Gefitinib in Combination With Gemcitabine and
Cisplatin in Advanced Non–Small-Cell Lung Cancer:
A Phase III Trial—INTACT 1
Giuseppe Giaccone, Roy S. Herbst, Christian Manegold, Giorgio Scagliotti, Rafael Rosell, Vincent Miller,
Ronald B. Natale, Joan H. Schiller, Joachim von Pawel, Anna Pluzanska, Ulrich Gatzemeier, John Grous,
Judith S. Ochs, Steven D. Averbuch, Michael K. Wolf, Pamela Rennie, Abderrahim Fandi,
and David H. Johnson
ABSTRACT
Purpose
The purpose of this study was to determine whether the addition of the epidermal growth factor
receptor tyrosine kinase inhibitor gefitinib (Iressa, ZD1839; AstraZeneca, Wilmington, DE) to standard
first-line gemcitabine and cisplatin provides clinical benefit over gemcitabine and cisplatin alone in
patients with advanced or metastatic non–small-cell lung cancer (NSCLC). Gefitinib has demonstrated
encouraging efficacy in advanced NSCLC in phase II trials in pretreated patients, and a phase I trial of
gefitinib in combination with gemcitabine and cisplatin showed favorable tolerability.
Patients and Methods
This was a phase III randomized, double-blind, placebo-controlled, multicenter trial in chemotherapy-
naive patients with unresectable stage III or IV NSCLC. All patients received up to six cycles of
chemotherapy (cisplatin 80 mg/m
2
on day 1 and gemcitabine 1,250 mg/m
2
on days 1 and 8 of the 3-week
cycle) plus either gefitinib 500 mg/d, gefitinib 250 mg/d, or placebo. Daily gefitinib or placebo was
continued until disease progression. End points included overall survival (primary), time to progression,
response rates, and safety evaluation.
Results
A total of 1,093 patients were enrolled. There was no difference in efficacy end points between the
treatment groups: for the gefitinib 500 mg/d, gefitinib 250 mg/d, and placebo groups, respectively,
median survival times were 9.9, 9.9, and 10.9 months (global ordered log-rank [GOLrank] P ⫽ .4560),
median times to progression were 5.5, 5.8, and 6.0 months (GOLrank; P ⫽ .7633), and response rates
were 49.7%, 50.3%, and 44.8%. No significant unexpected adverse events were seen.
Conclusion
Gefitinib in combination with gemcitabine and cisplatin in chemotherapy-naive patients with advanced
NSCLC did not have improved efficacy over gemcitabine and cisplatin alone. The reasons for this remain
obscure and require further preclinical testing.
J Clin Oncol 22:777-784. © 2004 by American Society of Clinical Oncology
INTRODUCTION
Platinum-based combination chemother-
apy constitutes standard treatment for pa-
tients with advanced or metastatic non–
small-cell lung cancer (NSCLC) and a
good performance status. Chemotherapy
has demonstrated modest but significant
improvements in survival rates over best
supportive care [1]. However, the progno-
sis for patients receiving platinum-based
chemotherapy as first-line treatment for
advanced NSCLC remains poor and side
effects are considerable [2]; therefore,
novel agents are urgently needed for this
disease. One of the most widely used plat-
inum-based combinations is gemcitabine
and cisplatin. In two phase III studies of
chemotherapy-naive patients with ad-
vanced NSCLC, the response rates for pa-
tients receiving gemcitabine and cisplatin
were 30.4% to 40.6%, median times to
From the Free University Medical Cen-
ter, Amsterdam, the Netherlands; Tho-
raxklinik, Heidelberg; Asklepios Fach-
kliniken, Munich-Gauting; Krankenhaus
Grosshansdorf, Grosshansdorf, Ger-
many; S. Luigi Gonzaga Hospital, Turin,
Italy; Hospital Germans Trias i Pujol,
Barcelona, Spain; Regional Center of
Oncology, Lodz, Poland; AstraZeneca,
Alderley Park, United Kingdom; Univer-
sity of Texas M.D. Anderson Cancer
Center, Houston, TX; Memorial Sloan-
Kettering Cancer Center, New York,
NY; Cedars-Sinai Comprehensive Can-
cer Center, Beverly Hills, CA; the Uni-
versity of Wisconsin Hospital and Clin-
ics, Madison, WI; AstraZeneca
Pharmaceuticals, Wilmington, DE; and
Vanderbilt-Ingram Cancer Center, Nash-
ville, TN.
