IntroductionBackground and rationale

This trial will evaluate the benefit of combining the anticonvulsant, pregabalin (PGB), with exogenous administration of the pineal hormone, melatonin (MLT), to treat chronic pain in fibromyalgia and its associated sleep disturbance. Fibromyalgia is a complex disorder characterised by a constellation of symptoms including chronic widespread pain, sleep disturbance, fatigue, irritable bowel syndrome, depressed mood, cognitive dysfunction, functional disability and impaired quality of life.1 2 Currently recommended interventions for the management of fibromyalgia include exercise, psychological therapies and medications such as antidepressants and anticonvulsants, including PGB.3Combination therapy Pain management involves various psychological, physical, pharmacological and invasive interventions.4 5 Even though drug therapy remains an important strategy, current drugs reduce pain, on average, by only 25–40%, with meaningful relief in only 40–60% of patients, in part due to incomplete efficacy and dose-limiting adverse effects (AEs).6–9 In addressing these limitations, evidence suggests that combining mechanistically different analgesic drugs may improve pain relief and other outcomes.10–14 In fact, treatment pattern studies suggest that approximately half of patients concurrently receive two or more analgesics.15 16 Currently, drug combinations are most often being prescribed in the absence of supportive evidence and several consensus reports and reviews emphasise the need for more research on combination drug therapy.17 18 Thus, evaluating the safety and efficacy of potentially beneficial analgesic combinations has been identified as an important gap highly relevant to improving patient care.

Pain and sleep

Chronic pain, including fibromyalgia1 2 and neuropathic pain,19 20 is frequently associated with sleep disturbance. In fibromyalgia patients, evidence indicates that a night of poorer sleep is followed by a more painful day, and a more painful day is followed by a night of poorer sleep.21 Furthermore, survey data from persons with lived experience – particularly those suffering from fibromyalgia – demonstrate that ratings for sleep quality (ie, falling asleep and staying asleep) are among the highest of the 20 outcome domains.22 Thus, interactions between pain and sleep are such that effective fibromyalgia treatment requires attention to both pain and sleep.23–25 So, in addition to treating one of the most prominent secondary features of chronic pain, pain therapies that also improve sleep are expected to have more favourable efficacy26–29 and, furthermore, combining two interventions that reduce pain and improve sleep through independent mechanisms is expected to have superior efficacy over monotherapy.

Aims and hypotheses

Based on emerging evidence, we hypothesise that a combination of the anticonvulsant drug, PGB, and the pineal hormone, MLT, has superior efficacy vs either monotherapy for fibromyalgia. PGB is the only anticonvulsant approved by Health Canada for fibromyalgia30 and has been shown to reduce pain in multiple neuropathic pain and fibromyalgia RCTs.31 Multiple PGB RCTs show reduced sleep disturbance via a direct positive effect on sleep maintenance (ie, staying asleep).26 The pineal hormone, MLT, has been shown to reduce pain in both preclinical and clinical studies (including four fibromyalgia trials,27 29 32 33 in addition to multiple RCTs demonstrating efficacy for primary insomnia and delayed sleep phase syndrome (ie, falling asleep).34 Thus, we hypothesise that a PGB–MLT combination has superior efficacy versus either monotherapy, because of (1) evidence of favourable analgesic interactions between these two classes of agents—including differing mechanisms of pain reduction and sleep improvement; (2) RCT evidence of superior efficacy of other PGB-containing, and MLT-containing, combinations; and (3) compounded benefits of concurrently reducing both pain and sleep disturbance.

Our primary objective is to conduct a double-blind, randomised, double-dummy trial comparing a MLT–PGB combination to each monotherapy for fibromyalgia. During each of 3 treatment periods, with a double-blind, randomised crossover design, study medication doses will be increased to maximal tolerated dose. The primary trial outcome is pain and other outcomes include sleep, physical function, mood, and AEs.

Methods: design, participants, interventions and outcomesTrial design

The proposed trial has been designed according to IMMPACT and CONSORT recommendations and involves a double-blind, randomised, double-dummy, three-period crossover design to compare melatonin, PGB, and a MLT–PGB combination in participants with fibromyalgia (figure 1).

