Non-Pharmacological Self-Management Strategies for Chemotherapy-Induced Peripheral Neuropathy in People with Advanced Cancer: A Systematic Review and Meta-Analysis

Non-pharmacological self-management interventions for chemotherapy-induced peripheral neurotherapy (CIPN) are of clinical interest; however, no systematic review has synthesized the evidence for their use in people with advanced cancer. Five databases were searched from inception to February 2022 for randomized controlled trials assessing the effect of non-pharmacological self-management interventions in people with advanced cancer on the incidence and severity of CIPN symptoms and related outcomes compared to any control condition. Data were pooled with meta-analysis. Quality of evidence was appraised using the Revised Cochrane Risk of Bias Tool for Randomized Trials (RoB2), with data synthesized narratively. Grading of Recommendations, Assessment, Development and Evaluations (GRADE) was applied to assess the certainty of the evidence. Thirteen studies were included, which had a high (69%) or unclear (31%) risk of bias. Greatest confidence was found for physical exercise decreasing CIPN severity (SMD: −0.89, 95% CI: −1.37 to −0.41; p = 0.0003; I2 = 0%; n = 2 studies, n = 76 participants; GRADE level: moderate) and increasing physical function (SMD: 0.51, 95% CI: 0.02 to 1.00; p = 0.04; I2 = 42%; n = 3 studies, n = 120; GRADE level: moderate). One study per intervention provided preliminary evidence for the positive effects of glutamine supplementation, an Omega-3 PUFA-enriched drink, and education for symptom self-management via a mobile phone game on CIPN symptoms and related outcomes (GRADE: very low). No serious adverse events were reported. The strongest evidence with the most certainty was found for physical exercise as a safe and viable adjuvant to chemotherapy treatment for the prevention and management of CIPN and related physical function in people with advanced cancer. However, the confidence in the evidence to inform conclusions was mostly very low to moderate. Future well-powered and appropriately designed interventions for clinical trials using validated outcome measures and clearly defined populations and strategies are warranted.


Introduction
Chemotherapy is a commonly used treatment in the advanced stages of cancer as a cancer control strategy, where improvements to symptom management, quality of life, and survival are prioritized [1,2]. However, chemotherapy is often associated with debilitating side effects such as chemotherapy-induced peripheral neuropathy (CIPN), which can lead cost savings in people with cancer [17][18][19]. Self-management strategies can be initiated independently from an internal decision (e.g., massage) or with support from a health professional (e.g., an exercise program) but include only those interventions that the patient is responsible for administering (i.e., not acupuncture) [20,21]. To date, no systematic reviews or clinical guidelines have assessed non-pharmacological self-management strategies specifically for advanced cancer patients, where chemotherapy use and CIPN is common, co-morbidities are high, and daily functioning and overall quality of life are often more valued than enhanced treatment outcomes or survival [22].
Therefore, in people with advanced cancer, this systematic review and meta-analysis of randomized controlled trials sought to determine the effect of non-pharmacological self-management interventions on the incidence and severity of CIPN symptoms and CIPN-related outcomes (quality of life, physical function, sleep, fatigue, gastrointestinal symptoms, nutrition status, psychosocial and financial outcomes, and adverse effects) compared to any control condition.

