VOLUME 28 NUMBER 3 JANUARY 2 2 JOURNAL OF CLINICAL ONCOLOGY S P E C I A L A R T I C L E From the Academic Department of Biochemistry, Royal Marsden Hospital; Cancer Research UK Centre for Epidemiology, Mathematics and Statistics, Queen Mary University of London; Clinical Trials Group, Royal Free and University College London Medical School; Department of Cancer Medicine, Charing Cross Hospital, London; Clinical Trials and Statistics Unit, The Institute of Cancer Research, Surrey; Clinical Trials Service Unit and Epidemiological Studies Unit, University of Oxford, Oxford, United Kingdom; Mayo Clinic, Rochester, MN; International Breast Cancer Study Group, Bern, Switzerland; Department of Surgical Oncology, University of Newcastle, Newcastle, Australia; Department of Breast Medical Oncology, M. D. Anderson Cancer Center, Houston, TX; Department of Medicine, Istituto Europeo di Oncologia, Milan; University and National Cancer Research Institute, Genoa, Italy; Department of Surgery, Medical University of Vienna; Division of General Surgery, Vienna Medical School, Vienna, Austria; OB/GYN/Breast Unit, University Hospital Frankfurt, Frankfurt, Gernany. Submitted March 7, 29; accepted August 26, 29; published online ahead of print at www.jco.org on November 3, 29. Written on behalf of the Aromatase Inhibitor Overview Group. The Early Breast Cancer Trialists Collaborative Group secretariat is funded by the core Clinical Trial Service Unit and Epidemiological Studies Unit, support from Cancer Research UK, and by the Medical Research Council. National Health Service funding was provided to the Royal Marsden National Institute for Health Research Biomedical Research Centre. Authors disclosures of potential conflicts of interest and author contributions are found at the end of this article. Corresponding author: Mitch Dowsett, PhD, Royal Marsden Hospital, 237 Fulham Rd, London SW3 6JJ, United Kingdom; e-mail: mitch.dowsett@ icr.ac.uk. 29 by American Society of Clinical Oncology 732-83X//283-59/$2. DOI:.2/JCO.29.23.274 Meta-Analysis of Breast Cancer Outcomes in Adjuvant Trials of Aromatase Inhibitors Versus Mitch Dowsett, Jack Cuzick, Jim Ingle, Alan Coates, John Forbes, Judith Bliss, Marc Buyse, Michael Baum, Aman Buzdar, Marco Colleoni, Charles Coombes, Claire Snowdon, Michael Gnant, Raimund Jakesz, Manfred Kaufmann, Francesco Boccardo, Jon Godwin, Christina Davies, and Richard Peto A B S T R A C T Purpose To conduct meta-analyses of randomized trials of aromatase inhibitors (s) compared with tamoxifen either as initial monotherapy (cohort ) or after 2 to 3 years of tamoxifen (cohort 2). Materials and Methods Data submitted to the Early Breast Cancer Trialists Collaborative Group were used in separate meta-analyses of two cohorts. Primary analyses involve postmenopausal women with tumors reported to be estrogen receptor positive. Log-rank P values are two-sided. Results Cohort comprised 9,856 patients with a mean of 5.8 years of follow-up. At 5 years, therapy was associated with an absolute 2.9% (SE.7%) decrease in recurrence (9.6% for v 2.6% for tamoxifen; 2P.) and a nonsignificant absolute.% (SE.5%) decrease in breast cancer mortality (4.8% for v 5.9% for tamoxifen; 2P.). Cohort 2 comprised 9,5 patients with a mean of 3.9 years of follow-up. At 3 years from treatment divergence (ie, approximately 5 years after starting hormonal treatment), therapy was associated with an absolute 3.% (SE.6%) decrease in recurrence (5.% for v 8.% for tamoxifen since divergence; 2P.) and an absolute.7% (SE.3%) decrease in breast cancer mortality (.7% for v 2.4% for tamoxifen since divergence; 2P.2). There was no convincing heterogeneity in the proportional recurrence reduction with respect to age, nodal status, tumor grade, or progesterone receptor status and no indication of an increase in nonbreast deaths with s in either cohort. Conclusion s produce significantly lower recurrence rates compared with tamoxifen, either as initial monotherapy or after 2 to 3 years of tamoxifen. Additional follow-up will provide clearer information on long-term survival. J Clin Oncol 28:59-58. 29 by American Society of Clinical Oncology INTRODUCTION is a selective estrogen receptor (ER) modulator that acts (along with its metabolites) largely or wholly by competitive binding to the receptor protein. In the 75% to 8% of patients with early breast cancer who have ER-positive disease, treatment with 5 years of tamoxifen immediately and substantially reduces local, contralateral, and distant recurrence rates and reduces 5-year breast cancer mortality, with little effect on the aggregate of all other causes of death. 2,3 However, there is little or no effect on breast cancer outcomes in ER-negative disease. The early results from the first clinical trials of tamoxifen showed only a small absolute gain in 5-year survival, 3 but the additional mortality reductions during years 5 to 9 and to 4 were each about as great as the mortality reduction during years to 4 (the 5 years while tamoxifen was still being taken). 2 Hence, the absolute reduction in breast cancer mortality produced by 5 years of tamoxifen was almost three times as great 5 years after diagnosis as it had been only 5 years after diagnosis. The development of well-tolerated inhibitors of aromatase, the enzyme that synthesizes estrogens from androgens, has provided an alternative means of depriving breast tumors of stimulation by endogenous estrogens in postmenopausal women whose ovaries no longer produce significant amounts of the hormone 4,5 and in premenopausal women in whom ovarian function has been suppressed or the ovaries have been removed. In recent years, several large randomized trials in early breast cancer with 5 29 by American Society of Clinical Oncology 59 Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Dowsett et al years of tamoxifen as the control arm have reported on the use of one of three third-generation aromatase inhibitors (s) anastrozole, letrozole, or exemestane; other large third-generation trials have yet to be reported. 