Intervention Review

You have free access to this content

Interventions for preventing neuropathy caused by cisplatin and related compounds

  1. James W Albers1,*,
  2. Vinay Chaudhry2,
  3. Guido Cavaletti3,
  4. Ross C Donehower4

Editorial Group: Cochrane Neuromuscular Group

Published Online: 31 MAR 2014

Assessed as up-to-date: 4 MAR 2013

DOI: 10.1002/14651858.CD005228.pub4


How to Cite

Albers JW, Chaudhry V, Cavaletti G, Donehower RC. Interventions for preventing neuropathy caused by cisplatin and related compounds. Cochrane Database of Systematic Reviews 2014, Issue 3. Art. No.: CD005228. DOI: 10.1002/14651858.CD005228.pub4.

Author Information

  1. 1

    University of Michigan, Department of Neurology, Ann Arbor, USA

  2. 2

    Johns Hopkins Outpatient Center, Neurology, Baltimore, Maryland, USA

  3. 3

    Università degli Studi di Milano-Bicocca, Dipartimento di Chirurgia e Medicina Interdisciplinare, Monza, Italy

  4. 4

    Johns Hopkins University, Division of Medical Oncology, Baltimore, USA

*James W Albers, Department of Neurology, University of Michigan, 1C325/0032 University Hospital, 1500 E. Medical Center Drive, Box 0316, Ann Arbor, MI 48109-0032, USA. jwalbers@umich.edu.

Publication History

  1. Publication Status: New search for studies and content updated (conclusions changed)
  2. Published Online: 31 MAR 2014

SEARCH

 
Characteristics of included studies [ordered by study ID]
Argyriou 2006

MethodsProspective, randomised, open label with blind assessment


Participants30 participants with variety of cancers scheduled to receive cisplatin-based regimens plus other chemotherapy drugs (including 5 who received docetaxel)

14 participants received vitamin E (vit E) supplementation and 16 participants served as controls (no vit E)


InterventionsVit E 600 mg/day during chemotherapy and for 3 months after completion of treatment


Outcomes
  1. Sural SNAP amplitude showed a significant decline in the control group relative to the vit E group
  2. Superficial peroneal SNAP amplitude showed a significant decline in the control group relative to the vit E group
  3. Ulnar SNAP amplitude showed a significant decline in control group relative to the vit E group
  4. The overall incidence of peripheral neuropathy differed significantly (5/16 vit E vs 13/19 no vit E)


NotesThe trial authors concluded that 1. vit E was well tolerated and effectively protected against cisplatin neurotoxicity, a result supported by electrophysiology, although 2. due to the lack of a placebo group, they were unable to rule out the possibility that bias contaminated participant self reported information


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but the "simple" randomisation method was not described

Allocation concealment (selection bias)Low riskOpaque, sealed envelopes containing random numbers (I or II)

Blinding (performance bias and detection bias)
Participant
High riskNot done, as participants received vit E or no vit E

Blinding (performance bias and detection bias)
Observer
Low riskBoth neurologists were blinded to group assignment until study completion

Incomplete outcome data (attrition bias)
All outcomes
Low riskEvaluated neurotoxicity on all participants completing the trial and evaluated efficacy using intention-to-treat analyses

Selective reporting (reporting bias)Low riskYes, as above

Other biasLow riskNo evidence of other bias identified

Argyriou 2006a

MethodsProspective, randomised, open label with blind assessment


ParticipantsInitially participants with colon cancer scheduled to receive FOLFOX-4, which included oxaliplatin 85 mg/m² every 2 weeks for 12 courses. 20 participants were randomised to treatment with oxcarbazepine and 20 to no treatment (16 participants in each group completed the study for n = 32 total)


InterventionsOxcarbazepine, initiated at 150 mg/day and titrated over 4 weeks to 600 mg twice daily and maintained for 20 weeks)


Outcomes
  1. Sural SNAP amplitude showed a significant decline in the control group relative to the oxcarbazepine group
  2. Superficial peroneal SNAP amplitude showed a significant decline in the control group relative to the oxcarbazepine group
  3. Ulnar SNAP amplitude showed no significant group difference
  4. The overall incidence of peripheral neuropathy differed significantly (9/20 oxcarbazepine vs 16/20 no treatment).
  5. The severity of neuropathy (total neuropathy score) significantly favoured the oxcarbazepine group


NotesThe trial authors concluded that oxcarbazepine might be able to protect against oxaliplatin-induced neuropathy. They acknowledged that the small sample size and the lack of a placebo group were limitations of the study design, but that larger placebo-controlled trials were warranted


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomisation method not specifically described, other than that there was a 1:1 ratio

Allocation concealment (selection bias)Unclear riskDescribed as "concealed", known only to the randomisation co-ordinator

Blinding (performance bias and detection bias)
Participant
High risk"Open label" trial

Blinding (performance bias and detection bias)
Observer
Low riskThe evaluators and the senior oncologist were blinded to participant group assignment

Incomplete outcome data (attrition bias)
All outcomes
Low riskInvestigators evaluated neurotoxicity in all participants completing the trial and measured efficacy using intention-to-treat analyses

Selective reporting (reporting bias)Low riskYes, as above

Other biasUnclear riskNo evidence of other bias identified

Arrieta 2011

MethodsProspective, randomised, placebo-controlled trial


ParticipantsHistologically-confirmed stage IIIB/IV NSCLC treated with cisplatin 80 mg/m² AND paclitaxil 175 mg/m²


InterventionsRetinoic acid (ATRA), 20 mg/m²/day beginning 1 week prior to chemotherapy initiation and continuing until completion of two courses

45 intervention and 47 control participants


OutcomesNCSs including sural, median, and ulnar sensory response (and peroneal motor) amplitude and latency measures, with quantitative results converted to a non-parametric "grade of damage," ranked from zero to three for latency and amplitude measures


