Oral Ketamine vs Placebo in Patients With Cancer-Related Neuropathic Pain

This multicenter randomized clinical trial compares oral ketamine with placebo for treating neuropathic pain in patients with cancer.

• To compare initial treatment benefit (at day 4 of assessment period of 16 days) using the sensory component of the SF-MPQ.
• To compare difference in overall pain between the study arms based on the VAS score.
• To compare difference in neuropathic pain between the study arms based on the LANSS pain scale.
• To assess worst pain score (index neuropathic site) between the two arms.
• To compare patient distress between the two arms based on NCCN Distress Thermometer.
• To assess the side-effects and tolerability of trial drug.
• To assess the effect of intervention on quality of life scores (based on Euroqol thermometer), anxiety and depression (based on HAD scale) and opioid requirements.
• To assess the effect of intervention on sensory changes (as assessed by Quantitative Sensory Testing).
• To assess the effect of intervention on Breakthrough Cancer Pain (BTcP) using the Breakthrough Pain Questionnaire.

TRIAL DESIGN
This is a randomised double-blind trial comparing ketamine with placebo in patients with malignant neuropathic pain. Following a run-in period where opioid analgesia dose will be optimised, ketamine or placebo will be administered orally four times a day. The dose will be increased as per the titration schedule (Appendix I of protocol) and dose increments will cease when pain allows or before then if toxicity is unacceptable.
The trial will be carried out in 4 stages (run-in, titration, assessment and run-out).
The trial is stratified (using minimisation) for centre, age, baseline McGill sensory pain score, gender and type of previous adjuvant analgesic for neuropathic pain.

Study Stages
The trial will be carried out in 4 stages (see chart 1 on next page): Version 3: 4 January 2012

Chart 1 -Trial Stages
Patients with neuropathic pain and cancer. Pain score > 4/10 on VAS (0-10) Assessed for inclusion in ketamine trial.
Given patient information sheet and, if after 24 hours, agree to take part: patient consented

Run-in Period
To optimise analgesia and drug side effects

SAMPLE SIZE CALCULATIONS
The primary endpoint is improvement in malignant neuropathic (assessed using the sensory component of the SF-MPQ). A key secondary endpoint is the initial treatment benefit ("success" rate at day 4 of assessment period).
The proposed sample size of 107 patients per arm will provide at least 80% power to detect an increase in the "success" (patients who remain on study with no significant change in opioid dose and have a 5-point drop in the sensory component of the SF-MPQ.) rate at day 16 of 20% on Ketamine compared to placebo over a range of possible placebo "success rates". These figures are based on the logrank test and a 5% twosided level of statistical significance. A sample size of 107 patients per arm will also provide 80% power to detect an increase in the initial treatment benefit from 40% on placebo to 60% on ketamine at the 5% 2-sided level of statistical significance using the chi-squared test. This sample size also means that the minimum power to detect an absolute improvement of 20% in initial treatment benefit is 80% whatever the placebo "success" rate.
As patients who drop-out of the trial early (either during the titration or assessment period) will be treated as "failures" there is no need to recruit extra patients to compensate for this.

Primary efficacy parameters
Efficacy of therapy will be evaluated by comparing the treated and control groups of patients in terms of the improvement in neuropathic pain as assessed using the sensory component of the SF-MPQ.
The primary trial end-point is time to treatment "failure".
A treatment "failure" is defined as • failure to achieve a 5-point drop in the sensory component of the SF-MPQ, from the end of the run in period (prior to randomisation) to any one of the assessment time points (end of titration period, day 4, day 8, day 12, day 16) • a significant change in prescribed background opioid dose (defined as greater than 30% increase in Version 3: 4 January 2012 • a withdrawal from the trial during the titration or assessment period for any reason (e.g. lack of efficacy, side effects) Conversely a treatment "success" is a patient who remains in the trial with no significant change in prescribed background opioid dose and has a 5-point drop in the sensory component of the SF-MPQ.

