Miraclemediatv.com

⅐ N U M B E R 5 ⅐ F E B R U A R Y 1 0 2 0 0 7 From the International Breast Cancer Study Five Years of Letrozole Compared With Tamoxifen As Initial Adjuvant Therapy for Postmenopausal Women With of Eastern Switzerland, Kantonsspital, St Endocrine-Responsive Early Breast Cancer: Update of Study Alan S. Coates, Aparna Keshaviah, Beat Thu¨rlimann, Henning Mouridsen, Louis Mauriac, John F. Forbes, Robert Paridaens, Monica Castiglione-Gertsch, Richard D. Gelber, Marco Colleoni, Istva´n La´ng, Lucia Del Mastro, Ian Smith, Jacquie Chirgwin, Jean-Marie Nogaret, Tadeusz Pienkowski, Andrew Wardley, Erik H. Jakobsen, Karen N. Price, and Aron Goldhirsch Harvard School of Public Health; Frontier Previous analyses of the Breast International Group (BIG) 1-98 four-arm study compared initial therapy with letrozole or tamoxifen including patients randomly assigned to sequential treatment Cancer Group, Institut Bergonié, Bordeaux, whose information was censored at the time of therapy change. Because this presentation may unduly reflect early events, the present analysis is limited to patients randomly assigned to the University Hospital Gasthuisberg, Catholic University of Leuven; Jules Bordet Insti- continuous therapy arms and includes protocol-defined updated results.
tute; Brussels, Belgium; European Institute Patients and Methods
Four thousand nine hundred twenty-two of the 8,028 postmenopausal women with receptor- Research Institute, Genoa, Italy; National Institute of Oncology, Budapest, Hungary; positive early breast cancer randomly assigned (double-blind) to the BIG 1-98 trial were assigned to 5 years of continuous adjuvant therapy with either letrozole or tamoxifen; the remainder of Christie Hospital National Health Service women were assigned to receive the agents in sequence. Disease-free survival (DFS) was the Trust, South Manchester University Hospi- At a median follow-up time of 51 months, we observed 352 DFS events among 2,463 women receiving letrozole and 418 events among 2,459 women receiving tamoxifen. This reflected an 18% reduction in the risk of an event (hazard ratio, 0.82; 95% CI, 0.71 to 0.95; P ϭ .007). No predefined subsets showed differential benefit. Adverse events were similar to previous reports.
Patients on tamoxifen experienced more thromboembolic events, endometrial pathology, hot flashes, night sweats, and vaginal bleeding. Patients on letrozole experienced more bone fractures, arthralgia, low-grade hypercholesterolemia, and cardiovascular events other than Conclusion
The present updated analysis, which was limited to patients on monotherapy arms in BIG 1-98, Cancer Research, National Cancer Institute yields results similar to those from the previous primary analysis but more directly comparable with results from other trials of continuous therapy using a single endocrine agent.
J Clin Oncol 25:486-492. 2007 by American Society of Clinical Oncology years of the other. Previously reported results, INTRODUCTION
called the primary core analysis, compared initial Authors’ disclosures of potential con-flicts of interest and author contribu- Several large studies attest to the value of third- tamoxifen and letrozole and included all available generation aromatase inhibitors in the adjuvant data from patients randomly assigned to the systemic therapy of postmenopausal women with monotherapy arms and data from the sequential Address reprint requests to International endocrine-responsive early breast cancer,1-6 and therapy arms censored at the time of the therapy Center, Effingerstrasse 40, CH-3008 Bern, such therapy has been recommended as part of switch.3 Such an analysis of all four arms is valid Switzerland; e-mail: alan.coates@ibcsg.org.
the standard care in this patient group.7-9 The but difficult to compare with other studies be- Breast International Group (BIG) 1-98 study is a cause it gives considerable weight to early events.
four-arm trial comparing 5 years of monotherapy Because prospective meta-analyses of aromatase with tamoxifen or with letrozole with sequences inhibitor trials are planned (J. Cuzick, J. Bliss, R.D.
of 2 years of one of these agents followed by 3 Gelber, personal communication, September 2006), Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved. Letrozole v Tamoxifen for Early Breast Cancer
Fig 1. Consort diagram of Breast Interna-
tional Group 1-98 trial. The shaded boxes, are not included in this analysis. L, letro-zole; T, tamoxifen.
