| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Divisions of Endocrinology and Metabolism (S.L.G.) and Geriatric Medicine (S.L.G., N.M.R.), Department of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15213; and Center for Biostatistics in AIDS Research (R.A.P.), Harvard School of Public Health, Boston, Massachusetts 02215
Address all correspondence and requests for reprints to: Susan L. Greenspan, M.D., University of Pittsburgh, Osteoporosis Prevention and Treatment Center, Kaufmann Medical Building, Suite 1110, 3471 Fifth Avenue, Pittsburgh, Pennsylvania 15213-3221. E-mail: griffithsd{at}msx.dept-med.pitt.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Biochemical markers of bone turnover are associated with changes in bone mass (5, 6) and fracture risk in postmenopausal women with osteoporosis (7). We have previously shown that early changes in biochemical markers of bone turnover in elderly women on alendronate therapy are associated with changes in bone mass after 2.5 yr of therapy (8). Few data are available on the ability of biochemical markers of bone turnover to predict changes in bone mass in elderly women on HRT, or combination therapy with hormone replacement and alendronate. The aim of this study was to examine whether short-term changes in markers of bone formation and resorption could predict long-term changes in bone mineral density (BMD) at the spine and hip in elderly women on HRT, alendronate, or combination therapy for 3 yr.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We recruited 573 community-dwelling women age 65 yr and older from the greater Boston area (9). Women were excluded if they had a history of any illness (renal failure, hepatic failure, active malignancy, current hyperthyroidism, hyperparathyroidism, or malabsorption) or were currently taking medications (glucocorticoids, anticonvulsants, or excess thyroid hormone) that affect bone mineral metabolism. Women who had been treated with an osteoporosis medication (bisphosphonates, HRT, raloxifene, or calcitonin) within a year of screening were also excluded. We also excluded women with a contraindication for HRT or bisphosphonates. Candidates were excluded if they had a femoral neck BMD value of 0.9 g/cm2 or greater (i.e. 0 SD of mean peak BMD using the Hologic database before the third National Health and Nutrition Examination Survey database) (10). The protocol was approved by the institutional review board at Beth Israel Deaconess Medical Center in Boston, MA. Prospective participants were advised of the nature of the study and provided written informed consent before participation.
The study design, screening, and run-in phase were previously described in detail (9). Briefly, this was a randomized, double-blind, 2 x 2 factorial design. We compared the effect of alendronate or placebo combined with HRT or placebo. The study was conducted from January 1996 through May 2001. Women who had a hysterectomy were given 0.625 mg/d of conjugated equine estrogen (Premarin) or matching placebo; women with an intact uterus received 0.625 mg/d of conjugated equine estrogen with medroxyprogesterone 2.5 mg/d (Prempro) or matching placebo (drug and matching placebo provided by Wyeth-Ayerst Laboratories, Philadelphia, PA). Women received alendronate sodium (Fosamax) or matching placebo (drug and matching placebo provided by Merck Research Laboratories, Rahway, NJ). We calculated daily calcium intake using a validated food frequency questionnaire (11). Participants received a supplement containing 600 mg calcium carbonate per tablet and 200 IU vitamin D per tablet (Caltrate Plus D; Wyeth-Ayerst) to increase their total calcium intake to more than 1000 mg/d. Finally, all participants received a daily multivitamin containing 400 IU per tablet so that their daily vitamin D intake was 400–800 IU.
Of the 573 women screened, 485 entered a 3-month, open-label run-in phase before randomization to minimize discontinuation after randomization. Reasons for discontinuance during the run-in phase have been previously reported (12). During the run-in phase, participants received HRT, alendronate placebo, a multivitamin, and calcium where necessary. Three hundred seventy-three women age 65–90 yr completed the run-in phase and were randomized to one of four treatment arms: 1) placebo only (hormone replacement placebo and alendronate placebo), 2) HRT only (HRT and alendronate placebo), 3) alendronate only (hormone replacement placebo and alendronate), and 4) combination therapy (HRT and alendronate). All participants continued taking calcium and vitamin D supplements throughout the study.
Outcome variables
BMD. BMD of the hip (total hip, femoral neck, trochanter, intertrochanter, and Ward’s triangle), lumbar spine (posteroanterior and lateral), and radius (ultradistal, mid-third, and one third distal radius) were measured by dual-energy x-ray absorptiometry (QDR-4500A bone densitometer; Hologic Inc., Bedford, MA) at baseline, randomization, and 6-month intervals for 3 yr. The coefficient of variation (CV) of BMD in elderly women on our densitometer was 1.7% for the lateral spine, 1.5% for the posteroanterior lumbar spine, 1.2% for the total hip, and 1.9% for the femoral neck (13, 14). Quality control for BMD scans was performed by Synarc, Inc. (San Francisco, CA) (15).
