| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Medical Oncology Clinical Research Unit (J.E.-W., J.Z.) and Surgical Oncology Branch (D.D.), National Cancer Institute, and Clinical Center (J.C.R., D.V., D.L., S.G., C.C.), National Institutes of Health, Bethesda, Maryland 20892; and Genome Institute of Singapore (E.T.L.), Genome, Singapore 138672
Address all correspondence and requests for reprints to: Jennifer Eng-Wong, Medical Oncology Clinical Research Unit, National Cancer Institute, Building 10, Room 12N226, 9000 Wisconsin Avenue, Bethesda, Maryland 20892. E-mail: engwongj{at}mail.nih.gov.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Objective: We evaluated the effect of raloxifene in premenopausal women at increased risk for breast cancer on bone mineral density (BMD).
Design: This was a phase II clinical trial.
Setting: This study was conducted at an academic medical center.
Participants: Thirty-seven premenopausal women at increased risk for breast cancer enrolled in the trial. Thirty subjects began treatment and 27 were evaluable.
Intervention: Raloxifene (60 mg daily) and elemental calcium (500 mg daily) were given for 2 yr. Subjects were followed up off medications for 1 yr.
Main Outcome Measure: The primary end point was the intrasubject percent change in BMD at 1 yr measured by dual-energy x-ray absorptiometry.
Results: The mean baseline lumbar spine density was 1.027 g/cm2. Lumbar spine density decreased 2.3% at 1 yr (P < 0.00001) and 3.5% at 2 yr (P < .00001). Percent change from yr 2 to 3 was +1.4%. The mean baseline total hip bone density was 0.905 g/cm2. Total hip density decreased 0.3% at 1 yr and 1.0% at 2 yr (P = 0.033). Percent change from yr 2 to 3 was +1.7%.
Conclusions: Raloxifene use is associated with a decrease in BMD in premenopausal women at increased risk for breast cancer. The clinical significance of this decrease is unknown and is attenuated with stopping raloxifene.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The SERMs raloxifene and tamoxifen exhibit agonist and antagonist estrogenic effects. In the breast they have antagonist effects and reduce the incidence of breast hyperplasia and invasive breast cancer (2, 3). Both agents have agonist effects on lipid profiles, decreasing low-density lipoprotein (LDL) and total cholesterol levels (4, 5). While these drugs are similar, they also have unique effects. Raloxifene does not stimulate endometrial growth as tamoxifen does (6). Moreover, the effects in pre- and postmenopausal women can be different. For example, while tamoxifen increases bone mineral density (BMD) in postmenopausal women, it has the opposite effect in premenopausal women (7). Because bone loss occurs with increasing age, agents that may hasten this process should be evaluated carefully when they are being considered for use in the prevention setting.
We assessed the effects of raloxifene in premenopausal women at increased risk for invasive breast cancer. Our primary goal was to determine the effect of raloxifene on BMD as measured by dual-energy x-ray absorptiometry (DEXA) scan after 1 yr on drug. Our hypothesis was that raloxifene would cause minimal bone loss in premenopausal women. Secondary end points assessed include: effects on serum bone turnover markers, lipids, fibrinogen, subject reported side effects, and quality of life. Effects on the IGF pathway and leptin have been previously reported (8). Reports on the effects on hormones and gynecological changes are in preparation (Premkumar, A., P. Stratton, personal communication) (9).
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Premenopausal women at high risk for invasive breast cancer were enrolled on protocol 98-C-0123, a phase II trial of raloxifene in premenopausal women at high risk for developing invasive breast cancer from December 1998 to January 2002. This study was approved by the National Cancer Institute Institutional Review Board, and all participants provided written informed consent. Eli Lilly & Co. (Indianapolis, IN) sponsored the trial. Eligible subjects included women between the ages of 23 and 47 yr with an increased risk of breast cancer by at least one of the following factors: Gail model risk assessment 1.7% or greater over 5 yr (10), a family history consistent with hereditary breast cancer (11), a diagnosis of lobular carcinoma in situ, atypical ductal hyperplasia, or locally treated ductal carcinoma in situ. Subjects were required to have had regular menstrual cycles (defined as 26–35 d) for the 6 months preceding enrollment in the trial. Premenopausal status was additionally verified by a FSH level less than 20 mIU/ml measured during the early follicular or luteal phase. After enrolling on the trial and before starting raloxifene, menstrual regularity was again confirmed during a 1- to 2-month run-in period during which menstrual cycle length was verified by patient report, and ovulation was confirmed by luteal phase progesterone levels (2–20 ng/ml). Women who were not ovulatory during the run-in period were taken off study. Exclusion criteria included the use of any other hormonal medications (birth control pills, hormone replacement therapy, or tamoxifen) or a history of deep venous thrombosis or pulmonary embolism. Women were required to use nonhormonal birth control methods throughout the course of the study.