we present here an analysis of data derived only from patients The scheduled update of the BIG 1-98 trial was defined in the proto- randomly assigned to continuous tamoxifen or letrozole on the col and reviewed by the BIG 1-98 Data and Safety Monitoring Committee.
monotherapy arms of study BIG 1-98. The analysis is based on a This analysis is restricted to patients who were randomly assigned to the PATIENTS AND METHODS
The design and conduct of the study have been described elsewhere.3 Briefly, BIG 1-98 is a randomized, phase III, double-blind trial comparing the follow- ing four options: monotherapy with letrozole or with tamoxifen for 5 years, orsequential administration of tamoxifen for 2 years followed by letrozole for 3 years, or sequential administration of letrozole for 2 years followed by tamox-ifen for 3 years. The trial was conducted in postmenopausal women with estrogen receptor– and/or progesterone receptor–positive operable invasive breast cancer. From March 1998 to March 2000, patients were randomlyassigned to one of the following two arms: monotherapy with letrozole (2.5 mg daily) or tamoxifen (20 mg daily). From April 1999 to May 2003, patients were Years After Random Assignment
randomly assigned to all four arms (Fig 1). Hormone receptor status was based Table 1. Efficacy End Points
Years After Random Assignment
Fig 2. Kaplan-Meier estimates of (A) disease-free survival (DFS) and (B) overall survival
(OS) comparing letrozole with tamoxifen. The median follow-up time is 51 months, and1,785 patients have been observed for at least 5 years.
Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved. Coates et al
A End Point Analyses (N = 4,922)
Hazard Ratio
P
B Subgroup Analyses (DFS)
Hazard Ratio
P
Fig 3. Cox model results of (A) primary,
secondary, and exploratory end points and(B) subgroups with primary end point (disease-free survival [DFS]). The size of boxes is inversely proportional to the SEof the hazard ratio. The dashed vertical line is placed at 0.82, which is the hazardratio estimate for the overall analysis of the primary study end point. L, letrozole;T, tamoxifen; ER, estrogen receptor; Radiotherapy not administered (n = 1,546) Chemotherapy not administered (n = 3,690) monotherapy arms. The primary end point was disease-free survival Statistical Analysis
(DFS), which was defined as the time from random assignment to the Log-rank tests stratified by two-arm or four-arm random assignment earliest time of invasive recurrence in local, regional, or distant sites; a new option and chemotherapy use (based on randomized chemotherapy stratum) invasive breast cancer in the contralateral breast; any second (nonbreast) were used to compare the two groups,10 and Kaplan-Meier estimates were malignancy; or death from any cause. Protocol-specified secondary end calculated.11 Cox proportional hazards regression (adjusting for random points included overall survival, which was defined as the time from assignment option and chemotherapy use) was used to adjust for various random assignment to death from any cause, and systemic DFS, which wasdefined as the time from random assignment to systemic recurrence (ig- prognostic factors.12 Cumulative incidence calculations were performed to noring local and contralateral breast events, second (nonbreast) malig- control for competing risks.13 Fisher’s exact tests were used to compare the nancy, or death). The following three exploratory end points were analyzed percentage of patients with adverse events.14 to facilitate comparison with other published studies: DFS as definedearlier but ignoring second (nonbreast) malignancies; time to recurrence Role of Coordinating Group, Trial Steering Committee, and
ignoring second (nonbreast) malignancies and censoring deaths without Funding Source
recurrence; and time to distant recurrence additionally ignoring local, The International Breast Cancer Study Group was responsible for study regional, and contralateral breast recurrence. Death without prior cancer design and coordination, data collection and management, medical review, event is a type of DFS event defined as any death that occurs withoutevidence of breast cancer recurrence or second primary cancer at any time data analysis, and reporting (including the decision to publish). The ethics during or after completion of trial treatment. An adverse event by defini- committees and required health authorities of each participating institution tion occurs or begins during trial treatment or within 28 days of trial approved the study protocol, and all patients gave written informed consent.
treatment completion, regardless of prior recurrence.
The independent International Breast Cancer Study Group Data and Safety Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved. Letrozole v Tamoxifen for Early Breast Cancer
Monitoring Committee reviewed safety semiannually throughout the courseof the trial in addition to the two predefined interim and final efficacy analyses at 261, 433, and 779 DFS events, respectively.
