1. 1
© The Author 2015. Published by Oxford University Press on behalf of the European Society for Medical
Oncology. All rights reserved. For permissions, please email: journals.permissions@oup.com.
The Effect of Filgrastim or Pegfilgrastim on Survival Outcomes
of Patients With Cancer Receiving Myelosuppressive
Chemotherapy
G. H. Lyman1,*
, M. Reiner2
, P. K. Morrow3
, J. Crawford4
1
Hutchinson Institute for Cancer Outcomes Research, Fred Hutchinson Cancer
Research Center and the Department of Medicine, University of Washington, Seattle,
WA, USA
2
Global Biostatistical Science, Amgen Inc., Thousand Oaks, CA, USA
3
Hematology/Oncology, Amgen Inc., Thousand Oaks, CA, USA
4
Department of Medicine, Duke University School of Medicine, and Duke Cancer
Institute, Durham, NC, USA
*
Correspondence to: Dr. Gary H. Lyman, Hutchinson Institute for Cancer Outcomes
Research, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, M3-
B232, Seattle, WA 98109-1024, USA; phone: +1 206-667-6670; fax: +1 206-667-5977;
email: glyman@fhcrc.org
Annals of Oncology Advance Access published April 7, 2015

2. 2
ABSTRACT
Background: Primary prophylaxis with granulocyte colony-stimulating factor (G-CSF) is
associated with higher chemotherapy relative dose intensity, which may lead to
improved outcomes; however, the association between G-CSF primary prophylaxis and
overall survival (OS) is not well characterized. This study assessed the effect of G-CSF
primary prophylaxis on patient outcomes in randomized, controlled, registrational clinical
trials of filgrastim and pegfilgrastim.
Patients and methods: Three placebo-controlled and two noninferiority clinical trials of
filgrastim and/or pegfilgrastim in patients receiving myelosuppressive chemotherapy for
lung, breast, or colorectal cancer were included. Median OS, 6- and 12-month survival
rates, and hazard ratios (HRs; unadjusted Cox model with 95% confidence intervals
[CIs]) were estimated for patients receiving ≥1 dose of filgrastim, pegfilgrastim, or
placebo. Comparisons were based on a log-rank test. A fixed-effect meta-analysis
assessed the effect of primary prophylaxis with filgrastim/pegfilgrastim on OS in the
placebo-controlled trials.
Results: In patients with lung cancer receiving filgrastim versus placebo, median OS
was 14.1 versus 11.1 months (HR, 0.81; 95% CI, 0.48–1.35; P=0.412); in patients who
crossed over to filgrastim from placebo after cycle 1, median OS was 16.9 months (HR,
0.75; 95% CI, 0.43–1.28; P=0.286). Median OS was inestimable in at least one
treatment arm in the other studies because of the small number of OS events. Where
estimable, 6- and 12-month survival rates were generally greater among patients
receiving filgrastim/pegfilgrastim versus placebo. In the meta-analysis of placebo-

3. 3
controlled studies comparing G-CSF primary prophylaxis with placebo in the as-treated
analysis sets, the HR (95% CI) for OS was 0.77 (0.58–1.03).
Conclusions: In this retrospective analysis, OS point estimates were greater among
patients receiving filgrastim versus placebo, but the differences were not statistically
significant. Further studies evaluating patient outcomes with G-CSF prophylaxis are
warranted.
Abstract word count: 287 (limit, 300)
Clinical trial registration: NCT00035594, NCT00094809
Keywords: pegfilgrastim, filgrastim, G-CSF, overall survival, neutropenia
Key message: In this analysis of registrational clinical trials of filgrastim and
pegfilgrastim, patients with cancer receiving myelosuppressive chemotherapy with
prophylactic G-CSF support had greater OS point estimates compared with placebo, and
the hazard ratio for OS was <1 in a meta-analysis of the placebo-controlled studies in
favor of G-CSF vs placebo. The results were not statistically significant.

