- Email: [email protected]
Long-term effects of anastrozole on bone mineral density: 7-year results from the ATAC trial
original article
Annals of Oncology 22: 857–862, 2011 doi:10.1093/annonc/mdq541 Published online 7 October 2010
Long-term effects of anastrozole on bone mineral density: 7-year results from the ATAC trial R. Eastell1, J. Adams2, G. Clack3, A. Howell4, J. Cuzick5, J. Mackey6, M. W. Beckmann7 & R. E. Coleman8* 1 Academic Unit of Bone Metabolism, University of Sheffield, Sheffield; 2Diagnostic Radiology, Imaging Science and Biomedical Engineering, University of Manchester, Manchester; 3Clinical Development, AstraZeneca, Cheshire; 4The Christie NHS Foundation Trust, University of Manchester, Manchester; 5Cancer Research UK Centre for Epidemiology, Mathematics and Statistics, Queen Mary University of London, Wolfson Institute of Preventive Medicine, London, UK; 6Division of Medical Oncology, Department of Oncology, University of Alberta Cross Cancer Institute, Edmonton, Alberta, Canada; 7Department of Obstetrics and Gynecology, Universita¨t ErlangenNu¨rnberg, Erlangen, Germany; 8Cancer Research Centre, Weston Park Hospital, Sheffield, UK
Received 24 March 2010; revised 9 June 2010; accepted 30 July 2010
patients. In total, 50 patients had evaluable data.
Results: Following anastrozole treatment, the lumbar spine median BMD increased by 2.35% (P = 0.04) and 4.02% (P = 0.0004) at years 6 and 7, while total hip median BMD increased by 0.71% (P = 0.3) and 0.5% (P = 0.8). After tamoxifen treatment, lumbar spine median BMD decreased by 0.79% (P = 0.2) and 0.30% (P = 0.9) at years 6 and 7, while total hip median BMD decreased by 2.09% (P = 0.0003) and 2.52% (P = 0.0002). Patients with a normal BMD or who were osteopenic at 5 years did not become osteoporotic. Conclusions: Anastrozole treatment-related bone loss did not continue into the off-treatment follow-up period. The recovery in lumbar spine BMD and absence of further loss at the hip is consistent with the reduction in the annual rate of fracture observed after treatment cessation in the main ATAC trial. Key words: anastrozole, aromatase inhibitors, bone, bone mineral density, breast cancer, tamoxifen
introduction In postmenopausal women, low estradiol levels are associated with increased bone loss, decreased bone mineral density (BMD) and an increased risk of fracture [1, 2]. As such, cancer treatment-induced bone loss (CTIBL) is an increasing problem for postmenopausal women diagnosed with breast cancer, with spinal and hip fractures resulting in a decreased quality of life [3]. Compared with the selective estrogen receptor modulator, tamoxifen, the third-generation aromatase inhibitors (AIs), such as anastrozole, letrozole and exemestane, have demonstrated superior efficacy and better overall safety in the adjuvant treatment of postmenopausal women with estrogen receptor-positive early breast cancer compared with tamoxifen [4–7]. Although there is general agreement that an AI should be part of adjuvant endocrine therapy, the debate continues over immediate use of an AI versus a sequence/switch strategy [6, 7]. The AIs act by reducing circulating estrogen levels in postmenopausal women. Consequently, one of the major concerns is their effect on bone health and their potential to *Correspondence to: Prof. R. Eastell, Academic Unit of Bone Metabolism, Sorby Wing, Northern General Hospital, University of Sheffield, Sheffield S5 7AU, UK. Tel: +44-114-271-4705; Fax: +44-114-261-8775; E-mail: [email protected]
increase the incidence of cancer treatment-induced osteoporosis and risk of fracture [5, 8–10]. In contrast, tamoxifen has bonesparing effects, probably related to its partial estrogen agonist activity leading to decreased bone resorption [11]. Due to their potentially deleterious effect on BMD, it is important to understand the effects of long-term AI therapy. Therefore, the ‘Arimidex’, Tamoxifen, Alone or in Combination (ATAC) trial (International Standard Randomised Controlled Trial Number 18233230) included a prospectively designed bone sub-protocol to specifically assess the effects of long-term treatment with anastrozole and tamoxifen on bone. Analyses of these data after a median follow-up of 1, 2 [12] and 5 years [8] showed that anastrozole significantly reduced BMD over the 5year treatment period compared with tamoxifen. Here, we present the first assessments of BMD data following completion of 5 years’ initial AI treatment from the ATAC bone sub-study.
patients and methods study design The ATAC trial compared the efficacy and tolerability of anastrozole (1 mg/ day), alone or in combination with tamoxifen and tamoxifen (20 mg/day)
ª The Author 2010. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]
original article
anastrozole and tamoxifen on bone mineral density (BMD) following 5 years of treatment.
Patients and methods: Lumbar spine and total hip BMD were assessed at years 6 and 7 in a total of 71 eligible
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Background: This ‘Arimidex’, Tamoxifen, Alone or in Combination (ATAC) trial sub-study examined the effects of
original article in postmenopausal women with localised early breast cancer. The combination therapy arm was discontinued following the 33-month median follow-up analysis due to lack of additional benefit over tamoxifen alone. Subsequently, all analyses have focused on the monotherapy arms only [13]. The bone sub-study was a randomised, double-blind, multicentre study carried out within the main ATAC trial. Details on the design and the methodology of this sub-study have been described previously [8, 12]. The sub-study was approved by the relevant ethics committees and international review boards and was conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonisation/Good Clinical Practice.
patients
study assessments The follow-up assessed the within-group change in BMD (grams per square centimetre) at the lumbar spine and total hip following treatment completion at 5 years as assessed by dual-energy X-ray absorptiometry, using General Electric Lunar Corp (Madison, WI) or Hologic Inc. (Bedford, MA) densitometers. Measurements were made at 6 and 7 years during the 2-year follow-up period after completion of treatment. The 5-year BMD scans were used as a new baseline. Analysis was carried out locally with quality control checked centrally by Bio-Imaging Technologies Inc. (Newton, PA).
