Predictive modeling of response to repetitive transcranial magnetic stimulation in treatment-resistant depression.

Identifying predictors of treatment response to repetitive transcranial magnetic stimulation (rTMS) remain elusive in treatment-resistant depression (TRD). Leveraging electronic medical records (EMR), this retrospective cohort study applied supervised machine learning (ML) to sociodemographic, clinical, and treatment-related data to predict depressive symptom response (>50% reduction on PHQ-9) and remission (PHQ-9 < 5) following rTMS in 232 patients with TRD (mean age: 54.5, 63.4% women) treated at the University of California, San Diego Interventional Psychiatry Program between 2017 and 2023. ML models were internally validated using nested cross-validation and Shapley values were calculated to quantify contributions of each feature to response prediction. The best-fit models proved reasonably accurate at discriminating treatment responders (Area under the curve (AUC): 0.689 [0.638, 0.740], p < 0.01) and remitters (AUC 0.745 [0.692, 0.797], p < 0.01), though only the response model was well-calibrated. Both models were associated with significant net benefits, indicating their potential utility for clinical decision-making. Shapley values revealed that patients with comorbid anxiety, obesity, concurrent psychiatric medication use, and more chronic TRD were less likely to respond or remit following rTMS. Patients with trauma and former tobacco users were more likely to respond. Furthermore, delivery of intermittent theta burst stimulation and more rTMS sessions were associated with superior outcomes. These findings highlight the potential of ML-guided techniques to guide clinical decision-making for rTMS treatment in patients with TRD to optimize therapeutic outcomes.

The use of theta burst stimulation in patients with schizophrenia - A systematic review.

Transcranial magnetic stimulation (TMS) can offer therapeutic benefits and provide value in neurophysiological research. One of the newer TMS paradigms is theta burst stimulation (TBS) which can be delivered in two patterns: continuous (cTBS - inducing LTD-like effects) and intermittent (iTBS - inducing LTP-like effects). This review paper aims to explore studies that have utilized TBS protocols over different areas of the cortex to study the neurophysiological functions and treatment of patients with schizophrenia. PubMed was searched using the following keywords "schizophrenia", "schizoaffective", or "psychosis", and "theta burst stimulation". Out of the 90 articles which were found, thirty met review inclusion criteria. The inclusion criteria included studying the reported effect (clinical, physiological, or both) of at least one session of TBS on human subjects, and abstracts (at minimum) must have been in English. The main target areas included prefrontal cortex (12 studies - 10 dorsolateral prefrontal cortex (DLPFC), 2 dorsomedial prefrontal cortex (DMPFC)) vermal cerebellum (5), and temporo-parietal cortex (8). Other target areas included inferior parietal lobe (2), and motor cortex (3). TBS neurophysiological effect was explored in 5 studies using functional magnetic resonance image (fMRI), magnetic resonance spectroscopy (MRS), electroencephalography (EEG), electromyography (EMG) and positron emission topography (PET) scan. Overall, TBS can offer great therapeutic potential as it is well-tolerated, feasible, and has few, if any, adverse effects. TBS may be targeted to treat specific symptomatology, as an augmenting intervention to pharmacotherapy, or even improving patient's insight into their diagnosis.

Isolating sensory artifacts in the suprathreshold TMS-EEG signal over DLPFC.

