Published Ahead of Print on October 24, 2011 as 10.1200/JCO.2010.33.9101 The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2010.33.

9101

JOURNAL OF CLINICAL ONCOLOGY

O R I G I N A L

R E P O R T

Capecitabine Versus Classical Cyclophosphamide, Methotrexate, and Fluorouracil As First-Line Chemotherapy for Advanced Breast Cancer
Martin R. Stockler, Vernon J. Harvey, Prudence A. Francis, Michael J. Byrne, Stephen P. Ackland, Bernie Fitzharris, Guy Van Hazel, Nicholas R.C. Wilcken, Peter S. Grimison, Anna K. Nowak, M. Corona Gainford, Akiko Fong, Lisa Paksec, Tatiana Sourjina, Diana Zannino, Val Gebski, R. John Simes, John F. Forbes, and Alan S. Coates
A B S T R A C T

Martin R. Stockler, Peter S. Grimison, M. Corona Gainford, Tatiana Sourjina, Diana Zannino, Val Gebski, R. John Simes, and Alan S. Coates, National Health and Medical Research Council Clinical Trials Centre; Nicholas R.C. Wilcken, Westmead and Nepean Hospitals, Sydney; Stephen P. Ackland, Calvary Mater Hospital; Akiko Fong, Lisa Paksec, and John F. Forbes, University of Newcastle, Newcastle, New South Wales; Prudence A. Francis, Peter MacCallum Cancer Centre, Melbourne, Victoria; Michael J. Byrne and Anna K. Nowak, Sir Charles Gairdner Hospital, Perth; Guy Van Hazel, The Mount Hospital, Perth, Western Australia, Australia; Vernon J. Harvey, Auckland Hospital, Auckland; and Bernie Fitzharris, Christchurch Hospital, Christchurch, New Zealand. Submitted December 2, 2010; accepted August 24, 2011; published online ahead of print at www.jco.org on October 24, 2011. Written on behalf of the Australian New Zealand Breast Cancer Trials Group. Supported in part by Roche International and Roche Australia, which provided drugs and research grants; the National Health and Medical Research Council of Australia; and the Cancer Institute New South Wales. Presented in part at the 29th Annual San Antonio Breast Cancer Symposium, December 14-17, 2006, San Antonio, TX, and 43rd Annual Meeting of the American Society of Clinical Oncology, June 1-5, 2007, Chicago, IL. Authors disclosures of potential conicts of interest and author contributions are found at the end of this article. Clinical Trials repository link available on JCO.org. Corresponding author: Martin R. Stockler, MBBS, MSc, FRACP, NHMRC Clinical Trials Centre, University of Sydney, Locked Bag 77, Camperdown NSW 1450, Australia; e-mail: martin.stockler@sydney.edu.au. 2011 by American Society of Clinical Oncology 0732-183X/11/2999-1/$20.00 DOI: 10.1200/JCO.2010.33.9101

Purpose We compared oral capecitabine, administered intermittently or continuously, versus classical cyclophosphamide, methotrexate, and uorouracil (CMF) as rst-line chemotherapy for women with advanced breast cancer unsuited to more intensive regimens. Patients and Methods Three hundred twenty-three eligible women were randomly assigned to capecitabine administered intermittently (1,000 mg/m2 twice daily for 14 of every 21 days; n 107) or continuously (650 mg/m2 twice daily for 21 of every 21 days; n 107), or to classical CMF (oral cyclophosphamide 100 mg/m2 days 1 to 14 with intravenous methotrexate 40 mg/m2 and uorouracil 600 mg/m2 on days 1 and 8 every 28 days; n 109). The primary end point was quality-adjusted progression-free survival (PFS); secondary end points included PFS, overall survival (OS), objective tumor response, and adverse events. Intermittent and continuous capecitabine were to be compared rst and, if similar (P .05), combined for denitive comparisons versus CMF. Results Quality-adjusted PFS (P .2), objective tumor response rate (20%; P .8), and PFS (median, 6 months; hazard ratio [HR], 0.86; 95% CI, 0.67 to 1.10; P .2) were similar in women assigned capecitabine versus CMF. OS was longer in women assigned capecitabine rather than CMF (median, 22 v 18 months; HR, 0.72; 95% CI, 0.55 to 0.94; P .02). Febrile neutropenia, infection, stomatitis, and serious adverse events were more common with CMF; hand-foot syndrome was more common with capecitabine. Conclusion Capecitabine improved OS by being similarly active, less toxic, and more tolerable than CMF. Capecitabine is a good rst-line chemotherapy option for women with advanced breast cancer who are unsuited to more intensive regimens. J Clin Oncol 29. 2011 by American Society of Clinical Oncology

INTRODUCTION

Most trials of rst-line chemotherapy for advanced breast cancer have tested whether intensive regimens improve outcomes in women t enough to tolerate them. However, intensive regimens may be unsuitable for many women with advanced breast cancer (eg, those who are older; who have signicant comorbidities; and perhaps whose disease is indolent, of low volume, and/or bone dominant). Intensive regimens may also be unsuitable for women who want to minimize treatment toxicity (eg, those who relapse after intensive regimens of adjuvant chemotherapy). Few trials have focused on this important and substantial subgroup unsuited for intensive rst-line chemotherapy.