Secondary efficacy parameters
a) The initial treatment benefit (at day 4 of assessment period of 16 days) using the sensory component of the SF-MPQ.
b) The difference in overall pain between the study arms based on the VAS score.
c) The difference in neuropathic pain between the study arms based on the Leeds Assessment of Neuropathic Symptoms and Signs (LANSS) pain scale.
d) The worst pain score (index neuropathic site) in the previous 24 hours (between the two arms) at study baseline and then during study assessment period.

Procedures for assessing efficacy parameters
The above tools are applied at the following time-points:- • VAS pain score -Daily throughout run in, titration and assessment period.
• LANSS -Start of run in period, end of run in period (prior to randomisation), and day 1, 4, 8, 12 and 16 of assessment period.
• Opioid use (24 hour oral morphine equianalgesic equivalent) -Start of run-in period, end of run-in period (prior to randomisation) and daily during titration and assessment periods.
• Euroqol thermometer -Quality of Life -End of run in period (prior to randomisation) and day 1, 4, 8, 12 and 16 of assessment period.
• Distress -using NCCN Distress Thermometer End of run in period (prior to randomisation) and day 1, 4, 8, 12 and 16 of assessment period.
• Breakthrough Cancer Pain -using the BTPQ End of run in period and day 16 of assessment period.
• Sensory testing -using QST End of run in period and day 16 of assessment period. Version 3: 4 January 2012

Adverse events
All adverse events (AEs) occurring during the run-in, titration, assessment and run-out stages or up to 30 days after the administration of the last trial treatment for any individual patient will be graded (CTCAE v3.0) and recorded on the relevant CRF. An AE is any untoward medical occurrence or experience in a patient that occurs following the administration of the trial medication regardless of the dose or causal relationship.

Serious adverse events and serious adverse drug reactions
A serious adverse events (SAE) is defined as an event that: − results in death − is life-threatening (the patient was at immediate risk of death at the time reaction was observed) − requires hospitalisation or prolongation of hospitalisation − results in persistent or significant disability/incapacity − is a congenital anomaly/birth defect − is considered medically significant by the investigator The assessment of causality is made by the investigator using the definitions in §6.4 of the protocol.

Definition of a serious adverse event
A serious adverse reaction (SAR) is an SAE that may be related to trial treatment. SAEs that will be considered related will include any SAE that is documented as possibly, probably or definitely related to protocol treatment.

Definition of Suspected Unexpected Serious Adverse Reactions
A suspected unexpected serious adverse reaction (SUSAR) is any serious adverse reaction that is unexpected. Unexpected is any reaction that is not a known reaction listed in the IMPD/Summary of Product Characteristics (SmPC).

DEFINITION OF POPULATIONS
As this trial has a pre-randomisation run-in period, the following analysis populations will be defined from all patients that are registered onto the study.

Intention-to-treat population
The intention-to-treat (ITT) population includes all patients randomised onto the study.

Eligible study population
The eligible study population is all patients randomised onto the study excluding patients with gross eligibility deviations at either registration or randomisation. Patients with gross eligibility deviations will be determined in consultation with the Chief Investigator.

Per protocol population
The per protocol (PP) population includes all eligible patients who, after randomisation, take trial treatment adequately as per protocol (which will be determined retrospectively in a blinded fashion). Version 3: 4 January 2012

Run-in stage population
The run-in stage population includes all patients registered onto the study, i.e. all those consenting to participate in the run-in stage.

Safety population
The safety population includes all patients who, after randomisation, start trial treatment (whether they receive ketamine or placebo).

7
TABULATIONS, FIGURES AND LISTINGS Throughout this section the tabulations consist of counts and percentages per category unless otherwise stated. Listings will include the patient identifier, study arm and the specific data for that listing.

Study Recruitment a)
A plot of actual cumulative recruitment against time since study opened will be provided. The target recruitment rate line will be superimposed on this plot. b) A table of recruitment by centre will also be provided showing the date the centre opened, the date the first and last patients were recruited and the number recruited. Centres that have opened but not recruited patients will also be incorporated in this table.

