Articles

FOLFIRI plus cetuximab versus FOLFIRI plus bevacizumab as

first-line treatment for patients with metastatic colorectal
cancer (FIRE-3): a randomised, open-label, phase 3 trial
Volker Heinemann, Ludwig Fischer von Weikersthal, Thomas Decker, Alexander Kiani, Ursula Vehling-Kaiser, Salah-Eddin Al-Batran,
Tobias Heintges, Christian Lerchenmüller, Christoph Kahl, Gernot Seipelt, Frank Kullmann, Martina Stauch, Werner Scheithauer, Jörg Hielscher,
Michael Scholz, Sebastian Müller, Hartmut Link, Norbert Niederle, Andreas Rost, Heinz-Gert Höffkes, Markus Moehler, Reinhard U Lindig,
Dominik P Modest, Lisa Rossius, Thomas Kirchner, Andreas Jung, Sebastian Stintzing

Summary
Background Cetuximab and bevacizumab have both been shown to improve outcomes in patients with metastatic Lancet Oncol 2014; 15: 1065–75
colorectal cancer when added to chemotherapy regimens; however, their comparative effectiveness when partnered Published Online
with first-line fluorouracil, folinic acid, and irinotecan (FOLFIRI) is unknown. We aimed to compare these agents in August 1, 2014
http://dx.doi.org/10.1016/
patients with KRAS (exon 2) codon 12/13 wild-type metastatic colorectal cancer.
S1470-2045(14)70330-4
See Comment page 1040
Methods In this open-label, randomised, phase 3 trial, we recruited patients aged 18–75 years with stage IV,
Department of Medical
histologically confirmed colorectal cancer, an Eastern Cooperative Oncology Group (ECOG) performance status of Oncology & Comprehensive
0–2, an estimated life expectancy of greater than 3 months, and adequate organ function, from centres in Germany Cancer Center, University
and Austria. Patients were centrally randomised by fax (1:1) to FOLFIRI plus cetuximab or FOLFIRI plus bevacizumab Hospital Grosshadern, Munich,
(using permuted blocks of randomly varying size), stratified according to ECOG performance status, number of Germany
(Prof V Heinemann MD,
metastatic sites, white blood cell count, and alkaline phosphatase concentration. The primary endpoint was objective D P Modest MD, L Rossius MD,
response analysed by intention to treat. The study has completed recruitment, but follow-up of participants is ongoing. S Stintzing MD);
The trial is registered with ClinicalTrials.gov, number NCT00433927. Gesundheitszentrum
St Marien, Amberg, Germany
(L Fischer von Weikersthal MD);
Findings Between Jan 23, 2007, and Sept 19, 2012, 592 patients with KRAS exon 2 wild-type tumours were randomly Oncological Practice,
assigned and received treatment (297 in the FOLFIRI plus cetuximab group and 295 in the FOLFIRI plus bevacizumab Ravensburg, Germany
group). 184 (62·0%, 95% CI 56·2–67·5) patients in the cetuximab group achieved an objective response compared (T Decker MD); Department of
Medicine IV, Klinikum Bayreuth
with 171 (58·0%, 52·1–63·7) in the bevacizumab group (odds ratio 1·18, 95% CI 0·85–1·64; p=0·18). Median
GmbH, Bayreuth, Germany
progression-free survival was 10·0 months (95% CI 8·8–10·8) in the cetuximab group and 10·3 months (9·8–11·3) in (Prof A Kiani MD); Oncological
the bevacizumab group (hazard ratio [HR] 1·06, 95% CI 0·88–1·26; p=0·55); however, median overall survival was Practice, Landshut, Germany
28·7 months (95% CI 24·0–36·6) in the cetuximab group compared with 25·0 months (22·7–27·6) in the (U Vehling-Kaiser MD);
Department of Hematology
bevacizumab group (HR 0·77, 95% CI 0·62–0·96; p=0·017). Safety profiles were consistent with the known side-
and Oncology, Krankenhaus
effects of the study drugs. The most common grade 3 or worse adverse events in both treatment groups were Nordwest, Frankfurt/Main,
haematotoxicity (73 [25%] of 297 patients in the cetuximab group vs 62 [21%] of 295 patients in the bevacizumab Germany
group), skin reactions (77 [26%] vs six [2%]), and diarrhoea (34 [11%] vs 40 [14%]). (Prof S-E Al-Batran MD);
Department of Medicine II,
Städtisches Klinikum Neuss,
Interpretation Although the proportion of patients who achieved an objective response did not significantly differ Neuss, Germany
between the FOLFIRI plus cetuximab and FOLFIRI plus bevacizumab groups, the association with longer overall (Prof T Heintges MD);
survival suggests that FOLFIRI plus cetuximab could be the preferred first-line regimen for patients with KRAS Oncological Practice, Münster,
Germany (C Lerchenmüller MD);
exon 2 wild-type metastatic colorectal cancer.
Haematology and Oncology,
Städtisches Klinikum
Funding Merck KGaA. Magdeburg, Magdeburg,
Germany (C Kahl MD);
Oncological Practice, Bad
Introduction cancer.3 However, a survival benefit associated with the Soden, Germany (G Seipelt MD);
Fluorouracil with folinic acid and irinotecan (FOLFIRI) addition of bevacizumab to standard first-line infusional Department of Medicine I,
is a frequently used chemotherapy regimen for the first- fluorouracil-based regimens has not yet been shown. Klinikum Weiden, Germany
line treatment of metastatic colorectal cancer. Results of The FIRE-3 trial of the Arbeitsgemeinschaft (Prof F Kullmann MD);
Oncological Practice, Kronach,
the phase 3 CRYSTAL trial showed that the addition of Internistische Onkologie (AIO) was designed to explore Germany (M Stauch MD);
the EGFR antibody cetuximab to FOLFIRI in this setting whether cetuximab or bevacizumab was a more effective Department of Internal
improved clinical outcome in patients whose tumours partner for FOLFIRI in the fi rst-line treatment of Medicine I and Comprehensive
did not have mutations at KRAS codons 12 and 13.1,2 metastatic colorectal cancer. Patients were initially Cancer Center, Medical
University of Vienna, Vienna,
Addition of the VEGF-A antibody bevacizumab to a first- recruited without regard to KRAS tumour mutation Austria
line regimen including irinotecan, bolus fluorouracil, status. However, following reports that cetuximab was (Prof W Scheithauer MD);
and folinic acid was also shown in a phase 3 trial to not active in patients with tumour KRAS exon 2 Department of Surgery,
improve outcome in patients with metastatic colorectal mutations (codon 12 or 13),4,5 enrolment was restricted Klinikum Chemnitz gGmbH,