statistical analysis Lumbar spine and total hip BMD were analysed separately using SAS version 8.2. All analyses and summaries were based on treatment first received. Only data for patients for whom there were evaluable BMD measurements at all time points are presented: for the anastrozole-treated patients, n = 21 and n = 23 for the lumbar spine BMD and total hip BMD analysis, respectively. In patients treated with tamoxifen, n = 27 for both lumbar spine and total hip measurements. The difference between the 6- and 7-year scans and the 5-year scans was calculated and reported as median percentage change from year 5 with associated interquartile ranges. In addition, these differences were analysed independently using a paired t-test on the log-transformed BMD values. The results of the statistical analysis were reported as back-transformed geometric means of the ratio of year 5 to year 6 and year 5 to year 7, respectively, with the 95% confidence interval and the P-value for each.
results patients In total, 308 postmenopausal women were recruited to the ATAC bone sub-protocol, and 108 patients in the monotherapy treatment arms completed the bone sub-study (57 patients
858 | Eastell et al.
treated with anastrozole and 51 patients treated with tamoxifen). Eligibility for entry into the extension protocol required documented 5-year scan data and 100 patients had these data (52 anastrozole and 48 tamoxifen). Of these, 4 anastrozole and 1 tamoxifen patients were recorded as osteoporotic at the end of the treatment period and were thus ineligible for entry into the extension protocol [8]. Of the 95 patients who were not osteoporotic and did have recorded scan data at the end of the 5-year study (48 anastrozole and 47 tamoxifen), the total number of osteopenic patients was 62 (35 anastrozole and 27 tamoxifen) and, of these, 25 and 18, respectively, entered the extension sub-study. Of the 33 patients with normal bone scans (13 anastrozole and 20 tamoxifen), 8 and 20, respectively, entered the extension sub-study. There were, therefore, 24 patients with recorded data who did not enter the extension protocol (15 anastrozole and 9 tamoxifen); 10 and 9, respectively, for each treatment group were osteopenic patients, while 5 anastrozole patients and 0 tamoxifen patients had normal BMD measurements [8]. Thus, the bone status of those patients who did not enter the extension study was similar for each treatment group. Overall, 71 patients consented to and were eligible for inclusion in the post-treatment follow-up extension sub-study. A total of 32 patients from the anastrozole group and 38 patients from the tamoxifen group completed the 6-year follow-up visit and 27 and 33 patients completed the 7-year follow-up visit, respectively (Figure 1). Patient demographics in the 71 patients consenting to the off-treatment follow-up were reasonably well balanced across treatment groups (supplementary online table) and were comparable to the entire population included in the 5-year analysis [8].
changes in BMD following completion of treatment From years 5–6, lumbar spine BMD increased in 15/23 (65.2%) of patients in the anastrozole group and in 12/27 (44.4%) patients in the tamoxifen group. By year 7, the proportion of patients reporting an increase in lumbar spine BMD had gone up to 20/23 (87.0%) in the anastrozole group and remained at 12/27 (44.4%) in the tamoxifen group. For total hip BMD, the number of anastrozole-treated patients reporting an increase in BMD was the same at years 6 and 7 (14/21, 66.7%), whereas only 4/27 (14.8%) and 6/27 (22.2%) tamoxifen-treated patients reported an increase in total hip BMD at years 6 and 7, respectively. Compared with the 5-year values, median lumbar spine BMD increased by +2.35% at year 6 (interquartile range 25.34 to 8.19, P = 0.04) and +4.02% at year 7 (interquartile range 26.04 to 14.01, P = 0.0004) in anastrozole-treated women. In the tamoxifen-treated group, there was a decrease in BMD of 20.79% at year 6 (interquartile range 210.61 to 4.35, P = 0.2) and 20.3% at year 7 (interquartile range 27.43 to 10.22, P = 0.9) (Figure 2A). Median total hip BMD for anastrozole-treated women stabilised during the off-treatment follow-up period with +0.71% at year 6 (interquartile range 29.42 to 4.63, P = 0.3) and +0.5% at year 7 (interquartile range 28.74 to 4.27, P = 0.8) compared with year 5. In the tamoxifen group, median BMD
Volume 22 | No. 4 | April 2011
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Patients eligible for the off-treatment follow-up were limited to those who participated in the monotherapy arms of the bone sub-protocol who remained recurrence free, were not osteoporotic at 5 years, had evaluable 5-year bone scans for both lumbar and total hip, gave written informed consent to further follow-up and did not possess any of the criteria for withdrawal. Patients with osteoporosis (defined as a T-score less than 22.5 at the lumbar spine or total hip, according to World Health Organization criteria) and those taking bisphosphonates were excluded from this extension study; those with osteopenia (T-score between 21 and 22.5) were included at the investigators’ discretion. In line with the study protocol, no additional drugs or calcium/vitamin D supplements were offered to the patients during the extra 2-year follow-up period.
Annals of Oncology
original article
Annals of Oncology
had decreased by 22.09% at year 6 (interquartile range 24.34 to 5.70, P = 0.0003) and 22.52% at year 7 (interquartile range 29.5 to 8.02, P = 0.0002) (Figure 2B). There was a statistically significant increase within the anastrozole 1 mg treatment group at years 6 and 7 compared with year 5 (Table 1), while the small decreases in lumbar spine BMD within the tamoxifen 20 mg treatment group at years 6 and 7 were not statistically significant (Table 1). Similarly, the change in total hip BMD at years 6 and 7 compared with year 5 showed a slight, but statistically significant, loss within the tamoxifen 20 mg treatment group (Table 1), while there was a small increase in total hip BMD within the anastrozole 1 mg treatment group at year 6 (and a marginal decrease at year 7) that were not statistically significant (Table 1). Geometric mean BMD values for lumbar spine and total hip from the start of treatment (year 0) are illustrated in Figure 2C and D, respectively. These show a return of bone after the 5year treatment period but not to the baseline levels. In this extension study, no woman who had a normal BMD or who was osteopenic at year 5 became osteoporotic at years 6 or 7, in either treatment group (Table 2). Some patients who were osteopenic at treatment completion (anastrozole n = 9; tamoxifen group n = 4) did not have a bone scan recorded at year 7 (Table 2). Although the reason for this is unclear, a likely explanation is that these patients dropped out of the study after 5 years to receive bisphosphonate treatment.
