Prevention and treatment of chemotherapy-induced peripheral neuropathy

INTRODUCTION — Chemotherapy-induced peripheral neuropathy (CIPN) is a common adverse effect of cancer therapy that can have a profound impact on quality of life and survivorship [1]. Long-term neurotoxicity is an important issue for the growing number of cancer survivors, with the highest number of affected patients having been treated for breast and/or colon cancer. CIPN may also adversely affect oncologic outcomes by forcing dose modifications and/or premature treatment discontinuation. (See "Overview of long-term complications of therapy in breast cancer survivors and patterns of relapse", section on 'Neurologic morbidity' and "Approach to the long-term survivor of colorectal cancer", section on 'Oxaliplatin-induced peripheral neuropathy'.)

The incidence of CIPN varies according to the chemotherapeutic agent, dose, duration of exposure, and method of assessment. The agents with the highest incidence are the platinum drugs, especially cisplatin and oxaliplatin, taxanes, vinca alkaloids, and bortezomib. The incidence, risk factors, pathogenetic mechanisms, and clinical characteristics of CIPN from these classes of agents as well as other chemotherapeutic drugs are discussed in detail elsewhere. (See "Overview of neurologic complications of platinum-based chemotherapy" and "Overview of neurologic complications of conventional non-platinum cancer chemotherapy".)

This topic review will cover approaches to prevention and treatment of CIPN focusing mainly on platinum drugs, taxanes, vinca alkaloids, and bortezomib. An overview of neurologic complications with platinum and non-platinum chemotherapy drugs, and recommendations for dose modification for platinum and non-platinum chemotherapeutic drugs when neuropathy develops during therapy are provided elsewhere. (See "Overview of neurologic complications of platinum-based chemotherapy" and "Overview of neurologic complications of conventional non-platinum cancer chemotherapy".)

OVERVIEW OF CLINICAL FEATURES AND NATURAL HISTORY

Chronic neurotoxicity — Regardless of the specific drug, chronic CIPN has similar distinguishing clinical features that help to differentiate it from other neuropathies (table 1) (see "Overview of neurologic complications of conventional non-platinum cancer chemotherapy" and "Overview of neurologic complications of platinum-based chemotherapy"):

●CIPN is typically dose-dependent and cumulative.

●With most chemotherapy drugs, CIPN typically has a symmetric, distal, "stocking and glove" distribution.

●CIPN predominantly consists of sensory rather than motor symptoms, and motor nerve function usually remains unchanged during treatment. An exception is thalidomide, which is associated with weakness and tremor in 30 to 40 percent. Paclitaxel can also be associated with motor neuropathy (mainly proximal) in up to 14 percent of patients (table 1). (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Thalidomide and related agents' and "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Clinical features, incidence, and risk factors'.)

●Autonomic neuropathy is rare with the exception of the vinca alkaloids, particularly vincristine, which commonly cause constipation (table 1). (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy".)

In many cases, the natural history of CIPN is gradual improvement over time after chemotherapy is stopped, although it may continue to worsen over a few months with some drugs (most notably cisplatin and oxaliplatin) before it starts to improve. This has been labeled as being a "coasting phenomenon." As examples:

●Even if the drug is discontinued, cisplatin- and oxaliplatin-related CIPN, on average, continues to worsen for approximately three months. While it has been claimed that neuropathy may even begin after therapy is discontinued, a close look at this phenomenon illustrates that the slope of worsening neuropathy is the same in the two to three months before the chemotherapy is stopped as it is in the two to three months after it has been stopped. This is consistent with the conclusion that the full manifestation of CIPN, after a dose of cisplatin or oxaliplatin, takes two to three months [2]. CIPN eventually improves in most patients, although recovery is often incomplete. (See "Overview of neurologic complications of platinum-based chemotherapy", section on 'Clinical and electrophysiologic manifestations'.)

●With paclitaxel, after completing treatment, approximately one-half of patients improve over a period of four to six months [3,4]. However, severe neuropathy can persist. In one study, up to 80 percent of patients still had neuropathic symptoms up to two years after completing treatment; approximately 25 percent reported severe symptoms of numbness and/or discomfort in their hands and feet [3]. While some patients report that CIPN worsens for a while after paclitaxel is stopped, on average, paclitaxel CIPN begins to improve in the first two to three months after it is stopped, in contrast to what occurs with oxaliplatin [2]. (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Clinical features, incidence, and risk factors'.)

●Vincristine neuropathy is usually reversible but improvement is gradual and may take up to several months. (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Vincristine'.)

●Symptoms of bortezomib neuropathy usually improve or completely resolve after three to four months following discontinuation of treatment; in one study, 64 percent with grade 2 or worse neuropathy experienced improvement or resolution of symptoms, compared with baseline, at a median of 110 days [5]. (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Bortezomib'.)

Persistent sensory and motor neuropathy is a significant clinical problem in cancer survivors that can result in chronic pain, limited mobility, and disruption of fine motor sensations, proprioception, and balance, and can contribute to functional limitations [6,7]. The deleterious impact on proprioceptive input in particular can significantly alter postural stability, balance, and gait, increasing the risk of falls.

Diagnostic criteria for chemotherapy-induced peripheral neuropathic pain are available (table 2). (See "Overview of cancer pain syndromes", section on 'Chemotherapy-related neuropathy'.)

Acute neurotoxicity — Some chemotherapeutic drugs that are associated with CIPN, notably oxaliplatin and paclitaxel, can cause an acute neurotoxicity syndrome, which is clinically distinct from CIPN, not necessarily peripheral, noncumulative, and usually improves within days of each dose, recurring with subsequent doses:

●Oxaliplatin-induced acute neurotoxicity is characterized by unique motor and sensory symptoms (cold sensitivity, throat discomfort, and muscle cramps, perioral numbness) that occur within hours to days of receiving the drug; it, on average, is less prominent with the first dose than with subsequent doses, but on average, does not appear to worsen cumulatively after the second dose. While it classically has been thought that this acute neuropathy completely resolves after a few days, data support the view that it does not completely resolve between oxaliplatin doses, at least with repeat dosing every two weeks. Patients with the most severe acute symptoms are at higher risk to develop more severe chronic neuropathy [2]. (See "Overview of neurologic complications of platinum-based chemotherapy", section on 'Acute neurotoxicity'.)

●Patients receiving paclitaxel can develop an acute pain syndrome. The timing is similar to that seen with oxaliplatin-induced acute neuropathy. This syndrome, in the distant past had classically been termed as paclitaxel-associated arthralgias and myalgias, but data first published in 2012 support the view that this is a form of an acute neuropathy, as opposed to a disorder of joints and/or muscles [8-10]. This problem is much more prominent in patients receiving higher doses on an every-three-week schedule than with lower weekly doses. This phenomenon also occurs with docetaxel, although less commonly. (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Clinical features, incidence, and risk factors'.)

PREVENTIVE APPROACHES — There are no established agents that can be recommended for the prevention of chemotherapy-induced peripheral neuropathy (CIPN) in patients with cancer undergoing treatment with potentially neurotoxic agents. This position is consistent with updated year 2020 guidelines from the American Society of Clinical Oncology (ASCO) for prevention and management of CIPN [11], and joint guidelines from the European Society for Medical Oncology/European Oncology Nursing Society/European Association of Neuro-Oncology (ESMO/EONS/EANO) [12]. For patients receiving oxaliplatin, we suggest interspersing a non-oxaliplatin-containing "maintenance" chemotherapy regimen with the oxaliplatin regimen when appropriate in the clinical setting (palliative chemotherapy). For patients receiving bortezomib, weekly rather than twice weekly treatment schedules and subcutaneous as compared with intravenous administration of bortezomib are preferred, as they are associated with less frequent and less severe neurotoxicity.

Given the absence of effective preventive agents, clinicians should carefully assess the benefits of agents known to cause CIPN against the risks of developing long-term irreversible neuropathy, especially among patients with underlying neuropathy and with conditions that predispose to neuropathy (eg, diabetes mellitus and/or a personal or family history of hereditary neuropathies).

Pharmacologic agents and physical strategies — Despite CIPN being relatively distinct from other neuropathies, many of the interventions chosen to undergo evaluation for prevention (and treatment) of CIPN are those with demonstrated efficacy in other common neuropathic pain conditions (eg, painful diabetic neuropathy, postherpetic neuralgia). Nearly all of these studies have been conducted in patients receiving platinum drugs (cisplatin, carboplatin, or oxaliplatin), taxanes, and/or vinca alkaloids.

Interventions with potential but unproven benefit — Although proof of benefit has not been established, available data suggest that exercise, cryotherapy, and compression therapy with or without cryotherapy may in part prevent symptoms of CIPN, and appear to reasonably safe, although clinicians should be aware of frostbite risk. Ganglioside-monosialic acid (GM1) was reported to be effective in one Chinese study but confirmation in larger trials and other ethnic groups is needed.

Cryotherapy, compression therapy, or both — For patients receiving weekly therapy with paclitaxel (and some patients receiving oxaliplatin), cryotherapy (eg, by using frozen socks and gloves before, during, and after drug infusion) may be useful to diminish objective and subjective symptoms of CIPN. However, as the data are less than robust, confirmation in additional larger studies is needed before this approach can be recommended. This conclusion is consistent with the updated ASCO and joint ESMO/EONS/EANO guidelines on prevention and treatment of CIPN [11,12].

The role of compression therapy, either with or without cryotherapy, remains uncertain. The following data are available:

●Cryotherapy alone – A few uncontrolled reports suggested potential utility for frozen gloves/socks worn during drug infusion to prevent/mitigate neurotoxicity related to weekly paclitaxel in women with breast cancer [13-15].Benefit was further supported by data from a prospective trial in which 40 women with breast cancer receiving weekly paclitaxel wore frozen gloves and socks on their dominant side before, during, and after each infusion (90 minutes total); symptoms on the treated side were compared with the untreated (nondominant) side [16]. The main endpoint was any decline in tactile sensation as assessed by the Semmes-Weinstein monofilament test, an outcome measure that is routinely used for diabetic neuropathy (picture 1), at a cumulative dose of 960 mg/m²; secondary outcome measures included clinician-rated thermosensory deficits using a thermal stimulator at 3 and 48°C, vibration perception using a 128 Hz tuning fork, performance speed as assessed by a grooved pegboard test, abnormal electrophysiologic signs as assessed by electrophysiologic testing of the median nerve, and current perception thresholds using a neurometer. Subjective symptoms were also assessed using a patient neuropathy questionnaire in which individual items were scored as A (no neuropathy), B (mild neuropathy), C (moderate neuropathy not interfering with activities of daily living [ADL]), D (moderate neuropathy that interferes with ADL), or E (severe neuropathy). Grades D and E were scored as severe neuropathy. Overall, 36 patients were treated to the target paclitaxel dose and were analyzed; none dropped out due to cold intolerance.

Significant differences were reported for both objective and subjective signs of CIPN. On the control (nondominant) side, 81 and 64 percent of patients had a decrease in tactile sensation in their hands and feet, respectively, while on the intervention (dominant) side, only 28 and 25 percent were reported to have had a decrease in tactile sensation in their hands and feet, respectively. Subjects reported less severe neuropathy on the intervention side (for the hand, 3 versus 42 percent; for the foot, 3 versus 36 percent), and there were differences in perception of warmth and reaction speed but no differences in vibration threshold or electrophysiologic testing between the intervention and control sides.

