Effectiveness of a Therapeutic Exercise Program to Improve the Symptoms of Peripheral Neuropathy during Chemotherapy: Systematic Review of Randomized Clinical Trials

Background: Therapeutic exercise has an important role to manage chemotherapy-induced peripheral neuropathy symptoms. However, there is little evidence of its effectiveness. Objective: To synthesize the evidence regarding therapeutic exercise during chemotherapy to improve peripheral neuropathy symptoms. Databases: PubMed, CINAHL, Cochrane Library, PEDro, ScienceDirect, Scopus, Web of Science and BIREME. Methodology: Randomized clinical trials were included. GRADE was used to synthesize evidence and an inverse variance model for meta-analysis. Results: Up to May 2022, 2172 references were analyzed and 14 studies that evaluated 1094 participants were included. The exercises were highly effective in improving pain threshold and moderately effective in improving peripheral neuropathy symptoms at the 8-week follow-up and the 4–24 weeks. Furthermore, the evidence was low in improving thermal threshold, tactile and vibratory sensitivity. Conclusion: Therapeutic exercise generates a significant reduction in peripheral neuropathy symptoms in patients in short- and long-term follow-up with a moderate level of evidence quality.


Introduction
Cancer is a global health issue with an increasing incidence and mortality. It is estimated that there will be 18.1 million new cases worldwide and 9.6 million deaths due to this disease [1].
Cancer therapy involves various treatments, such as surgery, chemotherapy, radiation therapy, immunotherapy and hormone therapy [2]. Chemotherapy, being one of the most widely used therapies, uses cytotoxic drugs with the aim of damaging the genetic material of neoplastic cells and preventing their replication [3]. This therapy is not selective; therefore, it damages both tumor cells and healthy cells. As a result of the damage, many side effects are manifested either in the short term such as nausea, neuropathy and fatigue, or in the long term such as premature menopause and cardiac and cognitive dysfunction [4].
Chemotherapy-induced peripheral neuropathy is one of the most debilitating side effects of chemotherapy, since the manifestation of symptoms is linked to the delivered dose of the different chemotherapeutic agents such as platinum compounds, taxanes vinca alkaloids, proteasome inhibitors and epothilones, among others [5].
The prevalence of developing peripheral neuropathy one month after completing chemotherapy is around 68%. The symptoms develop mainly in the hands and feet, with Life 2023, 13, 262 2 of 16 sensory alterations associated with numbness and paresis, motor generating balance and balance problems and autonomous problems with orthostatic hypotension.
Peripheral neuropathy can be evaluated objectively using quantitative sensory tests by performing clinical examinations, as well as subjective measurements such as questionnaires, scales and evaluations of nerve function [6,7]. However, there is no evaluation guideline that is used as a "gold standard", which generates a great limitation for proper clinical applicability for these patients [5].
Conservative management of chemotherapy-induced peripheral neuropathy involves physical exercise with different types of training, whether aerobic, endurance, motor sensory or balance. Only one systematic review was found, in which it evidenced that a training plan combined with resistance, strength and motor sensory exercises, which should last 36 weeks, at moderate intensities, with a frequency of 2 to 5 days a week and a duration 60 min has been effective in reducing symptoms of chemotherapy-induced peripheral neuropathy [8].
However, studies are lacking that address specific exercise programs for this particular condition [9] and that their results provide clinical applicability. While many studies were found specifying different types of training and subjective measurements of symptoms caused by chemotherapy-induced peripheral neuropathy, many did not identify specific objective and subjective measurements such as the perception of peripheral neuropathy, pressure pain threshold and thermal, tactile and vibration sensitivity.
Considering the above, this systematic review aims to synthesize the evidence regarding interventions with therapeutic exercises during chemotherapy to improve the symptoms produced by peripheral neuropathy with respect to the variables described above.

Study Designation
This systematic review was written according to the preferred reporting model for these study types and meta-analysis (PRISMA) and the recommendations of the Cochrane Collaborations for systematic reviews [10]. The review was registered in PROSPERO with the following number: CRD42020188275. On 9 April 2020, the search began in different databases, regarding various components of the research question considering population, intervention, comparison and results, to identify the knowledge gap; the search was performed until May 2022.