5 Although the trials considered in these meta-analyses have generally found reduced recurrence rates compared with 5 years of tamoxifen, none on its own has yet clearly demonstrated a reduction in breast cancer mortality, and none possesses adequate statistical power to do so or to yield reliable subgroup analyses of the effects on recurrence. Therefore, the trialists have collaborated through the Oxford Early Breast Cancer Trialists Collaborative Group (EBCTCG) overview process to pool data from all currently available trials of versus tamoxifen, using the methods of previous EBCTCG metaanalyses. When results from individual trial publications are amalgamated in a systematic review, no allowance can be made for the different duration of follow-up in each study or for the differing definitions of some end points, and subgroup analyses may be reported inconsistently. For example, the primary end point of most trials in this meta-analysis was disease-free survival but variably included or excluded occurrence of a new nonbreast cancer. Individual patient meta-analysis as carried out here allows each of these issues to be addressed. The trials have been grouped into two cohorts, and these cohorts have been analyzed separately. Cohort consists of the trials comparing 5 years of an with 5 years of tamoxifen, both starting soon after surgery, 6,7 and cohort 2 consists of the trials in which tamoxifen therapy is switched after 2 to 3 years to 2 to 3 years of an compared with 2 to 3 more years of tamoxifen (making a total of 5 years of hormonal therapy in both groups). 8- Some of the trials included a small proportion of patients with ER-negative or ER untested tumors. These have both been excluded from the present analyses because of the lack of any material effect of 5 years of tamoxifen in ER-negative disease. 2 Trials comparing treatment with an after 5 years of tamoxifen versus no continuation of any hormonal therapy are not covered by this overview. The database for this analysis was locked before updates from some of the trials were presented at the San Antonio Breast Cancer Symposium in December 28. Those updates are consequently not included in this report. an event rate ratio of.), and a broken vertical line is used to denote the average treatment effect in a meta-analysis of all the available evidence. In the present report, the broken vertical line is made particularly bold in the figures that present multiple subgroup analyses. This is to help emphasize tests of heterogeneity between subgroups, thereby reducing (but not eliminating) the relevance of tests of the null hypothesis within particular subgroups.,2 Another amendment is to present, among other things, analyses of time to first distant recurrence, ignoring any prior locoregional or contralateral recurrences, as well as analyses of various types of recurrence as a first event. (Recurrence was defined as the first reappearance of breast cancer at any site and thus included second primary breast cancer or locoregional or distant recurrence of the original cancer.) As before, 2 log-rank analyses are presented for any death, death without recurrence, and, by log-rank subtraction, death with recurrence. Because the incidence of recurrence was monitored carefully in these trials, the final amendment is that deaths from an unknown cause without any recorded recurrence are taken as non breast cancer deaths. Analyses are presented entirely separately for the trials of approximately 5 years of versus the same duration of tamoxifen (cohort ) and for the switching trials of some years of tamoxifen and then some years of versus the same duration of tamoxifen (cohort 2). In the switching trials, the analyses are of time since the treatments diverged, which is approximately 2 to 3 years after originally starting adjuvant tamoxifen shortly after diagnosis. All but one of the switching trials randomly assigned patients after the first few years of tamoxifen (to switch or continue). The remaining trial (ABCSG VIII trial) 8 randomly assigned patients at the start of adjuvant tamoxifen between 2 years of tamoxifen and then 3 years of and, as control, 5 years of tamoxifen. The analyses of this one trial are of time since the treatments would have diverged, and any patients who experienced recurrence, death, or loss to follow-up within the first 2 years are thus excluded from numerators, denominators, and all analyses. Keeping to the convention of previous EBCTCG analyses, the first year of adjuvant treatment is referred to as year. 2,3 To help allow for the fact that some trials are more mature than others, the follow-up in cohort trials is subdivided into years to (first 2 years of treatment), 2 to 4 (last 3 years of treatment), and 5 (after treatment). For cohort 2, follow-up is subdivided into years to 2 (the 3 years of randomly assigned treatment) and 3 (after treatment) after the treatments were to diverge; these divisions equate to years 2to4and5, respectively, from the start of adjuvant hormonal treatment because of the initial 2 years of treatment of all patients in this cohort with tamoxifen. For both cohorts, Kaplan-Meier graphs are plotted to 8 years from the start of adjuvant hormonal treatment. The lock date for the data analysis was September 3, 26 for all trials. MATERIALS AND METHODS Trial identification and data handling procedures have been described previously. 