Treatment of Estrogen Deficiency Symptoms in Women Surviving Breast Cancer, Part 5

May 1, 1999

There are several million breast cancer survivors worldwide. In the United States, 180,000 women were

ABSTRACT: There are several million breast cancer survivors worldwide. In the United States, 180,000 women were diagnosed with breast cancer in 1997, and approximately 97,000 of these women have an extremely low chance of suffering a recurrence of their cancer. With an average age at diagnosis of 60 years and a 25-year expected duration of survival, the current number of breast cancer survivors in the United States may approach 2.5 million women. Since breast cancer is now being detected at an earlier stage than previously and since adjuvant chemotherapy may cause ovarian failure, an increasing number of women are becoming postmenopausal at a younger age after breast cancer treatment. This conference was convened in September 1997 to consider how menopausal breast cancer survivors should be treated at the present time and what future studies are needed to develop improved therapeutic strategies. A total of 47 breast cancer experts and 13 patient advocates participated. The proceedings of the conference are being published in six installments in successive issues of oncology. This fifth part examines the potential role of antiestrogens and selective estrogen receptor modulators (SERMs) in breast cancer patients being treated for estrogen deficiency symptoms. [ONCOLOGY 13(5):721-735, 1999]

Proceedings of a ConferenceHeld at the Boar's Head Inn, Charlottesville,Virginia, September 21-23, 1997


Biology of SERMs and Strategies for Use

V. Craig Jordan, phd: In the 1970s, interest in the development of anticancer therapies intensified. Since a link between estrogens and breast cancer was known, therapy with antiestrogens was postulated to be an appropriate intervention. This was confirmed with the antiestrogen tamoxifen (Nolvadex), which has become the standard treatment for estrogen receptor (ER) positive breast cancer.

Studies of Tamoxifen

While investigating the toxicity of tamoxifen in the 1980s, my group found that, despite the importance of estrogen for preserving bone density, antiestrogens have an estrogenic effect on bone-an effect that I describe as an estrogenic “tickle.” Consequently, these agents maintain bone density. These observations changed the strategic development of the antiestrogens and also changed their name; this drug group is now known as the selective estrogen receptor modulators (SERMs).

All of the first clinical trials used 1 year of tamoxifen therapy, which did not produce survival advantages. Now we know that 5 years of tamoxifen may be the optimal duration for producing survival advantages in women with breast cancer. However, with longer durations of therapy it was essential to evaluate the entire physiologic system. It was of particular interest to determine, initially under laboratory conditions, whether antiestrogens would do harm to bone, but the opposite was found. In fact, both tamoxifen and a related compound, raloxifene (Evista), demonstrated target site–specific effects. They acted like estrogens at certain sites in the body and like antiestrogens at other sites.

In 1989, I suggested that the estrogenic properties of tamoxifen or other antiestrogens might be exploited therapeutically.[1] In the future, tamoxifen (or any antiestrogen) might be considered as a substitute for estrogen. Multiple goals were identified: reduce the risk of endometrial cancer, lower the risk of breast cancer, and prevent bone loss. For the first time it was possible to envision targeted antiestrogens having multifaceted effects throughout the body. The drugs would be able to switch on or off different biological actions. A clinical trial by the Wisconsin group, published in The New England Journal of Medicine,[2] demonstrated that tamoxifen was not harmful to bone, and, in fact, had a statistically significant beneficial effect.

All of the studies conducted in the 1980s confirmed this general pattern of the biological behavior of these compounds in women. The overview analysis of breast cancer clinical trials clearly demonstrated that tamoxifen produces survival advantages in women: Breast cancer patients live longer if they take up to 5 years of tamoxifen.[3] It is the only agent that produces a decrease of nearly 40% in contralateral breast cancer. So, it is antiestrogenic with respect to this parameter. However, it also has an estrogenic “tickle” that has some effect on lowering cholesterol and preserving bone density.

