Treatment With Once-Weekly Alendronate 70 mg Compared With Once-Weekly Risedronate 35 mg in Women With Postmenopausal Osteoporosis: A Randomized Double-Blind Study

Authors


  • Dr Rosen received funding from Aventis, Eli Lilly and Company, Merck, Novartis, NPC Pharmaceuticals, and Wyeth. Dr Hochberg served as a consultant for Aventis, GlaxoSmithKline, Merck, Novartis, Procter & Gamble, and Roche. Dr Bonnick owns stock in Procter & Gamble and received funding from Merck. Dr McClung served as a consultant for Aventis, Eli Lilly and Company, Merck, Procter & Gamble, and Roche. Dr Broy served as a consultant for Aventis, Eli Lilly and Company, Merck, Novartis, and Procter & Gamble. Dr Kagan served as a consultant on the medical GYN advisory board. Drs Chen, Petruschke, Thompson, and de Papp are employees of Merck. Dr Miller received grants from Amgen, Aventis, Eli Lilly and Company, Merck, Novartis, Pfizer, Pharmacia, Procter & Gamble, and Roche, and served as a consultant for Aventis, Eli Lilly and Company, Merck, and Procter & Gamble.

Abstract

Once-weekly alendronate 70 mg and once-weekly risedronate 35 mg are indicated for the treatment of postmenopausal osteoporosis. These two agents were compared in a 12-month head-to-head trial. Greater gains in BMD and greater reductions in markers of bone turnover were seen with alendronate compared with risedronate with similar tolerability.

Introduction: The nitrogen-containing bisphosphonates, alendronate and risedronate, are available in once-weekly (OW) formulations for the treatment of postmenopausal osteoporosis. A 12-month, head-to-head study was performed to compare these agents in the treatment of postmenopausal women with low BMD.

Materials and Methods: A total of 1053 patients from 78 U.S. sites were randomized to OW alendronate 70 mg (N = 520) or risedronate 35 mg (N = 533), taken in the morning after fasting. Endpoints included BMD changes over 6 and 12 months at the hip trochanter, total hip, femoral neck, and lumbar spine (LS); percent of patients with predefined levels of change in trochanter and LS BMD at 12 months; and change in biochemical markers of bone turnover at 3, 6, and 12 months. Tolerability was evaluated by adverse experience (AE) reporting.

Results: Significantly greater increases in hip trochanter BMD were seen with alendronate (3.4%) than risedronate (2.1%) at 12 months (treatment difference, 1.4%; p < 0.001) as well as 6 months (treatment difference, 1.3%; p < 0.001). Significantly greater gains in BMD were seen with alendronate at all BMD sites measured (12-month difference: total hip, 1.0%; femoral neck, 0.7%; LS, 1.2%). Significant differences were seen as early as 6 months at all sites. A greater percentage of patients had ≥0% (p < 0.001) and ≥3% (p < 0.01) gain in trochanter and spine BMD at 12 months with alendronate than risedronate. Significantly greater (p < 0.001) reductions in all biochemical markers of bone turnover occurred with alendronate compared with risedronate by 3 months. No significant differences were seen between treatment groups in the incidence of upper gastrointestinal AEs or AEs causing discontinuation.

Conclusions: In this 12-month, head-to-head trial of alendronate and risedronate, given in accordance with the approved OW regimens for treatment of osteoporosis in postmenopausal women, alendronate produced greater gains in BMD and greater reductions in markers of bone turnover than risedronate. The greater antiresorptive effect of alendronate was seen as early as 3 months, and the tolerability profiles were similar.

INTRODUCTION

OSTEOPOROSIS IS DEFINED as a skeletal disorder characterized by compromised bone strength, predisposing to an increased risk of fracture.(1) Postmenopausal osteoporosis is further characterized by an increased rate of bone turnover, with resorption exceeding formation, leading to declines in bone mass and strength. Bone strength is determined by several parameters, including BMD, biochemical markers (BCMs) of bone turnover, and other bone qualities. In clinical practice, bone strength is best assessed by measurements of BMD. In untreated populations, BMD is an important predictor of fracture risk; at any given age, the lower the BMD, the greater the fracture risk.(2-4) Some studies suggest that elevated levels of BCMs also independently contribute to fracture risk.(5-7) Antiresorptive agents increase BMD and decrease fracture risk by reducing bone turnover and restoring the balance between bone resorption and bone formation.

