Summary

Supportive measures are discussed in Peripheral neuropathies: supportive measures and rehabilitation.

Chemotherapy-induced peripheral neuropathy is associated with a variety of compounds at conventional dose and others at high dose. Neuropathy is dose limiting in several, and the dose is carefully tracked in most instances. Either a single high dose or a cumulative dose over a treatment course may be critical for limitation of toxicity. Reversibility in most instances depends on the degree of axonal loss. Chemotherapy-induced peripheral neuropathy is shown to notably affect quality of life and increases fall risk 3-fold (185). The neuropathy severity also depends on inherent factors, such as preexisting neuropathy, medication metabolism, renal or hepatic function for drug clearance, and idiosyncratic reactions. The mechanisms of neurotoxicity vary among drugs and may or may not invoke the same mechanism as the intended therapeutic effect. Many agents, for example platins, interfere with axonal transport and produce axonal injury, but many other mechanisms are known or suspected, including ganglionic nerve cell toxicity.

Neuropathy severity grading is well accepted but skewed to more severe grades. There are several chemotherapy-induced peripheral neuropathy classification systems in use, all with similar descriptions, usually employing four categories (I through IV). Groups providing scales include the Eastern Cooperative Oncology Group (ECOG), National Cancer Institute (NCI), NCI-Canada Common Toxicity Criteria (NCIC-CTC), World Health Organization (WHO), and Ajani. NCI adds a 5th category – death. Grade I is generally thought to be tolerable and inconsequential but includes descriptions such as subjective but not objective weakness, paresthesia, and reduced or absent reflexes. Grade II, also considered to be mild and tolerable in some but not all studies, includes descriptions such as sensory alterations or paresthesia interfering with function, symptomatic weakness interfering with function but not activities of daily living, mild or moderate sensory loss and paresthesia, severe paresthesia, or mild weakness. Most neurologists would consider this group to have a significantly symptomatic neuropathy. The Total Neuropathy Score and Functional Assessment of Cancer-Gynecologic Oncology Group, neurotoxicity (FACT/GOG-Ntx) are more promising complimentary tools to assess chemotherapy-induced peripheral neuropathy severity and impact. The FACT/GOG-Ntx is a subjective measure of chemotherapy-induced peripheral neuropathy-related quality of life. Total Neuropathy Score clinical versions incorporate both subjective measures and objective examinations of nerve function (136). Awareness of the significance and persistence of these symptoms and their effects on quality of life is increasing (148). Grade III is generally considered to be significant and includes neuropathy that interferes with activities of daily living, is intolerable, requires bracing, and has other disabling features. Obviously, these classification schemes cloud older reports of neuropathy incidence, recovery, and severity. A table of the most important classes and agents is presented (Table 1).

The search for an effective preventative agent to limit or protect from neurotoxicity is a longstanding goal. Many agents show promise in cell culture or animal models but ultimately fail in human trials. Many attempted human trials are for already approved agents, supplements, and nontraditional therapies. One challenge is to enroll recently diagnosed cancer patients awaiting chemotherapy to enter a parallel trial that requires blinding, supplementary tests, potential side effects questionnaires, and additional visits. For that reason, many studies offered supplement, vitamin, and alternative medical therapies. Attempts to find a useful agent is discussed in appropriate subsections.

Table 1. Important Chemotherapy Classes and Agents

Taxanes

Historical note and nomenclature. Currently available taxanes include paclitaxel, docetaxel, and cabazitaxel, which are all potent chemotherapeutic agents. Paclitaxel was originally derived from the bark of the Western yew tree, Taxus brevifolia, requiring a process that killed the tree. Current production involves a plant cell fermentation process of a Taxus cell line. Docetaxel is a semisynthetic derivative of a precursor molecule isolated from the needles of the more common European yew tree, Taxus baccata. Taxanes are FDA approved for use in ovarian and non-small-cell lung cancers and are used for various other malignancies. Docetaxel is approved in Europe for hormone-refractory prostate cancer. Despite similarities, paclitaxel produces a more severe neuropathy than docetaxel (150). Cabazitaxel was approved in the U.S. in 2010 for advanced prostate cancer; however, the drug seems to have no reduction in neuropathy incidence. Fourteen percent of patients in the TROPIC trial developed neuropathy, but all had prior docetaxel (85); the incidence was similar in an earlier trial (269).

The vehicle polyoxyethylated castor oil may contribute to the toxicity of paclitaxel. Polyoxyethylated castor oil-free versions of paclitaxel are available and enable an increased dose of paclitaxel without increasing the toxicity (180); however, chemotherapy-induced peripheral neuropathy rates are similar. A variety of other novel taxanes are under development, but none offer proven reduction in neurotoxicity at present.

Clinical manifestations. Paclitaxel induces a progressive, predominantly sensory neuropathy. Symptoms often occur within days after the first dose and include pain and paresthesia; nearly all patients experience pain or discomfort by a third treatment cycle of paclitaxel (117; 159). Initial symptoms and signs are numbness and painful paresthesia of the hands and feet. Ankle jerks, and then other reflexes, may be lost. Transient myalgia is also fairly common after each dose and frequently affects the thigh muscles; this symptom usually resolves within days.

Sensory loss, which progresses with cumulative doses, occurs typically in a stocking-glove distribution, but more multifocal involvement can occur. Both small- and large-fiber sensory functions are affected, but clinical ataxia is rare. Perioral numbness can occur. Muscle strength is frequently preserved or only minimally affected. Progression is dependent on the total cumulative dose. Neuropathy is not usually the dose-limiting factor using conventional exposures, but nearly all patients will develop neuropathy at sufficient dose.

A predominantly proximal idiosyncratic muscle weakness is reported in a minority of patients receiving paclitaxel or docetaxel (118). The course is variable and reversible. Autonomic neuropathy can also develop with high-dose therapy or in patients with preexisting neuropathy. Cranial neuropathies, including bifacial neuropathy, are reported (202).

Docetaxel appears to have clinical manifestations that are similar to but milder than paclitaxel. Rare reports of Lhermitte sign and severe motor neuropathy with higher cumulative or individual doses of docetaxel exist (348; 115). Nab-paclitaxel is an FDA-approved albumin-bound, 130-nm particle form of paclitaxel that has lesser hypersensitivity reactions and seems to have a similar or possibly a slightly higher likelihood to induce neuropathy; however, comparatively higher paclitaxel doses are administered (132). Additional studies support the higher neuropathy incidence in the Nab-paclitaxel group, 10% versus 3% of grade 3 or 5 chemotherapy-induced peripheral neuropathy in a large cohort of over 1200 patients with breast cancer (346). A metanalysis of 19 clinical trials (2878 cancer patients) tracked neuropathy incidence in patients treated with Nab- and conventional paclitaxel; the overall incidence was 51% --high grade 12.4% (265).

Etiology and pathophysiology. The probable neuropathic effect of paclitaxel is related to its ability to cause accumulation and aggregation of microtubules (130). These stabilized microtubules are rendered nonfunctional. Taxanes bind to the beta subunit of tubulin. Promotion of assembly and inhibition of disassembly of microtubules could potentially affect axonal transport, axonal sprouting, or nerve regeneration. The distal symmetric symptoms suggest axonal dysfunction, but the simultaneous involvement of hands and feet as well as perioral numbness continues to suggest an additional dorsal root ganglia neuronopathy, but additional mechanisms of neurotoxicity are speculative. In experimentally induced paclitaxel neuropathy, development of mechanical allodynia and hyperalgesia may be mediated via protein kinase C epsilon (92). Detailed studies using quantitative sensory testing, sensory maps, and peg board testing suggest taxanes preferentially affect A-beta and A-delta myelinated fibers but spare unmyelinated C fibers (99). Nav1.7 is an important channel in nociceptive dorsal root ganglion pain signaling; mutations in this gene are associated with congenital pain disorders (305). Paclitaxel increases Nav1.7 expression in rat and human sensory neurons (04). The agent also affects Nav1.7 axonal trafficking and vesicular velocity, which was studied using an advanced optical imaging paradigm that can view channel-containing vesicles (04).

Epidemiology. The development and severity of paclitaxel neuropathy depends on the cumulative dose; however, toxicity occurs over a wide dosage range, reported to be as low as 5 mg/m2 (96). Additionally, individual doses greater than 250 mg/m2 will also produce symptoms. Neuropathy usually begins with cumulative doses over 1500 mg/m2. Other risk factors may include high infusion rates, preexisting neuropathy, and older age. When co-administered with cisplatin, both paclitaxel and docetaxel induce a dose-limiting neuropathy, which can develop rapidly and appear more severe than with either the taxane or cisplatin independently. Over 60% of patients receiving taxanes develop some sensory symptoms (grade 1 sensory neuropathy). The ICON4 trial reported that 20% of patients receiving paclitaxel and platinum-compound based chemotherapy will develop grade 2 or higher sensory neuropathy (262). Other combination therapies are also associated with neurotoxicity to some degree; however, it is difficult to tease out responsible agents versus combination effects. Doxifluridine, doxorubicin, epirubicin, gemcitabine, 5-fluorouracil, leucovorin, tegafur, ifosfamide, and trastuzumab are just a few examples. Genetic variations may also make taxane-induced neuropathy more likely. Single nucleotide polymorphisms (SNPs) are identified that may predispose to toxicity (319; 144). Genome-wide analyses of lymphoblastoid cell lines exposed to paclitaxel have found similar SNP profiles and identified that RFX2, which encodes a member of the DNA-binding regulatory factor, may be a specific marker of potential neurotoxicity (365). Some have proposed that neuropathy development may be a biomarker for improved survival despite dose limitations (247). However, more extensive studies found no evidence of efficacy differences between patients who developed neuropathy and others who did not (296). Single nucleotide polymorphisms (SNPs) may predispose to the development of neuropathy in certain patients. An ABCB1 3435 TT genotype and CYP3A4 392 AA/AG genotypes increase the risk of neuropathy in patients with breast cancer treated with paclitaxel and docetaxel (192), and a CYP3A4 defective variant is associated with development of paclitaxel-induced neuropathy (295). A case citing a possible interaction between clopidogrel and paclitaxel, whereby the first caused poor clearance of the latter, leading to a higher risk of neuropathy, is reported (27). An asymptomatic but underlying defect in Charcot-Marie-Tooth genes has been assessed. Boora and colleagues found an association between paclitaxel chemotherapy-induced peripheral neuropathy with three nonsynonymous, recurrent single nucleotide variants in the rare ARHGEF10 gene (36). The gene is typically associated with autosomal dominant severe hypomyelinating Charcot-Marie-Tooth disease. Similarly, a Mayo group studied 269 neurologically asymptomatic cancer patients receiving paclitaxel and enrolled in a large consortium (Alliance) while undergoing prospective neuropathy assessments; 49 Charcot-Marie-Tooth genes were analyzed. Increased mutations were found in the periaxin (PRX) and ARHGEF10 genes; in the latter, of several targets, the rs9657362 SNP had the strongest effect (odds ratio of 4.8, p=4×10−4) (30). Mitofusin 2 depletion also may play a role in paclitaxel-induced allodynia development in mice (373). A single nucleotide polymorphisms (SNP) SCN9A-rs13017637 was a significant predictor of grade 2 or higher paclitaxel chemotherapy-induced peripheral neuropathy in a cohort of Japanese breast cancer patients (326). In an international cohort, taxane-induced neuropathy symptoms appear to be more common in black patients than in white patients (41% vs. 23%) (256). Other than lymphedema, other toxicity conditions had a similar incidence. SET-binding factor 2 (SBF2) mutations are associated with an increased chemotherapy-induced peripheral neuropathy severity in African American patients (295). An SBF2 knockdown sensory stem cell model found enhanced paclitaxel injury to cell viability, neurite growth, and sodium current inhibition (81).

An Australian group studied the effect of weekly paclitaxel monotherapy in 83 breast cancer patients (337). A combination of functional assessments, total neuropathy score clinical measure, sural nerve conduction studies, and von Frey hairs were employed. About 60% of patients had numbness or tingling at the 6-week mark and 86% at week 12; many had associated functional impairment and reduced sural evoked amplitude. Half of the affected patients noted persistent symptoms 6 months after therapy completion; however, many improved after cessation in clinical and physiological measures. Of note, about half of patients had dose reduction or cessation because of chemotherapy-induced peripheral neuropathy symptoms or manifestations (337). A population-based cohort of 884 Swedish breast cancer survivors and 1768 controls was compared using a detailed mailed questionnaire asking about chemotherapy-induced peripheral neuropathy symptoms (108). Post-taxane average was 3.6 years. Both sensory and motor symptoms were more common in cancer patients; other common neuropathy risk factors, older age, paclitaxel exposure, and weight were independent predictors. Tingling and numbness in distal legs were most common in the chemotherapy group, but half or more reported some motor, sensory, or autonomic symptoms.

