2013EdelmanRandomizedphaseIIstudyofixabe

Published Ahead of Print on April 15, 2013 as 10.1200/JCO.2012.45.3282
The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2012.45.3282
JOURNAL OF CLINICAL ONCOLOGY
O R I G I N A L
R E P O R T
Randomized Phase II Study of Ixabepilone or Paclitaxel Plus
Carboplatin in Patients With Non–Small-Cell Lung Cancer
Prospectively Stratified by Beta-3 Tubulin Status
Martin J. Edelman, Claus-Peter Schneider, Chun-Ming Tsai, Heung-Tae Kim, Elisabeth Quoix,
Alexander V. Luft, Remigiusz Kaleta, Pralay Mukhopadhyay, Ovidiu C. Trifan, Laura Whitaker,
and Martin Reck
Martin J. Edelman, University of Maryland Greenebaum Cancer Center, Baltimore, MD; Claus-Peter Schneider,
Zentralklinik Bad Berka, Bad Berka;
Martin Reck, Hospital Grosshansdorf,
Grosshansdorf, Germany; Chun-Ming
Tsai, Taipei Veterans General Hospital,
Taipei, Taiwan; Heung-Tae Kim,
National Cancer Center, Goyang-Si,
Korea; Elisabeth Quoix, Nouvel Hopital
Civil, Strasbourg, France; Alexander V.
Luft, Leningrad Regional Clinical Hospital, St Petersburg, Russia; and Remigiusz Kaleta, Pralay Mukhopadhyay,
Ovidiu C. Trifan, and Laura Whitaker,
Bristol-Myers Squibb, Wallingford, CT.
Published online ahead of print at
www.jco.org on April 15, 2013.
Supported by Bristol-Myers Squibb.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this
article.
Clinical trial information: NCT00723957.
Corresponding author: Martin J. Edelman, MD, University of Maryland
Greenebaum Cancer Center, Division of
Hematology/Oncology, 22 South Green
St, Baltimore, MD 21201-1595; e-mail:
[email protected].
© 2013 by American Society of Clinical
Oncology
0732-183X/13/3199-1/$20.00
A
B
S
T
R
A
C
T
Purpose
Retrospective studies have reported that tumor expression of the beta-3 tubulin (␤3T) isoform is
an unfavorable prognostic factor in non–small-cell lung cancer (NSCLC) treated with tubulininhibiting chemotherapy. Ixabepilone is a tubulin-inhibiting agent with low susceptibility to multiple
resistance mechanisms including ␤3T isoform expression in several tumor models. This randomized phase II study evaluated ixabepilone-based chemotherapy in stage IIIb/IV NSCLC,
compared with paclitaxel-based chemotherapy. Tumor specimens were prospectively evaluated for ␤3T expression.
Patients and Methods
Patients were stratified by ␤3T status (positive v negative) and randomly assigned at a ratio of 1:1
to receive ixabepilone (32 mg/m2) and carboplatin (area under concentration-time curve [AUC], 6)
or paclitaxel (200 mg/m2) and carboplatin (AUC, 6) for up to six cycles. The primary end point was
progression-free survival (PFS) in the ␤3T-positive subgroup.
Results
Ninety-five patients (␤3T positive, 52; ␤3T negative, 43) received ixabepilone plus carboplatin; 96
patients (␤3T positive, 49; ␤3T negative, 47) received paclitaxel plus carboplatin. No significant
differences in median PFS were observed between arms for either subgroup (␤3T positive, 4.3
months in both arms; ␤3T negative, 5.8 v 5.3 months). Ixabepilone did not significantly improve
overall survival (OS) for the ␤3T-positive subset or the overall population. Adverse events were
similar between the two arms and comparable with those in previous studies.
Conclusion
There was no predictive value of ␤3T in differentiating clinical activity of ixabepilone- or
paclitaxel-containing regimens. Ixabepilone did not improve PFS or OS in patients with ␤3Tpositive tumors. ␤3T-positive patients had worse PFS relative to ␤3T-negative patients, regardless
of treatment; hence, ␤3T expression seems to be a negative prognostic factor, but not a predictive
factor, in advanced NSCLC treated with either ixabepilone or paclitaxel platinum-based doublets.
J Clin Oncol 31. © 2013 by American Society of Clinical Oncology
DOI: 10.1200/JCO.2012.45.3282
INTRODUCTION
Currently accepted first-line chemotherapy for stage
IV non–small-cell lung cancer (NSCLC) includes a
platinum-based agent in combination with vinorelbine, gemcitabine, or a taxane.1,2 The combinations
of platinum plus pemetrexed and platinum, paclitaxel, and bevacizumab have recently demonstrated
some advantage over other regimens in nonsquamous carcinoma.3,4 In an attempt to individualize
treatment to improve patient survival and effectiveness of therapy, clinical research has been focused on
identification of biologic markers that are either
prognostic and/or predictive of response to specific
cytotoxic drugs (eg, DNA repair enzymes, ribonucleotide subunits, and beta-3 tubulin [␤3T]).5,6 Increased ␤3T expression is associated with an
increased rate of metastasis, and it is thought that
expression levels of ␤3T might be predictive for response to chemotherapy, especially taxanes.7 Several
studies have indicated that ␤3T overexpression
might be a common mechanism for taxane resistance in advanced tumors.8-10
Epothilones are a novel class of antimicrotubule agents, which, similarly to taxanes, bind to and
stabilize microtubules, thereby resulting in mitotic
© 2013 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
Copyright 2013 by American Society of Clinical Oncology
1
Edelman et al
arrest at G2/M transition.11 Ixabepilone, a semisynthetic, metabolically stable derivative of epothilone B, has shown clinical activity in
several tumor types, especially in tumors resistant to other cytotoxic
agents.12,13 Of interest, several preclinical studies have demonstrated
activity of ixabepilone in taxane-resistant disease. The taxane-resistant
breast cancer model Pat-21 demonstrates overexpression of ␤3T but is
sensitive to an epothilone.14 Additional studies have demonstrated
activity with ixabepilone in the ␤3T-overexpressing NSCLC tumor
models H1155 and LX-1, which are refractory to docetaxel and vinorelbine.15 In a phase II study of patients with advanced NSCLC
previously treated with a platinum-containing regimen, single-agent
ixabepilone demonstrated clinically relevant activity.16 Significantly,
responses were seen in patients who had progressed after taxanecontaining therapy.