All was not favorable however, particularly with regard to the risk of endometrial cancer associated with tamoxifen treatment. We believed that there was a link between tamoxifen and human endometrial cancer growth. The experiment used a breast cancer implanted on one side of an athymic animal and an endometrial cancer on the other side. Animals were then treated with estrogen and tamoxifen to see whether tamoxifen controls the growth of both tumors. Tamoxifen completely controlled the growth of the breast cancer but caused a significant increase in the growth of the endometrial cancer. Based on this information, our group was the first to urge that the association between tamoxifen and endometrial cancer be investigated further in the clinic.

The fact that tamoxifen had been successfully used to treat endometrial cancer did not exclude the possibility of an increased incidence of endometrial tumors during prolonged tamoxifen therapy for breast cancer. A large cohort of patients taking long-term tamoxifen therapy had to be monitored to resolve this issue.

For the past decade there has been intense interest in this topic, and in 1996 the International Agency for Research on Cancer reviewed all of the relevant data[4]. The agency’s report concluded that tamoxifen does produce rat liver tumors and is associated with an increased incidence of endometrial cancer. The drug was classified as a carcinogen. However, the agency stressed that no woman being treated with tamoxifen for breast cancer should have that treatment stopped because of the conclusions of the working group, since her risk of developing endometrial cancer was far lower than the benefits that a woman with breast cancer might derive from tamoxifen therapy.

Nevertheless, the observation did change the potential use of tamoxifen in the general population. Because of its ability to induce endometrial cancer, tamoxifen was no longer considered a candidate for the prevention of osteoporosis in postmenopausal women without breast cancer. In 1989, our group dropped its plans to test tamoxifen as a preventive agent in high-risk women. We felt that this focus was too narrow, and restated our new strategy as follows: “Important clues were being garnered about the effects of tamoxifen on bones and lipids. So, it was possible that derivatives could be targeted with applications to retard osteoporosis and atherosclerosis. In addition the ubiquitous application of novel compounds to prevent diseases associated with the progressive changes of menopause may, as a side effect, prevent breast cancer.”[1]

What was required was an agent that would switch on or off the estrogen effect at different sites in a woman’s body. The drug should not induce hot flashes; should act as an estrogen in

the brain; should prevent myocardial infarction and breast cancer; should maintain bone density and prevent osteoporosis; and should reduce the risk of endometrial cancer. Raloxifene has many of these desirable properties.

Studies of Raloxifene

Raloxifene has gone through a variety of developmental stages. Initially proposed as a treatment for breast cancer, its development for this use was not pursued, primarily because tamoxifen was so well-established, and major advantages of raloxifene were difficult to identify. However, scientists at Eli Lilly later found that raloxifene was almost exclusively an antiestrogen in the uterus. Raloxifene exhibited much lower uterotropic properties compared to tamoxifen.[5]

While our group was studying the effects of raloxifene on bone in 1987, we also found that the drug was able to prevent breast cancer in animal tumor models.[6,7] This was a critical experiment to develop a new preventive agent. Scientists at Eli Lilly and elsewhere confirmed our 1987 studies, showing that raloxifene maintains bone density and reduces cholesterol in rats. Most importantly, raloxifene was shown to have less estrogenic activity than tamoxifen; in fact, the estrogenic effect of tamoxifen on the uterus can be inhibited in rats by raloxifene.[7]

Studies in humans performed by John Termine and colleagues at Eli Lilly showed that raloxifene produces an increase in bone density, as compared with a decrease in bone density with the control arm.[8] These data thus demonstrated the translation of findings from the laboratory to the clinic.

With regard to the endometrium, we concluded from studies using transvaginal ultrasonography that endometrial thickness in subjects taking raloxifene is indistinguishable from that in subjects taking placebo.[9] That is an encouraging, different effect from that of tamoxifen in that organ.[10] Raloxifene is now available clinically as an FDA-approved drug for the prevention of osteoporosis.

Interim results from a phase II study of raloxifene in patients with advanced breast cancer who were ER positive, provided by Dr. Bill Gradishar at Northwestern and scientists at Eli Lilly, indicated that raloxifene is clearly not an estrogen in the breast.[11] In this study, 18 evaluable patients who had received no previous endocrine therapy were treated with raloxifene. Three showed partial responses for a median of 18 months, and five had prolonged stable disease for 10 months, for an overall response rate of 44%. This is what would be expected for an antiestrogen in breast cancer

Preliminary data obtained from more than 10,000 women who have participated in raloxifene trials in recent years show a 55% reduction in the cumulative incidence of breast cancer and a decrease in endometrial cancer among the groups receiving active treatment.[12] I should emphasize that these early finding must be followed up for the next 3 or 4 years to get a good indication of whether the prevention of breast cancer is a particularly useful effect of this drug.