2 Information was sought from all trials that had started by the year 2 in which two arms differed for some period of time only in one having an and the other having tamoxifen. The only two such trials that are known to be unavailable are, first, the Austrian Breast and Colorectal Cancer Study Group (ABCSG) XII trial, which started in 999 and randomly assigned,83 premenopausal women, whose ovarian function was being suppressed by goserelin, to 3 years of (anastrozole) or 3 years of tamoxifen and, second, the two switching arms from the Breast International Group (BIG) -98/International Breast Cancer Study Group (IBCSG) 8-98 trial, which started in 998. This trial randomly assigned 6,93 women four ways (between 5 years of [letrozole], 5 years of tamoxifen, switching from tamoxifen to, and switching from to tamoxifen); 3,95 of women were randomly assigned to the two sequencing arms. The statistical methods and the format of presentation of the results are as described in the most recent EBCTCG report, 2 with minor amendments. One amendment is to the format of the forest plots in which squares and horizontal lines are used to describe event rate ratios ( v tamoxifen) in various subgroups. As before, a solid vertical line is used to denote no effect (ie, RESULTS Cohort : 5 Years of Versus 5 Years of Two trials (Arimidex,, Alone or in Combination [ATAC] and BIG -98/IBCSG 8-98; Appendix Fig A, online only) contribute data, including a total of 9,856 women and more than 55, woman-years of follow-up (mean, 5.8 years). Mean and median follow-up times were 7. and 7.9 years, respectively, for ATAC and 4.3 and 4. years, respectively, for BIG -98. Most of the recurrences thus far recorded occurred during the first 5 years, while both groups were still on active treatment (Fig A). Use of an rather than tamoxifen was associated with a highly significant 23% (SE 5%) proportional reduction in these early recurrences (2P.; Appendix Fig AA). There seemed to be greater proportional decreases in isolated local recurrence (3%; SE %; 2P.3; Appendix Fig AB) and in contralateral disease (4%; SE 2%; 2P.9; Appendix Fig AC) than in distant recurrence either as a first event (6%; SE 6%; 2P.9; Appendix Fig AD) or at any time (8%; SE 6%; 2P.2; Appendix Fig AE), although this apparent 5 29 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Meta-Analysis of Aromatase Inhibitors Versus A Recurrence (%, ± SE) 5 4 3 2 5-year gain, 2.9% (SE,.7%) 8-year gain, 3.9% (SE,.%) Log-rank 2P <. 2.6% 9.6% 9.2% 5.3% B Breast Cancer Mortality (%, ± SE) 5 4 3 2 5-year gain,.% (SE,.5%) 8-year gain,.5% (SE,.8%) Log-rank 2P =. 5.9%.5%.% 4.8% C 5 8 Time (years) Recurrence rates (% / year) and log-rank analyses Years Years 2 4 Years 5+.69 (63 / 9,647) 2.3 (26 /,297) 2.33 (6 / 6,879) 2.46 (234 / 9,5) 2.8 (37 /,938) 2.78 (8 / 6,478) Rate ratio, from.67 (SE,.8).8 (SE,.8).83 (SE,.) (O E) / V -38.4 / 96.6-29.5 / 37.9-5.7 / 83. Death Without Recurrence (%, ± SE) 5 4 3 2 5-year loss,.2% (SE,.5%) 8-year loss,.3% (SE,.8%) Log-rank 2P =.9 4.4% 4.2% 5 8 Time (years) 9.% 8.8% Death rates (% / year) and log-rank analyses Years Years 2 4 Years 5+.54 (52 / 9,647). (24 /,297).6 ( / 6,879).59 (56 / 9,5).2 (2 /,938).62 (5 / 6,478) Rate ratio, from.92 (SE,.9).7 (SE,.4).99 (SE,.4) (O E) / V -2. / 26.8 4. / 58.3 -.4 / 52.9 D 5 8 Time (years) Death rates (%/year: total rate rate in women without recurrence) & log-rank analyses Years Years 2 4 Years 5+.59 (SE,.8).26 (SE,.).78 (SE,.6).57 (SE,.8).6 (SE,.2).79 (SE,.6) Rate ratio, from. (SE,.9).77 (SE,.). (SE,.3) (O E) / V.2 / 27.5-2.5 / 8.2.4 / 6.5 Any Death (%, ± SE) 5 4 3 2 5-year gain,.8% (SE,.6%) 8-year gain,.2% (SE,.%) Log-rank 2P =.3 9.6% 8.8% 5 8 Time (years) 8.% 7.8% Death rates (% / year) and log-rank analyses Years Years 2 4 Years 5+.2 (9 / 9,743) 2.3 (272 /,757) 3.29 (24 / 7,286).5 ( / 9,656) 2.57 (297 /,569) 3.26 (232 / 7,7) Rate ratio, from.97 (SE,.3).89 (SE,.8). (SE,.9) (O E) / V -.9 / 54.3-6.4 / 38.5. / 4.4 Fig. Life-table curves of (A) recurrence; (B) breast cancer mortality; (C) death without recurrence; and (D) any death, for estrogen receptor positive patients in trials of approximately 5 years of aromatase inhibitor () versus tamoxifen. O, observed; E, expected; V, variance. heterogeneity of effect between recurrence sites was not statistically significant ( 2 2 5., P.8). The proportional reduction in the risk of recurrence was not significantly different during the first 2 years and during the last 3 years of treatment (Fig 2A; in both periods, it is separately significant), and there seemed to be some further risk reduction in subsequent years, although this further reduction was not, on its own, statistically significant. At 8 years after the start of hormonal therapy, the absolute difference in recurrence risk was 3.9% (SE.%; 5.3% for v 9.2% for tamoxifen). Subgroup analyses of the recurrence results with respect to progesterone receptor (PgR) status, age, nodal status, and tumor grade are shown in Figures 2B to 2E. The global test of heterogeneity between all of these subgroups was not statistically significant (2P.), so undue emphasis on any one of them may yield misleading conclusions. There was no apparent heterogeneity with respect to age, nodal status, or grade between the proportional reductions in recurrence, but there was with respect to PgR status, which was known in 8,745 patients (89%). The proportional reduction in recurrence ( v tamoxifen) was 4% (SE 9%) in the 22% of patients categorized as ER positive/ www.jco.org 29 by American Society of Clinical Oncology 5 Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Dowsett et al Events/Women events Allocated Allocated Log-rank Variance Ratio of annual event rates Category tamoxifen O E of O E : (A) Time period (trend χ 2 = 2., 2P >., NS) Years 63/4,954 234/4,92 38.4 96.6.67 (SE.8) (3.3%) (4.8%) Years 2 4 26/4,67 37/4,557 29.5 37.9.8 (SE.8) (5.6%) (6.7%) Years 5+ 6/2,939 8/2,83 5.7 83..83 (SE.) (5.4%) (6.4%) (B) PgR status (χ 2 = 5.6, 2P =.2) PgR poor 8/958 82/937 36.9 72.5.6 (SE.9) (2.3%) (9.4%) PgR+ 38/3,452 443/3,398 37.3 2.8.83 (SE.6) (.%) (3.%) PgR unknown 85/544 96/567 4.4 43.6.9 (SE.4) (5.6%) (6.9%) (C) Entry age (trend χ 2 =.4, 2P >., NS) Age < 5 25/225 33/22 5.9 3.9.65 (SE.22) (.%) (6.3%) 5 59 75/,68 228/,674 3. 98.3.74 (SE.9) (.4%) (3.6%) 6 69 222/,898 266/,94 26.4 8.8.8 (SE.8) (.7%) (4.%) 7+ 62/,5 94/,22 2. 86..78 (SE.) (4.%) (7.3%) (D) Nodal status (N/N v all N+ χ 2 =.6, 2P >., NS) N/N 8/,937 7/2,2 23.9 7.2.72 (SE.) (6.