Study drugs
Subjects were prescribed raloxifene 60 mg/d for 2 yr and elemental calcium 500 mg once daily in the form of calcium carbonate. Raloxifene was supplied by Eli Lilly & Co. Calcium was obtained from commercial sources. Compliance was assessed by interviews; subjects maintained calendars and random serum drug levels.
BMD assessment
BMD was assessed with DEXA scans of the lumbar spine and total hip. All DEXA scans were performed on the same scanner (QDR4500; Hologic, Bedford MA). To assess instrument precision, BMD measurement of an anthropomorphic spine phantom was performed daily, and the coefficient of variation of the instrument was less than 0.4% as evaluated over each 6-month period. Values that were outside the manufacturer’s guidelines of ± 1.5% prompted immediate correction by the manufacturer’s service department. Measurements were collected before starting raloxifene; at 6, 12, and 24 months on the study drug; and again 1 yr after stopping the study drug. At all time points, the lumbar spine was measured twice at the same sitting, whereas the total hip was measured once. For the short-term precision determination, duplicate measurements of the lumbar spine with repositioning of the patient between each scan yielded a coefficient of variation of 1.1%.
Laboratory analyses
Serum for bone turnover markers (BTMs) was collected at baseline and 12 months on the drug. Samples were drawn during the early follicular phase of the menstrual cycle. Serum was stored frozen at –70 C until all samples were collected. Samples were then batched, and paired samples were evaluated. Bone-specific alkaline phosphatase was determined by immunoradiometric assay (Access analyzer; Beckman, Fullerton, CA) and serum N-telopeptides (NT) by ELISA (Nordik Bioscience, Herlev, Denmark). Both assays were performed at Esoterix, Inc. (Calabasas Hills, CA). Osteocalcin was evaluated by a two-site chemiluminescence immunoassay (Nichols Advantage analyzer, Nichols Institute Diagnostics, San Clemente, CA). Interassay coefficients of variation were all less than 10%.
Fasting serum for lipids and fibrinogen were collected at baseline; after 3, 12, and 24 months on the drug; and after stopping raloxifene for 1 yr. Samples were collected and run on the clinic visit day at the National Institutes of Health Clinical Center Laboratory. Total cholesterol and triglycerides were evaluated with an enzymatic assay using a Beckman Coulter reagent (Beckman Coulter, Fullerton, CA). High-density lipoprotein (HDL) was determined using a direct homogeneous assay using Beckman Coulter reagent; LDL was analyzed using a direct homogeneous assay using Genzyme N-geneous reagent (Genzyme, Cambridge, MA). Apolipoprotein A-I and B levels were determined using a nephelometric method using IMMAGE analyzer (Beckman Coulter). Fibrinogen was measured with a functional clotting time assay.
Adverse events
Subjects were asked to track adverse events on their calendars and were further evaluated at all clinic visits for adverse events. These are reported as per common toxicity criteria 2.0 grading criteria (12).
Quality-of-life measurements
Two instruments were used to assess the effect of raloxifene on quality of life: the Menopause Specific Quality of Life Questionnaire (MENQOL) and MOS 36-Item Short Form Health Survey (SF-36). Both tools are self-administered questionnaires and were completed by subjects before starting raloxifene; at 6, 12, and 24 months on the study drug; and again 1 yr after stopping the study drug. The MENQOL consists of four domains scored separately: vasomotor, psychosocial, physical, and sexual. The SF-36 consists of eight subscales summarized as the physical component score and the mental component score. Raw scores are converted to normal-based scoring to allow for comparison with average populations (13).
Statistical analysis
For BMD, each subject was compared with her own baseline to calculate the intrasubject percent change. Significance of change in BMD was assessed using the Wilcoxon signed rank test. The primary outcome was the percentage change from baseline in the lumbar spine BMD of the subjects after 1 yr of raloxifene therapy. We considered the agent to be acceptable for further study if the probability of a 7% BMD loss was less than 5% and unacceptable if it was greater than 20%. We had error probabilities of less than 20% for mistakenly rejecting an acceptable treatment and less than 10% for not rejecting an unacceptable treatment.