Given the antitumor activity of ixabepilone in taxane-resistant
tumors and the observation that tumors overexpressing ␤3T respond
poorly to taxanes, we hypothesized that ixabepilone may show clinical
activity superior to that of paclitaxel when administered as initial
combination treatment with carboplatin to patients with NSCLC who
have tumors that overexpress ␤3T. Here, we present the results of a
phase II study of ixabepilone plus carboplatin versus paclitaxel plus
carboplatin in patients with advanced NSCLC who were prospectively
analyzed and stratified based on ␤3T status.
PATIENTS AND METHODS
Study Design and Treatment
This was a randomized, international, multicenter, open-label, phase II
study of ixabepilone in combination with carboplatin, compared with paclitaxel and carboplatin, in patients with NSCLC who had not received any
systemic therapy for their cancer. Patients were randomly assigned at a ratio of
1:1 and stratified by tumor expression level of ␤3T (positive v negative), disease
stage (stage IIIb v IV), presence of brain metastases (yes v no), and study site to
one of the following arms: ixabepilone administered as a 3-hour intravenous
infusion at a starting dose of 32 mg/m2 on day 1 of a 21-day cycle, consistent
with a previous phase II trial of single-agent ixabepilone,16 followed by carbo-
platin administered at a dose calculated to produce an area under the
concentration-time curve of 6 mg/mL/min on day 1 of a 21-day cycle; or
paclitaxel administered as a 3-hour intravenous infusion at a starting dose of
200 mg/m2 on day 1 of a 21-day cycle followed by carboplatin (area under
concentration-time curve, 6) administered on day 1 of a 21-day cycle. The
recommended phase II dose of ixabepilone and carboplatin was determined as
30 mg/m2 in a previously reported phase I study17; however, a dose of 32
mg/m2 was used in this study, consistent with that used as a second-line single
agent in a previous phase II study.16 All randomly assigned patients were to be
treated for a maximum of six cycles or until disease progression or unacceptable toxicity.
The protocol was conducted in accordance with Good Clinical Practice,
as defined by the International Conference on Harmonisation and in accordance with the ethical principles underlying European Union Directive 2001/
20/EC and US Code of Federal Regulations, Title 21, Part 50. Before initiation,
this study was approved by the institutional review board or independent
ethics committee at each site. All patients provided written informed consent.
Patients
Eligible patients were men and women age ⱖ 18 years who had histologically confirmed NSCLC (stage IIIb with pleural effusion, stage IV disease, or
recurrent disease after surgery and/or radiation therapy), Karnofsky performance status of 70 to 100, and life expectancy of at least 3 months. Availability
of paraffin-embedded tissue (to measure expression levels of ␤3T) and measurable disease by RECIST (version 1.0) guidelines with at least one target
lesion outside of a previous radiotherapy field were required for study entry.
Patients with known uncontrolled brain metastases were excluded from the
study, and those with known metastases were required to have a baseline
imaging scan within 4 weeks of random assignment; patients with peripheral
neuropathy grade ⬎ 1 were also excluded. Radiation therapy or major surgery
and focal/palliative radiation therapy or minor surgery were to be completed 4
weeks and 1 week before random assignment, respectively. Patients with
inadequate hematologic (absolute neutrophil count [ANC] ⬍ 1,500 cells/␮L,
hemoglobin ⬍ 9.0 g/dL, and platelets ⬍ 100,000 cells/␮L), hepatic (total
bilirubin ⬎ upper limit of normal), or renal function (serum creatinine ⱖ 1.5
mg/dL; calculated creatinine clearance based on Cockcroft-Gault ⬍ 50 mL/
min) were excluded.
Study Assessments
␤3T expression levels in tissue specimens were assessed using immunohistochemistry (IHC) staining. Tissue sections that either had available
Random assignment
(N = 197)
Ixabepilone + carboplatin
Received treatment
Did not receive treatment
(n = 98)
(n = 95)
(n = 3)
Paclitaxel + carboplatin
Received treatment
Did not receive treatment
((n
n=9
99)
8)
(n = 96)
(n = 3)
Fig 1. CONSORT diagram; patient disposition. AE, adverse event.
Completed treatment
Still on treatment
Discontinued treatment
Disease progression
Study drug toxicity
AE unrelated to study drug
Death
Patient request
Maximum clinical benefit
Withdrawal of consent
Other
2
(n = 47)
(n = 1)
(n = 27)
(n = 4)
(n = 1)
(n = 3)
(n = 3)
(n = 4)
(n = 4)
(n = 1)
© 2013 by American Society of Clinical Oncology
Completed treatment
Still on treatment
Discontinued treatment
Disease progression
Study drug toxicity
AE unrelated to study drug
Death
Patient request
Maximum clinical benefit
Withdrawal of consent
Other
(n = 46)
(n = 0)
(n = 25)
(n = 6)
(n = 4)
(n = 4)
(n = 3)
(n = 5)
(n = 2)
(n = 1)
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
Ixabepilone Plus Carboplatin in Advanced NSCLC
Table 1. Baseline Patient Demographics and Clinical Characteristics
Ixabepilone
Plus
Carboplatin
(n ⫽ 98)
Characteristic (overall population)
Age, years
Median
Range
Sex
Male
Female
Ethnicity
White
Asian
Other
Tumor histology
Adenocarcinoma
Bronchioloalveolar carcinoma
Large-cell carcinoma
Squamous cell carcinoma
Other/unknown
Karnofsky performance status, %
100
90
80
70
Not reported
␤3Tⴱ
Positive
Negative
Disease stageⴱ
IIIb
IV
Presence of brain metastasisⴱ
No
Yes
Smoking status
Current
Former
Never
Unknown
No.
