TMS can selectively activate and condition two different sets of excitatory synaptic inputs to corticospinal neurons in humans

Background: Current protocols or repetitive transcranial magnetic stimulation (rTMS) induce mixed facilitatory and inhibitory effects. More selective, quasi-monophasic
high-frequency stimulators now become available. We sought to investigate the impact of current direction and pulse width on intermittent theta burst stimulation (iTBS)
effects on human motor cortex excitability. Also, we estimated strength-duration time constants from motor threshold and input-output (IO) curves for PA and AP
orientations.
Methods: We stimulated the dominant hand representation of the motor cortex in 15 healthy subjects, using “unidirectional biphasic” pulses generated by a controllable
TMS machine (cTMS-3, Rogue Resolutions Ltd., Cardiff, UK), connected to a standard figure-8 coil. iTBS was applied conventionally, using 20 sequences of 2 seconds
iTBS (10 bursts at 5 Hz burst repetition frequency, each burst consisting of 3 pulses of 80 % AMT intensity repeated at 50 Hz frequency). In separate sessions pulses
differing in current direction and shape were applied: a) posterio-anterior (PA) current direction in the brain, 75 μs (iTBS_PA75). b) AP current direction, 45 μs
(iTBS_AP45). Before and for 30 minutes after iTBS, we monitored the modulation of motor evoked potential (MEP) amplitude from the dominant first dorsal interosseus
using conventional, monophasic, suprathreshold pulses generated by a Magstim 2002 stimulator, inducing PA currents in the brain, at 0.2 Hz frequency. In an additional
study on ten healthy subjects, we investigated the effect the two coil orientations with three different pulse widths (30, 60 and 120 μs) on the IO curve and the latency of
the motor evoked potentials (MEPs).
Results: iTBS_AP45 yielded a pronounced and slightly delayed inhibition of MEP amplitude in all but one subjects, it was unrelated to the MEP latency differences.
iTBS_PA75 had a variable and inconsistent effect that was in part related to the latency differenceAP-LM , in that long latency differences were correlated with the
induction of inhibition rather than facilitation. We found a longer time constant for AP than PA orientation. MEP latencies yielded an interaction between pulse width and
orientation, due mainly to longer onset latencies following AP stimuli of short duration.
Conclusions: Current direction influences the outcome of iTBS, with a preference for AP currents. PA and AP stimuli activate the axons of neurones with different time
constants. Those activated by AP pulses excite corticospinal outputs with a longer latency than those activated by PA pulses. AP pulses of short duration recruit long
latency inputs most selectively.