BACKGROUND:There is paucity of data from randomized controlled trials supporting the use of peripheral nerve stimulation, a well-established therapy for the treatment of chronic pain. This study was undertaken, in part, to provide randomized controlled trial data in support of patient access to appropriate peripheral nerve stimulation therapy. The COMFORT study is the first large, postmarket, multicenter randomized controlled trials investigating the use of a Food and Drug Administration-cleared micro-implantable pulse generator (IPG) for treating chronic pain via peripheral nerve stimulation therapy. METHODS:Consented, eligible subjects were randomized to either the active arm, which received peripheral nerve stimulation and conventional medical management, or the control arm, which received conventional medical management alone and were allowed to cross over to the active arm, after 3 months. Pain and patient-reported outcomes were captured. Therapy responders were subjects who achieved at least a 50% reduction in pain scores compared with baseline. We are reporting the 12-month results of this 36-month study. RESULTS:At 12 months, the responder rate was 87% with a 69% average reduction in pain compared with baseline (7.5±1.2 to 2.3±1.7; p<0.001). Statistical significance was achieved for all patient-reported outcomes. There was an excellent safety profile with no serious adverse device effects or reports of pocket pain. A majority of subjects used unique programming options and found this device easy to use and comfortable to wear. CONCLUSIONS:These 12-month results are consistent with previously reported 6-month outcomes from this study, showing durability of peripheral nerve stimulation treatment with the micro-IPG system; subjects realized sustained large reduction in pain and improvement in patient-reported outcomes following treatment with this micro-IPG system. TRIAL REGISTRATION NUMBER:NCT05287373.
INTRODUCTION:Sacroiliac joint (SIJ) pain comprises up to 30% of cases of mechanical low back pain (LBP), the leading cause of disability worldwide. Despite sacral lateral branch cooled radiofrequency ablation (CRFA) showing efficacy in clinical trials, there is a lack of comparative-effectiveness long-term follow-up. METHODS:In this randomized, multicenter, comparative-effectiveness study, 210 patients with injection-confirmed SIJ pain who responded to prognostic lateral branch blocks were randomly assigned to receive CRFA of the L5 dorsal ramus and S1-S3/4 lateral branches or standard medical management (SMM) consisting of pharmacotherapy, physical therapy, injections, and integrative therapies. Patients were followed up at 1, 3, 6, 9, and 12 months, with participants reporting unsatisfactory SMM outcomes being allowed to crossover (XO) and receive CRFA at 3 months. The primary outcome measure was the mean change in average LBP score on a 0-10 Numeric Rating Scale (NRS), with secondary outcomes including measures of quality of life (QoL) and function. A responder was defined as a participant who experienced a ≥30% or ≥2-point decrease in average daily NRS pain score coupled with a score ≥5 out of 7 (moderately better) on the Patient Global Impression of Change scale. RESULTS:At 12 months, the mean NRS pain score declined from a baseline of 6.4±1.4 to 3.5±2.6, with 57.4% (35/61) of participants in the randomized CRFA cohort experiencing a ≥2-point or 30% decrease in average LBP from baseline. In the crossover cohort, 35/63 (55.6%) subjects had the same experience 12 months following the XO procedure; in the XO group, the mean LBP decreased from 6.1±1.5 to 3.4±2.5. Patients also experienced clinically meaningful improvements in QoL via EuroQoL-5D-5L at 12 months (mean change of +0.22±0.27 in the originally-treated CRFA group and +0.21±0.33 in the XO group). Oswestry Disability Index (ODI) scores also improved by 12.4%±14.7 (CRFA) and 13.7%±17.1 (XO) from baseline at study-end. No serious adverse events related to the CRFA procedure were reported. CONCLUSION:CRFA in patients with SIJ pain provided clinically significant and sustained improvements for 12 months following a single CRFA treatment, regardless of previous SMM treatment. TRIAL REGISTRATION NUMBER:NCT03601949.
BACKGROUND:Spinal cord stimulation (SCS) for chronic pain has undergone many advancements over the years, including the introduction of novel stimulation waveforms, expansion to treat new pain conditions, and the more recent implementation of evoked compound action potential (ECAP) controlled, closed-loop (CL) SCS. The ECAP-a measure of neural activation-may be used as a biosignal for adjusting stimulation amplitudes to avoid under- and overstimulation associated with spinal cord movement relative to the SCS lead. CONTENT OVERVIEW:Using ECAPs in a CL system relies on successfully isolating the neural response signal from the relatively large stimulation artifact. A study with the first commercialized CL-SCS device investigated pain relief with low-frequency stimulation within a therapy window described as ranging from initial perception of sensation to discomfort (paresthesia window). In contrast, another commercial system with CL technology includes additional functionality to support consistent therapy with higher frequency stimulation set at or near the perception threshold. This more recently approved CL-SCS system leverages advanced artifact suppression and flexible CL control algorithms to afford dose control for conventional, high-frequency (up to 1200 Hz), and contemporary SCS waveforms. CONCLUSIONS:Here, we discuss the history and motivation for CL systems in SCS and the challenges faced when implementing such technology.
