| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Third Department of Internal Medicine, Charles University, Faculty of Medicine (D.M., J.J.S.), 128 00 Prague, Czech Republic; and Area Medical Center Vienna, Eli Lilly & Co. (B.R.B., I.P.), 1030 Vienna, Austria
Address all correspondence and requests for reprints to: Dr. Jan J. Stepan, Third Department of Internal Medicine, Charles University, Faculty of Medicine, U Nemocnice 1, 128 00 Prague, Czech Republic. E-mail: jstepan{at}vol.cz.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Design, Patients, and Interventions: Ninety-nine ambulatory women who were diagnosed with postmenopausal osteoporosis and treated with alendronate (10 mg/d) for a mean period of 43 months were randomized to double-blind raloxifene (60 mg/d; n = 33), placebo (n = 33), or continuation of open-label alendronate (n = 33) for 12 months. Patients continued their assigned treatment in a subsequent 12-month, open-label extension phase. All patients received supplemental calcium (500 mg/d) and vitamin D (800 IU/d).
Main Outcome Measures: BMD (lumbar spine, total femur, femoral neck, distal forearm, and total body) and biochemical markers (serum intact amino-terminal propeptide of type I procollagen, type 1 collagen cross-linked C-telopeptide, and osteocalcin) were measured at baseline and follow-up visits.
Results: Discontinuation of alendronate therapy resulted in a decrease in lumbar spine BMD at 12 months (–2.66%; P < 0.05), but did not change total femur BMD (+0.35%; nonsignificant). Raloxifene and alendronate, compared with discontinuation, prevented lumbar spine BMD loss (–0.75% and –0.54% at 12 months, respectively; P < 0.05). Raloxifene and alendronate caused a similar increase in total femur BMD at 12 months (1.45% and 1.56%; both P < 0.05 vs. baseline; nonsignificant vs. discontinuation). Patients, who discontinued alendronate therapy experienced an increase in bone turnover. Bone turnover increases were less pronounced in patients taking raloxifene and were absent in those who continued alendronate. Of the three groups, mean bone turnover in raloxifene patients was the closest to premenopausal mean values.
Conclusions: BMD preservation and increase were most pronounced in patients continuing alendronate. Raloxifene treatment, compared with placebo, demonstrated beneficial effects on BMD and bone turnover after discontinuation of long-term alendronate therapy.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Raloxifene, a selective estrogen receptor modulator, is another antiresorptive drug for the treatment of osteoporosis, which induces less suppression of bone remodeling than alendronate (8, 9, 10, 11). Overall, the BMD increase that occurs with raloxifene treatment is less marked than that with alendronate (8). Similar to alendronate, raloxifene effectively prevents new vertebral fractures in postmenopausal women with osteoporosis after 3–4 yr of treatment (12, 13, 14). Raloxifene therapy, however, has not been proven to reduce the risk of nonvertebral fractures (12). No data are yet available on the effect of raloxifene subsequent to alendronate therapy.
The aim of this study, therefore, was to investigate BMD and biochemical markers of bone turnover of an osteoporosis treatment strategy beginning with an agent characterized by stronger suppression of bone turnover (alendronate), continued as long as fracture prevention effect is proven, and subsequently switched to another agent with less marked suppression (raloxifene). This treatment strategy was compared with two control groups: continued alendronate treatment and discontinuation of all antiresorptive therapy except calcium and vitamin D.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The study participants were recruited from ambulatory women in our clinic. In our database, we identified 125 patients with postmenopausal osteoporosis who had been treated with alendronate (10 mg/d) for at least 3 yr at the time of study initiation. BMD and biochemical marker measurements were performed as part of our clinical routine. Before the study, these patients did not participate in any clinical trial. Before initiation of alendronate treatment, osteoporosis was confirmed by BMD measurement (T score less than –2.5, by dual-energy x-ray absorptiometry) at the lumbar spine or proximal femur.
Ninety-nine patients, who fulfilled inclusion and exclusion criteria and agreed to participate, were enrolled in the clinical study. Historical data on BMD and biochemical bone marker measurements for this subset were recorded and analyzed. These data are reported as results of the observational period (–43- to 0-month measurement period). The baseline of the study is considered the end of this observational period. According to the protocol, the duration of the double-blind/open-label, randomized core study period was 1 yr (0- to 12-month measurement period). The protocol was amended thereafter, and the patients continued in an open-label extension phase and kept their treatment assigned at randomization (12- to 24-month measurement period).