Volume 22 | No. 4 | April 2011
discussion The primary objective of this follow-up analysis of the ATAC bone sub-study was to provide important additional information on the skeletal health effects of anastrozole after cessation of treatment. In patients entering the follow-up extension of the bone subprotocol, analysis of the two treatment groups at year 7 demonstrated that the greater bone loss seen with anastrozole over tamoxifen during the 5-year active treatment period did not continue into the off-treatment follow-up, and there was evidence of partial recovery in BMD at the lumbar spine and no further loss in BMD at the hip in the anastrozole group. Caution must be taken, however, concerning the generalisability of these findings, due to the low number of patients completing the 7-year follow-up period. In the sub-set of patients completing the off-treatment follow-up with normal BMD at 5 years, or identified as osteopenic at 5 years, none became osteoporotic after adjuvant treatment. The recently published results of the main ATAC study [14] showed similar fracture rates in the anastrozole- and tamoxifen-treated patients during the off-treatment period. These findings are consistent with the recovery in BMD confirmed after treatment completion. Several trials have indicated that other AIs (letrozole and exemestane) can be associated with an increased risk of fractures [15–18] and increased rates of bone loss during the
doi:10.1093/annonc/mdq541 | 859
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Figure 1. Patient disposition (completion or discontinuation).
original article
Annals of Oncology
treatment phase [16, 17, 19]. Similarly, the ATAC bone subprotocol study showed anastrozole-related BMD loss during the treatment phase [8] with a concomitant increased rate of fractures [14], both of which subsequently recovered following cessation of treatment. The ATAC bone sub-study is the first to present long-term follow-up data on BMD beyond 5 years of adjuvant AI therapy. Current treatment guidelines recommend the inclusion of an AI for optimal adjuvant endocrine therapy [20–22]. Bisphosphonates have been shown to prevent and treat osteoporosis in postmenopausal women and to reduce CTIBL. However, it has been reported that their use in the context of adjuvant AI therapy need only be initiated when the risk of fragility fractures is high, such as a previous diagnosis of osteoporosis, age >75 years and additional risk factors for fracture, BMD T-score less than 22.0, or an on-treatment fragility fracture has occurred [4, 23, 24].
860 | Eastell et al.
Despite the relatively low number of evaluable patients, this ATAC bone sub-study reinforces the suggestion that, for patients in whom pre-existing osteopenia is excluded at the start of treatment, or where osteopenia is identified at the end of long-term anastrozole treatment, further monitoring or preventative bone-loss strategies are no longer necessary beyond those normally used for all postmenopausal women [8]. The latest results from the main ATAC trial and bone subprotocol show that, on completion of anastrozole treatment, fragility fracture rates decrease back to levels similar to those observed with tamoxifen, and fractures occurring after treatment completion are not associated with patients becoming osteoporotic after AI treatment. Overall, the prevention of CTIBL in long-term adjuvant AI breast cancer therapy remains a high priority [25]. What these study results demonstrate is that, in contrast to the beneficial effects of anastrozole on breast cancer recurrence which extend
Volume 22 | No. 4 | April 2011
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Figure 2. Median percent change (6interquartile ranges) in bone mineral density (BMD) for (A) lumbar spine and (B) total hip, for the off-treatment follow-up period (years 6 and 7) compared with year 5; mean BMD for (C) lumbar spine and (D) total hip, for the 5-year adjuvant treatment (years 1–5) and off-treatment (years 6 and 7) periods compared with the start of treatment (year 0).
original article
Annals of Oncology
Table 1. Primary analysis of change in lumbar spine and total hip BMD from 5 to 6 and from 5 to 7 years (primary analysis population with data at 5, 6 and 7 years) Within treatment comparisons
Year 5 n gmean (g/cm2)
Year 7 or 6 versus year 5 Ratio of gmean(s) Two-sided 95% CI
P value
23 27
1.042 1.044
23 27
1.025 1.055
1.0164 0.9899
1.0009–1.0321 0.9767–1.0033
0.0385 0.1339
23 27
1.064 1.054
23 27
1.025 1.055
1.0386 0.9995
1.0185–1.0591 0.9813–1.0181
0.0006 0.9588
21 27
0.912 0.934
21 27
0.910 0.949
1.0020 0.9837
0.9864–1.0177 0.9751–0.9923
0.7961 0.0007
21 27
0.907 0.927
21 27
0.910 0.949
0.9964 0.9763
0.9809–1.0122 0.9628–0.9900
0.6384 0.0016
a
Data presented for patients who had evaluable BMD measurements at all visits. BMD, bone mineral density; CI, confidence interval; gmean, geometric mean; n, number of patients.
Table 2. T-scores shift table for lumbar spine and total hip for the overall off-treatment follow-up period (years 5–7, primary analysis population) Status at treatment completion
Shift to
Anastrozole 1 mg (N = 33)
Tamoxifen 20 mg (N = 38)
Normal bone
Normal bone Osteopenic Osteoporotic Not recordeda Normal bone Osteopenic Osteoporotic Not recordeda
8 0 0 0 2 14 0 9
14 3 0 3 3 11 0 4
Osteopenic
Osteoporosis and osteopenia were defined as a T-score less than 22.5 or a T-score between 21 and 22.5 at the lumbar spine or total hip, respectively. T-score is the minimum of the mean lumbar spine and mean total hip T-scores. If either of the BMD measurements was missing, the classification was done on the basis of the single measurement. No patient was osteoporotic at entry into the extension study. a No total hip and lumbar spine bone mineral density measurements due to missing dual-energy X-ray absorptiometry scan or patient discontinued. BMD, bone mineral density.
substantially beyond the cessation of treatment [14], anastrozole-associated BMD loss begins to resolve immediately after treatment cessation and any bone loss associated with anastrozole can be monitored and managed as needed [25–29]. In conclusion, this study is the first to quantify the long-term impact on BMD of AI treatment in women with early breast cancer. Anastrozole-related bone loss appears to be manageable and, therefore, any risk to bone health should be weighed
Volume 22 | No. 4 | April 2011
against its overall efficacy and tolerability profile for the treatment of hormone-sensitive early-stage postmenopausal breast cancer.
funding AstraZeneca, who provided support for the conduct of the study, data collection and project management, to RE; AstraZeneca and Aventis to JC.
acknowledgements We gratefully acknowledge the participating patients and study investigators for their invaluable contribution. We thank Varinia Mun˜oz, PhD, of Complete Medical Communications, who provided medical writing support funded by AstraZeneca.
disclosure RE has acted as a consultant for and received research funding from AstraZeneca. GC is an AstraZeneca employee and holds shares in the company. AH is a member of the Speaker’s Bureau at AstraZeneca. JC is a statistical consultant for AstraZeneca and has received research funding from AstraZeneca and Aventis. JA, JM, MWB and REC have no conflicts of interest to declare.