Clinical characteristics. Weight was measured as kilograms at baseline and every 6 months (Acme Digital In-Bed Scale, model 0501; Acme Medical Scale Co., San Leandro, CA). Height was measured to the nearest millimeter three times per visit and documented as the mean of the three (Harpenden Stadiometer; Holtrain Ltd., Crymych, Dyfed, UK). Body mass index was calculated as kilograms per square meter.
Biochemical markers of bone turnover and measures of bone mineral metabolism. A marker of bone resorption, urine N-telopeptide cross-linked collagen type 1 (NTx), was assessed with an ELISA (Osteomark7, nmol/liter of bone collagen equivalents/creatinine; Ostex International, Inc., Seattle, WA) (intraassay CV, 5–19%). Urine was collected upon second void at 0600–0800 h after an overnight fast and was frozen at –20 C. Serum tests for bone formation were intact osteocalcin (Novocalcin, ng/ml; Quidel Corp., Mountain View, CA) (intraassay CV, 4.8–10.0%) and bone-specific alkaline phosphatase (BSAP) (Alkphase-B, U/liter; Quidel) (intraassay CV, 3.9–5.8%). PTH was measured by allegro immunoradiometric assay (Nichols Allegro, pg/ml; San Juan Capistrano, CA) (intraassay CV, 1.8–3.4%). Serum was drawn after an overnight fast and stored at –80 C. All assays were simultaneously run at the end of the study by a single laboratory technician. Additional serum measurements included calcium, albumin, phosphate, and 25-hydroxyvitamin D.
Statistical analysis. Descriptive statistics are presented as mean ± SD unless otherwise noted. Comparisons over time within subjects were assessed using the Wilcoxon signed rank test, and comparisons between groups at specific points used the Kruskal-Wallis test for comparisons between the three active treatment groups. If the analysis showed significant differences between these three groups, then the Wilcoxon rank sum test was used for comparisons of pairs of groups. We used a Bonferroni correction to adjust for the three preplanned, pairwise comparisons (combined vs. alendronate, combined vs. HRT, and alendronate vs. HRT). Correlations were assessed using Spearman rank correlation. We used multiple regression to assess the importance of percent changes in markers adjusted for baseline values to determine the relative importance of the two. Receiver operating characteristic (ROC) curves were calculated defining responders as a change of greater than –1% from baseline at 3 yr, as done previously (9, 16). ROC curves were calculated in all arms of the study, including the placebo group, to have adequate numbers of nonresponders for analysis. Comparisons between areas under the curve (AUC) for different markers were assessed using the nonparametric method of DeLong et al. (17). Analysis was limited to participants with the necessary data (e.g. BMD changes at month 36 and biomarker data at baseline and 6 months), without imputation of missing data. We considered P < 0.05 statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
At baseline, there were no statistically significant differences in clinical characteristics, BMD, or biochemical markers of bone turnover between the control and treatment groups (Table 1
). As previously described (9), the mean BMD was in the low bone mass (osteopenic) range [determined by World Health Organization criteria (18)], and 34% of participants had osteoporosis. Retention was 90% for all participants, and compliance (defined as taking 80% of both medications) ranged from 51–63% in the four groups (9). After 3 yr, BMD increased by 3.0–5.9% at the total hip, 4.1–7.6% at the trochanter, 7.1–10.4% at the posteroanterior spine, and 7.3–11.8% at the lateral spine in the three active treatment groups (all P < 0.001), as previously reported (9).
|
After 3 yr, urinary NTx was 58.9 ± 23.2% below baseline in the combination group, 47.5 ± 39.3% below baseline in the alendronate group, and 23.0 ± 40.6% below baseline in the hormone replacement group (Fig. 1; all P < 0.001). There were no differences from baseline in the placebo group. Urinary NTx decreased significantly more in women on combination therapy compared with women on hormone replacement (P < 0.001), and this difference was observed as early as 6 months. There were no significant differences between the combination therapy and alendronate only groups (P > 0.10 at all time points). In addition, the mean change in urinary NTx of the women in the alendronate group was lower than that of the women in the hormone replacement group from month 6 on (all P < 0.01). BSAP decreased 49.1 ± 17.7% in the combination group, 39.2 ± 22.1% in the alendronate group, 31.7 ± 23.5% in the hormone replacement group, and 15.5 ± 21.1% in the placebo group after 3 yr (all P < 0.001). Significant decreases were evident in all participants on active treatment as early as 6 months (Fig. 1). The greatest decrease occurred in the women on combination therapy (P < 0.001 vs. HRT; P < 0.05, combined vs. alendronate only therapy). The mean percent decrease in BSAP in women on alendronate was not significantly different from the mean percent decrease in women on hormone replacement. The mean percent decrease in serum osteocalcin in women on active treatment was also seen as early as 6 months (Fig. 1). The decreases were similar in women on hormone replacement and combination therapy. Decreases in women on alendronate were smaller throughout the study than in women on combination therapy; at most time points, decreases in women on alendronate were smaller than in women on hormone replacement. There were no statistically significant differences in the percent change at 6 months between women on combined continuous hormone replacement and those who were on unopposed estrogen for the three markers (all P > 0.20).