Materials and Methods
This systematic review and meta-analysis is registered with the International Prospective Register of Systematic Reviews (PROSPERO ID: 308341) and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement [23].
MEDLINE (Ovid), Embase, Web of Science, CINAHL (EBSCOhost), and Cochrane CENTRAL databases were searched from inception to 7 February 2022 using keywords and controlled vocabulary terms (Supplementary Table S1) based on the following: ("nonpharmacological intervention" AND "chemotherapy-induced peripheral neuropathy" AND ("advanced cancer" OR "metastatic cancer") AND "randomi?ed controlled trial"). To identify relevant articles not found in the other databases, the first 200 records as sorted by relevance ranking were taken from Google Scholar on 7 February 2022 [24]. A snowballing technique was also used, whereby Google Scholar, database search updates, and reference lists of included studies and relevant literature were assessed to find additional studies not located in the original search strategy up until 5 April 2022.
Automated de-duplication of articles and manual text-mining [25] was conducted by one investigator (MC) using Endnote software (EndNote 20, Clarivate, Version 20.2, London, UK, 2021). During text-mining, irrelevant terms were searched in titles and abstracts to identify studies for exclusion such as 'systematic review', 'protocol', 'mice', 'conference', and 'medication' [25]. Screening of remaining titles, abstracts and full texts was then completed by two investigators independently (MC and (PMY or AS or OAA)) using Covidence software (Covidence Systematic Review Software, Veritas Health Innovation, Version, Melbourne, Australia, 2021). Screening conflicts were resolved by discussion among the investigators. Studies examining the effect of any non-pharmacological selfmanagement intervention on incidence or severity of CIPN symptoms in people with advanced cancer were included. Advanced cancer was defined as any hematological cancer or any solid tumor with locally or systemically advanced disease (The Union for International Cancer Control Tumor, Nodes, and Metastases Classification of Malignant Tumors (UICC TNM) Stage of III-IV or equivalent) [26]. Self-management strategies were defined as interventions that were administered by the person with cancer or their caregivers, including those initiated with support from a health professional as well as initiated independently [20,21]. Eligibility criteria is presented in Table 1. The primary outcomes were the incidence and severity of CIPN symptoms as measured by any tool at any timepoint. Secondary outcomes were chosen based on existing literature on the factors related to CIPN, including quality of life, physical function, sleep, fatigue, gastrointestinal symptoms, nutrition status, psychosocial and financial outcomes, and adverse effects measured using any tool at any timepoint. Data pertaining to study, participant, and intervention characteristics, as well as outcomes of interest, were extracted independently by one investigator (MC) and checked for accuracy by a second investigator (OAA). Where outcome data were missing or incompletely reported, investigators contacted the authors via email. Individual study quality assessment was conducted independently by two investigators (MC and OAA) using the Revised Cochrane Risk of Bias Tool for Randomized Trials (RoB2) [27]. The RoB2 comprises five domains with an overall risk of bias judgment calculated as low risk of bias, some concerns, or high risk of bias [27]. Disagreements in data extraction and quality assessment were managed by discussion among investigators.
Certainty in the body of evidence was determined using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach [28] via GRADEpro GDT software (GRADEpro Guideline Development Tool, McMaster University and Evidence Prime, 2021). Four levels of certainty for the estimated effect of each outcome were possible: very low (very little confidence in estimated effect), low (limited confidence), moderate (moderately confident), and high (very confident) [28].
Where two or more studies reported the same intervention and sufficient incidence or mean and variance data for the same outcome, data were pooled by meta-analysis using Review Manager (RevMan) software [RevMan Software, The Cochrane Collaboration, Version 5.4.1, The Cochrane Collaboration, Oxford, UK, 2020]). Using the standard random effects method, categorical outcomes were reported as odds ratios (OR) using the Mantel-Haenszel test, and continuous variables were pooled using the inverse variance test and reported as mean differences (MD), where the same tool and scale were used, or standardized mean differences (SMD), where different measurement tools or scales for the same outcome were used. Interpretation of effects sizes for SMDs was: 0.2 represents a small effect, 0.5 a moderate effect, and 0.8 a large effect [29]. Heterogeneity was evaluated with the I 2 statistic, whereby >50% represented substantial heterogeneity [30]. The cut-off for statistical significance was considered at a p-value of <0.05. Results were synthesized in tabular and narrative format.

Search Results and Study Quality
Thirteen studies [31][32][33][34][35][36][37][38][39][40][41][42][43] published between 2014 and 2020 were included ( Figure 1). As indicated in Figure 2, 62% of studies had high risk of bias [32,33,[35][36][37]39,40,42] and the remaining 38% had unclear risk of bias [31,34,38,41,43]. One main reason for bias was the lack of outcome assessor blinding and the use of subjective outcome measurement tools that could have influenced outcome assessment due to the knowledge of the intervention received. However, this is an acknowledged limitation in most exercise, diet, and lifestyle research [44]. Other common sources of potential bias were the lack of information regarding allocation concealment procedures, failure to consider missing data with methods such as intention-to-treat and imputation, selective outcome reporting, and inadequate reporting of whether the study was conducted in accordance with a retrospective protocol. Publication bias was unable to be assessed due to the small number of studies in each meta-analysis. GRADE level of evidence for all outcomes was very low to low.

Study and Population Characteristics
Intervention

Effect on CIPN Symptoms and Related Outcomes
Six studies (46%) [31,32,34,[39][40][41] were analyzed in 18 meta-analyses. Table 3 summarizes meta-analysis results, and forest plots for non-significant findings are presented in Supplementary Figures S1-S3. Justifications for GRADE ratings are given in Supplementary  Table S3. Table 3. Results from meta-analyses that were conducted to evaluate the effect of nonpharmacological self-management interventions on chemotherapy-induced peripheral neuropathy symptoms and related outcomes in people with advanced cancer.

