Air trapping measured at HRCT predicts benralizumab response better than blood eosinophil count in severe asthma

Clinical benefits of anti-IL5Rα biological therapy are heterogeneous among severe eosinophilic asthma patients. Small airway disease has been associated with asthma severity but treatment response is poorly described. We aimed to characterise benralizumab-responding phenotypes via the non-invasive, in vivo, endotypic analysis of QTCT data, and determine the prognostic value of air-trapping measures for clinical improvement. This prospective, two-centre, phase 2 open-label study (<ext-link>NCT03976310</ext-link>) included 59 SEA patients with a blood eosinophil count (BEC) ≥ 300/mm³ and ≥2 exacerbations who initiated benralizumab (30mg SC). 44 had high quality images used to generate expiratory/inspiratory Parametric Response Maps (PRM), and to analyse the small airway disease (SAD) component. PRM SAD was age-corrected given the known impact of physiological aging. ROC curve analyses were performed using the Youden index. Patients were grouped according to highly ambitious clinical improvement 12 months after benralizumab initiation (0 or 1 mild exacerbations, ΔFEV1 > 300 ml and ΔACQ5>0.5), as non-responders (n=13) and responders (n=31). BEC provided a reasonable prediction of responder status. With an AUC of 0.571 (0.359-0.782), for a 365/mm³ cut-off it showed a 62% sensitivity and 63% specificity. Yet, age-adjusted PRM SAD provided a better prediction of response to benralizumab with an AUC of 0.728 (0.557-0.899), showing 84.6% sensitivity and 60% specificity in predicting responders. Our results suggest that distal airway impairment may be a good indicator of responding phenotypes. Whether PRM SAD improves in parallel with clinical benefit will be critical to address.