Doi:10.1016/s0140-6736(05)67059-6

Articles
Switching of postmenopausal women with endocrine-
responsive early breast cancer to anastrozole after 2 years’
adjuvant tamoxifen: combined results of ABCSG trial 8 and
ARNO 95 trial
Raimund Jakesz, Walter Jonat, Michael Gnant, Martina Mittlboeck, Richard Greil, Christoph Tausch, Joern Hilfrich, Werner Kwasny,Christian Menzel, Hellmut Samonigg, Michael Seifert, Guenther Gademann, Manfred Kaufmann, on behalf of the ABCSG and the GABG* Summary
Background Tamoxifen has been the standard adjuvant treatment for postmenopausal women with hormone-
Lancet 2005; 366: 455–62
responsive early breast cancer for more than 20 years. However, the third-generation aromatase inhibitor *Study groups listed at end of
anastrozole has proven efficacy and tolerability benefits compared with tamoxifen when used as initial adjuvant
therapy. We investigate whether women who have received a period of adjuvant tamoxifen would benefit from being See Comment page 431
switched to anastrozole.
Austrian Breast and Colorectal
Cancer Study Group (ABCSG),
Vienna Medical University,
Methods We present a combined analysis of data from two prospective, multicentre, randomised, open-label trials Vienna General Hospital,
with nearly identical inclusion criteria. Postmenopausal women with hormone-sensitive early breast cancer who had
Waehringer Guertel 18-20,
completed 2 years’ adjuvant oral tamoxifen (20 or 30 mg daily) were randomised to receive 1 mg oral anastrozole Vienna A-1090, Austria
(n=1618) or 20 or 30 mg tamoxifen (n=1606) daily for the remainder of their adjuvant therapy. The primary endpoint
was event-free survival, with an event defined as local or distant metastasis, or contralateral breast cancer. Analysis Prof M Mittlboeck PhD,
was by intention to treat.
Prof R Greil MD, C Tausch MD,W Kwasny MD,Prof C Menzel MD, Findings 3224 patients were included in analyses. At a median follow-up of 28 months, we noted a 40% decrease in
the risk for an event in the anastrozole group as compared with the tamoxifen group (67 events with anastrozole vs Prof M Seifert MD); and German
110 with tamoxifen, hazard ratio 0·60, 95% CI 0·44–0·81, p=0·0009). Both study treatments were well tolerated.
Adjuvant Breast Cancer Group
(GABG), University of
There were significantly more fractures (p=0·015) and significantly fewer thromboses (p=0·034) in patients treated
Frankfurt, Frankfurt am Main,
with anastrozole than in those on tamoxifen.
Germany (Prof W Jonat MD,
Prof J Hilfrich MD,
Interpretation These data lend support to a switch from tamoxifen to anastrozole in patients who have completed
2 years’ adjuvant tamoxifen.
Introduction
tamoxifen therapy, therefore, seems to be imposed by Prof Raimund Jakesz Breast cancer is the most common female cancer. It is the limitations of the drug rather than by the optimum raimund.jakesz@meduniwien.
diagnosed in more than a million women worldwide and duration of therapy. In particular, the relapse pattern accounts for more than 400 000 deaths yearly. More than for low-risk and intermediate-risk tumours indicates 110 cases per 100 000 of the population are diagnosed in that adjuvant treatment should continue after 5 years, Germany and Austria every year.1 The incidence of with overview results suggesting that there is a 1·5–2% breast cancer increases with age, and about three- yearly risk of recurrence of breast cancer in years 5–15 quarters of the women affected are postmenopausal. In these individuals, about 80% of tumours are hormone- The 15-year outcome of some oestrogen-receptor positive tumours might be worse than that of oestrogen- For more than 20 years, the anti-oestrogen tamoxifen receptor negative lesions.12 The administration of has been the established endocrine adjuvant therapy tamoxifen beyond the optimum time of efficacy might, after surgery for postmenopausal women with early therefore, result in side-effects without a concomitant breast cancer. 5 years is generally judged the optimum duration for treatment,3 since tamoxifen therapy The limitations of tamoxifen have led to a search for beyond 5 years seems to confer no extra benefit in alternative endocrine therapies with increased efficacy terms of disease-free survival.4,5 However, several side- and fewer long-term complications. The third- effects are inherent with long-term tamoxifen generation aromatase inhibitors anastrozole, letrozole, treatment. The partial oestrogenic activity of tamoxifen and exemestane are highly selective for aromatase and in some tissues leads to an increased risk of inhibit 97–99% of oestrogen synthesis from this endometrial cancer and thromboembolic events over source.13,14 Results of trials such as the ATAC study15 have the course of treatment.6–8 Tamoxifen resistance can shown the improved efficacy and tolerability of also develop.9 The 5-year standard for adjuvant anastrozole over tamoxifen, and data now support the www.thelancet.com Vol 366 August 6, 2005
Articles
use of 5 years’ anastrozole as adjuvant therapy for Eligible patients were postmenopausal women aged postmenopausal women with early breast cancer.
