1L CrizoVSQT ALK+

The n e w e ng l a n d j o u r na l of m e dic i n e
Original Article
First-Line Crizotinib versus Chemotherapy

in ALK-Positive Lung Cancer
Benjamin J. Solomon, M.B., B.S., Ph.D., Tony Mok, M.D.,
Dong‑Wan Kim, M.D., Ph.D., Yi‑Long Wu, M.D.,
Kazuhiko Nakagawa, M.D., Ph.D., Tarek Mekhail, M.D.,
Enriqueta Felip, M.D., Ph.D., Federico Cappuzzo, M.D., Jolanda Paolini, B.Sc.,
Tiziana Usari, B.Sc., Shrividya Iyer, Ph.D., Arlene Reisman, M.P.H.,
Keith D. Wilner, Ph.D., Jennifer Tursi, M.Sc., and Fiona Blackhall, M.D., Ph.D.,
for the PROFILE 1014 Investigators*
A BS T R AC T
BACKGROUND
The efficacy of the ALK inhibitor crizotinib as compared with standard chemo- From Peter MacCallum Cancer Centre,
therapy as first-line treatment for advanced ALK-positive non–small-cell lung can- Melbourne, VIC, Australia (B.J.S.); State
Key Laboratory of South China, Hong
cer (NSCLC) is unknown. Kong Cancer Institute, Department of
Clinical Oncology, Chinese University of
METHODS Hong Kong, Shatin (T. Mok), and Guang-
dong Lung Cancer Institute, Guangzhou
We conducted an open-label, phase 3 trial comparing crizotinib with chemother- (Y.-L.W.) — both in China; Seoul National
apy in 343 patients with advanced ALK-positive nonsquamous NSCLC who had University Hospital, Seoul, South Korea
received no previous systemic treatment for advanced disease. Patients were ran- (D.-W.K.); Kinki University, Osaka, Japan
(K.N.); Florida Hospital Cancer Institute,
domly assigned to receive oral crizotinib at a dose of 250 mg twice daily or to Orlando (T. Mekhail); Vall d’Hebron In-
receive intravenous chemotherapy (pemetrexed, 500 mg per square meter of body- stitute of Oncology, Vall d’Hebron Uni-
surface area, plus either cisplatin, 75 mg per square meter, or carboplatin, target versity Hospital, Barcelona (E.F.); Istituto
Toscano Tumori, Livorno (F.C.), and Pfizer
area under the curve of 5 to 6 mg per milliliter per minute) every 3 weeks for up Oncology, Milan (J.P., T.U., J.T.) — both
to six cycles. Crossover to crizotinib treatment after disease progression was per- in Italy; Pfizer Oncology (S.I.) and Pfizer
mitted for patients receiving chemotherapy. The primary end point was progres- Global Innovative Pharma Business (A.R.)
— both in New York; Pfizer Oncology, La
sion-free survival as assessed by independent radiologic review. Jolla, CA (K.D.W.); and the Christie Hos-
pital and Institute of Cancer Sciences,
RESULTS Manchester University, Manchester, Unit-
ed Kingdom (F.B.). Address reprint re-
Progression-free survival was significantly longer with crizotinib than with che- quests to Dr. Solomon at the Depart-
motherapy (median, 10.9 months vs. 7.0 months; hazard ratio for progression or ment of Medical Oncology, Peter Mac-
death with crizotinib, 0.45; 95% confidence interval [CI], 0.35 to 0.60; P<0.001). Callum Cancer Centre, St. Andrew’s Pl.,
East Melbourne, VIC 3002, Australia; or
Objective response rates were 74% and 45%, respectively (P<0.001). Median overall at ben.solomon@petermac.org.
survival was not reached in either group (hazard ratio for death with crizotinib,
*A complete list of the investigators in
0.82; 95% CI, 0.54 to 1.26; P = 0.36); the probability of 1-year survival was 84% with the PROFILE 1014 trial is provided in
crizotinib and 79% with chemotherapy. The most common adverse events with the Supplementary Appendix, available
crizotinib were vision disorders, diarrhea, nausea, and edema, and the most com- at NEJM.org.
mon events with chemotherapy were nausea, fatigue, vomiting, and decreased Drs. Solomon and Mok contributed equal-
appetite. As compared with chemotherapy, crizotinib was associated with greater ly to this article.
reduction in lung cancer symptoms and greater improvement in quality of life. This article was updated on October 15,
2015, at NEJM.org.
CONCLUSIONS N Engl J Med 2014;371:2167-77.
Crizotinib was superior to standard first-line pemetrexed-plus-platinum chemo- DOI: 10.1056/NEJMoa1408440
therapy in patients with previously untreated advanced ALK-positive NSCLC. Copyright © 2014 Massachusetts Medical Society.
(Funded by Pfizer; PROFILE 1014 ClinicalTrials.gov number, NCT01154140.)
n engl j med 371;23 nejm.org December 4, 2014 2167

