Advertisement

Long-Term Outcomes of Restorative Neurostimulation in Patients With Refractory Chronic Low Back Pain Secondary to Multifidus Dysfunction: Two-Year Results of the ReActiv8-B Pivotal Trial

Open AccessPublished:December 18, 2021DOI:https://doi.org/10.1016/j.neurom.2021.10.011

      Abstract

      Background

      Impaired neuromuscular control and degeneration of the multifidus muscle have been linked to the development of refractory chronic low back pain (CLBP). An implantable restorative-neurostimulator system can override the underlying multifidus inhibition by eliciting episodic, isolated contractions. The ReActiv8-B randomized, active-sham-controlled trial provided effectiveness and safety evidence for this system, and all participants received therapeutic stimulation from four months onward.

      Objective

      This study aimed to evaluate the two-year effectiveness of this restorative neurostimulator in patients with disabling CLBP secondary to multifidus muscle dysfunction and no indications for spine surgery.

      Materials and Methods

      Open-label follow-up of 204 participants implanted with a restorative neurostimulation system (ReActiv8, Mainstay Medical, Dublin, Ireland) was performed. Pain intensity (visual analog scale [VAS]), disability (Oswestry disability index [ODI]), quality-of-life (EQ-5D-5L), and opioid intake were assessed at baseline, six months, one year, and two years after activation.

      Results

      At two years (n = 156), the proportion of participants with ≥50% CLBP relief was 71%, and 65% reported CLBP resolution (VAS ≤ 2.5 cm); 61% had a reduction in ODI of ≥20 points, 76% had improvements of ≥50% in VAS and/or ≥20 points in ODI, and 56% had these substantial improvements in both VAS and ODI. A total of 87% of participants had continued device use during the second year for a median of 43% of the maximum duration, and 60% (34 of 57) had voluntarily discontinued (39%) or reduced (21%) opioid intake.

      Conclusions

      At two years, 76% of participants experienced substantial, clinically meaningful improvements in pain, disability, or both. These results provide evidence of long-term effectiveness and durability of restorative neurostimulation in patients with disabling CLBP, secondary to multifidus muscle dysfunction.

      Clinical Trial Registration

      The study is registered on clinicaltrials.gov with identifier NCT02577354.

      Keywords

      Introduction

      Most cases of acute low back pain resolve spontaneously without treatment, but for chronic low back pain (CLBP), the prognosis is not favorable.
      • Itz C.J.
      • Geurts J.W.
      • van Kleef M.
      • Nelemans P.
      Clinical course of non-specific low back pain: a systematic review of prospective cohort studies set in primary care.
      Patients with CLBP often endure impaired quality of life, depression, anxiety, and sleep disturbance.
      • Breivik H.
      • Collett B.
      • Ventafridda V.
      • Cohen R.
      • Gallacher D.
      Survey of chronic pain in Europe: prevalence, impact on daily life, and treatment.
      ,
      • Alsaadi S.M.
      • McAuley J.H.
      • Hush J.M.
      • Maher C.G.
      Prevalence of sleep disturbance in patients with low back pain.
      Most CLBP patients suffer from mechanical/musculoskeletal pain that is predominantly nociceptive in nature and have no indication for spine surgery.
      • Deyo R.A.
      • Weinstein J.N.
      Low back pain.
      • Förster M.
      • Mahn F.
      • Gockel U.
      • et al.
      Axial low back pain: one painful area—many perceptions and mechanisms.
      • Saito T.
      • Steinke H.
      • Miyaki T.
      • et al.
      Analysis of the posterior ramus of the lumbar spinal nerve: the structure of the posterior ramus of the spinal nerve.
      • Bogduk N.
      On the definitions and physiology of back pain, referred pain, and radicular pain.
      The multifidus muscles are the most important stabilizers of the lumbar spine and play a crucial role in providing segmental stability in response to changes in posture and protection against sudden perturbations.
      • Ward S.R.
      • Eng C.M.
      • Gottschalk L.J.
      • Kim C.W.
      • Garfin S.R.
      • Lieber R.L.
      The architectural design of the lumbar multifidus muscle supports its role as stabilizer.
      • Kim C.W.
      • Gottschalk L.J.
      • Eng C.
      • Ward S.R.
      • Lieber R.L.
      The multifidus muscle is the strongest stabilizer of the lumbar spine.
      • Rosatelli A.L.
      • Ravichandiran K.
      • Agur A.M.
      Three-dimensional study of the musculotendinous architecture of lumbar multifidus and its functional implications.
      Mechanical CLBP is often associated with impaired neuromuscular control and degeneration of the lumbar multifidus muscles.
      • Kim C.W.
      • Gottschalk L.J.
      • Eng C.
      • Ward S.R.
      • Lieber R.L.
      The multifidus muscle is the strongest stabilizer of the lumbar spine.
      ,
      • Teichtahl A.J.
      • Urquhart D.M.
      • Wang Y.
      • et al.
      Fat infiltration of paraspinal muscles is associated with low back pain, disability, and structural abnormalities in community-based adults.
      • Freeman M.D.
      • Woodham M.A.
      • Woodham A.W.
      The role of the lumbar multifidus in chronic low back pain: a review.
      • Shahidi B.
      • Hubbard J.C.
      • Gibbons M.C.
      • et al.
      Lumbar multifidus muscle degenerates in individuals with chronic degenerative lumbar spine pathology.
      Persistent back pain–induced inhibition and disruption of proprioceptive signaling have also been linked to long-term motor cortex reorganization.
      • Meier M.L.
      • Vrana A.
      • Schweinhardt P.
      Low back pain: the potential contribution of supraspinal motor control and proprioception.
      Results of motor control exercise programs specifically targeting the multifidus muscle are mixed.
      • Chou R.
      • Deyo R.
      • Friedly J.
      • et al.
      Nonpharmacologic therapies for low back pain: a systematic review for an American College of Physicians clinical practice guideline.
      ,

      Skelly AC, Chou R, Dettori JR, et al. Noninvasive Nonpharmacological Treatment for Chronic Pain: A Systematic Review; 2018. https://doi.org/10.23970/AHRQEPCCER209.

      The isolated muscle contractions required to reverse impaired neuromuscular control are difficult to achieve voluntarily, especially in the presence of underlying inhibition and degeneration of the multifidus muscle.
      • Tsao H.
      • Druitt T.R.
      • Schollum T.M.
      • Hodges P.W.
      Motor training of the lumbar paraspinal muscles induces immediate changes in motor coordination in patients with recurrent low back pain.
      ,
      • Hodges P.W.
      • Danneels L.
      Changes in structure and function of the back muscles in low back pain: different time points, observations, and mechanisms.
      To overcome these limitations to rehabilitation, a restorative neurostimulation system (ReActiv8, Mainstay Medical, Dublin, Ireland) was developed to electrically stimulate the medial branch of the L2 dorsal ramus nerve to elicit isolated multifidus muscle activation.
      • Russo M.
      • Deckers K.
      • Eldabe S.
      • et al.
      Muscle control and non-specific chronic low back pain.
      ,
      • Deckers K.
      • De Smedt K.
      • Mitchell B.
      • et al.
      New therapy for refractory chronic mechanical low back pain—restorative neurostimulation to activate the lumbar multifidus: one year results of a prospective multicenter clinical trial.
      A recent double-blind, randomized, sham-controlled pivotal trial provided safety and effectiveness evidence for premarket approval from the United States Food and Drug Administration (FDA) in 2020.
      • Gilligan C.
      • Volschenk W.
      • Russo M.
      • et al.
      An implantable restorative-neurostimulator for refractory mechanical chronic low back pain: a randomized sham-controlled clinical trial.
      The objective of the prospective, observational analyses reported here was to evaluate the two-year effectiveness of this restorative neurostimulator in patients with disabling CLBP secondary to multifidus muscle dysfunction and no indications for spine surgery.

