Because of the technological evolution and the consequent possibility of electrically stimulating more specific anatomic structures, such as the dorsal root ganglion (DRG), there has been a shift in indications for neurostimulation from predominantly back and leg pain towards other indications, such as postoperative neuropathic pain and more dermatome-specific neuropathic and nonneuropathic pain syndromes like foot pain, groin pain, etc.
The DRG is an important structure involved in relaying sensory information from the periphery to the central nervous system. Stimulating the DRG functions as a low-pass filter at the level of the T-junction for spontaneous action potentials in chronic neuropathic pain. The DRG is also internally organized, allowing for the treatment of very specific targets. The development of a neurostimulation system specifically targeting the DRG, the main indications for the procedure, and the results of clinical are described in this chapter.
Keywordschronic pain, dorsal root ganglion, indications, mechanism of action, neurostimulation
Electricity has been used for the management of pain since ancient times. Perhaps most notably, the Greeks used the electric ray to numb the pain of childbirth and operations. In the Roman literature, we find referral to the use of torpedo fish for the treatment of headache and gout. Over the past 100 years, electricity has been used in different forms and applications for the management of pain.
The publication of the gate control theory by Melzack and Wall, which postulates that activity in large diameter Aβ afferent fibers attenuates spinal ascending pain transmission by activating inhibitory interneurons in the dorsal horn of the spinal cord, is believed to be the starting point for the development of spinal cord stimulation (SCS). In 1967 Shealy reported the first cases of refractory cancer pain treated with dorsal column stimulation. In 1989 SCS was approved by the US Food and Drug Administration (FDA). Since then, SCS has been used for the management of different chronic pain syndromes.
In a search for improving clinical outcomes and reducing complications, electrodes and pulse generators have progressively improved. The electrodes, initially applied directly to the dorsal column in a subdural location, were later inserted epidurally and even ventrally. The shape, size, and number of contacts were adapted to allow percutaneous lead placement and programmable stimulation. The available data suggest benefit for the use of SCS for failed back surgery syndrome (FBSS), complex regional pain syndrome (CRPS), angina pectoris, and nonsurgically correctable chronic critical leg ischemia.
Clinically, tonic SCS targets specific segmental levels to produce paresthesias overlapping the painful area. However, the anatomic specificity is not exact, which makes the localization of the lead in the spinal canal difficult. Finding the exact balance between paresthesia coverage of the painful area and the acceptable level of tingling sensation may prove to be difficult. For example, this is the case when paresthesia is required in the foot, hand, or groin. Inadequate coverage is one of the major reasons for SCS treatment failure. Even when coverage is initially optimal, movement of the lead over time has been reported with consequent loss of effect.
Patients treated with tonic SCS may experience position-related changes in the perception of neurostimulation. Changes in position from upright to supine or prone may provoke an uncomfortable increase in stimulation or a loss of paresthesia. These changes in stimulation are mainly due to changes in the thickness of the cerebrospinal fluid (CSF) layer interposing the spinal cord and the epidurally placed electrodes. Adapting stimulation parameters according to position can mitigate this problem. New stimulation paradigms, including the high-frequency SCS (10 KHz) and burst stimulation allow SCS without the need for paresthesias. Some reports demonstrate good results with these devices for neuropathic pain in the limbs. Although tonic SCS has been reported to be effective in the management of different chronic neuropathic pain syndromes, difficulty in finding the correct lead placement and stimulation parameters has prompted the search for other stimulation targets.
Dorsal Root Ganglion “Gatekeeper” in Pain Transmission
The dorsal root ganglion (DRG) has been compared with a gatekeeper, a railway marshaling station, and a highway intersection, among others. All of these terms indicate the key role the DRG plays in relaying sensory information from the periphery to the central nervous system (CNS).
Anatomy of the Dorsal Root Ganglion
The right and left paired “mixed” spinal nerves carry autonomic, motor, and sensory information between the periphery and the spinal cord. These spinal nerves are composed of afferent sensory dorsal axons (dorsal root) and motor ventral efferent axons (ventral root). They emerge from the intervertebral neural foramina between adjacent vertebral segments. The dorsal sensory roots exit the neural foramina to form the DRG.
