ANTIESTROGENIC tamoxifen, widely used for treatment

0021-972X/98/$03.00/0 Vol. 83, No. 4 Journal of Clinical Endocrinology and Metabolism Printed in U.S.A. Copyright 1998 by The Endocrine Society Comparison of Effects of Tamoxifen and Toremifene on Bone Biochemistry and Bone Mineral Density in Postmenopausal Breast Cancer Patients MERJA B. MARTTUNEN, PÄIVI HIETANEN, AILA TIITINEN, AND OLAVI YLIKORKALA Departments of Obstetrics and Gynecology (M.B.M., A.T., O.Y.) and Oncology (P.H.), Helsinki University Central Hospital, FIN-00290 Helsinki, Finland ABSTRACT Antiestrogens are used in the treatment, and sometimes even in the prophylaxis, of breast cancer. Tamoxifen is the most commonly used antiestrogen, but toremifene is gaining in popularity. We compared here the effects of tamoxifen and toremifene on bone metabolism and density in 30 postmenopausal patients with breast cancer, who were randomized to receive tamoxifen (20 mg/day, n 16) or toremifene (40 mg/day, n 14) for 1 yr. Biochemical markers of bone resorption [urinary hydroxyproline, serum cross-linked carboxyterminal telopeptide of type I collagen, urinary cross-linked aminoterminal telopeptide of type I collagen (NTx)] and bone formation [serum bone-specific alkaline phosphatase, osteocalcin, and aminoterminal and carboxyterminal propeptide of type I procollagen] were assessed before treatment and at 6 and 12 months of the antiestrogen regimen. Bone mineral density (BMD) in the lumbar spine and proximal femur (neck, trochanter, and Ward s triangle) was measured using dualenergy x-ray absorptiometry before treatment and at 12 months of treatment. Urinary NTx decreased after 6 months use of tamoxifen (mean fall: ANTIESTROGENIC tamoxifen, widely used for treatment (1), and even for prophylaxis, against breast cancer (2, 3), evidently operates through blocking the estrogen receptors in target tissues (4), although several other biochemical mechanisms of action are possible (5). Tamoxifen also has a number of estrogen agonistic effects that become apparent, e.g. in blood lipids (6) and endometrium (7). One additional benefit of long-term tamoxifen use in postmenopausal women is the bone preservation that has been documented in several placebo-controlled studies (8 12). Toremifene is a derivative of tamoxifen (11), which holds breast cancer treatment potential similar to that of tamoxifen (13 15). Moreover, toremifene seems to possess estrogenagonist effects, e.g. on blood lipids (16, 17) and endometrium (18), but no data exist on the effect of toremifene on bone metabolism. Therefore, we designed this trial to compare the effects of tamoxifen and toremifene on bone biochemical parameters and density of postmenopausal women with breast cancer. Received August 28, 1997. Revision received December 5, 1997. Accepted December 29, 1997. Address all correspondence and requests for reprints to: Merja B. Marttunen, M.D., Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Haartmaninkatu 2, FIN-00290 Helsinki, Finland. 33%) and of toremifene (mean fall: 16%). Use of tamoxifen was associated with a significant decrease in osteocalcin (mean fall: 25%) and aminoterminal propeptide of type I procollagen (mean fall: 22%), whereas toremifene failed to influence these markers. Tamoxifen increased BMD, on average, by 2% in the lumbar spine, 1% in the femoral neck, and 5% in Ward s triangle. Toremifene failed to increase BMD at any site measured, and in contrast, a slight trend toward a fall ( 0.3 to 0.9%) in BMD was seen in patients treated with toremifene. Falls in urinary NTx, from baseline to 6 months, correlated significantly with changes in the lumbar spine BMD (r 0.57, P 0.0002) in the whole patient series. We conclude that tamoxifen (20 mg/day) increases BMD in postmenopausal breast cancer patients, whereas toremifene (40 mg/day) merely prevents the increasing age-associated fall in BMD. More prolonged studies on bone