sign in sign up
<<
Home , Brainstem, Hypothalamus, Medial lemniscus, Midbrain, Neuroanatomy, Neuromodulation

-Relieving Mechanisms in 10

Vikram Sengupta, Sascha Qian, Ned Urbiztondo, and Nameer Haider

Introduction painful disease state being treated. Although clinical evi- dence will be provided, more exhaustive summaries of the Since its inception more than 50 years ago, the forms and clinical data will be reserved for later chapters in Part VI, applications of modern neuromodulation have undergone each of which is dedicated to the clinical aspects of a particu- tremendous expansion. The International Neuromodulation lar form of neuromodulation. Society defines neuromodulation as “the alteration of activity through targeted delivery of a , such as elec- trical stimulation or chemical agents, to specific neurological Background and Historical Perspective sites in the body,” most commonly to reduce pain or improve neurologic function. All forms of neuromodulation are The earliest documented use of electrical current for the reversible. is the most common form of treatment of pain was around 63 AD, when the Mesopotamian neuromodulation technology used today, and it refers to the physician, Scribonius Largus, discovered that shocks deliv- use of electrical or electromagnetic stimuli upon target tis- ered by the electrical torpedo could relieve bodily aches sues to elicit a therapeutic response. Although the focus of and . In the eleventh century, the Islamic philosopher this chapter will be on neurostimulation, other non-electrical Avicenna used cranial shocks delivered by the electric catfish therapies, such as intrathecal drug delivery systems, may fall to treat epilepsy. In the 1600s, the natural philosopher, into the category of neuromodulation. William Gilbert, reported using the magnetic lodestone to On August 10, 2017, the CDC recommended that the opi- treat and psychiatric illness. oid epidemic should be declared a national emergency. In In 1745, Ewald Georg von Kleist and Pieter van light of these circumstances, neuromodulation, which is rela- Musschenbroek independently invented the world’s first tively safe, effective, and validated in the treatment of chronic capacitor, known as the Leyden jar. Subsequently in 1747, neuropathic pain, is emerging as an important alternative to Jean Jallabert used electrical stimulation to increase therapy. Conversely, chronic high-dose opioid ther- flow, and to provoke muscle contraction and growth, thereby apy is not validated, is often ineffective, and carries a sub- restoring function to the paralyzed limb of a locksmith. Also stantial risk of dependence, abuse, , and other in the mid-eighteenth century, Benjamin Franklin used his morbidity and mortality. High-dose medication is also theo- electrostatic generator to treat paralysis and various painful rized to induce opioid-induced , and these drugs conditions. Some of his treatments achieved transient may therefore increase the of pain. improvement. However, his use of high voltages caused burns In this chapter, we will introduce the broad and ever-­ and nerve injuries, leading him to abandon the practice. expanding panoply of neuromodulation and its applications, The true inception of modern neurostimulation can be describing what is known of the mechanisms of action traced to three major developments in the mid-twentieth cen- through which it works. Although similar electrical stimuli tury: (1) the advent and success of the implantable cardiac may often be delivered, the prevailing mechanism of action pacemaker in the late 1950s; (2) the 1965 publication of will vary based on the pathophysiology of the underlying Melzack and Wall’s seminal article, Pain Mechanisms: A New Theory, in which they first proposed the gate control theory of pain, postulating that activation of large-diameter V. Sengupta · S. Qian (*) · N. Urbiztondo · N. Haider fibers could block transmission of pain signals Spinal & Skeletal Pain Medicine, Utica, NY, USA conveyed by small diameter fibers; and (3) the first clinical e-mail: [email protected]; [email protected] application of gate control theory by the neurosurgeon

© Springer Nature Switzerland AG 2019 79 T. R. Deer et al. (eds.), Deer’s Treatment of Pain, https://doi.org/10.1007/978-3-030-12281-2_10 80 V. Sengupta et al.

Norman Shealy in 1967, when he implanted the first spinal and thinly myelinated Aδ fibers, thereby closing a concep- cord stimulator (SCS) to successfully treat chronic neuro- tual gate to painful afferent stimuli located in the . pathic pain. Their model originated from several observations on the sen- In 1958, Medtronic released the first implantable pace- sory neural circuitry located in a crescent-shaped zone cap- maker, followed in 1958 with the first battery-operated wear- ping the dorsal horn known as the substantia gelatinosa (SG). able pacemaker. The revenue from these early successes was Pseudounipolar of the dorsal root then used to subsidize and development that repur- (DRG) conveyed painful stimuli along Aδ and C fibers from posed pacemaker technology to meet the growing demand the periphery into the SG where they formed (1) excitatory for devices specialized for neural applications. Indeed, the with secondary sensory neurons of the spinotha- first was a Medtronic cardiac pace- lamic tract known as tract cells (TCs) and (2) inhibitory syn- maker with leads modified for intrathecal placement. apses with SG (INs) that in turn formed In the decades that followed, neuromodulation blossomed inhibitory synapses with TCs. These fibers therefore pro- into one of the most fascinating and dynamic fields of medi- moted transmission of painful signals via TCs by way of (1) cine, with new improvements and applications emerging direct excitation and (2) disinhibition. every year. Monopolar platinum leads have evolved into Meanwhile, Aβ fibers carrying non-nociceptive signals multi-contact titanium leads. Paddle leads have been devel- bound for the dorsal columns (DC) were also observed to oped for insertion by surgical laminotomy. Internal pulse send collateral projections into the SG where they formed generator (IPG) batteries are smaller and longer-lived. excitatory synapses with TCs and INs, both of them excit- Rechargeable versions are now available. Every year, pro- atory. It had been empirically demonstrated that early direct grams multiply, while the list of validated and experimental excitation of TCs at the Aβ-TC was quickly over- applications expands. driven by TC inhibition caused by excitatory signaling at the Aβ-IN synapse. Hence, the net effect of Aβ signaling in the dorsal horn was to block nociceptive transmission to second- The Gate Control Theory of Pain ary sensory neurons (TCs) of the through a mechanism of (IN)-mediated inhibition, In 1965, Melzack and Wall first articulated the gate control thereby “closing the gate.” theory of pain modulation in their seminal article Pain On the basis of these observations, Melzack and Wall Mechanisms: A New Theory, published in the journal speculated that by administering exogenous electrical stimu- Science. They hypothesized that stimulation of large - lation to Aβ fibers, the spinothalamic gate could be closed as ated Aβ sensory fibers carrying touch and vibratory informa- therapy for neuropathic pain. Figure 10.1 depicts a schematic tion to the dorsal horn could block the transmission of of the gate control circuit similar to the one presented in the nociceptive stimuli conveyed by small unmyelinated C fibers original article by Melzack and Wall.

