Interventions for preventing neuropathy caused by cisplatin and related compounds
Abstract
Background
Cisplatin and several related antineoplastic drugs used to treat many types of solid tumours are neurotoxic, and most patients completing a full course of cisplatin chemotherapy develop a clinically detectable sensory neuropathy. Effective neuroprotective therapies have been sought.
Objectives
To examine the efficacy and safety of purported chemoprotective agents to prevent or limit the neurotoxicity of cisplatin and related drugs.
Search methods
On 4 March 2013, we searched the Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL, MEDLINE, EMBASE, LILACS, and CINAHL Plus for randomised trials designed to evaluate neuroprotective agents used to prevent or limit neurotoxicity of cisplatin and related drugs among human patients.
Selection criteria
We included randomised controlled trials (RCTs) or quasi‐RCTs in which the participants received chemotherapy with cisplatin or related compounds, with a potential chemoprotectant (acetylcysteine, amifostine, adrenocorticotrophic hormone (ACTH), BNP7787, calcium and magnesium (Ca/Mg), diethyldithiocarbamate (DDTC), glutathione, Org 2766, oxcarbazepine, or vitamin E) compared to placebo, no treatment, or other treatments. We considered trials in which participants underwent evaluation zero to six months after completing chemotherapy using quantitative sensory testing (the primary outcome) or other measures including nerve conduction studies or neurological impairment rating using validated scales (secondary outcomes).
Data collection and analysis
Two review authors assessed each study, extracted the data and reached consensus, according to standard Cochrane methodology.
Main results
As of 2013, the review includes 29 studies describing nine possible chemoprotective agents, as well as description of two published meta‐analyses. Among these trials, there were sufficient data in some instances to combine the results from different studies, most often using data from secondary non‐quantitative measures. Nine of the studies were newly included at this update. Few of the included studies were at a high risk of bias overall, although often there was too little information to make an assessment. At least two review authors performed a formal review of an additional 44 articles but we did not include them in the final review for a variety of reasons.
Of seven eligible amifostine trials (743 participants in total), one used quantitative sensory testing (vibration perception threshold) and demonstrated a favourable outcome in terms of amifostine neuroprotection, but the vibration perception threshold result was based on data from only 14 participants receiving amifostine who completed the post‐treatment evaluation and should be regarded with caution. Furthermore the change measured was subclinical. None of the three eligible Ca/Mg trials (or four trials if a single retrospective study was included) described our primary outcome measures. The four Ca/Mg trials included a total of 886 participants. Of the seven eligible glutathione trials (387 participants), one used quantitative sensory testing but reported only qualitative analyses. Four eligible Org 2766 trials (311 participants) employed quantitative sensory testing but reported disparate results; meta‐analyses of three of these trials using comparable measures showed no significant vibration perception threshold neuroprotection. The remaining trial reported only descriptive analyses. Similarly, none of the three eligible vitamin E trials (246 participants) reported quantitative sensory testing. The eligible single trials involving acetylcysteine (14 participants), diethyldithiocarbamate (195 participants), oxcarbazepine (32 participants), and retinoic acid (92 participants) did not perform quantitative sensory testing. In all, this review includes data from 2906 participants. However, only seven trials reported data for the primary outcome measure of this review, (quantitative sensory testing) and only nine trials reported our objective secondary measure, nerve conduction test results. Additionally, methodological heterogeneity precluded pooling of the results in most cases. Nonetheless, a larger number of trials reported the results of secondary (non‐quantitative and subjective) measures such as the National Cancer Institute Common Toxicity Criteria (NCI‐CTC) for neuropathy (15 trials), and these results we pooled and reported as meta‐analysis. Amifostine showed a significantly reduced risk of developing neurotoxicity NCI‐CTC (or equivalent) ≥ 2 compared to placebo (RR 0.26, 95% CI 0.11 to 0.61). Glutathione was also efficacious with an RR of 0.29 (95% CI 0.10 to 0.85). In three vitamin E studies subjective measures not suitable for combination in meta analysis each favoured vitamin E. For other interventions the qualitative toxicity measures were either negative (N‐acetyl cysteine, Ca/Mg, DDTC and retinoic acid) or not evaluated (oxcarbazepine and Org 2766).
Adverse events were infrequent or not reported for most interventions. Amifostine was associated with transient hypotension in 8% to 62% of participants, retinoic acid with hypocalcaemia in 11%, and approximately 20% of participantss withdrew from treatment with DDTC because of toxicity.
Authors' conclusions
At present, the data are insufficient to conclude that any of the purported chemoprotective agents (acetylcysteine, amifostine, calcium and magnesium, diethyldithiocarbamate, glutathione, Org 2766, oxcarbazepine, retinoic acid, or vitamin E) prevent or limit the neurotoxicity of platin drugs among human patients, as determined using quantitative, objective measures of neuropathy. Amifostine, calcium and magnesium, glutathione, and vitamin E showed modest but promising (borderline statistically significant) results favouring their ability to reduce the neurotoxicity of cisplatin and related chemotherapies, as measured using secondary, non‐quantitative and subjective measures such as the NCI‐CTC neuropathy grading scale. Among these interventions, the efficacy of only vitamin E was evaluated using quantitative nerve conduction studies; the results were negative and did not support the positive findings based on the qualitative measures. In summary, the present studies are limited by the small number of participants receiving any particular agent, a lack of objective measures of neuropathy, and differing results among similar trials, which make it impossible to conclude that any of the neuroprotective agents tested prevent or limit the neurotoxicity of platinum drugs.
PICOs
Plain language summary
Interventions for preventing nerve damage caused by cisplatin and other tumour‐inhibiting platinum drugs
Review question
We reviewed the evidence about the effect of treatments to prevent or reduce damage to nerves from the anticancer (chemotherapy) drug cisplatin or other platinum‐containing drugs.
Background
Cisplatin and other related platinum‐containing drugs that are used to treat solid tumours are toxic to the peripheral nervous system. Most people who complete a full course of cisplatin chemotherapy develop a sensory neuropathy (damage to nerves that carry sensation). Symptoms can include tingling in the extremities and numbness. The neuropathy may only partially recover or not recover at all when the chemotherapy is stopped. To try to reduce the toxicity of platinum drugs, researchers have looked for therapies to protect the nerves.
Study characteristics
We carried out a wide search for studies of treatments to prevent this type of nerve damage. We identified a total of 29 clinical trials, which involved almost 3000 participants who were receiving platinum‐containing anticancer drugs (mostly cisplatin, oxaliplatin and carboplatin) for various types of cancer (mainly colon, ovary, and lung cancers). The nine treatments studied were: amifostine (seven trials), calcium and magnesium (four trials), glutathione (seven trials), Org 2766 (four trials) and vitamin E (three trials). There was one trial each of acetylcysteine, diethyldithiocarbamate (DDTC), oxcarbazepine, and retinoic acid. We chose an objective clinical test of sensation to report as our preferred measure of the effects of treatment. Only seven of the studies used this measure. Nine reported the results of nerve conduction studies which are another objective measure of nerve function. Most of the studies used a subjective assessment of neuropathy, such as the National Cancer Institute‐Common Toxicity Criteria (NCI‐CTC) neuropathy grading scale.
Key results and quality of the evidence
Most of the included studies were fairly well performed, where it was possible to obtain this information.
Based on the combined results that generally described secondary and non‐quantitative measures such as the NCI‐CTC neuropathy grading scale, modest but promising (borderline statistically significant) results favoured the use of amifostine, calcium and magnesium, and glutathione to reduce the neurotoxicity of cisplatin and related chemotherapies. Three studies of vitamin E could only be studied individually but the results of each imply some mild subjective benefits. Nevertheless, given the limitations of the studies, such as small numbers of participants, lack of objective measures of neuropathy, and differing results among similar trials, the data remain insufficient to conclude that any of the neuroprotective agents tested prevent or limit the neurotoxicity of platinum drugs. Most of the treatments were not associated with adverse events. Amifostine infusions were associated with temporary low blood pressure in a significant number of cases, and retinoic acid with low levels of calcium in the blood. About one‐fifth of people treated with DDTC stopped taking it because of harmful effects.
Amifostine, calcium and magnesium, vitamin E, and glutathione require further well designed trials to clarify if they are effective or not.
This is an updated review. The evidence is current to 4 March 2013.
Authors' conclusions
Background
Cisplatin (cis‐diaminodichloroplatinum) was the first heavy metal used as an antineoplastic agent. It has been used since the early 1970s to treat several kinds of solid tumours, including lung, ovary, testis, bladder, head and neck, and endometrium (Mollman 1990; Prestayko 1979). Typical dosage regimes vary from 50 to 100 mg/m² given intravenously every three to four weeks, usually for about six cycles, based on clinical response and toxicity. Alternative schedules include 20 mg/m² daily for five days, or 20 mg/m² given weekly for about six weeks. Cisplatin is known to be toxic to the nervous system (Cavaletti 2004;Mollman 1990; van der Hoop 1990; Von Hoff 1979); it exhibits preferential uptake in the dorsal root ganglia and produces a dose‐related large fibre sensory neuropathy (neuronopathy). The sensory neuropathy most often becomes evident after a cumulative cisplatin dose of at least 300 mg/m², but occasional patients, especially those with risk factors, those with pre‐existing neuropathy, and those who are receiving combination chemotherapy, may develop symptoms after lower cumulative doses (Roelofs 1984; Windebank 1994). Most people completing a full course of cisplatin chemotherapy develop a symptomatic and clinically detectable sensory neuropathy. Symptoms, including unpleasant distal paraesthesias (tingling in the extremities) and numbness, may appear as soon as one month after initiating treatment. Lhermitte's symptom (an electric shock‐like sensation on bending the neck), likely caused by centripetal degeneration of posterior columns, has also been described. Associated signs include evidence of large fibre sensory loss (reduced vibration and joint position sensations) and diminished or absent muscle stretch reflexes (Roelofs 1984; Thompson 1984). Sensory ataxia (inco‐ordination) may be disabling in those who have severe neuropathy. Small fibre sensation is spared or mildly diminished (decreased pin‐pain sensation). Strength is generally normal. Symptoms and signs are symmetric and usually worse distally. Despite the high frequency of neuropathy among people treated with cisplatin, relatively few people develop functional impairment sufficient to interfere with activities of daily living (ADL). Nevertheless, neurotoxicity is a major reason that cisplatin is discontinued and the cumulative dosage limited, potentially reducing its chemotherapeutic efficacy. In an attempt to reduce the toxicity of platinum drugs, second and third generation tumour‐inhibiting platinum compounds including carboplatin, oxaliplatin, nedaplatin, and lobaplatin have emerged. Carboplatin lacks the nephrotoxicity of cisplatin, and neurotoxicity also is thought to be considerably less than that associated with cisplatin. Carboplatin is used to treat ovarian, small cell lung, and refractory testicular cancers. Oxaliplatin is now FDA and European Union approved for use in the treatment of metastatic colon cancer; neurotoxicity is the major dose‐limiting adverse‐event (Grothey 2003; Grothey 2004).
Quantitative measures of vibration perception threshold (VPT) and sensory nerve action potential (SNAP) amplitude are complementary means of evaluating large sensory fibres. Both measures have been used to document the development and progression of cisplatin‐induced neuropathy (Wald 1994). Sequential nerve conduction studies among people receiving cisplatin chemotherapy demonstrate progressive reduction of SNAP amplitudes, with little or no change in motor nerve conduction studies and needle electromyography (EMG). Reduction of the SNAP amplitude is thought to be a relatively sensitive indication of early cisplatin‐induced neurotoxicity. This objective change occurs early in the course of cisplatin neuropathy, often before development of symptoms or signs of sensory neuropathy. It is generally believed that most patients receiving cisplatin show a sequential decline in SNAP amplitude relative to baseline (Wald 1994; Wald 1995). A decline which exceeds 40% from baseline is used by some as the physiological criterion for sensory neuropathy associated with chemotherapy toxicity (Molloy 2001). Sural nerve biopsy shows evidence of Wallerian‐like axonal degeneration affecting the large myelinated fibres, with scant evidence of regeneration, consistent with damage at the level of the dorsal root ganglia. Unmyelinated fibres are spared (Krarup‐Hansen 1993). Cisplatin‐induced neuropathy stabilises or improves after discontinuing cisplatin (Elderson 1989), but clinical and electrodiagnostic deterioration often progresses for a few weeks after completing cisplatin treatment (called 'coasting'). With higher cumulative dosages, however, residual nerve damage may persist for long periods of time.
Experimental models of cisplatin neurotoxicity confirm that the neuronal cell body within the dorsal root ganglion is the site of injury. Cultured rat embryo dorsal root ganglion models have been used to study the mechanisms of cisplatin neurotoxicity (Windebank 1994). Cisplatin reproducibly inhibits axonal growth in a dose‐dependent manner that includes concentrations similar to those known to produce human toxicity. The mechanism of action is thought to be related to platinum binding to DNA and interfering with DNA synthesis. Cisplatin produced abnormalities in the nucleoli of spinal root ganglion cells of large and small neurons (Tomiwa 1986). The hypothesis that cisplatin‐induced neuropathy may result from nuclear and nucleolar changes in the sensory ganglion cell body was confirmed in the rat (Cavaletti 1992). After chronic cisplatin administration, the spinal ganglia and peripheral nerves showed severe damage of the spinal ganglia neurons with predominant involvement of the nucleus and nucleolus associated with a decrease in the cell size after chronic cisplatin administration. Changes also occurred in the sciatic and peroneal nerves with the features of axonopathy (Cavaletti 1992). These changes described in rats also have been confirmed in mice (Carozzi 2010). In addition, Schmidt 1995 showed in a mouse model that nerve growth factor (NGF) exerts a major effect on the metabolism of transmitters associated with nociception, pain and sensation in cervical dorsal root ganglia in various models of toxic neuropathy, including the neuropathy produced by cisplatin. Cisplatin specifically induced a significant decrease in the number of large‐ and medium‐sized neurons in the dorsal root ganglia, indicating neuronal atrophy, a finding that correlated with a highly significant loss of neuropeptides in cervical dorsal root ganglia in mice (Schmidt 1995).
