Burst Spinal Cord Stimulation as Compared With L2 Dorsal Root Ganglion Stimulation in Pain Relief for Nonoperated Discogenic Low Back Pain: Analysis of Two Prospective Studies

Introduction: Chronic discogenic low back pain (CD-LBP) is caused by degenerated disks marked by neural and vascular ingrowth. Spinal cord stimulation (SCS) has been shown to be effective for pain relief in patients who are not responsive to conventional treatments. Previously, the pain-relieving effect of two variations of SCS has been evaluated in CD-LBP: Burst SCS and L2 dorsal root ganglion stimulation (DRGS). The aim of this study is to compare the effectivity in pain relief and pain experience of Burst SCS with that of conventional L2 DRGS in patients with CD-LBP. Materials and Methods: Subjects were implanted with either Burst SCS ( n = 14) or L2 DRGS with conventional stimulation ( n = 15). Patients completed the numeric pain rating score (NRS) for back pain and Oswestry disability index (ODI) and EuroQoL 5D (EQ-5D) questionnaires at baseline, and at three, six, and 12 months after implantation. Data were compared between time points and between groups. Results: Both Burst SCS and L2 DRGS signi ﬁ cantly decreased NRS, ODI, and EQ-5D scores as compared with baseline. L2 DRGS resulted in signi ﬁ cantly lower NRS scores at 12 months and signi ﬁ cantly increased EQ-5D scores at six and 12 months. Conclusions: Both L2 DRGS and Burst SCS resulted in reduction of pain and disability, and increased quality of life in patients with CD-LBP. L2 DRGS provided signi ﬁ cantly increased pain relief and improvement in quality of life when compared with Burst SCS. Clinical Trial Registration: The clinical trial registration numbers for the study are NCT03958604 and NL54405.091.15.


INTRODUCTION
Chronic discogenic low back pain (CD-LBP) is a condition caused by a damaged or degenerated intervertebral disk (IVD). This degeneration is marked by loss of disk height, ingrowth of sensory neurons, and development of an inflammatory environment inside the disk, causing chronic pain. [1][2][3][4][5] Recently, several promising publications applied neurostimulation for nonoperative LBP, which included patients with CD-LBP, but did not specifically target patients with CD-LBP. [6][7][8][9][10] These studies used conventional spinal cord stimulation (SCS) (con-SCS) 9,10 in addition to dorsal root ganglion stimulation (DRGS). [6][7][8] Although these studies have shown con-SCS to be effective, novel forms of stimulation could provide enhanced pain relief. In 2010, De Ridder et al 11 introduced a passive recharge burst pattern waveform that was free of paresthesia sensation (referred to in this article as Burst SCS). It was in 2017 that Burst SCS was found to be superior to con-SCS in patients with chronic pain of the trunk and/or limbs. 12 Using the Burst waveform in an SCS pilot study for the treatment of CD-LBP, our group showed promising improvements in pain and function, resulting in significant reduction of back and leg pain in patients with CD-LBP and decrease in the level of disability. 13 DRGS is a direct spin-off of SCS, which places an electrical field near the somata of afferent nerve fibers. 14 Receiving its Conformité Européenne mark in 2013 and US Food and Drug Administration approval in 2015, DRGS was designed for the treatment of focal pain, although there have been multiple indications of the efficacy of DRGS in multidermatomal pain syndromes, including LBP. [15][16][17] DRGS has also been used for nonoperative LBP and was studied prospectively in the treatment of CD-LBP. 6 Con-SCS is believed to use the pain gate theory mechanistically, with antidromic stimulation of dorsal columns Aβ-fibers to modulate nociceptive signaling in the dorsal horn. 18 DRGS is thought to modulate nociceptive signals locally at the dorsal root ganglion (DRG) for focal pain, and multidermatomal coverage potentially though convergence and orthodromic propagation of signals into the dorsal horn. 18,19 Innervation of the lumbar disks has been shown to travel through the sympathetic nervous system, converging at the L2 level. [20][21][22][23][24][25] Huygen et al 15 first studied L2 DRGS lead placement in LBP with good but mixed results because seven patients had >50% pain relief, whereas five had poor results. We refined the study criteria from these data and proposed L2 DRGS for CD-LBP, excluding patients who responded to medial branch blocks and sacroiliac joint injections and who subsequently had a positive discogram result. This prospective study used DRGS at L2 to treat CD-LBP, resulting in significant, maintained improvements in pain, function, and quality of life. 6 The increased availability of new waveforms for pain treatment in CD-LBP warrants an evaluation of their effectivity in helping to choose the optimal therapy. In previous research, we studied the efficacy of both L2 DRGS and Burst SCS in independent studies for the treatment of CD-LBP, using the same inclusion and exclusion criteria in similar patient populations. 6,13 The aim of this analysis is to compare the effectivity of Burst SCS with that of L2 DRGS in pain relief, function, and quality of life in patients with CD-LBP.