Gate control mechanism of pain

Aα or Aβ Large diameter + +

- Action SG T system ()

- +

Aδ or C Small diameter

Fig. 10.1 Gate control schematic depicting large-diameter mechano- known as the tract cells (TCs) to form the gate circuitry. The net effect receptive Aα and Aβ fibers and small-diameter δA and C fibers carrying of input from the large fiber afferents is tract cell inhibition, which pain and temperature signals, all entering the dorsal horn from the closes the gate to painful signals. The net effect of the small fiber affer- periphery in parallel, each synapsing on inhibitory interneurons of the ents is TC excitation, which opens the gate to painful signals bound for substantia gelatinosa (SG) and secondary sensory spinothalamic the brain 10 Pain-Relieving Mechanisms in Neuromodulation 81

Central Neuromodulation and Mechanisms the ipsilateral precentral when the pain returned, lead- of Action ing to the conclusion that both gyri were involved in mediat- ing the pain. Although the exact mechanism of action is  (DBS) unclear, MCS is known to inhibit pain signal transmission in patients with central neuropathic pain. Positron-emission Deep brain stimulation (DBS) is a form of neuromodulation tomography (PET) scans using radiolabeled as an in which current is delivered through electrodes stereotacti- index of cerebral blood flow have implicated the ventral, lat- cally implanted into the parenchyma of subcortical brain eral, and medial as key structures. Other locations structures. DBS was first introduced by Benabid et al. when that experience increased blood flow under MCS are the in 1985 they successfully treated symptoms of Parkinson’s anterior cingulate, the orbitofrontal cortices, the anterior disease (PD) by combining thalamotomy with neurostimula- insula, and the upper . It has therefore been hypoth- tion of the bilateral ventral intermediate (VIM) thalamic esized that MCS diminishes the perception of pain both by nuclei. In addition to PD, DBS has subsequently been used modulating the affective-emotional component of chronic to control symptoms of dystonia, epilepsy, depression, pain and by activating descending inhibition of pain impulses obsessive-compulsive disorder (OCD), and cluster through the brainstem. headaches. Although the exact mechanism of action remains elusive, DBS is to act via three mechanisms: (1) by inhibi- Transcranial Magnetic Stimulation tion, (2) by depolarization blockade, and (3) by adjustment of neural activity. In the first model, neurostimulation depo- Transcranial magnetic stimulation (TMS) is the use of the larizes in the target brain structure, thereby causing the electromagnetic field generated by an electrical current to release of inhibitory . In the second model, generate an electrical current in the target area of the brain by neurostimulation depresses activity in the target by the physical process of electromagnetic induction. A coil hyperpolarizing neurons and therefore the electrical thresh- placed over the scalp induces an electrical field, which can old needed to produce an . Finally, DBS may then be manipulated either to excite or inhibit the target tis- induce a more regular and consistent firing in target tissue sue. Theoretical mechanisms of action for TMS on chronic where electrical activity has become disordered by disease. pain include modulation of descending inhibition circuits in DBS has been used to effectively treat a wide swath of addition to effects on limbic circuitry. One significant limita- neuropsychiatric diseases with varying and often unknown tion for TMS is the transient duration of analgesia; it is cur- cellular pathophysiology. For each clinical entity, a constel- rently unclear whether TMS results in any permanent lation of functionally linked neuroanatomic loci has been neuroplastic changes in pain processing. identified. Many single structures, however, have been shown to play crucial roles in multiple diseases. Although there is not yet FDA approval for the treatment of pain, research in Spinal Cord Stimulation (SCS) DBS has shown it to be effective in the treatment of several types of centrally mediated pain syndromes. For example, Overview the was a promising target for treatment of Spinal cord stimulation is neurostimulation via electrodes cluster headaches. Indeed, vigorous hypothalamic activation placed into the epidural space, either percutaneously or by is known to occur during cluster , short lasting uni- surgical laminotomy, and positioned to overlie the dorsal lateral neuralgiform headache with conjunctival injection columns of the spinal cord. SCS is FDA approved for use in and tearing, hemicrania continua, and paroxysmal hemicra- chronic intractable pain of the trunk and/or limbs, post-­ nias. Although all of these entities have been treated success- laminectomy syndrome, complex regional pain syndrome I fully with DBS, the hypothalamus has been largely and II (CRPS), chronic radiculopathy, painful neuropathies, abandoned as a target due to the high rate of morbidity asso- chronic refractory angina, and peripheral ischemic limb ciated with hypothalamic lead deployment. pain, but it has been used to successfully treat many other pain entities. A more comprehensive list of the indications can be found in Table 10.1. SCS is thought to act via three  Stimulation (MCS) broad classes of mechanisms, (1) neurophysiological, (2) neurochemical, and (3) vascular. Electrical stimulation of the cerebral motor cortex via over- Neurophysiologic mechanisms are generally attributed lying electrodes is known as motor cortex stimulation to at the level of the dorsal horn and to (MCS). MCS has been used in the treatment of burning cen- supraspinal brainstem or thalamocortical mechanisms. tral pain by removing the contralateral and Neurochemical mechanisms are thought to restore the 82 V. Sengupta et al.