Prevention of cisplatin neurotoxicity
In most instances of cisplatin‐induced neuropathy, the nerves only partially recover or do not recover at all. For this reason, effective neuroprotective therapies have been sought. As therapeutic strategies were developed to maximise cisplatin effectiveness, such as minimising renal insufficiency, cumulative dose‐related neuropathy emerged as a major dose‐limiting toxicity (Gandara 1991). Attempts to modulate cisplatin dose schedules did not influence the intensity of the resultant neurotoxicity (Hilkens 1995). Additional means of optimising the therapeutic index of cisplatin were sought, such as coadministration of chemoprotective or rescue therapies to reduce adverse side effects without reversing antitumour activity. An ACTH (4‐9) analogue Org 2766, glutathione (GSH), amifostine, and various neurotrophic growth factors have all been tried in clinical and experimental models to prevent cisplatin‐induced neurotoxicity, and reviews exist of the clinical pharmacology and therapeutic efficacy of several chemoprotectants (e.g. Links 1999). Acetyl‐L‐carnitine has recently been reported to have a neuroprotective effect in cisplatin and paclitaxel models (Pisano 2003). Both GSH and vitamin E have been tested in human trials with platinum drugs (Cascinu 1995; Cascinu 2002; Pace 2003; Smyth 1997). Although the precise mechanism of cisplatin‐induced neurotoxicity is unknown, various agents have been proposed to protect the peripheral nervous system from such neurotoxicity. For example, amifostine, an organic thiophosphate that acts as a scavenger of oxygen free‐radicals, shows selective protection of normal tissues (cytoprotective) against toxicities induced by radiation, alkylating agents, and platinum compounds without influencing the antitumour effects of these treatments (Planting 1999). Diethyldithiocarbamate (DDTC) is a chelating agent thought to bind and remove tissue‐bound platinum without interfering with the antitumour activity of cisplatin (Gandara 1995). Glutathione (GSH) is a nucleophilic sulphur‐containing tripeptide thiol thought to permit delivery of higher doses of cisplatin without producing the expected neurotoxicity, perhaps by preventing the initial accumulation of platinum adducts in the dorsal root ganglia (Cascinu 1995; Cascinu 2002). The neuropeptide Org 2766 is thought to potentially ameliorate cisplatin neuropathy by exerting trophic effects and enhancing endogenous nerve repair mechanisms (as opposed to directly protecting against cisplatin neurotoxicity) (van Gerven 1994). Vitamin E is an antioxidant thought to protect against cisplatin‐induced oto‐ and nephro‐toxicities and potentially to protect against neurotoxicity (Pace 2003). Acetylcysteine (N‐acetylcysteine, NAC) is a nutritional supplement thought to increase whole blood concentrations of glutathione, a useful agent for preventing the initial accumulation of platinum adducts and clinical oxaliplatin‐induced neurotoxicity (Lin 2006). BNP7787 (disodium 2, 2' dithio‐bisethane sulfonate) was developed as a putative chemoprotective agent (Miller 2008). Calcium and magnesium infusions were proposed to act as chelators of oxalate, thereby reducing the effect of oxalate on voltage‐gated sodium channels (Gamelin 2004). Oxcarbazepine resembles the antiepileptic drug carbamazepine and also blocks voltage‐sensitive sodium channels and certain calcium channels. Oxcarbazepine was developed as a neuroprotective against oxaliplatin‐induced neuropathy, which is thought to reflect the alteration of voltage‐gated sodium channels by oxalate, a metabolite of oxaliplatin (Argyriou 2006a).
Description of the intervention
Several different interventions have been described as being effective in preventing platinum‐induced neuropathy. A complex description of each intervention is beyond the scope of this review.
-
Acetylcysteine (NAC) is a nutritional supplement thought to increase whole blood concentrations of glutathione, a useful agent for preventing the initial accumulation of platinum adducts and clinical oxaliplatin‐induced neurotoxicity (Lin 2006).
-
Amifostine is an organic thiophosphate described as a broad‐spectrum cytoprotective agent potentially able to protect normal tissues from cytotoxic effects of chemotherapeutic agents (Kanat 2003).
-
Calcium and magnesium infusions may chelate oxalate (a metabolite of oxaliplatin), thereby reducing the effect of oxalate on voltage‐gated sodium channels (Ishibashi 2010).
-
Diethyldithiocarbamate (DDTC) is a heavy metal chelating agent and has been reported to reduce the incidence and severity of cisplatin‐induced neuropathy (Gandara 1995).
-
Glutathione (GSH). The mechanism of neurotoxicity from platinum drugs is thought to be the accumulation of platinum within the dorsal root ganglia. The thio nucleophilic region of reduced GSH has high affinity for heavy metals and may be able to prevent the accumulation of platinum in the dorsal root ganglia.
-
Org 2766 is an ACTH (4‐9) analogue which has been reported to reduce the incidence and severity of cisplatin‐induced polyneuropathy (van der Hoop 1990).
-
Oxcarbazepine resembles the antiepileptic drug carbamazepine and also blocks voltage‐sensitive sodium channels and certain calcium channels. It was developed as a neuroprotective against oxaliplatin‐induced neuropathy, which is thought to reflect the alteration of voltage‐gated sodium channels by oxalate, a metabolite of oxaliplatin (Argyriou 2006).
-
Retinoic acid. Retinoids play a role in a variety of biological functions; all‐trans retinoic acid (ATRA) binds and activates subtypes of retinoid acid receptors and stimulates nerve growth factor (NGF) and the expression of its receptor. ATRA prevents and reverses neuropathy‐associated morphological changes in diabetic animals (Arrieta 2011).
-
Vitamin E, an antioxidant protecting against free‐radical injury, may be protective against cisplatin‐induced polyneuropathy. Based on the observation that deficiency of vitamin E produces a pattern of peripheral neuropathy similar to cisplatin‐induced toxic neuropathy and the fact that decreased plasma vitamin E levels are observed in patients with cisplatin neuropathy, trials of vitamin E for prevention of toxic neuropathy have been done (Pace 2003).
Why it is important to do this review
Whilst platinum‐containing chemotherapy can be highly effective for the treatment of a number of cancers, its use and dosing is not infrequently limited by neurotoxicity. As well as limiting choice and efficacy of chemotherapy agents the disability caused by the neuropathy can be devastating and irreversible. Agents that limit or prevent toxicity from platinum containing chemotherapy could be very important adjuncts to oncology regimens.
Objectives
To examine the efficacy and safety of purported chemoprotective agents to prevent or limit the neurotoxicity of cisplatin and related drugs.
Methods
Criteria for considering studies for this review
Types of studies
We included all randomised or quasi‐randomised controlled human trials in which the efficacy of any form of chemoprotective treatment used to prevent or limit the neurotoxicity of cisplatin (or related oncologic platinum compounds including oxaliplatin or carboplatin) was compared with placebo, no treatment, or other treatments.
Types of participants
Adult participants of either sex undergoing chemotherapy with cisplatin (or related oncologic platinum compounds including oxaliplatin or carboplatin) as an antineoplastic medication.
Types of interventions
We included in the review any form of chemoprotective treatment, such as acetyl‐L‐carnitine, acetylcysteine (NAC), adrenocorticotrophic hormone (ACTH), amifostine, BNP7787, calcium and magnesium, Org 2766, glutathione, oxcarbazepine, vitamin E, and growth factors, used to prevent or limit cisplatin‐induced neurotoxicity.
Types of outcome measures
Primary and secondary outcomes were evaluated in the zero to six month period after completing or discontinuing chemotherapy. When more than one evaluation occurred during this interval, we selected the one closest to three months after completing chemotherapy treatment.
Primary outcomes
The primary outcome measure was the change in quantitative sensory testing (QST) results (e.g., vibration perception threshold (VPT). Measures of VPT have been used in several studies of cisplatin toxicity. Although VPT does not have the specificity of nerve conduction study results, the sensitivity of QST is likely similar to nerve conduction study results.
Secondary outcomes
Secondary outcome measures included nerve conduction study results (SNAP amplitude) and measures of neurological impairment. Any other available measure of impairment was considered, but priority was given to those clinical, functional, or electrodiagnostic measures that used a validated scale. Examples of secondary outcome measures included the following.
-
SNAP amplitude.
-
Clinical impairment measured by neurological examination, using a validated scale.
-
Functional measures of activities of daily living (ADL).
-
Information from toxicity rating scales.
-
Serious adverse events of chemoprotective treatment, which were fatal, life threatening, or require prolonged hospital admission.
The National Cancer Institute‐Common Toxicity Criteria for Peripheral Sensory Neuropathy (NCI‐CTC for neuropathy) is an example of a neurotoxicity scale commonly used in clinical practice and frequently used in the studies we reviewed. Despite the wide use of the NCI‐CTC, it is not a validated scale, which is a major limitation (Cavaletti 2010a; Cavaletti 2013). The NCI‐CTC as used in many of the studies we reviewed exists in slightly different forms. One form has the following grades: 1) Asymptomatic: loss of deep tendon reflexes or paraesthesias; 2) Moderate symptoms: limiting instrumental ADL; 3) Severe symptoms: limiting self care ADL; 4) Life‐threatening consequences: urgent intervention indicated; and 5) Death. The second form uses grades as follows: 1) Asymptomatic: loss of deep tendon reflexes or paraesthesias (including tingling), but not interfering with function; 2) Sensory alteration or paraesthesias (including tingling) interfering with function, but not ADL; 3) Sensory alteration or paraesthesias interfering with ADL; and 4) Disabling. In the studies we reviewed, few participants developed grade 4 neuropathy, using either scale, and we elected to emphasise neuropathy based on participants showing grades ≥2 from either scale. In our opinion, such grading likely represents similar degrees of neuropathy, and separation based on which of the scales used limited the analyses we could perform (which was already severely limited by the few studies using our prespecified end points).
Search methods for identification of studies
Electronic searches
On 4 March 2013, we searched the Cochrane Neuromuscular Disease Group Specialized Register, CENTRAL (2013, Issue 2 in The Cochrane Library), MEDLINE (January 1966 to February 2013), EMBASE (January 1980 to February 2013), LILACS (January 1982 to February 2013), and CINAHL Plus (January 1937 to February 2013).
Our search strategy was not restricted by language. In the initial and updated review, we did not approach pharmaceutical companies, but, in a few instances, we did contact the authors of the trials for additional information. We did not formally screen the references of the selected papers.
The detailed search strategies are in the appendices: Appendix 1 (CENTRAL); Appendix 2 (MEDLINE); Appendix 3 (EMBASE); Appendix 4 (CINAHL Plus); and Appendix 5 (LILACS).
Data collection and analysis
Selection of studies
Both for the original review and updates, two review authors (JA and VC or GC) independently screened the titles and abstracts of references to select potential studies for inclusion The review authors obtained the full text of all potentially relevant studies for independent assessment by two authors for each article (among JA, GC, VC, and RD). The authors decided which studies fitted the inclusion criteria and graded their risk of bias. The two authors assessing each article resolved any disagreements about inclusion by communication, resulting in consensus.
Data extraction and management
Two review authors performed data extraction for each article and a third author (JA) checked the data extraction to identify differences of opinion. We began by separating the studies into treatment categories: for example, amifostine, glutathione, Ca/Mg, DDTC (diethyldithiocarbamate), Org 2766, oxcarbazepine, ATRA and vitamin E. We assigned two authors to each of the categories (combining categories with few studies, such as DDTC and vitamin E). We next tabulated whether or not individual studies showed dose‐response efficacy (if more than one dose was studied). We summarised the types of outcomes reported (QST, nerve conduction studies, clinical, functional measure). We included consideration of serious adverse events and determined if information on cost benefits was available.
Assessment of risk of bias in included studies
We assessed risk of bias in included studies using the Cochrane Collaboration 'Risk of bias' tool, described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008, updated Higgins 2011). We considered the adequacy of studies in the following domains: sequence generation, allocation concealment, blinding (subject, observer and assessor), incomplete outcome data addressed, selective reporting, and other sources of bias. Assessments were 'low risk of bias', 'high risk of bias', and 'unclear risk of bias' (where there was insufficient information to make a judgment or when the criterion did not apply to the study). At least two review authors evaluated each study and consensus agreement was reached.
Data synthesis
For studies using a common outcome measure, we calculated a treatment effect across trials and expressed the results as risk ratios (RR). For continuous outcomes, where meta‐analysis was possible, we calculated a mean difference and expressed the results with confidence intervals.
When the results indicated unexplained heterogeneity, we undertook these tests using a random‐effects model; if not, we used a fixed‐effect model. When there was unexplained heterogeneity, we investigated its source by repeating the analysis after elimination of trials that were at high or unclear risk of bias on each of the risk of bias domains, paying particular attention to allocation concealment. We investigated heterogeneity beginning with visual inspection of forest plots to determine if the confidence intervals (CIs) were not overlapping or there were outliers to suggest statistical heterogeneity. As a general guideline, we used a Chi‐squared statistic greater than the degrees of freedom to indicate heterogeneity, recognising that a non‐significant result does not necessarily mean the absence of heterogeneity when the studies are small/few. As a rule of thumb, we used I2, describing heterogeneity as moderate for 30% to 60%, substantial for I2 of 50% to 90% and considerable for I2 of 75% to 100%. When there was unexplained heterogeneity, we repeated the analyses using a random‐effects model to see whether the results differed. The possible outcomes were: 1. effect estimate is the same, 2. Effect estimate is similar, with different CIs (we used the random‐effects model), or 3. Effect estimate is different, in which case we performed sensitivity analyses repeating the analyses excluding studies at high risk of bias (e.g. on allocation concealment), large studies, unpublished studies, or other studies, depending on any suspected source of heterogeneity.
Results
Description of studies
The initial search identified 135 articles. From these, we selected 39 articles for complete review. We identified 15 articles fulfilling inclusion criteria. The 2010 search identified 275 articles (including many duplications and previously identified), from which we selected 11 new studies for complete review, of which we included five (a total of 20 included studies).
The 2013 search produced the following results: Cochrane Neuromuscular Disease Group Specialized Register 76 papers (19 new), CENTRAL 79 papers, MEDLINE 263 papers (34 new), EMBASE 193 papers (41 new), CINAHL Plus 50 papers (14 new), and LILACS 1 paper. The 2013 search resulted in selection of 21 articles for complete review, from which eight trials fulfilled inclusion criteria. One additional retrospective article was included (for reasons described below) and two articles describing the results of meta‐analyses were excluded but selected for the Discussion, for comparison to our interpretation of the data. We documented reasons for excluding studies (see Characteristics of excluded studies). A total of 29 studies (2906 participants) are included in the review. The combined articles involved the following potential chemoprotectant treatments:
-
acetylcysteine (NAC) (one study);
-
amifostine (seven studies);
-
calcium and magnesium (four studies);
-
diethyldithiocarbamate (DDTC) (one study);
-
glutathione (GSH) (seven studies);
-
Org 2766 (four studies);
-
oxcarbazepine (one study);
-
retinoic acid (one study); and
-
vitamin E (three studies).
The characteristics of the individual studies, including the 'Risk of bias' assessments, are summarised in the table Characteristics of included studies.
Acetylcysteine (N‐acetylcysteine, NAC)
Our 2010 update identified one article fulfilling the selection criteria and reporting the results of a randomised controlled trial (RCT) designed to evaluate the protective effects of NAC against oxaliplatin‐induced neurotoxicity (Lin 2006). This study, described as a "pilot study" involved only a small number of participants. Oxaliplatin dose was discontinued or dose reduced to 75% of previous dose if the participants developed NCI‐CRC grade 3 or 4 neuropathy or had National Cancer Institute (NCI) grade 1 neuropathy respectively. Additionally, calculation of the score was arbitrarily modified (sum of worst neurological toxicity divided by number of assessable participants at each dose level). Trial participants received:
-
oxaliplatin for colorectal cancer (five NAC‐treated and nine placebo control participants) (Lin 2006).
Participants received oral NAC (1200 mg) one and one half hours before each oxaliplatin administration, occurring every two weeks for up to 12 treatment cycles.
