Combination Analgesic Development for Enhanced Clinical Efficacy (CADENCE Trial): Study Protocol for a Double-Blind, Randomized, Placebo-Controlled Crossover Trial of an Alpha-Lipoic Acid – Pregabalin Combination for the Treatment of Fibromyalgia Pain

Background Fibromyalgia is a clinical disorder commonly presenting with chronic widespread pain as well as sleep disturbance, fatigue, depression, and cognitive dysfunction. There is an urgent need for treatment strategies that provide better pain relief and fewer adverse effects (AEs). Efforts to develop rational combinations of specific fibromyalgia treatments have demonstrated potential for measurable improvements in pain relief, quality of life, and health care utilization. More than half of fibromyalgia patients receive 2 or more analgesics but current combination use is based on limited evidence. As an early proof-of-concept project from the Canadian Institutes of Health Research–Strategy on Patient-Oriented Research Chronic Pain Network, this trial protocol is expected to advance the field by rigorously evaluating a new treatment combination for fibromyalgia. Objective We will test the hypothesis that analgesic combinations containing at least one nonsedating agent would be as safe but more effective than either monotherapy because of additive pain relief without increasing overall AEs. Pregabalin (PGB), a sedating anticonvulsant, is proven effective for fibromyalgia, and the antioxidant, alpha-lipoic acid (ALA), one of the only nonsedating systemic agents proven effective for neuropathic pain, is currently being evaluated in fibromyalgia. Thus, we will conduct a clinical trial to compare a PGB+ALA combination to each monotherapy for fibromyalgia. Methods Using a double-blind, double-dummy, crossover design, 54 adults with fibromyalgia will be randomly allocated to 1 of 6 sequences of treatment with PGB, ALA, and PGB+ALA combination. During each of 3 different treatment periods, participants will take 2 sets of capsules containing (1) ALA (or placebo) and (2) PGB (or placebo) for 31 days, followed by an 11-day taper/washout period. The primary outcome will be mean daily pain intensity (0 to 10 scale) at maximal tolerated doses (MTDs) during each period. Secondary outcomes, assessed at MTD, will include global improvement, adverse events, mood, and quality of life. Results This trial attained ethics approval March 6, 2017 (Queen’s University Health Sciences and Affiliated Teaching Hospitals Research Ethics Board protocol number ANAE-313-17), and recruitment is set to start in August 2017. Conclusions This trial will provide rigorous evidence comparing the efficacy of a PGB-ALA combination to PGB alone and ALA alone in the treatment of fibromyalgia. Trial Registration International Standard Randomized Controlled Trial Number ISRCTN14939460; https://www.isrctn.com/ ISRCTN1493946 (Archived by WebCite at http://www.webcitation.org/6sFqAjxkt)


5.
This is the third submission of a previous proposal.

Research Approach:
andadverseeffects associated with thecombinationofthesedating anticonvulsantdrug,pregabalin(PGB),and the non-sedatingantioxidantagent,alpha-lipoicacid (ALA) versus each agent alone for the treatment of chronic . neuropathic pain The primary outcome will be daily pain from the maximal tolerable dose (day 39-45) for each treatment period using a numerical rating scale. The applicants propose a randomized double-blindtrial,using a double-dummy,balanced Latin Square crossover design in which patients will be allocated to one of 6 treatment sequences of the three treatments: ALA, PGBand ALA-PGBcombination.
induration weeks for a total of 24 per patient. The applicants propose a sample size of 84 patients to allow for the cross-over design and loss to follow up. They project that the trial will take 34 months to complete.
The rationale for this study is that chronic pain is very common and has significant social and economic impact and 50% of patients with chronic neuropathic pain use combination therapies that are not well studied or have significant side effects. The investigators postulate that using combination therapy of a non-sedating agent along with a sedating agent will maximize efficacy while minimizing unwanted side effects. ALA is the only non-sedating drug shown to be effective in neuropathic pain. PGB has been shown to be very effective in neuropathic pain but is quite sedating at the higher more effective doses. The applicants provide a compelling argument as to why this specific combination needs to be studied. Concerns had been previously raised as to why a placebo group was not used but this reviewer agrees with the applicants that a placebo group would not be ethical for a combination therapy pain trial. The use of a cross-over trial versus a parallel design trial is justified by the postulate that the specific mechanism for pain in individual patients differs and that therefore it would make more physiological sense for each patient to serve as their own control. In addition, this design would obviously decrease the needed sample size. There has been significant discussion as to whether a minimal clinically important difference of 1 is justified. Based on the rationale provided by the applicants and the pain literature, this reviewer feels that it is. It was previously raised that there had not been enough background work to justify this particular The objectiveofthisapplication istocompare the efficacy overall 3-period,active treatment-controlled Each of the three treatment periods will be 8 weeks study, however, I believe that between the available literature and the collateral work in this general area, this study is justified. A minor concern is the choice to use an 8 week treatment period instead of a 12 week one. While the applicants explain that 8 weeks is a compromise between the time needed to show a difference versus extending the length of the trial for participants, I still feel that given the literature supporting a 12 week trial period per regimen that the applicants should seriously consider changing this.