The nitrogen-containing bisphosphonates, alendronate and risedronate, have been shown to reduce the risk of both spine and nonspine fractures by reducing bone turnover, increasing bone mass, and improving bone strength.(8-11) In clinical research, the antifracture efficacy of these agents is correlated with the reduction in BCMs of bone turnover and increases in BMD.(12-19) The extent to which changes in BMD and biochemical markers account for changes in fracture risk remains an area of much debate. Whereas it has been suggested by some that changes in BMD may explain only a small proportion of vertebral fracture risk reduction,(13,(19-22) others have shown that most of the antifracture effects of antiresorptive agents are accounted for by changes in BMD.(18) It is likely that changes in BMD, when taken together with changes in markers of bone turnover, can account for a substantial proportion of the fracture reduction seen with antiresorptive agents.(15,16)

Over the past decade, bisphosphonate therapy has become the mainstay of osteoporosis treatment.(2,(8-11) Both alendronate and risedronate are available in a once-weekly (OW) formulation and are approved for the treatment and prevention of postmenopausal osteoporosis. The weekly formulations of these two bisphosphonates were approved based on demonstration of comparable gains in BMD as seen with the daily formulations of each drug.(8-11,(23,(24) Comparable reductions in bone turnover were also seen with the daily and weekly formulations of each drug.(23,24)

Head-to-head comparison is the best method by which to examine the relative efficacy of two agents.(25,26) In a prior direct comparison trial of 549 postmenopausal women, Hosking et al.(27) showed that OW alendronate (70 mg) produced significantly greater increases in spine and hip BMD and reductions in markers of bone turnover than daily risedronate (5 mg). In this study, risedronate was dosed 2 h after a meal, consistent with the existing European Union (EU) label at the time of the study. These findings may have been because of a greater antiresorptive effect of alendronate or decreased bioavailability of risedronate when dosed postmeal. Both medications were generally well tolerated, with no significant difference in discontinuations because of upper gastrointestinal (GI) adverse experiences (AEs).

This study of OW alendronate and OW risedronate in postmenopausal women with low BMD was conducted to evaluate the relative changes in BMD and bone turnover markers, with the approved dose of each drug using standard morning oral dosing. In addition, the study was designed to compare the upper GI tolerability of these two agents.

MATERIALS AND METHODS

Patient enrollment

Fosamax Actonel Comparison Trial (FACT) was a double-blind, randomized, active-controlled, multicenter study conducted at 78 sites within the United States. It was conducted in accordance with consideration for the protection of patients, as outlined in the Declaration of Helsinki, and was approved by the appropriate institutional review boards. All patients gave written informed consent before undergoing any examination or study procedure.

Eligible patients were community-dwelling, ambulatory, postmenopausal (at least 6 months) women ≥40 years of age (≥25 years if surgically menopausal). Those who had low BMD, defined by a BMD of ≥2.0 SD below young normal mean bone mass in at least one of four sites (total hip, hip trochanter, femoral neck, or postero-anterior [PA] lumbar spine [L1-L4]), based on the normative database of the manufacturer of the individual densitometer and who met prespecified entry criteria were randomized and enrolled. The BMD entry criteria were derived from the FDA-approved product labeling of alendronate and risedronate and are consistent with the National Osteoporosis Foundation (NOF) treatment guidelines.(2,(28,(29) Patients were otherwise required to be in good general health, with hip and spinal anatomy suitable for DXA. Patients with a history of abnormalities of the esophagus that delay esophageal emptying, such as stricture or achalasia, were excluded, as were patients unable to remain upright for 30 minutes after dosing. No other gastrointestinal exclusions were specified in the protocol. Patients with hypocalcemia (serum calcium < 8.5 mg/dl), hypovitaminosis D [serum 25(OH)D < 10 ng/ml], or metabolic bone diseases other than postmenopausal osteoporosis were excluded. Women were excluded if they had taken estrogen, estrogen analogues, tibolone, or anabolic steroids within 6 months; estrogen use for ≤1 week at least 3 months before study entry was allowed, as was the use of vaginal estrogen cream (≤2 g, up to two times weekly). Prior use of any bisphosphonate within 1 year or for ≥2 years within 5 years was a reason for exclusion, as was the use of any parathyroid hormone within the past year. Treatment with fluoride and treatment with glucocorticoids for >1 month with >7.5 mg of prednisone or its equivalent, daily within 6 months of randomization, and treatment with immunosuppressants were not allowed.