Submitted August 1, 2003; accepted
December 2, 2003.
Supported by a grant from AstraZen-
eca, Wilmington, DE.
Authors’ disclosures of potential con-
flicts of interest are found at the end of
this article.
Address reprint requests to Giuseppe
Giaccone, MD, Vrije Universiteit Medi-
cal Center, Department of Oncology,
De Boelelaan 1117, 1081 Amsterdam,
the Netherlands; e-mail: g.giaccone@
vumc.nl.
© 2004 by American Society of Clinical
Oncology
0732-183X/04/2205-777/$20.00
DOI: 10.1200/JCO.2004.08.001
JOURNAL OF CLINICAL ONCOLOGY
ORIGINAL REPORT
VOLUME 22 䡠 NUMBER 5 䡠 MARCH 1 2004
777
Downloaded from ascopubs.org by Universita Studi Di Torino on February 8, 2019 from 130.192.222.029
Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
progression were 5.6 to 6.9 months, and overall survival
times were 8.7 to 9.1 months [3,4].
The epidermal growth factor receptor (EGFR) has been
shown to play an important role in the growth and survival
of many solid tumors, including NSCLC. Activation of
EGFR enhances the processes responsible for tumor growth
and progression, including the promotion of proliferation,
angiogenesis, and invasion/metastasis and inhibition of ap-
optosis (Fig 1) [5-7]. Gefitinib (Iressa, ZD1839; AstraZen-
eca, Wilmington, DE) is an orally active EGFR tyrosine
kinase inhibitor (EGFR-TKI) that blocks the signal trans-
duction pathways implicated in the proliferation and sur-
vival of cancer cells [7].
Two phase II trials of gefitinib monotherapy in patients
with pretreated advanced NSCLC demonstrated encourag-
ing antitumor activity (objective response rates, 11.8% to
18.4%) and symptom relief (symptom improvement rates,
40.3% to 43.1%) and good tolerability [8-11]. This favor-
able tolerability profile, coupled with its mechanism of ac-
tion, which is distinct from that of cytotoxic agents, pro-
vides a strong rationale for use of gefitinib in combination
with standard cytotoxic regimens. This rationale is sup-
ported by data from preclinical studies in which gefitinib
potentiated the efficacy of various cytotoxic drugs against a
range of human solid tumor types, both in vitro and in vivo
[12,13]. In particular, synergy was observed when gefitinib and
cisplatin were used in combination, whereas no synergy could
be demonstrated between gemcitabine and gefitinib [13].
On the basis of phase I trials of gefitinib monotherapy,
two doses were identified for further study based on phar-
macokinetics and clinical activity. Gefitinib 250 mg/d is
higher than the lowest dose at which clinical responses were
Fig 1. Epidermal growth factor recep-
tor signal transduction. Adapted with
permission [5].
Giaccone et al
778
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from ascopubs.org by Universita Studi Di Torino on February 8, 2019 from 130.192.222.029
Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
seen (150 mg/d), and 500 mg/d is the highest dose level that
is well tolerated for a long period by most patients. A feasi-
bility study of daily oral gefitinib (250 and 500 mg/d) in
combination with gemcitabine and cisplatin demonstrated
a manageable and predictable tolerability profile with no
evidence of any clinically significant pharmacokinetic inter-
actions between gefitinib and cisplatin or gemcitabine [14].
Encouraging antitumor activity was seen in a range of solid
tumors at both gefitinib doses in this study [14].
