VOLUME 26 䡠 NUMBER 12
JOURNAL OF CLINICAL ONCOLOGY
From the Beatson Oncology Centre,
Glasgow, United Kingdom; University of
Sydney and Sydney Cancer Centre,
Sydney, Australia; Hospital Clı́nico San
Carlos, Madrid; Hospital Marques De
Valdecilla, Santander, Spain; Vienna
University Medical School, Vienna,
Austria; Tel Aviv Sourasky Medical
Center, Tel Aviv, Israel; Cancer Care
Manitoba, St Boniface General Hospital,
Winnipeg, Manitoba; Cross Cancer
Institute, Edmonton, Alberta; CHUQ
L’Hotel-Dieu De Quebec, Quebec,
Canada; Russian Cancer Research
Center, Moscow, Russian Federation;
Chang-Gung Memorial Hospital, Taipei,
Taiwan; F. Hoffmann-La Roche, Basel,
Switzerland; and the Memorial Sloan
Kettering Cancer Center, New York, NY.
Submitted October 30, 2007; accepted
December 19, 2007.
Supported by Roche.
Presented in part at the 31st European
Society of Medical Oncology Congress,
Istanbul, Turkey, September 29-
October 3, 2006; the American Society
of Clinical Oncology Gastrointestinal
Cancers Symposium, Orlando, FL,
January 19-21, 2007; and the 43rd
Annual Meeting of the American Soci-
ety of Clinical Oncology, Chicago, IL,
June 1-5, 2007.
Authors’ disclosures of potential con-
flicts of interest and author contribu-
tions are found at the end of this
article.
Corresponding author: Jim Cassidy,
MD, Glasgow University, Garscube
Estate, Switchback Rd, Glasgow, Scot-
land, G61 1BD; e-mail: j.cassidy@
beatson.gla.ac.uk.
© 2008 by American Society of Clinical
Oncology
0732-183X/08/2612-2006/$20.00
DOI: 10.1200/JCO.2007.14.9898
2006 © 2008 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.
䡠 APRIL 20 2008
O R I G I N A L R E P O R T
Randomized Phase III Study of Capecitabine Plus
Oxaliplatin Compared With Fluorouracil/Folinic Acid
Plus Oxaliplatin As First-Line Therapy for Metastatic
Colorectal Cancer
Jim Cassidy, Stephen Clarke, Eduardo Dı́az-Rubio, Werner Scheithauer, Arie Figer, Ralph Wong, Sheryl Koski,
Mikhail Lichinitser, Tsai-Shen Yang, Fernando Rivera, Felix Couture, Florin Sirzén, and Leonard Saltz
A B S T R A C T
Purpose
To evaluate whether capecitabine plus oxaliplatin (XELOX) is noninferior to fluorouracil. folinic acid,
and oxaliplatin (FOLFOX-4) as first-line therapy in metastatic colorectal cancer (MCRC).
Patients and Methods
The initial design of this trial was a randomized, two-arm, noninferiority, phase III comparison of
XELOX versus FOLFOX-4. After patient accrual had begun, the trial design was amended in 2003
after bevacizumab phase III data became available. The resulting 2 ⫻ 2 factorial design randomly
assigned patients to XELOX versus FOLFOX-4, and then to also receive either bevacizumab or
placebo. We report here the results of the analysis of the XELOX versus FOLFOX-4 arms. The
analysis of bevacizumab versus placebo with oxaliplatin-based chemotherapy is reported sepa-
rately. The prespecified primary end point for the noninferiority analysis was progression-
free survival.
Results
The intent-to-treat population comprised 634 patients from the original two-arm portion of the
study, plus an additional 1,400 patients after the start of the amended 2 ⫻ 2 design, for a total of
2,034 patients. The median PFS was 8.0 months in the pooled XELOX-containing arms versus 8.5
months in the FOLFOX-4 – containing arms (hazard ratio [HR], 1.04; 97.5% CI, 0.93 to 1.16). The
median overall survival was 19.8 months with XELOX versus 19.6 months with FOLFOX-4 (HR,
0.99; 97.5% CI, 0.88 to 1.12). FOLFOX-4 was associated with more grade 3/4 neutropenia/
granulocytopenia and febrile neutropenia than XELOX, and XELOX with more grade 3 diarrhea and
grade 3 hand-foot syndrome than FOLFOX-4.
Conclusion
XELOX is noninferior to FOLFOX-4 as a first-line treatment for MCRC, and may be considered as
a routine treatment option for appropriate patients.
J Clin Oncol 26:2006-2012. © 2008 by American Society of Clinical Oncology
aliplatin, has shown an overall response rate (ORR)
INTRODUCTION
of 55% and median overall survival (OS) of 19.5
FOLFOX-4, a bi-weekly schedule of bolus and infu- months in a large phase II study.6
sional fluorouracil (FU)/folinic acid (FA) plus oxali- Based on these phase II results, we started a
platin (Eloxatin; Sanofi-Aventis, Bridgewater, NJ), is two-arm, randomized, noninferiority, phase III
a widely used regimen for first-line treatment of study (XELOX-1; NO16966A) comparing XELOX
metastatic colorectal cancer (MCRC).1 Capecitab- with FOLFOX-4 in the first-line treatment of
ine (Xeloda; Hoffmann-La Roche Inc, Nutley, NJ) is MCRC. After the pivotal phase III data for bevaci-
an oral fluoropyrimidine with similar efficacy to bo- zumab became available,9 the protocol was
lus FU/FA as first-line treatment for MCRC2-4 and amended to a randomized, 2 ⫻ 2 factorial design
as adjuvant therapy for stage III colon cancer.5 Cape- with two coprimary objectives. The primary end
citabine has been tested in combination with oxali- point was progression-free survival (PFS). The first
platin in a variety of different schedules.6-8 XELOX, coprimary objective, which is the focus of this re-
which consists of a 21-day intermittent schedule of port, was to evaluate the noninferiority of XELOX
capecitabine combined with a 3-weekly dose of ox- with or without bevacizumab versus FOLFOX-4
 XELOX v FOLFOX-4 in First-Line Colorectal Cancer
with or without bevacizumab. The second coprimary objective was
to evaluate for the superiority of bevacizumab versus placebo when
combined with oxaliplatin-based chemotherapy (ie, XELOX
or FOLFOX-4), and will be presented as a separate paper.10
PATIENTS AND METHODS
Study Design
NO16966 commenced as a two-arm, open-label, randomized phase III
comparison of XELOX versus FOLFOX-4. The protocol was amended to a
placebo-controlled, double-blind (for bevacizumab), randomized, 2 ⫻ 2 fac-
torial design, and bevacizumab or placebo were added to both treat-
ment groups.
The study was performed in accordance with the Declaration of Helsinki
and Good Clinical Practice Guidelines. Written informed consent was ob-
tained from all patients participating in the study. Approval of the protocol was
obtained from an independent ethics committee or institutional review board
of each site.
Patient Population
Patients age ⱖ 18 years with histologically confirmed colorectal cancer,
unresectable metastatic disease (one or more unidimensionally measurable
lesions), an Eastern Cooperative Oncology Group (ECOG) performance sta-
tus of ⱕ 1 and a life expectancy of longer than 3 months were eligible. No prior
systemic therapy for metastatic disease or prior oxaliplatin or bevacizumab
were allowed. Radiation therapy or surgery for metastatic disease was permit-
ted if completed at least 4 weeks before randomization and if untreated mea-
surable disease remained. Patients were required to have adequate
hematologic/clotting, hepatic, and renal function. Pregnant or breast-feeding
women were excluded. Other key exclusion criteria were: clinically significant
