Propensity score matching/reweighting analysis comparing intravenous golimumab to infliximab for ankylosing spondylitis using data from the GO-ALIVE and ASSERT trials

Abstract

Objective To compare the relative efficacy of intravenous golimumab (GOL IV) and infliximab (IFX) for active ankylosing spondylitis (AS). Methods Propensity score (PS) methods were used to compare the efficacy of GOL IV 2 mg/kg and IFX 5 mg/kg using individual patient data (IPD) from the active arms of the phase 3 GO-ALIVE and ASSERT studies. Outcomes included the proportion of patients with a ≥ 20% improvement in the Assessment of Spondyloarthritis International Society Criteria (ASAS20), change from baseline in Bath Ankylosing Spondylitis Functional Index (BASFI) score, and change from baseline in C-reactive protein (CRP) levels from weeks 4–52. Results Before matching, 105 patients were treated with GOL IV and 201 patients were treated with IFX. After matching on all covariates, 118 patients were included in the ASAS20 analysis, 96 in the BASFI analysis, and 160 in the CRP analysis. After matching, GOL IV showed significantly greater improvement in ASAS20 response than IFX for weeks 28–44 (e.g., OR = 9.05 [95% CI 1.62–50.4] at week 44) and was comparable in change from baseline in BASFI scores and CRP levels to IFX at all time points. Results were robust for inclusion of different sets of covariates in scenario analyses. Conclusions This is the first analysis of its kind to leverage clinical trial data to compare two biologics using PS methods in the treatment of active AS. Overall, GOL IV was associated with greater improvement in ASAS20 response than IFX in patients with AS at 28, 36, and 44 weeks of follow-up. Key Points• Although intravenous golimumab (GOL IV) and infliximab (IFX) are the only two IV-based tumor necrosis factor (TNF) inhibitors with demonstrated phase 3 clinical efficacy in patients with ankylosing spondylitis (AS), no study has evaluated their comparative efficacy in a head-to-head trial.• Propensity score matching was used to derive indirect treatment comparisons of GOL IV and IFX for ≥ 20% in the Assessment of Spondyloarthritis International Society Criteria (ASAS20), change in Bath Ankylosing Spondylitis Functional Index (BASFI), and change in C-reactive protein (CRP) using individual patient data from the GO-ALIVE and ASSERT phase 3 trials.• Propensity score matched indirect comparisons showed improved relative efficacy of GOL IV compared to IFX; after matching for up to 16 baseline covariates, GOL IV was associated with significantly greater odds of ASAS20 response at weeks 28, 36, and 44 than IFX as well as equivalent changes from baseline in BASFI and CRP.• This novel application of propensity score matching using data from phase 3 trials, the first analysis of its kind in AS, allowed adjustment for important imbalances in prognostic factors between trials to generate estimates of comparative efficacy between GOL IV and IFX in the absence of a head-to-head trial between these treatments.

Indirect comparisons of siponimod with fingolimod and ofatumumab in multiple sclerosis: assessing the feasibility of propensity score matching analyses

“…Propensity score matching (PSM) is a common approach for comparing the efficacy of two or more treatments using observational data [24][25][26][27] . More recently, PSM has also been used to derive ITCs using individual patient data (IPD) from multiple RCTs evaluating therapies with similar therapeutic indications 25,26 .…”

“…Propensity score matching (PSM) is a common approach for comparing the efficacy of two or more treatments using observational data [24][25][26][27] . More recently, PSM has also been used to derive ITCs using individual patient data (IPD) from multiple RCTs evaluating therapies with similar therapeutic indications 25,26 . Propensity score methods leverage IPD for both patient populations being compared to correct for differences between populations that are expected to bias comparisons.…”

Comparative efficacy of ofatumumab versus oral therapies for relapsing multiple sclerosis patients using propensity score analyses and simulated treatment comparisons

“…This can be addressed using ITC methods that use individual patient data (IPD) from RCTs to adjust for between-trial differences, such as PS analyses using inverse probability treatment weighting (IPTW) and simulated treatment comparisons (STCs). 16 , 29 , 30 These methods are used in different situations depending on the available data: PS analyses are used when IPD are available for trials for both therapies, whereas STCs are used when IPD are available from trials for one therapy and only summary-level data (SLD) are available for trials for the comparator therapy. Population-adjustment techniques such as STC are increasingly being used in health technology assessment (HTA) applications.…”

Analyzing Tuberculosis Reactivation in Patients with Rheumatoid Arthritis and Ankylosing Spondylitis Treated with Biological Therapy Using Machine Learning Methods

“…A study published in 2020 [6], compared the relative efficacy of infliximab (IFX) at 5 mg/kg and intravenous golimumab (GOL IV) at 2 mg/kg in ankylosing spondylitis patients. The nearest neighbor matching algorithm was used to match patients from the GOL IV group with patients from the IFX group.…”