Correlation between lumbar spine and BTM was evaluated using the exact Spearman rank correlation method. For lipid evaluations, each value was compared with the subject’s baseline, and repeated-measures ANOVA was used to test for significance. Results were corrected for multiple comparisons using the Hochberg method. For quality-of-life measures, significance was determined by the Wilcoxon signed rank test for the MENQOL. SF-36 scores were normalized to the general U.S. population and analyzed using repeated-measures ANOVA.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Thirty-seven women enrolled in the trial, and of these, seven did not start the drug. During the run-in phase, these seven subjects withdrew for the following reasons: two for intercurrent medical issues, one refused further evaluations after signing consent, one for logistical reasons, one decided on a surgical prevention intervention, one for anovulation during the run-in period, and one for work-up of a suspicious breast nodule. Of the 30 remaining subjects, one stopped raloxifene after 22 d and one was lost to follow-up (Fig. 1
). One subject underwent menopause at age 47 yr (after 12 months in the study), and her data were censored at that time. Thus, for the primary end point of BMD at 1 yr, 27 subjects are available for analysis. Baseline characteristics for the 30 subjects who started raloxifene are shown in Table 1. Average FSH and luteal phase progesterone for these subjects were 6.4 mIU/ml and 13.3 ng/ml, respectively.
|
|
BMD
The seven subjects who did not start the drug did not significantly differ in their baseline BMD measurements from those who did (data not shown). The mean baseline lumbar spine T score was –0.2. The mean lumbar and total hip bone density were 1.027 g/cm2 (interquartile range 0.94–1.11 g/cm2) and 0.905 g/cm2 (interquartile range 0.84–0.97 g/cm2), respectively. Twenty-seven women completed a DEXA after 1 yr on the drug, 26 after 2 yr on the drug, and 15 after 1 yr off raloxifene. Average intrasubject percent change from baseline at the lumbar spine was –2.3% at 1 yr and – 3.5% at 2 yr (Fig. 2, P < 0.00001 at both time points). After stopping raloxifene for 1 yr, BMD was –2.9% from baseline but increased 1.4% from yr 2. Change from baseline at total hip was –0.3% at 1 yr, – 1.0% at 2 yr (P 0.033), and 0.5% after 1 yr off raloxifene but increased 1.7% from yr 2 to 3.
|
Twenty-seven sets of BTMs were available for analysis. After 12 months on raloxifene, all BTMs increased as BMD declined. Median change for bone-specific alkaline phosphatase, NT, and osteocalcin were 11, 10, and 18%, respectively. Only osteocalcin was significantly correlated with change in BMD (r = –0.51, P = 0.016 after correction for multiple comparisons).
Lipids and fibrinogen
Use of raloxifene was associated with favorable changes in the lipid profile. HDL significantly increased over the course of taking raloxifene. Mean baseline HDL was 59.9 mg/dl. The mean increase from baseline was 3.9, 6.3, and 2.9 mg/dl after 3, 12, and 24 months on raloxifene (P = 0.019). HDL declined after stopping raloxifene (–3.6 mg/dl from 24 to 36 months, P = 0.042). Mean baseline apolipoprotein B was 95.0 mg/dl. Apolipoprotein B declined while on raloxifene (mean change –5.6, – 2.5, and –10.4% after 3, 12, and 24 months on raloxifene, respectively, P = 0.011) and decreased nonsignificantly after stopping it (mean change –1.4% 24 to 36 months, P = 0.74). Mean baseline fibrinogen was 293 mg/dl. Fibrinogen declined while on raloxifene (mean change –30, – 10, and –17 mg/dl after 3, 12, and 24 months on raloxifene, respectively, P = 0.011) and increased significantly after stopping it (24- to 36-month mean change 43 mg/dl, P = 0.0002). No statistically significant change in total cholesterol, LDL, triglycerides, or apolipoprotein A-1 was seen. No subject began therapy with lipid-lowering agents over the course of the study.
Adverse events
All adverse events were grade 1 or 2, except for a concurrent diagnosis of two primary malignancies in one subject (papillary thyroid cancer and follicular lymphoma), and one basal cell skin cancer in one subject. The most common adverse events at least possibly attributed to study medication were change in menses (includes change in cycle length and change in menstrual flow), myalgias (mostly leg cramps), and hot flashes in 86, 64, and 57% of the subjects, respectively. Other effects often seen with hormonal therapy and reported by subjects in this study included breast pain (25%), headache (25%), and vaginal discharge (11%). Subjects had a 1.6% increase in BMI after 2 yr on raloxifene and a 3.6% increase from the end of yr 2 to 1 yr after stopping the drug.