cardiovascular disease; clinically detectable ascites; use of full-dose anticoagu-
lants or thrombolytics; known CNS metastases; serious nonhealing wound,
ulcer or bone fracture; known bleeding diathesis or coagulopathy; and pro-
teinuria ⱖ 500 mg/24 hours.
Treatment Plan
Patients were assigned to treatment using an interactive voice response
system. Randomization was stratified by region, ECOG performance status,
liver as a metastatic site, number of metastatic sites (organs), and alkaline
phosphatase level.
XELOX consisted of a 2-hour intravenous infusion of oxaliplatin 130
mg/m2 on day 1 plus oral capecitabine 1,000 mg/m2 twice daily for 2 weeks in
a 3-week cycle. The first dose of capecitabine was given in the evening of day 1
and the last dose on the morning of day 15. The FOLFOX-4 regimen has been
previously described.1 After amendment of the study protocol, bevacizumab
or placebo was added at 7.5 mg/kg every third week to XELOX, and at 5 mg/kg
every second week to FOLFOX-4. Bevacizumab or placebo was given as a 30-
to 90-minute intravenous infusion on day 1 of each cycle before oxaliplatin.
Treatment was continued until disease progression (PD) or for 48 weeks
(ie, up to 16 cycles of XELOX or 24 cycles of FOLFOX-4), whichever came first
(study treatment phase). Patients who completed the 48-week treatment phase
without PD were eligible to continue treatment until PD in a poststudy treat-
ment phase. Patients whose tumor became operable, and for whom resection
was performed, were allowed to enter the poststudy treatment phase.
If one of the regimen components was discontinued, treatment could be
continued with the remaining components as follows: capecitabine or FU/
FA ⫾ bevacizumab/placebo could be given after discontinuing oxaliplatin;
XELOX, FOLFOX-4, capecitabine, or FU/FA could be given after discontinu-
ing bevacizumab/placebo; bevacizumab/placebo could be given after discon-
tinuing XELOX, FOLFOX-4, capecitabine or FU/FA; or patients were not
permitted to continue on oxaliplatin with or without bevacizumab/placebo.
Assessments
Medical history, physical examination, chest x-ray, ECG, and carcino-
embryonic antigen measurement were performed within 21 days before start-
ing treatment. Assessments of vital signs, ECOG performance status, height,
www.jco.org
Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.
weight, and routine blood analysis (hematology and chemistry) were per-
formed within 7 days of starting treatment. During treatment, physical exam-
ination, hematology, and biochemistry analyses were repeated on day 1 of
every treatment cycle.
Tumor assessments (computed tomography scan, magnetic reso-
nance imaging) were made within 28 days before starting study treat-
ment and repeated after every two XELOX cycles and every three
FOLFOX-4 cycles (ie, every sixth week in both arms and at the end of
treatment). Response Evaluation Criteria in Solid Tumors guidelines11
were used to define all responses. Confirmation of response was re-
quired after a minimum of 4 weeks. Tumor responses were assessed by
investigators and also by an independent response review committee.
After completion of study treatment, patients were followed every 3
months until PD and/or death.
Patients were evaluated for adverse events during therapy and until 28
days after the last study drug dose. Adverse events were graded according to
National Cancer Institute Common Toxicity Criteria, version 3.
Statistical Analysis
The intent-to-treat (ITT) patient population included all patients who
underwent randomization and signed the informed consent form. The eligible
patient population (EPP) was the ITT population minus patients who did not
receive at least one dose of study drug, and those patients who violated major
protocol inclusion/exclusion criteria. As requested by regulatory authorities,
the primary analysis of noninferiority was conducted in this EPP population.
All patients receiving at least one dose of study drug were included in the
safety analysis.
As a first step, the analysis of pooled XELOX-containing compared with
pooled FOLFOX-4 – containing arms was performed. If positive, an interac-
tion test was performed on PFS to check for any interaction between the
different treatment components (FOLFOX-4, XELOX, bevacizumab, nonbe-
vacizumab). Independent of the interaction test, a clinical assessment of treat-
ment effect was also performed. An interaction could be ruled out if the
statistical interaction test was not significant and the clinical assessment re-
vealed no clinically relevant difference. If an interaction was ruled out, the
pooled analysis remained the primary analysis. If an interaction could not be
ruled out, then results in the bevacizumab and nonbevacizumab treatment
subgroups would have had to be considered.
PFS, the primary study end point, was defined as the time from the date
of randomization to the first documentation of disease progression by the
investigators, or death from any cause. In order to retain an experiment-wise
two-sided type I error of 5%, the noninferiority test for PFS was studied with a
two-sided significance level of 2.5% (the remaining 2.5% were used for the
superiority testing). Noninferiority was concluded if the upper limit of the
97.5% CI of the hazard ratio (HR) was ⱕ 1.23. The noninferiority margin of
1.23 corresponds to retention of at least 50% of the benefit that oxaliplatin plus
FU/FA has shown over FU/FA alone in the first-line treatment of MCRC. To
confirm the robustness of the primary analysis, Cox models including key
demographic and baseline characteristics (ie, sex, age, race, use of adjuvant
therapy), and stratification variables were performed on PFS.
For the further assessment of noninferiority, a preplanned exploratory
analysis of XELOX and bevacizumab versus FOLFOX-4 and bevacizumab and
a post hoc exploratory analysis of the XELOX versus FOLFOX-4 arms in the
two-arm part of the study were also performed.
The secondary efficacy end points were OS, ORR, duration of response,
and time to treatment failure. For ORR, noninferiority of pooled XELOX
versus pooled FOLFOX-4 arms was concluded if the lower limit of the two-
sided 97.5% CI for the odds ratio was higher than 0.66. The remaining second-
ary efficacy analyses were considered descriptive only.
PFS, OS, and time to treatment failure were analyzed using Cox models
and presented as Kaplan-Meier estimates with HR and 97.5% CIs. Overall
response was assessed by logistic regression, and presented with 97.5% CIs.
Duration of overall response was presented as Kaplan-Meier estimates with
97.5% CIs.
The analysis of NO16966 was event driven. The final analysis was to be
done when 1,200 PFS events had occurred in the EPP, this approach ensuring
90% power at an ␣ level of 2.5. All the results presented, except OS, are based
© 2008 by American Society of Clinical Oncology 2007
 Cassidy et al