Docetaxel neuropathy also occurs over a wide range of individual doses (10 to 115 mg/m2) and cumulative doses (50 to 750 mg/m2) (254). Neuropathies have been reported with nonplatinum-based docetaxel-combination therapies including co-administration with capecitabine, gemcitabine, melphalan, ifosfamide, and epirubicin to name a few.

Discovery of useful biomarkers to identify CIDP risk factors are ongoing. Blood measures of neurofilament light chain level (sNfL) and selected other compounds are promising biomarkers for CIPN, notably for paclitaxel and platins (179; 237; 47; 21; 60).

An in vitro model found paclitaxel effects on the voltage-gated sodium channel NaV1.8, which is preferentially expressed in DRG neurons (19).

Prevention. There is currently no standard treatment for the prevention of paclitaxel neuropathy, but neuroprotection studies have uncovered some potential agents. None, however, are commonly employed at present. Most research in this field has concentrated on platin-induced neuropathy, discussed later. However, none of the agents to be discussed are used in oncology practice. Many preventative agents tested include benign compounds or vitamin supplements, which increase the chance of patient participation; examples include vitamin E and glutathione. Most human trials, even randomized ones, evaluated small and heterogeneous cohorts. Many agents that showed promise in cell culture or animal models are not tested in humans. A practice guideline of the American Society of Clinical Oncology concluded that no agent was recommended in 2014 despite review of 48 potential randomized trials (147).

The sterile alpha and TIR motif containing 1 (SARM1) gene is a prodegenerative mediator of axonopathy implicated in various forms of chemotherapy-induced peripheral neuropathy. Selective and irreversible isothiazole SARM1 inhibitors showed protective effects in rodent and human in vitro axon and sciatic axotomy models (38). A SARM1 knockout mouse model found gene-dose dependent prevention of tail sensory nerve axonal potential amplitude loss. Inhibitors minimized paclitaxel-induced intraepidermal nerve fiber density reduction and provided partial protection of axonal function assessed by sensory nerve action potential amplitude and mechanical allodynia (38). A Drosophila larva nociceptive neuron model found that paclitaxel reduced endosome-mediated trafficking of integrins; integrins are known to support neuronal maintenance and potentially protect against paclitaxel-induced cellular toxicity (301).

Acetyl-L-carnitine, a member of the family of carnitines that plays an essential role in intermediary metabolism, has shown a neuroprotective effect in animal models of paclitaxel and cisplatin-induced neuropathies without interfering with antineoplastic effects of either medication (127). Human trials also suggest a protective effect (33; 218). Enhancement of neuronal nerve growth factor response via histone acetylation, a mechanism involved in gene expression, is the proposed mechanism for limiting neurotoxicity. Acetyl-L-carnitine also plays a role in regulation of cellular levels of acetyl-CoA, and acetylation of tubulin is thought to contribute to neuroprotection (33).

Vitamin E supplementation has demonstrated partial neuroprotection in patients receiving paclitaxel, cisplatin, or their combination. In a pilot, randomized, open label with blind assessment, controlled trial, neurotoxicity occurred in 73% of controls but in only 25% of patients treated with 600 mg/day of vitamin E during chemotherapy and for 3 months following its cessation. Exact mechanisms of protection are unknown; however, prior studies have shown diminished vitamin E levels in cancer patients (14). The same authors performed a similar phase II trial recruiting only patients that received paclitaxel (15). The results were similar, but no other groups have attempted to replicate these results.

Amifostine, an approved hematopoietic progenitor cell protectant in bone marrow, had been shown to reduce neurotoxicity in early clinical trials (292). Proposed mechanisms for amifostine in neuroprotection include scavenging of free radicals, decreasing DNA-DNA interstrand crosslinks induced by alkylating agents, and decreasing the formation of platinum-DNA adducts. The protective effect in non-neoplastic cells is thought to be the result of more efficient alkaline phosphatase activity, resulting in increased dephosphorylation of amifostine to its active free thiol form. Available clinical trials, however, suggest a lack of neuroprotection in patients on paclitaxel regimens as measured by nerve conduction studies, quantitative sensory testing, and composite neuropathy scores. These results may not be unexpected because taxanes are thought to cause neuropathy by disrupting axoplasmic flow, whereas amifostine acts to protect DNA (259). Pretreatment with amifostine also appears to reduce the severity of neurotoxicity in patients treated with carboplatin and docetaxel (187). Other phase II trials, however, did not corroborate this effect (246).

Non-randomized, controlled studies have shown that glutamine given in combination with paclitaxel may have a protective benefit, resulting in less weakness, less loss of vibration sense, and less toe numbness than in controls (316). Pretreatment with oral glutamate in a rat model delayed the onset of neuropathy without affecting its antineoplastic efficacy (43). Glutamine, a neutral amino acid that serves as a substrate for nucleotide synthesis in dividing cells, plays a critical role in providing nitrogen precursors used in the synthesis of pyrimidines and purines. It becomes depleted in cancer patients as a result of cachexia. This depletion may significantly impact those non-neoplastic cells that are rapidly dividing (293). Further randomized controlled studies are needed to further elaborate the protective effects of glutamine (09).

Other agents that have prevented experimentally induced paclitaxel neuropathy include Org 2766, nerve growth factor, prosaptides, and neurotrophic peptides, which facilitate nerve regeneration (52). Two antibiotics, radicicol and geldanamycin, have been shown to protect against the neurotoxic effect of both paclitaxel and vincristine in dorsal root ganglion cultures (291). This neuroprotective effect appears to be mediated via the inhibition of heat shock protein 90, which is involved in the induction of apoptosis. Although animal studies suggested a protective role of recombinant human leukemia inhibitory factor (rhuLIF), placebo-based, randomized, controlled studies have not shown protection against paclitaxel or carboplatin-induced neuropathy in humans (84). PINK1 mitigates paclitaxel-induced sensory dendrite alterations and restores mitochondrial homeostasis in a Drosophila larvae model of chemotherapy-induced peripheral neuropathy (181). 3-hydroxyflavone (3HF) was found to suppress paclitaxel-induced mRNA expression of several inflammatory cytokines in a rat chemotherapy-induced peripheral neuropathy model, including tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), interleukin-6 (IL-6), calcitonin gene-related peptide (CGRP), and substance P. However, no behavior improvements were identified in the rats (345). Choline-fenofibrate treatment reduced or prevented appearance of paclitaxel-induced mechanical allodynia and sensory action potential reduction in a mouse model; various inflammatory markers were also improved (49). An Italian group studied the effect of human IVIg on paclitaxel-induced chemotherapy-induced peripheral neuropathy in rats. The treatment reduced behavioral markers of mechanical allodynia but did not mitigate sensory nerve conduction physiological measures (238). Nasal administration of mesenchymal stem cells is another approach that seems to limit chemotherapy-induced peripheral neuropathy in mice (40). The commonly used angiotensin receptor type 1 antagonist, losartan, also appears to attenuate mechanical allodynia and some inflammatory markers in a rat paclitaxel chemotherapy-induced peripheral neuropathy model (170).

A study found blocking interleukin-20-induced signaling in a mouse paclitaxel-induced peripheral neuropathy model can suppress neuroinflammation and restore calcium homeostasis, which appears to alleviate paclitaxel-induced peripheral neuropathy (71). Furthermore, pretreatment with an interleukin-6-neutralizing antibody resulted in the prevention of paclitaxel-induced neuropathy in mice (156). The AMP-activated protein kinase activators metformin and narciclasine were shown to reduce paclitaxel-induced mechanical hypersensitivity in a mouse model and these and other activators are potential prevention candidates (160).

A randomized, double-blind, placebo-controlled trial of 206 breast cancer patients treated with taxanes were treated with ganglioside-monosialic acid or placebo (318). For Functional Assessment of Cancer Treatment Neurotoxicity subscale score after four cycles of chemotherapy, the symptom scale findings were significantly lower in the treatment group. Lithium has been suggested as a preventative agent though it is known to cause toxic neuropathy. A brief, placebo-controlled trial using 5 days of 300 mg lithium for a taxane treatment cycle in 36 Iranian breast cancer patients found no significant benefit in neuropathy symptoms or electrophysiology rates (251). A randomized trial of 73 paclitaxel-treated breast cancer patients found significantly less (about half) developed grade 2 or worse CIDP when pretreated with metformin versus placebo (20). NfL levels also differed between treatment groups.

Alternative therapies are also of interest to many patients. A study of 180 patients treated with various neurotoxic chemotherapy agents were randomized to wear frozen gloves during chemotherapy to prevent neuropathy, though a third could not tolerate the intervention. Treated patients had reduced hand sensory symptoms but the overall chemotherapy-induced neuropathies scales were not significantly different (24). A similar study in Singapore evaluated 38 patients, though 80% interrupted therapy because of intolerance to the cold therapy (255). Some that tolerated the therapy noted modest benefits. Hilotherapy that provides a more constant cool temperature improved patient-reported neuropathic symptoms compared to frozen gloves (76). A meta-analysis of various cryotherapy modalities in 2250 patients in nine trials found that therapy could reduce grade 2 neuropathy incidence primarily from patient-reported outcomes but not motor findings (165). Patients treated with cryotherapy had less neuropathy-promoted taxane dose reduction.

However, a consensus review determined that no agents are currently proven or recommended to prevent chemotherapy-induced peripheral neuropathy. The group specifically discouraged N-acetyl cysteine use (211).

Differential diagnosis and workup. Patients with malignancies may have other unrelated causes of neuropathy that must be considered. Other common causes of neuropathy, such as diabetes, alcohol use, and vitamin B12 deficiency, might be the underlying etiology of the patient’s symptoms or potentially make patients more sensitive to the neuropathic effects of chemotherapy. A careful history must include questions about symptoms prior to the initiation of chemotherapy, concomitant medical illnesses, and concurrent medications. Other drug-induced neuropathies must be considered; patients with malignancies may take isoniazid, pyridoxine, dapsone, metronidazole, or nitrofurantoin for associated infections. Compromise of the patient’s nutritional status can lead to nutritional neuropathies (thiamine, cobalamin, pyridoxine deficiency, or impaired glucose tolerance). Paraneoplastic neuropathy is also a risk, especially in patients with small-cell lung cancer as well as breast, gynecologic, or gastrointestinal malignancies. Finally, perhaps 10% to 50% of patients with advanced cancer associated with cachexia develop a late sensorimotor neuropathy.

Electrophysiologic abnormalities in taxane neuropathy include reduced amplitude or absent sensory nerve responses initially, with motor responses following later. With continued progression, conduction velocities may be slowed but commensurate with axonal loss. Eventually, denervation potentials may be seen in distal muscles. If paraneoplastic or metabolic neuropathy is a possibility (especially with lung cancer), anti-Hu antibodies, B12 levels, thyroid function studies, and 2-hour glucose tolerance studies can be measured. CSF evaluation is occasionally helpful.

Sural nerve biopsy has shown variable results: nerve fiber loss without evidence of regeneration, axonal atrophy, or secondary demyelination (287); or axonal neuropathy with preferential large, myelinated fiber loss and evidence of axonal regeneration (115). The presence of regeneration may be variable depending on the timing of the biopsy, as it is known that recovery will eventually occur.

Prognosis. Paclitaxel neuropathy improves months after therapy is terminated, but resolution may not be complete. A coasting phenomenon may occur after discontinuation of therapy (352). Docetaxel neuropathy, being considerably milder, has a better prognosis for complete recovery.

Management. Quantitative sensory testing, specifically vibratory threshold, has been suggested as a useful monitoring technique, presaging the development of paclitaxel neuropathy but not docetaxel neuropathy (150). These techniques are still used for experimental tracking but not typically employed for routine clinical care. However, skin biopsy for epidermal nerve fiber density is increasingly utilized. If neuropathy is recognized early in paclitaxel treatment, changing to docetaxel can result in a decrease or resolution of the neuropathy (281). Monitoring vibration thresholds and deep tendon reflexes in patients on cisplatin and paclitaxel combination therapy may help predict the final severity of the neuropathy and may provide a gauge to adjust for toxicity (58). Typical agents to treat neuropathic pain (such as tricyclic antidepressants, duloxetine, pregabalin, tramadol, and gabapentin) are employed to treat the painful qualities of taxane and other chemotherapy-induced neuropathies. However, there are little published data of controlled trials. One study suggests that nortriptyline has, at best, a modest effect despite widespread use (140). Transdermal lidocaine and nutritional supplements such as evening primrose oil, alpha-lipoic acid, and capsaicin may also play a limited role in patients with chemotherapeutic neuropathies; however, controlled studies are needed (279). A systematic literature review of published randomized clinical trials found 17 that met review criteria (356). Included were some trials of agents not available in many countries such as crocin, GM1, and tetrodotoxin. Agents that showed some benefit over placebo were similar to non-chemotherapy-induced peripheral neuropathy cohorts, such as pregabalin, duloxetine, and venlafaxine. Case numbers were deemed too small in the gabapentin trial reviewed.

