Revisiting SCS: stimulation or conduction block? - Pathos

SCS involves the implantation of electrodes in the epidural space. It was first performed in 1967 by Shealy et al., who performed thoracic laminectomy and implanted platinum plates sutured to the dura mater. This procedure became so widely used in the US during the second half of the 1960s that its indiscriminate use led to unsatisfactory results and a decline in interest in the method until the early 1970s [Ray 1978; North 1993]. During this period, the percutaneous insertion of temporary electrodes was proposed to test patients before definitive implantation [Hosobuchi et al. 1972; Erickson 1975]. Subsequently, the percutaneous technique, which was previously only used for testing, was adopted for definitive implantation (Zumpano & Saunders, 1976; North et al., 1977). Initially proposed solely for analgesic purposes, spinal cord stimulation (SCS) was later recognised as capable of interfering with various autonomic functions. It has been used to control urinary retention in paraplegics (Nashold et al., 1972), bladder incontinence (Cook et al., 1976; Augustinsson et al., 1982), angina pectoris (Murphy & Giles, 1987), and Raynaud's phenomenon in scleroderma (Zina et al., 1992; Sciacca et al., 1992). It has also been used to prevent damage from cerebral vasospasm.

What is known for certain is that SCS exerts an analgesic effect in certain types of neuropathic pain, but not in nociceptive pain. It also has an autonomic-vasodilatory effect, which is useful in treating peripheral artery disease and angina pectoris. Furthermore, due to its sympatholytic effect, SCS is an important contrast to CRPS-1 in terms of pathogenesis.

The hypotheses formulated to explain the mechanisms of action of SCS are summarised below: the theory of activation of large myelinated afferents; the theory of conduction block; and the theory of glial activation block.

Although the gate control theory continues to be discussed in this context as if it were a religious tradition, it is only implicated in the mechanism of action of SCS with regard to Aβ fibres that run parallel to the dorsal horn of the spinal cord and activate the inhibitory interneuron of the gelatinous substance. In reality, this has a modest analgesic effect that only affects nociceptive pain conducted by C fibres. This is almost irrelevant in the case of SCS as it does not affect the control of neuropathic pain from axonal neuropathy, which is more important.

Regardless of the gate control theory, the theory of the activation of large myelinated afferents remains the most widely accepted of the theories proposed. According to this theory, activation of Aβ fibres in the posterior columns of the spinal cord evokes orthodromic impulses. Upon reaching the nuclei of the gracile and cuneate fascicles, these impulses activate supraspinal centres, from which descending inhibitory pathways originate in the dorsolateral funiculus (DLF). Antidromic impulses travel along collaterals directed to the superficial layers of the dorsal horn of the spinal cord (Figure 1). In both pathways, Aβ fibres activate inhibitory interneurons in the dorsal horn of the spinal cord. The descending pathways in the DLF activate the 5-HT inhibitory receptors of the second neuron, which explains the analgesic effect in axonal neuropathic pain. The collaterals directed to the superficial layers of the dorsal horn of the spinal cord activate the inhibitory interneurons of the gelatinous substance, as well as the GABA_B receptors in the second neuron and the GABA_B receptors in the neuron of sympathetic origin. This contributes to the analgesic effect in axonal neuropathic pain.

At a time when the analgesic effect of spinal cord stimulation (SCS) was widely accepted to be the result of the activation of large myelinated afferents, Larson et al. [1974, 1975] challenged this theory. These authors reported the analgesic efficacy of electrostimulation of the anterior half of the spinal cord (see Figure 2) and hypothesised that pain relief from SCS was due to conduction block in the spinothalamic tract rather than activation of the posterior columns. Following this line of thinking, Hoppenstein (1975a, 1975b) achieved pain relief with a current intensity 30 times lower when the electrodes were placed in front of the lateral spinothalamic tract than when they were placed posteriorly. He also observed that, in the former case, pain relief was contralateral to the stimulated site.

In line with the observations of Larson and Hoppenstein, I have personally noted that patients undergoing percutaneous cervical cordotomy often report the temporary disappearance of contralateral pain as soon as the electrode penetrates the lateral spinothalamic tract during sensory stimulation at 75 Hz, even before the lesion is reached. During the same period, Campbell and Taub (1973) and Ignelzi and Nyquist (1976) formalised the hypothesis that, rather than adding impulses to the nerve, electrostimulation reduces the number of impulses, causing a conduction block.

Supporting the conduction block hypothesis, Campbell (1980) observed that increasing the intensity of electrical stimulation of a peripheral nerve first raises the excitation threshold for tactile stimuli, followed by nociceptive stimuli, until the skin is completely anaesthetised. Based on these observations, Campbell ruled out the possibility that neurostimulation activates large-calibre fibres and considered it more likely that it causes conduction block in all afferent nerve fibres. Furthermore, due to their greater surface-to-volume ratio, small-calibre fibres would be blocked at a lower stimulation frequency than that required for larger-calibre fibres.

Campbell defined this phenomenon as a conduction block related to the stimulation frequency. In agreement with Adelman and Fitzhugh (1975), he considered it to be due to the accumulation of K⁺ ions around the axon, which reduces the conductance of the Na⁺ channels. In practice, the stimulating current would collide with the electrical stimulus travelling orthodromically along the nerve fibres, blocking it (the collision current theory). Campbell also observed that, when treating lumbosacral pain, the electrodes can be placed behind the cauda equina rather than the spinal cord. From this position, it is unlikely that the electrodes will activate the posterior cords; however, it is likely that they will block the conduction of the nociceptive fibres of the first neuron before they enter the spinal cord. Indeed, the conduction block theory can explain the analgesic effects of electrostimulating peripheral nerves or the anterolateral quadrant of the spinal cord or cauda equina, but not the posterior cords.

Compared to the approach that was almost blindly accepted until around 15 years ago, SCS must now be re-examined in light of new technological advances, such as high-frequency SCS and Burst SCS. These techniques have the advantage of being effective in treating lumbar pain (which is often not clearly defined, but is likely to be of a nociceptive nature), and they are easier to implement as they do not require the patient to cooperate by searching for paresthesia. This means that they can be performed under general anaesthesia.