The DRG is located bilaterally on the distal end of the dorsal root in the anterolateral epidural space. In humans, there are 8 paired cervical, 12 paired thoracic, 5 paired lumbar, and 4 paired sacral DRGs. The morphology of the DRG is longer and wider in more caudal locations. It typically lies within the intervertebral foramen, behind the vertebral artery, and is protected by the vertebrae. In healthy individuals the DRGs from L1 to L4 are mostly foraminal, and L4 DRG lies foraminal. At L5 the DRG lies intraforaminally or in a minority of the cases intraspinally. At S1, most DRGs are intraspinal.
The DRG consists of a collection of bipolar cell bodies of primary sensory neurons surrounded by glial cells and the axons of the DRG sensory cells that form the primary afferent sensory nerve. The DRG neurons are pseudounipolar neurons composed of two branches, a distal and a proximal process, connected by an outgrowth cell body. The DRG contains the largest proportion of peripheral sensory nerves in the body. These cells are responsible for transducing visceral and somatic sensory information from the periphery and transmitting this information to the CNS. The cell bodies actively participate in the signaling process by sensing certain molecules and manufacturing other molecules that modulate the sensory transduction process. They are surrounded by layers of satellite glial cells (SGCs). The DRG is not protected by a blood nerve barrier, allowing small and large molecules and even macrophages to cross the SGC wrap of the DRG neuron.
SGCs form a functional unit of the sensory neuron within the DRG, and “instruct the nervous system what to do.” Stimulation of the DRG neuron triggers a delayed and long-lasting response via a pathway between glia, called the “sandwich synapse” (SS). The receptors on the SGCs that bind such molecules as chemokines, cytokines, adenosine-5′-triphosphate (ATP) and bradykinins participate in the transmission process within the DRG. Glial cells change in morphology and biochemical function after a peripheral afferent fiber injury and participate actively in central and peripheral nervous system processes.
Role of the Dorsal Root Ganglion in Impulse Propagation
Primary sensory neurons start at the peripheral receptive field of the neuron, and their cell body resides in the DRG. A stimulus in the periphery alters the firing of the neuron and propagates into the CNS, beginning in the dorsal horn of the spinal cord and ending in the relevant portions of the thalamus and brain. The pseudopolar neurons of the DRG lie within the DRG and have axons that split peripherally and centrally away from the soma ( Fig. 75.1 ). At the branch point, the so-called T-junction of the DRG can act as an impediment to electrical impulse from the nociceptor to the dorsal root entry zone; specifically, it can alter the propagation of the electrical pulse by acting as a low-pass filter to electrical information from the periphery ( Fig. 75.2 ).
Thermomechanical reception occurs in small unmyelinated, nociceptive C-fiber cells in the DRG. These cells contain substance P or calcitonin gene-related peptide (CGRP), which function as neuromodulators and neurotransmitters. The terminals of the larger, myelinated A-fiber neurons are low-threshold mechanoreceptors.
When primary sensory neurons are injured, Schwann cells and SGCs in the DRG release proinflammatory mediators, such as eicosanoids, bradykinins, serotonin, neurotrophins, cytokines (e.g., interleukins, tumor necrosis factor [TNF]-α), interferons, growth factors, chemokines, ATP, and reactive oxygen species.
Ion channels and receptors that are located in the primary sensory neurons have three functions: transduction, transmission, and modulation of sensory information. Transduction of noxious information to electrical signals at the peripheral terminals of the DRG include transient receptor potential channels, Na + channels, acid-sensing ion channels, and ATP-sensitive receptors. Propagation of action potentials involves Na + and K + channels, whereas the modulation of synaptic transmission is performed by voltage-gated Ca ++ channels and glutamate receptors. The latter are expressed on presynaptic membranes at the terminal of the primary afferents of the dorsal horn.
