OBJECTIVES: In the absence of head-to-head data between the FAc and DEX implants conduct an ITC to generate estimates of comparative efficacy and safety to inform a NICE submission.

METHODS: An SLR synthesized clinical evidence describing the safety and efficacy of FAc and DEX. The network of evidence was formed by the FAME (Individual Patient Data) and MEAD (Aggregated Data) registration trials, linked via a common comparator (Sham). Naive population adjusted and matched methods were considered. A naive analysis was done to generate estimates of comparative safety and efficacy using an adaptation of the Bucher method. To resolve trial differences identified in the naive analysis an Adjusted-ITC was done. Censoring algorithms to account for key trial differences were applied. A MAIC was then done to account for potential effect modifiers (EM) with reweighting so that the distribution of EMs at baseline matched the comparator trial population. Following MAIC reweighting, the distribution of patient weights was examined to assess population overlap and compute the effective sample size (ESS). Efficacy and safety estimands for FAc versus sham were recalculated based on the MAIC re-weighted ITC cohort (+/- censoring) and used to generate comparative estimates versus DEX. These formed the base case analysis. Second matching scenario was explored and a sensitivity analysis adjusting for all EMs regardless of the level of trial imbalance was done.

RESULTS: Matching on baseline imbalance EMs did not substantially reduce the ESS of the FAME ITC cohort. Neither did matching on all EMs. Base-case analyses adjusted for imbalances in EM variables between the two trials demonstrated the broad equivalence of FAc and DEX in key outcomes identified.

CONCLUSIONS: Despite criticism, the MAIC methodology provides a robust approach to address evidence gaps. These results underpinned a positive NICE recommendation in 2024 for the reappraisal of FAc (TA613).