www.thelancet.com/oncology Vol 15 September 2014 1065

 Articles
Chirurgische Onkologie,
Chemnitz, Germany
(J Hielscher); Department of
Medicine, Klinikum Stuttgart
Krankenhaus Bad Cannstatt,
Stuttgart, Germany
(M Scholz MD); Oncological
Practice, Ansbach, Germany
(S Müller MD); Department of
Medicine I, Westpfalz-Klinikum
GmbH, Kaiserslautern,
Germany (Prof H Link PhD);
Department of Oncology,
Haematology and Palliative
Care, Klinikum Leverkusen
gGmbH, Leverkusen, Germany
(Prof N Niederle MD); Medical
Clinic V (Haematology and
Oncology), Klinikum
Darmstadt, Darmstadt,
Germany (A Rost MD); Klinikum
Fulda, Tumorklinik, Fulda,
Germany (Prof H-G Höffkes MD);
Medical Department 1,
Johannes-Gutenberg
Universität Mainz, Mainz,
Germany (Prof M Moehler MD);
Klinik für Innere Medizin II,
Abteilung Hämatologie/
Onkologie,
Universitätsklinikum Jena,
Jena, Germany (R U Lindig MD);
and Institute of Pathology,
University of Munich, Munich,
Germany (L Rossius,
Prof T Kirchner MD, A Jung PhD)
Correspondence to:
Prof Volker Heinemann, Klinikum
Grosshadern, University of
Munich, Department of Medical
Oncology and Comprehensive
Cancer Center, D-81377 Munich,
Germany
volker.heinemann@med.uni-
muenchen.de
1066
by protocol amendment to patients without such
mutations. Subsequently, an unplanned subgroup
analysis was done to compare treatment efficacy in
patients with KRAS exon 2 mutations enrolled before
the protocol amendment.6
Particular tumour mutations occurring in exons 3 or 4
of KRAS or exons 2–4 of NRAS have recently also been
shown to be negative predictors of outcome in patients
receiving first-line therapy with the EGFR antibody
panitumumab combined with a regimen of folinic acid,
fluorouracil, and oxaliplatin (FOLFOX4).7,8 When patients
with such mutations were excluded from the KRAS
exon 2 wild-type population, a more pronounced survival
benefit associated with the addition of panitumumab to
FOLFOX4 was reported than before the exclusion of
these patients.7
Our objective in this study was therefore to compare
cetuximab or bevacizumab, plus FOLFIRI, in FIRE-3
trial patients without tumour KRAS exon 2 mutations.
Methods
Study design and patients
In this two-group, open-label, multicentre, randomised,
phase 3 trial, we recruited patients from hospitals,
outpatient clinics, and private practices in Germany and
Austria. Eligible patients were those aged 18–75 years
with stage IV, histologically confirmed, adenocarcinoma
of the colon or rectum, an Eastern Cooperative Oncology
Group (ECOG) performance status of 0–2, an estimated
life expectancy of greater than 3 months, and adequate
organ function (white blood cell count ≥3·0 × 10⁹ cells per L,
with neutrophils ≥1·5 × 10⁹ cells per L, platelets
≥100 × 10⁹ per L and haemoglobin ≥5·6 mmol/L
[corresponding to 9 g/dL]; serum bilirubin ≤1·5 × upper
limit of normal [ULN]; alanine aminotransferase and
aspartate aminotransferase ≤2·5 × ULN or ≤5 × ULN in
the presence of liver metastases, and serum creatinine
≤1·5 × ULN), and who had not had surgery within the
4 weeks before the start of study treatment. The presence
of at least one measurable reference lesion according to
Response Evaluation Criteria in Solid Tumors (RECIST)
version 1.09 was also required. Due to the emerging
evidence on the negative predictive value of KRAS exon 2
mutations10,11 and the subsequent changes to the label of
cetuximab, a protocol amendment was submitted on
Oct 7, 2008, and inclusion was restricted to only patients
with KRAS exon 2 wild-type tumours. Key exclusion
criteria included known or suspected brain metastases;
previous treatment with an EGFR-targeting agent or
bevacizumab; previous chemotherapy for colorectal
cancer, excluding adjuvant therapy completed at least
6 months before trial enrolment; or receipt of any
experimental drug treatment within 30 days before
enrolment. Patients were also excluded if they had
clinically relevant coronary heart disease, myocardial
infarction within the past 12 months or a risk of
uncontrolled arrhythmia; acute or subacute intestinal
obstruction or a history of chronic inflammatory disease
or chronic diarrhoea; symptomatic peritoneal carcino-
matosis; serious, non-healing wounds, ulcers or bone
fractures; uncontrolled hypertension; pronounced
proteinuria (nephrotic syndrome); arterial thrombo-
embolisms or severe haemorrhages within 6 months
before study enrolment (except a bleeding tumour before
tumour resection surgery); haemorrhagic diathesis or
thrombotic tendency; a pre-existing dihydropyrimidine
dehydrogenase deficiency; a pre-existing glucuronidation
defect (Gilbert-Meulengracht syndrome); a history of
secondary malignancy within the past 5 years, except for
basalioma or carcinoma in situ of the cervix if treated
with curative intent; or been receiving therapeutic
anticoagulation therapy.
The sponsor of the study was the Klinikum
Grosshadern, University of Munich. Randomisation,
data management, and primary data analyses were done
by a contract research organisation (ClinAssess GmbH,
Leverkusen, Germany). The trial was done in accordance
with the protocol and in compliance with the Declaration
of Helsinki. The protocol was approved by the ethics
committees of all participating centres. All patients
provided written informed consent before trial entry.
Randomisation and masking
Patients were randomly assigned in a 1:1 ratio to receive
either FOLFIRI plus cetuximab or FOLFIRI plus
bevacizumab. Randomisation was done centrally by fax
using permuted blocks of randomly varying size, with
stratification according to ECOG performance status (0–1
or 2), number of metastatic sites (one or more than one),
white blood cell count (<8 × 10⁹ cells per L or ≥8 × 10⁹ cells
per L) and alkaline phosphatase concentration (<300 units
per L or ≥300 units per L). Neither participants,
physicians giving the intervention, the radiologists
assessing the outcome, nor the statisticians were masked
to group assignment.
Procedures
On day 1 of each 14-day treatment cycle, patients received
either cetuximab (initial dose 400 mg per m² of body-
surface area infused over 120 min, and 250 mg per m² of
body-surface area infused over 60 min weekly, thereafter)
plus FOLFIRI (comprising a 60–90-min infusion of
irinotecan at a dose of 180 mg per m² of body-surface
area; a 120-min infusion of racemic folinic acid at a dose
of 400 mg per m² of body-surface area; fluorouracil as an
intravenous bolus of 400 mg per m² of body-surface area,
and then a continuous 46-h infusion of 2400 mg per m²
of body-surface area) or FOLFIRI with bevacizumab
(5 mg per kg of bodyweight infused initially over 90 min,
2 weeks later over 60 min, and over 30 min every 2 weeks
thereafter with the first dose administered after chemo-
therapy and all subsequent doses administered either
before or after chemotherapy). Treatment was continued
until disease progression, unacceptable toxic effects
www.thelancet.com/oncology Vol 15 September 2014
 Articles

occurred, a complete response was achieved, surgical
resection became possible, or the patient requested or
the physician decided that therapy should be withdrawn.
Dose modifications of chemotherapy and cetuximab
were permitted according to protocol-defined criteria.
Cetuximab therapy was stopped immediately and
permanently if a patient developed a grade 3 or 4 allergic
or hypersensitivity reaction. No recommendations for
reducing the bevacizumab dose were made, although
patients with KRAS exon 2 wild-type tumours who
received at least one dose of study treatment. No reliable
data for FOLFIRI plus bevacizumab were available at the
time of the trial design; however, in a trial of bevacizumab For the study protocol see
plus irinotecan, folinic acid, and bolus fluorouracil (IFL),3 http://www.klinikum.uni-
muenchen.de/CCCLMU-
45% of patients achieved an objective response. We Krebszentrum-Muenchen/
assumed that FOLFIRI would be a better regimen than download/inhalt/studien/fire3/
IFL, therefore we estimated that 50% of patients in our en/FIRE3_EN_translation_
FOLFIRI plus bevacizumab group would achieve an protocolTLO.pdf

treatment could be delayed in the event of related grade 3
toxicities.
After an initial baseline assessment within 14 days of
the start of study treatment, investigators assessed
tumour response by CT after 6 and 12 weeks of treatment
and then every 10 weeks thereafter until a final assess-
ment at the end of treatment. Subsequently, follow-up
assessments were done every 3 months until the patient
died, or for a maximum of 5 years. Adverse events were
recorded continuously from enrolment to the end of the
final study visit and were classified and graded according
to the National Cancer Institute Common Terminology
Criteria for Adverse Events version 3.0.
Tumour mutation status at KRAS exon 2 (codons
12 and 13), exon 3 (codon 61), and exon 4 (codon 146),
and NRAS exon 2 (codons 12 and 13), exon 3 (codons
59 and 61), and exon 4 (codons 117 and 146) was assessed
using a pyrosequencing approach, as described in the
objective response. On the basis of this assumption, we
calculated that 568 randomly assigned patients would
provide the trial with a power of 80% to detect a
significantly higher response rate (62%13,14) in the
FOLFIRI plus cetuximab group at a one-sided Fisher’s
exact test significance level of 2·5%. In a further planned
analysis, response was also assessed in a per-protocol
population, defined as all patients who completed at least See Online for appendix
752 randomly assigned
380 assigned to receive FOLFIRI plus cetuximab 372 assigned to receive FOLFIRI plus bevacizumab
7 terminated study before 10 terminated study before
treatment start treatment start

