Invited reviewThe physiological basis of transcranial motor cortex stimulation in conscious humans
Introduction
Direct recording from the pyramidal tract in cats and primates provided the first evidence of the action of single pulse electric stimulation of the motor cortex (Adrian and Moruzzi, 1939, Patton and Amassian, 1954, Kernell and Chien-Ping, 1967). In these experiments, the skull was removed and electric stimulation given through a small probe directly to the surface of the cortex. In the monkey, these showed that monopolar anodal stimulation (i.e. a focal anode on motor cortex referred to a large distant cathode) was more effective than cathodal stimulation (Hern et al., 1962). At threshold, anodal stimulation recruited a single descending volley in the pyramidal tract, known as a D wave (Patton and Amassian, 1954). At higher intensities of stimulation this volley was followed by other volleys with a periodicity of approximately 1.5 ms, which were termed I waves (Patton and Amassian, 1954). Patton and Amassian (1954) showed that the later volleys disappeared after removal of the cortical grey matter and were the first affected by cooling of the cortex. They suggested that the initial volley was produced by direct stimulation of the pyramidal tract axons (hence D=direct wave), whereas the later volleys were produced by synaptic activation of the same pyramidal tract neurons (hence I=indirect wave). The reason for the periodicity of the I-wave input has never been fully resolved. It may depend to some extent on reverberating activity in synaptic circuits in the cortex, or it may depend upon the membrane properties of the pyramidal tract neuron which cause it to fire repetitively after a large synchronous depolarising input.
After a few abortive attempts by others, Merton and Morton (1980) were the first to succeed in electrically stimulating the human motor cortex through the intact scalp (transcranial electrical stimulation, TES). As in the animal experiments, anodal stimulation was superior to cathodal stimulation but the method was uncomfortable. This was because only a small fraction of the applied current flowed through the resistance of the skull and scalp into the brain, while the rest travelled between the electrodes on the surface causing local pain and contraction of scalp muscles. The development of transcranial magnetic stimulation (TMS) (Barker et al., 1985) overcame these problems of discomfort by using a magnetic field to carry the electrical stimulus across the scalp and skull to the brain.
In the initial experiments with TES and TMS in humans many of the details relating to the mechanism of action of the two forms of stimulation on the cortex had to be deduced indirectly by examining the form of the EMG responses evoked in contralateral hand muscles. For example, on the basis of latency measurements and single motor unit studies, Day et al. (1989) proposed that TES with anodal stimulation activated the pyramidal tract outflow in a way very similar to that described in the monkey. They found that TES tended to fire single motor units at preferred intervals with an interpeak latency of approximately 1.5 ms. They presumed the first of these peaks was due to arrival of synaptic input from a D-wave whereas the later peaks were due to arrival of input from I-waves. In contrast to TES, they also found that TMS over the hand area tended to recruit the second motor unit peak rather than the first and they proposed that TMS tended to activate pyramidal neurons indirectly via synaptic inputs rather than at the axon of pyramidal tract neurones.
Direct evidence for the action of TMS and TES on the human cortex was only provided later from a small number of experiments in which descending volleys were recorded from the surface of the spinal cord during surgery on the spinal column after TMS or TES of the cortex. The first of these studies was performed by Boyd et al. (1986) who recorded from an epidural lead inserted during surgery on the spinal column for scoliosis. They showed that the response to TES consisted of a series of waves travelling down the spinal cord at 60–80 m/s a second, and they presumed that the first of these would be equivalent to the D-wave described in animal experiments. Later work by Berardelli et al., 1990, Thompson et al., 1991, Burke et al., 1993 examined the question of the difference between TMS and TES by recording the responses in the spinal cord after each form of stimulation. After some debate, in which it was realised that the level of anaesthesia had pronounced effects on recruitment of I-waves by either form of stimulation, it was generally agreed that at least for the human hand area, TES tended to recruit D-waves at threshold whereas TMS tended to recruit I-waves. However as the intensity of each was increased, then both recruited D- and I-waves. At very high intensities, Burke et al. (1993) showed that the part of the pyramidal tract axon activated by TES shifted deeper into the scalp to the approximate level of the cerebral peduncles and brain-stem. They termed the initial volley set up at these sites as the D2 and D3 waves, respectively.