NotesPeriod of evaluation (ie. 6 weeks to the end of second course of chemotherapy) was very likely inadequate. Also, the investigators did not separate motor and sensory electrophysiology, which is physiologically nonsensical


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskTable of random numbers

Allocation concealment (selection bias)Low riskNone detected

Blinding (performance bias and detection bias)
Participant
Low riskBlinded

Blinding (performance bias and detection bias)
Observer
Low riskNone apparent

Incomplete outcome data (attrition bias)
All outcomes
Low riskNone

Selective reporting (reporting bias)Low riskNone identified

Other biasHigh riskSignificant differences between study and control groups at baseline

Bogliun 1996

MethodsProspective, randomised, placebo-controlled


Participants54 participants with ovarian cancer treated with cisplatin (50 mg/m² and 75 mg/m² x 3 weeks); 27 participants received GSH and 27 participants served as controls


InterventionsGSH 2.5 g before cisplatin


OutcomesFor GSH vs control group:

  1. VPT: 2- to 3-fold increase vs 7- to 10-fold increase
  2. Sural amplitude decreases of 12% and 35% for low and high dose CDDP (GSH) and of 58% and 68% for high and low dose CDDP (control)
  3. NIS changed > 12 points in 5/19 vs 8/16
  4. NSS symptoms of neuropathy 14/19 vs 16/16


Notes1. No statistics used because of small numbers
2. Trend towards less peripheral neuropathy with GSH in all measures. The authors concluded that, despite the small numbers and lack of statistical analyses, neuroprotection was not complete but there was a trend toward less severe toxicity after cotreatment with GSH


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskUnclear

Blinding (performance bias and detection bias)
Participant
High riskRandomised but not blinded

Blinding (performance bias and detection bias)
Observer
High riskSame. The analyses of data were performed without knowledge of treatment

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskFew subjects lost to follow-up

Selective reporting (reporting bias)High riskNo formal statistical analysis presented

Other biasLow riskNone identified

Cascinu 1995

MethodsProspective, randomised, placebo-controlled


Participants50 participants with ovarian cancer treated with cisplatin (40 mg/m² every week x 9 weeks); 25 participants received GSH and 25 participants served as controls


InterventionsGSH 1.5 g/m² or placebo before cisplatin


Outcomes1. Sural SNAP: significant change in control group only

2. Ulnar SNAP: significant change in control group only

3. Median SNAP: significant change in control group only

4. At 9 weeks, 0 participants in the GSH group developed neurotoxicity by WHO criteria vs 16/18 in the control group and at 15 weeks, the figures were 4/24 in the GSH group vs 16/18 in the control group


NotesLong-term follow-up unclear. The trial authors believed that the results established GSH as a promising and partially effective agent for the prevention of cisplatin-induced and oxaliplatin-induced neurotoxicity


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandom, computer-generated

Allocation concealment (selection bias)Low riskAdequate, concealed envelopes

Blinding (performance bias and detection bias)
Participant
Low riskPlacebo-controlled

Blinding (performance bias and detection bias)
Observer
Low riskThe same blinded evaluator examined all participants

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on assessable participants

Selective reporting (reporting bias)Unclear riskBased on assessable participants

Other biasUnclear riskNone identified

Cascinu 2002

MethodsProspective, randomised, placebo-controlled


Participants52 participants with colorectal cancer treated with oxaliplatin (100 mg/m² x 2 every 2 weeks); 26 participants received GSH and 26 placebo-treated participants served as controls


InterventionsGSH 1.5 g/m² or normal saline as placebo before administration of oxaliplatin

Oxaliplatin 100 mg/m² x 2 every 2 weeks


OutcomesNeurotoxicity (NCI CTC)

Neurophysiology (sural sensory nerve conduction)


NotesGSH reduced symptoms and signs of neuropathy significantly. GSH was partially protective on measures of sural amplitude and latency. The trial authors believed that the results established GSH as a promising and partially effective agent for the prevention of cisplatin-induced and oxaliplatin-induced neurotoxicity


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskUsed cards from a computer-generated list in sealed envelopes

Allocation concealment (selection bias)Low riskUsed cards from a computer-generated list in sealed envelopes

Blinding (performance bias and detection bias)
Participant
Low riskDouble-blinded RCT

Blinding (performance bias and detection bias)
Observer
Low riskDouble-blinded RCT

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskAnalyses based on evaluable participants

Selective reporting (reporting bias)Unclear riskAnalyses based on evaluable participants

Other biasLow riskNone identified

Chay 2010

MethodsProspective, randomised, placebo-controlled


Participants27 participants (19 intervention and 8 control) with histologically verified colorectal cancer receiving oxaliplatin and either XELOX (including oxaliplatin 130 mg/m² on day 1 every 21 days, received by 22 of 27 participants) or FOLFOX (including oxaliplatin um 85 mg/m² on day 1 every 2 weeks, received by 5 of 22 participants) for 12 cycles of chemotherapy; 78% of participants received > 600 mg/m² of oxaliplatin-based chemotherapy


InterventionsCalcium 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 (Arm A) or placebo (Arm B)


OutcomesNeuropathy was measured by clinical assessment and NCS

NCS included secondary outcomes of sural, superficial peroneal, median, ulnar, and radial sensory nerves, plus median, ulnar, radial, posterior tibial, and peroneal motor nerves. The results were considered abnormal if each of the studies was outside the normal values of the laboratory controls.

Incidence of subjective neuropathy and grade 3 numbness in both groups were compared, as were differences in all objective neuropathy.


