The importance of considering differences in study and patient characteristics before undertaking indirect treatment comparisons: a case study of siponimod for secondary progressive multiple sclerosis.

Background:Indirect treatment comparisons (ITCs) provide valuable evidence on comparative efficacy where head-to-head clinical trials do not exist; however, differences in patient populations may introduce bias. Therefore, it is essential to assess between-trial heterogeneity to determine the suitability of synthesizing ITC results. We provide an illustrative case study in multiple sclerosis (MS) where we assess the feasibility of conducting ITCs between siponimod and interferon beta-1b (IFN beta-1b) and between siponimod and ocrelizumab. Methods:We assessed the feasibility of conducting ITCs using standard unadjusted methods (e.g. Bucher or network meta-analysis [NMA]) as well as matching-adjusted indirect comparisons (MAICs) using individual patient data (IPD) from the siponimod (EXPAND) trial, based on guidance from NICE. Time to confirmed disability progression (CDP) at 3 or 6 months was assessed. Results:Bucher ITCs and NMAs, which rely on summary-level data, were not able to account for important cross-trial differences. Comparisons between siponimod and IFN beta-1b were feasible using MAIC; the HRs (95% CI) for CDP-6 and CDP-3 were 0.55 (0.33-0.91) and 0.82 (0.42-1.63), respectively. ITCs were not feasible between siponimod and ocrelizumab because study designs and patient populations were too dissimilar to conduct a reliable ITC. Conclusions:This study highlights the importance of conducting a detailed feasibility assessment before undertaking ITCs to illuminate when excessive between-trial heterogeneity would cause biased results. MAIC was performed for siponimod and IFN beta-1b in the absence of a head-to-head trial and was considered a more valid approach than a traditional ITC to examine comparative effectiveness.