Form Return
The following tables are only provided as part of interim analysis reports.

Patient progress A CONSORT
(1) style flow diagram showing the progress of patients through the phases of the study (enrolment, run-in, intervention allocation, titration, assessment, run-out and data analysis) will be presented.

Eligibility/ population membership
As this trial involves a pre-randomisation run-in period, there are eligibility criteria for both registration and randomisation. These will be considered separately below.

Registration eligibility
All the data presentations in this section are based on the run-in population.
a) The factors considered for the inclusion and exclusion criteria on the registration form will be tabulated.
b) The waiver details of patients given a waiver to be registered onto the study will be listed, including details of whether or not the patient was subsequently randomised onto the trial.

Randomisation eligibility
a) For patients who were registered but not randomised onto the study, the reason for not being randomised will be listed. Version 3: 4 January 2012 b) The factors considered for the inclusion and exclusion criteria on the randomisation form will be tabulated.
c) The waiver details of patients given a waiver to be randomised onto the trial will be listed.

Population membership a)
Membership of the eligibility population ("eligible"/"ineligible") will be tabulated by treatment arm. b) Ineligible patients will be listed together with the reasons for ineligibility.
c) Membership of the per protocol population ("included"/ "excluded") will be tabulated by treatment arm. d) Membership of the safety population ("included"/ "excluded") will be tabulated by treatment arm. e) Patients excluded from the safety population will be listed together with the reason for not starting trial treatment and whether or not the patient was eligible.

Trial accrual
Patient accrual by centre [stratification factor] and whether the patient was randomised or did not participate beyond registration will be tabulated. A graph of the cumulative number of patients randomised against the target number will also be presented.

Run-in period
As a summary of the run-in period of the trial, the following will be tabulated: a) Worst Index Site Pain Score (VAS) at start and end of run-in period (mean, standard deviation, median inter-quartile (iq) range and range).

b) Number of days in run-in (mean, standard deviation, median inter-quartile (iq) range and range).
c) Details (name, route, prescribed dose, unit, schedule, start date and stop date (or "continuing") of adjuvant analgesic use throughout the run-in period will be listed. d) Future study participation will be tabulated. For patients not being considered for randomisation, a listing of the reasons will be produced along with a listing of specifications for "other" reasons. e) Summary of end of run-in pain score data from EuroQoL, Distress Thermometer, HADS, SF-MPQ, LANSS (mean, standard deviation, median inter-quartile (iq) range and range).
f) The number of patients completing the end of run-in period HADS, SF-MPQ and LANSS will be tabulated. Reasons for not completing these will be listed along with specifications for "other" reasons.

Baseline data at randomisation
As a summary of patient characteristics at randomisation, the following details will be tabulated by treatment The mean, standard deviation, median inter-quartile (iq) range and range will be calculated for a) and c).
As a summary of patient characteristics before starting allocated intervention, the following pre-treatment data will be tabulated by treatment arm: a) Weight (mean, standard deviation, median inter-quartile (iq) range and range). Version 3: 4 January 2012 b) Ethnicity, with specification for "other".
c) The site of the primary tumour, with specification for "other". d) Presence of metastases, with the site and specification for "other" if applicable. e) Cancer history: Prior chemotherapy, prior radiotherapy and prior hormonal therapy, with listing of details.
f) History of painful neuropathy and history of chronic pain will be tabulated: No; Yes, not problematic in the last 3 months; Yes, problematic in the last 3 months. For patients where these have been problematic in the last 3 months details will be listed. g) Whether or not the patient has a history of drug or alcohol abuse. For patients with history further details will be listed.
h) Whether or the patient is currently alcohol or drug dependent.
i) Details of current medical conditions will be listed.
k) Site, primary method of diagnosis, and association with metastatic site.
l) Worst score in the past 24 hrs for index neuropathic pain will be summarised (mean, standard deviation, median inter-quartile (iq) range and range). m) All questions on the BTPQ will be tabulated (frequencies, mean, standard deviation, median interquartile (iq) range, range -as appropriate).