Inclusion criteria specified ambulatory postmenopausal women, 50–80 yr of age, and previous treatment with alendronate (10 mg/d) for more than 3 yr. All patients filled their prescription regularly and were reportedly compliant. Subjects were excluded from the study for any of the following reasons: bone disorders other than primary osteoporosis, endocrine and malignant diseases, uterine and ovarian abnormalities, clinically severe postmenopausal symptoms that required estrogen therapy, a history of thromboembolic disorders, severe chronic diseases, or treatment with any agent that might influence bone turnover. Ethics committee approval and informed consent from all patients were obtained before any study procedures were performed. The primary objective of the core study was to explore the difference in vertebral BMD change from baseline after 12 months within and between the therapy groups. Assuming an SD of 3.5, 33 patients in each group would have been needed to detect a 2.4% difference in lumbar spine BMD between two groups with 80% power. The secondary objectives included the investigation of changes in total hip and femoral neck BMD and biochemical markers of bone remodeling between the different treatment arms.
After treatment with alendronate, at the end of the observational period, patients were randomized to double-blind raloxifene (60 mg/d; n = 33), placebo (n = 33), or continuing open-label alendronate (n = 33) for the core study period (12 months). In the extension phase (subsequent 12 months), the patients remained on their assigned therapy, but were treated open label. All patients received calcium (500 mg/d) and vitamin D (800 IU/d) over the entire treatment period. At each visit, participants were assessed for vital signs and adverse clinical experiences. Routine laboratory analyses (blood counts, chemistry, and routine urinary analysis) were performed at baseline and after the first and second years.
Bone densitometry measurements
BMD measurements were performed at the beginning of the observational period (–43-month measurement), at baseline of the core study period (zero time point), at the end of the core study period (12-month measurement), and at the end of the extension phase (24-month measurement). BMD was determined using the Hologic, Inc., QDR 4500 A (Waltham, MA) densitometer. Normative values provided by Hologic, Inc. (National Health and Nutrition Examination Survey III normative values for the proximal femur), were used for the determination of T scores (comparison with a gender-matched, young, normal reference population). The short-term precision in vivo errors for the lumbar spine (L1–L4), total femur, femoral neck, distal forearm, and total body were 0.7%, 0.9%, 1.9%, 0.9%, and 1.5%, respectively; the long-term precision in vitro error was 0.31%. No correction for drift was required during the study.
Biochemical markers of bone turnover
Blood specimens were collected in the morning after an overnight fast. Biochemical markers of bone turnover were measured in the observational period (–43 month), at baseline (0 month), and at 6-, 12-, and 24-month visits. The serum concentrations of intact amino-terminal propeptide of type I procollagen (PINP) was assessed by RIA (Procollagen Intact PINP, Orion Diagnostica, Espoo, Finland). The assay is not sensitive to the small molecular weight degradation products of the propeptide. The within-run imprecision was less than 5%, and the between-run imprecision was less than 7% at concentrations between 20 and 90 ng/ml. The concentrations of type 1 collagen cross-linked C-telopeptide (CTX) and N-MID osteocalcin (OC, amino-terminal and mid fragment, amino acids 1–43) in plasma were assessed using electrochemiluminesence-based immunoanalysis (Elecsys 1010 Analyzer, Roche, Mannheim, Germany). The within-run imprecision of the CTX was less than 5% for samples greater than 500 pg/ml, less than 7% for samples between 200 and 500 pg/ml, and less than 10% for very low CTX samples. The between-run imprecision results were less than 7% for samples greater than 500 pg/ml and less than 9% for samples between 200 and 500 pg/ml. The detection limit was less than 10 pg/ml. The within-run imprecision for the OC was less than 5%, and between-run imprecision was less than 6% at concentrations between 11 and 40 ng/ml.
One hundred and forty healthy premenopausal women (mean age, 39.5 yr; range, 28–46 yr; mean BMD T score at the lumbar spine, –0.1 ± 1.2 SD; at the total hip, 0.1 ± 0.9 SD) were recruited and measured separately at the same study center. The upper (ULN) and lower (LLN) limits of normal were taken to be the mean ± 2 SD as follows: PINP, 33.0 (19.3–56.4) ng/ml; CTX, 267 (156–458) pg/ml; and OC, 20.6 (11.6–36.4) ng/ml. Least significant changes were 22% for PINP, 16.2% for OC, and 32.2% for CTX.
