Didattica.unibocconi.it

Adjuvant Chemotherapy Followed by Goserelin Versus
Either Modality Alone for Premenopausal Lymph
Node–Negative Breast Cancer: A Randomized Trial
International Breast Cancer Study Group (IBCSG)1 Background: Although chemotherapy and ovarian function
Breast cancer is the most frequent non-cutaneous malignancy suppression are both effective adjuvant therapies for pa-
diagnosed among women in the Western world (1). The majority tients with early-stage breast cancer, little is known of the
of breast cancers are diagnosed at an operable stage, i.e., as a efficacy of their sequential combination. In an International
primary tumor without or with axillary lymph node metastases Breast Cancer Study Group (IBCSG) randomized clinical
but not widespread metastatic disease. Despite the considerable trial (Trial VIII) for pre- and perimenopausal women with
number of putative prognostic factors that have been described lymph node–negative breast cancer, we compared sequential
for breast cancer, the status of the axilla remains the most chemotherapy followed by the gonadotropin-releasing hor-
important prognostic factor (2). Up to 80% of patients with mone agonist goserelin with each modality alone. Methods:
newly diagnosed breast cancer in countries with mammographic From March 1990 through October 1999, 1063 patients
screening programs do not have involvement of axillary lymph stratified by estrogen receptor (ER) status and radiotherapy
plan were randomly assigned to receive goserelin for 24
Despite undergoing radical surgery, some groups of patients with lymph node–negative disease have recurrent disease at a months (n ؍ 346), six courses of “classical” CMF (cyclophos-
rate exceeding 40% (3), possibly because of previously unde- phamide, methotrexate, 5-fluorouracil) chemotherapy (n ؍
tected micro-metastases that later become overt. The aim of 360), or six courses of classical CMF followed by 18 months
systemic adjuvant therapy is the eradication or prevention of of goserelin (CMF 3 goserelin; n ؍ 357). A fourth arm (no
disease progression or recurrence after surgery. Women receiv- adjuvant treatment) with 46 patients was discontinued in
ing systemic adjuvant therapy have shown a clinically signifi- 1992. Tumors were classified as ER-negative (30%), ER-
cant improvement in both disease-free survival (DFS) and over- positive (68%), or ER status unknown (3%). Twenty percent
all survival (OS) (4 – 6). Systemic adjuvant therapies include of patients were aged 39 years or younger. The median
cytotoxic chemotherapy and endocrine therapies. Suppressed follow-up was 7 years. The primary outcome was disease-
ovarian function, which reduces or eliminates estrogen produc- free survival (DFS). Results: Patients with ER-negative tu-
tion, was the first adjuvant treatment studied in clinical trials mors achieved better disease-free survival if they received
CMF (5-year DFS for CMF ؍ 84%, 95% confidence interval
[CI] ؍ 77% to 91%; 5-year DFS for CMF 3 goserelin ؍
1Writing Committee: Monica Castiglione-Gertsch, Anne O’Neill, Karen N.
88%, 95% CI ؍ 82% to 94%) than if they received goserelin
Price, Aron Goldhirsch, Alan S. Coates, Marco Colleoni, M. Laura Nasi, Marco alone (5-year DFS ؍ 73%, 95% CI ؍ 64% to 81%). By
contrast, for patients with ER-positive disease, chemother-
Affiliations of Writing Committee members: International Breast Cancer Study apy alone and goserelin alone provided similar outcomes
Group (IBCSG) Coordinating Center and Inselspital, Bern, Switzerland (MCG); (5-year DFS for both treatment groups ؍ 81%, 95% CI ؍
IBCSG Statistical Center, Dana-Farber Cancer Institute, Harvard School of 76% to 87%), whereas sequential therapy (5-year DFS ؍
Public Health and Frontier Science and Technology Research Foundation, Bos-ton, MA (AO, KNP, MB, RDG); IBCSG Scientific Committee, European 86%, 95% CI ؍ 82% to 91%) provided a statistically non-
Institute of Oncology, Milan, Italy, and Oncology Institute of Southern Swit- significant improvement compared with either modality
zerland, Bellinzona, Switzerland (AG); IBCSG Scientific Committee, University alone, primarily because of the results among younger
of Sydney and The Cancer Council Australia, Sydney, Australia (ASC); European women. Conclusions: Premenopausal women with ER-
Institute of Oncology, Milan (MC); IBCSG Coordinating Center, Bern (MLN).
negative (i.e., endocrine nonresponsive), lymph node–nega-
Correspondence to: Monica Castiglione-Gertsch, MD, IBCSG Coordi- tive breast cancer should receive adjuvant chemotherapy.
nating Center, Effingerstrasse 40, CH-3008 Bern, Switzerland (e-mail: monica.
castiglione@siak.ch).
For patients with ER-positive (i.e., endocrine responsive)
See “Appendix” for the names and affiliations of the participants and authors disease, the combination of chemotherapy with ovarian func-
of the International Breast Cancer Study Group Trial VIII.
tion suppression or other endocrine agents, and the use of
See “Notes” following “References.” endocrine therapy alone should be studied. [J Natl Cancer
Inst 2003;95:1833– 46]
Journal of the National Cancer Institute, Vol. 95, No. 24, Oxford UniversityPress 2003, all rights reserved.
Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 involving premenopausal women. Suppressed ovarian function with last normal menstrual period within 1 year, 2) aged 52 years was achieved by surgical castration or by irradiation of the or younger with last normal menstrual period within 3 years, 3) ovaries. More recently, gonadotropin-releasing hormone agonist aged 55 years or younger with hysterectomy but no bilateral drugs such as goserelin have been used. Suppressed ovarian oophorectomy (for patients aged older than 45 years, biochem- function can also result from the use of cytotoxic agents. Patients ical confirmation of ovarian function was requested), or 4) who experienced amenorrhea, a consequence of suppressed biochemical evidence of continuing ovarian function (for doubt- ovarian function, after chemotherapy had longer disease-free survival than patients who maintained ovarian function in some All patients had a histologically proven unilateral breast studies (7–11) but not in others (12–14). Thus, a controversy cancer of stage T , T , T , T , T , N , or M [according to the exists regarding the use of agents that suppress ovarian function staging system of the Union Internationale Contre le Cancer after chemotherapy (4). Five years of tamoxifen, the most com- 1987 (20)], with either ER-positive or ER-negative primary mon endocrine therapy used in the adjuvant setting, has been tumors. The ER-unknown status was allowed only if ER deter- shown to be effective for reducing the risk of recurrent disease mination was not possible because of the lack of tumor material.
and death in premenopausal and postmenopausal patients with Steroid hormone receptor concentrations in the primary tumors endocrine-responsive breast tumors (5).
were determined by standard methods (21,22). ER concentra- In 1990, the International Breast Cancer Study Group tions of at least 10 fmol/mg of cytosol protein by ligand-binding (IBCSG) initiated a clinical trial (Trial VIII) for premenopausal assay were considered positive; lower values were considered and perimenopausal patients with lymph node–negative breast negative. Determination of steroid hormone receptor status by cancer to examine the role of adjuvant treatment using chemo- immunohistochemistry was allowed later in the study. Conse- therapy, ovarian suppression with goserelin, or the sequential quently, ER status for 33% of the patients was determined by combination of both modalities. Here, we provide the first report immunohistochemistry, and participating center values for pos- of results after a median follow-up of 7 years for women Surgery to remove the primary tumor was either a total mastectomy with axillary clearance or a conservative procedure (quadrantectomy or lumpectomy) with axillary lymph node dis- ATIENTS AND METHODS
section. Radiotherapy was recommended after breast-conserving Study Design
surgery and was postponed until the end of chemotherapy, ifapplicable (23). Staging before randomization included chest From March 1990 through October 1999, 1111 premeno- x-ray, contralateral mammogram, bone scintogram (if clinically pausal and perimenopausal patients were randomly assigned to indicated), and hematologic, liver, and renal function tests.
receive no adjuvant systemic treatment, six 28-day courses of Clinical, hematologic, and biochemical assessments were re- “classical” CMF chemotherapy (in which one course consisted quired every 3 months for the first year, every 6 months for the of oral cyclophosphamide at 100 mg/m2 on days 1–14, intrave- second year, and yearly thereafter. Modified World Health Or- nous methotrexate at 40 mg/m2 on days 1 and 8, and intravenous ganization toxicity grading criteria were used (24). Mammogra- 5-fluorouracil at 600 mg/m2 on days 1 and 8), 24 monthly phy was performed yearly. The data management and medical subcutaneous implants of goserelin (3.6 mg) every 28 days, or staff reviewed all study records (initial data, treatment, toxicity, six 28-day courses of classical CMF followed by 18 monthly and recurrence) and conducted regular site visit audits. In par- implants of goserelin. Systemic adjuvant therapy was to begin ticular, the study chair (M. Castiglione-Gertsch) reviewed the within 6 weeks of primary surgery. For the sequential treatment records for all grade 3 or worse toxicities.