Study design

The study was designed as a 12-month BMD trial with an additional 12-month extension during which patients were maintained on blinded study medicine. Patients who met all study entry criteria were enrolled and sequentially assigned an allocation number independent of study site. Patients were randomized to one of two treatment groups using a computer-generated, blinded schedule provided by the sponsor. Patients and investigators were blinded to treatment; each patient received two bottles of study medication (one active and one placebo). Group 1 patients received alendronate 70 mg OW (Fosamax; Merck, Whitehouse Station, NJ, USA) and risedronate-matching placebo, whereas group 2 patients received risedronate 35 mg OW (Actonel; Procter & Gamble Pharmaceuticals, Cincinnati, OH, USA) and alendronate-matching placebo. All study medications were taken with a full glass of water (6-8 oz) after fasting on arising for the day while remaining in an upright position for 30 minutes before the first food or beverage of the day. Patients were to begin treatment within 7 days of randomization and take their dose on the same day each week. Missed doses were to be taken the morning after the patient realized the dose was missed, with patients returning to their normal scheduled day with the next dose. To determine compliance, patients recorded medication use in a compliance worksheet over the 12 months of treatment, which was validated with tablet counts of returned medication at each study visit. In addition to study medication, all patients were instructed to consume, for the duration of the study, 1000 mg of elemental calcium and 400 IU of vitamin D daily, either from dietary sources or a supplement (Os-cal 500 + D; SmithKline Beecham, Pittsburgh, PA, USA) provided by the sponsor with their noon or evening meals.

Assessment of outcomes

BMD was measured by DXA at the PA lumbar spine and proximal femur at baseline and 6 and 12 months using Hologic or Lunar densitometers. BMD was measured on the same machine during baseline and scheduled visits to the investigational site through month 12. All BMD analyses were performed by a central analysis facility (Bio-Imaging Technologies, Newtown, PA, USA) blinded to treatment allocation. Measurements of BMD phantoms by each DXA detected no significant machine drifts during the 1-year study as determined by the central analysis facility.

Laboratory parameters used to evaluate changes in bone turnover were urinary N-telopeptide of type I human collagen (NTx) corrected for creatinine (Cr; Ortho Vitros, Ortho Clinical Diagnostics, Amersham, UK) and serum C-telopeptide (CTx; Roche Elecsys, as measured on the Elecsys 2010 automated analyzer, Mannheim, Germany) to assess rate of bone resorption, and serum bone-specific alkaline phosphatase (BSALP) (Access OSTASE Assay; Beckman-Coulter, Fullerton, CA, USA) and serum N-terminal propeptide of type 1 procollagen (P1NP; INTACT P1NP; Orion Diagnostica, Espoo, Finland) to assess rate of bone formation. Stored samples were analyzed by a central laboratory (Quest Diagnostics, Collegeville, PA, USA) in batches by patient at the end of the study. It was recommended that all biochemical markers of bone turnover be obtained while the subjects were fasting. Urinary NTx was measured from second morning void urine samples.

Efficacy and safety evaluations

The primary efficacy endpoint was the mean percent change from baseline in hip trochanter BMD at 12 months. This site was chosen as the primary endpoint because of the rapid and large gains in BMD seen at this site in response to bisphosphonate therapy, and the excellent precision of BMD measurements at this site. Secondary BMD endpoints included mean percent change from baseline in total hip, femoral neck, and lumbar spine BMD at 12 months and mean percent change in all BMD endpoints at 6 months. The percentage of patients with predefined increases of hip trochanter and lumbar spine BMD ≥0% from baseline and those with ≥3% and ≥5% increases from baseline at 12 months were also analyzed, as was the percentage of patients with BMD loss ≥3%. Percent change from baseline in biochemical markers of bone turnover (NTx, CTx, BSALP, and P1NP) at 3, 6, and 12 months was included as secondary efficacy endpoints.