appendix pp 3–4.
76 with tumour KRAS exon 2 67 with tumour KRAS exon 2
Outcomes mutations or unknown status mutations or unknown status
The primary endpoint was the proportion of patients
who had an objective response (complete or partial 297 without tumour KRAS exon 2 mutations who 295 without tumour KRAS exon 2 mutations who
response) according to RECIST version 1.0, as assessed received study treatment received study treatment
by the study investigators. Secondary endpoints included (intention-to-treat population) (intention-to-treat population)

progression-free survival (time from randomisation to

disease progression or death from any cause), overall 205 assessable for other tumour RAS* 202 assessable for other tumour RAS*
survival (time from randomisation to death from any mutations (RAS-assessable mutations (RAS-assessable
population) population)
cause), depth of remission (maximum percentage 171 wild-type at all loci 171 wild-type at all loci
change in tumour size compared with baseline), 34 mutated 31 mutated
secondary resection of liver metastases with curative
intent, and safety and tolerability. 42 not assessable for response 24 not assessable for response
The proportion of patients who had an objective 1 early death 4 early deaths
response was chosen as the primary endpoint because no 13 allergic reaction 20 other reasons
28 other reasons
data on median overall survival for FOLFIRI in com-
bination with either bevacizumab or cetuximab in the
first-line treatment of metastatic colorectal cancer were 255 without tumour KRAS exon 2 271 without tumour KRAS exon 2
available at the time of the study design. Furthermore, it mutations assessable for response† mutations assessable for response†
(per-protocol population) (per-protocol population)
was expected that cetuximab and bevacizumab would
differ to the greatest extent with respect to this measure
of treatment efficacy. Another consideration was that Patient status at clinical cutoff: Patient status at clinical cutoff:
responses by RECIST might be an appropriate surrogate 158 dead 185 dead
129 alive, treatment terminated 97 alive, treatment terminated
endpoint for overall survival, specifically with regard to 10 alive, on treatment 13 alive, on treatment
an anti-proliferative agent such as cetuximab.12
Figure 1: Study profile
Statistical analysis FOLFIRI=fluorouracil, folinic acid, and irinotecan. *Other RAS mutations investigated included KRAS exons 3 and 4
The primary analysis of response was done in the and NRAS exons 2–4. †Predefined in the study protocol as having received three cycles of chemotherapy and one
intention-to-treat population, defined as all randomised post-baseline CT scan.

www.thelancet.com/oncology Vol 15 September 2014 1067

 Articles

three cycles of study treatment and for whom at least one
radiological examination was done after baseline.
We assessed progression-free and overall survival
using the Kaplan-Meier method15 and compared them
with log-rank tests at a significance level of 5%. We
calculated odds ratios (ORs) and hazard ratios (HRs) for
FOLFIRI plus cetuximab versus FOLFIRI plus
bevacizumab. We assessed toxicity in patients that
received any study treatment. In retrospective analyses,
we further assessed treatment efficacy according to
tumour RAS mutation status.
In additional exploratory analyses, we investigated
differences between treatment groups with respect to
follow-up time, treatment exposure, response categories,
post-study therapy, and the incidence of adverse events
using two-sided Fisher’s exact tests at a significance level
of 5%. We visualised the heterogeneity of treatment
effect for overall survival across subgroups defined by
baseline characteristics using a forest plot. We did all
statistical analyses using SAS version 9.2.
This trial is registered with ClinicalTrials.gov, number
NCT00433927.
Role of the funding source
The funder of the study had no role in study design, data
collection, data analysis, or data interpretation. The
funder had no direct role in the writing of the report, but
supported this process through a writing grant to the
lead investigator (VH), permitting the commissioning of
medical writing services. The funder reviewed the
manuscript before journal submission. The cor-
responding author had full access to all the data in the
study and had the final responsibility for the decision to
submit for publication.

Intention-to-treat population RAS-assessable subgroup of intention-to-treat population

FOLFIRI plus cetuximab FOLFIRI plus bevacizumab FOLFIRI plus cetuximab FOLFIRI plus bevacizumab
(n=297) (n=295) (n=205) (n=202)
Sex
Men 214 (72%) 196 (66%) 145 (71%) 133 (66%)
Women 83 (28%) 99 (34%) 60 (29%) 69 (34%)
Age (years) 64·0 (38·0–79·0) 65·0 (27·0–76·0) 64·0 (41·0–76·0) 65·0 (33·0–76·0)
≤65 158 (53%) 160 (54%) 112 (55%) 105 (52%)
>65 139 (47%) 135 (46%) 93 (45%) 97 (48%)
≥70 90 (30%) 69 (23%) 60 (29%) 50 (25%)
ECOG performance status
0 154 (52%) 158 (54%) 98 (48%) 105 (52%)
1 136 (46%) 133 (45%) 102 (50%) 94 (47%)
2 7 (2%) 4 (1%) 5 (2%) 3 (1%)
Laboratory values
Leucocyte count ≥8 × 10⁹ per L 129 (43%) 118 (40%) 89 (43%) 80 (40%)
Alkaline phosphatase ≥300 U/L 40 (13%) 39 (13%) 25 (12%) 24 (12%)
Site of primary tumour
Colon 168 (57%) 177 (60%) 121 (59%) 120 (59%)
Rectum 115 (39%) 106 (36%) 73 (36%) 75 (37%)
Colon and rectum 9 (3%) 12 (4%) 8 (4%) 7 (3%)
Unknown 5 (2%) 0 3 (1%) 0
Number of metastatic sites
1 119 (40%) 123 (42%) 90 (44%) 90 (45%)
≥2 174 (59%) 171 (58%) 113 (55%) 112 (55%)
Unknown 4 (1%) 1 (<1%) 2 (1%) 0
Metastatic sites
Liver 241 (81%) 240 (81%) 172 (84%) 168 (83%)
Liver only 93 (31%) 94 (32%) 75 (37%) 69 (34%)
Liver not affected 52 (18%) 54 (18%) 31 (15%) 34 (17%)
Previous treatment
Surgery 249 (84%) 252 (85%) 178 (87%) 179 (89%)
Adjuvant chemotherapy 66 (22%) 56 (19%) 35 (17%) 41 (20%)
Radiotherapy 39 (13%) 40 (14%) 19 (9%) 28 (14%)

Data are number (%) or median (range). FOLFIRI=fluorouracil, folinic acid, and irinotecan. ECOG=Eastern Cooperative Oncology Group.

Table 1: Baseline characteristics of the intention-to-treat and RAS-assessable populations

