Using Brain Imaging to Improve Spatial Targeting of Transcranial Magnetic Stimulation for Depression

Abstract

Transcranial magnetic stimulation (TMS) is an effective treatment for depression but is limited in that the optimal therapeutic target remains unknown. Early TMS trials lacked a focal target and thus positioned the TMS coil over the prefrontal cortex using scalp measurements. Over time, it became clear that this method leads to variation in the stimulation site and that this could contribute to heterogeneity in antidepressant response. Newer methods allow for precise positioning of the TMS coil over a specific brain location, but leveraging these precise methods requires a more precise therapeutic target. We review how neuroimaging is being used to identify a more focal therapeutic target for depression. We highlight recent studies showing that more effective TMS targets in the frontal cortex are functionally connected to deep limbic regions such as the subgenual cingulate cortex. We review how connectivity might be used to identify an optimal TMS target for use in all patients and potentially even a personalized target for each individual patient. We address the clinical implications of this emerging field and highlight critical questions for future research.

Proof of concept study to develop a novel connectivity-based electric-field modelling approach for individualized targeting of transcranial magnetic stimulation treatment

“…Our model adds to the work suggesting that rsFC networks can inform TMS targeting (29), and our results are consistent with the findings that individualized rsFC maps may be most informative (30). Previous research has shown that regions of the dlPFC that are more strongly anticorrelated with sgACC tend to show better clinical efficacy when targeted with therapeutic TMS (51, 56), and the closer the stimulation site is to this optimally anticorrelated dlPFC location, the better the clinical outcome (54, 55, 57). Importantly, active rTMS to the l dlPFC has been shown to reduce anticorrelation between the dlPFC and sgACC (53, 58, 59), and prospective targeting based on individualized dlPFC – sgACC connectivity leads to a greater reduction in symptoms compared to traditional targeting (53, 54).…”

Individualized precision targeting of dorsal attention and default mode networks with rTMS in traumatic brain injury-associated depression

“…To improve upon these group-level targets, individualized connectivity measurements have also been used to identify patient-specific stimulation sites [8][9][10] . These targets may be superior to normative "anti-group mean sgACC" targets 6,7,11 .…”

“…Recent studies have attempted to identify rTMS targets based on functional connectivity (FC) with "seed" regions deeper in the brain 3 . At the group level, antidepressant efficacy of rTMS is related to normative anti-correlation between the stimulation target and subgenual anterior cingulate cortex (sgACC), suggesting that treatment may be suppressing activity in sgACC and the limbic system [4][5][6][7] . To improve upon these group-level targets, individualized connectivity measurements have also been used to identify patient-specific stimulation sites [8][9][10] .…”

The Use of Tractography-Based Targeting in Deep Brain Stimulation for Psychiatric Indications

“…The advantages of large normative datasets include an improved signal-to-noise ratio (SNR), potential use of state-of-the art equipment (i.e., Human Connectome Project) ( Van Essen et al, 2013 ), and allowing for a more universal scientific language and comparison between centers ( Horn and Blankenburg, 2016 ). However, given the considerable inter-individual variability in brain structure and connectivity, applying normative imaging at the patient level might prevent the ability to “personalize” neuromodulation treatments ( Fox et al, 2013 ; Cash et al, 2020 ). For instance, some authors have suggested that the inter-individual variability of tract positioning within the ALIC will require subject-specific high-resolution DTI in order to personalize targets ( Makris et al, 2016 ; Nanda et al, 2017 ).…”