Titration period
As a summary of the delivery of study therapy in the titration period of the trial, the following will be tabulated by study arm: a) Number of days in titration will be summarised (mean, standard deviation, median inter-quartile (iq) range and range). b) Dose level at end of titration.
c) The number of days for which a patient took less than 4 doses of study medication, with listing of reason.
d) The number of days for which a patient took more than 4 doses of study medication, with listing of reason.
e) The number of times a dose reduction took place for each patient, with listing of reason.
f) Whether or not the patient is continuing to the assessment period will be tabulated: Yes; No, patient request; No, investigator decision; No, toxicity; No, other. Further specification for the "No, other" category will be listed.

Assessment period
As a summary of the delivery of study therapy in the assessment period of the trial, the following will be tabulated by study arm: a) Number of days in assessment will be summarised (mean, standard deviation, median inter-quartile (iq) range and range). b) A cross tabulation of dose level at beginning of assessment and dose level at end of assessment.
c) The number of days for which a patient took less than 4 doses of study medication, with listing of reason.
d) The number of days for which a patient took more than 4 doses of study medication, with listing of reason. Version 3: 4 January 2012 e) The number of times a dose reduction took place for each patient, with listing of reason.
f) Whether or not the patient completed the 16 day assessment period. g) Whether or not the patient is continuing to the run-out period will be tabulated: Yes; No, patient request; No, investigator decision; No, toxicity; No, other. Further specification for the "No, other" category will be listed.

Run-out period
As a summary of the delivery of study therapy in the run-out period of the trial, the following will be tabulated by study arm: a) Number of days in run-out will be summarised (mean, standard deviation, median inter-quartile (iq) range and range).
b) The number of days for each patient where the actual dose taken was less than protocol dose, with listing of reason.
c) Whether or not the patient completed the run-out period.

Opioid Use
a) Number of patients taking any opioid medication will be tabulated by study period (run-in, titration, assessment) and study arm (Arm A, Arm B, Not allocated).
b) Details (name, route, prescribed dose, unit, schedule, start date and stop date (or "continuing") of opioid use will be listed by study arm and study period.
c) The normal number of breakthrough doses of opioid medication per day in the run-in period will be tabulated (mean, standard deviation, median inter-quartile (iq) range and range).
d) The proportion of days during the run-in period that a patient required breakthrough opioids will be summarised (mean, standard deviation, median inter-quartile (iq) range and range).
e) The proportion of days during the run-in period that a patient required more than 2 doses above normal breakthrough opioids will be summarised (mean, standard deviation, median inter-quartile (iq) range and range).

Unblinding
The following will be tabulated by study arm: a) Number of patients unblinded, with listing of reason.
b) System used to perform unblinding. c) Number of cases of unblinding discussed with Chief Investigator, with listing of reasons where there was no discussion.

Run-in period
Based on the run-in stage population: Opioid toxicity will be tabulated, with a listing of details (short name, worst grade observed, outcome and whether or not the event was a SAE) and a specification of "other" toxicities.

Adverse events
The data presentations in this section relate to the titration, assessment and run-out periods and are based on the safety analysis population. All adverse events will be graded according to the NCI CTCAE v3. Version 3: 4 January 2012 a) The worst grade of each listed AE will be tabulated by study arm.
b) The worst grade of "other" toxicities that graded grade 2 by at least 5% of patients will be tabulated.
c) The details (Patient identifier, study arm, short name for AE, CTC grade, relationship to study drug and whether or not a SAE) for all adverse events graded as 3/4 will be listed.

Serious adverse events/reactions/SUSARs
Details of serious adverse/events/reactions and SUSARs will be provided in a separate document output from the CTU's pharmacovigilance database.