Table 10.1 Uses and indications for neurostimulation Modality Common indications Other indications Deep brain stimulation Parkinson’s disease *Δ Epilepsy Δ (DBS) Essential *Δ Depression Obsessive-compulsive disorder *Δ Headaches Dystonia *Δ Addiction disorder Severe Vagal nerve stimulation Drug-refractory epilepsy * Δ (VNS) Unipolar and bipolar depression * Hemicrania continua Δ * Medication-­overuse headache Δ Spinal cord stimulation Chronic intractable pain of the trunk and/or limbs, including Intercostal (SCS) post-laminectomy syndrome * Spinal cord injury Complex regional pain syndrome I and II * Arachnoiditis Chronic radicular pain * Central pain Painful neuropathies avulsion Refractory angina Δ Postherpetic neuralgia Peripheral ischemic limb pain, refractory or not amenable to pain surgical bypass Δ Abdominal pain Lower extremity pain secondary to complex regional pain Perineal pain (DRG) stimulation syndrome I and II *Δ Phantom limb pain Chronic radicular pain Chronic postsurgical knee pain Chronic visceral pain neuropathic pain Painful diabetic neuropathy Peripheral nerve of peripheral nerve etiology *Δ, including Craniofacial pain stimulation (PNS) occipital neuralgia Cluster headache Postherpetic neuralgia Peripheral nerve field Axial stimulation (PNFS) Abdominal pain Pelvic pain Atypical facial pain Single asterisk denotes current US and Drug Administration (FDA)-approved indications Single triangle denotes an indication which has received Conformité Européenne (CE) mark

­balance of inhibitory and excitatory neurotransmitters both In the case of SCS, antidromic impulses travel in a caudad at the segmental level as in the case of gamma-aminobutyric direction, back into the dorsal horn where they act in a spinal acid (GABA) and glutamate and via descending inhibitory segmental manner by activating inhibitory neurons which pathways that release , , and norepi- then block transmission of nociceptive signals from C and Aδ nephrine into the dorsal horn. Finally, vascular mechanisms fibers to secondary neurons known as tract cells. This spinal are most pertinent in vascular diseases such as peripheral segmental mechanism is thought to be the primary mecha- arterial disease and chronic refractory angina where SCS is nism by which classical SCS causes relief of neuropathic thought to relieve pain by restoring the supply of oxygen pain. through and by reducing sympathetic tone. Orthodromic impulses are responsible for modulating supraspinal neurophysiologic mechanisms responsible for Neurophysiologic Mechanisms of SCS SCS-mediated pain control. These impulses are also Stimulation of large myelinated primary sensory neurons in responsible for the paresthesia experienced by patients the dorsal cord provokes action potentials that originate at a undergoing SCS. node of Ranvier and propagate bidirectionally away from the It is important to note that the paresthesias in and of them- stimulus. Antidromic impulses travel in the opposite direc- selves are not likely causing pain relief. In light of recent tion of natural physiologic impulses. Orthodromic impulses computational models conceptualizing the thalamus as a par- travel in the same direction as natural physiologic impulses. allel processor for ascending sensory signals, one might In the case of SCS, orthodromic impulses are those that speculate that the paresthesia generated in classic SCS causes travel in a rostral direction toward supraspinal structures a profusion of neutral sensory signals that “drown out” the where Aβ fibers from the dorsal column synapse in the cune- perception of painful stimuli through lateral inhibition and ate and gracile nuclei of the medulla, with subsequent direct signal competition in the thalamus. This, however, is impulses traveling to the thalamus and unlikely because classic tonic stimulation typically produces (PAG) (Figs. 10.2 and 10.3). mostly subthreshold presynaptic potentials in the thalamus, 10 Pain-Relieving Mechanisms in Neuromodulation 83

Thalamus Ventral posterolateral (VPL)

Periaqueductal gray

Pons

Internal arcuate fibers decussate and ascend in the medial Medulla

Primary afferent neurons (AG fibers) Primary afferent Spinothalamic neurons tract (Aδ & C fibers)