arm, the first goserelin implant was scheduled to be given on day28 of the sixth course of CMF. Informed consent was required End Points and Statistical Considerations
according to the criteria established within the individual coun-tries. The protocol was reviewed and approved by institutional Disease-free survival was defined as the length of time from the date of randomization to any recurrent disease (including In April 1992, on the basis of results from other trials (15– ipsilateral breast recurrence), the appearance of a second pri- 18), randomization to the no-adjuvant-treatment control arm was mary cancer (including contralateral breast cancer), or death, discontinued. At that time, the trial had accrued 205 patients, 46 whichever occurred first. Overall survival was defined as the of whom had been randomly assigned to the control arm. The length of time from the date of randomization to death from any results for this small initial cohort have been previously pub- lished (19). This article reports the results of comparisons be- Disease-free survival and overall survival percentages, stan- tween the three active adjuvant-treatment arms.
dard errors, and treatment effect comparisons were obtained Randomization was conducted centrally (at the coordinating from the Kaplan–Meier method (25), Greenwood’s formula centers in Bern, Switzerland, and Sydney, Australia) after strat- (26), and log-rank tests (27), respectively. Cox proportional ification according to estrogen receptor (ER) status (negative, hazards regression models (28) were used to control for prog- positive, or unknown), whether radiotherapy was planned after nostic features, to estimate relative risks (RRs) and 95% confi- breast-conserving surgery (yes or no), and by participating in- dence intervals (CIs) for the treatment comparisons, and to test stitution (see Appendix). The permuted blocks randomization for interactions between potential predictive factors and treat- schedule was produced by use of pseudorandom numbers gen- ment effects. To check assumptions of proportionality, curves of the log of the cumulative hazard for each value of a covariate Pre- or perimenopausal status was defined as having one of adjusted for other covariates in the model were plotted and the following sets of characteristics: 1) aged older than 52 years assessed visually to determine if the vertical shift between the Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 curves was constant over time. The data appeared to meet theassumptions of proportionality in all cases, with the exception ofage for the goserelin alone versus CMF alone comparison in theER-negative cohort. Adding an interaction term for age and timein the model for this treatment comparison did not change thetreatment effect estimate. All probability values were obtainedfrom two-sided tests. Results are reported at a median follow-upof 7 years.
Treatment– covariate interactions were studied by use of the nonparametric Subpopulation Treatment Effect Pattern Plot(STEPP) methodology (29,30). STEPP involves defining severaloverlapping subgroups of patients on the basis of a covariate ofinterest and studying the resulting pattern of the treatment ef-fects estimated within each subgroup. In this article, patient age Fig. 1. Flow chart of enrollment and assessability for the primary analysis for
at study entry was the covariate of interest, and the treatment patients enrolled in the International Breast Cancer Study Group (IBCSG) TrialVIII. CMF ϭ cyclophosphamide at 100 mg/m2 orally on days 1–14, methotrex- effects estimated within each age subgroup were measured in ate at 40 mg/m2 intravenously on days 1 and 8, and 5-fluorouracil at 600 mg/m2 terms of 5-year disease-free survival percentages, both overall intravenously on days 1 and 8, repeated for six 28-day courses. mos. ϭ months.
and for cohorts defined according to ER status.
The intention to perform separate analyses according to ER status was specified in the original protocol. After the closure of the no-adjuvant-treatment control arm in 1992, the study was Disease-Free Survival and Overall Survival
redesigned to assess whether six courses of CMF followed by 18implants of goserelin improved results relative to six courses of Overall, no differences were observed among the three treat- CMF alone (80% power to detect an improvement in 5-year DFS ment groups (CMF chemotherapy, goserelin, CMF chemother- from 80% to 88%) and whether 24 implants of goserelin and six apy followed by goserelin) in terms of disease-free survival (Fig.
courses of CMF were comparable (95% chance to reject equiv- 2, A, and Table 2) or overall survival (Fig. 2, B). However, alence if goserelin [72% 5-year DFS] was less effective than differences among the treatment groups were suggested for CMF [80% 5-year DFS]). Two hundred twenty-four events were subpopulations defined according to ER status. Disease-free required; 228 were observed at the time of this analysis.
survival for patients with ER-negative tumors who received The Data and Safety Monitoring Committee reviewed accrual CMF alone (5-year DFS ϭ 84%, 95% CI ϭ 77% to 91%) or and safety data twice a year. Two predetermined interim efficacy CMF followed by goserelin (5-year DFS ϭ 88%, 95% CI ϭ analyses were performed (in December 1997 and June 2000), 82% to 94%) was greater than that for patients with ER-negative and study continuation was recommended on both occasions. In tumors who received only goserelin (5-year DFS ϭ 73%, 95% 1998, a protocol amendment restricted enrollment to patients CI ϭ 64% to 81%) (Fig. 2, C, and Table 2). By contrast, with ER-positive tumors on the basis of evidence from other disease-free survival estimates for patients with ER-positive trials that ovarian ablation might not be effective for patients tumors who received CMF alone (5-year DFS ϭ 81%, 95% CI ϭ 76% to 87%) or who received goserelin alone (5-year DFS ϭ81%, 95% CI ϭ 76% to 87%) were equivalent, whereas there Patient Eligibility and Characteristics
was a modest, statistically nonsignificant advantage associatedwith the sequential administration of CMF followed by goserelin(5-year DFS ϭ 86%, 95% CI ϭ 82% to 91%) (Fig. 2, D, and Of the 1111 patients randomly assigned, 46 were assigned to Table 2). Unplanned, retrospective subgroup analyses according the no-adjuvant-treatment arm and 1065 were assigned to one of to age suggested that, among women with ER-negative tumors, the three adjuvant-treatment arms (Fig. 1). Two patients enrolled the superiority of the CMF-containing regimens compared with from a noncompliant participating center were excluded from all goserelin alone was seen for both older and younger women, analyses. Of the remaining 1063 patients, 20 (1.9%) patients did whereas for women with ER-positive tumors, the advantage of not meet protocol eligibility criteria for the following reasons: CMF followed by goserelin was seen only for younger women postmenopausal status (n ϭ 11), in situ disease only (n ϭ 2), resection margins involved with tumor (n ϭ 3), prior malig- STEPP analyses were used to evaluate the differences in nancy (n ϭ 2), lymph node–positive disease (n ϭ 1), and treatment effects in terms of 5-year disease-free survival accord- medical unsuitability (n ϭ 1). However, all 20 ineligible patients ing to age (Fig. 3). For this sliding-window STEPP analysis, are included in the intent-to-treat analyses.
each subpopulation contained approximately 165 patients, and The characteristics of the 1063 assessable patients who were each subsequent subpopulation was formed moving from left to enrolled in active treatment arms are shown in Table 1. The right by dropping approximately 30 patients with the lowest age median age was 45 years (range ϭ 28 –58 years). Thirty percent and adding approximately 30 patients with the next higher age.
(315) of the patients had primary tumors classified as ER- The x coordinate indicates the median age for the patients in negative (11% ER absent and 19% ER low), 68% (720) were each subpopulation. The y coordinate indicates the 5-year classified as ER-positive, and 3% (28) were classified as ER- disease-free survival percent estimated using the Kaplan–Meier unknown. The median number of axillary lymph nodes exam- method on data from patients in each subpopulation. The results ined was 16 (range ϭ 5– 60 lymph nodes).
for the entire study population show that, without separation of Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 Table 1. Patients’ characteristics according to treatment*
All patients
ER-negative cohort
ER-positive cohort
*CMF ϫ 6 ϭ cyclophosphamide at 100 mg/m2 on days 1–14, orally; methotrexate at 40 mg/m2 on days 1 and 8, intravenously; and 5-fluorouracil at 600 mg/m2 on days 1 and 8, intravenously; repeated for six 28-day courses. Goserelin ϫ 24 ϭ goserelin at 3.6 mg by subcutaneous implant monthly for 24 months. For the sequential combination therapy, CMF ϫ 6 was followed by goserelin ϫ 18. 3 ϭ followed by; ER ϭ estrogen receptor; RT ϭ radiotherapy.