%) (8.5%) N 3 83/734 9/72 6.9 42.3.85 (SE.4) (.3%) (3.%) Fig 2. Subgroup analyses of recurrences by (A) time period; (B) progesterone receptor (PgR) status; (C) entry age; (D) nodal status; and (E) tumor grade, for estrogen receptor positive patients in trials of approximately 5 years of aromatase inhibitor () versus tamoxifen. Dashed vertical line corresponds to zero heterogeneity of effect. NS, not significant; unspec., unspecified; O, observed; E, expected. N4+ 324/,866 372/,779 36.8 67.4.8 (SE.7) (7.4%) (2.9%) N+ unspec. 45/254 65/244 2. 26.5.63 (SE.6) (7.7%) (26.6%) N unknown 4/63 22/56 4.5 8.9.6 (SE.26) (8.6%) (4.%) (E) Tumor grade (trend χ 2 =.3, 2P >., NS) Well differentiated 63/,86 93/,27 4.6 38.4.68 (SE.3) (5.3%) (7.6%) Moderately 265/2,47 33/2,388 4.4 45.2.76 (SE.7) (.%) (3.9%) Poorly 8/854 29/823 23.8 92..77 (SE.9) (2.2%) (25.4%) Grade unknown 75/497 88/474 9.8 39.3.78 (SE.4) (5.%) (8.6%) Total 584/4,954 72/4,92 83.6 37.5.769 (SE.49) (.8%) (4.7%) 2P <. 99% or 95% confidence intervals Global heterogeneity (B E): χ 2 = 6.9, P >., NS 4.5..5 2. better better Treatment effect 2P <. PgR poor and only 7% (SE 6%) in patients categorized as ER positive/pgr positive. Although the apparent heterogeneity between these two proportional risk reductions is formally statistically significant (2P.2), in view of the nonsignificant global test of heterogeneity, the total number of subgroups considered with respect to this and other factors, and the PgR-associated effect being significant in one of the two trials (ATAC) but not the other (BIG -98/IBCSG 8-98), the apparent heterogeneity with respect to PgR status could be 52 29 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Meta-Analysis of Aromatase Inhibitors Versus Deaths/Women Al deaths Year code Allocated Allocated Log-rank Variance Ratio of annual death rates and study name Treatment comparison tamoxifen O E of O E : (A) Mortality with recurrence (χ 2 =.2, 2P >., NS) 96J CTC ATAC, UK (Ana v Tam) 5 yr 223/2,548 24/2,54.6.3 98M IBCSG 8-98/BIG -98 (Let v Tam) 5 yr 2/2,46 27/2,388 9.3 57.8 Subtotal 335/4,954 367/4,92 9.9 69..89 (SE.7) (6.8%) (7.5%) reduction 2P >., NS (B) Mortality without recurrence (χ 2 =., 2P >., NS) 96J CTC ATAC, UK (Ana v Tam) 5 yr 23/2,548 2/2,54 2.4.8 98M IBCSG 8-98/BIG -98 (Let v Tam) 5yr 73/2,46 73/2,388.9 36.2 Subtotal 286/4,954 273/4,92.5 38.. (SE.9) (5.8%) (5.6%) increase 2P >., NS (C) Any death (χ 2 =.3, 2P >., NS) 96J CTC ATAC, UK (Ana v Tam) 5 yr 436/2,548 44/2,54 8.2 23. 98M IBCSG 8-98/BIG -98 (Let v Tam) 5 yr 85/2,46 2/2,388.2 94. Subtotal 62/4,954 64/4,92 8.4 37..94 (SE.6) (2.5%) (3.%) reduction 2P >., NS 99% or 95% confidence intervals.5..5 2. better better Fig 3. Trial-specific forest plot of (A) mortality with recurrence; (B) mortality without recurrence; and (C) any death, for estrogen receptor positive patients in trials of approximately 5 years of aromatase inhibitor () versus tamoxifen (Tam). O, observed; E, expected; NS, not significant; CTC, Cancer Trials Centre; ATAC, Arimidex,, Alone or in Combination; UK, United Kingdom; Ana, anastrozole; IBCSG, International Breast Cancer Study Group; BIG, Breast International Group; Let, letrozole. largely or wholly a chance finding. If so, the overall proportional reduction in recurrence rate ( v tamoxifen) may provide the best guide to the treatment effect in each of the subgroups in Figures 2B to 2E. Thus far, there has been only a small and nonsignificant proportional reduction (%; SE 7%) in breast cancer mortality in cohort (Figs B and 3A). The absolute difference between and tamoxifen in breast cancer mortality was.% (SE.5%) at 5 years. Although it seemed to be smaller (only.5%; SE.8%) at 8 years (Fig B), the mean duration of follow-up is only 5 years, so this could change with the accumulation of further data. Of the,26 deaths from any cause, 559 (44%) were of women who had not experienced a recurrence and had died of an unrelated cause. There was no apparent difference between and tamoxifen in these non breast cancer deaths (Figs C and 3B). Hence, the difference in overall mortality was also nonsignificant (Figs D and 3C). Cohort 2: 2 to 3 Years of for All Women and Then 2 to 3 Years of Versus Four trials (German Adjuvant Breast Cancer Group/Arimidex- Nolvadex, Intergroup Exemestane Study/BIG 2-97, Italian Anastrozole Trial (ITA), and ABCSG VIII) contribute data, including a total of 9,5 women with more than 33, woman-years of follow-up after the time of switching (mean, 3.9 years); mean and median follow-up times were 6. and 6. years, 4.4 and 4.4 years, 5.9 and 6. years, and 3. and 2.9 years, respectively. The analyses of recurrence and death are calculated from the time the allocated treatments would differ. (The time from diagnosis is indicated in the graphs as being approximately 2 years greater than this.) There was an overall 29% (SE 6%) proportional reduction in recurrence, with no significant heterogeneity between the contributing trials (Appendix Fig A2A, online only). The absolute cumulative risk of recurrence was reduced by 3.% (SE.6%) 3 years after switching (ie, approximately 5 years after diagnosis) and by 3.6% (SE.