Here we present the results from a multinational, ran-
domized, double-blind, placebo-controlled phase III study.
The Iressa NSCLC Trial Assessing Combination Treatment
(INTACT) 1 investigated the efficacy and safety of gefitinib
(250 and 500 mg once daily) versus placebo in combination
with cisplatin plus gemcitabine in chemotherapy-naive pa-
tients with advanced NSCLC. The primary objective was to
assess whether gefitinib increases overall survival relative to
placebo in combination with cisplatin and gemcitabine. A
second trial of identical design (INTACT 2) has been con-
ducted to investigate gefitinib in combination with pacli-
taxel plus carboplatin, another commonly used combina-
tion in patients with advanced NSCLC [15].
PATIENTS AND METHODS
Patients were randomly assigned to one of three treatment groups
in a double-blind manner. All patients received gemcitabine and
cisplatin in combination with gefitinib 500 mg/d, gefitinib 250
mg/d, or placebo. Patients were further stratified by dynamic
randomization [16] according to weight loss in previous 6 months
(ⱕ 5% v ⬎ 5%), disease stage (III v IV), performance status (0 to
1 v 2), and measurable disease (yes v no).
Treatment
Gefitinib or placebo was administered orally, once daily.
Chemotherapy was administered in 3-week cycles for a total of six
cycles: intravenous gemcitabine 1,250 mg/m
2
for 30 minutes on
days 1 and 8; intravenous cisplatin 80 mg/m
2
after gemcitabine
administration on day 1 only. Subsequently, patients continued
on gefitinib or placebo until disease progression. One gefitinib/
placebo dose reduction (500 to 250 mg/d or 250 to 100 mg/d) was
allowed per patient. In the event of grade 3 or 4 adverse events not
thought to be due to disease progression or gemcitabine and
cisplatin, gefitinib administration could be interrupted for a max-
imum of 14 days.
Eligibility Criteria
The inclusion criteria included histologically/cytologically
confirmed NSCLC, locally advanced stage III disease not curable
with surgery or radiotherapy or stage IV disease, aged ⱖ 18 years,
and WHO performance status of 0 to 2. Patients were not eligible
for this trial if they had previously received chemotherapy (prior
surgery or localized radiation were allowed); were hypersensitive
to mannitol, corticosteroids, H
2
-antagonists, antihistamines or
agents formulated with polyoxyethylated castor oil; had received
radiotherapy within the last 2 weeks; had unresolved toxicity from
previous radiation therapy or incomplete healing from previous
surgery; had preexisting motor or sensory neurotoxicity (National
Cancer Institute Common Toxicity Criteria ⱖ grade 2); showed
evidence of severe or uncontrolled systemic disease; had recent
conditions requiring medication or uncontrolled significant active
infections; had an absolute neutrophil count less than 2,000/mm
3
,
WBCs less than 4,000/mm
3
, or platelets less than 100,000/mm
3
;
had serum bilirubin greater than 1.25 times the upper limit of
reference range (ULRR), ALT or AST greater than 2.5 times ULRR
(five times ULRR in the presence of liver metastases), serum
creatinine greater than 1.5 times ULRR, or creatinine clearance
less than 60 mL/min; were pregnant or breast-feeding; had other
coexisting malignancies or malignancies diagnosed within the last
5 years with the exception of basal-cell carcinoma or cervical
cancer in situ; or had mixed NSCLC plus small-cell lung cancer.
Patients with stable brain metastasis or spinal-cord compression
were eligible. All patients signed a written informed consent form,
and trial approval was obtained from the ethics committee at each
trial center. The study followed the Declaration of Helsinki [17]
and good clinical practice guidelines.