Literature Search
An electronic search of various articles indexed in the following databases was performed: PubMed, CINAHL Plus, Cochrane Central Register of Controlled Trials, Physiotherapy Evidence Database, Science Direct, Scopus and Web of Science y BIREME. The search strategy was adapted for each database. In PubMed, a combination of words was used: ("Neoplasms" [Mesh]) OR cancer AND (exercise) OR "Resistance Training" [Mesh] AND chemotherapy AND ("Peripheral Nervous System Diseases" [Mesh] OR peripheral neuropathy OR pressure pain threshold OR thermal sensitivity OR tactile sensitivity OR vibration sensitivity).
The Start program (version 3.4 BETA, sourced by the Research Laboratory in Software Engineering (LaPES) of the Federal University of São Carlos, Brazil) was used, which was held in the selection of titles, abstracts and full text, considering the analysis between evaluators and consensus criteria. Two independent reviewers (V.T. and D.O.) performed the selection process and two (C.S. and I.L.) participated in the consensus.

Inclusion Criteria
This systematic review included only randomized clinical trials that included therapeutic exercise with a variable related to symptoms of peripheral neuropathy in patients who underwent chemotherapy. The language of the publications was unlimited and should Life 2023, 13, 262 3 of 16 contain a pre-post comparison exercise and that the beginning of the training program was during chemotherapy.

Evaluation of Methodological Quality of the Studies
To Physiotherapy Evidence Database, a PEDro (www.pedro.org.au, accessed on 1 May 2022) scale was used to assess the methodological quality of the studies based on the Delphi list [11]. The studies that were included in this database were previously qualified; if there were no studies, they were manually evaluated by two examiners with a possible consensus by two evaluators. Clinical trials with scores greater than or equal to 6 were considered high methodological, 4 to 5 were rated as moderate quality and lastly, less than or equal to 3 were classified as low methodological quality [12,13].
Data of participants and methodology of the studies were extracted using a standardized form adapted from the Cochrane Collaboration model [10]. Moreover, effect size (ES) with a 95% confidence interval (CI) for continuous outcomes in each comparison group was calculated and the values before and after the intervention were considered. The treatment was classified as small (<0.3), moderate (between 0.4 and 0.7) and large (>0.8) according to Cohen's index interpretation [14].
The results of the primary studies were interpreted according to the effectiveness of the training programs to improve the symptoms of perception of peripheral neuropathy and increase the pain threshold to pressure, thermal, tactile and vibratory sensitivity. They were considered positive when comparing the intervention and the control groups presented a statistically significant improvement in the primary outcomes.
The preventive effects of therapeutic exercise were rated with an equal sign when there was no difference between the pre and post intervention. Finally, studies that presented a significant reduction in the variables of interest of the intervention were classified as having no effect. The effect size was calculated for studies that presented descriptive ideas, represented by means and standard deviation for the main variables such as pressure, thermal, tactile and vibratory pain threshold.
The synthesis of evidence for each intervention was calculated with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) that considers high, moderate, low or very low evidence level based on the following factors: limitations, indirectness, inconsistency, imprecision and lastly bias risk. In this review, the GRADEpro software (https://gradepro.org, accessed on 1 May 2022) [15] was used to create tables with the synthesis of evidence.

Statistical Analysis
A meta-analysis was performed using means and standard deviations from each selected clinical trial. The difference of standardized means and the 95% confidence interval were calculated using an inverse variance model of random effects for the meta-analysis, considering the data after the intervention. Data heterogeneity between studies were assessed using the I statisitic 2 . The p values were calculated and statistical significance was set at <0.05. Statistical analysis was performed using the RevMan five-revision management software (version 5.3, 11-13 Cavendish Square, London, UK).

Results
The studies obtained based on the search with the keywords totaled 2.172 articles, which were examined by title and abstracts; after reading the full text of 42 studies, 14 articles were selected that met the inclusion criteria. The flow diagram used is presented in Figure 1.