Novartis, the manufacturer of letrozole, provided drug distribution and financial support and imposed no restrictions on the investigators with respect to trial data. The article was prepared by the authors, who had full access to thedata and who made final decisions on content, and the steering committee(including a minority membership of Novartis employees) reviewed the article Recurrence (%)
Proportion Breast Cancer
Between March 1998 and May 2003, 8,028 women were randomly assigned to receive 5 years of adjuvant endocrine therapy with letro- Years After Random Assignment
zole, tamoxifen, or a sequence of these agents. Of these women, 4,933patients were allocated to continuous therapy with either letrozole or tamoxifen (Fig 1). Eleven patients (seven assigned to letrozole and four assigned to tamoxifen) withdrew consent, declined all therapy, and submitted no follow-up data after random assignment. The re- sults of the remaining 4,922 patients are reported here.
Patient characteristics were well-balanced and similar to those reported for the primary core analysis.3 Details are available in Appendix Table A1 (online only). Median follow-up time of themonotherapy arms for this updated analysis was 51 months, which Malignancy (%)
is substantially longer than that of the primary core analysis be- Proportion Second
cause of the exclusion of patients on the sequential arms censoredat therapy switch.
Approximately half of the patients were still receiving study ther- apy, whereas more than 1,200 had completed 5 years of treatment.
Years After Random Assignment
More patients in the letrozole group discontinued trial treatment early as a result of an adverse event (12.3% of patients on letrozole and 11.1% of patients on tamoxifen), whereas more patients in the tamox- ifen group discontinued treatment early as a result of disease progres- sion (7.9% of patients on letrozole and 11.5% of patients on tamoxifen; Appendix Table A2, online only).
Efficacy
Details of events by therapy arm are listed in Table 1. The protocol-defined primary end point was DFS. Altogether, 352 DFS events were recorded among the 2,463 patients assigned to letrozole, and 418 DFS events were recorded among the 2,459 patients assigned Proportion Death Without
a Prior Cancer Event (%)
to tamoxifen (Fig 2). The hazard ratio for DFS was 0.82 (95% CI, 0.71 to 0.95; P ϭ .007), and the 5-year DFS survival estimates were 84.0% and 81.1% for letrozole and tamoxifen, respectively (Fig 2). Hazard Years After Random Assignment
ratios for the secondary end points defined earlier were similar to that Fig 4. Cumulative incidence of (A) breast relapse, (B) second (nonbreast) malig-
of the primary end point of DFS (Fig 3A), although the hazard ratios nancy, and (C) death without prior cancer event comparing letrozole with tamoxifen.
for overall survival and systemic DFS were not statistically significant.
We explored various protocol-defined subgroups to identify whether there was any apparent difference in the relative efficacy of n ϭ 8), and other (letrozole, n ϭ 39; tamoxifen, n ϭ 31). Causes of letrozole on DFS compared with the overall benefit observed. No death without prior cancer event included cerebrovascular accident subgroups showed significantly different relative efficacy; in particu- (letrozole, n ϭ 8; tamoxifen, n ϭ 3), thromboembolic event (letrozole, lar, no significant heterogeneity was observed by nodal involvement n ϭ 3; tamoxifen, n ϭ 3), cardiac (letrozole, n ϭ 12; tamoxifen, n ϭ 7), status or progesterone receptor status (Fig 3B).
sudden death of unknown cause (letrozole, n ϭ 7; tamoxifen, n ϭ 11), Cumulative incidence analyses of breast cancer recurrence, sec- and other causes (letrozole, n ϭ 30; tamoxifen, n ϭ 24).
