Neuromodulation Therapy for Chemotherapy-Induced Peripheral Neuropathy: A Systematic Review

Chemotherapy-induced peripheral neuropathy (CIPN) is a debilitating and painful condition in patients who have received chemotherapy. The role of neuromodulation therapy in treating pain and improving neurological function in CIPN remains unclear and warrants evidence appraisal. In compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, we performed a systematic review to assess change in pain intensity and neurological function after implementation of any neuromodulation intervention for CIPN. Neuromodulation interventions consisted of dorsal column spinal cord stimulation (SCS), dorsal root ganglion stimulation (DRG-S), or peripheral nerve stimulation (PNS). In total, 15 studies utilized SCS (16 participants), 7 studies utilized DRG-S (7 participants), and 1 study utilized PNS (50 participants). Per the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) criteria, there was very low-quality GRADE evidence supporting that dorsal column SCS, DRG-S, and PNS are associated with a reduction in pain severity from CIPN. Results on changes in neurological function remained equivocal due to mixed study findings on thermal sensory thresholds and touch sensation or discrimination. Future prospective, well-powered, and comparative studies assessing neuromodulation for CIPN are warranted.


Introduction
Chemotherapy-induced peripheral neuropathy (CIPN) is a common and debilitating condition in cancer patients who have received chemotherapy. CIPN is predominantly a sensory neuropathy localized in a stocking-glove distribution. Sensory symptoms may range from allodynia to hypoesthesia and may be accompanied by motor and autonomic deficits [1], resulting in increased fall risk and decreased quality of life [2,3]. The severity of CIPN is dependent on the type of chemotherapy agent and its associated neurotoxicity profile, cumulative dose, and duration of exposure [4][5][6]. Symptoms generally improve with discontinuation of the chemotherapy agent over time, although complete recovery to baseline is not always attained. It is estimated that 50-90% of patients who receive chemotherapy go on to develop CIPN in the acute setting, and 30-40% of patients continue to suffer from CIPN chronically [1,4,6].
Current pharmacologic therapy for CIPN includes a multimodal strategy consisting of the following regimen: acetaminophen, non-steroidal anti-inflammatory drugs (NSAIDs), topical agents (e.g., lidocaine and capsaicin), anti-convulsants (e.g., gabapentinoid and carbamazepine), anti-depressants (e.g., selective norepinephrine reuptake inhibitors and tricyclic anti-depressants), opioids, and cannabinoids [1,7]. However, their effectiveness may be limited. Low therapeutic index and adverse effect profiles from some medications may present additional challenges. Neuromodulation therapy has emerged as a viable treatment option for many refractory cases of chronic pain, including CIPN [8].
This review focuses on dorsal column spinal cord stimulation (SCS), dorsal root ganglion stimulation (DRG-S), and peripheral nerve stimulation (PNS). Dorsal column SCS involves the placement of leads in the epidural space overlying the dorsal column of the spinal cord [8]. Various stimulation waveforms, including tonic stimulation, high-frequency (10-kHz) stimulation, burst stimulation, differential target multiplexed stimulation, and closed-loop stimulation, can be delivered via dorsal column SCS [9][10][11]. DRG-S involves lead placement in the epidural space overlying the dorsal root ganglion [12,13]. PNS involves lead placement in close proximity to a target nerve that innervates the location of the painful area [14,15].
Evidence from landmark trials has highlighted the efficacy of SCS, DRG-S, and PNS in the treatment of neuropathic pain conditions, revealing superior pain relief, physical functionality, mental health, and patient satisfaction compared to conventional medical management alone [10,[16][17][18][19]. Studies have also highlighted that animal models with CIPN may benefit from stimulation. For instance, a study on rats with paclitaxel-induced mechanical and cold hypersensitivity revealed that SCS therapy significantly inhibited the development of mechanical and cold hypersensitivity compared to rats receiving paclitaxel alone or sham SCS [20]. To date, there is Food and Drug Administration (FDA) approval supporting SCS therapy for painful diabetic neuropathy (PDN) [21,22], failed back surgery syndrome [23], non-surgical refractory back pain [24], and complex regional pain syndrome (CRPS) [25]. DRG-S is approved for use in CRPS, while PNS is approved for the treatment of chronic pain, post-surgical pain, or post-traumatic pain located in the back, extremities, or head/neck.
In this systematic review, we appraise the current literature on the use of SCS, DRG-S, and PNS for the treatment of pain related to CIPN. The primary objective is to assess changes in pain intensity. The secondary objective is to assess changes in neurological function.