Figure 1

Study design for the clinical trial of a melatonin-pregabalin (MLT–PGB) combination for fibromyalgia. Sequence 1, combination-PGB-MLT; Sequence 2, MLT-combination-PGB; Sequence 3, MLT-PGB-combination; Sequence 4, PGB-MLT-combination; Sequence 5, PGB-combination-MLT; Sequence 6, combination-MLT-PGB; BL, baseline; MTD, maximal tolerated dose; T-WO, taper/washout.

Patient and public involvement

Chris DeBow, our patient partner, provided critical input to the first draft leading to the final manuscript version.

Eligibility: inclusion criteria

Target population: Participants ≥18 years and older meeting the 2016 American College of Rheumatology diagnostic criteria for fibromyalgia2 will be considered. Eligible participants will have daily moderate pain (≥3/10) for at least 3 months, a normal liver function test (ALT no greater than 20% above normal), and creatinine clearance >50 mL/min. Participants will require sufficient cognitive function and language comprehension skills – regardless of their native language – for questionnaire completion and communication with trial personnel (and translator, if necessary). Women of childbearing potential will be required to receive a highly effective form of contraception.

Eligibility: exclusion criteria

Prospective participants with any major organ system disease, psychiatric, suicidal ideation or substance abuse disorder, hypersensitivity to any of the study medications, or other painful condition as severe as their fibromyalgia pain – as well as shift workers who work during typical sleeping hours – will be excluded. Prospective participants with any other abnormalities or conditions that, in the judgement of the investigators would interfere with the protocol, will be excluded.

Study setting

This will be a single-centre outpatient trial conducted at an academic healthcare institution: Providence Care Hospital, Queen’s University, Kingston, Ontario, Canada. The planned start date for this trial is 1 December 2024 and the planned end date is 30 November 2028.

Randomisation and blinding

This will be a randomised crossover trial, using a double-dummy, balanced Latin Square crossover design in which participants will be allocated to one of six sequences of the three treatments: PGB, MLT, and PGB+MLT (figure 1). At the beginning of the trial, a pharmacist at the Kingston Health Sciences Centre will prepare a concealed allocation schedule randomly assigning the six sequences, in variable block sizes of 12–16, to a consecutive series of numbers. On enrollment, each participant will be assigned to the next consecutive number, and the corresponding series of study medications will be dispensed.

All study medications will be encapsulated, in gelatin capsules with drug-specific colours, in an identical fashion. Treatment codes for each participant will be generated by the pharmacist and will not be disclosed to study personnel or participants until completion of all participants in the trial. If needed in case of emergency, individual treatment codes will be disclosed by an investigational pharmacist to a non-study healthcare practitioner. Outcome measures will be evaluated and recorded by trial personnel who will be blinded. As an assessment of blinding to the treatment group, each participant and the study coordinator will complete a blinding questionnaire at the end of treatment.

InterventionsPGB

PGB is an anticonvulsant, which blocks the α−2-δ subunit of N-type voltage-gated calcium channels, resulting in reduced release of excitatory neurotransmitters.35–38 PGB has been approved by Health Canada, and recommended as a first-line treatment, for neuropathic pain39 40 and fibromyalgia.30 41 42 Safety and efficacy of PGB have been evaluated in dozens of high-quality RCTs31 including studies from our group.43–46 A recent meta-analysis of PGB reported on 19 RCTs involving 7003 participants with postherpetic neuralgia, DPN, central neuropathic pain, and fibromyalgia.31 Relevant to this proposal, review of sleep outcomes in PGB trials support a positive effect of PGB on disturbed sleep through a direct positive effect on sleep maintenance.26 Side effects associated with 600 mg/day of PGB from these studies included somnolence in 15–25% and dizziness in 27–46% and led to treatment discontinuation in 18–28%.31

MLT

MLT (N-acetyl-5-methoxytryptamine), a hormone secreted by the pineal gland, has been implicated in several homeostatic functions including sleep, modulation of circadian rhythms and mood, enhanced immunity, and antioxidant free radical scavenging.47 Relevant to pain, accumulating preclinical as well as clinical evidence suggests antinociceptive effects of MLT by acting at MT1 and MT2 membrane receptors (to reduce cyclic AMP) but possibly also by (1) indirectly activating opioid receptors, (2) inhibiting production of pro-inflammatory cytokines, (3) activating GABAA receptors and (4) through its antioxidant effects.48 Administration of exogenous sources of MLT has been approved by Health Canada as a natural health product for the improvement of sleep and the prevention and/or reduction of jet lag.49 Supporting this, a recent meta-analysis identified 19 studies (1683 participants) of MLT to treat primary adult sleep disorders and reported efficacy for primary insomnia and delayed sleep phase syndrome.34 Relevant to pain, a recent review of MLT in fibromyalgia reported a ‘positive effect’ in all studies; however, these were of mixed quality and evidence was limited to only 4 studies involving 98 participants.32