Thus, we hope that raloxifene, the first of the SERMs (or targeted antiestrogens), will be accepted as a reasonable alternative to hormone replacement therapy (HRT). The primary hope is to prevent osteoporosis by preventing bone loss. Like tamoxifen, raloxifene decreases cholesterol. Since tamoxifen reduces fatal myocardial infarction, in a small Scottish study, clinical studies are being developed to determine whether raloxifene has a similar effect. The goal is to have a preventive maintenance therapy that would be ideal for reducing osteoporosis, coronary heart disease, breast cancer, and endometrial cancer.


In response to questions, Dr. Jordan indicated that insufficient data exist to establish whether raloxifene acts differently on ER-alpha than on ER-beta. It is unclear whether antiestrogens worked only in the presence of estrogen. Dr. Jordan also expressed some concerns about the use of progestins to block the effect of tamoxifen on the uterus, based on animal studies suggesting that progestins may interrupt the anti–breast cancer effects of tamoxifen.

A representative from Eli Lilly reported that the company’s adverse event monitoring system has not uncovered any adverse effects of raloxifene on central nervous system function. An active program is prospectively studying this issue in more detail.

Effects of Tamoxifen on Bone

Trevor Powles, MD, PhD: Bone loss is a serious problem in the treatment of breast cancer. Long-term survival rates in women with inoperable breast cancers are now approaching 70% and greater. More and more patients, particularly premenopausal women, are receiving adjuvant chemotherapy, and most of these women will experience ovarian ablation from it. Osteoporosis in later life is a serious, debilitating illness that must be addressed in these patients.

At the outset of the tamoxifen program, it was unclear whether this drug would act as an antiestrogen or an estrogen on bone. Using dual-energy x-ray absorptiometry, our group demonstrated in 1996 that the bone loss normally seen in postmenopausal women was reversed by tamoxifen, and that women receiving tamoxifen showed a significant bone gain out to 2 years.[13] In women treated with tamoxifen who also received HRT, the addition of hormone replacement produced no detrimental interaction on bone. (This was similar to observations of no adverse effects on cholesterol, fibrinogen, and antithrombin-3.) With regard to lumbar spine and femoral hip bone density, adding HRT to tamoxifen produced a small, but real, additional beneficial effect.

In premenopausal women , transient bone loss was seen for the first 2 years of tamoxifen therapy. This is an antiestrogenic effect in women who have high levels of estrogen, which bring out the antagonistic effect of tamoxifen. Various mechanisms at the receptor level may account for this. For example, after 2 years the receptor system may be able to downregulate in order to minimize that loss.

A remaining question relates to the effects of chemotherapy-induced chemical castration. When a premenopausal woman is given chemotherapy, she suddenly becomes postmenopausal. She will lose 2%, 3%, or 4% of bone during the first 2 to 3 years, but then the curve flattens out, so that the receptors may upregulate in order to recognize the reduced amount of estrogen. Under those circumstances, an effect may occur that is opposite to what would normally happen in premenopausal women. In other words, tamoxifen may act as if the woman were postmenopausal (which she has become as a result of chemotherapy) and enhance bone preservation. However, additional data are required to confirm this hypothesis.


Michael Kleerekoper, MD, commented that bone density at all measured sites predicts fractures three times better than cholesterol predicts heart attacks. The problem with bone density assessment relates to the serial nature of measurements. For some therapeutic agents, such as fluoride, there is a dichotomy between its effect on bone density and its effect on fractures. However, certainly at baseline, bone density is an effective technology for predicting fractures. It does not really matter whether the hip, spine, forearm, heel, or finger is chosen as the measured site. The data have been very consistent across epidemiologic studies.