Although surgical interventions have reported some success in treating painful chemotherapy-induced neuropathies, anecdotal reports of improvement of neuropathic symptoms with multiple nerve decompressions should be considered with caution until more controlled studies are conducted (88).

Vinca alkaloids

Historical note and nomenclature. Vincristine, vindesine, and vinblastine are extracts of periwinkle plants that act as mitotic inhibitors. Vinblastine was first used in 1958 for Hodgkin lymphoma; vincristine was introduced in 1962 (200). Vincristine is now used in Hodgkin and non-Hodgkin lymphomas, acute leukemia, rhabdomyosarcoma, neuroblastoma, and Wilm tumor. Vinblastine is now also used in the treatment of non-Hodgkin lymphomas, mycosis fungoides, testicular carcinoma, Kaposi sarcoma, and histiocytosis X. Vinorelbine, a semi-synthetic vinca alkaloid, was introduced in 1985 and is approved for use for non-small cell lung carcinoma and has also been used in ovarian, cervical, and breast carcinomas (82).

Clinical manifestations. Neuropathy is the dose-limiting side effect of all vinca alkaloids, with vincristine being the most severe. Side effects usually occur by 4 to 5 weeks. Dysesthesia in the fingers may precede those in the feet.

Loss of ankle jerks is the initial sign and earliest manifestation, occurring by 2 to 3 weeks; total areflexia frequently develops. Pinprick and temperature sensations are the most affected, whereas vibration sense is less affected; proprioception often remains normal.

Distal symmetric motor weakness occurs with continued exposure and is considered the definitive dose-limiting side effect. Weakness may develop rapidly. Clumsiness of the hands or muscle cramps, especially after exercise, may be the harbinger of motor impairment. Footdrop or wristdrop can be initial manifestations.

Dysautonomia can occur within days of vincristine exposure. It often begins with constipation and may progress to paralytic ileus and include urinary retention, orthostatic hypotension, and impotence.

Cranial neuropathy (including extraocular muscles, laryngeal neuropathy, and optic neuropathy) has been reported but is considered rare (338). Esophageal motor dysfunction with dysphagia is also reported.

Etiology and pathophysiology. The cause of the neuropathy is most likely from impaired microtubule function involved in axonal transport. Both myelinated and unmyelinated fibers are affected. This effect also impairs regeneration of nerves injured by other causes, such as trauma (261).

The chemotherapeutic effect of the vinca alkaloids is via their high affinity binding to tubulin, which causes a disruption of the mitotic spindles and arrest of cells in metaphase. In an animal model of vincristine neuropathy, there is a clear disorganization of the axonal microtubule cytoskeleton, with accumulation of neurofilaments in the soma of spinal ganglion cells, suggesting impaired anterograde axonal transport (339). Exposure of neurites in vitro induces axonal degeneration that appears to be mediated via calcium-induced activation of the cysteine protease, calpain (359). Pathology shows primary axonal degeneration with atrophy of axons. Neuropathic pain in the animal model appears to be mediated by nociceptor hyper-responsiveness via second messenger pathways involving protein kinase A, protein kinase C, and nitric oxide (08). Second messenger signaling for primary afferent nociceptor sensitization may show sexual dimorphism in the development of neuropathic pain (168). An estrogen-dependent effect on protein kinase C epsilon may contribute to the development of vincristine-induced hyperalgesia. Serotonin transporter-deficient mice are relatively protected from mechanical allodynia and heat hyperalgesia in experimental vincristine neuropathy (141).

Vincristine-induced neuropathy appears to include inflammatory mechanisms. There is evidence that this process is driven by activation of the NLRP3 inflammasome (NLR family pyrin domain containing 3) and release of interleukin 1β (314). A patient with severe vincristine induced neuropathy was screened for genetic susceptibility traits (155). A complex pattern of polymorphisms was found in various genes, but the significance was not clear. Other genome-wide association studies of large cohorts using similar methods found no clear susceptibility genes. A study of neuropathy in 90 pediatric oncology patients that developed vincristine induced chemotherapy-induced peripheral neuropathy found seven polymorphisms in the cohort, some in known hereditary neuropathy genes (349). A group created a machine learning model using genome-wide association studies generated polymorphisms to predict which patients would develop vincristine-induced chemotherapy-induced peripheral neuropathy (371).

Enteric neuropathy is an increasingly recognized aspect of neuropathy in general and CIPN in some instances. The contractility of colonic smooth muscle strips from cisplatin or vincristine treated rats found significant loss of contractility in both that was more marked with cisplatin (210).

Epidemiology. Approximately one half of patients develop a sensorimotor neuropathy at normal doses of 1.4 mg/m2 per week (54). The interval between doses appears to be important in the development of neuropathy: intervals of 1 week are more toxic than 3 weeks, even at the same cumulative dose. Up to one third of patients develop autonomic symptoms. For vinorelbine, neuropathy, mostly mild, has been reported in 14% to 48% of patients receiving a total dose of 30 mg/m2 (368).

Treatment may uncover previously asymptomatic hereditary neuropathies, specifically Charcot-Marie-Tooth disease type IA and possibly hereditary neuropathy with liability to pressure palsies (133; 169), both of which are caused by defects in the PMP-22 gene. Such patients are predisposed to developing neuropathy with increased motor involvement at low or even single doses. Weakness is more pronounced in patients with preexisting peripheral neuropathy, but asymptomatic cases can be unmasked by vincristine treatment. Despite clear black box warnings, cases continue to emerge. Typically, hereditary neuropathy is not suspected in advance (182). In order to prevent this toxicity, medulloblastoma treatment regimens without vincristine were used in a known Charcot-Marie-Tooth 1A disease child (29). Most other forms of Charcot-Marie-Tooth likely have a proportionate risk based on the level of underlying neuropathy; however, a patient with a novel early growth response 2 (EGR2) gene defect who was previously asymptomatic is reported to have excessive sensitivity to vincristine therapy and unmasking of demyelinating neuropathy (252). Therefore, it is imperative to take a detailed personal and family history of patients and examine for overt signs of undiagnosed hereditary neuropathy prior to commencing vinca alkaloid chemotherapeutic regimens. A study found that a polymorphism in the CEP72 gene was associated with increased risk and severity of vincristine-associated neuropathy (93). Voltage-clamp recordings of small and medium dorsal root ganglion neurons from vincristine-treated mice that developed allodynia revealed a significant upregulation of TTX-S Na+ current in medium but not small neurons (70). The increase in TTX-S Na+ current density is thought to be mediated by Nav1.6 channels.

Vincristine-induced neuropathy is also evident in a significant percentage of treated children (341). Severity appears to be increased in children with acute lymphoblastic leukemia with associated hepatic involvement (344). Quality-of-life measures are also significantly worse in children with chemotherapy-induced peripheral neuropathy compared to others (350).

Prevention. There is no clearly established means of preventing the toxic neuropathy, but there are some promising possibilities. Given the severity and early development of neuropathy in patients with Charcot-Marie-Tooth disease, prescreening for Charcot-Marie-Tooth disease might be considered in patients with a family history of neuropathy or a suspicious phenotype.

Several agents have been shown to prevent experimental vincristine neuropathy. Similar to taxane and platinum-based compound-induced neuropathies, acetyl-L-carnitine, has been shown in animal models to be protective against vincristine-induced neuropathies (127). However, human data are lacking. Oral glutamate appears to ameliorate the neurotoxic effects of vincristine (162; 42). Amifostine has reduced neurotoxicity in early trials; however, later trials with paclitaxel-induced neuropathy did not corroborate this benefit (259). Coadministration of colony-stimulating factor increases the severity of vinca-induced toxic neuropathy (364). Nerve growth factor and insulin-like growth factor type 1 have also been shown to prevent the experimentally induced neuropathy (75). The immunosuppressive drug FK506 (tacrolimus) appears to restore normal regenerative ability of nerves exposed to vincristine (261). Although previous studies suggested a protective effect of the ACTH analogue Org 2766 in vincristine-treated animal models and cisplatin-treated ovarian cancer patients, a randomized, double-blind, placebo-controlled study evaluating its effects on vincristine-induced neuropathy in Hodgkin and non-Hodgkin lymphoma failed to detect any benefit (184). Other agents include radicicol and geldanamycin. Poly(ADP-ribose) polymerase (PARP)1/2 are nuclear enzymes activated on DNA damage, and PARP1/2 inhibition provides resistance against DNA damage. Selective PARP1/2 inhibitors (ABT-888 and related analogues) significantly attenuated development of mechanical allodynia and reduced poly ADP-ribose (PAR) activation in rat skin after coadministration with vincristine (44). Another novel PARP inhibitor compound 4a (analog of ABT-888) attenuates neuropathy in a mouse model treated with cisplatin and oxaliplatin, discussed later (323). Tropisetron, a 5-HT3 receptor antagonist, significantly suppressed vincristine-related neuropathy in a rat model (22). Genes that reduce Wallerian degeneration are also known, notably the WldS (slow Wallerian degeneration gene). Deletion of SARM1 (sterile alpha and TIR motif containing protein 1) protects mice against both axotomy and vincristine exposure compared to control animals (125). Other targets and agents are emerging in mouse models of taxane neuropathy. An oral Epac inhibitor reduces mouse allodynia and epidermal nerve fiber loss in a mouse model of paclitaxel-induced chemotherapy-induced neuropathies (307). Zinc supplements inhibit TRPV1 and may alleviate mouse induced paclitaxel neuropathy (217). Another mouse model found the neurokinin-1 receptor antagonist nefopam significantly decreased a mechanical allodynia marker (paw withdrawal threshold latency) (201). A rat model found oxytocin and liraglutide, a GLP‐1 analogue, reduced signs of vincristine‐induced neuropathy, lessened lipid peroxidation, and decreased nerve growth factor expression (109). A study of 23 pediatric patients with acute lymphoblastic leukemia who received vincristine found a high percentage of chronic inflammatory demyelinating polyneuropathy based on clinical and electrophysiology measures (17). All patients received supplemental pyridoxine and pyridostigmine that was claimed to be helpful, but no comparison group was offered. Oxytocin was found to lessen neurite damage in cultured dorsal root ganglia neurons and lessen mechanical hyperalgesia in treated mice (380). Coenzyme Q10 was possibly helpful in the prevention of vincristine neuropathy in a rat model (107). Cholecystokinin type 2 receptors (CCK2R) are involved in nociceptive modulation. CCK2R inhibitors, proglumide, and Ly225910 were administered to vincristine treated mice. CCK24 was highly upregulated in the dorsal root ganglia of vincristine treated mice. Pretreatment is associated with reduced allodynia measures and neuropathy prevention. The agents did not inhibit cytotoxicity of human cancer cell lines (28).

Differential diagnosis and workup. Of patients who develop an autonomic neuropathy, paraneoplastic autonomic neuropathy must be considered in those with small cell lung cancer, adenocarcinoma of the lung, testicular cancer, pancreatic cancer, and Hodgkin lymphoma (317). Rarely, patients receiving vincristine develop an associated fulminant neuropathy (severe quadriparesis) that is indistinguishable from Guillain-Barré syndrome (248). Again, it is important to obtain a detailed family history regarding possible hereditary neuropathies, as these may be unmasked during vinca alkaloid regimens.

Electrophysiologic testing is consistent with an axonal neuropathy. There are decreased sensory nerve action potential and compound motor action potential amplitudes on nerve conduction studies. The distal motor latency may be slightly prolonged in less than half of patients. EMG shows signs of denervation and decreased recruitment in distal muscles. Sural nerve biopsy displays axonal degeneration with minor segmental demyelination. Axonal regeneration occurs by a month after discontinuing the medication (234). Significantly diminished conduction velocities with relative sparing of compound motor action potential amplitudes may suggest an underlying hereditary neuropathy. If atypical motor involvement exists, genetic testing for Charcot-Marie-Tooth disease and HNPP may be warranted.

Prognosis. Clinical improvement follows cessation of the drug. Muscle weakness usually recovers rapidly; sensory symptoms may take several months. Despite marked improvement of clinical signs and symptoms, sensory nerve action potential amplitude recovery may be minor, sometimes averaging only 15%. There can be residual sensory symptoms and diminished or absent deep tendon reflexes in up to two thirds of patients.

Management. The only effective management of vinca-induced neuropathy is reduction of the dose, which usually reverses most major signs and symptoms without necessarily requiring a discontinuation of the drug. Venlafaxine inhibits hyperalgesia in a rat model of painful vincristine neuropathy (224). There is at least one case report of the successful utilization of plasma exchange for vinblastine overdose with severe neuropathy as a feature; however, this is an unusual circumstance (312). (Also see “taxanes” above.) Otherwise symptomatic treatment is similar to other neuropathic pain agents discussed in other sections.

