PI-075 - MODEL-BASED META-ANALYSIS OF PLACEBO-RESPONSE FOR MONTHLY MIGRAINE DAYS IN MIGRAINE PREVENTION CLINICAL TRIALS FROM LITERATURE

C. Grant¹, C. Lim², J. Liu³, T. Fullerton⁴, J. Hughes⁴; ¹Mayo Clinic Graduate School of Biomedical Sciences, Rochester, MN, United States, ²Pfizer, Boston, MA, USA, ³Pfizer Inc, Groton, CT, United States, ⁴Pfizer, Groton, CT, USA.

Background: Migraines are experienced by over 1 billion people worldwide, with debilitating effects for those who suffer from them regularly. Clinical trials studying the preventive treatment of episodic or chronic migraine are often placebo-controlled, with the size of the placebo effect differing greatly between studies and over study duration. This study aimed to quantitatively describe placebo effects vs. treatment time in double-blinded migraine prevention studies using model-based meta-analysis (MBMA).

Methods: Literature monthly migraine day (MMD) data were collected following PRISMA guidelines from studies investigating therapies for prevention of migraine. Only double-blinded clinical trials published after 1995 in adult populations with episodic or chronic migraine were considered as part of the literature search. Longitudinal MBMA was performed on placebo data using a non-linear mixed effects approach. Effects of baseline MMDs, active comparator, and publication year were evaluated as covariates. Final models were used to simulate the placebo-induced change from baseline MMDs over 6 months of treatment.

Results: Of the 250 articles identified by the literature search, 213 articles were removed during abstract screening and full text review, with the main reason for exclusion being due to monthly migraine days not being reported in the study. Data used in the analyses were collated from 37 clinical trials from episodic and chronic migraine populations. The final MBMA described both the longitudinal changes in average placebo-response as well as the heterogeneity in response size (Figure 1). The average time to steady-state (95% CI) was identified as 3.7 months (3.2, 4.3). Covariate analysis identified baseline MMDs as a significant covariate, with an increase in baseline of 20% being associated with a 11.4% increase in placebo effect. For a treatment arm with 11 MMDs at baseline, the model predicted an average change from baseline MMDs at steady-state of -2.88 (-3.40, -2.36) MMDs.

Conclusion: Given the increasingly competitive landscape of anti-migraine medications for migraine prevention, this MBMA provides a valuable description of longitudinal placebo response in patients with episodic or chronic migraine for guiding development of novel migraine therapeutics.