Introduction

With the introduction of new medical therapies, there is an inevitable growth of fundamental knowledge and improvements in outcomes. Dorsal root ganglion stimulation (DRG-S) is a novel form of neuromodulatory therapy, that, rather than placing the electrical field over the dorsal columns of the spinal cord, as with conventional, tonic spinal cord stimulation (t-SCS), is placed near the cell nuclei of the afferent neurons of the dorsal root ganglia (DRG). Given its novelty, the corresponding body of evidence is evolving compared to longer-employed therapies. The pivotal ACCURATE study, published in 2017, demonstrated superiority of DRG-S to t-SCS in the treatment of complex regional pain syndrome (CRPS) and causalgia.1 Additional studies have shown results differing from past t-SCS work, from potential treatment indications to functional and psychological outcomes. Supporting basic science work is also demonstrating that on a cellular, mechanistic level, DRG-S is functioning in a different manner.

Stimulating at the DRG is currently accepted as an alternative form of neuromodulation used for complex regional pain syndrome I and II and dermatomal pain syndromes, although additional benefits compared to t-SCS are becoming apparent. Targeting the somata of the pseudo-unipolar afferent nerve fibers allows access to all nerve fiber types. Surrounded by dura mater and only partially by a thin layer of cerebrospinal fluid (CSF), DRG-S utilizes a fraction of the energy required. Both modalities utilize a tonic waveform; however, DRG-S has efficacy when delivered in a subthreshold, paresthesia-free manner, at frequencies as low as 4 Hz, and when delivered intermittently.2–4 None of these are possible with conventional t-SCS.

Early basic science work identified filtering of afferent fibers at the DRG as the basic mechanism of action.5–7 However, filtering afferent signaling does not explain the effects of DRG-S on broad, multi-dermatomal conditions with a single lead placement.8,9 These effects suggest orthodromic effects from neurostimulation, an accepted mechanism of t-SCS, but rarely discussed with DRG-S. Single lead coverage of low back pain led to a theory that stimulating the DRG of the cutaneous branches of a dermatome led to inhibition at points of convergence centrally in the dorsal horn (DH).10 Additional antidromic effects from DRG-S on the sympathetic nervous system, namely its ability to influence neuroinflammation and tissue perfusion in peripheral vascular disease, are only beginning to be explored.11–14

While these potential clinical benefits, alternative mechanisms, and potential expanded indications are worthy of excitement, a corresponding growth of prospective randomized research is still developing. This best-practice manuscript is meant to serve as a guide to the current state of the use of DRG-S, utilizing peer-reviewed literature, a reliance on clinical evidence, experience, and expert opinion, to assist practitioners in the application of this therapy in the safest, most efficacious manner, and to ensure the highest level of patient care. This builds on previous work by Deer et al, with an update of recent evidence.15

Anatomy and Function of the DRG

The DRG is the bundle of cell bodies of sensory afferent neurons situated at the intervertebral foramen of each of the 31 pairs of human spinal nerves. Sensory information travels centrally along a mixed spinal nerve to the DRG, where afferent axons bifurcate to give rise to a stem axon and a central axon, the T-junction.16

At the DRG, the perineurium encasing the spinal nerve root is replaced by a thicker layer of dura mater, with the arachnoid space starting at the proximal segment of the DRG.17,18 Thereafter, the dorsal and ventral roots travel in the subarachnoid space to their appropriate vertebral levels to split into nerve rootlets, entering at separate levels of the spinal cord. See Figure 1.

Figure 1 Representative sections through immune-stained lumbar DRG showing the CSF occupying sub-arachnoid space (SAS) stopping at the DRG as a fold-like recess (A and B arrows). The SAS is in direct contact with DRG between the spinal nerve roots (A and Barrow heads) and in the angle between the dorsal root and surface of DRG . The scheatic drawing (C) summarizes the position of the CSF delimited by the arachnoid (dotted green line) in relation to DRG. Continuous black and dashed red lines indicate dura and pia mater, respectively. Reprinted from Ann Anat, 205, Joukal M, Klusáková I, Dubový P. Direct communication of the spinal subarachnoid space with the rat dorsal root ganglia. 9–15, Copyright 2016, with permission from Elsevier.18

The DRG also contains non-neuronal cells and tissues vital to supporting and protecting the functions of the nucleus. Each neuronal cell body is encased by satellite glial cells. These cells support the somata and have interconnecting gap junctions which are involved with the sensitization of adjacent neurons.19,20 See Figure 2. In addition to connective tissue, there is a group of immune cells consisting mainly of macrophages and lymphocytes, blood vessels, and bundles of sympathetic nerve fibers.21

Figure 2 (A) Schematic drawing of a DRG neuron cell bodies (N1-5) which are enveloped by a sheath of satellite glial cells (SGC 1+2, black and separated by connective tissue). (B) Scanning electron micrograph showing an intricate meshwork of sympathetic fibers forming a terminal Dogiel’s nest around the cell body of a rat DRG neuron. Scale bar, 10 μm. Images reprinted from Nascimento AI, Mar FM, Sousa MM. The intriguing nature of dorsal root ganglion neurons: linking structure with polarity and function. Prog Neurobiol. 2018;168:86–103.21

Function of Somata Within the DRG

The DRG has an active role in the processing of stimuli as well as up- and down-regulation of pain. After inflammation, partial nerve injury, or direct compression of the DRG,22,23 a portion of its afferent cell bodies may become sufficiently hyperexcitable and generate ectopic action potentials (APs).24 These ectopic APs are relayed to the T-junction where they may collide with incoming APs to decrease neural input.25 Action potentials can also be initiated in adjacent, unaffected C-fibers in the DRG and propagate centrally to cause signs and symptoms associated with neuropathic pain.21,23,26 Pain-transmitting C-fibers are particularly prone to the effects of ectopic APs.5,7,27 See Figure 3A.