While the Semmes-Weinstein monofilament test is the most widely used test to diagnose loss of protective sensation, which is one of the components of CIPN, it does not necessarily assess the more common symptoms of sensory neuropathy, such as paresthesia, tingling, numbness, loss of position and/or vibratory sense, and pain [17].

Data from more recent trials have been mixed:

•A larger trial, involving 180 patients, randomized patients to have their upper extremities treated with cryotherapy, or not, in patients receiving paclitaxel. Cryotherapy-treated patients had less hand neuropathy symptoms [18].

•Another 42-patient randomized trial of extremity cryotherapy versus not in patients receiving paclitaxel concluded that cryotherapy might be a useful approach but needed further evaluation [19].

•In another trial, 79 gynecologic oncology patients receiving paclitaxel-based chemotherapy were randomly assigned to cold therapy (frozen gloves/socks with a cold pack inside worn on both hands and both feet starting 15 minutes before until 15 minutes after each paclitaxel infusion), versus none; the incidence and impact of symptomatic CIPN was evaluated using the FACT/GOG-Neurotoxicity (Ntx) scale at each chemotherapy cycle and at the one month follow-up after chemotherapy completion [20]. There was a significant difference in the CIPN incidence from the first cycle of therapy that favored the cold therapy, and the benefits were greater with each subsequent chemotherapy cycle (overall incidence of any grade CIPN for treated versus control group 48.7 versus 100 percent). At one month post completion of chemotherapy, the overall incidence of CIPN of any grade was diminished in the treated group, but remained the same in control (26 versus 100 percent). At all time points, the scores for FACT/GOG-Ntx were significantly better in the treated group. Four patients discontinued cold therapy because of pain, but there were no significant other adverse effects of therapy.

•In contrast to the above reports suggesting that cryotherapy was a promising approach and was reasonably well tolerated, two other pilot trials noted that the cryotherapy was not very well tolerated by a substantial number of patients and did not appear to be effective for decreasing CIPN [21,22]. One reason may be the different means of administering cryotherapy in the various studies.

●Compression therapy with or without cryotherapy – Several trials have explored compression therapy with or without cryotherapy, but the benefits of compression therapy remain uncertain:

•In a small phase II trial of wearing one size too small surgical gloves (two gloves per hand) on only one hand for 90 minutes in 42 patients receiving nabpaclitaxel in an attempt to reduce the flow of blood during chemotherapy, less neuropathy (as measured by the CTCAE and by the subjective Patient Neurotoxicity Questionnaire) was seen in the double-gloved hand, compared with the other (ungloved) hand [23]. Measured fingertip temperatures were decreased on the gloved hand.

•The utility of combining cryotherapy and compression therapy was evaluated in a very small single-arm trial involving 13 patients [24]. The author compared the data from these 13 patients with that of 20 patients who were involved in a prior study conducted by the same investigators, who were treated via a continuous-flow limb hypothermia using a garment on one lower extremity while the other lower extremity served as a control [25]. The limb subjected to combined treatment appeared to do better than did limbs exposed to either therapy alone.

•Cryotherapy and compression therapy were directly compared in a 38-patient trial of patients receiving nabpaclitaxel, in which patients used cryotherapy on one hand and compression therapy on the other [26]. The outcomes appeared similar in each study arm.

•The recently reported CONTRoL trial assigned 100 patients receiving paclitaxel to get extremity cryotherapy, compression therapy, or neither (NCT03873272) [27]. In a preliminary report presented at the 2021 San Antonio Breast Cancer Symposium, the cryotherapy group fared less well than the other two groups, both in terms of toxicity from the prevention approach and also in the amount of CIPN that patients developed.

Notably, in the spring of 2022, the NCI approved cooperative group phase III clinical trial conducted by ALLIANCE/SWOG that will randomize patients receiving a taxane to one of three arms: compression therapy, combined compression and cryotherapy, or a control group (the STOP CIPN trial, SWOG 2205). Hopefully this trial, which is not yet open for patient enrollment, will define the true benefit of this approach.

Exercise — Adequately powered phase III trials with rigorous assessment of neurotoxicity by both clinicians and patients using standardized and validated neuropathy rating scales are needed before it can be concluded that exercise reduces either the frequency or severity of CIPN [11]. Nevertheless, given the myriad benefits associated with exercise, including its potential to mitigate CIPN, it is reasonable to suggest exercise to patients receiving potentially neurotoxic chemotherapy, as supported by ASCO and ESMO/EONS/EANO guidelines [12].

Early reports suggest a possible protective effect of exercise on CIPN related to taxanes, platinum drugs, and vinca alkaloids:

●An early trial, in which 61 patients undergoing chemotherapy for lymphoma were assigned to chemotherapy alone or with a 36-week intervention of sensorimotor, endurance, and strength training twice a week, concluded that exercise was associated with significantly improved quality of life (QOL; the primary endpoint), significantly reduced peripheral neuropathy-related deep sensitivity (as assessed by blinded application of a tuning fork to the metatarsophalangeal and malleolus medialis by the clinician), and a better activity level and balance [28].

●A potential benefit for exercise to prevent or mitigate CIPN was also suggested in a secondary analysis of a multicenter phase III trial in which 619 cancer patients receiving chemotherapy were randomly assigned to chemotherapy with or without a standard exercise program (EXCAP; an individualized, moderate-intensity, home-based, six-week progressive walking and resistance exercise program). The primary endpoint of the main trial was fatigue, and results have not yet been published. However, a subset analysis of 355 patients enrolled in this trial (185 receiving chemotherapy alone and 170 receiving exercise plus chemotherapy) who were receiving a taxane-, platinum-, or vinca alkaloid-containing chemotherapy regimen examined the impact of exercise on CIPN [29]. The degree of CIPN was assessed by patient report of the intensity of numbness and tingling, and of hot/cold sensation of the hands or feet, measured on a scale from 0 (no problems) to 10 (as bad as you can imagine). As assessed by these nonstandard neurotoxicity assessment methods, exercise appeared to significantly reduce CIPN for patients receiving vinca alkaloids, platinum, or taxanes; the benefit seemed to be limited to older individuals. The authors postulated that the lack of impact in younger individuals might be attributable to the need for a higher "dose" of exercise needed to prevent CIPN in younger, leaner, and more fit individuals [30].

●Other published data also provide some support for exercise [28,31-34]. All of these data should be viewed as preliminary and hypothesis-generating, and not definitive. A cooperative group trial is under development to address this topic.

Additional data on exercise in patients with established CIPN is provided below. (See 'Exercise' below.)

Acupuncture — One small randomized sham-controlled trial of weekly electroacupuncture for prevention of taxane-induced CIPN in 63 women with early breast cancer failed to demonstrate benefit; furthermore, the electroacupuncture group had a slower recovery of neuropathy after chemotherapy was stopped [35].

Until further data are available, we agree with the updated 2020 ASCO guidelines, which concluded that no recommendations can be made regarding the use of acupuncture to prevent CIPN [11]. Updated guidelines from ESMO/EONS/EANO discourage the use of acupuncture to prevent CIPN [12].

Additional studies on the use of acupuncture and electroacupuncture to treat established CIPN are discussed below. (See 'Acupuncture' below.)

GM1 — Until further data are available, no recommendations can be made regarding the use of GM1 as a neuroprotective agent, an approach that is consistent with ASCO guidelines [11]. GM1 is not available in the United States.

Ganglioside-monosialic acid (GM1) is a monosialo-glycosphingolipid with neuroprotective properties in vitro [36,37]. The following data are available in patients receiving potentially neurotoxic chemotherapy:

●A retrospective evaluation of chemotherapy-induced neuropathy in patients who took GM1 versus those who did not revealed that patients who had had GM1 had less neuropathy [38]. A randomized, but not blinded, phase II Chinese trial also suggested that GM1 might reduce the severity of oxaliplatin-induced neurotoxicity [39].

●A subsequent randomized Chinese trial of GM1 (80 mg intravenously, daily for three days, starting the day before chemotherapy) versus placebo was conducted in 206 patients with early stage breast cancer planning to receive a taxane-based adjuvant chemotherapy regimen (predominantly docetaxel, with or without cyclophosphamide and epirubicin). The primary endpoint was the physician-assessed Functional Assessment of Cancer Therapy (FACT) Neurotoxicity subscale score after four cycles of chemotherapy [40,41]. Secondary endpoints included neurotoxicity evaluated by two other scales: the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE; v4.0) and the Eastern Cooperative Oncology Group (ECOG) neuropathy scale [42].

Treatment with GM1 reduced both the severity and incidence of taxane-induced peripheral neuropathy as reflected by a higher FACT Neurotoxicity subscale score (43.27 [95% CI 43.05-43.49] versus 34.34 [95% CI 33.78-34.89]). In addition, fewer patients had grade 1 or worse neurotoxicity as assessed by the NCI CTCAE; 14 versus 100 percent) or ECOG neurotoxicity scales (26 versus 98 percent). Patients receiving GM1 had a lower incidence of chemotherapy dose reductions/delays, and fewer patients developed taxane-associated acute pain syndrome. (See 'Acute neurotoxicity' above.)

There was no suggestion of an increase in grade 3 or 4 adverse events in the GM1 group, but information on long-term outcomes, especially oncologic outcomes, was not provided. Interestingly, the neuropathy in the control group totally resolved three months after taxane completion, something that is not common in other trials. The authors of this manuscript suggested that the trial be replicated in a non-Chinese population, a conclusion that was also made by the authors of an associated editorial [43]. Work is ongoing to try to replicate this trial.

Glucocorticoids and NSAIDs for taxane-associated acute pain syndrome

●Glucocorticoids – While some data suggest that a more prolonged course of glucocorticoid administration might decrease paclitaxel-associated acute pain syndrome, they are insufficient to support a recommendation for this approach at this time and we suggest not pursuing this approach.

There are conflicting data regarding a potential effect of glucocorticoids in attenuating acute neuropathy related to paclitaxel or docetaxel:

•In 1999, a study was published that involved 46 patients receiving paclitaxel who, despite the use of nonsteroidal anti-inflammatory medications, had developed substantial trouble related to taxane-induced acute pain syndrome (labeled as arthralgias/myalgias in those days). With the next dose of paclitaxel, they were also given low-dose oral prednisone (10 mg twice daily, starting 24 hours after the completion of chemotherapy and continued for a total of five days). In this uncontrolled trial, 39 of 46 patients (85 percent) reported substantially less pain than they had with the first cycle [44]. (See 'Acute neurotoxicity' below.)

•Twenty years later, in 2019, another manuscript was published. This one reported on a retrospective study that was designed to evaluate risk factors for taxane-induced acute pain syndrome in 121 breast cancer patients receiving nanoparticle albumin-bound paclitaxel (nabpaclitaxel) every three weeks. There appeared to be a significant protective effect from the administration of dexamethasone for up to three days after chemotherapy (odds ratio [OR] 0.133, 95% CI 0.0235-0.7450) [45]. Of note, while all patients received dexamethasone on day 1, only 5 of the 121 patients actually received dexamethasone on days 2 and 3.