PGB–MLT combination

Hypothesised superiority of a PGB-MLT combination is based on RCT data indicating that combinations of two mechanistically different treatments show superior efficacy vs either single agent10–14 43 50–52 and also based on relevant evidence suggesting favourable antinociceptive interactions between these drugs.53–55 Review of sleep outcomes, from RCTs of PGB for pain treatment, has revealed a positive benefit on disturbed sleep through a direct effect on sleep maintenance (ie, staying asleep). Exogenous administration of the pineal hormone, MLT, has been reported to reduce pain in both preclinical and clinical settings (including four fibromyalgia trials), in addition to other RCT evidence of efficacy for primary insomnia and delayed sleep phase syndrome (ie, falling asleep). Thus, we hypothesise that a PGB-MLT combination has superior efficacy versus monotherapy because of: (1) specific evidence of favourable interactions between these 2 classes of agents – including different mechanisms of pain reduction and sleep improvement; (2) RCT evidence of superior efficacy of other PGB-containing and MLT-containing combinations29 33; and (3) expected substantial benefits of concurrently targeting both pain and sleep. Of note, a recent systematic review published by our group,10 found no clinical trials (nor any publications since then) comparing the combination of PGB (or gabapentin) plus MLT to either monotherapy for chronic widespread pain pointing to this important knowledge gap.

Pharmacokinetic considerations

Characterising the effect of each drug on the pharmacokinetics of the other is vital to properly understand the utility of a combination given the potential for both pharmacodynamic and pharmacokinetic interactions.56 57

PGB

After oral administration, PGB is almost completely absorbed exhibiting linear pharmacokinetics and no metabolites have been described.58 The excretion of PGB is through the urine. After repeated dosing, the peak steady-state is reached between 24 and 48 hours later and the protein binding is practically null. Thus, the potential for drug–drug interaction is minimal. PGB required some dose adjustment (lower dose) in patients with renal impairment.58 59 As with other sedating drugs, PGB can potentiate the effects of other sedatives such as benzodiazepines.60

MLT

After oral administration, MLT undergoes first-pass hepatic metabolism,111 with subsequent secondary renal metabolism and elimination.61 62 Review of studies involving varying doses indicates that oral administration of MLT results in a Tmax (time to max concentration) ranging from 15 to 210 min, and a half-life ranging from 28 to 126 min.63 A small interaction study reported that caffeine increased MLT bioavailability,64 while another study reported that cigarette smoking reduced it.65 Other small interaction studies suggested that MLT bioavailability was increased by coadministration with oral contraceptives66 and the antidepressant, fluvoxamine.67

PGB–MLT combination

Based on the above information about each agent, we expect no clinically significant pharmacokinetic interactions between PGB and MLT. However, we have been unable to find any published formal pharmacokinetic or pharmacodynamic studies evaluating interactions between melatonin and pregabalin (or gabapentin).

Study treatment dosing schedule

Study participants will be randomised, in a double-blind fashion, to one of six possible sequences (eg, sequence 1: combination>MLT>PGB) such that each participant progresses through each of three 6 week treatment periods. Treatment periods will conclude with a 7 day dose taper and a 4 day complete washout. During each period, participants will receive two sets of capsules: (1) ‘MLT’ capsules – may contain MLT 3 mg or inert matching placebo and (2) ‘PGB’ capsules—may contain pregabalin 75 mg or inert matching placebo. During the ‘MLT-PGB combination’ period, set one will contain MLT (3 mg capsules) and set two will contain PGB (75 mg capsules). During the ‘MLT alone’ period, set one will contain MLT (3 mg capsules) and set two will contain inert matching placebo. During the ‘PGB alone’ period, set one will contain inert placebo and set two will contain PGB (75 mg capsules). MLT (and MLT placebo) study drug administration will occur only in the evenings, whereas PGB (and PGB placebo) study drug administration will occur according to twice daily dosing (morning and evening).