Amifostine
Our search identified nine articles describing a total of eight studies fulfilling the selection criteria and reporting the results of RCTs designed to evaluate the protective effects of amifostine against chemotherapy neurotoxicity. Two articles with different first authors (Kemp 1996; Rose 1996), described the same trial and were considered together for this review (Kemp 1996). One of the studies identified in the 2013 review (DeVos 2005), described a meta‐analysis of studies involving amifostine infusions for the prevention of neurotoxicity related to carboplatin (with paclitaxel); we included a review of that meta‐analysis in the Discussion. The seven eligible trials included 372 amifostine‐treated participants and 371 control participants. These trials included participants being treated with:
-
cisplatin for advanced head and neck cancer (37 amifostine‐treated participants and 37 control participants) (Planting 1999);
-
paclitaxel and carboplatin for non‐small cell lung cancer (19 amifostine‐treated and 19 control participants) (Kanat 2003);
-
paclitaxel and carboplatin for ovarian cancer (93 amifostine‐treated and 94 control participants) (Lorusso 2003);
-
cisplatin and cyclophosphamide for advanced ovarian cancer (122 amifostine‐treated and 120 control participants) (Kemp 1996);
-
oxaliplatin (FOLFOX‐4) for advanced or relapsed colorectal or gastric cancer with variable metastatic disease to liver, lung, and other sites (46 amifostine‐treated and 46 control participants) (Lu 2008);
-
carboplatin with paclitaxel for advanced ovarian carcinoma (45 amifostine‐treated and 45 control participants (DeVos 2005); and
-
radiotherapy and cisplatin for cervical cancer (10 amifostine‐treated participants and 10 control participants) (Gallardo 1999)
The amifostine regimens varied only slightly among trials, and all participants randomised to the amifostine groups received intravenous amifostine immediately before chemotherapy at doses of 740 mg/m² (DeVos 2005; Planting 1999), 910 mg/m² (Kanat 2003; Kemp 1996; Lorusso 2003), or 500 mg/m² (Lu 2008). All trials involved scheduled treatments of up to six cycles of chemotherapy, either weekly (Planting 1999), at three‐week intervals (DeVos 2005; Kanat 2003; Kemp 1996), or in two‐ or four‐week cycles (Lu 2008).
Calcium and magnesium (Ca/Mg)
The 2010 and 2013 updates identified three articles fulfilling the selection criteria and reporting the results of an RCT to evaluate the protective effects of calcium and magnesium infusions against oxaliplatin neurotoxicity (Chay 2010; Grothey 2011; Ishibashi 2010). Grothey 2011 and Ishibashi 2010 included participants being treated with six cycles of FOLFOX, which includes oxaliplatin (85 mg/m²) every two weeks. The Chay 2010 trial included participants being treated with up to eight cycles of either XELOX, which includes oxaliplatin 130 mg/m² on day one every three weeks (received by 22 or 27 participants) or FOLFOX‐4, which includes oxaliplatin (85 mg/m²) every two weeks. This study was terminated prematurely because of negative results from the CONcePT trial. The initial three eligible trials included 76 Ca/Mg‐treated participants and 78 control participants. One additional article describing a retrospective, non‐randomised study of Ca/Mg used at the discretion of treating physicians (as part of a large multicentre RCT of different chemotherapy drugs) was included because neither the participants nor the investigators appeared aware that Ca/Mg was being (or was going to be) investigated (Knijn 2011). When data from this retrospective study were included, there were 627 Ca/Mg‐treated participants and 259 control participants. The 2013 review also identified two papers that described meta‐analysis of studies involving Ca/Mg infusions for the prevention of oxaliplatin‐related neurotoxicity (Ao 2012; Wen 2013); we included these studies in the Discussion.
-
oxaliplatin for colorectal cancer (9 Ca/Mg‐treated and 10 control participants) (Chay 2010);
-
oxaliplatin for colon cancer (50 Ca/Mg‐treated and 52 control participants) (Grothey 2011);
-
oxaliplatin for metastatic colorectal cancer (17 Ca/Mg‐treated and 16 control participants) (Ishibashi 2010); and
-
oxaliplatin plus capecitabine, and bevacizumab with or without cetuximab for previously untreated colorectal cancer (551 Ca/Mg‐treated and 181 control, at the discretion of the treating physician) (pseudo‐random and retrospective) (Knijn 2011).
In two studies, investigators administered calcium gluconate 1 g (10 mL of 10% calcium gluconate) + 15% magnesium sulfate 1 g diluted into 100 mL of normal saline infused over 15 minutes before and after oxaliplatin infusion (Chay 2010; Grothey 2011). In addition to oxaliplatin FOLFOX chemotherapy includes administration of 5‐fluorouracil and l‐leucovorin with each of six, two‐week treatment cycles. Knijn 2011 infused 2.25 mmol calcium glubionate plus 4 mmol magnesium chloride in 100 mL glucose over 15 minutes, before and after oxaliplatin infusion (retrospective analyses). Ishibashi 2010 administered Ca/Mg (850 mg/720 mg, respectively) before and after infusion of oxaliplatin or placebo.
Diethyldithiocarbamate (DDTC)
Our search identified only one article fulfilling the selection criteria and reporting the results of an RCT designed to evaluate the protective effects of DDTC against cisplatin neurotoxicity (Gandara 1995). The trial included participants being treated with six courses of 100 mg/m² cisplatin every four weeks at 1.6 g/m²:
-
cisplatin for ovarian, small cell lung cancer (SCLC), and non‐small cell lung cancer (NSCLC) (data available for 96 DDTC‐treated participants and 99 control (placebo‐treated) participants (Gandara 1995).
DDTC, at 1.6 g/m², was administered 15 minutes before the start of cisplatin. In addition to cisplatin, other chemotherapies used were etoposide 100 mg/m² on days 1, 2, and 3 for SCLC and NSCLC, and cyclophosphamide 750 mg/m² for ovarian carcinoma as six courses every four weeks.
Gluthathione (GSH)
Our search identified seven articles fulfilling the selection criteria that reported the results of RCTs designed to evaluate the protective effects of GSH against platinum drugs (cisplatin and oxaliplatin) neurotoxicity. The seven prospective, randomised, and placebo‐controlled studies included 193 GSH‐treated participants and 194 control participants. The two platinum drugs used were cisplatin (153 and 155 respectively in each group and oxaliplatin (40 in the GSH group and 39 in the control group). The dose of cisplatin was variable (from 40 to 100 mg/m²) and oxaliplatin 100 mg/m² for every one to two weeks for three to nine weeks.
These trials included participants being treated with:
-
cisplatin for ovarian cancer (27 GSH‐treated participants and 27 control participants) (Bogliun 1996);
-
cisplatin for ovarian cancer (25 GSH‐treated participants and 25 control participants) (Cascinu 1995);
-
oxaliplatin for colorectal cancer (26 GSH‐treated participants and 26 control participants) (Cascinu 2002);
-
cisplatin for relapsing ovarian cancer (16 GSH‐treated participants and 17 control participants) (Colombo 1995);
-
oxaliplatin for colorectal cancer (14 GSH‐treated participants and 13 control participants) (Milla 2009);
-
cisplatin for non‐small cell lug cancer (NSCLC) and head and neck cancer (11 GSH‐treated participants and 9 control participants) (Schmidinger 2000); and
-
cisplatin for ovarian cancer (74 GSH‐treated participants and 77 control participants) (Smyth 1997).
The GSH dose was variable from 1.5 g/m² to 5 g before chemotherapy. The end points used were a toxicity score NCI/World Health Organization (WHO)) in all participants; functional scales (Hospital Anxiety and Depression (HAD) and Rotterdam scales) in 151 participants; Neuropathy Impairment Scale/Neuropathy Symptom Scale (NIS/NSS) in 54 participants; sensory nerve conduction studies (sural, median, or ulnar) in 156 participants; motor nerve conduction studies in 20 participants; and vibration perception threshold (VPT) testing in 54 participants.
Org 2766
Our search identified four articles fulfilling the selection criteria and reporting the results of RCTs designed to evaluate the protective effects of Org 2766 against cisplatin neurotoxicity. The four trials included 188 Org 2766‐treated participants and 123 control participants. These trials included participants being treated with:
-
cisplatin and cyclophosphamide for epithelial ovarian cancer (33 Org 2766‐treated participants and 22 control participants) (van der Hoop 1990);
-
cisplatin and cyclophosphamide for epithelial ovarian cancer (7 Org 2766‐treated and 11 control participants) (Hovestadt 1992);
-
cisplatin and different combinations of etoposide, bleomicin and ifosphamide for testicular and adenocarcinoma of unknown primary (19 Org 2766‐treated and 23 control participants) (van Gerven 1994); and
-
cisplatin and cyclophosphamide for epithelial ovarian cancer (129 Org 2766‐treated and 67 control participants) (Roberts 1997).
The Org 2766 regimens largely varied among trials in the range 0.25 to 4.0 mg/kg. All participants randomised to the Org 2766 groups received subcutaneous Org 2766 before and one hour after cisplatin in one trial (Roberts 1997), before and 24 hours after cisplatin in two trials (Hovestadt 1992; van der Hoop 1990) and during the daily course of cisplatin administration in one trial (van Gerven 1994). All trials involved scheduled treatments of up to nine cycles of chemotherapy, at three (Hovestadt 1992; van der Hoop 1990; van Gerven 1994) or three‐to‐four week intervals (Roberts 1997). Cisplatin was administered at different doses/cycle: 75 mg/m² (Hovestadt 1992; van der Hoop 1990), 75 to 100 mg/m² (Roberts 1997) or 100 mg/m² (van Gerven 1994).
Oxcarbazepine (OXC)
Our search identified one article fulfilling the selection criteria and designed to evaluate the protective effects of OXC against cisplatin neurotoxicity (Argyriou 2006a). Argyriou 2006a was prospective and randomised but without a placebo control arm (that is, control participants were untreated). The trial participants received:
-
cisplatin for colon cancer (1782 mg cumulative OXC‐treated participants and 1750 mg cumulative control participants) (Argyriou 2006a).
Argyriou 2006a included 40 patients being treated with cisplatin for colon cancer, of whom 20 were randomised to receive OXC, 600 mg twice a day, and 20 were randomised to receive no treatment. The cumulative oxaliplatin dose after 12 courses was similar in both groups (1782 mg OXC versus 1750 mg control). Ultimately, 16 participants in each group completed the trial, and data from the 32 participants were available for analysis. Neurological symptoms and neurophysiological testing (sural, superficial peroneal, and ulnar sensory nerve action potential (SNAP) and peroneal motor responses) were recorded at baseline and after 4, 8, and 12 courses of chemotherapy (that is, the last neurophysiological tests were performed at completion of 24 weeks of treatment).
Retinoic acid, all‐trans (ATRA)
Our search identified one article fulfilling the selection criteria and designed to evaluate the protective effects of ATRA against cisplatin neurotoxicity (Arrieta 2011). Arrieta 2011 was a prospective, randomised, placebo‐controlled study that included 92 participants being treated with cisplatin and paclitaxel for stage IIIB/IV NSCLC cancer, of whom 45 were randomised to receive ATRA, 20 mg/m²/day one week prior to chemotherapy until completion of two courses. The length of follow‐up was only until immediately after the second course of therapy (three weeks). Motor and sensory response amplitudes were recorded, but the electrophysiological results were summarised as "grade of damage," ranked from zero to three for latency and amplitude measures. In addition, clinical examination results and NCI‐CTC version 3.0 were recorded for each participant. The trial participants received:
-
cisplatin and paclitaxel for stage IIIB/IV NSCLC cancer (45 ATRA‐treated participants and 47 control participants) (Arrieta 2011).
Vitamin E
Our original search identified only one article fulfilling the selection criteria of RCT to assess the protective effect of vitamin E against cisplatin neurotoxicity (Pace 2003). In that study, the cumulative cisplatin dose was > 300 mg/m² (Pace 2003). Alpha tocopherol (vitamin E) was used at a dose of 300 mg/day starting before cisplatin and continuing up to three months after cisplatin treatment. Unmasked neurologists recorded neurological symptoms and signs and investigators performed neurophysiological testing (sural and median sensory nerves) at baseline, after three cycles, and after completion of chemotherapy. A cumulative toxicity score was assigned. The 2010 update identified a second trial of similar design (Argyriou 2006). Argyriou 2006 included six courses of cisplatin chemotherapy in various doses. The treatment protocols in both studies were specific for the individual cancers, and all included chemotherapy drugs in addition to cisplatin. Of note, five participants with gastric cancer also received docetaxel, another potential neurotoxicant (two participants in the vitamin E group and three control participants). Thirty‐five participants were enrolled; of whom 30 completed the trial. Of these, 14 participants received vitamin E (600 mg/day) starting before cisplatin and continuing up to three months after cisplatin treatment. Unmasked neurologists recorded neurological symptoms and signs and investigators performed neurophysiological testing (sural and median sensory nerves) at baseline, after three cycles, and after completion of chemotherapy. Both trials were prospective and randomised studies without a placebo control arm (control group participants were untreated). Both studies included participants being treated with cisplatin for various solid tumours (lung, ovarian, rhinopharynx, gastric, testicular, oesophagus, ethmoid, and tongue cancer). The 2010 update identified a third RCT involving the use of vitamin E (Pace 2010), but the substantial participant dropout rate prevented intention‐to‐treat analyses. The 2013 update identified one additional study, a randomised, double‐blind, placebo‐controlled trial investigating use of vitamin E among 189 evaluable participants with breast, lung, and other cancers being treated with cisplatin (8), carboplatin (2), oxaliplatin (50), taxanes (109), or combinations of each (20), with the inclusion of a large number of participants receiving taxanes confounding (Kottschade 2011). In this study, participants were randomly assigned to vitamin E (dl‐alpha‐tocopherol) 300 mg by mouth twice daily or matching placebo twice daily, initiated within four days of the first chemotherapy treatment and continued throughout and for one month beyond the last chemotherapy treatment (Kottschade 2011). The three trials included 130 vitamin E‐treated participants and 134 control participants. These trials included participants being treated with:
-
cisplatin for various solid tumours (lung, ovarian, rhinopharyngeal, gastric, testicular, oesophagus, ethmoid, and tongue) (13 vitamin E‐treated participants and 14 control participants) (Pace 2003);
-
cisplatin chemotherapy for various solid tumours (lung, ovarian, rhinopharyngeal, gastric, testicular, oesophagus, ethmoid, and tongue) (14 vitamin E‐treated participants and 16 control participants) (Argyriou 2006); and
-
cisplatin (8), carboplatin (2), oxaliplatin (50), taxanes (109), or combinations of each (20) for the treatment of breast, lung, and other cancers (96 vitamin E‐treated participants and 93 control participants) (Kottschade 2011).
Risk of bias in included studies
The 'Risk of bias' information for each study is summarised in Figure 1.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Red (‐) = high risk of bias; yellow (?) = unclear risk of bias; green (+) = low risk of bias.
Acetylcysteine (N‐acetylcysteine, NAC)
The single trial of NAC had adequate baseline and follow‐up NCI neurotoxicity assessment; although electrophysiological assessment, the nerve conduction study data, was only described qualitatively (Lin 2006), indicating no electrophysiological change in the NAC group without available data for the control group. The method of randomisation and of allocation concealment was unclear. It also was unclear whether the study was blinded to participants, investigators, or outcome assessors.
Amifostine
All seven trials had adequate baseline and follow‐up assessments that occurred after completing chemotherapy within the intended zero to six‐month window. All trials also included assessment following the last treatment cycle, with adequate follow‐up to that point. Two trials included additional assessments up to three months after the last treatment cycle (Kemp 1996; Planting 1999). The method of randomisation was deemed secure and allocation concealment at low risk of bias for two trials (Lorusso 2003; Lu 2008), but unclear for four trials. DeVos 2005 performed a phase II evaluation, with a control group that did not receive placebo, and participants received both carboplatin and taxanes. One additional trial was an open pilot study in which the participants were randomised and blinded but the control group did not receive placebo an the investigators were not blinded (Gallardo 1999). None of the trials clearly masked the participant or the observer, and only one study stated that the outcome assessor was blinded to the treatment group or to adverse events reported by the participant (Kemp 1996). The adequacy of analyses was deemed at low risk of bias for four studies but unclear for one (Planting 1999).