metabolism, comparing these two antiestrogens, are needed; but even now, clinicians should be aware of these differences between tamoxifen and toremifene. (J Clin Endocrinol Metab 83: 1158 1162, 1998) Subjects and Methods With the permission of the local ethics committee, we studied 32 postmenopausal (more than 6 months since their last menstrual period, FSH 40 IU/L) patients with breast cancer. These patients had undergone surgery for stage II breast cancer 6 8 weeks before entering our study. The cancer had spread to the axillary nodes, but a thorough clinical workup showed no other metastases, and no metastases were found during 12 months follow-up. As a part of a large multicenter trial to compare the antitumor activity of tamoxifen and toremifene, the patients were randomized (by sealed envelopes) to start either tamoxifen (20 mg/day) or toremifene (40 mg/day), which are thought to exert similar antiestrogenic effects (13 15). In addition, after the initiation of antiestrogens, all patients received local radiation therapy for 5 weeks. Two patients randomized to toremifene group were excluded because of the condition possibly affecting bone metabolism (1 patient with hypothyroidism, another one with hypertension and use of diuretics). Therefore, 30 patients (16 in the tamoxifen group and 14 in the toremifene group) were accepted for the final study (Table 1). Twelve patients (5 in the tamoxifen group, 7 in the toremifene group) had used various forms of hormone replacement therapy (HRT), until the diagnosis of breast cancer, for 1 15 (mean: 7.7) yr, and HRT in these women had been stopped 6 10 weeks before the start of antiestrogen regimen. No study patient had any other conditions or medications known to affect bone turnover (thyroid or renal disease, use of corticosteroids, diuretics, or excess use of alcohol), but 5 women (1 on tamoxifen, 4 on toremifene) smoked 5 30 (average: 17) cigarettes per day. Patients followed a normal Finnish diet containing calcium (approximately 800-1500 mg daily). Serum and spot urine samples were collected after on overnight fast before initiation of treatment, and 6 and 12 months later. Before each 1158

TAMOXIFEN AND TOREMIFENE EFFECTS ON BONE 1159 TABLE 1. Clinical characteristics of the study population Tamoxifen Toremifene P value n 16 14 NS Age (yr) 61.9 8.7 57.4 8.4 NS Height (cm) 161 4.6 164 7.5 NS BMI (kg/m 2 ) 25.7 4.3 26.0 3.3 NS Time since menopause (yr) 13.1 8,1 9.5 6.8 NS Previous use of HRT 5 7 NS Smoking 1 4 NS Bone resorption markers: Hydroxyproline/creatinine ( mol/mmol) 37.5 27.5 31.9 15.3 NS ICTP ( g/l) 3.9 1.5 4.1 1.8 NS NTx (nmol/mmol) 78.7 31.6 90.6 43.7 NS Bone formation markers: Bone alkaline phosphatase ( g/l) 49.0 18.1 46.1 22.2 NS Osteocalcin ( g/l) 5.4 2.8 5.8 2.9 NS PINP ( g/l) 46.6 15.4 52.8 19.7 NS PICP ( g/l) 110.4 31.4 128.9 45.0 NS Bone Mineral Density (g/cm 2 ): Lumbar spine L1-4 0.897 0.141 0.985 0.130 NS Femur neck 0.767 0.123 0.798 0.130 NS Trochanter 0.661 0.091 0.679 0.093 NS Ward s triangle 0.552 0.119 0.598 0.138 NS Values are mean SD. sampling, the patients were advised to follow a gelatine-restricted diet for 48 h. The samples were kept frozen ( 80 C) until assayed as follows: Bone resorption Bone resorption was evaluated biochemically by measurement of urinary output of hydroxyproline (HOP), with high-performance liquid chromatography (19); the intraassay coefficient of variation of this method was 8.9%. The urinary cross-linked aminoterminal telopeptide of type I collagen (NTx) was measured with an enzyme-linked immunosorbent assay using a monoclonal antibody directed against the N- telopeptide of type I collagen isolated from human urine (20). The intraassay variation in this method was 6.2%. To avoid errors caused by differences in urine dilution, both HOP and NTx data are expressed against creatinine, which was assessed by a routine laboratory method. The concentration of cross-linked