol. We expect that information emerging from these studies will allow us to 1) optimize NIBS effects across ranges of stimulation parameters and brain states, within individual subjects, and in terms test-retest reliability, 2) to collect pilot data in healthy volunteers to establish feasibility and for power analysis for future patient-oriented hypothesis- driven protocols, and 3) to train new fellows in the use of these different methods.

We specifically propose to begin this development protocol with five substudies that span the NIBS areas in need of further research, examining stimulus intensity, stimulus timing and frequency, stimulus targeting, and methods for controlling brain state. These explorations will provide a useful initial foray into NIBS parameter space, and will in addition help develop means to more effectively dose TMS outside of motor cortex and reduce variability in TMS effects, to test new methods of targeting TMS, to develop the usefulness of EEG in TMS, and to understand the relationship of TMS timing and endogenous oscillations.

The protocol consists of the below substudies:

Substudy 1 (Using TMS-evoked potential (TEP) to dose TMS outside of the motor cortex)

This study aims to begin an exploration of the feasibility of using EEG to evaluate the response to single pulses of TMS over differing cortical brain areas, as a novel method of individualizing the dose of TMS in a site-specific fashion. The local cortical response represents an evoked wave of neuronal activity in the immediate vicinity of the stimulating coil and could provide a useful dosing measure outside the motor cortex. In this first step, at two cortical locations (motor cortex and occipital visual cortex) we explore the relationship of TMS evoked potentials with functional evoked responses (MEPs from motor cortex, visual discrimination in visual cortex) and with fMRI BOLD response to TMS at different intensities applied to both cortical locations.

Substudy 2 (The value of electric field modeling in TMS localization)

The study question for this substudy is as follows: Does inclusion of electric field modeling to TMS targeting increase the efficacy of TMS?

The overall goal is to compare an electric field guided coil placement method with (f)MRI-guided and scalp-guided NIBS targeting approaches. Including the latter condition additionally allows for the substudy to address the only study that compared scalp-based and fMRI-based targeting (Sack et al., 2009). The Sack et al. study was designed as between groups, using very small Ns (of 5): the NIBS field would greatly benefit from a replication with a larger group and a within-subjects design. In this first substudy, we will test the efficacy of three E-field modeling approaches to TMS targeting, with the intention of using the best method in a subsequent substudy to compare TMS targeting methods.

Substudy 3 (Controlling ongoing cortical state during NIBS with neurofeedback)

The study question for this substudy is as follows: Does controlling brain states using neurofeedback result in greater intra- and inter-individual variability in responses to TMS?

This study aims to explore the use of real-time fMRI neurofeedback to control the activated state of the amygdala while it is targeted transynaptically with short trains of 5 Hz rTMS.

Substudy 4 (Using cTMS EEG to understand mechanism and to optimize theta-burst stimulation (TBS))

The study question for this substudy is as follows: What are the effects of TBS on EMG across different waveforms of TMS pulses?

In this substudy, we aim to parametrically examine the effects of intertrial interval on TBS response, and to harness the flexible control of pulse parameters using the cTMS device (Rogue Research Inc., Canada) to investigate the effects of TBS.

Substudy 5 (using TMS as a probe of the fronto-striatal network, a key circuit implicated in reward processing)

We seek to establish whether TMS can reach, in a dose-dependent way, a targeted subcortical region transynaptically which is too distant for effective stimulation directly. Specifically, we will test whether TMS, delivered to the dorsolateral prefrontal cortex (DLPFC) or the pre-supplementary motor area (pre-SMA), results in changes in activation of the striatum, in a dose- dependent fashion. This hypothesis will be tested with the perturbation-imaging procedure of TMS/fMRI interleaving, i.e. while participants receive TMS during both resting state and task based fMRI.

Study Population

Up to 180 healthy volunteers, age 18 and older. The number is the sum of requested participants (N=25) in the 5 substudies, with an additional 25 given an anticipated 20% drop-out rate.

Design

Healthy adult volunteers will participate typically using repeated measures design, given the goals of exploring parametric NIBS effects and to establish reliability in those effects, although between groups designs might be required in some cases (e.g., when learning is involved). Experiments will be carried out in two phases: with Phase I involving screening and baseline procedures and Phase II being the experimental NIBS sessions. Phase I will include consenting and screening. It will generally include an MRI session due to the need for at least a structural MRI to be used for neuronavigation in NIBS targeting Phase I may also include introduction and training in some behavioral task or tasks, as well as baseline measures of, EEG, and EMG, as needed. In Phase II the experimental NIBS sessions will be performed with each participant. The number of sessions will be variable, in general between 1 and 4, lasting about 1-3 hours each, depending on the exploratory question. NIBS may be given in conjunction with EEG, EMG, and/or fMRI, either simultaneously or in a pre-post manner. The NIBS stimulation parameters will never exceed safe ranges, as established in international consensus safety reports. Experimental control comparisons may be between sham and active NIBS, and/or in the latter case being NIBS to different scalp locations or at different times, or in the case of parametric studies, differing values of the parameter in question.

The experience in the NIBS field over the past three decades provides no evidence that there is an upper limit on the number of NIBS sessions an individual can safely participate in. This conclusion is primarily based on the many thousands of individuals who have received TMS treatment for depression according to the FDA-cleared labeling, in which six weeks of daily weekday treatment (30 sessions involving 3000 pulses of rTMS per session) are provided, with no new, unexpected adverse events reported. Thus, the amount of participation will only be limited in two ways: first, subjects participate in one experimental session per day. A single session, which generally lasts between 1-3 hours, may last no longer than 8 hours to allow for the initial testing paradigm followed by retests or performing other components of the same substudy later in the day, with appropriate rest breaks and meal breaks during long sessions. Second, subjects can participate in only one substudy at a time.

Outcome measures:

Substudy 1: MEP and TEP amplitudes and latencies; fMRI BOLD changes; visual discrimination threshold

Substudy 2: Pre-post change in behavioral performance (RT and accuracy) and in parietal fMRI BOLD response across conditions.

Substudy 3: Individual and group BOLD fMRI signal in the amygdala, and task-related functional connectivity changes between the amygdala and cortical networks associated with emotion processing.

Substudy 4: Amplitude and latency of MEP.

Substudy 5: change in fMRI BOLD response in the striatum as the result of surface cortical TMS.