Mechanisms and mode of action of spinal cord stimulation in chronic neuropathic pain

3

ABSTRACT

Tonic spinal cord stimulation (SCS) has been used as a treatment for chronic neuropathic pain ever since its discovery in late 1960s. Despite its clinical successes in a subset of chronic neuropathic pain syndromes, several limitations such as insufficient pain relief and uncomfortable paresthesias have led to the development of new targets, the dorsal root ganglion, and new stimulation waveforms, such as burst and high frequency. The aim of this review is to provide a brief overview of the main mechanisms behind the mode of action of the different SCS paradigms. Tonic SCS mainly acts via a segmental spinal mechanism where it induces GABA-release from inhibitory interneurons in the spinal dorsal horn. Tonic SCS concurrently initiates neuropathic pain modulation through a supraspinal-spinal feedback loop and serotonergic descending fibers. Mechanisms of stimulation of the DRG as well as those related to new SCS paradigms are now under investigation, where it seems that burst SCS not only stimulates sensory, discriminative aspects of pain (like Tonic SCS) but also emotional, affective, and motivational aspects of pain. Initial long-term study results on closed-loop SCS systems hold promise for improvement of future SCS treatment.

KEYWORDS:

• Spinal cord stimulation
• pain modulation
• tonic
• burst
• high-frequency
• mechanism of action
• dorsal horn
• nociception
• serotonin
• GABA

1. Introduction

Spinal cord stimulation (SCS) is indicated as a treatment option for multiple chronic neuropathic pain syndromes, including failed back surgery syndrome (FBSS), complex regional pain syndrome (CRPS), and diabetic polyneuropathy (PDPN) [1]. Conventional or Tonic SCS treatment requires the implantation of stimulation electrodes in the epidural space, placed on top of the dorsal column and at the level at which the nerve roots associated with the painful dermatomes enter the spinal cord. Shealy and colleagues [2] first introduced spinal cord stimulation in 1967 and it has since then evolved in a treatment option for chronic neuropathic pain [1].

2. Tonic SCS

The stimulation paradigm introduced by Shealy et al. is referred to as conventional or tonic stimulation. Tonic stimulation is applied with a frequency ranging from 40 to 80 Hz and a pulse width varying between 200 and 500 µs and at intensities which induce tingling sensations or paresthesias. Preclinically, stimulation intensity or amplitude is programmed at 66–90% of the intensity that generates regular muscle twitches in the animal. The concept of tonic SCS is based on the Gate Control Theory by Melzack and Wall [3]. The electrical pulses delivered by SCS activate Aβ fibers in the dorsal column which, via antidromic transmission, activate inhibitory interneurons in the spinal dorsal horn. These interneurons modulate incoming nociceptive input from Aδ and C fibers and release the inhibitory neurotransmitter γ-aminobutyric acid (GABA), thereby ‘closing the gate.’ The ‘closed gate’ prevents transmission of nociceptive signals to the brain and thereby inhibits pain sensation.

2.1. Spinal mechanisms of tonic SCS: role of GABA

Preclinical studies have established the involvement of the neurotransmitter GABA and the inhibitory GABAergic interneurons in the mechanism underlying tonic SCS mediated analgesia. Extracellular GABA concentrations in the dorsal horn of neuropathic rats were shown to be increased during SCS [4]. Furthermore, Janssen et al. showed reduced intracellular GABA immunoreactivity in the dorsal horn of rats with Partial Sciatic Nerve Ligation (PSNL) after 30 minutes of tonic SCS [5]. From this, it is concluded that tonic SCS induced GABA release into the extracellular space in the spinal dorsal horn and that this is a pivotal mechanism underlying the pain-relieving mechanism of tonic SCS. Intrathecal pharmacological studies have further elucidated and detailed the involvement of this GABAergic mechanism in tonic SCS, demonstrating in particular the GABA_B receptor to be very important [6,7]. Importantly these preclinical findings have been translated into the clinic, demonstrating that the synergistic effect of administering a subclinical dose of the GABA_B receptor agonist baclofen and tonic SCS turned SCS non-responders into responders [8].