on the main clinical cutoff date (January 31, 2006). In order to provide the
most up-to-date results for OS, data presented are based on a more recent
clinical cutoff date (January 31, 2007).
RESULTS
Patient Population
Between July 2003 and May 2004, 634 patients were randomly
assigned in the two-arm portion of the study. Between February 2004
and February 2005, an additional 1,401 patients were randomly as-
signed in the 2 ⫻ 2 factorial part of the study. One patient, who was
randomly assigned twice, was excluded from the FOLFOX-4-
bevacizumab arm (and not treated) for not providing informed con-
sent, but was later included and treated at another center in the
XELOX and placebo arm.
Overall, 2,034 patients made up the ITT population from the
following regions: Europe (n ⫽ 1,048); Canada (n ⫽ 343); Oceania
(n ⫽ 188); United States (n ⫽ 178); central/eastern Asia (n ⫽ 163);
South America (n ⫽ 65); and South Africa (n ⫽ 49). The EPP popu-
lation comprised a total of 1,904 patients (Fig 1). The safety popula-
tion comprised 1,304 patients in the nonbevacizumab-treated groups
and 694 patients in the bevacizumab-treated groups. Baseline demo-
graphic and clinical characteristics were well balanced between treat-
ment arms (Table 1).
Both a clinically relevant and statistically significant (P ⫽ .7025)
treatment interaction was ruled out. Therefore, the pooled analysis of
all patients treated with XELOX (ie, XELOX, XELOX and placebo,
and XELOX and bevacizumab) versus FOLFOX-4 (ie, FOLFOX-4,
FOLFOX-4 and placebo, FOLFOX-4 and bevacizumab) was the pri-
mary analysis of noninferiority.
The median PFS (ITT analysis) was 8.0 months in the pooled
XELOX-containing arms versus 8.5 months in the FOLFOX-4 –
containing arms (HR, 1.04; 97.5% CI, 0.93 to 1.16), the upper limit of
the 97.5% CI being below the predefined noninferiority margin of
1.23 (Fig 2). Similar results were observed in the EPP with a median
PFS of 7.9 months in the pooled XELOX-containing arms versus 8.5
months in the FOLFOX-4 – containing arms (HR, 1.05; 97.5% CI,
0.94 to 1.18).
The two exploratory analyses were further supportive of the main
study findings. Median PFS with XELOX and bevacizumab was 9.3
months versus 9.4 months with FOLFOX-4 and bevacizumab (HR,
1.01; 97.5% CI, 0.83 to 1.23; ITT). In the two-arm part of the study, the
median PFS with XELOX was 7.3 months versus 7.7 months with
FOLFOX-4 (HR, 0.96; 97.5% CI, 0.80 to 1.16; ITT).
The noninferiority results for PFS were further supported by
subgroup analyses defined by demographic and baseline characteris-
tics and stratification variables (Fig 3). The HRs for PFS in the ITT
population were similar across the different subgroups.