BACKGROUND CONTEXT:Mechanical chronic low back pain (CLBP) associated with lumbar multifidus muscle dysfunction is a major cause of long-term disability. Available nonoperative and interventional treatments often provide limited and transient benefit. Restorative neurostimulation is intended to activate the lumbar multifidus to restore neuromuscular control and address the underlying pathophysiology of mechanical CLBP. PURPOSE:The purpose of this extension of the original randomized controlled trial was to track outcomes in a crossover group for the subsequent treatment year. Participants in the original treatment group were also followed for an additional year, allowing long-term clinical follow-up. STUDY DESIGN/SETTING:RESTORE is a postmarket, multicenter, open-label randomized controlled trial performed at 25 clinical sites in the United States. PATIENT SAMPLE:A total of 203 participants were randomized 1:1 to restorative neurostimulation or optimal medical management. At the completion of the randomized phase, 80/96 (83%) eligible control participants elected to cross over to restorative neurostimulation treatment. These participants and the original treatment group were followed for an additional year. OUTCOME MEASURES:Oswestry Disability Index (ODI), back pain (Numerical Rating Scale [NRS]), and health-related quality of life (EuroQol 5-Dimension 5-Level [EQ-5D-5L]) were assessed at 18 and 24 months postrandomization. Outcomes included responder rates based on minimal clinically important change thresholds, pain remission, composite responder definitions, work ability, subject global impression of change, opioid use, and number needed to treat. METHODS:This prespecified delayed-start crossover analysis included a nonrandomized comparison of outcomes between early and delayed-start treatment groups, with within-group analyses serving as the primary basis for inference regarding crossover efficacy. Longitudinal outcomes for the treatment group were analyzed using mixed models for repeated measures. Crossover group outcomes were analyzed using completer analyses. RESULTS:The treatment group improvements were sustained through 24 months, with mean changes from baseline of -24.6±17.1 in ODI, -4.2±2.4 in NRS, and +0.189±0.163 in EQ-5D-5L (all p<.001). After crossover, participants demonstrated significant improvements after 1 year of active therapy, with mean changes from baseline of -21.1±14.8 in ODI, -3.8±2.2 in NRS, and +0.191±0.144 in EQ-5D-5L (all p<.001), comparable in magnitude to those observed in the treatment group at the same timepoint. At 12 months, responder rates in the treatment group were 70% for ODI ≥15 point improvement, 53% for ≥50% NRS reduction, 52% for NRS remission, and 72% for the composite endpoint, compared with 17%, 6%, 6%, and 12% in the control group, respectively; after crossover, corresponding responder rates at 12 months of active therapy were 64%, 61%, 60%, and 73%, consistent in magnitude with those observed in the treatment group at the equivalent timepoint of therapy. Corresponding numbers needed to treat ranged from 1.7 to 2.2 and remained similarly low when the crossover outcomes were compared with control. CONCLUSIONS:These findings reinforce the role of restorative neurostimulation as an effective therapeutic option for patients with multifidus dysfunction-related CLBP who have exhausted conservative care and suggest that earlier intervention may reduce prolonged disability and loss of quality of life.
Background:With the continued advancement of artificial intelligence (AI), large language models (LLMs) such as GPT-4 may assist clinicians in evaluating patient candidacy for spinal cord stimulation (SCS). We compared a general-purpose, non-fine-tuned LLM (GPT-4), an expert multidisciplinary team (MDT), and a clinician-input, rule-based e-Health decision-support tool. The study focused exclusively on decision agreement and did not assess clinical outcomes (eg, pain relief or device retention). Methods:This single-center, retrospective cohort was conducted at Fondazione Istituto G. Giglio (Cefalù, Italy) and included 93 consecutive adults referred to the MDT for SCS evaluation between January 2022 and March 2024. The MDT issued binary recommendations ("proceed" vs "do not proceed") as the reference standard. The e-Health tool generated "yes", "maybe", or "no" outputs from structured clinician-entered data. GPT-4 was applied zero-shot, using a single standardized prompt on anonymized vignettes within an offline environment. The primary endpoint was agreement (weighted κ) among MDT, e-Health, and GPT-4; sensitivity/specificity analyses explored three interpretations of "maybe". Results:The MDT recommended SCS for 91.4% of patients, compared with 54.8% for the e-Health tool and 46.2% for GPT-4. Agreement was moderate for MDT vs e-Health (κ = 0.51) and e-Health vs GPT-4 (κ = 0.46), and fair for MDT vs GPT-4 (κ = 0.29). GPT-4 demonstrated a more conservative profile, favoring specificity over sensitivity. Conclusion:A non-fine-tuned GPT-4 approximated but did not replicate MDT decision-making, functioning as a high-specificity, low-sensitivity filter. A layered workflow combining rule-based tools with expert oversight and targeted LLM adaptation may best optimize SCS candidate selection.