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org on October 26, 2024. For personal use only.
No other uses without permission. Copyright © 2014 Massachusetts Medical Society. All rights reserved.
R
earrangements of the anaplastic and higher numbers indicating increasing dis-
lymphoma kinase (ALK) gene are present ability)15; and adequate hepatic, renal, and bone
in 3 to 5% of non–small-cell lung cancers marrow function (as defined in the study proto-
(NSCLCs).1,2 They define a distinct subgroup of col). Patients with treated brain metastases were
NSCLC that typically occurs in younger patients eligible if the metastases were neurologically
who have never smoked or have a history of light stable for at least 2 weeks before enrollment and
smoking and that has adenocarcinoma histo- the patient had no ongoing requirement for glu-
logic characteristics.3-5 cocorticoids. All patients provided written in-
Crizotinib is an oral small-molecule tyrosine formed consent before enrollment.
kinase inhibitor of ALK, MET, and ROS1 kinases.6
In phase 1 and 2 studies, crizotinib treatment re- Study Oversight
sulted in objective tumor responses in approxi- The protocol was approved by the institutional
mately 60% of patients with ALK-positive NSCLC review board or independent ethics committee at
and in progression-free survival of 7 to 10 each participating center and complied with the
months.7-9 In a randomized phase 3 trial involving International Ethical Guidelines for Biomedical
patients with advanced ALK-positive NSCLC who Research Involving Human Subjects, Good Clin-
had received previous platinum-based chemother- ical Practice guidelines, the Declaration of Hel-
apy, crizotinib showed efficacy superior to that sinki, and local laws. The study was designed by
of single-agent second-line chemotherapy with the sponsor (Pfizer) and by members of the
either pemetrexed or docetaxel.10 However, the PROFILE 1014 steering committee (see the Sup-
efficacy of crizotinib as initial treatment for plementary Appendix). The sponsor collected and
patients with newly diagnosed advanced ALK-posi- analyzed the data in conjunction with the au-
tive NSCLC as compared with the existing stan- thors, all of whom had full access to the data. The
dard-of-care, platinum-based double-agent chemo- manuscript was written by the first two authors,
therapy,11,12 is unknown. with medical writing support from ACUMED (Ty-
We report the results of an ongoing interna- therington, United Kingdom, and New York) fund-
tional, multicenter, randomized, open-label, phase ed by the sponsor. All the authors vouch for the
3 study (PROFILE 1014) that compares crizotinib accuracy and completeness of the data and for
treatment with pemetrexed-plus-platinum chemo- the fidelity of this report to the study protocol.
therapy with respect to efficacy, safety, and pa- The protocol and statistical analysis plan are avail-
tient-reported outcomes in patients with previously able at NEJM.org.
untreated advanced ALK-positive NSCLC.
Study Design and Treatment
Patients were randomly assigned, in a 1:1 ratio, to
Me thods
receive oral crizotinib, at a dose of 250 mg twice
Patients daily, or intravenous chemotherapy (pemetrexed, at
Patients were eligible for enrollment if they had a dose of 500 mg per square meter of body-surface
histologically or cytologically confirmed locally area, plus either cisplatin, at a dose of 75 mg per
advanced, recurrent, or metastatic nonsquamous square meter, or carboplatin, target area under the
NSCLC that was positive for an ALK rearrange- curve of 5 to 6 mg per milliliter per minute) ad-
ment (as determined centrally with the use of a ministered every 3 weeks for a maximum of six
Vysis ALK Break Apart FISH Probe Kit [Abbott cycles. The choice of platinum chemotherapy was
Molecular])7,13 and if they had received no previ- made by the investigator. Randomization was
ous systemic treatment for advanced disease. stratified according to ECOG performance status
Other eligibility criteria included an age of 18 years (0 or 1 vs. 2), Asian or non-Asian race, and pres-
or older; measurable disease as assessed accord- ence or absence of brain metastases. Treatment
ing to the Response Evaluation Criteria in Solid was continued until RECIST-defined disease pro-
Tumors (RECIST), version 1.114 (summarized in gression, development of unacceptable toxic ef-
Table S1 in the Supplementary Appendix, available fects, death, or withdrawal of consent. Continu-
with the full text of this article at NEJM.org); an ation of crizotinib beyond disease progression was
Eastern Cooperative Oncology Group (ECOG) per- allowed for patients who had been randomly as-
formance status of 0, 1, or 2 (on a scale of 0 to 5, signed to crizotinib if the patient was perceived
with 0 indicating that the patient is asymptomatic by the investigator to be having clinical benefit.
2168 n engl j med 371;23 nejm.org December 4, 2014
Crizotinib vs. Chemother apy in Lung Cancer
Patients in the chemotherapy group who had dis- tion factors were used for between-group com-
ease progression as confirmed by independent ra- parisons of progression-free survival and overall
diologic review could cross over to crizotinib treat- survival; stratified Cox regression models were
ment if safety screening criteria were met. applied to estimate hazard ratios. As prespecified
The primary end point was progression-free in the protocol, overall survival was also ana-
survival (the time from randomization to RECIST- lyzed with the rank-preserving structural failure
defined progression, as assessed by independent time model20-22 to explore the effect of crossover
radiologic review, or death). Secondary end points to crizotinib in the chemotherapy group. All
included the objective response rate, overall sur- analyses in the chemotherapy group, with the
vival, safety, and patient-reported outcomes. exception of the analysis of overall survival, in-
cluded only data collected before crossover to
Assessments crizotinib. We used a two-sided stratified Co-
Tumor assessment was performed during screen- chran–Mantel–Haenszel test to compare the ob-
ing (within 28 days before randomization), every jective response rate between treatment groups.
6 weeks during treatment, and at the post-treat- Safety evaluations were performed in the as-
ment follow-up visits (which were scheduled ev- treated population, which included all patients
ery 6 weeks) until RECIST-defined progression. who received at least one dose of study medica-
For patients who crossed over to crizotinib treat- tion. Safety results were not adjusted for the
ment or continued crizotinib treatment beyond shorter duration of treatment in the chemother-
progression, assessments continued to be per- apy group. Patient-reported outcomes were eval-
formed every 12 weeks. Brain or bone lesions that uated in patients in the intention-to-treat popu-
were detected at the time of screening were evalu- lation who also had a baseline assessment and
ated in all subsequent tumor assessments (i.e., at least one post-baseline assessment. Additional
every 6 weeks). In all patients, brain and bone details of the statistical methods are provided in
scanning was repeated every 12 weeks to moni- the Supplementary Appendix.
tor for new lesions. All scans were submitted for
central independent radiologic review by radiolo- R e sult s
gists who were unaware of the group assignments.
Adverse events were classified and graded ac- Patients
cording to Common Terminology Criteria for Ad- Between January 2011 and July 2013, a total of
verse Events, version 4.0. Patient-reported out- 343 patients underwent randomization — 172 to
comes were assessed with the use of the European crizotinib and 171 to chemotherapy (intention-to-
Organisation for Research and Treatment of Can- treat population) (Fig. S1 in the Supplementary
cer (EORTC) quality-of-life core questionnaire Appendix). Three patients underwent random-
(QLQ-C30),16,17 the corresponding lung cancer ization but received no study treatment, leaving
module (QLQ-LC13),18 and the EuroQol Group 340 patients in the as-treated population — 171
5-Dimension Self-Report Questionnaire (EQ-5D).19 patients in the crizotinib group and 169 in the
chemotherapy group (with 91 patients receiving
Statistical Analysis pemetrexed−cisplatin and 78 receiving peme-
We estimated that with 229 events of progression trexed−carboplatin). At the time of data cutoff,
or death, the study would have 85% power to the median duration of follow-up for overall sur-
detect a 50% improvement in progression-free survival was 17.4 months for patients assigned to
vival with crizotinib versus chemotherapy (from crizotinib and 16.7 months for those assigned to
6 months to 9 months), at a one-sided alpha chemotherapy. The baseline characteristics in the
level of 0.025. The prespecified number of events intention-to-treat population were well balanced
for the primary end point was reached in No- between the groups (Table 1).
vember 2013; the data cutoff date was November
30, 2013. Efficacy end points were measured in Efficacy
the intention-to-treat population, which includ- The median progression-free survival was 10.9
ed all patients who underwent randomization. months (95% confidence interval [CI], 8.3 to 13.9)
The Kaplan–Meier method was used to estimate among patients in the crizotinib group, as com-
time-to-event end points. Two-sided log-rank pared with 7.0 months (95% CI, 6.8 to 8.2) among
tests stratified according to baseline stratifica- patients in the chemotherapy group (hazard ratio