Calcium and magnesium (Ca/Mg)
The three trials had adequate baseline and follow‐up assessments that occurred after completing chemotherapy within the intended zero to six month window, with assessment occurring immediately after the sixth treatment cycle (10 weeks). The method of randomisation was deemed secure, and allocation of concealment and subject, observer, and outcome assessor blinding appeared at low risk of bias. An unavoidable limitation of the Ishibashi 2010 study was that enrolment was discontinued after 33 participants were entered into the study because the interim analyses showed poorer results in the Ca/Mg group (early termination of enrolment). Similarly, the Chay 2010 study was terminated prematurely because of reports of negative results from the CONcePT trial, and the Grothey 2011 study was terminated prematurely because of reports from another trial suggesting that Ca/Mg decreased treatment efficacy, data subsequently found to be incorrect.
Diethyldithiocarabamate (DDTC)
In the single DDTC study (Gandara 1995), the methods of randomisation and allocation concealment were unclear, although we judged the participant blinding to be at low risk of bias. Observer and outcome assessor blinding were considered at low risk of bias. A large number of withdrawals for progressive disease or toxicity (40 in the intervention arm and 71 in the control arm) were noted.
Glutathione (GSH)
All seven trials had adequate baseline and follow‐up assessments that occurred after completing chemotherapy within the intended zero to six‐ month window. The method of randomisation was deemed secure, allocation concealment was also at low risk of bias, and blinding was done for participants and observers in three trials (Cascinu 1995; Cascinu 2002; Smyth 1997), but not for the remaining three trials (Bogliun 1996; Colombo 1995; Milla 2009; Schmidinger 2000). Outcome blinding was at low risk of bias for four trials (Bogliun 1996; Cascinu 1995; Cascinu 2002; Smyth 1997), and unclear in the other three (Colombo 1995; Milla 2009; Schmidinger 2000). The adequacy of analyses was deemed adequate for five studies (Bogliun 1996; Cascinu 1995; Cascinu 2002; Milla 2009; Smyth 1997), and inadequate for two (Colombo 1995; Schmidinger 2000).
Org 2766
All four trials had adequate baseline and follow‐up assessments that occurred after completing chemotherapy within the intended zero to six‐month window. One trial (Hovestadt 1992) included additional assessments up to 24 months after the last treatment cycle. The method of randomisation was deemed secure and allocation concealment at low risk of bias for one trial (van der Hoop 1990), but unclear for the remaining three trials. Participant and outcome assessor blinding was at low risk of bias in three trials (Hovestadt 1992; Roberts 1997; van der Hoop 1990), and unclear in one (van Gerven 1994), while observer blinding was at low risk of bias in two trials (Roberts 1997; van der Hoop 1990) and unclear in two (Hovestadt 1992; van Gerven 1994). The adequacy of analyses was deemed adequate for one study (Roberts 1997), but unclear for two trials (Hovestadt 1992; van Gerven 1994), and inadequate for one (van der Hoop 1990).
Oxcarbazepine (OXC)
Out of 40 participants enrolled in the single OXC study (Argyriou 2006a), eight dropped out (four in each arm), mainly for disease progression plus two for adverse symptoms related to OXC treatment. The specific method of randomisation was unclear, but the details were known only to the randomisation coordinator and concealed from the data analysts. Participants were not blinded, as this was an open label study. The dropout rate was not considered excessive, and the analyses were performed both on an intention‐to‐treat and completion‐of‐trial basis.
Retinoic acid, all‐trans (ATRA)
There were no withdrawals and no participants were lost to follow‐up. We deemed the risk for bias for all measures to be low.
Vitamin E
Out of 47 participants enrolled in the initial vitamin E study (Pace 2003), 20 participants dropped out, mainly for disease progression. Twenty‐seven participants, all of whom completed six cycles of cisplatin chemotherapy and received a cumulative dose of cisplatin greater than 300 mg/m², were available for analysis. Individual doses and combinations of chemotherapy varied. Methods of randomisation, allocation concealment, and blinding were unclear. Insufficient details were provided on withdrawals, dropouts, and losses to follow‐up and we thought the analysis at unclear risk of bias in this respect. A modified total neuropathy scale was used, but this modified version is not a validated measure. In the second study (Argyriou 2006). Five of the 37 enrolled participants dropped out because of death or disease progression. Thirty participants completed the chemotherapy protocol; the mean total cisplatin dose was comparable in the vitamin E and control groups (120.6 ± 5.2 mg and 121.9 ± 3.4 mg, respectively). The precise randomisation method was not described but the allocation concealment appeared to be at low risk of bias, with details of the randomisations known only to the randomisation co‐ordinator. In this open label study, the subjects were not blinded but those performing the evaluations were blinded, as were the outcome assessors. Out of 207 initial enrollees, there were 18 cancellations (7 ineligible and 11 withdrawals, mainly due to refusal of further treatment) (Kottschade 2011). One hundred and twenty‐seven participants completed the study treatment protocol but a total of 189 participants completed a sufficient portion of the protocol to be available for the end point analyses. The study was a randomised, double‐blind, placebo‐controlled trial. Allocation concealment was not specifically described but there was no indication that it was inadequate. The inclusion of a large number of participants receiving taxanes was confounding.
Effects of interventions
The results of the review for each potential chemoprotective agent follow.
Acetylcysteine (N‐acetylcysteine, NAC)
Primary outcome measure
No primary outcome measure was available in the single NAC study (Lin 2006).
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
Sensory (sural SNAP) amplitude, distal latency, and conduction velocity, and motor (median compound muscle action potential (CMAP)) amplitude, distal latency, conduction velocity, and F wave latency were measured at baseline and after 4, 8, and 12 cycles of chemotherapy. However, the data were only reported for the NAC group, indicating no significant deterioration over the full trial (something the authors attributed to NAC), without reporting similar data for the control group.
(2) Clinical impairment on neurological examination using a validated scale
A neurologist performed a complete neurological examination but the report does not provide the data, aside from noting (in the Discussion) that oxaliplatin‐induced neurotoxicity is characterised by a rapid‐onset acute sensory neuropathy.
(3) Functional activities of daily living
Not an outcome
(4) Information from toxicity rating scales
Toxicity due to possible sensory neuropathy was assessed every two weeks using the National Cancer Institute‐Common Toxicity Criteria (NCI‐CTC). After 12 cycles of treatment, the incidence of ≥ Grade 1, 2, and 3 neurotoxicity was 80%, 20%, and 0% among the five participants in the NAC group, respectively, and 100%, 89%, and 33% in the nine participants in the control group (P = 0.01).
Adverse effects attributed to the study intervention
No adverse neurotoxicity was attributed to the NAC intervention.
Details of other outcomes not specified in the protocol
None described
Amifostine
Primary outcome measure
Quantitative sensory testing (QST) was used as an outcome measure in a single amifostine trial (Planting 1999), although only mean baseline and post‐treatment values and non‐parametric analyses were reported. This trial involved 74 participants with advanced neck cancer who were treated with six cycles of weekly cisplatin (70 mg/m²). Cisplatin neurotoxicity was measured by VPT recordings made from the second metatarsal bone of each hand (test of large sensory axons and receptors). An increased threshold indicates worsening sensory performance. Three measurements of the VPT (in μm of skin displacement) were recorded for each hand and the mean taken as the VPT for each side. Participants receiving pretreatment with amifostine (740 mg/m²) prior to cisplatin administration (37 participants) showed significantly less subclinical neurotoxicity as measured by the VPT relative to control participants receiving cisplatin only (37 participants). Although the groups were similar in terms of most measures, a slightly smaller number of participants (borderline significant) in the amifostine group relative to the control group completed six cycles of cisplatin (20 versus 28; P = 0.07). The VPT analyses, which were based on participants in whom a three‐month post‐treatment VPT value was available, included 14 participants in the amifostine group and 20 participants in the control group. The mean VPT increased at three months compared to baseline values for both groups, but the increase was smaller for the amifostine group compared to the control group (left hand, 0.15 μm versus 0.48 μm, mean difference (MD) 0.33 μm (95% CI ‐0.01 μm to 0.67 μm); right hand 0.18 μm versus 0.40 μm, MD 0.12 μm (95% CI ‐0.03 μm to 0.27 μm). The authors explained that the borderline significant group difference for the right hand VPT reflected an imbalance at baseline for the groups and limited data.
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
This secondary outcome measure was used in a single trial evaluating the protective effect of amifostine against neurotoxicity of carboplatin (area under the curve according to the Calvert formula) and paclitaxel (175 mg/m²) among participants with non‐small cell lung cancer (Kanat 2003). Chemotherapy‐induced neurotoxicity was assessed by serial SNAPs recorded from the right sural, left ulnar, and right median sensory nerves. Evaluations were performed at baseline and immediately after the sixth cycle of treatment. Participants who received pretreatment with amifostine (910 mg/m²) prior to chemotherapy (19 participants) were compared to control participants who received chemotherapy only (19 participants). The mean SNAP amplitudes for both groups were comparable at baseline and also after six cycles of chemotherapy. There was no significant decline in the mean amplitudes for either group after treatment.
(2) Clinical impairment on neurological examination using a validated scale
There was no use of a uniform or standardised neurological examination scale among the studies, but several clinical scales were used that were based on descriptions of conventional neurological symptoms or signs. For example, in Planting 1999, clinical neurotoxicity "grade 1" developed in 4 of 37 participants in the cisplatin plus amifostine group and 5 of 37 participants in the cisplatin only group (risk ratio (RR) 0.80, 95% CI 0.23 to 2.75) (see Analysis 1.1). Conversely, paraesthesias "grade 2" (reflecting an adverse sensory symptom outcome) developed in 8 of 19 participants in the carboplatin and paclitaxel plus amifostine group compared to 18 of 19 in the carboplatin and paclitaxel only group (RR 0.59, 95% CI 0.36 to 0.98) in Kanat 2003. Neurotoxicity as evaluated by a clinical examination and scored as "grade 3‐4" (not further described) was reported among 93 participants receiving carboplatin and paclitaxel plus amifostine at any of the scheduled evaluations in 19 of 508 evaluations (3.7%) versus in 37 of 514 evaluations (7.2%) among participants receiving carboplatin and paclitaxel only (Lorusso 2003). Our review protocol was interested in results after treatment was completed. Data recorded after the last treatment cycle were not available in the Lorusso et al. article (Lorusso 2003), but, despite there being a similar proportion of subjects in each group who continued to show an adverse effect, the neurotoxicity comparisons for the last three cycles of chemotherapy no longer showed a substantial group difference (10 of 244 evaluations (4%) in the amifostine group and 23 of 245 evaluations (9%) in the control group). Furthermore, the paper reports the data as adverse neurotoxicity occurring during any of the treatment cycles and therefore includes multiple observations from individual participants, not independent observations, precluding further evaluation.
(3) Functional activities of daily living
Based on a sensorimotor neurotoxicity score, Kanat 2003 reported a minor ("grade 2") decrease in the activities of daily living (ADL) score in 2 of 19 participants in the carboplatin and paclitaxel plus amifostine group compared to 9 of 19 participants in the carboplatin and paclitaxel only group (RR of 0.22, 95% CI 0.06 to 0.90) (see Analysis 2.1).
(4) Information from toxicity rating scales
Kemp 1996 reported that the incidence of neurotoxicity, defined as symptoms of peripheral neuropathy or as a decrease in the neurological function daily activity score, was related to the cumulative dose of cisplatin among participants with ovarian cancer who were treated with cisplatin and cyclophosphamide. By treatment cycle five, there was a significant difference between the groups, favouring the amifostine group (P = 0.015). Following the last treatment cycle, the NCI‐CTC neurologic toxicity rating (grades 0, 1, 2, or 3) was significantly reduced (P = 0.029) by pretreatment with amifostine (122 participants pre‐treated with amifostine, 120 control participants). The risk of developing any level of neurotoxicity (grade 1, 2, or 3) was almost significantly reduced by pretreatment with amifostine relative to no treatment (RR of 0.81, 95% CI 0.66 to 1.00) (see data for Kemp 1996, Analysis 3.1). None of the participants in the amifostine group but two of the participants in the control group discontinued cisplatin because of neurotoxicity. Lu 2008 used the NCI‐CTC (2.0) scale to evaluate neurotoxicity. The occurrence rate of an acute neurotoxicity during treatment was significantly lower in the amifostine group relative to the control (glutamine) group: 6.5% versus 95.7% (P < 0.001). The occurrence rates of grade I–II and grade III–IV peripheral neurotoxicity after chemotherapy were significantly lower in amifostine group than in control group (10.9% versus 73.9%, P < 0.001; 2.2% versus 19.6%, P = 0.007). The results at trial end were combined by grades 0, 1‐2, and 3‐4 and could not be evaluated using the more frequently used definition of neuropathy based on a ≥ grade 2 rating. Nevertheless, after treatment cycle 6, the occurrence rates of all grades of neuropathy were significantly lower in the amifostine group relative to the control (glutamine) group: grade 0 (no neuropathy), 38 of 44 versus 6 of 43; grades 1‐2, 6 of 44 versus 30 of 43; grade 3‐4, 0 of 44 versus 7 of 43 (P < 0.001) (See Lu 2008 data in Analysis 3.1). The incidence of acute cold paraesthesias was also lower in the intervention group than the control group. The frequency of regimen change because of chemotherapy‐related neurotoxicity was significantly lower in the amifostine group than in the control group (4.3% versus 23.9%, P = 0.007). The overall response rates of evaluable participants were 44.4% in the amifostine group and 38.5% in the control group (P = 0.66). DeVos 2005 performed evaluations at the end of six cycles of chemotherapy using a subjective neuropathy‐related symptom score (for which none of the items reached statistical significance) and the NCI‐CTC neurotoxicity rating. Participants who experienced an NCI‐CTC rating of ≥ grade 2 at the end of six cycles (plus those terminated early because of neurotoxicity) included 7 of 45 participants in the amifostine group and 17 of 45 in the control group. Those who had a NCI‐CTC grade 3 included 4 of 45 participants in the amifostine group and 5 of 45 control participants. Gallardo 1999 used a neurotoxicity scale, presumably the NCI‐CTC grading scale for neuropathy. That study included 20 participants with locally advanced, histologically diagnosed, cervical cancer to investigate the feasibility of five‐day scheduling of amifostine with radiotherapy and cisplatin, randomised to receive amifostine (10 participants) or no amifostine (10 participants) before infusion of cisplatin.