Results
The studies obtained based on the search with the keywords totaled 2.172 articles, which were examined by title and abstracts; after reading the full text of 42 studies, 14 articles were selected that met the inclusion criteria. The flow diagram used is presented in Figure 1.
Eight of the studies obtained a score equal to or greater than six; therefore, they were classified as having high methodological quality [8,17,18,22,23,[26][27][28]. On the other hand, three studies [7,19,20] reached a score between four and five and were categorized as moderate quality. Finally, three studies [16,21,24] obtained a score of three and were therefore cataloged as having low methodological quality. None of the studies scored on the blind allocation criteria for both participants and therapists. In addition, only five studies
Eight of the studies obtained a score equal to or greater than six; therefore, they were classified as having high methodological quality [8,17,18,22,23,[26][27][28]. On the other hand, three studies [7,19,20] reached a score between four and five and were categorized as moderate quality. Finally, three studies [16,21,24] obtained a score of three and were therefore cataloged as having low methodological quality. None of the studies scored on the blind allocation criteria for both participants and therapists. In addition, only five studies [17,20,25,27,28] reported that the evaluators who measured at least one key result were blinded. Only six studies [7,19,23,25,26,28] presented results for all subjects who received treatment or were assigned to the control group, or when this could not be, data for at least one key outcome were analyzed by intention to treat (Table 1).   Table 2 describes the main characteristics of the 14 included studies, of which 1094 people participated, having an average age range of 19 to 79 years old, with solid and hematological cancer diagnoses. The comparison was made between a control group, which was based on standard care mainly focused on evaluations and education to patients, and an intervention group, where the main type of exercises that were performed were aerobic training, strengthening of both lower and upper limbs and balance exercises, that began from the first day of chemotherapy treatment or weeks after it. Among the most used equipment in these were the elastic bands, treadmill and pedometer. From 14 studies, five mentioned that intervention was supervised by a physiotherapist [16,17,27,28], one by a nurse [23], one by a sport scientist [21], one by a certified investigator by the ACSM [7], one by an exercise physiologist or oncology nurse [19] and one by a specialist in prescribing exercises for cancer patients [18]. Four studies did not mention the profession of the trainer [20,22,24,26].

Characteristics of Included Studies
The training frequency ranged from 2 to 7 days a week with a duration of 5 to 60 min depending on the exercise performed, a moderate intensity depending on the chemotherapy cycle in which the patients were, and the duration of the protocol varied between 4 and 56 weeks. A physiotherapist supervised 23% of the programs, while the others were supervised by a professional trained to carry out this work. The main variables analyzed were peripheral neuropathy perception, pressure pain threshold and thermal, tactile and vibratory sensitivity. The follow-up was carried out in a period where the minimum range was 4 weeks and the maximum was approximately 36 weeks. The effect size range varied from 0.05 to 0.93 for peripheral neuropathy perception; as for pressure pain threshold, it ranged from 0.44 to 0.57, thermal threshold was 0.06 to 0.28, tactile sensitivity presented a value of 0.01 and vibratory sensitivity presented a value of 0.15.

Synthesis of Evidence
The evidence regarding therapeutic exercises to improve symptoms of peripheral neuropathy, pressure pain threshold, thermal threshold, and tactile and vibratory sensitivity was synthesized according to GRADE with follow-up times of 8 weeks (Table 3) and between 4 weeks and 24 weeks (Table 4). For the synthesis of evidence, two studies were excluded [16,23] because they did not present the necessary data to form part of this analysis, such as control group, intervention, follow-up and the duration of the protocol.

Follow-Up 4 to 24 Weeks
Eight studies [7,17,18,[20][21][22][26][27][28] evaluated the peripheral neuropathy perception using questionnaires and scales; with a total of 747 participants in these trials, 52.7% were part of the control group, while 47.2% belonged to the exercise group. According to the factors that can lower the level of quality of the evidence, a score of not serious was presented for risk bias, indirect evidence, imprecision and publication bias, while the inconsistency was categorized as serious since four of the studies were classified as serious and one as very serious, while three of them were not serious. Finally, the therapeutic exercises of strengthening, aerobics, equilibrium and balance presented moderate evidence to improve the symptoms of peripheral neuropathy during chemotherapy for cancer treatment.

Follow-Up at 8 Weeks
Seven studies [7,18,21,22,26,27] evaluated the peripheral neuropathy perception through scales and questionnaires, which included a total of 538 participants of which 53.3% were from the control group and 46.6% from the exercise group. According to factors that may lower the level of quality of the evidence, risk of bias, imprecision and publication bias were categorized as non-serious.
The inconsistency was serious since three studies were classified as serious, one study was classified as very serious and two studies as not serious. Eight weeks of therapeutic exercises to strengthen the upper and lower limbs, equilibrium, aerobics and balance, presented moderate evidence to improve the symptoms of peripheral neuropathy during chemotherapy for cancer treatment.

Pressure Pain Threshold
Two studies [17,19] with a follow-up range of 16 to 18 weeks in which pressure pain threshold was evaluated with an algometer had a total of 254 participants, where 33.8% belonged to the control group and 66.1% to the group with exercises. None of the factors that may lower the level of quality of the evidence were considered serious or very serious. A 16-18-week follow-up with therapeutic nerve gliding, stretching, aerobic and interval resistance exercises presented high evidence for increasing the pressure pain threshold in the trapezium, quadriceps and gluteal areas.