In reply to a question regarding a possible increase in venous thromboembolism in women treated with tamoxifen plus HRT, Dr. Powles said that he had data on 2,500 women with fairly short follow-up (10 years). Additional time would be needed to demonstrate the absence of an increased incidence of thromboembolism or thrombophlebitis in women treated with tamoxifen alone or combined with HRT.

A participant asked for clarification regarding the combined effects of tamoxifen and estrogen replacement. Dr. Powles responded that if one gives tamoxifen to premenopausal women who have peak endogenous estradiol levels of ³ 1,000 pM/L, the treatment reduces the risk of breast cancer relapse and mortality. Regarding postmenopausal women, it is difficult to see why the use of an estrogen patch, which produces estradiol levels of 200 pM/L, should be different from its use in premenopausal women.

Dr. Powles commented that tamoxifen has been intensely studied for the past 10 years and has been used since 1971. Extensive data have been collected on the millions of women who have received the drug over the last 5 years. Thus, a great deal is known about the actions and side effects of tamoxifen.

Dr. Jordan observed that in premenopausal women, treatment with tamoxifen produces a clear-cut antitumor action despite a huge overcompensation in the production of estrogen. In a postmenopausal woman, adding a small amount of estrogen is really of no consequence with respect to safety. One needs to monitor patients for adverse effects, but one should take into account the huge amount of estrogen circulating endogenously in premeno- pausal women receiving tamoxifen.

Nelson B. Watts, MD, raised the issue of potential discrepancies between follow-up bone density measurements and the antifracture effect. The only situation that he knew of where there had been an improvement in bone density and failure to show an improvement in fracture rate was with the use of fluoride. With the antiresorptive agents-the estrogens, calcitonin, and the bisphosphonates-there seems to be a greater antifracture effect than can be accounted for by the changes in bone mass. Dr. Watts called the improvement in bone mass seen with tamoxifen reassuring, and suggested that the bone density measurements may underestimate the antifracture effect.

Effect of Raloxifene on Bone

Felicia Cosman, MD: Clinical trials involving raloxifene are extensive. Various components of these trials have examined multiple organ systems, including the breast, uterus, heart and bone. Major osteoporosis prevention studies started in about 1993. Osteoporosis treatment trials began soon afterward.

Most of the data I will present were derived from a planned, 2-year interim analysis of a 3-year study.[14] One of the major end points was bone density. Substudies that examined calcium balance and histomorphometric analysis to address the issue of quality of bone will also be mentioned. Fracture efficacy end points in the treatment studies will be assessed from a preliminary analysis. Fracture end points were not included in the prevention studies since the number of patients would have to have been enormous to include this end point in a healthy population in a short-term investigation.

A small calcium dynamics study was conducted by Heaney and Draper[15] over a 6-month period. They examined a placebo group, a 60-mg raloxifene group, and a conjugated estrogen

(Premarin)–medroxyprogesterone regimen group. Both the raloxifene and the HRT regimen were able to induce a positive calcium balance. Both regimens also reduced bone resorption, as well as bone accretion, although the bone accretion was not decreased to as great an extent as was bone resorption. In general, raloxifene had less potent effects on these two bone turnover indices than did estrogen.

In terms of bone quality or histomorphometry, a limited amount of data are available from Dr. Larry Raisz’s group in Connecticut.[16] They examined 11 paired biopsies obtained before and after a 60-mg dose of raloxifene and 13 paired biopsies obtained from patients who received a 150-mg dose. These were small biopsies and could not be subjected to the type of extensive analysis usually performed on biopsy specimens.

The data basically demonstrated no abnormalities in bone mineralization. There were no abnormalities in either osteoblast or osteoclast cells and no evidence whatsoever to suggest that the quality of the bone was abnormal. This finding is considered to be important since an agent such as fluoride can get into the crystal of the bone and perhaps impair bone quality, while at the same time increasing bone mineral content. A separate ongoing histomorphometry study is being performed in postmenopausl women before and 2 years after randomization to 60 or 120 mg/d of raloxifene vs placebo. Results to date show normal bone quality with no osteomalacia, cell damage, woven bone, or marrow fibrosis.[16]