1068 www.thelancet.com/oncology Vol 15 September 2014

 Articles

Results 295 patients in the bevacizumab group had received

Between Jan 23, 2007, and Sept 19, 2012, 752 patients third-line treatment (appendix p 5). However, many
from 116 centres (appendix p 1–2) in Germany patients are still alive and in follow-up, and these data are
(110 centres) and Austria (six centres) were randomly immature.
assigned to a treatment group. 592 patients with KRAS In the primary analysis, complete or partial responses
exon 2 wild-type tumours who received treatment were were reported in 184 (62%) of 297 patients in the
subsequently included in the intention-to-treat popu- cetuximab group and 171 (58%) of 295 patients in the
lation; 297 were assigned to FOLFIRI plus cetuximab and bevacizumab group (OR 1·18, 95% CI 0·85–1·64; p=0·18;
295 to FOLFIRI plus bevacizumab (figure 1). The table 2). The proportion of patients who achieved a
baseline characteristics of the treatment groups were complete response was higher in the cetuximab than in
similar (table 1). With more exclusions in the FOLFIRI the bevacizumab group (table 2). In the predefined per-
plus cetuximab group, the per-protocol population of protocol population, the proportion of patients who
assessable patients consisted of 526 patients; 255 in the achieved a response was significantly higher in the
cetuximab group and 271 in the bevacizumab group. cetuximab group than in the bevacizumab group (184
The data cutoff for the current analysis was April 17, 2013. [72%] of 255 vs 171 [63%] of 271, respectively; p=0·017).
The median duration of follow-up was 33·0 months (IQR Progression-free survival was similar between the
19·0–55·4) for the cetuximab group and 39·0 months treatment groups (figure 2A): median progression-free
(22·5–56·9) for the bevacizumab group. The median survival was 10·0 months (95% CI 8·8–10·8) in the
duration of treatment was 4·8 months (IQR 2·6–7·7) in cetuximab group and 10·3 months (9·8–11·3) in the
the cetuximab group versus 5·3 months (2·8–8·3) in the bevacizumab group (HR 1·06, 95% CI 0·88–1·26;
bevacizumab group regarding the use of all planned study p=0·55). Median overall survival was 28·7 months
agents, and was 6·8 months (3·7–12·4) versus 8·0 months (95% CI 24·0–36·6) in the cetuximab group compared
(4·9–13·1), respectively, regarding any treatment. The with 25·0 months (22·7–27·6) in the bevacizumab group
median number of cycles administered was lower in the (HR 0·77, 95% CI 0·62–0·96; p=0·017; figure 2B).
cetuximab group than in the bevacizumab group (10 Subgroup analyses of overall survival suggested a
[6·0–17·0] vs 12 [6·0–17·0]; p=0·014). particular benefit for patients in the cetuximab group who
Second-line anticancer therapy was administered to had rectal cancer and for those with leucocyte counts of
204 (78%) of 260 living patients in the cetuximab group less than 8 × 10⁹ cells per L (appendix p 7). The proportion
and 191 (76%) of 250 living patients in the bevacizumab of patients who discontinued treatment because they were
group, with a similar proportion of patients in each
treatment group receiving an oxaliplatin-based regimen FOLFIRI plus cetuximab FOLFIRI plus bevacizumab
(130 [64%] of 204 vs 120 [63%] of 191, respectively). The Intention-to-treat population 297 295
proportion of patients crossing over to a regimen Objective response 184 (62%; 56·2–67·5) 171 (58%; 52·1–63·7)
including bevacizumab (95 [47%] of 204 in the cetuximab Complete response 13 (4%) 4 (1%)
group) or an EGFR antibody (79 [41%] of 191 in the Partial response 171 (58%) 167 (57%)
bevacizumab group) was also similar. Similar proportions Stable disease 53 (18%) 85 (29%)
of patients received treatment beyond progression Progressive disease 21 (7%) 16 (5%)
without changing the antibody: 15% (31 of 204 patients) Not evaluable 39 (13%) 23 (8%)
in the cetuximab group and 17% (33 of 191) in the RAS wild-type subgroup 171 171
bevacizumab group. Similar results for second-line
Objective response 112 (65%; 57·9–72·6) 102 (60%; 51·9–67·1)
treatment without antibodies were also noted; 13 (6%) of
Complete response 9 (5%) 2 (1%)
204 of patients in the cetuximab group and 11 (6%) of 191
Partial response 103 (60%) 100 (58%)
in the bevacizumab group received fluorouracil mono-
Stable disease 26 (15%) 50 (29%)
therapy, and 53 (26%) of 204 and 58 (30%) of 191,
Progressive disease 10 (6%) 8 (5%)
respectively, received fluorouracil and oxaliplatin in
Not evaluable 23 (13%) 11 (6%)
combination. Bevacizumab was given either in com-
RAS mutant* subgroup 34 31
bination with fluorouracil (nine [4%] of 204 patients vs
Objective response 13 (38%; 22·2–56·4) 18 (58%; 39·1–75·5)
nine [5%] of 191 patients), fluorouracil and oxaliplatin (60
Complete response 0 2 (6%)
[29%] vs 22 [12%]), or fluorouracil and irinotecan (25 [12%]
Partial response 13 (38%) 16 (52%)
vs one [<1%]). EGFR antibodies (cetuximab or
panitumumab) were given as single agents in ten (5%) of Stable disease 12 (35%) 9 (29%)

204 patients in the cetuximab group and 11 (6%) of 191 in Progressive disease 5 (15%) 1 (3%)

the bevacizumab group, or in combination with Not evaluable 4 (12%) 3 (10%)

irinotecan in four (2%) and 29 (15%) patients, respectively,
or oxaliplatin plus fluorouracil in 13 (6%) and 35 (18%)
patients, respectively. At the time of analysis, 107 (36%) of
297 patients in the cetuximab group and 118 (40%) of
Data are n, n (%; 95% CI) or n (%), unless otherwise specified. FOLFIRI=fluorouracil, folinic acid, and irinotecan. *KRAS
exon 2 wild-type; other RAS mutation in KRAS exon 3 or 4 or NRAS exon 2, 3, or 4.
Table 2: Best overall response to treatment in the intention-to-treat and RAS populations

www.thelancet.com/oncology Vol 15 September 2014 1069

 Articles

A Intention-to-treat population B Intention-to-treat population

100 FOLFIRI plus cetuximab 100
Events Median 95% CI
FOLFIRI plus bevacizumab
158 28·7 months 24·0–36·6
185 25·0 months 22·7–27·6
Events Median 95% CI
75 75
Progression-free survival (%)

250 10·0 months 8·8–10·8

242 10·3 months 9·8–11·3

Overall survival (%)

50 50

25 25

0 0
0 12 24 36 48 60 72 0 12 24 36 48 60 72
Number at risk Number at risk
FOLFIRI plus cetuximab 297 100 19 10 5 3 0 FOLFIRI plus cetuximab 297 218 111 60 29 9 0
FOLFIRI plus bevacizumab 295 99 15 6 4 0 0 FOLFIRI plus bevacizumab 295 214 111 47 18 2 0

C RAS wild-type population D RAS wild-type population

100 100
Events Median 95% CI Events Median 95% CI
144 10·4 months 9·5–12·2 91 33·1 months 24·5–39·4
143 10·2 months 9·3–11·5 110 25·6 months 22·7–28·6

75 75
Progression-free survival (%)

Overall survival (%)

50 50

25 25

0 0
0 12 24 36 48 60 72 0 12 24 36 48 60 72
Time since randomisation (months) Time since randomisation (months)
Number at risk Number at risk
FOLFIRI plus cetuximab 171 64 14 8 4 2 0 FOLFIRI plus cetuximab 171 128 71 39 20 6 0
FOLFIRI plus bevacizumab 171 57 8 3 1 0 0 FOLFIRI plus bevacizumab 171 127 68 26 9 1 0

Figure 2: Kaplan-Meier estimates of progression-free and overall survival

In the intention-to-treat (A and B) and RAS wild-type (C and D) populations, according to treatment group. FOLFIRI=fluorouracil, folinic acid, and irinotecan.

deemed to be eligible for secondary resection of metastatic
disease with curative intent was similar in both treatment
groups (36 [12%] of 297 in the cetuximab group and
40 [14%] of 295 in the bevacizumab group).
In the KRAS exon 2 wild-type population, other RAS
mutations were identified in the tumours of 65 (16%) of
407 patients assessable at all tested loci (figure 1). Baseline
characteristics (table 1) and post-study therapy (data not
shown) were similar for the intention-to-treat and RAS-
assessable treatment groups. In patients with tumours
that were wild-type at all tested RAS loci, the proportions
of patients achieving a response were similar between
groups (OR 1·28, 95% CI 0·83–1·99; p=0·32; table 2);
progression-free survival was also similar between the
treatment groups (median progression-free survival
10·4 months [95% CI 9·5–12·2] in the cetuximab group
vs 10·2 months [9·3–11·5] in the bevacizumab group;
HR 0·93 [95% CI 0·74–1·17]; p=0·54). However, a marked
advantage in overall survival was noted for patients in the
cetuximab group compared with the bevacizumab group
(median overall survival 33·1 months [95% CI 24·5–39·4]
vs 25·6 months [22·7–28·6]; HR 0·70, 95% CI 0·53–0·92;
p=0·011; figure 2C and 2D).
For patients with other tumour RAS mutations (ie,
RAS mutant, but wild-type at KRAS exon 2), there was no
significant difference in the proportion of patients who
had a response to treatment between those in the
cetuximab group and those in the bevacizumab group
(OR 0·45, 95% CI 0·17–1·21; p=0·14; table 2).
31 progression events were noted in the other RAS
mutant population in the cetuximab group, as were 22 in
the other RAS mutant population in the bevacizumab
group; median progression-free survival was 6·1 months
(95% CI 5·3–8·5) in the cetuximab group and