Outcome
Pooled  Incidence of CIPN was significantly lower with a strength, endurance, and sensorimotor training program in comparison to standard care (n = 1 study [33]; n = 61 participants with lymphoma; GRADE level: very low). Severity of CIPN was significantly less with strength and endurance training with or without sensorimotor training when compared to standard care or written exercise guidelines via meta-analysis (SMD: −0.89, 95% CI: −1.37 to −0.41; p = 0.0003; I 2 = 0%; n = 2 studies [31,34], n = 76 participants with lung or colorectal cancer; GRADE level: moderate; Figure 3). Physical function was significantly higher with exercise (strength and endurance training with or without sensorimotor training, or walking) when compared to standard care or written exercise guidelines via meta-analysis (SMD: 0.51, 95% CI: 0.02 to 1.00; p = 0.04; I 2 = 42%; n = 3 studies [31,32,34], n = 120 participants with lung or gastrointestinal cancer; GRADE level: moderate; Figure 3). Quality of life [31] and emotional wellbeing [33] were significantly higher with exercise in one of three studies (33%), but no significant effect was found with meta-analysis (n = 76 participants with lung or colorectal cancer; GRADE level: very low). Significant improvements in pain [31], sleep [33], fatigue [33], diarrhea [33], and financial problems [33] were found with exercise in one of two studies (50%), but data were unable to be pooled by meta-analysis (n = 107 participants with lymphoma or lung cancer; GRADE level: very low). One study found improvements in nutrition status and lean body mass with walking (n = 44 participants with gastrointestinal cancer; GRADE level: very low). No serious adverse events were reported with exercise interventions [32][33][34].

Nutrition Supplements
Incidence of CIPN was significantly lower with glutamine supplementation compared to standard care (n = 1 study [38]; n = 86 participants with colorectal cancer; GRADE level: very low), whereas participants prescribed curcumin had a significantly higher incidence of CIPN in comparison to standard care (n = 1 study [36]; n = 27 participants with colorectal cancer; GRADE level: very low). CIPN severity was significantly lower with consumption of an Omega-3 PUFA-enriched nutrition drink on an isocaloric diet compared to an isocaloric diet alone (n = 1 study [37]; n = 112 participants with lung cancer; GRADE level: very low).
Quality of life and incidence of chemotherapy dose reduction was significantly improved with glutamine consumption compared to standard care (n = 1 study [38]; n = 86 participants with colorectal cancer; GRADE level: very low). Fatigue severity, appetite loss, and body weight maintenance were significantly improved with the consumption of an Omega-3 PUFA-enriched drink on an isocaloric diet compared to an isocaloric diet alone (n = 1 study [37]; n = 112 participants with lung cancer; GRADE level: very low). Curcumin supplementation was associated with a significantly higher cancer treatment response rate and length of survival but also led to a significantly increased incidence of diarrhea (n = 1 study [36]; n = 27 participants with colorectal cancer; GRADE level: very low). No serious adverse events were reported with nutrition supplements [36][37][38].

Japanese Herbal Medicine
The incidence of CIPN was significantly reduced with the consumption of Japanese herbal medicine in two studies [39,40], while the remaining study [41] found a significantly increased incidence with the intervention. However, when meta-analyzed, Japanese herbal medicine had no significant association with the likelihood of CIPN of Grade 1, 2, 3 or ≥2 (n = 3 studies [39][40][41]; n = 283 participants with colorectal cancer; GRADE level: very low). Sensitivity analysis according to the type of herbal supplement (Goshajinkigan vs. ninjin'yoeito) did not result in significant findings.
Relative dose intensity of oxaliplatin was significantly higher with consumption of Japanese herbal medicine in two [39,41] of three studies [39][40][41]; however, there was no significant effect with meta-analysis (n = 2 studies [39,41]; n = 238 participants with colorectal cancer; GRADE level: very low). No serious adverse events were reported with the consumption of Japanese herbal medicine [39][40][41].

Technology-Facilitated Education for Symptom Self-Management
The incidence of CIPN was significantly lower in participants who received education for symptom self-management via a mobile phone game in comparison to a symptom management booklet (n = 1 study [43]; n = 76 participants with breast cancer; GRADE level: very low). Time to CIPN symptom response was lower in participants who received education via automated telephone voice technology incorporating symptom monitoring compared to cognitive-behavioral nurse-administered symptom management via telephone (35 vs. >55 days); however, statistical significance was not tested (n = 1 study [42]; n = 47 participants with breast cancer; GRADE level: very low).
Quality of life and medication compliance was significantly higher in participants who received education for symptom self-management via a mobile phone game in comparison to a symptom management booklet; however, results for physical function, nausea, and psychological health favored the control (n = 1 study [43]; n = 76 participants with breast cancer; GRADE level: very low). No serious adverse events were reported [43].