%) 6 years after switching (ie, approximately 8 years after diagnosis; Fig 4A). There was a significantly greater risk reduction during the first 3 years after switching (ie, while treatment with or tamoxifen was continuing) than during subsequent years (ie, after cessation of treatment; 2P.3 for heterogeneity; Fig 5A). Again, there seemed to be greater proportional reductions in isolated local recurrence (4%; SE 3%; 2P.2; Appendix Fig A2B) and contralateral disease (35%; SE 7%; 2P.3; Appendix Fig A2C) than in distant recurrence either as the first event (24%; SE 7%; 2P.; Appendix Fig A2D) or at any time (24%; SE 7%; 2P.7; Appendix Fig A2E), but this apparent heterogeneity of effect between different sites of recurrence was not significant. Subgroup analyses of the recurrence results with respect to PgR status, age, nodal status, and tumor grade are shown in Figures 5B to 5E. The global test of heterogeneity for these four subgroup analyses was not significant. There was no apparent heterogeneity with respect to any of the four subgroups in the proportional reductions in recurrence. PgR status was known in 8,84 patients (9%). The proportional reduction in recurrence was 37% (SE 2%) in women with ER-positive/PgR-poor disease and 2% (SE 8%) in women with ER-positive/PgR-positive disease, but these proportional risk reductions are not significantly different from each other. There was a significant reduction in breast cancer mortality (proportional reduction in mortality with recurrence 22%; SE 9%; 2P.2; Fig 6A). The absolute difference in breast cancer mortality was.7% (SE.6%) approximately 5 years after diagnosis and.7% (SE.8%) approximately 8 years after diagnosis (Figs 4B and 6A), but the latter result in particular may change somewhat when more women have been observed to this www.jco.org 29 by American Society of Clinical Oncology 53 Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Dowsett et al A C Recurrence (%, ± SE) 5 4 3 2 3-year gain, 3.% (SE,.6%) 6-year gain, 3.6% (SE,.%) Log-rank 2P <. 6.% 2.6% 8.% 7.9% 2.4% 6.3% 5.%.7% 3 6 3 6 ( ) ( 2) ( 5) ( 8) ( ) ( 2) ( 5) ( 8) Time Since Allocated Treatments Differed (years) (and approximate time since diagnosis) Recurrence rates (% / year) and log-rank analyses Years 2 ( 2 4) Years 2 4 ( 5 7) Years 5+ ( 8+).68 (87 /,34) 2.78 (47 / 5,298) 3.2 (23 / 76) 2.76 (33 /,962) 2.99 (49 / 5,7) 3.87 (27 / 697) Rate ratio, from.6 (SE,.7).92 (SE,.).85 (SE,.27) (O E) / V -6. / 8.4-6. / 7.9-2. / 2. Death Without Recurrence (%, ± SE) 5 4 3 2 3-year gain,.4% (SE,.3%) 6-year gain,.7% (SE,.8%) Log-rank 2P =. 3-year gain,.7% (SE,.3%) 6-year gain,.7% (SE,.8%) Log-rank 2P =.2 Death rates (%/year: total rate rate in women without recurrence) & log-rank analyses Years 2 ( 2 4) Years 2 4 ( 5 7) Years 5+ ( 8+).54 (SE,.7).6 (SE,.7).7 (SE,.39).79 (SE,.8).83 (SE,.8).8 (SE,.48) Rate ratio, from.68 (SE,.4).88 (SE,.4).65 (SE,.34) (O E) / V -3.9 / 36.5-5.8 / 45.5-2.5 / 5.7 3.%.8% 5.8% 4.4% 2.% 5.%.7% 3.3% 3 6 3 6 ( ) ( 2) ( 5) ( 8) ( ) ( 2) ( 5) ( 8) Time Since Allocated Treatments Differed (years) (and approximate time since diagnosis) Death rates (% / year) and log-rank analyses Years 2 ( 2 4) Years 2 4 ( 5 7) Years 5+ ( 8+).57 (64 /,34).4 (55 / 5,298).2 (8 / 76).7 (77 /,962).22 (6 / 5,7).58 ( / 697) Rate ratio, from.82 (SE,.5).85 (SE,.7).68 (SE,.38) (O E) / V -7. / 35. -4.9 / 28.5 -.8 / 4.7 B D Breast Cancer Mortality (%, ± SE) Any Death (%, ± SE) 5 4 3 2 5 4 3 2 Time Since Allocated Treatments Differed (years) (and approximate time since diagnosis) 3-year gain,.% (SE,.5%) 6-year gain, 2.2% (SE,.%) Log-rank 2P =.4 Time Since Allocated Treatments Differed (years) (and approximate time since diagnosis) Death rates (% / year) and log-rank analyses Years 2 ( 2 4) Years 2 4 ( 5 7) Years 5+ ( 8+). (25 /,3) 2.59 (44 / 5,566) 2.2 (7 / 77).47 (66 /,3) 2.95 (6 / 5,455) 3.22 (25 / 776) Rate ratio, from.75 (SE,.).87 (SE,.).66 (SE,.25) (O E) / V 2.9 / 7.5.8 / 74.2-4.3 /.3 Fig 4. Life-table curves of (A) recurrence; (B) breast cancer mortality; (C) death without recurrence; and (D) any death, for estrogen receptor positive patients in trials of 2 to 3 years of tamoxifen and then 2 to 3 years of aromatase inhibitor () versus tamoxifen. O, observed; E, expected; V, variance. point. Of the 638 deaths from any cause, 276 (43%) were of women who had not experienced a recurrence and had died of an unrelated cause. There was a nonsignificant reduction in such deaths (Figs 4C and 6B). The absolute difference in overall mortality was.% (SE.5%) approximately 5 years after diagnosis (Figs 4D and 6C). It seemed to grow larger over the next few years and was 2.2% (SE.%; 2P.4) approximately 8 years after diagnosis, but this too may change with further follow-up in these trials and with the availability of more trials. DISCUSSION Each cohort in this overview included between 9, and, patients, and in aggregate, the two cohorts included almost 9, 54 29 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Meta-Analysis of Aromatase Inhibitors Versus Events/Women events Allocated Allocated Log-rank Variance Ratio of annual event rates Category tamoxifen O E of O E : (A) Time since treatments differed* (trend χ 2 = 8.7, 2P =.3) Years 2 ( 2 4) 87/4,58 33/4,57 6. 8.4.6 (SE.7) (4.%) (6.7%) Years 3+ ( 5+) 7/2,978 76/2,854 7.9 84..9 (SE.) (5.7%) (6.2%) (B) PgR status (χ 2 =.6, 2P >., NS) PgR poor 64/74 7/764 9. 4..63 (SE.2) (9.%) (4.%) PgR+ 24/3,375 294/3,34 3.7 3.5.79 (SE.8) (7.%) (8.8%) PgR unknown 53/429 78/42 6. 3.5.6 (SE.4) (2.4%) (9.4%) (C) Entry age (trend χ 2 =., 2P >., NS) Age < 5 2/9 5/3.2 6.5 (.%) (.5%) 5 59 4/,423 4/,393 7.9 6.8.75 (SE.) (8.%) (.%) 6 69 45/,93 26/,869 38.7 87.6.64 (SE.9) (7.6%) (.6%) 7+ 86/,63 7/,5 9.9 46.5.8 (SE.3) (8.%) (9.6%) (D) Nodal status (N/N v all N+ χ 2 =., 2P >., NS) N/N 96/2,399 23/2,44 3.9 54..77 (SE.2) (4.%) (5.%) N 3 34/565 55/53 2.3 2.4.56 (SE.6) (6.%) (.4%) N4+ 2/24 39/44 8. 3.9.56 (SE.2) (6.%) (27.%) N+ unspec. 56/92 27/897 3.9 86.4.7 (SE.9) (7.%) (23.%) Fig 5. Subgroup analyses of recurrences by (A) time period; (B) progesterone receptor (PgR) status; (C) entry age; (D) nodal status; and (E) tumor grade, for estrogen receptor positive patients in trials of 2 to 3 years of tamoxifen and then 2 to 3 years of aromatase inhibitor () versus tamoxifen. Dashed vertical line corresponds to zero heterogeneity of effect. (*) Values in parentheses are approximate time since diagnosis. O, observed; E, expected; unspec., unspecified; NS, not significant. N unknown 5/58 55/52 2. 26.2 (.%) (.6%) (E) Tumor grade (trend χ 2 =., 2P >., NS) Well differentiated 46/788 49/786.7 22.8 (5.8%) (6.2%) Moderately 8/2,676 248/2,73 37.7 3.6.69 (SE.8) (6.8%) (9.2%) Poorly 88/529 88/492 4.6 4.4.89 (SE.5) (6.6%) (7.9%) Grade unknown 42/55 94/526 25.3 32.5.46 (SE.2) (8.2%) (7.9%) Total 357/4,58 479/4,57 68.9 22.4.7 (SE.6) (7.9%) (.6%) 2P <. 