      Materials and Methods

      Data for this secondary analysis were obtained from the cohort of 204 patients enrolled at 26 multidisciplinary centers in the United States, Australia, and Europe in the randomized, sham-controlled, double-blind pivotal trial. All patients were receiving therapeutic stimulation from four months onward. Details regarding patient eligibility, study design, implant procedure, and medium-term results through one year have been previously published.
      • Gilligan C.
      • Volschenk W.
      • Russo M.
      • et al.
      An implantable restorative-neurostimulator for refractory mechanical chronic low back pain: a randomized sham-controlled clinical trial.

      Patients

      Study participants were adults with a diagnosis of disabling mechanical CLBP (ie, a seven-day recall of average LBP of ≥6.0 and ≤9.0 cm on the 10-cm visual analog scale [VAS] and Oswestry disability index [ODI] of ≥21 and ≤60 points on a scale from 0 to 100). Mechanical CLBP was defined as low back pain without significant radicular symptoms. Participants had low back pain on at least half of the days in the previous year, were nonresponsive to a minimum of 90 days of nonsurgical medical management including medication and physical therapy, and had a positive prone instability test result (provocative pain test using posterior-anterior pressure on individual lumbar vertebrae that improves with activation of the posterior lumbar musculature) consistent with impaired neuromuscular control of the multifidus muscle and consequent lumbar segmental instability.
      • Hicks G.E.
      • Fritz J.M.
      • Delitto A.
      • Mishock J.
      Interrater reliability of clinical examination measures for identification of lumbar segmental instability.
      Full eligibility criteria are provided in the Supplementary Data.

      Trial Design and Oversight

      The conduct of the trial complied with the FDA regulations, ISO 14155, International Conference on Harmonization, and the Declaration of Helsinki. Local institutional review board or ethics committee approval was obtained at each site, and all participants provided written informed consent. Results are reported in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines.
      • Moher D.
      • Hopewell S.
      • Schulz K.F.
      • et al.
      CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials.
      The study is registered on clinicaltrials.gov with identifier NCT02577354.

      Procedures

      All participants received the implanted restorative neurostimulation system (Fig. 1).
      Figure thumbnail gr1
      Figure 1Implantation procedure and materials. Stimulation leads were placed bilaterally near the L2 medial branch of the dorsal ramus nerve as it crosses the L3 transverse process and distally fixated to the L2/3 intertransversarii using flexible tines. The proximal sections were tunneled subcutaneously to the surgically created pocket in the gluteal or lower lumbar region where they were connected directly to the implantable pulse generator.
      During the open-label phase of the trial, all devices were programmed to deliver therapeutic stimulation at a frequency of 20 Hz, a pulse width of 214 μsec, and participant-specific pulse amplitudes and electrode configurations to elicit tonic multifidus contractions for 10 seconds twice per minute. All participants were instructed and trained to deliver two 30-minute stimulation sessions per day while prone or lying on their side using their wireless activator.

      Outcomes

      Prespecified outcome measures included the seven-day recall of average low back pain on the 10-cm VAS,
      • Price D.D.
      • McGrath P.A.
      • Rafii A.
      • Buckingham B.
      The validation of visual analogue scales as ratio scale measures for chronic and experimental pain.
      ODI,
      • Fairbank J.C.
      • Pynsent P.B.
      The Oswestry disability index.
      EQ-5D index (EuroQol quality-of-life survey [EQ-5D-5L]),
      • Herdman M.
      • Gudex C.
      • Lloyd A.
      • et al.
      Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L).
      percentage of pain relief (PPR), subject global impression of change (SGIC),
      • Hurst H.
      • Bolton J.
      Assessing the clinical significance of change scores recorded on subjective outcome measures.
      LBP resolution which we defined as VAS ≤ 2.5 cm, treatment satisfaction questionnaire (TSQ), clinical global impression of change (CGI),

      Guy W. ECDEU assessment manual for psychopharmacology. 1976. http://www.archive.org/details/ecdeuassessmentm1933guyw.

      and medication usage. These outcomes were assessed and compared with baseline at six months and one and two years, and annual follow-ups are scheduled for a total of five years.
      Ongoing safety reporting included serious device- or procedure-related adverse events (AEs) that were actively solicited and documented at each visit and reported and coded according to the Medical Dictionary for Regulatory Activities, version 19.1. The Clinical Events Committee (CEC) adjudicated all AEs.

      Data Analysis

      Descriptive statistics including mean and standard deviation or standard error of the mean, 95% confidence intervals (CIs), median, first and third quartiles (Q1 and Q3) were used to summarize continuous variables. Binary outcomes were represented as counts and proportions.
      To reduce potential bias caused by incomplete follow-up, imputation for missing data was stratified based on the reason for missingness. Baseline observation carried forward, or “failure” for binary outcomes, was used for participants withdrawn for reported lack of efficacy at any time or for permanent explant after infection. For those withdrawn for other reasons (ie, precautionary device removal for magnetic resonance imaging [MRI], resolution of pain, a relocation, or otherwise lost to follow-up) or random missed visits, the mixed-effects model repeated measures (MMRM) approach was used to provide implicit imputations of missing data for continuous outcomes.
      • Molenberghs G.
      • Verbeke G.
      Linear Mixed Models for Longitudinal Data.
      To evaluate mean changes from baseline, 95% CIs and adjusted paired t-tests derived from MMRM contrasts were used. Two-sided p values less than 0.05 were considered statistically significant.
      To estimate the proportion of patients achieving “success” for the defined binary outcome variables, multiple imputation (MI) was used for overall estimates of success by visit with associated 95% confidence limits.
      • Rubin D.B.
      Multiple Imputation for Nonresponse in Surveys.
      ,
      • Little R.J.A.
      • Rubin D.B.
      Statistical Analysis with Missing Data.
      Analyses were performed using SAS version 9.3 (SAS Institute Inc, Cary, NC).