Efficacy Response Rates and Overall Survival

The cutoff dates were January 31, 2006, for PFS and January 31, ORR for the two regimens, as assessed both by investigators
2007, for OS. The median follow-up times at the respective cutoff (XELOX 47% v FOLFOX 48%) and the independent response review
dates were 17.7 and 29.7 months. committee (37% in each arm), were essentially the same (Table 2).

Initial 2-arm design Protocol amendment Revised 2x2 factorial design

Allocation

Randomized (n = 634) Randomized (n = 1401)

ITT Analysis

Allocated to XELOX Allocated to FOLFOX-4 Allocated to XELOX + Allocated to XELOX + Allocated to FOLFOX-4 + Allocated to FOLFOX-4 +
(n = 317) (n = 317) placebo (n = 350) bevacizumab (n = 350) placebo (n = 351) bevacizumab (n = 350)

Included in ITT (n = 317) Included in ITT (n = 317) Included in ITT (n = 350) Included in ITT (n = 350) Included in ITT (n = 351) Included in ITT (n = 349)
Did not receive treatment
No informed consent(n = 1)

Included in EPP population Included in EPP population Included in EPP population Included in EPP population Included in EPP population Included in EPP population
(n = 303) (n = 294) (n = 327) (n = 337) (n = 326) (n = 317)
Excluded (n = 14) Excluded (n = 23) Excluded (n = 23) Excluded (n = 13) Excluded (n = 25) Excluded (n = 33)
EPP Analysis

No measurable lesion (n = 7)
Prior treatment for
advanced disease (n = 6)
Did not receive ≥ 1 dose of
study treatment (n = 1)
No measurable lesion (n = 15)
Prior treatment for advanced
disease (n = 3)
Did not receive ≥ 1 dose of
study treatment (n = 4)
MCRC not histologically
confirmed (n = 1)
Evidence of CNS disease
(n = 1)
Received prior oxaliplatin
(n = 1)
No measurable lesion (n = 9)
Prior treatment for advanced
disease (n = 6)
Did not receive ≥ 1 dose of
study treatment (n = 7)
Severe renal impairment
(n = 1)
No measurable lesion (n = 8)
Prior treatment for
advanced disease (n = 2)
Did not receive ≥ 1 dose of
study treatment (n = 1)
MCRC not histologically
confirmed (n = 1)
Evidence of CNS disease
(n = 1)
No measurable lesion (n = 11)
Prior treatment for advanced
disease (n = 4)
Did not receive ≥ 1 dose of
study treatment (n = 9)
MCRC not histologically
confirmed (n = 1)
Evidence of CNS disease
(n = 1)
Positive pregnancy test (n = 1)
No measurable lesion
(n =13)
Prior treatment for advanced
disease (n = 7)
Did not receive ≥ 1 dose of
study treatment (n =15)
MCRC not histologically
confirmed (n = 2)
No informed consent (n = 1)

Fig 1. CONSORT diagram. XELOX, capacitabine and oxaliplatin; FOLFOX-4, infused fluorouracil, folinic acid, and oxaliplatin; ITT, intent-to-treat; EPP, eligible patient
population; MCRC, metastatic colorectal cancer.

2008 © 2008 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.
 XELOX v FOLFOX-4 in First-Line Colorectal Cancer

Table 1. Baseline Patient Characteristics (intent-to-treat population)

FOLFOX-4 ⫹ FOLFOX-4 ⫹ XELOX ⫹ XELOX ⫹
FOLFOX-4 Placebo Bevacizumab XELOX Placebo Bevacizumab

Characteristic No. % No. % No. % No. % No. % No. %

No. of patients 317 351 349 317 350 350
Sex
Male 204 64 186 53 205 59 194 61 205 59 213 61
Female 113 36 165 47 144 41 123 39 145 41 137 39
Age, years
Median 62 60 60 61 61 61
Range 24-83 26-83 19-82 24-84 18-83 18-86
ECOG performance status
0 163 51 211 60 198 57 160 50 207 59 207 59
1 154 49 138 40 147 43 157 50 143 41 142 41
2 0 0 0 0 0 0 0 0 0 0 1 ⬍1
Primary tumor site
Colorectal 17 5 25 7 28 8 30 9 30 9 32 9
Colon 200 63 232 66 223 64 204 64 233 67 236 67
Rectal 100 32 94 27 98 28 83 26 87 25 82 23
Stage at first diagnosis
Local regional 144 45 141 40 128 37 133 42 138 39 122 35
Metastatic 173 55 210 60 221 63 184 58 212 61 228 65
No. of metastatic sites
0 1 ⬍1 1 ⬍1 1 ⬍1 0 0 0 0 0 0
1 118 37 142 41 150 43 127 40 155 44 134 38
2 121 38 122 35 132 38 106 33 112 32 121 35
3 47 15 65 19 44 13 55 17 58 17 64 18
ⱖ4 30 10 21 6 22 6 29 9 25 7 31 9
Alkaline phosphatase
Abnormal 135 43 147 42 146 42 132 42 149 43 156 45
Normal 182 57 201 58 199 58 183 58 200 57 191 55
Prior adjuvant therapy
No 234 74 266 76 261 75 229 72 259 74 274 78
Yes 83 26 85 24 88 25 88 28 91 26 76 22
Median time from start of adjuvant therapy 613.0 913.0 813.0 687.0 843.0 752.5
to randomization, days

Abbreviations: ECOG, Eastern Cooperative Oncology Group; FOLFOX-4, infused fluorouracil, folinic acid, and oxaliplatin; XELOX, capecitabine and oxaliplatin.