NotesThe planned follow-up was for a total of 3 years, with review every 3 months for the first year, then every 4 months the second year, and then every 6 months in subsequent years

Due to the trial termination, only 19 participants reached the end of the treatment for study. A total of 13 participants completed the planned 6 months of adjuvant chemotherapy. No participants in the treatment arm terminated chemotherapy prematurely while 3 participants in the placebo arm terminated treatment because of grade 3 neurotoxicity

Overall 22 out of 27 participants experienced neuropathy. The subjective neuropathy rate was 77% in Arm A and 86% in Arm B (P = 0.6). At the end of treatment, three participants in Arm A and 0 in Arm B had grade 3 numbness (P = 0.09). There was no significant difference in neuropathy between arms, whether during or at the end of treatment. The median objective neuropathy score was 6 in Arm A and 0 in Arm B (P = 0.02), indicating that the Ca/Mg treatment group had worse objective neuropathy scores than the placebo group
Conclusion: premature closure of this study


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskStates that the trial is randomised but the method is not described

Allocation concealment (selection bias)Low riskNo details

Blinding (performance bias and detection bias)
Participant
Low riskNo details, but participants said to be blinded

Blinding (performance bias and detection bias)
Observer
Low riskObserver blinded

Incomplete outcome data (attrition bias)
All outcomes
Low riskTrial terminated early

Selective reporting (reporting bias)Low riskNone identified

Other biasLow riskNone identified

Colombo 1995

MethodsProspective, randomised, placebo-controlled study


ParticipantsWomen with relapsed ovarian carcinoma treated with weekly cisplatin


Interventions16 receiving glutathione (GSH) and 17 controls


OutcomesProvided results of sural sensory NCS amplitudes (mean), and used the NCI-CTC neurotoxicity scale


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
Unclear riskNot described

Blinding (performance bias and detection bias)
Observer
Unclear riskNot described

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNot described

Selective reporting (reporting bias)High riskNo details as to neurological assessment provided. No P values given for any significance (although calculable from the data provided). No objective data provided except mean sural amplitude without significance values

Other biasLow riskNone identified

DeVos 2005

MethodsRandomised phase II study


ParticipantsA randomised phase II study of paclitaxel with carboplatin with/without amifostine, as first line treatment of 90 women with advanced ovarian carcinoma


InterventionsAmifostine, 740 mg/m² preceding chemotherapy


OutcomesSymptoms of neurotoxicity and standard NCI-CTC rating


NotesThere are several problems with the study:

  1. The two groups are not the same. The ones with worse tumour type are in control group. More participants in the intervention arm had treatment delays
  2. It is not clear if the authors are making a case for a positive effect of amifostine in this study. They are assuming that it works for cisplatin and using pooled data to show that it works.
  3. Too many subgroups tested. In over all there is no difference, for grade 1, intervention group is worse, for grade 2 to 3 the control group is worse
  4. Toxicity is worse in the intervention group.
  5. Clearly no subjective change or even QOL change but no clear data provided to see if CTC of neurotoxicity was based on objective criteria.
  6. This is more about taxanes given the modest neurotoxicity of carboplatin in most patients


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
Unclear riskNot described. Unclear if participants were blinded, as control group did not receive placebo

Blinding (performance bias and detection bias)
Observer
High riskNot done

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk10 did not complete (5 each arm), 1 received low dose; information not clear on other 46

Selective reporting (reporting bias)Low riskNone detected

Other biasLow riskNone identified

Gallardo 1999

MethodsAn open, single-blinded pilot study to investigate the feasibility of 5-day scheduling of amifostine with radiotherapy and cisplatin (total dose 400 mg/m²)


Participants20 women with locally advanced, histologically diagnosed cervical cancer, treated with radiotherapy randomised to receive cisplatin at 20 mg/m² for 5 days in 2 cycles (beginning day 1 and day 22) and 100 mg/m² for 2 cycles with amifostine before cisplatin infusion (Group A) or cisplatin in the same dose without amifostine (Group B)


InterventionsAmifostine 825 mg/m² before cisplatin infusion


OutcomesNeurotoxicity grade (presumably NCI-CTC neurotoxicity)


NotesNeurotoxicity scale not specified, but from NCI


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskNot specified

Allocation concealment (selection bias)High riskPilot

Blinding (performance bias and detection bias)
Participant
Unclear riskParticipants

Blinding (performance bias and detection bias)
Observer
High riskObserver not blinded

Incomplete outcome data (attrition bias)
All outcomes
Low riskNot applicable

Selective reporting (reporting bias)Low riskNone detected

Other biasUnclear riskSame

Gandara 1995

MethodsProspective, randomised, placebo-controlled


Participants214 participants with ovarian cancer, SCLC, or NSCLC treated with cisplatin (100 mg/m² every 4 weeks x 6); 106 participants received DDTC and 108 participants received placebo. Data were available for n = 195 participants from the combined lung and ovarian cancer populations (placebo, 99 participants; DDTC, 96 participants)


InterventionsDDTC 1.6 g/m² 15 min before cisplatin


OutcomesNCI toxicity rating scale showed 13 of 96 (13%) in the DDTC group and 12 of 99 (12%) in the placebo group developed neuropathy (P = not significant)


NotesAdverse effects were reported in all study participants with severe adverse events in 27 control group and 30 of DDTC arm. The authors concluded that DDTC is not effective against protecting cisplatin-induced peripheral neuropathy


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskPresumably, but not described

Blinding (performance bias and detection bias)
Participant
Low riskPlacebo-controlled

Blinding (performance bias and detection bias)
Observer
Low riskDescribed as blinded, including a blinded interim safety analysis, Independent statistical analyses

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on evaluable participants but a small percentage were lost to follow-up

Selective reporting (reporting bias)Unclear riskSame

Other biasLow riskNone detected

Grothey 2011

MethodsProspective, randomised, placebo-controlled, double-blind


Participants102 participants with colon cancer; 50 received Ca/Mg and 52 received placebo


InterventionsCa/Mg


OutcomesNCI-CTC adverse event score for neuropathy


NotesData appear to support the hypothesis that Ca/Mg infusions (marginally) decrease oxaliplatin-induced chronic sensory motor neuropathy but not acute cold-induced paraesthesias and cramps. However, the study was not completed for 6 months and had early closure (due to reports of treatment interference)