†ER status was determined by a ligand-binding assay for 67% of the patients and by immunohistochemistry for the other 33% of the patients (21,22). For the ligand-binding assay, ER concentrations of at least 10 fmol/mg of cytosol protein were considered positive. For the immunohistochemistry, participating center values were used.
‡Tumor grade was determined at each participating site (31).
§Methodologically, for tumors graded as grade 1, it is possible that the ER-negative classification might be false.
Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 Fig. 2. Kaplan–Meier plots of disease-free survival (DFS) (panel A) and overall
or older in the ER-positive cohort (panel F). The number of patients, number of
survival (OS) (panel B) for 1063 pre- and perimenopausal women with lymph
DFS events, 5-year DFS % and 95% confidence interval (CI) for each treatment node–negative breast cancer enrolled in the International Breast Cancer Study group, and the relative risk of an event (recurrent disease, second malignancy, or Group (IBCSG) Trial VIII according to randomized treatment group at a median death), 95% CI, and P value for each pairwise treatment comparison for DFS are follow-up of 7 years. Also shown are Kaplan–Meier plots of DFS for 315 shown in Table 2. For the overall survival in panel B, the 5-year OS % are 95%
patients in the estrogen receptor (ER)–negative cohort (panel C), for 720
(95% CI ϭ 93% to 97%; 35 deaths) for goserelin alone, 93% (95% CI ϭ 90% patients in the ER-positive cohort (panel D), for 126 patients aged 39 years or
to 95%; 37 deaths) for CMF alone, and 95% (95% CI ϭ 93% to 97%; 27 deaths) younger in the ER-positive cohort (panel E), and for 594 patients aged 40 years
Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 Table 2. Disease-free survival (DFS) according to treatment*
*CMF ϫ 6 ϭ cyclophosphamide at 100 mg/m2 on days 1–14, orally; methotrexate at 40 mg/m2 on days 1 and 8, intravenously; and 5-fluorouracil at 600 mg/m2 on days 1 and 8, intravenously; repeated for six 28-day courses. Goserelin ϫ 24 ϭ goserelin at 3.6 mg by subcutaneous implant monthly for 24 months. For thesequential combination therapy, CMF ϫ 6 was followed by goserelin ϫ 18. CI ϭ confidence interval; ER ϭ estrogen receptor; 3 ϭ followed by.
†For each analysis, the relative risk is the risk of an event (recurrent disease, second malignancy, or death [Table 4]) for the first cohort listed compared with that for the second cohort listed. A value greater than 1.00 indicates an increased risk of an event for the first cohort listed.
‡All statistical tests were two-sided.
the analysis according to ER status, there was no clear pattern of status of the primary tumor, multiple regression analyses of treatment differences according to age (Fig. 3, A). By contrast, disease-free survival were conducted separately for the ER- the CMF-containing regimens provide superior disease-free sur- negative and ER-positive cohorts (Table 3). Factors for treat- vival across all age groups for patients with ER-negative tumors ment, age, primary therapy, tumor size, and tumor grade were (Fig. 3, B), whereas the benefit of the sequential regimen for included in all models. Treatment differences remained statisti- patients with ER-positive disease increased substantially as the cally significant for the ER-negative cohort, even after adjust- median age of the patient subpopulation decreased below ment for other factors. For the ER-positive cohort, age, primary approximately age 43 years (Fig. 3, C). Fig. 3, C, also illustrates treatment (increased risk of an event for breast-conserving sur- that the equivalent outcome for CMF alone and goserelin alone gery without radiotherapy), and tumor grade were prognostically Interactions between the magnitude of treatment differences and ER status were assessed using Cox proportional hazards Incidence of Amenorrhea
models. Despite low statistical power, tests for interactions sug-gested that, compared with the CMF-containing regimens, gos- The percentage of patients who reported no menses during erelin alone was less effective for the ER-negative cohort than each month after randomization according to treatment group is for the ER-positive cohort (interactions: P ϭ .13 for goserelin shown in Fig. 4. The percentage of patients who reported no compared with the CMF– goserelin sequence and P ϭ .17 for menses during each month from the no-adjuvant-therapy control group is provided as an estimate for the natural rate of cessation Considering the suggestive, statistically nonsignificant differ- of menses associated with increasing age.
ences in treatment effect and the current approach of tailoring For patients aged 39 years or younger (Fig. 4, A), goserelin adjuvant therapy according to the steroid hormone receptor induced amenorrhea within 2 months of study entry for 90% of Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 the end of six courses of CMF. Among patients in whomgoserelin was not given after CMF, menses resumed in approx-imately 15%, although amenorrhea continued in approximately35%– 40% of patients throughout the 36-month period of obser-vation. Among patients who received goserelin after CMF, vir-tually all achieved amenorrhea during the 18-month goserelintreatment period. Interestingly, resumption of menses after ces-sation of goserelin was slower in patients who had receivedinitial CMF chemotherapy than in those who did not receiveCMF chemotherapy, although menses did return in approxi-mately 40% of patients by the end of the 36-month follow-upperiod—the same percentage as among patients who had re-ceived goserelin alone.
The pattern of incidence of amenorrhea over time was dif- ferent for patients aged 40 years or older at the time of studyentry (Fig. 4, B). The median age at study entry for this patientcohort was 46 years. Chemotherapy-induced amenorrhea wasobserved sooner and in a larger percentage of patients than wasobserved in the younger cohort. More than 90% of patients whoreceived six courses of CMF achieved amenorrhea by the end ofchemotherapy. Although menses resumed in a few patients whodid not receive goserelin after chemotherapy, a high incidence ofamenorrhea was observed during the entire 36-month follow-upperiod, regardless of whether goserelin was used. The incidenceof amenorrhea after completion of goserelin alone was the same(approximately 55%) as that observed for the no-adjuvant-therapy group during the third year of follow-up.
Sites of Treatment Failure
Of the 1063 patients, 228 (21.4%) had recurrent disease or died (Table 4). For the ER-negative cohort, the percentage ofpatients with visceral metastases was lower for the CMFgroup than for the goserelin alone group (difference ϭ 4.6%,95% CI ϭ Ϫ2.3% to 11.5%), and the percentage of patients withlocal recurrences was lower for the CMF followed by goserelingroup than for the goserelin alone group (difference ϭ 4.6%,95% CI ϭ Ϫ1.4% to 10.6%). For the ER-positive cohort, thepercentage of patients with local recurrences was lower for theCMF or CMF followed by goserelin groups than for the gosere-lin alone group (difference with CMF ϭ 4.3%, 95% CI ϭ 0.0%to 8.6%, and difference with CMF followed by goserelin ϭ3.0%, 95% CI ϭ Ϫ1.5% to 7.5%).
Fig. 3. Subpopulation Treatment Effect Pattern Plots (STEPP) showing 5-year
disease-free survival (DFS) percentage according to randomized treatment group
CMF Treatment and Toxicity
and age for all patients (panel A), the estrogen receptor (ER)–negative cohort
(panel B), and the ER-positive cohort (panel C) of women enrolled in the
International Breast Cancer Study Group (IBCSG) Trial VIII. For this sliding-
Among the 717 patients randomly assigned to receive six window STEPP analysis, each subpopulation contained approximately 165 pa- courses of CMF (either alone or followed by goserelin), 646 tients, and each subsequent subpopulation was formed moving from left to right (90%) completed all six courses, 55 (8%) received at least one by dropping approximately 30 patients with the lowest age and adding approx- but fewer than six courses, and 16 (2%) received no chemother- imately 30 patients with the next higher age. The x coordinate indicates the apy. Patient compliance was similar among treatment groups median age for the patients in each subpopulation. The y coordinate indicates the5-year DFS percentage estimated using the Kaplan–Meier method on data from regardless of whether CMF was followed by goserelin. Grade 3 or worse toxicities (primarily leukopenia, neutropenia, and nau-sea/vomiting) were experienced by 18.8% of the patients during the patients and within 3 months for virtually all patients. Am- CMF, including three life-threatening toxicities (two pulmonary enorrhea continued until the end of treatment (i.e., at 24 embolisms and one cerebrovascular accident). There were no months), when menses resumed in all but a few patients. By treatment-related deaths. Of the 701 patients who received at contrast, chemotherapy-induced amenorrhea was achieved more least one course of CMF, 18.7% reported alopecia requiring slowly and was observed in approximately 50% of patients by Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 Table 3. Multiple regression analyses of disease-free survival for estrogen receptor (ER)-negative and ER-positive cohorts*
Results of multiple regressions defined according to treatment comparison ER-negative
ER-positive
*CMF ϫ 6 ϭ cyclophosphamide at 100 mg/m2 on days 1–14, orally; methotrexate at 40 mg/m2 on days 1 and 8, intravenously; and 5-fluorouracil at 600 mg/m2 on days 1 and 8, intravenously; repeated for six 28-day courses. Goserelin ϫ 24 ϭ goserelin at 3.6 mg by subcutaneous implant monthly for 24 months. For thesequential combination therapy, CMF ϫ 6 was followed by goserelin ϫ 18. CI ϭ confidence interval; 3 ϭ followed by; Rx group ϭ treatment group indicatedat the top of the column; BCS ϭ breast-conserving surgery; RT ϭ radiotherapy.