80 years or younger (ABCSG trial 8) or 75 years or However, tamoxifen is still a useful and ubiquitous younger (ARNO 95) with histologically verified, locally treatment option, and by employing a strategy of radically treated invasive or minimally invasive breast switching therapy from tamoxifen to an aromatase cancer without previous chemotherapy, hormone inhibitor, the unnecessary longer-term side-effects of therapy, or radiotherapy. Postmenopausal status was tamoxifen might be obviated and the complications of assumed for patients whose last menstruation took place long-term tamoxifen therapy avoided. Data indicate a at least 12 months before study entry, for those who had positive effect on recurrence-free survival when undergone bilateral ovariectomy, or for whom follicle- switching from tamoxifen to an aromatase inhibitor.16,17 stimulating hormone and luteinising hormone The most recent technical assessment from the concentrations indicated postmenopausal status. All American Society of Clinical Oncology (ASCO)18 patients had endocrine-responsive tumours—ie, with recommends that optimum adjuvant therapy for concentrations of oestrogen receptors or progesterone postmenopausal women should now include the use of receptors of more than 10 fmol/mg cytosol protein, or an aromatase inhibitor, either as initial treatment or were oestrogen-receptor or progesterone-receptor after 2–5 years’ treatment with tamoxifen, to reduce the positive as assessed histochemically. ABCSG trial 8 risk of tumour recurrence. The aim of the Austrian included patients with G1 and G2 ductal carcinoma and Breast and Colorectal Cancer Study Group (ABCSG) Gx lobular tumours, whereas patients with ductal trial 8/Arimidex-Nolvadex (ARNO) 95 combined carcinoma of any grade were recruited to ARNO 95.
analysis was to assess whether switching to anastrozole Tumours were graded according to the Bloom and after 2 years’ tamoxifen treatment is more effective than the standard 5 years’ adjuvant tamoxifen therapy. Inclusion criteria common to both trials were absence of preoperative chemotherapy, hormone therapy, or radiotherapy, tumour infiltration of up to ten (ABCSG Patients
trial 8) or nine (ARNO 95) lymph nodes, and absence of This study is a prospectively-planned, event-driven organ metastases. Exclusion criteria across both trials combined analysis of two trials—ABCSG trial 8 and the were indeterminate menopausal status (or menopausal ARNO 95 trial by the German Adjuvant Breast Cancer status maintained by medication), presence of secondary Group (GABG)—both of which were prospective, malignant disease, tumour infiltration of skin or breast multicentre, randomised, open-label studies and had muscle (T4 tumours), and presence of other broadly similar inclusion criteria and outcome measures.
concomitant serious medical conditions—eg, thoseinvolving bone marrow function, the central nervoussystem, uncompensated cardiac insufficiency, or 3901 patients randomised in ABCSG trial 8 uncontrolled local or systemic infection. Although hormone replacement therapy was not excluded in the procotol, it was considered as explicitly contraindicated in both countries for patients receiving adjuvant breastcancer treatment.
mastectomy or breast-conserving surgery with axiliary lymph-node dissection or sentinel lymph-node biopsy (with or without subsequent radiotherapy), followed by 275 treatment discontinuations for other adjuvant tamoxifen therapy started within 6 weeks (ABCSG trial 8) or 4 weeks (ARNO 95) of surgery or 10 last follow-up date within 2 years of For both studies, patients had to complete 2 years’ adjuvant oral tamoxifen therapy in accordance with local guidelines (20 mg daily for ABCSG trial 8, and 20 mg or30 mg daily for ARNO 95; patients administered 30 mgcontinued on that dose unless otherwise indicated).
Women were randomised before beginning treatmentwith tamoxifen in ABCSG trial 8 and within 2 years oftamoxifen treatment in ARNO 95. The accrual period for this combined analysis was January, 1996, to August, 2003. Since the randomisation processes in Austria andGermany differed, the timepoint of 2 years post-surgery Figure 1: Trial profile
was used as a starting point for this analysis.
www.thelancet.com Vol 366 August 6, 2005
Articles
All patients provided written informed consent and during randomisation. For the common analysis, we both studies were done in accordance with the calculated that 278 events would be required for the Declaration of Helsinki. ABCSG trial 8 and ARNO 95 final analysis to detect a hazard ratio (HR) of 0·7 for were approved by the relevant ethics committees in event-free survival between the treatment groups with a power of 80% and a two-sided significance level of0·05. Interim analyses were planned on reaching 139 Procedures
and 209 events, using a significance level of 0·001 Randomisation for ABCSG trial 8 was done centrally at (stopping boundary) to maintain a significance level of the ABCSG randomisation centre, Vienna, Austria.
