Spectral TMS (s-TMS)

The therapeutic benefit of repetitive Transcranial Magnetic Stimulation (rTMS) for the treatment of Major Depressive Disorder (MDD) is thought to depend upon engagement of brain functional networks (FNs). Engagement is dependent upon an interaction between the rTMS stimulation frequency and the preferred oscillatory frequency of the target network for that individual. We report here on a novel method to interrogate the left dorsolateral prefrontal cortex (DLPFC) treatment target to identify the optimal frequency for engagement of the frontoparietal control network (FCN) for each individual. 23 subjects with MDD were stimulated with 40 pulse rTMS trains at frequencies varying from 3-17 Hz at 0.2 Hz increments, plus a separate intermittent theta burst (iTBS) stimulation, for a total of 70 frequency interrogations. FN engagement was assessed using continuous high-density array TMS-EEG recordings and measurement of the change in the density of the spectral correlation coefficient (SCC) between left DLPFC and other brain areas following each interrogation. All subjects had one or more frequency bands that showed high SCC values across multiple FNs, regardless of the frequency of interrogation. Those subjects who showed clinical response to 10 Hz rTMS showed increases in SCCD over FCN connections across in response to 10 Hz interrogation, but not in response to interrogation in other bands. There was a strong association between the increase in SCCD in FCN and degree of improvement in depressive symptoms after both 2 and 6 weeks of treatment. These findings suggest that SCC density may be useful for identifying rTMS stimulation frequencies that are associated with therapeutic benefit in MDD. This study will examine the feasibility using these findings to determine if SCC-determined rTMS treatment can benefit subjects with Major Depressive Disorder.

INTRODUCTION AND RATIONALE

Repetitive Transcranial Magnetic Stimulation (rTMS) has been reported to relieve symptoms of Major Depressive Disorder (MDD) when administered at a number of different stimulation frequencies: 1, 5, 10, 15, 18, and 50 Hz (theta burst stimulation, or TBS) all have reported to provide clinical benefit. It is difficult to compare the efficacy of these different frequencies of stimulation because of differences in study design and limitations of sample size, but there is significant evidence of efficacy for each of these frequencies when applied to one or more stimulation targets.1 There have not been head-to-head studies to compare different stimulation frequencies in individual subjects to determine whether individuals have similar clinical responses to more than one frequency of stimulation.

One feature that is shared by all frequencies of rTMS is that while stimulation is ongoing, both neuronal spiking and ongoing oscillations are synchronized to the frequency of stimulation (Fröhlich and McCormick, 2010 ; Thut et al., 2011 ). This modulation of cerebral oscillatory activity is hypothesized to underlie the therapeutic effects of rTMS for MDD. , Entrainment of brain oscillations has been shown to modify brain activity in a variety of functional networks (FNs), with changes in oscillations associated with alterations in task performance, local oscillatory activity, connectivity patterns and in the case of motor networks, corticospinal coupling.

Different frequencies of rTMS stimulation have distinct effects on FN engagement and task performance. These differential effects are consistent with the fact that FNs have one or more preferred resonant frequencies that mediate connectivity both within and across networks (Hacker et al., 2017 , , , The effects of rTMS stimulation depends upon the interaction between the frequency of stimulation and the ongoing oscillatory activity of the target FN(s): the oscillatory pattern of the specific brain region being stimulated (as measured with electroencephalography [EEG]) is a major determinant of rTMS engagement with the associated FN(s)3, and the effects on network performance.

The frontoparietal control network (FCN) is of particular interest in MDD because it is dysregulated in MDD, with the degree of dysfunction related to severity of depressive symptoms. The most commonly used neuroanatomic rTMS stimulation target is left dorsolateral prefrontal cortex (DLPFC), which is a critical hub of the FCN. Furthermore, the FCN plays a crucial role integrating the function of multiple other FNs. , , 10 Hz is the most commonly rTMS stimulation used to stimulate the left DLPFC target, but it has not been established that this frequency is optimal for engaging the FCN. Multiple stimulation frequencies have differential effects on the FCN, changing FCN interactions with default mode network (DMN) from excitatory to inhibitory. This finding is consistent with the fact that adjusting the frequency of stimulation affects which nodes within a single FN are engaged, , the extent to which rTMS stimulation engages local vs. distant network modules, and selectively alters the engagement between a FN and an affiliated brain region. FCN connectivity is known to be mediated by multiple frequencies ranging from 2 Hz (delta band) though 20 Hz (beta band) (Johnson et al., 2019 ; Lopez et al., 2019 ; Cooper et al., 2015 ; 18).

The present study was performed to determine whether it was possible to identify differential effects of rTMS treatment frequency on engagement of the FCN during rTMS treatment of MDD, through examination of frequency coupling across a broad spectrum of stimulation frequencies within the FCN. It was recently reported that changes in whole-brain spectral connectivity in the frequency band (alpha spectral correlation, or SC) from pre- to post- the first session of 10 Hz rTMS applied to left DLPFC predicted outcome from a course of 30 rTMS treatments. In the current study, we examined spectral correlation across a broad frequency spectrum following a series of rTMS "interrogations" ranging from 3 Hz to TBS. This allowed us to generate a spectral correlation coefficient (SCC) across this frequency spectrum and determine which frequencies of stimulation would lead to increased SCC within the FCN. This study hypothesizes that: 1) each subject's frequency response pattern would show one or more distinct frequencies of stimulation that generated increased SCC; 2) these frequency response patterns would differ across subjects, but be highly reproducible within subjects over a course of rTMS treatment; and, 3) those subjects with concentrated increases in FCN SCC (increased SCC density) following 10 Hz interrogation would be more likely to show clinical response to 10 Hz rTMS treatment.

STUDY OBJECTIVE

The primary objective of this study is to demonstrate feasibility of rTMS treatment for depression at stimulation frequencies other than 10 Hz. Objectives:

1. to determine if each subject's frequency response pattern shows one or more distinct frequencies of stimulation that generated increased SCC;
2. to determine if these frequency response patterns differ across subjects; and,
3. To gather pilot data for an extramural funding application to investigate the efficacy of rTMS treatment at individualized stimulation frequencies as an alternative to the standard 10 Hz protocol.