In patients with severe asthma with eosinophilia in reslizumab clinical trials, high peripheral blood eosinophil levels are associated with low FEV1 reversibility

Table 3 Clinical asthma endpoints

	Placebo		Reslizumab
	Baseline	52 weeks	Baseline	52 weeks	Treatment effect	p value
≥ 400 EOS < 14% reversibility group
					RR (reslizumab vs placebo) (95% CI)
≥ 1 exacerbation, n (%)	NA	28 (53.8%)	NA	26 (40.6%)
Adjusted exacerbation rate (over 52 weeks)^a (95% CI)	NA	1.52 (0.84, 2.72)	NA	0.91 (0.48, 1.72)	0.600 (0.330, 1.090)	0.0937

		Change from baseline at Week 52		Change from baseline at Week 52	52-week treatment difference (reslizumab vs placebo) (95% CI)
N	52	44	64	52
ACQ-7 score^b	2.511 (0.117)	− 0.781 (0.158)	2.464 (0.111)	− 1.189 (0.159)	− 0.408 (− 0.753, − 0.064)	0.0206
AQLQ score^b	4.259 (0.153)	0.872 (0.198)^c	4.518 (0.140)	1.425 (0.200)	0.554 (0.148, 0.959)	0.0079
ASUI score^b	0.634 (0.028)	0.118 (0.029)^c	0.690 (0.024)	0.193 (0.029)^c	0.075 (0.013, 0.137)	0.0174

Overall population excluding the ≥ 400 EOS < 14% reversibility group
					RR (reslizumab vs placebo) (95% CI)
≥ 1 exacerbation, n (%)	NA	209 (49.3%)	NA	125 (30.3%)
Adjusted exacerbation rate over 52 weeks^a (95% CI)	NA	1.80 (1.37, 2.36)	NA	0.80 (0.61, 1.05)	0.446 (0.346, 0.575)	< 0.0001

		Change from baseline at Week 52		Change from baseline at Week 52	52-week treatment difference (reslizumab vs placebo) (95% CI)
N	424	371	413	367
ACQ-7 score^b	2.708 (0.041)	− 0.805 (0.065)	2.637 (0.043)	− 1.141 (0.064)	− 0.336 (− 0.470, − 0.201)	< 0.0001
AQLQ score^b	4.182 (0.053)	0.874 (0.077)^d	4.297 (0.053)	1.171 (0.076)^c	0.297 (0.150, 0.445)	< 0.0001
ASUI score^b	0.630 (0.010)	0.145 (0.012)^e	0.641 (0.010)	0.198 (0.012)^f	0.053 (0.028, 0.078)	< 0.0001

ACQ Asthma Control Questionnaire, AQLQ Asthma Quality of Life Questionnaire, ASUI Asthma Symptom Utility Index, CI confidence interval, EOS blood eosinophil level, LS least square, NA not applicable, SE standard error
^aAdjusted exacerbation rates and confidence intervals based on Negative Binomial regression model adjusted for baseline use of oral corticosteroids (yes or no) and geographical region (US or other)
^bBaseline data are mean (SE). All change from baseline data are LS mean (SE)
^cData unavailable for one patient
^dData unavailable for three patients
^eData unavailable for five patients
^fData unavailable for four patients

ISSN: 1710-1492