Spinal cord

Decussation in anterior white commissure

Fig. 10.2 Ascending sensory pathways: the dorsal column and spino- the substantia gelatinosa or the nucleus proprius. These tract cells give thalamic tract. Dorsal column (blue): the axons of pseudounipolar neu- off axons which decussate via the anterior white commissure and then rons of the DRG carrying fine touch, vibration, pressure, and travel to the anterolateral quadrant of the spinal cord where fibers carry- along from the periphery along Aα and Aβ fibers enter ing pain and temperature enter the lateral spinothalamic tract and fibers the tracts of the ipsilateral dorsal column and travel to the carrying crude touch and firm pressure enter the anterior spinothalamic medulla where they synapse on second-order neurons in the gracile and tract. These axons then travel up the spinal cord, through the rostral cuneate nuclei. These secondary neurons then decussate as internal ventromedial medulla, and into the thalamus where they synapse with arcuate fibers to form the and travel through the third-order neurons in several thalamic nuclei, including the medial and form synapses with third-order sensory neurons in the thalamus. dorsal, ventral posterior lateral, and ventral posterior medial nuclei. Axons coming from the body synapse in the ventral posterolateral These signals are then conveyed to the , the primary nucleus and those from the head synapse in the ventral posteromedial somatosensory cortex, and the , respectively. These thala- nucleus. Axons of thalamic third-order neurons then travel to the pri- mocortical projections are functionally divided into two subsystems mary somatosensory cortex. Spinothalamic tract (red): the axons of known as the direct system, for the conscious appreciation of pain, and pseudounipolar sensory neurons of the dorsal root ganglion carrying the indirect system, which is responsible unconscious processing of pain and temperature stimuli along Aδ and C fibers project into the pain. The indirect system is divided into spino-reticulo-thalamo-­cortical dorsal horn and typically rise one to two spinal segments via Lissauer’s circuits responsible for and the spino-mesencephalic- tract where they synapse on secondary neurons known as tract cells in limbic circuits responsible for the affective impact of pain 84 V. Sengupta et al.

ab Spinal cord stimulation Descending inhibitory pathways Periaqueductal gray

Midbrain

Orthodromic Locus impulses coeruleus (LC) to brainstem

Spinal cord stimulator leads Pons

Aδ fiber

Nucleus raphe magnus (NRM)

Medulla Antidromic to other spinal levels

Serotonergic projection

DRG cell Spinal cord

cd

Noadrenergic from LC A & or Antidromic Aβ fiber signals from NRM from SCS to dorsal column

A & or C fiber 2˚ sensory- tract cell of the spino- thalamic pathway Descending inhibitory Gate control pathways mechanism To spinothalamic Substantia tract gelatinosa

Fig. 10.3 (a) Spinal cord section with SCS leads overlying the dorsal ceruleus (LC) and the (NRM) are located, columns triggering an ascending orthodromic impulse bound for the respectively. Through caudad projections that descend along the dorsal medulla and a descending antidromic impulse descending along the columns and terminate in the dorsal horn, these nuclei exert supraspinal same Aβ fiber and into the dorsal horn, thereby closing the gate. (b) neurochemical pain control by secreting inhibitory substance P, norepi- Gate control neurocircuitry in situ. (c) Descending supraspinal inhibi- nephrine, and serotonin into the SG, thereby stabilizing the gate. (d) tory pathways originating in the periaqueductal gray (PAG) and the Inhibitory fibers synapsing on a tract cell in the dorsal horn medulla where noradrenergic and serotonergic neurons of the locus 10 Pain-Relieving Mechanisms in Neuromodulation 85 suggesting that the need for overlap is more of an incidental Neurophysiologic Mechanisms of SCS: Tonic phenomenon which at most can serve as a perceptual marker Versus Burst Stimulation Patterns that the antidromic signals closing the gate in the dorsal horn Classical SCS utilizes tonic stimulation with single pulses in have been correctly mapped to the affected dermatomal dis- sequence, each with the same pulse width, pulse rate, and tribution. Indeed, in order to close the gate on a particular amplitude. However, single action potentials do not typically distribution of aberrant nociceptive C and Aδ fibers, βA trigger enough central neurons to cause modulation at the fibers that directly form circuits with those neurons in the level of the thalamus. Indeed, these central neurons must dorsal horn and therefore originate in the same dermatome integrate the input from multiple presynaptic neurons in a must be stimulated. process known as spatial summation in order to reach thresh- However, there are supraspinal mechanisms through old. In this context, threshold is defined as the membrane which classical SCS is thought to achieve pain relief. One voltage required to elicit an action potential. Hence, tonic experimental study performed by El-Khoury achieved pain pulse stimulation is most likely filtered out at the level of the relief by stimulation of sites rostral to various spinal cord thalamus as noise, only rising to thalamocortical circuits as lesions, thereby preventing the contribution of an antidromic isolated somatosensory signals. Therefore, classical SCS process to analgesia. This finding implicates higher-order most likely exerts its main effects by affecting gating in the pain relief caused by modulation of sensory-discriminatory dorsal horn. elements of pain at the level of the brainstem and also In contrast, it has been observed that through a process ­modulation of affective-emotional elements of pain at the known as temporal summation or facilitation, short clusters level of thalamocortical circuits. of action potentials known as “bursts” can lead to the stack- Ultimately, the experimental and clinical data suggest that ing of intracellular calcium concentration in a single presyn- classical SCS relieves neuropathic pain predominantly aptic bouton causing sufficient release to through modulation of gating mechanisms at the spinal seg- monosynaptically trigger a postsynaptic central neuron, mental level. Indeed, one experimental pain model found thereby leading to meaningful signaling either through long-­ that DC SCS at the level where an injured sciatic nerve term potentiation (LTP) or long-term depression (LTD). One entered the spinal cord was more effective than more rostral large prospective, RCT known as the SUNBURST trial stimulation. This notion has been empirically supported found that burst stimulation as described by DeRidder is through the mapping of the most effective sites for electrode generally more effective than tonic stimulation and is associ- placement. ated with higher patient satisfaction. Burst stimulation received FDA approval in October 2016. Although its long-­ Neurophysiologic Mechanisms of SCS: term efficacy has yet to be established, burst stimulation is a Subthreshold Versus Suprathreshold Stimulation promising form of SCS which may lead to a new generation In the context of this topic, any neurostimulation delivered at of paresthesia-free spinal cord stimulators acting both amplitude that fails to create paresthesia is considered to be through pre-existing mechanisms and by a novel supraspinal subthreshold. Likewise, any electrical stimulus with an mechanism, possibly at the level of the thalamus. amplitude sufficient to generate paresthesia is considered to be suprathreshold. Neurochemical Mechanisms of SCS Subthreshold SCS at high frequency (10 kHz) has been In addition to the neurocircuitry-mediated gating mechanism shown to reduce the excitability of lamina I pain projection described above, antidromic and orthodromic impulses from neurons when compared to sham in an in vivo model, but its SCS also relieve neuropathic pain by spinal segmental and exact mode of action remains unclear. Although this form of supraspinal mechanisms by restoring the balance of inhibi- SCS has performed well in several large prospective clinical tory and excitatory molecules in the neurochemical milieu of trials, as of the writing of this chapter, the PROCO trial has the dorsal horn to stabilize the gate via suppression of hyper- just released data suggesting there is no clinical benefit con- excitable nociceptive neurons. ferred by high-frequency stimulation. At a spinal segmental level, antidromic impulses propa- In addition to the mechanisms described above, supra- gating along Aβ fibers enter the dorsal horn provoking threshold SCS may also work through orthodromic signals release of inhibitory GABA, and reducing the release of glu- that activate frank motor activity in peripheral motor tamate, thereby restoring a neurochemical balance that more in patients with low-thoracic SCS systems. Efferent motor effectively suppresses the hyperexcitability of wide dynamic traffic has been observed even with low-intensity stimulation range (WDR) neurons that is thought to contribute substan- and is thought to be mediated by monosynaptic facilitation tially to central sensitization. These findings have been cor- of spinal motor strong enough to generate ortho- roborated by a number of biochemical and anatomical dromic action potentials. studies of the dorsal horn. 86 V. Sengupta et al.