Figure 3 (A) Illustration representing filtering at the DRG in the physiologic state. Note the diameter difference between the peripheral and central axon (B). The electrical field applied to the DRG decreases propagation of sensory input. (C) Represents the orthodromic and antidromic propagation of action potentials induced by DRG-S.

APs are initiated in the peripheral terminals of DRG neurons, although there are some small-fiber afferent sympathetic fibers that are capable of sending APs back down the same axon. This process is known as the dorsal root reflex (DRR).28 The DRR is utilized in peripheral vascular innervation, where the release of substance P and calcitonin gene-related peptide (CGRP) from the peripheral terminals contributes to vasodilation and neurogenic inflammation.29 This system allows for the effects of antidromic propagation of APs with DRG-S. See Figure 3B and C.

By increasing or decreasing afferent signaling through ectopic APs, the DRG is positioned to serve as the “gatekeeper” of information from the external and internal environments to the CNS.

DRG Size

The size of the DRG is correlated to the number of neurons it contains and varies depending on its vertebral level.30 In the cervical region, DRG size increases from the very small and absent in >70% at the level C1 to the largest at the C8 level.31 The C8 DRG also has a larger volume (177 mm3) than the adjacent T1 DRG (144 mm3), and close to double the volume of the C5 DRG despite a smaller receptive field.30 Similarly, the lumbar DRG size increases from the L1 to L5 levels, from approximately 4 mm × 4.5 mm at L1 to 5.5 mm × 10 mm at L5. Conversely, the size of sacral DRG decreases from S1 to S4 with the S1 DRG measuring roughly 6.5×13 mm compared to 5×3 mm at S4.32

Foramina

The DRG and nerve root exit the spinal canal through the intervertebral foramina, which also contain arteries, veins, ligaments, and epidural fat. The anterior border of the foramen is formed by the posterior aspect of the adjacent vertebral bodies, intervertebral disc, and the lateral expansion of the posterior longitudinal ligament. The dorsal border is the superior and inferior articular processes and facet joint. The roof and floor of the neural foramen are formed by the pedicles of the respective levels.

The size of the lumbar foramen ranges in transverse and sagittal measurement from 8.5 mm wide and 18 mm tall at L1 to 10×22.5 mm at L5.33

The DRG usually lies in the superior portion of the foramen close to the rostral pedicle.34 In the thoracic and lumbar levels, the DRG is positioned in close approximation to the vertebral pedicle, with the L5 DRG found intraforaminal in roughly 75% of cases, with only 6% found extraforaminal.35 In contrast, the sacral DRG was more likely to be found within the spinal canal than the neuroforamen. The S1 DRG is intraforaminal in up to 60% of cases, at S2 DRG the majority are in the spinal canal (50–85%), and at the S3 and S4 all the DRG are within the canal.35,36

The neural tissue and vessels are suspended in the foramen by transforaminal and intraforaminal ligaments. The 5 transforaminal ligaments are thick, rigid, and often unyielding, and their function is to support the intraforaminal structures. Measuring 2–5 mm and crisscrossing the foramen, these ligaments occupy a significant portion of the foraminal area. The finer intraforaminal ligaments suspend the neural and vascular tissues and prevent them from contacting the transforaminal ligaments and periosteum, enabling electrophysiological transmissions.37 These ligaments in addition to bony overgrowth may be potential obstacles to DRG-S lead placement. See Figure 4.

Figure 4 The various lumbar intervertebral and intraforaminal ligaments, nerve root, and vessels. The thicker transforaminal ligaments may impede foraminal access. Reproduced with permission from Akdemir G. Thoracic and lumbar intraforaminal ligaments. J Neurosurg Spine. 2010;13(3):351–355.37

Vasculature of the DRG

Blood supply to the DRG is provided by the spinal branch of the dorsal trunk of the segmental arteries. The radiculomedullary branches provide blood supply to the spinal nerve, running alongside the ventral aspect of the DRG and nerve roots.38 After piercing the DRG, the vessel forms a subcapsular capillary plexus, then penetrates deeper, branching into intraganglionic vessels, forming a dense capillary network, that directly interacts with sensory neurons.39 Capillaries in the DRG are fenestrated and lack a blood–brain barrier, allowing blood-borne molecules to directly enter the DRG and interact with neuronal and non-neuronal cells.40 The regulation of blood flow to the DRG is achieved by muscular sphincters which adjust flow to functional and metabolic demands.41

Peripheral nerves generally have a tight blood–nerve interface, similar to the blood–brain barrier, especially in comparison to the intensely vascularized and highly permeable vascular interface of the DRG.39 This permits a three-fold increase of blood flow within the DRG as compared to a peripheral nerve.42

This porous barrier allows for systemic states such as diabetes mellitus and permits blood-borne toxic substances such as chemotherapeutic agents access to sensitive neuronal tissue. The alpha-motoneurons which are housed within the tight blood–brain barrier of the CNS, and not the DRG, are less susceptible to such injury. This microstructural feature, specific for the DRG, likely explains why anti-neoplastic and anti-HIV agents preferentially induce a peripheral sensory polyneuropathy.39

Multiple changes in the DRG are seen in diabetic peripheral neuropathy (DPN): changes in microvascular blood flow,43 histologic changes such as thickening of the perineural cell basement membrane,44 changes in metabolic and immunologic processes, and a severity-dependent decrease in DRG volume.45–50 These pathophysiological changes seen in DPN support the notion that DPN originates at the DRG rather than the peripheral nerve itself.51,52

These observations suggest the DRG plays a central and primary role in the pathogenesis of sensory polyneuropathies, further supporting its role as a therapeutic target.