†Relative risk for each analysis is the risk of an event (recurrent disease, second malignancy, or death [Table 4]) for the first cohort listed compared with that for the second cohort listed. A value greater than 1.00 indicates an increased risk of an event for the first cohort listed.
‡All statistical tests (Wald test for single covariate; likelihood ratio test for multiple covariates) were two-sided. Models included indicator variables for unknown §Methodologically, for tumors graded as grade 1, it is possible that the ER-negative classification might be false.
Goserelin Treatment and Toxicity
patients assigned to the no-adjuvant-treatment group was lessthan that for patients in the three treatment groups combined, but Among the 346 patients assigned to goserelin alone, 304 with the small number of patients, the difference was not statis- (88%) received at least 22 implants, 36 (10%) received fewer tically significant (P ϭ .19). Five-year disease-free survival implants because of recurrent disease (5%) or other reasons percentages (95% CI; sample size) were 61% (95% CI ϭ 47% (5%), and six (2%) received no goserelin. Among the 357 to 75%; n ϭ 46) for no treatment, 73% (95% CI ϭ 62% to 84%; patients assigned to goserelin following CMF, 289 (81%) re- n ϭ 63) for goserelin alone, 79% (95% CI ϭ 67% to 91%; n ϭ ceived at least 17 implants, 37 (10%) received fewer implants 43) for CMF alone, and 81% (95% CI ϭ 71% to 92%; n ϭ 53) because of recurrent disease (3%) or other reasons (7%), and 31 for CMF followed by goserelin. The corresponding values for (9%) received no goserelin. Grade 3 or worse toxicities (primar- 5-year disease-free survival percentages (95% CI; sample size) ily weight gain) were experienced by 3.9% of the 666 patients were 46% (95% CI ϭ 19% to 73%; n ϭ 13), 64% (95% CI ϭ who received at least one goserelin implant (4.7% in the gos- 44% to 84%; n ϭ 22), 89% (95% CI ϭ 74% to 98%; n ϭ 18), 89% erelin alone group and 3.1% in the goserelin following CMF (95% CI ϭ 74% to 98%; n ϭ 18) for the ER-negative cohort and group). One life-threatening (suicidal) depression was reported 67% (95% CI ϭ 50% to 84%; n ϭ 30), 73% (95% CI ϭ 58% to during goserelin treatment (after 6 months of CMF and four 88%; n ϭ 34), 70% (95% CI ϭ 52% to 89%; n ϭ 24), 81% (95% CI ϭ 68% to 95%; n ϭ 32) for the ER-positive cohort.
Comparisons With No Adjuvant Treatment
DISCUSSION
The median follow-up for the 205 patients who were enrolled IBCSG Trial VIII for premenopausal and perimenopausal before April 2, 1992, was 10.4 years. Disease-free survival for women with lymph node–negative breast cancer began in 1990, Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 Ovarian function suppression was the first adjuvant systemic treatment studied for patients with early-stage breast cancer(4,32–34). Chemotherapy is effective adjuvant therapy for pre-menopausal women (6). For several years, the effects of cyto-toxic agents on ovarian function were studied, but the interpre-tation of results remains controversial (7–14). The associationbetween chemotherapy-induced amenorrhea and outcome, how-ever, was confounded in retrospective analyses with chemother-apy dose intensity and duration. Although induction of amenor-rhea was found to be an important indicator of improvedoutcome for chemotherapy regimens that were less dose intenseand of shorter duration (7,8), the effect of amenorrhea on out-come was less evident when an intensive chemotherapy regimenwas used (12). Recently, however, a randomized trial showedstatistically significant increases in both the incidence of amen-orrhea and disease-free survival for very young patients (i.e.,aged 39 years or younger) with ER-positive tumors who re-ceived high-dose chemotherapy with peripheral blood progeni-tor cell support compared with standard doses of chemotherapy(35).
The incidence of amenorrhea was studied meticulously in IBCSG Trial VIII. The spontaneous amenorrhea rate in our olderpatients (i.e., aged 40 years or older) was approximately 50% at3 years (Fig. 4, B, untreated group). We observed that the onsetof ovarian function suppression was delayed slightly for patientswho received chemotherapy relative to those who received go-serelin alone. However, despite this delay, the treatment out-come was similar for all three groups of older patients withendocrine-responsive disease (Fig. 2, F), whereas the combina-tion of the two modalities was better than the individual modal-ities for younger patients (i.e., those aged 39 years or younger)(Fig. 2, E). Although the underlying mechanism associated withthis observation is unclear, it is possible that if chemotherapycompletely suppressed ovarian function in the older patients,then subsequent treatment with goserelin may not have had theopportunity to improve outcome. This possibility might alsoexplain the observation in the Early Breast Cancer Trialists’Collaborative Group overview (4), which included mostlywomen aged 40 years or older and showed a lack of benefit ofoophorectomy when administered in addition to chemotherapy.
Fig. 4. Percentage of patients enrolled in the International Breast Cancer Study
By contrast, for younger women in IBCSG Trial VIII, resump- Group (IBCSG) Trial VIII with amenorrhea during each month from random- tion of menses following completion of goserelin was slower ization according to treatment. Panel A shows the results for patients aged 39
and occurred less often during the 36-month follow-up for years or younger and panel B shows the results for patients aged 40 years or
older.
women who received initial CMF chemotherapy than for thosewho received no CMF chemotherapy. Consequently, prolonged when tamoxifen was not routinely used for premenopausal pa- amenorrhea in a higher percentage of patients treated with tients and when it was unclear whether ovarian function sup- chemo-endocrine therapy than in those treated with endocrine pression might be effective exclusively for patients with therapy or chemotherapy alone may have contributed to the endocrine-responsive disease (i.e., ER-positive tumors). Al- prolonged disease-free survival associated with the combination though overall the differences in disease-free survival among the therapy observed in the younger cohort.
three treatment groups studied in IBCSG Trial VIII were not Ovarian ablation (4), tamoxifen (5), and polychemotherapy statistically significant, differential effects were observed when (6) have all been shown to improve disease-free survival and the analyses were conducted separately for the ER-negative and overall survival, and their combined use has been the subject of ER-positive cohorts. As expected today, for patients with ER- continuing investigation. The combination of tamoxifen and negative tumors, those who received CMF alone or followed by ovarian function suppression was better than either treatment goserelin had better disease-free survival than those who re- individually for premenopausal patients with advanced breast ceived goserelin alone. By contrast, for patients with ER- cancer (36). The combination of tamoxifen plus goserelin was positive tumors, the observed results for CMF alone and for better than goserelin alone in the adjuvant setting following six goserelin alone were equal, and the sequential use of CMF courses of CAF (cyclophosphamide, adriamycin, fluorouracil) followed by goserelin provided a statistically nonsignificant chemotherapy (37). Among the several studies (38 – 41) that benefit, primarily because of the results among younger women.
have compared adjuvant chemotherapy with endocrine therapies Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 Table 4. Sites of first treatment failure according to treatment*
% of total at a median follow-up of 7 years *CMF ϭ six courses of cyclophosphamide at 100 mg/m2 on days 1–14, orally; methotrexate at 40 mg/m2 on days 1 and 8, intravenously; and 5-fluorouracil at 600 mg/m2 on days 1 and 8, intravenously. Goserelin ϭ goserelin at 3.6 mg by subcutaneous implant monthly for 24 months. For the sequential combination therapy,six courses of CMF was followed by 18 monthly implants of goserelin. ER ϭ estrogen receptor.
that consisted of 5 years of tamoxifen and 2 or 3 years of with IBCSG Trial VIII, both trials were conducted almost ex- gonadotropin-releasing hormone (Gn-RH) agonist, the largest clusively in patients with lymph node–positive disease. The has been the Austrian Breast and Colorectal Cancer Study Group results of our study are similar to those of the Zoladex Early Breast (ABCSG) Trial 5 (41), which yielded better results with com- Cancer Research Association (ZEBRA) trial, which compared go- bination endocrine treatment than with chemotherapy alone.
serelin for 2 years with CMF (intravenous, days 1 and 8, for six Unfortunately, no trial has yet been conducted in the adjuvant courses) in 1640 patients and had more than 7 years of follow-up setting to compare tamoxifen plus ovarian function suppression (45). The second trial (46), a multicenter study conducted in with tamoxifen alone, either with or without chemotherapy.