      Results

      Study Population

      The demographic and baseline characteristics of the 204 participants are summarized in Table 1. Participants had a mean age of 47 ± 9 years, and 54% were women. The mean duration of CLBP was 14 ± 11 years (range from 7 months to 50 years) from the onset of the first occurrence, and the mean percentage of days with LBP in the previous year was 97 ± 8%. The mean VAS was 7.3 ± 0.7 cm, the mean ODI was 39 ± 10, and the mean EQ-5D-5L index was 0.585 ± 0.174. All participants had undergone physical therapy with an average of 31 ± 52 sessions. Of all participants, 12% had undergone medial branch rhizotomy (>1 year before enrollment), 49% had received spinal injections (>30 days before enrollment), and 37% were taking opioid analgesics for LBP.
      Table 1Baseline Characteristics of the Study Population at Baseline.
      CharacteristicParticipants

      Combined group (N = 204)

      Mean ± SD or n/N (%)
      Age (y)47 ± 9
      Female sex110/204 (54)
      Body mass index (BMI)
      The BMI is the weight in kilograms divided by the square of the height in meters.
      28 ± 4
      Pain duration from onset of the first occurrence (y)14.2 ± 10.6
      Percent of days with low back pain in the past year97 ± 8
      Leg pain associated with back pain53/204 (26)
      Previous medial branch rhizotomy25/204 (12)
      Months from most recent rhizotomy44.4 ± 74.7
      Previous injection procedure99/204 (49)
      Number of previous physical therapy sessions31 ± 52
      Medications for low back pain
       At least one medication for low back pain162/204 (79)
       NSAIDs98/204 (48)
       Opioid-analgesics76/204 (37)
       Simple analgesics42/204 (21)
       Muscle relaxants16/204 (8)
       Anticonvulsants18/204 (9)
       Other (≤5%)24/204 (12)
      VAS score for low back pain
      Scores on the VAS for average recall low back pain over the past seven days range from 0 to 10, with higher scores indicating more severe pain.
      7.3 ± 0.7
      ODI score
      Scores on the ODI range from 0 to 100, with higher scores indicating more severe disability.
      39 ± 10
      EQ-5D-5L index
      Scores on the European Quality of Life with five dimensions and five levels (EQ-5D-5L) index range from −0.5 to 1, with higher scores indicating better quality of life.
      0.585 ± 0.174
      The BMI is the weight in kilograms divided by the square of the height in meters.
      Scores on the VAS for average recall low back pain over the past seven days range from 0 to 10, with higher scores indicating more severe pain.
      Scores on the ODI range from 0 to 100, with higher scores indicating more severe disability.
      § Scores on the European Quality of Life with five dimensions and five levels (EQ-5D-5L) index range from −0.5 to 1, with higher scores indicating better quality of life.

      Participant Disposition

      Longitudinal follow-up data were available for 190 of 204 participants (93%) at six months, 176 of 204 (86%) at one year, and 156 of 204 (79%) at two years (Fig. 2).
      Figure thumbnail gr2
      Figure 2CONSORT flow diagram for participant disposition.
      At the two-year follow-up, ten participants had missed their follow-up visit, and 38 participants had been withdrawn from the study after permanent system explant (31) or otherwise lost to follow-up (7). One participant in whom the system was explanted for infection was reimplanted after the infection had cleared (Fig. 2).

      Two-Year Outcomes

      Completed-Cases Analysis (n = 156)

      Key efficacy outcomes progressively improved over time, and changes from baseline were statistically significant and clinically meaningful at all follow-ups (p < 0.0001; Figs. 3 and 4, Table 2).
      • Dworkin R.H.
      • Turk D.C.
      • Wyrwich K.W.
      • et al.
      Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations.
      • Chou R.
      • Loeser J.D.
      • Owens D.K.
      • et al.
      Interventional therapies, surgery, and interdisciplinary rehabilitation for low back pain: an evidence-based clinical practice guideline from the American Pain Society.
      • Ostelo R.W.J.G.
      • Deyo R.A.
      • Stratford P.
      • et al.
      Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change.
      • Walters S.J.
      • Brazier J.E.
      Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D.
      By two years, the mean average LBP VAS had improved by −4.8 ± 0.2 cm (95% CI −5.2 to −4.5; p < 0.0001), and 72% of participants had a ≥50% reduction in VAS, with an average reduction of 85%, 62% of participants had a ≥70% VAS reduction, and 67% had resolution of CLBP (VAS ≤ 2.5 cm), with an average residual VAS of 0.97 cm. The mean ODI score decreased by −21.4 ± 1.3 (95% CI −24.0 to −18.7; p < 0.0001), and 62% of participants had a ≥20-point ODI reduction, with an average reduction of 32 points. The mean EQ-5D-5L index improved by 0.218 ± 0.017 (95% CI 0.184 to 0.253; p < 0.0001). The proportion of participants with a reduction in LBP VAS of ≥50% and/or ODI of ≥20 points without an increase in either was 77%. The proportion who exceeded these cut-offs in both VAS and ODI was 57%. Within the cohort of participants with two-year follow-up data, 57 of 156 (37%) were taking opioid analgesics at baseline, 34 of 57 (60%) had voluntarily discontinued (39%) or reduced (21%) opioid use, and 1 of 57 (2%) had increased dosage.
      Figure thumbnail gr3
      Figure 3Mean ratings over time for a. low back pain VAS, b. Oswestry disability index, and c. EQ-5D-5L index. All changes from baseline p < 0.0001. Error bars represent the standard error of the mean.
      Figure thumbnail gr4
      Figure 4Response rates at common clinical importance thresholds for a. VAS (reduction ≥50% and 70%, and absolute VAS ≤ 2.5 cm), and b. ODI (≥20 points) and composites of VAS and ODI (≥50% and/or 20 points, ≥50% and/or 20 points). Solid lines represent completed cases, and dashed lines represent imputation for missing data (N = 204).
      Table 2Outcomes Reported for Completers and All Participants With Stratified Imputation for Missing Data.
      AnalysisBaseline

      Mean ± SD
      6 mo

      Mean (SE) or % (n/N)

      (95% CI)
      For continuous outcomes p < 0.0001 for two-sided t-test if change from baseline differs from 0.
      1 y

      Mean (SE) or % (n/N)

      (95% CI)
      For continuous outcomes p < 0.0001 for two-sided t-test if change from baseline differs from 0.
      2 y

      Mean (SE) or % (n/N)

      (95% CI)
      For continuous outcomes p < 0.0001 for two-sided t-test if change from baseline differs from 0.
      N = 204N = 190N = 204N = 176N = 204N = 156N = 204
      LBP VAS (cm)7.3 ± 0.73.7 (0.2)3.9 (0.2)3.0 (0.2)3.4 (0.2)2.4 (0.2)3.1 (0.2)
       Change in VAS (cm)−3.6 (0.2)

      (−3.9, −3.3)
      −3.4 (0.2)

      (−3.8, −3.1)
      −4.3 (0.2)

      (−4.7, −3.9)
      −3.9 (0.2)

      (−4.3, −3.6)
      −4.8 (0.2)

      (−5.2, −4.5)
      −4.2 (0.2)

      (−4.6, −3.8)
       Change in VAS (%)−48.6 (2.7)

      (−53.9, −43.3)
      −47.1 (2.6)

      (−52.3, −41.9)
      −58.9 (2.6)

      (−64.1, −53.6)
      −54.3 (2.7)

      (−59.5, −49.0)
      −66.7 (2.6)

      (−71.7, −61.6)
      −58.1 (2.7)

      (−63.4, −52.8)
       ≥30% improvement in VAS66.1 (125/189)

      (59.4, 72.9)
      63.2 (3.5)

      (56.5, 70.0)
      73.9 (130/176)

      (67.4, 80.4)
      66.9 (3.4)

      (60.3, 73.6)
      82.6 (128/155)

      (76.6, 88.6)
      71.6 (3.3)

      (65.1, 78.1)
       ≥50% improvement in VAS52.9 (100/189)

      (45.8, 60.0)
      51.0 (3.6)

      (44.0, 58.0)
      63.6 (112/176)

      (56.5, 70.7)
      58.0 (3.5)

      (51.1, 65.0)
      71.6 (111/155)

      (64.5, 78.7)
      62.1 (3.5)

      (55.1, 69.0)
       ≥70% improvement in VAS33.9 (64/189)