4. 4
Introduction
Severe neutropenia (SN) and febrile neutropenia (FN) are dose-limiting toxicities of
myelosuppressive chemotherapy that may lead to chemotherapy dose delays or dose
reductions [1-3]. Prophylactic granulocyte colony-stimulating factor (G-CSF) therapy has
been shown to reduce the incidence of SN and FN in patients receiving chemotherapy
for nonmyeloid malignancies and has been associated with increased chemotherapy
relative dose intensity (RDI) compared with the absence of G-CSF support [4-9].
Retrospective analyses of randomized controlled trials (RCTs) have shown an
association between higher chemotherapy RDI and improved outcomes in patients
receiving cytotoxic chemotherapy [10, 11]. A recent meta-analysis of RCTs assessing
G-CSF primary prophylaxis in patients receiving myelosuppressive chemotherapy
demonstrated a reduction in all-cause mortality among patients receiving G-CSF support
[12]; however, the effect of G-CSF primary prophylaxis on overall survival (OS) has not
been well characterized in clinical trials. This retrospective analysis assessed the effect
of G-CSF primary prophylaxis on OS in patients receiving myelosuppressive
chemotherapy and filgrastim, pegfilgrastim, or placebo in five registrational RCTs for
which outcome data were available [5-9].
Methods
Studies and patients
An overview of the five registrational RCTs is presented in Table 1. The placebo-
controlled trials were a phase 3 trial of filgrastim in small cell lung cancer (SCLC) [5], a
phase 3 trial of pegfilgrastim in breast cancer (NCT00035594) [6], and a phase 2 trial of

5. 5
pegfilgrastim in colorectal cancer (NCT00094809) [7]. Two phase 3 noninferiority trials
compared pegfilgrastim with filgrastim in breast cancer (referred to as noninferiority
study 1 [8] and noninferiority study 2 [9]).
Calculation of summary statistics for exposure to pegfilgrastim/filgrastim was based on
the total dose for each patient over four cycles of chemotherapy (pegfilgrastim dose =
single dose × 4; filgrastim dose = single dose × days of dosing × 4). Similarly, the
tabulation of the number of doses reflects the total number of doses received by a
patient in four cycles. For patients who crossed over from placebo to G-CSF after cycle
1, summary statistics for exposure and the total number of doses were calculated based
on a maximum of three cycles.
Meta-analysis of OS from Cox proportional hazard (PH) models in the three placebo-
controlled trials was performed using a fixed-effect model; no measurable heterogeneity
was detected (I2
= 0%). Results of the meta-analysis at the level of individual patient
data (IPD) with a stratified Cox PH model were identical.
Data collection and analysis
The original publications of the placebo-controlled registrational trials reported
OS/deaths for the intent-to-treat (ITT) analysis set. In this analysis, the treatment arms
for OS/deaths were determined by the treatment actually received (as-treated set).
Exposure to pegfilgrastim/filgrastim and OS are reported separately for patients who
crossed over from placebo to G-CSF according to study design. The meta-analysis of
the three placebo-controlled trials used the as-treated analysis sets.

6. 6
Incidence of severe neutropenia and febrile neutropenia
Severe neutropenia was defined as grade 3/4 neutropenia (absolute neutrophil count
[ANC] <1.0 × 109
cells/L) or grade 4 neutropenia (ANC <0.5 × 109
cells/L). Febrile
neutropenia was defined as a temperature ≥38.2ºC with ANC <1.0 × 109
cells/L [5, 7] or
ANC <0.5 × 109
cells/L [6, 8, 9] on the same day [5-9] or the day after [6, 7] completion
of chemotherapy. The incidence and duration of SN and FN were evaluated for the
safety analysis set in the placebo-controlled studies. Differences between the G-CSF
and placebo arms in the incidence of SN and FN were evaluated by calculating the
relative risk (RR), where RR <1.0 indicated a lower event rate for the G-CSF arm versus
the placebo arm; P values were calculated using the chi-square test. The incidence of
SN and FN was assessed only in cycle 1 because patients randomized to placebo in the
lung and breast cancer studies who developed FN could cross over after cycle 1 to
receive open-label filgrastim or pegfilgrastim, respectively [5, 6]. Blood collection for the
assessment of ANC was performed 3 days per week in the lung cancer study [5],
allowing reliable estimates of the duration of neutropenia. Because blood was collected
only once per week in the breast and colorectal cancer studies [6, 7], duration of
neutropenia could not be assessed.
Antibiotic use
Differences in intravenous and oral antibiotic use for any cause between the G-CSF and
placebo arms were evaluated by calculating RR, where RR <1.0 indicated a lower use
rate for the G-CSF arm relative to the placebo arm; P values were based on the chi-
square test. Deaths by extent of disease were summarized descriptively.