Relation to the Sympathetic Chain

The sympathetic chain serves as a conduit for small fiber nerves to travel broad distances to attach to blood vessels or to enter remote mixed nerve roots to get to their end organ targets. DRGs at the levels of T1–L2 contain sympathetic afferent fibers. The DRG offers distinct characteristics that could make it an optimal target for neuromodulation.

Sensory information is relayed through somatic afferent C- and Aδ-fibers, passing through the DRG to the dorsal horn. Sympathetic fibers are relayed to the hypothalamus, which controls the autonomic response. Efferent signals travel down the dorsal longitudinal fasciculus to reach sympathetic preganglionic neurons in the intermediolateral nucleus of the spinal cord. Preganglionic fibers exit through the ventral root and enter the sympathetic chain through the white ramus communicans (WRC) to synapse onto sympathetic postganglionic neurons located in the paravertebral sympathetic ganglia. The postganglionic fibers, which are unmyelinated C-fibers, exit the sympathetic chain through the grey ramus communicans (GRC) to join the mixed spinal nerve at that level, and go on to their targets.53

Sympathetic tone is predicated on the frequency of AP firing, with lower frequencies seen in basal tone, and higher frequencies in excitatory states, such as vasoconstriction.54 In turn, antidromic propagation of APs at very low frequencies could provide a normalizing effect on the vasculature and decrease neurogenic inflammation.14,28,29,53,55,56

Sympathetic preganglionic WRC neurons can synapse with anywhere from 1:4 to 1:32 postganglionic GRC neurons,57 and as such the sympathetic nervous system has been described as an amplifier of neural transmission. This feature allows for potential diffuse messaging and neuromodulation from a single DRG.58

Brief History of DRG-S Technology

The DRG was identified as a target for dermatomal pain control since the first reported ganglionectomy in the 1970s.59 Until 2010, given the risk of deafferentation pain, the most promising DRG-involved pain control technique pulsed radiofrequency, which ultimately only demonstrated modest efficacy in the cervical spine.60

Prior to 2011, applying a dosed electrical field over the DRG was reported for pain in only two case reports and involved use of a t-SCS system. In 1995, Wright and Collition placed SCS leads at the L2 DRG based on Nakamura et al’s concept of sympathetic convergence and reported satisfactory results over 8 months.61,62 In 2011, the second case report described an SCS lead over the C2 DRG for post-herpetic neuralgia (PHN).63 Alo et al described lumbar and sacral nerve root stimulation with leads placed transforaminally or sacrally via the sacrococcygeal hiatus, where they likely captured the DRG that lie within the sacral canal.32,64

The DRG emerged as a potential target for neuroelectrical modulation secondary to work beginning in the 1980s.65,66 It was in September 2004 that the first patent was filed for a novel device to stimulate the DRG specifically.67 See Figure 5. At this time, t-SCS on the dorsal columns was the standard and sole option, relying on paresthesia that was prone to position-related changes in stimulation. It was technologically primitive by today’s standards, combined with less robust lead configurations leading to less coverage for low back and even less so for the distal extremities. DRG-S was conceptualized to meet some of these shortcomings.

Figure 5 An original patent image demonstrating an implantation concept with an electrode being implanted into a dorsal root ganglia. Reproduced from Kim DH, Imran MA. Method and system for stimulating a dorsal root ganglion. 2009:1–72. Available from: https://patents.google.com/patent/US7580753B2/en. Creative Commons.67

Applying a lead on the DRG required several modifications from the existing t-SCS systems. These included a smaller lead diameter, increased flexibility, a reduced contact size, and a novel technique for device implantation. In addition, device programming needed to allow finer adjustments as dosing needs were substantially lower than previously utilized.

The proof-of-concept, first-in-human work was performed on three patients in 2008. This experience helped mold the next phase of lead and delivery engineering, IPG modification, and lead placement methodology. Once a reproducible device was available, a pilot study was conducted in the spring of 2009 that demonstrated a statistically significant pain improvement over 72 hours.68 The next phase of research was moved to Europe and Australia in 2010. Two small multicenter studies demonstrated safety and efficacy of targeting the DRG for foot pain and FBSS,69,70 and this led to regulatory approvals in Europe and Australia in 2011.

In 2012, Deer et al published the 2009 study data, where 8 of 10 trialed subjects achieved pain relief while reporting no adverse events, using average settings of 68 Hz, pulse width of 200 µs, and 0.800 mA.71 Shortly after that publication, in a multicenter European and Australian prospective trial of DRG-S, Liem et al again demonstrated safety and efficacy of DRG-S in 32 patients followed for 6 months.72 Eighty-nine percent of patients with neuropathic foot pain achieved significant pain relief, an area that is often challenging to cover with t-SCS.

Given the results of early studies, a prospective, multicenter, randomized comparative effectiveness trial of 152 subjects diagnosed with CRPS I and II of the lower extremities, comparing DRG-S to t-SCS, named the ACCURATE trial, was initiated.1 As with the Liem trial, this trial included pain conditions or painful regions such as the foot, knee, and inguinal areas that have historically been more challenging to treat, even with t-SCS, and used average settings of approximately 20 Hz, 300 µs, and 0.900 mA. Further programming advancements followed, as by 2019 it was identified that DRG-S was effective without the need for paresthesia recruitment, and by the next year it was revealed that DRG-S was effective at frequencies as low as 4 Hz and when dosed intermittently.2–4

DRG-S Mechanism of Action

DRG-S places an electrical field over the nuclei of primary afferent neurons, allowing for modulation of prior to signal propagation within the spinal cord and potentially inhibiting at synapses with second-order neurons in the DH.

