INTRODUCTION: In the absence of head-to-head trials, the long-term relative effi cacy of asthmabiologics can be estimated through indirect treatment comparison after matching baselinepatient characteristics from available single-arm long-term extension trials.
METHODS: We employed unanchored matching-adjusted indirect comparison, re-weightingindividual patient data for dupilumab from the TRAVERSE trial ( N =1,368, NCT02134028) and itsparent randomized controlled trials (RCTs) to match aggregate tezepelumab data from theDESTINATION trial ( N =475, NCT03706079) and its parent RCTs for prognostic factors andtreatment effect modifi ers. Outcomes included annualized exacerbation rate (AER) for allasthma exacerbations, AER for asthma exacerbations leading to hospitalization and/oremergency room (ER) visits, and change from baseline (CFB) in pre-bronchodilator forcedexpiratory volume in 1s (pre-BD FEV ) from randomization (in parent RCTs) to Week 100 andWeek 104 for dupilumab and tezepelumab, respectively. Sensitivity analysis (SA) explored theuse of selected key covariates from the primary analysis ( Table ).
RESULTS: After weighting, the effective sample size of TRAVERSE was 271. Compared totezepelumab, dupilumab showed a signifi cantly lower AER for all asthma exacerbations (meandifference [MD]: −0.269; p <0.0001) and a comparable AER for asthma exacerbations leading tohospitalization and/or ER visits (MD: 0.006; p =0.62), with similar results in SA. Dupilumab alsoshowed greater improvement (MD: 0.064L; p =0.07) in pre-BD FEV than tezepelumab, with asignifi cantly higher CFB in SA (MD: 0.153L; p <0.0001).
CONCLUSIONS: In the matched cohort, long-term dupilumab treatment demonstrated lower AERand higher pre-BD FEV than tezepelumab. Limitations included potential unmeasuredcovariates and that the results may not be generalizable to real-life practice.