Compliance with Ethics Guidelines
This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors. The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO CRD42022338500).

Search Strategy
This study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [26]. As such, a comprehensive search of several databases from each database's inception to 16 May 2022 in the English language was conducted, along with the creation of a hand-searched reference list of identified publications. The databases included Ovid MEDLINE(R), Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, and Scopus. The search strategy was designed and conducted by an experienced librarian (L.J.P.) with input from the study's principal investigator (R.S.D.). Controlled vocabulary supplemented with keywords was used to search spinal cord stimulation and peripheral nerve stimulation for CIPN in humans. The actual strategy listing all search terms used and how they are combined is available in the Supplementary Tables S1-S3.

Study Selection
Included studies abided by the following criteria: any study design that involved humans undergoing neuromodulation (SCS, DRG-S, PNS) for the treatment of CIPN. Exclusion criteria consisted of the following: review or meta-analysis articles. Of note, we included abstracts in this review. Two authors (M.Y.J. and M.M.) independently selected abstracts along with full-text articles from the above-listed databases, while a third author (R.S.D.) resolved any discrepancies.

Data Extraction
The following data on the characteristics of each study were extracted: study design, funding source, mean age and number of participants, type of intervention and lead placement location, and stimulation settings. Relevant data on the primary and secondary outcomes of interest were also extracted. The primary outcome of interest was participantreported changes in pain intensity related to their CIPN after neuromodulation therapy. Secondary outcomes of interest included participant-reported changes in neurological function (e.g., change in motor or sensory deficit). For each included study, two authors (M.Y.J. and M.M.) independently extracted all relevant data, with a third author (R.S.D.) arbitrating any disputes.

Assessment of Risk of Bias
Two independent reviewers (Y.F.H. and R.S.D.) assessed the risk of bias for the included studies using the Newcastle-Ottawa Quality Assessment Scale for case-control and cohort studies. There was no randomized controlled trial captured in this review. For observational studies, bias was assessed based on the following domains: Selection (representativeness of the exposed cohort, selection of the non-exposed cohort, ascertainment of exposure, demonstration that outcome of interest was not present at the start); Comparability (comparability of cohorts on the basis of the design or analysis); and Exposure/Outcome (assessment of outcome, follow-up long enough for outcomes to occur, adequacy of follow-up of cohorts). A maximum of one star for each numbered item within the Selection domain and Exposure/Outcome domain can be awarded per study. A maximum of two stars can be awarded for the Comparability domain. A higher number of stars indicates a lower risk of bias for each respective domain.

Quality of Evidence
GRADEpro software (McMaster University and Evidence Prime, Inc., Kraków, Poland, 2022; http://gradepro.org, accessed on 21 June 2022) was used to assess the quality of evidence for each outcome per GRADE (Grading of Recommendations, Assessment, Development, and Evaluations) criteria. Randomized controlled trials are categorized as high-level evidence. This can be downgraded based on the risk of bias, inconsistency, indirectness, imprecision, and publication bias.

Evidence Supporting Neuromodulation Therapy for Pain Relief
SCS therapy for CIPN-related pain was evaluated across 15 case studies [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42] and reported in Table 1. A total of 3 recent case reports reported improvement in pain intensity by 80% during the trial phase, which was maintained after permanent SCS implantation, with visual analog scale (VAS) scores ranging between 0/10 and 2/10 at 2 years follow-up [28,35,36]. Cata et al. [41] reported SCS implantation in two patients with CIPN. Of these, 1 patient reported a VAS score of 4.5/10 at baseline, which improved to 0/10 after SCS therapy, and substantial pain relief was maintained at 4 months follow-up. The second patient had initial improvement of pain from 4.6/10 at baseline to 0/10 during the trial period and after implantation but reported a slight increase in VAS score to 3.6/10 at the 2-week follow-up. Regardless, the patient reports satisfaction with SCS therapy and improvement in gait and leg flexibility [41]. Panchal et al. reported the use of a wireless SCS device that provided more than 90% improvement in pain intensity from CIPN [33].