Discussion
This systematic review is the first to synthesize evidence on the non-pharmacological self-management interventions for CIPN and related outcomes in people with advanced cancer. The strongest evidence was found for physical exercise with medium to large beneficial effects on CIPN severity and physical function in adults, with moderate certainty in the estimated effect size [31,34]. Some evidence with very low certainty in the effect, and supported by one study only, was found in adults for the positive effects of glutamine supplementation on CIPN incidence, quality of life, and incidence of chemotherapy dose reduction [38]; an Omega-3 PUFA-enriched nutrition drink for CIPN severity, fatigue, appetite loss, and body weight maintenance [37]; and a symptom self-management mobile phone game for CIPN incidence, quality of life, and medication compliance [43]. Findings are inconclusive as to whether CIPN symptoms are improved by glutamatic acid supplementation in children [35] and curcumin supplementation [36], Japanese herbal medicine consumption [39][40][41], and an automated telephone symptom management system [42] in adults. There were no serious adverse events reported with any non-pharmacological self-management interventions [31][32][33][34][35][36][37][38][39][40][41][42][43].
Most evidentiary support exists for physical exercise for CIPN in people with advanced cancer, which has also been reported in the general cancer population. This is consistent with the findings from a recent systematic review [45] of 16 physical exercise interventions for CIPN in people at any stage of cancer, where exercise also improved physical function. The present review found additional benefits of exercise on CIPN symptom severity, which is unique to our review and perhaps unique to people with advanced cancer, and was not a finding in the general cancer population [45]. These beneficial effects of exercise on CIPN are supported by mechanistic studies. For example, exercise counteracts the deleterious effects of chemotherapy on the nervous system central to CIPN by reducing inflammation, suppressing pain pathways, and enhancing neuroprotective factors involved in the development, survival, and function of neurons [46,47]. In addition, exercise has broader health benefits for people with advanced cancer, such as improved quality of life, fatigue, body composition, psychosocial function, and sleep quality, and it appears safe when implemented on advice from a health professional [48]. Although strong recommendations cannot be made with the available evidence for the use of exercise for CIPN symptoms and related outcomes, it appears to be a safe, cost-effective, and viable adjuvant to chemotherapy to promote general health and wellbeing as well as CIPN prevention and management. Future well-powered randomized controlled trials are needed to confirm the efficacy and determine optimal exercise regimens, including exercise type and frequency, which is largely heterogeneous in the existing literature.
An emerging area of research for the prevention and management of CIPN in advanced cancer is the use of nutritional supplements. Consistent with the findings of this review, a recent meta-analysis in the general cancer population found that participants consuming Omega-3 PUFA supplements were 80% less likely to experience CIPN [49]. However, the findings of our current review were based on a study supplementing Omega-3 PUFA in an oral nutrition supplement drink [37]. Therefore, it cannot be confirmed whether the beneficial effects on CIPN resulted from Omega-3 PUFA or one or a combination of other nutrients with known benefits on nerve functioning and suggested modest benefits on CIPN in other cancer populations, such as amino acids, B vitamins, vitamin E, and magnesium [16,[50][51][52]. Our current review did find additional potential benefits of the amino acid glutamine for CIPN in advanced cancer populations [38]. Elsewhere, a review of five studies found oral glutamine to benefit neuropathic pain in participants with cancer of any stage; however, the clinical efficacy of glutamine for CIPN was judged to be unable to justify the additional daily cost [53]. Thus, future research on dietary interventions should consider cost-effectiveness in conjunction with the efficacy of isolated compounds and prioritize assessment of essential nutrients rather than non-nutrients such as curcumin or Japanese herbal medicine, for which there is less convincing evidence. No clinical recommendations can yet be made for nutrition supplements for CIPN in advanced cancer patients. However, clinical practice should focus on correcting nutritional deficiencies prior to, during, and after chemotherapy, especially for the aforementioned nutrients with suggested involvement in CIPN or biochemical parameters that have been found to be deficient in those with CIPN, such as vitamin D, hemoglobin, and albumin [54,55].
Our review also suggests that technology could assist in the education for selfmanagement of CIPN. In the general cancer population, healthcare interventions delivered via a mobile device (mHealth) have been associated with improved pain, fatigue, psychological distress, and sleep outcomes [56]. mHealth interventions have also shown potential to reach a large population due to ease of access, resulting in economic benefits to patients and healthcare systems [56]. In people with advanced cancer, improvements in cancer-related pain were observed after mHealth enabled psychoeducation that would otherwise only be accessible when implemented by a specialist [57]. Technology aside, the research emphasizes the requirement for self-management strategies in people with advanced cancer to be individualized and multi-faceted to consider older age, depression, impaired physical functioning, and low literacy, which may lead to greater difficulty in symptom self-management [17]. In addition, self-management strategies should be embedded within the person's support network and include strong partnerships with healthcare professionals, caregivers, and relatives [17]. Future research is warranted to determine the key elements of self-management strategies and the potential role of technology to facilitate implementation specifically for CIPN in people with advanced cancer.