Table 1. Baseline Characteristics in the Intention-to-Treat Population.* Figure 1 (facing page). Progression-free and Overall
Survival.
Crizotinib Chemotherapy Panel A shows Kaplan–Meier estimates of progres-
Characteristic (N = 172) (N = 171)
sion-free survival in the intention-to-treat population.
Age — yr There were 100 events of progression or death with
crizotinib (89 progression events as assessed by inde-
Median 52 54
pendent radiologic review and 11 deaths without docu-
Range 22–76 19–78 mented progression) and 137 events with chemothera-
Male sex — no. (%) 68 (40) 63 (37) py (132 progression events as assessed by
independent radiologic review and 5 deaths without
Race — no. (%)† documented progression). The median progression-
White 91 (53) 85 (50) free survival was 10.9 months with crizotinib as com-
Asian 77 (45) 80 (47) pared with 7.0 months with chemotherapy. The rate of
progression-free survival at 18 months was 31% (95%
Other 4 (2) 6 (4) CI, 23 to 39) in the crizotinib group and 5% (95% CI, 2
Smoking status — no. (%) to 10) in the chemotherapy group. Panel B shows Ka-
plan–Meier estimates of overall survival in the inten-
Never smoked 106 (62) 112 (65)
tion-to-treat population. Because the rate of death
Former smoker 56 (33) 54 (32) from any cause at the time of data cutoff was relatively
Current smoker 10 (6) 5 (3) low (26%; 90 of the 343 patients who underwent ran-
domization), the median overall survival was not
Histologic characteristic of tumor — no. (%) reached in either group. Of the 171 patients randomly
Adenocarcinoma 161 (94) 161 (94) assigned to chemotherapy, 120 (70%) subsequently re-
ceived crizotinib treatment. Of the 172 patients as-
Nonadenocarcinoma 11 (6) 10 (6)
signed to crizotinib, 21 (12%) subsequently received
ECOG performance status — no. (%)‡ platinum-based chemotherapy. This analysis was not
0 or 1 161 (94) 163 (95) adjusted for crossover. Tick marks on the curves in
Panels A and B indicate censoring of data. Panel C
2 10 (6) 8 (5)
shows hazard ratios and 95% confidence intervals for
Extent of disease — no. (%) the treatment effect on progression-free survival in
Locally advanced 4 (2) 3 (2) subgroups of the intention-to-treat population defined
according to prespecified stratification factors and
Metastatic 168 (98) 168 (98) baseline characteristics. Race was self-reported. East-
Time since first diagnosis — mo ern Cooperative Oncology Group (ECOG) performance
status scores range from 0 to 5, with higher scores in-
Median 1.2 1.2
dicating increasing disability; an ECOG performance
Range 0–114.0 0–93.6 status of 0 indicates that the patient is fully active, 1
Brain metastases present — no. (%) 45 (26) 47 (27) that the patient is ambulatory but restricted in strenu-
ous activity, and 2 that the patient is ambulatory and
* There were no significant differences between the groups in any of the charac- capable of self-care but is unable to work. Data for
teristics listed in this table. ECOG performance status were missing for 1 patient.
† Race was self-reported.
‡ The Eastern Cooperative Oncology Group (ECOG) performance status was
assessed at the time of screening; the score was not reported for one patient
in the crizotinib group. Scores range from 0 to 5, with higher scores indicat- sponse was 11.3 months and 5.3 months, re-
ing increasing disability; an ECOG performance status of 0 indicates that the spectively. The best percentage change from base-
patient is fully active, 1 that the patient is ambulatory but restricted in strenu-
ous activity, and 2 that the patient is ambulatory and capable of self-care but line in target lesions and the best overall response
is unable to work. in individual patients are shown in Figure S2 in
the Supplementary Appendix. Intracranial lesions
progressed or new intracranial lesions developed
for progression or death with crizotinib, 0.45; in 25 patients in the crizotinib group and in 26
95% CI, 0.35 to 0.60; P<0.001) (Fig. 1A). The haz- patients in the chemotherapy group (15% each).
ard ratio favored crizotinib across most subgroups There was no significant difference in overall
defined according to stratification factors and survival between patients in the crizotinib group
other baseline characteristics (Fig. 1C). and those in the chemotherapy group at the time
The objective response rate was significantly of the progression-free survival analysis (hazard
higher with crizotinib than with chemotherapy ratio for death with crizotinib, 0.82; 95% CI, 0.54
(74% [95% CI, 67 to 81] vs. 45% [95% CI, 37 to 53], to 1.26; P = 0.36) (Fig. 1B) — probably owing to
P<0.001) (Table 2). The median duration of re- the relatively low rate of death from any cause
A Progression-free Survival B Overall Survival