Quality of life
Twenty-eight subjects completed MENQOL assessment at baseline, 26 subjects after 6 and12 months on raloxifene, 23 subjects at 2 yr, and 14 subjects 1 yr after stopping raloxifene. In the vasomotor domain, subjects reported a statistically significant increase in symptoms from an average score of 1.32 (on a 1–8 scale; lower numbers correlate with a better quality of life) to 1.74 at 12 months and 2.14 at 24 months (corrected P = 0.0078). This change in vasomotor symptoms declined 1 yr after stopping raloxifene (mean score 1.71). No significant changes in the psychosocial, physical, or sexual domains were reported. A similar number of subjects completed the SF-36 at the evaluated time points. The distributions of the results are skewed toward better-than-average physical and mental health, compared with the general U.S. population. No statistically significant changes in the physical component score or mental component score are seen while on raloxifene or after stopping it.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The differential effects of SERMs on bone in the pre- and postmenopausal setting are not fully understood. In vitro and animal data implicate the importance of bone type, presence of estrogen, and different interactions between raloxifene and ER
vs. ERß (15, 16, 17). It is also possible that raloxifene interacts with so-called orphan receptors to mediate the effect on bone. For example, the ER-related receptor (ERR)- does not bind estradiol, yet the transcription of ERR is up-regulated by estradiol and blocked by faslodex in an osteoblast cell model, suggesting that the ERR expression is controlled through ERs (18).
It is useful to consider the magnitude of the change in BMD and BTM we observed in the context of other hormonal interventions and physiological change with age. The changes in BMD in our trial are less dramatic than what has been observed in premenopausal populations treated with GnRH analogs. Six months of GnRH agonist use resulted in a 3–5% decrease in lumbar BMD from baseline to end of treatment, accompanied by a greater increase (
50%) in osteocalcin and NT. With the cessation of treatment, BMD improved over the next year and osteocalcin levels returned to baseline (19, 20). In comparison physiological bone changes in perimenopausal years are accompanied by no statistically significant change in BMD, compared with premenopausal controls, and a lesser increase in osteocalcin (20%) (21, 22). Although our subjects continued to lose bone at 24 months, the relationship between longer-term use of a SERM and ongoing loss of BMD is unknown. Subjects on tamoxifen for 3 yr experienced no further decline in lumbar spine from yr 2–3, although there was further decline in the hip (7).
We reported on a number of secondary outcomes to provide a more complete picture of how raloxifene affects premenopausal women. The studies of tamoxifen on similar end points in like cohorts allow for some comparisons. Raloxifene increased HDL and decreased apolipoprotein and fibrinogen. These findings suggest a possible cardioprotective effect of raloxifene. Similar trends are reported with raloxifene in postmenopausal women, yet not all the same lipoproteins are affected (23, 24). The Raloxifene Use for the Heart trial will examine the effect of raloxifene on cardiac events in postmenopausal women so that the clinical importance of these changes on risk factors is better understood (25). Of note, similar changes have been seen with tamoxifen, and yet no reduction in cardiovascular disease was detected (26).
Change in menses was the most commonly reported event and included both shorter and longer cycles, heavier and lighter menstrual flow. It is difficult to ascertain how much of this change is attributable to raloxifene because women of this age range typically experience menstrual cycle changes with increasing frequency as they approach menopause. Hot flashes are reported in 9–28% of postmenopausal women on raloxifene (27, 28), which is less than we observed (57%). The large discrepancy may be due to increased attention to this symptom because none of our premenopausal subjects had hot flashes at baseline. Although common toxicity criteria 2.0 grading criteria for hot flashes include only mild (grade 1) and moderate (grade 2) categories, it is important to note that intolerable hot flashes did not occur and hot flashes did not cause any subject to withdraw from the study. The increase in hot flashes is consistent with the change in the vasomotor domain on the MENQOL instrument.
There are limited comparable data on quality of life in premenopausal women. In the National Surgical Adjuvant Breast and Bowel Project P-1 trial, the similar age group (ages 35–49 yr) on tamoxifen did not score differently on the SF-36 from those subjects on placebo (29). However, for the specific symptoms of hot flashes and vaginal discharge, the younger cohort had the greatest proportional increase of these events with tamoxifen use. This is in keeping with our finding of an increase in vasomotor symptoms on the MENQOL and the 63% of women who developed hot flashes while on trial; however, no increase in vaginal discharge was seen in our subjects. The quality-of-life assessments indicate that these agents are well tolerated and have minimal adverse impact, which resolves when the agent is stopped.