Statistical methods
A commonly used modified intention to treat approach was employed in the analysis; that is, all patients who had at least one dose of therapy, a baseline measurement, and at least one posttreatment observation were included in the analysis. All patients were analyzed according to their original group assignment. The study was designed to enroll 100 patients.
BMD and bone biomarker measurements during the 3-yr alendronate observation period were analyzed using a repeated measures model with a term for time and an unstructured covariance matrix. The mean and SD range values for bone biochemical markers were determined on the log scale and then exponentiated to return to the original measurement scale (i.e. geometric mean). T scores were calculated as follows for each patient measurement: measured value – premenopausal mean value/premenopausal SD. Comparisons between the patient population at each time point and the normal population were made using Wilcoxon rank-sum tests.
At baseline (month 0) of the study period, patient demographics are presented for the three randomized treatment groups. The mean percent change from baseline (for BMD) and mean bone biomarker levels (log transformed; see above) with 1 SD ranges are presented. Within- and between-group comparisons were made using a repeated measures model including terms for baseline level of response, treatment, time, and treatment by time interaction with an unstructured covariance structure. The proportions of patients with minimum or maximum values over the 6- to 24-month period outside the normal range were compared between treatment groups using Fisher’s exact test.
All tests were two-sided at the 0.05 level of significance. Unadjusted pairwise comparisons between treatments are reported only when a significant therapy or therapy-time effect indicated a difference between therapies. Analyses were performed on version 8.2 of the SAS System for Windows (SAS Institute, Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
One hundred patients were randomized into the study. One patient randomized to raloxifene decided to discontinue the study due to personal reasons before any postbaseline measurements were taken. Two patients taking alendronate discontinued between the 12- and 24-month assessments (one because of allergic skin reactions, another because of gastric pain), but both completed the 24-month visit measurements. All other patients completed the 2-yr core and extended study periods (Fig. 1
). Two completers showed low compliance with raloxifene (<30%) in the core study, both discontinued raloxifene entirely and remained on only calcium and vitamin D in the extension phase. After the core study period, one patient in the placebo arm switched to raloxifene. All other patients were more than 80% compliant with the study medication.
|
Before randomization into the core study, patients were treated with alendronate (10 mg/d) for 43 ± 7 months (Fig. 1
). These –43 and 0 month BMD and biochemical marker values compared with those in healthy premenopausal women (T scores) are summarized in Table 1. The BMD increased significantly during alendronate treatment (P < 0.05). Before treatment initiation (–43 months), mean serum PINP, plasma CTX, and OC were significantly increased compared with reference values in premenopausal women (P < 0.05; Table 1). The majority of patients were in a stage of high bone turnover; 33, 99, and 80% of the patients had values above the ULN for PINP, CTX, and OC, respectively. The mean of biochemical bone markers indicated an effective, significant suppression of bone remodeling after approximately 6 months of alendronate treatment (–70.5%, –84.1%, and –67.3% for PINP, CTX, and OC, respectively; P < 0.05) and at the end of the observational period (Table 1).
|
Baseline characteristics of patients randomized into three treatment groups (placebo, alendronate, and raloxifene) are summarized in Table 2; baseline biochemical bone marker concentrations are displayed in Fig. 2. At randomization, no important differences were observed between the groups with respect to age, weight, height, years since menopause, duration of alendronate pretreatment, BMD, or mean biochemical markers of bone remodeling, although nonvertebral fractures were less common in the alendronate group. Moreover, percent gains in BMD at the different measurement sites during the previous 43-month observational period were also not different (data not shown). Compared with baseline (0 month visit), at the end of the core study period (12 months), lumbar spine BMD decreased only in the placebo group (P < 0.05). In the alendronate and raloxifene groups, lumbar spine BMD remained unchanged, and both were superior to the decrease in the placebo group (P < 0.05). Compared with baseline, total femur and femoral neck BMD increased significantly in both alendronate- and raloxifene-treated patients (P < 0.05). Overall, the changes in total femur and femoral neck BMD during the core period were not significantly different among the three groups. Compared with both baseline and placebo, total body BMD increased significantly in the alendronate and raloxifene groups (P < 0.05). The increases were not different between the two active treatment groups (Fig. 2). Distal forearm BMD increased significantly (P < 0.05) and similarly in all three groups (2.83%, 3.29%, and 3.13% for placebo, alendronate, and raloxifene, respectively).
|
|
Total body BMD remained significantly increased and superior to placebo group values in both alendronate- and raloxifene-treated patients (P < 0.05; Fig. 2).
