Effect of Alendronate and MK-677 (a Growth Hormone Secretagogue)

[musclebeauty]

GH increases bone turnover and stimulates  osteoblast activity. We hypothesized that administration of MK-677, an  orally active GH secretagogue, together with alendronate, a potent  inhibitor of bone resorption, would maintain a higher bone formation  rate relative to that seen with alendronate alone, thereby generating  greater enhancement of bone mineral density (BMD) in women with  postmenopausal osteoporosis. We determined the individual and combined  effects of MK-677 and alendronate administration on insulin-like growth  factor I levels and biochemical markers of bone formation (osteocalcin  and bone-specific alkaline phosphatase) and resorption [urinary  N-telopeptide cross-links (NTx)] for 12 months and BMD for 18 months.

In  a multicenter, randomized, double blind, placebo-controlled, 18-month  study, 292 women (64–85 yr old) with low femoral neck BMD were randomly  assigned in a 3:3:1:1 ratio to 1 of 4 daily treatment groups for 12  months: MK-677 (25 mg) plus alendronate (10 mg); alendronate (10 mg);  MK-677 (25 mg); or a double dummy placebo. Patients who received MK-677  alone or placebo through month 12 received MK-677 (25 mg) plus  alendronate (10 mg) from months 12–18. All other patients remained on  their assigned therapy. All patients received 500 mg/day calcium.

The  primary results, except for BMD, are provided for month 12. MK-677,  with or without alendronate, increased insulin-like growth factor I  levels from baseline (39% and 45%; P < 0.05 vs. placebo). MK-677 increased osteocalcin and urinary NTx by 22% and 41%, on the average, respectively (P < 0.05 vs.  placebo). MK-677 and alendronate mitigated the reduction in bone  formation compared with alendronate alone based on mean relative changes  in serum osteocalcin (−40% vs. −54%; P < 0.05, combination vs. alendronate) and reduced the effect of alendronate on resorption (NTx) as well (−52% vs. −61%; P < 0.05, combination vs. alendronate). MK-677 plus alendronate increased BMD at the femoral neck (4.2% vs. 2.5% for alendronate; P  < 0.05). However, similar enhancement was not seen with MK-677 plus  alendronate in BMD of the lumbar spine, total hip, or total body  compared with alendronate alone. GH-mediated side effects were noted in  the groups receiving MK-677, although adverse events resulting in  discontinuation from the study were relatively infrequent. In  conclusion, the anabolic effect of GH, as produced through the GH  secretagogue MK-677, attenuated the indirect suppressive effect of  alendronate on bone formation, but did not translate into significant  increases in BMD at sites other than the femoral neck. Although the  femoral neck is an important site for fracture prevention, the lack of  enhancement in bone mass at other sites compared with that seen with  alendronate alone is a concern when weighed against the potential side  effects of enhanced GH secretion.

Issue Section: Article  

OSTEOPOROSIS IS A common and important cause of morbidity and mortality among postmenopausal women (1–3). Virtually all agents currently available to treat osteoporosis are primarily antiresorptive in mechanism (4–6).  In contrast, several lines of evidence suggest that GH has a  stimulatory effect on bone remodeling and could be useful in the  treatment of osteoporosis due to its anabolic properties. GH stimulates  osteoblast differentiation and proliferation in vitro (7).  Depending on the species and cell lines, GH also increases osteoblast  production of insulin-like growth factors I and II (IGF-I and IGF-II) (7)  both of which are mitogenic, increase human osteoblast differentiation,  and are probably important local regulators of bone remodeling (8). Furthermore, GH has been shown to stimulate bone formation and increase the strength of cortical bone in aged rats (9).