Limitations and Future Research
Conclusions of this review have been drawn from a limited number of heterogeneous studies, evident by most outcomes having a GRADE rating of very low. Reasons for downgrading the level of evidence included substantial heterogeneity, small sample size, inadequate blinding, unclear random allocation concealment, lack of intention-to-treat analysis, and selective outcome reporting. Sources of heterogeneity were varying chemotherapy regimens, cancer types, and interventions, which were often reported with minimal detail. Furthermore, CIPN assessment methods were mostly subjective, not always validated, and differed greatly between the studies, which is problematic as a wide variation in CIPN outcomes has been noted with different measurement tools [58]. This suggests that the current tools measure different aspects of CIPN and thus might need to be implemented in combination to gather an accurate evaluation of CIPN. In addition, this systematic review was stringently conducted and reported according to best-practice guidelines [23,29] but only included studies in English and, therefore, may not have captured valuable research from some countries.
Future studies are needed to enhance the confidence in findings by increasing the body of evidence as well as the quality of the conduct of research. Randomized controlled trials should be well-designed with prescriptive interventions that seek to establish cause-effect and dose-response relationships. Randomized controlled trials should also be well-powered and use validated outcome measures and blinding where possible. Data should be reported in full and include adverse events, secondary outcomes related to CIPN, and compliance with self-management interventions. In people with advanced cancer, studies should prioritize exercise interventions and consider nutritional interventions and interactive technology-facilitated education for self-management support strategies administered alone or in combination. Additional suggested areas of research include those which have shown benefit in the general cancer population but are yet to be tested in advanced cancer. This includes low-cost and easily accessible interventions of massage, heat therapy, and meditation [20,59]. Future studies should also explore how non-pharmacological self-management interventions may best be implemented in conventional cancer care.

Conclusions
The strongest evidence with the most certainty was found for physical exercise as a safe and viable adjuvant to chemotherapy treatment for the prevention and management of CIPN and related physical function in people with advanced cancer. Nutrition supplements of glutamine and an Omega-3 PUFA-enriched drink showed some benefit, and the use of interactive technology may facilitate education for self-management of CIPN; however, certainty of these effects was very low. The confidence in the evidence to inform conclusions was mostly very low to moderate, which warrants future clinical trials. Trials should be rigorously designed and reported to include adequate sample size, clearly defined populations and interventions, and use valid outcome measures.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu14122403/s1, Table S1: Systematic search strategy to identify randomized controlled trials that examined the effect of non-pharmacological interventions on chemotherapy-induced peripheral neuropathy symptoms and related outcomes; Table S2: Characteristics and findings of studies that examined the effect of self-administered non-pharmacological interventions on chemotherapy-induced peripheral neuropathy symptoms and related outcomes; Figure S1: Japanese herbal medicine had no significant association with the likelihood of CIPN of Grade 1, 2, 3 or ≥2 (n = 2-3 studies; n = 226-271 participants; GRADE level: very low). Sensitivity analysis according to the type of herbal supplement (Goshajinkigan vs. ninjin'yoeito) did not result in significant findings; Figure S2: Physical exercise had no significant effect on quality of life, endurance, and emotional or social wellbeing (n = 2 studies; n = 76 participants; GRADE level: very low); Figure S3: Japanese herbal medicine had no significant effect on fatigue, nausea, vomiting, diarrhea, anorexia, relative dose reduction in oxaliplatin, neutropenia, nor thrombocytopenia (n = 2-3 studies; n = 231-283 participants; GRADE level: very low); Table S3: Grading of Recommendations, Assessment, Development and Evaluation (GRADE) for each outcome examining the effect of selfadministered non-pharmacological interventions on chemotherapy-induced peripheral neuropathy symptoms and related outcomes.

Conflicts of Interest:
The authors declare no conflict of interest.