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomised but not described

Allocation concealment (selection bias)Unclear riskPresumably, but not described

Blinding (performance bias and detection bias)
Participant
Low riskPlacebo-controlled

Blinding (performance bias and detection bias)
Observer
Unclear riskNot described. Independent statistical analyses

Incomplete outcome data (attrition bias)
All outcomes
Low riskBased on evaluable subjects but small percentage lost to follow-up

Selective reporting (reporting bias)Low riskNone detected

Other biasUnclear riskStudy closed early for "Treatment interference"

Hovestadt 1992

MethodsProspective, randomised, placebo-controlled


Participants18 women with epithelial ovarian cancer treated with cisplatin (75 mg/m²) and cyclophosphamide (750 mg/m²) every 3 weeks up to 9 cycles; 7 participants received Org 2766 and 11 participants served as controls


InterventionsOrg 2766 low dose (0.25 mg/m²) and high dose (1 mg/m²), immediately before cisplatin and 24 hours later


OutcomesMean VPT in placebo-treated participants was higher than in the low- and high-dose Org 2766-treated participants at 1 month, intermediate 1 to 4 months after treatment, and higher again after 4 to 12 and 12 to 24 months)


NotesDescriptive analysis only without formal statistical tests; the number of evaluable participants decreased during the study. The authors suggested that treatment with Org 2766 to prevent a cisplatin-induced neuropathy should possibly be continued up to 4 months after the last cycle of cisplatin


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskPresumably done, but not described

Blinding (performance bias and detection bias)
Participant
Low riskPlacebo-controlled

Blinding (performance bias and detection bias)
Observer
Low riskPlacebo-controlled

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on evaluable participants, but substantial dropout

Selective reporting (reporting bias)Unclear riskSame

Other biasLow riskNone identified

Ishibashi 2010

MethodsProspective, randomised, placebo-controlled


Participants33 participants with metastatic colorectal cancer treated with FOLFOX, which included oxaliplatin, 85 mg/m² every 2 weeks for 6 cycles; 17 Ca/Mg-treated and 16 control participants


InterventionsCa (850 mg) and Mg (720 mg) infusions, before and after infusion of oxaliplatin or placebo


OutcomesNCI-CTC and Debiopharm Neurotoxicity Scale (DEB-NTS)


NotesFOLFOX chemotherapy includes administration of 5-fluorouracil and l-leucovorin with each of 6, 2-week treatment cycles


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not specifically described

Allocation concealment (selection bias)Low riskCentralised allocation via random number assignment

Blinding (performance bias and detection bias)
Participant
Low riskDescribed as "double-blind" study

Blinding (performance bias and detection bias)
Observer
Low riskDescribed as "double-blind" study

Incomplete outcome data (attrition bias)
All outcomes
Low riskEarly termination of enrolment because of relatively "poorer" performance in the Ca/Mg group at interim analyses

Selective reporting (reporting bias)Unclear riskYes, based on data to time of termination

Other biasUnclear riskNo evidence of other bias identified

Kanat 2003

MethodsProspective, randomised


Participants38 participants with NSCLC treated with paclitaxel (175 mg/m²) and carboplatin (AUC = 6) were randomised; 19 participants received amifostine and 19 participants served as controls


InterventionsAmifostine 910 mg/m² every 3 weeks x 6


OutcomesThe mean SNAP amplitudes for both groups were comparable at baseline and also after 6 cycles of chemotherapy, there being no decline (clinically or statistically apparent) in the mean amplitudes for either group after treatment. 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 (P = 0.018)


NotesThe trial authors concluded that the addition of amifostine to carboplatin and paclitaxel may prevent or reduce the incidence of neurotoxicity in the treatment of NSCLC, but they cautioned in the text that the data from their trial were insufficient to support use of amifostine as a neuroprotective agent for paclitaxel and carboplatin chemotherapy


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskUnclear, as not described

Blinding (performance bias and detection bias)
Participant
High riskRandomised but no placebo

Blinding (performance bias and detection bias)
Observer
Unclear riskNot described

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskAll data included

Selective reporting (reporting bias)Unclear riskNo subjects dropped out of study

Other biasLow riskNone detected

Kemp 1996

MethodsProspective, randomised


Participants242 participants with advanced ovarian cancer treated with cisplatin (100 mg/m²) and cyclophosphamide (1000 mg/m²) every 3 weeks x 6 cycles; 122 participants received amifostine, 120 participants served as controls


InterventionsAmifostine 910 mg/m² pretreatment


OutcomesNeurotoxicity, defined as symptoms of peripheral neuropathy or 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 (Kemp 1996). By treatment cycle 5, there was a significant difference (P = 0.015) between the groups, favouring the amifostine group. Following the last treatment cycle, the NCI neurological 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)


NotesThe authors concluded that pretreatment with amifostine reduced the neurotoxicity associated with cisplatin chemotherapy, based on reduction of the NCI toxicity rating scale


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskPresumably, but not described

Blinding (performance bias and detection bias)
Participant
High riskRandomised but not placebo-controlled

Blinding (performance bias and detection bias)
Observer
Low riskEvaluations performed by a third party blinded to treatment or adverse events

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskUnclear if difference in dropouts between groups

Selective reporting (reporting bias)Low riskUsed intent to treat analyses

Other biasLow riskNone identified

Knijn 2011

MethodsRetrospective analyses without randomisation (intervention at discretion of treating physicians)


Participants755 previously untreated colorectal cancer patients who had been randomised to received capecitabine, oxaliplatin and bevacizumab or the same with the addition of cetuximab (CBC). 6 cycles over 6 months. 732 participants were evaluable


InterventionsCa/Mg, IV 2.25 mmol calcium glubionate plus MgCl 4 mMol in 100 mL glucose over 15 minutes, before and after oxaliplatin infusion retrospectively given at the discretion of the treating physician


OutcomesNCI-CTC v 3.0


NotesVery confusing retrospective dosing. Only reduced grade 1 neurotoxicity but not grade ≥ 2; and only early not late?


