Supplementary materials: Comparison of indirect treatment methods in migraine prevention to address differences in mode of administration

These are peer-reviewed supplementary materials for the article 'Comparison of indirect treatment methods in migraine prevention to address differences in mode of administration' published in the Journal of Comparative Effectiveness Research.1. MethodsSI Figure 1: Migraine evidence networks for (A) episodic migraine and (B) chronic migraineSI Figure 2: Eligibility criteria for the literature reviewSI Figure 3: PRISMA diagram for the literature reviewSI Figure 4: Observed placebo responses prior to indirect treatment comparisonsFixed effects Bayesian network meta-analysis (NMA)Random effects NMAFixed effects NMRSurvey of expert opinion2. Results: Random EffectsStandard NMA in Episodic MigraineSI Table 1.Standard NMA in Chronic MigraineSI Table 2: Model assessment for fixed and random effects for standard NMA in chronic migraine.Placebo Response Regression in Episodic MigraineSI Table 3: Model assessment for fixed and random effects for placebo response regression in episodic migraine.Placebo Response Regression in Chronic MigraineSI Table 4: Model assessment for fixed and random effects for placebo response regression in chronic migraine.SI Table 5: Random effects results for estimated differences in change from baseline in MMDs at 12 weeks in episodic migraine.SI Table 6: Random effects results for estimated differences in change from baseline in MMDs at 12 weeks in chronic migraine.Aim: Indirect treatment comparisons (ITCs) are anchored on a placebo comparator, and the placebo response may vary according to drug administration route. Migraine preventive treatment studies were used to evaluate ITCs and determine whether mode of administration influences placebo response and the overall study findings. Materials & methods: Change from baseline in monthly migraine days produced by monoclonal antibody treatments (subcutaneous, intravenous) was compared using fixed-effects Bayesian network meta-analysis (NMA), network meta-regression (NMR), and unanchored simulated treatment comparison (STC). Results: NMA and NMR provide mixed, rarely differentiated results between treatments, whereas unanchored STC strongly favors eptinezumab over other preventive treatments. Conclusion: Further investigations are needed to determine which ITC best reflects the impact of mode of administration on placebo.