Originality of proposal
This particular combination of therapies (and mechanisms of action) has not been studied before and has the potential to be very effective.

Applicant
The PI is a professor in the Departments of Anesthesia and Biomedical and Molecular Sciences at the University of Toronto and is the current recipient of the SEAMO Clinician Scientist Development Program Research Salary Award and a Mid-Career Investigator Award Research award from CIHR. He currently holds 9 grants and is the PI on 6 all of which are related to pain research. He has held an additional 5 grants in the past 5 years. He has supervised a large number of post docs and students and has received both funding and awards for his mentorship. He has given a large number of invited talks throughout the world on pain management and is clearly considered an expert in the field.
The co-applicants on this proposal have expertise in psychology, psychiatry, epidemiology and biostatistics and neurosciences. This team has worked together on other projects in the past and clearly has the expertise to conduct this study.

Environment for research
The applicant clearly has an excellent environment for research as demonstrated by his ability to perform and

Impact of the research
Chronic pain is a huge problem. However, this study focuses on one specific subgroup of this population and one specific combination of therapy. The intervention has the potential to significantly impact the lives of those with chronic neuropathic pain but the broader societal impact is not clear.

Budget
Although the cost per patient is rather high, the budget is well justified and the overall cost of the project is very reasonable for the amount of work required. The team is very strong with a solid record. It has the ability to complete the trial as proposed. The area of research is important and the possible impact on management is substantial. It is difficult to produce high quality trials in chronic pain management. Inclusion criteria are clear. Treatment allocation and study reporting is clear. Excellent measurement of recruitment period base on population and previous experience. Thus trial is possible within timeframe outlined. Weakness: The trial period of 8 weeks is not sufficient. Although important for maintaining compliance (as the applicants appropriately describe), it remains insufficient. There is no justification that the results of a standard 12 trial is equivalent to 8 weeks. The applicants are looking for statistical significance, however clinical significance will change practice patterns. A change of 0.5-1 on a VAS is not sufficient to affect clinical change.
How will the maximal tolerated dose be evaluated? How will the study coordinator determine that this threshold has been reached and that the side effects are medication related? There are many outcome measures reported. There is no information on how they will impact the reporting of the results or outcomes. The sample size is not based on clinically important changes in pain but on statistically significant measures. The sample size is also based on "there is at least one pair of study treatments that are different". This outcome is different than the primary research question. It is unclear how the researchers will account for the participants who are noncompliant o do not fill out their diaries (to more than 50%), How will a sensitivity analysis be interpreted. This is a three period cross-over trial lasting 24 weeks (with twice-weekly contact with patients) looking at ALA, pregabalin and their combination. Interest is in all three pairwise comparisons. While cross-over trials are difficult to conduct and patient dropout can compromise validity, the study team has successfully completed cross-over trials before, which have been published in high-impact journals (e.g. NEJM, Lancet). The investigators position it as a proof-of-concept trial (in places referred to as a phase IIb trial), but its influence on clinical practice is also discussed. It is not entirely clear whether the investigators believe that this trial has the potential to produce evidence that warrants a change in practice without further research, or whether it is a precursor to a more definitive trial.

Major concerns
My main concern is dropout. The investigators estimating 30% dropout based on previous experience (81 patients to end up with 54 patients). This is likely realistic, but is also high and raises questions about generalizability of the results. There is some inconsistency in describing how missing data will be handled. The analysis plan states that "All patients will be included in the analyses. When data from only one period are available, sensitivity analysis including all patients will also be performed…" Which statement is correct? Since analysis will use a linear mixed model, there should be no problem including all patients even without assuming extreme values for other periods. Analysis will be based on mean pain score from the last 7 days, however as different patients will have different numbers of measurements, this will result in an outcome that is heteroscedastic (means based on more measurements will have smaller variances), and hence run counter to the assumptions of the linear mixed model. Is there any reason why one cannot just use the individual pain scores for each day, for each patient, without averaging?

Minor points
Section 2.16 states that the aim is to test the null hypothesis that there is no difference in pain intensity for any of the treatments vs. the alternative that there is at least one pair of treatments that are different. However, the analysis plan describes pairwise testing, and this is also what the sample size calculation is based on. Pairwise testing seems appropriate, but consistency in the protocol as to the main objective would be helpful. The summary cites NNTs but does not say what the outcome is, i.e. number needed to treat to achieve ? what This is not described until the main proposal. The fact that pain trials reveal that statistically significant differences vary between 0.5 to 1 points is not adequate evidence that 1 point is a clinically meaningful difference (response to reviews, proposal page 10).