Safety was monitored during the study with the recording of clinical and laboratory AEs. Investigators evaluated patients for AEs during study visits, and patients could report AEs in person or by phone at any time during the study.

Statistical analysis

The primary hypothesis of the study was that OW alendronate 70 mg would produce a greater mean percent increase from baseline in hip trochanter BMD than OW risedronate 35 mg at 12 months. The primary hypothesis was tested using an intention-to-treat analysis that was modified to include all patients randomized, who had taken at least one dose of study drug, and had both a baseline and at least one post-randomization BMD measurement. The modified intention-to-treat (MITT) analysis used a last-observation-carried-forward approach for missing values. ANOVA using a linear model, including factors for study center and treatment group, was used to analyze the effect of treatment on mean percent change from baseline in hip trochanter BMD data at 6 and 12 months. Treatment differences and associated 95% CIs were estimated from the ANOVA model. A sample size of 366 evaluable patients per group had 90% power to detect a treatment difference in the change from baseline in hip trochanter BMD of 1.2 percentage points between alendronate and risedronate. The treatment difference was considered to be statistically significant if the p value was <0.05. Statistical analyses were performed using Statistical Analysis Systems Version 8.02 (SAS Institute, Cary, NC, USA). A similar primary analytical approach was used for all secondary BMD analyses.

A per-protocol (PP) approach was used as a secondary approach to analyze BMD changes. The PP approach excluded protocol violators (i.e., those patients who took <85% study drug, patients who took prohibited medications during the course of the trial, or patients who violated the protocol in a significant manner as specified in the data analysis plan) and noncompleters based on predefined criteria. All protocol violators were identified before study unblinding, and no data were carried forward in the PP approach. The PP approach was used as the primary approach to analyze the biochemical markers, because they can change rapidly because of the above-mentioned protocol violations, interruption of study therapy, or concurrent illness.(30) Because the biochemical marker data were distributed in a log normal fashion, the analysis was performed on the log-transformed data. Results were back-transformed and reported as percent decrease from baseline. The mean percent change in urine NTx, CTx, serum BSALP, and P1NP was analyzed in a similar fashion to the BMD analyses. MITT analyses were also performed as a secondary approach to the analysis of biomarkers.

The comparison of the percentage of patients with improvement in hip trochanter and lumbar spine BMD of ≥0%, ≥3%, and ≥5% and BMD loss of ≥3% between two groups was analyzed by a Mantel-Haenszel test controlling for study center.

No multiplicity adjustment procedure was applied to the efficacy analysis of the primary endpoint. However, predefined multiplicity adjustment procedures were applied for the efficacy analysis of secondary endpoints.(31) The closed-testing approach was applied to adjust for multiplicity among secondary BMD efficacy endpoints of the percent change from baseline at month 12 in PA lumbar spine, total hip, and femoral neck. Under this approach, the order of statistical testing was PA lumbar spine BMD, total hip BMD, and femoral neck BMD. The Hochberg procedure was applied for a multiplicity adjustment within either bone resorption markers or bone formation markers.(31)

The safety analysis included all patients who received at least one dose of study medication in either treatment group. Differences in proportions of patients with any AEs, serious AEs, and discontinuations because of AEs were analyzed using Fisher's exact test. The treatment groups were also compared for the proportion of patients with upper GI AEs using Fisher's exact test.

RESULTS

Patient disposition

A total of 1759 patients were screened at 78 study sites in the United States (Fig. 1). A total of 1053 patients (520 alendronate and 533 risedronate) were randomized, and 1042 patients (515 alendronate and 527 risedronate) received at least one dose of study medication. The completion rate for the first 12 months of the study was similar in the two treatment groups (alendronate, 84.2%; risedronate, 85.2%).

Figure FIG. 1.

Patient disposition.20

A similar percentage of alendronate (89.2%) and risedronate (90.2%) patients were included in the MITT analysis. A similar percentage of alendronate (70.2%) and risedronate (70.4%) patients were included in the biomarker PP analysis. All 1042 patients who received at least one dose of study medication were included in the safety analysis.