0·05 for the final analysis. The number of events Randomisation for ARNO 95 was done at the needed to trigger the first interim analysis was reached Department of Medical Biometry and Statistics, in April, 2004 (143 observed events). The Steering Freiburg, Germany. The computer-assisted randomi- Committee decided to reassess all data for accuracy. As sation schedules for ABCSG trial 8 were based on the stopping boundary for event-free survival was minimisation as a dynamic algorithm designed to reached at this analysis, the independent data counteract imbalance between treatments, taking monitoring committee decided, in November, 2004, to stratification factors into account. For ARNO 95, these recommend close of recruitment and to publish the schedules were based on block randomisation. Patients were randomised to either continue tamoxifen or to Analyses were by intention to treat. Data are presented switch to oral anastrozole (1 mg per day) for 3 years after in Kaplan-Meier curves,21 and tested by log-rank tests.22,23 completion of 2 years’ adjuvant therapy. In Austria, this HRs and their corresponding 95% CIs were estimated randomisation was done within 6 weeks after surgery, by the proportional-hazards regression model of Cox.24 whereas in Germany it was done at any time between Main analyses were based on the first corresponding surgery and 6 weeks after completion of the first 2 years’ event per patient. In additional sensitivity analyses, first events only were considered. Thus, for analysis of Patients received a physical examination and were distant recurrence-free survival, the first observed monitored for safety and tolerability. Case report forms distant metastasis was included in the main analysis. In were provided for documentation and adhered to as per the sensitivity analysis, however, patients who did not protocol. In Austria, the monitoring took place at have a distant metastasis as first cancer-related event— 3-monthly intervals throughout the first year of eg, secondary cancer, locoregional event, or contralateral randomised therapy, at 6-monthly intervals in the event—were censored at the first observed event.
second and third year, and yearly thereafter. In All analyses were done with SAS (version 8.02).
Germany, assessment of patients was done at 6-monthly Adverse events were only counted once per patient, and intervals. There were no observation-free intervals.
Gynaecological examinations, thoracic X-rays, skeletal Tamoxifen (n=1606)
Anastrozole (n=1618)
ultrasound, and standard mammography were done as appropriate (at least yearly) to identify the presence of disease recurrence (locoregional, contralateral, or distant metastatic tumour [lymph node or organ]). The assessments done at each visit were common to, and the number of patients with data was similar in, both treatment groups. Events were confirmed histologically, cytologically, or, where not clinically obvious, by the various radiological screening methods used at the The primary endpoint was event-free survival, defined as time to relapse at any site or incidence of contralateral breast cancer. Distant recurrence-free survival and tolerability issues were also compared.
Statistical analysis
In Austria, patients were allocated to the treatment groups according to the method of Pocock and Simon,20 stratifying for the following prognostic factors: age,tumour grade, tumour stage, nodal status, and Data are number (%) unless otherwise indicated. ER=oestrogen receptor; PgR=progesterone receptor.
participating centres grouped into federal states. In Table 1: Baseline characteristics
Germany, only the participating centre was considered www.thelancet.com Vol 366 August 6, 2005
Articles
Role of the funding source
The study designs were developed by the ABCSG and the GABG. The management of the trial has beenundertaken by the ABCSG and GABG with funding and organisational support from the trial sponsors:AstraZeneca in Austria and the GABG in Germany. The ABCSG statistician analysed all data. AstraZeneca funded editorial assistance in the form of technical preparation of references, figures, tables, technicalediting for English language, formatting of the report toLancet style, and administrative support. AstraZeneca had no role in data interpretation. The corresponding author had full access to all the data in the study and hadfinal responsibility for the decision to submit for Numbers at risk (estimated proportion of events)
Results
Figure 1 shows the trial profile. 3224 patients (2262 in
Figure 2: Kaplan-Meier curves of event-free survival
ABCSG trial 8 and 962 in ARNO 95) were randomised 0 timepoint=2 years after surgery. SD at 3 years: tamoxifen=0·81,anastrozole=0·65 to either continue tamoxifen (n=1606) or switch toanastrozole (n=1618). Median follow-up was 28 months are described with absolute frequencies and proportions.
(95% CI 26–30) after initial treatment with tamoxifen.
Differences in the adverse event rates were estimated The treatment groups were well balanced in terms of with exact odds ratios (OR) and corresponding 95% CIs.
age, nodal status, tumour stage, tumour grade, Exact ORs stratified by country were calculated for the oestrogen-receptor and progesterone-receptor status, five types of serious adverse events available for Austrian and previous surgery (table 1). At the time of disclosure and German patients (myocardial infarct, embolism, of trial data, 882 (55%) patients assigned anastrozole thromboses, fractures, and endometrial cancer). The and 884 (55%) assigned tamoxifen had completed exact calculations were done with StatXact (version 6).
All p values are two-sided, and a p less than 0·05 was Event-free survival was higher in patients who took anastrozole than in those who continued treatment with HR (95% CI)
Receptor (ER/PgR) Positive/positive (n=2519) Figure 3: HR (95%CI) for anastrozole versus tamoxifen stratified by nodal status, tumour grade, age, and hormone receptor status
ER=oestrogen receptor; PgR=progesterone receptor.
www.thelancet.com Vol 366 August 6, 2005
Articles
Tamoxifen (n=1597)
Anastrozole (n=1602)
OR (95% CI), p
Data are number (%) unless otherwise indicated.