France, compared adjuvant chemotherapy (anthracycline-based However, this question is now being addressed by the global for 77% of the patients) plus ovarian suppression (either ovarian Suppression of Ovarian Function Trial (SOFT; coordinated by irradiation or triptorelin for 3 years) with adjuvant chemotherapy the IBCSG on behalf of the Breast International Group and the alone in 926 patients. After 10 years of follow-up, the results North American Breast Cancer Intergroup). SOFT compares showed similar disease-free survival and overall survival for the tamoxifen alone versus ovarian function suppression (by either two treatment groups, which led the investigators to conclude the Gn-RH analog triptorelin or bilateral oophorectomy or ovar- that adjuvant chemotherapy and ovarian function suppression ian irradiation) plus tamoxifen versus ovarian function suppres- have a similar mechanism of antitumor activity (46). Because sion plus exemestane (a steroidal aromatase inhibitor) for pa- the mean age of the patients included in the French trial was 43 tients with steroid hormone receptor–positive tumors who years and results were not provided for the youngest cohort, it is remain premenopausal after adjuvant chemotherapy or for not possible to assess the consistency of these findings with whom tamoxifen alone is considered a reasonable treatment option (42,43). The complementary Tamoxifen and Exemestane IBCSG Trial IX found that CMF followed by tamoxifen was Trial (TEXT) compares the Gn-RH analog triptorelin plus ta- more effective than tamoxifen alone for postmenopausal patients moxifen versus triptorelin plus exemestane for patients who with lymph node–negative disease (47). In the present study, the receive the Gn-RH analog with or without chemotherapy from trial results confirm and extend the finding that chemotherapy is the start of their adjuvant therapy program (42,43). Thus, the more effective than endocrine therapy in terms of disease-free roles of ovarian function suppression and of an aromatase in- survival for patients with ER-negative tumors. The definition of hibitor are being prospectively studied in the adjuvant setting for ER-negative status on the basis of ligand-binding assay, how- ever, represents a mixture of definitely endocrine-nonresponsive tumors (i.e., an ER-absent cohort with no steroid hormone IBCSG Trial VIII was designed at a time when the Early receptor expression) and those with some modest responsiveness Breast Cancer Trialists’ Collaborative Group overview analyses to endocrine manipulations (i.e., an ER-low cohort with low indicated that tamoxifen was not likely to be effective for steroid hormone receptor expression). Thus, if the ER-absent women younger than 50 years (44) and, thus, tamoxifen was not cohort were to be considered separately, the differences in included in the trial. Despite the absence of tamoxifen in the outcome favoring the CMF-containing regimens compared with study design, IBCSG Trial VIII is valuable for exploring the goserelin alone may be greater than those observed. Further- relationships among hormone receptor expression, short- more, disease classified as ER-negative but progesterone recep- duration ovarian function suppression, endocrine effects of che- tor (PgR)-positive may also have some endocrine responsive- motherapy, amenorrhea, and age. Two other trials (45,46) of ness, because PgR status may be the dominant indicator for adjuvant ovarian function suppression in premenopausal patients endocrine responsiveness among premenopausal women (3). A started accrual around the same time as IBCSG Trial VIII and project is underway that will assess ER and PgR expression also did not contain tamoxifen as a treatment option. By contrast using quality-controlled immunohistochemical methodology in Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 a single laboratory for patients enrolled in IBCSG Trial VIII and mone receptor–positive tumors who receive ovarian function will assess outcomes separately for the ER- and PgR-absent and suppression from the start of their adjuvant therapy program ER- or PgR-positive cohorts (48). This investigation will also clarify the relationship between steroid hormone receptor ex- In addition to amenorrhea and ovarian function suppression, pression, response to treatment, and the degree of overexpres- cytotoxic chemotherapy may also have direct effects on sion of c-erbB-2. The association with c-erbB-2 expression is endocrine-responsive organs (52). Furthermore, the increased important because the addition of ovarian ablation to tamoxifen use of steroids as antiemetics and as supportive drugs for several therapy has been shown to be effective, compared with no old and new chemotherapy regimens (53) may provide addi- adjuvant treatment, for patients with tumors overexpressing tional antitumor effects for patients with endocrine-responsive tumors. However, disease-free survival for women younger than Very few of the premenopausal women enrolled in clinical age 35 years with ER-positive tumors treated with either che- trials that tested polychemotherapy (6) are young enough to motherapy alone or with tamoxifen alone is statistically signif- resist the effect of cytotoxic chemotherapy on ovarian endocrine icantly worse than that for older premenopausal women (54 – function, and thus allow the efficient testing of the role of further 56). Thus, it is important that alternative treatment approaches ovarian suppression. In fact, we observed a most intriguing such as ovarian function suppression with or without chemo- finding in our subgroup of women younger than 40 years.
therapy be studied in young patients. In addition, serum endo- Despite an earlier induction of ovarian function suppression with crine level profiles and novel technologies should be developed goserelin than with chemotherapy for this cohort of patients, it to investigate endocrine effects of treatments in tumor stroma was the women who received chemotherapy followed by gos- and adjacent tissue, because resistance to endocrine therapies erelin who had better disease-free survival. However, caution and to the endocrine effects of chemotherapy may be related to must be used when assessing the validity of results based on a mechanisms that involve additional components of the tumor retrospective subset analysis (50). Consequently, our observa- microenvironment and not just to events in the tumor cells.
tion should not alter current patient care, but rather emphasizes Endocrine therapies are important in the adjuvant treatment the relevance of current studies of chemotherapy and endocrine of young patients with endocrine-responsive early-stage breast cancer. Because the diagnosis of breast cancer in young women Premenopausal women with endocrine-responsive tumors, is rare, widespread collaboration will be important to the suc- especially those at low risk of recurrent disease, may not require cessful conduct of relevant clinical trials such as the ongoing chemotherapy provided they receive adequate endocrine ther- Breast International Group/North American Breast Cancer In- apy. To investigate this issue, the IBCSG conducted a random- tergroup SOFT, TEXT, and PERCHE studies (42,43).
ized clinical trial in premenopausal women with lymph node–positive disease who received combined endocrine therapy with APPENDIX
ovarian ablation (or suppression) and tamoxifen (51). In IBCSGTrial 11–93, four cycles of adjuvant chemotherapy (AC; doxo- International Breast Cancer Study Group—Trial VIII partic-
rubicin at 60 mg/m2 or epirubicin at 90 mg/m2 plus cyclophos- ipants and authors: Scientific Committee: A. Goldhirsch, A. S.
Coates (co-chairs). Foundation Council: J. Collins (president), B.
phamide at 600 mg/m2, every 21 days) and ovarian function Thu¨rlimann (vice president), H.-J. Senn (treasurer), S. Holmberg, J.
suppression (goserelin, bilateral oophorectomy, or ovarian irra- Lindtner, A. Veronesi, H. Cortés-Funes. Coordinating Center,
diation) and 5 years of tamoxifen (20 mg/day) was compared Bern, Switzerland: M. Castiglione-Gertsch (chief executive officer
with endocrine therapy (ovarian function suppression and ta- and study chair), M. L. Nasi (studies coordinator), C. Jenatsch, G.
moxifen) alone. The study was small, with only 174 patients Egli, M. Rabaglio, A. Saurer, R. Maibach, M. Iannino Gerber, A.
randomly assigned from May 1993 through November 1998.
Hiltbrunner. Statistical Center, Harvard School of Public Health
Ninety-five percent of the patients had one to three lymph nodes and Dana-Farber Cancer Institute, Boston, MA: R. Gelber (group
involved, and 53% of the patients had only one lymph node statistician), A. O’Neill, K. Price (scientific director), M. Bonetti, H.
involved. The median age was 45 years. After a median Peterson, D. Zahrieh, M. Zelen, S. Gelber, A. Keshaviah, S. Li. Data
follow-up of 4.4 years, the 4-year disease-free survival Ϯ stan- Management Center, Frontier Science and Technology Research
Foundation, Amherst, NY: R. Hinkle, M. Isley, L. Blacher (direc-
dard error was 87% Ϯ 4% for the group that received AC and tor), S. Lippert, J. Celano. Pathology Office: B. Gusterson, R.