Capecitabine is an oral prodrug that is converted to uorouracil by an enzymatic pathway ending with thymidine phosphorylase, which is present at higher levels in sensitive tumor cells than in normal cells. Capecitabine showed signicant activity with an objective tumor response rate of 20% in a single-arm trial of women with heavily pretreated breast cancer.1 Small randomized trials in advanced breast cancer have suggested that capecitabine has comparable activity to paclitaxel and intravenous CMF administered every 3 weeks.2,3 However, classical CMF is substantially more effective than this every-3-week intravenous regimen.4 A continuous regimen of capecitabine at 666 mg/m2 twice daily had similar activity and
2011 by American Society of Clinical Oncology

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

Copyright 2011 by American Society of Clinical Oncology

Stockler et al

less toxicity than an intermittent regimen of 1,250 mg/m2 twice daily for 14 of every 21 days in advanced colorectal cancer.5 The duration of chemotherapy seems important in advanced breast cancer. We previously showed that continuing chemotherapy until progression was superior to administering it for three cycles, then stopping and administering an additional three cycles at each sign of progression.6 A meta-analysis of randomized trials showed that administering chemotherapy for a longer duration resulted in better survival than administering the same chemotherapy for a shorter duration.7 Capecitabine provides a convenient way of administering chemotherapy over a protracted period. Oral chemotherapy was attractive to many patients surveyed with a range of advanced cancers.8 The aim of this trial was to determine whether treatment with capecitabine was preferable to that with classical CMF and whether capecitabine administered continuously (21 of every 21 days) was preferable to the same total dose administered intermittently (14 of every 21 days) in women with advanced breast cancer unsuited to more intensive rst-line chemotherapy. We hypothesized that quality-adjusted progression-free survival (PFS) would be better with capecitabine than with classical CMF and with continuous rather than intermittent capecitabine, through a combination of equivalent or better tumor control, equivalent or lesser toxicity, and greater patient acceptability.
PATIENTS AND METHODS
Patients The protocol described the target population as women with advanced breast cancer where chemotherapy for advanced disease is being considered for the rst time, and where more intensive chemotherapy is not considered more appropriate, without further specication. Key eligibility criteria included a pathologic diagnosis of breast cancer, suitability for chemotherapy with either CMF or capecitabine, performance status (PS; Eastern Cooperative Oncology Group [ECOG]) of 0 to 3, neutrophil count of 1.5 109/L or greater, platelet count of 75 109/L or greater, calculated creatinine clearance of 30 mL/min or greater, and serum bilirubin less than 50 mol/L. Measurable disease was not required. Key exclusion criteria were previous chemotherapy for advanced breast cancer, adjuvant chemotherapy within the last 6 months, an indication for chemotherapy more intensive than CMF or capecitabine, and GI disease precluding oral chemotherapy. All patients provided signed, written informed consent. The protocol was approved by the human research ethics review committees at all participating institutions. Random assignment was performed centrally by telephone and balanced dynamically for treating institution, ECOG PS (0 to 1 v 2 to 3), presence of liver and/or brain metastases, planned use of bisphosphonates, and planned use of prednisonewithCMF.9 Participantswererandomlyassignedtointermittentcapecitabine, continuous capecitabine, or the classical Bonadonna regimen of oral cyclophosphamide with intravenous methotrexate and uorouracil (ie, CMF). Treatments Intermittent capecitabine was started at 1,000 mg/m2 twice daily on days 1 through 14 and repeated every 3 weeks. The dose was increased to 1,250 mg/m2 twice daily in the absence of adverse events after two cycles. Continuous capecitabine was started at 650 mg/m2 twice daily on days 1 through 21 and repeated every 3 weeks with no dose escalation. CMF consisted of oral cyclophosphamide 100 mg/m2 on days 1 through 14, with methotrexate 40 mg/m2 and uorouracil 600 mg/m2 administered intravenously on days 1 and 8 and repeated every 4 weeks. Methotrexate doses were reduced for patients with a calculated creatinine clearance of 30 to 70 mL/min. Prednisolone 40 mg/m2/d on days 1 through 14 was permitted with CMF. Doses were based on surface area calculated with the lesser of ideal or actual body weight. Treatment was delayed until toxicities other than alopecia had resolved to grade 1 or lower. Dose reductions of approximately 25% were recommended for prespecied toxicities. Concurrent treatment with trastu2
2011 by American Society of Clinical Oncology