SCS impulses propagated orthodromically along rostral reciprocally interacting brain structures is known as the dorsal column projections activate descending supraspinal “pain matrix.” The pain matrix consists of the anterior cingu- inhibitory pathways originating in the periaqueductal gray late cortex (ACC), posterior cingulate cortex (PCC), prefron- (PAG) and medulla where aminergic neurons of the locus tal cortex (PFC), parietal somatosensory cortex, insula, ceruleus (LC) and the nucleus raphe magnus (NRM). thalamus, hypothalamus, , , and peri- Through caudad projections that descend along the dorsal aqueductal gray (PAG). It is involved both in modulating the columns and terminate in the dorsal horn, these nuclei exert sensory-discriminatory element of pain at the level of the supraspinal neurochemical pain control by secreting inhibi- brainstem and in modulating the affective-emotional ele- tory , serotonin, and substance P into the SG, ments of pain at cortical and subcortical levels. thereby further stabilizing the gate. SCS has been shown to influence the activity of neurons in many of these structures, including the thalamus and Vascular Mechanisms of SCS somatosensory cortices. One PET study of nine patients In the case of vascular diseases such as chronic refractory with chronic neuropathic lower extremity pain revealed angina and peripheral disease, SCS acts by peripheral increases in cerebral blood flow to the contralateral thala- vasodilation and downregulation of efferent sympathetic mus, bilateral parietal association cortices, the anterior cin- activity. SCS has been shown to increase blood flow in der- gulate gyrus, and the , all components of matomes corresponding to the spinal segmental level of the pain matrix. In the thalamus and parietal association cor- stimulation. This observation by Cook et al. was subse- tices, SCS is thought to affect pain threshold. In the ACC quently applied in the treatment of peripheral vascular dis- and prefrontal areas, it may regulate the affective-emotional ease and chronic refractory angina. Furthermore, Linderoth aspects of pain. and Meyerson found that these effects were mediated both Another study used fMRI to examine the cortical and sub- by antidromic activation of small-diameter fibers and by cortical effects of SCS on ten patients with chronic pain, inhibition of sympathetic outflow. Later studies revealed that concluding that SCS reduces the affective component of pain low-intensity SCS triggers antidromic action potentials by reducing the connectivity between somatosensory and along Aβ fibers which, acting indirectly through interneu- limbic structures of the pain matrix, thereby disconnecting rons, trigger antidromic action potentials in unmyelinated the purely sensory signals from structures known to attach afferent fibers that then migrate to the periphery where they affective-emotional salience to the pain. cause the release of a powerful vasodilator known as calcito- nin gene-related protein (CGRP) at peripheral nerve termi- SCS for Chronic Visceral Abdominal Pain nals. CGRP then activates endothelial nitric oxide synthesis Chronic visceral abdominal pain evolves by similar mecha- and release, resulting in vascular smooth muscle relaxation nisms to those involved in peripheral sensitization in which and an increase in blood flow to the affected limb. tonic nociceptive stimulation leads to enlargement of recep- In the treatment of angina pectoris, SCS reduces anginal tive fields, increased recruitment of peripheral fibers, increase pain by redistributing coronary blood flow and by decreasing in the number of suprathreshold responses to sensory inputs, myocyte oxygen demand. Clinically, this is observed as an and spontaneous activation of neurons that are typically increase in the time to angina during exercise, an increased silent in response to nociceptive stimuli. The dorsal column resistance to critical ischemia, and an anti-arrhythmic effect has also been shown to transmit some visceral nociceptive achieved through modulation of cardiac neurons. Notably, afferents and can also amplify visceral pain. SCS does not mask the perception of true myocardial infarc- Overall, the mechanism by which SCS achieves visceral tion. SCS has been shown to reduce cardiac nociceptive pain abdominal pain relief is quite similar to that in somatic pain transmission, stabilize the cardiac conduction system, reduce control. One unique mechanism, however, is that SCS sup- infarct size through coronary flow redistribution, and reduce presses the visceromotor response caused by colonic disten- sympathetic outflow. In addition to reduced infarct size, the tion, thereby reducing the discomfort caused by beneficial effect of SCS manifests as a reduced magnitude of gastrointestinal hypermotility and dysmotility. Both thoracic ST segment changes during active ischemia and fewer atrial and lumbar lead placement in rat models have been shown to arrhythmias. ablate the visceromotor response. Clinically, SCS has been used to successfully treat mesen- Higher-Order Supraspinal Mechanisms: teric ischemic pain, esophageal dysmotility, gastroparesis, The Pain Matrix IBS, chronic pancreatitis, familial Mediterranean , post-­ In addition to the brainstem and subcortical structures dis- traumatic splenectomy, generalized chronic abdominal pain, cussed above, SCS is thought to interact with and modulate a and chronic pelvic pain. However, there is a paucity of pro- broader constellation of higher-order structures implicated in spective, randomized controlled trial data to support wide- the modulation of pain. This group of functionally related, spread use at this time. 10 Pain-Relieving Mechanisms in Neuromodulation 87