Human aging is associated with declining serum concentrations of GH and IGF-I (10–12). This reduction may contribute to the decrease in bone mass that accompanies normal aging (13).  Recombinant human GH (rhGH) increases markers of bone turnover,  suggesting an overall increase in bone remodeling, in healthy and  osteoporotic elderly women and GH-deficient (GHD) adults (14–19). Increased bone turnover has also been shown in GHD adults treated with rhGH based on histomorphometric measures (20).  Although stimulation of skeletal dynamics did not result in increased  trabecular bone volume, cortical thickness increased significantly.  Whereas GH alone decreased bone mineral density (BMD) in GHD adults  after 1 yr of treatment (21), continued treatment with rhGH increased BMD by 18 months in these patients (22).  Initial decreases in bone mass after GH administration were ascribed to  the hormone’s effect to accelerate both sides of the bone balance  equation, formation and resorption, whereas the effect with continued  administration was a net anabolic increase in bone density (23).

Despite these effects in GHD individuals, GH administration has not consistently increased bone mass in the elderly (24–26). In one study, GH given for 6 months increased lumbar spine density by 1.6% in men older than 60 yr of age (24).  In another study, administration of GH for 6 months increased bone  mineral content by 0.9% in elderly men, although the researchers  described the clinical consequence of this increase as unknown (25).  Furthermore, administration of rhGH for 12 months to frail elderly men  and women resulted in increased bone turnover with no increase (at an  average daily dose of 0.003 mg/kg·day or less) or a decrease (at an  average daily dose of >0.006 mg/kg·day) in BMD (26).

MK-677 is an orally active nonpeptide spiropiperidine previously demonstrated to be functionally indistinguishable in vitro and in vivo (27) from GH-releasing peptide, a relatively selective GH secretagogue (28–30).  MK-677 enhances the pulsatile release of GH, resulting in sustained  elevations in IGF-I, and is well tolerated after oral administration in  animals, healthy young men, and older men and women (27, 31–33).  Furthermore, administration of MK-677 to elderly women for 9 weeks  increased serum osteocalcin, a marker of bone formation, on the average  by 29%, and urinary N-telopeptide cross-links (NTx), a marker of bone  resorption, on the average by 25% (34).

Alendronate is a potent nitrogen-containing bisphosphonate (35, 36) that increases bone mass (37, 38)  and reduces the incidence of vertebral and other fractures, including  those of the hip, in women with postmenopausal osteoporosis (6, 37).  Alendronate quickly acts to decrease bone resorption, reaching a  plateau effect within 3 months based on a reduction in urinary NTx (39).  This decrease in bone resorption is followed by a subsequent secondary  reduction in bone formation, which plateaus within 3–6 months, as shown  by a reduction in bone formation markers such as osteocalcin and  bone-specific alkaline phosphatase (BSAP). This sequence of events is  anticipated due to the well established coupling of bone resorption and  formation (40).  It was hypothesized that combining administration of a net anabolic  agent such as a GH secretagogue and a bone resorption inhibitor such as  alendronate might allow uncoupling of the indirect suppressive influence  of alendronate on bone formation. If administration of MK-677 with  alendronate resulted in less suppression of bone formation and similar  effects on bone resorption relative to the effects of alendronate alone,  combination treatment may increase bone mass beyond that seen with  alendronate alone. This would be expected to result in a decreased risk  of fractures associated with osteoporosis.

We determined therefore  the individual and combined effects of chronic administration of MK-677  and alendronate on IGF-I levels, biochemical markers of bone formation  and resorption, and BMD in women with postmenopausal osteoporosis. The  percent change from baseline in serum osteocalcin and urinary NTx were  the primary and secondary end points of the study, respectively. The  percent change from baseline of the femoral neck BMD was the  prespecified key BMD end point based on the balance of cortical and  trabecular bone at this site.

Subjects and Methods

Subjects

Two-hundred and ninety-two women (mean age, 72.1 yr; range, 64–85 yr)  were selected for participation at 10 study centers. To be eligible for  the study, subjects had to be postmenopausal (without menses for at  least 4 yr), with a femoral neck BMD at least 2.0 sd below the mean peak value for healthy young women (<0.695 g/cm2 as measured by Hologic, Inc., Waltham, MA; model 1000W, 2000, or 4500), but no more than 3.0 sd  below the age-specific mean. Other than osteoporosis, the patients were  in good health. Patients with any fracture attributed to osteoporosis  or any disease or drug therapy (including any GH, bisphosphonate,  fluoride, glucocorticoid, or estrogen therapy within the past 6 months  or bisphosphonate treatment at any time) potentially affecting bone  metabolism were excluded. The following were additional exclusion  criteria: abnormal renal function, elevated fasting glucose, a history  of cancer or major upper gastrointestinal mucosal erosive disease, or  low 25-hydroxyvitamin D levels. The women were recruited by direct  mailings or telephone contacts and advertisements in the media. Ethical  review committee approval was obtained at each participating site, and  written informed consent was obtained from each subject.