A pooled analysis of the available data from four studies (DeVos 2005; Kanat 2003; Kemp 1996; Lu 2008) was possible for the equivalent of NCI‐CTC grades ≥1 (all grades of neuropathy). The pooled analysis showed a significantly reduced risk of developing any level of neurotoxicity favouring the amifostine group relative to the placebo group, with an RR of 0.66 (95% CI 0.57 to 0.76) (random‐effects model because of substantial heterogeneity) (Analysis 3.1). A pooled analysis of the available data from three studies (DeVos 2005; Gallardo 1999; Kanat 2003) was possible for NCI‐CTC grades ≥2. Despite use of slightly different versions of the NCI‐CTC, inclusion of taxanes by DeVos 2005, and slightly different evaluation times (e.g., end of the treatment cycle versus three months later), there was a significantly reduced risk favouring the amifostine group relative to the placebo group, showing an RR of 0.26 (95% CI 0.11 to 0.61) (random‐effects model because of substantial heterogeneity) (Analysis 3.2, Figure 2). This analysis included a small number of participants (74 received amifostine) and included data from an open pilot study (Gallardo 1999), although similar results were found after exclusion of the Gallardo et al. study results. An additional pooled analysis of the available data from three studies was possible for the equivalent of NCI‐CTC grades ≥3 (Kemp 1996; Lu 2008; DeVos 2005). Despite use of slightly different versions of the NCI‐CTC and slightly different evaluation times (e.g., end of treatment cycle versus three months later), there was a non‐significant reduced risk favouring the amifostine group relative to the placebo group, showing an RR of 0.54 (95% CI 0.2 to 1.29) (random‐effects model because of substantial heterogeneity) (Analysis 3.3, Figure 3).
Forest plot of comparison: 2 Amifostine ‐ neurotoxicity rating, outcome: 2.2 Neurotoxicity rating ≥ 2.
Forest plot of comparison: 2 Amifostine ‐ neurotoxicity rating, outcome: 2.3 Neurotoxicity rating ≥ 3.
Adverse effects attributed to the study intervention
Hypotension occurred during 133 of 508 amifostine infusions (26%, 95% CI 23.5 to 23.8%) in Lorusso 2003. However, the degree of hypotension was generally mild and well tolerated and led to a reduction of the amifostine dose from 910 mg/m² to 740 mg/m² in subsequent cycles after only 18 of the 133 occurrences (14%). Similarly, Kanat 2003 reported that hypotension occurred in 5 of 19 participants (26%) during the first or second administration of amifostine; lowering the subsequent dose from 910 mg/m² to 740 mg/m² for these participants again eliminated the problem. In Planting 1999, hypotension was observed during amifostine infusion in 17 of 36 participants (47%) and in 45 out of 184 cycles (24%), despite use of a lower dose of amifostine (740 mg/m²) compared to the other three trials. The hypotension was described as grade 3 in two participants (three cycles) and grade 4 in three participants (four cycles), but considered to be of clinical relevance in only one participant. In Kemp 1996, amifostine was well tolerated; the principal adverse side effect was a transient decrease in blood pressure observed in 75 of 122 (62%) of participants during amifostine administration. In addition, emesis occurred in 96% of participants in the amifostine group relative to 88% of participants in the control group. Lu 2008 reported that the incidence of leucopenia, thrombocytopenia, and anaemia were all less in the amifostine group relative to the placebo group, but not significantly so. For grade 3‐4 haemotoxicity, the occurrence rates of hypoleukaemia were 8.7% versus 17.4% in the amifostine versus placebo groups, respectively (P = 0.216). The incidence of transient hypotension was higher in the amifostine group (occurring in 4 of 46 participants during infusion). DeVos 2005 reported that nausea and vomiting occurred more frequently in participants receiving amifostine than in those not receiving amifostine (moderate nausea 21% versus 29%, and severe nausea 2% versus 6%, P = 0.007). Amifostine infusion was temporarily halted in five participants for 10 cycles due to hypotension (DeVos 2005). Gallardo et al. reported that 1 of 10 amifostine‐treated participants needed temporary interruption of amifostine due to hypotension, and 8 of 10 participants receiving amifostine developed hypocalcaemia during treatment (Gallardo 1999).
Details of other outcomes not specified in the protocol
Progression of disease (ovarian cancer) was similar in the amifostine and control groups after an average of 24 months of follow‐up (Lorusso 2003). Amifostine did not compromise the antitumour effect of cisplatin in the treatment of ovarian cancer (Kemp 1996). However, amifostine also did not result in a higher dose intensity of cisplatin (Kemp 1996; Planting 1999). Similarly, amifostine did not affect the response to oxaliplatin chemotherapy for digestive tract tumours (Lu 2008).
Calcium and magnesium (Ca/Mg)
Primary outcome measure
No primary outcome measure was available in the any of the Ca/Mg studies (Chay 2010; Grothey 2011; Ishibashi 2010; Knijn 2011).
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
SNAP amplitudes were measured in the Chay 2010 study, but the results were reported as "abnormal" if each of the nerve conduction study measures was outside the normal values for their laboratory controls. At the end of treatment, six of seven participants receiving Ca/Mg compared to zero of nine participants receiving placebo experienced neuropathy, indicating that the treatment group had worse objective scores relative to the placebo group (RR 16.25, 95% CI 1.07 to 247.19) (Analysis 4.1).
(2) Clinical impairment on neurological examination using a validated scale
Not reported, aside from mention by Grothey 2011 that acute muscle spasms associated with oxaliplatin were significantly reduced (P < 0.01)
(3) Functional activities of daily living
Not reported
(4) Information from toxicity rating scales
The NCI‐CTC and Debiopharm Neurotoxicity Scale (DEB‐NTS) were used to asses the development and severity of neurotoxicity in the Ishibashi 2010 study. According to the NCI‐CTC criteria after six cycles of treatment, the incidence of ≥ Grade 1, 2, and 3 neurotoxicity were 100%, 6%, and 6% in the Ca/Mg group, respectively, and 94%, 6%, and 0% in the control group, there being no significant difference between groups. Similarly, according to the DEB‐NTS criteria after six cycles of treatment, the incidence of ≥ Grade 1, 2, and 3 neurotoxicity was 100%, 71%, and 6% in the Ca/Mg group, respectively, and 94%, 56%, and 0% in the control group, there being no significant difference between groups. The authors did not comment on the acute neuropathy symptoms. Chay 2010 included an acute subjective sensory neuropathy rating using the NCI‐CTC (grades 0, 1, 2, or 3) during and at the end of treatment. Overall, the subjective neuropathy rate was 77% in the Ca/Mg group and 86% in the placebo group (P = 0.6). At the end of treatment, three participants receiving Ca/Mg and none receiving placebo reported grade 3 numbness (P = 0.02). Assessment of grade 3 toxicity favoured the placebo group but not significantly (P = 0.09), with no significant treatment group difference for dysaesthesias or time to onset of neuropathy. Grothey 2011 included a subjective cumulative sensory neurotoxicity (sNT) rating (grade ≥2 endpoint), the oxaliplatin‐specific sNT. In the 102 participants available for analysis, Ca/Mg decreased the incidence of sNT ≥ grade 2 on the oxaliplatin‐specific sNT scale (RR 0.56, 95% CI 0.33 to 0.94) (Analysis 5.1), representing a significant minor effect. No effect on acute, cold‐induced sNT was found.
The pooled results from the three available studies reporting NCI‐CTC neuropathy ≥ grade 2 neurotoxicity at the end of chemotherapy showed no significant risk reduction associated with the use of Ca/Mg (Chay 2010; Grothey 2011; Ishibashi 2010), RR 0.84 (95% CI 0.44 to 1.60, random‐effects model) (Analysis 5.2, Figure 4).
Forest plot of comparison: 5 Ca/Mg ‐ sensory neuropathy (SNP), outcome: 5.1 Chronic NCI‐CTC ≥ grade 2.
Adverse effects attributed to the study intervention
No adverse neurotoxicity was attributed to the Ca/Mg intervention (Chay 2010; Grothey 2011; Ishibashi 2010).
Details of other outcomes not specified in the protocol
The authors in Ishibashi 2010 reported in their final analyses that Ca/Mg infusions did not influence antitumour activity among participants. However, their interim analyses had indicated that "the results of treatment were poor" in the Ca/Mg group relative to the control group, resulting in discontinuation of subsequent enrolment. The authors of Chay 2010 concluded that Ca/Mg infusions failed to reduce the rate of oxaliplatin‐induced acute and cumulative sensory neuropathy, although there appeared to be a trend toward a benefit in the patient’s perception of numbness during chemotherapy, a perception that was not supported by the authors' nerve conduction findings.
Diethyldithiocarabamate (DDTC)
Primary outcome measure
No primary outcome measure was available in the single DDTC study (Gandara 1995).
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
SNAP amplitudes were not measured.
(2) Clinical impairment on neurological examination using a validated scale
Not reported
(3) Functional activities of daily living
Not reported
(4) Information from toxicity rating scales
NCI toxicity rating scale showed 13 of 96 participants (13%) in the DDTC arm and 12 of 99 participants (12%) in the control group developed neuropathy (RR 1.12, 95% CI 0.54 to 2.32) (see Analysis 6.1).
Adverse effects attributed to the study intervention
Adverse effects were reported in all study participants, with severe adverse events in 27 control group participants and 30 participants in the DDTC group. Twenty‐two participants in the DDTC group and nine in control group withdrew because of toxicity.
Details of other outcomes not specified in the protocol
None described
Glutathione (GSH)
Primary outcome measure
QST was used as an outcome measure in only one study (Bogliun 1996). Although the manuscript does not state the site of recording the VPT, a two‐ to three‐fold increase in vibratory threshold occurred in the GSH group versus a seven‐ to 10‐fold increase in the control group three months after chemotherapy. One of the Cochrane authors (GC) who is a co‐author on this paper was able to retrieve the original information as follows. QST, measured at cisplatin doses of 50 mg/m² and 75 mg/m², changed from baseline to end of induction to end of consolidation from approximately 1.5 to 1.6 and 4.1 respectively in the GSH group, and changed from approximately 1.6 to 8.3 and 9.4 respectively in the control group. No additional information was available.
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
Sural SNAP amplitude decreased by a greater amount in the control arm than in the GSH arm (58% to 68% controls versus 12% to 35% GSH) among participants receiving < 150 mg/m² or > 150 mg/m², respectively (Bogliun 1996). Similarly, Cascinu 1995 documented a significant reduction of sural, median, and ulnar SNAP amplitudes in the control group versus the GSH arm. At week 15, the sural SNAP decreased from 13.3 ± 4.1 µv to 8.0 ± 1.68 µv in the control arm (n = 18) and from 10.8 ± 5.84 µv to 9.0 ± 6.48 µv in the GSH arm (n = 24). Median and ulnar SNAP amplitudes also reduced by significant amounts only in the control groups. In the oxaliplatin trial, the sural amplitude decreased in the placebo group (n = 8) from 11.0 ± 6.92 µv to 7.20 ± 5.05 µv in the placebo arm (P = 0.05) but did not change in the GSH group (n = 10) (9.1 ± 6.34 µv to 8.7 ± 5.50 µv; P = non‐significant) after eight cycles of therapy (Cascinu 2002). Colombo 1995 recorded sensory responses at baseline and at examination end, after nine weeks of weekly chemotherapy. The sural SNAP amplitude changed from 13.2 ± 8.5 µv to 7.6 ± 3.3 µv in the placebo group (n = 16) and from 13.1 ± 8.2 µv to 9.7 ± 5.9 µv in the GSH group (n = 15). A combined group analysis of post‐treatment scores produced a MD in favour of GSH, with 95% CIs that allowed the possibility of no effect (Analysis 7.1).
No SNAP amplitudes were measures in the other studies (Bogliun 1996; Cascinu 2002; Milla 2009; Schmidinger 2000).
(2) Clinical impairment on neurological examination using a validated scale
At three months post chemotherapy, 5 of 19 participants in the GSH group and 8 of 16 in the control group changed in their neurological disability score (NDS) by more than 12 points (RR 0.53, 95% CI 0.21 to 1.29) (see Analysis 8.1) (Bogliun 1996). In the same study, the neuropathy symptom score (NSS) developed in 14 of 19 participants in the GSH arm and all 16 participants in the control group (RR 0.75, 95% CI 0.56 to 0.99) (see Analysis 9.1). After nine weeks of chemotherapy, at study end, Colombo 1995 reported that 2 of 16 GSH participants and 4 of 15 placebo participants became symptomatic for sensory neuropathy (Analysis 9.1). The pooled results from the Bogliun 1996 and Colombo 1995 data showed a borderline significant decreased risk of developing neuropathy based on the NSS favouring the GSH group relative to the control group (RR 0.69, 95% CI 0.50 to 0.97) (Analysis 9.1). Colombo 1995 performed clinical neurological examinations but did not use a validated scale; Schmidinger 2000 also performed clinical neurological examinations and, although the investigators did not use a validated scale, they found no cases of peripheral neuropathy in either the GSH or the control group.
(3) Functional activities of daily living
Participants who received GSH reported an improvement in their quality of life. They also reported significantly less tingling in hands and feet and GSH allowed more cycles of cisplatin to be administered because of less toxicity (Smyth 1997).
(4) Information from toxicity rating scales
Bogliun 1996 assessed the NSS (symptoms) and NDS (signs) neuropathy scores, indicating a trend toward less severe neurotoxicity after cotreatment with GSH, although neuroprotection was not complete. Four of 24 participants in the GSH group developed neurotoxicity (3 grade I and 1 grade II by WHO criteria) and 16 of 18 in the control group (3 grade I; 10 grade II; 2 grade III and 1 grade IV) developed neuropathy by WHO neurotoxicity grade criteria (RR 0.19, 95% CI 0.08 to 0.47) (see Analysis 10.1) (Cascinu 1995). Grade 3 or 4 neurotoxicity by NCI‐CTC grading was seen in 0 of 21 participants at 8 weeks and 1 of 10 participants at 12 weeks in the GSH group and in 5 of 18 participants at 8 weeks and 6 of 8 participants at 12 weeks in the control group (RR 0.13, 95% CI 0.02 to 0.89) (see Analysis 11.1) (Cascinu 2002). In Schmidinger 2000, no change in WHO toxicity was noted in either the GSH or the control group, although the authors noted that no finding of peripheral neuropathy was observed in any of their participants. In Smyth 1997 after six cycles, neuropathy (NCI‐CTC) was seen in 39% of participants (24 grade 1, and 5 grade 2) in the GSH group and 49% of participants in the control group (32 grade 1, 4 grade 2, and 2 grade 3). Mean increase in Hospital Anxiety and Depression functional score was 0.8 in the GSH arm and 2.5 in the control arm. In addition, 45 of 47 had better Rotterdam scores in the GSH group. In Milla 2009, neurologic adverse effects were assessed using the neurosensory section of the NCI‐CTC, version 3. At the end of treatment, only moderate neurotoxicity was reported in the GSH arm (50% grade 1 and 50% grade 2), whereas in the placebo arm the neurotoxicity was more severe (69% grade 2 and 31% grade 3), a difference considered statistically significant (Mann‐Whitney test; P = 0.0037).
Three trials utilised WHO or NCI‐CTC grade ≥2 neurotoxicity at end of treatment (the most frequently used subjective measure of neuropathy in available studies) after 12 cycles of oxaliplatin (Cascinu 1995; Cascinu 2002; Milla 2009). The pooled data significantly favoured the GSH group relative to the placebo group and showing a protective RR of 0.29 (95% CI 0.10 to 0.85, random‐effects model) (see Analysis 12.1, Figure 5). Although all three studies showed a favourable effect using this subjective rating, the data showed substantial heterogeneity (I2 = 73%) and the studies that contributed data to the combined analysis involved a small number of participants (11 of 48 participants receiving GSH showing ≥2 neurotoxicity relative to 34 of 39 control participants showing a similar grade of neurotoxicity).
Forest plot of comparison: 11 GSH ‐ NCI NT rating 2‐4 at treatment end, outcome: 11.1 Chronic NCI ≥ grade 2.