Mechanisms of SCS Non-responsiveness are slim and flexible, allowing their placement without com- Non-responsiveness to SCS in neuropathic pain is poorly pression of the DRG itself. understood. In the setting of CRPS, it occurs in roughly 1/3 Neuroanatomically, there are a number of factors that make of patients and has been reliably connected to the severity of the DRG an apt target for the neuromodulation of pain. First, . It is thought that severe allodynia arises from it is a site where the bulk of ascending pain inputs, most of extreme depletion of the inhibitory neurotransmitter GABA them bound for the spinothalamic tract, as well as neutral in the dorsal horn. This has been demonstrated in a rat model mechanoreceptive stimuli bound for the dorsal column, con- where experimental depletion of inhibitory neurotransmit- verge before somatotopically diverging to ventral and dorsal ters caused severe allodynia that could not be alleviated by zones of the spinal cord. Its superficial dorsal location along a SCS. In another experimental study, the intrathecal delivery relatively thin overlying layer of and CSF and low of the GABA-B was shown to convert SCS surrounding tissue impedances render the DRG more suscep- non-responders into SCS responders. tible to stimulation with lower energy levels that average The timing of SCS placement also appears to play a cru- around 15% of the power output required for dorsal column cial role in its chances of success. Indeed, it has been clini- stimulation. The convergence of all sensory pathways at this cally observed that SCS is most effective when started point allows the stimulation of spinothalamic fibers that are 7 months to 1 year after the onset of neuropathic pain. typically inaccessible because they travel in close proximity to Central sensitization appears to play a role in the evolution the descending motor neurons of the . of SCS non-responsiveness, as it is likely a progressive pro- Targeting a single nerve root permits greater anatomic speci- cess mediated in part by the depletion of neurotransmitters ficity for pain control, allowing the stimulator to focus on a essential to SCS responsiveness. One study found that SCS foot or toe rather than an entire limb. Finally, paresthesia fields works in part by reversing GABA depletion and glutamate in DRG tend not to vary with position, a benefit that occurs excess known to occur during early central sensitization. because the lead is deployed into a bony vertebral foramen Another found that GABA supplementation soon after cen- that is oriented at a 90-degree angle to the spinal cord. tral cord injury could reverse hyperalgesia but could not do DRG stimulation is thought to actively reduce the net so later in the process. This model of central sensitization as nociceptive activity to the spinal cord by restoring the neural a progressive disease process is borne out in the natural his- filtration function served by the normal DRG. Notably, this tory of neuropathic pain syndromes. For example, after a mechanism is distinct from the gate control mechanism by period of rapid progression lasting around 1 year, the symp- which activation of large-diameter sensory fibers blocks toms of CRPS I typically stabilize, thereafter progressing transmission of pain signals in small-diameter fibers at the quite slowly or not at all. level of the dorsal horn. Indeed, DRG stimulation appears to The level of mechanical allodynia is a clinical predictor of normalize peripheral input before it arrives at the gate. SCS non-responsiveness. This observation may be related to DRG stimulation has demonstrated 1-year efficacy in the the fact that mechanical allodynia is conveyed not by C or Aδ treatment of a number of clinical entities, including radicular fibers, but by large myelinated βA fibers, only a small frac- pain, failed back surgery syndrome, complex regional pain tion of which are activated by SCS. syndrome of the lower extremities, and chronic postsurgical Finally, SCS failure may be attributable to suboptimal pain. The targeted nature of DRG stimulation allows the physical and technical factors, including but not limited to treatment of pain in subdermatomal distributions with pre- the inaccurate placement of electrodes and excessive thick- cisely sculpted paresthesia fields. Examples include the ness of the highly conductive intervening dorsal CSF layer, treatment of foot and post-herniorrhaphy groin pain. Other which can cause dorsal root fiber activation through lateral case reports have described DRG stimulation as therapy for dispersion of the SCS current. The target site for stimulation , postherpetic neuralgia, phantom limb pain, depends on the . Caudally, the dorsal column visceral pain, somatic body wall pain, upper extremity pain, afferents are located at the surface of the cord, but more ros- complex regional pain syndrome of the knee, and postsurgi- trally, they course ventrally, which typically makes caudal cal knee pain. sites more sensible locations for electrode placement.