Pre-clinical and clinical work is beginning to unravel mechanistic and therapeutic mechanisms underlying DRG-S, and differences compared to t-SCS are becoming evident. On the cellular level, the inhibitory neurotransmitter gamma-aminobutyric acid (GABA) is elevated in t-SCS but not with DRG-S, neither in the dorsal horn or in the DRG itself.73,74 Furthermore, Aδ-fiber stimulation at 1 Hz causes long-term depression in the DH,75 a likely mechanism underpinning the delayed washout of DRG-S at 1 Hz compared to 20 Hz and 1000 Hz,76 and clinically reflected in the ability to cycle DRG-S.

Impeding the transduction of afferent pain signaling and of ectopic AP transmission is the most commonly accepted mechanism of DRG-S, and this has been supported in basic science research.5,7,22 However, the effects of DRG-S on broad, multi-dermatomal conditions with a single lead placement would depend on the orthodromic propagation of stimulation, an accepted mechanism of t-SCS, but rarely discussed regarding DRG-S.8,10,77,78

It has been postulated that DRG-S activates Aδ-, Aβ-, and C-fiber low threshold mechanoreceptor (LTMRs) fibers, which utilize the endogenous opioid system (EOS) to modulate the touch and pain processes at frequencies clinically utilized with DRG-S.75,79,80 This mechanism is thought to underlie the ability of DRG-S to modulate nociceptive or mixed-pain syndromes.

Additionally, DRG-S has demonstrated promising antidromic effects on the sympathetic nervous system in the treatment of peripheral vascular disease, blood pressure reduction, as well as decreasing neuroinflammation.11,13,14,81

Narrative Literature Review of Peer-Reviewed Clinical Findings in Dorsal Root Ganglion Stimulation Methods Development Process

As part of its mission to improve patient care and access to advanced neuromodulation techniques, the American Society of Pain and Neuroscience formed the Best Practices Work Group, consisting of ASPN members who were chosen for their clinical expertise, familiarity with the current peer-reviewed literature, research capabilities, diversity of practice, and previous publications. All authors disclosed any conflicts of interest and recused themselves from any section impacted. The primary author served as a conflict-free referee of any bias in the paper.

DRG-S has been approved in the United States since 2016 for the diagnosis of CRPS I and causalgia. The ACCURATE study remains the lone multicenter, randomized controlled DRG-S trial. Many of the remaining reports averaged results of pooled diagnoses and non-standardized lead configurations, making interpretation of diagnosis-specific results and recommendations for DRG-S best practices challenging. Nonetheless, the available data per condition from these papers demonstrate fragmented evidence of very good outcomes for a wide variety of diagnoses. Successful DRG-S relies on lead placement, and we see mixed results from inaccurate lead placement.82 Currently, only a handful of prospective papers are diagnosis and lead placement specific, and these tended to have small numbers.9,77 Currently, a post-approval real-world study is ongoing in the United States, the TARGET study, with FDA monitoring to identify best uses of the therapy and continuing improvements in safety.83

In the last year there have been three systematic reviews on DRG-S demonstrating CRPS and causalgia as the only diagnoses to potentially qualify for grade A evidence with high certainty.84–86 To reduce redundancy, we will present the literature in a narrative format, based on the current body of evidence, per diagnosis/body part.

Literature Search, Evidence Ranking, and Consensus Development

A comprehensive search of several databases from 2010 to August 2022, limited to English language and excluding animal studies, was made. The databases included Ovid MEDLINE® and Epub ahead of print, in-process, and other non-indexed citations, Ovid Embase, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus. The search strategy was designed and conducted by an experienced librarian with input from three of the study investigators (JMH, TJL, KBC). Controlled vocabulary supplemented with keywords was used to search for studies describing DRG-S. Additional keywords included “ganglia”, “spinal ganglia”, “neuromodulation”, “electrostimulation”, “neurostimulation”, “dorsal root”, and “drg”. For the purposes of grading, we defined a case report as any study involving ≤2 patients and a case series as any study with ≥3 patients.

Identified peer-reviewed literature was critiqued using the United States Preventive Services Task Force (USPSTF) criteria for quality of evidence,87 with modifications for neuromodulation studies (see Table 1). After USPSTF letter grading was assigned, the working subgroup then assigned the “level of certainty regarding benefit” as described in Table 2.

Table 1 Quality of Evidence Ranking Using United States Preventative Services Task Force Criteria Modified for DRG-S

Table 2 Levels of Certainty Regarding Net Benefit

Work groups were convened to evaluate the literature and examine the evidence for the topics developed by the lead authors in outline form. Recusal was performed for any section where the author had a direct conflict of interest. After the literature search was completed and separated by body part treated, outcomes were collected and cited. Diagnosis-specific data from publications with multiple diagnoses were extracted when available. If data per diagnosis were not provided, the data were excluded.

As a consensus guideline, this document provides recommendations regarding practices for DRG-S, and should not be construed as a standard of care. Best practices were based on several factors, including peer-reviewed evidence, and, regardless of the strength of evidence, requires interpretation for clinical application.

Results Complex Regional Pain Syndrome (CRPS) Type I

The treatment of CRPS and causalgia with DRG-S is accepted as a potential first-line neuromodulatory therapy, and the majority of DRG-S literature is focused on the treatment of CRPS type I and II. Aside from the ACCURATE study, there are 9 case reports and a presence in virtually all the DRG-S publications.1,88–94 Published in 2017, the ACCURATE multicenter prospective, randomized, controlled trial evaluated the safety and efficacy of DRG-S compared to traditional SCS in subjects with CRPS type I or causalgia at three and 12 months. The percentage of patients who achieved greater than 50% reduction in VAS at three months with DRG-S was 81.2% versus 55.7% in the SCS arm and 74.2% in the DRG-S group versus 53.0% with SCS at 12 months. DRG-S demonstrated superiority compared to SCS for CRPS I and II at 3 (p < 0.0004) and 12 months (p < 0.0004). DRG-S subjects also experienced significantly less paresthesia in non-painful regions.1