Platins

Historical note and nomenclature. Cisplatin (cisdiamminedichloroplatinum), the first platinum analogue, was first described in 1844 as Peyrone’s chloride; however, its cytotoxic effects were not discovered until 1965. It became available for clinical use in 1978. Cisplatin binds DNA and is widely used for the treatment of ovarian, testicular, and bladder cancer. In an effort to curb the renal, gastrointestinal, and ototoxicity of cisplatin, other platinum analogues such as carboplatin and oxaliplatin were developed. Carboplatin has a similar therapeutic profile but is slightly less toxic. Oxaliplatin is the most recent platin compound and is used primarily in colorectal and other cancers (274). Other platins approved in other countries but not the United States include nedaplatin (Japan), heptaplatin (South Korea), and lobaplatin (China) (167). The next generation of platin agents includes alternative approaches or structures including oral (satraplatin), incorporation of targeting agents, platinum scaffolds that elicit biological effects distinct from current platinum drugs, and nanoscale drug delivery devices. Lipoplatin, a liposomal formulation of cisplatin, progressed well in clinical trials and could become the next platinum-based drug (167). A formal review of platin-induced peripheral neurotoxicity considers all aspects of the topic (313).

Clinical manifestations. Cisplatin produces a predominantly sensory neuropathy characterized by a severe loss of proprioception. Large fiber function is more affected than small fiber function; Lhermitte sign may herald the development of neuropathy. Weakness and motor neuropathy are not expected. Almost all patients develop diminished vibratory sense, even before any notable symptoms. Symptoms often begin with paresthesia and numbness of the feet. Temperature sensation is relatively preserved. When severe, gait ataxia and pseudoathetotic arm posturing may appear. Symptoms often worsen transiently after therapy is discontinued (coasting effect) but eventually remit. Pain and muscle weakness (mild at worst) are uncommon (275). Gastroparesis caused by autonomic dysfunction may occur, but autonomic neuropathy in general is not prominent. Carboplatin neuropathy is generally considered to be less severe than cisplatin neuropathy.

Tinnitus and high frequency hearing loss are relatively common clinical effects caused by ototoxicity. Neuropathies of bulbar cranial nerves and plexopathy are reported (35).

Oxaliplatin induces two clinically important neuropathic syndromes. After prolonged therapy, oxaliplatin will cause a peripheral neuropathy that is almost identical to that caused by cisplatin but typically resolves sooner (62). Lhermitte sign and urinary retention are reported after prolonged therapy (324). The other clinical syndrome is unique to oxaliplatin and occurs 30 to 60 minutes after infusion. It is characterized by spontaneous cold-induced paresthesias and painful dysesthesia. These symptoms can be accompanied by jaw and eye pain, possibly due to muscle cramps and tetanic spasms, and by alterations in vision and voice; additional less common symptoms include shortness of breath, fasciculations, and dysphagia (214). These symptoms resolve fully within several days but recur after each infusion. This phenomenon can be mitigated by calcium gluconate magnesium sulfate infusions (191). A Danish group carefully studied patients that received oxaliplatin or docetaxel 5 or more years prior to assess residual neuropathy (26). Patients underwent a thorough examination, including interview, neurologic examination, questionnaires, assessment tools, nerve conduction studies, quantitative sensory testing, motor unit number estimation, and corneal confocal microscopy. Patients were divided into no, possible, probable, and confirmed chemotherapy-induced peripheral neuropathy. Sixty-three patients agreed to participate: 28 received docetaxel and 35 received oxaliplatin. Forty‐one percent had confirmed chemotherapy-induced peripheral neuropathy, 34% possible or probable chemotherapy-induced peripheral neuropathy, and 22% did not have chemotherapy-induced peripheral neuropathy. Most affected had primarily sensory neuropathy. Although many were labelled as mild, about 40% had residual neuropathy more than 5 years after treatment that affected quality of life (26).

Etiology and pathophysiology. Chronic neurotoxicity is thought to be the result of platinum compounds accumulating in the dorsal root ganglia, leading to shrinking or loss of dorsal root ganglia neurons and a resultant sensory neuronopathy. Both cisplatin and oxaliplatin cause a chronic neuropathy associated with cumulative dose and length dependence; however, cisplatin is more likely to damage large, myelinated nerve fibers. The unique early reversible neurotoxicity seen after oxaliplatin infusion may be from interference with axonal ion conductance and neural excitability (03).

Platinum compounds are supposed to exert their cytotoxic effects on malignantly transforming cells by forming DNA adducts that block DNA replication and transcription, leading to cell cycle entry and eventual cell death and induction of apoptosis. The blood-brain barrier blocks platinum, but high levels are measurable in dorsal root ganglia and sural nerve axons, which is likely why motor fibers are primarily spared (188). Platinum also accumulates in sensory dorsal root ganglia after oxaliplatin (363). Downstream effects of DNA adducts may be potential targets to inhibit Wallerian degeneration or apoptosis. Activation of calpain by the gene Sarm1, a key regulator of Wallerian degeneration, appears to be required for neurotoxicity (63). Mice that lack the Sarm1 gene do not develop peripheral neuropathy as evaluated by both behavioral and pathological measures. A mouse dorsal root ganglia model assessing the transcriptomic profile of cisplatin treated cells found upregulation pathways involved in DNA damage response and products involved in apoptosis pathways (51). In addition, upregulation of the gene SIRT2, a gene involved in nucleotide excision repair, in a knock-in mouse model, protected mice against cisplatin neuropathy. The genetic effects were evident in neuronal but not lung cancer cell models (376).

There is also secondary degeneration of the posterior columns, which likely accounts for the Lhermitte sign. Oxaliplatin differs from other platinum compounds in that it contains a 1,2-diaminocyclohexane carrier ligand leading to the formation of bulkier platinum–DNA adducts, causing increased DNA inhibition and further apoptosis, yet other than the acute syndrome, typically less long-term neurotoxicity (366). Uncovering toxicity mechanisms in the hope of finding preventative targets continues. Immune and neuroinflammatory changes associated with chemotherapy-induced peripheral neuropathy from oxaliplatin and paclitaxel are reported, including neuronal injury marker activating transcription factor-3 and certain chemokines (221). Mitochondrial damage is assumed to play a critical role in cisplatin-induced peripheral neuropathy. Significantly reduced expression levels of mitofusin 2 (MFN2), implicated in axonal Charcot Marie Tooth disease, are reported in nerves of cisplatin-exposed animals (34). Based on multiple assays of rat dorsal root ganglia neuron cultures, cisplatin and oxaliplatin appear to affect mitochondrial function as well (199). A rat chemotherapy model was used to measure real-time blood mitochondrial DNA (mtDNA) levels and complex I enzyme activity, and findings were correlated with pain-like behavior signs (340). Systemic oxaliplatin significantly increased mtDNA levels in whole blood prior to pain development. Furthermore, paclitaxel- and bortezomib-treated animals displayed significantly higher levels of mtDNA at the peak of mechanical hypersensitivity. Complex I activity was not altered by any of the three chemotherapeutic agents. The findings suggest a potential role for circulating mtDNA levels as a noninvasive, predictive chemotherapy-induced peripheral neuropathy biomarker. However, axonal transport disruption is an implicated mechanism in platin induced chemotherapy-induced peripheral neuropathy. A review details the current evidence (300).

Oxaliplatin is rapidly biotransformed to platinum derivatives and oxalate immediately after infusion. These compounds are distributed in three compartments. Of the 3, the plasma-bound and ultrafilterable platinum are thought to be responsible for its acute effects (203). The acute toxicity syndrome of oxaliplatin is sometimes referred as an acute channelopathy because of oxaliplatin’s ability to cause prolonged opening of sodium channels leading to a hyperexcitable state. Adelsberger and colleagues demonstrated this in sensory nerves, whereas others have electrophysiologically demonstrated hyperexcitability in the form of repetitive nerve discharges, multiple motor unit potentials, and neuromyotonic discharges (03; 366; 203). Neuromyotonia may be idiopathic or immune-mediated or may be the side effect of drugs, toxins, or radiotherapy and is the result of hyperexcitability of the peripheral nerve axon. Prolonged opening of sodium channels or diminished conduction across potassium channels may be responsible. These effects lead to continued neurotransmitter release, repetitive nerve firing, and muscle contraction. Oxaliplatin most likely exerts it toxic effects on voltage gated sodium channels via chelation of unbound calcium, possibly by the oxalate metabolite (137). The effect is prevented in vitro by sodium but not potassium channel blockage (362). Topiramate shows promise abating this hyperexcitability side effect in a rat model (06).

Interactions between transient receptor potential ankyrin 1 (TRPA1), a channel expressed in nociceptors, and the Sigma-1 receptor were studied using various physiological and chemical methods including patch clamp recordings. The Sigma-1 inhibitor S1RA was found to inhibit TRPA1 expression and function (225). The agent also showed benefit in a rat oxaliplatin model and might represent a new therapeutic target (225). Sigma-1 had previously shown some benefit in a randomized phase IIa trial of oxaliplatin treated colorectal cancer patients. The treated patients showed improved quantitative cold thresholds but similar total neuropathy score, quality of life measures, and electrophysiology (46).

Epidemiology. Cisplatin neuropathy usually begins after several courses of doses greater than 75 mg/m2. After cumulative doses greater than 300 mg/m2, almost all patients exhibit some signs of neuropathy, beginning with diminished vibration sensation. After cumulative doses greater than 600 mg/m2, neuropathy is moderate to severe in a significant number of patients. However, mild neuropathy from prior treatment does not predispose to a more severe neuropathy with re-exposure if re-exposure cumulative doses are kept below 420 mg/m2 (351). The combination of cisplatin and docetaxel produces more severe neuropathy than either agent alone (149). The presence of neuronal autoantibodies appears not to predispose patients to the development of cisplatin neuropathy (122).

In a total of nine combined studies (682 patients) of oxaliplatin by the National Cancer Institute and the World Health Organization, the risk of severe acute neuropathy was 10% after six cycles (cumulative dose 780 mg/m2) and 50% after nine cycles (cumulative dose 1170 mg/m2) (62). Neuropathy completely resolved in 78% after discontinuation. Eighty-two percent experienced symptom regression by 3 to 4 months (median follow-up); 41% had complete disappearance of symptoms by 6 to 8 months (median follow-up). Chronic neuropathy is also cumulative and most commonly occurs in individuals who have received total doses greater than 540 to 850 mg/m2, about nine to 10 cycles of 85 mg/m2 or six cycles of 130 mg/m2 (62). Some studies suggest that the incidence of neurotoxicity increases when oxaliplatin is co-administered with fluorouracil and leucovorin; however, this effect has not been consistently demonstrated (62). Studies by Mattioli and colleagues suggest that bifractionation of oxaliplatin infusion may decrease long-term neurotoxicity, reporting only 6.4% grade 3 neuropathy out of 77 treated patients (231). A large, genome-wide association study of 1379 colon cancer patients receiving therapy, including oxaliplatin, compared subjects based on neuropathy severity (172). No single nucleotide polymorphisms met significance for neuropathy severity or neuropathy improvement.

Prevention. Multiple agents have been attempted for prevention of platinum-induced neuropathy. Both animal and human models have demonstrated the neuroprotective effects of acetyl-L-carnitine in cisplatin therapy (see “taxanes” above). Vitamin E also appears to offer protection in cisplatin therapy. Calcium and magnesium infusions have shown protective effects against oxaliplatin neurotoxicity. In a randomized controlled study of pre- and post-treatment with calcium gluconate and magnesium chloride, subjects received a cumulative dose of 910 mg/m2 whereas controls received 650 mg/m2. The incidence of neuropathy in the treated group was 27% versus 75% in the control group. Pharyngeal dysesthesia occurred in 1.6% versus 26% (123). Gabapentin, given in daily doses of 200 to 300 mg in seven patients, both treated and prevented neuropathic symptoms for up to 14 cycles (226). Carbamazepine, given at 400 to 600 mg per day, has also shown protective effects in oxaliplatin therapy (204).

Pretreatment with amifostine, an organic thiophosphate, has been demonstrated to ameliorate the development of platin neuropathy (176). Side effects of amifostine therapy include hypotension, nausea, and vomiting. Glutathione, possibly by reducing oxidative stress of accumulating metabolites in the dorsal root ganglia, modestly decreased cisplatin neurotoxicity in women with ovarian cancer (308). Another randomized controlled study of colorectal cancer patients demonstrated the neuroprotective effects of glutathione in patients receiving oxaliplatin (53).