7. 7
Overall survival
Kaplan-Meier estimates of OS (the time from study day 1 until death) and 6- and 12-
month survival rates (where estimable) were compared between arms using an
unadjusted log-rank test. Maximum follow-up time on the phase 3 breast cancer trial [6]
was 159 days, insufficient to estimate 6- and 12-month survival. Median survival time
was not reached for the two noninferiority trials. Unadjusted Cox PH models were used
to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). A fixed-effect,
study-level meta-analysis was performed to estimate OS among the three placebo-
controlled trials. In the absence of heterogeneity, the results from the IPD analysis
(stratified on study) were identical.
Results
Cycle 1 incidence of severe neutropenia and febrile neutropenia
In the ITT analysis sets of the three placebo-controlled studies, the incidence of grade
3/4 neutropenia in cycle 1 was significantly lower among patients receiving G-CSF
versus placebo (Table 2). The incidences of grade 3/4 and grade 4 neutropenia were
significantly lower among patients in the lung cancer study receiving filgrastim versus
placebo (P=0.0014 and P=0.0001, respectively), as was the incidence of FN (P<0.001).
Likewise, the incidences of grade 3/4 and grade 4 neutropenia were significantly lower
among patients in the breast cancer study receiving pegfilgrastim versus placebo (both
P<0.0001), as was the incidence of FN (P<0.001). In the colorectal cancer study, the
incidence of grade 3/4 neutropenia was significantly lower among patients receiving
pegfilgrastim versus placebo (P=0.0027).

8. 8
Cycle 1 antibiotic use
Oral and intravenous antibiotic use (any cause) in cycle 1 was significantly reduced
among patients in the lung cancer study receiving filgrastim versus placebo (P=0.0048)
and among patients in the breast cancer study receiving pegfilgrastim versus placebo
(P=0.0007; Table 2). The difference in antibiotic use in the colorectal cancer study,
however, was not statistically significant (P=0.580).
Exposure to G-CSF and chemotherapy
Across all studies, the median exposure (mg) to pegfilgrastim or filgrastim was similar
(Table 3). In the two placebo-controlled trials, in which crossover from placebo to G-
CSF was allowed by study design [5, 6], the median exposure (mg and number of
doses) was similar and consistent among patients receiving G-CSF for a total of four
cycles beginning in cycle 1 and patients who crossed over from placebo and received G-
CSF for a total of three cycles beginning in cycle 2. In the noninferiority studies [8, 9],
mean exposure was higher in pegfilgrastim treatment arms compared with filgrastim
treatment arms.
The studies included in this analysis reported limited information on chemotherapy RDI,
with no information on dose reductions or delays reported in the placebo-controlled lung
cancer study or the first noninferiority breast cancer study [5, 9]. In the placebo-
controlled trial in colorectal cancer [7], a nonsignificant difference in the incidence of
chemotherapy dose delays or reductions was observed between patients receiving
pegfilgrastim versus placebo (33.4% vs 45%; P=0.06). In the placebo-controlled trial in
breast cancer [6], 80% and 78% of patients in the pegfilgrastim and placebo groups,
respectively, received their planned chemotherapy dose on time; these results were
expected because patients who developed FN then crossed over to pegfilgrastim. The

9. 9
second noninferiority breast cancer study reported similar chemotherapy dose
administration between the filgrastim and pegfilgrastim arms, with 90% of patients
receiving chemotherapy according to the planned schedule [8].
Overall survival and 6- and 12-month survival
There were generally fewer OS events among patients receiving G-CSF versus placebo
in the placebo-controlled studies; there were generally fewer OS events among patients
receiving pegfilgrastim versus filgrastim in the noninferiority studies (Table 4; Figure 1).
In the lung cancer study, the point estimate for median OS was 3 months longer among
patients receiving filgrastim only versus placebo only, but the difference was not
statistically significant (P=0.412 for HR). Among patients in the placebo arm receiving
open-label filgrastim after cycle 1 as a result of FN, the point estimate for median OS
was almost 6 months longer versus placebo only, but the difference was not statistically
significant (P=0.286 for HR). In the other studies, median OS could not be estimated
because of the small number of OS events.
To evaluate the relationship between G-CSF primary prophylaxis and survival, a fixed
effect model meta-analysis of the placebo-controlled studies was performed. In the as-
treated analysis sets comparing G-CSF primary prophylaxis with placebo, the HR (95%
CI) for OS was 0.77 (0.58–1.03).
In the placebo-controlled studies, where estimable, 6- and 12-month survival rates were
generally higher among patients receiving G-CSF (beginning in cycle 1 or cycle 2)
versus placebo (Table 4). In the noninferiority studies, 6- and 12-month survival rates,
where estimable, were comparable among patients receiving pegfilgrastim versus
filgrastim.