ond nonbreast cancer, and death without cancer event are shown inFigure 4. Types of second primary cancers included endometrial Safety
(letrozole, n ϭ 4; tamoxifen, n ϭ 16), colon (letrozole, n ϭ 10; Prespecified adverse events, including cholesterol values (90.8% tamoxifen, n ϭ 13), lung (letrozole, n ϭ 5; tamoxifen, n ϭ 8), ovarian nonfasting), were collected every 6 months while on study.3 The (letrozole, n ϭ 2; tamoxifen, n ϭ 6), renal (letrozole, n ϭ 3; tamoxifen, analysis population for safety included 4,895 patients, excluding 27 Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved. Coates et al
patients who did not receive any trial treatment. More patients DISCUSSION
receiving letrozole, compared with patients receiving tamoxifen,reported at least one adverse event of any grade (2,292 patients v All reported trials show that modern aromatase inhibitors reduce the 2,165 patients, respectively) and at least one life-threatening or risk of relapse of early breast cancer among postmenopausal women fatal adverse event (113 of 2,448 patients [4.6%] v 92 of 2,447 with endocrine-responsive early breast cancer, whether in direct com- patients [3.8%], respectively). Table 2 lists the worst grade of adverse parison to the standard agent tamoxifen or as extended therapy after events by type. Safety profiles of letrozole and tamoxifen in this up- completion of tamoxifen. Uncertainty persists about the optimal time dated analysis of monotherapy are generally consistent with the pre- to introduce aromatase inhibitor therapy (whether initially as in the viously reported results of BIG 1-98. Similar to the primary core present analysis of BIG 1-98 and in the Arimidex, Tamoxifen, Alone or results,3 patients on tamoxifen experienced significantly more throm- in Combination [ATAC] trial1; after approximately 2 years of tamox- boembolic events, endometrial pathology, hot flashes, night sweats, ifen2,4,5; or after completion of 5 years of tamoxifen6). Models seeking and vaginal bleeding. Patients on letrozole experienced significantly to compare these approaches have been proposed,15 but direct ran- more bone fractures, arthralgia, low-grade cholesterol elevation, and domized comparisons are as yet lacking. One such comparison will cardiovascular events other than ischemic heart disease and cardiacfailure. The relatively higher recording of low-grade cholesterol eleva- become available when data from the sequential therapy arms of BIG tion on letrozole may be largely an artifact reflecting a cholesterol- lowering effect of tamoxifen. Although the overall incidence of cardiac Meanwhile, a meta-analysis of aromatase inhibitor trials adverse events did not differ significantly between the two treatments, has been proposed. To present data more directly comparable a trend for higher grade cardiac events on letrozole compared with with other studies, we conducted this updated analysis of BIG 1-98 limited to the patients assigned to 5 years of continuous Note that grade 5 adverse events in Table 2 include deaths that therapy with either letrozole or tamoxifen. The initial report of occurred within 4 weeks of receiving trial treatment, whereas deaths BIG 1-98 included patients randomly assigned to the two se- without prior cancer event (which are summarized in Results, under quential arms, censoring follow-up at the time of therapy Efficacy, and Table 1) can occur at any time. The difference in number switch. Although this allowed optimal examination of the early of deaths without recurrence (60 of 2,463 of patients on letrozole, efficacy results, as the trial matures, the inclusion of these extra 2.4%; 48 of 2,459 patients on tamoxifen, 2.0%) was not statistically patients progressively biases the overall results by stressing events in the first 2 years of therapy. The present analysis avoids Table 2. Worst Grade of Adverse Events
NOTE. Adverse events were recorded during or within 28 days after stopping trial treatment. The adverse events reported in the table were recorded using specific check boxes on the case report forms, except for arthralgia and myalgia, which were recorded in an “other” category and, thus, may be underestimated. Gradeswere determined based on the National Cancer Institute Common Toxicity Criteria (version 2.0) if available; otherwise, grades were determined using criteria definedby a senior International Breast Cancer Study Group oncologist in the protocol. Fisher’s exact P values are reported for the comparison of any grade and are notadjusted for multiple comparisons.
Abbreviations: CVA, cerebrovascular accident; TIA, transient ischemic attack.
ءGrade not defined per the National Cancer Institute Common Toxicity Criteria (version 2.0).
†Fisher’s exact P Ͻ .001 for incidence of grade 3 to 5 cardiac event.
‡Other cardiovascular events included cardiovascular disorder not otherwise specified, aneurysm, aortic aneurysm rupture, aortic dilation, aortic stenosis, arteriosclerosis, atherosclerosis (obliterans), femoral arterial stenosis, hypertensive angiopathy, iliac artery stenosis, and intermittent claudication.
Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved. Letrozole v Tamoxifen for Early Breast Cancer
this potential problem and yet provides an adequately powered Disclosures of Potential Conflicts of Interest section in Information comparison with more prolonged follow-up of the two contin- uous therapy arms. Nevertheless, more prolonged follow-up of Employment: N/A Leadership: N/A Consultant: Henning Mouridsen,
Novartis; Louis Mauriac, Novartis; John F. Forbes, Novartis; Robert
this and other adjuvant trials of aromatase inhibitors is indi- Paridaens, Novartis; Istva´n La´ng, Novartis; Lucia Del Mastro, Novartis; Ian cated; at the time of this analysis, more than 1,000 patients Smith, Novartis; Andrew Wardley, Novartis; Aron Goldhirsch, Novartis remain on therapy on each arm. It will also be important to Stock: Beat Thu¨rlimann, Novartis Honoraria: Henning Mouridsen,
see whether the early benefits of aromatase inhibitors are Novartis; Louis Mauriac, Novartis; John F. Forbes, Novartis; Robert maintained with prolonged follow-up. Tamoxifen benefits are Paridaens, Novartis; Istva´n La´ng, Novartis; Lucia Del Mastro, Novartis; Ian known to persist for years after the completion of therapy.16 Smith, Novartis; Jacquie Chirgwin, Novartis; Andrew Wardley, Novartis;
Aron Goldhirsch, Novartis Research Funds: Aparna Keshaviah, Novartis;
The conclusions of the present analysis are essentially con- Louis Mauriac, Novartis; Robert Paridaens, Novartis; Monica firmatory of the primary core analysis,3 but the present data Castiglione-Gertsch, Novartis; Richard D. Gelber, Novartis; Jacquie should provide a more easily understood and unbiased basis for Chirgwin, Novartis; Tadeusz Pienkowski, Novartis; Karen N. Price, Novartis; comparison with other studies. The most relevant comparison Aron Goldhirsch, Novartis Testimony: N/A Other: N/A
is with the hormone receptor–positive cohort in the first updateof the ATAC trial, which compared 5 years of tamoxifen with AUTHOR CONTRIBUTIONS
anastrozole. At a median follow-up time of 47 months, ATACinvestigators reported 635 DFS events and a hazard ratio of 0.82 Conception and design: Alan S. Coates, Aparna Keshaviah, Beat
(95% CI, 0.65 to 0.93) favoring anastrozole. Definitions of end Thu¨rlimann, Henning Mouridsen, John F. Forbes, MonicaCastiglione-Gertsch, Richard D. Gelber, Ian Smith, Karen N. Price, points varied slightly between ATAC and BIG 1-98, with ATAC ignoring second (nonbreast) primaries in its definition of DFS Administrative support: Alan S. Coates, Monica Castiglione-Gertsch,
and BIG 1-98 ignoring ductal carcinoma in situ in its definition of DFS. In the present analysis of BIG 1-98 at a median Provision of study materials or patients: Aparna Keshaviah, Beat
follow-up time of 51 months, after ignoring second (nonbreast) Thu¨rlimann, Henning Mouridsen, Louis Mauriac, John F. Forbes, primary events, 671 DFS events were reported, with a hazard Robert Paridaens, Monica Castiglione-Gertsch, Marco Colleoni, Istva´nLa´ng, Lucia Del Mastro, Ian Smith, Jacquie Chirgwin, Jean-Marie ratio of 0.83 (95% CI, 0.71 to 0.96) favoring letrozole. The Nogaret, Tadeusz Pienkowski, Andrew Wardley, Erik H. Jakobsen, Karen hazard ratio for the exploratory end point of time to recurrence was identical between the present analysis of BIG 1-98 (Fig 3A) Collection and assembly of data: Alan S. Coates, Aparna Keshaviah,
and the aforementioned ATAC analysis, and neither study shows a Beat Thu¨rlimann, Henning Mouridsen, Louis Mauriac, John F. Forbes, significant difference in overall survival. This analysis adds to the body Robert Paridaens, Monica Castiglione-Gertsch, Marco Colleoni, Istva´n of data supporting a role for the inclusion of an aromatase inhibitor in La´ng, Lucia Del Mastro, Ian Smith, Jacquie Chirgwin, Jean-MarieNogaret, Tadeusz Pienkowski, Andrew Wardley, Erik H. Jakobsen, Karen the adjuvant therapy of postmenopausal women with receptor- Data analysis and interpretation: Alan S. Coates, Aparna Keshaviah,
Henning Mouridsen, Louis Mauriac, John F. Forbes, Monica
Castiglione-Gertsch, Richard D. Gelber, Ian Smith, Andrew Wardley,
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS
OF INTEREST
Manuscript writing: Alan S. Coates, Aparna Keshaviah, Richard D.