MATERIALS AND METHODS
This study analyzed data from two separate prospective studies with similar CD-LBP populations, using identical inclusion and exclusion criteria. 6,13 The Burst SCS study was conducted from June 2019 to December 2021 and the L2 DRGS study from November 2015 to November 2017, both in the Rijnstate Hospital, The Netherlands. The trials were conducted by the same staff. Both trial protocols were approved by the local medical ethics committee Arnhem-Nijmegen (trial reference numbers: NL67172.091.18 and NL54405.091.1), and all participants gave written informed consent.
Pre-and postprocedure protocols were identical in the studies and are available in these works. 6,13 Implant procedures of DRGS and SCS were according to international protocol. 26 For the Burst SCS study, patients (n = 14) were implanted using two eightcontact "Octrode" epidural leads and programmed with the BurstDR stimulation paradigm (Proclaim™, Abbott; Plano, TX). Briefly, the left electrode tip was placed at T8, and the right tip was placed at T9; no intraoperative testing was performed. For the L2 DRGS study, patients (n = 15) were implanted with a four-contact 50-cm MN 20550-50 lead (Spinal Modulation, Inc, Menlo Park, CA) placed over the L2 DRG and an Axium Mn 20200 internal pulse generator (Spinal Modulation, Inc). During intraoperative testing, all patients reported paresthesia coverage of the painful region.
Burst SCS stimulation settings were 40 Hz interburst frequency, 500 Hz intraburst frequency of five pulses, 1-millisecond pulse width, and 1.3 to 0.6 mA with on/off cycling. Amplitudes were adjusted to 60% of threshold levels. Average and individual burst cycling settings are presented in Supplementary Data Tables S1 and S2. L2 DRGS stimulation settings ranged from 4 to 40 Hz frequency and between 100 and 420 milliseconds pulse width, and then reduced to subthreshold levels. Individual and average DRGS stimulation parameters are displayed in Supplementary Data Tables S3 and S4. For both studies, patients completed the numeric pain rating score (NRS) as an index of pain, 27 at baseline, after implantation trial, and, with a positive trial, at three, six, and 12 months after implantation. As a measure of disability, the Oswestry Disability Index (ODI) 28 was completed, and as a measure of quality of life and health status, the EuroQoL 5D (EQ-5D)-3L or EQ-5D-5L 29,30 was used. Both were completed at baseline and at three, six, and 12 months after implantation. Only data on matching follow-up moments between the two studies are reported in this study.
Differences between time points of Burst SCS and L2 DRGS treatments were assessed using (non) parametric unpaired t-tests (GraphPad Prism). Differences between patient demographics were assessed using unpaired t-tests or chi-square test. Unless otherwise stated, error is displayed as SEM. Only data regarding patients who received permanent implants were evaluated.

DISCUSSION
This comparison of two prospective studies using neuromodulation for treatment of patients with CD-LBP indicates that L2 DRGS may be a better long-term treatment option than Burst SCS for CD-LBP. There may be inherent challenges that influence validity when comparing clinical studies for similar diagnoses, such as inclusion and exclusion criteria, age, sex, and cultural/social differences within patient populations. However, these two prospective pilot studies used a nearly identical inclusion criterion, patient populations, were performed by the same team, and measured the same outcomes. Both studies indicated a clinically significant efficacy in pain relief, function based on the ODI, and improvements in quality of life.

Pain Relief
Both Burst SCS and L2 DRGS were effective in the treatment of pain in patients with CD-LBP at all follow-up time points. 6,13 Although both groups showed early success, at 12 months, the L2 DRGS group revealed significantly improved pain relief of 22.9 on the NRS, on which Burst SCS pain relief only reached 42.5. In addition, pain relief associated with L2 DRGS remained stable after three months, whereas with Burst SCS, a trend of decreasing pain relief was shown after three months, eventually leading to statistical significance between the two treatments at 12 months ( Fig. 1) (Table 2).