zumab was not allowed. Treatment was to continue until disease progression, patient intolerance, or unacceptable toxicity without any specied maximum duration. Management after disease progression was at the supervising clinicians discretion. Use of subsequent chemotherapy was recorded, but its dose, duration, and outcome were not. Assessments Patients were assessed and completed quality-of-life assessments before each cycle of chemotherapy: every 3 weeks for those receiving capecitabine and 4 weeks for CMF. Adverse events were rated with the National Cancer Institute Common Toxicity Criteria, version 2.0.10 Computerized axial tomographic scans of the chest and abdomen and skeletal radionucleide scans were performed at baseline and every 12 weeks until progression. Quality-of-life assessments were completed every 4 weeks from treatment cessation to progression. The primary outcome was quality-adjusted PFS. Secondary outcomes included PFS (from random assignment to progression or death), overall survival (OS; random assignment to death), and objective tumor response (complete plus partial response). The need for palliative radiation or change in chemotherapy was taken to indicate disease progression. Objective tumor response was assessed according to RECIST (Response Evaluation Criteria in Solid Tumors) 1.0 and required conrmatory imaging at least 1 month after initial documentation.11 Statistical Considerations The planned sample size of 465 patients (155 per arm) accrued over 3 years and observed for 1 additional year was designed to give 80% power to detect a 36% improvement in median PFS from 6 to 8.15 months (hazard ratio [HR], 0.74) comparing the combined capecitabine groups with the CMF group. Recruitment was stopped by the trial management committee after 325 patients had been accrued over 4 years because of diminishing support, without knowledge of any efcacy results. Planned comparisons for each end point were performed in two steps. The two capecitabine groups were compared rst, and if the P value was greater than .05, they were combined for a denitive comparison with the CMF group. If comparison of the two capecitabine groups yielded a P value of .05 or less, then the denitive analysis for that end point was to be of the three separate groups. KaplanMeier curves for PFS and OS were constructed and compared using the log-rank testfortheprimaryanalysesoftime-to-eventdata.Theinuenceofbaselinefactors on treatment effects was assessed by testing for interactions with treatment in Cox proportional hazards models for PFS and OS. All analyses were by intention to treat. All P values and CIs are two sided. Quality adjusted PFS for each group was the product of its mean utility score and the area under its time-to-progression curve truncated at 36 months.12 Utility scores from random assignment to progression were calculated and combined for each treatment group.13,14 CIs and P values for differences between groups were calculated with bootstrap methods. Detailed descriptions of the quality-of-life instruments, data, analyses, and results will be published separately.

RESULTS

We recruited 325 women from 34 centers between July 2001 and June 2005 (Fig 1). Two were judged ineligible and excluded from the primary analyses: one assigned continuous capecitabine had her diagnosis changed from metastatic breast cancer to nonsmall-cell lung cancer after random assignment but before start of treatment; the other assigned intermittent capecitabine was administered trastuzumab concurrently. All reported analyses are based on 323 eligible women with a median follow-up of 3.3 years. Inclusion of the ineligible women did not materially alter the results or conclusions. Baseline characteristics were well balanced among the three treatment groups (Table 1). The median age was 62 years; 23% were age 70 years or older, and 25% were age 55 years or younger. Almost half of the women had liver and/or brain metastases, and 13% had poor PS (ECOG 2 or 3). Approximately one third had received adjuvant chemotherapy at least 6 months previously, mostly with CMF.
JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

First-Line Capecitabine in Advanced Breast Cancer

Randomly allocated (N = 325)

Allocated to intermittent capecitabine (n = 108)

Allocated to continuous capecitabine (n = 108)

Allocated to classical CMF (n = 109)

Ineligible, concurrent trastuzumab (n = 1)

Ineligible, metastatic lung cancer (n = 1)

Ineligible, none (n = 0)

Fig 1. CONSORT diagram. CMF, cyclophosphamide, methotrexate, and uorouracil.

Still on treatment (n = 4) Stopped treatment for (n = 100) progression Stopped treatment (n = 2) before progression Lost to follow-up (n = 1)

Still on treatment Stopped treatment for progression Stopped treatment before progression Lost to follow-up

(n = 7) (n = 98) (n = 0) (n = 2)

Still on treatment (n = 1) Stopped treatment for (n = 102) progression Stopped treatment (n = 2) before progression Lost to follow-up (n = 4)

Analyzed (n = 107)

Analyzed (n = 107)

Analyzed (n = 109)

Approximately two thirds had hormone receptorpositive tumors, all of whom had undergone previous endocrine therapy. Objective tumor responses were equally frequent in the three treatment groups (P .5; Table 2) and in those assigned capecitabine versus

CMF (P .8). Conrmed complete or partial responses were documented in 63 participants (20%), and disease control at 6 months (conrmed response or stable disease at 6 months) was documented in 151 (47%).

Table 1. Baseline Patient Demographics and Clinical Characteristics Capecitabine Intermittent Characteristic Total eligible patients Age, years 50 50-59 60-69 70 Liver and/or brain metastases Interval from BC diagnosis to advanced BC diagnosis 4 years ECOG PS 0 1 2 3 Hormone receptor status ER or PR positive ER and PR negative or unknown Prior adjuvant chemotherapy None CMF AC plus CMF AC Taxane Prior endocrine therapy None Any No. 107 8 31 44 24 48 60 30 64 11 2 66 41 66 21 11 7 2 23 83 % 100 7 29 41 22 45 56 28 60 10 2 62 38 62 20 10 7 2 21 78 Continuous No. 107 12 38 35 22 51 56 39 55 10 3 72 35 67 21 10 5 4 19 88 % 100 11 36 33 21 48 52 36 51 9 3 67 33 62 20 9 5 4 18 82 No. 109 23 23 34 29 51 61 40 54 13 2 70 39 69 23 8 5 4 23 86 CMF % 100 21 21 31 27 47 56 37 50 12 2 64 36 63 21 7 5 4 21 79 No. 323 43 92 113 75 150 177 109 173 34 7 208 115 202 65 29 17 10 65 257 Total % 100 13 28 35 23 46 55 34 54 11 2 64 36 63 20 9 5 3 20 80

Abbreviations: AC, doxorubicin and cyclophosphamide; BC, breast cancer; CMF, cyclophosphamide, methotrexate, and uorouracil; ECOG, Eastern Cooperative Oncology Group; ER, estrogen receptor; PR, progesterone receptor; PS, performance status.

www.jco.org

2011 by American Society of Clinical Oncology

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

Stockler et al

Table 2. Tumor Response Capecitabine Intermittent Response Total patients CR PR SD, months At 3 At 6 Progressive disease or inevaluable Objective tumor response (CR PR) Objective disease control (CR PR SD at 6 months) No. 107 4 19 42 29 42 23 52 % 100 4 18 39 27 39 22 49 Continuous No. 107 0 21 49 32 37 21 53 % 100 0 20 46 30 35 20 50 No. 109 1 18 45 27 45 19 46 CMF % 100 1 17 41 25 41 18 42 No. 323 5 58 136 88 124 63 151 Total % 100 2 18 42 27 38 20 47

NOTE. No prespecied comparison was statistically signicant. Abbreviations: CMF, cyclophosphamide, methotrexate, and uorouracil; CR, complete response; PR, partial response; SD, stable disease.