Evidence Supporting Neuromodulation Therapy for Pain Relief
SCS therapy for CIPN-related pain was evaluated across 15 case studies [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42] and reported in Table 1. A total of 3 recent case reports reported improvement in pain intensity by 80% during the trial phase, which was maintained after permanent SCS implantation, with visual analog scale (VAS) scores ranging between 0/10 and 2/10 at 2 years followup [28,35,36]. Cata et al. [41] reported SCS implantation in two patients with CIPN. Of these, 1 patient reported a VAS score of 4.5/10 at baseline, which improved to 0/10 after SCS therapy, and substantial pain relief was maintained at 4 months follow-up. The second patient had initial improvement of pain from 4.6/10 at baseline to 0/10 during the trial period and after implantation but reported a slight increase in VAS score to 3.6/10 at the 2-week follow-up. Regardless, the patient reports satisfaction with SCS therapy and improvement in gait and leg flexibility [41]. Panchal et al. reported the use of a wireless SCS device that provided more than 90% improvement in pain intensity from CIPN [33].   DRG-S for CIPN-related pain was evaluated in seven case reports [43][44][45][46][47][48][49]. Of these, 3 recent case studies reported 100% improvement in pain with VAS scores ranging between 0/10 to 1/10, and this substantial relief was maintained between 5 months to 3 years post-implant [44,48,49]. Finney et al. and Yelle et al. reported 50% and 60% improvement in pain intensity, respectively, after DRG-S [46,47]. However, notable complications have been reported with DRG-S use. For example, Sindhi et al. [43] described a male patient who suffered sharp stabbing 10/10 pain intensity at the glans of the penis after penile mass excision and chemoradiotherapy treatment. The patient reported >60% improvement in pain after DRG-S. However, after 5 months, there was a severe increase in pain which was not responding to reprogramming, which was attributed to slight dorsal migration of the DRG leads. Similarly, Rao et al. [45] reported >75% improvement in pain with the DRG-S trial, although immediately after permanent implantation, the patient suffered sudden lower extremity numbness. This was attributed to the right DRG lead traversing the dura anterior to the spinal cord and compressing the right S1 nerve root before exiting the thecal sac. This lead was removed immediately, and the patient underwent re-implantation one month later with adequate coverage of the right posterior calf, although with persistent right foot pain.
The impact of PNS in alleviating neuropathic pain was evaluated in 1 retrospective study of 50 participants. Sacco et al. [27] performed percutaneous auricular neurostimulation (PANS) therapy which stimulated peripheral nerves in the ear. All 18 participants who were available for quantitative assessment of pain intensity reported improvement in pain VAS scores from a mean of 8.11 to 3.77. Of the remaining 32 participants available for qualitative pain assessment, 59% reported marked improvement, 12.5% had minimal improvement, and 29% reported no improvement.