Several other agents have also shown promise. In a rat model, pretreatment with oral glutamate delayed the onset of neuropathy without affecting its antineoplastic efficacy (43). Nerve growth factor (12) and recombinant human glial growth factor (330) have also shown promise in preventing experimentally induced neuropathy. Neurotrophin-3 has been successfully introduced by gene therapy via both adenovirus and plasmid vectors in a mouse model; this produced some protection against the development of neuropathy (270). Further randomized controlled studies are warranted. Albers and colleagues updated a Cochrane review of available preventative agents for platin (mostly cisplatin) neurotoxicity (05). The trials examined included multiple separate studies of amifostine, glutathione, and Org 2766 of varying sophistication and numbers; a form of quantitative sensory testing was considered to be adequately objective. Single small studies for acetylcysteine (14 participants), diethyldithiocarbamate (195 participants), calcium and magnesium (33 participants), and oxcarbazepine (32 participants) and the two eligible trials involving vitamin E (57 participants) did not perform quantitative sensory testing. In all, data from 1537 participants were included. In brief, the primary conclusion was that none of these trials provided sufficient objective evidence of prevention of platin-induced neuropathy. Amifostine remains the only FDA-approved medication in the U.S. for the prevention of cisplatin toxicity, primarily renal. However, the data remain doubtful that it also prevents associated neuropathy. Monosialotetrahexosylganglioside (GM1) was used for a similar rationale as the taxane neuropathy prevention report discussed earlier. In a phase 3 Chinese study, 196 patients with colorectal cancer receiving a protocol including oxaliplatin were randomized to receive GM1 or placebo pretreatment (357). There were no significant differences in objective markers of new grade 2 or worsened neuropathy but symptom scores were better in the GM1 group. The antiinflammatory agent L-carnosine was used in 65 very similar colorectal cancer patients. The treated patients showed significantly decreased NF-κB (27%) and TNF-α (36.6%) levels/activity versus placebo; the neuropathy grading score was also improved (374). Reduced caspase-3 levels suggested a beneficial antiapoptotic effect, but overall patient numbers were not high enough to prove a lower oxaliplatin neuropathy incidence.

Numerous agents showed initial promise in mouse or cell line models of cisplatin or oxaliplatin neuropathy but were not formally tested in human trials. Examples include rosmarinic acid (13), rat derived adipose stem cells (90), agmatine (97), and ninjin'yoeito and ginseng extracts (320). Salmon calcitonin reduced oxaliplatin-induced cold and mechanical allodynia in a rat model (11). Calcium channel blockers also show promise in attenuation of this common side effect. A review of a large cohort of patients with colon cancer found that acute neuropathy symptoms were less in treated patients (329). The molecule APX2009 may be effective at preventing or even possibly reversing the effects of cisplatin and oxaliplatin-induced neurotoxicity in cultured sensory neurons (175). Ethoxyquin appears to protect sensory neuron culture and rat dorsal root ganglia neurons against cisplatin toxicity (379). A mouse dorsal root ganglia model found senescent-like neuronal cells based on various cellular markers and gene expression that were more vulnerable to platin effects. Depletion of these cells or a pharmacological “senolytic” agent, ABT263, made the culture resistant to platin toxicity (02). An oral diabetes agent, alogliptin, a dipeptidyl peptidase 4 (DPP-4) inhibitor shows promise in limiting oxaliplatin toxicity in PC12 cell models and mechanical allodynia and sciatic nerve axonal degeneration rat models (299). A randomized controlled trial of ART-123, a recombinant human soluble thrombomodulin, in 79 patients with colon cancer receiving chemotherapy, including oxaliplatin, found a slight reduction in grade 2 or higher neuropathy in treated patients (186). MS-275, an MHC class I histone deacetylase inhibitor, appears to prevent chemotherapy-induced peripheral neuropathy in mouse and cell line models and may have additional antineoplastic properties (194). The FDA-approved adenosine A2A receptor antagonist istradefylline (KW6002) protected against cisplatin-induced nephrotoxicity and neuropathic pain in mice with or without tumors (86). Omeprazole was also found to suppress oxaliplatin-induced chemotherapy-induced peripheral neuropathy in a rodent model (241). Zonisamide pretreatment showed some toxicity benefit in a rat dorsal root ganglia neuronal culture model (325). Pterostilbene and celecoxib showed similar prevention oxaliplatin behavioral neuropathy signs in a rat model (01). The agent’s antiinflammatory, antioxidative, and antiapoptotic effects are suggested as the mechanism.

However, calmangafodipir failed to show patient-reported neuropathy symptom benefit in two randomized placebo trials of oxaliplatin treated colorectal cancer patients (268). The study was terminated early because of unexpected calmangafodipir hypersensitivity reactions (3.6%). No agent is currently agreed to be effective in chemotherapy-induced peripheral neuropathy prevention including platin induced cases. A systematic metaanalysis of multiple studies is reported in 2022 (266).

Infusion of donor hepatic mitochondria (mitotherapy) to a rat model of oxaliplatin-induced CIPN significantly prevented mechanical sensory loss (Von Frey hairs), and allodynia (acetone test) presented a potential new treatment strategy for further research (219).

Differential diagnosis and workup. Please see corresponding “taxanes” section. Also, because cisplatin is nephrotoxic, neuropathy due to renal failure might be considered in the appropriate clinical situation.

Electrodiagnostic testing shows decreased sensory nerve action potential amplitudes. Conduction velocities are largely normal. Compound motor action potentials are usually normal. EMG is normal in pure neuropathy but can show evidence of tetany in patients who develop hypomagnesemia from cisplatin-induced nephrotoxicity (16). If evidence of tetany is found, calcium and magnesium levels as well as renal function testing should be considered.

During the post-infusion toxicity phase of oxaliplatin, neuromyotonic discharges and complex repetitive discharges can be seen. Lehky and colleagues investigated 22 patients within 24 to 48 hours of oxaliplatin infusion. Motor conductions and needle electromyography prior to infusion were entirely normal, whereas post-infusion studies revealed repetitive compound motor action potential after discharges on average in 1.2 of 4 motor nerves studied. Needle EMG at the same time revealed neuromyotonic discharges or multiplet motor unit potentials on average in 2.1 of 3 muscles sampled.

Nerve biopsy has shown mixed axonal degeneration (of both myelinated and unmyelinated fibers) and segmental demyelination (280).

Prognosis. Patients with mild neuropathy and cumulative doses less than 500 mg recover after cessation of the drug. Recovery commonly occurs after a period of coasting (progression after stopping the drug) and may take up to 8 months (106). However, neuropathy may appear after cessation of the drug. Patients with profound large fiber loss have poor recovery. In a study of long-term survivors older than 12 years of age who received cisplatin therapy, 38% of patients had asymptomatic neuropathy, 28% had symptomatic neuropathy, and 6% had disabling neuropathy (315). Oxaliplatin-induced neuropathy typically occurs after cumulative doses greater than 540 mg/m2. Neuropathy clinically completely resolved in approximately 40% of patients by 8 months, but about 20% of patients have persistent, significant symptoms (111). Liposomal oxaliplatin may have less toxicity in general but seems to have the same severity of neuropathy.

Management. Vibratory threshold may be useful in monitoring the development of neuropathy (106). Avoidance of cold temperatures including cold foods should be encouraged in the acute phase after oxaliplatin infusion (62). Discontinuation of treatment should be considered when neuropathy develops.

Regarding symptomatic treatment, anecdotal reports of symptomatic improvement of oxaliplatin-induced paresthesia with venlafaxine and topiramate have been published (102). Nortriptyline provided moderate improvement over placebo in a randomized controlled study investigating symptomatic relief of cisplatin-induced paresthesia (140). (Also see corresponding “taxanes” section.)

Suramin

Historical note and nomenclature. Suramin, a polysulfonated naphthylurea, is a persistently experimental antineoplastic agent being tested against prostate cancer, adrenal cancer, hypernephroma, T-cell leukemia, and lymphoma. It is toxic to the adrenal gland and the nervous system. The agent was developed in 1916 as an anti-protozoal and anti-helminth agent. There have been no recent advances to suggest eventual approval. The agent is shown to have antiviral activity against enterovirus A (276). It shows promise against various other viral infections, including Chikungunya and possibly Ebola.

Clinical manifestations. Neuropathy is the dose-limiting side effect. More than 50% of patients develop a length-dependent axonal sensorimotor neuropathy (67). Symptoms begin with paresthesia and are followed by weakness, often predominantly proximal. When severe (about 10%) symptoms of bulbar weakness, respiratory weakness, and dysautonomia can occur (193). Additionally, an ascending demyelinating neuropathy similar to Guillain-Barré syndrome may be seen (310).

Etiology and pathophysiology. The etiology of the neuropathy is unknown, but disruption in glycolipid transport and metabolism is suspected. Exposure to suramin induces ceramide accumulation, exhibited as lamellar inclusion bodies, and leads to apoptosis (129). Interruption of calcium signaling via purinergic receptor blockade may also compromise calcium-dependent intracellular pathways (274). Human sural nerve biopsy has shown lymphocytic infiltration in the demyelinating form of neuropathy (67).

There are several suggested mechanisms of action of suramin. Suramin blocks the action of DNA-primase, DNA-polymerase-alpha, and reverse transcriptase. It also inhibits many growth factors including platelet-derived growth factor, insulin-like growth factor type 1, and nerve growth factor. In vitro studies implicate the influx of extracellular calcium as playing a role in some capacity. This may be related to blockage of P2 purinergic receptors found on dorsal root ganglia neurons and Schwann cells (274). Schwann cells and dorsal root ganglia neurons accumulate characteristic glycolipid lysosomal inclusions in an animal model (285).

Epidemiology. The incidence of neuropathy has not been studied extensively, but neuropathy appears to be common; in one study, 12 of 22 patients developed an axonal length-dependent neuropathy and 3 of 22 developed a demyelinating neuropathy (67). Peak plasma concentrations greater than 350 µg/ml appear to be a significant risk factor for the development of neuropathy (274).

Prevention. Limiting the peak serum concentration may be beneficial because neuropathy most commonly develops in patients with peak concentrations greater than 300 µg/ml (193). Similar to platinum, vincristine, and taxane neurotoxicity, neurotrophins such as nerve growth factor have been shown to prevent suramin neurotoxicity in experimental models (367).

Differential diagnosis and workup. Acute inflammatory demyelinating polyneuropathy can occur in association with Hodgkin disease and lung cancer and should be considered in the differential diagnosis in patients with that clinical picture. A vasculitic neuritis has been reported in association with lymphoma as well as kidney and prostate cancer. Although sural nerve biopsy shows vasculitic changes in some cases of suramin-induced demyelinating neuropathy, an independent, tumor-associated vasculitis should be kept in the differential diagnosis.

Serial electrophysiologic studies can monitor progression. Electrophysiologic studies show features consistent with either a demyelinating neuropathy or an axonal neuropathy. Findings may be time-dependent; multiple conduction blocks appear early and fibrillation potentials later (193). Nerve biopsy is appropriate if vasculitic neuropathy secondary to prostate cancer is suspected. Sural nerve biopsy otherwise shows axonal degeneration.

Management. Experience is somewhat limited, but most patients will partially improve on cessation of treatment. The demyelinating form may have a better prognosis. Patients with demyelinating features benefit from suramin discontinuation and possibly plasmapheresis (67).

Alkylating agents (iphosphamide, hexamethylmelamine, procarbazine)

Historical note and nomenclature. Iphosphamide is a cyclophosphamide analog used to treat germ cell cancers, cervical cancer, lung cancer, lymphoma, and sarcoma. Hexamethylmelamine and its derivatives, such as altretamine, are a group of experimental alkylating agents used in highly refractory tumors (74). Procarbazine is used in treating oligodendrogliomas as part of PCV chemotherapy.

Clinical manifestations. In general, the alkylating agents produce more central neurotoxicity than peripheral neurotoxicity. For ifosfamide and procarbazine, encephalopathy is more frequent than neuropathy. The reverse is true for hexamethylmelamine.

Iphosphamide peripheral neurotoxicity usually manifests as severe pain and paresthesias of the hands or the feet, or both, 10 to 14 days after the initial treatment. The severe symptoms usually last hours then abate and return to baseline over weeks (317). Exacerbation of pre-existing neuropathy also occurs. High-dose iphosphamide has been reported to cause an axonal sensorimotor neuropathy with painful soles and limb weakness that did not totally resolve at 18 months after treatment (119).

Hexamethylmelamine induces a mild sensorimotor axonal polyneuropathy that is reversible on cessation of the drug. Altretamine, a hexamethylmelamine derivative, has been used in combination with paclitaxel without any apparent additive neuropathic effect (258). Procarbazine induces a mild sensory neuropathy at high doses, manifested by distal paresthesias, depressed deep tendon reflexes, and myalgias. These are also reversible on cessation of therapy.

Etiology and pathophysiology. Iphosphamide antineoplastic effect results from covalent binding to and cross-linking of nucleic acid chains. It is postulated that the site of injury is the dorsal root ganglion in ifosfamide toxicity (317).

Epidemiology, prevention, and diagnosis. Preexisting neuropathy is a risk factor for the development of clinical symptoms. De novo neuropathy occurs at high doses of ifosfamide, usually greater than 14 gm/m2. The incidence of neuropathy may approach 20%. Pyridoxine may reduce the severity of hexamethylmelamine neuropathy (375). However, high-dose pyridoxine can, unfortunately, also independently cause a severe sensory polyneuropathy. Electrodiagnostic studies may document the neuropathy.

Prognosis and management. Prognosis is usually good with reversal of the neuropathy with discontinuation of treatment. Ifosfamide-induced painful paresthesias can be managed with adjuvant analgesics although narcotics do not seem to be helpful (317).