Details of other outcomes not specified in the protocol
Oliguria occurred in 21 of 27 participants in cisplatin alone and 10 of 27 participants in the cisplatin with GSH group (RR 0.48, 95% CI 0.28 to 0.81) (see Analysis 13.1) (Bogliun 1996). Participants in the GSH group also required fewer haemotransfusions and showed fewer incidences of thrombocytopenia and anaemia than did participants in the control group (Cascinu 1995).
Adverse effects attributed to the study intervention
None described
Details of other outcomes not specified in the protocol
None described
Org 2766
Primary outcome measure
QST was used as an outcome measure in all four trials. Cisplatin neurotoxicity was measured by VPT recordings made from the second metacarpal bone of each hand (Hovestadt 1992; van der Hoop 1990; van Gerven 1994), or from the index finger and great toe (Roberts 1997). An increased threshold indicates worsening sensory performance. Three measurements of the VPT (in μm of skin displacement) were recorded for each hand and the mean taken as the VPT for each side in one study (van der Hoop 1990). Two trials (Hovestadt 1992; van Gerven 1994) referred to van der Hoop 1990 in describing their methodology, but that study did not provide a clear description of the method actually used. The final study did not report details of the QST methodology used (Roberts 1997).
Participants receiving treatment with Org 2766 1 mg/m² (n = 16) before and 24 hours after cisplatin administration showed significantly less increase in the VPT than did the 22 placebo‐treated participants (mean value after 4th cycle 0.50 μm versus 1.61 μm, P < 0.005; mean value after 6th cycle 0.88 μm versus 5.87 μm, P < 0.005) in van der Hoop 1990. In the same study, the administration of Org 2766 at the dose of 0.25 mg/kg had no effect. In Hovestadt 1992, only an exploratory, descriptive statistical analysis without formal tests for significance was performed, due to the low number of participants enrolled in the study (seven in the placebo group, five treated with Org 2766 at the dose of 0.25 mg/kg and six at the dose of 1 mg/kg). Mean values in placebo‐treated participants were higher than in the low‐ and high‐dose Org 2766‐treated participants one month after cisplatin treatment (mean 3.7 versus 2.9 versus 1.1 μm, respectively), intermediate one to four months after treatment (8.1 versus 14.6 versus 2.5 μm), and higher again after 4 to 12 and 12 to 24 months (4.8 versus 3.6 versus 2.0 μm, and 2.9 versus 0.6 versus 0.8 μm). However, the number of evaluable participants decreased markedly during the study and the final evaluation included only nine participants (for whom no indication about treatment is available). van Gerven 1994 reported VPT changes after the fourth cycle and three to five months later. Mean values obtained after the fourth cycle (0.95 μm in the placebo group versus 0.45 μm in the Org 2766‐treated group, number of participants nine versus six respectively) and three to five months later (4.03 versus 1.85 μm, number of participants 12 versus eight) were statistically compared with two different methods: the difference in slopes between Org 2766 and placebo‐treated groups was significant by ANOVA (P value = 0.04), but not by Wilcoxon's rank sum test (P = 0.06). In that study, VPTs became abnormal three to five months after treatment in nine out of 12 placebo‐treated participants and in four of the eight participants treated with Org 2766 (RR 0.67, 95% CI 0.31 to 1.43) (see Analysis 14.1). In Roberts 1997, VPT was assessed separately for the index finger and for the great toe. After the blind was broken, 174 participants (59 in the placebo group, 52 in the 2 mg/kg Org 2766 group, and 57 in the 4 mg/kg Org 2766 group) were evaluable. At each time point (i.e., at each cisplatin cycle up to six courses and monthly for three months after treatment withdrawal) no difference was observed between groups. The four Org 2799 studies all performed QST and evaluated our primary outcome measure, VPT. Three of the trials (Roberts 1997; van der Hoop 1990; van Gerven 1994), measured VPT at the index finger or hand, and evaluated comparable doses of Org 2766 (1 mg or 2 mg). Based on comparison of the Org 2766 treatment and placebo groups at three to five months, the combined data from the three trials showed no significant group difference at the follow‐up QST examination (MD ‐1.77 μm, 95% CI ‐4.78 to 1.23 μm, random‐effects model) (see Analysis 15.1).
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
This secondary outcome measure was not used in any of the selected trials.
(2) Clinical impairment on neurological examination using a validated scale
There was no use of a uniform or standardised neurological examination scale among the studies, but clinical scales were used that were based on descriptions of conventional neurological symptoms or signs. In one trial (van der Hoop 1990) neurological examination was based on a series of signs and symptoms evaluated after four and six chemotherapy courses, resulting in a "sum score." Using this score a difference was observed only for the high dose Org 2766 treatment at the evaluation performed after the 6th course versus the placebo‐treated participants (P = 0.03). Two other trials used non‐validated neurological scales and the authors performed no statistical comparison (Hovestadt 1992; van Gerven 1994). Roberts 1997 stated that neurological evaluation was used to assess inclusion/exclusion criteria, but provided no data about the severity of the neuropathy. In this trial, repeated neurological examinations failed to demonstrate any significant difference between Org 2766‐ and placebo‐treated participants up to three months after cisplatin treatment withdrawal.
(3) Functional activities of daily living
This secondary outcome measure was not used in any of the selected trials.
(4) Information from toxicity rating scales
In one study (van Gerven 1994), authors reported that two participants in the placebo group and one in the Org 2766 group discontinued cisplatin treatment because of peripheral neurotoxicity (P = not significant). The other three trials reported no details about neurotoxicity‐induced treatment withdrawal.
Adverse effects attributed to the study intervention
No adverse effects were attributed to Org 2766 in any trial.
Details of other outcomes not specified in the protocol
Progression of disease (ovarian cancer) was similar in the Org 2766 and control groups after an average of 24 months of follow‐up (Roberts 1997). Org 2766 did not compromise the antitumour effect of cisplatin in the treatment of ovarian cancer (van der Hoop 1990).
Oxcarbazepine (OXC)
Primary outcome measure
QST was not used as a primary outcome measure in the single OXC study available (Argyriou 2006a).
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
This secondary outcome measure was reported for sural, superficial peroneal, and ulnar SNAP amplitudes. A modest but significant decline in SNAP amplitude pre‐ to post‐ treatment was reported for the control group relative to the OXC group for the sural nerve (14.5 ± 6.1 to 8.3 ± 6.1 µv versus 13 ± 6.8 to 11.5 ± 7.1 µv) and superficial peroneal nerve (9.2 ± 3.2 to 6.8 ± 4.8 µv versus 9.3 ± 4.0 to 8.8 ± 4.4 µv) but not for the ulnar nerve (9.9 ± 3.4 to 7.0 ± 4.3 µv versus 9.5 ± 4.2 to 8.2 ± 4.3 µv ) (or for ulnar or peroneal motor amplitudes; motor nerves not thought to be affected by oxaliplatin) (Argyriou 2006a). However, group comparisons of SNAP amplitudes at post‐treatment (at six months, after 24 cycles) showed no significant differences (Analysis 16.1, Analysis 16.2, Analysis 16.3).
(2) Clinical impairment on neurological examination using a validated scale
Based on a NSS and the NDS, the incidence of oxaliplatin‐induced neuropathy was decreased among those who completed the trial in the OXC group relative to the control group (five of 16 versus 12 of 16) (RR 0.42, 95% CI 0.19 to 0.91) (see Analysis 17.1), and the between‐group comparisons of the NSS (0.6 ± 0.9 versus 1.5 ± 1.3) and NDS (5 ± 8.2 versus 20.0 ± 23.1) differed significantly, both favouring the OXC group. Also, the severity of oxaliplatin‐induced neuropathy was significantly lower in the OXC group versus controls, based on the total neuropathy scores (MD ‐7.10, 95% CI ‐11.98 to ‐2.22) (see Analysis 17.2). None of the participants receiving OXC reported negative sensory symptoms versus three controls; five participants receiving OXC reported positive sensory symptoms versus 12 controls.
(3) Functional activities of daily living
This secondary outcome measure was not used in the selected trial.
(4) Information from toxicity rating scales
None described
Adverse effects attributed to the study intervention
Adverse chemotherapy effects were described but none was attributed to OXC (there was a similar frequency in both groups).
Details of other outcomes not specified in the protocol
None reported
Retinoic acid
Primary outcome measure
QST was not performed in Arrieta 2011.
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
Arrieta 2011 utilised nerve conduction study results summarised by converting the quantitative results to a non‐parametric zero to three ranking of "damage," summated for latency and amplitude measures. These electrophysiological results were not separated into their motor and sensory components, something physiologically nonsensical. Significant deterioration was reported for the summated sensory response amplitude only for the placebo group, but this group demonstrated worse function relative to the intervention group at baseline. Baseline differences between the intervention and control groups are troublesome and not explained. Also, the period of evaluation (i.e. six weeks to the end of the second course of chemotherapy) was likely to have been inadequate.
(2) Clinical impairment on neurological examination using a validated scale
No validated scale of clinical impairment on neurological examination was measured (Arrieta 2011).
(3) Functional activities of daily living
Arrieta 2011 did not measure any functional ADL.
(4) Information from toxicity rating scales
Arrieta 2011 utilised the NCI‐CTC rating scale for assessment of neuropathy. NCI‐CTC neuropathy grades ≥2 were present in 23 of 45 ATRA participants and in 37 of 47 placebo participants (RR 0.75, 95% CI 0.55 to 1.02) (Analysis 18.1). As noted previously, the period of evaluation (i.e., six weeks to the end of second course of chemotherapy) was likely to have been inadequate.
Adverse effects attributed to the study intervention
Five of 45 participants in the intervention group but no participants in the control group developed hypertriglyceridaemia grade 3 and 4 (Arrieta 2011).
Details of other outcomes not specified in the protocol
None described
Vitamin E
Primary outcome measure
QST was not performed in any of the vitamin E studies (Argyriou 2006; Kottschade 2011; Pace 2003).
Secondary outcome measures
(1) Nerve conduction measures of sensory response amplitudes
In Pace 2003, after six cycles of treatment, four of 13 participants in the vitamin E group had at least one abnormal finding among the median sensory or sural sensory amplitude, whereas 11 of 14 participants in the control group had at least one abnormal amplitude (RR 0.39, 95% CI 0.17 to 0.93) (see Analysis 19.1). Nerve conduction measures of sensory response amplitudes showed that the median SNAP amplitudes were reduced by a lesser degree in those taking vitamin E relative to control participants (vitamin E group: baseline 15.1 ± 9.2 µv, six months later 12.0 ± 0.6 [sic] µv; control group: baseline 15.0 ± 9.2 µv, six months later 8.7 ± 5.3 µv (P < 0.01)). By comparison, the sural amplitudes showed little effect of vitamin E (vitamin E group: baseline 15.5 ± 6.3 µv, six months later 13.7 ± 5.5 µv; control group: baseline 14.5 ± 8.5 µv, six months later 13.6 ± 9.2 µv, P = not significant). In the second vitamin E trial (Argyriou 2006), after six cycles of treatment and three months after treatment ended, all SNAP amplitudes (sural, superficial peroneal, and ulnar) had deteriorated, although the change from baseline was significantly less in the vitamin E group relative to controls. In contrast, the peroneal and ulnar motor amplitudes and conduction velocities showed no significant group differences in terms of change from baseline as reported in the trial report. However, group comparisons of the SNAP amplitudes at post‐treatment (at end of treatment and three months later) showed no significant differences, although the comparison of the superficial peroneal SNAP amplitude for the vitamin E and control groups showed slightly better performance for the vitamin E group at the end of treatment (8.5 ± 6.6 versus 4.9 ± 3.6 µv; P = 0.07; MD 3.60 µv, 95% CI ‐0.35 to 7.55 µv), although the CIs just allowed the possibility of no effect (Analysis 20.1). There was little difference between vitamin E and control groups in post‐treatment ulnar SNAP amplitude (Analysis 20.2). When the sural nerve results at end of treatment were combined for the two studies, no significant treatment effect existed (MD 2.01 µv, 95% CI ‐1.60 to 5.61 µv) (see Analysis 20.3, Figure 6). Kottschade 2011 did not perform nerve conduction studies.
Forest plot of comparison: 20 Vitamin E ‐ SNAP amplitude, outcome: 20.1 Sural amplitude uv after 6 cycles.
(2) Clinical impairment on neurological examination using a validated scale
Although both selected trials measured clinical impairment on neurological examination (Argyriou 2006; Pace 2003), neither trial used a validated scale. Argyriou 2006 reported that the incidence of clinical neuropathy was lower in the vitamin E group than in the control group among participants completing the study (3 of 14 versus 11 of 16, RR 0.31, 95% CI 0.11 to 0.90) (see Analysis 21.1). Similar results were reported for the intention‐to‐treat analysis (5 of 16 versus 13 of 19, RR 0.46, 95% CI 0.21 to 1.00) (see Analysis 21.2). The severity of clinical neuropathy, as judged by the mean score on a modified peripheral neuropathy scale at trial end, also showed a significant difference favouring the vitamin E group (4.99 ± 1.33 versus 10.47 ± 10.62, P = 0.04) (MD ‐5.48, 95% CI ‐10.73 to ‐0.23) (see Analysis 22.1).
(3) Functional activities of daily living
The vitamin E studies (Argyriou 2006; Kottschade 2011; Pace 2003), did not measure any functional ADL and did not provide any information on standardised toxicity rating scales.
(4) Information from toxicity rating scales
In Pace 2003, the incidence of neurotoxicity, measured by a modified version of the total neuropathy score, was significantly lower in the group with vitamin E supplementation (4 of 13 participants) compared to the group without vitamin E (12 of 14 participants; RR of developing symptoms or signs of neurotoxicity was 0.36, 95% CI 0.15 to 0.83) (see Analysis 23.1). In addition, the severity of neuropathy, measured with neurotoxicity scores, was higher (worse) in participants without vitamin E supplementation compared to those who received vitamin E (4.7 versus 2.0, P < 0.01).
Argyriou 2006, using a measure presumably derived from a modified Peripheral Neuropathy (PNP) score, reported results similar to those of Pace 2003, including mean PNP scores in the vitamin E group versus the control group of 4.99 ± 1.33 and 10.47 ± 10.62, respectively (P = 0.023). The pooled results from these two studies identified neurotoxicity after completion of chemotherapy in 9 of 29 participants receiving vitamin E versus 25 of 33 controls (P = 0.002); with an RR of 0.41 (95% CI 0.23 to 0.73, fixed‐effect model) (see Analysis 23.1, Figure 7).
Forest plot of comparison: 24 Vitamin E ‐ Clinical impairment, outcome: 24.1 Total neuropathy score after 6 cycles.
Kottschade 2011 reported the incidence of neuropathy using the NCI‐CTC adverse event (NCI‐CTCAE) score for neuropathy (a scale of one to four), as well as by a neuropathy‐specific questionnaire developed by the North Central Cancer Treatment Group (NCCTG). The primary endpoint was a ≥2 grade sensory neuropathy using the NCI‐CTC 3.0 criteria, with the percentage of grade 2+ sensory neuropathic toxicity. Although this endpoint produced a negative result, the inclusion of a large number of participants (109, 58% of the total) treated with taxanes versus those treated with cisplatin, oxaliplatin, or carboplatin made it impossible to combine the results of this study with other studies using similar measures because it was impossible to separate out those subjects who did not receive taxanes.