carboxyterminal telopeptide of type I collagen (ICTP) in serum was determined by RIA (Telopeptide ICTP, Orion Diagnostica, Espoo, Finland) (21), and the intraassay coefficient of variation for this measurement ranged from 3 9%. Bone formation Bone formation was assessed by measurement of bone-specific alkaline phosphatase in serum, which reflects the activity of bone-forming osteoblasts (22). This was performed by immunoradiometric assay (Tandem-R Ostase, Hybritec Europe, Liege, Belgium). The intraassay variation in this method was below 7%. The concentration of serum osteocalcin was measured by an immunoradiometric assay using antibodies against human osteocalcin (Osteocalcin FEIA, Farmacia CAP System, Uppsala, Sweden). The intraassay coefficient of variation was 7%. The serum aminoterminal (PINP) and carboxyterminal (PICP) propeptide of type I procollagen were determined by RIA (Procollagen Intact PINP RIA Kit, Procollagen PICP RIA Kit, Orion Diagnostica) (23). The intraassay coefficient of variation for the measurement of PINP was 5 8%; and for measurement of PICP, it was 3%. Serum and urine samples of all patients were assessed for a given marker in the same batch of assays to eliminate the effect of interassay variation. Bone mineral density (BMD) BMD in the lumbar spine (LI-LIV) and in different sites of the proximal femur (the femoral neck, the trochanter major region, and Ward s triangle) was measured by dual-energy x-ray absorptiometry (Hologic QDR-1000, Waltham, MA) before treatment and 12 months after the start of treatment. Data are given as density against area (g/cm 2 ) (24). The intraassay coefficient of variation with this method is 0.5% in lumbar spine and 1% in femoral neck, in our department. All data are expressed as the mean sd. The data of the changes in BMD and biochemical markers of bone metabolism during treatment were first subjected to ANOVA; and if this showed a difference, the significance of the difference was tested by paired Student s t test. Comparison between the groups was performed with the unpaired Student s t test. Correlations between two variables were calculated with the Spearman nonparametric correlation analysis. Results Before initiation of the trial, the two study groups were comparable in age, height, body mass index, time since menopause, and smoking (Table 1). The groups also were comparable, before initiation of antiestrogens, with respect to BMD and biochemical markers of bone metabolism (Table 1). Both tamoxifen and toremifene reduced urinary output of NTx at 6 months (mean fall: 33% in the tamoxifen group and 16% in the toremifene group) and 12 months (mean fall: 39% in the tamoxifen group and 16% in the toremifene group) (Fig. 1); no significant changes were seen in the other variables for bone resorption. Use of tamoxifen was accompanied by falls in osteocalcin (mean 25%) and PINP (mean 22%) at 6 months, and this led to a significantly lower level of osteocalcin in women taking tamoxifen than in those taking toremifene, both at 6 and at 12 months of treatment (Fig. 1). In addition, the levels of PINP and PICP fell in women on tamoxifen, causing significantly lower levels of these markers in women on tamoxifen at 6 and 12 months of treatment (Fig. 1). Neither tamoxifen nor toremifene caused any significant changes in HOP, bone-specific alkaline phosphatase, and ICTP (data not shown). Use of tamoxifen was accompanied by increases in BMD (Fig. 2), on average, 2% (P 0.05) in the lumbar spine, 1% in the femoral neck, and 5% in Ward s triangle (P 0.05). Use of toremifene was accompanied either by no change or by small trends toward a fall in BMD, which were on average