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)High riskNot randomised; at discretion of treating physicians

Allocation concealment (selection bias)High riskAs above

Blinding (performance bias and detection bias)
Participant
High riskNot done

Blinding (performance bias and detection bias)
Observer
High riskNot done

Incomplete outcome data (attrition bias)
All outcomes
High riskUnknown, but inadequate handling of unavailable data

Selective reporting (reporting bias)Unclear riskUnknown

Other biasUnclear riskNot detected

Kottschade 2011

MethodsProspective, randomised, placebo-controlled


ParticipantsColorectal cancer (but also breast, lung, and other cancers) treated with taxanes (109), cisplatin (8), carboplatin (2), oxaliplatin (50) or combination (20); total 189 evaluable (of 207 enrolled)


InterventionsVitamin E 300 mg twice daily or placebo


OutcomesNCI-CTC NCI-CTC adverse event score for neuropathy 3.0


NotesDifferent types of malignancy were not of comparable distribution in intervention and control groups, although the chemotherapy treatments were similar. The inclusion of a large number of participants receiving taxanes was confounding


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandom number sequence

Allocation concealment (selection bias)Unclear riskNo information provided

Blinding (performance bias and detection bias)
Participant
Low riskSubject blinded

Blinding (performance bias and detection bias)
Observer
Unclear riskNo information provided

Incomplete outcome data (attrition bias)
All outcomes
Unclear risk18 cancellations/withdrawals out of 207 participants

Selective reporting (reporting bias)Low riskNone identified

Other biasLow riskNone identified

Lin 2006

MethodsProspective, randomised, placebo-controlled


Participants14 participants with colorectal cancer treated with oxaliplatin (85 mg/m² every 2 weeks for 12 cycles, plus weekly fluorouracil and leucovorin); 5 participants received N-acetylcysteine (NAC) and 9 served as placebo controls


InterventionsNAC 1200 mg one and one-half hours before each oxaliplatin treatment or placebo


OutcomesSensory (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. Clincal neurotoxicity was assessed every 2 weeks using the NCI-CTC rating scale. After 12 cycles of treatment, the incidence of ≥ Grade 1, 2, and 3 neurotoxicity was 80%, 20%, and 0% among the 5 participants in the NAC group, respectively, and 100%, 89%, and 33% in the control group (P = 0.01)


NotesThe nerve conduction data were reported only for the NAC group, indicating no significant deterioration over the full trial


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
Unclear riskDescribed only as randomised and placebo-controlled

Blinding (performance bias and detection bias)
Observer
Unclear riskPresumably, but not described

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskNot described, but all data evaluated

Selective reporting (reporting bias)Unclear riskReported nerve conduction results from the NAC group but not the control group

Other biasUnclear riskNo evidence of other bias identified

Lorusso 2003

MethodsProspective, randomised


Participants187 participants with ovarian cancer treated with carboplatin and paclitaxel every 21 days for 6 cycles; 93 participants received amifostine and 94 participants served as controls


InterventionsAmifostine 910 mg/m² pretreatment


OutcomesAmifostine appeared to be protective against neurotoxicity (grade 3 to 4 neurotoxicity 3.7% vs 7.2%; P = 0.02)


NotesThe authors concluded that amifostine could exert some protection from the cumulative toxicity associated with carboplatin and paclitaxel


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer generated list

Allocation concealment (selection bias)Low riskRandomisation performed centrally to avoid bias

Blinding (performance bias and detection bias)
Participant
High riskNot placebo controlled

Blinding (performance bias and detection bias)
Observer
Unclear riskPerformed by "dedicated" physicians but unclear if blinded

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on evaluable participants but small proportion of missing data

Selective reporting (reporting bias)Unclear riskBased on evaluable participants

Other biasLow riskNone identified

Lu 2008

MethodsProspective, randomised, placebo-controlled


ParticipantsPatients (n = 92; 46 in each group) with digestive tract tumours (colorectal or gastric carcinoma) treated with an oxaliplatin regimen (FOLFOX 4): oxaliplatin 85 mg/m² every 2 weeks; 4-week cycle repeated for 6 cycles


InterventionsAmifostine, 500 mg/m² every 4 weeks for 6 cycles, or placebo


OutcomesNCI-CTC


NotesMore haematological toxicity in PC+A arm; however mean values were similar


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskRandomised using random number table

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
Unclear riskNot described

Blinding (performance bias and detection bias)
Observer
Unclear riskNot described

Incomplete outcome data (attrition bias)
All outcomes
Low risk7 withdrawals in the AMIF group; 10 withdrawals in the placebo group; well described

Selective reporting (reporting bias)Low riskNone detected

Other biasLow riskNone detected

Milla 2009

MethodsProspective, randomised, placebo-controlled


Participants27 participants with colorectal cancer treated with FOLFOX4, which includes oxaliplatin (85 mg/m² with each 2 week cycle up to 12 cycles) plus other drugs (none known to be neurotoxic); 14 participants received GSH and 13 participants received saline placebo


InterventionsGSH; 1500 mg/m² or placebo before each oxaliplatin treatment


OutcomesAt the end of treatment, only moderate neurotoxicity 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). The difference was considered statistically significant (Mann-Whitney test; P = 0.0037)


Notes-


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskNot described

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
Unclear riskDescribed only as randomised and placebo-controlled

Blinding (performance bias and detection bias)
Observer
Unclear riskPresumably but not described

Incomplete outcome data (attrition bias)
All outcomes
Low riskNo dropouts (all data included)

Selective reporting (reporting bias)Unclear riskNot described but data from all subjects included

