Table 4.

Comparison of mean AOC of ANC in patients who received pegfilgrastim on the same day as chemotherapy versus the next day in cycle 1 and cycle 3

AOC threshold	Cycle	Day of pegfilgrastim use	Mean ± SD AOC (n a) 109/L × day	Crude Mean AOC Difference (95 % CI) 109/L × day	Adjusted Mean AOCb Difference (95 % CI) 109/L × day
<0.5 × 109/L	Cycle 1	Same day	0.65 ± 0.65 (95)	0.31 (0.17, 0.43)	0.30 (0.16, 0.43)
		Next day	0.34 ± 0.49 (96)
	Cycle 3	Same day	0.74 ± 0.50 (30)	0.47 (0.28, 0.64)	0.43 (0.23, 0.61)
		Next day	0.27 ± 0.33 (33)
	Cycles 1 and 3c			0.33 (0.20, 0.45)	0.36 (0.22, 0.49)
<1.0 × 109/L	Cycle 1	Same day	1.85 ± 1.71 (95)	0.75 (0.41, 1.04)	0.73 (0.37, 1.05)
		Next day	1.10 ± 1.31 (96)
	Cycle 3	Same day	2.23 ± 1.18 (30)	1.21 (0.74, 1.65)	1.11 (0.63, 1.55)
		Next day	1.02 ± 0.96 (33)
	Cycles 1 and 3c			0.82 (0.50, 1.12)	0.88 (0.54, 1.20)

Mean AOC difference = mean AOC in same-day patients−mean AOC in next-day patients; analyses in cycle 1 include data from eligible patients in Amgen studies 20020134, 20020778, 2003012, and 20030123 and in cycle 3 include data from eligible patients in Amgen studies 20020778 and 20030122

Bold indicates that 95 % CIs for crude or adjusted mean AOC difference do not include 0

AOC area over the curve; ANC absolute neutrophil count; CI confidence interval; ECOG Eastern Cooperative Oncology Group; SD standard deviation

^aThe n for each parameter (of each treatment arm) is the number of patients eligible for the corresponding statistics in cycle 1 or cycle 3

^bCovariates included in the adjusted model were age and ECOG performance status

^cLinear mixed-effect regression model was used to calculate mean AOC difference in cycles 1 and 3 combined. Within-study and within-patient correlations were controlled for in the analysis. Bootstrap procedure was used to derive 95 % CIs