Evidence Supporting Neuromodulation Therapy for Neuropathy and Neurological Deficits
The role of SCS in improving neuropathy and neurological deficits from CIPN was reported in eight case reports [28,33,35,36,[39][40][41][42] consisting of nine patients (Table 1). Mixed results were reported regarding temperature sensation. Kamdar et al. [35] reported a reduction in cold hypersensitivity after SCS therapy. However, Cata et al. [41] reported no change in thermal thresholds after SCS therapy for one of their evaluated patients. Outcomes for touch sensation also revealed equivocal and mixed results. Kamdar et al. [35] reported an unchanged sensory exam post-implant, while Cata et al. [41] reported improved touch detection in both of their patients after SCS therapy. Furthermore, Cata et al. reported that sharpness detection was only improved in one of the two patients [41]. Six studies consisting of seven participants reported improved ability to participate in daily activities and more mobility with reasons attributed to improved gait [35,41], more stability [35], better flexibility of legs [41], ambulation with less pain [33], decreased dependence on assistance from others [39], and improvement of hand function [36,42].
Change in neurological function following DRG-S was evaluated in six studies [43,46,48] consisting of six patients. Two studies reported results on touch sensation, with one study [48] citing no change in sensation to light touch or pinprick sensation (n = 1), while another study [45] reporting worsened lower extremity numbness after DRG-S. Four studies consisting of four participants reported improved ability to participate in daily activities and more mobility, with reasons being attributed to improved ability to wear shoes and exercise [44,48] and reduced pain while working [43] or walking [43,46].
The impact of PNS on neurological deficits was evaluated in 1 retrospective study of 50 participants [27]. In this study utilizing PANS therapy, participants reported notable qualitative improvements in numbness, gait, and balance. One patient reported adverse outcomes due to the intolerance of intermittent pulsing.

Bias Assessment
The risk of bias was assessed for one observational study [27]. Since the remaining studies were case reports/series, they were considered high risk for bias and not amenable to appraisal with bias assessment tools. Using the Newcastle-Ottawa Quality Assessment Scale, the study by Sacco et al. [27] received two stars for the Selection domain (maximum possible four stars) and two stars for the Exposure/Outcome domain (maximum possible three stars). There were no stars assigned to the Comparability domain, indicating a high risk of bias for that respective domain.

Quality of Evidence
An evidence profile table and summary of findings table with quality of evidence for each outcome are summarized in Tables 2 and 3. Overall, there was very low quality of evidence that neuromodulation interventions (SCS, DRG-S, PNS) were effective in reducing pain intensity from CIPN. Results were equivocal regarding the role of neuromodulation interventions in improving neurological function, with some reporting improvements, some reporting no change, and some reporting deteriorating neurological deficits. For both outcomes, there was a high risk of bias, inconsistency, indirectness, impression, and potential publication bias in the included studies. A total of 22 studies were case reports, and 12 studies were non-peer-reviewed abstracts presented at conferences. a Included studies consisted of 22 case reports and 1 retrospective review. There was a high risk of bias in patient selection, comparability, and assessment of outcomes. b High heterogeneity was present in between and within the studies. c Some studies used dorsal column spinal cord stimulator, and some studies used dorsal root ganglion stimulator. Variations in the targeted location of chemotherapy-induced peripheral neuropathy. d Success rates widely varied between the studies. e 12/23 studies were case reports presented in conferences. f Studies reported on different neurological function outcomes. There was no consistency in the functional testing scale used. "⊕" indicates very low certainty, "⊕⊕" indicates low certainty, "⊕⊕⊕" indicates moderate certainty, and "⊕⊕⊕⊕" indicates high certainty. Only 10 studies assessed neurological function. Of these, 6 of 10 studies reported improved gait after neuromodulation. Two studies reported improved hand motor function. Four studies reported improved sensory thresholds. Only one study reported worsening lower extremity numbness. a Included studies consisted of 22 case reports and 1 retrospective review. There was a high risk of bias in patient selection, comparability, and assessment of outcomes. b High heterogeneity was present in between and within the studies. c Some studies used dorsal column SCS, some studies used DRG-S, and some studies used PNS. There was variation in the targeted location of CIPN. d Success rates widely varied between the studies. e 12/23 studies were case reports presented in conferences. f Studies reported on different neurological function outcomes. There was no consistency in the functional testing scale used. GRADE Working Group grades of evidence -Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect. "⊕" indicates very low certainty, "⊕⊕" indicates low certainty, "⊕⊕⊕" indicates moderate certainty, and "⊕⊕⊕⊕" indicates high certainty.

