Neurostimulation as a potential modality for transitioning patients with chronic refractory pain syndromes off opioid analgesics: A systematic review

INTRODUCTION

Opioid analgesics are drugs used in practice for pain management. These medications exert analgesia by agonizing mu-opioid receptors, therefore decreasing painful sensations. Common opioid analgesics utilized in medicine include oxycodone, hydrocodone, fentanyl, morphine, codeine, and hydromorphone. Opioids are reserved for arduous pain refractory to other pharmacological options because opioids can lead to abuse, dependence, respiratory depression, overdose, and death.^[1] In 2021, there were a reported 80,816 opioid overdose deaths.^[2] Opioids are often used illicitly outside practitioners’ prescribing for self-treatment, euphoria, or both.^[1]

Despite their effectiveness, opioids pose a substantial public health risk warranting a need to investigate alternative pain-management methods.

Neurostimulation refers to the use of electricity to stimulate nerves or brain regions implicated in painful sensations.^[3,4] Electricity is transmitted to target regions through a medical device, and said medical device is either applied externally or surgically implanted.^[3,5] External neurostimulators include a transcutaneous electrical nerve stimulator (TENS) unit – which applies electricity to painful areas^[6] – or transcranial magnetic stimulation (TMS) – a device which applies magnetic energy to specific brain regions.^[5] Surgically implanted neurostimulators include deep brain stimulation – which is a series of electrodes which stimulate target brain regions^[7] – or spinal cord stimulators (SCS) – a device which stimulates components of the spinal cord.^[3] Neurostimulators have demonstrated a promising niche in pain management, but their utility is limited.

Neurostimulators and opioids will intersect in this review because as the opioid epidemic continues in the United States (US), the need to transition eligible patients off opioids becomes greater. At present, there are a multitude of neurostimulators indicated for pain management available in the US, giving patients numerous avenues for pain management. As of August 18, 2022, more than 1400 neurostimulators for treating pain have been authorized by the Food and Drug Administration (FDA) according to the 510(k) Premarket Notification, de novo, and Premarket Approval databases.^[7-11] The authorized neurostimulators were identified by pinpointing all relevant product codes and then searching each database under each product code. Product codes are used by the FDA to help identify a generic category of a medical device.^[12,13] This repository of neurostimulators demonstrates a growing field of nonpharmacotherapeutic options for patients with acute or chronic pain. Not all neurostimulators are equal; each neurostimulator has its own unique attributes, effectiveness profile, and safety profile. As a result, some neurostimulators may have greater potential for helping patients than others, thereby creating the theoretical possibility that there is at least one effective neurostimulator for everyone somewhere on the market. If a neurostimulator is effective for a patient, it would hopefully provide pain relief to an extent such that opioid consumption is decreased or eliminated.

This review will henceforth serve as a literature review identifying clinical studies which address the effects neurostimulation has on opioid consumption. Collecting this information could suggest a role for neurostimulation in opioid cessation or opioid dose reduction. Transitioning from opioids to neurostimulation is a gap in care which requires more understanding. This review seeks to bridge that gap.

METHODS

Four searches were conducted on PubMed and the Cochrane trials database between August 18, 2022, and August 22, 2022, yielding 62 unique results. On PubMed, the searches were “neurostimulation and opioid consumption,” “(“Implantable Neurostimulations” [Mesh]) AND “Analgesics, Opioid”[Mesh],” and “(“Neurostimulators”) AND (“Opioids”).” In Cochrane trials, the following search was conducted: “opioid in Title Abstract Keyword AND neurostimulator in Title Abstract Keyword – (Word variations have been searched).” Among the 62 results, 35 were relevant to neurostimulator-induced non-migraine pain control based on titles alone. These 35 entries were narrowed further by reading abstracts and/or full-body texts to eliminate articles which either did not meet inclusion criteria or met exclusion criteria. Each entry was evaluated for bias using the Cochrane Risk of Bias 2 tool. Articles were included if they were retrospective or prospective clinical studies, randomized controlled trials, or multi-case reports which assessed effects of neurostimulation on opioid consumption in humans. Articles were excluded (1) if neurostimulators were not being used for pain, (2) if they were commentaries or single case reports (one patient), (3) if they were protocols, or (4) if they were about neurostimulation used for acute pain (i.e., post-operative). After narrowing results based on inclusion-exclusion criteria, the remaining articles were reviewed and key points extracted for inclusion in the results.