Progressive disease was documented in 296 (92%) of the 323 women. PFS was similar in those assigned capecitabine intermittently versus continuously (median, 6 months; HR, 0.97; 95% CI, 0.73 to 1.28; P .8) and in those assigned capecitabine versus CMF (median, 6 months; HR, 0.86; 95% CI, 0.67 to 1.10; P .20; Fig 2A). Poor PS was the only baseline factor signicantly associated with PFS, and adjustment for it in a multivariable Cox model did not affect the size or statistical signicance of the effect of treatment on PFS (Table 3). Quality-adjusted PFS, the primary end point for the trial, was similar in those assigned capecitabine versus CMF (mean, 8.8 v 7.6 months; 95% CI for difference, 0.75 to 3.0; P .20). OS analyses included 237 deaths (73%) among the 323 women. OS was similar in those assigned capecitabine intermittently versus continuously (HR, 0.86; 95% CI, 0.62 to 1.12; P .4). OS was substantially longer in those assigned capecitabine versus CMF (median, 22 v 18 months; HR, 0.72; 95% CI, 0.55 to 0.94; P .02; Fig 2B). Longer survival was also associated with presence of hormone receptors, good PS, and absence of liver or brain metastases (Table 3). Adjustment for these baseline factors did not materially alter the size or statistical signicance of the survival benet associated with capecitabine. Weak interaction effects reected somewhat greater survival benets for capecitabine over CMF in the absence of previous adjuvant chemotherapy (P .07) or with a longer interval from diagnosis to treatment (P .10). Adverse events occurred equally frequently in those assigned capecitabine intermittently or continuously (Table 4). Hand-foot syndrome was more frequent with capecitabine, whereas uncomplicated neutropenia, febrile neutropenia, stomatitis, and miscellaneous other grade 3 or 4 adverse events were more frequent with CMF. Serious adverse events occurred in signicantly fewer patients assigned capecitabine than CMF (21% v 35%; P .02), particularly those related to neutropenia (Table 5). Dose reductions were equally frequent with capecitabine administered intermittently and continuously. The dose of capecitabine was reduced by one level (to 75% of starting dose) in 82 (38%) of 214 women and by two levels (to 50% of starting dose) in another 33 women (15%). Dose escalation of intermittent capecitabine (from 1,000 mg/m2 twice daily to 1,250 mg/m2 twice daily) was attempted in 11 (10%) of 107 women, six of whom experienced an adverse event requiring a dose reduction within two cycles. The average duration of chemotherapy was longer in those assigned capecitabine than in those assigned CMF (9 months, 12.1 cycles v 6 months, 5.5 cycles; Appendix Table A1, online only). Women as4
2011 by American Society of Clinical Oncology

signed capecitabine rather than CMF were more likely to continue chemotherapy beyond 6 months (40% v 21%; P .001) and beyond 12 months (18% v 6%; P .005). Chemotherapy was continued until progression in signicantly more women allocated capecitabine than CMF (77% v 51%; P .001; Appendix Table A2, online only).

A
Progression-Free Survival (proportion)

1.0
Capecitabine CMF Median 6 7

0.8

Hazard Ratio (95% CI) 0.86 (0.67 to 1.10) 1.00 Log-rank P = .2

0.6

0.4

0.2

12

18

24

30

36

42

48

Time Since Random Allocation (months)

No. at risk Capecitabine CMF 214 109 51 19 18 3 8 1 Hazard Ratio (95% CI) 0.72 (0.55 to 0.94) 1.00 Log-rank P = .02

B
Overall Survival (proportion)

1.0
Capecitabine CMF Median 22 18

0.8

0.6

0.4

0.2

12

18

24

30

36

42

48

Time Since Random Allocation (months)

No. at risk Capecitabine CMF 214 109 149 72 77 29 31 8 13 4

Fig 2. Progression-free and overall survival. CMF, cyclophosphamide, methotrexate, and uorouracil.
JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