A number of case reports have reported significant pain relief in patients with CRPS with utilization of DRG-S88,90–92 One case report described improvements in the physical manifestations of CRPS in a patient with leg pain after a spinal cord injury.93 Patient preference for DRG-S is suggested for treating CRPS to have a preference for DRG-S over SCS when trialed with both treatment options89 or after inadequate relief with SCS.94,95

Utilizing a multi-waveform SCS system, Al Kaisy et al retrospectively compared t-SCS, 1K SCS, and burst-SCS on the DRG.96 CRPS was the indication in 19 out of 39 (49%) patients. Thirty-two of 39 patients proceeded to implant after trial (82%) with 28/32 responders (87.5%) at 6 weeks and 21/32 (66%) at a mean of 18 months' follow-up. A burst protocol was preferred in the majority, 78% (25/32), of patients. As the leads and SCS system used in this study were not designed to treat the DRG, the inability to fine-tune frequency and amplitude may have played a role in this preference over t-SCS.

Causalgia and Post-Surgical Pain Etiologies

The benefits of DRG-S for post-surgical pain syndromes (PSPS) and causalgia include better dermatomal coverage, proximity to the sympathetic chain, and utilizing points of convergence in the DH for broader coverage.10,14 The evidence regarding DRG-S for PSPS/causalgia (including post-hernia, post-arthroplasty, failed back surgery syndrome, post-amputation, post-abdominal, and pelvic surgery) is reported in 3 prospective and 3 retrospective case series, a subset of the ACCURATE RCT, and 11 case reports.97–106

Given the limited, low-quality evidence SCS has demonstrated for the following indications, compared to the positive results in the DRG-S studies, there is an opportunity for DRG-S to be considered as a first-line treatment option for intractable, non-responsive PSPS and causalgia.

Groin, Post-Herniorrhaphy, and Abdominal Wall Pain

Post-herniorrhaphy pain stems from distal dermatomal receptive fields, an optimal condition for DRG-S to provide improved paresthesia-free coverage.4,107 Given the limited evidence for SCS and PNS for inguinal and post-herniorrhaphy pain108–113 and positive results from the prospective and retrospective case series, and case reports discussed below, DRG-S is positioned as a reasonable consideration for first-line application.101,105,114

Morgalla et al reported on a prospective case series on 34 consecutive patients who previously had undergone inguinal hernia repair with PSPS from ilioinguinal or iliohypogastric nerve injury.101 Thirty of 34 trialed patients had >50% pain improvement and underwent implantation. Most patients (25/30) had implantation at L1 and L2 on the effected side. At three months' follow-up, VAS decreased a mean 75.5% (30/30 patients) which was maintained at 63.5% at three years (11/30 reached this time period). At three years, 72.7% of patients (8/11) reported more than a 50% decrease in pain. Patients had improvements in Beck Depression Inventory (BDI), Pain Disability Index (PDI), Pain Catastrophizing Scale (PCS), and Brief Pain Inventory (BPI), along with 73.3% reducing their opioid requirements after implantation.

In a retrospective review, Schu et al reported on 29 patients with chronic groin pain.105 Within this cohort, ten of 13 patients trialed with DRG-S for post-herniorrhaphy chronic pain went on to implantation, with a mean follow-up time of 17.4 weeks. Eighty percent of patients reported >50% VAS reduction with a mean VAS reduction of 76.8%. Five patients had follow-up longer than six months, with the mean follow-up time 30.6 weeks and reduction in VAS of 74.3%.

In 2014, Zuidema et al reported on a 3-patient case series using paresthesia mapping and leads placed at the T12, L2, and T11 and L2 levels. At between 2 and 3 months patients had between 90% and 100% pain relief.114 Finally, it is worth noting that within the ACCURATE trial 14 patients were diagnosed with causalgia in an ilioinguinal nerve distribution.1

Mol et al retrospectively followed 5 patients suffering from anterior cutaneous nerve entrapment syndrome (ACNES) treated with DRG-S for 12 months.115 These patients with non-surgical caused nerve entrapment pain in anterior abdominal wall had 3 patients sustaining >50% pain relief, one 37% relief, and one patient without improvement at 12 months' follow-up. Another case series included 9 patients treated for ACNES with DRG-S followed for 3–30 months.116 At 3 months 8 of 9 experienced >50% pain relief, with four of nine reaching 30 months with maintained improvements. In another case report Akuamoah et demonstrated impressive improvements with DRG-S placed for hernia pain after an extreme lateral interbody fusion (XLIF) surgery, and then later adding leads for the same failed XLIF.117 VAS improved from hernia/flank pain from a VAS of 10 to 1.5 at 18 months and the T12 and S1 leads improved low back and leg pain from a 9/10 to a 3.5/10, ODI from 78 to 29, and EQ-5D from −0.11 to 0.59 at 6 months post lead addition.

Post-Surgical Joint Pain

Post-surgical joint pain or causalgia of the post-surgical joint is a PSPS that has garnered little attention with SCS compared with DRG-S.118–120 There have been 2 prospective studies, 1 retrospective, 1 case series, 3 case reports, and 1 subset analysis of the ACCURATE study representing 66 patients with post joint replacement pain.99,102,104,121–124 The evidence DRG-S has demonstrated in post-surgical joint pain/causalgia compared to SCS places DRG-S as the authors' preferred neuromodulatory option in refractory post-surgical joint pain/causalgia.

Hunter et al included 12 post-total knee arthroplasty (TKA) patients in a retrospective review from a larger cohort of 217 patients.104 During the trial period, the post-TKA patients experienced an approximately 70% decrease in their pain. Eight patients would go on to receive an implantation of a permanent system. The authors recommended ipsilateral L3 and L4 DRG stimulation for optimal results. Long-term results were not described.