10. 10
Discussion
Chemotherapy-induced FN can often result in chemotherapy dose delays or dose
reductions, which may result in reduced chemotherapy RDI and compromised outcomes
[13-16]. Meta-analyses and pooled analyses have provided some evidence of an
association between G-CSF primary prophylaxis and improved survival in patients
receiving myelosuppressive chemotherapy [4, 12, 17]. In this retrospective analysis of
RCTs of filgrastim and pegfilgrastim [5-7], patients receiving G-CSF primary prophylaxis
had greater median OS point estimates and greater 6- and 12-month survival rates
versus placebo, but the benefits were not statistically significant. Similar nonsignificant
increases in the median OS point estimate and 6- and 12-month survival rates were
seen among patients with SCLC who were randomized to placebo but then crossed over
to open-label filgrastim after experiencing FN during cycle 1. In the fixed-effect meta-
analysis of the placebo-controlled studies (Cox PH model with no heterogeneity), the HR
for OS was <1 in favor of G-CSF primary prophylaxis compared with placebo; although
the result was not statistically significant, it may be considered hypothesis generating.
In contrast with the original reports of the registrational studies [5-9], this analysis
assessed OS data from patients in the as-treated analysis data set rather than the ITT
data set to ensure that the OS estimate reflected the actual treatment received by
patients. Results of our survival analysis are consistent with those in previous reports [4,
12, 17, 18]. In a meta-analysis of RCTs (N=59), patients receiving G-CSF primary
prophylaxis with myelosuppressive chemotherapy had reduced mortality risk versus
patients without G-CSF support (RR, 0.93; P<0.001) [12]. In an earlier meta-analysis of
17 RCTs, patients receiving primary prophylaxis with filgrastim, pegfilgrastim, or
lenograstim had lower risk of infection-related mortality (RR, 0.55; P=0.018) and early
mortality during chemotherapy treatment (RR, 0.60; P=0.002) versus patients without G-

11. 11
CSF support [4]. Evidence of improved survival has also been observed with G-CSF
prophylaxis in patients receiving myelosuppressive chemotherapy for myeloid
malignancies [17, 18]. Notably, a higher 5-year survival rate was observed with primary
prophylaxis with filgrastim or lenograstim versus no prophylaxis in a subgroup of patients
with acute lymphoblastic leukemia (44% versus 27%; P=0.03) [17]. Additionally, patients
receiving filgrastim or pegfilgrastim prophylaxis during the first cycle (P=0.018) or any
cycle (P=0.04) of chemotherapy for acute myeloid leukemia survived longer than
patients who did not receive prophylaxis [18]. Collectively, our retrospective analysis
and these studies provide evidence that G-CSF primary prophylaxis may improve OS in
patients receiving myelosuppressive chemotherapy.
In all three placebo-controlled studies, exposure to G-CSF was generally similar
between active treatment arms, and the incidence of grade 3/4 neutropenia in cycle 1
was significantly lower among patients receiving G-CSF compared with those receiving
placebo. The incidences of grade 4 neutropenia, FN, and antibiotic use in cycle 1 were
also lower in the G-CSF arms of the placebo-controlled studies, but the differences were
not statistically significant in the colorectal cancer study. The apparent treatment effect
of G-CSF primary prophylaxis was likely reduced in the placebo-controlled SCLC [5] and
breast cancer [6] studies, which allowed patients in the placebo arm to cross over to
secondary prophylaxis with G-CSF after cycle 1 as a result of FN. This crossover design
limited the ability to evaluate the impact of G-CSF on survival outcomes in these trials.
This analysis had several limitations. First, the number of OS events was small, and the
studies were neither designed nor powered to assess OS [5-9]. Second, patients who
crossed over from the placebo arm to the G-CSF arm in cycle 2 did so after experiencing
FN in cycle 1 and cannot be assumed to have the same risk of FN as patients who

12. 12
remained on the assigned therapy. In the lung cancer study, OS in the placebo-only arm
may have been shortened by early deaths/withdrawals compared with the crossover arm
in which patients survived to receive subsequent therapy. Third, follow-up time was not
equivalent among the studies (60–240 days in the placebo-controlled breast cancer
study and noninferiority study 1 versus 900–1000 days in the other studies). Fourth, the
studies differed in the tumor types and chemotherapy regimens and in how and when
neutropenia was assessed. Lastly, because of limited availability of data, we were not
able to determine in this analysis if patients receiving G-CSF experienced fewer
chemotherapy dose delays or higher RDIs that may have correlated with improved
outcomes. Despite these limitations, many of which were likely to bias against the G-
CSF-containing arm, point estimates for OS were greater among patients receiving G-
CSF primary prophylaxis compared with placebo. However, the differences were not
statistically significant.
Current guidelines recommend initiating primary prophylaxis with G-CSF in patients at
high risk for FN in the first chemotherapy cycle or when needed for the on-time delivery
of full-dose chemotherapy [19-22]. Secondary prophylaxis is less desirable because
many patients may experience neutropenic complications before receiving G-CSF
support. Prospective assessment of the effects of G-CSF primary prophylaxis on long-
term outcomes in patients with cancer receiving myelosuppressive chemotherapy is
warranted.