1160 MARTTUNEN ET AL. JCE&M 1998 Vol 83 No 4 FIG. 1. Biochemical markers of bone resorption (NTx) and formation (osteocalcin, PINP, and PICP) before and during 6 and 12 months use of tamoxifen (n 16) or toremifene (n 14) treatment in postmenopausal breast cancer patients. Box plots show 10th, 25th, 50th (median), 75th, and 90th percentiles of levels of biochemical markers. a, P 0.05 from baseline; b, P 0.01 from baseline; c, P 0.05 between groups; d, P 0.01 between groups. 0.7% in the lumbar spine, 0.3% in the femoral neck, and 0.9% in the trochanteric region (Fig. 2). In women on tamoxifen and toremifene, changes in BMD differed significantly (P 0.05) (Fig. 2). There were no significant relationships between the changes in BMD and prestudy BMD in either group or in the whole patient series. In the whole study group, patients age correlated negatively with basal lumbar BMD (r 0.55, P 0.004). A similar negative correlation emerged between patients age and BMD measured at different sites of the femur (r 0.40 to 0.54, P 0.01 0.03). Moreover, the time spent in menopause correlated negatively with BMD (no data shown). Body mass index and BMD, at any site measured, showed no relationship. Moreover, no differences in BMD or in biochemical markers emerged between women who had (n 12) or had not (n 18) used HRT (data not shown). The five smokers had higher BMD in the lumbar spine (1.062 0.138 vs. 0.913 0.13 g/cm 2, P 0.05), in the femoral neck (0.91 0.111 vs. 0.756 0.108 g/cm 2, P 0.01), and in Ward s triangle (0.683 0.153 vs. 0.552 0.114 g/cm 2, P 0.05) than did nonsmokers (n 25). The effect of toremifene on BMD was not altered by smoking, because BMD in smokers, at all sites measured, tended to show a smaller change than did BMD in nonsmokers during toremifene intake, but the small number of women involved precludes a more detailed analysis. Discussion Tamoxifen and toremifene compete for preference in treatment of breast cancer (13 15). Because these agents are often used for years, comparative data on their effects on bone are interesting and of potential clinical importance. Our comparison of these agents presents strong evidence that tamoxifen (20 mg/day) has more favorable effects on bone than does toremifene (40 mg/day) for the postmenopausal breast

TAMOXIFEN AND TOREMIFENE EFFECTS ON BONE 1161 FIG. 2. Change in BMD (% of initial) at 12 months in tamoxifen- or toremifene-treated breast cancer patients. Box plots show 10th, 25th, 50th (median), 75th, and 90th percentiles of changes in BMD. a, P 0.05 from baseline; b, P 0.05 between groups. cancer patients who constitute the large majority of those on an antiestrogen regimen today (25). After surgery and radiation therapy, our patients used only antiestrogens as an adjuvant therapy, common practice in the present-day oncology. We assessed BMD in the lumbar spine and in different sites of the proximal femur, where estrogen-sensitive changes in BMD are most rapidly seen (26, 27). Because almost no data exist on the biochemical effects of antiestrogens on bone, we assessed several biochemical markers that are thought to reflect bone degradation or formation (28, 29). This allowed us to estimate whether changes in BMD at 12 months of treatment could have been predicted by changes in biochemical markers 6 months earlier. We can confirm the previous finding that tamoxifen increases BMD in patients with breast cancer (8 11); this increase was significant in the lumbar spine and Ward s triangle; but also, other sites in the proximal femur showed a clear trend toward an increase in BMD after use of tamoxifen for 12 months. This increase was preceded by significant falls in the urinary output of NTx and in serum levels of osteocalcin and PINP, 6 months earlier. This implies that the bone-restoring effect of tamoxifen causes biochemical changes at least 6 months before rises in BMD become detectable. This theory is supported by a significant relationship between individual changes in NTx, ICTP, and PINP, and those in BMD. Because toremifene is chemically and pharmacologically closely related to tamoxifen (30, 13 15), we expected that the dose of tamoxifen (20 mg/day) and