Adverse effects attributed to the study intervention
No information was provided pertaining to the adverse effects attributed to the study interventions in Pace 2003. Overall adverse effects experienced were judged unlikely to be due to vitamin E supplementation in Argyriou 2006. Similar results were report for the Kottschade 2011 study; namely, no adverse effects were attributed to vitamin E.
Details of other outcomes not specified in the protocol
None described.
Discussion
Summary of main results
Acetylcysteine (N‐acetylcysteine, NAC)
The single eligible study evaluating the use of N‐acetylcysteine (NAC) against the neurotoxicity of oxaliplatin was described as a pilot study and included only a small number of participants (five receiving NAC and nine controls) (Lin 2006). In addition, uncertainty about the randomisation methods and presence of blinding (participants or investigators), report only of subjective toxicity scales, and inclusion of nerve conduction results from the NAC group but not the control group are issues that limit interpretation of the preliminary study results.
Amifostine
The eligible studies evaluating the use of amifostine as a neuroprotective agent against the neurotoxicity of cisplatin and other chemotherapy agents are inconclusive in demonstrating efficacy, primarily because few studies utilised quantitative measures of neurotoxicity. Although the studies were generally well done, participant masking was unclear, perhaps because amifostine was given intravenously in conjunction with interval chemotherapy and felt to be of little interest to the recipient. The authors of one study acknowledged that the trial was not conceived as a double‐blind study, but that the physicians assessing non‐haematological toxicity usually were not the same physicians involved in administering treatment and they therefore could not be influenced by the evaluation of symptoms and signs associated with chemotherapy (Lorusso 2003). It is unknown whether this belief about the low likelihood of inadvertent influence on judgment is correct. Paclitaxel was used together with carboplatin in two of the amifostine trials that were included (Kanat 2003; Lorusso 2003).
Planting 1999 was the single study that included quantitative sensory testing (QST), our primary outcome measure, among their assessment instruments. That trial showed a favourable outcome in terms of amifostine neuroprotection, but the subclinical result was based on only 14 participants in the amifostine group and 20 participants in the control group and the results were not particularly robust, showing statistical but unclear clinical significance. Among the secondary outcome measures we selected, the only other quantitative measure used in any of the four studies was related to evaluation of peripheral nerve electrophysiology. A single study utilised electrophysiological measures, recording sensory nerve action potential (SNAP) amplitudes from two upper extremity sensory nerves and one lower extremity sensory nerve at baseline and after completing chemotherapy (Kanat 2003). This study, which also included a relatively small number of participants (19 participants in each study arm), found that the quantitative measures of large fibre sensory axons failed to identify evidence of significant amifostine neuroprotection. Unfortunately, the study also showed an unexpected and almost implausible low level of neurotoxicity among the control participants who received carboplatin and paclitaxel but not amifostine, rendering the amifostine efficacy result uninterpretable.
Despite the lack of quantitative measures showing amifostine efficacy, the authors of all seven trials concluded that pretreatment with amifostine reduced, prevented, or at least exerted some protection from the cumulative neurotoxicity associated with cisplatin or carboplatin and paclitaxel. This conclusion reflected the results obtained from the various neurological examination scales or neurotoxicity scales, most often the National Cancer Institute‐Common Toxicity Criteria (NCI‐CTC) neuropathy rating. The neurotoxicity rating scale was not specified by Gallardo 1999, but presumably represented the NCI‐CTC scale. For the most part, the other scales utilised were of unknown sensitivity or specificity, and the clinical relevance of the results uncertain. For example, it is unclear whether any of the sensory symptoms or signs resulted in substantial functional impairment or persisted. Some of the neurotoxicity results were reported in terms of each evaluation, rather than for each participant, suggesting fluctuation during the trial. Such fluctuation suggests inclusion of non‐specific symptoms, as opposed to the persistent distal predominant and symmetrical sensory symptoms and signs characteristic of most toxic neuropathies. One exception is the functional activities of daily living (ADL) scale reported by Kanat 2003. Although based on a small number of participants (19 participants in each study arm), the results of this sensorimotor neurotoxicity score suggested a small but statistically significant decrease in the ADL in 2 of 19 participants in the amifostine group compared to 9 of 19 participants in the control group (risk ratio (RR) 0.22, 95% confidence interval (CI) 0.6 to 0.90). Another exception reflected the use of the NCI neurological toxicity rating among 242 participants with ovarian cancer treated with cisplatin and cyclophosphamide (Kemp 1996). The significant decrease in the NCI rating among participants pre‐treated with amifostine compare to control participants was an impressive finding, despite the lack of QST or electrophysiological evaluations. The limited availability of clinical trials utilising conventional QST and nerve conduction study measures of peripheral nerve function is disappointing, at least in part because of the relative simplicity of the peripheral nervous system evaluation relative to other neurological functions, such as behavior. Similarly, De Vos 2005 concluded that amifostine showed minor but significant activity in diminishing neurotoxicity (but not in preventing paclitaxel plus carboplatin‐induced bone marrow toxicity) without interfering with chemotherapy efficacy.
The largely positive results based on subjective neuropathy ratings generally led to conclusions that amifostine could reduce the incidence and severity of peripheral neurotoxicity caused by cisplatin or oxaliplatin chemotherapy (e.g., Kemp 1996; and Lu 2008). These conclusions were supported by the pooled data involving secondary neurotoxicity rating scales, most often based on the NCI‐CTC neuropathy scale. Although the additional studies identified in the 2013 update permitted pooling of data, there still were only a small number of studies that could be included, and those studies involved a relatively small number of participants. For example, the pooled data from three studies involving the equivalent of NCI‐CTC grades ≥2 neurotoxicity included only 148 participants, of which 10 of 74 receiving amifostine showed ≥ grade 2 neurotoxicity relative to 39 of 74 control participants showed a similar grade of neurotoxicity (DeVos 2005; Gallardo 1999; Kanat 2003). In this analysis, the Kanat 2003 results represented outlying (positive) values relative to the other studies. That the pooled data showed substantial heterogeneity is perhaps not surprising considering the different cancers and primary treatment protocols described in these studies. Similar results were obtained from pooled analysis of the available data from three other studies using the equivalent of NCI‐CTC grades ≥3 although the results were no longer statistically significant (De Vos 2005; Kemp 1996; Lu 2008), demonstrating how sensitive the results are to inclusion of data from different trials. In this analysis, the Lu 2008 results represent very outlying (positive) values relative to the other studies.
In terms of direct relevance to chemotherapy, the expectation that reduced neurotoxicity would result in increased dosing was not clearly realised in any of the trials reviewed, although Lu 2008 reported that the proportion of chemotherapy schedule adjustment because of chemotherapy‐induced neurotoxicity was significantly lower in the amifostine group relative to the placebo group (4.3% versus 23.9%, P = 0.007).The expectation of the beneficial result of increased dosing is based on recognition that neurotoxicity is the primary dose‐limiting adverse effect attributed to cisplatin.
De Vos 2005 conducted a randomised phase II study of paclitaxel and carboplatin with and without amifostine and, as part of their discussion, pooled the results of three studies, all involving chemotherapy with paclitaxel and carboplatin (De Vos 2005; Kanat 2003; Leong 2003). We had excluded from our review the data from the Leong et al. study because participants received only two doses of carboplatin. We included the data from the De Vos et al. study, but noted that we were unable to separate the effects of carboplatin versus paclitaxel. The pooled results reported by De Vos et al. identified an odds ratio (OR) for developing ≥ grade 2 neurotoxicity of 0.30 (95% CI 0.15, 0.63; P < 0.05, random‐effects model. This compared to our pooled results (after change the RR to an OR), which showed an OR of 0.08 (95% CI 0.01to 0.79, random‐effects model, P = 0.03) for developing ≥ grade 2 neurotoxicity. Notwithstanding the different studies included, the two analyses based on pooled data (our analysis and the one reported by De Vos et al.) suggested potential amifostine protection against chemotherapy‐induced neurotoxicity.
In conclusion, the results of the available trials suggest the possibility of potential amifostine neuroprotection against cisplatin and other chemotherapy drugs. However, data involved a relatively small number of trials and participants, did not include improvement of primary measures or objective quantitative secondary measures, and the neuroprotection appears to be of small magnitude. Given this, in our opinion the overall efficacy results, while promising, remain inconclusive.
Calcium and magnesium (Ca/Mg)
The studies of Ca/Mg infusions as a chemoprotective agent against oxaliplatin neurotoxicity included subjective NCI‐CTC grading for neuropathy and Debiopharm Neurotoxicity Scale (DEB‐NTS) or similar neurotoxicity grading criteria (Chay 2010; Grothey 2011; Ishibashi 2010). Enrolment in all three studies was terminated prematurely, owing to reports that treatment results were poorer in the Ca/Mg group than in the control group. in This was according to the authors' interim analysis in the Ishibashi 2010 study, results not confirmed in the final analyses and, in the Chay 2010 study, according to reports from other studies indicating negative results. Grothey 2011 was terminated early because of reports of treatment interference by the study medication (Grothey 2011).The early discontinuations resulted in a small sample size and limited the data available to determine if Ca/Mg infusions had neuroprotective potential. The NCI‐CTC rating of Grothey 2011, the largest of the available studies, reported a borderline significant result favouring Ca/Mg infusions for preventing neuropathy, defined as an NCI‐CTC grading ≥2 (OR 0.42; 95% CI 0.17 to 0.99). However, the combined available data using the same grading scale did not support a significant beneficial effect. In a non‐randomised retrospective analyses, Knijn 2011 reported that early ≥ NCI‐CTC grade 2 neurotoxicity (occurring during six cycles of oxaliplatin chemotherapy) occurred in 218 of 551 (40%) Ca/Mg‐treated participants versus 81 of 181 (45%) non‐Ca/Mg‐treated participants. Similarly, late ≥ grade 2 neurotoxicity present at the last cycle before going off of study was present in 148 of 551 (27%) Ca/Mg‐treated participants versus 62 of 181 (34%) non‐Ca/Mg‐treated participants. The retrospective Knijn 2011 described a decrease in the frequency of late ≥ grade 2 neurotoxicity present at the last cycle before going off of study in Ca/Mg‐treated participants relative to those who did not receive Ca/Mg. When we pooled these positive results with the other studies reporting ≥ grade 2 neurotoxicity at the end of chemotherapy, we found a modest but statistically significant result favouring the use of Ca/Mg, OR 0.68 (95% CI 0.49 to 0.94) (Chay 2010; Grothey 2011; Ishibashi 2010; Knijn 2011).
Our limited pooled results showed a non‐significant reduced risk of developing a NCI‐CTC ≥ grade 2 favouring Ca/Mg (RR 0.84, 95% CI 0.62 to 1.05), results different from those two published meta‐analysis involving the efficacy of Ca/Mg infusions (Ao 2012; Wen 2013). Ao 2012 analyses included one study requiring translation that we are waiting to review (Dong 2010), a double‐blind study involving oxaliplatin‐induced neuropathy in which 4 of 20 Ca/Mg‐treated participants developed chronic neuropathy versus 11 of 27 control participants (a borderline significant difference). The pooled analyses performed in Ao 2012 reported an OR of 0.44 (95% CI 0.23 to 0.85, fixed‐effect model), indicating a significant result favouring treatment with Ca/Mg. After converting our analyses to an OR and using a fixed effects analyses, we showed a non‐significant OR of 0.58 (95% CI 0.27 to 1.21), results still different from those reported by Ao et al. Similarly, Wen 2013 reported the results of a second meta‐analysis involving oxaliplatin‐related neurotoxicity. Their analysis is potentially flawed because of the variability of the studies included, in which the study by Chay et al. was the only RCT included (it had a non‐significant outcome for total cumulative subjective sensory neurotoxicity). Their analyses reached significance only after including the non‐randomised trials of Knijn 2011 (discussed immediately above) and Gamelin 2004. We did not include data from either the Knijn et al. or the Gamelin et al. studies in our analyses because both represented retrospective non‐randomised studies (and because it is a Cochrane Neuromuscular Disease Group policy to not include data from non‐randomised studies in the results section). When we investigated the effect of pooling the retrospective data from Knijn et al. with the data we reported in the results section, we found a borderline significant RR of 0.80 (95% CI 0.62 to 1.05) favouring the use of Ca/Mg, a result qualitatively similar to the results of the two published meta‐analyses. Regardless, the conclusion of Wen 2013a that Ca/Mg infusions tend to decrease the incidence of oxaliplatin‐induced cumulative neurotoxicity and thus enhance patients' tolerance to oxaliplatin is based primarily on subjective data derived from retrospective studies, not RCTs.
In summary, the results of the best available trials selected for analyses are difficult to evaluate because of the small number of subjects, due primarily to early termination of several important studies (for reasons unrelated to the study intervention or treatment protocols). Based on the remaining data however, our results of the available RCT data do not show suggest statistically significant effect of Ca/Mg in preventing oxaliplatin‐induced neurotoxicity. Inclusion of results from several retrospective non‐randomised studies suggest more positive, borderline significant results favouring the use of Ca/Mg. At present, however, the limited data are not convincingly positive in favour of Ca/Mg neuroprotection and in our opinion, the overall efficacy results are promising but inconclusive.
Diethyldithiocarabamate (DDTC)
The single study of DDTC as a neuroprotective agent suffers from having no measures of neurotoxicity other than subjective reporting (NCI). To gain full appreciation of the magnitude of the difference between the two arms (DDTC and placebo), one should probably add those withdrawn for toxicity, to patient request, and adverse experience (Gandara 1995).
Glutathione (GSH)
Overall, six out of seven studies reported a significant protective effect of GSH. All measures of peripheral neuropathy favoured the GSH group, including the measure of VPT (one study), sural SNAP amplitudes (four studies) and improvement in functional measures or various neurotoxicity rating scales (six studies). Even though the overall effect of GSH appears to be beneficial and protective, the variable dosages used with different malignancies and different combinations of chemotherapy, high drop out rate, predominant reliance on subjective measures, limited statistical analyses, and lack of long‐term follow‐up, make the overall effect of GSH difficult to judge.
Org 2766
Overall, the few eligible studies evaluating the use of Org 2766 as a neuroprotective agent against the neurotoxicity of cisplatin are inconclusive in demonstrating efficacy. A major concern is that the total number of participants enrolled in the studies is rather low (188 Org 2766‐treated and 123 control participants) and, moreover, participants are not homogenously distributed among the four trials, since one of them (Roberts 1997) included 68% of the Org 2766‐treated and 54% of the control participants. All the trials included QST, our primary outcome measure, among their assessment instruments, while none of the studies reported an evaluation of peripheral nerve electrophysiology or effects on ADL. Neurological examination was based on non‐validated scales in all the four trials.
The first study suggesting a protective effect of Org 2766 (van der Hoop 1990) is based on an inadequate statistical analysis. In fact, analysis was performed after six cycles of cisplatin on only 28 participants out of the 55 admitted to the study, while the others were not eligible or had not yet received the planned chemotherapy cycles. Intermediate analysis (i.e. after four cycles of cisplatin) was performed on 39 participants. The authors of the second study (Hovestadt 1992) admitted that the number of participants was too low to allow a reliable formal statistical analysis. The third study was the only one performed mostly on males (22 men versus 1 woman). The authors used two different methods of statistical analysis and the results were conflicting (van Gerven 1994). The largest study had adequate subject, outcome assessor and observer blinding and also the statistical analysis was adequate (Roberts 1997). Instead of providing evidence of protection induced by Org 2766, the authors suggested that high doses of the compound might even increase the rate of change and degree of neuropathy induced by cisplatin (P value > 0.05).