First-Line Capecitabine in Advanced Breast Cancer

Table 3. Effects of Treatment and Baseline Characteristics on PFS and OS PFS Univariable Characteristic Assigned treatment (v CMF) Capecitabine (combined) Intermittent capecitabine Continuous capecitabine Age, years (v 50-69) 50 70 Hormone receptor positive (ER or PR positive v ER and PR negative) Interval from rst BC diagnosis to advanced BC diagnosis 4 years ECOG performance status (v 1) 0 2-3 Previous adjuvant chemotherapy (v none) CMF CMF AC other Liver and/or brain metastases Hemoglobin 120 g/L HR 0.86 0.84 0.88 1.12 0.94 0.89 0.87 0.85 1.52 1.06 0.99 1.11 1.08 95% CI 0.67 to 1.10 0.63 to 1.11 0.66 to 1.16 0.81 to 1.56 0.71 to 1.25 0.70 to 1.13 0.69 to 1.08 0.66 to 1.09 1.06 to 2.17 0.80 to 1.41 0.69 to 1.41 0.89 to 1.40 0.86 to 1.36 P .2 .2 .4 .5 .7 .3 .2 .2 .02 .7 .9 .4 .5 0.84 1.5 0.65 to 1.08 1.05 to 2.15 .17 .03 Multivariable (capecitabine v CMF) HR 0.85 95% CI 0.66 to 1.09 P .19 HR 0.72 0.67 0.78 0.93 1.19 0.74 0.9 0.69 2.2 0.93 1.03 1.34 0.93 Univariable 95% CI 0.55 to 0.94 0.49 to 0.92 0.57 to 1.06 0.65 to 1.35 0.88 to 1.61 0.57 to 0.97 0.69 to 1.16 0.51 to 0.92 1.52 to 3.18 0.68 to 1.29 0.70 to 1.51 1.04 to 1.73 0.72 to 1.20 P .02 .01 .11 .7 .3 .03 .4 .01 .001 .7 .9 .02 .6 0.66 2.14 0.49 to 0.89 1.48 to 3.11 .006 .001 0.69 0.53 to 0.90 .006 OS Multivariable (capecitabine v CMF) HR 0.71 95% CI 0.55 to 0.94 P .01

1.4

1.09 to 1.82

.01

Abbreviations: AC, doxorubicin and cyclophosphamide; BC, breast cancer; CMF, cyclophosphamide, methotrexate, and uorouracil; ECOG, Eastern Cooperative Oncology Group; ER, estrogen receptor; HR, hazard ratio; OS, overall survival; PFS, progression-free survival; PR, progesterone receptor.

A posthoc exploratory analysis with a time-dependent Cox model was performed to assess the effects of treatment on PFS during the rst 6 months separately from the effects beyond 6 months. During the rst 6 months, PFS was similar in those assigned capecitabine or CMF (HR, 1.15; 95% CI, 0.82 to 1.63; P .4); beyond 6 months, PFS was signicantly longer in those assigned capecitabine (HR, 0.62; 95% CI, 0.44 to 0.88; P .007).

Most women (91%) received second or subsequent lines of chemotherapy, including capecitabine in approximately one third of women originally assigned CMF and CMF in approximately one eighth of women originally assigned capecitabine (Appendix Table A3, online only). Other cytotoxic drugs and trastuzumab were used similarly often after progression in the three treatment groups. The effects and duration of second and subsequent lines of chemotherapy

Table 4. Adverse Events Capecitabine (No. of patients) Intermittent Adverse Event Fatigue Nausea Vomiting Diarrhea Stomatitis Hand-foot syndrome Tearing or conjunctivitis Alopecia Infection Fever without neutropenia Febrile neutropenia Neutropenia Leukopenia Anaemia Thrombocytopenia Other Grade 1-4 72 71 41 57 36 66 33 14 46 9 1 23 31 34 6 93 Grade 3-4 2 2 2 6 0 15 1 0 6 0 0 1 0 2 3 29 Continuous Any Grade 70 67 48 64 39 70 28 9 45 7 0 32 38 29 12 100 Grade 3-4 4 4 4 8 0 17 0 0 3 0 0 1 0 3 3 32 CMF (No. of patients) Any Grade 62 70 32 43 43 4 34 54 44 14 11 56 74 47 32 91 Grade 3-4 7 1 3 6 6 0 0 0 9 1 11 28 5 4 3 42 P (capecitabine v CMF) Grade 3-4 .1 .4 1.0 .8 .001 .001 1.0 1.0 .2 .3 .001 .001 .004 .5 1.0 .08

Abbreviation: CMF, cyclophosphamide, methotrexate, and uorouracil.

www.jco.org

2011 by American Society of Clinical Oncology

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

Stockler et al

Table 5. Serious Adverse Events Capecitabine Event Type Infection, fever, neutropenia Infection without neutropenia Infection with neutropenia Febrile neutropenia Fever without neutropenia Thrombosis/embolism GI event Nausea and vomiting Diarrhea Mucositis GI bleeding Dehydration Pain Cardiac event Dyspnea Other Total Intermittent Continuous CMF Total (n 107) (n 107) (n 109) (N 323)

11 0 1 0 6 2 4 0 0 0 5 2 0 16 47

6 1 0 2 4 7 6 0 0 1 1 4 2 4 38

12 4 12 1 6 6 3 4 3 1 0 0 0 15 67

29 5 13 3 16 15 13 4 3 2 6 6 2 35 152

Abbreviation: CMF, cyclophosphamide, methotrexate, and uorouracil.

were not collected. The estimated benet of capecitabine on OS was unaffected by subsequent use of trastuzumab (interaction P 0.3).
DISCUSSION