100 Hazard ratio for progression 100
Crizotinib
Progression-free Survival (%)
or death in the crizotinib group,
80 0.45 (95% CI, 0.35–0.60) 80
Overall Survival (%)

P<0.001 (two-sided stratified log-rank test) Chemotherapy
60 60
40 40
Crizotinib
Hazard ratio for death in the crizotinib
20 20 group, 0.82 (95% CI, 0.54–1.26)
Chemotherapy P=0.36 (two-sided stratified log-rank test)
0 0
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
Months Months
No. at Risk No. at Risk
Crizotinib 172 120 65 38 19 7 1 0 Crizotinib 172 152 123 80 44 24 3 0
Chemotherapy 171 105 36 12 2 1 0 0 Chemotherapy 171 146 112 74 47 21 4 0
C Progression-free Survival, According to Subgroup

Subgroup No. of Patients Hazard Ratio (95% CI)
Crizotinib vs. chemotherapy 343 0.45 (0.35–0.60)
Age
≥65 yr 55 0.37 (0.17–0.77)
<65 yr 288 0.51 (0.38–0.68)
Sex
Male 131 0.54 (0.36–0.82)
Female 212 0.45 (0.32–0.63)
Race
Non-Asian 186 0.53 (0.36–0.76)
Asian 157 0.44 (0.30–0.65)
Smoking status
Smoker or former smoker 125 0.64 (0.42–0.97)
Nonsmoker 218 0.41 (0.29–0.58)
Time since diagnosis
>1 yr 35 0.14 (0.04–0.51)
≤1 yr 308 0.52 (0.40–0.68)
ECOG performance status
2 18 0.19 (0.05–0.76)
0 or 1 324 0.47 (0.36–0.62)
Adenocarcinoma
Yes 322 0.49 (0.37–0.64)
No 21 0.37 (0.12–1.10)
Type of disease
Metastatic 336 0.48 (0.37–0.63)
Locally advanced 7 0.54 (0.07–3.91)
Brain metastases
Yes 92 0.57 (0.35–0.93)
No 251 0.46 (0.34–0.63)
0.01 0.1 1.0 10
Crizotinib Better Chemotherapy

Better
(26%; 90 of the 343 patients who underwent ran- tinib group and 79% (95% CI, 71 to 84) in the
domization) and the fact that 70% of the pa- chemotherapy group. After adjustment for cross-
tients in the chemotherapy group crossed over to over with the rank-preserving structural failure
crizotinib treatment. The probability of 1-year time model, the hazard ratio for death with crizo-
survival was 84% (95% CI, 77 to 89) in the crizotinib was 0.60 (95% CI, 0.27 to 1.42) as calcu-

Table 2. Response to Treatment in the Intention-to-Treat Population.*

tails are provided in the Supplementary Appen-
dix. Other systemic therapies received during
Crizotinib Chemotherapy follow-up are listed in Table S2 in the Supple-
Response (N = 172) (N = 171)
mentary Appendix. The baseline characteristics
Type of response — no. (%) of the patients and the efficacy outcomes in
Complete response 3 (2) 2 (1) subgroup analyses of crizotinib versus individual
Partial response 125 (73) 75 (44) chemotherapy regimens were similar to those in
Stable disease 29 (17) 63 (37)
the analysis of the overall population (Table S3
and Fig. S4 in the Supplementary Appendix).
Progressive disease 8 (5) 21 (12)
Could not be evaluated† 7 (4) 10 (6) Safety and Adverse Events
Objective response rate — % (95% CI)‡ 74 (67–81) 45 (37–53) The median duration of treatment was 10.9
Time to response — mo§ months (range, 0.4 to 34.3) in the crizotinib
Median 1.4 2.8 group (a median of 16 cycles started [range, 1 to
Range 0.6–9.5 1.2–8.5 50]) and 4.1 months (range, 0.7 to 6.2) in the
chemotherapy group (a median of 6 cycles of
Duration of response — mo¶
chemotherapy started [range, 1 to 6]). The most
Median 11.3 5.3
common adverse events of any cause for which
95% CI 8.1−13.8 4.1−5.8 the incidence was at least 5 percentage points
* Tumor responses were assessed with the use of Response Evaluation Criteria
higher in the crizotinib group than in the che-
in Solid Tumors (RECIST), version 1.1, and were confirmed by independent motherapy group were vision disorder (occurring
radiologic review. in 71% of the patients), diarrhea (in 61%), and
† Responses could not be evaluated in 4 patients in each group because of early
death.
edema (in 49%); and the events for which the
‡ P<0.001 for the comparison between the two groups. The 95% confidence in- incidence was at least 5 percentage points higher
terval was calculated with the use of the exact method based on the F distri- in the chemotherapy group than in the crizo-
bution.
§ The time to tumor response was calculated from the date of randomization to
tinib group were fatigue (occurring in 38% of
the date of the first documentation of a partial or complete response as deter- the patients), anemia (in 32%), and neutropenia
mined by independent radiologic review. (in 30%) (Table 3). Most adverse events in the
¶ The duration of response was calculated from the date of the first documenta-
tion of a partial or complete response to the date of RECIST-defined progres-
two treatment groups were grade 1 or 2 in sever-
sion or death, with the use of the Kaplan–Meier method. ity. Grade 3 or 4 elevations of aminotransferase
levels occurred in 24 patients in the crizotinib
group (14%) and in 4 patients in the chemo-
lated with the Wilcoxon test (Fig. S3A in the therapy group (2%), but these elevations were
Supplementary Appendix) and 0.67 (95% CI, managed primarily with dose interruptions or
0.28 to 1.48) as calculated with the log-rank test dose reductions. Four hepatic events resulted in
(Fig. S3B in the Supplementary Appendix), indi- permanent discontinuation of treatment in the
cating that crossover may have confounded the crizotinib group: three events involved elevated
results of the primary overall survival analysis. aminotransferase levels only (one event of grade
Among patients randomly assigned to crizo- 3 elevation of both alanine and aspartate amino-
tinib, 65 of 89 patients with progressive disease transferase levels and one event each of grade 2
(73%) continued to receive crizotinib beyond and grade 3 elevation of the alanine aminotrans-
disease progression for a median of 3.1 months ferase level), and one event involved a grade 2
(range, 0.7 to 22.6). A total of 21 patients as- drug-induced liver injury that met the criteria for
signed to crizotinib (12%) subsequently received Hy’s law23 (elevated aminotransferase and total
platinum-based chemotherapy. At data cutoff, 79 bilirubin levels without evidence of cholestasis
patients who had been randomly assigned to [i.e., no elevated serum alkaline phosphatase
crizotinib (46%) and 62 patients assigned to level]) (see the Supplementary Appendix). An ad-
chemotherapy who had crossed over to crizo- ditional case that met the criteria for Hy’s law
tinib (36%) were still receiving crizotinib thera- occurred in a patient in the chemotherapy group
py. Eighteen patients in the chemotherapy group after crossover to crizotinib. No deaths from
who had progressive disease did not receive fol- hepatic dysfunction occurred. Grade 3 or 4 neu-
low-up therapy with crizotinib; additional de- tropenia occurred in 11% of patients in the
Table 3. Adverse Events from Any Cause in the As-Treated Population.*
Crizotinib Chemotherapy
Adverse Event (N = 171) (N = 169)†
Any Grade Grade 3 or 4 Any Grade Grade 3 or 4
number of patients (percent)