Demographics and baseline characteristics

The demographics and baseline characteristics of each treatment group are provided in Table 1. Subjects were postmenopausal women who were primarily white (95.3%), with a mean age of 64.5 years and mean time since menopause of 18.5 years. The use of calcium and vitamin D supplementation during the study was similar between treatment groups, as was the baseline serum 25-hydroxyvitamin D level (Table 1). Twenty-five percent of patients had a prior history of any upper GI disorder at baseline, and 41% used concomitant non-steroidal anti-inflammatory drugs (NSAIDs) during the study. Patients' baseline BMD and biochemical marker values are presented in Table 1.

Table Table 1.. Demographics and Baseline Characteristics
original image

Efficacy

Primary endpoint: hip trochanter BMD:

A significantly greater increase in hip trochanter BMD was observed at 12 months with alendronate compared with risedronate (3.4% versus 2.1%; p < 0.001; Fig. 2A). The treatment difference was 1.4% (95% CI: 0.8%, 1.9%). More rapid gains in BMD were seen with alendronate than risedronate; the difference in hip trochanter BMD was significant as early as 6 months (treatment difference 1.3% [95% CI: 0.8%, 1.8%; p < 0.001]). The increases from baseline at 6 and 12 months were significant with both alendronate and risedronate (p < 0.001). The results of the PP analyses of hip trochanter BMD were consistent with the findings of the MITT analyses (data not shown; p < 0.001 both at 6 and 12 months).

Figure FIG. 2.

Mean percent change from baseline to month 12 ±SE (modified intention-to-treat approach). (A) Hip trochanter BMD. (B) Total hip BMD. (C) Femoral neck BMD. (D) Lumbar spine BMD. *p < 0.05; **p < 0.01; ***p < 0.001.20

Secondary BMD endpoints:

At 12 months, a significantly greater increase in BMD was observed among patients treated with alendronate compared with risedronate at the total hip (2.2% versus 1.2%, treatment difference 1.1% [95% CI: 0.7%, 1.4%; p < 0.001]), femoral neck (1.6% versus 0.9%, treatment difference 0.7% [95% CI: 0.1%, 1.2%; p = 0.005]), and lumbar spine (3.7% versus 2.6%, treatment difference 1.2% [95% CI: 0.7%, 1.6%; p < 0.001]; Figs. 2B–2D). Significant differences between treatment groups were seen as early as 6 months at the total hip (treatment difference 0.9% [95% CI: 0.5%, 1.2%; p < 0.001]), femoral neck (treatment difference 0.5% [95% CI: 0.04%, 1.0%; p = 0.035]), and lumbar spine (treatment difference 0.7% [95% CI: 0.3%, 1.1%; p = 0.002]; Figs. 2B–2D). Increases from baseline in BMD were significant for both alendronate and risedronate at the total hip, femoral neck, and lumbar spine at 6 (p < 0.001) and 12 months (p < 0.001).

The percentage of patients who had a ≥3% increase in hip trochanter BMD at 12 months was 51.1% for alendronate and 40.7% for risedronate (treatment difference of 10.3% [95% CI: 4.0, 16.7, p = 0.002]; Fig. 3A). Similarly, a significantly greater percentage of alendronate- than risedronate-treated patients either maintained or gained BMD (≥0%) at the hip trochanter at 12 months (84.5% versus 67.8%, respectively; p < 0.001). Fewer alendronate than risedronate patients (4.7% versus 11.0%; p = 0.001) experienced a loss of BMD ≥3% at the hip trochanter. At the lumbar spine, 60.1% of alendronate patients and 41.0% of risedronate patients (treatment difference of 19.1% [95% CI: 12.9, 25.3; p < 0.001]) had gains in BMD ≥3% at 12 months (Fig. 3B). Similarly, significantly more alendronate than risedronate patients either maintained or gained BMD (≥0%) at the lumbar spine at 12 months (87.3% versus 75.6%, respectively; p < 0.001). Significantly fewer (p = 0.008) alendronate (1.3%) than risedronate (4.1%) patients experienced a loss of BMD ≥3% at the lumbar spine.