%
60.0
29.0-80.0
%
60.0
34.0-85.0
73.5
26.5
67
32
67.7
32.3
68
30
0
69.0
30.6
0.0
76
22
1
76.8
22.2
1.0
51
1
3
38
5
52.0
1.0
3.1
38.8
5.1
55
0
5
34
5
55.6
0.0
5.1
34.3
5.1
9
35
41
13
0
9.2
35.7
41.8
13.3
0.0
10
43
35
10
1
10.1
43.4
35.4
10.1
1.0
53
45
54.1
45.9
51
48
51.5
48.5
7
91
7.1
92.9
8
91
8.1
91.9
89
9
90.8
9.2
90
9
90.9
9.1
49
28
18
3
50.0
28.6
18.4
3.1
43
31
22
3
43.4
31.3
22.2
3.0
No.
%
Age, years
Median
60.0
Range
29.0-80.0
Sex
Male
40
75.5
Female
13
24.5
Karnofsky performance status, %
100
5
9.4
90
15
28.3
80
24
45.3
70
9
17.0
Not reported
0
0.0
Tumor histology
Adenocarcinoma
30
56.6
Bronchioloalveolar carcinoma
1
1.9
(continued in next column)
www.jco.org
No.
72
26
Ixabepilone
Plus
Carboplatin
(n ⫽ 53)
Characteristic (␤3T-positive patients)
Paclitaxel
Plus
Carboplatin
(n ⫽ 99)
Paclitaxel
Plus
Carboplatin
(n ⫽ 51)
No.
%
60.0
43.0-80.0
32
19
62.7
37.3
7
22
16
5
1
13.7
43.1
31.4
9.8
2.0
33
0
64.7
0.0
Table 1. Baseline Patient Demographics and Clinical
Characteristics (continued)
Ixabepilone
Plus
Carboplatin
(n ⫽ 53)
Characteristic (␤3T-positive patients)
Large-cell carcinoma
Squamous cell carcinoma
Other/unknown
Paclitaxel
Plus
Carboplatin
(n ⫽ 51)
No.
%
No.
%
2
17
3
3.8
32.1
5.7
2
13
3
3.9
25.5
5.9
Abbreviation: ␤3T, beta-3 tubulin.
ⴱ
Stratification factor.
formalin-fixed, paraffin-embedded tumor tissue or required rebiopsy of tumor tissue were submitted to Dako North America (Carpinteria, CA) for ␤3T
IHC analysis. Formalin-fixed, paraffin-embedded tissue blocks were sectioned
and adhered to positively charged glass slides. ␤3T staining was assessed using
a prototype pharmacodiagnostic IHC assay that was optimized by Dako
(Glostrup, Denmark). The assay was based on previously reported IHC assays
for ␤3T.8,18 Tissue sections that contained ⱖ 50% of tumor cells staining at ⱖ
2⫹ intensity (on scale of 0, 1⫹, 2⫹, and 3⫹, where 0 was no staining and 3⫹
was maximum intensity) were scored as ␤3T positive, whereas all others were
scored as ␤3T negative. Endothelial cells within tissue sections consistently
stained at an intensity ⱖ 2⫹ with this assay and thus served as an internal
positive control. ␤3T testing was performed before random assignment, and
␤3T status was used as a stratification factor at random assignment.
Tumor evaluations (physical examination and imaging studies) were
performed at baseline and every 6 weeks during treatment until documented
disease progression. Response was assessed according to revised RECIST (version 1.0), with a repeat assessment to confirm response performed at ⱖ 4
weeks. Patients with objective responses who discontinued treatment for reasons other than disease progression were to undergo regular tumor assessment
until disease progression.
All patients underwent physical examination, including vital signs and
performance status, serum chemistries (lactate dehydrogenase, blood urea
nitrogen or urea, creatinine, ALT, AST, alkaline phosphatase, and total bilirubin), and hematology (complete blood count plus differential and platelets)
before random assignment. Thereafter, hematology was evaluated weekly, and
serum chemistry was evaluated before each cycle. Creatinine clearance based
on Cockroft-Gault was calculated before each cycle. Toxicities were evaluated
continuously during the treatment phase, and patients who experienced any
drug-related toxicity were observed at least every 4 weeks during the follow-up
phase until the toxicities were resolved, were stabilized, returned to baseline, or
were deemed irreversible.
Adverse events, laboratory values, and other symptoms were graded
according to the National Cancer Institute Common Toxicity Criteria (version
3.0). Toxicities requiring a dose reduction were ANC ⬍ 500/␮L lasting ⱖ 5
days, febrile neutropenia, platelets ⬍ 25,000/␮L, grade 4 hemoglobin, grade 3
nausea/vomiting, grade 3 stomatitis, grade 3 diarrhea, neuropathy (grade 2
lasting ⬎ 7 days or grade 3 lasting ⬍ 7 days), and other grade ⱖ 3 toxicities.
Patients experiencing grade 3 neuropathy lasting ⱖ 7 days or grade 4 neuropathy were to discontinue treatment with ixabepilone and paclitaxel. Retreatment criteria included ANC ⬎ 1500/␮L, platelets ⬎ 100,000/␮L, and
resolution of treatment-related nonhematologic toxicity to baseline or
grade ⱕ 1 (excluding grade 2 alopecia and grade 2 fatigue). Patients with grade
2 neuropathy were not re-treated until improvement to grade 1 or 0. Patients
who did not recover from a treatment-related toxicity to baseline or grade ⱕ 1
within 3 weeks of scheduled re-treatment were required to discontinue using
the drug causing the delay or discontinue study participation. Patients requiring more than two dose reductions from any of the study drugs were also
required to discontinue study participation. No dose re-escalations were permitted during the study.