•Benefit for prolonged dexamethasone was also suggested in a subsequent retrospective study of 30 patients receiving every-three-week paclitaxel plus carboplatin for a variety of malignancies who also received dexamethasone 8 mg daily on days 2 and 3 for antiemetic prophylaxis; their outcomes were compared with a cohort of 30 comparable patients treated with this regimen in the years prior to the routine use of prolonged dexamethasone at the treating institution [46]. The frequency of all-grade acute pain syndrome was approximately 70 percent; it was similar in both groups. However, there was a lower incidence of severe symptoms (grade 2 or worse, 14 versus 40 percent) and a shorter duration of pain (5.8 versus 4.3 days) in the dexamethasone-treated patients.

•A 2021 publication evaluated, in a prospective multicenter randomized study, the addition of tapering dexamethasone (4 mg daily for two days followed by 2 mg daily for two days) after dexamethasone premedication (8 mg twice daily for three days) or dexamethasone premedication alone in 130 breast cancer patients commencing docetaxel treatment [47]. They concluded that their data did not provide substantial support for the use of tapering dexamethasone for the trying to prevent acute docetaxel neuropathy.

While some of the above data suggest that a more prolonged course of glucocorticoid administration might decrease taxane-associated acute pain syndrome, they are insufficient to support a recommendation for such. Given the known toxicities of dexamethasone and the lack of good data to support its use in this situation, we recommend that it not be used for this. Of note, this topic was not mentioned in either the first or second versions of ASCO CIPN guidelines because the ASCO guidelines only look at prospective randomized trial data, and the last trial described above was published after the last ASCO guideline.

●NSAIDs – Etoricoxib, a selective COX-2 inhibitor nonsteroidal anti-inflammatory drug (NSAID) was studied for preventing the taxane-associated acute pain syndrome in a phase II randomize trial [48]. Although prophylactic etoricoxib was associated with a decreased incidence and degree of the taxane-associated acute pain syndrome, the control group did not receive a placebo. This drug is not approved in the United States (licensing application was rejected in 2007 because of the risk of myocardial infarction, strokes, and deaths) although it is available by prescription in Canada and in many other countries. Additional study is needed before it can be concluded that this is a beneficial prophylactic treatment.

Agents which are not recommended — The following agents are not recommended for prophylaxis of CIPN, a position that is supported by ASCO and joint ESMO/EONS/EANO guidelines [11,12].

Acetyl-L-carnitine — The chemoprotective effect of acetyl-L-carnitine has been studied in patients treated with cisplatin [49,50] and taxanes [50-52]. Given a report from a randomized placebo-controlled trial that prophylactic use of acetyl-L-carnitine actually worsened taxane-related neuropathy [52,53], use of this agent should be specifically avoided [11,12,54].

Anticonvulsants

●Two trials addressing the neuroprotective benefit of carbamazepine/oxcarbazepine anticonvulsants in patients treated with oxaliplatin have come to opposite conclusions as to benefit [55,56].

●The utility of other anticonvulsants, such as gabapentinoids, has also not been established:

•Two small placebo-controlled randomized trials of pregabalin suggest that it is of no benefit; one was conducted in patients receiving paclitaxel, and the other was conducted in patients treated with oxaliplatin [57,58].

•Potential benefit for gabapentin was suggested in a small 40-patient placebo-controlled randomized trial conducted in women receiving adjuvant paclitaxel chemotherapy for breast cancer [59]. The experimental group received gabapentin 300 mg orally three times daily on days 1 through 14 of every 21-day treatment cycle. Compared with placebo, the rate of patient-reported grade 2 and 3 neuropathy was significantly lower in the gabapentin group, and there was less decline in nerve conduction velocity (NCV) after four cycles of paclitaxel in this group as well (eg, 18 versus 61 percent decline in NCV for the sural nerve). However, given the small sample size, the use of an outdated method of giving adjuvant chemotherapy, the use of clinician-reported rather than patient-reported outcomes, as well as questions as to the statistical analysis, we view this trial as less than definitive, and additional larger prospective studies are needed before concluding that there is merit to this approach.

Larger definitive studies are needed to confirm efficacy and clarify risks. The updated 2020 ASCO and joint ESMO/EONS/EANO guidelines concluded that clinicians should not offer these agents to prevent CIPN [11,12].

Antidepressants — Three trials investigating the neuroprotective effects of antidepressants on CIPN came to disparate conclusions:

●In a double-blind trial, 114 patients beginning chemotherapy with vinca alkaloids, platinum derivatives, or taxanes were randomly assigned to amitriptyline (100 mg daily) or placebo for the duration of chemotherapy. The severity of neuropathy was scored using the qualitative National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) scale, and patients graded neuropathic symptoms on a numeric visual analog scale twice weekly [60]. With short median follow-up (median 21 months for the amitriptyline group and 19 months for the controls), there were no significant differences between the groups.

●On the other hand, benefit for prophylactic venlafaxine was noted in a double-blind trial in which 48 patients with oxaliplatin-induced acute neurotoxicity were randomly assigned to venlafaxine (37.5 mg extended release twice daily from day 2 to 11) or placebo [61]. Neurotoxicity was assessed using a numeric rating scale for pain intensity and other neuropathic symptoms (using the neuropathic pain symptom inventory), and a qualitative oxaliplatin-specific neurotoxicity scale of Levi [62].

Significantly more patients treated with venlafaxine reported full relief of acute neurotoxicity (31 versus 5 percent), and at three months, when all patients were no longer receiving oxaliplatin, significantly fewer patients receiving venlafaxine had chronic neuropathy (rates of grade 3 neurotoxicity were 0 versus 33 percent), and significantly more had no neuropathy (39 versus 6 percent).

However, benefit for venlafaxine could not be shown in a similarly designed pilot randomized trial of venlafaxine to prevent neurotoxicity in 50 patients treated with oxaliplatin [63]. Although there was a trend toward benefit for venlafaxine when evaluated by the oxaliplatin-specific neuropathy scale [64] and by some acute neuropathy measures (throat discomfort and discomfort swallowing cold liquids) for the first two oxaliplatin doses, these trends were outweighed by the lack of any trend towards benefit in all other assessments, including the CIPN20 sensory subscale, clinician-completed NCI CTCAE scores, or in the cumulative administered dose of oxaliplatin, which was identical in both arms.

The 2014 Cochrane meta-analysis of interventions to prevent neuropathy in patients receiving platinum-type agents did not address the benefits of antidepressants for CIPN [65]. The 2020 systematic review of neuroprotectants from ASCO concluded that larger definitive studies are needed to confirm efficacy and clarify risks, and that clinicians should not offer amitriptyline or venlafaxine for prevention of CIPN until additional supporting data become available [11]. This conclusion differs from the 2014 ASCO guideline, which had suggested potential benefit from venlafaxine [66].

Of note, based on animal tumor data suggesting that duloxetine may be useful for preventing CIPN, an Alliance cooperative group placebo-controlled trial has been initiated to test the value of this drug for preventing oxaliplatin-induced CIPN.

Chemoprotectants

Amifostine — Amifostine is a frequently studied neuroprotective agent, mostly in the distant past. Amifostine is an organic thiophosphate that theoretically diminishes chemotherapy-related neurotoxicity by donating a protective thiol group, an effect that is highly selective for normal but not malignant tissue [67].

Multiple randomized trials have addressed the benefit of amifostine in preventing neurotoxicity from platinum-compounds (both cisplatin and carboplatin) and taxanes [68-78] with mixed results.

A year 2014 Cochrane meta-analysis of interventions for preventing neuropathy caused by cisplatin and related compounds (including carboplatin, given in conjunction with paclitaxel) included seven amifostine trials [65]. In a pooled analysis of data from three trials with the outcome of NCI CTCAE ≥grade 2 neuropathy [72,73,78], amifostine showed a significantly reduced risk of developing neurotoxicity ≥grade 2 (table 3) (risk ratio [RR] 0.26, 95% CI 0.11-0.61). However, the authors concluded that the data were insufficient to recommend amifostine, in part because only one trial used quantitative objective measures of neuropathy as an endpoint.

In 2014, a systematic review of neuroprotectants from ASCO [66] of six trials of amifostine for prevention of taxane-associated neuropathy [68,69,71,73,74,77] concluded that evidence of benefit was inconsistent across studies, counterbalanced by toxicities such as nausea, vomiting, and lightheadedness, and that clinicians should not offer amifostine for the prevention of CIPN to patients with cancer undergoing treatment with neurotoxic agents. This position was reiterated in the updated year 2020 ASCO guidelines [11], and also in the 2020 joint ESMO/EONS/EANO guidelines [12].

Nimodipine — Nimodipine is a calcium channel antagonist that provides protection from cisplatin-related neuropathy in animal models. Benefit in humans could not be confirmed in a double-blind, placebo-controlled trial of nimodipine in 51 patients with ovarian cancer who started treatment prior to initiating cisplatin, which was prematurely terminated because of increased nausea/vomiting and poor treatment compliance [79]. An analysis of neurotoxicity scores in 40 patients revealed significantly worse outcomes in the nimodipine group, suggesting that the drug exacerbated rather than prevented neurotoxicity.

The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN concluded that nimodipine should not be offered for prevention of CIPN to patients receiving treatment with neurotoxic agents [11], a position that was also taken by the 2020 joint ESMO/EONS/EANO guidelines [12].

RhuLIF — A single randomized, double-blind phase II trial that randomly assigned 117 patients receiving carboplatin and paclitaxel to placebo, low-dose, or high-dose recombinant human leukemia inhibitory factor (rhuLIF) failed to show any differences in standardized composite peripheral nerve electrophysiology scores, the primary endpoint, or in any secondary neurologic variables [80].

The updated 2020 ASCO systematic review concluded that rhuLIF should not be offered for prevention of CIPN to patients with cancer who are receiving treatment with neurotoxic agents [11].

Neurotropin — A possible protective effect for neurotropin was suggested in a small randomized phase II pilot trial in which 80 patients receiving oxaliplatin for stage II or III colon cancer were randomly assigned to neurotropin (nerve growth factor) or placebo [81]. As assessed by NCI CTCAE v4.0 neurotoxicity grading, there were no differences in the rate of grade 1 neurotoxicity (100 percent in each group), but significantly more patients in the control group had grade ≥2 neurotoxicity (61 versus 21 percent). Confirmation of these data in a larger prospective study with objective quantitative endpoints is needed.

Diethyldithiocarbamate — Diethyldithiocarbamate (DDTC) blocks cisplatin-induced toxicities in animal models without interfering with antitumor activity. Unfortunately, benefit in patients treated with cisplatin could not be shown in a single placebo-controlled randomized trial of 221 patients with lung or ovarian cancer who were receiving cisplatin [82]. Furthermore, patients receiving DDTC received lower cumulative doses of cisplatin, being more likely to be withdrawn early due to chemotherapy-related toxicity.

A year 2014 Cochrane meta-analysis of interventions for preventing neuropathy caused by cisplatin and related compounds included this trial and concluded that methodologic issues (subjective reporting of neuropathy using qualitative measures, and the elimination of data from patients withdrawn for toxicity, specific request, and adverse experience) rendered the trial results inconclusive [65]. The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN concluded that DDTC should not be offered for prevention of CIPN to patients with cancer receiving treatment with neurotoxic agents [11], a position also endorsed by the 2020 joint ESMO/EONS/EANO guidelines [12].