Bortezomib and other proteasome inhibitors

Historical note and nomenclature. Bortezomib is a reversible proteosome inhibitor that was FDA approved in 2003 for monotherapy in the treatment of refractory multiple myeloma. It is also approved for use in numerous other countries. It is used in combination with other chemotherapeutic drugs for treatment of non-Hodgkin lymphoma and malignant melanoma, especially thalidomide, discussed later. Carfilzomib, a similar proteasome inhibitor that has different pharmacology and pharmacokinetics from bortezomib was FDA approved in the United States in July 2012 for the treatment of multiple myeloma. Although yet to be determined, this agent may have a lower rate of neuropathy, but it causes significant cardiotoxicity (335). Phase II studies suggest a neuropathy incidence of 12% (grade I and II) (304). More severe grades are not yet appreciated. A second-generation oral proteasome inhibitor, ixazomib, which may have less toxicity, was approved late in 2015. It is not yet clear if the neuropathy risk is lower (294). Refractory lupus nephritis may be another use, but neuropathy is also a limiting factor (297).

Clinical manifestations. Bortezomib most commonly causes a dose-related, length-dependent, mixed fiber, axonal, sensory greater than motor polyneuropathy that typically resolves within weeks of discontinuation but is an important treatment-limiting side effect (277; 353). Comparing 43 different phase 3 trials using the agent found that overall neuropathy incidence ranged from 8.4% to 80.5% (median = 37.8%) and severe neuropathy (grade 3 to 4) ranged from 1% to 33.2% (median = 8%) (206). Dosage adjustments are made based on the emergence of neuropathy. Small fiber manifestations include burning dysesthesia or paresthesia and numbness. In addition, reports of paralytic ileus and hypotensive episodes may be manifestations of autonomic involvement (163). Demyelinating or inflammatory neuropathy with notable weakness is also increasingly recognized as a possible complication (288; 332; 306). A large study of 955 newly diagnosed multiple myeloma patients compared carfilzomib or bortezomib for transplant-ineligible patients. The progression-free survival was the same but grade 2 or worse neuropathy incidence was lower in the carfilzomib group (2.5% vs. 35.1%) (112). A Swedish study of 457 patients with multiple myeloma who received combinations of bortezomib, lenalidomide, carfilzomib, or thalidomide reviewed state health system recorded clinical characteristics. They found 102 (22%) were clinically labeled to have chemotherapy-induced peripheral neuropathy—most commonly associated with bortezomib (84%) (250). They noted the higher costs and health care utilization in affected patients compared to others.

Etiology and pathophysiology. The mechanism of the neuropathic effect is unknown. Bortezomib functions as an inhibitor of the 26S proteosome, which is present in all mammalian cells. The nucleosome’s function is primarily catalytic; thus, inhibition of this catalytic activity results in a cascade of events arresting the cell cycle and inducing apoptosis (163). This may explain the predominance of sensory involvement because of preferential dorsal root ganglia toxicity. Schwann cells also appear to be a target in rat and mouse models (59; 302). Interference with endoplasmic reticulum function may promote abnormal myelination (302). There remain conflicting data about whether the induced neuropathy is strictly dose-dependent; some subjects develop more severe neuropathy than anticipated at conventional dosage. As a result, genetic studies have attempted to identify markers of increased susceptibility to bortezomib-induced neuropathy; promising results are emerging (45; 114). However, the neuropathy ascertainment methods were criticized in the Broyl study (56). Similar to other agents, existing neuropathy increases the likelihood of more significant toxic neuropathy (195). In addition, itraconazole may increase the likelihood of neuropathy development (143). There is also early evidence that subcutaneous bortezomib may induce slightly lower rates of sensory neuropathy than traditional intravenous administration (355; 208). Obesity is also a risk factor for overall neuropathy incidence compared to overweight or normal/underweight patients (245). Furthermore, a genome-wide association study demonstrated that a locus in the PKNOX1 gene and one near the CBS gene were associated with bortezomib-induced neuropathy and may be a possible future therapeutic target (222). One important bortezomib mechanism of action for tumor and possibly neuronal survival is inhibition of nuclear factor kappa B (NFκB), a transcription factor involved in cell survival and proliferation. In a transgenic mouse model that selectively blocks the NFκB pathway in neurons, animals with impaired NFκB activation developed significantly less severe neuropathy than wild type animals (07). Bortezomib significantly increased the expression of NOD-like receptor family pyrin domain containing 3 (NLRP3) and phosphorylated signal transducer and activator of transcription-3 (STAT3) in dorsal root ganglion (DRG). Inhibition of NLRP3 by intrathecal siRNA prevented allodynia in a rat model (207). Another study showed that low vitamin D levels are associated with a higher severity of bortezomib-induced neuropathy in myeloma patients, and, therefore, levels should be monitored and repleted if low (358). A successful rat model was developed that was attenuated by ethosuximide and phenyl N-tert-butylnitrone (a reactive oxygen species scavenger), suggesting involvement of calcium, antioxidants, or radical scavengers in bortezomib-induced painful neuropathy (101). Rat models show increased tumor necrosis factor alpha levels correlates with neuropathy development (378). Low pretreatment brain-derived neurotrophic factor levels were also a neuropathy risk factor (18). A microRNA molecule (miR-155) was applied to a rat model of bortezomib neuropathy to assess response to neuropathic pain development. Various inflammatory markers and mechanical and cold sensitivity behavioral markers lessened in the treatment group (100). Mitochondrial toxicity is also evident in both bortezomib and carfilzomib, but that pathway does not appear to be the primary toxic neuropathy explanation (164). Accumulation of delta 2 tubulin (D2) in sensory nerve in culture and rodent models may be a primary bortezomib effect; D2 is also a hallmark in human bortezomib chemotherapy-induced peripheral neuropathy (267). A review of multiple possible and potential bortezomib neurotoxic mechanisms cites many of these and other pathways (372).

Prokineticins, a proinflammatory cytokine family of agents, may drive epigenetic mechanisms important for chemotherapy-induced peripheral neuropathy development based on mouse spinal cord studies of various genetic regulation and cytokine levels in bortezomib-treated animals (284). Bortezomib may have differing effects at high and low doses. Despite clear toxic sensory neuropathy at higher doses, lower doses have immunomodulatory effects that are beneficial for a Lewis rat model of experimental autoimmune neuritis (183). Microarray studies of mRNA expression found the upregulation of genes responsible for regulating immunological and apoptotic processes when comparing patients with multiple myeloma with and without chemotherapy-induced peripheral neuropathy after a bortezomib, thalidomide, and dexamethasone regimen (216). Human induced pluripotent stem cell (hiPSC)-derived cultured motor neurons and sensory neurons displayed distal-to-proximal neurite degeneration when exposed to bortezomib. One group found this process coincided with disruptions in mitochondrial function and energy homeostasis, similar to those described in rodent models of bortezomib-induced neuropathy (309). This process was blocked by exogenous nicotinamide adenine dinucleotide (NAD+), a mediator of the SARM1-dependent axon degeneration pathway (309).

Baseline neuropathy prior to chemotherapy initiation is a recognized chemotherapy-induced peripheral neuropathy development risk factor. A Chinese group retrospectively studied 155 multiple myeloma patients (98). Those with baseline neuropathy not only had higher chemotherapy-induced peripheral neuropathy rates than others as expected but also had higher relapse rates and worse survival rates.

Analysis of a rat model of bortezomib-induced neuropathy found impaired Schwann cell lysosomal function rather than autophagy may be a primary factor in bortezomib-induced CIPN (370).

Epidemiology. In 228 patients in two phase II trials examined during the FDA safety review of bortezomib at the standard 1.3 mg/m2 dose, peripheral neuropathy occurred in 37%. Of those 37%, 14% had grade 3 neuropathy and none had grade 4. Paresthesia and dysesthesia occurred in 23% (173). Chaudhry and colleagues prospectively studied 27 myeloma patients receiving a combination of bortezomib and thalidomide (66). Three subjects had mild neuropathy at baseline; however, 26 developed mild-to-moderate neuropathy during treatment. The physiology was axonal in all but three cases. A meta-analysis of 34 trials (6492 patients) that assessed the risk of neuropathy in patients treated with bortezomib concluded that 34% are affected (all grades); patients with lymphoma had a higher risk than myeloma patients (265). A study comparing low versus higher/conventional bortezomib dose in elderly patients with multiple myeloma found similar cancer response rates but lower toxicity incidence including chemotherapy-induced neuropathies (72). In addition, a comparison study adding daratumumab in a protocol to allow lower bortezomib dosing found similar response and lower neuropathy rates; however, the bortezomib routes (subcutaneous vs. intravenous) were a possible explanation (61).

A review of the FDA adverse event reporting system found that neuropathy was reported in 2.1%, 5.0%, and 10.9% of adverse events after carfilzomib, ixazomib, and bortezomib, respectively (240). The differences support lower neuropathy rates in the newer generation proteosome inhibitors. A meta-analysis of studies reporting on patients treated with ixazomib tabulated 1440 patients in various studies (69). Efficacy for progression-free survival was good; the relative risk of neuropathy was 1.48. A retrospective study of 52 newly diagnosed multiple myeloma patients in China found lower rates of chemotherapy-induced peripheral neuropathy in patients receiving ixazomib as opposed to bortezomib in addition to lenalidomide and dexamethasone (157).

Based on queried data from the Multiple Myeloma Research Foundation (MMRF) patient registry, one group found similar reported chemotherapy-induced peripheral neuropathy rates of patients treated with lenalidomide versus bortezomib. Neuropathy rates also were not correlated with prognosis (220).

Differential diagnosis and evaluation. Electrodiagnostic testing typically reveals a length-dependent, axonal, sensory greater than motor polyneuropathy. Demyelinating findings suggest M-protein related polyneuropathy, an association with multiple myeloma in many cases, but the emergence of demyelinating and inflammatory neuropathies from bortezomib complicate this distinction. Careful monitoring for neuropathy is necessary to guide ongoing treatment.

Management. Although there are no randomized, controlled studies investigating symptomatic treatment for bortezomib-related neuropathic symptoms, typical agents used in the treatment of neuropathic pain may be beneficial. Symptomatic treatment of dysautonomia may also be needed. A small open-label phase 2 study of 25 patients with multiple myeloma found a cocktail including docosahexaenoic acid, α-lipoic acid, and vitamins for 6 months showed low neuropathy rates and no bortezomib treatment discontinuation (228). Inclusion of the oral, novel mechanism agent, selinexor, in a chemotherapy cocktail for previously treated multiple myeloma allows for a lower dose of bortezomib and subsequently lower chemotherapy-induced peripheral neuropathy rates (322).

Thalidomide and lenalidomide

Historical note and nomenclature. Thalidomide is a synthetic glutamic acid derivative that contains a phthalimide ring and a glutarimide ring. The drug was first used in 1957 as a safe, mild sedative and as an antiemetic for pregnant women; however, it was never FDA-approved for use in the U.S. It was withdrawn from clinical use because of teratogenicity and neuropathy in 1961. Thalidomide was later reintroduced for a number of disorders, including dermatologic disorders such as leprosy, cutaneous lupus erythematosus, and Behçet syndrome; HIV-associated conditions, such as aphthous ulcers and wasting syndrome; and chemotherapy against a number of malignancies, especially multiple myeloma (230). The drug was approved for erythema nodosum leprosum approximately 20 years after the original withdrawal.

Clinical manifestations. Thalidomide produces a predominantly sensory neuropathy, frequently initial leg and foot dysesthesia and a tight feeling around the ankles or the feet (120). Examination shows distal sensory and reflex loss, with little time between the development of symptoms and signs. Muscle cramps, constipation, somnolence, dizziness, and fatigue are common adverse effects (116).

Thalidomide is a known teratogen that causes dysmelic deformities of the limbs, likely as a result of a toxic embryonic neuropathy.

Etiology and pathophysiology. Although it is clear that thalidomide causes an axonal neuropathy, the mechanism of this effect is unknown.

Thalidomide has both immunomodulatory and anti-inflammatory effects, many of which are due to its selective inhibition of tumor necrosis factor alpha in monocytes. It also reduces helper T-cells and increases suppressor T-cells, enhances interleukins 4 and 5, and inhibits gamma interferon (50). In the embryo, thalidomide exposure results in a failure of dorsal root ganglion cell maturation (233). The possibility that the dorsal root ganglion cells are the target in exposure-induced neuropathy remains open but might explain the relatively poor recovery after drug withdrawal (128). Other studies have suggested a dying back axonopathy (65).