Table 3: Serious adverse events by treatment group
In women in whom disease progressed, distant metastases accounted for 62% (n=110) of recurrences (figure 4). Metastases arose in 3% of anastrozole-treated Numbers at risk (estimated proportion of distant recurrences)
patients and in 5% of patients treated only with tamoxifen (HR 0·61, 0·42–0·87, p=0·0067), indicating a 39% decrease in risk of metastases for womenswitching to anastrozole. When looking at distant Figure 4: Kaplan-Meier curves of distant recurrence-free survival
metastases as first events only, the univariate model 0 timepoint=2 years after surgery. SD at 3 years: tamoxifen=0·76,anastrozole=0·52.
gives an HR of 0·54 (0·37–0·80, p=0·0016).
Contralateral or ipsilateral recurrence accounted for tamoxifen (figure 2). 67 events were noted in the only 16% (n=28) and 23% (n=41) of recurrences, anastrozole group and 110 events in the tamoxifen respectively. More recurrences were observed in the group. In the combined analysis, there was an HR of tamoxifen group than in the anastrozole group (table 2).
0·60 (95% CI 0·44–0·81, p=0·0009) in favour of 59 individuals in the tamoxifen group and 45 in the anastrozole at 3 years post-switch for the occurrence of an event. With respect to first events only, the HR was Overall survival at 3 years post-switch was slightly 0·59 (0·44–0·81, p=0·0008). Event-free survival 3 years higher in patients who switched to anastrozole (97%) after switching was 92·7% (SD 0·81) for the tamoxifen than in those who continued on tamoxifen (96%), group and 95·8% (0·65) for the group switched to though this difference was not significant (p=0·16; anastrozole, corresponding to an absolute benefit at Table 3 shows the incidence of serious adverse events Figure 3 shows the risk of recurrence of cancer by treatment group. There were significantly more stratified by nodal status, tumour grading, age, and fractures (p=0·015) and significantly fewer thromboses receptor status. Although the 95% CIs of the subgroups (p=0·034) in patients treated with anastrozole than in overlap, and so the differences are not significant, the those treated with tamoxifen. There was also a trend data suggest that women with G1, G2, and Gx lobular towards fewer emboli (p=0·064) and endometrial tumours responded better to anastrozole than to cancers (p=0·069) in patients treated with anastrozole.
tamoxifen than did those with G3 tumours. For all The incidence of predefined adverse events in ABCSG patients, irrespective of tumour grading, the advantage trial 8 is shown in table 4. No adverse events were of switching to anastrozole over continuing with prespecified in the study protocol of ARNO 95. There tamoxifen was not affected by nodal status, age at were significantly more reports of nausea (p=0·0162) surgery, or receptor positivity, although there is a (non- and a trend towards more reports of bone pain significant) suggestion that the benefit of anastrozole in (p=0·0546) in the anastrozole group than in the negative patients is greater (figure 3). Tamoxifen
Anastrozole
Tamoxifen
Anastrozole
(n=1117)
(n=1120)
(95% CI), p
Number of patients who had had a previous event of a different type is given inparentheses.
Data are number (%) unless otherwise indicated.
Table 2: Recurrences and deaths by treatment group
Table 4: Predefined adverse events by treatment group in ABCSG trial 8
www.thelancet.com Vol 366 August 6, 2005
Articles
Discussion
tamoxifen17 were more likely than patients who Our data show that, in postmenopausal women with continued on tamoxifen to have arthralgia and early breast cancer, switching to anastrozole after osteoporosis at a follow-up of 30·6 months. Results 2 years’ tamoxifen treatment results in reduced rates of from a bone substudy of that trial showed that after disease recurrence, particularly with respect to distant 1 year, exemestane was associated with significantly metastases. There are two possible explanations for this greater reductions in the lumbar spine and total hip finding: tamoxifen resistance might be overcome by a bone-mineral density (BMD) than tamoxifen. The change in treatment; or aromatase inhibitors might decrease in BMD was rapid—within 6 months of simply be a better treatment option, since they reduce switching to exemestane—and by the end of the first peripheral oestrogen concentrations to extremely low year, the BMD loss was similar to that seen with other levels, whereas tamoxifen is a partial agonist.
The number of women in the combined analysis who In a placebo-controlled trial28 of the effect of had G3 tumours was small, yet nearly a third of exemestane on BMD in postmenopausal women with recurrences arose in this group. Overall, patients with early breast cancer, the aromatase inhibitor modestly G1, G2, or Gx lobular tumours responded better to increased bone loss from the femoral neck.