      (27.1, 40.6)
      33.2 (3.4)

      (26.5, 39.9)
      46.6 (82/176)

      (39.2, 54.0)
      43.0 (3.6)

      (36.1, 50.0)
      61.9 (96/155)

      (54.3, 69.6)
      54.3 (3.7)

      (47.1, 61.5)
       LBP resolution (VAS ≤ 2.5 cm)39.2 (74/189)

      (32.2, 46.1)
      38.3 (3.5)

      (31.4, 45.1)
      51.7 (91/176)

      (44.3, 59.1)
      47.7 (3.5)

      (40.7, 54.6)
      66.5 (103/155)

      (59.0, 73.9)
      57.6 (3.6)

      (50.5, 64.7)
      ODI39.1 ± 10.321.9 (1.1)22.7 (1.0)19.0 (1.4)20.7 (1.0)17.6 (1.2)20.2 (1.0)
       Change in ODI−17.0 (1.1)

      (−19.2, −14.8)
      −16.4 (1.0)

      (−18.4, −14.4)
      −19.9 (1.2)

      (−22.3, −17.6)
      −18.4 (1.0)

      (−20.4, −16.4)
      −21.4 (1.3)

      (−24.0, −18.7)
      −18.9 (1.0)

      (−21.0, −16.8)
       Change in ODI (%)−43.0 (2.8)

      (−48.5, −37.4)
      −41.5 (2.7)

      (−46.8, −36.1)
      −50.5 (2.9)

      (−56.3, −44.8)
      −46.4 (2.8)

      (−51.8, −41.0)
      −54.3 (3.2)

      (−60.6, −48.0)
      −47.5 (2.8)

      (−53.0, −42.0)
       ≥20-point improvement in ODI48.1 (91/189)

      (41.0, 55.3)
      46.7 (3.5)

      (39.8, 53.7)
      57.4 (101/176)

      (50.1, 64.7)
      53.4 (3.5)

      (46.5, 60.3)
      61.3 (95/155)

      (53.6, 69.0)
      54.8 (3.6)

      (47.7, 61.9)
      Composite of VAS and ODI
       ≥50% improvement in VAS and/or ≥20 points ODI63.5 (120/189)

      (56.6, 70.4)
      60.4 (3.5)

      (53.6, 67.2)
      73.3 (129/176)

      (66.8, 79.8)
      67.4 (3.4)

      (60.8, 74.0)
      77.3 (119/154)

      (70.7, 83.9)
      67.4 (3.5)

      (60.4, 74.3)
       ≥50% improvement in VAS and ≥20 points ODI37.8 (71/188)

      (30.8, 44.7)
      36.8 (3.4)

      (30.0, 43.5)
      47.7 (84/176)

      (40.3, 55.1)
      44.0 (3.6)

      (37.0, 51.1)
      56.5 (87/154)

      (48.7, 64.3)
      49.9 (3.6)

      (42.8, 57.1)
      EQ-5D-5L index0.585 ± 0.1740.765 (0.010)0.758 (0.011)0.780 (0.012)0.762 (0.011)0.798 (0.013)0.768 (0.011)
       Change in EQ-5D-5L index0.180 (0.014)

      (0.153, 0.207)
      0.173 (0.011)

      (0.151, 0.194)
      0.198 (0.016)

      (0.167, 0.229)
      0.177 (0.011)

      (0.156, 0.199)
      0.218 (0.017)

      (0.184, 0.253)
      0.183 (0.011)

      (0.161, 0.205)
      PPR (%)55.0 (2.5)

      (50.1, 59.9)
      53.3 (2.5)

      (48.4, 58.2)
      65.7 (2.4)

      (60.9, 70.5)
      60.7 (2.5)

      (55.8, 65.6)
      72.1 (2.4)

      (67.3, 77.0)
      62.3 (2.5)

      (57.3, 67.3)
      SGIC “Better” or “Much better”57.4 (109/190)

      (50.3, 64.4)
      55.1 (3.5)

      (48.2, 62.0)
      71.6 (126/176)

      (64.9, 78.3)
      65.9 (3.4)

      (59.3, 72.5)
      78.6 (121/154)

      (72.1, 85.1)
      68.6 (3.4)

      (61.9, 75.2)
      TSQ “Definitely satisfied”64.7 (123/190)

      (57.9, 71.5)
      62.8 (3.4)

      (56.0, 69.5)
      78.2 (136/174)

      (72.0, 84.3)
      71.8 (3.2)

      (65.5, 78.1)
      80.0 (124/155)

      (73.7, 86.3)
      68.3 (3.4)

      (61.6, 75.1)
      CGI “Much better”56.8 (108/190)

      (49.8, 63.9)
      55.0 (3.6)

      (48.0, 62.0)
      73.3 (129/176)

      (66.8, 79.8)
      67.5 (3.4)

      (60.8, 74.1)
      77.6 (118/152)

      (71.7, 84.3)
      66.6 (3.6)

      (59.6, 73.7)
      Continuous outcome estimates from mixed model repeated measures regression models adjusted for baseline, all other binary outcomes analyzed with MI for missing data. Statistics are expressed as % (n/N) for binary outcomes and N, mean (standard error) for continuous outcomes.
      For continuous outcomes p < 0.0001 for two-sided t-test if change from baseline differs from 0.

      Imputed Analysis (N = 204)

      A side-by-side comparison of the completed-cases analysis (n = 156) and the imputed intention-to-treat (ITT) analysis (N = 204) is provided in Table 2. Reported outcomes remained statistically significant (p < 0.0001) and clinically meaningful at all follow-ups.

      Device Use

      Over 94% of participants continued delivering treatment between one and two years. During the 60 days leading to the six-month follow-up, the median device use was 88% (Q1 73%, Q3 96%) of the maximum duration possible (number of days times the maximum duration of 60 minutes daily); for the 90 days leading to the 12-month follow-up, it was 77% (Q1 48%, Q3 90%); and for the 90 days leading to the 24-month follow-up, it was 42% (Q1 6.5%, Q3 75%) (Fig. 5).
      Figure thumbnail gr5
      Figure 5Interquartile ranges of device use as a percentage of the maximum cumulative duration permitted.

      Safety Analysis

      Device- or procedure-related serious AEs (SAEs) are summarized in Table 3 by follow-up interval. Events through the one-year visit have been discussed previously.
      • Gilligan C.
      • Volschenk W.
      • Russo M.
      • et al.
      An implantable restorative-neurostimulator for refractory mechanical chronic low back pain: a randomized sham-controlled clinical trial.
      No additional device- or procedure-related SAEs were reported, and no lead migrations were observed. Overall, 45 of 204 participants (22%) underwent a total of 47 surgical interventions, during which 32 systems were removed (16%), one system was reimplanted (<1%), four pulse generators were repositioned (2%), and ten participants had their leads replaced (5%). Reasons for system removal were lack of efficacy (n = 18), infection (n = 6), safety precaution before MRI scan (n = 6), resolution of LBP (n = 1), and relocation to a remote area without device follow-up infrastructure (n = 1). Seven unrelated SAEs were reported for seven participants (3%) during the first year, and six unrelated SAEs were reported for six participants (3%) during the second year of the follow-up. All 13 events were reviewed by the CEC and adjudicated as unrelated to the device or procedure.
      Table 3Device- and Procedure-Related SAEs and Surgical Interventions.
      Type of event and reason0–6 mo6–12 mo12–24 mo
      Events