Higher frequency in crizotinib group
Vision disorder‡ 122 (71) 1 (1) 16 (9) 0
Diarrhea 105 (61) 4 (2) 22 (13) 1 (1)
Edema§ 83 (49) 1 (1) 21 (12) 1 (1)
Vomiting 78 (46) 3 (2) 60 (36) 5 (3)
Constipation 74 (43) 3 (2) 51 (30) 0
Elevated aminotransferases§ 61 (36) 24 (14) 22 (13) 4 (2)
Upper respiratory infection§ 55 (32) 0 21 (12) 1 (1)
Abdominal pain§ 45 (26) 0 20 (12) 0
Dysgeusia 45 (26) 0 9 (5) 0
Headache 37 (22) 2 (1) 25 (15) 0
Pyrexia 32 (19) 0 18 (11) 1 (1)
Dizziness§ 31 (18) 0 17 (10) 2 (1)
Pain in extremity 27 (16) 0 12 (7) 0
Higher frequency in chemotherapy group
Fatigue 49 (29) 5 (3) 65 (38) 4 (2)
Neutropenia§ 36 (21) 19 (11) 51 (30) 26 (15)
Stomatitis§ 24 (14) 1 (1) 34 (20) 2 (1)
Asthenia 22 (13) 0 41 (24) 2 (1)
Anemia§ 15 (9) 0 54 (32) 15 (9)
Leukopenia§ 12 (7) 3 (2) 26 (15) 9 (5)
Thrombocytopenia§ 2 (1) 0 31 (18) 11 (7)
Similar frequency in the two treatment groups
Nausea 95 (56) 2 (1) 99 (59) 3 (2)
Decreased appetite 51 (30) 4 (2) 57 (34) 1 (1)
Cough§ 39 (23) 0 33 (20) 0
Neuropathy§ 35 (20) 2 (1) 38 (22) 0
Dyspnea§ 30 (18) 5 (3) 26 (15) 4 (2)
* Adverse events are listed here if they were reported in 15% or more of patients in either treatment group; rates were
not adjusted for differences in treatment duration. Higher frequency indicates a difference of 5 percentage points or
more between groups; similar frequency indicates a difference of less than 5 percentage points between groups.
† Only events that occurred before crossover to crizotinib are included.
‡ The category of vision disorder comprised a cluster of adverse events including (in descending order of frequency in
the crizotinib group) visual impairment, photopsia, blurred vision, vitreous floaters, reduced visual acuity, diplopia, and
photophobia.
§ This item comprised a cluster of adverse events that may represent similar clinical symptoms or syndromes.
crizotinib group and in 15% in the chemothera- tinib group had interstitial lung disease, result-
py group, with no cases of febrile neutropenia ing in permanent discontinuation of crizotinib
reported with crizotinib and two with chemo- treatment.
therapy. Other grade 3 or 4 adverse events from Adverse events from any cause that were as-
any cause are shown in Table S4 in the Supple- sociated with permanent discontinuation of treat-
mentary Appendix. Two patients (1%) in the crizo- ment occurred in 12% of the patients in the