In conclusion, although the results of the first trial (van der Hoop 1990) suggested the possibility of potential Org 2766 neuroprotection, effects appear to be of small magnitude or not convincingly positive in favour of Org 2766 neuroprotection, particularly in view of the results reported in the most recent trial (Roberts 1997). Furthermore, the combined data from the three trials using the same measure showed no significant group difference at the follow‐up QST examination (mean difference ‐1.77 95% CI ‐4.78 to 1.23). The overall efficacy results are negative.
Oxcarbazepine (OXC)
The only study evaluating the efficacy of OXY for prophylaxis against oxaliplatin‐induced neuropathy reported a favourable effect (Argyriou 2006a). The results were based on an open label evaluation (randomised but not placebo‐controlled), a small sample size, and without quantitative sensory testing as a primary outcome measure. However, validated clinical instruments (NSS and NDS) and appropriate neurophysiological measures were incorporated and showed several significant group differences, all favouring the OXC group. The significant nerve conduction results involved a change in the baseline to six month recordings for the lower extremity SNAP amplitude measures (sural and superficial peroneal), but not the ulnar sensory or peroneal motor measures, results consistent with those expected to represent the most sensitive indicators of an oxaliplatin‐induced neuropathy. Comparisons of the mean SNAP amplitudes for the treatment versus control groups post treatment (six‐month recordings after 24 cycles) showed no significant differences in any of the SNAP amplitudes, however. Although the significant neurophysiological group differences based on the change from baseline to six month records were modest and of uncertain clinical importance, the overall results support further investigation of OXC in a larger placebo‐controlled randomised trial.
Retinoic acid
The results from the single study involving retinoic acid as a neuroprotective agent are limited by methodological issues resulting in uninterpretable nerve conduction study results, unbalanced and unexplained baseline group differences, use of the NCI‐CTC grading as the only measure of neuropathy, and an inadequate follow‐up interval (Arrieta 2011). The NCI‐CTC assessment of neuropathy grades ≥2 were present in 23 of 45 ATRA participants and in 37 of 47 placebo participants, showing a borderline significant difference favouring the ATRA treatment group over the placebo group (RR 0.75 95% CI 0.55 to 1.02) (Analysis 18.1).
Vitamin E
Although the results involving vitamin E as a neuroprotective agent are encouraging, methodology issues, the small size of the study, the use of multiple chemotherapeutic regimens (including taxane in the largest available study), lack of blinding, and lack of primary outcome measures make the data less than convincing. This conclusion is despite the statistically significant results from the pooled data involving subjective measures of neuropathy (NCI‐CTC neurotoxicity rating). The changes noted in median SNAP but not the in sural SNAP amplitudes and the use of a non‐validated toxicity measure suggest that additional more definitive studies are needed. In our opinion the overall efficacy results, while promising, remain inconclusive.
Comment
The quality and characteristics of the trials reviewed were quite variable, and included different measures of neuropathy (qualitative and subjective), different durations of follow‐up, and different analyses. The duration of follow‐up must be sufficient to identity cisplatin‐induced sensory nerve deterioration, and therefore should extend beyond the last cisplatin treatment. How long after the last treatment is open to debate, as all toxic neuropathies demonstrate some progression after exposure ceases, and patients with cisplatin‐induced neuropathy can show improvement (depending on the initial severity) after cisplatin is discontinued. We included all evaluations performed zero to six months after the last treatment, selecting the evaluation closest to three months after treatment to the extent possible. We believe that a two to three month interval after treatment is completed is biologically reasonable. In all, 15 trials were included in our initial review, a further five trials in our 2010 review, and an additional 9 trials in our 2013 update. The combined trials involved nine separate, unrelated potential neuroprotective agents and included many disparate measures of neuropathy, resulting in sufficient data to combine the results for only a few measures, most of which were secondary measures such as the NCI‐CTC neurotoxicity rating. Based on our review, we feel that the evaluation of agents intended to prevent cisplatin‐induced sensory neuropathy should include validated measures, and not necessarily be limited to the primary measures we initially identified (tests of QST at the index finger and great toe and nerve conduction study evaluation of SNAP amplitudes in the median sensory and sural nerves). Before performing our review, considerable consideration was given to two measures (QST or nerve conduction studies) competing for the primary endpoint. We selected QST as the primary endpoint, in part, because it had been used in several prominent trials. While QST is an excellent quantitative measure of the endpoint of interest (sensation), the SNAP amplitude has the advantage of providing information about the actual cisplatin target, the peripheral sensory nerve, independent of patient co‐operation or motivation. Sufficient information exists about both measures to perform power calculations to determine the number of subjects required to detect a meaningful group difference. A difficulty we experienced in performing our review related to the limited data available, even when QST or SNAP amplitude recordings were performed. Most studies provided only descriptive statistics (e.g., mean, SD) reflecting the baseline examination and the follow‐up examination for treatment and control groups, without information about change (mean, SD) between baseline and follow‐up examinations. Inclusion of the latter facilitates comparisons with subsequent studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study. Red (‐) = high risk of bias; yellow (?) = unclear risk of bias; green (+) = low risk of bias.
Forest plot of comparison: 2 Amifostine ‐ neurotoxicity rating, outcome: 2.2 Neurotoxicity rating ≥ 2.
Forest plot of comparison: 2 Amifostine ‐ neurotoxicity rating, outcome: 2.3 Neurotoxicity rating ≥ 3.
Forest plot of comparison: 5 Ca/Mg ‐ sensory neuropathy (SNP), outcome: 5.1 Chronic NCI‐CTC ≥ grade 2.
Forest plot of comparison: 11 GSH ‐ NCI NT rating 2‐4 at treatment end, outcome: 11.1 Chronic NCI ≥ grade 2.
Forest plot of comparison: 20 Vitamin E ‐ SNAP amplitude, outcome: 20.1 Sural amplitude uv after 6 cycles.
Forest plot of comparison: 24 Vitamin E ‐ Clinical impairment, outcome: 24.1 Total neuropathy score after 6 cycles.
Comparison 1 Amifostine ‐ clinical impairment, Outcome 1 Clinical impairment at 3 months.
Comparison 2 Amifostine ‐ functional activities of daily living, Outcome 1 Functonal activities of daily living as measured at 3 months.
Comparison 3 Amifostine ‐ neurotoxicity rating, Outcome 1 Neurotoxicity rating ≥ 1.
Comparison 3 Amifostine ‐ neurotoxicity rating, Outcome 2 Neurotoxicity rating ≥ 2.
Comparison 3 Amifostine ‐ neurotoxicity rating, Outcome 3 Neurotoxicity rating ≥ 3.
Comparison 4 Calcium/magnesium ‐ nerve conduction study (NCS) results, Outcome 1 Abnormal sensory nerve action potential.
Comparison 5 Calcium/magnesium ‐ sensory neuropathy (SNP), Outcome 1 Oxaliplatin specific sensory neurotoxicity ≥ grade 2.
Comparison 5 Calcium/magnesium ‐ sensory neuropathy (SNP), Outcome 2 Chronic NCI‐CTC ≥ grade 2.
Comparison 5 Calcium/magnesium ‐ sensory neuropathy (SNP), Outcome 3 SNP grade 3.
Comparison 6 Diethyldithiocarbamate ‐ National Cancer Institute (NCI) neurotoxicity rating scale, Outcome 1 NCI toxicity for neuropathy.
Comparison 7 Glutathione ‐ nerve conduction study (NCS) results, Outcome 1 SSC amplitude.
Comparison 8 Glutathione ‐ clinical impairment, Outcome 1 Impairment at 3 months.
Comparison 9 Glutathione ‐ Neuropathy Symptom Score (NSS), Outcome 1 NSS symptoms of neuropathy at 3 months.
Comparison 10 Glutathione ‐ World Health Organization (WHO) evidence of neurotoxicity, Outcome 1 WHO neurotoxicity during trial.
Comparison 11 Glutathione ‐ chronic National Cancer Institute (NCI) toxicity rating, Outcome 1 NCI toxicity at 12 weeks.
Comparison 12 Glutathione ‐ National Cancer Institute (NCI) neurotoxicity rating 2‐4 at treatment end, Outcome 1 Chronic NCI ≥ grade 2.
Comparison 13 Glutathione ‐ other outcomes, Outcome 1 Oliguria.
Comparison 14 Org 2766 ‐ qualitative vibration position testing (VPT), Outcome 1 Abnormal VPT at 3 to 5 months.
Comparison 15 Org 2766 (1 or 2 mg)‐ vibration perception testing (VPT) finger or hand, Outcome 1 VPT hand 3 to 5 months post treatment.
Comparison 16 Oxcarbazepine ‐ sensory nerve action potential (SNAP) amplitude, Outcome 1 SNAP amplitude: sural.
Comparison 16 Oxcarbazepine ‐ sensory nerve action potential (SNAP) amplitude, Outcome 2 SNAP amplitude: superficial peroneal.
Comparison 16 Oxcarbazepine ‐ sensory nerve action potential (SNAP) amplitude, Outcome 3 SNAP amplitude: ulnar.
Comparison 17 Oxcarbazepine ‐ neurotoxicity rating, Outcome 1 Neuropathy after 12 cycles.
Comparison 17 Oxcarbazepine ‐ neurotoxicity rating, Outcome 2 Severity of neuropathy (TNS).
Comparison 18 Retinoic acid ‐ National Cancer Institute (NCI) neurotoxicity rating, Outcome 1 NCI‐CTC grade ≥ 2.
Comparison 19 Vitamin E ‐ qualitative sensory nerve conduction study (NCS) amplitudes, Outcome 1 Abnormal median or sural sensory amplitude.
Comparison 20 Vitamin E ‐ sensory nerve action potential amplitude, Outcome 1 Superficial peroneal amp 3 mos s/p treatment.
Comparison 20 Vitamin E ‐ sensory nerve action potential amplitude, Outcome 2 Ulnar SNAP amp 3 mos s/p treatment.
Comparison 20 Vitamin E ‐ sensory nerve action potential amplitude, Outcome 3 Sural amplitude uv after 6 cycles.
Comparison 21 Vitamin E ‐ incidence of neuropathy, Outcome 1 Incidence of peripheral neuropathy (completed trial).
Comparison 21 Vitamin E ‐ incidence of neuropathy, Outcome 2 Incidence of peripheral neuropathy (intention to treat).
Comparison 22 Vitamin E ‐ modified peripheral neuropathy (PNP) score, Outcome 1 Modified peripheral neuropathy (PNP) score.
Comparison 23 Vitamin E ‐ clinical impairment, Outcome 1 Total neuropathy score after 6 cycles.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Clinical impairment at 3 months Show forest plot | 1 | 74 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.8 [0.23, 2.75] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Functonal activities of daily living as measured at 3 months Show forest plot | 1 | 38 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.22 [0.06, 0.90] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Neurotoxicity rating ≥ 1 Show forest plot | 4 | 457 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.66 [0.57, 0.76] |
| 2 Neurotoxicity rating ≥ 2 Show forest plot | 3 | 148 | Risk Ratio (M‐H, Random, 95% CI) | 0.26 [0.11, 0.61] |
| 3 Neurotoxicity rating ≥ 3 Show forest plot | 3 | 419 | Risk Ratio (M‐H, Random, 95% CI) | 0.54 [0.22, 1.29] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Abnormal sensory nerve action potential Show forest plot | 1 | 16 | Risk Ratio (M‐H, Fixed, 95% CI) | 16.25 [1.07, 247.19] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Oxaliplatin specific sensory neurotoxicity ≥ grade 2 Show forest plot | 1 | 102 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.56 [0.33, 0.94] |
| 2 Chronic NCI‐CTC ≥ grade 2 Show forest plot | 3 | 153 | Risk Ratio (M‐H, Random, 95% CI) | 0.84 [0.44, 1.60] |
| 3 SNP grade 3 Show forest plot | 1 | 19 | Risk Ratio (M‐H, Fixed, 95% CI) | 7.7 [0.45, 131.36] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 NCI toxicity for neuropathy Show forest plot | 1 | 195 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.12 [0.54, 2.32] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 SSC amplitude Show forest plot | 2 | 73 | Mean Difference (IV, Fixed, 95% CI) | 1.79 [‐0.53, 4.11] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Impairment at 3 months Show forest plot | 1 | 35 | Risk Ratio (M‐H, Random, 95% CI) | 0.53 [0.21, 1.29] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 NSS symptoms of neuropathy at 3 months Show forest plot | 2 | 66 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.69 [0.50, 0.97] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 WHO neurotoxicity during trial Show forest plot | 1 | 42 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.19 [0.08, 0.47] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 NCI toxicity at 12 weeks Show forest plot | 1 | 18 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.13 [0.02, 0.89] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Chronic NCI ≥ grade 2 Show forest plot | 3 | 87 | Risk Ratio (M‐H, Random, 95% CI) | 0.29 [0.10, 0.85] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Oliguria Show forest plot | 1 | 54 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.48 [0.28, 0.81] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Abnormal VPT at 3 to 5 months Show forest plot | 1 | 20 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.67 [0.31, 1.43] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 VPT hand 3 to 5 months post treatment Show forest plot | 3 | 175 | Mean Difference (IV, Random, 95% CI) | ‐1.77 [‐4.78, 1.23] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 SNAP amplitude: sural Show forest plot | 1 | 32 | Mean Difference (IV, Fixed, 95% CI) | 3.20 [‐1.39, 7.79] |
| 2 SNAP amplitude: superficial peroneal Show forest plot | 1 | 32 | Mean Difference (IV, Fixed, 95% CI) | 2.00 [‐1.19, 5.19] |
| 3 SNAP amplitude: ulnar Show forest plot | 1 | 32 | Mean Difference (IV, Fixed, 95% CI) | 1.20 [‐1.78, 4.18] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Neuropathy after 12 cycles Show forest plot | 1 | 32 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.42 [0.19, 0.91] |
| 2 Severity of neuropathy (TNS) Show forest plot | 1 | 40 | Mean Difference (IV, Fixed, 95% CI) | ‐7.1 [‐11.98, ‐2.22] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 NCI‐CTC grade ≥ 2 Show forest plot | 1 | 92 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.75 [0.55, 1.02] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Abnormal median or sural sensory amplitude Show forest plot | 1 | 27 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.39 [0.17, 0.93] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Superficial peroneal amp 3 mos s/p treatment Show forest plot | 1 | 30 | Mean Difference (IV, Fixed, 95% CI) | 3.60 [‐0.35, 7.55] |
| 2 Ulnar SNAP amp 3 mos s/p treatment Show forest plot | 1 | 30 | Mean Difference (IV, Fixed, 95% CI) | 1.50 [‐2.71, 5.71] |
| 3 Sural amplitude uv after 6 cycles Show forest plot | 2 | 57 | Mean Difference (IV, Fixed, 95% CI) | 2.01 [‐1.60, 5.61] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Incidence of peripheral neuropathy (completed trial) Show forest plot | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.31 [0.11, 0.90] |
| 2 Incidence of peripheral neuropathy (intention to treat) Show forest plot | 1 | 35 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.46 [0.21, 1.00] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Modified peripheral neuropathy (PNP) score Show forest plot | 1 | 30 | Mean Difference (IV, Fixed, 95% CI) | ‐5.48 [‐10.73, ‐0.23] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Total neuropathy score after 6 cycles Show forest plot | 2 | 62 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.41 [0.23, 0.73] |