Although the data were useful, they were still limited because the patients were anaesthetised and also because there is a limited amount of time available to make recordings during surgery. However, the realisation that descending volleys could also be recorded from epidural electrodes implanted chronically in the spinal cord for the relief of pain lead to a series of studies on descending volleys in conscious human subjects. The first of these were performed by Kaneko et al. (1996) and by Nakamura et al. (1996) and confirmed the D- and I-wave hypothesis of TMS and TES of the hand area. From 1998 onwards, we have also made a series of studies using the same technique and these are summarised here.
Section snippets
Method of recording epidural volleys
The details of the recording methods are given in the papers cited below. In brief, all the recordings are taken from patients who had spinal cord stimulators implanted for the treatment of intractable dorsolumbar pain. The electrode (model 3487A Medtronic, Minneapolis) was implanted percutaneously in the epidural space either at C1–C2 level or at the thoracic level (over T10 in most of the patients). Recordings were made of descending activity 2–3 days after implantation during the trial
Transcranial electrical stimulation (TES)
As reported by others, the lowest intensities of anodal stimulation at AMT recruit an early descending volley, which has a latency of 2–2.6 ms when recorded from the high cervical cord (Fig. 1) (Di Lazzaro et al., 1998a). In humans, this latency is compatible with direct activation of the pyramidal tract axons just below the grey matter, and hence this wave is referred to as a D wave. At higher stimulus intensities, starting from about 150% AMT, a later small wave appears in some of the
Effects of voluntary contraction and fatigue on descending volleys
One of the advantages of being able to record descending volleys from conscious subjects is that we can examine the effect of voluntary contraction, and also study some of the interactions between pairs of pulses that are often used in EMG studies. Voluntary contraction has no effect on the amplitude of the D-wave recruited either by anodal (Fig. 3) or cathodal TES (Di Lazzaro et al., 1999b) or on the D wave evoked by LM magnetic stimulation. However, voluntary contraction does increase the
Paired-pulse TMS interactions
Paired pulse experiments are designed to give insight into the nature of the cortical circuitry activated by TMS. A variety of different methods exist to examine the connections within the motor cortex itself or connections to the motor cortex from other parts of the central nervous system. Direct recordings of the effects on descending volleys have not only confirmed the mechanisms of these effects, but in some cases, also revealed an unexpected selectivity for different components (D, I1, I2,
Stimulation of the leg area of motor cortex
Previous studies using surface and single unit EMG recording suggest that magnetic stimulation over the leg area may recruit neurones in a different way to those in the hand area. Priori et al. (1993) found that the latency of surface and single unit EMG responses in the tibialis anterior muscle was the same for both vertex electric and magnetic stimulation of the leg area. They assumed that electric stimulation activated corticospinal axons in the subcortical white matter, and therefore they
Conclusions
In general, the data recorded from the epidural space of the spinal cord has confirmed many of the conclusions about recruitment of D and I wave volleys that had been proposed on the basis of indirect measurements of EMG activity. However, a number of additional and unexpected findings have emerged.
The data confirm that:
- 1.
At threshold, the first descending volley evoked by anodal TES of the hand area has a shorter latency than that evoked by TMS.
- 2.
The initial volley after anodal TES is likely to be
References (51)
- et al.
Non-invasive magnetic stimulation of human motor cortex
Lancet
(1985) - et al.
Descending volley after electrical and magnetic transcranial stimulation in man
Neurosci Lett
(1990) - et al.
Comparison of descending volleys evoked by transcranial magnetic and electric stimulation in conscious humans
Electroenceph clin Neurophysiol
(1998) - et al.
Direct demonstration of the effect of lorazepam on the excitability of the human motor cortex
Clin Neurophysiol
(2000) - et al.
Descending volleys evoked by transcranial magnetic stimulation of the brain in conscious humans: effects of coil shape
Clin Neurophysiol
(2002) - et al.
Direct demonstration of long latency cortico-cortical inhibition in normal subjects and in a patient with vascular parkinsonism
Clin Neurophysiol
(2002) - et al.
Motor thresholds in humans: a transcranial magnetic stimulation study comparing different pulse waveforms, current directions and stimulator types
Clin Neurophysiol
(2001) - et al.
Effect of stimulus intensity and voluntary contraction on corticospinal potentials following transcranial magnetic stimulation
J Neurol Sci
(1996) - et al.
Direct and indirect activation of human corticospinal neurons by transcranial magnetic and electrical stimulation
Neurosci Lett
(1996) - et al.
Transcranial electric and magnetic stimulation of the leg area of the human motor cortex: single motor unit and surface EMG responses in the tibialis anterior muscle
Electroenceph clin Neurophysiol
(1993)