crizotinib group and in 14% of those in the che-

Figure 2 (facing page). Overall Change from Baseline
motherapy group (before crossover); the corre- in Global Quality of Life, Functioning Domains, and
sponding rates of adverse events deemed by the Symptoms.
investigator to be related to treatment that were Panel A shows the overall change from baseline in
associated with permanent discontinuation were global quality of life (QOL) and functioning domains
5% and 8%. One case of fatal pneumonitis, con- as assessed with the use of the European Organisation
for Research and Treatment of Cancer Quality-of-Life
sidered to be related to crizotinib treatment, oc-
Questionnaire (QLQ-C30). Panels B and C show the
curred in a patient who had crossed over from overall change from baseline in symptoms as assessed
chemotherapy. Grade 5 adverse events of any cause with the QLQ-C30 and the corresponding module for
are shown in Table S5 in the Supplementary Ap- lung cancer (QLQ-LC13), respectively. Patient-reported
pendix. With the exception of the fatal pneumo- outcomes were assessed at baseline, on days 7 and 15
of cycle 1, on day 1 of every subsequent cycle, and at
nitis, described above, that occurred after cross-
the end of treatment. Scores on each scale ranged
over to crizotinib, no deaths were reported that from 0 to 100. For global quality of life and functioning
were deemed by the investigators to be related to domains, higher scores indicate better global quality of
treatment. life or functioning, and hence positive changes (up-
ward bars) indicate improvement from baseline; for
Patient-Reported Outcomes symptoms, higher scores indicate greater severity of
symptoms, and hence negative changes (downward
Baseline scores on the QLQ-C30, QLQ-LC13, and bars) indicate improvement from baseline. A change of
EQ-5D are summarized in Table S6 in the Sup- 10 points or more is considered to be a clinically
plementary Appendix. There was a significantly meaningful change. An asterisk indicates P<0.001, and
greater overall improvement from baseline in a dagger P<0.05 for the comparison between treat-
ment groups. In Panel C, the mean changes from the
global quality of life among patients who re-
baseline score in dysphagia and in pain in the chest
ceived crizotinib than among those who received with chemotherapy were 0.10 and −0.05, respectively.
chemotherapy (P<0.001) (Fig. 2A, and see the
Results section in the Supplementary Appendix
for additional details). Crizotinib was also associ- scale) with crizotinib than with chemotherapy
ated with a significantly greater overall improve- (P = 0.002).
ment from baseline in physical, social, emo-
tional, and role functioning domains (P<0.001) Discussion
(Fig. 2A).
There was a significantly greater overall re- This study showed the superiority of first-line
duction from baseline with crizotinib than with therapy with crizotinib over pemetrexed-plus-
chemotherapy in the symptoms of pain, dyspnea, platinum chemotherapy in patients with previ-
and insomnia as assessed with the use of the ously untreated advanced ALK-positive NSCLC.
QLQ-C30 (Fig. 2B) and in the symptoms of dys- Initial treatment with crizotinib significantly pro-
pnea, cough, chest pain, arm or shoulder pain, longed progression-free survival as compared with
and pain in other parts of the body as assessed chemotherapy consisting of pemetrexed plus cis-
with the use of the QLQ-LC13 (Fig. 2C) (P<0.001 platin or carboplatin. These results were inde-
for all comparisons) (see the Results section in pendent of the type of platinum treatment admin-
the Supplementary Appendix for additional de- istered, the performance status of the patient,
tails). Patients treated with crizotinib also had a the patient’s race, and the presence or absence
significantly greater delay in the worsening of of brain metastases. Crizotinib treatment was
lung-cancer symptoms (a composite of cough, also associated with a significantly higher re-
dyspnea, or pain in the chest) than did patients sponse rate and significantly greater improve-
treated with chemotherapy (hazard ratio for wors- ments in patient-reported measures of physical
ening of symptoms with crizotinib, 0.59; 95% CI, functioning, key lung-cancer symptoms (cough,
0.45 to 0.77; P < 0.001; estimated probability of dyspnea, chest pain, and fatigue), and global
being event-free at 6 months, 38% vs. 22%) (Fig. quality of life.
S5 in the Supplementary Appendix). A signifi- The standard of care for newly diagnosed
cantly greater improvement from baseline was NSCLC has generally been platinum-based dou-
observed in EQ-5D general health status scores ble-agent chemotherapy,11 except in the case of
(as assessed with the use of a visual-analogue NSCLC that is positive for an epidermal growth
A Global Quality of Life and Functioning Domains (QLQ-C30)

Improvement
10
*
Mean Change from Baseline Score
* *
5 * *
†
0
−5
−10
−15
ng
ng
tio e
L
ng
ng
tio ole
tio al
tio al
tio al
ng
nc itiv
O
ni
ni
nc oci
nc on
nc sic
ni
ni
Q
nc R
ni
Fu og n
Fu oti
Fu Phy
Fu S
bal
Em
C
lo
Fu
G
B Symptoms (QLQ-C30)
15
*
10
5 †
Reduction in Symptoms
0 * * * * *
−5
−10
−15
−20
ng
ue
in
ea
ss
ea
Vo and
ni
io
Pa
Lo
pn
rrh
tig
iti
at
m
ys
Fa
tip
so
a
ia
se
tit
D
D
In
ns
pe
au
Co
Ap
N
C Symptoms (QLQ-LC13)
10
5 †
* * † * * *
0
Reduction in Symptoms
−5
−10
−15
−20
a
sis
ia
e
y
ou m
er
es
er
e
gi
th
pa l
ug
ec
ro ra
ou
Sh Ar
pn
ld
ty
Ch
ha
wh
eu he
Co
op
op
M
ys
or in
p
se
N rip
in
Al
ys
D
em
in
El
r
n
D
Pe
So
Pa
i
H
Pa
in
Pa
factor receptor (EGFR) mutation, for which ran- apy.24-28 For tumors with nonsquamous histologic
domized trials have shown superior efficacy of characteristics, cisplatin−pemetrexed has been
EGFR tyrosine kinase inhibitors over chemother- shown to be superior to cisplatin−gemcitabine.12

