Dependency of non-invasive brain stimulation effects on real-time EEG-based measurements of instantaneous excitability in human motor cortex

Abstract

Non-invasive brain stimulation (NIBS) is a promising tool to modulate brain networks. Among NIBS techniques, transcranial magnetic stimulation (TMS) can activate cortical neurons in a spatiotemporal scale of millimeters and milliseconds and it has proved to have potential not only for electrophysiological, cognitive and behavioral research but also for the treatment of neurological and psychiatric disorders. Nevertheless, electrophysiological responses to TMS as well as clinical improvement of TMS treated patients is often characterized by a high variability across sessions and subjects, which limits the application of TMS in clinical settings. A significant part of this variability may be ascribable to the ongoing brain activity at the time of the stimulation. To experimentally verify the advantage of brain state-dependent TMS, we built a TMS set-up that is able to trigger TMS pulses in real-time on the basis of the ongoing electrical brain activity concurrently registered with electroencephalography (EEG). We used this real-time EEG-triggered TMS set-up to assess dependency of electrophysiological responses to TMS of the left human motor cortex on the phase of the sensorimotor 8 – 12 Hz oscillation, i.e. μ-oscillation, in healthy participants. We synchronized single TMS pulses to the negative peak, positive peak or random phase of the μ-oscillation to assess μ-phase-dependent corticospinal and cortical excitability. Moreover, we repeatedly applied 200 triplets of TMS at 100 Hz, i.e. repetitive TMS (rTMS), either at the negative peak, positive peak or random phase of the μ-oscillation, to assess dependency of rTMS-induced long-lasting changes in corticospinal and cortical excitability on μ-phase. We measured corticospinal excitability with the amplitude of the motor evoked potentials (MEPs) recorded in a muscle of the right hand and cortical excitability with the amplitude of TMS-evoked potentials (TEPs) and the power of TMS-induced oscillations recorded with EEG. Furthermore, we used part of the data acquired for the assessment of μ-phase-dependent cortical excitability to evaluate the differences in cortical responses between real TMS at intensities above and below resting motor threshold (RMT) and a realistic sham stimulation. Larger MEPs were obtained with single TMS pulses applied at the negative vs. positive peak or random phase of the ongoing μ-oscillation. Moreover, a long-lasting (>30 minutes) increase of MEP amplitude was achieved with negative peak synchronized rTMS, while no significant change with respect to the pre-intervention MEP amplitude was observed after positive peak synchronized or random phase rTMS. Single TMS pulses applied at the negative vs. positive peak of the μ-rhythm produced also enhanced TEPs. In particular, the negative deflection occurring approximately 100 ms after TMS application was consistently larger in the negative vs. positive peak condition both for above and below RMT intensities. Importantly, evoked responses elicited by realistic sham TMS were not modulated by the phase of the μ-oscillation. Negative peak triggered burst-rTMS did not significantly change any of the investigated EEG measures, despite the reported long-lasting significant increase in MEPs. Finally, TMS above and below RTM resulted in TEPs with larger amplitudes and in a significant modulation of the power of induced oscillatory activity compared to evoked potentials and induced oscillations elicited by realistic sham. In conclusion, we demonstrated that the negative vs. positive peak of the sensorimotor μ-rhythm represents a state of higher excitability of the motor system and that rTMS synchronized to this high excitability state leads to long-term potentiation (LTP)-like effects on more than 90% of the tested participants. These findings support the notion that brain state-dependent TMS can reduce inter-subject variability with respect to “excitability enhancing” NIBS protocols that are blind to the ongoing brain activity (52-61% LTP-like responders). We also showed that TEPs are sensitive to instantaneous excitability fluctuations, while complex molecular and synaptic changes induced at the cortical level by rTMS may not always be detectable at the macroscopic level by EEG. Finally, cortical responses to real TMS of the motor cortex very likely reflect direct activation of neurons, rather than sensory evoked activity. These findings are relevant for the improvement of NIBS-based therapies of brain network disorders. Future studies are needed to investigate whether oscillations in other brain areas of clinical relevance also influence responses to TMS similarly to what we observed in the motor cortex for the μ-rhythm.