references 1. Eastell R. Aromatase inhibitors and bone. J Steroid Biochem Mol Biol 2007; 106: 157–161. 2. McCloskey E. Effects of third-generation aromatase inhibitors on bone. Eur J Cancer 2006; 42: 1044–1051. 3. Hoff AO, Gagel RF. Osteoporosis in breast and prostate cancer survivors. Oncology (Williston Park) 2005; 19: 651–658.
doi:10.1093/annonc/mdq541 | 861
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Lumbar spinea 5–6 years Anastrozole 1 mg Tamoxifen 20 mg 5–7 years Anastrozole 1 mg Tamoxifen 20 mg Total hipa 5–6 years Anastrozole 1 mg Tamoxifen 20 mg 5–7 years Anastrozole 1 mg Tamoxifen 20 mg
Year 7 or 6 n gmean (g/cm2)
original article
862 | Eastell et al.
18. McCloskey EV, Hannon RA, Lakner G et al. Effects of third generation aromatase inhibitors on bone health and other safety parameters: results of an open, randomised, multi-centre study of letrozole, exemestane and anastrozole in healthy postmenopausal women. Eur J Cancer 2007; 43: 2523–2531. 19. Lønning PE, Geisler J, Krag LE et al. Effects of exemestane administered for 2 years versus placebo on bone mineral density, bone biomarkers, and plasma lipids in patients with surgically resected early breast cancer. J Clin Oncol 2005; 23: 5126–5137. 20. Goldhirsch A, Glick JH, Gelber RD et al. Meeting highlights: international expert consensus on the primary therapy of early breast cancer 2005. Ann Oncol 2005; 16: 1569–1583. 21. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology V.2. 2008. Breast Cancer. http://www.nccn.org/ (February 2010, date last accessed). 22. Winer EP, Hudis C, Burstein HJ et al. American Society of Clinical Oncology technology assessment on the use of aromatase inhibitors as adjuvant therapy for postmenopausal women with hormone receptor-positive breast cancer: status report 2004. J Clin Oncol 2005; 23: 619–629. 23. Hadji P, Body JJ, Aapro MS et al. Practical guidance for the management of aromatase inhibitor-associated bone loss. Ann Oncol 2008; 19: 1407–1416. 24. Reid DM, Doughty J, Eastell R et al. Guidance for the management of breast cancer treatment-induced bone loss: a consensus position statement from a UK Expert Group. Cancer Treat Rev 2008; 34: S3–S18. 25. Brufsky A. Management of cancer-treatment-induced bone loss in postmenopausal women undergoing adjuvant breast cancer therapy: a Z-FAST update. Semin Oncol 2006; 33: S13–17. 26. Theriault RL. Strategies to prevent chemotherapy-induced bone loss in women with breast cancer. Clin Breast Cancer 2005; 5 (Suppl): S63–S70. 27. Bundred NJ, Campbell ID, Davidson N et al. Effective inhibition of aromatase inhibitor-associated bone loss by zoledronic acid in postmenopausal women with early breast cancer receiving adjuvant letrozole: ZO-FAST Study results. Cancer 2008; 112: 1001–1010. 28. Confavreux CB, Fontana A, Guastalla JP et al. Estrogen-dependent increase in bone turnover and bone loss in postmenopausal women with breast cancer treated with anastrozole. Prevention with bisphosphonates. Bone 2007; 41: 346–352. 29. Van Poznak C, Hannon R, Clack G. Managing cancer treatment-induced bone loss: 24-month results from the study of anastrozole with the bisphosphonate risedronate (SABRE). Presented at the San Antonio Breast Cancer Symposium, San Antonia, TX, 10–14 December 2008.
Volume 22 | No. 4 | April 2011
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
4. Brufsky AM. Managing bone loss in women with early-stage breast cancer receiving aromatase inhibitors. Clin Breast Cancer 2007; 8 (Suppl 1): S22–S34. 5. Perez EA. Safety of aromatase inhibitors in the adjuvant setting. Breast Cancer Res Treat 2007; 105 (Suppl 1): 75–89. 6. Buzdar A, Chlebowski R, Cuzick J et al. Defining the role of aromatase inhibitors in the adjuvant endocrine treatment of early breast cancer. Curr Med Res Opin 2006; 22: 1575–1585. 7. Dowsett M, Cuzick J, Ingle J et al. Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol 2010; 28: 509–518. 8. Eastell R, Adams JE, Coleman RE et al. Effect of anastrozole on bone mineral density: 5-year results from the anastrozole, tamoxifen, alone or in combination trial 18233230. J Clin Oncol 2008; 26: 1051–1057. 9. Brufsky AM. Cancer treatment-induced bone loss: pathophysiology and clinical perspectives. Oncologist 2008; 13: 187–195. 10. Chowdhury S, Pickering LM, Ellis PA. Adjuvant aromatase inhibitors and bone health. J Br Menopause Soc 2006; 12: 97–103. 11. Marttunen MB, Hietanen P, Tiitinen A, Ylikorkala O. Comparison of effects of tamoxifen and toremifene on bone biochemistry and bone mineral density in postmenopausal breast cancer patients. J Clin Endocrinol Metab 1998; 83: 1158–1162. 12. Eastell R, Hannon RA, Cuzick J et al. Effect of an aromatase inhibitor on BMD and bone turnover markers: 2-year results of the Anastrozole, Tamoxifen, Alone or in Combination (ATAC) trial (18233230). J Bone Miner Res 2006; 21: 1215–1223. 13. The ATAC Trialists’ Group. Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: first results of the ATAC randomised trial. Lancet 2002; 359: 2131–2139. 14. The ATAC Trialists’ Group. Effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: 100-month analysis of the ATAC trial. Lancet Oncol 2008; 9: 45–53. 15. Coates AS, Keshaviah A, Thu¨rlimann 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 1-98. J Clin Oncol 2007; 25: 486–492. 16. 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 2007; 369: 559–570. 17. Coleman RE, Banks LM, Girgis SI et al. Skeletal effects of exemestane on bonemineral density, bone biomarkers, and fracture incidence in postmenopausal women with early breast cancer participating in the Intergroup Exemestane Study (IES): a randomised controlled study. Lancet Oncol 2007; 8: 119–127.