The median OS in the ITT population was 19.8 months in the pooled 0.99 (97.5% CI, 0.80 to 1.23; online-only Fig A1) and in the two-arm
XELOX arms and 19.6 months in the pooled FOLFOX-4 arms, with a part of the study the HR of XELOX versus FOLFOX-4 was 0.90 (97.5%
corresponding HR of 0.99 (97.5% CI, 0.88 to 1.12; Fig 4). The HR of CI, 0.74 to 1.10). The details of OS and other descriptive secondary end
XELOX and bevacizumab versus FOLFOX-4 and bevacizumab was points are presented in Table 2.

Second-Line Therapy
FOLFOX-4/FOLFOX-4+placebo/FOLFOX-4+bevacizumab
There were no major imbalances between the treatment groups
1.0
n = 1017; 826 events with respect to the use of second-line therapy: XELOX-containing
XELOX/XELOX+placebo/XELOX+bevacizumab
Proportion of Patients

0.9 n = 1017; 813 events arms, 51%, and FOLFOX-4 – containing arms, 53%. The most com-
0.8 mon agents used were: irinotecan (41% with FOLFOX-4 v 39% with
HR = 1.04 [97.5% CI: 0.93–1.16] (ITT)
0.7 HR = 1.05 [97.5% CI: 0.94–1.18] (EPP)
0.6
XELOX); FU (29% v 25%); capecitabine (12% v 7%); cetuximab
Upper limit below < 1.23 (non-inferiority margin)
0.5 (11% v 9%); and bevacizumab (5% v 6%).
0.4
0.3
0.2
Safety
0.1 8.0 8.5
The median dose intensities (ratio of dose received to dose
planned) of FU, capecitabine, oxaliplatin, and bevacizumab were ⱖ
0 5 10 15 20 25 30
0.89 in all treatment arms. The median number of cycles administered
Time (months) was 10 in the FOLFOX-4/FOLFOX-4 and placebo groups (range, one
to 24), 12 in the FOLFOX-4-bevacizumab group (range, one to 25),
Fig 2. Progression-free survival (intent-to-treat population [ITT]). FOLFOX-4,
infused fluorouracil, folinic acid, and oxaliplatin; XELOX, capacitabine and oxali- seven in the XELOX/XELOX and placebo groups (range, one to 16),
platin; EPP, eligible patient population; HR, hazard ratio. and eight in the XELOX and bevacizumab group (range, one to 17).

www.jco.org © 2008 by American Society of Clinical Oncology 2009

Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.
 Cassidy et al

n Table 2. Analysis of Efficacy (intent-to-treat population)

FOLFOX-4/FOLFOX-4 ⫹ XELOX/XELOX ⫹
ITT population 2034
Placebo/FOLFOX-4 ⫹ Placebo/XELOX ⫹
End Point Bevacizumab Bevacizumab
ECOG PS 0 1146 No. of patients 1,017 1,017
ECOG PS ≥ 1 881 Primary
Median progression-free 8.5 8.0
survival, months
1 metastatic site 826
Hazard ratio 1.04
≥ 2 metastatic sites 1205 97.5% CI 0.93 to 1.16ⴱ
Secondary
Abnormal alkaline phosphatase 865 Overall response rate, %
Normal alkaline phosphatase 1156 Investigator assessed 48 47
Odds ratio 0.94
97.5% CI 0.77 to 1.15†
Liver metastases No 483 IRC assessed 37 37
Yes 1151 Odds ratio 1.00
97.5% CI 0.81 to 1.23
Female 827 Median overall survival, 19.6 19.8
months‡
Male 1207
Hazard ratio 0.99
97.5% CI 0.88 to 1.12
< 65 years 1294 Median time to treatment 6.3 5.9
≥ 65 years 740 failure, months§
Hazard ratio 1.08
97.5% CI 0.97 to 1.20
No prior adjuvant chemotherapy 1523 Median duration of 7.6 7.5
Prior adjuvant chemotherapy 511 response, months
Hazard ratio 1.00
97.5% CI 0.85 to 1.18
Asian or Pacific Islander 215
White 1759 Abbreviations: FOLFOX-4, fluorouracil, folinic acid, and oxaliplatin; IRC, indepen-
dent response review committee; XELOX, capecitabine and oxaliplatin.
ⴱ
Non-inferiority was concluded if the upper limit of the 97.5% CI was ⬍ 1.23.
0.2 0.4 0.6 1 2 3 4 56 †Non-inferiority was concluded if the lower limit of the 97.5% CI was ⬎ 0.66.
Risk Ratio and 95% CI ‡Cut-off date of January 31, 2007.
§Safety population.