1070 www.thelancet.com/oncology Vol 15 September 2014

 Articles

FOLFIRI plus cetuximab (n=297) FOLFIRI plus bevacizumab (n=295)

Grades 1–2 Grade 3 Grade 4 Grade 5 Grades 1–2 Grade 3 Grade 4 Grade 5
Haematotoxicity* 188 (63%) 62 (21%) 11 (4%) 0 205 (69%) 47 (16%) 14 (5%) 1 (<1%)
Skin reaction†‡ 181 (61%) 72 (24%) 5 (2%) 0 125 (42%) 6 (2%) 0 0
Acneiform exanthema/rash† 180 (61%) 50 (17%) 0 0 23 (8%) 0 0 0
Liver toxicity*† 179 (60%) 18 (6%) 3 (1%) 0 161 (55%) 17 (6%) 2 (<1%) 0
Fatigue* 147 (49%) 2 (<1%) 0 0 158 (54%) 4 (1%) 0 0
Diarrhoea* 136 (46%) 31 (10%) 3 (1%) 0 145 (49%) 33 (11%) 7 (2%) 0
Nausea* 133 (45%) 10 (3%) 0 0 170 (58%) 14 (5%) 0 0
Pain* 133 (45%) 16 (5%) 0 0 150 (51%) 21 (7%) 0 0
Stomatitis* 114 (38%) 11 (4%) 0 0 120 (41%) 10 (3%) 2 (<1%) 0
Infection† 112 (38%) 20 (7%) 5 (2%) 0 117 (40%) 21 (7%) 1 (<1%) 2 (<1%)
Hypocalcaemia 99 (33%) 4 (1%) 1 (<1%) 0 50 (17%) 6 (2%) 1 (<1%) 0
Hypomagnesaemia 93 (31%) 10 (3%) 2 (<1%) 0 41 (14%) 1 (<1%) 1 (<1%) 0
Paronychia* 93 (31%) 14 (5%) 3 (1%) 0 27 (9%) 0 0 0
Alopecia* 88 (30%) 3 (1%) 0 0 106 (36%) 5 (2%) 0 0
Hypokalaemia 87 (29%) 20 (7%) 2 (<1%) 0 46 (16%) 7 (2%) 2 (<1%) 0
Desquamation*† 85 (29%) 19 (6%) 1 (<1%) 0 32 (11%) 2 (<1%) 0 0
Obstipation* 74 (25%) 2 (<1%) 1 (<1%) 0 67 (23%) 3 (1%) 0 0
Hand-foot syndrome*† 69 (23%) 9 (3%) 1 (<1%) 0 40 (14%) 2 (<1%) 0 0
Vomiting* 66 (22%) 7 (2%) 0 0 87 (29%) 10 (3%) 0 0
Polyneuropathy† 63 (21%) 0 0 0 65 (22%) 1 (<1%) 0 0
Bleeding/haemorrhage† 61 (21%) 2 (<1%) 0 0 83 (28%) 1 (<1%) 0 0
Oedema* 49 (16%) 3 (1%) 0 0 27 (9%) 1 (<1%) 0 0
Hypertonia* 44 (15%) 18 (6%) 1 (<1%) 0 93 (32%) 20 (7%) 0 0
Fever (without neutropenia grade 3–4)* 43 (14%) 2 (<1%) 0 0 42 (14%) 0 1 (<1%) 0
Nephrotoxicity* 40 (13%) 4 (1%) 0 0 56 (19%) 3 (1%) 1 (<1%) 0
Decreased appetite 39 (13%) 4 (1%) 0 0 38 (13%) 3 (1%) 0 0
Weight decreased 30 (10%) 2 (<1%) 0 0 36 (12%) 1 (<1%) 0 0
Thrombosis (any)† 10 (3%) 18 (6%) 0 0 16 (5%) 17 (6%) 1 (<1%) 0
Thromboembolic event 7 (2%) 8 (3%) 7 (2%) 0 4 (1%) 7 (2%) 9 (3%) 1 (<1%)
Infusional-related allergic reaction† 11 (4%) 8 (3%) 4 (1%) 0 1 (<1%) 0 0 0
Hyperglycaemia 3 (1%) 6 (2%) 0 0 2 (<1%) 2 (<1%) 0 0
Infection with neutropenia 3 (1%) 4 (1%) 2 (<1%) 0 9 (3%) 6 (2%) 1 (<1%) 2 (<1%)
Syncope 1 (<1%) 4 (1%) 2 (<1%) 0 1 (<1%) 1 (<1%) 1 (<1%) 0
Dyspnoea 17 (6%) 4 (1%) 1 (<1%) 0 17 (6%) 2 (<1%) 1 (<1%) 0
Fever (with neutropenia grade 3–4, without clinical 0 4 (1%) 1 (<1%) 0 6 (2%) 2 (<1%) 1 (<1%) 0
infection)*
Ileus 0 2 (<1%) 3 (1%) 0 1 (<1%) 2 (<1%) 0 0
Anaphylactic reaction 1 (<1%) 1 (<1%) 3 (1%) 0 0 0 0 0
General physical health deterioration 0 3 (1%) 1 (<1%) 0 3 (1%) 0 0 0
Hypotension 10 (3%) 4 (1%) 0 0 3 (1%) 1 (<1%) 0 0
γ-glutamyltransferase increased 3 (1%) 2 (<1%) 1 (<1%) 0 2 (<1%) 3 (1%) 0 0
Anaemia 1 (<1%) 2 (<1%) 0 0 1 (<1%) 0 0 0
Blood product transfusion 0 2 (<1%) 0 0 0 0 0 0
Dehydration 2 (<1%) 2 (<1%) 0 0 4 (1%) 3 (1%) 1 (<1%) 0
Diabetes mellitus 3 (1%) 2 (<1%) 0 0 1 (<1%) 0 0 0
Hyponatraemia 10 (3%) 2 (<1%) 0 0 9 (3%) 3 (1%) 0 0
Abscesses fistulae 3 (1%) 1 (<1%) 0 0 13 (4%) 2 (<1%) 1 (<1%) 0
Activated partial thromboplastin time prolonged 2 (<1%) 1 (<1%) 0 0 1 (<1%) 1 (<1%) 0 0
Arrhythmias* 7 (2%) 0 1 (<1%) 0 7 (2%) 4 (1%) 0 1 (<1%)
Ascites 2 (<1%) 1 (<1%) 0 0 0 0 0 0
Blood lactate dehydrogenase increased 2 (<1%) 1 (<1%) 0 0 4 (1%) 0 0 0
(Table 3 continues on next page)

www.thelancet.com/oncology Vol 15 September 2014 1071

 Articles

FOLFIRI plus cetuximab (n=297) FOLFIRI plus bevacizumab (n=295)