Osteoporosis Prevention Trials

These studies[14] involved healthy women between the ages of 45 and 60 years who had recently become postmenopausal (ie, onset of menopause between 2 and 8 years prior to entry into the study). All subjects had spinal bone density measurements indicating more bone than required to meet the criteria for osteoporosis (ie, a score of 2.5 standard deviations or more below the mean for young normal women). Two major multicenter studies were initiated, one in Europe (11 sites) and the other in North America (9 sites). Both involved 550 to 600 women. The mean ages of patients in the European and North American studies were 55 and 54 years, respectively; patients in both studies had been postmenopausal for a similar number of years (4½ to 5 years); and a similar percentage of patients in both studies had low bone mineral density, but again, not in the osteoporotic range.

The major end points examined in these prevention studies were biochemical markers of bone metabolism at 3-month intervals and bone density measurements every 6 months. The markers of bone metabolism were considered to be important for two reasons. First, they help explain the mechanisms of the drugs. Second, several large epidemiologic studies have shown bone turnover markers to be independently predictive of risks of fracture, especially in the hip. Consequently, decreases in these markers, particularly with antiresorptive agents, could explain, in part, why changes in bone density do not have to be that pronounced in order to have a fairly significant effect on fracture risks. Measurement of bone density, of course, is a major predictor of fracture outcome.

Upon enrollment, all patients were given a calcium supplement and then randomized to one of four arms: 30, 60, or 150 mg of raloxifene, or placebo. Most of the data that I am presenting are from the 60-mg arm, which was the target dose for approval based on multiple efficacy end points from the European study.[17] A slight decrease in urinary C-telopeptide, a selective, specific marker of bone resorption, was observed in the calcium control (placebo) group (Figure 1). In contrast, a much larger diminution in C-telopeptide was seen with raloxifene. The levels remained substantially below baseline; between 30% to 40% throughout the 2-year study (Figure 1).

Since bone formation is coupled to bone resorption, bone formation rates also decrease secondarily, once bone resorption is decreased. Consequently, osteocalcin, a major indicator of bone formation, was reduced between 25% and 30% in the patients treated with raloxifene. The osteocalcin nadir occurred at about 12 months, but there was a substantial reduction throughout the 2-year study.

In the raloxifene group, there was an increase in bone mass, particularly within the first 6 months, as is common with antiresorptive agents (Figure 2). The biggest increments were seen early on, and the ultimate bone mass accrual was not large but was certainly highly significant-approximately 1½%. This compared with a nearly 1% loss in the placebo group. The magnitude of the increment in bone mass was fairly similar to that seen with tamoxifen.[13]

In the femoral neck there was a more delayed increase in bone mass, and the total increment was approximately 1% (Figure 2). A slightly greater loss occurred in the placebo group in the femoral neck region. In the trochanter of the hip, which contains slightly more cancellous bone, raloxifene had a slightly greater effect. This was actually the biggest increment of all of the bone density bone sites measured; approximately a 2% increase with raloxifene, as compared with a ¾% loss in the placebo group. Averaging out the femoral neck and trochanter measurements, the total hip showed an increment of about a 1½% in the raloxifene group vs a 1% loss in the placebo group.

The magnitude of the increases in total hip and spine bone density was similar. This is in contradistinction to what is usually seen with estrogen or the bisphosphonates. With the latter agents, there is a 3:2 or 2:1 increment ratio in the spine vs the hip.

Thus, the effect of raloxifene was considered somewhat unusual, and it was occurring for unclear reasons. In the body as a whole, a 1½% increase was seen in the raloxifene groups vs 1% loss in the placebo arm. If one looks at the degree of increments in the several sites compared to losses in the calcium-alone control group, all of them are between 2 and 2½% at the spine, total hip, and total body. All of these data are from the European study.[17] The North American study showed similar data, but the effects were slightly truncated.[14] A substantial dose-response effect was not observed, at least between 60 and 150 mg/d. It should be emphasized that 60 mg/d is the raloxifene dose approved in the United States.