© 2013 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
3
Edelman et al
A
Ixa + carb
Pac + carb
Progression-Free
Survival (proportion)
1.0
0.8
Stratified log-rank P = .632
0.6
0.4
0.2
0
2
98
99
Efficacy
The primary end point of the study was a comparison of PFS
between patients with ␤3T-positive tumors receiving carboplatin plus
paclitaxel versus carboplatin plus ixabepilone (Fig 2). For the subgroup of patients with ␤3T-positive tumors, median PFS was 4.3
months in both treatment arms (Fig 2C; ixabepilone plus carboplatin:
80% CI, 3.2 to 4.9; paclitaxel plus carboplatin: 80% CI, 4.0 to 5.4; HR,
1.04; 80% CI, 0.78 to 1.41; P ⫽ .853). Therefore, the primary end point
of the study was not met. In the overall patient population, median
4
© 2013 by American Society of Clinical Oncology
10
12
14
16
22
26
69/98
75/99
15
11
4
4
2
1
0
1
0
0
Median (80% CI)
HR (80% CI)
5.3 (4.3 to 5.7)
5.1 (4.3 to 5.6)
0.92 (0.73 to 1.15)
Ixa + carb
Pac + carb
Stratified log-rank P = .35
0.6
0.4
0.2
2
4
6
8
10
12
14
16
1
0
0
0
0
0
Time (months)
45
48
36
36
24
28
11
11
No. Progressed/
No. Randomly Assigned
Ixa + carb
Pac + carb
Ixa + carb over pac + carb
27/45
35/48
9
4
3
3
Median (80% CI)
HR (80% CI)
5.8 (5.3 to 6.8)
5.3 (4.4 to 5.8)
0.78 (0.55 to 1.10)
Ixa + carb
Pac + carb
1.0
0.8
Stratified log-rank P = .853
0.6
0.4
0.2
0
No. at risk
Ixa + carb
Pac + carb
Group
52
57
0.8
No. at risk
Ixa + carb
Pac + carb
C
8
1.0
0
Progression-Free
Survival (proportion)
Patients and Treatment
A total of 197 patients with treatment-naive NSCLC were randomly assigned to receive either ixabepilone plus carboplatin (n ⫽ 98)
or paclitaxel plus carboplatin (n ⫽ 99) at 33 sites across 10 countries
from December 2, 2008, to February 23, 2010. Of these, 191 patients
were treated as follows: 95 with ixabepilone plus carboplatin and 96
with paclitaxel plus carboplatin (Fig 1). The baseline demographics
and disease characteristics were well balanced between the groups,
including the subgroup of patients with ␤3T-positive tumors (Table
1). Patients in the ixabepilone plus carboplatin arm received a median
of six cycles (range, one to six) of ixabepilone as well as carboplatin.
Patients in the paclitaxel plus carboplatin received a median of five
cycles (range, one to six) of paclitaxel and carboplatin.
76
75
Ixa + carb
Pac + carb
Ixa + carb over pac + carb
Group
RESULTS
6
No. Progressed/
No. Randomly Assigned
Group
B
4
Time (months)
No. at risk
Ixa + carb
Pac + carb
Progression-Free
Survival (proportion)
Statistical Analysis
The primary objective of this study was to test the hypothesis that
progression-free survival (PFS) in the ixabepilone plus carboplatin arm would
be superior to PFS in the paclitaxel plus carboplatin arm for the ␤3T-positive
subgroup. The sample size was calculated to provide 90% power to detect a
statistically significant difference in PFS between the two arms for the ␤3Tpositive subgroup using a two-sided ␣ ⫽ 0.2 log-rank test, if the true hazard
ratio (HR) were 0.58 (assuming median PFS of 3.2 v 5.5 months). Secondary
study objectives included comparing PFS for the entire population between
the two arms, estimating tumor response rates and survival in both arms, and
evaluating safety of the ixabepilone plus carboplatin combination. Patients
with ␤3T-negative tumors were also enrolled onto the study to estimate the
treatment effect in this subgroup and evaluate the predictive and prognostic
value of the biomarker.
PFS (time from random assignment to date of progression or death) was
estimated using Kaplan-Meier methodology. For the ␤3T-positive subgroup,
comparison of PFS between treatment arms was carried out using a two-sided
␣ ⫽ 0.2 log-rank test stratified by disease stage (stage IIIb v IV) and presence of
brain metastases (yes v no). An unadjusted Cox proportional hazards model
stratified by disease stage (stage IIIb v IV) and presence of brain metastases (yes
v no) was used to compute PFS HR and the associated two-sided 80% CI. In
addition, one-sided P values for log-rank tests and one-sided 90% CIs for HRs
were also reported. Two-sided 80% CIs for median PFS were calculated for
each treatment arm. The 95% CIs for the medians were also reported.
Similar methods were used for comparing PFS between the two arms for
the entire population; however, the log-rank test and unadjusted Cox proportional hazards model were stratified by ␤3T expression level (positive v negative), disease stage (stage IIIb v IV), and presence of brain metastases (yes v no).
For the ␤3T-negative subgroup, median PFS for each treatment arm and HRs
along with 80% CIs were reported.
Response and disease control rates and respective 80% CIs were estimated for each arm for the ␤3T-positive subgroup, ␤3T-negative subgroup,
and overall population. Similar analyses, as for PFS, were also conducted for
overall survival (OS) in the ␤3T-positive and -negative subsets and the overall population.
2
4
6
8
10
12
14
16
1
1
0
1
0
0
Time (months)
53
51
40
39
28
29
11
15
No. Progressed/
No. Randomly Assigned
Ixa + carb
Pac + carb
Ixa + carb over pac + carb
42/53
40/51
6
7
1
1
Median (80% CI)
HR (80% CI)
4.3 (3.2 to 4.9)
4.3 (4.0 to 5.4)
1.04 (0.78 to 1.41)
Fig 2. Progression-free survival in (A) overall patient population, (B) beta-3
tubulin (␤3T) –negative patients, and (C) ␤3T-positive patients. Carb, carboplatin;
HR, hazard ratio; Ixa, ixabepilone; Pac, paclitaxel.