Grades 1–2 Grade 3 Grade 4 Grade 5 Grades 1–2 Grade 3 Grade 4 Grade 5
(Continued from previous page)
Bronchospasm 0 1 (<1%) 0 0 0 0 0 0
Carcinoembryonic antigen increased 0 0 1 (<1%) 0 0 0 0 0
Cardial-ischemic event 0 0 1 (<1%) 0 0 1 (<1%) 1 (<1%) 0
Cardiomyopathy 0 1 (<1%) 0 0 0 0 0 0
Cholangitis 0 0 1 (<1%) 0 0 1 (<1%) 0 0
Convulsion 0 1 (<1%) 0 0 1 (<1%) 0 0 0
Device dislocation 0 1 (<1%) 0 0 0 0 0 0
Device leakage 0 1 (<1 %) 0 0 0 0 0 0
Device related infection 0 1 (<1%) 0 0 0 1 (<1%) 0 0
Dizziness 29 (10%) 0 1 (<1%) 0 31 (11%) 0 0 0
Electrolyte imbalance 0 1 (<1%) 0 0 0 0 0 0
Fracture 0 1 (<1%) 0 0 0 0 0 0
Gastrointestinal perforation 0 1 (<1%) 0 0 0 2 (<1%) 0 0
Hemiplegia 0 1 (<1%) 0 0 0 0 0 0
Hepatic failure 0 0 1 (<1%) 0 0 0 0 0
Hyperkalaemia 13 (4%) 1 (<1%) 0 0 18 (6%) 2 (<1%) 0 0
International normalised ratio decreased 5 (2%) 1 (<1%) 0 0 3 (1%) 1 (<1%) 0 0
Iron deficiency anaemia 0 1 (<1%) 0 0 0 0 0 0
Neoplasm malignant 0 1 (<1%) 0 0 0 0 0 0
Nervous system disorder 0 1 (<1%) 0 0 0 0 0 0
Neurotoxicity (motoric)* 9 (3%) 0 1 (<1%) 0 8 (3%) 3 (1%) 0 0
Neutropenia 3 (1%) 1 (<1%) 0 0 2 (<1%) 0 0 0
Pleurisy 0 1 (<1 %) 0 0 1 (<1%) 0 0 0
Ureteric stenosis 0 1 (<1%) 0 0 0 0 0 0
Urinary retention 1 (<1%) 1 (<1%) 0 0 0 0 0 0
Visual impairment 3 (1%) 1 (<1%) 0 0 4 (1%) 0 0 0
Weight increased 7 (2%) 1 (<1%) 0 0 12 (4%) 3 (1%) 0 0
Wound healing complications 5 (2%) 1 (<1%) 0 0 4 (1%) 4 (1%) 0 0
Abdominal wall haematoma 0 0 0 0 0 0 1 (<1%) 0
Amenorrhoea 0 0 0 0 0 1 (<1%) 0 0
Cardiac failure 0 0 0 0 0 3 (1%) 0 0
Cerebral ischaemia 0 0 0 0 0 0 1 (<1%) 0
Dyspnoea exertional 5 (2%) 0 0 0 8 (3%) 1 (<1%) 0 0
Flatulence 7 (2%) 0 0 0 6 (2%) 1 (<1%) 0 0
Hemiparesis 0 0 0 0 1 (<1%) 1 (<1%) 0 0
Hyperhidrosis 9 (3%) 0 0 0 10 (3%) 1 (<1%) 0 0
Hypermagnesaemia 3 (1%) 0 0 0 3 (1%) 1 (<1%) 0 0
Hypernatraemia 0 0 0 0 1 (<1%) 0 1 (<1%) 0
Inflammatory marker increased 1 (<1%) 0 0 0 0 1 (<1%) 0 0
Leucopenia 2 (<1%) 0 0 0 1 (<1%) 0 1 (<1%) 0
Ligament rupture 0 0 0 0 0 1 (<1%) 0 0
Medical device complication 2 (<1%) 0 0 0 1 (<1%) 1 (<1%) 0 0
Mental disorder 24 (8%) 0 0 0 22 (7%) 3 (1%) 0 0
Metastasis 0 0 0 0 0 1 (<1%) 0 0
Oncological complication 0 0 0 0 0 0 1 (<1%) 0
Proteinuria 8 (3%) 0 0 0 5 (2%) 1 (<1%) 0 0
Sepsis 0 0 0 0 0 0 1 (<1%) 1 (<1%)
Speech disorder 1 (<1%) 0 0 0 1 (<1%) 1 (<1%) 0 0
Subileus 1 (<1%) 0 0 0 2 (<1%) 2 (<1%) 0 0
(Table 3 continues on next page)