|
In the active treatment groups, the percent change in biochemical markers at 6 months was associated with the 3-yr percent change in BMD at the hip and spine (Table 2). For instance, the improvement in hip BMD at 3 yr was negatively correlated with the percent change at 6 months in NTx (r = –0.428; P < 0.001), BSAP (r = –0.457; P < 0.001), and osteocalcin (r = –0.219; P < 0.001). Similar associations were observed between the 6-month percent changes in NTx, osteocalcin, and BSAP and the 3-yr changes in BMD at the femoral neck, trochanter, and posteroanterior and lateral spine (Table 2) for women on active therapy. We observed qualitatively similar results for the association of the absolute change in markers at 6 months and the absolute change in BMD at 3 yr. There were no consistent, significant differences in the strength of these associations between the active treatment groups.
|
When participants on active treatment were separated into three tertiles based on decreases in NTx at 6 months, those in the tertile with the largest decreases in NTx had the greatest gain in BMD at the total hip and trochanter (approximately three times the increase in the participants with the smallest change in NTx), and approximately twice as large at the posteroanterior spine (Fig. 2). Furthermore, when we grouped participants into tertiles according to the change in BSAP response at 6 months, those with the greatest decrease in BSAP had the greatest increase in BMD at the total hip, trochanter, and posteroanterior spine compared with those with the smallest change in BSAP at 6 months (Fig. 2). Similar findings were observed for osteocalcin.
|
). A 30% decrease in NTx at 6 months was associated with a predicted average increase in posteroanterior spine BMD of 6% for women on HRT, a 7% increase for women on alendronate, and an 8.5% increase for women on combination therapy at 3 yr (Table 3). Similar associations were found for a 30% decrease in BSAP (Table 3).
|
). At all three sites, the AUC was highest for NTx and lowest for osteocalcin. At the total hip, AUC ranged from 59.7–77.9%. NTx was a significantly better predictor of an increase in hip BMD at 3 yr than osteocalcin (P < 0.01) but not significantly different from BSAP. Similarly, at the trochanter, the AUC ranged from 59.7–77.9%. NTx was a better predictor than both BSAP (P < 0.05) and osteocalcin (P < 0.01). The data also suggested that BSAP was a better predictor than osteocalcin (P = 0.06). At the posteroanterior spine, the AUC ranged from 65.6–75.0%. There were no significant differences between the three markers in accuracy, although again, NTx had the highest accuracy. Similar results (not shown) were found when cutoffs of 0 or 3% were used.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We have previously shown that 6-month changes in markers of bone turnover are associated with long-term changes in the BMD of elderly women on alendronate for 2.5 yr (8). The present study confirms our previous findings and extends them for a duration of 3 yr. Ravn and colleagues examined the response of biochemical markers of bone turnover to alendronate therapy in younger women in the Early Postmenopausal Intervention Cohort (EPIC) study (19, 20, 21). The investigators noted associations similar to those in our elderly cohort. In addition, Rosen et al. (22) examined early changes in biochemical markers of bone turnover in young postmenopausal women who were randomized to HRT or calcium. They also observed that changes in biochemical markers at 6 months were associated with changes in BMD at 1 yr. Similar observations have been reported by others (23, 24, 25). Evio et al. (26) recently examined changes in markers and BMD in elderly women on HRT, alendronate, or combination therapy over 2 yr. In contrast to our study, they did not find any extra gain in bone mass with combination therapy but did note a greater decrease in NTx and a marker of bone formation (P1NP) in combination therapy vs. monotherapy, which occurred by 1 yr. The lack of significant differences in BMD between these groups may be due to the smaller sample size or higher dropout rate (i.e. 23%, double that of our study) and a shorter study duration (2 yr compared with 3 yr in our study).