PFS was 5.3 months (80% CI, 4.3 to 5.7) for ixabepilone plus carboplatin and 5.1 months (80% CI, 4.3 to 5.6) for patients treated with
paclitaxel plus carboplatin (Fig 2A; HR, 0.92; 80% CI, 0.73 to 1.15;
P ⫽ .632). For the ␤3T-negative subgroup of patients, median PFS was
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
Ixabepilone Plus Carboplatin in Advanced NSCLC
Overall Survival
(proportion)
1.0
Ixa + carb
Pac + carb
Censored
Censored
0.8
0.6
0.4
0.2
0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30 32
Time (months)
No. at risk
Positive
104 94 87 77 61 51 44 36 27 22 21 15 10
Negative
93 83 74 64 54 45 41 34 30 28 18 11 9
Group
No. of Deaths/
No. Randomly Assigned
Positive
Negative
Positive over negative
80/104
57/93
Median (80% CI)
7
1
4
0
2
0
0
0
HR (80% CI)
10.6 (9.2 to 12.4)
13.1 (10.7 to 14.5)
1.16 (0.92 to 1.46)
Fig 3. Overall survival by beta-3 tubulin status (positive v negative). Carb, carboplatin;
HR, hazard ratio; Ixa, Ixabepilone; Pac, paclitaxel.
5.8 months (80% CI, 5.3 to 6.8) for ixabepilone plus carboplatin and
5.3 months (80% CI, 4.4 to 5.8) for paclitaxel plus carboplatin (Fig 2B;
HR, 0.78; 80% CI, 0.55 to 1.10; P ⫽ .350).
OS, a secondary end point of the trial, also did not favor the
ixabepilone arm in the ␤3T-positive subset. In this subgroup, the
median OS for ixabepilone plus carboplatin was 10.61 months (80%
CI, 7.79 to 12.42); it was 11.37 months (80% CI, 9.23 to 13.93) for
patients treated with paclitaxel plus carboplatin (HR, 1.60; 80% CI,
1.20 to 2.10; P ⫽ .053). For the ␤3T-negative subgroup, the median
OS for ixabepilone plus carboplatin was 16.92 months (80% CI, 14.26
to 23.56); it was 9.40 months (80% CI, 8.25 to 12.29) for patients
treated with paclitaxel plus carboplatin (HR, 0.70; 80% CI, 0.50 to
0.90; P ⫽ .123). In the overall population, the median OS was 13.04
months (80% CI, 10.74 to 14.26) for ixabepilone plus carboplatin; it
was 10.15 months (80% CI, 9.00 to 12.29) for paclitaxel plus carboplatin (HR, 1.10; 80% CI, 0.90 to 1.40; P ⫽ .622).
Because of the observed differences in OS, noted in the ␤3Tnegative subgroup, additional exploratory analyses were conducted to
understand these results. In particular, a multivariate Cox regression
analysis was conducted, adjusting for the following prognostic factors:
sex, age, performance status, ethnicity, and smoking status. When
adjusted using the Cox regression model, the HR was 0.777 (80% CI,
0.528 to 1.144; P ⫽ .4030) in the ␤3T-negative subset. In the ␤3Tpositive subset, the HR was 1.497 (80% CI, 1.083 to 2.070; P ⫽ .1106)
using the multivariate analysis; in the overall population, the HR was
1.035 (80% CI, 0.817 to 1.311; P ⫽ .8526). ␤3T positivity has been
described as a negative prognostic factor. We evaluated PFS and OS in
patients who were ␤3T positive (n ⫽ 104) versus patients who were
␤3T negative (n ⫽ 93). PFS for ␤3T-positive patients was 4.27 months
(80% CI, 3.61 to 5.42); it was 5.59 months (80% CI, 4.63 to 6.47) for
␤3T-negative patients (HR, 1.32; 80% CI, 1.06 to 1.64). The OS for
␤3T-positive patients was 10.6 months (80% CI, 9.2 to 12.4); it was
13.1 months (80% CI, 10.7 to 14.5) for ␤3T-negative patients (HR,
1.16; 80% CI, 0.92 to 1.46; Fig 3).
The response rate for the ixabepilone plus carboplatin arm was
numerically lower than that for the paclitaxel plus carboplatin arm in
the subpopulation of patients with ␤3T-positive tumors (17%; 95%
CI, 8 to 30 v 29%; 95% CI, 18 to 44; Table 2). Similar results were seen
in the overall population, with response rates of 21% (95% CI, 14 to
31) and 28% (95% CI, 20 to 38) in the corresponding arms. An
identical response rate of 27% (95% CI, 15 to 42) was noted for the two
treatment arms in the subpopulation of patients with ␤3T-negative
tumors. An overall disease control rate of 77% to 78% was observed in
both treatment arms in the ␤3T-positive and -negative subpopulations as well as in the overall patient population (Table 2). A summary
of efficacy results is presented in Appendix Tables A1 and A2 (online only).
Safety
Adverse events were similar between the two arms and comparable to those seen in previous studies (Table 3). Treatment-related
adverse events were mostly mild to moderate (grade 1 or 2) in severity.
No grade 4 nonhematologic adverse events were reported.
Table 2. Response Rates
␤3T Positive
Ixabepilone
Plus
Carboplatin
(n ⫽ 53)
␤3T Negative
Paclitaxel Plus
Carboplatin
(n ⫽ 51)
Ixabepilone
Plus
Carboplatin
(n ⫽ 45)
Overall Population
Paclitaxel Plus
Carboplatin
(n ⫽ 48)
Ixabepilone
Plus
Carboplatin
(n ⫽ 98)
Paclitaxel Plus
Carboplatin
(n ⫽ 99)
Response
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
Tumor response
CR
PR
SD
PD
NDⴱ
Response rate†
Disease control rate‡
1
8
32
10
2
9
41
1.9
15.1
60.4
18.9
3.8
17.0
77.4
2
13
24
7
5
15
39
3.9
25.5
47.1
13.7
9.8
29.4
76.5
0
12
23
4
6
12
35
0.0
26.7
51.1
8.9
13.3
26.7
77.8
0
13
24
7
4
13
37
0.0
27.1
50.0
14.6
8.3
27.1
77.1
1
20
55
14
8
21
76
1.0
20.4
56.1
14.3
8.2
21.4
77.6
2
26
48
14
9
28
76
2.0
26.3
48.5
14.1
9.1
28.3
76.8
Abbreviations: ␤3T, beta-3 tubulin; CR, complete response; ND, not determined; PD, progressive disease; PR, partial response; SD, stable disease.
ⴱ
Data unavailable because of early discontinuation or death or because patient was never treated in study.