Compared with baseline, the mean serum concentrations of all three biochemical markers of bone turnover measured (PINP, CTX, and OC) in the alendronate group remained unchanged during the core and extension periods. At the first follow-up visit, 6 months after discontinuation of alendronate, the mean values of PINP and CTX increased significantly in both placebo and raloxifene groups (P < 0.05). Only in the placebo group did the mean CTX value increase above the premenopausal mean (P < 0.05). In both placebo and raloxifene groups, mean bone marker values did not change significantly during the additional 18-month follow-up. At study end, mean levels of CTX and OC in the placebo group exceeded those in the alendronate group (P < 0.05; Fig. 3).
|
Compared with pretreatment values (–43 month), all biochemical bone markers in all treatment groups at each follow-up measurement until the end of the study remained significantly suppressed (data not shown).
The treatments were well tolerated. Two patients in the placebo (leg cramp, 1; upper gastrointestinal symptoms, 1) four in the alendronate (upper gastrointestinal symptoms, 2; bone pain, 2), and eight in the raloxifene group (lower back pain, 2; thrombophlebitis, 1; hemorrhoids, 1; leg cramp, 1; breast pain, 1; headache, 1; degenerative changes of the eye veins, 1) developed treatment-emergent signs and symptoms (P = 0.126). These treatment-emergent signs and symptoms did not result in either hospitalization or discontinuation, except for the two patients taking alendronate who discontinued treatment approximately 2 months before the end of the extension phase due to gastric pain and allergic skin reaction, respectively. The numbers of patients with incidental nonvertebral fractures during 24 months in the placebo, alendronate, and raloxifene groups were two, one, and one, respectively.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The BMD increase during continuous alendronate treatment for approximately 5.5 yr observed in this study is in agreement with previous reports of large studies (1, 2, 5). The BMD changes and the onset of subsequent bone loss after discontinuation of alendronate therapy in other studies seem to be related to the population, vitamin D supplementation, and duration and dose of alendronate (15). One-year discontinuation after 3 yr of alendronate treatment resulted in a small decrease in both the lumbar spine and total femur BMD (16). Similar observations have been made in other studies (17, 18, 19, 20, 21). Similar results have also recently been published in a large subset of patients with up to 5-yr follow-up in the Fracture Intervention Trial comparing alendronate and placebo after 3–6 yr of alendronate treatment (6, 7). Thus, a certain degree of bone loss resumes after discontinuation of alendronate.
In patients receiving continuous alendronate treatment, biochemical markers remain suppressed compared with premenopausal values (1, 2, 5, 22). From the data published by Fink et al. (23), we calculated that approximately 52% and 66% of the patients taking alendronate in their study should have had values suppressed below their local normal premenopausal range for PINP and CTX after 4 months. This is similar to the percentage of individuals with suppressed PINP and CTX values (52% each) in our study. Other markers of bone turnover, such as serum osteocalcin, type I procollagen C-terminal propeptide, bone alkaline phosphatase, and urinary free deoxypyridinoline, are reported to decrease during alendronate therapy less markedly, to close to the premenopausal mean (23, 24, 25). Obviously, the marked bone turnover reduction, in some cases below the normal premenopausal range, is an important characteristic of the clinical effectiveness of alendronate. Greater reductions in bone turnover with alendronate therapy are associated with fewer hip, nonspine, and vertebral fractures after 3–4 yr of alendronate treatment (26). In contrast, in a few alendronate-treated individuals, reduction of bone turnover much below the lower limit of the normal premenopausal range may eventually compromise the therapy benefit (27).
The increase in bone resorption indicated by the elevated mean CTX concentration after discontinuation of alendronate in our study was similar to that previously reported (15). As a consequence of discontinuation of treatment, in a marked proportion of patients the CTX level exceeded the ULN premenopausal range. Increases in different bone markers after alendronate discontinuation appear to depend on the dose and the duration of the treatment (15, 16). Although the clinical consequences of increased bone resorption after discontinuation of alendronate treatment need to be tested, preliminary data suggest that discontinuation, compared with continuation of alendronate therapy, is associated with a greater risk of clinical vertebral fractures (7). Moreover, increased levels of markers of bone turnover appear to be associated with increased risk of fractures in untreated postmenopausal women. Patients with bone marker concentrations above a threshold, e.g. in the upper tertile or above the ULN premenopausal range (more than the mean ± 2 SD), suffer disproportionately from impaired bone strength and increased fracture risk (28, 29, 30, 31). Thus, discontinuation of alendronate treatment without additional intervention may be inadequate to manage osteoporosis in a large portion of women with postmenopausal osteoporosis.