Proposed titrations and dose ceilings are similar to those used in previous RCTs of MLT and of PGB and will be identical in all treatment periods to maintain double-blinded conditions. Our experience and success with dose titration of analgesic drugs in eight trials is extensive and our practice of regular participant follow-up by the study coordinator allows participants to safely reach their maximal tolerated dose. The first 25 days of the treatment period will involve escalation towards the schedule’s ceiling dose, or the individual participant’s MTD-maximal tolerated dose. Dose titration will stop when participants are unable or unwilling to tolerate the related AEs or on achieving complete pain relief. The 1 week from days 25 to 31 will involve this fixed MTD of the study drugs. For the ‘MLT/PGB combination’, and ‘MLT alone’ treatments, the highest possible ceiling dose of MLT will be 12 mg/day. For the ‘PGB alone’ and ‘MLT/PGB combination’ treatments, the highest possible ceiling dose of PGB will be 450 mg/day. Following the fixed-dose phase at MTD there will be a 7 day dose taper. A study coordinator will contact participants by telephone at least once a week to evaluate AEs, guide study drug titration, and encourage compliance. Study participants will be encouraged to contact a study physician, as needed, who will be available 24 hours a day. With each dosage increase of both study medications, if mild to moderate treatment-emergent AEs (eg, sedation, dizziness) are encountered, participants will be asked if they can tolerate continuing at that dose for another 2–3 days. If so, this dosage will be continued with the expectation that tolerance to side effects will occur. If side effects are severe, intolerable or do not improve, both study medications will be decreased to the next lowest possible dose and an increase will be attempted one more time at the next scheduled dose increase. If this again results in intolerable side effects, both study drugs will be decreased back to the previous dose, which will be defined as the MTD for that individual.

Cointerventions and rescue medication

Participants will be weaned off any pre-trial MLT, PGB and/or gabapentin over at least 7 days. Participants perceiving benefit from stable pre-trial doses of opioids (<90 mg morphine equivalents), antidepressant drugs, nonsteroidal anti-inflammatory agents or acetaminophen may continue these at a steady dose for the duration of the study. As per a crossover design, concomitant medications are unlikely to bias trial results because they will be equally represented across all three treatments. However, participants will be required to avoid any longer lasting procedural pain therapies (eg, nerve blocks, drug infusions) during the entire study because they may be unevenly distributed in the trial and could skew the results. If participants received any such therapies, this would be considered a trial violation, and the participant will be withdrawn from the trial.

During taper and washout periods, participants may take acetaminophen, ≤2.6 g/day, for inadequate pain relief. This rescue medication will be taken after the MTD phase and thus not affect the primary outcome measure of pain during the MTD phase of each treatment.

Outcome measures

The primary outcome will be the mean of daily ‘average’ pain intensity ratings from the last 7 days at MTD, of each treatment period. Pain intensity is self-rated each morning as ‘average pain over the last 24 hours’ using a 0–10 Likert numerical rating scale with the anchors: 0 = ‘no pain’; 10 = ‘worst pain imaginable’.68 Following informed consent, participants receive study teaching on how to consistently rate their pain intensity, twice daily, as per the daily pain diary. Our process of regular monitoring and encouragement of complete daily diary completion during weekly follow-up calls to participants has ensured excellent compliance and data completeness in our seven previous and recent trials. Secondary outcomes include change from treatment period baseline to MTD week, Fibromyalgia Impact Questionnaire,69 MOS-Sleep Scale,70 global pain relief, Brief Pain Inventory,71 Beck Depression Inventory-II,72 Beck Anxiety Inventory,73 the short-form McGill Pain Questionnaire,74 the SF-36 Quality of Life Survey,75 MTDs of MLT and PGB, frequency/severity of other treatment-emergent AEs, blinding questionnaires, acetaminophen consumption and study drug pill counts. Participant safety will be monitored through vigilant and judicious drug titration. Any occurrences of major adverse events will be tracked as secondary outcomes and also reported to the Queen’s University Research Ethics Board and Health Canada.