Thermal Threshold
Two studies [7,20] with a follow-up that ranged from 6 to 19 weeks and where thermal threshold was evaluated through questionnaires and scales included 486 participants, of which 51.4% were from the control group and 48.5% from the exercise groups. According to the factors that can lower the level of quality of the evidence, the risk of bias, imprecision, inconsistency and publication bias were categorized as non-serious; however, the indirect evidence was stated as very serious, since both included studies using tools that are not objective for measurement, such as quantitative sensory testing and a numerical scale of 0 to 10.
Along with the above, a follow-up of 6-19 weeks with therapeutic exercises on a vibration platform for the whole body, aerobics and strengthening exercises in the upper and lower limbs presented low evidence to improve the thermal threshold in the lower limb area.

Tactile Sensitivity
Two studies [18,20] with a follow-up range of 12 to 19 weeks evaluated the effects of the therapeutic exercises on tactile sensitivity with questionnaires and Esthesiometer; a total of 158 participants were included, of which 50.6% were from the control and 49.3% were from the exercise group. Regarding the factors that can reduce the level of quality of the evidence, the risk of bias, indirect evidence and publication bias were not serious, while the inconsistency was categorized as serious because two studies were categorized as serious, as well as imprecise because the included studies considered less than 200 participants. Thus, a 12-19-week follow-up with therapeutic exercises to strengthen the lower limbs, aerobics and with a whole-body vibration platform presented low evidence to improve tactile sensitivity in the lower limb area.

Vibratory Sensitivity
Five studies [17,18,22,24,25] with a follow-up range of 4 to 36 weeks evaluated the effectiveness of therapeutic exercises on the vibratory threshold using a quantitative sensitive test and diapason; a total of 183 participants were included, of which 53% were from the control group and 46.9% from the exercise group. According to the factors that can lower the level of quality of the evidence, the risk of bias, indirect evidence and publication bias were not serious, while inconsistency was classified as serious. Only one was not serious, as well as imprecise because the included studies considered less than 200 participants. A follow-up of 4-36 weeks of therapeutic exercises to strengthen the lower limbs, aerobics, balance and motor sensory showed low evidence to increase the vibratory threshold in the lower limb area, such as phalangeal metatarsus, medial malleolus and phalanges, as well as upper limbs such as hands and wrists.

Tactile Sensitivity
Two studies [18,20] with a follow-up range of 12 to 19 weeks evaluated the effects of the therapeutic exercises on tactile sensitivity with questionnaires and Esthesiometer; a total of 158 participants were included, of which 50.6% were from the control and 49.3% were from the exercise group. Regarding the factors that can reduce the level of quality of the evidence, the risk of bias, indirect evidence and publication bias were not serious, while the inconsistency was categorized as serious because two studies were categorized as serious, as well as imprecise because the included studies considered less than 200 participants. Thus, a 12-19-week follow-up with therapeutic exercises to strengthen the lower limbs, aerobics and with a whole-body vibration platform presented low evidence to improve tactile sensitivity in the lower limb area.

Vibratory Sensitivity
Five studies [17,18,22,24,25] with a follow-up range of 4 to 36 weeks evaluated the effectiveness of therapeutic exercises on the vibratory threshold using a quantitative sensitive test and diapason; a total of 183 participants were included, of which 53% were from the control group and 46.9% from the exercise group. According to the factors that can lower the level of quality of the evidence, the risk of bias, indirect evidence and publication bias were not serious, while inconsistency was classified as serious. Only one was not serious, as well as imprecise because the included studies considered less than 200 participants. A follow-up of 4-36 weeks of therapeutic exercises to strengthen the lower limbs, aerobics, balance and motor sensory showed low evidence to increase the vibratory threshold in the lower limb area, such as phalangeal metatarsus, medial malleolus and phalanges, as well as upper limbs such as hands and wrists.
The sensitivity analysis revealed that heterogeneity was influenced by the studies by Bland et al. 2019 [18] and Mijwel et al. 2019 [19]. There were no changes in the results in favor of therapeutic exercise compared to the control groups and the DME was reduced to −0.40 with changes in heterogeneity from moderate to significant (I 2 = 45%; p = 0.09).