2.2. Supraspinal mechanisms of tonic SCS: role of serotonin (5-HT)

Adding to the spinal or segmental effects of tonic SCS, multiple studies have shown that SCS also modulates pain through a supraspinal-spinal feedback loop (Figure 1). Early evidence for the role of descending projections in the mechanism underlying tonic SCS was based on a study by Saadé et al., where tonic SCS rostral to a dorsal column lesion inhibited dorsal horn neurons in segments below this lesion [9]. Additionally, fMRI studies showed that tonic SCS altered activation of supraspinal areas associated with the lateral spinothalamic tract [10]. Hence, tonic SCS results in the activation of supraspinal areas and modulate incoming nociceptive signaling at the spinal levels through their descending projections. The rostral ventromedial medulla (RVM) in the brainstem is involved in the descending modulation of pain via serotonergic input to the dorsal horn [11]. Various studies by Song et al. showed that tonic SCS modulates RVM activity [12], that it increased serotonin content in the dorsal horn of SCS responding rats and that intrathecal administration of serotonin markedly improved the effect of tonic SCS in non-responding rats [13]. Like serotonin, norepinephrine is involved in mediating tonic SCS induced antinociception as well. SCS was shown to increase the synthesis of norepinephrine in the locus coeruleus [14]. Clinically, co-administration of SCS and the serotonin-norepinephrine reuptake inhibitor duloxetine improved McGill pain questionnaire outcomes and willingness to repeat SCS surgery, although it did not significantly increase NRS outcomes [15]. In addition to the activation of descending feedback, orthodromic activation of Aβ-fibers causes the sensation of paresthesias. Paresthesias are tingling sensations perceived in the area that is innervated by the activated Aβ-fibers and clinicians use them to map the stimulation to cover the painful areas.

Figure 1. Spinal nociceptive network and mechanisms of SCS. Electrical stimulation of the dorsal column activates Aβ-fibers. Antidromic activation of Aβ-fibers results in the activation of GABAergic interneurons in the dorsal horn, causing inhibition of pain transmission via nociceptive-specific projection neurons and wide dynamic range neurons (gate control theory). Orthodromic activation of Aβ-fibers initiates the supraspinal feedback loop that modulates pain transmission in the dorsal horn through descending serotonergic and noradrenergic projections. (adapted from Smits et al. [16])

2.3. Limitations of tonic SCS

The aforementioned paresthesias are sometimes deemed very uncomfortable by patients, making paresthesias a potential drawback of tonic SCS. Also, besides its clinical successes in treating a selection of neuropathic pain syndromes such as FBSS, CRPS, and PDPN [1], not all patients treated with tonic SCS experience sufficient pain relief. In general, the reported success rate for tonic SCS is roughly 50% pain relief on the VAS/NRS scale in approximately 50–70% of patients [17]. Furthermore, the effect of tonic SCS seems to decrease over the years as shown in various long-term follow-up studies [18]. All these limitations have urged the tonic SCS-field to develop new SCS paradigms and select possible new targets.

3. Novel SCS target: dorsal root ganglion (DRG)

Recent technological advancements have allowed application of SCS to new targets, such as the dorsal root ganglion (DRG). DRG neurons transmit sensory information from the periphery to the central nervous system and pathological changes are observed after nerve injury, contributing to chronic neuropathic pain [19]. DRG somas are able to modulate pain transmission at their connection to the axon, the T-junction. This T-junction can (1) block transmission of action potentials (APs), (2) it can actively participate in the propagation of APs, and (3) it can act as a low pass filter, selectively allowing the propagation of APs to the dorsal horn [20]. Clinical studies have shown that DRG stimulation is as successful in relieving pain as SCS targeting the dorsal column [21,22]. Despite the clinical success, analgesic mechanisms of DRG stimulation are still largely unknown. Preclinical studies have shown GABA release in the dorsal horn not to be involved (as with Tonic SCS of the Dorsal columns see section 2.1) [23] but a modulatory role via local GABA in the DRG has been suggested [24]. Furthermore, changes in BOLD response in brain areas associated with pain [25] and reduced neuronal excitability in the DRG were seen in response to DRG stimulation [26].