13. 13
Acknowledgments
The authors acknowledge James Balwit, MS, and Benjamin Scott, PhD, whose work
was funded by Amgen Inc., for assistance in writing this manuscript.
Funding
This study was funded by Amgen Inc. There were no grants or applicable grant
numbers for this study.
Disclosures
GHL is the primary investigator on a research grant from Amgen Inc. to the Fred
Hutchinson Cancer Research Center. MR and PKM are employees of and own stock in
Amgen Inc. JC has served as a consultant for and has received research funding from
Amgen Inc.

14. 14
References
1. Lyman GH, Dale DC, Crawford J. Incidence and predictors of low dose-intensity
in adjuvant breast cancer chemotherapy: a nationwide study of community
practices. J Clin Oncol 2003; 21: 4524-4531.
2. Lyman GH, Dale DC, Friedberg J et al. Incidence and predictors of low
chemotherapy dose-intensity in aggressive non-Hodgkin's lymphoma: a
nationwide study. J Clin Oncol 2004; 22: 4302-4311.
3. Lyman GH. Chemotherapy dose intensity and quality cancer care. Oncology
(Williston Park) 2006; 20: 16-25.
4. Kuderer NM, Dale DC, Crawford J, Lyman GH. Impact of primary prophylaxis
with granulocyte colony-stimulating factor on febrile neutropenia and mortality in
adult cancer patients receiving chemotherapy: a systematic review. J Clin Oncol
2007; 25: 3158-3167.
5. Crawford J, Ozer H, Stoller R et al. Reduction by granulocyte colony-stimulating
factor of fever and neutropenia induced by chemotherapy in patients with small-
cell lung cancer. N Engl J Med 1991; 325: 164-170.
6. Vogel CL, Wojtukiewicz MZ, Carroll RR et al. First and subsequent cycle use of
pegfilgrastim prevents febrile neutropenia in patients with breast cancer: a
multicenter, double-blind, placebo-controlled phase III study. J Clin Oncol 2005;
23: 1178-1184.
7. Hecht JR, Pillai M, Gollard R et al. A randomized, placebo-controlled phase II
study evaluating the reduction of neutropenia and febrile neutropenia in patients
with colorectal cancer receiving pegfilgrastim with every-2-week chemotherapy.
Clin Colorectal Cancer 2010; 9: 95-101.

15. 15
8. Green MD, Koelbl H, Baselga J et al. A randomized double-blind multicenter
phase III study of fixed-dose single-administration pegfilgrastim versus daily
filgrastim in patients receiving myelosuppressive chemotherapy. Ann Oncol
2003; 14: 29-35.
9. Holmes FA, O'Shaughnessy JA, Vukelja S et al. Blinded, randomized,
multicenter study to evaluate single administration pegfilgrastim once per cycle
versus daily filgrastim as an adjunct to chemotherapy in patients with high-risk
stage II or stage III/IV breast cancer. J Clin Oncol 2002; 20: 727-731.
10. Lyman GH. Impact of chemotherapy dose intensity on cancer patient outcomes.
J Natl Compr Canc Netw 2009; 7: 99-108.
11. Lyman G, Poniewierski M, Wogu A et al. Association of survival with
chemotherapy intensity, myelosuppression, and supportive care in patients with
advanced solid tumors. J Clin Oncol 2013; 31: 6534.
12. Lyman GH, Dale DC, Culakova E et al. The impact of the granulocyte colony-
stimulating factor on chemotherapy dose intensity and cancer survival: a
systematic review and meta-analysis of randomized controlled trials. Ann Oncol
2013; 24: 2475-2484.
13. Budman DR, Berry DA, Cirrincione CT et al. Dose and dose intensity as
determinants of outcome in the adjuvant treatment of breast cancer. The Cancer
and Leukemia Group B. J Natl Cancer Inst 1998; 90: 1205-1211.
14. Kwak LW, Halpern J, Olshen RA, Horning SJ. Prognostic significance of actual
dose intensity in diffuse large-cell lymphoma: results of a tree-structured survival
analysis. J Clin Oncol 1990; 8: 963-977.