†CR ⫹ PR/No. of patients.
‡CR ⫹ PR ⫹ SD/No. of patients.
www.jco.org
© 2013 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
5
Edelman et al
Table 3. Toxicity
Ixabepilone Plus Carboplatin (n ⫽ 95)
Any Grade
Grade 3
Paclitaxel Plus Carboplatin (n ⫽ 96)
Grade 4
Any Grade
Grade 3
Grade 4
Hematologic AE
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
Neutropenia
Anemia
Thrombocytopenia
Leukopenia
74
89
64
74
82.2
98.9
71.1
82.2
19
11
10
28
21.1
12.2
11.1
31.1
35
4
4
3
38.9
4.4
4.4
3.3
73
84
49
65
79.3
90.3
52.7
69.9
29
0
1
15
31.5
0.0
1.1
16.1
22
0
0
0
23.9
0.0
0.0
0.0
Ixabepilone Plus Carboplatin (n ⫽ 95)
Any Grade
Nonhematologic AE
ⴱ
Neurologic
Peripheral sensory neuropathy†
Skin
Alopecia
Pruritus
GI
Nausea
Vomiting
Diarrhea
Upper abdominal pain
Constipation
General/administration site
Fatigue
Asthenia
Pain
Pyrexia
Metabolic
Decreased appetite
Musculoskeletal/connective tissue
Arthralgia
Myalgia
Grade 3
Paclitaxel Plus Carboplatin (n ⫽ 96)
Grade 4
Any Grade
Grade 3
Grade 4
No.
%
No.
%
No.
%
No.
%
No.
%
No.
%
35
36.8
0
0.0
0
0.0
54
56.3
7
7.3
0
0.0
45
3
47.4
3.2
NA
0
0.0
NA
0
0.0
55
6
57.3
6.3
NA
1
1.0
NA
0
0.0
39
16
13
5
5
41.1
16.8
13.7
5.3
5.3
1
1
2
0
0
1.1
1.1
2.1
0.0
0.0
0
0
0
0
0
0.0
0.0
0.0
0.0
0.0
29
8
9
1
6
30.2
8.3
9.4
1.0
6.3
1
1
1
0
0
1.0
1.0
1.0
0.0
0.0
0
0
0
0
0
0.0
0.0
0.0
0.0
0.0
29
11
6
6
30.5
11.6
6.3
6.3
1
1
0
0
1.1
1.1
0.0
0.0
0
0
0
0
0.0
0.0
0.0
0.0
24
14
4
2
25.0
14.6
4.2
2.1
2
2
0
0
2.1
2.1
0.0
0.0
0
0
0
0
0.0
0.0
0.0
0.0
29
30.5
1
1.1
0
0.0
24
25.0
1
1.0
0
0.0
15
15
15.8
15.8
0
0
0.0
0.0
0
0
0.0
0.0
22
27
22.9
28.1
2
2
2.1
2.1
0
0
0.0
0.0
Abbreviations: AE, adverse event; NA, not applicable.
ⴱ
Treatment-related AEs in ⱖ 5% of patients in either treatment arm.
†Includes the following Medical Dictionary for Regulatory Activities (version 9.1) terms: neuralgia, neuropathy peripheral, paraesthesia, peripheral sensory
neuropathy, neurotoxicity, and polyneuropathy.
Within 30 days of receiving the study drug, there were three
deaths in the ixabepilone plus carboplatin arm and five deaths in the
paclitaxel plus carboplatin arm. The three deaths in the ixabepilone
plus carboplatin arm resulted from cardiac events (one ventricular
fibrillation, one heart failure, and one sudden cardiac arrest). Given
the high incidence of coexisting cardiac disease in the lung cancer
population, these events were most likely not related to the drug.
In the ixabepilone plus carboplatin group, 30% and 38% of
patients had at least one dose reduction of ixabepilone and/or carboplatin, respectively. In the paclitaxel plus carboplatin arm, at least 23%
and 25% of patients had at least one dose reduction of paclitaxel
and/or carboplatin, respectively.
DISCUSSION
Cytotoxic chemotherapy has unequivocally been associated with benefit in NSCLC. In advanced disease, chemotherapy improves both
quality and quantity of life, albeit modestly. Despite these benefits, it is
clear that a substantial number of patients do not obtain benefit and
experience toxicity as a consequence of treatment. Although new
6
© 2013 by American Society of Clinical Oncology
so-called targeted agents have also demonstrated benefit, it is likely
that classical cytotoxic agents will remain as cornerstones of treatment
for the foreseeable future. Therefore, there is an increasing effort to
optimize the use of these agents by identifying groups most likely to
benefit from specific agents.
Taxanes (paclitaxel and docetaxel) are commonly used in chemotherapy regimens for all stages of disease. These agents function by
binding to tubulin and preventing disassembly among other mechanisms of action. High expression of specific tubulin subtypes, most
notably the beta-3 isoform, has been associated with resistance to
taxanes.8-10 In preclinical models, epothilones have demonstrated activity despite high ␤3T expression13 and are poor substrates for the
multidrug resistance transporter, another potential mechanism of
taxane resistance.
Our trial failed to confirm the hypothesis that a regimen based on
ixabepilone, the first epothilone to enter routine clinical use, would
demonstrate a higher level of activity in ␤3T-positive disease compared with taxane-based treatment. We did prospectively confirm
previous retrospective data indicating that ␤3T expression is an adverse prognostic and/or predictive marker for regimens containing
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
Ixabepilone Plus Carboplatin in Advanced NSCLC
antitubulin agents.8,19 Future studies will be needed to determine if a
nonantitubulin-based regimen (eg, without taxanes, vinca alkaloids,
or epothilones) can be selected based on ␤3T expression (ie, predictive
marker) or whether ␤3T expression is simply an adverse prognostic
marker for NSCLC.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS
OF INTEREST
Although all authors completed the disclosure declaration, the following
author(s) and/or an author’s immediate family member(s) indicated a
financial or other interest that is relevant to the subject matter under
consideration in this article. Certain relationships marked with a “U” are
those for which no compensation was received; those relationships marked
with a “C” were compensated. For a detailed description of the disclosure
categories, or for more information about ASCO’s conflict of interest policy,
please refer to the Author Disclosure Declaration and the Disclosures of
Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: Remigiusz Kaleta, Bristol-Myers
Squibb (C); Pralay Mukhopadhyay, Bristol-Myers Squibb (C); Ovidiu C.
