Chinese Medical Journal 2003;116(8):1127-1133
Effects of raloxifene hydrochloride on bone mineral density, bone metabolism and serum lipids in postmenopausal women: a randomized clinical trial in Beijing

Correspondence to:Zheng Shurong,Department of Obstetrics and Gynaecology, Peking University First Hospital, Beijing 100034, China (Tel: 86-10-66170137. Fax:86-10-66170137. E-mail:cyzh@public.fhnet.cn.net)
Keywords
raloxifene;bone mineral density;metabolism;lipids;postmenopause
Abstract
Objective To determine the effects of raloxifene hydrochloride (RLX) on bone mineral density (BMD), bone metabolism markers and serum lipids in healthy postmenopausal women in Beijing.
Methods A multicenter, randomized, double-blind, placebo-controlled study was conducted in a total of 204 healthy postmenopausal women (age 59.5±5.0 years and weight 62.8±8.7 kg) treated with either RLX 60 mg (n=102) or placebo (n=102) daily for 12 months. BMD, serum lipids, and bone markers were measured before and after drug administration.
Results Compared with placebo, RLX produced a significant increase in both total lumbar spine and total hip BMD. For the lumbar spine, percentage increase in total BMD was 2.3% with RLX compared with a decrease of 0.1% with placebo (P<0.001). Corresponding values for total hip BMD were a 2.5% increase for RLX and a 1.1% increase for placebo (P=0.011). For biochemical markers of bone metabolism, serum osteocalcin and C-telopeptide, percentage decreases were 27.65% and 24.02% in RLX-treated subjects. Corresponding values in placebo were a 10.64% decrease and a 15.75% increase (RLX compared with placebo, both P<0.001). For total cholesterol and low-density lipoprotein cholesterol levels, percentage decreases were 6.44% and 34.58% in the RLX-treated group. Corresponding values in placebo-treated patients were a 1.44% increase and a 19.07% decrease (RLX compared with placebo, both P<0.001). No differences were found for high-density lipoprotein cholesterol or triglyceride levels between the two groups. Only 5 subjects discontinued early owing to an adverse event (3 in the RLX group and 2 in the placebo group).
Conclusions This study confirms that RLX exerts positive effects on the skeleton, increasing BMD and decreasing biochemical markers of bone metabolism, and has a positive effect on the overall serum lipid profile in postmenopausal women in China.
With diminished ovarian function and decreased estrogen levels, postmenopausal women are at risk for osteoporosis, which leads to possible serious and debilitating fractures. Estrogen, combined with progestin in hormone replacement therapy (HRT), is commonly prescribed to women to alleviate postmenopausal vasomotor symptoms and protect against the development of osteoporosis.[1,2] However, the long-term use of estrogen is required to potentially realize that these clinical benefits also increases the risks of endometrial cancer and breast cancer.[2-4] The increased cancer risks along with undesirable side effects including scheduled or unscheduled bleeding and spotting may result in decreased compliance with therapy.[5,6] More recent therapies with improved tolerability and reduced safety concerns may improve compliance, and aid in the prevention and treatment of osteoporosis.

Raloxifene hydrochloride (RLX), a benzothiophene derivative, is a selective estrogen receptor modulator (SERM) able to act as a tissue specific estrogen receptor agonist or antagonist.[7] Raloxifene therapy has been shown to be effective in preventing[8] and treating[9-11] osteoporosis in postmenopausal women. It is distinguished from SERM compounds of the triphenylethylene class including tamoxifen and toremifene by its inability to stimulate the endometrium.[8,12] Additionally, RLX has been shown to reduce the risk of invasive breast cancer in postmenopausal women with osteoporosis after 3 years of treatment with RLX,[13] and continues to reduce this risk after 4 years of treatment.[14] Like estrogen therapy, RLX favorably alters various biochemical markers of cardiovascular risk by decreasing low-density lipoprotein cholesterol, fibrinogen, and lipoprotein(a), and by increasing high-density lipoprotein-2 cholesterol without elevating triglycerides.[15] Unlike estrogen, RLX does not increase the inflammatory marker, C-reactive protein.[16] In osteoporotic women at increased risk for cardiovascular events, RLX may significantly reduce the incidence of coronary and cerebrovascular events, compared with the early increased risk observed with estrogen and HRT.[17]

While osteoporosis is currently a major cause of mortality, morbidity, and medical expense worldwide, these problems will escalate in the future as a result of an aging population. Given the prediction that over half of all hip fractures in the world will occur in Asian populations within the next 50 years,[18] studies in this population are warranted. The purpose of this study was to determine the effect of RLX on bone mineral density (BMD), bone metabolism markers, and lipids in healthy Chinese postmenopausal women.