4. Novel SCS paradigms and their mechanisms

4.1. Introduction to new SCS paradigms

In addition to defining new targets for successful SCS treatment, new SCS waveforms, like Burst and High Frequency (HF) have been developed and implemented in SCS treatment. Both HF and burst SCS are paresthesia-free paradigms as stimulation is usually below sensory threshold. The lack of paresthesia allows double-blind placebo-controlled clinical trials, a study format that has not been possible with tonic SCS. Clinical studies using burst or HF-SCS have shown similar analgesic efficacy as with tonic SCS, although there is still a need for more randomized clinical studies to further substantiate possible superiority of these novel paradigms. With the development of these new waveforms, Burst- and HF-SCS, the concept of charge delivery and how this affects the nervous system has gained interest within the field. Rather than focussing on the individual parameters such as amplitude and pulse width that determine the amount of charge, the amount of charge delivery or ‘dosing’ may be of importance in pain modulation [27]. Additionally, preventing charge buildup by either active or passive recharge may also differentially influence the pain transmission neurons in the spinal dorsal horn [28] but as of now, this has not proven to have any implications for effect on pain behavior [29].

4.2. High-frequency SCS

High-frequency stimulation delivers tonic pulses to the dorsal column with frequencies ranging from 1 to 10 kHz, thereby delivering more charge per second as compared to tonic SCS. Even though tonic and HF SCS both deliver pulses in a tonic waveform, the vast difference in frequency and energy delivery between the two paradigms seems to result in the activation of different neuronal mechanisms. It is suggested that HF SCS does not activate Aβ axons in the dorsal column, which may explain the absence of paresthesias. Despite several studies on the mechanisms underlying HF SCS, it is far from clear and the mechanisms of action need to be elucidated. Currently, there are three main working hypotheses [27,30]: (1) HF SCS induces a depolarization block, prohibiting the propagation of action potentials; (2) HF SCS induces desynchronization which may result in pseudo-spontaneous or stochastic neuronal activity in the spinal ‘gate’; (3) HF SCS may induce temporal summation, where multiple, on their own insufficient, impulses induce neuronal activation within a certain time frame.

4.3. Burst SCS

Burst SCS differs from tonic and HF SCS as pulses are delivered to the dorsal column in a cluster of pulses of a high frequency, separated by a longer time duration (the inter-pulse interval). Compared to tonic SCS, burst has a lower charge per pulse, whereas the charge per second is considerably higher [31]. The charge per burst has shown to be important for inhibition of firing of the pain transmission neurons in the dorsal horn of rats [32].

Interestingly, the mechanism underlying Burst SCS, as with tonic SCS, is shown to include the activation of GABAergic interneurons in the spinal dorsal horn [33]. Pharmacological interventions and intrathecal administration of GABA_A and GABA_B antagonists were shown to abolish the pain-relieving effect of both tonic and Burst SCS [33] thereby strongly suggesting GABA signaling to be involved in the mechanism underlying burst SCS. Furthermore, a delayed wash-in effect is noted in preclinical studies on the effect of Burst SCS as compared to tonic SCS in chronic neuropathic animals [34]. From this, it is suggested that the delayed wash-in effect with Burst SCS is due to the activation and involvement of supraspinal areas. Indeed, both EEG and imaging studies have shown that burst SCS activates supraspinal areas that are involved in motivation and emotion to a larger extent than tonic SCS [35,36]. It is suggested that burst SCS activates both the medial and lateral spinothalamic tract whereas tonic SCS only activates the lateral [37]. The former is more involved with emotional, affective components of pain and the latter is more involved in sensory, discriminative aspects.