OS was signicantly longer with capecitabine than CMF, despite similar rates of objective tumor response, disease control, PFS over the rst 6 months, and quality-adjusted PFS. Capecitabine was better tolerated than CMF and less likely to be stopped for reasons other than progression, resulting in longer treatment continuation, better PFS beyond 6 months, and better OS beyond 12 months. The effects of capecitabine, both benecial and adverse, were similar whether the same total dose was administered intermittently or continuously. We chose quality-adjusted PFS as the primary end point and PFS as the primary measure of efcacy because of concerns that subsequent chemotherapy would obscure any effects on OS. However, OS is the most direct, reliable, and compelling measure of patient benet. OS and PFS, with both capecitabine and CMF, were similar or better in this trial than in comparable contemporary trials.15 This is probably because of both the nature of our study population and the efcacy of our treatments (assigned and subsequent). Wespeculatethatlongertreatmentdurationwithcapecitabinecompared with CMF caused the observed improvements in OS and PFS beyond 6 months without an effect on objective tumor response. The larger improvement in OS compared with PFS raises the possibility that subsequentchemotherapywasmoreeffectiveaftercapecitabinethanafter CMF. This might have occurred if there was less tumor resistance induced by exposure to one drug rather than three or if subsequent chemotherapy was more deliverable after capecitabine because it caused less cumulative bone marrow dysfunction than CMF. The possible importance of chemotherapy duration is highlighted by apparently contradictory results from a trial of capecitabine versus
6
2011 by American Society of Clinical Oncology

CMF or doxorubicin and cyclophosphamide (AC) as adjuvant chemotherapy for women age 65 years or older with early breast cancer (CALGB [Cancer and Leukemia Group B] 49907).16 In that trial, relapse-free survival and OS were signicantly shorter in women assigned capecitabine. The CALGB trial had a larger sample size than our trial (600 v 325 participants) but fewer relapses (95 v 296) and deaths (62 v 237). The different age ranges and stages of disease do not explain the divergent results. Similar doses of capecitabine and CMF and identical average durations of CMF (6 months) were used in both trials. Use of AC in 56% of the control group of the adjuvant trial is unlikely to have affected the results, because four cycles of AC administered every 3 weeks has produced results similar to those with six cycles of classical CMF in the adjuvant setting.17,18 The average duration of treatment with capecitabine was the most notable difference: 18 weeks in the adjuvant trial versus 36 weeks in our trial. The results of our trial should not be extrapolated to the adjuvant setting. We started intermittent capecitabine at a dose of 1,000 mg/m2 twice daily, 20% lower than the recommended starting dose, and 50% of women had dose reductions for toxicity, yet objective tumor response rates, PFS, and OS were similar or better than those with CMF. This suggests that even substantial dose reductions do not adversely affect the anticancer activity of capecitabine and may contribute to the improved tolerability we observed. Other studies using doses of capecitabine lower than those recommended have reached similar conclusions.19 These data raise the hypothesis that starting capecitabine at a dose lower than 1,250 mg/m2 twice per day may be preferable; a randomized trial comparing starting doses is needed for a denitive answer to this question. The main strengths of our trial were its inclusive eligibility criteria, consistency of results across subgroups, and extended follow-up. The relative effects of capecitabine versus CMF were similar in all relevant clinical subgroups. Extended follow-up resulted in events being observed in the majority of participants, giving the trial sufcient power to detect moderate treatment effects, despite stopping early. Posthoc power calculations show the trial as conducted provided 66% power to detect the PFS effect specied by the protocol. The main limitations of our trial are the ambiguity of broad, pragmatic eligibility criteria and the use of an unfashionable control regimen. We chose to include the large and heterogeneous subset of women with advanced breast cancer for whom more intense rst-line chemotherapy was considered unsuitable. Although it is difcult to dene this population precisely, we believe it is an easily identied and substantial proportion of women encountered in routine clinical practice but poorly represented in clinical trials. Our choice in 2000 of classical CMF as the comparator is open to criticism, because it is no longer regarded as a standard of care and may be considered a suboptimal control. However, other regimens that might be considered standard in this setting (eg, AC, single-agent anthracycline, or single-agent taxane) have comparable efcacy.6,20-22 Furthermore, many women now present with metastatic disease after adjuvant therapy that included an anthracycline and/or taxane. The proportion of patients treated with trastuzumab after progressionwassimilarinthecapecitabineandCMFgroupsanddidnotaffectthe estimated treatment effect. More intensive regimens may be preferable for women with extensive, rapidly progressing, or life-threatening visceral disease.23-28 However, these regimens are also more toxic and may be less suitable for women with advanced breast cancer who have signicant comorbidities, have had bad experiences with adjuvant chemotherapy, or have breast cancer that is indolent or of low volume and are considering a less intense rst-line regimen.
JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

First-Line Capecitabine in Advanced Breast Cancer

In conclusion, capecitabine improved OS by being similarly active, less toxic, and more tolerable than classical CMF. Capecitabine is a good option for women unsuited to more intensive regimens as rst-line chemotherapy for advanced breast cancer.

AUTHORS DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Although all authors completed the disclosure declaration, the following author(s) indicated a nancial 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 ASCOs conict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: Vernon J. Harvey, F. Hoffman-La Roche (U); Stephen P. Ackland, Roche (U); Guy Van Hazel, Roche (C) Stock Ownership: None

Honoraria: Martin R. Stockler, Roche; Vernon J. Harvey, F. Hoffman-La Roche, Chugai Pharmaceuticals; Prudence A. Francis, Roche; Stephen P. Ackland, Roche; Nicholas R.C. Wilcken, Roche; Anna K. Nowak, Roche; Val Gebski, Roche; John F. Forbes, AstraZeneca Research Funding: Martin R. Stockler, Roche; Vernon J. Harvey, F. Hoffman-La Roche; Anna K. Nowak, Roche; Diana Zannino, Roche; R. John Simes, Roche Expert Testimony: None Other Remuneration: Prudence A. Francis, Roche; Alan S. Coates, Roche