      n
      Patients

      n/N (%)
      Events

      n
      Patients

      n/N (%)
      Events

      n
      Patients

      n/N (%)
      Device- and procedure-related SAEs
       Infection (resolved)66/204 (2.9)
       Intraprocedural upper airway obstruction (resolved)11/204 (0.5)
       Nonradicular patch of numbness on thigh (ongoing)11/204 (0.5)
      Surgical interventions and reasons
      Patients may have had more than one procedure; therefore, the total does not equal the sum of the categories.
       System removal88/204 (3.9)1111/204 (5.4)1313/204 (6.4)
      Reported lack of efficacy11/204 (0.5)88/204 (3.9)99/204 (4.4)
      Infection
      One patient was reimplanted after the infection cleared.
      66/204 (2.9)
      Facilitate MRI11/204 (0.5)33/204 (1.5)22/204 (1.0)
      Participant relocation11/204 (0.5)
      LBP pain relief11/204 (0.5)
       Reimplant post-infection
      One patient was reimplanted after the infection cleared.
      11/204 (0.5)
       Revision55/204 (2.5)55/204 (2.5)55/204 (2.5)
      Lead replacement33/204 (1.5)33/204 (1.5)44/204 (2.0)
      Pulse generator repositioning22/204 (1.0)22/204 (0.9)11/204 (0.5)
      Patients may have had more than one procedure; therefore, the total does not equal the sum of the categories.
      One patient was reimplanted after the infection cleared.

      Discussion

      Restorative neurostimulation is indicated for patients with refractory mechanical CLBP secondary to multifidus muscle dysfunction and no pathology seen on MRI that is clearly identified and is likely the cause of the CLBP that is amenable to surgery.
      Before enrollment, all participants had failed conventional medical management, which included at least physical therapy and medication for LBP. Most participants had undergone one or more interventional procedures, and over a third were on chronic opioids. Published studies on this condition consistently report that these patients very rarely experience spontaneous, substantial improvements in their pain and disability.
      • Itz C.J.
      • Geurts J.W.
      • van Kleef M.
      • Nelemans P.
      Clinical course of non-specific low back pain: a systematic review of prospective cohort studies set in primary care.
      ,
      • Kongsted A.
      • Kent P.
      • Axen I.
      • Downie A.S.
      • Dunn K.M.
      What have we learned from ten years of trajectory research in low back pain?.
      • Chen Y.
      • Campbell P.
      • Strauss V.Y.
      • Foster N.E.
      • Jordan K.P.
      • Dunn K.M.
      Trajectories and predictors of the long-term course of low back pain: cohort study with 5-year follow-up.
      • Dunn K.M.
      • Campbell P.
      • Jordan K.P.
      Long-term trajectories of back pain: cohort study with 7-year follow-up.
      • Costa Lda C.
      • Maher C.G.
      • McAuley J.H.
      • et al.
      Prognosis for patients with chronic low back pain: inception cohort study.
      • Tamcan O.
      • Mannion A.F.
      • Eisenring C.
      • Horisberger B.
      • Elfering A.
      • Müller U.
      The course of chronic and recurrent low back pain in the general population.
      • Costa L.O.P.
      • Maher C.G.
      • Latimer J.
      • et al.
      Motor control exercise for chronic low back pain: a randomized placebo-controlled trial.
      Longitudinal follow-ups demonstrated a progressive recovery trajectory that is consistent with restoration of multifidus neuromuscular control. At the two-year follow-up, durable and clinically substantial benefits had accrued in all predefined outcome measures (p < 0.0001). Although the study did not directly compare restorative neurostimulation with other available therapies, the improvements were more substantial than those seen with other available therapies for this patient population, such as motor control exercises.
      • Saragiotto B.T.
      • Maher C.G.
      • Yamato T.P.
      • et al.
      Motor control exercise for chronic non-specific low-back pain.
      ,
      • Mueller J.
      • Niederer D.
      Dose-response-relationship of stabilisation exercises in patients with chronic non-specific low back pain: a systematic review with meta-regression.
      At baseline, average pain was severe (VAS of 7.3 cm) and disability marginally severe (ODI of 39.1), but after two years of treatment, average pain and disability were mild (2.4 cm and 17.6, respectively). Individual clinical benefits were reflected in the “responder” analysis of the core outcome domains, in which 72% of participants showed a substantial (≥50%) pain reduction,
      • Dworkin R.H.
      • Turk D.C.
      • Wyrwich K.W.
      • et al.
      Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations.
      with an average improvement from baseline within this cohort of 84%. A total of 67% of participants reported LBP resolution with an average residual VAS of 0.97 cm, and 61% of participants reported a substantial (≥20 points) ODI improvement,
      • Chou R.
      • Loeser J.D.
      • Owens D.K.
      • et al.
      Interventional therapies, surgery, and interdisciplinary rehabilitation for low back pain: an evidence-based clinical practice guideline from the American Pain Society.
      with an average reduction of 32 points from baseline.
      Pain and disability are interdependent symptoms of the underlying etiology and codeterminants of a patient’s well-being or health state.
      • Dworkin R.H.
      • Turk D.C.
      • Farrar J.T.
      • et al.
      Core outcome measures for chronic pain clinical trials: IMMPACT recommendations.
      Improvements in each or both of these outcome domains are recognized as treatment success by patients, physicians, and regulators.
      • Chiarotto A.
      • Boers M.
      • Deyo R.A.
      • et al.
      Core outcome measurement instruments for clinical trials in nonspecific low back pain.

      Food and Drug Administration. Guidance for Industry on Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims; 2009.

      • Pham T.
      • Van Der Heijde D.
      • Lassere M.
      • et al.
      Outcome variables for osteoarthritis clinical trials: the OMERACT-OARSI set of responder criteria.
      A total of 77% of participants experienced a substantial improvement in LBP VAS (≥50%) and/or ODI (≥20 points), and 57% experienced such substantial improvements in both. The average EQ-5D-5L utility score had increased from 0.585 to 0.798, which closely approaches the age-matched US population norm of 0.815.
      • Jiang R.
      • Janssen M.F.B.
      • Pickard A.S.
      US population norms for the EQ-5D-5L and comparison of norms from face-to-face and online samples.

      Medication Use

      Insufficient relief from existing treatments commonly leads to prescription of chronic opioid therapy for patients with CLBP, despite frequent poor outcomes.
      • Qaseem A.
      • Wilt T.J.
      • McLean R.M.
      • et al.
      Noninvasive treatments for acute, subacute, and chronic low back pain: a clinical practice guideline from the American College of Physicians.
      Even though participants were considered refractory to pain medication, 76 of 204 (37%) were receiving chronic opioid therapy for their CLBP at baseline. Of the 57 of 156 participants who used opioids at baseline and had a two-year follow-up, 60% had either voluntarily discontinued use or decreased consumption, and only one patient had increased intake. These results suggest that the treatment helps patients voluntarily abandon or reduce opioid consumption. Similar reductions were reported for other LBP medications, including nonsteroidal anti-inflammatory drugs (NSAIDs), simple analgesics, and muscle relaxants, providing further support of clinical benefit.