adjuvant therapy than did those with G3 tumours, as Management of the increased risk of fractures caused by expected. Undifferentiated tumours generally have a less BMD loss includes regular BMD screening. For patients pronounced response to endocrine therapy and could, therefore, be expected to progress more readily; the osteoporosis—eg, advanced age, smoking status, family 5-year survival rate of patients with undifferentiated history, and high body-mass index—the administration tumours at diagnosis is about 20% lower than that of of bisphosphonates could be considered as a patients with highly-differentiated or moderately- Overall, the published work indicates that there are difference in response, patients with undifferentiated potential benefits to switching from tamoxifen to an tumours would be less likely to receive adjuvant aromatase inhibitor after 2 years, and that patients could benefit from the antitumour effects of tamoxifen in the short term while avoiding the complications of long- populations with a good prognosis (about three-quarters term tamoxifen therapy. However, to date, the studies of patients were node-negative and a similar proportion have been structured such that the analyses relate only to received breast conservation surgery), we did not expect the period of switched treatment. Patients in whom to see a survival difference at this stage. Furthermore, cancer recurs at an early stage or who do not survive to longer follow-up is needed to show a significant the end of the initial tamoxifen phase are, therefore, difference in overall survival in a trial between two excluded, and the randomised population is selected active treatments than in a trial of an active treatment from patients with tumours that show a good response to endocrine therapy. As such, the results of this The contrasting safety profiles of anastrozole and investigation and other switching trials apply only to tamoxifen are well known. We noted significantly more those women who have successfully completed fractures and significantly fewer thromboses in patients 2–3 years’ adjuvant therapy for early breast cancer. They treated with anastrozole than in those who received only are not applicable to newly diagnosed patients, and tamoxifen. However, we also noted a non-significant should not be used to support a treatment strategy of tendency towards fewer emboli and endometrial starting with tamoxifen with the intention of changing to cancers in women on anastrozole. The ATAC trial15 has an aromatase inhibitor after 2 or more years. Overall, already provided evidence of the long-term safety and however, the results of these studies show the efficacy tolerability of anastrozole treatment, and no new safety advantages attached to treatment with an aromatase concerns arose during this analysis. As expected, the inhibitor, despite the qualitative differences cross-trial fracture rate in the group switched to anastrozole was comparisons reveal as to the magnitude of such higher than in the group who received continuous advantages, definitions of predefined adverse events, or tamoxifen. However, the fracture rate in the anastrozole group was lower than that seen at a similar point in the The benefits of reduced recurrence of cancer when anastrozole group of the ATAC trial.26 This finding switching adjuvant therapy to an aromatase inhibitor could suggest a continued protective effect of tamoxifen before progression on tamoxifen might be related to on bone in the ABCSG trial 8/ARNO 95 patients; cellular changes within the tumour in response to anastrozole-treated patients in the ATAC trial had tamoxifen treatment. The effect of switching from one received no previous treatment with tamoxifen.
endocrine treatment to another after less than 2 years However, data from another aromatase inhibitor, needs further investigation, as does switching from exemestane, do not support this hypothesis, since primary therapy with an aromatase inhibitor to other patients switched to exemestane after 2–3 years’ treatment modalities. Although even less thoroughly www.thelancet.com Vol 366 August 6, 2005
Articles
understood than tamoxifen resistance, resistance to Krankenhaus Linz); R Fuegger (Krankenhaus der Elisabethinen Linz); aromatase inhibitors over 5 years of exposure also leads B Gebhard (Landeskrankenhaus Voecklabruck); O Grabner to the recurrence of the original disease.29 (Landeskrankenhaus Rottenmann); C Groeger (KrankenhausNeunkirchen); H Haberfellner (Landeskrankenhaus Schaerding); Switching treatment to an aromatase inhibitor offers A Haid (Landeskrankenhaus Feldkirch); K Haider (Krankenhaus Wiener the opportunity to continue adjuvant therapy for longer Neustadt); B Hartmann (Krankenhaus Neunkirchen); H Hausmaninger than 5 years, since problems of tolerability that arise (St Johanns Spital Salzburg); CC Hinterbuchinger (Landeskrankenhaus from the partial agonist effects of tamoxifen are Kirchdorf); F Hofbauer (Krankenhaus Oberpullendorf); H Hofmann(Landeskrankenhaus Feldbach); W Horvath (Krankenhaus Guessing); circumvented. In one study,30 extended adjuvant therapy J Karner (Kaiser Franz Josefs Spital Wien); C W Kopf (Krankenhaus der with letrozole, another non-steroidal aromatase Barmherzigen Brueder Linz); A Kretschmer (Krankenhaus inhibitor, conferred a significant benefit in terms of Waidhofen/Thaya); E Kubista (Medizinische Universitaet Wien); disease-free survival after 5 years’ tamoxifen therapy.