For the spine and hip, very similar results were found in the North American and European studies with respect to the absence of a dose-response effect.[14,17] Raloxifene was certainly far better than placebo. In the placebo group, 33% of women lost more than 2.5% of bone over a 2-year period, a rate of loss of about 1% per year, compared to the raloxifene-treated subjects. With raloxifene, 12% of patients lost about 1% of bone per year in the spine. Very similar data were obtained in the hip region, where approximately 12% of subjects lost about 1% of bone per year.

PEPI Trial

The results of the osteoporosis prevention trials can be compared with those of the Postmenopausal Estrogen/Progestin Intervention (PEPI) trial,[18] which included different estrogen regimens in postmenopausal women. In that trial, 3% of women who received conjugated estrogens lost more than 1% of bone per year. The corresponding numbers for the raloxifene group were about 3% for the spine and about 7% for the hip-closer for the hip than for the spine.

Data from the osteoporosis prevention trials also can be compared with data from a study of alendronate (Fosamax), 5 mg/d.[19] In this study, the effect of alendronate on the lumbar spine-an actual gain of 3.5%-was larger than that seen for raloxifene or tamoxifen in a similar patient population. However, the effects of alendronate on the total hip and total body (1.9% and 0.7%, respectively) were quite similar to the results seen with raloxifene.

At this point, raloxifene appears to be a reasonable alternative for osteoporosis prevention, with beneficial effects seen throughout the skeleton on bone mass and on bone turnover. To evaluate the efficacy of raloxifene for osteoporosis treatment, however, we need to examine the drug’s effects on fracture occurrence.

The Multiple Outcomes of Raloxifene Evaluation (MORE) is being performed in 7,705 postmenopausal women up to 80 years old (mean age, 66.7 years). All women are receiving supplemental calcium and vitamin D. At baseline, all had osteoporosis (as determined by bone density), and some also had osteoporotic vertebral deformity at baseline. After 2 years of treatment, new vertebral fracture incidence was reduced by approximately 44% with either 60 or 120 mg/d of raloxifene, as compared with placebo. However, there was no significant difference between raloxifene and placebo in the 90 women with nonspinal fractures. Bone mass at the hip and spine increased 2% to 3% with raloxifene vs placebo.[20]


In response to questions, Dr.Cosman said that the dropout rates were very similar between the placebo and raloxifene groups. There were very few side effects. The drug was very well tolerated and very easy to take. Effects on the brain are being examined.

In response to a question regarding the combination of two antiresorptive agents, Dr. Cosman said that she could only speculate, because no data have been published. In the FACET study, a Merck-sponsored study assessing the combination of alendronate plus HRT, it appears that there may be added benefit of the combination with respect to increasing bone density, even though each drug is effective individually.[21] The effect on fracture incidence is unknown.

Thus, there may be some value to combining two antiresorptive agents, particularly those that are less effective than either alendronate or HRT. There may also be a benefit in terms of the additive effect on bone turnover per se.

A number of relatively small studies have examined the combination of calcitonin and etidronate (Didronel), calcitonin and HRT, etidronate and HRT, Dr. Cosman added. Every one of those studies has shown some additive benefit of the combination studied. This occurs, however, with an economic burden and additional side effects. Very few data regarding any of these combined modalities are available in premenopausal women.

Jerilynn Prior, md, commented that any woman who is at risk for losing bone or for having a greater rate of bone loss should add a maximum amount of tolerable calcium to her diet or should take a calcium supplement. She would recommend 2,000 mg/d of calcium and an increase in vitamin D to 800 IU/d. Dr. Cosman said that the optimal dose of calcium really is unknown at this time. Based on available data, the new recommendations of the National Academy of Science state that 1,200 mg/d probably is the best daily calcium dose in the postmenopausal population. Dr. Cosman would not advocate 2,000 mg daily, as that dose has no additional proven benefit.

Dr. Cosman noted that the comment regarding calcium plus vitamin D was very apropos. A recent study published by Dawson-Hughes and colleagues showed that the combined effect of calcium plus vitamin D in an older population (> 65 years old) was highly effective in reducing osteoporotic fractures.[22] The reduction observed, about 60%, is better than any drug currently available for osteoporosis therapy. Thus, Dr.

Cosman believes that calcium plus vitamin D should be a mainstay of therapy, particularly for the older patient. The hope then is to obtain additional benefit from the available pharmacologic agents.


