Résumé
ABSTRACT Biologics are effective for severe asthma, but not all patients benefit equally. There is an urgent need to understand which biologic works best for which patient. We systematically searched for predictors of response to biologics (except omalizumab) for severe asthma in four bibliographic databases and two trial registries from 1990 to 2024. Two reviewers screened records, extracted data, and assessed risk of bias using a modified CASP checklist. Data were synthesized narratively, and certainty of evidence assessed using the modified GRADE framework. Comparable studies were meta‐analyzed using a random‐effects model. From 5853 records, 21 studies were identified investigating predictors of anti‐IL5/5Rα, 4Rα, and anti‐TSLP response. We found predominantly ‘moderate’ to ‘high’ quality evidence that raised blood eosinophil counts (≥ 300 cells/μL), FeNO levels (> 40 ppb), lack of or low OCS dose (< 10 mg/day), and better asthma control predict biologic response. Evidence for the predictive value of other characteristics was limited and mostly ‘low’ quality. Key reasons for downgrading the evidence were heterogeneous response definitions and imprecision. No data were identified for the pediatric population or biologics targeting the non‐T2 pathway. Outside of traditional inflammatory and clinical variables, there is an unmet need for universally applicable predictors of biologic response for severe asthma.