Annals of Oncology
Annals of Oncology 22: 857–862, 2011 doi:10.1093/annonc/mdq541 Published online 7 October 2010
Long-term effects of anastrozole on bone mineral density: 7-year results from the ATAC trial R. Eastell1, J. Adams2, G. Clack3, A. Howell4, J. Cuzick5, J. Mackey6, M. W. Beckmann7 & R. E. Coleman8* 1 Academic Unit of Bone Metabolism, University of Sheffield, Sheffield; 2Diagnostic Radiology, Imaging Science and Biomedical Engineering, University of Manchester, Manchester; 3Clinical Development, AstraZeneca, Cheshire; 4The Christie NHS Foundation Trust, University of Manchester, Manchester; 5Cancer Research UK Centre for Epidemiology, Mathematics and Statistics, Queen Mary University of London, Wolfson Institute of Preventive Medicine, London, UK; 6Division of Medical Oncology, Department of Oncology, University of Alberta Cross Cancer Institute, Edmonton, Alberta, Canada; 7Department of Obstetrics and Gynecology, Universita¨t ErlangenNu¨rnberg, Erlangen, Germany; 8Cancer Research Centre, Weston Park Hospital, Sheffield, UK
Received 24 March 2010; revised 9 June 2010; accepted 30 July 2010
patients. In total, 50 patients had evaluable data.
Results: Following anastrozole treatment, the lumbar spine median BMD increased by 2.35% (P = 0.04) and 4.02% (P = 0.0004) at years 6 and 7, while total hip median BMD increased by 0.71% (P = 0.3) and 0.5% (P = 0.8). After tamoxifen treatment, lumbar spine median BMD decreased by 0.79% (P = 0.2) and 0.30% (P = 0.9) at years 6 and 7, while total hip median BMD decreased by 2.09% (P = 0.0003) and 2.52% (P = 0.0002). Patients with a normal BMD or who were osteopenic at 5 years did not become osteoporotic. Conclusions: Anastrozole treatment-related bone loss did not continue into the off-treatment follow-up period. The recovery in lumbar spine BMD and absence of further loss at the hip is consistent with the reduction in the annual rate of fracture observed after treatment cessation in the main ATAC trial. Key words: anastrozole, aromatase inhibitors, bone, bone mineral density, breast cancer, tamoxifen
introduction In postmenopausal women, low estradiol levels are associated with increased bone loss, decreased bone mineral density (BMD) and an increased risk of fracture [1, 2]. As such, cancer treatment-induced bone loss (CTIBL) is an increasing problem for postmenopausal women diagnosed with breast cancer, with spinal and hip fractures resulting in a decreased quality of life [3]. Compared with the selective estrogen receptor modulator, tamoxifen, the third-generation aromatase inhibitors (AIs), such as anastrozole, letrozole and exemestane, have demonstrated superior efficacy and better overall safety in the adjuvant treatment of postmenopausal women with estrogen receptor-positive early breast cancer compared with tamoxifen [4–7]. Although there is general agreement that an AI should be part of adjuvant endocrine therapy, the debate continues over immediate use of an AI versus a sequence/switch strategy [6, 7]. The AIs act by reducing circulating estrogen levels in postmenopausal women. Consequently, one of the major concerns is their effect on bone health and their potential to *Correspondence to: Prof. R. Eastell, Academic Unit of Bone Metabolism, Sorby Wing, Northern General Hospital, University of Sheffield, Sheffield S5 7AU, UK. Tel: +44-114-271-4705; Fax: +44-114-261-8775; E-mail: [email protected]
increase the incidence of cancer treatment-induced osteoporosis and risk of fracture [5, 8–10]. In contrast, tamoxifen has bonesparing effects, probably related to its partial estrogen agonist activity leading to decreased bone resorption [11]. Due to their potentially deleterious effect on BMD, it is important to understand the effects of long-term AI therapy. Therefore, the ‘Arimidex’, Tamoxifen, Alone or in Combination (ATAC) trial (International Standard Randomised Controlled Trial Number 18233230) included a prospectively designed bone sub-protocol to specifically assess the effects of long-term treatment with anastrozole and tamoxifen on bone. Analyses of these data after a median follow-up of 1, 2 [12] and 5 years [8] showed that anastrozole significantly reduced BMD over the 5year treatment period compared with tamoxifen. Here, we present the first assessments of BMD data following completion of 5 years’ initial AI treatment from the ATAC bone sub-study.
patients and methods study design The ATAC trial compared the efficacy and tolerability of anastrozole (1 mg/ day), alone or in combination with tamoxifen and tamoxifen (20 mg/day)
ª The Author 2010. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: [email protected]
original article
anastrozole and tamoxifen on bone mineral density (BMD) following 5 years of treatment.
Patients and methods: Lumbar spine and total hip BMD were assessed at years 6 and 7 in a total of 71 eligible
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Background: This ‘Arimidex’, Tamoxifen, Alone or in Combination (ATAC) trial sub-study examined the effects of
original article in postmenopausal women with localised early breast cancer. The combination therapy arm was discontinued following the 33-month median follow-up analysis due to lack of additional benefit over tamoxifen alone. Subsequently, all analyses have focused on the monotherapy arms only [13]. The bone sub-study was a randomised, double-blind, multicentre study carried out within the main ATAC trial. Details on the design and the methodology of this sub-study have been described previously [8, 12]. The sub-study was approved by the relevant ethics committees and international review boards and was conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonisation/Good Clinical Practice.
patients
study assessments The follow-up assessed the within-group change in BMD (grams per square centimetre) at the lumbar spine and total hip following treatment completion at 5 years as assessed by dual-energy X-ray absorptiometry, using General Electric Lunar Corp (Madison, WI) or Hologic Inc. (Bedford, MA) densitometers. Measurements were made at 6 and 7 years during the 2-year follow-up period after completion of treatment. The 5-year BMD scans were used as a new baseline. Analysis was carried out locally with quality control checked centrally by Bio-Imaging Technologies Inc. (Newton, PA).
statistical analysis Lumbar spine and total hip BMD were analysed separately using SAS version 8.2. All analyses and summaries were based on treatment first received. Only data for patients for whom there were evaluable BMD measurements at all time points are presented: for the anastrozole-treated patients, n = 21 and n = 23 for the lumbar spine BMD and total hip BMD analysis, respectively. In patients treated with tamoxifen, n = 27 for both lumbar spine and total hip measurements. The difference between the 6- and 7-year scans and the 5-year scans was calculated and reported as median percentage change from year 5 with associated interquartile ranges. In addition, these differences were analysed independently using a paired t-test on the log-transformed BMD values. The results of the statistical analysis were reported as back-transformed geometric means of the ratio of year 5 to year 6 and year 5 to year 7, respectively, with the 95% confidence interval and the P-value for each.