Although there were associations of baseline marker values with the percent change in BMD at 6 months, in models using both the baseline marker values and the percent change in marker values at 6 months, baseline values were not associated with the change in BMD after adjustment for the percent change at 6 months. This suggests that the association of baseline values of markers with the percent change in BMD at 3 yr was a result of the association of baseline marker values with the 6-month percent changes in markers, rather than being an independent predictor of BMD change.
In our study, we found the trends in BSAP (a marker of bone formation) and NTx (a marker of bone resorption) to be similar. That is, the women who were on alendronate alone had a significantly greater decrease in BSAP and NTx than the women on hormone replacement alone. However, osteocalcin, another marker of bone formation, was lower in the women on hormone replacement than the women on alendronate alone. Although BSAP and osteocalcin are both markers of bone formation, BSAP is an isoenzyme localized in the plasma membrane of the osteoblast and is not affected by renal function (27). Osteocalcin, a noncollagenous protein synthesized by the osteoblast, is excreted by the kidney (27). The physiological function of these markers is still largely unexplained. However, differences in the location, excretion, or function could potentially account for some of the differences noted above.
In clinical practice, it is important to determine which patients will respond to therapy. We previously reported that 92% of women in this study responded to alendronate (9) [defined as having a total hip BMD change of greater than –1% at 3 yr (28)]. Using this definition, 82% of women responded to hormone replacement, and 95% responded to combination therapy. Other investigators have defined therapeutic response differently, using either a 0 or +3% cutoff for change in BMD (19).
One of the limitations of this study is that we included women with both osteopenia and osteoporosis, whereas other studies have examined a more homogenous cohort of osteoporotic patients. Second, because we had a 3-month run-in phase, where patients who were intolerant of hormone replacement were allowed to discontinue the study before randomization, these results may not be generalizable to all elderly women. However, there are several strengths of this study. It was conducted in a single center, using a single bone densitometer. All markers were collected in the morning in a fasting state to avoid potential errors caused by the circadian rhythm of biochemical markers (29). In addition, all biochemical markers of bone turnover were batch assayed in the same laboratory by a single technician. Furthermore, this is the first 3-yr, placebo-controlled study that compares two different types of antiresorptive therapies with the combination to determine the ability of biochemical markers of bone turnover to predict long-term outcomes. Finally, we had an acceptable compliance and retention rate of 90% for a 3-yr study involving hormone replacement in elderly women.
In summary, we have found that 6-month changes in biochemical markers of bone turnover can predict 3-yr changes in bone mass at the spine and hip in elderly women on hormone replacement, alendronate, and combination therapy. Future studies are needed to determine whether changes in these markers after therapy are associated with fracture reduction or long-term BMD outcomes for individual patients.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Wyeth-Ayerst Laboratories (Philadelphia, PA) provided the Premarin and Prempro, matching placebo, and Os-Cal Plus D, and Merck Research Laboratories (Rahway, NJ) provided the alendronate and matching placebo used in this study.
First Published Online February 15, 2005
Abbreviations: AUC, Area under the curve; BMD, bone mineral density; BSAP, bone-specific alkaline phosphatase; CV, coefficient of variation; HRT, hormone replacement therapy; NTx, N-telopeptide cross-linked collagen type 1; ROC, receiver operating characteristic.
Received June 11, 2004.
Accepted February 4, 2005.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. K. Ivaska, J. Lenora, P. Gerdhem, K. Akesson, H. K. Vaananen, and K. J. Obrant Serial Assessment of Serum Bone Metabolism Markers Identifies Women with the Highest Rate of Bone Loss and Osteoporosis Risk J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2622 - 2632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Ardeshirpour, P. Dann, D. J. Adams, T. Nelson, J. VanHouten, M. C. Horowitz, and J. J. Wysolmerski Weaning Triggers a Decrease in Receptor Activator of Nuclear Factor-{kappa}B Ligand Expression, Widespread Osteoclast Apoptosis, and Rapid Recovery of Bone Mass after Lactation in Mice Endocrinology, August 1, 2007; 148(8): 3875 - 3886. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J-J Body, M Lichinitser, S Tjulandin, P Garnero, and B Bergstrom Oral ibandronate is as active as intravenous zoledronic acid for reducing bone turnover markers in women with breast cancer and bone metastases Ann. Onc., July 1, 2007; 18(7): 1165 - 1171. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Ivaska, K. Pettersson, A. Nenonen, K. Uusi-Rasi, A. Heinonen, P. Kannus, and H. K. Vaananen Urinary Osteocalcin Is a Useful Marker for Monitoring the Effect of Alendronate Therapy Clin. Chem., December 1, 2005; 51(12): 2362 - 2365. [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 |