88% Ϯ 4% for the endocrine therapy-alone group (RR for the Bettelheim, R. Reed, G. Viale, E. Mallon. Quality of Life Office: J.
addition of AC ϭ 1.22, 95% CI ϭ 0.53 to 2.81; P ϭ .63), Bernhard, C. Hu¨rny, H. Gusset, N. Mathys, B. Cliffe. The Ontario
suggesting that further study of the role of chemotherapy is Cancer Treatment and Research Foundation, Toronto Sunny-
warranted in this setting. Today, virtually all premenopausal brook Regional Cancer Centre, Toronto, Ontario, Canada: K.
women with lymph node–positive, steroid hormone receptor– Pritchard, D. Sutherland, C. Sawka, G. Taylor, R. Choo, C.
positive disease receive chemotherapy, despite the absence of Catzavelos, K. Roche. National Institute of Oncology, Budapest,
evidence showing that it is necessary for all such women.
Hungary: I. La´ng, E. Hitre, E. Juhos, I. Szamel. Centro di Riferi-
Endocrine therapy alone with ovarian function suppression and mento Oncologico, Aviano, Italy: D. Crivellari, S. Monfardini, E.
tamoxifen or an aromatase inhibitor may be sufficient to achieve Galligioni, M. D. Magri, A. Veronesi, A. Buonadonna, S. Massarut, excellent outcomes without chemotherapy, especially for pa- C. Rossi, E. Candiani, A. Carbone, R. Volpe, M. Roncadin, M.
Arcicasa, F. Coran, S. Morassut. Spedali Civili and Fondazione
tients at low risk of recurrent disease. This question is being Beretta, Brescia, Italy: E. Simoncini, G. Marini, P. Marpicati, M.
investigated in the Premenopausal Endocrine Responsive Che- Braga, P. Grigolato, L. Lucini. General Hospital, Gorizia, Italy: S.
motherapy (PERCHE) trial, which compares ovarian function Foladore, L. Foghin, G. Pamich, C. Bianchi, B. Marino, A. Murgia, suppression plus chemotherapy followed by tamoxifen or ex- V. Milan. European Institute of Oncology, Milan, Italy: A. Gold-
emestane versus ovarian function suppression and tamoxifen or hirsch, M. Colleoni, G. Martinelli, L. Orlando, F. Nolé, A. Luini, R.
exemestane without chemotherapy for patients with steroid hor- Orecchia, G. Viale, F. Peccatori, F. de Braud, A. Costa, S. Zurrida, Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 P. Veronesi, V. Sacchini, V. Galimberti, M. Intra, U. Veronesi.
Wales, Australia: M. Friedlander, B. Brigham, C. Lewis. Royal
Ospedale Infermi, Rimini, Italy: A. Ravaioli, D. Tassinari, G.
Adelaide Hospital, Adelaide, Australia: I. N. Olver, P. G. Gill, A.
Oliverio, F. Barbanti, P. Rinaldi, L. Gianni, G. Drudi. Ospedale S.
Taylor, D. Keefe. University of Sydney, Dubbo Base Hospital and
Eugenio, Rome, Italy: M. Antimi, M. Minelli, V. Bellini, R. Porzio,
Royal Prince Alfred Hospital, Sydney, Australia: J. Beith, M.
E. Pernazza, G. Santeusanio, L. G. Spagnoli. Ospedale S. Bortolo,
Boyer, A. S. Coates, R. J. Simes, A. Sullivan, M. H. N. Tattersall.
Vicenza, Italy: M. Magazu, V. Fosser, P. Morandi, G. Scalco, M.
W. P. Holman Clinic, Launceston, Australia: D. Boadle, I. Byard,
Balli, M. Gion, S. Meli, G. Torsello. The Institute of Oncology,
Ljubljana, Slovenia: J. Lindtner, D. Erzen, E. Majdic, B. Stabuc, A.
Plesnicar, R. Golouh, J. Lamovec, J. Jancar, I. Vrhoved, M. Kram-
REFERENCES
berger. Groote Schuur Hospital and University of Cape Town,
Cape Town, Republic of South Africa: D. M. Dent, A. Gudgeon,
(1) Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. Cancer E. Murray, I. D. Werner, P. Steynor, J. Toop, E. McEvoy. Sandton
statistics, 2003. CA Cancer J Clin 2003;53:5–26.
Oncology Center, Johannesburg, Republic of South Africa: D.
(2) Goldhirsch A, Glick JH, Gelber RD, Coates AS, Senn HJ. Meeting high- Vorobiof, M. Chasen, G. Fotheringham, G. de Muelenaere, B. Sku- lights: international consensus panel on the treatment of primary breast dowitz, C. Mohammed, A. Rosengarten. Madrid Breast Cancer
cancer. Seventh International Conference on Adjuvant Therapy of Primary Group, Madrid, Spain: H. Cortés-Funes, C. Mendiola, J. Hornedo,
Breast Cancer. J Clin Oncol 2001;19:3817–27.
R. Colomer, F. Cruz Vigo, P. Miranda, A. Sierra, F. Martinez-Tello, (3) Colleoni M, Gelber S, Coates AS, Castiglione-Gertsch M, Gelber RD, A. Garzon, S. Alonso, A. Ferrero. West Swedish Breast Cancer
Price K, et al., for the International Breast Cancer Study Group. Influence Study Group, Göteborg, Sweden: C. M. Rudenstam, A. Wallgren,
of endocrine-related factors on response to perioperative chemotherapy for S. Ottosson-Lönn, R. Hultborn, G. Colldahl-Ja¨derström, E. Cahlin, patients with node-negative breast cancer. J Clin Oncol 2001;19:4141–9.
J. Mattsson, S. B. Holmberg, L. Ivarsson, O. Ruusvik, L. G. Niklas- (4) Early Breast Cancer Trialists’ Collaborative Group. Ovarian ablation in son, S. Dahlin, G. Karlsson, B. Lindberg, A. Sundba¨ck, S.
early breast cancer: overview of the randomised trials. Lancet 1996;348: Bergegårdh, H. Salander, C. Andersson, M. Heideman, Y. Hessman, O. Nelzén, G. Claes, T. Ramhult, J. H. Svensson, P. Liedberg, M.
(5) Early Breast Cancer Trialists’ Collaborative Group. Tamoxifen for early Suurku¨la, S. Persson. Swiss Group for Clinical Cancer Research
breast cancer: an overview of the randomised trials. Lancet 1998;351:1451– 67.
(SAKK) member institutions—Inselspital, Bern, Switzerland:
(6) Early Breast Cancer Trialists’ Collaborative Group. Polychemotherapy for M. F. Fey, M. Castiglione-Gertsch, E. Dreher, H. Schneider, S. Aebi, early breast cancer: an overview of the randomised trials. Lancet 1998; J. Ludin, G. Beck, A. Haenel, J. M. Lu¨thi, H. J. Altermatt, M.
Nandedkar, K. Buser. Kantonsspital, St. Gallen, Switzerland: H. J.
(7) Pagani O, O’Neill A, Castiglione-Gertsch M, Gelber RD, Goldhirsch A, Senn, B. Thu¨rlimann, Ch. Oehlschlegel, G. Ries, M. Töpfer, U.
Rudenstam CM, et al. Prognostic impact of amenorrhoea after adjuvant Lorenz, O. Schiltknecht, B. Spa¨ti, A. Ehrsam, M. Bamert, W. F.
chemotherapy in premenopausal breast cancer patients with axillary node Jungi. Istituto Oncologico della Svizzera Italiana, Bellinzona,
involvement: results of the International Breast Cancer Study Group Switzerland: F. Cavalli, O. Pagani, H. Neuenschwander, L. Bronz,
(IBCSG) trial VI. Eur J Cancer 1998;34:632– 40.
C. Sessa, M. Ghielmini, T. Rusca, P. Rey, J. Bernier, E. Pedrinis, T.
(8) Goldhirsch A, Gelber RD, Castiglione M. The International Breast Cancer Gyr, L. Leidi, G. Pastorelli, G. Caccia, A. Goldhirsch. Kantonsspi-
Study Group. The magnitude of endocrine effects of adjuvant chemother- tal, Basel, Switzerland: R. Herrmann, C. F. Rochlitz, J. F. Harder,
apy for premenopausal breast cancer patients. Ann Oncol 1990;1:183– 8.