METHODS

Subjects and design
Ambulatory postmenopausal women between 50-70 years of age, who were free of severe or chronically disabling conditions, had their last menstrual period at least 2 years before beginning the study, and had a T-score for femoral neck or lumbar spine BMD measurements in the range of -2.5 to 1.0 (-2.5
Individuals were excluded from the study for any of the following reasons: known, suspected, or history of carcinoma of the breast or estrogen-dependent neoplasia (e.g. endometrial carcinoma); history of cancer within the previous 5 years; history of deep venous thrombosis (DVT); requirement for high-dose heparinization (>7500 U/d); bone disorders or treatment with any drug impacting upon bone metabolism; acute or chronic liver disease (bilirubin >34 μmol/L, alanine transaminase >100 U/L, or alkaline phosphatase >300 U/L); impaired kidney function (serum creatinine >177 μmol/L); or abnormal uterine bleeding of an unknown etiology.

This was a multicenter, double-blind, placebo-controlled, and randomized (block randomization by manufacturer) study conducted at 3 investigative sites (Peking University First Hospital, Beijing Hospital and General Hospital of the PLA, which are all general hospitals) in Beijing, China. At baseline, a total of 204 subjects were randomly assigned to either a RLX or placebo treatment group. One hundred and two subjects were randomly assigned to the placebo group, while 102 subjects were assigned to the RLX group (60 mg/d). Each investigative site enrolled 68 subjects, of which 34 were randomly assigned to receive placebo and 34 to receive RLX 60 mg/d.

From Visit 1, each subject was supplemented with Calcichew D, containing 500 mg of elemental calcium and 200 IU of vitamin D (produced by Nycomed Pharma SA, Norway), 1 tablet/d until the end of study.

Subjects were administered either RLX or an identically appearing placebo (provided by Eli Lilly & Company), 1 tablet/d from Visit 2 and continued for 12 months.

Case report forms were 100% verified versus source data by study monitors. Data were collected by a validated AS400 mainframe database system using both double entries and logic check.

Measures
Bone mineral density determination
Dual-Energy X-Ray Absorptiometry (DEXA) was used to determine BMD. Peking University First Hospital used a HOLOGIC QDR-4500, testing 4 sites (L1-4) of the lumbar spine at the posteroanterial position, and 4 sites of the left hip (femoral neck, trochanter, inter-trochanter, and Ward’s triangle). Beijing Hospital and the General Hospital of the PLA used a LUNAR DPX-L, testing 4 sites (L1-4) of the lumbar spine at the posteroanterial position and 3 sites of the right hip (femoral neck, trochanter, and Ward’s triangle). The coefficient of variation (CV) of DEXA for the three hospitals was 1%.

Laboratory tests
Laboratory tests included those for biochemical markers of bone metabolism, biochemistry, and serum lipids.

The biochemical markers of bone metabolism, serum osteocalcin and serum C-telopeptide (CTX), were analyzed by a central laboratory (National Center for Clinical Laboratory of the Ministry of Health) using one-step ELISA analysis (kits provided by Osteometer Biotech, Denmark). For osteocalcin, both within and inter-group CV was <7.0%. For CTX, the within group CV was <5.4%, while the inter-group CV was <8.1%.