Trifan, Bristol-Myers Squibb (C); Laura Whitaker, Bristol-Myers Squibb
(C) Consultant or Advisory Role: Martin J. Edelman, Bristol-Myers
Squibb (C); Elisabeth Quoix, Bristol-Myers Squibb (C); Martin Reck, Eli
REFERENCES
1. Schiller JH, Harrington D, Belani CP, et al:
Comparison of four chemotherapy regimens for
advanced non-small cell lung cancer. N Engl J Med
346:92-98, 2002
2. Kelly K, Crowley J, Bunn PA Jr, et al: Randomized phase three trial of paclitaxel plus carboplatin
versus vinorelbine plus cisplatin in the treatment of
patients with advanced non–small-cell lung cancer:
A Southwest Oncology Group trial. J Clin Oncol
19:3210-3218, 2001
3. Scagliotti GV, Parikh P, von Pawel J, et al:
Phase III study comparing cisplatin plus gemcitabine
with cisplatin plus pemetrexed in chemotherapynaive patients with advanced-stage non–small-cell
lung cancer. J Clin Oncol 26:3543-3551, 2008
4. Sandler A, Gray R, Perry MC, et al: Paclitaxelcarboplatin alone or with bevacizumab for non-smallcell lung cancer. N Engl J Med 355:2542-2550, 2006
5. Han JY, Lim HS, Shin ES, et al: Comprehensive analysis of UGT1A polymorphisms predictive
for pharmacokinetics and treatment outcome in
patients with non–small-cell lung cancer treated
with irinotecan and cisplatin. J Clin Oncol 24:22372244, 2006
6. Rosell R, Danenberg KD, Alberola V, et al:
Ribonucleotide reductase messenger RNA expression and survival in gemcitabine/cisplatin-treated
Lilly (C), F. Hoffman-La Roche (C), AstraZeneca (C), Daiichi Sankyo
(C), Bristol-Myers Squibb (C) Stock Ownership: Remigiusz Kaleta,
Bristol-Myers Squibb; Pralay Mukhopadhyay, Bristol-Myers Squibb;
Ovidiu C. Trifan, Bristol-Myers Squibb; Laura Whitaker, Bristol-Myers
Squibb Honoraria: Martin J. Edelman, Bristol-Myers Squibb;
Chun-Ming Tsai, Eli Lilly, AstraZeneca, Roche, Boehringer Ingelheim,
Pfizer; Martin Reck, Eli Lilly, F. Hoffman-La Roche, Daiichi Sankyo,
AstraZeneca Research Funding: Martin J. Edelman, Bristol-Myers
Squibb Expert Testimony: None Other Remuneration: None
AUTHOR CONTRIBUTIONS
Conception and design: Martin J. Edelman, Heung-Tae Kim, Pralay
Mukhopadhyay, Ovidiu C. Trifan, Laura Whitaker, Martin Reck
Provision of study materials or patients: Martin J. Edelman,
Claude-Peter Schneider, Chun-Ming Tsai, Heung-Tae Kim, Elisabeth
Quoix, Alexander V. Luft, Martin Reck
Collection and assembly of data: Martin J. Edelman, Claus-Peter
Schneider, Chun-Ming Tsai, Elisabeth Quoix, Remigiusz Kaleta, Ovidiu
C. Trifan, Laura Whitaker, Martin Reck
Data analysis and interpretation: Martin J. Edelman, Chun-Ming Tsai,
Alexander V. Luft, Pralay Mukhopadhyay, Ovidiu C. Trifan, Martin Reck
Manuscript writing: All authors
Final approval of manuscript: All authors
advanced non-small cell lung cancer patients. Clin
Cancer Res 10:1318-1325, 2004
7. Se`ve P, Dumontet C: Is class III beta-tubulin a
predictive factor in patients receiving tubulin-binding
agents? Lancet Oncol 9:168-175, 2008
8. Se`ve P, Mackey J, Isaac S, et al: Class III
beta-tubulin expression in tumor cells predicts response and outcome in patients with non-small cell
lung cancer receiving paclitaxel. Mol Cancer Ther
4:2001-2007, 2005
9. Mozzetti S, Ferlini C, Concolino P, et al: Class
III beta-tubulin overexpression is a prominent mechanism of paclitaxel resistance in ovarian cancer
patients. Clin Cancer Res 11:298-305, 2005
10. Paradiso A, Mangia A, Chiriatti A, et al: Biomarkers predictive for clinical efficacy of taxol-based
chemotherapy in advanced breast cancer. Ann Oncol 16:iv14-iv19, 2005 (suppl 4)
11. Bollag DM, McQueney PA, Zhu J, et al:
Epothilones, a new class of microtubule-stabilizing
agents with a taxol-like mechanism of action. Cancer Res 55:2325-2333, 1995
12. Lee FY, Borzilleri R, Fairchild CR, et al: BMS247550: A novel epothilone analog with a mode of
action similar to paclitaxel but possessing superior
antitumor efficacy. Clin Cancer Res 7:1429-1437,
2001
13. Lee FY, Smykla R, Johnston K, et al: Preclinical efficacy spectrum and pharmacokinetics of ixabepilone. Cancer Chemother Pharmacol 63:201212, 2009
14. Jordan MA, Miller H, Ni L, et al: The Pat-21
breast cancer model derived from a patient with
primary taxol resistance recapitulates the phenotype
of its origin, has altered beta-tubulin expression and
is sensitive to ixabepilone. J Clin Oncol 47, 2006
(suppl; abstr LB-280)
15. Dumontet C, Jordan MA, Lee FF: Ixabepilone:
Targeting betaIII-tubulin expression in taxaneresistant malignancies. Mol Cancer Ther 8:17-25,
2009
16. Vansteenkiste J, Lara PN Jr, Le Chevalier T, et
al: Phase II clinical trial of the epothilone B analog,
ixabepilone, in patients with non–small-cell lung
cancer whose tumors have failed first-line platinumbased chemotherapy. J Clin Oncol 25:3448-3455,
2007
17. Plummer R, Woll P, Fyfe D, et al: A phase I
and pharmacokinetic study of ixabepilone in combination with carboplatin in patients with advanced
solid malignancies. Clin Cancer Res 14:8288-8294,
2008
18. Se`ve P, Isaac S, Tre´dan O, et al: Expression of
class III ␤-tubulin is predictive of patient outcome in
patients with non-small cell lung cancer receiving
vinorelbine-based chemotherapy. Clin Cancer Res
11:5481-5486, 2005
19. Reiman T, Lai R, Veillard AS, et al: Crossvalidation study of class III beta-tubulin as a predictive marker for benefit from adjuvant chemotherapy
in resected non-small-cell lung cancer: Analysis of
four randomized trials. Ann Oncol 23:86-93, 2012
■ ■ ■
www.jco.org
© 2013 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
7
Edelman et al
Acknowledgment
We thank patients and investigators for their participation in the trial. We also thank Sudha Vemuri, PhD, Bristol-Myers Squibb, for her
writing and editorial support. The authors take full responsibility for the content of this publication and confirm that it reflects their viewpoint
and scientific expertise. We also wish to acknowledge StemScientific, funded by Bristol-Myers Squibb, for providing editorial support in the
preparation of this article before submission.