Discussion
With respect to the quantitative analysis of this systematic review, the findings of the meta-analysis show that a therapeutic exercise program of 4 to 56 weeks generates significant changes, reducing the symptoms of peripheral neuropathy in subjects with cancer compared to the control group with short-term and long-term follow-up. However, given the significant heterogeneity presented, these results should be viewed with caution.
Of the main studies selected, two of them reported improvement of pressure pain threshold after a therapeutic exercise program [17,19]. The study by Mijwel et al., 2018 [13] reported that resistance exercise associated with HIIT significantly improves muscle strength and reduces pain sensitivity; these studies had a moderate effect size with a range of 0.44 to 0.57 and with a high level of evidence.
Regarding the evaluation of neuropathy symptoms through questionnaires, six studies [7,18,21,22,[26][27][28] presented moderate evidence with an effect size of 0.27 to 0.47, categorizing them as small to moderate. Three of these studies [7,26,28] reported improvement of peripheral neuropathy symptoms, while two other studies [21,22] kept their symptoms. There was one study [27] that reported an improvement in neuropathic pain symptoms while maintaining generalized pain symptoms after a therapeutic exercise program. The studies that evaluated neuropathic symptoms through thermal threshold [7,20], tactile sensitivity [18,20] and vibratory sensitivity [17,18,22,24,25] presented a low level of evidence.
In regard to the studies that evaluated tactile sensitivity [18,20] and thermal threshold [7,20], no significant differences were obtained. The effect size of the studies [7,20] that evaluated thermal was 0.28 and 0.06, considered small. Likewise, the study [20] that evaluated tactile sensitivity had a small effect size of 0.01. The study that evaluated the vibratory threshold [22] presented a small effect size with a value of 0.15. Those results could be explained by the fact that symptoms of peripheral neuropathy induced by chemotherapy should be assessed by both objective methods and from a patient perspective [29]. Moreover, the heterogeneity of evaluation tools used in those studies negatively impacts these results.
The findings reported in this review are similar to the study by Dobson et al., 2014 [30] that evaluated the effects of balance and aerobic exercise training on neuropathy symptoms, sensory dysfunctions and increased peripheral nerve conduction velocity. However, the review by Dobson et al., 2014 [30] did not determine the quality of evidence and included studies that focus on diabetic neuropathy and its neuro-inflammatory etiology and did not incorporate studies where the neuropathy is the result of exposure to chemotherapy.
The present systematic review determines that strengthening, aerobic and motor sensory exercises for a total average time of 14 weeks (considering a standard deviation of 9 weeks) 2 to 5 days per week and with low to moderate intensities present a moderate level of evidence for the improvement of neuropathy symptoms. These facts coincide with the results of the study by Kneis et al., 2019 [31] that reported that resistance and equilibrium exercises reduced sensitivity symptoms, with an improvement in the physical function of the cancer survivors.
According to the above and considering that there is still controversy regarding the effectiveness of drugs in the management of peripheral neuropathy [32], multimodal exercise emerges as an effective and safe therapeutic tool to reduce peripheral symptoms induced by chemotherapy.
Some limitations of the present study should be considered: (1) three studies showed poor methodological quality [10,15,18]; (2) different therapeutic exercise protocols were used; (3) significant heterogeneity (I 2 = 86%) which is due to the different types of exercises between studies and different evaluation tools used, providing variability; (4) the selected studies used distant tools to evaluate peripheral neuropathy; (5) two studies [7,20] used methods to evaluate the peripheral neuropathy that were inadequate, since they lacked reliability and objectivity; (6) non-specificity of therapeutic exercise in terms of its dosage; (7) the beginning of the therapeutic exercise program was nonspecific since some studies began together with the chemotherapy [17,27], on the first day of chemotherapy [7,16], before starting it [18,23], during this therapy [24,25] or on subsequent days to the start of chemotherapy [18,19]. In addition, three studies [20,22,26] did not specify the beginning of the intervention protocol; (8) Three studies [20,22,24] did not identify any supervisor, which may limit the correct execution of the exercises; (9) different follow-up times between studies can generate a bias in the results.

Conclusions
In general, this systematic review and meta-analysis suggest that therapeutic exercise generates a significant reduction in peripheral neuropathy symptoms in patients in short and long-term follow-up with a moderate level of evidence quality. However, these results must be viewed with caution due to the significant heterogeneity of the studies analyzed. The available studies are diverse in terms of methodology, exercise dosage, and tools to assess peripheral neuropathy; therefore, further research is warranted. Future clinical trials must present adequate methodological quality and use valid and reliable evaluation methods.

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