Morgalla et al trialed 62 patients, with 51 proceeding to implant to treat causalgia of multiple regions, including 30 trialed with causalgia of the knee with 27 proceeding to implant.121 Other regions included hand, foot, leg, and PLP. Twenty-five patients reached 3 years' follow-up. Results were not separated per body part; however, overall improvements included a 50% reduction in pain (VAS of 8 to 4), BPI decreased 53% (6.8 to 3.2), BPI-PI 68% (5.7 to 1.8), and Beck Depression Index (BDI) decreased 24% (~10.5 to 8).

Kretzchmar et al retrospectively reported on 23 patients following DRG-S for post-surgical pain, including 11 post-knee surgeries and 2 post-total hip replacements.99 The authors did not report the data for these patients separately. At 36 months, the cohort demonstrated a VAS improvement of 73% (55 to 15), SF-12 MCS improvement of 25% (34 to 44), PCS improvement of 18% (42 to 50), and at 12 months a Quality-of-Life Impairment by Pain Inventory (QLIP) improvement of 102% (18 to 37). Leads for knee pain were placed at the L3 and L4 levels, and, for the hip, one patient was implanted at the L1, L2, and L3 levels and one patient at the L5 and S1 levels.

Martin et al reported on 12 patients implanted with a DRG-S, five of which had post-surgical knee pain.102 The average follow-up for all patients was 34 months (range 15 to 78 months). Of the five post-surgical patients, 3 have a single L3 lead, 1 a single L4 lead, and 1 patient has L3 and L4 leads. The average pain relief for the five post-surgical patients was 72.7% at an unknown mean time period.

A recent case series and a case report by Chapman et al, totaling 5 patients, reported on joint pain deemed not primarily neuropathic in nature treated with DRG-S.123 Joints included the hip, knee, and ankle. There were impressive improvements in VAS, quality of life as measured by the EQ-5, as well as joint-specific testing. Joint-specific questionnaires demonstrated impressive improvements.

In a paper detailing 3 novel implant techniques utilized on two patients suffering from post-TKR knee pain, Schultheis et al demonstrated a reduction in NRS of 50% over 12 months with leads at L3 and L4.124 There were improvements across multiple domains including multiple functional, psychological, and sleep domains, including a 24% improvement in PDI.

Failed Back Surgery Syndrome and Non-Surgical Low Back Pain

DRG-S for failed back surgery syndrome (FBSS) and non-surgical low back pain (NSLBP) has shown limited but impressive results in the treatment of pain, function, affect, and quality of life. SCS remains the gold standard neuromodulatory therapy for FBSS and NSLBP, although evidence is accumulating to consider DRG-S as a second-line therapy, or after failed SCS cases, as evidenced in the following reports:

Weiner et al carried out a prospective study using a wireless SCS system on the DRG for FBSS in 2016.106 They reported on 11 patients who underwent unilateral DRG-S ranging from L1 to L5 for 45 days. Stimulation settings of 500 µs pulse width with 100 Hz frequency were used, which are levels incongruent with today’s practice. Yet the overall pain reduction was 59.9%, with L2 experiencing the most significant relief at 73%.

One year later, Huygen et al published a prospective study evaluating L2 or L3 DRG-S for LBP secondary to FBSS in 12 patients.82 At least one lead was placed at the L2 or L3 and additional leads to provide paresthesia coverage in additional painful areas. Included were patients with non-radiating LBP and patients with low back and lumbar radicular pain. Average pain relief was 45.5% at 12 months, and more than half of the patients reported at least 50% pain improvement. Mood and quality of life were also improved with DRG-S at 12 months.

Based on the results of the previous study, Kallewaard et al followed up with a prospective study of 13 patients treated with DRG-S for axial LBP secondary to FBSS after lumbar discectomy.77 Eleven patients with non-radiating pain had a successful trial and were implanted. At one year, average pain was reduced by 72% (NRS 7.2 to 2.3), and quality of life measured with EQ-5D improved 38% (0.61 to 0.84), disability measured with the ODI improved by 58% (42.1 to 21.5), and a 94% improvement in mood was seen as measured with POMS (16.4 to 1).

Non-Surgical Low Back Pain

DRG-S for the treatment of NSLBP was studied by Kallewaard et al in a prospective, single-arm case series for discogenic LBP treated at the bilateral L2 level.9 L2 DRG-S for discogenic pain only was based on the notion of sympathetic convergence at the L2 level in conjunction with the results of Huygen et al.82 Fifteen patients who had a negative response to medial branch blocks and a concordant discogram were successfully trialed and proceeded to permanent implant. Patients reported a mean NRS reduction of 63.4% (n=15) at six months and 68.3% (n=14) at 12 months (p < 0.001). Impressively, the cohort demonstrated a 141% increase in mood (EQ-5D 0.34 to 0.82), a 58% improvement in ODI (46.1% to 19.2%), and a 101% improvement in POMS (19.2 to −0.11) at the 12-month endpoint.

Reports Containing Mixed FBSS and NSLBP

Outcomes of FBSS and NSLBP treated with DRG-S have been studied in several additional reports. The combined outcomes of the following papers demonstrate further lower-quality evidence for the use of DRG-S to treat both FBSS and NSLBP:

Chapman et al described outcomes of DRG-S at the T12 level for axial low back pain regardless of surgical intervention.8 As prior studies utilized L2 placement based on sympathetic convergence, the authors provided a review of literature on the potential mechanisms underlying efficacy of DRG-S at T12.10 The cohort included 7 patients with a history of lumbar fusion, 5 with a laminectomy/discectomy, and 5 patients with NSLBP, with 4 having failed prior stimulation. Ten patients also had S1 leads placed for lumbar radicular pain or sacroiliac joint pain. Patients experienced a 76% reduction in VAS (9.3 to 2.2), and a 78% improvement in ODI, a 180% improvement in EQ-5 (0.30 to 0.84), a 93% improvement in psychological status as measured by the MCS (SF-36 30.8 to 59.5), and a 108% improvement in PCS (SF-36 23.8 to 49.4). Moreover, 70% of patients reduced opioid consumption, with two stopping completely.