Gelber, Karen N. Price
Although all authors completed the disclosure declaration, the following Final approval of manuscript: Alan S. Coates, Aparna Keshaviah, Beat
authors or their immediate family members indicated a financial interest. Thu¨rlimann, Henning Mouridsen, Louis Mauriac, John F. Forbes, No conflict exists for drugs or devices used in a study if they are not being Robert Paridaens, Monica Castiglione-Gertsch, Richard D. Gelber, evaluated as part of the investigation. For a detailed description of the Marco Colleoni, Istva´n La´ng, Lucia Del Mastro, Ian Smith, Jacquie disclosure categories, or for more information about ASCO’s conflict of Chirgwin, Jean-Marie Nogaret, Tadeusz Pienkowski, Andrew Wardley, interest policy, please refer to the Author Disclosure Declaration and the Erik H. Jakobsen, Karen N. Price, Aron Goldhirsch women with primary breast cancer. N Engl J Med responsive early breast cancer to anastrozole after 2 REFERENCES
years’ adjuvant tamoxifen: Combined results of 3. Breast International Group 1-98 Collaborative
ABCSG trial 8 and ARNO 95 trial. Lancet 366:455- 1. Baum M, Buzdar A, Cuzick J, et al: Anastro-
Group: A comparison of letrozole and tamoxifen in zole alone or in combination with tamoxifen versus postmenopausal women with early breast cancer.
6. Goss PE, Ingle JN, Martino S, et al: Random-
tamoxifen alone for adjuvant treatment of post- ized trial of letrozole following tamoxifen as ex- menopausal women with early-stage breast cancer: 4. Boccardo F, Rubagotti A, Puntoni M, et al:
tended adjuvant therapy in receptor-positive breast Results of the ATAC (Arimidex, Tamoxifen, Alone or Switching to anastrozole versus continued tamox- cancer: Updated findings from NCIC CTG MA. 17.
in Combination) trial efficacy and safety update ifen treatment of early breast cancer: Preliminary results of the Italian Tamoxifen Anastrozole Trial.
7. Winer EP, Hudis C, Burstein HJ, et al:
2. Coombes RC, Hall E, Gibson LJ, et al: A
American Society of Clinical Oncology technology randomized trial of exemestane after two to three 5. Jakesz R, Jonat W, Gnant M, et al: Switch-
assessment on the use of aromatase inhibitors as years of tamoxifen therapy in postmenopausal ing of postmenopausal women with endocrine- adjuvant therapy for postmenopausal women with Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved. Coates et al
hormone receptor-positive breast cancer: Status prolonged observation of each patient: II. Analysis data: Recent advances and continuing controver- report 2004. J Clin Oncol 23:619-629, 2004 8. Goldhirsch A, Glick JH, Gelber RD, et al:
11. Kaplan EL, Meier P: Nonparametric estima-
15. Cuzick J, Sasieni P, Howell A: Should aro-
Meeting highlights: International expert consensu- tion from incomplete observation. J Am Stat Assoc matase inhibitors be used as initial adjuvant treat- son the primary therapy of early breast cancer 2005.
ment or sequenced after tamoxifen? Br J Cancer 12. Cox DR: Regression models and life tables.
9. Carlson RW, Brown E, Burstein HJ, et al: NCCN
16. Early Breast Cancer Trialists’ Collaborative
Task Force Report: Adjuvant therapy for breast cancer.
13. Gray RJ: A class of k-sample tests for com-
Group: Effects of chemotherapy and hormonal ther- J Natl Compr Canc Netw 4:S1-S26, 2006 (suppl 1) paring the cumulative incidence of a competing risk.
apy for early breast cancer on recurrence and 15- 10. Peto R, Pike MC, Armitage P, et al: Design
year survival: An overview of the randomised trials.
and analysis of randomized clinical trials requiring 14. Agresti A: Exact inference for categorical
Acknowledgment
We thank the patients, physicians, nurses, and data managers who participated in this clinical trial; Novartis; and the International Breast Appendix
The Appendix is included in the full-text version of this article, available online at www.jco.org. It is not included in the PDF version Information downloaded from jco.ascopubs.org and provided by M D ANDERSON HOSP on July 22, 2010 from Copyright 2007 by the American Society of Clinical Oncology. All rights reserved.

Source: http://www.miraclemediatv.com/MMTVMEDIA/17/1340_1282937986968.PDF