99% or 95% confidence intervals Global heterogeneity (B E): χ 2 = 2.7, P >., NS 4.5..5 2. better better Treatment effect 2P <. patients in only six trials. These relatively large numbers per trial reflected the expectation of only a small incremental improvement in recurrence-free survival given the relatively good prognosis of tamoxifen-treated, ER-positive patients. 2 Overview analysis remains of value for the characterization of the time dependence of the effects on recurrence; for the eventual effects on 5,, and 5-year survival; for subgroup analyses; and for identification of any uncommon favorable or unfavorable adverse effects (not addressed here). www.jco.org 29 by American Society of Clinical Oncology 55 Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Dowsett et al Deaths/Women Al deaths Year code Allocated Allocated Log-rank Variance Ratio of annual death rates and study name Treatment comparison tamoxifen O E of O E : (A) Mortality with recurrence (χ 2 = 2.8, 2P >., NS) 3 94V GABG/ARNO Germany Tam 2 yr, Ana v Tam 3 yr 5/447 22/455 3.4 9. 96Y Austrian BCSG VIII Tam 2 yr, Ana v Tam 3 yr * 5/,82 24/,87 4.2 9.5 98C IES / BIG 2-97 Tam 2 3 yr, Exe v Tam 3 2 yr 7/2,2 33/2,2 8.6 6.4 98R ITA Italy Tam 2 yr, Ana v Tam to 5 yr 2/22 24/224 6. 8.8 Subtotal 59/4,58 23/4,57 22.2 87.7.78 (SE.9) (3.5%) (4.5%) reduction 2P =.2 (B) Mortality without recurrence (χ 2 = 5., 2P >., NS) 3 94V GABG/ARNO Germany Tam 2 yr, Ana v Tam 3 yr 5/447 4/455 4.5 4.7 96Y Austrian BCSG VIII Tam 2 yr, Ana v Tam 3 yr * 56/,82 56/,87.4 27.6 98C IES / BIG 2-97 Tam 2 3 yr, Exe v Tam 3 2 yr 6/2,2 77/2,2 9. 34.2 98R ITA Italy Tam 2 yr, Ana v Tam to 5 yr 5/22 2/224.3.7 Subtotal 27/4,58 49/4,57 3.7 68.2.82 (SE.) (2.8%) (3.3%) reduction 2P =. (C) Any death (χ 2 = 2.6, 2P >., NS) 3 94V GABG/ARNO Germany Tam 2 yr, Ana v Tam 3 yr 2/447 36/455 7.9 3.8 96Y Austrian BCSG VIII Tam 2 yr, Ana v Tam 3 yr * 7/,82 8/,87 5.6 37. 98C IES / BIG 2-97 Tam 2 3 yr, Exe v Tam 3 2 yr 78/2,2 2/2,2 7.7 94.6 98R ITA Italy Tam 2 yr, Ana v Tam to 5 yr 7/22 26/224 4.8.5 Subtotal 286/4,58 352/4,57 36. 56..79 (SE.7) (6.3%) (7.8%) reduction 2P =.4 99% or 95% confidence intervals.5..5 2. better better Fig 6. Trial-specific forest plot of (A) mortality with recurrence; (B) mortality without recurrence; and (C) any death, for estrogen receptor positive patients in trials of 2 to 3 years of tamoxifen (Tam) and then 2 to 3 years of aromatase inhibitor () versus tamoxifen. (*) Random allocation at start of tamoxifen. O, observed; E, expected; NS, not significant; GABG, German Adjuvant Breast Cancer Group; ARNO, Arimidex-Nolvadex; Ana, anastrozole; Austrian BCSG, Austrian Breast and Colorectal Cancer Study Group; IES, Intergroup Exemestane Study; BIG, Breast International Group; ITA, Italian Anastrozole trial; Exe, exemestane. Most of the trials recruited patients known to be hormone receptor (ER or PgR) positive. Patients recorded as having ER-negative/ PgR-positive tumors, who were included in that hormone receptor positive definition, are commonly treated with hormone treatment, but the overview analyses of 5 years of tamoxifen versus no treatment show no significant benefit in this subgroup. 2 The analyses presented here therefore exclude them and are restricted to the ER-positive subset of patients. In recent years, progressively smaller proportions of ER-negative tumors have been reported as PgR positive, and increasingly, the evidence suggests that most tumors recorded as ERnegative/PgR-positive generally represent either a false-negative ER measurement or a false-positive PgR measurement. Therefore, it remains reasonable to treat such patients with endocrine therapy in case some can respond. As a sensitivity analysis, the recurrence reductions were recalculated for all patients with ER-positive and/or PgR-positive disease, and there was no material difference in the findings (data not shown). Each of the third-generation s used in the trials in this metaanalysis has been shown to achieve near-complete inhibition of aromatase in vivo and near-complete suppression of plasma estrogen levels. 3,4 Minor pharmacologic differences exist between the drugs, and there are ongoing trials to determine whether there are resultant differences between them in efficacy and tolerability; attempts to compare different s by indirect between-trial comparisons are inappropriate. However, the two cohorts of trials were analyzed separately because cohort includes patients who would experience relapse on tamoxifen in the first 2 to 3 years after surgery and cohort 2 does not. Direct comparison would be invalid because cohort 2 excludes patients with a particularly unfavorable outcome on tamoxifen. Cohort comprises data from ATAC at 95 months of median follow-up and from BIG -98 at 48 months of median follow-up. Accordingly, approximately two thirds of the statistical information for the overview results in cohort is provided by ATAC, and the preponderant influence of ATAC in cohort is particularly marked in longer follow-up (5 to 8 years). Updated results from BIG -98 5 have recently been published elsewhere. The use of 5 years of an immediately after surgery (cohort ) further reduced recurrence by approximately 23% relative to 5 years of tamoxifen. In cohort 2, when the switch to an is made after 2 to 3 years of tamoxifen, there seems to be a particularly marked reduction in risk of recurrence (by 4%) during the following 3 years. There is evidence that disease relapsing after tamoxifen treatment can show substantial differences in molecular phenotype from the disease at presentation. 