Another study measuring the effects of frequency on outcomes included 20 patients titrated to 4 Hz.2 Their cohort included 20 patients, of which 8 had a history of prior lumbar fusion, 2 with laminectomy/discectomy, and 10 treated for NSLBP. All patients had bilateral T12 leads to cover lower back pain and a combination of unilateral or bilateral S1 or other lumbar leads to cover lower extremity pain. Frequency was titrated from a mean of 16 Hz to 4 Hz over a mean period of 80 days with pre-DRG-S, pre-titration, and post-titration outcomes measured. Efficacy was maintained at 4 Hz, as their improvements from baseline of 77% in VAS (8.8 to 2.0), 72% (67 to 19) in disability, and a 145% improvement in mood (EQ-5D 0.33 to 0.81) were maintained at 366 days post-implant. The cohort also maintained an opioid reduction of 87 to 43 MME and a reduction in the number of interventional procedures performed over the equivalent pre-DRG-S period from 0.5 to 0.04 per month. The article was published with a sister review of literature article on the mechanisms underlying stimulation at 4 Hz.79

In additional work from this group, measuring changes in quantitative sensory testing (QST) measurements, 11 patients were followed: 6 with a history of FBSS and 5 with NSLBP.125 Leads were placed at T12 and S1 for all patients, with 2 patients requiring leads at L4. Quantitative sensory testing was performed in patients before trial lead placement and either before trial lead removal or at the one-month follow-up after permanent implant. Patients experienced a 77% reduction in VAS, a 62% improvement in ODI, an 80% improvement in EQ-5, a 24% improvement psychological status as measured by the MCS (SF-36 42 to 52), and a 104% improvement in PCS (SF-36 23 to 47). The paper also demonstrated a normalization of pressure pain threshold, mechanical detection threshold, and conditioned pain modulation, signs of DRG-S improving both the nociceptive and neuropathic components of LBP.

Huygen et al followed 56 patients treated with DRG-S for 1 year with mixed diagnoses. Twenty-five of these patients were treated for FBSS.126 Lead locations were not mentioned. Baseline NRS improved from 8 to 3.9 in FBSS patients at 1 year. The overall cohort also contained 24 patients with CRPS I and II, post-amputation pain, radicular pain, and others. At 12 months for the overall cohort, improvements included VAS reduction of 49% (8.0 to 4.1), BPI Pain Intensity improved 26%, BPI Pain severity improved 26%, mood as measured by POMS improved 52% (27.8 to 13.3), and EQ-5 improved 72% (0.36 to 0.62).

n summary, DRG-S has demonstrated an improvement in pain, function, affect, and quality of life for FBSS and NSLBP. Although yet to be reproduced in a larger-scale study, DRG-S shows concordant improvements in functional and psychological measures as compared to pain to a degree that is worthy of further studies.

Peripheral Neuropathy

The ability to directly modulate the Aδ and C small fibers that innervate the skin and subcutaneous tissue, at level Aβ collaterals, and taking advantage of convergence in the DH, would appear to make DRG-S an optimal therapy for pain associated with peripheral neuropathic (PN). There are only two small prospective studies and one retrospective study, and seven case reports of application of DRG-S for varying causes of PN.58,127–134 Reports included diabetic PN, painful small-fiber PN, idiopathic PN, polysensory PN, hereditary sensory and autonomic PN, PN associated with Lyme’s disease, and chemotherapeutic agent-induced PN. Follow-up in most cases was as short as 6 weeks to 6–12 months. Leads were typically placed at L4-S1, with less than half of patients implanted with S1 leads. Our authors agree that S1 lead placement is integral to success with distal lower extremity neuropathic pain, and there may be potential for improved results with an addition of an S1 lead.

Eldabe et al retrospectively reported on treating DPN, with 10 patients trialed with DRG-S, 7 undergoing device implant, and 2 patients directly converted without a trial undergoing device explant.127 In the remaining 5 patients, 4 reached 12-month follow-up, and there was a 64% reduction in VAS. Aside from one failure secondary to inability to place leads, placement was from primarily L5 to L2. The two additional failures had only single level leads placed at L5, and the explanted patient also only had a single L5 lead placed. Based on convergence at the S1 level, the authors concur that a single level S1 lead is often adequate for the treatment of peripheral diabetic neuropathy.

Koetsier et al in a prospective study treating polyneuropathy reported 8 of 9 patients having a successful trial, with 7 patients implanted.131 Leads were placed at the L5 in 95% of cases and S1 in 44% of cases. At 6 months' follow-up they found a 57% improvement in VAS, a 50% improvement in anxiety and depression as measured by HADS, a 60% improvement in BPI pain intensity, and a 20% improvement in quality of life as measured by the EQ-5D.

Falowski et al retrospectively evaluated 8 patients with DRG-S placed at L4-S1 for diabetic neuropathy and polyneuropathy.129 This small cohort demonstrated an 80% reduction in pain at 6 weeks, with seven out of eight patients decreasing their opioids or discontinuing them. Ho et al described a series on painful hereditary and idiopathic axonal polyneuropathy.132 Three of four patients trialed went to implantation with bilateral L5 and S1 leads placed. There was an average pain reduction of 74% at 6 months.

Post-Amputation

Eldabe et al published a retrospective study on eight patients utilizing DRG stimulation for PLP. Leads were placed at relevant DRG based on patient-reported pain distributions.103 The average follow-up was nine months, and the average VAS decreased from 83.5 mm to 38.9 mm (53.4%).