Figure FIG. 3.

(A) Percent of patients with ≥3% loss in BMD or ≥0%, ≥3%, or ≥5% gains in BMD from baseline to 12 months at the hip trochanter. (B) Percent of patients with ≥3% loss in BMD or ≥0%, ≥3%, or ≥5% gains in BMD from baseline to 12 months at the lumbar spine. **p < 0.01; ***p < 0.001.20

Biochemical markers of bone turnover

For all biochemical markers measured, the reduction in bone turnover at 12 months was significantly greater (p < 0.001) with alendronate than risedronate (Figs. 4A–4D). Significant differences in all markers of bone turnover were seen as early as 3 months. At 12 months, urinary NTx decreased by −52.8% in the alendronate group and −40.3% in the risedronate group (treatment difference −12.6 [95% CI: −16.6, −8.5; p < 0.001]; Fig. 4A); serum CTx decreased by −73.8% in the alendronate group and −54.7% in the risedronate group (treatment difference −19.1 [95% CI: −22.7, −15.6; p < 0.001]); serum BSALP decreased by −40.6% in the alendronate group and −28.1% in the risedronate group (treatment difference −12.5 (95% CI: −15.3, −9.6; p < 0.001]); and serum P1NP decreased by −63.9% in the alendronate group and −48.0% in the risedronate group (treatment difference −15.9 (95% CI: −18.9, −12.9; p < 0.001]). Reductions in NTx, CTx, BSALP, and P1NP from baseline were significant (p < 0.001) by 3 months and remained significant (p < 0.001) at 6 and 12 months with both alendronate and risedronate. The results of the MITT analyses of NTx, CTx, BSALP, and P1NP were consistent with the findings of the PP analyses for these endpoints.

Figure FIG. 4.

Changes in biochemical markers expressed as mean percent change from baseline ± SE at 6 and 12 months (per-protocol approach). (A) Urine NTx corrected for Cr (nmol bone collagen equivalents/mmol of Cr). (B) Serum CTx (ng/ml). (C) Serum BSALP (μg/liter). (D) Serum N-terminal P1NP (μg/liter). Between-group comparison of ALN vs. RIS: ***p < 0.001.20

Safety

The overall incidence of clinical AEs was not significantly different between treatment groups (76.5% alendronate versus 75.7% risedronate; p = 0.772; Table 2). To preserve patient and investigator blinding during the second year extension, AEs that occurred in relatively few patients (<0.5%) are not presented by treatment group in this publication. There were no significant differences between the treatment groups in the incidence of serious AEs (p = 0.653) or discontinuations because of AEs (p = 1.000; Table 2). One patient died of a myocardial infarction during the study.

Table Table 2.. Incidence of Adverse Experiences
original image

Upper GI AEs occurred in 22.5% and 20.1% of the patients in the alendronate and risedronate treatment groups, respectively, with no significant difference between treatments (p = 0.364; Table 2). There were no significant differences in the incidence of discontinuations because of upper GI AEs (p = 0.708) or serious upper GI AEs (p = 0.124). The most common upper GI AE reported during the trial was dyspepsia (8.2% versus 7.8% in the alendronate and risedronate groups, respectively). Gastroesophageal reflux disease (GERD) was reported in 2.7% of alendronate-treated patients and 3.0% of risedronate-treated patients.

The most common AEs reported (incidence ≥5% in either treatment group) were arthralgia, back pain, diarrhea, dyspepsia, headache, nasopharyngitis, nausea, pain in extremity, sinusitis, and upper respiratory tract infection, with no significant differences between treatment groups. Clinical fractures were uncommon during the study (incidence <5% in either treatment group). Fractures that occurred during the trial, whether or not they were associated with trauma, were reported by investigators as AEs, without requirements of radiographic confirmation or adjudication. Twenty-six fractures were reported in the alendronate group, and 20 fractures were reported in the risedronate group. There was no significant difference in the percent of patients reporting fractures as AEs between treatment groups.

Laboratory AEs were infrequent, with similar incidence in the alendronate and risedronate treatment groups.