      Device Use

      Device use remained high during the second year, with over 94% of participants delivering treatment. Among the 103 of 155 participants (67%) with resolution of CLBP, median device use was 48% (Q1 15%, Q3 74%), which was not materially different from the overall distribution. Eight participants did not administer any treatment during the second year, and four of them had resolution of LBP with a residual average VAS of 1.4 cm. A total of 16 participants had administered over 90% of the maximum permitted treatment amount, and 13 also had resolution of LBP with a residual average VAS of 0.42 cm. This illustrates that even though LBP resolution can be sustained in the absence of stimulation, participants may prefer to continue with therapy delivery despite resolution of their symptoms. Within the remission cohort, three participant profiles are thus emerging: those who stop or minimize stimulation and remain in remission, those who require occasional stimulation to remain in remission or to manage flare-ups, and those who regularly administer a high level of stimulation despite being in remission.

      Safety

      The overall incidence of related SAEs was 8 of 204 (3.9%; Table 3), including six post-surgery infections requiring system removal, all reported during the first four months of follow-up. The permanent system removal rate of 31 of 204 (15.2%) is in line with spinal cord stimulation reports over the same two-year period,
      • Wang V.C.
      • Bounkousohn V.
      • Fields K.
      • Bernstein C.
      • Paicius R.M.
      • Gilligan C.
      Explantation rates of high frequency spinal cord stimulation in two outpatient clinics.
      ,
      • Hagedorn J.M.
      • Lam C.M.
      • D’Souza R.S.
      • et al.
      Explantation of 10 kHz spinal cord stimulation devices: a retrospective review of 744 patients followed for at least 12 months.
      and the rate of participants requiring surgical intervention, 45 of 204 (20%), is comparable with published incidence data for other neuromodulation therapies for chronic pain.
      • Hayek S.M.
      • Veizi E.
      • Hanes M.
      Treatment-limiting complications of percutaneous spinal cord stimulator implants: a review of eight years of experience from an academic center database.
      • Eldabe S.
      • Buchser E.
      • Duarte R.V.
      Complications of spinal cord stimulation and peripheral nerve stimulation techniques: a review of the literature.
      • Shamji M.F.
      • Westwick H.J.
      • Heary R.F.
      Complications related to the use of spinal cord stimulation for managing persistent postoperative neuropathic pain after lumbar spinal surgery.
      Lead migration represents the most common AE reported in neurostimulation trials, occurring at the rates of 1.4% to 13.6%.
      • Hayek S.M.
      • Veizi E.
      • Hanes M.
      Treatment-limiting complications of percutaneous spinal cord stimulator implants: a review of eight years of experience from an academic center database.
      ,
      • Deer T.R.
      • Mekhail N.
      • Provenzano D.
      • et al.
      The appropriate use of neurostimulation of the spinal cord and peripheral nervous system for the treatment of chronic pain and ischemic diseases: the Neuromodulation Appropriateness Consensus Committee.
      No lead migrations were observed in this trial, demonstrating the effectiveness of the distal fixation tines.

      Strengths and Limitations

      The strength of this study is that it reports on a relatively large and homogeneous cohort of patients with severe and refractory CLBP with an extended follow-up duration of two years. The study demonstrates durable benefits in patients with a baseline pain duration longer than one year. However, because only three patients had a pain duration between 6 and 12 months, the benefits of earlier intervention remain to be studied.
      The main limitation is the absence of a long-term comparator because of therapy activation in the sham-control group after conclusion of the blinded phase at four months. Furthermore, studies with long follow-up durations will inherently have to account for missing data, particularly those for chronic pain conditions.
      • Kim Y.
      Missing data handling in chronic pain trials.
      Indiscriminate use of last observation carried forward has been criticized as a source of systematic bias in chronic pain trials,
      • Palmer R.H.
      Estimate at your peril: imputation methods for patient withdrawal can bias efficacy outcomes in chronic pain trials using responder analyses.
      and more appropriate methods have been recommended.

      Permutt TJ. E9(R1)Statistical Principles for Clinical Trials: Addendum: Estimands and Sensitivity Analysis in Clinical Trials. Food and Drug Administration, HHS. 2017:50433–50434.

      • McNicol E.
      • Ferguson M.
      • Bungay K.
      • et al.
      Systematic review of research methods and reporting quality of randomized clinical trials of spinal cord stimulation for pain.
      • Herbert R.D.
      • Kasza J.
      • Bø K.
      Analysis of randomised trials with long-term follow-up.
      To inform the interpretation of the complete-cases analyses (n = 156), we have provided supporting ITT analyses (N = 204) using transparent and conservative MMRM imputation, which was stratified based on likely randomness of, and reason for, missing data. The relatively small difference across all outcome measures between the completed-cases and imputed analyses and the statistical significance and clinical relevance of results in both (Table 2) instills confidence in the robustness of our data and the validity of the conclusions drawn.

      Conclusions

      The two-year results of the ReActiv8-B trial show durable, statistically significant, and clinically substantial benefits in a cohort of patients with severe, disabling CLBP and multifidus muscle dysfunction who were refractory to conservative care including physical therapy and medications. Participants demonstrated improvements in pain and disability that increased the longer they were treated. This recovery trajectory is consistent with restoration of neuromuscular control and structural muscle changes. The safety profile of the therapy was favorable compared with that of available implantable neurostimulators for the treatment of back pain.

      Acknowledgements

      The authors thank the subinvestigators, research coordinators, and nursing staff at the study sites for their contribution to site management and patient care. The authors thank Robert H. Dworkin, PhD, for his advice on interpretation and presentation of trial results, Teresa Yurik, MS, and Lisa Grant, MS, for their statistical advice and data analysis support, and Diane Burnside, BS, and Jason Shiroff, BS (Mainstay Medical Clinical Department), for management of the trial.

      Authorship Statements

      Christopher Gilligan, Richard Rauck, James Rathmell, Timothy Deer, Shivanand Lad, Jeffrey Fischgrund, Bruce Mitchell, Kristiaan Deckers, Kris De Smedt, Sam Eldabe, Marc Russo, Jean-Pierre Van Buyten, Ganesan Baranidharan, and Vivek Mehta contributed to the development of the protocol. Christopher Gilligan drafted the manuscript, and Sam Eldabe revised the manuscript. All authors reviewed and approved the manuscript before initial submission. All authors were clinical investigators on the trial, with the following exceptions: Richard Rauck served as chair of the Data Monitoring Committee, James Rathmell served as chair of the Clinical Events Committee, William Klemme served as independent MRI reviewer, and Jan Pieter Heemels provided editorial support.