6 If this occurs in subclinical disease during adjuvant therapy, it is possible that the disease may take on an enhanced sensitivity to treatment that could underpin this effect, but there are no substantive data to demonstrate this. 56 29 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Meta-Analysis of Aromatase Inhibitors Versus When examined in relation to time on treatment versus time after cessation of treatment, there was some evidence for an attenuation of the reduced risk of recurrence that was stronger for cohort 2. However, for both cohorts, the CIs of the hazard ratios for the period after cessation of therapy were wide, underscoring the importance of continued follow-up and the ongoing clinical trials examining extensions to the duration of therapy. The global test for heterogeneity of the subgroups was not significant in either cohort, although in cohort, the subset according to PgR status individually was significant. Previous trials have demonstrated that 5 years of tamoxifen is highly, and approximately equally, effective in both ER-positive/PgR-positive and ER-positive/PgR-poor disease. 2 The current data seem to suggest that patients reported to be PgR poor derive greater benefit from an. However, the two trials contributing these data differ substantially in their findings according to PgR status. In addition, central analyses of PgR (as opposed to the predominately locally performed receptor analyses in the overview) suggest similar proportional treatment benefit for PgR-positive and PgR-poor populations in both trials. 7,8 Thus, the apparent heterogeneity of effect with regard to PgR status in cohort may well be a chance finding. Widely acknowledged difficulties in reproducing PgR status may contribute to these variable findings. Cohort 2 showed no suggestion of increased effectiveness of s over tamoxifen in the PgR-poor group. The overview indicates that the enhanced efficacy of s over tamoxifen is proportionally similar across all cohorts examined. In the trials studying approximately 5 years of adjuvant tamoxifen versus no tamoxifen in ER-positive disease, the main effect of tamoxifen on breast cancer recurrence was seen in the first 5 years (absolute gain of 2.5% at 5 years and roughly constant thereafter), but the main effect on breast cancer mortality was seen later (absolute gain of 3.3% at 5 years and 9.% at 5 years). 2 In the present trials of s versus tamoxifen, the mean follow-up time thus far available is, in both cohorts, only about 6 years after diagnosis (mean follow-up duration after random assignment was 5.8 years in cohort and 3.9 years in cohort 2), and the absolute further recurrence reduction ( v tamoxifen) 5 years after diagnosis was approximately 3% and highly significant in both cohorts (2.9%, SE.7% in cohort ; and 3.%, SE.6% in cohort 2). Analogy with the recurrence and mortality patterns in the tamoxifen trials suggests that the absolute additional effects ( v tamoxifen) on breast cancer mortality 5 years after diagnosis would be expected to be approximately % in both cohorts, whereas the observed differences were.% (SE.5%) in cohort and.7% (SE.3%) in cohort 2, but with further follow-up, the breast cancer mortality differences may increase. Limited follow-up beyond 5 years after surgery is so far available from either cohort; at this stage, the data from cohort 2 but not from cohort provide evidence for such a further reduction in breast cancer mortality. As with other end points, the mortality results from cohorts and 2 should not be compared directly. There was no indication that the s were associated with an increase in non breast cancer deaths. This is reassuring regarding the overall safety of s relative to tamoxifen. The overall mortality reductions for cohort 2 but not cohort ( v tamoxifen) achieved statistical significance. As recently discussed, 9 the combination of safety analyses (integrated here as non breast cancer deaths) with efficacy analyses (breast cancer specific deaths) to a summary measure (overall mortality) can lead to substantial loss of information by mixing important signals with irrelevant noise. Analyses of cause-specific mortality have not been conducted at this time but will be an important component of later overviews. This overview provides clear evidence that third-generation s achieve modest absolute improvements in breast cancer end points with significant reductions in recurrence in both cohorts and in breast cancer specific mortality in cohort 2. Because the proportional benefits do not vary significantly with factors affecting prognosis, the absolute gain is greater for patients with poorer prognosis. There are also significant differences in adverse effects between tamoxifen and s that have not been reviewed here; use is associated with fewer endometrial cancers and thromboembolic events than tamoxifen but with more arthralgia and fractures. 2,2 The decision on whether to initiate treatment with an or tamoxifen or whether to switch to an after 2 to 3 years of tamoxifen rather than continuing with tamoxifen for 5 years depends on a careful evaluation of these factors in individual patients. This overview has provided greater confidence in the efficacy aspects of this evaluation. AUTHORS DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a U are those for which no compensation was received; those relationships marked with a C were compensated. For a detailed description of the disclosure categories, or for more information about ASCO s conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: Marc Buyse, IDDI (C); Michael Gnant, Merck (C) Consultant or Advisory Role: Mitch Dowsett, AstraZeneca (C), Novartis (C); Jack Cuzick, AstraZeneca (C); Michael Gnant, Novartis (C), Wyeth (C); Raimund Jakesz, Advisory Board Participation (C) Stock Ownership: Marc Buyse, IDDI Honoraria: Mitch Dowsett, AstraZeneca, Novartis; Jack Cuzick, AstraZeneca, Novartis, Eli Lilly; John Forbes, AstraZeneca, Novartis; Judith Bliss, Pfizer; Michael Baum, AstraZeneca; Aman Buzdar, Novartis, AstraZeneca; Marco Colleoni, Novartis; Michael Gnant, AstraZeneca, Novartis, Pfizer, Roche, sanofi-aventis, Amgen; Raimund Jakesz, AstraZeneca, Novartis; Manfred Kaufmann, AstraZeneca, Novartis, Pfizer Research Funding: Mitch Dowsett, AstraZeneca, Novartis; Jack Cuzick, AstraZeneca; Alan