1072 www.thelancet.com/oncology Vol 15 September 2014


FOLFIRI plus cetuximab (n=297)
Grades 1–2 Grade 3
(Continued from previous page)
Tumour perforation 0 0
Sixth nerve paralysis 0 0
*Adverse event terms pre-printed in the case report forms (all other adverse events were coded by Medical Dictionary for Regulatory Activities version 13.1 preferred terms). †Composite categories ‡Here, all skin
reactions are included, irrespective of whether they are already reported as separate preferred term.
Table 3: Adverse events
12·2 months (9·7–13·9) in the bevacizumab group
(HR 2·22, 95% CI 1·28–3·86; p=0·004). 22 deaths
occurred in the other RAS mutant population in the
cetixumab group as did 17 in the other RAS mutant
population in the bevacizumab group; median overall
survival was 16·4 months (95% CI 15·9–27·6) in the
cetixumab group versus 20·6 months (17·0–28·4) in the
bevacizumab group (HR 1·20, 95% CI 0·64–2·28;
p=0·57). In the combined population of patients with any
tumour RAS mutation, more patients in the bevacizumab
group had a response than did those in the cetuximab
group, and median progression-free survival was longer,
but these differences were not significant (appendix
pp 6, 8); overall survival was much the same between
treatment groups (appendix pp 6, 9).
The safety profiles in both treatment groups were
consistent with the known side-effects of the individual
study drugs (table 3). The incidence of grade 3 or worse
adverse events was similar between treatment groups,
noted in 211 (71%) of 297 patients in the cetuximab group
and 188 (64%) of 295 in the bevacizumab group. More
patients in the cetuximab group (46 [15%] of 297) than in
the bevacizumab group (31 [11%] of 295) discontinued
treatment due to drug-related toxicity. Five deaths, all
during treatment with FOLFIRI plus bevacizumab, were
reported to be related to adverse events: arrhythmia,
sepsis, thromboembolic event (one patient each), and
infection with neutropenia (two patients). Two of these
deaths (arrhythmia, and infection with neutropenia)
were deemed to be related to study treatment.
Discussion
To our knowledge, this is the first randomised phase 3
trial to compare the efficacy of FOLFIRI plus cetuximab
and FOLFIRI plus bevacizumab in the first-line
treatment of KRAS exon 2 wild-type metastatic colorectal
cancer (panel). In the primary analysis, the difference in
the proportion of patients who achieved an objective
response between the treatment groups was not
statistically significant; both targeted agents seemed to
be equally effective in terms of progression-free survival
when combined first-line with FOLFIRI. However,
overall survival was significantly longer in the FOLFIRI
plus cetuximab group than in the FOLFIRI plus
bevacizumab group, with a clinically meaningful
difference in median survival.
www.thelancet.com/oncology Vol 15 September 2014
Articles
FOLFIRI plus bevacizumab (n=295)
Grade 4 Grade 5 Grades 1–2 Grade 3 Grade 4 Grade 5
0 0 0 0 1 (<1%) 0
0 0 0 1 (<1%) 0 0
Clearly, secondary endpoints are exploratory measures
of outcome and results should be interpreted accordingly.
We note also that for overall survival, the number of
events is small, and confidence intervals are wide.
However, the pattern of efficacy outcomes in the FIRE-3
study is very similar to that reported for the randomised
phase 2 PEAK study,16 which compared the EGFR
antibody panitumumab with bevacizumab when added
to FOLFOX6 as first-line therapy. Overall survival for
patients with KRAS exon 2 wild-type tumours was
significantly longer in the FOLFOX6 plus panitumumab
group than in the FOLFOX6 plus bevacizumab group,
despite the absence of marked differences in progression-
free survival and objective response.16 By contrast, the
US-based randomised phase 3 CALGB 80405 study,
which compared first-line cetuximab or bevacizumab in
combination with FOLFOX or FOLFIRI, showed that
overall survival (the primary endpoint) was similar
Panel: Research in context
Systematic review
Randomised trials have shown that the addition of either the EGFR antibody cetuximab or
the VEGF-A antibody bevacizumab to first-line chemotherapy regimens improved clinical
outcome in patients with metastatic colorectal cancer.1–3 However, as confirmed by
electronic searches of PubMed using the terms “cetuximab”, “bevacizumab”, “colorectal”,
and “FOLFIRI” for articles published in English, with no restrictions on publication date
(last search was Feb 24, 2014), we found no articles reporting which of these two
antibodies provided the most patient benefit when added to the standard first-line
FOLFIRI regimen (fluorouracil, folinic acid, and irinotecan). Our phase 3 trial investigated
FOLFIRI plus cetuximab versus FOLFIRI plus bevacizumab as first-line treatment for
patients with KRAS exon 2 wild-type metastatic colorectal cancer. In an exploratory
retrospective analysis, we also investigated treatment outcome in patients with tumours
wild-type at further loci within KRAS and NRAS (RAS wild-type).
Interpretation
In relation to the primary endpoint, we noted that the proportion of patients who
achieved an objective response with FOLFIRI plus cetuximab in the intention-to-treat
population was not significantly higher than that achieved with FOLFIRI plus
bevacizumab. Although progression-free survival was also similar between the treatment
groups, overall survival was significantly longer for patients in the FOLFIRI plus cetuximab
group than for those in the FOLFIRI plus bevacizumab group. In the subgroup of patients
with RAS wild-type tumours, the overall survival benefit in favour of the FOLFIRI plus
cetuximab group was even more marked. Our data suggest that FOLFIRI plus cetuximab
should be the chosen first-line treatment regimen for patients with RAS wild-type
metastatic colorectal cancer.
1073
 Articles
between treatment groups in an initial analysis of the
KRAS exon 2 wild-type population.17 Most of the patients
in CALGB 80405 received first-line FOLFOX, so a
possible selection bias towards this regimen versus the
FOLFIRI regimen cannot be excluded. CALGB 80405
also allowed the inclusion of patients with locally
advanced but unresectable disease, whereas FIRE-3 was
restricted to patients with metastatic colorectal cancer.
Furthermore, data for treatment outcome in the RAS
wild-type population of CALGB 80405 have not yet been
presented. In view of these factors, any of which could
affect the apparent relative efficacy of the targeted agents,
it is difficult to draw conclusions from a cross-trial
comparison of FIRE-3 and CALGB 80405 outcome data.
In metastatic colorectal cancer, typically only 70–80% of
the first-line treatment population is exposed to later
lines of therapy. First-line treatment can therefore be
considered to be the most important phase of therapy.
Furthermore, the effects of first-line treatment on patient
outcomes might be greater than in any other line. For
example, even in trials with intensive second-line
regimens, absolute improvements in median overall
survival tend to be only slight.18–20
Several possible explanations exist as to why we noted a
significant difference in overall survival between the
treatment groups in our trial but noted no apparent
difference in progression-free survival or response. First,
the trial was open-label and did not include an
independent radiological review of response data, which
might have contributed to some bias in assessments.
Second, although the number of patients receiving
second-line therapy in each treatment group was similar,
and although the number of patients crossing over to the
alternative anti-VEGF or anti-EGFR therapy was similar,
the sequencing of targeted agents for patients in the
treatment groups was in many cases reversed. The
difference in overall survival between treatment groups
could therefore perhaps be related to changes in tumour
biology during first-line therapy. In this context, we note
that upregulation of VEGF in association with resistance
to cetuximab has been reported in experimental
models.21,22 In the clinical setting, such phenotypic
changes could favour second-line anti-VEGF treatment
after first-line cetuximab therapy. Additionally, less than
half (79 [41%] of 191) of patients in the FOLFIRI plus
bevacizumab group received second-line EGFR antibody
therapy, whereas this therapy was offered to all patients
included in the FOLFIRI plus cetuximab group.
Application of an active agent in a population selected for
potential activity of this agent might therefore be crucial.
Another consideration is that response to therapy might
not be captured adequately by RECIST when different
strategies of targeted therapy are used. This issue will be
addressed by an ongoing central review of imaging in
our trial, which will also assess RECIST-independent
parameters such as depth of response. These analyses
will be reported at a later date. To explore further the
1074
factors that might have affected overall survival in our
study, a clinical trial in which the intention-to-treat
population is randomly assigned to different sequential
treatment strategies might be informative.
Patient selection to favour possible sensitivity to anti-
EGFR agents was further examined by excluding patients
with tumour mutations in KRAS exons 3 or 4 or NRAS
exons 2, 3, or 4 from the intention-to-treat population.
For this all wild-type RAS population, the overall survival
advantage in the FOLFIRI plus cetuximab group was
even more pronounced than in the intention-to-treat
population of patients with KRAS exon 2 wild-type
tumours. Response and progression-free survival were
again similar between the treatment groups. By contrast,
patients with tumour RAS mutations other than KRAS
exon 2 seemed to derive a greater clinical benefit from
FOLFIRI plus bevacizumab than FOLFIRI plus
cetuximab. Median overall survival in the FOLFIRI plus
bevacizumab group was much the same in the intention-
to-treat population and the population of patients who
were wild-type at all tested RAS loci, providing further
support for the concept that the effect on overall survival
was generated by first-line use of cetuximab together
with FOLFIRI, rather than by later lines of treatment.
The overall survival advantage seen in the RAS wild-
type population in favour of FOLFIRI plus cetuximab
over FOLFIRI plus bevacizumab (HR 0·70, 95% CI
0·53–0·92; p=0·011) was in line with the overall survival
difference in the RAS wild-type population of the PEAK
study in favour of FOLFOX6 plus panitumumab over
FOLFOX6 plus bevacizumab (HR 0·63, 95% CI
0·39–1·02; p=0·058).16 Similar data emphasising the
importance of extended RAS analysis in relation to the
efficacy of anti-EGFR therapy were presented for
cetuximab in retrospective analyses of the CRYSTAL23
and OPUS24 trials, as well as for panitumumab in a
retrospective analysis of the PRIME study.7,8
In conclusion, results of this head-to-head trial did not
show that FOLFIRI plus cetuximab was associated with a
significantly higher proportion of objective responses than
was FOLFIRI plus bevacizumab in the first-line treatment
of patients with KRAS exon 2 wild-type or RAS wild-type
metastatic colorectal cancer. Although progression-free
survival was similar between treatment groups, overall
survival was markedly longer in the FOLFIRI plus
cetuximab group, especially for patients with RAS wild-
type tumours. Overall, the data suggest that adding
cetuximab rather than bevacizumab to first-line FOLFIRI
will provide the most benefit to these patients.
Contributors
VH, UV-K, HL, NN, H-GH, and SS were responsible for the study design.
VH, LFvW, TD, AK, UV-K, S-EA-B, TH, CL, CK, GS, FK, MSt, WS, JH,
MSc, SM, HL, NN, AR, MM, RUL, DPM, TK, and SS collected the data.
VH, TD, UV-K, HL, and MM analysed the data. VH, TD, UV-K, S-EA-B,
TH, FK, WS, MSc, HL, H-GH, MM, DPM, TK, and SS interpreted the data.
VH, UV-K, S-EA-B, WS, HL, H-GH, DPM, and SS wrote the manuscript.
LFvW is a member of the study protocol committee. LFvW, AK, UV-K,
S-EA-B, TH, CK, MSt, MSc, SM, NN, AR, and MM recruited patients.
www.thelancet.com/oncology Vol 15 September 2014