Fig 3. Subgroup analysis of progression-free survival (PFS) according to baseline

demographic and stratification variables (intent-to-treat [ITT] population). n,
number of patients; ECOG, Eastern Cooperative Oncology Group.
For the assessment of safety of XELOX versus FOLFOX-4, pa-
tients in the pooled XELOX/XELOX and placebo and pooled
FOLFOX-4/FOLFOX-4 and placebo arms were compared. For
the assessment of safety with addition of bevacizumab, patients
in XELOX and bevacizumab and FOLFOX-4 and bevacizumab
arms were compared.
A summary of adverse events associated with XELOX versus
FOLFOX-4 by National Cancer Institute Common Toxicity Criteria
grade is presented in Table 3. While the overall rates of grade 3/4
adverse events were fairly similar with both FOLFOX-4 and XELOX,
The addition of bevacizumab did not alter the similarities and
differences in safety profile between XELOX and FOLFOX-4 (online-
only Table A1). The tolerability profile of XELOX and FOLFOX-4 in
the United States versus all patients is shown in online-only Table A2.
Treatment was discontinued due to adverse events in 161 pa-
tients (25%) treated with FOLFOX-4, 170 patients (26%) treated with
XELOX, 104 patients (30%) treated with FOLFOX-4-bevacizumab,
FOLFOX-4/FOLFOX-4+placebo/FOLFOX-4+bevacizumab
n = 1017; 695 events
1.0 XELOX/XELOX+placebo/XELOX+bevacizumab
Proportion of Patients

grade 4 adverse events were more common with FOLFOX-4 (25% 0.9 n = 1017; 692 events
and 12%, respectively), this difference being predominantly due to 0.8
HR = 0.99 [97.5% CI: 0.88–1.12] (ITT)
0.7 HR = 1.00 [97.5% CI: 0.88–1.13] (EPP)
grade 4 neutropenia. 0.6
The rates of some individual adverse events also differed between 0.5
the two regimens. Whereas FOLFOX-4 was associated with more 0.4
0.3
grade 3/4 neutropenia/granulocytopenia (44% v 7%), febrile neutro- 0.2
penia (4.8% v 0.9%), and grade 3/4 venous thromboembolic events 0.1 19.6 19.8
(6.3% v 3.8%) than XELOX, XELOX was associated with more grade
0 5 10 15 20 25 30 35 40
3 diarrhea (19% v 11%) and grade 3 hand-foot syndrome (6% v 1%).
As would be expected, rates of grade 3/4 neurosensory toxicity were Time (months)
similar with both regimens (approximately 17%). Grade 3/4 cardiac
Fig 4. Overall survival (intent-to-treat population [ITT]). FOLFOX-4, infused
disorders were reported in nine FOLFOX-4 recipients (1.4%) and six fluorouracil folinic acid, and oxaliplatin; XELOX, capacitabine and oxaliplatin; EPP,
XELOX recipients (0.9%). eligible patient population; HR, hazard ratio.

2010 © 2008 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.
 XELOX v FOLFOX-4 in First-Line Colorectal Cancer

Table 3. Most Frequent (⬎ 20% of patients) Adverse Events (treatment-related and unrelated)
NCI-CTC Grade

FOLFOX-4/FOLFOX-4 ⫹ Placebo (n ⫽ 649) XELOX/XELOX ⫹ Placebo (n ⫽ 655)

1 2 3 4 1 2 3 4

Adverse Event No. % No. % No. % No. % No. % No. % No. % No. %
All events 627 97 576 89 454 70 161 25 615 94 572 87 449 69 81 12
Nausea 258 40 123 19 33 5 0 0 224 34 153 23 30 5 0 0
Diarrhea 179 28 141 22 70 11 3 ⬍1 171 26 126 19 123 19 9 1
Neutropenia 18 3 79 12 173 27 106 16 16 2 118 18 39 6 6 ⬍1
Fatigue 131 20 110 17 49 8 2 ⬍1 108 16 107 16 33 5 1 ⬍1
Vomiting 141 22 85 13 28 4 0 0 143 22 103 16 34 5 0 0
Paresthesia 165 25 54 8 25 4 0 0 137 21 74 11 31 5 0 0
Stomatitis 162 25 67 10 13 2 0 0 102 16 30 5 8 1 0 0
Anorexia 103 16 59 9 17 3 0 0 108 16 56 9 16 2 0 0
Hand-foot syndrome 45 7 13 2 8 1 — — 105 16 53 8 40 6 — —
Constipation 114 18 50 8 15 2 0 0 97 15 45 7 9 1 0 0
Pyrexia 119 18 45 7 9 1 0 0 82 13 33 5 6 ⬍1 0 0
Abdominal pain 81 12 66 10 24 4 1 ⬍1 75 11 63 10 36 5 0 0
Thrombocytopenia 36 6 92 14 20 3 2 ⬍1 23 4 78 12 39 6 7 ⬍1
Peripheral neuropathy 74 11 33 5 23 4 0 0 73 11 30 5 24 4 0 0
Asthenia 49 8 59 9 22 3 1 ⬍1 38 6 58 9 26 4 2 ⬍1

Abbreviations: FOLFOX-4, infused fluorouracil, folinic acid, and oxaliplatin; NCI-CTC, National Cancer Institute Common Toxicity Criteria; XELOX, capecitabine and
oxaliplatin.