R Lenzhofer (Krankenhaus Schwarzach); MG Lilgenau (KrankenanstaltSanatorium Hera Wien); G Loncsar (Krankenhaus Hainburg); The extended adjuvant approach is also being H P Ludwig (Wilhelminenspital Wien); G Luschin-Ebengreuth investigated with 3 years’ anastrozole therapy (compared (Medizinische Universitaet Graz); K Mach (Krankenhaus Oberwart); with no treatment) after the standard 5 years’ treatment P Magg (Krankenhaus Scheibbs); C Marth (Medizinische Universitaet Innsbruck); E Melbinger (Landeskrankenhaus Wolfsberg); E Nessler (Krankenhaus Dornbirn); W Neunteufel (Krankenhaus Dornbirn); Research indicates that 5 years of treatment with R Obwegeser (Medizinische Universitaet Wien); D Pacher (Privatklinik tamoxifen is no longer the optimum therapy for Villach); W Passath (Landeskrankenhaus Hartberg); S Poestlberger postmenopausal women with endocrine-responsive (Krankenhaus der Barmherzigen Schwestern Linz); F Pressl(Landeskrankenhaus Steyr); R Punzengruber (Krankenhaus Amstetten); early breast cancer. The results of the ATAC trial15 show H Rabl (Landeskrankenhaus Leoben); G Ralph (Landeskrankenhaus that 5 years of anastrozole as initial endocrine therapy is Bruck/Mur); A C Reichenauer (Krankenhaus der Barmherzigen Brueder better than tamoxifen for adjuvant monotherapy, and St Veit); G Reiner (Krankenhaus Mistelbach); K Renner several trials support changing adjuvant therapy to an (Sozialmedizinisches Zentrum Ost Wien); M Resinger (Krankenhausder Barmherzigen Brueder Eisenstadt); P Riss (Thermenklinikum aromatase inhibitor after initial treatment with Moedling); M Rottmann (Krankenanstalt Rudolfstiftung Wien); G Salem tamoxifen. Both of these findings are taken into account (Krankenhaus St Poelten); H Salzer (Wilhelminenspital Wien); in the ASCO technology assessment.18 Although further P Sandbichler (Krankenhaus St. Vinzenz Zams); R Schildberger investigation of the use of aromatase inhibitors is (Landeskrankenhaus Freistadt); M Schmid (Medizinische UniversitaetGraz); J Schueller (Krankenanstalt Rudolfstiftung Wien); H Seewann necessary to ascertain the ideal sequence and duration of (Landeskrankenhaus Fuerstenfeld); P Sevelda (Krankenhaus Lainz adjuvant endocrine therapy, this combined analysis Wien); B Spechtenhauser (Bezirkskrankenhaus Kufstein); G Steger confirms that postmenopausal women who receive (Medizinische Universitaet Wien); S Stengl (Krankenhaus der tamoxifen as adjuvant therapy should be switched to Barmherzigen Brueder Wien); H Stephan (LandeskrankenhausBregenz); M Stierer (Hanusch Krankenhaus Wien); S Taucher anastrozole after 2 years of treatment.
(Medizinische Universitaet Wien); J Thaler (Krankenhaus der Contributors
Barmherzigen Schwestern Wels); K Toegel (Krankenhaus Krems); R Jakesz is the principal investigator and chairs the writing and H Trapl (Krankenhaus Baden); K Unterrieder (Landeskrankenhaus steering committees. W Jonat, M Gnant, and M Kaufmann participated Villach); A Urbania (Landeskrankenhaus Klagenfurt); S A Wenzl-Eybl in the coordination of the trial and the preparation of trial results for (Krankenhaus der Barmherzigen Schwestern Ried); V Wette analysis. M Mittlboeck was responsible for the statistical analysis and (Krankenhaus der Barmherzigen Brueder St. Veit); B Zeh participated in trial design. All authors contributed to the design of the (Donauklinikum Tulln); and W Zeilmann (Kardinal study, participated in the overall operational management of the trial, Schwarzenberg’sches Krankenhaus Schwarzach).
contributed to data interpretation, and participated in the writing of GABG (ARNO 95) investigators (all institutes are in Germany)
A K Bergmann (Universitaetsklinikum Tuebingen); F von Bismarck(Klinikum Rechts der Isar, Muenchen); K Brunnert (Klinik fuer ABCSG trial 8/ARNO 95 trial writing committee
Prof R Jakesz (chairman of the ABCSG trial 8/ARNO 95 trial steering
Senologie und Plastische Chirurgie Osnabrueck); M Butterwegge committee, Vienna Medical University, Vienna, Austria); Prof W Jonat (Marienhospital Osnabrueck); O Camara (Universitaetsklinik Jena); (member of the ABCSG trial 8/ARNO 95 trial steering committee, M Carstensen (Albertinen Krankenhaus Hamburg); K Christl University Clinic Schleswig-Holstein, Kiel, Germany); Prof M Gnant (Krankenhaus Eggenfelden); M H R Eichbaum (Universitaetsklinikum (member of the ABCSG trial 8/ARNO 95 trial steering committee, Heidelberg); J Feltz-Suessenbach (Klinikum Schaumburg, Stadthagen); Vienna Medical University, Vienna, Austria); Prof M Mittlboeck R Fricker (Stadtkrankenhaus Hanau); M Friedrich (Universitaetsklinik (statistician, Vienna Medical University, Vienna, Austria); Prof R Greil Schleswig Holstein, Campus Luebeck); L Funk (St. Vincentius (Paracelsus Medical University, Salzburg, Austria); C Tausch (Linz, Krankenhaeuser Karlsruhe); E Goepel (Gynaekologische Austria); Prof J Hilfrich (Henriettenstiftung, Hannover, Germany); Gemeinschaftspraxis Hamburg); S Graeber (Universitaet Halle); W Kwasny (Wiener Neustadt Hospital, Wiener Neustadt, Austria); C A Hanusch (Frauenklinik vom Roten Kreuz Muenchen); L Heilmann Prof C Menzel (Salzburg Hospital, Salzburg, Austria); Prof H Samonigg (Stadtkrankenhaus Ruesselsheim); VJ Heilmann (Universitaetsklinik (Graz Medical University, Graz, Austria); Prof M Seifert (Vienna Ulm); T Hitschold (Stadtkrankenhaus Worms); A Hoenig Medical University, Vienna, Austria); Prof G Gademann (University of (Universitaetsklinik Wuerzburg); M Holbeck (Martin Luther Magdeburg, Magdeburg, Germany); Prof M Kaufmann (member of the Krankenhaus Schleswig); G Hopf (Elisabeth Krankenhaus Kassel); ABCSG trial 8/ARNO 95 trial steering committee); Johann Wolfgang T Horvath (Kreisklinik Albstadt); H Jank (DRK Kliniken Westend (Goethe University, Frankfurt, Germany).