DISCUSSION

In this study, OW alendronate 70 mg showed significantly greater increases in BMD at the hip trochanter, femoral neck, total hip, and lumbar spine compared with OW risedronate 35 mg; significant differences were seen as early as 6 months and were maintained at 12 months. Similarly, more patients treated with alendronate had either an apparent maintenance or gain in BMD at the hip trochanter and lumbar spine compared with those treated with risedronate, and fewer alendronate-treated patients had significant losses in BMD (≥3% loss) at these sites. Significantly greater reductions in all BCMs were seen at 3 months with alendronate compared with risedronate, and these differences were maintained over the 12-month study duration. Both study medications were generally well tolerated, with a similar incidence of upper GI AEs and discontinuations because of upper GI AEs.

Randomized, prospective active comparator studies are considered to provide the best evidence to evaluate the relative efficacy of two treatment options.(25) When direct comparative studies with clinically relevant outcomes (i.e., fractures) are not available, results of head-to-head studies with validated surrogate endpoints provide the best alternative evidence.(25,26) The results of this study, which included surrogate endpoints, are consistent with the findings of previous alendronate and risedronate studies and the only previous head-to-head study comparing treatment with alendronate and risedronate in postmenopausal women.(27,(32-34)

This study included OW dosing of both alendronate and risedronate, consistent with current prescribing information. Both medications were taken while fasting. The change in BMD at the hip trochanter was chosen as the primary study endpoint because of the combination of the magnitude of the change in BMD with OW dosing of these agents, excellent precision, and relative freedom from the effects of degenerative change, which often confound precision and interpretation of results at the PA lumbar spine.(23,(24,(35-37) The percentage of patients with changes in hip trochanter or lumbar spine BMD ranging from ≤−3% to ≥5% was evaluated as another predictor of relative antiresorptive potency. A 3% increase corresponds to the least significant change necessary to be 95% confident that the improvement in BMD for an individual patient is real, provided that the test precision does not exceed 1%.(38) Changes of this magnitude have been shown to predict substantial antifracture benefit with alendronate.(12) Significantly more risedronate than alendronate patients had an apparent loss of BMD at the hip trochanter and lumbar spine during 1 year of therapy. However, even patients with an apparent loss of BMD during antiresorptive therapy have been shown to have some reduction in fracture risk.(19)

Both markers of bone resorption (urine NTx and serum CTx), and markers of bone formation (BSALP and P1NP) were included as secondary endpoints in this study. Reduction of bone turnover is the main physiologic mechanism by which antiresorptive drugs increase BMD and reduce fracture risk.(39) By decreasing resorption activity at critically thinned trabeculae (“stress risers”), the microarchitecture of bone may be preserved and the fracture risk lessened.(40) Some prospective longitudinal studies have indicated that increased rates of bone turnover complement BMD for predicting fracture risk.(5,(30,(41,(42) Furthermore, changes in BCMs in response to antiresorptive therapy have been shown to predict vertebral fracture reduction at least as well as, if not better than, do changes in BMD.(14-16) Reduction of bone turnover occurs more rapidly than do changes in BMD during therapy and is likely to account for the early vertebral fracture protection reported with antiresorptive agents.

Several analyses using individual patient data have shown that reduction of biochemical markers correlates with decreased risk of both vertebral and nonvertebral fractures in clinical trials.(14-17,(43) In the analysis of data from 358 women taking risedronate and 335 women taking placebo who were enrolled in the VERT studies, a threshold effect was reported for reduction in vertebral fracture risk (i.e., there was a level beyond which further reduction of the resorption markers urinary CTx and NTx seemed to result in no further reduction in the risk of vertebral fracture).(15) The number of fracture cases among women with the largest reductions in turnover markers was not reported and is likely to have been small; therefore, the finding of a “threshold” would need to be confirmed in a larger study. Notably, Eastell et al.(15) did not observe a threshold effect for reduction in resorption markers relative to reduction in nonvertebral fracture risk. In the analysis of data from the FIT study by Bauer et al.,(16) which was based on a larger sample (n = 3105 women taking alendronate), there was no evidence of a threshold effect for either vertebral or nonvertebral fracture risk reduction. Each 1 SD reduction in 1-year change in BSALP was associated with fewer vertebral, nonvertebral, and hip fractures; a similar relationship was seen for both PINP and serum CTx, although this reached statistical significance only for spine fracture and not hip and nonspine fracture.