Somatotopy of the Dorsal Root Ganglion
The cell bodies within the DRG are somatotopically organized. The topographic distribution of sciatic and femoral nerve sensory neuronal somata in the L4 DRG of the adult rat was mapped after retrograde tracing. The tracers were applied to the proximal transected end of either nerve alone, or from both nerves in the same animal using separate tracers. Three-dimensional reconstructions of the distribution of labeled neurons were made from serial sections of the L4 DRG, which is the only ganglion that these two nerves share. The results showed that with little overlap, femoral nerve neurons distribute dorsally and rostrally, whereas sciatic nerve neurons distribute medially and ventrally ( Fig. 75.3 ). This finding indicates the existence of a somatotopical organization for the representation of different peripheral nerves in dorsal root ganglia of adult animals. Theoretically, this makes it possible for precision targeting.
Dorsal Root Ganglion as Target for the Treatment of Chronic Pain
The accessibility of the DRG and its critical role in the transmission and transduction of pain make it a preferred target for the treatment of chronic pain ( Fig. 75.4 ).
Dorsal Root Ganglionectomy
Dorsal root ganglionectomy was popular during the 1960s and 1970s. However, this technique has some limitations, which has curtailed its use. It requires open surgery and far lateral laminectomy and foraminotomy to expose the DRG. Even in settings of one level radicular pain, a single-level ganglionectomy is seldom efficacious because of the fact that peripheral pain afferents send fibers to multiple adjacent DRGs. It has therefore been suggested that to improve outcomes, the DRG that somatotopically matches the pain distribution as well as ganglia rostal and caudal should be resected. This makes the procedure a large, multilevel, and potentially morbid operation. The outcomes for ganglionectomy have mainly been reported in case series and retrospective studies for the management of neuropathic pain conditions such as CRPS, chest wall pain, monoradicular sciatica, occipital neuralgia, and cervicogenic headache with a neuropathic component. The quality of these studies does not allow one to formulate firm conclusions regarding the effectiveness of this intervention for the treatment of intractable pain. As alluded to earlier, a single-level ganglionectomy is not sufficient to control the pain because of convergence of sensory inputs. The complications of ganglionectomy include the occurrence of deafferentation pain, dysthesias, allodynia, and even CRPS.
Radiofrequency Techniques Adjacent to the Dorsal Root Ganglion
Radiofrequency (RF) treatment consists of the application of a high-frequency electrical current close to a nerve(s) or ganglion. The high-frequency current generates heat, which in turn results in destruction of the nerve tissue. RF treatment was initially thought to be selective for the small Aδ fibers. Several studies have evaluated conventional RF for the treatment of radicular pain. A well-designed randomized controlled trial failed to demonstrate the superiority of RF treatment compared with placebo for the management of lumbosacral radicular pain. However, a smaller, earlier randomized controlled trial evaluating RF treatment adjacent to the cervical DRG showed a positive effect compared with placebo. Another trial compared RF treatment at 67°C with RF treatment at 40°C and found clinical improvement in both groups.
The possibility of heat-induced large nerve lesions resulting in deafferentation pain led to the development of pulsed radiofrequency (PRF), whereby the highfrequency current is applied in short bursts followed by a longer silent period. This allows for dissipation of the heat, thereby maintaining the electrode tip below the neuroablative temperature of 42°C. In a small, double-blind study performed in patients with chronic cervical radicular pain, PRF treatment adjacent to the cervical DRG was found to be superior to sham RF. An uncontrolled, prospective study evaluating the treatment of lumbosacral radicular pain with PRF treatment adjacent to the lumbar DRG demonstrated a good clinical outcome in more than 50% of patients, lasting until 6 months after treatment.
Neurostimulation of the Dorsal Root Ganglion
Neurostimulation is a minimally invasive, reversible treatment for chronic pain. Over the years the value of SCS for the treatment of chronic neuropathic pain syndromes has been documented in randomized controlled trials and prospective and retrospective studies.
The drawbacks to this treatment option include the following:
Some patients experience only partly or even no reduction in their pain.
Achieving satisfactory paresthesia coverage of the painful area may be difficult.
Even when paresthesia coverage is initially achieved, the lead may be displaced over time.
Patients may experience position-related changes in stimulation that is bothersome and requires the stimulation program to be adjusted.