Berlin); M Johnscher (Bonifatius Hospital Lingen); G Kaltenecker(Staedtisches Klinikum Karlsruhe); C Karg (Kreiskrankenhaus ABCSG trial 8 investigators (all institutes are in Austria)
Waiblingen); T Liersch (Universitaetsklinik Goettingen); E Lindenlauf H Aigner (Krankenhaus Spittal/Drau); P Balcke (Krankenhaus (Jung Stilling Krankenhaus Siegen); B Lisboa (Universitaetsklinikum St Poelten); C Bosse (Krankenhaus Klosterneuburg); F Burger Eppendorf Hamburg); K Luermann (Klinikum Hoyerswerda); I Maier (Krankenhaus Horn); C Dadak (Medizinische Universitaet Wien); (Klinikum Rosenheim); G von Minckwitz (Universitaetsklinik E Forsthuber (Landeskrankenhaus Klagenfurt); M Fridrik (Allgemeines Frankfurt); B Mueller (Klinikum der Stadt Mannheim); C Mundhenke www.thelancet.com Vol 366 August 6, 2005
Articles
(Universitaetsklinik Schleswig Holstein, Campus Kiel); A Neff Geisler J, King N, Anker G, et al. In vivo inhibition of (Clemenshospital Muenster); T Reimer (Universitaetsklinik Rostock); aromatization by exemestane, a novel irreversible aromatase K Rensing (Universitaetsklinik Muenster); I Schrader inhibitor, in postmenopausal breast cancer patients.
(Henriettenstiftung Hannover); D Schulze (St Vincenz Krankenhaus Clin Cancer Res 1998; 4: 2089–93.
Paderborn); I Schulz-Im-Busch (St. Josephs Hospital Cloppenburg); Geisler J, Haynes B, Anker G, Dowsett M, Lønning PE. Influence K W Schweppe (Ammerland Klinik Westerstede); G Sergius (St Joseph of letrozole and anastrozole on total body aromatization and Krankenhaus Berlin); H K Sommer (Kreiskrankenhaus Neustadt); plasma estrogen levels in postmenopausal breast cancer patientsevaluated in a randomized, cross-over study. K Stahl (Prosper Hospital Recklinghausen); S Stein (Staedtische J Clin Oncol 2002; 20:
Krankenanstalten Idar-Oberstein); J Steller (Helios Research Center The ATAC Trialists’ Group. Results of the ATAC (Arimidex, Titisee-Neustadt); E Stickeler (Universitaetsklinik Freiburg); Tamoxifen, Alone or in Combination) trial after completion of R Stiglmayr (Siloah Krankenhaus Pforzheim); W Wiest (St Vincenz und 5 years’ adjuvant treatment for breast cancer. Lancet 2005; 365:
St Elisabeth Hospital Mainz); K J Winzer (Universitaetsklinikum Charité Berlin); H Wolf (Kreiskrankenhaus Leonberg); K Wollschlaeger Boccardo F, Rubagotti A, Puntoni M, et al. Switching to anastrozole (Universitaetsfrauenklinik Magdeburg); and B Zieren-Ernhardt versus continued tamoxifen treatment of early breast cancer: (Universitaetsklinik Charité Berlin).
preliminary results of the Italian Tamoxifen Anastrozole (ITA) Conflict of interest statement
trial. J Clin Oncol 2005; published online July 11, DOI: R Jakesz, W Jonat, and M Kaufmann have done research sponsored by AstraZeneca. M Gnant, M Mittlboeck, R Greil, C Tausch, J Hilfrich, Coombes RC, Hall E, Gibson LJ, et al. A randomized trial of W Kwasny, C Menzel, H Samonigg, M Seifert, and G Gademann exemestane after two to three years of tamoxifen therapy inpostmenopausal women with primary breast cancer. N Engl J Med declare that they have no conflict of interest. 2004; 350: 1081–92.
Acknowledgments
Winer EP, Hudis C, Burstein HJ, et al. American Society of We thank all patients in Austria and Germany for participating in this Clinical Oncology technology assessment on the use of aromatase investigation, the ABCSG and GABG trialists, the staff at the Vienna inhibitors as adjuvant therapy for postmenopausal women with and Frankfurt trial offices, the respective study site affiliates, and hormone receptor–positive breast cancer: status report 2004.