It has been suggested that very low levels of bone turnover could adversely affect bone strength by impairing the ability to repair bone microdamage.(44) In the analyses mentioned above, there was no evidence of increased fracture risk among those with the lowest turnover levels (or with the largest percent reductions in bone turnover) measured at 1 year.(15,16) Furthermore, in patients treated with up to 10 years of alendronate therapy or 7 years of risedronate therapy, in which sustained reduction of bone turnover was observed, there was no evidence of an excess risk of fracture.(45,46)

The relative contribution of increases in BMD and decreases in remodeling activity to fracture risk reduction in patients treated with bisphosphonates is currently an area of debate.(17,(22,(47) At least two different methods have been used to quantify the relative contribution of increases in BMD and decreases in remodeling activity, as measured by BCMs, to reduction in fracture risk. Analyses based on individual patient data have shown that for alendronate and risedronate, which are each proven to reduce both vertebral and nonvertebral fracture, changes in both BMD and BCMs were predictive of the reduction in fracture risk.(12,(15,(16) In both cases, the greater the increases in BMD and/or decreases in BCMs, the greater the reduction in fracture risk. In addition, meta-regression analyses based on summary information from several clinical trials have complemented the analyses based on individual patient data.(17-19) The shortcomings of meta-regression, however, have been discussed by others.(22,(47,(48) In summary, both BMD and BCMs are important determinants of fracture risk, and changes in these parameters are predictive of the antifracture efficacy of alendronate and risedronate.(12-19)

Among patients treated with OW alendronate or OW risedronate in this study, there was no significant difference in the incidence of overall AEs, including serious AEs and discontinuations because of AEs. Of particular interest in a trial of two bisphosphonates was the reporting of upper GI AEs. There was no significant difference between treatment groups in the number of patients who reported any upper GI AE, any serious upper GI AE, or upper GI AE causing discontinuation. Furthermore, there were no upper GI exclusions, other than those stated in the product labels; 41% of patients were concomitant NSAID users, and 25% had a history of upper GI disorders at baseline, consistent with the characteristics of postmenopausal women typically encountered in clinical practice.

This study has several strengths. The design of the study included patients who would be typical candidates for bisphosphonate therapy in the clinical practice setting. The sample size was large enough to make inferences on BMD at many sites and on a wide range of BCMs. The loss to follow-up was low. Despite its strengths, there are limitations of this study. Whether the differences in BMD and BCMs observed are attributable to inherent differences between the two drugs or to the differences in doses studied cannot be answered by this study. Importantly, this 1-year study of ∼1000 patients was not powered to differentiate between these two agents in terms of fracture risk reduction. The sample size for such a fracture study would depend on the expected difference between the two agents, the fracture incidence, and the desired power. Based on a 1-year incidence of vertebral fracture of 5%, a sample size of >50,000 would be required to detect a 10% difference with 90% power.(49) Whereas fracture studies are considered the optimal approach for evaluating antiresorptive agents, the necessary sample size for adequate power to evaluate (for superiority or equivalence) two active agents and the associated costs and logistics make such studies difficult to perform.

CONCLUSIONS

In this 12-month, randomized, head-to-head study of alendronate and risedronate, using the approved OW oral dosing regimens for treatment of osteoporosis in postmenopausal women, treatment with alendronate produced greater increases in BMD at all sites and greater reductions in all BCMs of bone turnover than risedronate. Differences in BMD and BCM were seen early and were maintained over 12 months. The evidence provided by these results is consistent with those of a previously reported head-to-head trial and meta-analyses.(27,34) Both study medications were well-tolerated, with no significant difference in upper GI AEs or discontinuations because of treatment.

Acknowledgements

The authors acknowledge the contributions to patient recruitment and data collection of the principal clinical investigators for the FACT study and the staff at their study sites. In addition, we thank Carol Skalky, Joanne Estojak, and Patti Shott of Merck & Co., Inc. for providing logistic and administrative support for the conduct of the study and the coordination of data processing.

Ancillary