S. Bartens, U. Eppenberger, J. Torhorst. Hôpital des Cadolles,
(9) Brincker H, Rose C, Rank F, Mouridsen HT, Jakobsen A, Dombernowsky Neuchaˆtel, Switzerland: D. Piguet, P. Siegenthaler, V. Barrelet,
P, et al. Evidence of a castration-mediated effect of adjuvant cytotoxic R. P. Baumann. University Hospital, Zu¨rich, Switzerland: B.
Pestalozzi, C. Sauter, D. Fink, M. Fehr, U. Haller, U. Metzger, P.
(10) Padmanabhan N, Howell A, Rubens RD. Mechanism of action of adjuvant Huguenin, R. Caduff. Centre Hospitalier Universitaire Vandois,
chemotherapy in early breast cancer. Lancet 1986;2:411– 4.
Lausanne, Switzerland: L. Perey, S. Leyvraz, P. Anani, F. Gomez,
(11) Minton SE, Munster PN. Chemotherapy-induced amenorrhea and fertility D. Wellman, G. Chapuis, P. De Grandi, P. Reymond, M. Gillet, J. F.
in women undergoing adjuvant treatment for breast cancer. Cancer Control Delaloye. Hôpital Cantonal, Geneva, Switzerland: P. Alberto, H.
Bonnefoi, P. Scha¨fer, F. Krauer, M. Forni, S. Diebold. Kantonsspi-
(12) Parulekar W, Trudeau ME, Shepherd L, Ottaway J, Day A, Franssen E, et tal Graubu¨nden, Chur, Switzerland: F. Egli, P. Forrer, A. Willi,
al. Incidence and prognostic impact of amenorrhea during adjuvant therapy R. Steiner, J. Allemann, T. Ru¨edi, A. Leutenegger, U. Dalla Torre.
in high risk premenopausal breast cancer patients: analysis of a National Australian New Zealand Breast Cancer Trials Group (ANZ
Cancer Institute of Canada Clinical Trials Group (NCIC CTG) phase III BCTG) member institutions—Operations Office, University of
study [abstract 97]. Proc ASCO 2001;20:25A.
Newcastle, Newcastle, Australia: J. F. Forbes, D. Lindsay, A.
(13) Bonadonna G, Rossi A, Valagussa P, Banfi A, Veronesi U. The CMF Wilson. Statistical Center, National Health and Medical Re-
program for operable breast cancer with positive axillary nodes: updated search Council Clinical Trials Centre, University of Sydney,
analysis on the disease-free interval, site of relapse and drug tolerance.
Camperdown, Australia: R. J. Simes, H. Dhillon. The Cancer
Council Victoria (previously Anti-Cancer Council of Victoria),
(14) Fisher B, Sherman B, Rockette H, Redmond C, Margolese R, Fisher ER.
Clinical Trials Office, Melbourne, Australia: J. Collins, R. Sny-
1-Phenylalanine mustard (L-PAM) in the management of premenopausal der, E. Abdi, A. Barling, R. Basser, W. I. Burns, M. Chipman, patients with primary breast cancer: lack of association of disease-freesurvival with depression of ovarian function. National Surgical Adjuvant J. Chirgwin, I. Davis, R. Holmes, M. Green, D. Hastrich, P. Gregory, Project for Breast and Bowel Cancers. Cancer 1979;44:847–57.
R. McLennan, L. Mileshkin, P. Mitchell, G. Richardson, M.
(15) Fisher B, Costantino J, Redmond E, Poisson R, Bowman D, Couture J, et Schwarz, I. Russell, C. Underhill, D. Reading, A. Zimet. Auckland
al. A randomized clinical trial evaluating tamoxifen in the treatment of Breast Cancer Study Group, Auckland, New Zealand: V. J.
patients with node-negative breast cancer who have estrogen-receptor Harvey, P. Thompson, D. Porter. Flinders Medical Centre, Bed-
positive tumors. N Engl J Med 1989;320:479 – 84.
ford Park, South Australia: T. Malden. Newcastle Mater Miseri-
(16) Fisher B, Redmond C, Dimitrov NV, Bowman D, Legault-Poisson S, cordiae Hospital Waratah, Newcastle, Australia: J. F. Forbes, J.
Wickerham DL, et al. A randomized clinical trial evaluating sequential Stewart, D. Jackson, R. Gourlay, J. Bishop, S. Cox, S. Ackland, A.
methotrexate and fluorouracil in the treatment of patients with node- Bonaventura, C. Hamilton, J. Denham, P. O’Brien, M. Back, S.
negative breast cancer who have estrogen-receptor-negative tumors.
Brae, R. Muragasu. Prince of Wales, Randwick, New South
Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 (17) Mansour EG, Gray R, Shatila AH, Osborne CK, Tormey DC, Gilchrist (37) Davidson N, O’Neill A, Vukov A, Osborne CK, Martino S, White D, et al.
KW, et al. Efficacy of adjuvant chemotherapy in high-risk node-negative Effect of chemohormonal therapy in premenopausal node (ϩ) receptor (ϩ) breast cancer. An intergroup study. N Engl J Med 1989;320:485–90.
breast cancer: an Eastern Cooperative Oncology Group phase III intergroup (18) Ludwig Breast Cancer Study Group. Prolonged disease-free survival after trial (E5188, INT-0101) [abstract 249]. Proc ASCO 1999;18:67a.
one course of perioperative adjuvant chemotherapy for node-negative (38) Roché H, Mihura J, de Lafontan B, Reme-Saumon M, Martel P, Dubois J, breast cancer. N Engl J Med 1989;320:491– 6.
et al. Castration and tamoxifen versus chemotherapy (FAC) for premeno- (19) Castiglione-Gertsch M, Gelber RD, O’Neill A, Coates AS, Goldhirsch A. The pausal, node and receptors positive breast cancer patients: a randomized International Breast Cancer Study Group. Systemic adjuvant treatment for trial with a 7 years median follow up [abstract 134]. Proc ASCO 1996;15:117.
premenopausal node-negative breast cancer. Eur J Cancer 2000;36:549 –50.
(39) Roché HH, Kerbrat P, Bonneterre J, Fargeot P, Fumoleau P, Monnier A, et (20) Hermanek P, Sobin LH, editors. TNM classification of malignant tumors.
al. Complete hormonal blockade versus chemotherapy in premenopausal 4th ed. Berlin (Germany): Springer-Verlag; 1987.
early-stage breast cancer patients (pts) with positive hormone-receptor (21) Davis BW, Zava DT, Locher GW, Goldhirsch A, Hartmann WH. Receptor (HRϩ) and 1–3 node-positive (Nϩ) tumors: results of the FASG 06 trial heterogeneity of human breast cancer as measured by multiple intratumoral [abstract 279]. Proc ASCO 2000;19:72a.
assays of estrogen and progesterone receptor. Eur J Cancer Clin Oncol (40) Boccardo F, Rubagotti A, Amoroso D, Mesiti M, Romeo D, Sismondi P, et al. Cyclophosphamide, methotrexate, and fluorouracil versus tamoxifen (22) Berger U, Wilson P, McClelland RA, Davidson J, Coombes RC. Correla- plus ovarian suppression as adjuvant treatment of estrogen receptor- tion of immunocytochemically demonstrated estrogen receptor distribution positive pre-/perimenopausal breast cancer patients: results of the Italian and histopathologic features in primary breast cancer. Hum Pathol 1987; Breast Cancer Adjuvant Study Group 02 randomized trial. J Clin Oncol (23) Wallgren A, Bernier J, Gelber RD, Goldhirsch A, Roncadin M, Joseph D, (41) Jakesz R, Hausmaninger H, Kubista E, Gnant M, Menzel C, Bauernhofer et al., for the International Breast Cancer Study Group. Timing of radio- T, et al. Randomized adjuvant trial of tamoxifen and goserelin versus therapy and chemotherapy following breast-conserving surgery for patients cyclophosphamide, methotrexate, and fluorouracil: evidence for the supe- with node-positive breast cancer. Int J Radiat Oncol Biol Phys 1996;35: riority of treatment with endocrine blockade in premenopausal patients with hormone-responsive breast cancer–Austrian Breast and Colorectal (24) Crivellari D, Bonetti M, Castiglione-Gertsch M, Gelber RD, Rudenstam Cancer Study Group Trial 5. J Clin Oncol 2002;20:4621–7.
CM, Thu¨rlimann B, et al., for the International Breast Cancer Study Group.