Compared with alendronate discontinuation, switching to raloxifene caused a smaller increase in bone resorption. Individually, fewer raloxifene than discontinuing patients had CTX increased above premenopausal level, and fewer raloxifene than alendronate patients had markers suppressed below premenopausal level.
Similar to other studies, we compared the concentrations of biochemical markers with normal values and ranges established in healthy premenopausal women (23, 24, 28, 29, 30, 32, 33). We also calculated mean T scores to quantify the relationship of bone turnover to the reference premenopausal range (24, 34). Our study indicates high bone turnover and predominance of bone resorption in patients before the initiation as well as after the discontinuation of alendronate treatment. As a result, discontinuation of alendronate was associated with bone loss. A switch to raloxifene, in contrast, maintained a balanced bone turnover and prevented BMD decrease.
A limitation of this study is the difference applied in the labeling of the clinical trial material between study arms; raloxifene and placebo were administered double blind, whereas alendronate was open label. This bias could potentially influence the study outcome in terms of compliance and measurement. In the extension period, for example, women informed they had been taking placebo had to continue with calcium and vitamin D therapy only. The investigators did not record any concomitant use of bone-active agents in this group and study period. The high overall compliance in all arms and the objective nature of the outcome measurements (dual-energy x-ray absorptiometry and biochemical analyses) suggest that the impact of the labeling and unblinding bias could be limited.
Strengths of this study include a low dropout rate and comparison of biochemical bone marker data with those of a relatively large local premenopausal reference population. The premenopausal reference values were similar to those in other studies (8, 23). Internationally validated premenopausal standards for biochemical markers, however, could add value to the interpretation of our findings on bone turnover. To draw more meaningful clinical conclusions, histomorphometric evaluations and/or calcium kinetics studies should also be performed. Although BMD changes with raloxifene have been shown to underestimate the vertebral fracture prevention effect (35), the impact of positive BMD and bone turnover changes by raloxifene on bone strength were not evaluated in our study. Studies with fracture end points are necessary to estimate the effectiveness and risks of the different treatment regimens of the postmenopausal osteoporosis.
In conclusion, discontinuation of long-term alendronate therapy resulted in the resumption of bone loss and an increase in bone turnover, in some cases to above the premenopausal reference value. BMD preservation and increase were most pronounced in patients continuing alendronate. Compared with discontinuation, switching from alendronate to raloxifene resulted in maintenance of and increased BMD and a new equilibrium between markers of bone resorption and formation, closer to the premenopausal mean.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BMD, Bone mineral density; CTX, collagen cross-linked C-telopeptide; LLN, lower limit of normal; OC, osteocalcin; PINP, propeptide of type I procollagen; ULN, upper limit of normal.
Received November 11, 2004.
Accepted December 2, 2005.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Qaseem, V. Snow, P. Shekelle, R. Hopkins Jr., M. A. Forciea, D. K. Owens, and for the Clinical Efficacy Assessment Subcommittee Pharmacologic Treatment of Low Bone Density or Osteoporosis to Prevent Fractures: A Clinical Practice Guideline from the American College of Physicians Ann Intern Med, September 16, 2008; 149(6): 404 - 415. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. MacLean, S. Newberry, M. Maglione, M. McMahon, V. Ranganath, M. Suttorp, W. Mojica, M. Timmer, A. Alexander, M. McNamara, et al. Systematic Review: Comparative Effectiveness of Treatments to Prevent Fractures in Men and Women with Low Bone Density or Osteoporosis Ann Intern Med, February 5, 2008; 148(3): 197 - 213. [Abstract] [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 |
), raloxifene (
), and alendronate (
) for 24 months. Before randomization, patients were treated with alendronate (10 mg/d) for 43 months. a, Significant change from baseline (P < 0.05); b, significantly different vs. placebo (P < 0.05); c, significantly different vs. raloxifene (P < 0.05).