16. 16
15. Bonadonna G, Valagussa P, Moliterni A et al. Adjuvant cyclophosphamide,
methotrexate, and fluorouracil in node-positive breast cancer: the results of 20
years of follow-up. N Engl J Med 1995; 332: 901-906.
16. Lepage E, Gisselbrecht C, Haioun C et al. Prognostic significance of received
relative dose intensity in non-Hodgkin's lymphoma patients: application to LNH-
87 protocol. The GELA. (Groupe d'Etude des Lymphomes de l'Adulte). Ann
Oncol 1993; 4: 651-656.
17. Giebel S, Thomas X, Hallbook H et al. The prophylactic use of granulocyte-
colony stimulating factor during remission induction is associated with increased
leukaemia-free survival of adults with acute lymphoblastic leukaemia: a joint
analysis of five randomised trials on behalf of the EWALL. Eur J Cancer 2012;
48: 360-367.
18. Bradley AM, Deal AM, Buie LW, van Deventer H. Neutropenia-associated
outcomes in adults with acute myeloid leukemia receiving cytarabine
consolidation chemotherapy with or without granulocyte colony-stimulating factor.
Pharmacotherapy 2012; 32: 1070-1077.
19. Smith TJ, Khatcheressian J, Lyman GH et al. 2006 update of recommendations
for the use of white blood cell growth factors: an evidence-based clinical practice
guideline. J Clin Oncol 2006; 24: 3187-3205.
20. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in
Oncology: Myeloid Growth Factors, version 2.2014. Available at:
http://www.NCCN.org. Accessed February 2, 2015.
21. Crawford J, Caserta C, Roila F, Esmo Guidelines Working Group. Hematopoietic
growth factors: ESMO Clinical Practice Guidelines for the applications. Ann
Oncol 2010; 21: v248-v251.

17. 17
22. Aapro MS, Bohlius J, Cameron DA et al. 2010 update of EORTC guidelines for
the use of granulocyte-colony stimulating factor to reduce the incidence of
chemotherapy-induced febrile neutropenia in adult patients with
lymphoproliferative disorders and solid tumours. Eur J Cancer 2011; 47: 8-32.

18. 18
Figure Legend
Figure 1. Kaplan-Meier curves for median overall survival in the safety analysis set
for (A) the phase 3 small cell lung cancer study, (B) the phase 3 breast
cancer study, (C) the phase 2 colorectal cancer study, (D) noninferiority
study 1 (phase 3 breast cancer), and (E) noninferiority study 2 (phase 3
breast cancer). Patients randomized to placebo were allowed to cross
over to receive open-label filgrastim (A) or pegfilgrastim (B) as a result of
febrile neutropenia during cycle 1. Note: The duration of follow-up differs
among the studies.

19. 19

20. 20

21. 21

22. 22

23. 23

24. Table 1. Overview of Study Designs in Registrational Trials of Filgrastim and Pegfilgrastim
Study Design Primary Endpoint
Chemotherapy
Regimen G-CSF Dosing
Phase 3 SCLC [5]
(N=211)
Double-blind,
randomized, placebo-
controlled
FN (temperature
≥38.2°C with ANC <1.0
× 109
cells/L)
Cyclophosphamide,
doxorubicin, etoposide
Filgrastim 230 μg/m2
daily up to 14 days per
cycle
Phase 3 breast [6]
(N=928)
Double-blind,
randomized, placebo-
controlled
FN (temperature
≥38.2°C with ANC <0.5
× 109
cells/L on same
day or 1 day after fever)
Docetaxel Pegfilgrastim 6 mg once
per cycle
Phase 2 CRC [7]
(N=252)
Double-blind,
randomized, placebo-
controlled
Incidence of grade 3/4
neutropenia (ANC <1.0
× 109
cells/L)
FOLFOX4, FOLFIRI, or
FOIL
Pegfilgrastim 6 mg per
cycle
Noninferiority 1: phase
3 breast [8]
(N=157)
Double-blind,
randomized,
noninferiority
Duration of grade 4
neutropenia (ANC <0.5
× 109
cells/L) in cycle 1
Doxorubicin, docetaxel Pegfilgrastim 6 mg/kg
once per cycle or
filgrastim 5 μg/kg daily
Noninferiority 2: phase
3 breast [9]
(N=310)
Double-blind,
randomized,
noninferiority
Duration of grade 4
neutropenia (ANC <0.5
× 109
cells/L) in cycle 1
Doxorubicin, docetaxel Pegfilgrastim 100 μg/kg
once per cycle or
filgrastim 5 μg/kg daily
ANC=absolute neutrophil count; CRC=colorectal cancer; FN=febrile neutropenia; FOIL=5-fluorouracil (5-FU), oxaliplatin,
irinotecan, leucovorin; FOLFIRI=5-FU, leucovorin, irinotecan; FOLFOX4=5-FU, leucovorin, oxaliplatin; G-CSF=granulocyte
colony-stimulating factor; SCLC=small cell lung cancer.