results patients In total, 308 postmenopausal women were recruited to the ATAC bone sub-protocol, and 108 patients in the monotherapy treatment arms completed the bone sub-study (57 patients
858 | Eastell et al.
treated with anastrozole and 51 patients treated with tamoxifen). Eligibility for entry into the extension protocol required documented 5-year scan data and 100 patients had these data (52 anastrozole and 48 tamoxifen). Of these, 4 anastrozole and 1 tamoxifen patients were recorded as osteoporotic at the end of the treatment period and were thus ineligible for entry into the extension protocol [8]. Of the 95 patients who were not osteoporotic and did have recorded scan data at the end of the 5-year study (48 anastrozole and 47 tamoxifen), the total number of osteopenic patients was 62 (35 anastrozole and 27 tamoxifen) and, of these, 25 and 18, respectively, entered the extension sub-study. Of the 33 patients with normal bone scans (13 anastrozole and 20 tamoxifen), 8 and 20, respectively, entered the extension sub-study. There were, therefore, 24 patients with recorded data who did not enter the extension protocol (15 anastrozole and 9 tamoxifen); 10 and 9, respectively, for each treatment group were osteopenic patients, while 5 anastrozole patients and 0 tamoxifen patients had normal BMD measurements [8]. Thus, the bone status of those patients who did not enter the extension study was similar for each treatment group. Overall, 71 patients consented to and were eligible for inclusion in the post-treatment follow-up extension sub-study. A total of 32 patients from the anastrozole group and 38 patients from the tamoxifen group completed the 6-year follow-up visit and 27 and 33 patients completed the 7-year follow-up visit, respectively (Figure 1). Patient demographics in the 71 patients consenting to the off-treatment follow-up were reasonably well balanced across treatment groups (supplementary online table) and were comparable to the entire population included in the 5-year analysis [8].
changes in BMD following completion of treatment From years 5–6, lumbar spine BMD increased in 15/23 (65.2%) of patients in the anastrozole group and in 12/27 (44.4%) patients in the tamoxifen group. By year 7, the proportion of patients reporting an increase in lumbar spine BMD had gone up to 20/23 (87.0%) in the anastrozole group and remained at 12/27 (44.4%) in the tamoxifen group. For total hip BMD, the number of anastrozole-treated patients reporting an increase in BMD was the same at years 6 and 7 (14/21, 66.7%), whereas only 4/27 (14.8%) and 6/27 (22.2%) tamoxifen-treated patients reported an increase in total hip BMD at years 6 and 7, respectively. Compared with the 5-year values, median lumbar spine BMD increased by +2.35% at year 6 (interquartile range 25.34 to 8.19, P = 0.04) and +4.02% at year 7 (interquartile range 26.04 to 14.01, P = 0.0004) in anastrozole-treated women. In the tamoxifen-treated group, there was a decrease in BMD of 20.79% at year 6 (interquartile range 210.61 to 4.35, P = 0.2) and 20.3% at year 7 (interquartile range 27.43 to 10.22, P = 0.9) (Figure 2A). Median total hip BMD for anastrozole-treated women stabilised during the off-treatment follow-up period with +0.71% at year 6 (interquartile range 29.42 to 4.63, P = 0.3) and +0.5% at year 7 (interquartile range 28.74 to 4.27, P = 0.8) compared with year 5. In the tamoxifen group, median BMD
Volume 22 | No. 4 | April 2011
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Patients eligible for the off-treatment follow-up were limited to those who participated in the monotherapy arms of the bone sub-protocol who remained recurrence free, were not osteoporotic at 5 years, had evaluable 5-year bone scans for both lumbar and total hip, gave written informed consent to further follow-up and did not possess any of the criteria for withdrawal. Patients with osteoporosis (defined as a T-score less than 22.5 at the lumbar spine or total hip, according to World Health Organization criteria) and those taking bisphosphonates were excluded from this extension study; those with osteopenia (T-score between 21 and 22.5) were included at the investigators’ discretion. In line with the study protocol, no additional drugs or calcium/vitamin D supplements were offered to the patients during the extra 2-year follow-up period.
Annals of Oncology
original article
Annals of Oncology
had decreased by 22.09% at year 6 (interquartile range 24.34 to 5.70, P = 0.0003) and 22.52% at year 7 (interquartile range 29.5 to 8.02, P = 0.0002) (Figure 2B). There was a statistically significant increase within the anastrozole 1 mg treatment group at years 6 and 7 compared with year 5 (Table 1), while the small decreases in lumbar spine BMD within the tamoxifen 20 mg treatment group at years 6 and 7 were not statistically significant (Table 1). Similarly, the change in total hip BMD at years 6 and 7 compared with year 5 showed a slight, but statistically significant, loss within the tamoxifen 20 mg treatment group (Table 1), while there was a small increase in total hip BMD within the anastrozole 1 mg treatment group at year 6 (and a marginal decrease at year 7) that were not statistically significant (Table 1). Geometric mean BMD values for lumbar spine and total hip from the start of treatment (year 0) are illustrated in Figure 2C and D, respectively. These show a return of bone after the 5year treatment period but not to the baseline levels. In this extension study, no woman who had a normal BMD or who was osteopenic at year 5 became osteoporotic at years 6 or 7, in either treatment group (Table 2). Some patients who were osteopenic at treatment completion (anastrozole n = 9; tamoxifen group n = 4) did not have a bone scan recorded at year 7 (Table 2). Although the reason for this is unclear, a likely explanation is that these patients dropped out of the study after 5 years to receive bisphosphonate treatment.