Epidemiology. The estimated incidence of thalidomide-induced neuropathy ranges from 21% to 50%, with women and elderly patients having the highest risk (257). There may be a pharmacogenetic predisposition to the development of neuropathy. In a study, cumulative dose and duration of therapy were unrelated to its development, and smoking may have a protective effect (142). Other studies suggest a somewhat lower risk (around 9%) and suggest cumulative dose and age but not preexisting neuropathy are potential risk factors (244). Other studies have indicated that the incidence of neuropathy is related to cumulative doses of thalidomide, with significant signs and symptoms developing at cumulative doses as low as 20 grams (57). The threshold amount to produce neuropathy might be even lower if pre-existing neuropathy exists (64). Additional work has identified various genetic polymorphisms that may increase the susceptibility of neuropathy development (166). The identified regions involved genes thought to govern repair mechanisms and inflammation in peripheral nerves. The polymorphisms identified were different from a separate study of patients with vincristine neuropathy.

Lenalidomide is a second-generation thalidomide analogue with more potent anti-inflammatory and antiangiogenic activities and less toxicity, including neurotoxicity. Phase 1 and phase 2 trials appeared to show minimal to no increase in neuropathy incidence (361). However, reports of neuropathy are known, but many patients received other neurotoxic agents, such as thalidomide, bortezomib, or vincristine (68; 87). An Italian group prospectively and carefully tracked 19 patients with multiple myeloma (83). Seven of the 19 had neuropathy at baseline attributed to prior chemotherapy. Half of the patients tracked developed sensory neuropathy though most was mild and did not affect the neuropathy scale employed (total neuropathy score, TNSc). Neuropathy did not correlate with lenalidomide cumulative dose or hematologic response (83). Combination of this agent and carfilzomib is shown to increase survival in patients with refractory multiple myeloma. Both are implicated in chemotherapy-induced neuropathies. Neuropathy rates were similar, approximately 3%, in groups with and without additional carfilzomib (303). The agent is often combined with bortezomib and dexamethasone; the grade 2 or higher neuropathy rate in a study of 458 patients was 17% (283). Thalidomide was a standard treatment of POEMS syndrome; however, that severe entity including significant neuropathy often was exacerbated by treatment. Lenalidomide is a promising alternative to avoid that complication (95). Three patients with neuropathy from POEMS syndrome were treated with lenalidomide and dexamethasone. The neuropathy notably improved on treatment for all (131).

Pomalidomide is the newest agent in this class that received accelerate FDA approval in 2013 for treatment of refractory multiple myeloma. However, lenalidomide remains standard therapy in most protocols. Neuropathy incidence is reported to be 12%. The agent is used in combination with dexamethasone, often combined with bortezomib in lenalidomide-refractory and proteasome inhibitor-exposed myeloma patients. However, lenalidomide remains the standard treatment option in this class. In a post-FDA approval open label study of 682 refractory patients, neuropathy of any grade was found in 17.9% of patients, 43.8% of whom had peripheral neuropathy at baseline; the median time to neuropathy onset of any grade was 1.7 months (91).

Prevention and prognosis. No known agents prevent thalidomide neurotoxicity; the goal of prevention lies in early detection. Limiting daily doses to 25 mg daily appears to be the most effective means of preventing the onset of neuropathy (23). Recovery is often poor to not at all. Early studies reported that 50% of patients made no recovery at all, and another 25% made only partial recovery after 4 to 6 years (121). Other studies do not negate that poor prognosis (244).

Management. Because the neuropathy is slow to resolve, if at all, the goal of management is the early detection of neuropathy. Manufacturers recommend monthly neurologic exams for the first 3 months and then periodically thereafter, along with electrophysiologic evaluation. Baseline sensory nerve action potentials should be obtained with repeat studies every 6 months, and more frequently should amplitudes drop by 30%. Discontinuation of thalidomide is recommended if baseline sensory nerve action potentials drop to 50% of their pre-treatment baseline (124). Studies have confirmed that sensory nerve action potential amplitudes are the best means of monitoring for thalidomide-induced neuropathy (57). Monitoring of F-waves, especially chronodispersion, may be useful in detecting the earliest changes (286).

Pregnancy. Thalidomide is a known teratogen that causes dysmelic deformities of the limbs, likely as a result of a toxic embryonic neuropathy. Because it is a thalidomide analog, lenalidomide is approved for use in women of childbearing age only using a Risk Evaluation and Mitigation Strategy so that only certified practitioners can prescribe the chemotherapy. Of interest, false positive pregnancy tests were found in 1% to 2% of treated patients (55).

Ixabepilone

The epothilones are a newer class of chemotherapeutic agents with currently low tumor resistance. The class includes the natural agents, epothilone B (patupilone) and epothilone D, which are not yet FDA approved, as well as the semisynthetic analog, ixabepilone (32; 321); ixabepilone was FDA approved in October 2007. Breast cancer is the primary indication, but studies have been completed or are ongoing for other cancer types, including ovarian and prostate cancer. The group binds to tubulin similarly to certain other chemotherapeutic agents, such as vinca alkaloids and taxanes, but at differing binding sites. Similar to taxanes, the drugs promote dysfunctional stabilization of microtubules but with a differing mechanism, in contrast to microtubular destabilizing agents such as vincristine, colchicine, and podophyllotoxin (77). The agent preferentially binds to the betaIII isotype of tubulin (212). Peripheral neuropathy from ixabepilone is a dose-limiting side effect. Mild-to-moderate (grade 1 to 2), predominately sensory neuropathy that improves or resolves is most common, but more severe grades (grade 3) occur rarely in monotherapy and at rates of 10% to 15% with combination therapy or in patients previously exposed to taxanes or capecitabine; grade 4 neuropathy appears to be very rare (89; 278; 334). Dose reduction may be adequate in many, but treatment discontinuation occurs as well; more dispersed treatment protocols may have lesser toxicity (333). Twenty-one percent of patients treated with ixabepilone plus capecitabine discontinued treatment because of sensory neuropathy in a large phase III trial (333). Severity increases with cumulative dosing, especially after an average of four treatment cycles. The overall neuropathy incidence varies depending on dose and coincident treatments but is as high as 67%. The reported reversibility of sensory neuropathy is surprising considering the experience with other microtubule targeting agents such as vincristine and taxanes. A meta-analysis of all available phase II and III trials found that overall rates of severe neuropathy (grade III/IV) in early previously untreated patients was approximately 1%; 24% of heavily treated metastatic breast cancer patients developed at least grade III neuropathy (347). When all neuropathy grades were considered, rates increased to 20% and 74%, respectively. In a study of 42 uterine cancer patients of which only 15% received prior chemotherapy, grade II chemotherapy-induced peripheral neuropathy was noted in 18%; unfortunately, the study found limited but insufficient tumor suppression efficacy (232). A study of skin punch biopsy found reduced epidermal nerve fiber density and ultrastructural axonal changes in 10 patients treated for breast cancer with ixabepilone. Abnormal mitochondrial appearance was also noted (103). One reported patient developed significant weakness associated with neuropathy after one treatment cycle (39). Carefully monitoring for neuropathy and timely dose adjustment or treatment discontinuation is advocated depending on the neuropathy severity (321). A review of multiple trials that assessed neuropathy incidence and severity is available (158). Another epothilone, sagopilone, has also been demonstrated to cause sensory neuropathy but is not FDA approved.

Other agents

Eribulin mesylate. The U.S. FDA approved this novel agent late in 2010 for use in refractory metastatic breast cancer, specifically in patients who have failed conventional treatment, including taxanes. The drug is a tubulin aggregating agent analogous to paclitaxel but differs from other drugs in its mechanism of action. The agent is in the halichondrin class of drugs, which has a high antimitotic activity against breast cancer and liposarcoma (328). The drug inhibits the microtubular growth phase without affecting the shortening phase, thus, causing tubulin sequestration into non-productive aggregates. The drug appears to have more microtubular stabilizing effects and paclitaxel more neurodegenerative effects on rat sciatic nerve preparations (25). The drug demonstrated less neuropathy than paclitaxel in preclinical studies but retained efficacy in resistant cell lines (78). Toxic neuropathy is a recognized side effect with use, but nearly all patients have received prior taxanes. Phase 2 studies concluded that the incidence of neuropathy exacerbation was small (73); however, grade 3 neuropathy occurred in 6.9% (79). The largest study found that neuropathy was the most common toxicity reason for treatment discontinuation, occurring in 24 of 503 patients (5%) (78). A study of eribulin mesylate and trastuzumab as first-line treatment for initially metastatic epidermal growth factor receptor 2-positive breast cancer patients in 52 women found grade 1 to 3 peripheral neuropathy in 66.7% of patients who had received prior trastuzumab and in 71.0% of patients who had not, supporting the neuropathy incidence despite a lack of prior chemotherapy-induced peripheral neuropathy-inducing agents (273). A meta-analysis of 4849 patients from 19 different trials found 27.5% neuropathy and 4.7% severe neuropathy incidence (264). Another literature review of 32 studies containing 6129 treated subjects found an overall neuropathy incidence of 28% (377). A comparison study of 110 women with breast cancer receiving either eribulin or vinorelbine adjunctive therapy found neuropathy occurred at a similar rate but the vinorelbine group showed signs suggestive of more autonomic neuropathy and slightly earlier onset after initial treatment (369). A phase 2 trial of initial treatment of 118 HER2-negative metastatic breast cancer patients comparing eribulin and gemcitabine versus paclitaxel and gemcitabine found lower neuropathy incidence and similar efficacy based on quality of life and neuropathy symptom scores (178). Eribulin combination or monotherapy is an option in patients with soft tissue sarcoma that have a contraindication for the standard chemotherapy agent, doxorubicin. One Japanese study of 28 patients found one discontinued treatment because of neuropathy (342).

Further studies support the lesser neurotoxicity of eribulin as compared to paclitaxel. A cohort of 99 breast cancer patients was randomized to various regimens, including paclitaxel plus carboplatin or eribulin plus carboplatin (229). Others received eribulin plus cyclophosphamide or eribulin plus capecitabine. In summary, the eribulin/carboplatin efficacy was equivalent to the paclitaxel/carboplatin cohort, but the incidence of neuropathy was notably less: 74% versus 32%. The other eribulin regimens that included no platin or taxane had 22% to 26% neuropathy incidence but lesser cancer efficacy. One case of demyelinating neuropathy and myokymia attributed to eribulin therapy has been reported (343). A multicenter phase IV study of eribulin in 170 Italian metastatic breast cancer patients found about half experienced a dose reduction for adverse reactions, most commonly neutropenia. Additionally, 14.7% developed chemotherapy-induced peripheral neuropathy sufficient to alter therapy (198). A cohort of 118 breast cancer patients were randomized to regimens including either eribulin or paclitaxel (253). Grade 1 or higher chemotherapy-induced peripheral neuropathy rates were significantly lower in the eribulin group. In an ongoing observational multicenter study of advance or metastatic breast cancer treated with eribulin, the incidence and progression of new, severe neuropathy after treatment was assessed (215). The group found roughly 6% had grade 3 or higher neuropathy, though many more had lesser neuropathy grades or exacerbation of prior neuropathy (15%). Many received prior chemotherapy regimens including taxanes, platins, or vinca alkaloids or other neuropathy risk factors.

Cetuximab. This biological agent was FDA approved in 2012 and is first-line treatment of certain head and neck cancers and for colorectal cancers that display certain genetic markers; the agent is an epidermal growth factor receptor inhibitor. Most treated patients also received conventional chemotherapy that included oxaliplatin or other platins; however, a slightly higher but not significant rate of sensory neuropathy was observed in cetuximab-treated patients. A neurotoxic link with this agent is not established, and neuropathy is not cited as a common form of toxicity. One phase III trial showed that less than 1% of patients treated with cetuximab (along with bevacizumab, leukovorin, and 5-FU) developed grade 3 neuropathy but did not address lesser grades (290). Another study did demonstrate that patients with advanced non-small cell lung cancer treated with paclitaxel and carboplatin concurrently with cetuximab had a higher rate of sensory neuropathy than those treated sequentially (146). There are case reports of cetuximab leading to a chemotherapy-induced peripheral neuropathy-like illness responsive to IVIG (31).

Cytosine arabinoside. Cytosine arabinoside (Ara-C) is used in the treatment of myelocytic leukemia and non-Hodgkin lymphoma. It is an analog of 2-deoxycytidine. Predominantly sensory peripheral neuropathy is seen with low dose Ara-C, whereas sensorimotor neuropathy occurs after high-dose therapy. Symptoms start from hours to 2 weeks after exposure and can take several forms. A pure sensory or sensorimotor neuropathy is most common. It can also produce a rapidly progressive ascending neuropathy. Bilateral brachial plexopathies have also been reported. A sensory neuropathy in one case manifested as pain in the toes and involuntary movement of the legs. Symptoms improve after discontinuing the medication but may persist for years (317).

High-dose Ara-C therapy causes a demyelinating polyneuropathy in approximately 1% of cases (260). This follows a progressive monophasic course, starting 2 to 3 weeks after initiation of therapy and produces severe motor weakness, even quadriparesis requiring ventilatory support. An acute cerebellar syndrome is the most common neurotoxic effect of Ara-C and is seen with intravenous administration. Intrathecal administration has been known to cause myelopathy and necrotizing encephalopathy.

