Abstract
Background: Severe eosinophilic asthma (SEA) remains morbid despite standard care. Lacking head-to-head trials, the comparative efficacy of biologics (mepolizumab, benralizumab, dupilumab, tezepelumab) is unclear. This network meta-analysis quantifies their relative efficacy and safety in SEA. Method: Following PRISMA-NMA, we searched databases (inception to July 23, 2025) for RCTs of adolescents/adults with SEA comparing these biologics against placebo or each other. Primary outcomes were annualized asthma exacerbation rate (AAER) and change in pre-bronchodilator FEV₁; secondary outcomes included change in post-bronchodilator FEV₁, ACQ/AQLQ scores, and safety. Bayesian random-effects network meta-analysis was performed, and treatment rankings were summarized using SUCRA. Certainty of evidence was assessed using GRADE. Results: Seventeen RCTs involving 7,173 participants were included. For AAER, tezepelumab ranked highest (SUCRA: 73.35%), followed by mepolizumab (72.65%), benralizumab (56.23%), and dupilumab (36.20%); the corresponding network estimates versus placebo were MD −1.23 (95% CrI: −2.99 to 0.58), −1.16 (95% CrI: −2.76 to 0.49), −0.75 (95% CrI: −1.54 to −0.17), and −0.37 (95% CrI: −2.13 to 1.42), respectively. For ΔPre-BD FEV₁, tezepelumab ranked highest (SUCRA: 79.64%), followed by dupilumab (54.42%), benralizumab (52.19%), and mepolizumab (45.71%). Benralizumab ranked highest for ΔPost-BD FEV₁ and ΔACQ-5, tezepelumab for ΔACQ-6, dupilumab for ΔAQLQ, and mepolizumab for lower SAE risk. For SAEs, network estimates versus placebo were OR 0.45 for mepolizumab (95% CrI: 0.23 to 0.72) and OR 0.58 for benralizumab (95% CrI: 0.35 to 0.84). Leave-one-out analyses showed stable rankings after exclusion of non-tezepelumab studies, whereas exclusion of Wechsler ME 2022 removed the tezepelumab node. GRADE certainty ranged from very low to high, and potential publication bias was detected for SAEs. Conclusion: This NMA suggests that biologics provide clinically relevant benefits for SEA, particularly in reducing exacerbations. Tezepelumab and mepolizumab had the highest ranking probabilities for AAER. Mepolizumab and benralizumab ranked favorably for lower SAE risk, but these safety estimates should be interpreted cautiously because of potential publication bias. Treatment rankings remain limited by low or very low certainty for many indirect active-treatment comparisons, underscoring the need for head-to-head trials.