Other biasUnclear riskNo evidence of other bias identified

Pace 2003

MethodsProspective, randomised, placebo-controlled (control participants were untreated)


Participants27 participants with various solid tumours (lung, ovarian, rhinopharyngeal, gastric, testicular, oesophagus, ethmoid, and tongue cancer) treated with cisplatin (cumulative dose > 300 mg/m²); 13 participants treated with vitamin E (vit E) and 14 participants served as controls


InterventionsVit E 300 mg/day starting before cisplatin and continuing up to 3 months after cisplatin treatment


OutcomesThe median SNAP amplitudes were reduced by a greater degree in those without vit E than those taking vit E (without vit E group: baseline 14.5 ± 8.5 µv; 6 months later 13.6 ± 9.2 µv; vit E group: baseline 15.5 ± 6.3 µv; 6 months later 13.7 ± 5.5 µv). The differences between sural SNAPs in the 2 groups were not significant


NotesOf 47 patients enrolled, 20 dropped out mainly for disease progression. Clinical impairment on neurological examination was measured by an non-validated scale. The trial authors concluded that supplemental vit E decreases the incidence and severity of neurotoxicity


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
High riskParticipants were not blinded

Blinding (performance bias and detection bias)
Observer
High riskNeurologists performing examinations were not blinded to treatment status

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on evaluable participants, although substantial drop out

Selective reporting (reporting bias)Unclear riskBased on evaluable participants, although substantial drop out

Other biasLow riskNone identified

Planting 1999

MethodsProspective, randomised, placebo-controlled


Participants74 participants with advanced head and neck cancer treated with cisplatin (70 mg/m² weekly); 37 participants received amifostine and 37 participants served as controls


InterventionsAmifostine 740 mg/m² immediately before cisplatin


OutcomesParticipants receiving amifostine showed significantly less subclinical neurotoxicity as measured by the VPT relative to controls


NotesThe trial authors explained that the borderline significant group difference for the right hand VPT reflected an imbalance at baseline for the groups and limited data. The authors concluded that amifostine reduced the risk of subclinical neurotoxicity caused by cisplatin but did not result in a higher cisplatin dose intensity


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskNot described

Blinding (performance bias and detection bias)
Participant
Unclear riskNot described

Blinding (performance bias and detection bias)
Observer
Unclear riskNot described

Incomplete outcome data (attrition bias)
All outcomes
Low riskBased on evaluable data, but there was limited drop out

Selective reporting (reporting bias)Unclear riskSame

Other biasLow riskNone identified

Roberts 1997

MethodsProspective, randomised, placebo-controlled


Participants196 participants with epithelial ovarian cancer treated with cisplatin (75 to 100 mg/m²) and cyclophosphamide (600 to 1000 mg/m²); 129 participants treated with Org 2766 and 67 participants served as controls


InterventionsOrg 2766 2 mg/m² or 4 mg/m²


OutcomesVPT increased during the study, independent of receiving Org 2766, with no difference in the rate of change or the degree of neuropathy


NotesNo evidence of efficacy in preventing or slowing development of neuropathy. The authors concluded that no benefit could be demonstrated using Org 2766 and suggested that high Org 2766 dose might eventually increase the peripheral neurotoxicity of cisplatin


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskRandomised but method not described

Blinding (performance bias and detection bias)
Participant
Low riskDecribed as blinded, placebo-controlled study

Blinding (performance bias and detection bias)
Observer
Unclear riskPresumably, but not described

Incomplete outcome data (attrition bias)
All outcomes
Low riskBased on evaluable participants, but limited drop out

Selective reporting (reporting bias)Low riskBased on evaluable participants, but limited drop out

Other biasLow riskNone identified

Schmidinger 2000

MethodsProspective, randomised, placebo-controlled


Participants20 participants with non-small cell lung cancer (NSCLC) and head and neck cancer treated with cisplatin (80 mg/m² x every 4 weeks); 11 participants treated with GSH and 9 participants served as controls


InterventionsGSH 5 g/m² before cisplatin or placebo


OutcomesNo difference in clinical or nerve conduction measurement between the 2 groups


NotesThe trial authors concluded that the combination of GSH and cisplatin was safe and the antitumour efficacy of cisplatin not impaired. They identified no difference between GSH-treated and control groups


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskRandomised but method not described

Blinding (performance bias and detection bias)
Participant
Unclear riskNot specifically described

Blinding (performance bias and detection bias)
Observer
Unclear riskNot specifically described

Incomplete outcome data (attrition bias)
All outcomes
Low riskBased on evaluable participants but only 1 participant dropped out (in control group)

Selective reporting (reporting bias)Low riskBased on evaluable participants but only 1 participant dropped out (in control group)

Other biasLow riskNone identified

Smyth 1997

MethodsProspective, randomised, placebo-controlled


Participants151 participants with ovarian cancer treated with cisplatin (100 mg/m² x every 3 weeks x 6 courses); 74 participants received GSH and 77 participants served as controls


InterventionsGSH 3 g/m² before cisplatin


OutcomesMean increase in Hospital Anxiety and Depression score was 0.8 in the GSH group and 2.5 in the placebo group (depression score)
45 of 47 in Rotterdam had better scores in GSH


NotesParticipants reported improvement in QOL if they received GSH
They had significantly less tingling in hands and feet
GSH allowed more cycles of cisplatin because of less toxicity. The authors concluded that participants receiving GSH had a clear improvement in quality of life with significantly less tingling in their hand and feet and received more cycles of cisplatin because of less toxicity


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Low riskComputer generated

Allocation concealment (selection bias)Low riskCentral randomisations

Blinding (performance bias and detection bias)
Participant
Unclear riskPresumably (placebo controlled), but not described

Blinding (performance bias and detection bias)
Observer
Unclear riskPresumably (placebo controlled), but not described

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on evaluable participants but substantial drop out