and 109 patients (31%) treated with XELOX and bevacizumab. The
most common reasons for treatment discontinuation were neurosen-
sory toxicity and diarrhea.
Treatment-related mortality up to 28 days after the last treat-
ment dose was documented in 11 patients (1.7%) treated with
FOLFOX-4, 14 patients (2.1%) treated with XELOX, six patients
(1.8%) treated with FOLFOX-4 and bevacizumab, and eight patients
(2.3%) treated with XELOX and bevacizumab. The respective 60-day
all-cause mortality rates were 2.3% (n ⫽ 15), 3.4% (n ⫽ 22), 1.5%
(n ⫽ 5), and 2.5% (n ⫽ 9).
DISCUSSION
This large multinational randomized phase III trial demonstrates that
XELOX is noninferior to FOLFOX-4 in terms of PFS, OS, and ORR in
the first-line treatment of patients with MCRC. With approximately
1,000 patients in each pooled treatment arm, the NO16966 trial is the
largest conducted to date to address this question. The overall finding
of noninferiority was also supported by exploratory analyses con-
ducted in the XELOX and bevacizumab versus FOLFOX-4 and bev-
acizumab subgroups and the original XELOX versus FOLFOX-4
arms. Collectively, these analyses show that the findings of the main
analysis are robust.
Previously, the efficacy and safety of capecitabine as single-agent
therapy was established based on comparisons with bolus regimens of
FU/FA.2-5 However, in current clinical practice, infusional regimens
of FU/FA are often preferred because they confer some benefits in
terms of toxicity and efficacy. This trial compared a capecitabine-
based regimen with a regimen that included a bolus and continuous
infusion of FU, and demonstrated that the efficacy of these two treat-
ment options in this combination therapy context is similar.
The efficacy of capecitabine and oxaliplatin combinations has
been tested in three other smaller phase III trials in the first-line
setting12-14 and in one trial in the second-line setting.15 While the
XELOX regimen was used in two first-line trials,12,14 a modified cape-
citabine and oxaliplatin regimen (CAPOX) was used in the third
first-line trial.13 FOLFOX-6 was used as the comparator regimen in
one trial,12 while two trials used other infusional FU and oxaliplatin
with (FUFOX) or without FA (FUOX) regimens. Ducreux et al12
showed noninferiority of XELOX versus FOLFOX-6 in terms of re-
sponse rate (42 v 46%). Median TTP (9.1 v 9.7 months) and overall
survival (19.5 v 18.8 months) were also similar in the two groups.
Dı́az-Rubio et al14 reported equivalent efficacy with XELOX and
FUOX, with a median TTP of 8.9 months with XELOX and 9.5
months with FUOX. In contrast, in the trial by Porschen et al,13 the
median PFS was 7.1 months with the nonstandard CAPOX regimen
and 8.0 months with FUFOX (HR, 1.17; 95% CI, 0.96 to 1.43; P ⫽
.117); the upper 95% CI exceeded the prospectively defined interval
for noninferiority (1.29). However, the study by Porschen et al13 was
underpowered, and it is possible that an imbalance in restaging inter-
vals in this trial, particularly during the first 20 weeks, may have biased
the results in favor of FUFOX. Finally, Rothenberg et al15 showed that
XELOX is noninferior to FOLFOX-4 in terms of PFS in the second-
line treatment of MCRC.
Safety data from this trial suggest that while the profile of adverse
events associated with both the XELOX and FOLFOX-4 regimens are
similar, there are differences in the rates at which these occur. The
XELOX regimen was associated with more grade 3 diarrhea, whereas
FOLFOX-4 was associated with more grade 3/4 neutropenia/granulo-
cytopenia and febrile neutropenia.
The FOLFOX-4 regimen involves two 22-hour infusions of FU
every 14 days, which require catheter placement and regular visits to

www.jco.org © 2008 by American Society of Clinical Oncology 2011

Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.
 Cassidy et al

the clinic. The FOLFOX-4 regimen was selected as the comparator in
this study for regulatory reasons but is still widely used globally. Al-
though modified FOLFOX regimens (such as mFOLFOX-6) have
replaced FOLFOX-4 in some countries and although these do afford
some greater convenience in terms of one 2-day infusion instead of
two 22-hour infusions, the XELOX regimen requires still fewer
planned office visits, with oxaliplatin administered every 21 days and
capecitabine taken orally. Of note, there was a slight increase in grade
3/4 thromboembolic events with FOLFOX-4, some of which may
have been attributable to catheter use.
In conclusion, XELOX is noninferior to FOLFOX-4, and may be
considered as a standard treatment in the first-line treatment
of MCRC.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS
OF INTEREST
Although all authors completed the disclosure declaration, the following
author(s) indicated a financial or other interest that is relevant to the subject
matter under consideration in this article. Certain relationships marked
with a “U” are those for which no compensation was received; those
relationships marked with a “C” were compensated. For a detailed
description of the disclosure categories, or for more information about
ASCO’s conflict of interest policy, please refer to the Author Disclosure
Declaration and the Disclosures of Potential Conflicts of Interest section in
Information for Contributors.
Employment or Leadership Position: Florin Sirzén, Roche (C)
Consultant or Advisory Role: Jim Cassidy, Roche (C); Stephen Clarke,
Roche (C); Eduardo Dı́az-Rubio, Roche (C), Sanofi-aventis (C); Arie
Figer, Roche (C); Ralph Wong, Roche (C), Sanofi-aventis (C) Stock
Ownership: None Honoraria: Jim Cassidy, Roche; Stephen Clarke,
Roche; Eduardo Dı́az-Rubio, Roche, Sanofi-aventis; Werner Scheithauer,
Roche, Sanofi-aventis; Arie Figer, Roche; Fernando Rivera, Roche; Felix
Couture, Roche Research Funding: Jim Cassidy, Roche; Eduardo
Dı́az-Rubio, Roche, Sanofi-aventis; Werner Scheithauer, Roche,
Sanofi-aventis; Arie Figer, Roche; Fernando Rivera, Roche; Leonard Saltz,
Roche, Genentech Expert Testimony: None Other Remuneration: None
AUTHOR CONTRIBUTIONS
Conception and design: Jim Cassidy, Florin Sirzén
Administrative support: Florin Sirzén
Provision of study materials or patients: Jim Cassidy, Stephen Clarke,
Eduardo Dı́az-Rubio, Werner Scheithauer, Arie Figer, Ralph Wong,
Sheryl Koski, Mikhail Lichinitser, Tsai-Shen Yang, Fernando Rivera,
Felix Couture, Leonard Saltz
Collection and assembly of data: Werner Scheithauer, Arie Figer,
Fernando Rivera, Florin Sirzén
Data analysis and interpretation: Jim Cassidy, Florin Sirzén
Manuscript writing: Jim Cassidy, Stephen Clarke, Eduardo Dı́az-Rubio,
Florin Sirzén
Final approval of manuscript: Jim Cassidy, Eduardo Dı́az-Rubio,
Werner Scheithauer, Arie Figer, Ralph Wong, Sheryl Koski, Mikhail
Lichinitser, Tsai-Shen Yang, Fernando Rivera, Felix Couture, Florin
Sirzén, Leonard Saltz