The mechanism of the neuropathic effect is unknown. Ara-C incorporates into DNA and inhibits DNA polymerase and, thus, blocks DNA synthesis. In the rapidly progressive form associated with high-dose therapy, autopsy revealed demyelination with intact axons, raising the possibility of a direct neurotoxic effect on Schwann cells.

The occurrence of neuropathy is rare, with a reported incidence of 0.6%. It can occur after a wide range of doses, from 600 mg to 36 gms/m2 (317) but is more commonly associated with high doses (37).

Electrophysiologic studies demonstrate features consistent with both axonal and demyelinating neuropathy. This is consistent with nerve biopsy results, which have shown axonal swelling and segmental demyelination (37). In a case report, carbamazepine-produced relief of pain in the toes and involuntary leg movements.

Misonidazole. Misonidazole (2-nitroimidazol) is a radiosensitizer first used in the 1970s. It is currently not used in the United States. Misonidazole sensitizes hypoxic cells to the effect of ionizing radiation therapy. Misonidazole produces a dose-limiting sensory neuropathy, which develops within weeks of the initiation of therapy. The neuropathy is characterized by painful burning dysesthesia of the feet. The neuropathy typically resolves on discontinuation of therapy. On examination, patients have loss of ankle jerks, diminished pain, touch, vibration, and position sense. Strength is normal (235). Misonidazole may also be associated with encephalopathy. Intravenous administration produces central neurotoxicity but spares the peripheral nervous system (134). The drug is chemically related to the antibiotic metronidazole, a well-recognized cause of predominantly sensory neuropathy.

The exact cause of the neuropathy is unknown. Animal studies show axonal degeneration, especially sensory terminals, in intramuscular fibers (135). Human biopsy material shows mixed segmental demyelination and remyelination with axonal loss of all fiber sizes. Increased neurofilaments are seen with axon swelling (317).

In a prospective study, 39% of patients treated with misonidazole developed neuropathy (236). This appears to occur with cumulative doses greater than 18 gm (94).

In electrodiagnostic testing, decreased sensory nerve action potential amplitude is the most notable finding. Compound motor action potential amplitudes are usually normal. Nerve conduction velocities may be normal (223). Sural nerve biopsy has shown fragmented and disorganized myelin, paranodal demyelination, and a reduction in the total number of myelinated fibers accompanied by axon loss (235). (18)F-misonidazole is in use as a PET marker of tumor hypoxia (110).

Etoposide. Etoposide, or VP-16, is a member of a class of compounds known as epipodophyllotoxins. Etoposide exerts its antineoplastic effect by stimulating DNA-topoisomerase II to cleave DNA. It is a semisynthetic derivative of podophyllotoxin that has shown efficacy in the treatment of small-cell lung cancer, germ cell tumors, Hodgkin disease, and diffuse histiocytic lymphoma. VP-16 produces a mild, occasionally moderate length-dependent sensorimotor neuropathy. Its occurrence is related to both total dose and dosing intervals. Intracarotid etoposide with carboplatin has been implicated in a case of severe optic neuropathy with total external ophthalmoplegia (197). Etoposide more commonly causes headaches, seizures, and somnolence.

The exact mechanism of neuropathy is unknown. Increasing age of the patient, pronounced weight loss, and pre-existing neuropathy may be predisposing factors in the development of neuropathic toxicity (331). When coadministered with paclitaxel, the paclitaxel dose is the limiting factor in the development of neuropathy.

In a study involving etoposide plus cisplatin, co-administration of melatonin-reduced neurotoxicity and improved efficacy (205). It is not clear whether this effect would hold for etoposide alone.

Nerve conduction studies are consistent with both demyelination and axonal damage. Electron microscopy of biopsy material shows degeneration of myelin lamellae (331).

Management is limited to discontinuation of chemotherapy and symptomatic treatment.

Arsenic trioxide. Heavy metals are a rare but important cause of potentially treatable neuropathy. Therapeutic uses of heavy metals are primarily historical, but one agent remains in use. Arsenic trioxide is used in the treatment of recurrent acute promyelocytic leukemia. Neuropathy is a recognized side effect. Two of 12 patients developed grade 2 or 3 neuropathy in one series and 1 of 11 in another (126). A case of severe irreversible axonal neuropathy after exposure in the setting of unsuspected thiamine deficiency is described (189). Severe additional cases continue to be reported (145). The agent remains in use for this condition. Addition of bortezomib for relapsed patients appears to be successful but neuropathy was a concern; 1 of 22 patients developed severe grade 3 neuropathy (190). A neuropathy quality-of-life instrument addressing sensory, motor, and autonomic symptoms found similar results when comparing 162 patients with acute promyelocytic leukemia treated with arsenic trioxide or conventional chemotherapy (104).

An open-label, low-dose but dose-escalating arsenic trioxide study of 11 patients with systemic lupus erythematosus found that one patient developed axonal neuropathy, but concomitantly with a hepatitis A infection (139). A large Australian study of 487 acute promyelocytic leukemia patients treated with arsenic trioxide found 23% had some form of neurotoxicity, about 90% CIPN, including some with grade 3 or 4 (209). Obesity was a strong risk factor in subgroup analysis that might require dosing adjustment.

Immune checkpoint inhibitors. Immune checkpoint inhibitors are successful examples of monoclonal antibodies that enhance immune function designed to target cytotoxic lymphocyte-associated protein 4 (CTLA-4) and programmed cell death-1 (PD-1) for various cancers, notably melanoma and small cell lung cancers but indication is expanding. Proimmune reactions are recognized complications, including various neuromuscular forms. Ipilimumab, nivolumab, and pembrolizumab are examples that are associated with presumed inflammatory reactions including focal or generalized neuropathy; examples include focal phrenic, facial, optic, other cranial, and generalized acute or subacute neuropathy (154; 138). Peripheral and neuromuscular conditions appear to be more common than central nervous system toxicity. A search of 3763 treated patients in a global pharmacovigilance database found 35 had induced neurologic conditions including 22 neuropathy cases (196). The cases of acute neuropathy mimicking Guillain-Barre syndrome show more common lymphocytes in spinal fluid and over half show axonal neuropathy (138). A review of cases reported to the FDA adverse event reporting system (FAERS) tabulated neurologic adverse events reported through December 2019. Not surprisingly, the number of events is increasing over time from expanded treatment use, increased cancer types, and availability. Events were usually within 4 months of treatment onset; the median onset of toxicity was 1.6 months (239). Of the 3619 patients reported to have neurologic adverse events, 1357 were cited to have neuropathy (2.7% of overall reports) (239). One melanoma case report found difficulty distinguishing ipilimumab-induced chronic inflammatory demyelinating polyneuropathy (CIDP) from leptomeningeal tumor extension (48). A 65-year-old woman with unresectable metastatic melanoma developed a subacute Guillain-Barré-like neuropathy after two doses of pembrolizumab that repeatedly recurred in a CIDP-like pattern. Her melanoma initially responded but the treatment was interrupted. Her prolonged recovery led to melanoma progression while off therapy. A switch to ipilimumab produced melanoma regression but no neuropathy exacerbation (249). Another case initially starting acutely to suggest AIDP but relapsing after 8 weeks supported a CIDP diagnosis triggered by nivolumab (327). A German study of a large cohort of treated patients for various malignancies found 31 had neurologic complications, about a third with direct cancer or vascular causes. A review found nine patients were diagnosed with acute autoimmune polyneuropathy mimicking Guillain-Barré syndrome (GBS) supported by CSF features and electrophysiology; six others had myasthenia gravis and others had myositis (243). Additional cases of Guillain-Barré syndrome-like processes are reported (171). A case report noted Ramsay Hunt syndrome and acute ataxic sensory neuropathy in a patient treated with nivolumab. The facial neuropathy improved after steroids and antiviral therapy, but the neuropathy did not; both facial and generalized neuropathy improved after IVIg (289). Additional cases of rare sensory ganglionopathy are reported (80). A 44-year-old man with metastatic melanoma developed severe asymmetric progressive weakness following 23 doses of pembrolizumab (298). Cases of myasthenia gravis, transverse myelitis, myositis, chronic inflammatory demyelinating neuropathy, and GBS encephalitis are uncommon but reported (242). The incidence of these reactions is still emerging. A retrospective review of 347 Mayo clinic patients treated with PD1 inhibitors found 10 (2.9%) developed some subacute neurologic process after a median of 5.5 treatment cycles (range 1 to 20) (174). Of the 10 identified, four developed neuropathies. A case of antineutrophil cytoplasmic antibody-associated vasculitic neuropathy is also described (354). A literature review of all reported neurologic complications of checkpoint inhibitors found 197 papers reporting on 255 patients (177). Sixteen percent had acute or chronic neuropathy, a similar percentage to all central nervous system manifestations.

Other monoclonal antibodies. Numerous monoclonal antibody-based chemotherapy drugs are approved or under study. Some may cause chemotherapy-induced peripheral neuropathy, but most patients received prior chemotherapy. Some examples include pertuzumab, approved in 2012 for refractory breast cancer treatment—additional toxic neuropathy is suspected, but the agent is typically combined with docetaxel. There is a similar issue with ado-trastuzumab emtansine, approved for adjuvant breast cancer treatment in 2019.

Brentuximab vedotin received accelerated FDA approval in 2011 and full approval in 2017 for refractory Hodgkin lymphoma. Neuropathy is a recognized side effect though many patients received prior platin or vincristine treatments. A study of patients with T-cell leukemia found neuropathy in two thirds of treated patients (44 of 66) and only 4% of cases on either methotrexate of bexarotene; however, the leukemia response was markedly better in the brentuximab group (271). Axonal neuropathy is attributed to a toxic effect of monomethyl auristatin E on axonal microtubules (227). An open label German study of 104 newly diagnosed Hodgkin lymphoma patients, all of whom received initial brentuximab but not vincristine, found roughly one third developed neuropathy, mostly grades I and II—all but one more severe grade III case is reported to resolve (105). A case of severe motor neuropathy is reported (263). Five-year follow-up of 58 patients with uncommon anaplastic large cell lymphoma treated with brentuximab vedotin found 33 developed neuropathy, although 16 had preexisting neuropathy at enrollment. At the 5-year mark, 30 are noted to have resolved or significantly improved; the drug showed significant efficacy (272). A Japanese post marketing surveillance study of brentuximab vedotin for Hodgkin and large cell lymphoma found peripheral sensory neuropathy was common and was reported in 39.1% (grade ≥ 3, 6.3%) (161). The large ECHELON-2 trial treating 452 patients in 17 countries for peripheral T-cell lymphoma found somewhat superior survival for brentuximab vedotin, another agent cocktail, compared to conventional CHOP therapy, which includes vincristine (153). However, neuropathy rates were similar—about half of treated patients in both groups.

A retrospective French study found additional demyelinating neuropathy patterns in three that met Guillain-Barré syndrome and nine that met CIDP criteria—many developed sensory ataxia and distal weakness (113). Many stopped therapy because of neuropathy. Despite prior neurotoxic chemotherapy or stem cell transplant, all had neuropathy onset after the brentuximab vedotin. A large retrospective multicenter study assessed neuropathy incidence in lymphoma patients attributed to brentuximab vedotin, including patients not eligible for prior published trials (41). Treatment was discontinued in 23% because of neuropathy, much higher than the pivotal trials that led to approval (6.6%).

Polatuzumab vedotin, a similar antibody drug conjugate that delivers the suspected neuropathic agent monomethyl auristatin E, is also associated with neuropathy. Rates are likely 55% to 72% of patients with non-Hodgkin lymphoma (213). A B-cell lymphoma trial found neuropathy is about 40% of patients as part of initial therapy (336). However, the treatment in combination with bendamustine and rituximab appears to be superior to conventional chemotherapy (10). A Taiwanese study of 40 patients with relapsing or refractory large B-cell lymphoma treated with polatuzumab vedotin combined with rituximab-bendamustine found a favorable response (360). Grade 1 to 3 neuropathy was reported in 25% of patients. Neuropathy resolution after treatment completion is said to occur in most patients. A large phase 3 trial of B-cell lymphoma patients compared cocktails containing polatuzumab vedotin versus vincristine in addition to rituximab, cyclophosphamide, doxorubicin, and prednisone (311). Survival rates were better in the polatuzumab group; however, any grade chemotherapy-induced peripheral neuropathy rates were surprisingly similar (44.3% vs. 50.4%).

Enfortumab vedotin, another antibody-drug conjugate, delivers the microtubule-disrupting agent monomethyl auristatin E to cells that express Nectin-4. It was approved by the FDA in December 2019 to treat metastatic urothelial cancer. A study of 125 patients previously treated with platinum chemotherapy and PD-1 inhibitors found about half developed additional neuropathy but the agent delivered a clinically meaningful response (282). Treatment is often combined with pembrolizumab. Neuropathy was common in various clinical trials, reported to affect 50% to 60% of patients, and is a common treatment limiting side effect (151; 152).