Peripheral Neuromodulation and Its Dorsal Root Ganglion (DRG) Stimulation Mechanisms of Action

With dorsal root ganglion (DRG) stimulation, pain control is Peripheral neuromodulation is possibly the most diverse and rap- achieved by placing an electrode, typically by an anterograde idly expanding area of neuromodulation. It is comprised of a percutaneous approach, into the of the neural variety of techniques including peripheral nerve stimulation, foramen directly over the dorsal root ganglion. These leads peripheral nerve field stimulation, transcutaneous electrical 88 V. Sengupta et al. nerve stimulation, and functional electrical stimulation. Since the pain in the trigeminal and cervical distributions. Therefore, first percutaneous stimulator was used for chronic headaches in electrical stimulation of the occipital nerves, which is superfi- 1999, peripheral neuromodulation has been used to treat an ever- cial and easily accessed branch of C2, can be used to modulate expanding variety of nerves and nerve plexuses to treat neuro- pain in those distributions. A PET study of eight patients with pathic, visceral, cardiac, abdominal, low back, and facial pain. chronic treated with ONS demonstrated changes in regional blood flow to the dorsal rostral pons, the anterior cin- gulate cortex, and the , all structures involved in central Peripheral Nerve Stimulation (PNS) pain processing. The clinical efficacy of ONS has been evalu- ated by three large controlled studies, the ONSTIM and PRISM Peripheral nerve stimulation is a neuromodulation technique trials, and another randomized study with mixed results. in which electrodes are placed percutaneously directly along the course of peripheral nerves and electrical current is Sphenopalatine Ganglion (SPG) Stimulation applied to alleviate chronic pain. The prevailing hypothesis for Cluster Headaches for the mechanism of pain-relieving action by PNS is via Another target of interest for PNS in the treatment of refrac- dorsal horn gate closure provoked by orthodromic action tory cluster headaches is the sphenopalatine ganglion (SPG), potentials that travel along Aβ fibers, entering the dorsal horn which lies in the pterygopalatine fossa providing postgangli- from the periphery. Just as with central neuromodulation, onic innervation and sensation to the facial, meningeal, and however, the mechanism likely varies based on the clinical cerebral blood vessels. Local ischemia caused by vasospasm entity being treated and its unique underlying . of these vessels is thought to underlie the pain caused by PNS may also cause analgesia by neurochemical mecha- cluster headaches. Furthermore, cluster headaches have been nisms by influencing the local concentration of neurotrans- successfully treated by SPG blockade and radioablation, and mitters, neuromodulators, and inflammatory molecules that animal models have shown that SPG stimulation can increase mediate the pain response. Animal studies have revealed that regional blood flow, thereby reversing vasospastic ischemia. repeated stimulation of peripheral nerves results in decreased C-fiber response to pain at the level of the spinal cord. Trigeminal, Facial, and Mandibular Nerve At low-frequency stimulation, PNS is believed both to mod- Stimulation for Facial Pain ulate downstream nociceptors and to promote favorable central PNS has been used to treat facial pain in a variety of condi- pain processing upstream. Low-frequency stimulation of Aδ tions and involves stimulation of the peripheral nerves that fibers in a rat model caused long-term depression of monosyn- provide sensory innervation to the painful area. Electrodes aptic and polysynaptic excitatory postsynaptic potentials in the have also been placed in the V1, V2, and V3 trigeminal dis- substantia gelatinosa. In a cat model, stimulation of the poste- tributions to achieve pain relief in . rior tibial and sciatic nerves resulted in decreased C-fiber Direct stimulation of the has also been response to pain at the level of the spinal cord, thereby impli- performed for trigeminal neuralgia, poststroke pain, periph- cating a spinal pathway. At ultrahigh stimulation frequencies, eral nerve injury, and postherpetic neuralgia. PNS is thought to cause direct blockade of the peripheral nerve. One case series studied PNS for poststroke facial pain and A promising development in the field of PNS has been the postherpetic neuralgia with five of seven in the poststroke discovery that, unlike SCS, it may have antinociceptive prop- group experiencing relief and none of those in the posther- erties extending beyond ischemic pain. Ellrich and Lamp petic neuralgia group experiencing pain relief. used laser infrared stimuli to activate Aδ and C fibers in the superficial radial nerves of subjects. Laser stimula- Single Nerve and Nerve Plexus Pain tion provoked painful prickling sensations, and associated Through the use of motor stimulation, ultrasound, and ana- cortical evoked potentials were observed. However, when tomical landmarks, percutaneous electrodes can now safely peripheral stimulation was applied, the same painful stimu- be placed along any number of painful peripheral nerves or lus provoked a diminished perception of pain and was asso- nerve plexuses. The first reports included supraorbital ciated with reduced evoked potentials and latencies. implantation, but an expanding variety of single peripheral nerves have now been reported including median, ulnar, sci- Headache: Occipital Nerve Stimulation atic, genitofemoral, and ilioinguinal. for Occipital Neuralgia, Chronic Migraine, and Cluster Headache The mechanism of action of ONS in occipital neuralgia and Peripheral Nerve Field Stimulation (PNFS) chronic migraine relies on the anatomic convergence of trigem- inal, dural, and cervical afferents in the brainstem. Many patients have neuropathic pain occurring in non-­ Experimentally, it has been shown that activation of afferents dermatomal distributions, instead of occurring in areas that from the caudal trigeminal nucleus at the level of C2 can induce either span multiple dermatomes or where the receptive fields 10 Pain-Relieving Mechanisms in Neuromodulation 89 of several peripheral nerves overlap. In such cases, SCS and Medial Branch of the Dorsal Rami of Spinal PNS can fail to capture the irregular pain patterns, while PNFS Nerves can at times provide relief. PNFS does not target a specific nerve or sensory distribution. Instead, the lead is positioned The medial branches of the dorsal rami of spinal nerves pro- within the tissue in order to stimulate the network of cutane- vide sensation to the facet joints and motor innervation of the ous afferents that delineate an area of pain. PNFS is thought to multifidi, small deep spinal muscles known to play a role cause analgesia either by orthodromic Aβ-mediated gating in local segmental spinal stability and proprioception. mechanisms in the dorsal horn or by provoking the release of Ultrasound and EMG studies have shown that multifidi atro- . By placing the electrode at the epicenter of pain, phy and fail to fire appropriately after back injuries. These relief can often be provided to an entire painful area. changes are thought to cause instability that renders subjects more susceptible to repeated back injuries and chronic pain. A new device is being developed to electrically stimulate Transcutaneous Electrical Nerve Stimulation these muscles with the hope of treating axial back pain by (TENS) restoring dynamic spinal stability.