AUTHOR CONTRIBUTIONS
Conception and design: Martin R. Stockler, Michael J. Byrne, Stephen P. Ackland, Bernie Fitzharris, Guy Van Hazel, Nicholas R.C. Wilcken, M. Corona Gainford, Val Gebski, R. John Simes, John F. Forbes, Alan S. Coates Provision of study materials or patients: Prudence A. Francis, Michael J. Byrne, Guy Van Hazel, Nicholas R.C. Wilcken, Anna K. Nowak Collection and assembly of data: All authors Data analysis and interpretation: All authors Manuscript writing: All authors Final approval of manuscript: All authors
20. Bishop JF, Dewar J, Toner C, et al: Initial paclitaxel improves outcome compared with CMFP combination chemotherapy as front-line therapy in untreated metastatic breast cancer. J Clin Oncol 17:2355-2364, 1999 21. Harvey V, Mouridsen H, Semiglazov V, et al: Phase III trial comparing three doses of docetaxel for second-line treatment of advanced breast cancer. J Clin Oncol 24:4963-4970, 2006 22. Piccart-Gebhart MJ, Burzykowski T, Buyse M, et al: Taxanes alone or in combination with anthracyclines as rst-line therapy of patients with metastatic breast cancer. J Clin Oncol; 26:1980-1986, 2008 23. Chan S, Friedrichs K, Noel D, et al: Prospective randomized trial of docetaxel versus doxorubicin in patients with metastatic breast cancer. J Clin Oncol 17:2341-2354, 1999 24. Paridaens R, Biganzoli L, Bruning P, et al: Paclitaxel versus doxorubicin as rst-line single agent chemotherapy for metastatic breast cancer: A European Organisation for Research and Treatment of Cancer randomized study with cross-over. J Clin Oncol 18:724-733, 2000 25. OShaughnessy J, Miles D, Vukelja S, et al: Superior survival with capecitabine plus docetaxel combination therapy in anthracycline-pretreated patients with advanced breast cancer: Phase III trial results. J Clin Oncol 20:2812-2823, 2002 26. Albain S, Nag SM, Calderillo-Ruiz G, et al: Gemcitabine plus paclitaxel versus paclitaxel monotherapy in patients with metastatic breast cancer and prior anthracycline treatment. J Clin Oncol 26: 3950-3957, 2008 27. Jones SE, Erban J, Overmoyer B, et al: Randomized phase III study of docetaxel compared with paclitaxel in metastatic breast cancer [see comment]. J Clin Oncol 23:5542-5551, 2005 28. Seidman AD, Berry D, Cirrincione C, et al: Randomized phase III trial of weekly compared with every-3-weeks paclitaxel for metastatic breast cancer, with trastuzumab for all HER-2 overexpressors and random assignment to trastuzumab or not in HER-2 nonoverexpressors: Final results of Cancer and Leukemia Group B protocol 9840. J Clin Oncol 26:1642-1649, 2008

REFERENCES
1. Blum JL, Jones SE, Buzdar AU, et al: Multicenter phase II study of capecitabine in paclitaxelrefractory metastatic breast cancer. J Clin Oncol 17:485-493, 1999 2. Talbot DC, Moiseyenko V, Van Belle S, et al: Randomized phase II comparing oral capecitabine (Xeloda) with paclitaxel in patients with metastatic/ advanced breast cancer pretreated with anthracyclines. Br J Cancer 86:1367-1372, 2002 3. OShaughnessy JA, Blum J, Moiseyenko V, et al: Randomized, open-label, phase II trial of oral capecitabine (Xeloda) vs. a reference arm of intravenous CMF (cyclophosphamide, methotrexate and 5-uorouracil) as rst-line therapy for advanced/metastatic breast cancer. Ann Oncol 12:1247-1254, 2001 4. Engelsman E, Klijn JC, Rubens RD, et al: Classical CMF versus a 3-weekly intravenous CMF schedule in postmenopausal patients with advanced breast cancer: An EORTC Breast Cancer Co-operative Group Phase III Trial (10808). Eur J Cancer 27:966-970, 1991 5. Findlay M, Van Cutsem E, Kocha W, et al: A randomized phase II study of Xeloda (capecitabine) in patients with advanced colorectal cancer. Proc Am Soc Clin Oncol 16, 1997 (suppl; abstr A798) 6. Coates A, Gebski V, Bishop JF, et al: Improving the quality of life during chemotherapy for advanced breast cancer: A comparison of intermittent and continuous treatment strategies. N Engl J Med 317:1490-1495, 1987 7. Stockler M, Wilcken N, Coates A: Chemotherapy for metastatic breast cancer: When is enough enough? Eur J Cancer 33:3147-3148, 1997 8. Liu G, Franssen E, Fitch MI, et al: Patient preferences for oral versus intravenous palliative chemotherapy. J Clin Oncol 15:110-115, 1997 9. Signorini DF, Leung O, Simes RJ, et al: Dynamic balancing for clinical trials. Stat Med 12:23432350, 1993 10. National Cancer Institute: Cancer Therapy Evaluation Program: Common Toxicity Criteria, Version 2.0,