25. Table 2. Incidence of Neutropenia and Antibiotic Use in the Safety Analysis Set During Cycle 1 (ITT Analysis)
Study
Grade 3/4
Neutropenia,a
n/N (%)
Grade 4
Neutropenia,b
n/N (%)
Febrile
Neutropenia,
n/N (%)
Antibiotic Use,c
n/N
(%)
Phase 3 SCLC [5]
(N=211)
Filgrastim
Placebo
84/93 (90.3)
101/101 (100)
RR, 0.90
(95% CI, 0.85–0.97)
P=0.0014
78/93 (83.9)
100/101 (99.0)
RR, 0.85
(95% CI, 0.77–0.93)
P=0.0001
28/92 (30.4)
57/102 (55.9)
RR, 0.54
(95% CI, 0.38–0.78)
P<0.001
44/100 (44.0)
68/107 (63.6)
RR,d
0.69
(95% CI, 0.53–0.90)
P=0.0048e
Phase 3 breast [6]
(N=928)
Pegfilgrastim
Placebo
39/467 (8.4)
358/461 (77.7)
RR, 0.11
(95% CI, 0.08–0.15)
P<0.0001
24/467 (5.1)
260/461 (56.4)
RR, 0.09
(95% CI, 0.06–0.14)
P<0.0001
2/467 (<1)
52/461 (11.3)
RR, 0.04
(95% CI, 0.01–0.15)
P<0.001
56/467 (12.0)
93/461 (20.2)
RR,d
0.59
(95% CI, 0.44–0.81)
P=0.0007
Phase 2 CRC [7]
(N=252)
Pegfilgrastim
Placebo
5/124 (4.0)
18/117 (15.4)
RR, 0.26
(95% CI, 0.10–0.68)
P=0.0027
4/124 (3.2)
6/117 (5.1)
RR, 0.63
(95% CI, 0.18–2.17)
P=0.4592
1/124 (<1)
2/117 (1.7)
RR, 0.47
(95% CI, 0.04–5.13)
P=0.552
16/124 (12.9)
18/117 (15.4)
RR,d
0.84
(95% CI, 0.45–1.57)
P=0.580
ANC=absolute neutrophil count; CI=confidence interval; CRC=colorectal cancer; G-CSF=granulocyte colony-stimulating factor;
ITT=intent-to-treat; RR=relative risk; SCLC=small cell lung cancer.
a
Grade 3/4 = ANC <1000 cells/mm3
.
b
Grade 4 = ANC <500 cells/mm3
.

26. c
Includes all intravenous and oral antibiotic use.
d
RR <1.0 indicates a lower event rate for the G-CSF arm relative to the placebo arm.
e
P value based on the chi-square test.

27. Table 3. Exposure to Pegfilgrastim/Filgrastim and Baseline Disease Characteristics in the Safety Analysis Set By Study
Study
Patients
Treated,
n Exposure to G-CSF, mg G-CSF Doses, n
Baseline Disease
Stage, %
Phase 3 SCLC [5]
Filgrastim only (n=100) 100 Median, 23.5
Range, 1.1–60.4
Median, 65
Range, 3.0–126
Limited: 27
Extensive: 73
Placebo then filgrastima
(n=73) 73 Median, 18.6
Range, 3.0–26.1
Median, 50
Range, 11–71
Limited: 29
Extensive: 71
Placebo only (n=34) 34 NA NA Limited: 23
Extensive: 77
Phase 3 breast [6]
Pegfilgrastim only (n=467) 467 Median, 24.0
Range, 6.0–36.0
Median, 4.0
Range, 1.0–4.0
Nonmetastatic: 19
Metastaticb
: 81
Placebo then pegfilgrastima
(n=89) 89 Median, 18.0
Range, 6.0–18.0
Median, 3.0
Range, 1.0–3.0
Nonmetastatic: 9
Metastatic: 91
Placebo only (n=465) 372 NA NA Nonmetastatic: 18
Metastaticb
: 82
Phase 2 CRC [7]
Pegfilgrastim (n=126) 124 Mean, 21.5
SD, 5.4
Median, 24.0
Range, 6.0–24.0
Stage 2: 0.8
Stage 3: 0.8
Stage 4: 95
Not assessed: 3