Selective reporting (reporting bias)Unclear riskAs above, analyses performed on relatively small subset of evaluable participants

Other biasLow riskNone identified

van der Hoop 1990

MethodsProspective, randomised, placebo-controlled


Participants55 participants with epithelial ovarian cancer treated with cisplatin (75 mg/m²) and cyclophosphamide (750 mg/m²); 33 participants received Org 2766 and 22 participants served as controls


InterventionsOrg 2766 low dose (0.25 mg/m²) and high dose (1 mg/m²), immediately before and after cisplatin


OutcomesIn the high dose Org 2766 group, the VPT increased about 2-fold, significantly less than the nearly 8-fold increase in the placebo group


NotesThe authors felt that Org 2766 prevented or attenuated cisplatin neuropathy


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Low riskDone before study, and treatment codes were not revealed to clinicians

Blinding (performance bias and detection bias)
Participant
Low riskPlacebo-controlled

Blinding (performance bias and detection bias)
Observer
Low riskTreatment codes were not revealed to clinicians

Incomplete outcome data (attrition bias)
All outcomes
Low riskIntention-to-treat analyses based on all enrolled participants

Selective reporting (reporting bias)Low riskIntention-to-treat analyses based on all enrolled participants

Other biasLow riskNone identified

van Gerven 1994

MethodsProspective, randomised, placebo-controlled


Participants42 participants with testicular and adenocarcinoma of unknown primary treated with cisplatin (100 mg/m² for at least 4 cycles) and with different combinations of etoposide, bleomicin and ifosphamide; 19 participants treated with Org 2766 and 23 participants served as controls


InterventionsOrg 2766 2 mg/d x 5 days or placebo


OutcomesBorderline significant difference in degree of VPT favouring Org 2766 vs placebo. Symptom scores were not significantly different.


NotesThe trial authors indicated that Org 2766 cannot completely prevent cisplatin neuropathy but may ameliorate nerve damage


Risk of bias

BiasAuthors' judgementSupport for judgement

Random sequence generation (selection bias)Unclear riskRandomised but method not described

Allocation concealment (selection bias)Unclear riskRandomised but method not described

Blinding (performance bias and detection bias)
Participant
Unclear riskNot described

Blinding (performance bias and detection bias)
Observer
Unclear riskNot described

Incomplete outcome data (attrition bias)
All outcomes
Unclear riskBased on participants completing 4 cycles of chemotherapy

Selective reporting (reporting bias)Low riskAs above, few eligible participants had missing data

Other biasLow riskNone identified

 
Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion

Ali 2009Review of existing studies

Ao 2012Meta-analysis of existing data

Argyriou 2005Unable to separate the effects of cisplatin from those of paclitaxel

Block 2005Review article

Bogliun 1992Neither the control group nor the intervention group developed neurotoxicity, so unable to assess efficacy

Cancer Care Ontario 2012Review article

Capizzi 1995Review article

Cassidy 1998Randomised, placebo-controlled trial that was discontinued prematurely because of nausea and vomiting, limiting analyses. Nevertheless, the available data suggested that nimodipine exacerbated (not prevented) neurotoxicity

Catalano 1999Abstract

Chen 2012Not a randomised controlled trial; analyses based on retrospective data

Culy 2001Review article

De Grandis 2007Review article

De Vos 2005Unable to separate the effects of carboplatin from those of paclitaxel

Durand 2003Reports of treatment of acute sensory symptoms, not prevention of neuropathy

Gamelin 2004Retrospective, non-randomised study

Gamelin 2008Letter

Gedlicka 2003Letter

Gispen 1992No details of study provided

Glover 1987Phase I trial

Glover 1989Phase I trial

Glover 2003Treatment schedule not comparable

Gradishar 2001Phase II study, not a randomised controlled trial

Grothey 2005Review article

Haigentz 2003Phase I trial

Heidenreich 1999Limited information

Hensley 2009Description of practice guideline

Hilkens 1995Study involved dosing of cisplatin

Hilpert 2005Unable to separate the effects of carboplatin from those of paclitaxel

Hochster 2007Letter

Kesari 2005Letter

Leong 2003Only two doses of carboplatin

Lissoni 1997Details of neuropathy not provided

Lissoni 1999Groups not comparable and previous therapy unknown

Lissoni 2002Not a randomised trial

Miller 2008Unable to separate the effects of cisplatin from those of docetaxel

Mollman 1988Study of risk factors associated with cisplatin neuropathy

Moore 2003Phase II, open label study without placebo or any comparison group

Pace 2005Letter

Pace 2010Substantial dropout rate, excluding intention-to-treat analyses

Penz 2001Pilot study, not controlled trial

Rick 2001Open label, unblinded trial

Wang 2007Randomised but neither blinded nor placebo-controlled

Wen 2013Meta-analysis of calcium/magnesium infusions for the prevention of oxaliplatin-related neurotoxicity in people with colorectal cancer

Wilkes 2007Review article

 
Characteristics of studies awaiting assessment [ordered by study ID]
Dong 2010

Methods

Participants

Interventions

Outcomes

NotesAwaiting translation

Pang 2010

Methods

Participants

Interventions

Outcomes

NotesAwaiting translation

Romano 2011

Methods

Participants

Interventions

Outcomes

NotesArticle not obtainable

 
Comparison 1. Amifostine - clinical impairment

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Clinical impairment at 3 months174Risk Ratio (M-H, Fixed, 95% CI)0.8 [0.23, 2.75]

 
Comparison 2. Amifostine - functional activities of daily living

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Functonal activities of daily living as measured at 3 months138Risk Ratio (M-H, Fixed, 95% CI)0.22 [0.06, 0.90]

 
Comparison 3. Amifostine - neurotoxicity rating

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Neurotoxicity rating ≥ 14457Risk Ratio (M-H, Fixed, 95% CI)0.66 [0.57, 0.76]