Summary of Evidence
This systematic review appraises the evidence of neuromodulation therapy for pain and neuropathy in CIPN. Overall, this review highlights that dorsal column SCS, DRG-S, and PNS are associated with a moderate to high reduction in pain severity from CIPN. Pain relief from dorsal column SCS ranged from 50-100% between 3 months and 2 years, whereas pain relief from DRG-S ranged from 50-100% between 1 month and 3 years. These findings are concordant with the literature highlighting the efficacy of neuromodulation therapy for a variety of neuropathic pain conditions, including failed back surgery syndrome [50] and PDN [51]. Furthermore, although the mechanism of nerve injury between CIPN and PDN differs, they both present similarly with distal axonopathy and symmetrical lengthdependent sensory neuropathy in a stocking-glove distribution [52]. Thus, it would seem rational that the response to neuromodulation would be similar between CIPN and PDN.
Efficacy of PNS therapy was highlighted in 50 participants, with reports of >50% pain relief among those who reported quantitative pain assessments. However, in this study [27], it is unclear how auricular nerve stimulation may alleviate CIPN which typically manifests in the distal extremities in a stocking-glove distribution. PNS implantation at the site of CIPN (e.g., lower extremity peripheral nerves or upper extremity peripheral nerves) was not described. Due to the very low-quality GRADE evidence supporting decreased pain intensity from neuromodulation interventions, additional prospective and powered trials are warranted. There are ongoing prospective clinical trials assessing the efficacy of dorsal column SCS for CIPN (ClinicalTrials.gov Identifier: NCT05411523).
In terms of change in neurological function and neuropathy assessment metrics, results from studies remained equivocal. Although most studies reported improvement in numbness, gait, balance, and ambulation, results were mixed with regard to thermal sensory thresholds, touch sensation, and touch discrimination. Similar to the primary outcome of pain intensity, there was very low-quality GRADE evidence on the outcome of neurological function.
Given the stocking-glove distribution of CIPN which frequently involves multiple nerve distributions, the authors recommend offering dorsal column SCS as the neuromodulation intervention of choice due to its ability to target broader areas of pain. DRG-S also has the capability of providing alleviation across multiple nerve distributions, partly due to its mechanism of cross-talk between dorsal root ganglions at different levels [13]. PNS may be considered if the patient's pain symptoms are primarily located in one to two nerve distributions. The decision to pursue a specific neuromodulation intervention should also weigh its respective adverse effect profile. Overall, studies have highlighted that SCS [53][54][55], DRG-S [56], and PNS [15] are safe interventions overall. Special considerations unique to patients with CIPN may include an immunocompromised status due to ongoing malignancy and/or chemotherapy, which is associated with an increased risk for infection. Coagulopathy due to the underlying malignancy or due to side effects from chemotherapy may increase the risk for hematoma, which is particularly worrisome in the setting of neuraxial hematoma from SCS or DRG-S placement.
Our review suggests that clinicians caring for patients with CIPN should consider neuromodulation options in their treatment algorithm, particularly due to the paucity of evidence supporting conservative measures and conventional pharmacotherapy. Conventional treatment options for CIPN are highlighted by the American Society of Clinical Oncology (ASCO) guidelines [57]. The ASCO guidelines confirmed that no pharmacotherapeutic agents are recommended for CIPN prevention. Certain strategies such as dose delaying, dose reduction, substitution, and stoppage of chemotherapy may be considered in patients who develop intolerable symptoms from CIPN. In terms of treatment after the occurrence of CIPN, duloxetine is the only current neuropathic agent with intermediatelevel evidence to support use in CIPN with moderate-level efficacy. Importantly, the ASCO guidelines recommended against the use of other commonly used neuropathic agents such as amitriptyline, gabapentin, pregabalin, and venlafaxine due to no evidence of efficacy or low-level efficacy. Importantly, many neuropathic medications may have a high prevalence of side effects such as somnolence, dizziness, difficulty in concentration, and other drugspecific side effects [7]. Further evidence is warranted for other conservative treatment options such as scrambler therapy, acupuncture, and exercise [57].
In summary, prior to offering neuromodulation options for the treatment of pain in CIPN, the authors recommend offering first-line conventional therapy such as multimodal pharmacologic therapy and physical therapy. Although pharmacologic therapy may consist of non-steroidal anti-inflammatory drugs, gabapentinoids, and anti-depressants, the literature supports that only duloxetine should be offered for treatment due to intermediate-level evidence for moderate efficacy. Of note, CIPN may also improve over time after completion of chemotherapy. In the setting of conventional treatment failure, pain specialists may offer neuromodulation options, including SCS, DRG-S, and PNS.