28. Placebo only (n=126) 117 NA NA Stage 2: 0
Stage 3: 1.6
Stage 4: 97
Not assessed: 1.6
Noninferiority 1: phase 3 breast [8]
Pegfilgrastim (n=80) 79 Mean, 23.5
SD, 2.5
Mean, 3.9
SD, 0.4
Stage 2: 25
Stage 3: 27
Stage 4: 48
Filgrastim (n=77) 76 Mean, 15.0
SD, 4.1
Mean, 41.5
SD, 8.2
Stage 2: 31
Stage 3: 27
Stage 4: 43
Noninferiority 2: phase 3 breast [9]
Pegfilgrastim (n=154) 150 Mean, 29.1
SD, 9.5
Mean, 3.8
SD, 0.7
Stage 2: 49
Stage 3: 26
Stage 4: 25
Filgrastim (n=156) 151 Mean, 15.5
SD, 4.4
Mean, 41.0
SD, 8.7
Stage 2: 56
Stage 3: 25
Stage 4: 19
CRC=colorectal cancer; G-CSF=granulocyte colony-stimulating factor; NA=not applicable; SCLC=small cell lung
cancer; SD=standard deviation.
a
Patients randomized to placebo who crossed over to receive open-label filgrastim/pegfilgrastim as a result of febrile neutropenia
during cycle 1.
b
Stages 3 and 4 combined.

29. Table 4. Kaplan-Meier Estimates of Overall Survival and 6- and 12-Month Survival Rates in the Safety Analysis Seta,b
Study
Events,
n/N (%)
Censored,c
n/N (%)
Median OS,
mo (95% CI)
6-Month
Survival,
% (95% CI)
12-Month
Survival,
% (95% CI)
Unadjusted HR
(95% CI)
Phase 3 SCLC [5]
Filgrastim only 52/100 (52.0) 48/100 (48.0) 14.1 (10.6–NE) 87.0 (80.4–93.6) 55.0 (45.2–64.8)
0.81 (0.48–1.35)
P=0.412
Placebo then
filgrastimd 39/73 (53.4) 34/73 (46.6) 16.9 (13.0–NE) 91.8 (85.5–98.1) 67.1 (56.3–77.9)
0.75 (0.43–1.28)
P=0.286
Placebo only 20/34 (58.8) 14/34 (41.2) 11.1 (9.0–NE) 76.5 (62.2–90.7) 44.1 (27.4–60.8) Reference
Phase 3 breast [6]
Pegfilgrastim only 5/467 (1.1) 462/467 (98.9) NE NE NE
0.40 (0.14–1.17)
P=0.094
Placebo then
pegfilgrastimd 4/89 (4.5) 85/89 (95.5) NE NE NE
1.62 (0.51–5.17)
P=0.414
Placebo only 10/372 (2.7) 362/372 (97.3) NE NE NE Reference
Phase 2 CRC [7]
Pegfilgrastime
47/124 (37.9) 77/124 (62.1) NE (19.0–NE) 88.6 (82.8–94.5) 80.0 (72.5–87.5)
0.81 (0.54–1.20)
P=0.292
Placebo 50/117 (42.7) 67/117 (57.3) 24.8 (19.0–NE) 85.7 (79.0–92.4) 72.0 (63.1–80.8) Reference
Noninferiority 1: phase
3 breast [8]
Pegfilgrastim 2/79 (2.5) 77/79 (97.5) NE 97.4 (93.8–100) NE
NE
P=0.608
Filgrastim 3/76 (3.9) 73/76 (96.1) NE 94.6 (88.2–100) NE Reference

30. Noninferiority 2: phase
3 breast [9]
Pegfilgrastim 18/150 (12.0) 132/150 (88.0) NE 98.6 (96.8–100) 94.4 (90.7–98.2)
NE
P=0.092
Filgrastim 30/151 (19.9) 121/151 (80.1) 32.8 (NE–NE) 97.3 (94.7–99.9) 94.6 (90.9–98.2) Reference
CI=confidence interval; CRC=colorectal cancer; HR=hazard ratio; NE=not estimable; OS=overall survival; SCLC=small cell lung
cancer.
a
Median survival = NE because number of events insufficient to reach the median.
b
Estimates of 6- and 12- month survival = NE because of insufficient time on study.
c
Patients who had not died at the time of the analysis were censored at the date they were last known to be alive.
d
Patients randomized to placebo who crossed over to receive open-label filgrastim/pegfilgrastim after cycle 1 as a result of febrile
neutropenia.
e
Includes three patients randomized to placebo who crossed over to receive open-label filgrastim/pegfilgrastim after cycle 1 as a
result of febrile neutropenia.