This study is important because it provides safety data on raloxifene in high-risk premenopausal women. The limitations of our trial are largely due to the small sample size and the lack of a control arm. Although women in the pre- and perimenopausal years can physiologically lose BMD (30, 31), our observation that bone loss stopped when the study drug was stopped makes this an unlikely explanation. The loss of bone appears to be a SERM drug class effect because tamoxifen in a placebo-controlled trial had a similar effect in premenopausal women at increased risk for breast cancer. Another potential shortcoming is that although we did give calcium supplements to our cohort, we did not prescribe vitamin D. Vitamin D deficiencies generally are described in older persons and younger populations without sun exposure. Given that our study participants were young, eating normal diets, and active, we anticipated that adequate levels of vitamin D would be obtained through diet and sun exposure during usual activities. However, we did not monitor vitamin D levels or intake. Finally, although all secondary end points were planned at the outset of the trial, they should be recognized as hypothesis generating rather than conclusive findings.
In conclusion, our study suggests that raloxifene has similar effects to tamoxifen in premenopausal women in regard to bone, lipids, side effects, and quality of life. Skeletal effects differ markedly by menopausal status, and this highlights the importance of evaluating agents in randomized clinical trials rather than extrapolating study findings from postmenopausal to premenopausal women. This trial does not establish raloxifene as a substitute for tamoxifen in premenopausal women at increased risk for breast cancer; however, it does provide evidence for future study of this agent. Finally, although breast cancer prevention agents are promising and needed, attention to the willingness of women at risk to take a chemopreventive agent is necessary. Recent literature suggests that women do not want to take tamoxifen (32, 33); thus, future research should explore more acceptable alternatives and combined interventions for prevention of both hormone receptor-positive and -negative breast cancers.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author disclosure summary: J.E.-W., J.C.R, D.V., D.L., S.G., D.D., C.C., and J.Z. have nothing to declare. E.T.L. consults for S*BIO and Eli Lilly.
First Published Online July 25, 2006
Abbreviations: BMD, Bone mineral density; BTM, bone turnover marker; DEXA, dual-energy x-ray absorptiometry; ER, estrogen receptor; ERR, ER-related receptor; HDL, high-density lipoprotein; LDL, low-density lipoprotein; MENQOL, Menopause Specific Quality of Life Questionnaire; SF-36, 36-Item Short Form Health Survey; NT, N-telopeptide; SERM, selective ER modulator.
Received December 27, 2005.
Accepted July 18, 2006.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
: a mediator of estrogen response in bone. J Clin Endocrinol Metab 90:3115–3121This article has been cited by other articles:
 |
|
 |
|
  K. Visvanathan, R. T. Chlebowski, P. Hurley, N. F. Col, M. Ropka, D. Collyar, M. Morrow, C. Runowicz, K. I. Pritchard, K. Hagerty, et al. American Society of Clinical Oncology Clinical Practice Guideline Update on the Use of Pharmacologic Interventions Including Tamoxifen, Raloxifene, and Aromatase Inhibition for Breast Cancer Risk Reduction J. Clin. Oncol., July 1, 2009; 27(19): 3235 - 3258. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Eng-Wong, J. Orzano-Birgani, C. K. Chow, D. Venzon, J. Yao, C. E. Galbo, J. A. Zujewski, and S. Prindiville Effect of Raloxifene on Mammographic Density and Breast Magnetic Resonance Imaging in Premenopausal Women at Increased Risk for Breast Cancer Cancer Epidemiol. Biomarkers Prev., July 1, 2008; 17(7): 1696 - 1701. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Greenlee, C. Atkinson, F. Z. Stanczyk, and J. W. Lampe A Pilot and Feasibility Study on the Effects of Naturopathic Botanical and Dietary Interventions on Sex Steroid Hormone Metabolism in Premenopausal Women Cancer Epidemiol. Biomarkers Prev., August 1, 2007; 16(8): 1601 - 1609. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. T Chatterton Jr, J. Zujewski, E. T Mateo, J. Eng-Wong, and V C. Jordan Effect of raloxifene on salivary sex steroid concentrations in premenopausal women J. Endocrinol., December 1, 2006; 191(3): 599 - 604. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A. Johnson The SERM of My Dreams. J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 3754 - 3756. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
