Table 4.

Multivariate logistic analyses^a of relationships between infusion rates (continuous variable) and incidence of immediate infusion-related reactions (narrow terms^b) in ramucirumab-treated patients

Factor	Comparison	OR (95% CI) for risk of immediate IRR	p value
Infusion rate	1-unit increase, mg/min	1.014 (0.999, 1.030)	0.071
Age	≥ 65 vs. < 65 years	0.831 (0.633, 1.090)	0.181
Sex	Female vs. male	1.127 (0.848, 1.497)	0.409
Region	Asia vs. rest of the world	2.151 (1.596, 2.898)	< 0.001
Premedication	Yes vs. no	0.519 (0.389, 0.691)	< 0.001
Chemotherapy	Yes vs. no	1.443 (1.033, 2.016)	0.032
Dosing regimen	10 mg/kg Q3W vs. 8 mg/kg Q2W	0.790 (0.565, 1.106)	0.170
	12 mg/kg Q2W vs. 8 mg/kg Q2W	0.456 (0.196, 1.061)	0.068
	6 mg/kg QW vs. 8 mg/kg Q2W	0.922 (0.215, 3.951)	0.913
	8 mg/kg D1D8 Q3W vs. 8 mg/kg Q2W	0.505 (0.289, 0.884)	0.017

CI confidence interval, D day, IRR infusion-related reaction, OR odds ratio, QW weekly, Q2W every 2 weeks, Q3W every 3 weeks

^aIncluding infusion rate as a continuous variable and adjusting for other factors that might affect the incidence of immediate infusion-related reactions (age, sex, premedication, chemotherapy, region, and dosing regimen). Baseline bodyweight was not included as a covariate because of its strong correlation with infusion rate

^bBroad- and narrow-scope preferred terms were used within the Standardized Medical Dictionary for Regulatory Activities (MedDRA) queries. “Narrow” preferred terms included terms that are highly likely to represent the condition of interest and were considered sufficient to identify immediate IRRs with reasonable precision and to appropriately reflect the incidence rate of immediate IRRs