Appendix
Table A1. Summary of PFS and ORR Results
␤3T Positiveⴱ
PFS, months
Median
80% CI
95% CI
Events
No.
%
HR
80% CI
One-sided 90% CI
Log-rank P
One sided
Two sided
Response rate
No.
%
80% CI
95% CI
␤3T Negativeⴱ
PFS, months
Median
80% CI
95% CI
Events
No.
%
HR
80% CI
One-sided 90% CI
Log-rank P
One sided
Two sided
Response rate
No.
%
80% CI
95% CI
8
© 2013 by American Society of Clinical Oncology
Ixabepilone Plus Carboplatin
(n ⫽ 53)
Paclitaxel Plus Carboplatin
(n ⫽ 51)
4.27
3.22 to 4.90
2.89 to 5.65
4.27
4.01 to 5.42
3.61 to 6.05
42
79.2
40
78.4
1.04
0.78 to 1.41
— to 1.41
.853
.5735
9
17.0
10.5 to 25.5
8.1 to 29.8
15
29.4
21.0 to 39.1
17.5 to 43.8
Ixabepilone Plus Carboplatin
(n ⫽ 45)
Paclitaxel Plus Carboplatin
(n ⫽ 48)
5.78
5.32 to 6.83
5.29 to 8.41
5.32
4.40 to 5.78
4.27 to 5.98
27
60.0
35
72.9
0.78
0.55 to 1.10
— to 1.10
.3500
.1750
12
26.7
18.1 to 36.9
14.6 to 41.9
(continued on following page)
13
27.1
18.7 to 37.0
15.3 to 41.8
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
Ixabepilone Plus Carboplatin in Advanced NSCLC
Table A1. Summary of PFS and ORR Results (continued)
Overall Population
PFS, months
Median
80% CI
95% CI
Events
No.
%
HR
80% CI
One-sided 90% CI
Log-rank P
One sided
Two sided
Response rate
No.
%
80% CI
95% CI
Ixabepilone Plus Carboplatin
(n ⫽ 98)
Paclitaxel Plus Carboplatin
(n ⫽ 99)
5.29
4.27 to 5.65
4.14 to 5.88
5.13
4.27 to 5.59
4.21 to 5.78
69
70.4
75
75.8
0.92
0.73 to 1.15
— to 1.15
.632
.316
21
21.4
16.1 to 27.7
13.8 to 30.9
28
28.3
22.4 to 34.9
19.7 to 38.2
Abbreviations: ␤3T, beta-3 tubulin; HR, hazard ratio; ORR, overall response rate; PFS, progression-free survival.
ⴱ
␤3T status based on internal variable region data.
www.jco.org
© 2013 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.
9
Edelman et al
Table A2. Summary of OS Results
␤3T Positiveⴱ
OS, months
Median
80% CI
95% CI
Events
No.
%
HR
80% CI
One-sided 90% CI
Log-rank P
Two sided
One sided
␤3T Negativeⴱ
OS, months
Median
80% CI
95% CI
Events
No.
%
HR
80% CI
One-sided 90% CI
Log-rank P
Two sided
One sided
Overall Population
OS, months
Median
80% CI
95% CI
Events
No.
%
HR
80% CI
One-sided 90% CI
Log-rank P
Two sided
One sided
Ixabepilone Plus Carboplatin
(n ⫽ 53)
Paclitaxel Plus Carboplatin
(n ⫽ 51)
10.61
7.79 to 12.42
7.16 to 13.96
11.37
9.23 to 13.93
8.15 to 15.41
46
86.8
34
66.67
1.60
1.20 to 2.10
— to 2.10
.053
.9735
Ixabepilone Plus Carboplatin
(n ⫽ 45)
Paclitaxel Plus Carboplatin
(n ⫽ 48)
16.92
14.26 to 23.56
13.04 to 27.96
9.40
8.25 to 12.29
7.98 to 13.40
23
51.1
34
70.8
0.70
— to 0.90
.123
.0615
Ixabepilone Plus Carboplatin
(n ⫽ 98)
Paclitaxel Plus Carboplatin
(n ⫽ 99)
13.04
10.74 to 14.26
9.99 to 14.52
10.15
9.00 to 12.29
8.28 to 13.40
69
70.4
68
68.7
1.10
0.90 to 1.40
— to 1.40
.622
.689
Abbreviations: ␤3T, beta-3 tubulin; HR, hazard ratio; OS, overall survival.
ⴱ
␤3T status is based on IVRS data
10
© 2013 by American Society of Clinical Oncology
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from jco.ascopubs.org on May 30, 2014. For personal use only. No other uses without permission.
Copyright © 2013 American Society of Clinical Oncology. All rights reserved.

Download Report