RESULTS

Six search entries met all criteria for inclusion in the results. [Table 1] illustrates reasons why search entries were excluded from the final analysis. The six results consisted of three prospective clinical trials,^[14-16] one retrospective study,^[17] one study comprising three case reports,^[18] and one prospective neurophysiological study which assessed mu-opioid receptor availability before and after TMS.^[19] There were 273 participants included in final analyses with 139/273 (50.9%) using opioids before enrollment into their respective studies. Five studies analyzed opioid consumption as a primary or secondary outcome, and the neurophysiological study analyzed the effects neurostimulation has on opioid receptors.

Upon evaluating the study results, 57/139 (41%) participants discontinued opioid use altogether, 37/139 (26.6%) participants decreased opioid use, 8/139 (5.8%) participants increased opioid use, and 37/139 (26.6%) participants reported no change in opioid use. De Caridi et al. quantified “decreased” opioid consumption as a >50% reduction;^[18] the remaining studies did not specify the amount by which participants decreased or increased their opioid doses. Assuming a decrease in opioid use translated to a reduction of 50% from baseline (range 10–90%) and an increase in opioid use meant an increase of 50% (range 10–90%), overall opioid use among the 139 opioid users decreased by an estimated 45.6% ± 13 after neurostimulation. It is noteworthy that the aforementioned results reflect participants who successfully completed a trial. There were 330 participants before each study’s completion. Forty-seven (47) participants withdrew before final analyses due to complications or lack of effectiveness, and 10 participants were lost at follow-up.^[14,16] This translates to 17.3% (57/330) of participants failing to complete a study.

On average, 30% of participants per study lowered their opioid dose from baseline following neurostimulation (range 21–43.5%). Some participants remained on their original opioid dose; this was reported in three studies. The mean rate of participants remaining on the same opioid dose after neurostimulation was 17.9% per study (range 0–38%) and 29.8% per study among the three studies which reported the result. The two studies which did not report any participants continuing the same dose had sample sizes of three and 14.^[15,18] Four studies reported some participants discontinued opioid use wholly after neurostimulation. On average, 46.4% of participants per study discontinued opioid use completely (range 0–71.4%), and 58% of participants per study among the four studies which reported the result discontinued opioid use altogether. McRoberts et al. did not report any participants who stopped taking opioids, and McRoberts et al. had a sample size of 23.^[16] These results are summarized in [Table 2].

[Table 3] highlights details of the six results. Among the six studies, five utilized SCS technology, and one used TMS. Lamusuo et al. used TMS technology to determine effects TMS has on mu-opioid receptor availability; participants in this trial were otherwise young (age range 21–32 years) and healthy.^[19] All five SCS studies reported statistically significant reductions in pain scores following SCS implantation. It is unclear what proportion of participants who reported substantial pain relief also decreased opioid use and to what extent. Various pain syndromes were treated across these studies including lower back pain due to multifidus dysfunction, cancer-related chest wall pain, phantom limb pain, failed back surgery syndrome, knee pain, and arm pain.^[14-18] The median sample size was 23 participants (range 3–156).

Only Gilligan et al. and McRoberts et al. reported adverse events following SCS implantation.^[14,16] Both trials reported infection, device migration, surgical revisions, and device removal due to complications or lack of effectiveness. McRoberts et al. published their study in 2013, and there were 45 adverse events reported with 32 participants (1.4 adverse events per participant).^[16] Gilligan et al. published their study in 2021, and there were 56 adverse events reported with 204 participants (0.28 adverse events per participant).^[14] The rate of adverse events per participant is 400% higher in McRoberts et al. than in Gilligan et al, therefore suggesting either an improvement in technology and/or technique over the past decade.