Combined transcranial magnetic stimulation and electroencephalography (TMS-EEG) is an effective way to evaluate neurophysiological processes at the level of the cortex. To further characterize the TMS-evoked potential (TEP) generated with TMS-EEG, beyond the motor cortex, we aimed to distinguish between cortical reactivity to TMS versus non-specific somatosensory and auditory co-activations using both single-pulse and paired-pulse protocols at suprathreshold stimulation intensities over the left dorsolateral prefrontal cortex (DLPFC). Fifteen right-handed healthy participants received six blocks of stimulation including single and paired TMS delivered as active-masked (i.e., TMS-EEG with auditory masking and foam spacing), active-unmasked (TMS-EEG without auditory masking and foam spacing) and sham (sham TMS coil). We evaluated cortical excitability following single-pulse TMS, and cortical inhibition following a paired-pulse paradigm (long-interval cortical inhibition (LICI)). Repeated measure ANOVAs revealed significant differences in mean cortical evoked activity (CEA) of active-masked, active-unmasked, and sham conditions for both the single-pulse (F(1.76, 24.63) = 21.88, p < 0.001, η2 = 0.61) and LICI (F(1.68, 23.49) = 10.09, p < 0.001, η2 = 0.42) protocols. Furthermore, global mean field amplitude (GMFA) differed significantly across the three conditions for both single-pulse (F(1.85, 25.89) = 24.68, p < 0.001, η2 = 0.64) and LICI (F(1.8, 25.16) = 14.29, p < 0.001, η2 = 0.5). Finally, only active LICI protocols but not sham stimulation ([active-masked (0.78 ± 0.16, P < 0.0001)], [active-unmasked (0.83 ± 0.25, P < 0.01)]) resulted in significant signal inhibition. While previous findings of a significant somatosensory and auditory contribution to the evoked EEG signal are replicated by our study, an artifact attenuated cortical reactivity can reliably be measured in the TMS-EEG signal with suprathreshold stimulation of DLPFC. Artifact attenuation can be accomplished using standard procedures, and even when masked, the level of cortical reactivity is still far above what is produced by sham stimulation. Our study illustrates that TMS-EEG of DLPFC remains a valid investigational tool.

Pre-Stimulus Power but Not Phase Predicts Prefrontal Cortical Excitability in TMS-EEG.

The cortical response to transcranial magnetic stimulation (TMS) has notable inter-trial variability. One source of this variability can be the influence of the phase and power of pre-stimulus neuronal oscillations on single-trial TMS responses. Here, we investigate the effect of brain oscillatory activity on TMS response in 49 distinct healthy participants (64 datasets) who had received single-pulse TMS over the left dorsolateral prefrontal cortex. Across all frequency bands of theta (4-7 Hz), alpha (8-13 Hz), and beta (14-30 Hz), there was no significant effect of pre-TMS phase on single-trial cortical evoked activity. After high-powered oscillations, whether followed by a TMS pulse or not, the subsequent activity was larger than after low-powered oscillations. We further defined a measure, corrected_effect, to enable us to investigate brain responses to the TMS pulse disentangled from the power of ongoing (spontaneous) oscillations. The corrected_effect was significantly different from zero (meaningful added effect of TMS) only in theta and beta bands. Our results suggest that brain state prior to stimulation might play some role in shaping the subsequent TMS-EEG response. Specifically, our findings indicate that the power of ongoing oscillatory activity, but not phase, can influence brain responses to TMS. Aligning the TMS pulse with specific power thresholds of an EEG signal might therefore reduce variability in neurophysiological measurements and also has the potential to facilitate more robust therapeutic effects of stimulation.

Differentiating transcranial magnetic stimulation cortical and auditory responses via single pulse and paired pulse protocols: A TMS-EEG study.

OBJECTIVE: We measured the neurophysiological responses of both active and sham transcranial magnetic stimulation (TMS) for both single pulse (SP) and paired pulse (PP; long interval cortical inhibition (LICI)) paradigms using TMS-EEG (electroencephalography). METHODS: Nineteen healthy subjects received active and sham (coil 90° tilted and touching the scalp) SP and PP TMS over the left dorsolateral prefrontal cortex (DLPFC). We measured excitability through SP TMS and inhibition (i.e., cortical inhibition (CI)) through PP TMS. RESULTS: Cortical excitability indexed by area under the curve (AUC(25-275ms)) was significantly higher in the active compared to sham stimulation (F(1,18) = 43.737, p < 0.001, η2 = 0.708). Moreover, the amplitude of N100-P200 complex was significantly larger (F(1,18) = 9.118, p < 0.01, η2 = 0.336) with active stimulation (10.38 ± 9.576 µV) compared to sham (4.295 ± 2.323 µV). Significant interaction effects were also observed between active and sham stimulation for both the SP and PP (i.e., LICI) cortical responses. Finally, only active stimulation (CI = 0.64 ± 0.23, p < 0.001) resulted in significant cortical inhibition. CONCLUSION: The significant differences between active and sham stimulation in both excitatory and inhibitory neurophysiological responses showed that active stimulation elicits responses from the cortex that are different from the non-specific effects of sham stimulation. SIGNIFICANCE: Our study reaffirms that TMS-EEG represents an effective tool to evaluate cortical neurophysiology with high fidelity.