Efficacy analyses
Primary and secondary efficacy analyses will be conducted on the ITT populations.
Box-and-whisker plots of change from baseline by study arm will also be produced for all of the variables listed above.
The number of patients at each time-point will be indicated on all box-and-whisker plots.

Primary efficacy analysis
The estimated difference in the "success" rate between the study arms at the day 16 assessment point will be tabulated along with the associated 95% confidence interval. A Kaplan-Meier plot of time to treatment failure, with 'Proportion of successes' as the y-axis, will be given. The p-value associated with the log-rank comparison of the survival curves will be given on this plot. This analysis will be performed twice: • Primary analysis: Where a significant change in opioid dose based only on the prescribed background opioid dose is used in the definition of a treatment "failure".
• Exploratory analysis: Where a significant change in opioid dose based on the prescribed background opioid dose plus recorded breakthrough (PRN) dose is used in the definition of a treatment "failure".
A secondary confirmatory analysis of the primary analysis will be based on a Cox regression model incorporating the study stratification factors as covariates. The p-value, hazard ratio and associated 95% confidence interval for study arm will be tabulated.
As part of the assessment of the primary end-point the number of patients with a significant change in opioid dose (>30% increase in 24 hour morphine equivalent daily dose during titration or assessment period) during the titration or assessment period, defining them as a treatment "failure", will be tabulated. This data will be reviewed and confirmed by the CI blinded to treatment arm.

Secondary efficacy analysis
a) The primary efficacy analysis will be repeated on the PP population.
b) The differences in the initial treatment benefit rates (using the "success" rate at the day 4 assessment point) between the study arms will be estimated and tabulated along with the associated 95% confidence interval. The p-value for the comparison will also be given.
c) The number of patients completing the pain score questionnaires (EuroQol, Distress, HADS, SF-MPQ, and LANSS) at the end of the titration period and days 4, 8, 12 and 16 of the assessment period will be tabulated. A listing of reasons for not completing the questionnaires will be given. Version 3: 4 January 2012 d) The adjusted standardised AUC (AUC calculated over the assessment/titration period, divided by number of days on assessment/titration and with the baseline value subtracted) of the following efficacy measures will be tabulated (mean, standard deviation, median inter-quartile (iq) range and range) by study arm. The difference in the median AUC will be given together with the associated 95% confidence interval and estimated p-value. All of the above will be produced for for the patients will be calculated in terms of morphine equivalent daily dose. This will be summarised by treatment arm (mean, standard deviation, median inter-quartile (iq) range and range). The estimated difference in the medians will be given together with the associated p-value and 95% confidence interval.
f) Side-effects and tolerability of trial drug -covered by §7.6.
g) The method for summarising the effect of intervention on sensory changes (as assessed by Quantitative Sensory Testing (QST)) will be defined after discussion with the Chief Investigator.
h) All questions on the BTPQ will be tabulated as a comparison of the proportion of patients improving between the arms.

Concomitant medication
Details of concomitant medication taken by patients will be listed.

Primary efficacy analysis
The primary comparison will be in terms of time to treatment "failure" (as defined in §4.1) between the study arms. This comparison will be made using the log-rank test. A 95% confidence interval for the difference in proportions at day 16 will be derived using "Method 10" as described in R.G. Newcombe (1) .

Secondary efficacy analyses
Initial treatment benefit, using the sensory component of the SF-MPQ, will be assessed by comparing the "success rates" at day 4 of assessment period between the treatment arms using the chi-square test. A 95% confidence interval for the difference in rates at day 4 will be derived using "Method 10" as described in R.G.
The analysis of VAS pain score will be complicated by missing data. Analysis of this data will use AUC techniques (3) . Missing values will be filled in using interpolation or last value carried forward as appropriate.
(Note that this is different from the multiple imputation approach stated in the protocol -the different assessment and titration period durations for the patients make multiple imputation impractical).
The adjusted standardised AUC will be compared between the arms using the Mann Whitney U-test.
Bootstrap methods will be used to estimate the median difference and associated 95% confidence interval.