ACTH analog ORG 2766 — The efficacy of the ACTH (corticotropin) analog ORG 2766 for prevention of cisplatin- or vincristine-associated neuropathy has been tested in six randomized trials [83-88]. A systematic review from ASCO [66] of all six trials concluded that the four smallest, all from the Netherlands, suggested benefit, but that two larger well-conducted trials found no evidence of benefit [83,85] and, in fact, one suggested that neuropathy might in fact be worse in the intervention group [83].

A year 2014 Cochrane meta-analysis of interventions for preventing neuropathy caused by cisplatin and related compounds (including carboplatin, given in conjunction with paclitaxel) [65] included four of the ORG 2776 trials [83,86-88]; they concluded that the overall efficacy results were negative. The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN concluded that ORG 2766 should not be offered for prevention of CIPN to patients receiving treatment with neurotoxic agents [11].

Calmangafodipir — A placebo-controlled, three-arm phase II trial evaluated calmangafodipir in subjects receiving oxaliplatin [89]. There was enough of a suggestion of benefit to lead to the initiation of patient accrual on two subsequent phase III trials (NCT04034355, and NCT03654729).

The updated 2020 ASCO and joint ESMO/EONS/EANO guidelines concluded that calmangafodipir should not be offered to patients to prevent CIPN [11,12]. Subsequently, updated results of the trial revealed that there was more neuropathy in the treatment arm than in the placebo arm, likely related to an unfavorable interaction between this drug and oxaliplatin [90].

L-carnosine — A nutraceutical product, L-carnosine, was evaluated in a randomized trial evaluating it versus a no-treatment control arm in patients receiving oxaliplatin [91]. The study reported very positive results, but such results need to be tempered given that no placebo was used.

Until additional data are available, the updated 2020 ASCO guidelines concluded that L-carnosine should not be offered to patients to prevent CIPN [11].

Metformin — Metformin was studied in a 40-patient randomized study (n = 40) as a neuroprotectant for oxaliplatin-induced neuropathy [92]. Grade 2 to 3 neuropathy (table 3) was less in the metformin group (60 versus 95 percent, p = 0.009), and this arm had better Functional Assessment of Cancer Therapy-Gynecologic Oncology Group Neurotoxicity subscale (FACT-GOG NTX)-12 scores (24.0 versus 19.2, p <0.001).

Until more information is available, the updated 2020 ASCO guidelines concluded that clinicians should not offer metformin for prevention of CIPN in patients receiving treatment with neurotoxic agents [11].

Vitamins, minerals, and dietary supplements — There is no consistent evidence to indicate benefit for any vitamin, mineral, or dietary supplement as a preventive strategy for CIPN.

●Calcium and magnesium infusions – Based on an early report suggesting benefit for intravenous calcium and magnesium (IV Ca/Mg) prior to and immediately following oxaliplatin infusion in patients with advanced disease [93], placebo-controlled phase III trials were initiated in patients receiving oxaliplatin for metastatic colorectal cancer (the CONcePT trial [94]) and in the adjuvant setting (the N04C7 trial [95] and two other smaller trials [96,97]). However, an analysis of the first 180 patients enrolled in the CONcePT trial found significantly lower response rates in patients treated with IV Ca/Mg compared with the control group [98], and this led to premature closure of these trials. The lower response rate in patients with metastatic colorectal cancer who were receiving IV Ca/Mg in conjunction with oxaliplatin was not subsequently confirmed [94,99]. Three of these prematurely closed trials did not show any significant neuropathy benefit from Ca/Mg [94,96,97], while one [95] suggested significantly less acute neurotoxicity with the use of Ca/Mg supplementation and fewer patients with ≥grade 2 chronic neurotoxicity.

The issue of neuroprotection with IV Ca/Mg was directly addressed in United States Intergroup trial N08CB, in which 353 patients with resected colon cancer undergoing adjuvant therapy with FOLFOX (oxaliplatin plus short-term infusional fluorouracil and leucovorin) were randomly assigned to one of three arms: IV Ca/Mg (1 g calcium gluconate, 1 g magnesium sulfate) before and after oxaliplatin, IV Ca/Mg before and placebo after, or placebo both before and after oxaliplatin [100]. The primary endpoint was cumulative sensory neurotoxicity as assessed by the sensory subscale of the European Organisation for Research and Treatment of Cancer (EORTC) QOL questionnaire to assess chemotherapy-induced peripheral neuropathy (QLQ-CIPN20) [101]; secondary endpoints included the NCI CTCAE for neurotoxicity (table 3) and an oxaliplatin-specific neurotoxicity scale; acute neuropathy data were also collected for five days after each oxaliplatin dose. There was no benefit for IV Ca/Mg in preventing or diminishing the severity of acute or chronic neurotoxicity, and supplementation did not allow for higher delivered doses of oxaliplatin or a lower chemotherapy discontinuation rate.

A year 2016 systematic review of calcium and magnesium infusions to prevent oxaliplatin-induced CIPN concluded that there was no benefit to supplemental calcium and magnesium in this setting [102]. The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN also concluded that supplemental calcium and magnesium supplementation are of no benefit and should not be offered to patients receiving oxaliplatin therapy [11], a position that was also taken in the joint ESMO/EONS/EANO guidelines [12].

●Glutamic acid derivatives – Glutamic acid is a nonessential amino acid that can be converted to form glutamine or levo-glutamine by endogenous reaction with ammonia NH2. Glutamate is a salt or anion of glutamic acid that can also be converted to glutamine by endogenous reaction with NH2. All three chemically-distinct substances have been studied as neuroprotectants.

Trials examining the neuroprotective benefits of supplementation with glutamine and glutamate have been conducted in patients treated with oxaliplatin, taxanes, and vincristine [103-106]. Results have been mixed:

•One trial conducted exclusively in patients receiving oxaliplatin suggested that oral glutamine significantly reduced the incidence and severity of peripheral neuropathy, but it was neither blinded nor placebo-controlled [103].

•A pilot trial investigating long-term supplementation with glutamate for preventing CIPN in 43 women with ovarian cancer receiving paclitaxel plus carboplatin only demonstrated benefit in the severity of pain sensation [104].

•Two trials addressing the use of prophylactic glutamic acid in patients receiving vincristine have been completed, and both suggest a modest degree of benefit:

-In the first trial, 87 patients with breast cancer receiving weekly vincristine were randomly assigned to glutamic acid (500 mg three times a day during the entire six weeks of treatment) versus placebo [106]. Patients receiving glutamic acid developed significantly fewer moderate to severe subjective paresthesias (19 versus 36 percent), and there were fewer patients who lost their Achilles tendon reflex during treatment (19 versus 42 percent). There was also less clinician-assessed neurotoxicity (as determined by a neurotoxicity score derived by adding the grade of each neurotoxic parameter for the weekly clinic visit in which maximum neurotoxicity occurred); score ≥6 in 21 versus 43 percent of the glutamic acid and placebo groups, respectively. However, there were no significant differences in the rates of constipation or loss of other deep tendon reflexes. Differences in the need for dose reduction due to treatment-related neurotoxicity and long-term neurotoxicity were not addressed.

-In the second trial, 94 children receiving vincristine for a childhood malignancy were randomly assigned to glutamic acid (500 mg three times daily for the four weeks during induction therapy with vincristine) versus placebo [105]. The patients, parents, and nurses were blinded as to randomization, but not the assessing clinicians. The onset of neurotoxicity was significantly earlier in the placebo group for loss of deep tendon reflexes, paresthesias, and an increased frequency of constipation. At week 4, mild or moderate paresthesias developed in 7.4 and 1.9 percent of the glutamate group versus 20 and 7.5 percent of the placebo group. Although there were no significant differences in overall constipation, by week 4, the number of patients with severe constipation was 0 versus 10 percent in the glutamic acid and placebo groups, respectively. As with the earlier study, differences in the need for dose reduction due to treatment-related neurotoxicity and long-term outcomes were not addressed.

Although these results suggest some benefit for prophylactic use of glutamic acid in patients receiving vincristine, the benefits are modest, and whether this translates to improved long-term neurologic outcomes or better disease control because of fewer patients requiring dose reduction is unclear. Data regarding the potential benefit of glutamine (a different product from glutamic acid, despite their similar names) treatment for patients with vincristine-related neuropathy are discussed below.

The use of glutamine in patients receiving platinum and related compounds was not addressed in the 2014 Cochrane meta-analysis [65]. The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN concluded that until larger definitive studies are available, clinicians should not offer glutamate for prevention of CIPN to patients receiving treatment with neurotoxic agents [11].

●Glutathione and N-acetyl cysteine – At least seven small randomized trials have addressed the protective effect of glutathione against CIPN with a platinum agent [107-113]. A 2014 Cochrane review of three of these trials with an endpoint of ≥grade 2 neurotoxicity according to the NCI CTCAE grading scale [107,110,111] concluded that glutathione significantly reduced neurotoxicity (RR 0.29, 95% CI 0.10-0.85). However, the authors concluded that glutathione could not be recommended as a neuroprotectant given the limited number of patients enrolled on these trials (n = 387), the small number of actual neurotoxicity events (11 versus 34 in the treatment and control groups, respectively, in the three trials subjected to meta-analysis), and the lack of objective quantitative neurotoxicity endpoints in all of the trials.

A later larger trial addressing benefit in 185 patients treated with paclitaxel/carboplatin failed to demonstrate any benefit for glutathione [114]. Given that carboplatin is the least neurotoxic of the platinum drugs, it would appear that most of the neuropathy from this regimen was provided by paclitaxel, not platinum agents, as described in the smaller trials discussed above.

One other pilot randomized study addressed the value of N-acetyl cysteine, an antioxidant known to increase serum glutathione concentrations, in 14 patients receiving oxaliplatin [115]. As with other trials testing the utility of glutathione, treatment-related toxicity was assessed using the NCI CTCAE grading criteria. After 12 cycles of therapy, one of the five patients receiving N-acetyl cysteine developed ≥grade 2 sensory neurotoxicity compared with eight of nine placebo-treated patients.

The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN concluded that glutathione should not be offered for prevention of CIPN to patients receiving treatment with paclitaxel plus carboplatin, and that N-acetyl cysteine should not be offered to patients receiving potentially neurotoxic chemotherapy [11]. This position was also taken in the 2020 joint ESMO/EONS/EANO guidelines [12].

●Goshajinkigan (Gosha Jinki Gan) – Several published trials examining the benefit of a Kampo medicine (Goshajinkigan, Gosha-Jinki-Gan) for prevention of chemotherapy neuropathy have had disparate results [116-119].

A meta-analysis of five trials of Goshajinkigan (397 patients) for prevention of CIPN found that when evaluated using the neurotoxicity criteria of Debiopharm [120], Goshajinkigan was associated with a reduced incidence of grade ≥1 (RR 0.43, 95% CI 0.27-0.66) and grade 3 (RR 0.42, 95% CI 0.25-0.71) CIPN, but not grade ≥2 CIPN [121]; there was no benefit when neuropathy was assessed using the NCI CTCAE grading scale (table 3), and no improved response to chemotherapy. The authors concluded that Goshajinkigan was unlikely to prevent CIPN in patients receiving neurotoxic chemotherapy.

Until further information is available, the updated 2020 ASCO guidelines concluded that clinicians should not offer Goshajinkigan for prevention of CIPN in patients receiving treatment with neurotoxic agents [11], a position that was also taken in the 2020 ESMO/EONS/EANO guidelines [12].