The biochemistry tests included total bilirubin, alkaline phosphatase, gamma glutamyl transferase (GGT), alanine transaminase (ALT), aspartate transaminase (AST), blood urea nitrogen (BUN), creatinine, uric acid, phosphorus, calcium, glucose (fasting), total protein, albumin, and creatine kinase (CK). Lipid analyses included total cholesterol, high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C, calculated), and triglycerides. Enzymatic analysis was performed in the central laboratory (National Center for Clinical Laboratory of the Ministry of Health) using a commercial assay (kit provided by Roche Diagnostic).

Statistical analysis
This study was designed to enroll approximately 200 patients. The patients were assigned by random allocation to one of two treatment groups. Primary comparison in this study showed difference between the RLX group and the placebo group with respect to lumbar spine BMD. The study was designed to have at least 85 patients completed per treatment group, which allowed for a 15% dropout rate and provided 89% power to detect a treatment difference in the percentage change in BMD of 2% over a 12-month period. This calculation assumed a two-sided, 0.05 level of significance, and also assumed that the observed standard deviation of percentage change in BMD per year was a maximum of 4%.

Statistical analysis was conducted using the intent-to-treat (ITT) principle. Baseline was defined as the last value before randomization. Endpoint was defined as the last post-baseline value obtained for a subject for a given measure (last value carried forward). The statistical analysis was carried out using SAS v8.1.

For continuous data, the means of the two groups were compared using analysis of variance (ANOVA). The nonparametric Wilcoxon’s rank sum test was also used. For categorical data, the proportions were compared using the Pearson’s chi-square test with one degree of freedom or the Fisher’s Exact test, as appropriate.

The primary efficacy variable was percentage change from baseline to endpoint in lumbar spine and hip BMD. Calculation: Lumbar spine BMD = Total lumbar BMC/Total lumbar area (Peking University First Hospital calculated L1-4, Beijing Hospital and General Hospital of the PLA calculated L2-4 BMC: bone mineral content). Total hip BMD=Total hip BMC/Total hip area (Peking University First Hospital calculated 4 sites, Beijing Hospital and General Hospital of the PLA calculated 3 sites). Percentage change from baseline to endpoint was the absolute change divided by the baseline value multiplied by 100. Mean absolute and percent changes from baseline to endpoint were evaluated using two-way analysis of variance (ANOVA), with therapy group, investigator and therapy-by-investigator interaction as fixed effects. Because of the non-normality or skewness of biochemical markers of bone metabolism and serum lipids, both the original response variable and the rank-transformed response variable were analyzed with the rank analyses considered primary.

RESULTS

Subjects
Of the 204 randomly assigned subjects, 195 (95.6%) completed the study, including 98 (96.1%) subjects in the placebo group and 97 (95.1%) subjects in the RLX group. There were no statistically significant differences between treatment groups for overall discontinuations (P=0.352). A total of 9 (4.4%) subjects discontinued early. Of these subjects, 5 discontinued early because of an adverse event (arthralgia in 1, paroxysmal headache in 1, cardiac neurosis in 1, endometrial carcinoma in 1, and cerebellar infarction in 1), 2 were lost to follow-up, 1 did not meet the entry inclusion criteria, and 1 had a personal conflict or other decision.

Demographic and other baseline characteristics did not differ significantly between the two treatment groups at baseline ( Table 1 ).

Efficacy
Bone mineral density
Lumbar spine and total hip BMD increased by 2.3% and 2.5% from baseline respectively in the RLX group (both within-group, P<0.05). The respective alterations in BMD from baseline for the placebo group were a 0.1% decrease for lumbar spine (within-group, P>0.05) and a 1.1% increase for total hip (within-group, P<0.05). A significant difference between the treatment groups was observed in the percentage change from baseline to endpoint for both total lumbar BMD (P<0.001) and total hip BMD (P=0.011). Based on each site of hip BMD, a significant difference between the treatment groups was observed in the percentage change from baseline to endpoint for trochanter, but no difference was found in Ward’s triangle or femoral neck ( Table 2 ).

Biochemical markers of bone metabolism
Serum CTX showed a median increase of 15.8% in the placebo group and a median decrease of 24.0% in the RLX group ( Table 3 ). The decrease in serum CTX in the RLX group was significantly different from the change in this variable for the placebo group at the end of the 12-month treatment (P<0.001, Table 3 ).