Given that most advanced ALK-positive NSCLCs phase 3 studies of first-line EGFR tyrosine ki-
have nonsquamous histologic characteristics, nase inhibitors versus chemotherapy in EGFR-
pemetrexed in combination with cisplatin or mutation–positive NSCLC, this finding is most
carboplatin was selected as the standard chemo- likely attributable to the confounding effects of
therapy for this trial. The efficacy of peme- crossover treatment.32 Of the 171 patients ran-
trexed-based first-line chemotherapy has since domly assigned to chemotherapy, 120 received
been documented in ALK-positive NSCLC,29,30 a crizotinib treatment during follow-up for sur-
finding that supports this selection. A potential vival. It should be noted that the median sur-
limitation of our study was that pemetrexed was vival had not been reached in either group, with
not continued beyond the planned six cycles of a median follow-up of 17 months.
pemetrexed-plus-platinum chemotherapy, since The safety profile of crizotinib was consistent
this was not considered to be a standard ap- with that reported earlier in patients with previ-
proach when the study was initiated. However, ously treated advanced ALK-positive NSCLC10 and
in a study of patients without disease progression differed from that observed with chemotherapy.
after four cycles of cisplatin−pemetrexed, main- The incidence of adverse effects in the two treat-
tenance pemetrexed therapy improved median ment groups was probably affected by the fact
progression-free survival over placebo by only that the duration of therapy with crizotinib was
1.3 months (4.1 months vs. 2.8 months) from longer than that with chemotherapy and that
the start of maintenance therapy.31 The way in crizotinib continued to be used in some patients
which the use of maintenance pemetrexed ther- beyond progression.33 Discontinuations of thera-
apy or other chemotherapy regimens would have py occurred in 5% of patients with crizotinib-
affected the results in the control group of the related adverse events and in 8% of patients with
current study is unclear. chemotherapy-related adverse events. More seri-
The magnitude of the improvement in pro- ous potential adverse events previously reported
gression-free survival observed in the current with crizotinib were hepatotoxic and pulmonary
study is similar to that observed in studies of toxic effects.10 In the current study, grade 3 or 4
EGFR-mutation–positive tumors treated with first- elevations of aminotransferase levels occurred in
line EGFR tyrosine kinase inhibitors.24-26 Although 14% of the patients in the crizotinib group and
formal comparison across studies cannot be made, could be managed with dose interruptions or
the efficacy of crizotinib in the first-line setting dose reductions. Two patients discontinued crizo-
(median progression-free survival, 10.9 months; tinib therapy because of interstitial lung disease,
objective response rate, 74%) appeared to be and one case of fatal pneumonitis was reported
greater than that seen with crizotinib in an oth- in a patient who had crossed over from chemo-
erwise similar patient population that had re- therapy to crizotinib.
ceived previous treatment with platinum-based In conclusion, in patients with previously un-
chemotherapy (median progression-free survival, treated ALK-positive NSCLC, crizotinib treatment
7.7 months; response rate, 65%).10 Initiating crizo- was superior to pemetrexed-plus-platinum chemo-
tinib as first-line therapy in patients whose tu- therapy with respect to progression-free survival,
mors test positive for ALK rearrangements maxi- objective response rate, reduction of lung-cancer
mizes the probability that these patients will symptoms, and improvement in quality of life.
benefit from ALK-directed therapy. Supported by Pfizer.
Disclosure forms provided by the authors are available with
Overall survival did not differ significantly the full text of this article at NEJM.org.
between the treatment groups at the time of this We thank the participating patients and their families; the
analysis, with a relatively small number of deaths research nurses, study coordinators, and operations staff;
Joanne Fitz-Gerald and Wendy Sacks at ACUMED (Tytherington,
reported (26%; 90 of the 343 patients who un- United Kingdom, and New York) for medical writing support;
derwent randomization). As seen in randomized and Abbott Molecular for support of the ALK FISH testing.
References
1. Soda M, Choi YL, Enomoto M, et al. al survey of phosphotyrosine signaling tients with non-small-cell lung cancer
Identification of the transforming EML4- identifies oncogenic kinases in lung can- who harbor EML4-ALK. J Clin Oncol
ALK fusion gene in non-small-cell lung cer. Cell 2007;131:1190-203. 2009;27:4247-53.
cancer. Nature 2007;448:561-6. 3. Shaw AT, Yeap BY, Mino-Kenudson M, 4. Camidge DR, Doebele RC. Treating
2. Rikova K, Guo A, Zeng Q, et al. Glob- et al. Clinical features and outcome of pa- ALK-positive lung cancer — early suc-
cesses and future challenges. Nat Rev 15. Oken MM, Creech RH, Tormey DC, et domised phase 3 trial. Lancet Oncol 2012;
Clin Oncol 2012;9:268-77. al. Toxicity and response criteria of the 13:239-46.
5. Blackhall FH, Peters S, Bubendorf L, Eastern Cooperative Oncology Group. Am 26. Sequist LV, Yang JC, Yamamoto N, et
et al. Prevalence and clinical outcomes for J Clin Oncol 1982;5:649-55. al. Phase III study of afatinib or cisplatin
patients with ALK-positive resected stage 16. Aaronson NK, Ahmedzai S, Bergman plus pemetrexed in patients with meta-
I-III adenocarcinoma: results from the B, et al. The European Organization for static lung adenocarcinoma with EGFR
European Thoracic Oncology Platform Research and Treatment of Cancer QLQ- mutations. J Clin Oncol 2013;31:3327-34.
Lungscape Project. J Clin Oncol 2014;32: C30: a quality-of-life instrument for use 27. Maemondo M, Inoue A, Kobayashi K,
2780-7. in international clinical trials in oncolo- et al. Gefitinib or chemotherapy for non–
6. Christensen JG, Zou HY, Arango ME, gy. J Natl Cancer Inst 1993;85:365-76. small-cell lung cancer with mutated
et al. Cytoreductive antitumor activity of 17. Fayers P, Bottomley A. Quality of life EGFR. N Engl J Med 2010;362:2380-8.
PF-2341066, a novel inhibitor of anaplas- research within the EORTC-the EORTC 28. Mitsudomi T, Morita S, Yatabe Y, et al.
tic lymphoma kinase and c-Met, in ex- QLQ-C30. Eur J Cancer 2002;38:Suppl 4: Gefitinib versus cisplatin plus docetaxel
perimental models of anaplastic large- S125-S133. in patients with non-small-cell lung can-
cell lymphoma. Mol Cancer Ther 2007;6: 18. Bergman B, Aaronson NK, Ahmedzai cer harbouring mutations of the epidermal
3314-22. S, Kaasa S, Sullivan M. The EORTC QLQ- growth factor receptor (WJTOG3405): an
7. Kwak EL, Bang YJ, Camidge DR, et al. LC13: a modular supplement to the open label, randomised phase 3 trial.
Anaplastic lymphoma kinase inhibition EORTC Core Quality of Life Question- Lancet Oncol 2010;11:121-8.
in non–small-cell lung cancer. N Engl J naire (QLQ-C30) for use in lung cancer 29. Camidge DR, Kono SA, Lu X, et al.
Med 2010;363:1693-703. clinical trials. Eur J Cancer 1994;30A:635- Anaplastic lymphoma kinase gene rear-
8. Camidge DR, Bang YJ, Kwak EL, et al. 42. rangements in non-small cell lung cancer
Activity and safety of crizotinib in pa- 19. EuroQol Group. EuroQol — a new fa- are associated with prolonged progres-
tients with ALK-positive non-small-cell cility for the measurement of health-relat- sion-free survival on pemetrexed. J Thorac
lung cancer: updated results from a phase ed quality of life. Health Policy 1990;16: Oncol 2011;6:774-80.
1 study. Lancet Oncol 2012;13:1011-9. 199-208. 30. Shaw AT, Varghese AM, Solomon BJ,
9. Kim D-W, Ahn M-J, Shi Y, et al. Re- 20. Robins JM, Tsiatis A. Correcting for et al. Pemetrexed-based chemotherapy in
sults of a global phase II study with crizo- non-compliance in randomized trials us- patients with advanced, ALK-positive
tinib in advanced ALK-positive non-small ing rank-preserving structural failure non-small cell lung cancer. Ann Oncol
cell lung cancer (NSCLC). J Clin Oncol time models. Commun Stat Theory Meth- 2013;24:59-66.
2012;30:Suppl. abstract. ods 1991;20:2609-31. 31. Paz-Ares L, de Marinis F, Dediu M, et
10. Shaw AT, Kim DW, Nakagawa K, et al. 21. Robins JM, Blevins D, Ritter G, Wul- al. Maintenance therapy with pemetrexed
Crizotinib versus chemotherapy in ad- fsohn M. G-estimation of the effect of plus best supportive care versus placebo
vanced ALK-positive lung cancer. N Engl J prophylaxis therapy for Pneumocystis ca- plus best supportive care after induction
Med 2013;368:2385-94. rinii pneumonia on the survival of AIDS therapy with pemetrexed plus cisplatin
11. Schiller JH, Harrington D, Belani CP, patients. Epidemiology 1992;3:319-36. for advanced non-squamous non-small-
et al. Comparison of four chemotherapy 22. White IR, Babiker AG, Walker S, Dar- cell lung cancer (PARAMOUNT): a dou-
regimens for advanced non–small-cell byshire JH. Randomization-based meth- ble-blind, phase 3, randomised controlled
lung cancer. N Engl J Med 2002;346:92-8. ods for correcting for treatment changes: trial. Lancet Oncol 2012;13:247-55.
12. Scagliotti GV, Parikh P, von Pawel J, et examples from the Concorde trial. Stat 32. Mok T, Yang JJ, Lam KC. Treating pa-
al. Phase III study comparing cisplatin Med 1999;18:2617-34. tients with EGFR-sensitizing mutations:
plus gemcitabine with cisplatin plus peme- 23. Temple R. Hy’s law: predicting seri- first line or second line — is there a dif-
trexed in chemotherapy-naive patients ous hepatotoxicity. Pharmacoepidemiol ference? J Clin Oncol 2013;31:1081-8.
with advanced-stage non-small-cell lung Drug Saf 2006;15:241-3. 33. Shaw AT, Solomon BJ, Mok T, et al.
cancer. J Clin Oncol 2008;26:3543-51. 24. Mok TS, Wu YL, Thongprasert S, et al. Effect of treatment duration on incidence
13. Abbott Molecular. Vysis ALK Break Gefitinib or carboplatin–paclitaxel in of adverse events (AEs) in a phase III study
Apart FISH Probe Kit package insert, 2011 pulmonary adenocarcinoma. N Engl J of crizotinib versus chemotherapy in ad-
(http://www.abbottmolecular.com/static/ Med 2009;361:947-57. vanced ALK-positive non-small cell lung
cms_workspace/pdfs/US/Vysis_ALK_ 25. Rosell R, Carcereny E, Gervais R, et cancer (NSCLC). Presented at the 15th
FISH_Probe_Kit_PI.pdf). al. Erlotinib versus standard chemothera- World Conference on Lung Cancer, Syd-
14. Eisenhauer EA, Therasse P, Bogaerts J, py as first-line treatment for European ney, October 27–30, 2013. abstract.
et al. New response evaluation criteria in patients with advanced EGFR mutation- Copyright © 2014 Massachusetts Medical Society.
solid tumours: revised RECIST guideline positive non-small-cell lung cancer
(version 1.1). Eur J Cancer 2009;45:228-47. (EURTAC): a multicentre, open-label, ran-
receive the journal’s table of contents each week by e-mail
To receive the table of contents of the Journal by e-mail