Hunter et al published a retrospective case series on 4 patients using selective radiofrequency stimulation to for predicting targets for neuromodulation in patients with post-amputation pain.135 Lead locations were in the range L3-5, and patients experienced 60%, 85%, 90%, and 90% relief after the trial period. No implant results were given.

Post-Herpetic Neuralgia

PHN is a dermatomal pain syndrome that has yielded mixed results when treated with DRG-S. Varicella zoster infection-induced apoptosis of afferent cell bodies within the DRG reduces the somata targeted by DRG-S, and likely decreasing resultant potential efficacy.

In literature, there are 8 reports with patients treated with DRG-S for PHN. Seven of 8 reports detailed trial and implant rates totaling 20 trialed patients, with 10 patients proceeding to device implant.71,121,136–141 Outcomes related to treatment were generally not detailed, aside from a patient with a 37.5% reduction in NRS at 12 months and one explanted due to lack of pain relief,136 and 3 patients in a case series with >50% pain relief at 12–18 months' follow-up.140 DRG-S was effective in a case of subacute pain from Herpes Zoster outbreak at 2 months' follow-up.142 Mansano et al document a case of DRG-S of the Gasserian ganglion for trigeminal neuralgia and a novel implant technique utilizing intraoral puncture with maxillary fixation.141 Previous attempts at this type of target were complicated by high migration rates, fractures from biting, and irritating tonic suprathreshold stimulation.

Our authors conclude that evidence of DRG-S efficacy in PHN is relatively poor and is dependent on the ability to capture unaffected somata. Additionally, the placement of leads at levels adjacent to the affected DRG may inhibit pain through collateral Aβ fibers, a theory that needs elucidation.

Pelvic and Visceral Pain

The evidence for DRG-S for chronic abdominal or pelvic pain is reported in 1 case series and 4 case reports.143–147 The first-described use of DRG-S for chronic pelvic pain was published in 2016. In their case report, the authors described a case involving the use L1 and L2 DRG-S for chronic pelvic girdle pain. At 6 months, the patient reported a decrease in NRS from 7/10 at baseline to 4/10.143

Hunter et al followed with a case report of seven patients who underwent bilateral L1 and bilateral S2 DRG-S for chronic pelvic pain of varying etiologies. At the time of publication, three patients had follow-up beyond six months post-implantation and were continuing to report pain reduction greater than 50%.144 In another case report, Zuidema et al reported on the use of S3 DRG-S for refractory perineal pain after right-sided Bartholin’s cyst surgical resection.98 At two weeks' follow-up, VAS improved almost 89%.

Two additional case reports with different pelvic pain etiologies have also been published. Giordano reported four-month results of bilateral L1 and S2 DRG-S for coccyx and pelvic pain, and the patient was reporting 70–80% relief.145 Similarly, Hassanain reported the 12-month results of a patient with testicular pain who had 70–80% pain relief.146

Kloosterman et al described the use of bilateral T11 DRG-S for pain following Roux-en-Y gastric bypass surgery.97 At six-month follow-up, the patient continued to report approximately 90% pain relief, and improvements in ODI (78% to 10%) and SF-36 (mental: 20.9 to 58 and physical: 15.5 to 58.1). Justiz et al utilized DRG-S at the bilateral T10 levels for the treatment of chronic abdominal pain due to hereditary pancreatitis, with VAS improving from 8 to 0 at 6-month follow-up.147

Mixed Etiologies

The initial study targeting the DRG was a prospective, multicenter, single-arm, pilot study including ten subjects with chronic intractable neuropathic pain of the trunk and/or limbs and followed for four weeks upon which the leads were removed.71 Patients received an average of 2.9 leads attached to an external generator. At baseline, the overall mean VAS improved 70% during the four-week study. The average decrease in back pain was 84%, and leg and foot pain improved 80% and 70%, respectively. It was noted that there were no changes in stimulation related to body position.

A follow-up 32-subject study treating a variety of painful etiologies treated with DRG-S and followed for six months was undertaken by Liem et al.72 At all time periods, more than half of subjects reported pain relief of at least 50%. At six months post-implantation, average overall pain ratings were 58% lower than at baseline (p < 0.001), and the proportion of subjects experiencing at least a 50% reduction in pain in the back, leg, and foot areas were 57%, 70%, and 89%, respectively.

Wensing et al followed 25 patients treated with DRG-S for post-surgical pain of multiple diagnosis for 2 years.100 Improvements were seen from a baseline VAS of 76 to 38 and 46, at 1 and 2 years, respectively. There were corresponding improvements in EQ-5 from a baseline 0.48 to 0.70 and 0.68 at 1 and 2 years. Improvements in physical function were measured with pain severity (6.9 at baseline, 4.4 at 1 year, 4.5 at 2 years) and interference (4.6 at baseline, 3.0 at 1 year, and 3.5 at 2 years). Although three patients underwent device explant over 2 years, 71% of patients were highly satisfied with the therapy at the endpoint.

DRG-S was utilized in a case of post-stroke lower extremity pain after a medullary infarction with 100% reduction in pain who was weaned off 32 mg hydromorphone at 6 months' follow-up.148

Piedade et al also studied the effects of frequency, where 19 patients were randomized to 20 Hz, 40 Hz, 60 Hz, 80 Hz, and sham.149 CRPS I and II were the most frequent pain etiology. Pre DRG-S mean VAS was 8.6 and reduced to 3.9 with therapy. Pain intensity returned to 3.7 at 20 Hz but increased with higher frequencies reaching 5.8 at 80 Hz. Significant differences among the groups were shown over VAS, MPQ, EQ-5D, and BDI, and it was concluded that 20 Hz was a more optimal frequency.

Sympathetically Mediated Conditions

The effects of DRG-S on the sympathetic nervous system (SNS) have been of brought to light in recent works.11,12,