Sample size

Statistical considerations underlying sample size calculations are mainly based on the null hypothesis (H0) that there is no difference in pain intensity during MLT-PGB combination treatment as compared with treatment with either single agent. Based on data from our previous trial in fibromyalgia pain,43 we estimated an intrasubject SD of 1.5 for the difference in average pain at MTD between the two treatment groups. In a previous trial, a mean difference of 1.3 in average pain at MTD was observed between PGB and duloxetine–PGB combination, which is equivalent to approximately 65% of a clinically important reduction in pain.76 To detect this difference for any of six pairwise comparisons between the treatment groups in this trial with 80% power and at an overall alpha level of 0.05, we calculated that a sample of 54 participants is required based on a formula in Qu and Zheng (2003)77 for the power of a high-order cross-over design, which is revised for the testing of a superiority hypothesis and with a Bonferroni adjustment for multiple comparisons. Our recent and current experiences suggest that approximately 8% of enrolled participants (with chronic pain) per treatment period drop out of similar clinical trials. Taking three treatment periods into consideration, we would expect the enrollment of 70 participants to yield approximately 54 study completers.

Statistical analysis

Statistical analyses for this RCT will be conducted by one of this proposal’s co-applicants, Dr. Dongsheng Tu. Considerations underlying the proposed analyses for this crossover trial are based on the null hypothesis (H0) that there is no difference in pain intensity during MLT–PGB combination treatment as compared with treatment with either single-agent. Participants who have completed at least two of the three periods (ie, for at least one pairwise comparison) will be included in an intent-to-treat analysis. When data from only one period are available, sensitivity analysis including all patients will also be performed by assuming some reasonable but extreme values for the remaining periods. Those receiving at least one dose of study drug will be included in the safety analysis.

Primary outcome analysis

The primary outcome—mean daily ‘average’ pain from the last 7 days (at MTD) of each treatment period—will be calculated as an average of pain scores as recorded in the pain dairy within the period of the last 7 days if more than 50% of the information (ie, at least 4 days) is not missing. Otherwise, the mean daily pain will be treated as missing. This is based on the often used ‘half rule’ used to summarise repeated responses, which has proven unlikely to introduce bias to trial results.78 Sensitivity analyses based on the average of all available pain scores will also be performed to confirm the results of the primary analysis. Although carryover effects are unlikely, we recognise this possibility. Therefore, a linear mixed model with sequence, period, treatment, the first-order carryover and baseline pain score as fixed effects and participant as random effect79 will be used to test whether there is any treatment difference among the three treatments and to estimate the least square mean of the mean daily pain intensity for each treatment, adjusting the carryover as well as period effects (ie, stability of pain levels). If the carryover effect is not significant, a reduced model excluding the carryover term will be refitted. If a significant overall treatment effect is found from the linear mixed (full or reduced) model analysis, the following pair-wise comparisons will be performed based on the least square means and SD from the linear mixed model: combination versus MLT alone or PGB alone, and MLT alone versus PGB alone. Sensitivity analyses will be performed using a pattern-mixture model based on patterns of missing data to check the robustness of results in the case that data may not be missing at random. A Fisher’s least significant difference procedure80 will be used to adjust the p values for these comparisons.

Secondary outcome analyses

The secondary outcomes will be analysed similarly except that (1) there is no need to average the scores over the last 7 days since there is only one measurement in the last week for the singular measures (ie, final week questionnaires) and (2) the scoring algorithms developed for the Brief Pain Inventory, the Beck Depression and Anxiety Inventories. The SF-36 and the Fibromyalgia Impact Questionnaire will be first used to derive the subscales or domains within these instruments, and the scores on these subscales or domains will be used as response variables in the linear mixed model analysis. Study data will be analysed once the study participants reached 54. No interim analysis is planned. Given the potential for sex/gender differences in treatment response, we will conduct sex/gender-based subgroup analyses for this trial. The main goal is to compare treatment-induced changes from baseline in pain, for women to men, while controlling for age and other factors. The analytical approach will be the same as that described above. The treatment by sex/gender interaction terms will be included in the nonparametric analysis of covariance model of Akritas et al, which may also include age and other factors, to formally assess the potential differential treatment effects in the sex/gender groups.81 We will also conduct subgroup analyses to evaluate differences in efficacy results according to specific concomitant medications taken by >20% of participants.

Ethics and dissemination

This trial has been registered with the International Standard Randomised Controlled Trial Number Registry (ISRCTN #18278231), has been granted ethical approval by the Queen’s University Health Sciences Research Ethics Board (Queen’s HSREB Protocol #6040998) and is currently under review for a Clinical Trial Application to Health Canada Natural and Non-prescription Health Products Directorate. All participants will provide written informed consent prior to trial participation. Following trial completion, results will be disseminated in one or more biomedical journal publications and presented at one or more scientific meetings.