 2 Neurotoxicity rating ≥ 23148Risk Ratio (M-H, Random, 95% CI)0.26 [0.11, 0.61]

 3 Neurotoxicity rating ≥ 33419Risk Ratio (M-H, Random, 95% CI)0.54 [0.22, 1.29]

 
Comparison 4. Calcium/magnesium - nerve conduction study (NCS) results

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Abnormal sensory nerve action potential116Risk Ratio (M-H, Fixed, 95% CI)16.25 [1.07, 247.19]

 
Comparison 5. Calcium/magnesium - sensory neuropathy (SNP)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Oxaliplatin specific sensory neurotoxicity ≥ grade 21102Risk Ratio (M-H, Fixed, 95% CI)0.56 [0.33, 0.94]

 2 Chronic NCI-CTC ≥ grade 23153Risk Ratio (M-H, Random, 95% CI)0.84 [0.44, 1.60]

 3 SNP grade 3119Risk Ratio (M-H, Fixed, 95% CI)7.7 [0.45, 131.36]

 
Comparison 6. Diethyldithiocarbamate - National Cancer Institute (NCI) neurotoxicity rating scale

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 NCI toxicity for neuropathy1195Risk Ratio (M-H, Fixed, 95% CI)1.12 [0.54, 2.32]

 
Comparison 7. Glutathione - nerve conduction study (NCS) results

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 SSC amplitude273Mean Difference (IV, Fixed, 95% CI)1.79 [-0.53, 4.11]

 
Comparison 8. Glutathione - clinical impairment

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Impairment at 3 months135Risk Ratio (M-H, Random, 95% CI)0.53 [0.21, 1.29]

 
Comparison 9. Glutathione - Neuropathy Symptom Score (NSS)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 NSS symptoms of neuropathy at 3 months266Risk Ratio (M-H, Fixed, 95% CI)0.69 [0.50, 0.97]

 
Comparison 10. Glutathione - World Health Organization (WHO) evidence of neurotoxicity

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 WHO neurotoxicity during trial142Risk Ratio (M-H, Fixed, 95% CI)0.19 [0.08, 0.47]

 
Comparison 11. Glutathione - chronic National Cancer Institute (NCI) toxicity rating

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 NCI toxicity at 12 weeks118Risk Ratio (M-H, Fixed, 95% CI)0.13 [0.02, 0.89]

 
Comparison 12. Glutathione - National Cancer Institute (NCI) neurotoxicity rating 2-4 at treatment end

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Chronic NCI ≥ grade 2387Risk Ratio (M-H, Random, 95% CI)0.29 [0.10, 0.85]

 
Comparison 13. Glutathione - other outcomes

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Oliguria154Risk Ratio (M-H, Fixed, 95% CI)0.48 [0.28, 0.81]

 
Comparison 14. Org 2766 - qualitative vibration position testing (VPT)

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Abnormal VPT at 3 to 5 months120Risk Ratio (M-H, Fixed, 95% CI)0.67 [0.31, 1.43]

 
Comparison 15. Org 2766 (1 or 2 mg)- vibration perception testing (VPT) finger or hand

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 VPT hand 3 to 5 months post treatment3175Mean Difference (IV, Random, 95% CI)-1.77 [-4.78, 1.23]

 
Comparison 16. Oxcarbazepine - sensory nerve action potential (SNAP) amplitude

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 SNAP amplitude: sural132Mean Difference (IV, Fixed, 95% CI)3.20 [-1.39, 7.79]

 2 SNAP amplitude: superficial peroneal132Mean Difference (IV, Fixed, 95% CI)2.00 [-1.19, 5.19]

 3 SNAP amplitude: ulnar132Mean Difference (IV, Fixed, 95% CI)1.20 [-1.78, 4.18]

 
Comparison 17. Oxcarbazepine - neurotoxicity rating

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Neuropathy after 12 cycles132Risk Ratio (M-H, Fixed, 95% CI)0.42 [0.19, 0.91]

 2 Severity of neuropathy (TNS)140Mean Difference (IV, Fixed, 95% CI)-7.1 [-11.98, -2.22]

 
Comparison 18. Retinoic acid - National Cancer Institute (NCI) neurotoxicity rating

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 NCI-CTC grade ≥ 2192Risk Ratio (M-H, Fixed, 95% CI)0.75 [0.55, 1.02]

 
Comparison 19. Vitamin E - qualitative sensory nerve conduction study (NCS) amplitudes

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Abnormal median or sural sensory amplitude127Risk Ratio (M-H, Fixed, 95% CI)0.39 [0.17, 0.93]

 
Comparison 20. Vitamin E - sensory nerve action potential amplitude

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Superficial peroneal amp 3 mos s/p treatment130Mean Difference (IV, Fixed, 95% CI)3.60 [-0.35, 7.55]

 2 Ulnar SNAP amp 3 mos s/p treatment130Mean Difference (IV, Fixed, 95% CI)1.50 [-2.71, 5.71]

 3 Sural amplitude uv after 6 cycles257Mean Difference (IV, Fixed, 95% CI)2.01 [-1.60, 5.61]

 
Comparison 21. Vitamin E - incidence of neuropathy

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Incidence of peripheral neuropathy (completed trial)130Risk Ratio (M-H, Fixed, 95% CI)0.31 [0.11, 0.90]

 2 Incidence of peripheral neuropathy (intention to treat)135Risk Ratio (M-H, Fixed, 95% CI)0.46 [0.21, 1.00]

 
Comparison 22. Vitamin E - modified peripheral neuropathy (PNP) score

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Modified peripheral neuropathy (PNP) score130Mean Difference (IV, Fixed, 95% CI)-5.48 [-10.73, -0.23]

 
Comparison 23. Vitamin E - clinical impairment

Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size

 1 Total neuropathy score after 6 cycles262Risk Ratio (M-H, Fixed, 95% CI)0.41 [0.23, 0.73]