every Wednesday evening, sign up at NEJM.org.


1L CrizoVSQT ALK+

Uploaded by

Copyright:

Available Formats

1L CrizoVSQT ALK+

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1L CrizoVSQT ALK+

Uploaded by

Copyright:

Available Formats

The n e w e ng l a n d j o u r na l of m e dic i n e

First-Line Crizotinib versus Chemotherapy

(Funded by Pfizer; PROFILE 1014 ClinicalTrials.gov number, NCT01154140.)

n engl j med 371;23 nejm.org December 4, 2014 2167

2168 n engl j med 371;23 nejm.org December 4, 2014

n engl j med 371;23 nejm.org December 4, 2014 2169

2170 n engl j med 371;23 nejm.org December 4, 2014

A Progression-free Survival B Overall Survival

Overall Survival (%)

C Progression-free Survival, According to Subgroup

Crizotinib Better Chemotherapy

n engl j med 371;23 nejm.org December 4, 2014 2171

Table 2. Response to Treatment in the Intention-to-Treat Population.*

2172 n engl j med 371;23 nejm.org December 4, 2014

Table 3. Adverse Events from Any Cause in the As-Treated Population.*

Any Grade Grade 3 or 4 Any Grade Grade 3 or 4

number of patients (percent)

n engl j med 371;23 nejm.org December 4, 2014 2173

crizotinib group and in 14% of those in the che-

2174 n engl j med 371;23 nejm.org December 4, 2014

A Global Quality of Life and Functioning Domains (QLQ-C30)

n engl j med 371;23 nejm.org December 4, 2014 2175

2176 n engl j med 371;23 nejm.org December 4, 2014

receive the journal’s table of contents each week by e-mail

To receive the table of contents of the Journal by e-mail

n engl j med 371;23 nejm.org December 4, 2014 2177

You might also like