Robustness of assessment of pulmonary endpoints in phase 3 trials with ARD-3150 in non-cystic fibrosis bronchiectasis (NCFB) patients with chronic Pseudomonas aeruginosa (PA) infections

Introduction: ARD-3150 is a once-daily inhaled formulation of liposomal and free ciprofloxacin tested in two 48 week randomized (2:1) placebo controlled Phase 3 trials (ORBIT-3 and ORBIT-4) in NCFB patients with chronic PA lung infections. Pulmonary exacerbation (PE) events were defined using either the O’Donnell definition (OD) (O’Donnell et al., Chest 1998), or OD plus a sputum change analysis (ODS) (Pasteur et al., Thorax 2010). Methods: Survival Analysis (SA) of the time to first PE and the counting processes extension of the Cox Model (CP) for all PEs and severe PEs were conducted using the two PE definitions. Statistical significance (S) and non-significance (N) at the p=0.05 level are noted. Results: With ODS, the delay in the median time to first PE in ORBIT-3 and ORBIT-4 was 44 (N) and 123 days (S), respectively, while the point estimate for the delay was 78 (N) and 72 days (S) with the OD criterion. The CP hazard ratio for ARD-3150 vs placebo for all PEs using ODS was 0.83 (N) and 0.58 (S) for ORBIT-3 and ORBIT-4, respectively, while using the OD definition it was 0.84 (N) and 0.61 (S). For severe PEs, the CP hazard ratio using ODS was 0.80 (N) and 0.40 (S) for ORBIT-3 and ORBIT-4, respectively, versus 0.81 (N) and 0.42 (S) with OD. Conclusion: The evaluation of the clinically important impact of ARD-3150 versus placebo on the number of PEs was not affected by the presence or absence of sputum changes, though it impacted the numeric value of the time to first PE.