Wolfgang Draxler and Karl Thomanek for editorial and biometrical J Clin Oncol 2005; 23: 619–29.
expertise. We also thank AstraZeneca for making the study medication Bloom HJ, Richardson WW. Histological grading and prognosis in available and Complete Medical Communications for editorial breast cancer: a study of 1409 cases of which 359 have been followed for 15 years. Br J Cancer 1957; 11: 359–77.
Pocock SJ, Simon R. Sequential treatment assignment with References
balancing for prognostic factors in the controlled clinical trial.
Ferlay J, Bray F, Pisani P, Parkin DM. GLOBOCAN 2002: cancer Biometrics 1975; 31: 103–15.
incidence, mortality and prevalence worldwide. IARC CancerBase, Kaplan EL, Meier P. Nonparametric estimation from incomplete version 2.0. Lyon: IARCPress, 2004. http://www-depdb.iarc.fr/ observations. J Am Stat Assoc 1958; 53: 457–81.
globocan/table2_sel.htm (accessed July, 2005).
Peto R, Peto J. Asymptotically efficient rank invariant procedures.
Anderson WF, Chatterjee N, Ershler WB, Brawley OW. Estrogen J R Stat Soc 1972; 135: 185–207.
receptor breast cancer phenotypes in the surveillance,epidemiology, and end results database. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. 76: 27–36.
50: 163–70.
Goldhirsch A, Wood WC, Gelber RD, Coates AS, Thürlimann B,Senn HJ. Meeting highlights: updated international expert Cox DR. Regression models and life tables. J R Stat Soc 1972; 34:
consensus on the primary therapy of early breast cancer.
J Clin Oncol 2003; 21: 3357–65.
Lundin J, Lundin M, Holli K, et al. Omission of histologic grading Peto R. Five years of tamoxifen: or more? from clinical decision making may result in overuse of adjuvant therapies in breast cancer: results from a nationwide study.
88: 1791–93.
J Clin Oncol 2001; 19: 28–36.
Fisher B, Dignam J, Bryant J, et al. Five versus more than five yearsof tamoxifen therapy for breast cancer patients with negative lymph The ATAC Trialists’ Group. Anastrozole alone or in combination nodes and estrogen receptor-positive tumors. with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: first results of the 88: 1529–42.
ATAC randomised trial. Lancet 2002; 359: 2131–39.
Early Breast Cancer Trialists’ Collaborative Group. Effects ofchemotherapy and hormonal therapy for early breast cancer on Coleman RE, Banks LM, Hall E, et al. Intergroup exemestane recurrence and 15-year survival: an overview of the randomised study: 1 year results of the bone sub-protocol. Breast Cancer Res trials. Lancet 2005; 365: 1687–717.
Treat 2005; 88: S35 (abstract).
Fisher B, Costantino JP, Redmond CK, Fisher ER, Wickerham DL, Lonning PE, Geisler J, Krag LE, et al. Effects of exemestane Cronin WM. Endometrial cancer in tamoxifen-treated breast cancer administered for 2 years versus placebo on bone mineral density, patients: findings from the National Surgical Adjuvant Breast and bone biomarkers, and plasma lipids in patients with surgically J Natl Cancer Inst 1994; 86: 527–37.
J Clin Oncol 2005; published online June 27, DOI: 10.1200/JCO.2005.07.097.
Jaiyesimi IA, Buzdar AU, Decker DA, Hortobagyi GN. Use oftamoxifen for breast cancer: twenty-eight years later. Goss P. Anti-aromatase agents in the treatment and prevention of 13: 513–29.
Cancer Control 2002; 9 (2 suppl): 2–8.
Clarke R, Leonessa F, Welch JN, Skaar TC. Cellular and molecular Goss PE, Ingle JN, Martino S, et al. A randomized trial of letrozole pharmacology of antiestrogen action and resistance. in postmenopausal women after five years of tamoxifen therapy for 53: 25–71.
N Engl J Med 2003; 349: 1793–802.
Early Breast Cancer Trialists’ Collaborative Group. Tamoxifen for Jakesz R, Samonigg H, Greil R, et al. Extended adjuvant treatment early breast cancer. (Cochrane Review). In: The Cochrane Library, with anastrozole: results from the Austrian Breast and Colorectal Issue 1. Chichester: John Wiley & Sons, 2004.
Cancer Study Group Trial 6a (ABCSG-6a). Proc Am Soc Clin Oncol
2005; 23: 10s (abstract).
Saphner T, Tormey DC, Gray R. Annual hazard rates of recurrence
for breast cancer after primary therapy. J Clin Oncol 1996; 14:
2738–46.
Morandi P, Middleton LP, Mirza NQ, et al. Occult breast cancer
(OBC), a retrospective analysis of clinico-pathological
characteristics and long-term outcome: the University of Texas MD
Anderson Cancer Center experience. Breast Cancer Res Treat 2004;
88: S165 (abstract).
www.thelancet.com Vol 366 August 6, 2005

Source: http://www.brustzentrum-hamburg.de/materialien/ABCSG8.pdf