Assessments
Overall survival and time to progression were assessed from
the date of randomization to the date of patient death and the date
of objective disease progression (death was considered a progres-
sion event in those patients who died before disease progression),
respectively. Patients without documented death or objective pro-
gression at the time of the final analysis were censored at the date
they were last known to be alive or of their last objective tumor
assessment, respectively. Tumor response was evaluated according
to Response Evaluation Criteria in Solid Tumors [18]. During the
trial and for 30 days after the last dose of gefitinib or placebo,
patients were monitored for adverse events, which were graded
according to the National Cancer Institute Common Toxicity
Criteria. Hematology and biochemistry assessments were per-
formed ⱕ 7 days before the date of randomization and at each
clinic visit.
Statistical Analysis
An Independent Data Monitoring Committee provided on-
going guidance and recommendations for patient management,
based on review of the formal interim efficacy analysis, and also
reviewed safety data on an ad-hoc basis, as determined by a steer-
ing committee. The role of the committee was to provide executive
oversight and supervision for the conduct of the trial, through
review of trial enrollment, protocol and clinical conduct, and
blinded safety data.
Gefitinib was compared with placebo on an intent-to-treat
basis with respect to overall survival. The study was designed to
have 90% power for a two-sided overall significance level test of
the hypothesis that gefitinib increases survival relative to placebo
given a hazard ratio of 1.33. Assuming a 1-year survival rate of
30% in the placebo arm, this hazard ratio equates to an increase
in median survival of 2.3 months for both gefitinib arms. The
final analysis of overall survival was planned to include 750
events. On the basis of the study design assumptions, 1,029
patients were required.
An adaptive survival analysis procedure was used at the final
analysis that tested either for a positive or negative gefitinib dose-
response relationship based on prospective criteria applied to the
observed data. A survival trend test (global ordered log-rank test
[GOLrank]) was to be used for a positive dose-response relation-
ship, whereas pairwise log-rank tests would be used for a mixed
dose-response relationship [19]. To preserve an overall two-sided
5% significance level, and to account for the use of a GOLrank test
Gefitinib Plus Gemcitabine/Cisplatin in NSCLC
www.jco.org 779
Downloaded from ascopubs.org by Universita Studi Di Torino on February 8, 2019 from 130.192.222.029
Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
at an interim analysis, simulations with the adaptive procedure
were used to calculate a nominal significance level of 4.4% for the
final analysis. According to prospective criteria for the adaptive
procedure, the final analysis used a GOLrank test to compare
survival between the treatment arms.
A posthoc multivariate analysis with eight prespecified prognos-
tic factors at trial entry (disease stage III v IV; performance status 0 or
1 v 2; weight loss in prior 6 months ⱕ 5% v ⬎ 5%; sex; histology;
presence or absence of metastases to bone, liver, or brain) was per-
formed to assess which variables predicted improved survival.
In a posthoc subgroup analysis, stratification and prognostic
factors (disease stage III v IV; performance status 0 or 1 v 2; weight
loss in prior 6 months ⱕ 5% v ⬎ 5%; presence or absence of
metastases to bone, liver, or brain) and subgroups of sex and
histology were examined in a univariate model. An unadjusted
Cox proportional hazard test was applied to the overall survival
data for each subgroup to estimate the hazard ratio and 95% CI
for the treatment comparisons of gefitinib 250 or 500 mg/d
versus placebo.
RESULTS
Patients
In total, 1,093 patients were enrolled from 155 centers
between May 2000 and March 2001. Most patients were
enrolled by European trial centers (n ⫽ 816; 74.7%), but
patients were also enrolled in North America (n ⫽ 139;
12.7%), Asia (n ⫽ 58; 5.3%), South America (n ⫽ 45;
4.1%), and South Africa (n ⫽ 17; 1.6%). Almost three
quarters of the patients were men and the median age was
approximately 60 years. Overall, 998 (90.4%) of the patients
were white. Most patients had stage IV disease (n ⫽ 757;
69.2%) or IIIB with malignant pleural effusion (n ⫽ 239;
21.9%). Most patients (n ⫽ 984; 90.0%) had a performance
status of 0 or 1. The most common histology types were
adenocarcinoma and squamous-cell carcinoma, seen in 504
(46.1%) and 328 patients (30.0%), respectively. The three
treatment arms were well balanced (Table 1). The median
follow-up duration was 15.9 months.