LFvW, AK, UV-K, S-EA-B, CL, CK, FK, and AJ reviewed the manuscript. AK
and MM treated patients. LR did practical work in the laboratory, including
gene sequencing and pyrogram analysis. AJ designed the mutation
analyses, supervised the practical work, and analysed pyrograms.
Declaration of interests
VH has received financial grants to undertake this study and prepare the
manuscript from Merck KGaA; honoraria for talks and participation in
advisory boards from Merck; and financial grants to undertake clinical
studies, honoraria for talks, and participation in advisory boards from
Roche, Amgen, and Sanofi. FK has received lecture honoraria from
Merck KGaA. WS has received lecture honoraria from Merck and
honoraria for lectures and advisory board meetings from Roche. AR has
received research funding paid to the study team from Merck Serono.
MM has received honoraria for lectures or presentations from Merck
and Roche. DPM has received research grants and fees from Merck and
Roche. TK has received research grants from Roche, Amgen, and Merck
Serono; has received honoraria for lectures and speakers’ bureaus from
Amgen, Merck Serono, and Roche-Ventana; and owns stock options in
Roche and Novartis. AJ has received honoraria for lectures and advisory
boards from Merck Serono and Amgen. SS has received personal fees
from Merck KGaA, Roche AG, Amgen GmbH, and Sanofi-Aventis. All
other authors declare no potential competing interests.
Acknowledgments
This study was funded by Merck KGaA. Jim Heighway of Cancer
Communications and Consultancy Ltd, Knutsford, UK, provided medical
writing services, which included initial drafting of the manuscript under
the guidance of VH and subsequent modification according to direction
from other authors.
References
1 Van Cutsem E, Kohne CH, Hitre E, et al. Cetuximab and
chemotherapy as initial treatment for metastatic colorectal cancer.
N Engl J Med 2009; 360: 1408–17.
2 Van Cutsem E, Kohne CH, Lang I, et al. Cetuximab plus irinotecan,
fluorouracil, and leucovorin as first-line treatment for metastatic
colorectal cancer: updated analysis of overall survival according to
tumor KRAS and BRAF mutation status. J Clin Oncol 2011;
29: 2011–19.
3 Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus
irinotecan, fluorouracil, and leucovorin for metastatic colorectal
cancer. N Engl J Med 2004; 350: 2335–42.
4 Bokemeyer C, Bondarenko I, Hartmann JT, et al. KRAS status and
efficacy of first-line treatment of patients with metastatic colorectal
cancer (mCRC) with FOLFOX with or without cetuximab: the OPUS
experience. Proc Am Soc Clin Oncol 2008; 26 (suppl): 4000 (abstr).
5 Van Cutsem E, Lang I, D’haens G, et al. KRAS status and efficacy in
the first-line treatment of patients with metastatic colorectal cancer
(mCRC) treated with FOLFIRI with or without cetuximab: the
CRYSTAL experience. Proc Am Soc Clin Oncol 2008;
26 (suppl): 2 (abstr).
6 Stintzing S, Fischer von Weikersthal L, Decker T, et al. FOLFIRI plus
cetuximab versus FOLFIRI plus bevacizumab as first-line treatment
for patients with metastatic colorectal cancer-subgroup analysis of
patients with KRAS: mutated tumours in the randomised German
AIO study KRK-0306. Ann Oncol 2012; 23: 1693–99.
7 Douillard JY, Oliner KS, Siena S, et al. Panitumumab-FOLFOX4
treatment and RAS mutations in colorectal cancer. N Engl J Med
2013; 369: 1023–34.
8 Douillard JY, Siena S, Cassidy J, et al. Final results from PRIME:
randomized phase 3 study of panitumumab with FOLFOX4 for
first-line treatment of metastatic colorectal cancer. Ann Oncol 2014;
25: 1346–55.
9 Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to
evaluate the response to treatment in solid tumors. European
Organization for Research and Treatment of Cancer, National
Cancer Institute of the United States, National Cancer Institute of
Canada. J Natl Cancer Inst 2000; 92: 205–16.
www.thelancet.com/oncology Vol 15 September 2014
Articles
10 De Roock W, Claes B, Bernasconi D, et al. Effects of KRAS, BRAF,
NRAS, and PIK3CA mutations on the efficacy of cetuximab plus
chemotherapy in chemotherapy-refractory metastatic colorectal
cancer: a retrospective consortium analysis. Lancet Oncol 2010;
11: 753–62.
11 Sartore-Bianchi A, Di Nicolantonio F, Nichelatti M, et al. Multi-
determinants analysis of molecular alterations for predicting
clinical benefit to EGFR-targeted monoclonal antibodies in
colorectal cancer. PLoS One 2009; 4: e7287.
12 Macedo LT, da Costa Lima AB, Sasse AD. Addition of bevacizumab
to first-line chemotherapy in advanced colorectal cancer: a
systematic review and meta-analysis, with emphasis on
chemotherapy subgroups. BMC Cancer 2012; 12: 89.
13 Folprecht G, Lutz MP, Schoffski P, et al. Cetuximab and
irinotecan/5-fluorouracil/folinic acid is a safe combination for the
first-line treatment of patients with epidermal growth factor
receptor expressing metastatic colorectal carcinoma. Ann Oncol
2006; 17: 450–56.
14 Raoul JL, Van Laethem JL, Peeters M, et al. Cetuximab in
combination with irinotecan/5-fluorouracil/folinic acid (FOLFIRI)
in the initial treatment of metastatic colorectal cancer: a multicentre
two-part phase I/II study. BMC Cancer 2009; 9: 112.
15 Kaplan EL, Meier P. Nonparametric estimation from incomplete
observations. J Am Stat Assoc 1958; 53: 457–81.
16 Schwartzberg LS, Rivera F, Karthaus M, et al. PEAK: a randomized,
multicenter phase II study of panitumumab plus modified
fluorouracil, leucovorin, and oxaliplatin (mFOLFOX6) or
bevacizumab plus mFOLFOX6 in patients with previously
untreated, unresectable, wild-type KRAS exon 2 metastatic
colorectal cancer. J Clin Oncol 2014; 32: 2240–47.
17 Venook AP, Niedzwiecki D, Lenz HJ, et al. CALGB/SWOG 80405:
phase III trial of irinotecan/5-FU/leucovorin (FOLFIRI) or
oxaliplatin/5-FU/leucovorin (mFOLFOX6) with bevacizumab (BV)
or cetuximab (CET) for patients (pts) with KRAS wild-type (wt)
untreated metastatic adenocarcinoma of the colon or rectum
(MCRC). Proc Am Soc Clin Oncol 2014; 32 (suppl): LBA3 (abstr).
18 Bennouna J, Sastre J, Arnold D, et al, for the ML18147 Study
Investigators. Continuation of bevacizumab after first progression
in metastatic colorectal cancer (ML18147): a randomised phase 3
trial. Lancet Oncol 2013; 14: 29–37.
19 Giantonio BJ, Catalano PJ, Meropol NJ, et al. Bevacizumab in
combination with oxaliplatin, fluorouracil, and leucovorin
(FOLFOX4) for previously treated metastatic colorectal cancer:
results from the Eastern Cooperative Oncology Group Study E3200.
J Clin Oncol 2007; 25: 1539–44.
20 Van Cutsem E, Tabernero J, Lakomy R, et al. Addition of aflibercept
to fluorouracil, leucovorin, and irinotecan improves survival in a
phase III randomized trial in patients with metastatic colorectal
cancer previously treated with an oxaliplatin-based regimen.
J Clin Oncol 2012; 30: 3499–506.
21 Ciardiello F, Bianco R, Caputo R, et al. Antitumor activity of
ZD6474, a vascular endothelial growth factor receptor tyrosine
kinase inhibitor, in human cancer cells with acquired resistance to
antiepidermal growth factor receptor therapy. Clin Cancer Res 2004;
10: 784–93.
22 Viloria-Petit A, Crombet T, Jothy S, et al. Acquired resistance to the
antitumor effect of epidermal growth factor receptor-blocking
antibodies in vivo: a role for altered tumor angiogenesis. Cancer Res
2001; 61: 5090–101.
23 Ciardiello F, Lenz H-J, Köhne CH, et al. Treatment outcome
according to tumor RAS mutation status in CRYSTAL study
patients with metastatic colorectal cancer (mCRC) randomized to
FOLFIRI with/without cetuximab. Proc Am Soc Clin Oncol 2014;
32 (suppl): 3506 (abstr).
24 Bokemeyer C, Köhne C-H, Ciardiello F, et al. Treatment outcome
according to tumor RAS mutation status in OPUS study patients
with metastatic colorectal cancer (mCRC) randomized to FOLFOX4
with/without cetuximab. Proc Am Soc Clin Oncol 2014;
32 (suppl): 3505 (abstr).
1075