      Supplementary Data

      References

        • Itz C.J.
        • Geurts J.W.
        • van Kleef M.
        • Nelemans P.
        Clinical course of non-specific low back pain: a systematic review of prospective cohort studies set in primary care.
        Eur J Pain. 2013; 17: 5-15https://doi.org/10.1002/j.1532-2149.2012.00170.x
        • Breivik H.
        • Collett B.
        • Ventafridda V.
        • Cohen R.
        • Gallacher D.
        Survey of chronic pain in Europe: prevalence, impact on daily life, and treatment.
        Eur J Pain. 2006; 10: 287-333https://doi.org/10.1016/j.ejpain.2005.06.009
        • Alsaadi S.M.
        • McAuley J.H.
        • Hush J.M.
        • Maher C.G.
        Prevalence of sleep disturbance in patients with low back pain.
        Eur Spine J. 2011; 20: 737-743https://doi.org/10.1007/s00586-010-1661-x
        • Deyo R.A.
        • Weinstein J.N.
        Low back pain.
        N Engl J Med. 2001; 344: 363-370https://doi.org/10.1056/NEJM200102013440508
        • Förster M.
        • Mahn F.
        • Gockel U.
        • et al.
        Axial low back pain: one painful area—many perceptions and mechanisms.
        PLoS One. 2013; 8e68273https://doi.org/10.1371/journal.pone.0068273
        • Saito T.
        • Steinke H.
        • Miyaki T.
        • et al.
        Analysis of the posterior ramus of the lumbar spinal nerve: the structure of the posterior ramus of the spinal nerve.
        Anesthesiology. 2013; 118: 88-94https://doi.org/10.1097/ALN.0b013e318272f40a
        • Bogduk N.
        On the definitions and physiology of back pain, referred pain, and radicular pain.
        Pain. 2009; 147: 17-19
        • Ward S.R.
        • Eng C.M.
        • Gottschalk L.J.
        • Kim C.W.
        • Garfin S.R.
        • Lieber R.L.
        The architectural design of the lumbar multifidus muscle supports its role as stabilizer.
        J Biomech. 2006; 39: S101
        • Kim C.W.
        • Gottschalk L.J.
        • Eng C.
        • Ward S.R.
        • Lieber R.L.
        The multifidus muscle is the strongest stabilizer of the lumbar spine.
        Spine J. 2007; 7: 76Shttps://doi.org/10.1016/j.spinee.2007.07.188
        • Rosatelli A.L.
        • Ravichandiran K.
        • Agur A.M.
        Three-dimensional study of the musculotendinous architecture of lumbar multifidus and its functional implications.
        Clin Anat. 2008; 21: 539-546https://doi.org/10.1002/ca.20659
        • Teichtahl A.J.
        • Urquhart D.M.
        • Wang Y.
        • et al.
        Fat infiltration of paraspinal muscles is associated with low back pain, disability, and structural abnormalities in community-based adults.
        Spine J. 2015; 15: 1593-1601https://doi.org/10.1016/j.spinee.2015.03.039
        • Freeman M.D.
        • Woodham M.A.
        • Woodham A.W.
        The role of the lumbar multifidus in chronic low back pain: a review.
        PM R. 2010; 2: 142-167https://doi.org/10.1016/j.pmrj.2009.11.006
        • Shahidi B.
        • Hubbard J.C.
        • Gibbons M.C.
        • et al.
        Lumbar multifidus muscle degenerates in individuals with chronic degenerative lumbar spine pathology.
        J Orthop Res. 2017; 35: 2700-2706https://doi.org/10.1002/jor.23597
        • Meier M.L.
        • Vrana A.
        • Schweinhardt P.
        Low back pain: the potential contribution of supraspinal motor control and proprioception.
        Neuroscientist. 2019; 25: 583-596https://doi.org/10.1177/1073858418809074
        • Chou R.
        • Deyo R.
        • Friedly J.
        • et al.
        Nonpharmacologic therapies for low back pain: a systematic review for an American College of Physicians clinical practice guideline.
        Ann Intern Med. 2017; 166: 480-492https://doi.org/10.7326/M16-2458
      1. Skelly AC, Chou R, Dettori JR, et al. Noninvasive Nonpharmacological Treatment for Chronic Pain: A Systematic Review; 2018. https://doi.org/10.23970/AHRQEPCCER209.

        • Tsao H.
        • Druitt T.R.
        • Schollum T.M.
        • Hodges P.W.
        Motor training of the lumbar paraspinal muscles induces immediate changes in motor coordination in patients with recurrent low back pain.
        J Pain. 2010; 11: 1120-1128https://doi.org/10.1016/j.jpain.2010.02.004
        • Hodges P.W.
        • Danneels L.
        Changes in structure and function of the back muscles in low back pain: different time points, observations, and mechanisms.
        J Orthop Sports Phys Ther. 2019; 49: 464-476https://doi.org/10.2519/jospt.2019.8827
        • Russo M.
        • Deckers K.
        • Eldabe S.
        • et al.
        Muscle control and non-specific chronic low back pain.
        Neuromodulation. 2018; 21: 1-9https://doi.org/10.1111/ner.12738
        • Deckers K.
        • De Smedt K.
        • Mitchell B.
        • et al.
        New therapy for refractory chronic mechanical low back pain—restorative neurostimulation to activate the lumbar multifidus: one year results of a prospective multicenter clinical trial.
        Neuromodulation. 2018; 21: 48-55https://doi.org/10.1111/ner.12741
        • Gilligan C.
        • Volschenk W.
        • Russo M.
        • et al.
        An implantable restorative-neurostimulator for refractory mechanical chronic low back pain: a randomized sham-controlled clinical trial.
        Pain. 2021; 162: 2486-2498https://doi.org/10.1097/j.pain.0000000000002258
        • Hicks G.E.
        • Fritz J.M.
        • Delitto A.
        • Mishock J.
        Interrater reliability of clinical examination measures for identification of lumbar segmental instability.
        Arch Phys Med Rehabil. 2003; 84: 1858-1864https://doi.org/10.1016/S0003-9993(03)00365-4
        • Moher D.
        • Hopewell S.
        • Schulz K.F.
        • et al.
        CONSORT 2010 explanation and elaboration: updated guidelines for reporting parallel group randomised trials.
        BMJ. 2010; 340: c869https://doi.org/10.1136/bmj.c869
        • Price D.D.
        • McGrath P.A.
        • Rafii A.
        • Buckingham B.
        The validation of visual analogue scales as ratio scale measures for chronic and experimental pain.
        Pain. 1983; 17: 45-56https://doi.org/10.1016/0304-3959(83)90126-4
        • Fairbank J.C.
        • Pynsent P.B.
        The Oswestry disability index.
        Spine (Phila Pa 1976). 2000; 25 ([discussion: 2952]): 2940-2952
        • Herdman M.
        • Gudex C.
        • Lloyd A.
        • et al.
        Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L).
        Qual Life Res. 2011; 20: 1727-1736https://doi.org/10.1007/s11136-011-9903-x
        • Hurst H.
        • Bolton J.
        Assessing the clinical significance of change scores recorded on subjective outcome measures.
        J Manipulative Physiol Ther. 2004; 27: 26-35https://doi.org/10.1016/j.jmpt.2003.11.003
      2. Guy W. ECDEU assessment manual for psychopharmacology. 1976. http://www.archive.org/details/ecdeuassessmentm1933guyw.