The median absolute changes in serum osteocalcin showed a decrease of 10.6% and 27.7% for the placebo and RLX groups, respectively. The decrease was significantly greater for the RLX group when compared with the placebo group at the termination of the 12-month treatment period (P<0.001, Table 3 ).

Serum lipids
Total cholesterol showed a median increase of 1.4% and a median decrease of 6.4% for the placebo and RLX groups respectively ( Table 4 ). This decrease in total cholesterol for the RLX group was significantly different from the change observed in the placebo group (P<0.001, Table 4 ). The within-group decrease in the RLX group was significant (P<0.05), whereas the within-group increase in total cholesterol in the placebo group was not significant.

The median absolute change for HDL-C was significantly different from baseline within both groups with a decrease of 11.4% and 12.7% in the placebo and RLX groups respectively at the end of the 12-month study ( Table 4 ). The median absolute change was not significantly different between the placebo and RLX groups (P=0.476, Table 4 ).

The within-group changes were significantly different for LDL-C in the placebo and RLX groups with a median decrease of 19.1% and 34.6%, respectively ( Table 4 ). The decrease in LDL-C in the RLX group was significantly greater than the change in this variable for the placebo group at the endpoint of the 12-month study (P<0.001, Table 4 ).

The median absolute changes in triglyceride concentrations increased by 5.2% and 7.2% in the placebo and RLX groups, respectively ( Table 4 ). The changes were not significant either within or between groups (P=0.405, Table 4 ).

Safety
The mean exposure to study drugs was 347.80 days and 343.81 days for the placebo and RLX groups respectively, with no significant difference between the two groups (P=0.478).

Adverse events
Of the 204 subjects randomized, 5 reported one or more serious adverse events during the study, 4 in the RLX group and 1 in the placebo group. There were no significant differences in the incidence of serious adverse events between the two therapy groups (P=0.174). The subjects in the RLX group reported endometrial carcinoma, cardiac neurosis, enlarged uterine fibroids, and vertebrobasilar blood supply insufficiency. There was no evidence to show that the 4 serious adverse events in the raloxifene-treated group were related to RLX. The subjects in the placebo group reported angina pectoris (coronary angiography was normal).

Eight subjects (7.8%) experienced hot flushes in the placebo group, while 13 subjects (12.9%) experienced this event in the RLX group. Eight subjects (7.8%) experienced leg cramps in the placebo group, while 13 subjects (12.9%) experienced this event in the RLX group. Neither hot flushes nor leg cramps showed significant differences between the placebo and RLX groups in the current investigation. Deep venous thrombosis was not reported in either group.

Clinical laboratory evaluation
Inorganic phosphorus was the only serum chemistry analyte to show a significant between-group difference (P=0.015). Significant between-group differences in change from baseline to endpoint were observed for albumin, alkaline phosphatase, total bilirubin, GGT, and cholesterol. There were no between-groups differences for ALT, AST or total protein.

DISCUSSION

The three hospitals involved in the trial used different types of DEXA for measuring BMD. Therefore, only the percentage changes in lumbar spine and hip BMD were analyzed. At 12 months, both lumbar spine and total hip BMD were significantly increased following RLX treatment when compared with placebo. While total lumbar spine BMD increased by 2.3% in the RLX group, and decreased by 0.1% in the placebo group. Total hip BMD increased by 2.5% in the RLX group and 1.1% in the placebo group. The magnitude of this increase in BMD is equivalent to that reported in a larger study.[8,19]

The current study shows that the biochemical markers of bone metabolism, serum osteocalcin and CTX decreased significantly in raloxifene-treated subjects compared with placebo. The results similar to those of others[19] have indicated a decrease in bone turnover. This result coupled with the increase in BMD, provides support for RLX acting as an anti-resorptive agent on the skeletal system to preserve bone mass and hence lower the risk of osteoporosis.