Stimulation of the DRG is an appealing alternative to SCS because stimulation at this site involves spinal and supraspinal mechanisms dependent on Aβ fibers. The DRG contains sensory neuron somata for all sensory modalities and fiber types, and stimulation of the sensory afferent units allows direct modulation of Aδ- and C-fibers. Sensory neuron stimulation diminishes neuronal excitability, thus interfering with pain pathogenesis.
In Vitro Studies
In vitro studies have shown that ganglion field stimulation at 30 V or more generates an immediate response and elevated Ca ++ levels throughout the treatment period, which returns to baseline, occurred within 60 seconds of terminating stimulation.
Ganglion field stimulation suppresses action potential firing, which is maintained throughout the stimulation period. The conduction velocity of slow-conducting fibers is significantly reduced after ganglion field stimulation. It has been speculated that clinical electrical stimulation of the DRG activates sensory neurons, causing Ca ++ influx that triggers Ca ++ -dependent processes, which in turn leads to decreased somatic excitability and increased filtration of high-frequency afferent action potential trains. This, combined with dorsal horn modulation of nociceptor input from activation of low-threshold units, may produce beneficial effects in humans with chronic pain.
Dorsal Root Ganglion Stimulation in Humans
The characteristics of the DRG and findings from in vitro pain models make it a logical target for neurostimulation. Case reports on cervicogenic headache, postherpetic neuralgia and discogenic pain describe significant pain reduction. However, the conventional equipment used for midline dorsal column stimulation is not readily adaptable to placement at the DRG. The leads must be manually curved, and the dimensions, rigidity, and spaces between contacts may compress the DRG, thereby recruiting unwanted targets, including the dorsal root entry zone (DREZ) and motor fibers. Moreover, traditional epidural equipment often results in motor stimulation. Therefore specialized and modified equipment are required.
The lead should be flexible and its form adapted to the curve of the DRG. The space between the contacts should allow adapted somatotopic stimulation and the system must enable the practitioner to make incremental changes in stimulation output so as to allow fine adjustments in current delivered to neural tissues. Considering the dermatomal and segmented anatomic distribution of sensory afferents, the pulse generator should allow multiple leads to be connected.
The St. Jude Medical Axium neurostimulator system consists of a stimulator device (an external trial neurostimulator [TNS] that is used for the trial period, followed by an implanted neurostimulator [INS] if successful), and up to four quadripolar percutaneous leads and wireless patient- and clinician-programmer devices. Both TNS and INS are constant voltage devices.
Under monitored sedation, leads are placed via an epidural approach, with access gained using the loss-of-resistance technique standard for this type of intervention. Leads are advanced in an anterograde or retrograde fashion and then steered under fluoroscopic guidance into the intervertebral foramen near the DRG. Appropriate lead position is determined through intraoperative device programming after the patient is awakened to confirm paresthesia overlap with the painful regions. If pain-paresthesia overlap is not achieved through programming, the leads are repositioned under fluoroscopy and programmed again. The DRG is in a consistent location anatomically; thus lead position should accurately reflect the ability to stimulate the ganglion. Due to the fact that cell bodies are present in the ganglion and not in the nerve root, and because many membrane alterations occur in the cell bodies of primary sensory neurons and not nerve roots, there are electrophysiologic differences between these structures. In part, generating an electrical field around a ganglion provides an enhanced ability to provide acute and specific subdermatomal coverage compared with a nerve root. Although prior investigators had tried DRG stimulation, there were limitations in both the logistics of lead placement, as well as the ability to provide tailored stimulation therapy.
There is a steep learning curve involved with the placement of DRG leads. Entry into the neuroforamen can be difficult, depending on the level at which one is implanting the lead. For example, entering the L4-5 or L5-S1 neuroforamen can be challenging in patients with foraminal stenosis or degenerative disc disease. Therefore many investigators perform the trial using an extension lead that can be converted to a permanent lead. Based on patient feedback on paresthesia coverage of the painful area during the procedure, the external nerve stimulator is programmed. The second phase consists of connecting the lead to the battery with or without an extension lead, depending on where the pocket site is (i.e., gluteal or abdominal). When strain relief loops are used, there is no need to anchor the lead ( Figs. 75.5–75.9 ).