Proposed Mechanism of Action
Several mechanisms of action have been proposed to explain the improvement of pain from neuromodulation. Although a commonly proposed mechanism is the gate control theory, new mechanisms have been proposed associated with unique waveform paradigms, including high-frequency, burst, and other waveforms [58]. In high-frequency SCS, the gate control mechanism may be activated without stimulating pathways for paresthesia [59]. In burst stimulation, both the lateral and medial aspects of the spinothalamic tract are activated [60]. Besides the gate control theory, analgesia from SCS may also be explained by its modulation of neurotransmitters (cholinergic, serotonergic, and opioidergic pathways), depression of wide dynamic range neurons, and activation of supraspinal levels from orthodromic dorsal column action potentials [61]. In PNS, modulation of the central nervous system involves the dorsal lateral prefrontal cortex, parahippocampal areas, and the anterior cingulate cortex [14,62].
The mechanisms explaining the improvement of neurological deficits from neuromodulation therapy remain unknown. A prior review by Khunda et al. proposed that the neurological improvement of neurogenic urinary and bowel disorders through neuromodulation may be due to the stimulation of afferent pathways from the genital area [63]. Recently, a study revealed that activity-dependent SCS could rapidly restore truncal and lower extremity motor function in patients with complete paralysis [64]. Future studies are warranted to elucidate the mechanisms of improvement in neurological function after neuromodulation therapy.

Limitations
Clinical heterogeneity was significant across studies that evaluated different neuromodulation interventions for CIPN. Even in studies that focused on one neuromodulation intervention (e.g., DRG-S), there was variable placement of lead location or variable waveform parameters. Furthermore, details regarding waveform settings were not provided in several studies. Due to this level of heterogeneity, outcome measurements were unable to be pooled. Only one observational study was captured, and the remaining evidence was limited to case series/reports. Although this review evaluated all neuromodulation interventions, the majority of studies focused on dorsal column SCS and DRG-S. Another reason for low-quality evidence was the absence of placebo-controlled randomized controlled trials. A prior review of placebo-controlled trials highlights the improvement in neuropathic pain syndromes from neuromodulation interventions [50]. However, challenges to performing placebo-controlled studies in neuromodulation are prominent and include difficulty in developing placebo or sham arms, issues with inadequate blinding, placebo and nocebo effects, and, most importantly, ethical concerns centered over the concept of equipoise [65]. Thus, with these significant limitations and very low-quality GRADE evidence, it is imperative that the findings from this systematic review be interpreted with caution. Clinical implications from this systematic review, whether in support of or against the efficacy of neuromodulation for improving pain and neurological function from CIPN, cannot be concretely concluded.

Future Directions
Future studies should conduct adequately powered and prospective studies assessing neuromodulation for CIPN. Although efficacy may be dependent on the type of waveform, as demonstrated in prior studies [66,67], included studies did not compare the efficacy between different waveforms for CIPN. Thus, comparative studies assessing different waveforms are warranted. Finally, primary care clinicians, oncologists, and hematologists rarely consider neuromodulation for the treatment of cancer-related pain. Therefore, dissemination of information and education for both physicians and patients via conference proceedings, social media coverage [68,69], and other avenues of information dissemination are important [70].

Conclusions
Our systematic review performed an evidence synthesis on neuromodulation interventions for the treatment of CIPN. The current evidence suggests that neuromodulation interventions, including SCS, DRG-S, and PNS, may lead to clinically meaningful pain relief in patients with CIPN. Improvement in gait, motor function, and sensory function was also highlighted in several studies. The GRADE certainty of evidence is limited due to a lack of prospective comparative studies, clinical and methodological heterogeneity, and low sample size.  Institutional Review Board Statement: Study did not require ethical approval as it was a systematic review analyzing previously published data.

Informed Consent Statement: Not applicable.
Data Availability Statement: Data is available upon request to the corresponding author.