1. Lerner LJ, Jordan VC: Development of antiestrogens and their use in breast cancer: B. F. Cain Memorial Lecture. Cancer Res 50:4177-4189, 1990.

2. Love RR, Mazess RB, Barden HS, et al: Effects of tamoxifen on bone mineral density in postmenopausal women with breast cancer.N Engl J Med 326:852-856, 1992.

3. Early Breast Cancer Trialists’ Collaborative Group: Tamoxifen for early breast cancer: An overview of the randomized trials. Lancet 351(9114):1451-1467, 1998.

4. International Agency for Research on Cancer (IARC): 1996 Tamoxifen. IARC Monogr 66:274-365, 1996.

5. Black LJ, Jones CD, Falcone JF: Antagonism of estrogen action with a new benzothiophene derived antiestrogen. Life Sci 32:1031-1036, 1983.

6. Jordan VC, Phelps E, Lindgren JU: Effects of antiestrogens on bone in castrated and intact female rats. Breast Cancer Res Treat 10:31-35, 1987.

7. Gottardis MM, Jordan VC: The antitumor actions of keoxifene (raloxifene) and tamoxifen in the N-nitrosomethylurea-induced rat mammary carcinoma model. Cancer Res 47:4020-4024, 1987.

8. Delmas PD, Bjarnason NH, Mitlak BH, et al: Effects of raloxifene on bone mineral density, serum cholesterol concentrations and uterine endometrium in postmenopausal women. N Engl J Med 337:1641-1647, 1997.

9. Boss SM, Huster WJ, Neild JA, et al: Effect of raloxifene hydrochloride on the endometrium of postmenopausal women. Am J Obstet Gynecol 177:1458-1464, 1997.

10. Kedar RB, Bourne TH, Powles TJ, et al: Effects of tamoxifen on uterus and ovaries of postmenopausal women in a randomized breast cancer prevention trial. Lancet 342:1318-1321, 1994.

11. Gradishar WJ, Glusman JE, Vogel CL,et al: Raloxifene HCl a new endocrine agent is active in estrogen receptor positive metastatic breast cancer (abstract). Breast Cancer Res Treat 46:53, 1997.

12. Jordan VC, Glusman JE, Eckert S, et al: Raloxifene reduces incident primary breast cancers: Integrated data from multicenter double blind, placebo controlled randomized trials in postmenopausal women (abstract). Breast Cancer Res Treat 50:227, 1998.

13. Powles TJ, Hickish T, Kanis JA, et al: 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, 1996.

14. Eli Lilly, Indianapolis, Indiana: Evista prescribing information.

15. Heaney RP, Draper MW: Raloxifene and estrogen: Comparative bone-remodeling kinetics. J Clin Endocrinol Metab 82:3245-3249, 1997.

16. Ott SM, Oleksik A, Lu Y, et al: Bone histomorphometric results of a 2-year randomized, placebo controlled trial of raloxifene in postmenopausal women. Bone 23(suppl):S295, 1998.

17. Delmas PD, Bjarnason NH, Mitlak BH, et al: Effects of raloxifene on bone mineral density, serum cholesterol concentration, and uterine endometrium in postmenopausal women. N Engl J Med 337:1641-1647, 1997.

18. PEPI Trial Group: Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women. JAMA 273:199-208, 1995.

19. Hosking D, Chilvers CED, Christiansen C, et al: Prevention of bone loss with alendronate in postmenopausal women under 60 years of age.N Engl J Med 338:485-492, 1998.

20. Ettinger B, Black D, Cumming S, et al: Raloxifene reduces the risk of incident vertebral fractures: 24-Month interim analysis (abstract). Osteoporosis Int 8(suppl 3):11, 1998.

21. Lindsay R, Cosman F, Cary DJ, et al: Effect of alendronate added to ongoing hormone replacement therapy in the treatment of postmenopausal osteoporosis (abstract). Osteoporosis Int 8(suppl 3):12, 1998.

22. Dawson-Hughes B, Harris S, Krall E, et al: Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age and older. N Engl J Med 37:670-675, 1997.