Efficacy
At each interim analysis, the Independent Data Moni-
toring Committee recommended that the trial continue. At
the time of this analysis, 732 and 628 events were observed
for survival and time to progression, respectively, with a
minimum follow-up of 12 months for overall survival and 6
months for all other end points. There was no statistically
significant difference in overall survival between each of the
gefitinib arms and the placebo arm. The median survival
times were 9.9 months for each of the gefitinib groups, and
10.9 months for the placebo group (GOLrank P ⫽ .4560;
Fig 2). One-year survival rates were 43% and 41% for the
500 mg/d and 250 mg/d gefitinib groups, respectively, and
44% for the placebo group. Similarly, median time to pro-
gression was 5.5 and 5.8 months for the 500 mg/d and 250
mg/d gefitinib groups, respectively, and 6.0 months in the
placebo group, with no statistically significant difference
between treatment arms (GOLrank P ⫽ .7633; Fig 3). In the
posthoc multivariate analysis, a performance status of 2,
weight loss, and bone and liver metastases were significant
(P ⬍ .05) predictors of worse survival outcome. In the
posthoc univariate analysis examining known NSCLC
prognostic factors and subgroups of sex, time on chemo-
therapy, and histology, no survival differences were seen
Table 1. Patient Demographics
Gefitinib
500 mg/d
(n ⫽ 365)
Gefitinib
250 mg/d
(n ⫽ 365)
Placebo
(n ⫽ 363)
Sex, %
Female 27.9 23.3 27.8
Male 72.1 76.7 72.2
Age, years
Median 61 59 61
Range 31-85 34-83 33-81
Disease stage, %
ⴱ
IIIa 3.0 1.6 1.9
IIIb 29.9 25.8 28.4
Without pleural effusion 5.5 6.3 6.6
With pleural effusion 24.4 19.5 21.8
IV 66.8 72.3 68.6
WHO performance status, %
ⴱ
0 32.1 34.0 33.9
1 58.1 56.4 55.6
2 9.6 9.6 9.6
Weight loss in previous 6 months, %
ⴱ
ⱕ 5% 62.7 64.1 63.9
⬎ 5% 37.0 35.6 35.0
Disease measurability, %
ⴱ
Measurable 94.8 94.8 93.4
Nonmeasurable 4.9 4.9 5.8
Histology, %
ⴱ
Squamous 28.8 32.1 29.2
Adenocarcinoma 43.3 48.5 46.6
Squamous and adenocarcinoma 2.7 1.4 1.4
Bronchoalveolar 1.1 0.8 0.3
Undifferentiated 11.5 9.3 11.8
Large cell 11.5 7.7 8.8
Race, %
White 91.0 90.4 89.8
Black 0.8 1.4 1.4
Asian 1.9 1.6 0.8
Hispanic 1.4 2.5 2.2
Oriental 4.9 3.6 5.8
Other 0.0 0.5 0.0
Metastatic disease, %†
Lung, other 54.9 53.4 51.0
Bone 30.3 34.8 37.8
Liver 21.7 22.3 20.5
Adrenal tissue 21.3 22.7 19.7
Lymph nodes 10.2 10.2 14.1
Skin or soft tissue 5.7 5.3 7.2
Brain 5.3 4.5 2.8
Other 7.8 4.9 4.8
ⴱ
Information was not available for all patients.
†Stage IV patients only.
Giaccone et al
780
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from ascopubs.org by Universita Studi Di Torino on February 8, 2019 from 130.192.222.029
Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
between treatments for any subgroups of patients [20].
Objective tumor response rates were 50.3% and 51.2% for
the 500 mg/d and 250 mg/d gefitinib groups, respectively,
and 47.2% in the placebo group (P ⫽ not significant). Very
few complete responses were seen (Table 2).
