Application of structured statistical analyses to identify a biomarker predictive of enhanced tralokinumab efficacy in phase III clinical trials for severe, uncontrolled asthma

Table 5 Analysis of key secondary endpoints by baseline periostin concentration in the STRATOS 1 all-comers population (full analysis set)^a

Change from baseline in	Periostin cut-off, ng/ml	Tralo Q2W vs. combined placebo^b, n	Treatment effect (95% CI)	p-value
FEV₁, %	> 27.4	117 vs. 120	6.61 (0.07, 13.14)	0.048
FEV₁, %	≤27.4	239 vs. 241	5.54 (1.00, 10.09)	0.017
ACQ-6 score	> 27.4	104 vs. 103	− 0.11 (− 0.35, 0.13)	0.372
ACQ-6 score	≤27.4	212 vs. 211	−0.20 (− 0.37, − 0.03)	0.022
AQLQ score	> 27.4	104 vs. 103	0.09 (−0.18, 0.36)	0.529
AQLQ score	≤27.4	210 vs. 210	0.18 (−0.01, 0.37)	0.060
Total Asthma Symptom Score	> 27.4	103 vs. 96	0.03 (−0.22, 0.27)	0.843
Total Asthma Symptom Score	≤27.4	210 vs. 216	−0.15 (− 0.32, 0.03)	0.098

ACQ-6 Asthma Control Questionnaire (6-item), AQLQ Standardised Asthma Quality of Life Questionnaire for 12 Years and Older, CI Confidence interval, FEV₁ Forced expiratory volume in 1 s, Q2W Every 2 weeks, Q4W Every 4 weeks, Tralo Tralokinumab
^aThe two treatment arms were compared with the pooled placebo group using repeated measures analyses. For percentage change in FEV₁ the model was treatment group + geographical region + age group + periostin group at baseline + number of exacerbations in the previous year + visit + treatment*visit + treatment*biomarker group using an unstructured variance-covariance matrix. For change in score the model was baseline score + treatment group + geographical region + age group + periostin group at baseline + number of exacerbations in the previous year + visit + treatment*visit + treatment*biomarker group using a compound symmetric variance-covariance matrix. Baseline was the last non-missing measurement recorded prior to randomisation (typically randomisation). No multiplicity adjustments were made and all p-values stated are nominal; values <0.05 are bolded
^bThe placebo treatment group is a combined treatment group (placebo Q2W + placebo Q4W)

ISSN: 1471-2466