The median trial duration was 52 weeks (range 2–104 weeks). The mean age among the studies was 62.2 years (range 38– 77 years). About 52.5% of all participants were female. Only McRoberts et al. identified race demographics of their trial population (all Caucasian, n = 23).^[16] Two studies revealed the mean body mass index of their participants, and this was 28.1 kg/m² ± 4 between both studies.^[14,17] In all studies, participants had failed to previously attain adequate pain relief from conventional modalities including, but not limited to medication therapy, intercostals/epidural steroid blocks, surgery, physical therapy, chiropractic medicine, TENS units, acupuncture, and massage. Two trials specified the mean duration of pain participants endured before receiving an implantable neurostimulator: 14 pain years ± 10.6^[14] and 13.9 pain years ± 12.2.^[16]

DISCUSSION

The results demonstrate that SCS has potential to reduce or stop opioid use among patients with chronic pain syndromes. Lamusuo et al. proposed that TMS at the M1/ S1 cortex may invoke a release of endogenous opioids, but this theory was merely a suggestion for future research.^[19] This theory along with the significant findings from the other five study results suggest neurostimulation may decrease or eliminate opioid use in pain patients by providing adequate pain relief and/or activating endorphins.^[14-19] All studies found significant decreases in opioid use and significant reductions in pain scores, but the linearity of pain score versus opioid consumption is uncertain in all but two studies. The two studies in which complete linearity could be found (n = 3 and n = 14) were not reporting linearity as an outcome; rather, regression was found coincidentally through post hoc analysis.^[15,18] Pain score changes and opioid consumption changes were assessed independently among the studies.

Seventy-two percent (72%) of participants from the trial published by Gilligan et al. (n = 156) reported 50% or greater reduction in pain, but only 60% decreased or stopped opioid use. In this case, 12% of participants who reported ≥ 50% pain reduction either continued their opioid dose or increased their opioid dose.^[14] McRoberts et al. (n = 23) reported 65.2% of participants experienced ≥ 30% reduction in pain at 1 year, but only 43.5% lessened their opioid dose.^[16] Barpujari and Erdek only assessed changes in morphine equivalent doses and not pain scores.^[17] It is unclear (1) how many participants improved their pain but did not decrease their opioid dose or (2) how many participants neither alleviated their pain nor reduced their opioid dose. It is worth noting that opioids are only available by prescription in the US. For a patient to decrease or stop their dose, the practitioner and patient would need to come to a mutual agreement, or the prescription would have as-needed instructions on the label, or the participant would need to decrease their dose against prescription instructions. Opioids also have a potential for misuse and addiction, thereby posing as a challenge for dose tapering for some participants, especially those with addictive tendencies.^[1,20]

It is difficult to draw conclusions from these results due to the varying ages and sample sizes of the studies. For instance, sample sizes of 3, 14, and 23 will likely not produce the same statistical power as a sample size of 156. However, two studies – including Gilligan et al. who had the sample size of 156 – were deemed to have an unclear risk of bias, therefore making the significance of their results harder to determine. Statistical significance aside, SCS overall has uncertainty surrounding its effectiveness. Implantable SCS first became available in 1981, and technology has advanced since then.^[21] Older technologies and techniques may not be as effective as more modern ones. Despite this, Ferraro et al. wrote in a 2022 review that safety and effectiveness of SCS remains largely uncertain. Ferraro et al. reviewed 15 randomized controlled clinical trials cumulatively enrolling 908 participants who received SCS for pain. Among these 15 trials analyzed by Ferraro et al., five were open-label – which the authors indicated had “inevitable limitations” – and 10 were blinded. The authors found six of the 10 blinded studies produced “small effects.”^[22]

The results of this review and reviews like that of Ferraro et al. reveal inconsistencies surrounding the safety and effectiveness of SCS. Garcia, Wray, and Kumar also acknowledged the information paucity surrounding SCS, writing “most of the studies that currently exist regarding SCS therapy are either small prospective studies or retrospective studies.”^[23] With a median sample size and duration of 23 and 1 year, respectively, results from this review may lack statistical power to ascertain solid long-term conclusions even though all five studies produced substantial results. In addition, this review sought to evaluate neurostimulation broadly but instead only reviewed SCS. Although all neurostimulators stimulate nerves with electricity,^[3] it is impossible to assume all neurostimulators are equal without further comparison. The neurophysiological study by Lamusuo et al. shows promising potential of TMS in replacing opioids for pain, but more research is needed.^[19] Finally, there were two open-label trials and one study presenting case reports, and these types of studies are not as powerful as randomized controlled trials.