●Omega-3 fatty acids and alpha lipoic acid – A single small, placebo-controlled, double-blind trial examined the neuroprotective benefit of omega-3 fatty acids (640 mg three times daily during chemotherapy) in 57 patients with breast cancer treated with taxanes [122]. Clinical and electrophysiologic studies (both performed by the same neurologist who was blinded as to the treatment group) were combined to provide a "total neuropathy score." Peripheral neuropathy developed during therapy in 9 of the 30 patients taking the fatty acids, compared with 16 of the 27 patients in the control group (30 versus 59 percent, OR 0.3, 95% CI 0.10-0.88). Among the patients who did develop neuropathy, there were no significant differences in severity between the groups. These benefits have not been independently confirmed in larger groups.

On the other hand, a placebo-controlled (but nonblinded) trial of alpha-lipoic acid in 243 patients receiving oxaliplatin or cisplatin-based chemotherapy regimens failed to demonstrate any benefit in preventing neurotoxicity, as assessed by qualitative neurotoxicity scales [123].

The updated 2020 ASCO systematic review of neuroprotectants for prevention of CIPN concluded that clinicians should not offer omega-3 fatty acids or alpha lipoic acid for prevention of CIPN to patients receiving treatment with neurotoxic agents [11], a position also taken in the joint ESMO/EONS/EANO guidelines [12].

●Vitamin E – Six trials of varying methodologic quality have evaluated the neuroprotective effect of vitamin E (300 to 800 mg daily) in patients treated with taxanes or platinum agents [124-128]:

•A protective effect of vitamin E intake during and after cisplatin-based chemotherapy was suggested in three very small randomized trials, two of which focused exclusively on cisplatin [124-126]. Two of these used an open-label control group without a placebo [124,125]. The third trial randomly assigned 108 patients receiving cisplatin for a variety of cancers to vitamin E (400 international units daily during and for three months following the discontinuation of cisplatin) or placebo [126]. Neurotoxicity was prospectively assessed prior to, during, and one month following the completion of therapy by a neurologist who was blinded as to the randomization, and by electrophysiologic testing prior to and following treatment. A cumulative neurotoxicity score was assigned to each patient based upon neuropathic signs and symptoms and electrophysiologic changes. The analysis was limited to the 41 patients (17 vitamin E and 24 placebo) who received at least 300 mg/m² of cisplatin. Patients who received vitamin E had significantly lower total neuropathy scores (1.4 versus 4.1) and a significantly lower incidence of grade 3 neurotoxicity (6 versus 42 percent). However, there were only 17 analyzed patients randomly assigned to vitamin E.

•A very small open-label trial of vitamin E (300 mg twice daily) versus no vitamin E in 37 patients receiving paclitaxel also suggested benefit [127]. Five patients withdrew early, and only 32 were evaluable for efficacy. The primary endpoint was clinical evaluation of paclitaxel-induced peripheral neuropathy (based upon a neurologic symptom score and neurologic disability score), and a secondary endpoint was a modified peripheral neuropathy score (mild 1 to 11; moderate 12 to 23; severe >24) consisting of combined clinical and electrophysiological evaluation. There was a significantly lower incidence of peripheral neuropathy in those receiving vitamin E (19 versus 63 percent), and there was also a significant difference in the modified peripheral neuropathy score (mean 2.25 [range 0 to 15] versus 11 [range 0 to 32], p = 0.01). Supplementation was well tolerated and had an excellent safety profile.

•A randomized study of 65 patients receiving oxaliplatin demonstrated no benefit for 400 mg per day of vitamin E [129].

•Additionally, the largest double-blind, placebo-controlled trial of 207 patients (38 percent treated with taxanes, 31 percent platinum-related agent, and 11 percent a combination) failed to find any protective effect of vitamin E supplementation (400 mg twice daily) [128]. A total of 189 patients were eligible for analysis. There was no significant difference between treatment arms in the incidence of clinician-rated ≥grade 2 sensory neuropathy (34 versus 29 percent for vitamin E and placebo, respectively), the primary endpoint, and no differences in the time to neuropathy onset, need for chemotherapy dose reductions due to neuropathy, or in patient-reported neuropathy assessments.

A 2016 meta-analysis that included all but the last of the published vitamin E trials concluded that, overall, there was a nonsignificant trend for less neuropathy in all of the studied patients (p = 0.07) and that there appeared to be a possible benefit for vitamin E in patients receiving cisplatin (p = 0.0002) [130]. However, until additional data are available, the updated 2020 ASCO and joint ESMO/EONS/EANO guidelines concluded that clinicians should not offer vitamin E as a neuroprotectant agent to patients receiving potentially neurotoxic chemotherapy [11,12].

●Vitamin B – Benefit for oral B group vitamins could not be shown in a randomized trial of 47 patients who received oxaliplatin, vincristine, or a taxane [131]. The year 2020 ASCO and joint ESMO/EONS/EANO guidelines concluded that clinicians should not offer vitamin B as a neuroprotectant agent to individuals receiving potentially neurotoxic chemotherapy [11,12].

●Minocycline – Two small randomized trials have failed to demonstrate benefit for minocycline in patients receiving oxaliplatin [132] or paclitaxel [133]. The year 2020 ASCO and joint ESMO/EONS/EANO guidelines concluded that clinicians should not offer minocycline as a neuroprotectant agent to individuals receiving potentially neurotoxic chemotherapy [11,12].

●All-trans retinoic acid – Retinoic acid reduces chemotherapy-induced neuropathy in animal models. Benefit in humans was suggested in a double-blind trial in which 95 patients receiving cisplatin plus paclitaxel chemotherapy for non-small cell lung cancer were randomly assigned to all-trans retinoic acid (ATRA) 20 mg daily or placebo [134]. As assessed immediately after the second cycle of chemotherapy (median follow-up three weeks), ATRA was associated with a borderline significant reduction in axonal degeneration (as demonstrated by NCV studies), and there was a trend toward less ≥grade 2 neuropathy as assessed qualitatively by the NCI CTCAE grading scale (56 versus 75 percent, p = 0.056). Confirmatory trials are not available.

This trial was included in the 2014 Cochrane meta-analysis, and the authors concluded that results from the nerve conduction studies were uninterpretable and that follow-up was inadequate to assess benefit [65]. The year 2020 ASCO guidelines concluded that clinicians should not offer ATRA as a neuroprotectant agent to individuals receiving potentially neurotoxic chemotherapy [11].

Potentially beneficial measures for specific drugs

Oxaliplatin

Limiting exposure to cold — The acute neurotoxicity that is seen frequently in the 72 to 96 hours after each infusion of oxaliplatin is often linked to cold exposure (drinking cold liquids, inhaling cold air, placing hands in the freezer). Avoidance of cold during this time frame should mitigate this toxicity to some extent, but not all symptoms (eg, perioral numbness, hand cramping) are related to cold.

However, there are emerging data and anecdotal experiences to support, but not prove, that using cryotherapy at the time of oxaliplatin administration may actually decrease both acute and chronic neuropathy symptoms. In one report, cryotherapy administered to the hands and feet during the infusion seemed to be as well tolerated in patients receiving oxaliplatin as it is in those receiving paclitaxel [18]. In addition, anecdotal reports have noted that patients receiving oral cryotherapy during oxaliplatin administration appeared to have less trouble with oral cold intolerance. Further data are necessary to determine whether cryotherapy is effective in patients receiving oxaliplatin. (See 'Cryotherapy, compression therapy, or both' above.)

Stopping and reintroducing oxaliplatin — As the clinical situation permits, if significant neuropathy develops during treatment, we suggest discontinuing the oxaliplatin and switching to a non-oxaliplatin-containing "maintenance" chemotherapy regimen to permit as much recovery as possible before reintroducing oxaliplatin if/when the cancer progresses on the non-oxaliplatin therapy.

Interspersing a non-oxaliplatin-containing "maintenance" chemotherapy regimen is a reasonable maneuver to prevent or delay the development of neuropathy in responding patients who have received three or four months of therapy with oxaliplatin but have no clinically significant neuropathy. Data from the OPTIMOX-1 and CONcePT trials in patients undergoing palliative chemotherapy for metastatic colorectal cancer suggest that this strategy decreases the risk of severe neuropathy without compromising antitumor efficacy. However, continued treatment with oxaliplatin is also an option in this setting, particularly in a responding patient who is tolerating chemotherapy well and has aggressive and/or bulky disease. Totally chemotherapy-free intervals have the potential to worsen outcomes in this setting and are not recommended.

The role of interrupted treatment schedules of oxaliplatin is discussed separately. (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Patients receiving oxaliplatin'.)

Lengthened infusion duration — The dose-limiting, cumulative neurotoxicity is not influenced by the duration of infusion or fractionation and is only dependent on the cumulative dose administered.

Lengthening the duration of the oxaliplatin infusion from two to six hours was evaluated in a randomized trial, in which 64 patients receiving adjuvant chemotherapy for colorectal or gastric carcinoma were randomly assigned to six- or two-hour infusions of oxaliplatin [135]. The overall percentage of patients with sensory neurotoxicity was not significantly decreased with the six-hour infusion (84 versus 93 percent with the two-hour infusion), although there was a significant decrease in the number of treatment cycles with grade 2 or greater neurotoxicity (6 versus 19 percent).

Combined with the lack of clear effect on cumulative neurotoxicity, the utility of this approach is limited by the logistical issues associated with a prolonged infusion and the lack of evidence as to whether this approach negatively influences the antitumor activity of oxaliplatin.

Bortezomib — Weekly, rather than twice-weekly treatment schedules, and subcutaneous, as compared with intravenous, administration of bortezomib are associated with less frequent and less severe neurotoxicity. We recommend subcutaneous rather than intravenous administration of bortezomib. We suggest weekly rather than twice weekly administration for most patients unless an urgent treatment response is needed initially. (See "Multiple myeloma: Treatment of first or second relapse", section on 'Refractory to lenalidomide and daratumumab' and "Multiple myeloma: Administration considerations for common therapies", section on 'Proteasome inhibitors'.)

In a phase III trial directly comparing weekly versus twice-weekly administration, the risk of grade 3 or 4 neurotoxicity was significantly less with once-weekly use (8 versus 28 percent), and the rate of treatment discontinuation because of peripheral neuropathy was also lower (5 versus 15 percent) [136].

Subcutaneous administration of bortezomib eliminates the high peak drug levels that occur with intravenous use. In a phase III trial directly comparing subcutaneous versus intravenous bortezomib (with a twice-weekly administration schedule), subcutaneous administration was as effective as intravenous use, and peripheral neurotoxicity of any grade (38 versus 53 percent), ≥grade 2 (24 versus 41 percent), and ≥grade 3 (6 versus 16 percent) were all significantly less with subcutaneous administration [137,138].

Paclitaxel — All patients initiating treatment with paclitaxel should be screened for vitamin D deficiency, and repleted if levels are low.