5. Novel SCS system: closed-loop SCS

The above-described SCS paradigms deliver current to the dorsal column by fixed input, meaning that all the stimulation parameters (i.e., frequency, pulse width, inter-pulse interval, and amplitude) are unchanged during stimulation. This way of stimulation can be referred to as open-loop stimulation. Conventional stimulation of the dorsal column generates Evoked Compound Action Potentials (ECAPs), which can be used to measure Aβ fiber recruitment [38]. A study by Parker and colleagues recorded ECAPs in chronic pain patients undergoing SCS and discovered that ECAP amplitude increases with increasing SCS current. In addition, the distinction between comfortable and uncomfortable stimulation intensity could be made based on ECAP shape [38]. This information contributed to the development of a closed-loop SCS system where the intensity of conventional stimulation paradigms is continuously adapted by measuring ECAPs, comparing them to a set point of comfortable stimulation and optimal pain relief, and changing input current (i.e. amplitude) by means of a feedback algorithm [39]. Long term (12 months) results of both a prospective, multicenter, single-arm study (Avalon trail) [40] and a double-blind RCT (Evoke trail) [41] on ECAP-controlled closed-loop SCS show sustained significant pain relief [40,41] and even superiority over open-loop SCS [41]. However, preclinical investigations into the exact mechanisms underlying ECAP-controlled closed-loop SCS still need to be performed. As this form of closed-loop SCS utilizes conventional stimulation settings, the molecular mechanisms of analgesia may be similar to tonic SCS.

6. Preclinical assessment and effectivity of SCS

Although the mechanisms underlying SCS in modulating neuropathic pain have been extensively studied in animals, it should, however, be noted that behavioral assessment of pain in animals is complicated and this may make the translation of findings to the human situation complicated. Whereas humans can communicate their pain levels, animals cannot do this. The most commonly used tests, e.g. Von Frey assessment, evaluate pain thresholds based on withdrawal reflexes. Since reflexes pertain to the spinal cord, the supraspinal influence on pain disregarded using these kinds of tests. As previously discussed, both tonic and burst SCS activate supraspinal areas and this should not be disregarded when assessing SCS effectivity. Cognitive-motivational aspects of pain, and thereby supraspinal processing can be assessed using the Mechanical Conflict-Avoidance System (MCAS). During testing, the animal has to move from a lit box to a dark box, thereby crossing a hallway with sharp, height-adjustable probes (0–5 mm). Exit latency, i.e. the time it takes for the animal to exit the lit box with all four paws, is an outcome measure of the MCAS. In the MCAS, the animal has two options: (1) escape an aversive setting (light box) by subjecting itself to the painful probes, or (2) remain in the aversive setting to avoid the painful probes. Choosing one or the other requires input from cortical areas to weight costs vs benefits. Both tonic and burst SCS showed reduced exit latencies compared to pre-SCS values [42].

From this, it concluded that (1) tonic SCS does activate not only the lateral spinothalamic tract and associated brain areas but also, to some extent, the medial (see also section 2.2) and (2) as Burst SCS showed a more prominent effect it significantly stimulates the medial spinothalamic tract and with that brain areas involved in motivational cognitive aspects of pain. The introduction of operant tests (MCAS, but also conditioned place preference) in the field of experimental pain and SCS does add to our understanding and increases translatability.

7. Conclusion

From this review, it is concluded that Tonic SCS is a valuable treatment option for neuropathic pain and mainly acts via a segmental spinal mechanism where it induces GABA-release from inhibitory interneurons in the spinal dorsal horn. At the same time, Tonic SCS also modulates neuropathic pain through a supraspinal-spinal feedback loop and serotonergic descending fibers. To overcome limitations of Tonic SCS, such as the fact that not all patients treated with tonic SCS experience sufficient pain relief but also the presence of uncomfortable paresthesias, the field of tonic SCS has been evolving and expanding to include additional targets, the DRG, multiple novel stimulation paradigms, such as HF and Burst SCS, and an ECAP-controlled closed-loop system. Mechanisms of stimulation at the DRG as well as related to new SCS paradigms are now under investigation, where it seems that Burst SCS not only stimulates sensory, discriminative aspects of pain (like Tonic SCS) but also emotional, affective, and motivational aspects of pain. In view of the latter, it becomes more and more important to include operant testing. Analysis of affective, emotional, and motivational aspects of pain, in addition to reflex-based sensory, discriminative aspects, is absolutely needed to study the effects of mechanism and mode of action of new SCS paradigms in the treatment of neuropathic pain.

Declaration of interest

The contents of the paper and the opinions expressed within are those of the authors, and it was the decision of the authors to submit the manuscript for publication.

Elbert A. Joosten is a consultant for Boston Scientific and Saluda and receives financial support for experimental research on SCS in neuropathic pain from Boston Scientific, Abbott, and Medtronic.

A reviewer on this manuscript has disclosed that they have IP on burstDR stimulation. The other peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.