Both total cholesterol and LDL-C levels decreased significantly in the RLX group compared with placebo. No differences were found in HDL-C or triglyceride levels between the two groups. The lowering of total cholesterol and LDL-C levels without a significant rise in triglycerides demonstrates an overall favorable effect on the serum lipid profile and provides evidence of a beneficial effect of RLX in preventing cardiovascular disease. In osteoporotic women at increased risk for cardiovascular events, RLX may significantly reduce the incidence of these events without an early increased risk as observed with estrogen and HRT.[17]

In conjunction with the positive efficacy results, the current study confirmed that RLX is tolerated well in postmenopausal women. The result is consistent with the safety profile demonstrated for RLX in previous clinical trials.[8,13] No deaths were reported during the period of the study and few subjects discontinued because of adverse events. There was one case of endometrial carcinoma in the RLX group. According to literatures, RLX does not stimulate the endometrium,[20] and also does not increase the incidence of endometrial carcinoma.[13] No differences in endometrial thickness were observed between the RLX and placebo groups in the osteoporosis prevention trials after 24 months, as assessed by transvaginal ultrasonography (TVU).[8] Therefore the case of endometrial carcinoma in our study is not causally related to RLX treatment.

In general, the significant alterations observed in laboratory analytes were largely expected or not considered clinically relevant. The significant decrease in inorganic phosphorus and alkaline phosphatase for the RLX group was expected based on previous experience with RLX and is consistent with the skeletal anti-resorptive effect of the drug. It has been previously reported that hot flushes, leg cramps, and DVT were adverse events related to RLX.[13] Consistent with other studies, the current study showed a higher incidence of hot flushes and leg cramps in the RLX group. However, the differences between the RLX and placebo groups were not statistically significant for these events. This is possibly due to the sample size, which was designed for primary efficacy variables and might not have had enough power to demonstrate the significant difference in the incidences of these two adverse events. DVT was not reported in the two groups. Moreover, there were no statistically significant differences between the two groups in the reporting of treatment-emergent adverse events by overall incidence or body system. The data are important from the osteoporosis prevention perspective. In younger postmenopausal women with relatively normal bone mass, the goal of intervention is to preserve bone mass and skeletal integrity. In the untreated state, bone loss in such women averages at least 1% per year.[21,22] Therefore, maintenance of bone mass by prevention of baseline loss through therapy will provide an advantage over time. In addition, there is evidence that discontinuation of virtually any effective antiresorptive therapy will lead to a return to baseline in BMD changes. Thus, in the effort to prevent osteoporosis, the tolerability of a particular therapy for long-term compliance may be more important than short-term increase in BMD.
RLX is also antiproliferative in the breast. Analysis of newly diagnosed breast cancer cases in the current RLX database (over 10 000 women, treated for a mean duration of 2.9 years) shows that breast cancer incidence is reduced by about half in raloxifene-treated subjects, compared with placebo subjects.[23] In a multicenter, randomized, double-blind multiple outcomes of raloxifene evaluation (MORE) trial, a total of 7705 women were followed up for a median of 40 months.[11] In this study, 13 cases of breast cancer were confirmed in the RLX group (n=5129) versus 27 cases in the placebo group (n=2576), resulting in a relative risk (RR) of 0.24, 95% confidence interval (CI), 0.13 to 0.44 (P<0.001). Four years after the initiation of the MORE trial, an additional mammogram was conducted with 61 cases of invasive breast cancer reported (RR 0.28, 95% CI, 0.17 to 0.46).[14]

In conclusion, RLX 60 mg/d increases lumbar and hip bone mineral density of postmenopausal women in Beijing, which provides a very effective and necessary therapy for the prevention of osteoporosis. The favorable effect on the serum lipid profile provides support for a beneficial effect of RLX in preventing cardiovascular disease. As such, this study shows that RLX is efficacious and well tolerated in healthy Chinese postmenopausal women.