Although there are no prospective trials examining whether administration of vitamin D prevents paclitaxel-related neuropathy, at least two studies support the view that low vitamin D levels are a risk factor [139,140]. In a preliminary report of a retrospective analysis of data from the phase III SWOG0221 trial, which compared different paclitaxel-containing regimens in patients with early stage breast cancer, 169 of 1116 patients (15.1 percent) experienced peripheral neuropathy and 376 (33.7 percent) had vitamin D deficiency (defined as a level ≤20 ng/mL) [140]. Vitamin D deficiency was associated with a higher neuropathy risk (19.3 versus 13.0 percent). The results were even more striking in Black Americans. Compared with White Americans, Black Americans were more likely to have vitamin D deficiency (78 versus 29 percent), and they also had a more than twofold higher risk of paclitaxel-related peripheral sensory neuropathy (29.3 versus 13.3 percent, OR 2.66, p <0.001). Adjusting for vitamin D deficiency decreased but did not eliminate the higher neuropathy risk in Black Americans. (See "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Clinical features, incidence, and risk factors'.)

Vincristine — Because of the high incidence of constipation, patients receiving vincristine should take prophylactic stool softeners and/or laxatives. (See "Chemotherapy-associated diarrhea, constipation and intestinal perforation: pathogenesis, risk factors, and clinical presentation", section on 'Vinca alkaloids'.)

To minimize the potential neurotoxic effects of vincristine, many (but not all) protocols recommend an upper limit of 2 mg on single 1.4 mg/m² doses, regardless of body surface area. (See "Treatment protocols for lymphoma" and "Overview of neurologic complications of conventional non-platinum cancer chemotherapy", section on 'Vincristine'.)

TREATMENT

Chronic neurotoxicity

Adjustments to drug dose, administration, or regimen — In general, patients with mild CIPN can continue to receive full doses; however, if symptoms increase in severity or the neuropathy interferes with function, the risk of potentially disabling neurotoxicity must be weighed against the benefit of continued treatment. For patients who develop more severe CIPN during active treatment, clinicians should assess and discuss with patients the appropriateness of dose reduction, delay, discontinuation, or the use of intermittent treatment schedules. Treatment delays or dose reduction can improve symptoms in some cases, notably with taxanes and bortezomib. Published guidelines for dose reduction for specific agents are presented elsewhere. (See "Overview of neurologic complications of platinum-based chemotherapy" and "Overview of neurologic complications of conventional non-platinum cancer chemotherapy".)

Sometimes, a patient can be switched to an alternative less neurotoxic agent, if one is available (eg, substitution of carboplatin for cisplatin). (See "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy", section on 'Carboplatin-based regimens' and "Initial management of advanced non-small cell lung cancer lacking a driver mutation", section on 'Carboplatin typically preferred over cisplatin'.)

An important point is that clinicians and patients may make different decisions for continuing neurotoxic chemotherapy based upon whether treatment is being administered with curative intent (eg, in the adjuvant setting), or in a palliative setting for incurable metastatic disease.

For patients with advanced colorectal cancer who develop oxaliplatin-related neuropathy, it may be reasonable to discontinue oxaliplatin temporarily while maintaining a fluoropyrimidine with or without bevacizumab (ie, intermittent therapy). Another option is to switch to an alternative non-oxaliplatin-containing regimen. (See "Systemic therapy for metastatic colorectal cancer: General principles", section on 'Continuous versus intermittent therapy'.)

In other settings, (eg, adjuvant therapy for stage III colon cancer), a shorter duration of oxaliplatin-based therapy (ie, three versus six months) may provide similar efficacy while reducing the risk of CIPN. (See "Adjuvant therapy for resected stage III (node-positive) colon cancer", section on 'Duration of therapy'.)

Symptomatic treatment

Physical therapy and rehabilitation — For patients with an abnormal gait and/or mobility deficits, referral for rehabilitation and targeted therapeutic exercise can enhance postural control and balance, essential elements in preventing falls. (See "Physical rehabilitation for cancer survivors", section on 'Neuropathy'.)

Specific interventions that are recommended — For patients with cancer who are experiencing CIPN, we suggest duloxetine. There is insufficient evidence to support a recommendation for any other treatment.

Duloxetine — Benefit for duloxetine in patients with painful CIPN was initially demonstrated in a multi-institutional, double-blind cross-over trial in which 231 patients with taxane- or platinum-related painful CIPN (59 percent attributed to oxaliplatin) and at least grade 1 sensory neuropathy and an average pain score >4/10 that persisted for three or more months, with stable use of analgesics, were randomly assigned to duloxetine (30 mg daily for one week, then 60 mg daily for four additional weeks) or placebo [141]. Pain score was measured weekly by the Brief Pain Inventory, Short Form (BPI-SF); other endpoints included quality of life (QOL) as assessed by the Functional Assessment of Cancer Therapy-Gynecologic Oncology Group Neurotoxicity (FACT-GOG NTX) subscale [142] and the European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 [143] instruments, as well as nonpainful symptoms. After the initial five weeks, there was a taper and washout period for two weeks, and all patients crossed over to the alternative treatment.

Individuals receiving duloxetine during their initial five-week period had a significantly larger average decrease in mean pain score than did those who initially received placebo (mean decrease 1.06 versus 0.34, p = 0.003). Furthermore, patients randomized to duloxetine also had a greater degree of improvement in functional and QOL scores, and more patients reported improved numbness and tingling in the feet (41 versus 23 percent) but not the hands (36 versus 34 percent). Using this schedule, the drug was well tolerated with no significant differences in toxicities noted between active and placebo arms. Exploratory subgroup analysis suggested that duloxetine might be more efficacious for oxaliplatin-induced rather than taxane-induced painful neuropathy. In a later exploratory analysis, oxaliplatin-treated patients with painful neuropathy and better baseline emotional functioning (less anxiety and depression) were four times more likely to respond to duloxetine [144].

The benefit of duloxetine for treating chemotherapy neuropathy has been addressed in three additional small trials:

●One trial randomized CIPN-suffering patients to receive venlafaxine, duloxetine, or a placebo [145]. Decreased neuropathy was seen in the venlafaxine and duloxetine groups, with more reduction in the duloxetine group versus the venlafaxine group (p <0.05).

●Additionally, in a small randomized Japanese trial of 34 patients with CIPN after oxaliplatin, paclitaxel, vincristine or bortezomib, duloxetine outperformed vitamin B12 [146].

●On the other hand, duloxetine was inferior to pregabalin in a randomized double-blind phase II trial of 80 patients with taxane-related CIPN [147].

Given these results, the updated 2020 ASCO guidelines suggest that clinicians may offer duloxetine to patients with CIPN [11], and 2020 joint ESMO/EONS/EANO guidelines recommend duloxetine for treatment of neuropathic pain in this setting [12]. Despite the recommendation for use of duloxetine, the magnitude of benefit is modest and much less than is desirable.

Interventions that may be beneficial that have limited harms — Although proof of benefit has not been provided, data suggestive of benefit support that three integrative medicine approaches (scrambler therapy, acupuncture/acupressure/reflexology, and exercise) may diminish established CIPN symptoms and appear to be reasonably safe. Further research is needed to better delineate the utility of these approaches before it can be concluded that they are or are not of benefit. Nevertheless, given the myriad benefits associated with exercise, including its potential to mitigate CIPN, it is reasonable to suggest exercise to patients receiving potentially neurotoxic chemotherapy. Given the low potential for harm, a trial of acupuncture, acupressure, or reflexology is reasonable for patients who desire it; however, treatment should end after a few sessions, unless there is demonstrable benefit. Where available, a trial of scrambler therapy is reasonable for patients who desire it.

Exercise — Three small trials support the view that exercise may be helpful for patients with established CIPN:

●In one trial, 45 patients suffering from CIPN were randomized to usual care versus a 10-week home-based muscle-strengthening and balancing exercise program [148]. The exercise group had a significant reduction in neuropathic pain scores (p <0.0001) and improvement in functional QOL (p = 0.0002), symptom QOL (p = 0.0003), and Global Health Status QOL (p = 0.004).

●In another small trial, patients with metastatic colorectal cancer were randomized to be in a waitlist control group versus to be in an exercise group [149]. The CIPN in the waitlist control group worsened, while it improved in those assigned to get exercise.

●An additional trial randomly assigned 50 cancer survivors with persisting CIPN endurance plus balance training versus endurance training only [150]. Endurance training reduced sensory symptoms in both groups, although neuropathy was not the primary outcome.

Despite these results, the updated 2020 ASCO guideline concluded that outside of the context of a clinical trial, no recommendation can be made about the use of exercise to treat CIPN [11]. Having said this, recommending exercise for patients is generally considered to be a good thing. Joint ESMO/EONS/EANO guidelines take the stance of recommending physical exercise when potentially neurotoxic chemotherapy is initiated rather than waiting until the onset of manifest CIPN [12].

Acupuncture — The available data are mixed with acupuncture:

●At least eight small trials suggest benefit from acupuncture [151-158].

●One trial of electroacupuncture was positive [159], while two (including one that utilized a sham control [35]) were negative [35,160].

●Two systematic reviews have also come to opposite conclusions:

•A year 2019 systematic review of three clinical trials (203 patients) concluded that there was insufficient evidence to recommend acupuncture for treatment of CIPN [161].

•On the other hand, a second systematic review and meta-analysis of six randomized trials (360 patients) concluded that acupuncture led to modest but significant improvements in pain scores as assessed by the mean change in the Brief Pain Inventory-short-form (BPI-SF) and/or Functional Assessment of Cancer Therapy-neurotoxicity (FACT-NTX) score from baseline (-1.21, 95% CI -1.61 to -0.82) [162].

A major problem is that there is a paucity of well-conducted, rigorous trials that include a sham acupuncture control group, which is necessary because of the strong placebo effect of both acupuncture and sham acupuncture. (See "Overview of the clinical uses of acupuncture", section on 'Research challenges'.)

Nevertheless, the limited data do not disprove the possibility of benefit for individual patients. Three available guidelines have come to different conclusions:

●The year 2020 ASCO guideline on CIPN concluded that no recommendation could be made about the use of acupuncture in patients with CIPN [11].

●A year 2020 joint ESMO/EONS/EANO guideline took a bit more positive standpoint and suggested consideration of acupuncture to treat CIPN symptoms [12].

●A year 2022 joint guideline on integrative therapies for pain in oncology patients from ASCO and the Society for Integrative Oncology (SIO) concluded that acupuncture may be offered to patients experiencing pain from CIPN, given that benefits likely outweigh harms, although characterizing the evidence quality as low, and the recommendation as weak [163].

Acupressure and reflexology — In contrast to acupuncture and electroacupuncture, which use thin needles to stimulate pressure points, acupressure and reflexology are both body work techniques which involve applying pressure to specific points on the body; acupressure involves the whole body, while reflexology involves the feet, hands, ears, and face [164]. Additional descriptions of these practices are provided separately. (See "Overview of complementary, alternative, and integrative medicine practices in oncology care, and potential risks and harm", section on 'Acupuncture, acupressure, and reflexology'.)

Two small trials with approximately 30 patients per arm have evaluated the effectiveness of reflexology for reducing CIPN symptoms [165,166].

●One trial in patients with multiple cancers compared reflexology foot massage twice daily for 20 minutes over six weeks compared with standard hospital care found improvements in sensory function but no improvement in pain related to CIPN [165]. Side effects were not reported.

●A second trial in 63 patients with gynecologic cancer tested a self-care foot reflexology approach with aromatherapy, three times weekly for 15 minutes on each foot, 18 sessions over a period of six weeks versus a waitlist control; the intervention group had significantly lower levels of CIPN symptoms, including pain, and less interference with daily activities. Side effects were not reported [166].

