The specific spatiotemporal evolution of TMS-evoked potentials reflects the engagement of cortical circuits

Transcranial Magnetic Stimulation (TMS) evokes electroencephalographic (EEG) responses that can last hundreds of milliseconds. Yet only the first 80 ms after the pulse are widely accepted to reflect genuine cortical responses to TMS, while later components have mainly been attributed to the effects of sensory co-stimulations. Here we reappraise this view by investigating the target-specificity of the spatiotemporal evolution of TMS-evoked potentials (TEPs). To this end, we compared TEPs elicited by TMS targeting the premotor and primary motor cortices in healthy subjects, under conditions designed to optimize the TMS effectiveness on the cortex while minimizing peripheral confounds. As a counterfactual, we conducted the same comparison on the EEG responses evoked by realistic sham TMS and high-intensity somatosensory scalp stimulation. We found that EEG responses to motor and premotor TMS can exhibit distinct spatiotemporal evolutions up to 300 ms both at the group and single-subject levels. These differences were absent or marginally present in both realistic sham TMS and high-intensity somatosensory scalp stimulation. Our findings suggest that, when effectiveness is optimized and peripheral confounds are controlled, TMS elicits specific long-lasting EEG responses that reflect genuine responses to the initial engagement of specific cortical targets. These results challenge previous assumptions and highlight how TMS-EEG can be reliably used to assess large-scale properties within corticothalamic networks.