Volume 22 | No. 4 | April 2011
discussion The primary objective of this follow-up analysis of the ATAC bone sub-study was to provide important additional information on the skeletal health effects of anastrozole after cessation of treatment. In patients entering the follow-up extension of the bone subprotocol, analysis of the two treatment groups at year 7 demonstrated that the greater bone loss seen with anastrozole over tamoxifen during the 5-year active treatment period did not continue into the off-treatment follow-up, and there was evidence of partial recovery in BMD at the lumbar spine and no further loss in BMD at the hip in the anastrozole group. Caution must be taken, however, concerning the generalisability of these findings, due to the low number of patients completing the 7-year follow-up period. In the sub-set of patients completing the off-treatment follow-up with normal BMD at 5 years, or identified as osteopenic at 5 years, none became osteoporotic after adjuvant treatment. The recently published results of the main ATAC study [14] showed similar fracture rates in the anastrozole- and tamoxifen-treated patients during the off-treatment period. These findings are consistent with the recovery in BMD confirmed after treatment completion. Several trials have indicated that other AIs (letrozole and exemestane) can be associated with an increased risk of fractures [15–18] and increased rates of bone loss during the
doi:10.1093/annonc/mdq541 | 859
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Figure 1. Patient disposition (completion or discontinuation).
original article
Annals of Oncology
treatment phase [16, 17, 19]. Similarly, the ATAC bone subprotocol study showed anastrozole-related BMD loss during the treatment phase [8] with a concomitant increased rate of fractures [14], both of which subsequently recovered following cessation of treatment. The ATAC bone sub-study is the first to present long-term follow-up data on BMD beyond 5 years of adjuvant AI therapy. Current treatment guidelines recommend the inclusion of an AI for optimal adjuvant endocrine therapy [20–22]. Bisphosphonates have been shown to prevent and treat osteoporosis in postmenopausal women and to reduce CTIBL. However, it has been reported that their use in the context of adjuvant AI therapy need only be initiated when the risk of fragility fractures is high, such as a previous diagnosis of osteoporosis, age >75 years and additional risk factors for fracture, BMD T-score less than 22.0, or an on-treatment fragility fracture has occurred [4, 23, 24].
860 | Eastell et al.
Despite the relatively low number of evaluable patients, this ATAC bone sub-study reinforces the suggestion that, for patients in whom pre-existing osteopenia is excluded at the start of treatment, or where osteopenia is identified at the end of long-term anastrozole treatment, further monitoring or preventative bone-loss strategies are no longer necessary beyond those normally used for all postmenopausal women [8]. The latest results from the main ATAC trial and bone subprotocol show that, on completion of anastrozole treatment, fragility fracture rates decrease back to levels similar to those observed with tamoxifen, and fractures occurring after treatment completion are not associated with patients becoming osteoporotic after AI treatment. Overall, the prevention of CTIBL in long-term adjuvant AI breast cancer therapy remains a high priority [25]. What these study results demonstrate is that, in contrast to the beneficial effects of anastrozole on breast cancer recurrence which extend
Volume 22 | No. 4 | April 2011
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Figure 2. Median percent change (6interquartile ranges) in bone mineral density (BMD) for (A) lumbar spine and (B) total hip, for the off-treatment follow-up period (years 6 and 7) compared with year 5; mean BMD for (C) lumbar spine and (D) total hip, for the 5-year adjuvant treatment (years 1–5) and off-treatment (years 6 and 7) periods compared with the start of treatment (year 0).
original article
Annals of Oncology
Table 1. Primary analysis of change in lumbar spine and total hip BMD from 5 to 6 and from 5 to 7 years (primary analysis population with data at 5, 6 and 7 years) Within treatment comparisons
Year 5 n gmean (g/cm2)
Year 7 or 6 versus year 5 Ratio of gmean(s) Two-sided 95% CI
P value
23 27
1.042 1.044
23 27
1.025 1.055
1.0164 0.9899
1.0009–1.0321 0.9767–1.0033
0.0385 0.1339
23 27
1.064 1.054
23 27
1.025 1.055
1.0386 0.9995
1.0185–1.0591 0.9813–1.0181
0.0006 0.9588
21 27
0.912 0.934
21 27
0.910 0.949
1.0020 0.9837
0.9864–1.0177 0.9751–0.9923
0.7961 0.0007
21 27
0.907 0.927
21 27
0.910 0.949
0.9964 0.9763
0.9809–1.0122 0.9628–0.9900
0.6384 0.0016
a
Data presented for patients who had evaluable BMD measurements at all visits. BMD, bone mineral density; CI, confidence interval; gmean, geometric mean; n, number of patients.
Table 2. T-scores shift table for lumbar spine and total hip for the overall off-treatment follow-up period (years 5–7, primary analysis population) Status at treatment completion
Shift to
Anastrozole 1 mg (N = 33)
Tamoxifen 20 mg (N = 38)
Normal bone
Normal bone Osteopenic Osteoporotic Not recordeda Normal bone Osteopenic Osteoporotic Not recordeda
8 0 0 0 2 14 0 9
14 3 0 3 3 11 0 4
Osteopenic
Osteoporosis and osteopenia were defined as a T-score less than 22.5 or a T-score between 21 and 22.5 at the lumbar spine or total hip, respectively. T-score is the minimum of the mean lumbar spine and mean total hip T-scores. If either of the BMD measurements was missing, the classification was done on the basis of the single measurement. No patient was osteoporotic at entry into the extension study. a No total hip and lumbar spine bone mineral density measurements due to missing dual-energy X-ray absorptiometry scan or patient discontinued. BMD, bone mineral density.
substantially beyond the cessation of treatment [14], anastrozole-associated BMD loss begins to resolve immediately after treatment cessation and any bone loss associated with anastrozole can be monitored and managed as needed [25–29]. In conclusion, this study is the first to quantify the long-term impact on BMD of AI treatment in women with early breast cancer. Anastrozole-related bone loss appears to be manageable and, therefore, any risk to bone health should be weighed
Volume 22 | No. 4 | April 2011
against its overall efficacy and tolerability profile for the treatment of hormone-sensitive early-stage postmenopausal breast cancer.
funding AstraZeneca, who provided support for the conduct of the study, data collection and project management, to RE; AstraZeneca and Aventis to JC.
acknowledgements We gratefully acknowledge the participating patients and study investigators for their invaluable contribution. We thank Varinia Mun˜oz, PhD, of Complete Medical Communications, who provided medical writing support funded by AstraZeneca.
disclosure RE has acted as a consultant for and received research funding from AstraZeneca. GC is an AstraZeneca employee and holds shares in the company. AH is a member of the Speaker’s Bureau at AstraZeneca. JC is a statistical consultant for AstraZeneca and has received research funding from AstraZeneca and Aventis. JA, JM, MWB and REC have no conflicts of interest to declare.
references 1. Eastell R. Aromatase inhibitors and bone. J Steroid Biochem Mol Biol 2007; 106: 157–161. 2. McCloskey E. Effects of third-generation aromatase inhibitors on bone. Eur J Cancer 2006; 42: 1044–1051. 3. Hoff AO, Gagel RF. Osteoporosis in breast and prostate cancer survivors. Oncology (Williston Park) 2005; 19: 651–658.
doi:10.1093/annonc/mdq541 | 861
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
Lumbar spinea 5–6 years Anastrozole 1 mg Tamoxifen 20 mg 5–7 years Anastrozole 1 mg Tamoxifen 20 mg Total hipa 5–6 years Anastrozole 1 mg Tamoxifen 20 mg 5–7 years Anastrozole 1 mg Tamoxifen 20 mg
Year 7 or 6 n gmean (g/cm2)
original article
862 | Eastell et al.