REFERENCES

1. Compston JE. Hormone replacement therapy for osteoporosis: clinical and pathophysiological aspects. Reprod Med Rev 1994;3:209-224.
2. Col NF, Eckman MH, Karas RH, et al. Patient-specific decisions about hormone replacement therapy in postmenopausal women. JAMA 1997;277:1140-1147.
3.Colditz GA. Relationship between estrogen levels, use of hormone replacement therapy, and breast cancer. J Natl Cancer Inst 1998;90:814-823.
4.Writing Group for the Women’s Health Initiative Investigators. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. Principal results from the women’s health initiative randomized controlled trial. JAMA 2002;288:321-333.
5. Rees MC. The need to improve compliance to HRT. Br J Obstet Gynaecol 1997;104:Suppl16:1-3.
6. Spencer CP, Cooper AJ, Whitehead MI. Management of abnormal bleeding in women receiving hormone replacement therapy. BMJ 1997;315:37-42.
7. Mitlak BH, Cohen FJ. In search of optimal long-term female hormone replacement: the potential of selective estrogen receptor modulators. Hormone Res 1997;48:155-163.
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 1997;337:1641-1647.
9. Delmas P, Ensrud K, Adachi J, et al. Efficacy of raloxifene on vertebral fracture risk reduction in postmenopausal women with osteoporosis. 4-year results from a randomized clinical trial. J Clin Endocrinol Metab 2002;87:3609-3617.
10. Ettinger B, Black DM, Mitlak BH, et al. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: results from a 3-year randomized clinical trial. JAMA 1999;282:637-645.
11. Lufkin EG, Whitaker MD, Nickelsen T, et al. Treatment of established postmenopausal osteoporosis with raloxifene: a randomized trial. J Bone Miner Res 1998;13:1747-1754.
12. Goldstein SR, Scheele WH, Rajagopalan SK, et al. A 12-month comparative study of raloxifene, estrogen, and placebo on the postmenopausal endometrium. Obstetr Gynecol 2000;95:95-103.
13. Cummings SR, Eckert S, Krueger KA, et al. The effect of raloxifene on risk of breast cancer in postmenopausal women: results from the MORE randomized trial. JAMA 1999;281:2189-2197.
14. Cauley JA, Norton L, Lippman ME, et al. Continued breast cancer risk reduction in postmenopausal women treated with raloxifene: 4-year results from the MORE trial. Breast Cancer Res Treat 2001;65:125-134.
15. Walsh BW, Kuller LH, Wild RA, et al. Effects of raloxifene on serum lipids and coagulation factors in healthy postmenopausal women. JAMA 1998;279:1445-1451.
16.Walsh BW, Paul S, Wild RA, et al. The effects of hormone replacement therapy and raloxifene on C-reactive protein and homocysteine in healthy postmenopausal women: a randomized, controlled trial. J Clin Endocrinol Metab 2000;85:214-218.
17. Barrett-Connor E, Grady D, Sashegyi A, et al. Raloxifene and cardiovascular events in osteoporotic postmenopausal women: four-year results from the MORE (Multiple Outcomes of Raloxifene Evaluation) randomized trial. JAMA 2002;287:847-857.
18. Cooper C, Campion G, Melton LJ III. Hip fractures in the elderly: a world-wide projection. Osteoporosis Int 1992;2:285-289.
19. Johnston CC, Bjarnason NH, Cohen FJ, et al. Long-term effects of raloxifene on bone mineral density, bone turnover, andserum lipid levels in early postmenopausal women: three-year data from 2 double-blind, randomized, placebo-controlled trials. Arch Intern Med 2000;160:3444-3450.
20. Boss SM, Huster WJ, Neild JA, et al. Effects of raloxifene hydrochloride on the endometrium of postmenopausal women. Am J Obstet Gynecol 1997;177:1458-1464.
21. Hurley D, Kholsa S. Subspeciality clinics: endocrinology, metabolism, and nutrition. Clin Proc 1997;72:943-949.
22. Ravn P, Hetland ML, Overgaard K, et al. Premenopausal and postmenopausal changes in bone mineral density of the proximal femur measured by dual-energy X-ray absorptiometry. J Bone Miner Res 1994;9:1975-1980.
23. Jordan VC, Glusman JE, Eckert S, et al. Incident primary breast cancers are reduced by raloxifene: integrated data from multicenter, double-blind randomized trials in 12 000 postmenopausal women. Program/Proceedings of the American Society of Clinical Oncology 34th Annual Meeting May 16-19, P122a. 1998: WB Saunders Company/Mack Printing Group.