The joint guideline on integrative therapies for pain in oncology patients from ASCO/SIO concluded that reflexology or acupressure may be offered to patients experiencing pain from CIPN, given that benefits likely outweigh harms, although characterizing the evidence quality as low, and the recommendation as weak [163].

Scrambler therapy — A device that delivers patient-specific electrocutaneous stimulation to the skin (called "scrambler therapy") has shown promise for treatment of neuropathic pain, including that associated with CIPN. Nonrandomized pilot trials suggest that scrambler therapy can reduce chemotherapy-induced neuropathy symptoms, even if symptoms have been present for >1 year [167-169]. Based on these trials, two prospective, randomized clinical trials were conducted to evaluate this approach for patients with established chronic CIPN, with disparate results:

●In the first trial, 50 patients with CIPN symptoms for at least three months and CIPN-related tingling or pain at 4 of 10 in severity the week prior to registration were randomly assigned to scrambler therapy versus transcutaneous electrical nerve stimulation (TENS) for two weeks [170]. Patient-reported outcomes were used to measure efficacy daily for two weeks during therapy and for eight weeks thereafter. Twice as many scrambler-treated patients had at least a 50 percent documented improvement from their baseline pain (tingling and numbness scores during the two treatment weeks of 36 to 56 percent compared with 16 to 28 percent in the TENS group). Differences between the groups were less apparent but still evident toward the end of the eight-week observation period. There were no apparent toxic effects.

●On the other hand, a second sham-controlled trial of scrambler therapy failed to discern benefit [171]. Thirty-five patients with cancer, CIPN of >3 months' duration, and pain or other CIPN symptoms (tingling, numbness) with an average daily rating of at least 4 of 10 were randomly assigned to 10 30-minute sessions of scrambler therapy on the dermatomes above the painful areas or to sham treatment on the back, typically at L3 to L5 (for leg pain) and C5 to C8 (for arm pain), where the nerve roots would enter the spinal cord. The primary endpoint was "average pain" after 28 days on a 10-point numeric rating scale. There were no significant differences between the two groups at days 10, 28, 60, or 90 with regard to average pain, scores on the CIPN-modified BPI, or scores on the motor and sensory subscales of the EORTC CIPN-20 instrument. Although individual responses were noted during scrambler treatment, most dissipated by day 30. The primary author of this trial feels that the scrambler treatments were not given as well as might be done and that improved scrambler treatment approaches do appear to provide significant benefit.

In total, scrambler therapy is an emerging treatment approach that appears to benefit some affected patients with CIPN, with minimal adverse effects [172]. Although updated 2020 ASCO guidelines concluded that a recommendation could not be made regarding the use of scrambler therapy [11], while stating that a trial of this relatively nontoxic approach, where available, is reasonable for patients who desire it. Joint ESMO/EONS/EANO guidelines, on the other hand, were more negative on this approach, specifically not recommending it [12].

Other interventions that remain investigational

Tricyclic antidepressants — Two small trials have failed to demonstrate benefit from the tricyclic antidepressants nortriptyline and amitriptyline for treatment of chemotherapy-induced neuropathic symptoms [173,174]:

●In a randomized, double-blind cross-over trial, 51 patients with cisplatin-induced painful neuropathy were randomly assigned to nortriptyline (escalating doses to a target maximum of 100 mg daily) or placebo [173]. After four weeks, there was a one-week washout period, and the patients switched to the alternative treatment. As assessed by a QOL questionnaire and visual analog scales, there were no significant differences between nortriptyline and placebo.

●Benefit for amitriptyline could also not be shown in a second double-blind trial in which 44 patients with neuropathic symptoms of severity ≥3/10 after treatment with a vinca alkaloid, platinum derivative, or taxane were randomly assigned to low-dose amitriptyline (starting at 10 mg daily, with escalating doses up to 50 mg daily) [174]. Efficacy was assessed by the EORTC QOL questionnaire (EORTC-QLQ-C30) and the neuropathic Pain Symptom Inventory. The study was prematurely terminated due to poor recruitment; amitriptyline provided no significant benefit for sensory neuropathic symptoms.

Updated 2020 guidelines from ASCO conclude that outside of the context of a clinical trial, no recommendation can be made about the use of tricyclic antidepressants to treat CIPN [11]. This conclusion differs from that in the 2014 guidelines, which suggested that tricyclic antidepressants were a reasonable choice for a therapeutic trial [66]. The 2020 joint ESMO/EONS/EANO guidelines, however, suggest tricyclic antidepressants as a reasonable option if duloxetine has failed or contraindications are present [12].

Oral mucosal cannabinoid extract — A randomized double-blind study of 18 patients with CIPN failed to show any benefit from use of nabiximols, an oral mucosal spray containing cannabinoids [175]. The updated 2020 ASCO guideline concluded that a recommendation could not be made as to the utility of nabiximols for treatment of CIPN [11].

Gabapentinoids — The available evidence from randomized trials does not support benefit from gabapentin or pregabalin in CIPN, despite their efficacy in treating other forms of neuropathy.

●Gabapentin – The following data are available:

•A potential benefit for gabapentin for CIPN was originally suggested in an ASCO abstract in 2000 [176]. The abstract included seven patients receiving oxaliplatin who developed neuropathy, and were started on gabapentin 100 mg twice a day. The dose could be increased to 100 mg three times a day if symptoms persisted. It was reported that neuropathy symptoms resolved despite patients receiving up to 14 doses of oxaliplatin. This work was never published as a manuscript, and it does not appear very believable that this low dose of gabapentin was as efficacious as was reported.

•Much higher doses of gabapentin were evaluated in a double-blind, placebo-controlled cross-over trial in which 115 patients with symptomatic CIPN and pain scores ≥4 on a 0 to 10 pain scale or neuropathy ≥1 on a 0 to 3 neuropathy scale were randomly assigned to gabapentin (target dose 2700 mg/day in three divided doses) versus placebo [177]. After the initial six weeks, cross-over occurred following a two-week washout period. Efficacy was assessed using a patient-reported average pain score per day, using a numerical rating scale, and using the qualitative Eastern Cooperative Oncology Group (ECOG) neuropathy scale. Changes in symptom severity were similar in both groups, and benefit for gabapentin could not be demonstrated.

•A small randomized placebo-controlled trial of gabapentin in approximately 50 children with vincristine-related neuropathic pain was negative [178]. The relatively low dose (mean 18 mg/kg per day) may have been contributory.

•A retrospective report included two sequential cohorts of patients who received similar oxaliplatin treatments for metastatic colorectal cancer; one cohort received gabapentin (initial dose 300 mg daily) to try to prevent neuropathy while the other did not [179]. The gabapentin dose in the second cohort was allowed to be increased to 600 mg three times daily. Results from this work revealed that similar degrees of neurotoxicity were seen on both arms and that there were no differences in the relative dose intensities of oxaliplatin, with similar oxaliplatin dose reductions and delays in the two groups. These authors concluded that gabapentin was not helpful for preventing/treating oxaliplatin-associated neuropathy.

●Pregabalin – The available data on pregabalin are as follows:

•A randomized, double-blind placebo-controlled trial to assess the efficacy of pregabalin in prevention/treatment of CIPN in patients with advanced colorectal cancer was terminated early when an interim analysis found that the conditional power to detect a difference in treatment groups was insufficient to warrant study continuation [180], suggesting that pregabalin is ineffective for the prevention/treatment of CIPN. While the details of this trial are not clear (the data do not appear to have been published), it appears that this was designed as a prevention trial in patients starting oxaliplatin. However, if there is no decrease in CIPN with a neuroprotective agent in a prevention trial setting, the agent would not likely be helpful in the treatment setting, given that it did not effectively treat the CIPN that developed. Additionally, another prospective, pilot, placebo-controlled trial evaluating pregabalin as a potential agent to decrease paclitaxel-induced neuropathy did not provide sufficiently promising data to proceed with a more formal phase III trial [57].

•Investigators from the Czech Republic reported on the use of pregabalin in 30 children (mean age of 13.5 years) with CIPN in an open-label trial design [181]. These patients had received a variety of neurotoxic chemotherapy drugs and had a pain score of 4 out of 10 when they entered the trial. The mean visual analog scale score decreased by 59 percent during eight weeks of pregabalin treatment, and 86 percent of the evaluable patients had long-lasting pain relief. It could be argued that unblinded data in children may be even more suspect to bias than in adults.

•Between 2012 and 2014, five manuscripts written in Japanese, with English-language abstracts, pertained to pregabalin therapy for chemotherapy-induced neuropathy. The first two of these were case reports that noted an improvement in chemotherapy neuropathy in single patients treated with pregabalin, both written by the same group of authors [182,183].

In 2013, another report noted that 13 patients, suffering from oxaliplatin-induced sensory neuropathy, were treated with pregabalin [184]. It appears that this report is based on a review of clinical records, as opposed to the result of a prospective trial experience. The authors noted improvement in 8 of the 13 that they attributed to pregabalin.

Another report in 2013, by Nihei et al, also appeared to be generated from a clinical practice review [185]. This report evaluated the use of pregabalin in 27 patients with oxaliplatin-induced neuropathy and 28 patients with paclitaxel-induced neuropathy, and compared these patients with others who were treated with drugs other than gabapentinoids. Overall, 41 percent of patients with oxaliplatin-induced neuropathy, and 29 percent of patients with paclitaxel-induced neuropathy responded to therapy with pregabalin, defined by a decrease of one grade of neuropathy. This was higher than what was observed in patients who received non-pregabalin drugs (10 and 12 percent, respectively).

•In the last of these five manuscripts, Nihei and colleagues described their use of pregabalin as the first-line treatment for oxaliplatin-induced neuropathy, in what appears to be a prospective trial format [186]. They reported that pregabalin was helpful in 33 percent of patients.

The updated 2020 ASCO guideline concluded that no recommendations could be made as to the use of gabapentinoids for treatment of CIPN [11]. This recommendation differs from the 2014 recommendation which suggested that gabapentinoids might be helpful for treatment of CIPN [66]. The 2020 joint ESMO/EONS/EANO guidelines suggest anticonvulsants such as gabapentinoids as a reasonable option if duloxetine has failed or contraindications to duloxetine use are present [12].

Some insurance companies require that patients with CIPN receive a gabapentinoid agent before allowing the use of duloxetine [187]. This practice is not consistent with ASCO guidelines [11].

Glutamine — Modest benefit for glutamine supplementation was suggested in a small randomized trial of young patients who developed neuropathy (sensory or motor) while receiving vincristine treatment for a variety of malignancies [188]. Upon identification of neuropathy, patients were randomly assigned to glutamine (6 g/m² per dose [maximum 10 g/dose]) twice daily or placebo for 21 days followed by a 21-day washout period. The main endpoints were motor and sensory neuropathy scores from the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) v3, neuropsychological assessment scores, and QOL assessment at three time points: baseline, following the initial 21-day supplementation period, and after the 21-day "washout" period. A significantly higher number of children progressed on the sensory neuropathy scale in the placebo compared with the glutamine group between baseline and the end of the 21-day supplementation period (p = 0.02), but this difference disappeared by the 21-day washout period. There were no differences in neuropsychiatric testing at any time point, and the beneficial impact of glutamine on self-reported QOL was only observed after the washout period. Data were not provided on the total vincristine dose received in either group or whether or not the use of glutamine prevented dose reductions or improved the outcome of therapy. While these data are intriguing, further studies are warranted to confirm the efficacy of glutamine for the treatment of vincristine-related neuropathy. This use was not considered in the ASCO guideline [11], and joint ESMO/EONS/EANO guidelines concluded that no recommendation for its use in adults could be made [12].