Duration of Therapy, Dose Intensity, and
Dose Adherence
The dose-intensity of both gemcitabine and cisplatin
was similar in all three treatment groups (Table 3), demon-
strating that chemotherapy dose intensity was not compro-
mised by the addition of gefitinib. There was a high overall
adherence to gefitinib; however, most gefitinib dose inter-
ruptions and reductions were seen in the gefitinib 500 mg/d
arm (the number was similar in the gefitinib 250 mg/d and
placebo arms) [Table 3]. Patients receiving gefitinib 250
mg/d or placebo had a longer therapy duration than those
receiving gefitinib 500 mg/d.
Safety and Tolerability
Most adverse events occurred during the combination
phase of the trial and many were consistent with the known
toxicities of the chemotherapy agents. Overall, the safety
data from the monotherapy period of the trial support the
gefitinib safety profile previously established in phase I and
II trials. The most commonly occurring adverse events were
gastrointestinal, skin-related, or hematologic in nature. Sta-
tistical analysis of prespecified adverse events during the
chemotherapy phase revealed no difference between treat-
ment arms except for diarrhea and skin events (diarrhea
P ⬍ .0001 for 500 mg/d v 250 mg/d or placebo; P ⫽ .0924 for
250 mg/d v placebo; defined skin events P ⬍ .0001 for all
comparisons [no adjustments were made to the P values to
take account of the multiple comparisons]), which are known
to be associated with gefitinib treatment. A clear dose-response
relationship was observed for these events. Interstitial lung
disease (ILD)–type events were experienced by three, one, and
three patients in the gefitinib 500 mg/d, gefitinib 250 mg/d,
and placebo arms, respectively, giving an overall incidence of
less than 1%. The incidence of ILD-type events and other
respiratory events that were possibly indicative of ILD is sum-
marized in Table 4. No difference in reports of symptoms
possibly related to, or indicative of, ILD (eg, dyspnea, increased
cough, pneumonia) was seen between the groups.
The most frequently occurring adverse events consid-
ered by the investigators to be related to gefitinib/placebo
treatment were rash, diarrhea, and acne, which were gener-
ally mild (grade 1 or 2; Table 5). No significant additive
toxicity was evident in this placebo-controlled setting.
Deaths and withdrawals owing to gefitinib/placebo-related
toxicity were low and balanced between the three treatment
arms. The proportion of patients withdrawn from treatment
because of adverse events of any cause was higher (23.0%) for
patients in the 500 mg/d group, compared with 14.5% and
11.3% for 250 mg/d and placebo groups, respectively. The
types of events leading to withdrawal were similar across the
three groups: diarrhea, nausea, vomiting, and acne-like rash.
DISCUSSION
In this study, gefitinib showed no survival benefit over
placebo when combined with gemcitabine and cisplatin in a
large population of chemotherapy-naive patients with ad-
vanced NSCLC. Furthermore, gefitinib did not improve
Fig 2. Kaplan-Meier estimates of overall survival in each treatment group.
(global ordered log-rank P ⫽ .4560).
Fig 3. Kaplan-Meier estimates of time to progression in each treatment
group. (global ordered log-rank P ⫽ .7633).
Table 2. Objective Tumor Responses in Each Treatment Group
(population assessable for tumor response)
Objective Tumor Response (%)
Gefitinib
500 mg/d
(n ⫽ 330)
Gefitinib
250 mg/d
(n ⫽ 336)
Placebo
(n ⫽ 324)
Complete response 2.1 3.3 0.9
Partial response 48.2 47.9 46.3
Response rate (complete plus
partial response)
50.3 51.2 47.2
Gefitinib Plus Gemcitabine/Cisplatin in NSCLC
www.jco.org 781
Downloaded from ascopubs.org by Universita Studi Di Torino on February 8, 2019 from 130.192.222.029
Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
![Fig 1. Epidermal growth factor receptor signal transduction. Adapted with permission [5].](/figures/fig-1-epidermal-growth-factor-receptor-signal-transduction-11xy84bb.webp)
