Data management plan
Case report forms (CRFs) for each subject screened and enrolled in this study will be completed directly in the Research Electronic Data Capture (REDCap) database to the extent possible without the use of a paper CRF. REDCap is hosted by Aarhus University. Source documents include medical records, pain diaries, and CRFs (paper or REDCap (eCRF)). The study personnel at each site will be trained in the study procedures. All investigators will have access to the final trial data set. After the study, anonymized data will be available in a data repository and will be available upon request subject to written agreement with a department head.
Statistics
With a minimally relevant between-phenotype group difference in total pain reduction of 1.25 NRS points, a standard deviation of 1.6 within phenotype groups [
5], 80% power, and a 5% risk of type I error, the sample size estimate is 27 + 27 patients for the primary objective. With a minimally relevant treatment versus placebo difference in total pain reduction of 1.5 NRS points, 80% power, and a 5% risk of type I error, the sample size estimate is 30 + 15 patients using a treatment:placebo ratio of 2:1 for the supportive objective. With an expected dropout rate of 1/6 (data available for intention-to-treat (ITT) analysis), the recruitment stops when 54 patients in each phenotype group have been randomized to lacosamide and placebo in a 2:1 ratio. Thus, we expect to randomize 108 patients. For the explorative outcome in case there is no phenotype difference, 72 patients randomized to lacosamide and 36 patients randomized to placebo in the total population in an ITT population will give > 90% power to find a minimally relevant treatment versus placebo difference in total pain reduction of 1.25 NRS points and 85% power to find a minimally relevant treatment versus placebo difference in total pain reduction of 1.0 NRS points.
Statistical analysis of the primary outcome will be performed by
t test and of the secondary outcomes by the Mann–Whitney
U test. Nondichotomous tertiary outcomes will be performed by
t test or Mann–Whitney
U test, where applicable. Since we do not expect differences in baseline between the two phenotypes [
5], a major impact of baseline pain intensity on the outcome, or a center effect, we do not plan to include these as covariates in the analyses.
For the primary outcome, the delta values from the average pain intensity in the baseline week to the last treatment week (last 7 days) will be used. Response rates and other dichotomous data are analyzed using Fisher’s exact test. For the primary (and supportive) objective, we are interested in the mechanistic aspects and in understanding whether the sensory phenotype “irritable nociceptor” is a prognostic biomarker for a sodium channel blocker in therapeutic doses. Therefore, the primary analysis for the primary objective is the PP population. Missing data will not be replaced. The PP population comprises those patients who complete at least 2 weeks on stable medication with at least 100 mg bid. Thus, if patients who fulfil the PP definition stop the treatment before the 12th week, the last seven pain scores on stable medication will be used for the primary analysis, and they will be invited for an additional visit identical to visit 3. All patients who have taken at least one study capsule will be encouraged to stay in the study, complete the diary, and come for a visit after 12 weeks.
As a secondary analysis, the ITT population will be used. The ITT analysis will also be used for the explorative outcome in the whole population if there is no phenotype difference. Given a substantial dose-dependent withdrawal rate due to adverse events, the EMA suggests a conservative responder analysis and that noncompleters are defined as nonresponders [
19]. Therefore, the baseline observation carried forward (BOCF) in the ITT population (all patients randomized) will be the primary imputation method, and the last observation carried forward (LOCF) will be done as a secondary imputation method. Patients will be asked to complete the pain diaries despite their withdrawal from trial medication to minimize the need for imputation.
Significance is considered at the 5% level. If there are changes to the original statistical plan, the type of change and the date of change will be documented, and the document will be signed by the sponsor.
Safety
Patients will record any adverse events in the pain diary and will be interviewed at each telephone call and study visit with open questions. The type of event, times of onset and termination, severity, and relationship with the treatment drug will be recorded.
Publication
Regardless of the outcome, the results (including positive, negative, and inconclusive results) of the trial will be published in a recognized international journal. ICMJE guidelines for authorship will be followed.