that of toremifene (40 mg/day), which are thought to be equipotent (13 15), to cause similar effects on bone. Therefore, it was a surprise that although the effects of tamoxifen and toremifene on biochemical markers are rather similar in direction, toremifene did not increase BMD (which, in effect, remained virtually unchanged during the 1-yr trial). For ethical reasons, we could not include a placebo group in our study; but from previous studies, it is well established that postmenopausal women, similar to those recruited in our study, who are using no HRT lose approximately 2 3% of BMD in the lumbar spine or femoral neck each year (31, 32, 26). Judging from these figures, we may assume that although toremifene did not increase BMD, it prevented an increasing age-related fall in BMD. Thus, toremifene (40 mg/day) too has a bone-preserving effect, but this is significantly weaker than that of tamoxifen (20 mg/day). It is noteworthy also that biochemical markers in women on toremifene did not indicate bonepreserving changes similar to those that occurred in women on tamoxifen, which agrees with the conclusion that toremifene has a weaker restoring effect on bone than does tamoxifen. The reason for this difference between tamoxifen and toremifene is unknown, but it may be possible that estrogen receptors in bone cells (33) are not similarly stimulated by the estrogenic agonistic effects of toremifene as by the effects of tamoxifen. No previous data exist on the effects of tamoxifen or toremifene on NTx or other biochemical bone markers. Because NTx is now regarded as one of the most reliable indices of bone degradation (29), our demonstration of a clear relationship between falls in NTx output at 6 months and changes in BMD, 6 months later, in breast cancer patients using tamoxifen or toremifene, may therefore be of clinical significance. In the clinical routine, it might be worthwhile to assess NTx output before and after 6 months of an antiestrogen regimen. A clear fall in NTx may be seen as a predictor of a future increase in BMD and may reassure both the patient and physician that her antiestrogen regimen may increase BMD. Clearly, our data may call for further comparisons on larger numbers of patients using various regimens of tamoxifen or toremifene for longer periods than 1 yr, but already at this phase, clinicians should be informed of this difference. References 1. Early Breast Cancer Trialists Collaborative Group. 1992 Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy. Lancet. 339:1 15:71 85. 2. Powles TJ. 1992 The case for clinical trials of tamoxifen for prevention of breast cancer. Lancet. 340:1145 1147. 3. Jordan VC. 1995 Alternate antiestrogens and approaches to the prevention of breast cancer. J Cell Biochem. 22:51 57. 4. Coezy E, Borgna JL, Rochefort H. 1982 Tamoxifen and metabolites in MCF-7 cells: correlation between binding to estrogen receptor and inhibition of cell growth. Cancer Res. 42:317 323. 5. Colletta AA, Benson JR, Baum M. 1994 Alternative mechanisms of action of anti-oestrogens. Breast Cancer Res Treat. 31:5 9. 6. Bagdade JD, Wolter J, Subbaiah PV, Ryan W. 1990 Effects of tamoxifen on plasma lipids and lipoprotein lipid composition. J Clin Endocrinol Metab. 70:1132 1135. 7. Neven P, Shepherd JH, Lowe DG. 1993 Tamoxifen and gynaecologist. Br J Obstet Gynecol. 100:893 897. 8. Love RR, Mazess RB, Barden HS, et al. 1992 Effects of tamoxifen on bone mineral density in post-menopausal women with breast cancer. N Engl J Med. 326:852 856. 9. Ward RL, Morgan G, Dalley D, Kelly PJ. 1993 Tamoxifen reduces bone turnover and prevents lumbar spine and proximal femoral bone loss in early postmenopausal women. Bone Miner. 22:87 94. 10. Grey AB, Stapleton JP, Evans MC, Tatnell MA, Ames RW, Reid IR. 1995 The effect of the antiestrogen tamoxifen on bone mineral density in normal late postmenopausal women. Am J Med. 99:636 641. 11. Powles TJ, Hickish T, Kanis JA, Tidy A, Ashley S. 1996 Effect of tamoxifen on bone mineral density measured by dual-energy X-ray absorptiometry in healthy premenopausal and postmenopausal women. J Clin Oncol. 14:78 84. 12. Kangas L. 1990 Introduction to toremifene. Breast Cancer Res Treat. [Suppl] 16:3 7. 13. Stenbygaard LE, Herrstedt J, Thomsen JF, Svendsen KR, Engelholm SA, Dombernowsky P. 1993 Toremifene and tamoxifen in advanced breast cancer - a double-blind crossover trial. Breast Cancer Res Treat. 25:57 63. 14. Hayes DF, Van Zyl JA, Hacking A, et al. 1995 Randomized comparison of