Topical treatments containing amitriptyline and ketamine with and without baclofen — The North Central Cancer Treatment Group (NCCTG) conducted a randomized trial in which 208 patients with CIPN (mostly after treatment with taxanes or oxaliplatin) were randomly assigned to topical treatment with a compounded pluronic lecithin organogel containing 10 mg of baclofen, 40 mg of amitriptyline, and 20 mg of ketamine in each 1.3 g measured dose versus placebo gel [189]. The primary endpoint was the sensory subscale of the EORTC QOL questionnaire (EORTC QLQ-CIPN20) at four weeks. Patients in the treatment group experienced a slight but statistically nonsignificant improvement in their sensory neuropathy over those treated with placebo (p = 0.053). However, there was a significant improvement in the motor subscales. The greatest improvements were related to symptoms of tingling, cramping, and shooting/burning pain in the hands and difficulty holding a pen.

On the other hand, benefit could not be shown for a slightly more potent cream containing 2 percent ketamine and 4 percent amitriptyline without baclofen in a double-blind, placebo-controlled trial involving 462 patients with CIPN from taxane or non-taxane chemotherapy of severity >4 on a scale of 1 to 10 [190]. Patients completed a seven-day daily pain, numbness and tingling diary one week prior to study entry, and at three and six weeks after enrollment; the average score at six weeks was the primary outcome. Twice daily use of the cream was not associated with a decrease in CIPN symptoms.

The updated 2020 ASCO guideline concluded that outside of the context of a clinical trial, no recommendation could be made about the use of topical amitriptyline and ketamine with or without baclofen [11]. This recommendation differs from the 2014 guidelines, which had noted that this approach was "promising" [66]. On the other hand, year 2020 joint guidelines from ESMO/EONS/EANO specifically recommend against use [12].

The use of these topical treatments for other types of neuropathic pain related to cancer is discussed elsewhere. (See "Cancer pain management: Role of adjuvant analgesics (coanalgesics)", section on 'Topical therapies'.)

Topical menthol — Case reports describe rapid symptomatic improvement in painful neuropathy symptoms using twice daily application of topical menthol (0.5 percent in calamine lotion) associated with bortezomib [191] and carboplatin [192], an approach that has been useful for the treatment of postherpetic neuralgia [193]. These findings were subsequently confirmed in a successful proof of concept study for topical 1 percent menthol in 51 patients with cancer-related nerve pain, predominantly CIPN (n = 35). Of the 38 evaluable patients, 31 (82 percent) had an improvement in BPI scores, while improvements in walking ability and sensation were also observed [194].

The use of topical menthol was not addressed in the updated 2020 ASCO guidelines [11], but the joint ESMO/EONS/EANO guidelines state that despite the lack of data from randomized trials, use of topical menthol could be "considered" given its low cost and lack of adverse effects [12].

Topical capsaicin — One pilot trial, involving 16 patients, supported that an 8% topical capsaicin cream product dramatically alleviated established CIPN [195]. Further work is necessary to confirm this result. The use of topical capsaicin was not addressed in the updated 2020 ASCO guidelines [11], but the joint ESMO/EONS/EANO guidelines state that use of capsaicin-containing patches could be "considered" in patients with CIPN [12].

Neurofeedback — An intriguing pilot study suggested potential benefit for electroencephalogram (EEG)-based neurofeedback. In this randomized controlled trial, 71 cancer survivors with CIPN (mostly females, predominantly treated with taxanes) were randomly assigned to EEG neurofeedback or to waitlist control [196]. Neurofeedback consisted of 20 sessions in which the participants watched and responded to their own EEGs while playing a game for 45 minutes per session; when they maintained their EEG waveform amplitude over a chosen threshold and inhibited less desirable waveforms, they were given rewarding feedback with a picture and a beep; the game paused when the participant did not match the thresholds preprogrammed into the software, and no auditory or visual feedback was given. Over time, the group completing the neurofeedback experienced significant improvements in pain scores when compared with the waitlist control group, and there were also improvements in numbness, symptom interference, physical functioning, and fatigue; these benefits persisted four months after treatment was completed. The treatment appeared to be very well tolerated, and there were no reports of adverse events.

The generalizability of these results is limited by the small patient sample, the lack of an active or sham control group, and the nonblinded assessment of outcomes. However, additional study of neurofeedback and other approaches to physiologic self-regulation (eg, relaxation training, yoga) are warranted.

Patients treated with bortezomib — Limited data suggest the possibility that bortezomib may cause an immune-mediated neuropathy with both motor and sensory involvement in addition to the usual neurotoxicity (a predominantly sensory neuropathy) and that this may be responsive to immunotherapy. A single report describes five patients with a peripheral neuropathy associated with severe motor involvement [197]. Electrodiagnostic studies indicated either a demyelinated or a mixed axonal-demyelinated neuropathy. The protein concentration in the cerebrospinal fluid was elevated, and magnetic resonance imaging (MRI) of the lumbar spine revealed enhancement of nerve roots in two patients. Four of the five patients appeared to have responded to intravenous immune globulin or glucocorticoids. These data are insufficient to base a recommendation for specific treatment.

Acute neurotoxicity — As noted above, acute neurotoxicity is predominantly seen in patients receiving oxaliplatin and the taxanes, paclitaxel and docetaxel. These are usually short-lived effects, and specific treatment may or may not be needed. (See 'Acute neurotoxicity' above.)

●Oxaliplatin – There are no known effective treatment approaches for acute oxaliplatin neurotoxicity. Preventive measures for oxaliplatin-induced neurotoxicity are discussed above. (See 'Oxaliplatin' above.)

●Taxanes – Taxane-induced acute pain syndrome may be quite debilitating, and consistently successful preventive agents/strategies have not been identified. (See 'Glucocorticoids and NSAIDs for taxane-associated acute pain syndrome' above.)

Optimal treatment is also not established. Nonsteroidal anti-inflammatory drugs (NSAIDs) are frequently used first-line, but the data to support benefit come entirely from anecdotal experience; there are no prospective trials [198-203]. Similarly, as discussed earlier in this section, reports on the benefits of glucocorticoid therapy are limited to anecdotal experience and uncontrolled series [44,199,203,204]. As an example, in 1999, a study was published that involved 46 patients receiving paclitaxel who, despite the use of nonsteroidal anti-inflammatory medications, had developed substantial trouble related to taxane-induced acute pain syndrome (labeled as arthralgias/myalgias in those days). With the next dose of paclitaxel, they were also given low-dose oral prednisone (10 mg twice daily, starting 24 hours after the completion of chemotherapy and continued for a total of five days). In this uncontrolled trial, 39 of 46 patients (85 percent) reported substantially less pain than they had with the first cycle [44].

Anecdotally, opioids do relieve the pain. (See 'Glucocorticoids and NSAIDs for taxane-associated acute pain syndrome' above.)

A systematic review of five studies (including two placebo-controlled randomized trials of glutathione or glutamine [114,205]) concluded that neither glutathione nor glutamine was superior to placebo.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Neuropathic pain" and "Society guideline links: Neuropathy" and "Society guideline links: Cancer pain".)

SUMMARY AND RECOMMENDATIONS

●Prevention

•There are no established agents that can be recommended for the prevention of chemotherapy-induced peripheral neuropathy (CIPN) in patients with cancer undergoing treatment with neurotoxic agents. Consistent with guidelines from the American Society of Clinical Oncology (ASCO), we recommend against the use of acetyl-L-carnitine (Grade 1B). We also suggest against the use of amifostine, calcium/magnesium for patients receiving oxaliplatin-based therapy, glutamic acid/glutamine, glutathione for patients receiving carboplatin/paclitaxel, Goshajinkigan, and vitamin E (Grade 2B). We suggest against the use of alpha lipoic acid, amitriptyline, calmangafodipir, cannabinoids, carbamazepine/oxcarbazepine, diethyldithiocarbamate, N-acetyl cysteine, nimodipine, omega-3 fatty acids, Org 2766, all-trans retinoic acid, recombinant human leukemia inhibitory factor (rhuLIF), venlafaxine, and vitamin B (Grade 2C). (See 'Agents which are not recommended' above.)

•No recommendations can be made regarding the benefit of acupuncture or ganglioside-monosialic acid (GM1) for preventing CIPN, or for the use of prolonged glucocorticoids to decrease paclitaxel or docetaxel-associated acute pain syndrome. (See 'Interventions with potential but unproven benefit' above.)

•Although proof of benefit has not been established, available data suggest that exercise, cryotherapy, and compression therapy with or without cryotherapy may, in part, prevent symptoms of CIPN, and appear to reasonably safe, although clinicians should be aware of frostbite risk. (See 'Cryotherapy, compression therapy, or both' above and 'Exercise' above and 'Acupuncture' above.)

•Oxaliplatin – For patients receiving oxaliplatin (see 'Oxaliplatin' above):

-When appropriate in the clinical setting, we suggest interspersing a non-oxaliplatin-containing "maintenance" chemotherapy regimen with the oxaliplatin regimen in patients undergoing palliative chemotherapy for metastatic colorectal cancer (Grade 2B).

-There are insufficient data to support a benefit from lengthening the duration of oxaliplatin infusion, and we suggest against its use (Grade 2B).

•Bortezomib – We recommend subcutaneous rather than intravenous administration (Grade 1A). In addition, we suggest weekly rather than twice-weekly administration for most patients unless an urgent treatment response is needed initially (Grade 2B). (See 'Bortezomib' above.)

•Vincristine – Because of the high incidence of constipation, patients receiving vincristine should take prophylactic stool softeners and/or laxatives. (See 'Vincristine' above.)

•Paclitaxel – Given available information from two independent database analyses, we suggest that patients initiating treatment with paclitaxel should be screened for vitamin D deficiency, and repleted if levels are low. (See 'Paclitaxel' above.)

●Treatment

•For patients with cancer who are experiencing CIPN following treatment with taxanes or platinum drugs, we suggest duloxetine (Grade 2B). We also suggest duloxetine for CIPN related to other neurotoxic chemotherapy agents (Grade 2C). (See 'Duloxetine' above.)

•There is insufficient evidence to support a recommendation for any other treatment including gabapentin/pregabalin. (See 'Interventions that may be beneficial that have limited harms' above.)

•Although proof of benefit has not been provided, data suggestive of benefit support that three approaches (scrambler therapy, acupuncture, and exercise) may diminish established CIPN symptoms and appear to be reasonably safe. Further research is needed to better delineate the utility of these approaches. Given the low potential for harm, a trial of acupuncture, acupressure, or reflexology is reasonable for patients who desire it. Similarly, although the data are conflicting, a trial of scrambler therapy, where available, is reasonable for patients with refractory symptoms who desire it. Exercise is also reasonable to recommend, based on preliminary evidence of benefit and its beneficial effects on other aspects of health. (See 'Exercise' above and 'Acupuncture' above and 'Scrambler therapy' above and 'Acupressure and reflexology' above.)