Producing a randomized controlled clinical trial with implantable neurostimulators is burdensome but possible.^[22] Implantable neurostimulators have inherent risks – such as lead failure, lead migration, absent effectiveness, infection, and the need for additional surgeries – as highlighted in [Table 3]. Surgically inserting a sham device may arguably have ethical considerations given the risks of surgery, the risks of complications, and the fact that participants with extensive unresolved pain are proceeding without treatment.^[24-26] Notwithstanding the liabilities of placebo-controlled implantable neurostimulation trials, Duarte et al. found 12 sham-controlled SCS clinical trials between February 2018 and January 2019. The median sample size among the 12 trials was 24 participants (range 10–68), and the median placebo duration was 2 weeks (range 12 h–26 weeks). Duarte et al. emphasized validity and replicability concerns of the 12 trials attributed to small sample sizes and short durations as well as issues with reporting transparency.^[27] Placebo-controlled non-invasive neurostimulator trials (like TMS or TENS units) could circumvent surgical and ethical problems of invasive neurostimulator trials.

Although not evaluated in this review, it is important to briefly highlight the safety and effectiveness of TMS and TENS units for completeness. Goudra et al. published a meta-analysis in 2017 which sought to assess the effectiveness of repetitive TMS from nine clinical trials. Across the nine trials, the mean reduction in pain score was −1.12 from TMS versus −0.28 from sham (P < 0.001). There were no significant adverse events reported. The duration and frequency of TMS was variable.^[28] Yang and Chang further validated claims of effectiveness in a systematic review of 106 trials by finding TMS demonstrates widespread effectiveness with few adverse events, but long-term effectiveness (>3 months) and the extent of pain relief remain unclear.^[29] Vance et al. wrote in a review that the data on TENS units’ effectiveness are conflicting but promising, and, like TMS, there are fewer adverse events reported versus opioids.^[30] The mechanism of action of TENS units is not fully elucidated, but anecdotal reports from patients suggest it can be effective.^[31,32] There have been reports of “TENS tolerance” where overuse of TENS units results in diminished effects.^[33]

Overall, neurostimulation has demonstrated at least some effectiveness for pain management and opioid dose reduction. All studies in this review produced statistically significant results favoring opioid dose reductions and/or pain management. There is a vast array of literature describing the effectiveness of invasive and non-invasive neurostimulators for pain, but many studies lack statistical power necessary to make sound suggestions for future use. This review did not investigate other neurostimulation devices outside of SCS which makes it difficult to produce any conclusions for any devices other than SCS (and minimally TMS). A deeper review utilizing more databases and less stringent inclusion-exclusion criteria may be necessary to identify more research. The dangers of implantable neurostimulators also identify the need to investigate noninvasive neurostimulators in more detail.

CONCLUSION

SCS successfully resulted in decreased opioid use across all the studies evaluated, and this reduction in opioid use can be attributed to better pain management. Despite the positive findings, the studies assessed in this review were mostly small in sample size, and several were either open-label or case reports. Due to the statistical limitations of the studies, it is difficult to draw a certain conclusion on the effectiveness for SCS, TMS, and other neurostimulators for pain. With the current epidemic in opioid overdose deaths, the use of medical devices for pain management may become an option for some patients. Neurostimulators like SCS have promising potential for the future of pain management, but more randomized, controlled research is needed to evaluate the long-term safety and effectiveness profiles of these devices. In the future, hopefully, there will be more succinct approaches for addressing which patients would be the most suitable candidates for specific types of neurostimulation.