REFERENCES
1. de Gramont A, Figer A, Seymour M, et al:
Leucovorin and fluorouracil with or without oxalipla-
tin as first-line treatment in advanced colorectal
cancer. J Clin Oncol 18:2938-2947, 2000
2. Van Cutsem E, Twelves C, Cassidy J, et al:
Oral capecitabine compared with intravenous flu-
orouracil plus leucovorin in patients with metastatic
colorectal cancer: Results of a large phase III study.
J Clin Oncol 19:4097-4106, 2001
3. Van Cutsem E, Hoff PM, Harper P, et al: Oral
capecitabine vs intravenous 5-fluorouracil and leuco-
vorin: Integrated efficacy data and novel analyses
from two large, randomized, phase III trials. Br J
Cancer 90:1190-1197, 2004
4. Hoff PM, Ansari R, Batist G, et al: Comparison of
oral capecitabine versus intravenous fluorouracil plus
leucovorin as first-line treatment in 605 patients with
metastatic colorectal cancer: Results of a randomized
phase III study. J Clin Oncol 19:2282-2292, 2001
5. Twelves C, Wong A, Nowacki MP, et al:
Capecitabine as adjuvant treatment for stage III
colon cancer. N Engl J Med 352:2696-2704, 2005
6. Cassidy J, Tabernero J, Twelves C, et al:
XELOX (capecitabine plus oxaliplatin): Active first-
line therapy for patients with metastatic colorectal
cancer. J Clin Oncol 22:2084-2091, 2004
7. Grothey A, Jordan K, Kellner O, et al: Capecit-
abine plus irinotecan (CAPIRI) vs capecitabine plus
oxaliplatin (CAPOX) as first-line therapy of advanced
colorectal cancer (ACRC): Updated results of a ran-
domized phase II trial. Eur J Cancer 1:S90, 2003
(suppl; abstr 295)
8. Scheithauer W, Kornek GV, Raderer M, et
al: Randomized multicenter phase II trial of two
different schedules of capecitabine plus oxalipla-
tin as first-line treatment in advanced colorectal
cancer. J Clin Oncol 21:1307-1312, 2003
9. Hurwitz H, Fehrenbacher L, Novotny W, et al:
Bevacizumab plus irinotecan, fluorouracil, and leuco-
vorin for metastatic colorectal cancer. N Engl J Med
350:2335-2342, 2004
10. Saltz L, Clarke S, Dı́az-Rubio E, et al: Efficacy
and safety of bevacizumab in combination with
oxaliplatin-based chemotherapy as first-line therapy
in metastatic colorectal cancer: A randomized phase
III study. J Clin Oncol doi:10.1200/JCO.2007.14.9930
11. Therasse P, Arbuck SG, Eisenhauer EA, et al:
New guidelines to evaluate the response to treat-
ment in solid tumors: European Organisation for
Research and Treatment of Cancer, National Cancer
Institute of the United States, National Cancer Insti-
tute of Canada. J Natl Cancer Inst 92:205-216, 2000
12. Ducreux M, Bennouna J, Hebbar M, et al:
Efficacy and safety findings from a randomized
phase III study of capecitabine (X) ⫹ oxaliplatin (O)
(XELOX) vs. infusional 5- FU/LV ⫹ O (FOLFOX-6) for
metastatic colorectal cancer (MCRC). J Clin Oncol
25:170s, 2007 (suppl; abstr 4029)
13. Porschen R, Arkenau H-T, Kubicka S, et al:
Capecitabine plus oxaliplatin compared with fluorou-
racil and leucovorin plus oxaliplatin: A randomized
comparison in metastatic colorectal cancer – a final
report of the AIO Colorectal Study Group. J Clin
Oncol 25:4217-4223, 2007
14. Dı́az-Rubio E, Tabernero J, Gómez-España A,
et al: Phase III study of capecitabine plus oxaliplatin
versus continuous infusion fluorouracil plus oxalipla-
tin as first-line therapy in metastatic colorectal can-
cer: Final report of the Spanish Cooperative Group
for the Treatment of Digestive Tumors trial. J Clin
Oncol 25:4224-4230, 2007
15. Rothenberg ML, Navarro M, Butts C, et al:
Phase III trial of capecitabine ⫹ oxaliplatin (XELOX)
vs. 5-fluorouracil (5-FU), leucovorin (LV), and oxali-
platin (FOLFOX4) as 2nd-line treatment for patients
with metastatic colorectal cancer (MCRC). J Clin
Oncol 25:171s, 2007 (abstr 4031)

■ ■ ■
Acknowledgment
The Acknowledgment is included in the full-text version of this article, available online at www.jco.org. It is not included in the PDF
version (via Adobe® Reader®).
Appendix
The Appendix is included in the full-text version of this article, available online at www.jco.org. It is not included in the PDF version
(via Adobe® Reader®).

2012 © 2008 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on October 5, 2008 . For personal use only. No other uses without permission.
Copyright © 2008 by the American Society of Clinical Oncology. All rights reserved.