1998. http://ctep.cancer.gov/protocolDevelopment/ electronic_applications/ctc.htm 11. Therasse P, Arbuck SG, Eisenhauer EA, et al: New guidelines to evaluate the response to treatment in solid tumours. J Natl Cancer Inst 92:205216, 2000 12. Glasziou PP, Simes RJ, Gelber RD: Quality adjusted survival analysis. Stat Med 9:1259-1276, 1990 13. Grimison PS, Simes RJ, Hudson HM, et al: Deriving a patient-based utility index from a cancerspecic quality of life questionnaire. Value Health 12:800-807, 2009 14. Grimison PS, Simes RJ, Hudson HM, et al: Preliminary validation of an optimally-weighted patient-based utility index by application to randomised trials in breast cancer. Value Health 12:967976, 2009 15. Wilcken N, Dear R: Chemotherapy in metastatic breast cancer: A summary of all randomised trials reported 2000-2007. Eur J Cancer 44:22182225, 2008 16. Muss HB, Berry DA, Cirrincione CT, et al: Adjuvant chemotherapy in older women with earlystage breast cancer. N Engl J Med 360:2055-2065, 2009 17. Fisher B, Brown AM, Dimitrov NV, et al: Two months of doxorubicin-cyclophosphamide with and without interval reinduction therapy compared with 6 months of cyclophosphamide, methotrexate, and uorouracil in positive-node breast cancer patients with tamoxifen-nonresponsive tumors: Results from the National Surgical Adjuvant Breast and Bowel Project B-15. J Clin Oncol 8:1483-1496, 1990 18. Fisher B, Anderson S, Tan-Chiu E, et al: Tamoxifen and chemotherapy for axillary node-negative, estrogen receptor-negative breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-23. J Clin Oncol 19:931-942, 2001 19. Zielinski C, Gralow J, Martin M: Optimising the dose of capecitabine in metastatic breast cancer: Confused, claried or conrmed? Ann Oncol 21:2145-2152, 2010

www.jco.org

2011 by American Society of Clinical Oncology

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1

Stockler et al

Acknowledgment Our heartfelt thanks go to the women, investigators, site staff, and coordinating center staff who participated in this study and to the Australian New Zealand Breast Cancer Trials Group Consumer Advisory Panel members who supported it: Auckland Hospital: V. Harvey; Border Medical Oncology: C. Underhill; Box Hill Hospital and Maroondah Hospital: J. Chirgwin; Calvary Mater Newcastle: S. Ackland; Christchurch Hospital: B. Fitzharris; Concord Repatriation General Hospital: A. Sullivan; Dubbo Hospital: P. Beale; Dunedin Hospital: D. Perez; Flinders Medical Centre: B. Koczwara; Frankston Hospital: V. Ganju; Geelong Hospital: R. McLennan; Lismore Base Hospital: A. Boyce; Monash Medical Centre: M. White; Mount Hospital: G. Van Hazel; Nepean Hospital: N. Wilcken; Peter MacCallum Cancer Centre: P. Francis; Prince of Wales Hospital: C. Lewis; Princess Alexandra Hospital: E. Walpole; Queen Elizabeth Hospital: K. Patterson; Royal Hobart Hospital: R. Lowenthal; Royal Melbourne Hospital and Western Hospital: M. Green; Royal Prince Alfred Hospital: M. Stockler; Sir Charles Gairdner Hospital: M. Byrne; St John of God Hospital, Bunbury: M. Buck; St John of God Hospital, Subiaco: D. Ransom; St Vincents Hospital, Melbourne: R. Snyder; St Vincents Hospital, Sydney: D. Dalley; Tamworth Hospital: D. Goldstein; Toowoomba Hospital and Tweed Heads Hospital: E. Abdi; Wellington Hospital: A. Simpson; and Westmead Hospital: N. Wilcken. Appendix

Table A1. Duration of Chemotherapy Capecitabine Intermittent Protocol Treatment (months) 0-6 7-11 12 No. 65 22 20 % 61 21 19 No. 63 26 18 Continuous % 59 24 17 No. 86 18 5 Standard CMF % 79 17 5

Abbreviation: CMF, cyclophosphamide, methotrexate, and uorouracil.

Table A2. Reasons for Ceasing Protocol Treatment Capecitabine Reason Progression SAE Unacceptable toxicity Patient intolerance Other Missing (had not stopped) Intermittent (n 107) 82 4 3 4 11 4 Continuous (n 107) 82 5 5 2 7 7 CMF (n 109) 55 8 8 16 27 2 Total 219 17 16 22 45 13

Abbreviations: CMF, cyclophosphamide, methotrexate, and uorouracil; SAE, serious adverse event.

Table A3. Treatment After Progression Capecitabine Treatment Chemotherapy Capecitabine (off protocol) CMF Doxorubicin/epirubicin Paclitaxel/docetaxel Vinorelbine Trastuzumab Other None Radiotherapy Endocrine therapy Intermittent (n 107) 6 11 35 34 17 5 27 10 42 31 Continuous (n 107) 7 15 39 36 21 11 26 7 39 18 CMF (n 109) 38 2 38 35 14 7 31 9 38 24

Abbreviation: CMF, cyclophosphamide, methotrexate, and uorouracil.

2011 by American Society of Clinical Oncology

JOURNAL OF CLINICAL ONCOLOGY

Information downloaded from jco.ascopubs.org and provided by at HOFFMANN-LA ROCHE AG on October 28, 2011 from Copyright 2011 American Society of Clinical Oncology. All rights reserved. 196.3.49.1