Coates, Novartis; Judith Bliss, Pfizer; Aman Buzdar, AstraZeneca; Marco Colleoni, Novartis; Charles Coombes, Pfizer; Michael Gnant, AstraZeneca, Novartis, sanofi-aventis, Roche, Amgen; Raimund Jakesz, Selected research projects Expert Testimony: None Other Remuneration: Alan Coates, Novartis AUTHOR CONTRIBUTIONS Conception and design: Mitch Dowsett, Jack Cuzick, Jim Ingle, Richard Peto Administrative support: Jon Godwin, Christina Davies Provision of study materials or patients: Alan Coates, John Forbes, Judith Bliss, Michael Baum, Aman Buzdar, Marco Colleoni, Charles Coombes, Claire Snowdon, Michael Gnant, Raimund Jakesz, Manfred Kaufmann, Francesco Boccardo Collection and assembly of data: Mitch Dowsett, Jack Cuzick, Jim Ingle, Alan Coates, John Forbes, Judith Bliss, Claire Snowdon, Raimund Jakesz, Jon Godwin, Christina Davies, Richard Peto www.jco.org 29 by American Society of Clinical Oncology 57 Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.
Dowsett et al Data analysis and interpretation: Mitch Dowsett, Jack Cuzick, Jim Ingle, Alan Coates, John Forbes, Judith Bliss, Marc Buyse, Michael Baum, Aman Buzdar, Marco Colleoni, Michael Gnant, Jon Godwin, Christina Davies, Richard Peto Manuscript writing: Mitch Dowsett, Jack Cuzick, Jim Ingle, Alan Coates, John Forbes, Judith Bliss, Marc Buyse, Michael Baum, Aman Buzdar, Marco Colleoni, Charles Coombes, Michael Gnant, Raimund Jakesz, Manfred Kaufmann, Francesco Boccardo, Jon Godwin, Richard Peto Final approval of manuscript: Mitch Dowsett, Jack Cuzick, Jim Ingle, Alan Coates, John Forbes, Judith Bliss, Marc Buyse, Michael Baum, Aman Buzdar, Marco Colleoni, Charles Coombes, Claire Snowdon, Michael Gnant, Raimund Jakesz, Manfred Kaufmann, Francesco Boccardo, Jon Godwin, Christina Davies, Richard Peto REFERENCES. Jordan VC: The science of selective estrogen receptor modulators: Concept to clinical practice. Clin Cancer Res 2:5-53, 26 2. Early Breast Cancer Trialists Collaborative Group: Effects of chemotherapy and hormonal therapy for early breast cancer recurrence and 5-year survival: An overview of the randomised trials. Lancet 365:687-77, 25 3. Early Breast Cancer Trialists Collaborative Group: Effects of adjuvant tamoxifen and of cytotoxic therapy on mortality in early breast cancer: An overview of 6 randomised trials among 28,896 women. N Engl J Med 39:68-692, 988 4. Dowsett M, Howell A: Breast cancer: Aromatase inhibitors take on tamoxifen. Nat Med 8:34-344, 22 5. Smith IE, Dowsett M: Aromatase inhibitors in breast cancer. N Engl J Med 348:243-2442, 23 6. Forbes JF, Cuzick J, Buzdar A, et al: Effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: -month analysis of the ATAC trial Arimidex,, Alone or in Combination (ATAC) Trialists Group. Lancet Oncol 9:45-53, 28 7. Coates AS, Keshaviah A, Thürlimann B, et al: Five years of letrozole compared with tamoxifen as initial adjuvant therapy for postmenopausal women with endocrine-responsive early breast cancer: Update of study BIG -98. J Clin Oncol 25:486-492, 27 8. Jakesz R, Jonat W, Gnant M, et al: Switching of postmenopausal women with endocrineresponsive early breast cancer to anastrozole after 2 years adjuvant tamoxifen: Combined results of ABCSG trial 8 and ARNO 95 trial. Lancet 366:455-462, 25 9. Boccardo F, Rubagotti A, Guglielmini P, et al: Switching to anastrozole versus continued tamoxifen treatment of early breast cancer: Updated results of the Italian Anastrozole (ITA) trial. Ann Oncol 7:vii-vii4, 26 (suppl 7). Coombes RC, Kilburn LS, Snowdon CF, et al: Survival and safety of exemestane versus tamoxifen after 2-3 years tamoxifen treatment (Intergroup Exemestane Study): A randomised controlled trial. Lancet 369:559-57, 27. Cuzick J: Forest plots and the interpretation of subgroups. Lancet 365:38, 25 2. Early Breast Cancer Trialists Collaborative Group: Treatment of Early Breast Cancer. Volume : Worldwide Evidence, 985-99. Oxford, United Kingdom, Oxford University Press, 99 3. Geisler J, Haynes B, Anker G, et al: Influence of letrozole and anastrozole on total body aromatization and plasma estrogen levels in postmenopausal breast cancer patients evaluated in a randomized, cross-over study. J Clin Oncol 2:75-757, 22 4. Geisler J, King N, Anker G, et al: In vivo inhibition of aromatization by exemestane, a novel irreversible aromatase inhibitor, in postmenopausal breast cancer patients. Clin Cancer Res 4:289-293, 998 5. Mouridsen HT, Giobbie-Hurder A, Goldhirsch A, et al: Letrozole therapy alone or in sequence with tamoxifen in women with breast cancer. N ENgl J Med 36:766-776, 29. Cancer Res 69:66s, 29 (suppl) 6. Gutierrez MC, Detre S, Johnston S, et al: Molecular changes in tamoxifen-resistant breast cancer: Relationship between estrogen receptor, HER-2, and p38 mitogen-activated protein kinase. J Clin Oncol 23:2469-2476, 25 7. Viale G, Regan MM, Maiorano E, et al: Prognostic and predictive value of centrally reviewed expression of estrogen and progesterone receptors in a randomized trial comparing letrozole and tamoxifen adjuvant therapy for postmenopausal early breast cancer: BIG -98. J Clin Oncol 25:3846-3852, 27 8. Dowsett M, Allred C, Knox J, et al: Relationship between quantitative estrogen and progesterone receptor expression and human epidermal growth factor receptor 2 (HER-2) status with recurrence in the Arimidex,, Alone or in Combination trial. J Clin Oncol 26:59-65, 28 9. Cuzick J: Primary end-points for randomized trials of cancer therapy. Lancet 37:256-258, 28 2. Lin NU, Winer EP: Advances in adjuvant endocrine therapy for postmenopausal women. J Clin Oncol 26:798-85, 28 2. Lønning PE, Geisler J: Indications and limitations of third-generation aromatase inhibitors. Expert Opin Investig Drugs 7:723-739, 28 58 29 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY Information downloaded from jco.ascopubs.org and provided by Bibliothek der MedUniWien (56) on April 26, 2 from 49.48.4.98. Copyright 2 by the American Society of Clinical Oncology. All rights reserved.