(42) Gelber RD, Castiglione-Gertsch M, Coates AS, Goldhirsch A, for the Burdens and benefits of adjuvant cyclophosphamide, methotrexate, and International Breast Cancer Study Group (IBCSG); Breast International fluorouracil and tamoxifen for elderly patients with breast cancer: the Group (BIG); North American Breast Intergroup. Tailored treatment in- International Breast Cancer Study Group Trial VII. J Clin Oncol 2000;18: vestigations for premenopausal women with endocrine responsive (ERϩ and/or PgRϩ) breast cancer: the open questions [abstract P103]. The Breast (25) Kaplan EL, Meier P. Nonparametric estimation from incomplete observa- tions. J Am Stat Assoc 1958;53:457– 81.
(43) Francis P, Fleming G, Nasi ML, Pagani O, Perez E, Walley B, for the (26) Greenwood M. The natural duration of cancer. London (U.K.): Her Maj- International Breast Cancer Study Group (IBCSG); Breast International esty’s Stationary Office; 1926. p. 1–26.
Group (BIG); North American Breast Intergroup. Tailored treatment in- (27) Mantel N. Evaluation of survival data and two new rank order statistics vestigations for premenopausal women with endocrine responsive (ERϩ arising in its consideration. Cancer Chemother Rep 1966;50:163–70.
and/or PgRϩ) breast cancer: the SOFT, TEXT, and PERCHE Trials (28) Cox DR. Regression models and life-tables (with discussion). J Royal Stat [abstract P104]. The Breast 2003;12 (Suppl 1):S44.
(44) Early Breast Cancer Trialists’ Collaborative Group. Systemic treatment of (29) Bonetti M, Gelber RD. A graphical method to assess treatment-covariate early breast cancer by hormonal, cytotoxic, or immune therapy. 133 ran- interactions using the Cox model on subsets of the data. Stat Med 2000; domised trials involving 31,000 recurrence and 24,000 deaths among 75,000 women. Lancet 1992;339:1–15, 71– 85.
(30) Bonetti M, Gelber RD, Goldhirsch A, Castiglione-Gertsch M, Coates AS, (45) Kaufmann M, Jonat W, Blamey R, Cuzick J, Namer M, Fogelman I, et al.
for The International Breast Cancer Study Group (IBCSG). Features that Survival analyses from the ZEBRA study: goserelin (Zoladex) versus CMF predict responsiveness to chemotherapy and endocrine therapies. The in premenopausal women with node-positive breast cancer. Eur J Cancer (31) Davis BW, Gelber RD, Goldhirsch A, Hartmann WH, Locher GW, Reed R, et al. Prognostic significance of tumor grade in clinical trials of adjuvant (46) Arriagada R, Le MG, Spielmann M, Mauriac L, Bonneterre J, Namer M, et therapy for breast cancer with axillary lymph node metastasis. Cancer al. Randomized trial of adjuvant ovarian suppression in 926 premenopausal patients with early breast cancer treated with adjuvant chemotherapy [ab- (32) Meakin JW, Hayward JL, Panzarella T, Allt WE, Beale FA, Bulbrook RD, et al. Ovarian irradiation and prednisone following surgery and radiother- (47) International Breast Cancer Study Group. Endocrine responsiveness and apy for carcinoma of the breast. Breast Cancer Res Treat 1996;37:11–9.
tailoring adjuvant therapy for postmenopausal lymph node-negative breast (33) Cole MP. Prophylactic compared with therapeutic x-ray artificial meno- cancer: a randomized trial. J Natl Cancer Inst 2002;94:1054 – 65.
pause. In: Joslin CA, Gleave EN, editors. The clinical management of (48) Goldhirsch A, Wood WC, Gelber RD, Coates AS, Thu¨rlimann B, Senn HJ.
advanced breast cancer. 2nd Tenovus workshop. Cardiff, Wales (U.K.): Meeting highlights: updated international expert consensus on the primary Alpha Omega Alpha Publishing; 1970. p. 2–11.
therapy of early breast cancer. J Clin Oncol 2003;21:3357– 65.
(34) Ravdin RG, Lewison EF, Slack NH, Dao TL, Gardner B, State D, et al.
(49) Love RR, Duc NB, Havighurst TC, Mohsin SK, Zhang Q, DeMets DL, et Results of a clinical trial concerning the worth of prophylactic oophorec- al. HER-2/neu overexpression and response to oophorectomy plus tamox- tomy for breast carcinoma. Surg Gynecol Obstet 1970;131:1055– 64.
ifen adjuvant therapy in estrogen receptor-positive premenopausal women (35) Basser R, O’Neill A, Martinelli G, Nasi ML, Peccatori F, Cinieri S, et al.
with operable breast cancer. J Clin Oncol 2003;21:453–7.
Randomized trial comparing up-front, multi-cycle, dose-intensive chemo- (50) Coates AS, Goldhirsch A, Gelber RD. Overhauling the breast cancer therapy (CT) versus standard dose CT in women with high-risk stage 2 or overview: are subsets subversive? Lancet Oncol 2002;3:525– 6.
3 breast cancer (BC): first results from IBCSG Trial 15–95 [abstract 20].
(51) International Breast Cancer Study Group. Randomized controlled trial of ovarian function suppression plus tamoxifen versus the same endocrine (36) Klijn JG, Blamey RW, Boccardo F, Tominaga T, Duchateau L, Sylvester therapy plus chemotherapy: is chemotherapy necessary for premenopausal R, et al., for the Combined Hormone Agents Trialists Group and the women with node-positive, endocrine responsive breast cancer? First re- European Organization for Research and Treatment of Cancer. Combined sults of International Breast Cancer Study Group Trial 11–93. The Breast tamoxifen and luteinizing hormone-releasing hormone (LHRH) agonist versus LHRH agonist alone in premenopausal advanced breast cancer: a (52) Morgan MW, O’Hare MJ. Cytotoxic drugs and the human adrenal cortex: meta-analysis of four randomized trials. J Clin Oncol 2001;19:343–53.
a cell culture study. Cancer 1979;43:969 –79.
Journal of the National Cancer Institute, Vol. 95, No. 24, December 17, 2003 (53) Nabholtz JM, Pienkowski T, Mackey J, Pawlicki M, Guastalla JP, Vogel C, tion, data management, and statistics was provided for the International et al. Phase III trial comparing TAC (docetaxel, doxorubicin, cyclophos- Breast Cancer Study Group by the Swedish Cancer Society, The Cancer phamide) with FAC (5-fluorouracil, doxorubicin, cyclophosphamide) in the Council Australia, the Australian New Zealand Breast Cancer Trials Group adjuvant treatment of node positive breast cancer (BC) patients: interim (NHMRC grants 890028, 920876, 950328, 980379, 141711), the Frontier analysis of the BCIRG 001 study [abstract 141]. Proc ASCO 2002;21:36A.
Science and Technology Research Foundation, the Swiss Group for Clinical (54) Aebi S, Gelber S, Castiglione-Gertsch M, Gelber RD, Collins J, Thu¨rli- Cancer Research (SAKK), the Swiss Cancer League, the American–Italian mann B, et al., for the International Breast Cancer Study Group. Is che- Cancer Foundation (AICF grants 101-98 and 101-99), the American Cancer motherapy alone adequate for young women with oestrogen-receptor- Society (grant RPG-90-013-08-PBP), and Public Health Service grant CA- positive breast cancer? Lancet 2000;355:1869 –74.
75362 from the National Cancer Institute, National Institutes of Health, (55) Goldhirsch A, Gelber RD, Yothers G, Gray RJ, Green S, Bryant J, et al.
Department of Health and Human Services. We also acknowledge support for Adjuvant therapy for very young women with breast cancer: need for the Cape Town participants from the Cancer Association of South Africa, for tailored treatments. J Natl Cancer Inst Monogr 2001;(30):44 –51.
the St. Gallen participants from the Foundation for Clinical Research of (56) Aebi S, Castiglione-Gertsch M. Adjuvant endocrine therapy for the very Eastern Switzerland, and for the Göteborg participants from the Swedish young patients. Corrected proof. The Breast (DOI:10.1016/S0960-9776 Society for Cancer Research (Cancerfonden). Astra-Zeneca provided the We thank the patients, physicians, nurses, and data managers who participate in the International Breast Cancer Study Group trials. We thank Rita Hinkle fordata management.
Initial support was provided by the Ludwig Institute for Cancer Research Manuscript received June 24, 2003; revised October 1, 2003; accepted and the Cancer League of Ticino. Continuing support for central coordina-

Source: http://didattica.unibocconi.it/mypage/upload/50573_20110618_010043_IBCS-2003.PDF