Unraveling the mysteries of motor cortical function in Parkinson disease.

In current models of Parkinson disease (PD), loss of nigrostriatal dopaminergic input results in disordered output from basal ganglia to thalamocortical pathways. This interferes with the function of sensory-motor cortex and likely causes many of the cardinal motor symptoms, including akinesia, bradykinesia, and rigidity. Treatment with L-dopa, pallidotomy, or deep brain stimulation of the subthalamic nucleus or globus pallidus internus removes this “noisy” signal, reduces interference with motor cortical function, and thus improves motor function. The model implies that it should be possible to detect abnormal cortical function in PD and that this should be resolved by L-dopa treatment or neurosurgery. A greater understanding of how bradykinesia occurs may come from studies that provide insight into how motor cortical function changes in PD. Neurophysiologic investigations using transcranial magnetic stimulation (TMS) provide the means to study the circuitry of the motor cortex. These have evolved from using single TMS pulses to excite motor output and contract muscles on the opposite side of the body to double-pulse methods, in which the first pulse excites cortical interneurons that in turn alter the size of themuscle twitch evoked by the second pulse. Different sets of interneurons can be investigated by changing the intensity or interval between pulses of stimulation. One of the first and most commonly used of these is the short interval intracortical inhibition (SICI) method, which probes the excitability of a GABAA-ergic input to the corticospinal pathway. These inhibitory circuits have a role in shaping voluntary motor output, since inhibition is reduced just before activation of the appropriate agonist muscle, whereas inhibition is increased onto projections to the antagonist muscle. Initial studies using this technique in PD clearly showed that SICI was reduced, and that this abnormality could be partly reversed with dopaminergic medications. This finding appeared to be consistent with other TMS studies suggesting that there were also abnormalities in GABAB-ergic networks in the motor cortex. However, not all subsequent studies were able to replicate these findings; moreover, reduced SICI was found not only in PD but also in dystonia and many other disorders. Clearly the situation is not as straightforward as it had once appeared. Moreover, unless the discrepancies are resolved, there will always be some doubt over the usefulness of paired pulse TMS designs. The article by Ni et al. in this issue of Neurology® is the culmination of a series of investigations from many laboratories that are slowly revealing the answers (figure). It turns out that SICI is not quite the simple test of GABAA inhibition that it was thought to be. It is contaminated by activity in a second set of excitatory interneurons that the first pulse causes to fire at very high frequencies of around 600 Hz. The result is that superimposed on the period of inhibition there are peaks of (relative) facilitation at intervals of about 1.5 msec. Although technically challenging, it is possible with some care to tease apart the opposing effects of inhibition and facilitation by subtle adjustments of the stimulus parameters. As it stands, the classic SICI paradigm that is still used all over the world is something of a curate’s egg: the “good” part examines GABAA-ergic pathways but the “bad” part contaminates them with unappetizing facilitation. To their credit, Ridding et al. had noted in their original publication that either decreased inhibition or an excess of excitation might cause the reduced SICI in PD, but they favored reduced inhibition as the explanation because it matched the prevailing idea of cortical GABA dysfunction in PD. Eight years ago, one of us produced evidence that they might have jumped to the wrong conclusion; the lack of SICI in PD was more likely to be due to excess facilitation than reduced inhibition. However, those experiments never directly demonstrated increased facilitation during the classic SICI paradigm. Instead, the study showed that, in the special case when the inhibitory pathway was studied in isolation, there were no deficits in PD. They hypothesized that reduced SICI in PD was caused by enhanced facilitation but in the absence of appropriate experiments to measure facilitation, this hypothesis was not proven. The article by Ni et al. resolves the question. By careful manipulation of the parameters of stimulation, they probed separately the inhibitory and facilitatory subcomponents of SICI. They found that the facilitatory