The latency of motor-evoked potentials (MEPs) can predict whether cTBS will exert an inhibitory or excitatory effect on the ipsilateral and contralateral primary motor cortex

• ¹ Université catholique de Louvain, Institute of Neuroscience (IONS), Belgium

Repetitive transcranial magnetic stimulation (rTMS) is increasingly used as a research tool to probe the function of a given brain structure, or as a non-invasive therapeutic approach in neurology and psychiatry. However, its effect is highly variable between individuals. For example, Hamada et al. (2012) showed that continuous theta burst stimulation (cTBS) delivered over the primary motor cortex (M1) increases M1 excitability in 58% of subjects, and decreases it in 42% of subjects. The authors found that the excitatory or inhibitory effect of cTBS on M1 can be predicted by the latency of motor-evoked potentials (MEPs), suggesting that this inter-individual variability could be related to differences in the neuronal circuits preferentially activated by TMS. Here, we further explore this hypothesis by characterizing the relationship between MEP latency and the effects of cTBS on M1 excitability, measured using both MEPs and TMS-evoked brain potentials (TEPs). The recording of TEPs can provide useful information about the responsiveness and connectivity between the brain region targeted by TMS and brain regions sampled using EEG (Bonato et al. 2006). cTBS was delivered over the left or right M1 using biphasic pulses delivered over the central sulcus in an anterior-posterior (A-P) or posterior-anterior (P-A) direction. MEPs and TEPs elicited by stimulation of both the ipsilateral and contralateral M1 were recorded before (T0), immediately after (T1) and 20 minutes after (T2) cTBS. Confirming the results of Hamada et al. (2012), we found a positive correlation between MEP latency at T0 and the effect of cTBS on excitability of the ipsilateral M1. MEPs of shorter latency were associated with an increase of MEP amplitude after cTBS (i.e. M1 facilitation), whereas MEPs of longer latency were associated with a decrease of MEP amplitude after cTBS (i.e. M1 inhibition). At T1, this correlation was significant for A-P pulses (r=-0.672, p=.017) but not for P-A pulses (r=-0.137, p=.671). At T2, this correlation was no longer significant (A-P pulses: r=-0.201, p=.552; P-A pulses: r=0.082, p=.800). Most interestingly, there was also a significant but reverse relationship between MEP latency and the effect of cTBS on excitability of the contralateral M1 which was significant for A-P pulses (T1: r=+0.693, p=.012; T2: r=0.623, p=.030) but not P-A pulses (T1: r=+0.31, p=.354; T2: r=-0.03, p=0.925). Furthermore, there was a highly significant relationship between MEP latency and the effect of cTBS on the N100 peak of TEPs elicited by stimulation of the contralateral M1 (T1: r=0.730, p=.007; T2: r=0.714, p=.009). In contrast there was no relationship between MEP latency and the effect of cTBS on the N100 peak elicited by stimulation of the ipsilateral M1 (T1: r=-0.143, p=.658; T2: r=-0.088, p=.785). Taken together, our results confirm that the inter-individual variability of the effect of cTBS on M1 excitability can be predicted by MEP latency, indicating that it is related to individual differences in the neuronal circuits activated by TMS. Furthermore, our results show that variability in MEP latency also predicts the remote effect of cTBS on the contralateral M1, as measured using MEPs and TEPs.