LATE-BREAKING ABSTRACT: Early identification of responders to reslizumab at 16 weeks using an algorithm derived from the pivotal clinical studies of severe eosinophilic asthma (SEA) patients

Abstract

Background: A tool to aid early identification of reslizumab (RES) responders (R) at week (wk)16 could guide continued therapy for patients (pts) with inadequately controlled severe eosinophilic asthma (SEA) Aims: To derive an algorithm using changes in clinical variables from baseline to wk16 of treatment, predictive of response at wk52 Methods: Data was derived from two 52-wk, pivotal trials of RES in 320 pts with GINA 4+5 SEA, eos ≥400 with mean of 1.85 clinical asthma exacerbations (CAE) in prior 12 months.Pts with 0 or 1 CAE at wk52 + at least one of: ≥10% forced expiratory volume in 1 second (FEV 1 ), or minimal clinically important difference (MCID) change in asthma control questionnaire-6 (ACQ6) or asthma quality of life questionnaire (AQLQ) were defined as R; those with ≥2 CAE as non-responders (NR) unless ≥10% FEV 1 + change ACQ6 or AQLQ by MCID, or 50% reduction from historical baseline. Multinomial logit model variables were: FEV 1 , ACQ6, AQLQ at baseline and wk16; historic CAE and CAE within 16wks of RES. Using linear scores derived from multiple logistic regression probabilities (P R, NR, I ) were calculated, allowing a decision rule for R, NR or indeterminate Results: The decision rule identified R with a P R >0.8 and NR with a P NR >0.25. The algorithm correctly predicted 228 pts (71%) as R, 28 pts (9%) as NR and 65 pts (20%) as indeterminate with 95% sensitivity and 68% specificity Conclusions: Assessment of FEV 1 , ACQ6, AQLQ and CAE status after 16wks of treatment predicted response at wk52 in 70% of RES-treated pts. This approach should help guide continuation of therapy Sponsor:Teva Pharmaceuticals Inc.