18. McCloskey EV, Hannon RA, Lakner G et al. Effects of third generation aromatase inhibitors on bone health and other safety parameters: results of an open, randomised, multi-centre study of letrozole, exemestane and anastrozole in healthy postmenopausal women. Eur J Cancer 2007; 43: 2523–2531. 19. Lønning PE, Geisler J, Krag LE et al. Effects of exemestane administered for 2 years versus placebo on bone mineral density, bone biomarkers, and plasma lipids in patients with surgically resected early breast cancer. J Clin Oncol 2005; 23: 5126–5137. 20. Goldhirsch A, Glick JH, Gelber RD et al. Meeting highlights: international expert consensus on the primary therapy of early breast cancer 2005. Ann Oncol 2005; 16: 1569–1583. 21. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology V.2. 2008. Breast Cancer. http://www.nccn.org/ (February 2010, date last accessed). 22. Winer EP, Hudis C, Burstein HJ et al. American Society of Clinical Oncology technology assessment on the use of aromatase inhibitors as adjuvant therapy for postmenopausal women with hormone receptor-positive breast cancer: status report 2004. J Clin Oncol 2005; 23: 619–629. 23. Hadji P, Body JJ, Aapro MS et al. Practical guidance for the management of aromatase inhibitor-associated bone loss. Ann Oncol 2008; 19: 1407–1416. 24. Reid DM, Doughty J, Eastell R et al. Guidance for the management of breast cancer treatment-induced bone loss: a consensus position statement from a UK Expert Group. Cancer Treat Rev 2008; 34: S3–S18. 25. Brufsky A. Management of cancer-treatment-induced bone loss in postmenopausal women undergoing adjuvant breast cancer therapy: a Z-FAST update. Semin Oncol 2006; 33: S13–17. 26. Theriault RL. Strategies to prevent chemotherapy-induced bone loss in women with breast cancer. Clin Breast Cancer 2005; 5 (Suppl): S63–S70. 27. Bundred NJ, Campbell ID, Davidson N et al. Effective inhibition of aromatase inhibitor-associated bone loss by zoledronic acid in postmenopausal women with early breast cancer receiving adjuvant letrozole: ZO-FAST Study results. Cancer 2008; 112: 1001–1010. 28. Confavreux CB, Fontana A, Guastalla JP et al. Estrogen-dependent increase in bone turnover and bone loss in postmenopausal women with breast cancer treated with anastrozole. Prevention with bisphosphonates. Bone 2007; 41: 346–352. 29. Van Poznak C, Hannon R, Clack G. Managing cancer treatment-induced bone loss: 24-month results from the study of anastrozole with the bisphosphonate risedronate (SABRE). Presented at the San Antonio Breast Cancer Symposium, San Antonia, TX, 10–14 December 2008.
Volume 22 | No. 4 | April 2011
Downloaded from http://annonc.oxfordjournals.org/ at colostateuniv on July 11, 2015
4. Brufsky AM. Managing bone loss in women with early-stage breast cancer receiving aromatase inhibitors. Clin Breast Cancer 2007; 8 (Suppl 1): S22–S34. 5. Perez EA. Safety of aromatase inhibitors in the adjuvant setting. Breast Cancer Res Treat 2007; 105 (Suppl 1): 75–89. 6. Buzdar A, Chlebowski R, Cuzick J et al. Defining the role of aromatase inhibitors in the adjuvant endocrine treatment of early breast cancer. Curr Med Res Opin 2006; 22: 1575–1585. 7. Dowsett M, Cuzick J, Ingle J et al. Meta-analysis of breast cancer outcomes in adjuvant trials of aromatase inhibitors versus tamoxifen. J Clin Oncol 2010; 28: 509–518. 8. Eastell R, Adams JE, Coleman RE et al. Effect of anastrozole on bone mineral density: 5-year results from the anastrozole, tamoxifen, alone or in combination trial 18233230. J Clin Oncol 2008; 26: 1051–1057. 9. Brufsky AM. Cancer treatment-induced bone loss: pathophysiology and clinical perspectives. Oncologist 2008; 13: 187–195. 10. Chowdhury S, Pickering LM, Ellis PA. Adjuvant aromatase inhibitors and bone health. J Br Menopause Soc 2006; 12: 97–103. 11. Marttunen MB, Hietanen P, Tiitinen A, Ylikorkala O. Comparison of effects of tamoxifen and toremifene on bone biochemistry and bone mineral density in postmenopausal breast cancer patients. J Clin Endocrinol Metab 1998; 83: 1158–1162. 12. Eastell R, Hannon RA, Cuzick J et al. Effect of an aromatase inhibitor on BMD and bone turnover markers: 2-year results of the Anastrozole, Tamoxifen, Alone or in Combination (ATAC) trial (18233230). J Bone Miner Res 2006; 21: 1215–1223. 13. The ATAC Trialists’ Group. Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: first results of the ATAC randomised trial. Lancet 2002; 359: 2131–2139. 14. The ATAC Trialists’ Group. Effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: 100-month analysis of the ATAC trial. Lancet Oncol 2008; 9: 45–53. 15. Coates AS, Keshaviah A, Thu¨rlimann 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 1-98. J Clin Oncol 2007; 25: 486–492. 16. 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 2007; 369: 559–570. 17. Coleman RE, Banks LM, Girgis SI et al. Skeletal effects of exemestane on bonemineral density, bone biomarkers, and fracture incidence in postmenopausal women with early breast cancer participating in the Intergroup Exemestane Study (IES): a randomised controlled study. Lancet Oncol 2007; 8: 119–127.
Annals of Oncology