Quality of Life
Both Burst SCS and L2 DRGS also showed clinically significant differences in quality of life in treating CD-LBP. 6,13 After parallel improvement up to the three-month point, DRGS resulted in significantly increased quality of life compared with Burst SCS at six and 12 months (Fig. 3) (Table 2).

Effects on Function and Disability
Both groups experienced statistically significant decreases in disability based on the ODI; there was no statistical significance between the treatments (Fig. 2) ( Table 2).

Longevity of Therapy
Our data show a gradual increase in pain and gradual decrease in quality of life in the Burst SCS cohort starting at three months in contrast to L2 DRGS. Although underlying causation is unclear, we surmise that L2 DRGS may be more robust against loss of efficacy. This is supported by clinical evidence that indicated that in a pooled analysis of 249 DRGS cases, only a few explants were owing to inadequate pain relief, contrasting with the higher rates of explantation secondary to inadequate pain relief in SCS literature. [31][32][33][34][35] Context Previous works have evaluated both SCS and DRGs in prospective clinical settings, providing evidence for the success of these interventions in CD-LBP. 8,10,36 Differences between patient populations make it difficult to directly compare these works with the data described in this study. Our report is a step toward evaluating the differences between these treatment options.
We hypothesize that the mechanisms of action that underlie DRGS at L2 are responsible for our findings. As stated earlier, innervation of the ventrolateral IVDs and vertebral bodies occurs through fibers running in the sympathetic chain that converge at the L2 DRG. [20][21][22][23] Stimulation at the L2 DRG allows direct modulation of fibers transmitting CD-LBP by increasing its natural signal filtering effects. 19,37,38 In contrast, modulation of the dorsal columns via SCS occurs through a circuitous, multisynaptic path, targeting nonnociceptive dorsal column fibers and antidromically modulating the spinal nociceptive network to effect sympathetic transmission. 18,39 This may result in less specific targeting and blocking of the nociceptive C and Aδ fibers involved in CD-LBP.

Limitations
Although this analysis presents what we believe is the first insight into efficacy of L2 DRGS compared with Burst SCS in the treatment of patients with nonoperated CD-LBP, its limitations must be noted. These are small prospective studies that cannot be compared with large randomized controlled trials, which are a more definitive means of comparing treatments. In addition, research has shown that lower Burst SCS amplitudes may result in increased pain relief. 40 However, because this has not yet been widely introduced in the clinic, this study reflects the current state of Burst SCS programming. Moreover, it is now common practice to standardize the cycling programs for Burst SCS, which was not previously the case. 41 Burst SCS has an established effect on the affective component of pain via the medial pathway, and we did not measure this. [42][43][44][45] Translational preclinical studies are needed to provide further mechanistic insight into the possible differences in mechanism of action between L2 DRGS and Burst SCS.

CONCLUSIONS
Comparison of two small-scale patient cohorts shows that L2 DRGS provided better long-term pain relief and increases in quality of life than did Burst SCS for patients with CD-LBP. Patients treated with L2 DRGS show significantly more pain relief at 12 months than do those treated with Burst SCS, and increased quality of life at six and 12 months.

COMMENTS
This is an interesting comparison between two different neuromodulation techniques to treat virgin low back pain, both provided by the same company so that at least industry bias is removed. The authors conclude that DRG of L2 is superior to BurstDR SCS. Although the data do suggest this, there are two caveats that limit the generalizability of this conclusion. The first caveat is that the study was not a priori set up to compare the two technologies, which limits the scientific value because there was no randomization of patients, the settings were different, and potentially, the clinical indications were not exactly the same. The second, more important caveat is that BurstDR SCS was provided in a nonstandardized way, and the data suggest that many patients could potentially have been overstimulated; that is, the amplitudes may have been too high. The progressive loss of efficacy in time hints at this. Unfortunately, the SCS amplitudes could not be provided by the authors, which precludes a subanalysis that correlates the efficacy of pain suppression of BurstDR with delivered amplitudes. Indeed, it has previously been shown that lower BurstDR amplitudes yield superior pain suppression. Therefore, the conclusion must be considered somewhat premature, and a true prospective randomized comparison study is required to determine whether one of the two techniques is superior. Nevertheless, even with its limitations, the study is beneficial in that it paves the way for much needed comparison studies between different neuromodulation approaches to help pain physicians decide what technique or technology may offer the patient most chances of successful pain suppression.