TENS is a form of neuromodulation in which electrical stim- ulation is delivered by adhesive leads placed on the over painful areas. More recently, TENS has emerged as a moder- Key Points ately effective treatment for pain. TENS is thought to act via –– The use of electrical stimulation to treat pain is a local, spinal, and supraspinal pathways. Low-frequency century-old practice. TENS (<10 Hz) appears to promote signaling via u-opioid, –– To understand the mechanisms of neuromodulation, GABA, serotonin, M1, and M3 receptors. The evolution of it is crucial to first understand the neurophysiologi- tolerance to TENS suggests the involvement of descending cal, neurochemical, and vascular processes and the inhibitory pathways as well. neuroanatomy pertinent to the transmission of pain, both in disease and in health. –– Spinal cord stimulation and other forms of neuro- Functional Electrical Stimulation (FES) modulation work through multiple mechanisms which can be categorized by type of process (neuro- Functional electrical stimulation (FES) is the transcutaneous physiological, neurochemical, and vascular) and application of controlled electrical stimulation to generate also by (local, spinal segmental, and contractions and functional movement in paralyzed muscles, supraspinal). thereby facilitating and improving the mobility of limbs, and –– The prevailing mechanism of action leading to other body functions lost due to injury, including respiratory, neurostimulation-mediated pain relief often varies sexual, bladder, and bowel function. This is an emerging tech- based on the pathophysiology of the underlying nology recently improved with brain-computer interfacing painful disease state. that is thought to reduce pain and improve function by work –– Gate control theory postulates that activation of by restoring the activity of deconditioned and unused large-diameter sensory fibers could block the trans- muscles. mission of pain signals conveyed by small-diameter fibers and is the target of action for multiple forms of central and peripheral neurostimulation. Vagal Nerve Stimulation (VNS) –– There is a preponderance of data to support the effi- cacy of neurostimulation and to corroborate its Vagal nerve stimulation (VNS) is electrical stimulation of underlying mechanisms of action, but much investi- the left vagal nerve. VNS has a well-established role in the gation still needs to be done. The most well-vali- treatment of epilepsy, but a possible application in the treat- dated form of neuromodulation is spinal cord ment of pain has emerged more recently. Its mechanism of stimulation. action is unknown, but is thought to be related to synchroni- –– Neurostimulation is likely to continue to expand in zation and desynchronization of vagal sensory afferents with importance over the coming years. cerebral activity. 90 V. Sengupta et al.

Recommended Reading 6. Lee S, Abd-Elsayed A. Some Non-FDA approved uses for neu- romodulation: a review of the evidence. Pain Pract. 2016;16(7): 935–47. 1. Ashkan K, Rogers P, Bergman H, Ughratdar I. Insights into 7. Liem L. Stimulation of the dorsal root ganglion. In: Stimulation of the mechanisms of deep brain stimulation. Nat Rev Neurol. the peripheral , vol. 29. Basel: Karger Publishers; 2017;13:548. 2016. p. 213–24. 2. Chang Chien GC, Mekhail N. Alternate intraspinal targets for 8. Melzack R, Wall PD. Pain mechanisms: a new theory. Surv spinal cord stimulation: a systematic review. Neuromodulation. Anesthesiol. 1967;11(2):89–90. 2017;20:629–41. 9. Shamji MF, De Vos C, Sharan A. The advancing role of neuromod- 3. Gildenberg PL. History of . In: A history of ulation for the management of chronic treatment-refractory pain. . Park Ridge: American Association of Neurological Neurosurgery. 2017;80(3S):S108–13. Surgeons; 1997. p. 439–64. 10. Shealy CN, Mortimer JT, Resnick JB. Electrical inhibition of pain 4. Goroszeniuk T, Pang D. Peripheral neuromodulation: a review. by stimulation of the dorsal columns: preliminary clinical reports. Curr Pain Headache Rep. 2014;18(5):412. Anesth Analg. 1967;46(4):489–91. 5. Imran TF, Malapero R, Qavi AH, Hasan Z, de la Torre B, Patel YR, 11. Vallejo R, Bradley K, Kapural L. Spinal cord stimulation in chronic Yong RJ, Djousse L, Gaziano JM, Gerhard-Herman MD. Efficacy pain: mode of action. Spine. 2017;42:S53–60. of spinal cord stimulation as an adjunct therapy for chronic refrac- tory angina pectoris. Int J Cardiol. 2017;227:535–42.