        • Molenberghs G.
        • Verbeke G.
        Linear Mixed Models for Longitudinal Data.
        1st ed. Springer, New York2000
        • Rubin D.B.
        Multiple Imputation for Nonresponse in Surveys.
        J. Wiley & Sons, Inc, 2004
        • Little R.J.A.
        • Rubin D.B.
        Statistical Analysis with Missing Data.
        2nd ed. J. Wiley & Sons, 2002
        • Dworkin R.H.
        • Turk D.C.
        • Wyrwich K.W.
        • et al.
        Interpreting the clinical importance of treatment outcomes in chronic pain clinical trials: IMMPACT recommendations.
        J Pain. 2008; 9: 105-121https://doi.org/10.1016/j.jpain.2007.09.005
        • Chou R.
        • Loeser J.D.
        • Owens D.K.
        • et al.
        Interventional therapies, surgery, and interdisciplinary rehabilitation for low back pain: an evidence-based clinical practice guideline from the American Pain Society.
        Spine (Phila Pa 1976). 2009; 34: 1066-1077https://doi.org/10.1097/BRS.0b013e3181a1390d
        • Ostelo R.W.J.G.
        • Deyo R.A.
        • Stratford P.
        • et al.
        Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change.
        Spine (Phila Pa 1976). 2008; 33: 90-94https://doi.org/10.1097/BRS.0b013e31815e3a10
        • Walters S.J.
        • Brazier J.E.
        Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D.
        Qual Life Res. 2005; 14: 1523-1532https://doi.org/10.1007/s11136-004-7713-0
        • Kongsted A.
        • Kent P.
        • Axen I.
        • Downie A.S.
        • Dunn K.M.
        What have we learned from ten years of trajectory research in low back pain?.
        BMC Musculoskelet Disord. 2016; 17: 220https://doi.org/10.1186/s12891-016-1071-2
        • Chen Y.
        • Campbell P.
        • Strauss V.Y.
        • Foster N.E.
        • Jordan K.P.
        • Dunn K.M.
        Trajectories and predictors of the long-term course of low back pain: cohort study with 5-year follow-up.
        Pain. 2018; 159: 252-260https://doi.org/10.1097/j.pain.0000000000001097
        • Dunn K.M.
        • Campbell P.
        • Jordan K.P.
        Long-term trajectories of back pain: cohort study with 7-year follow-up.
        BMJ Open. 2013; 3e003838https://doi.org/10.1136/bmjopen-2013-003838
        • Costa Lda C.
        • Maher C.G.
        • McAuley J.H.
        • et al.
        Prognosis for patients with chronic low back pain: inception cohort study.
        BMJ. 2009; 339: b3829https://doi.org/10.1136/bmj.b3829
        • Tamcan O.
        • Mannion A.F.
        • Eisenring C.
        • Horisberger B.
        • Elfering A.
        • Müller U.
        The course of chronic and recurrent low back pain in the general population.
        Pain. 2010; 150: 451-457https://doi.org/10.1016/j.pain.2010.05.019
        • Costa L.O.P.
        • Maher C.G.
        • Latimer J.
        • et al.
        Motor control exercise for chronic low back pain: a randomized placebo-controlled trial.
        Phys Ther. 2009; 89: 1275-1286https://doi.org/10.2522/ptj.20090218
        • Saragiotto B.T.
        • Maher C.G.
        • Yamato T.P.
        • et al.
        Motor control exercise for chronic non-specific low-back pain.
        Cochrane Database Syst Rev. 2016; : CD012004https://doi.org/10.1002/14651858.CD012004
        • Mueller J.
        • Niederer D.
        Dose-response-relationship of stabilisation exercises in patients with chronic non-specific low back pain: a systematic review with meta-regression.
        Sci Rep. 2020; 10: 16921https://doi.org/10.1038/s41598-020-73954-9
        • Dworkin R.H.
        • Turk D.C.
        • Farrar J.T.
        • et al.
        Core outcome measures for chronic pain clinical trials: IMMPACT recommendations.
        Pain. 2005; 113: 9-19https://doi.org/10.1016/j.pain.2004.09.012
        • Chiarotto A.
        • Boers M.
        • Deyo R.A.
        • et al.
        Core outcome measurement instruments for clinical trials in nonspecific low back pain.
        Pain. 2018; 159: 481-495https://doi.org/10.1097/j.pain.0000000000001117
      3. Food and Drug Administration. Guidance for Industry on Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims; 2009.

        • Pham T.
        • Van Der Heijde D.
        • Lassere M.
        • et al.
        Outcome variables for osteoarthritis clinical trials: the OMERACT-OARSI set of responder criteria.
        J Rheumatol. 2003; 30: 1648-1654
        • Jiang R.
        • Janssen M.F.B.
        • Pickard A.S.
        US population norms for the EQ-5D-5L and comparison of norms from face-to-face and online samples.
        Qual Life Res. 2021; 30: 803-816https://doi.org/10.1007/s11136-020-02650-y
        • Qaseem A.
        • Wilt T.J.
        • McLean R.M.
        • et al.
        Noninvasive treatments for acute, subacute, and chronic low back pain: a clinical practice guideline from the American College of Physicians.
        Ann Intern Med. 2017; 166: 514-530https://doi.org/10.7326/M16-2367
        • Wang V.C.
        • Bounkousohn V.
        • Fields K.
        • Bernstein C.
        • Paicius R.M.
        • Gilligan C.
        Explantation rates of high frequency spinal cord stimulation in two outpatient clinics.
        Neuromodulation. 2021; 24: 507-511https://doi.org/10.1111/ner.13280
        • Hagedorn J.M.
        • Lam C.M.
        • D’Souza R.S.
        • et al.
        Explantation of 10 kHz spinal cord stimulation devices: a retrospective review of 744 patients followed for at least 12 months.
        Neuromodulation. 2021; 24: 499-506https://doi.org/10.1111/ner.13359
        • Hayek S.M.
        • Veizi E.
        • Hanes M.
        Treatment-limiting complications of percutaneous spinal cord stimulator implants: a review of eight years of experience from an academic center database.
        Neuromodulation. 2015; 18: 603-609https://doi.org/10.1111/ner.12312
        • Eldabe S.
        • Buchser E.
        • Duarte R.V.
        Complications of spinal cord stimulation and peripheral nerve stimulation techniques: a review of the literature.
        Pain Med. 2016; 17: 325-336https://doi.org/10.1093/pm/pnv025
        • Shamji M.F.
        • Westwick H.J.
        • Heary R.F.
        Complications related to the use of spinal cord stimulation for managing persistent postoperative neuropathic pain after lumbar spinal surgery.
        Neurosurg Focus. 2015; 39: E15https://doi.org/10.3171/2015.7.FOCUS15260
        • Deer T.R.
        • Mekhail N.
        • Provenzano D.
        • et al.
        The appropriate use of neurostimulation of the spinal cord and peripheral nervous system for the treatment of chronic pain and ischemic diseases: the Neuromodulation Appropriateness Consensus Committee.
        Neuromodulation. 2014; 17 ([discussion: 550]): 515-550
        • Kim Y.
        Missing data handling in chronic pain trials.
        J Biopharm Stat. 2011; 21: 311-325https://doi.org/10.1080/10543406.2011.550112
        • Palmer R.H.
        Estimate at your peril: imputation methods for patient withdrawal can bias efficacy outcomes in chronic pain trials using responder analyses.
        Pain. 2012; 153: 1541https://doi.org/10.1016/j.pain.2012.04.024
      4. Permutt TJ. E9(R1)Statistical Principles for Clinical Trials: Addendum: Estimands and Sensitivity Analysis in Clinical Trials. Food and Drug Administration, HHS. 2017:50433–50434.

        • McNicol E.
        • Ferguson M.
        • Bungay K.
        • et al.
        Systematic review of research methods and reporting quality of randomized clinical trials of spinal cord stimulation for pain.
        J Pain. 2021; 22: 127-142https://doi.org/10.1016/j.jpain.2020.05.001
        • Herbert R.D.
        • Kasza J.
        • Bø K.
        Analysis of randomised trials with long-term follow-up.
        BMC Med Res Methodol. 2018; 18: 48https://doi.org/10.1186/s12874-018-0499-5