1162 MARTTUNEN ET AL. JCE&M 1998 Vol 83 No 4 tamoxifen and two separate doses of toremifene in postmenopausal patients with metastatic breast cancer. J Clin Oncol. 13:2556 2566. 15. Howell A, Downey S, Anderson E. 1996 New endocrine therapies for breast cancer. Eur J Cancer 32:A576 A588. 16. Gylling H, Pyrhönen S, Mäntylä E, Mäenpää H, Kangas L, Miettinen T. 1995 Tamoxifen and toremifene lower serum cholesterol by inhibition of delta 8-cholesterol conversion to lathosterol in women with breast cancer. J Clin Oncol. 13:2900 2905. 17. Saarto T, Blomqvist C, Ehnholm C, Taskinen MR, Elomaa I. 1995 Antiatherogenic effects of adjuvant antiestrogens; a randomized trial comparing the effects of tamoxifen and toremifene on plasma lipid levels in postmenopausal women with node-positive breast cancer. J Clin Oncol. 14:429 433. 18. Tomas E, Kauppila A, Blanco G, Apaja-Sarkkinen M, Laatikainen T. 1995 Comparison between the effects of tamoxifen and toremifene on the uterus in postmenopausal breast cancer patients. Gynecol Oncol. 59:261 266. 19. Turpeinen U, Pomoell UM. 1985 Liquid-chromatographic determination of total hydroxyproline in urine. Clin Chem. 31:828 830. 20. Hanson DA, Weis MAE, Bollen AM, Maslan SL, Singer FR, Eyre DR. 1992 A specific immunoassay for monitoring human bone resorption: quantitation of type I collagen cross-linked N-telopeptides in urine. J Bone Miner Res. 7:1251 1258. 21. Risteli J, Elomaa I, Niemi S, Novamo A, Risteli L. 1993 Radioimmunoassay for the pyridinoline cross-linked carboxyterminal telopeptide of type I collagen: a new serum marker of bone collagen degradation. Clin Chem. 39:635 640. 22. Epstein S. 1988 Serum and urinary markers of bone remodelling: assessment of bone turnover. Endocr Rev. 9:437 449. 23. Melkko J, Kauppila S, Niemi S, et al. 1996 Immunoassay for intact aminoterminal propeptide of human type I procollagen. Clin Chem. 42:947 954. 24. De Boer H, Blok GJ, Van Lingen A, Teule GJJ, Lips P, Van Der Veen EA. 1994 Consequences of childhood-onset growth hormone deficiency for adult bone mass. J Bone Miner Res. 9:1319 1326. 25. Jaiyesimi IA, Buzdar AU, Decker DA, Hortobagyi GN. 1995 Use of tamoxifen for breast cancer: twenty-eight years later. J Clin Oncol. 13:513 529. 26. Turner RT, Riggs BL, Spelsberg TC. 1994 Skeletal effects of estrogen. Endocr Rev. 15:275 300. 27. Christiansen C. 1996 Hormone replacement therapy and osteoporosis. Maturitas. [Suppl] 23:71 76. 28. Calvo MS, Eyre DR, Gundberg CM. 1996 Molecular basis and clinical application of biological markers of bone turnover. Endocr Rev. 17:333 368. 29. Garnero P, Shih WJ, Gineyts E, Karpf DB, Delmas PD. 1994 Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate treatment. J Clin Endocrinol Metab. 79:1693 1700. 30. Kangas L. 1992 Agonistic and antagonistic effects of antiestrogens in different target organs. Acta Oncol. 31:143 146. 31. Prince RL, Smith M, Dick IM, et al. 1991 Prevention of postmenopausal osteoporosis. A comparative study of exercise, calcium supplementation and hormone-replacement therapy. N Engl J Med. 325:1189 1195. 32. Hansen MA, Overgaard K, Riis BJ, Christiansen C. 1991 Role of peak bone mass and bone loss in postmenopausal osteoporosis: 12 year study. BMJ. 303:961 964. 33. Lindsay R. 1996 The oestrogen receptor in bone - evolution of our knowledge. Br J Obstet Gynecol. 103:16 19.