Study design

This was a multicenter, randomized, double blind, placebocontrolled,  parallel group, 6-month study with planned extensions from 6–12 and  12–18 months. After a 2-week, single blind placebo/calcium carbonate  (OSCAL 500, Marion Merrell Dow, Kansas City, MO) run-in period, 292  women were randomly assigned in a 3:3:1:1 ratio to 1 of 4 daily  treatment groups (Table 1).  The 4 treatment groups from months 0–12 were MK-677 (25 mg) plus  alendronate (10 mg); alendronate (10 mg); MK-677 (25 mg); and double  dummy placebo. Patients who received MK-677 or placebo through month 12  received MK-677 (25 mg) plus alendronate (10 mg) from months 12–18 while  retaining the study blind (Table 1). Patients in the other two groups continued their assigned therapy.

Table 1.

Treatments

n1 Rx months 0–122 Rx months 12–182 Group I 111 MK-677 (25 mg)/alendronate (10 mg) MK-677 (25 mg)/alendronate (10 mg) Group II 109 MK-677 placebo/alendronate (10 mg) MK-677 placebo/alendronate (10 mg) Group III 36 MK-677 (25 mg)/alendronate placebo MK-677 (25 mg)/alendronate (10 mg) Group IV 36 MK-677 placebo/alendronate placebo MK-677 (25 mg)/alendronate (10 mg) 1

Number enrolled in the study.

2

All patients receive OSCAL 500.

View Large

Each  patient received three tablets per day. Patients were instructed to  take alendronate (10 mg) or matching placebo orally once daily while in a  fasting state after rising in the morning (at least 30 min before  breakfast) and to avoid lying down for at least 30 min after dosing.  MK-677 (25 mg) or matching placebo was taken at least 30 min after  alendronate/alendronate placebo regardless of food intake. A 500-mg  elemental calcium supplement (as calcium carbonate, OSCAL 500, Marion  Merrell Dow) was ingested with dinner daily to ensure nutritional  adequacy of calcium for all patients. Compliance was monitored by pill  count and patient report of missed doses. Patients were evaluated every 4  weeks until month 3 and then at 6- to 12-week intervals, with a  possible total study participation of 18 months.

Biochemical analyses

Urine (fasting second morning voided specimen) chemistry values  (N-telopeptide cross-links and creatinine), special serum bone  biochemistry assessments (osteocalcin and bone-specific alkaline  phosphatase), and hormones (including IGF-I) were obtained at baseline  (after 2-week placebo run-in, before study drug) and at months 1, 3, 6,  9, and 12 of treatment. The primary comparison of biochemical markers of  bone turnover was after 12 months of treatment, allowing comparison  among the four original treatment groups (i.e. combination, alendronate, MK-677, or placebo).

Serum  IGF-I was measured by a competitive binding RIA after acid-ethanol  extraction (Endocrine Sciences, Inc., Tarzana, CA). At a mean serum  concentration of approximately 36.6–40.5 nmol/L, the within- and  between-assay coefficients of variation (CVs) were 5.9% and 8.2%,  respectively.

Osteocalcin was measured using an immunoradiometric  assay (CIS International, Pacific Biometrics, Seattle, WA) with  interassay CVs of 4.3% and 5.5% at serum concentrations of 1.5 and 3.4  nmol/L, respectively. BSAP was measured using an immunoradiometric assay  (Tandem-R Ostase, Hybritech, Inc., San Diego, CA) with an interassay CV  of 7.4%. The manufacturer reports these values in mass units; this is  the standard unit of expression in medical literature for BSAP. NTx were  measured using the Osteomark assay from Ostex International, Inc.  (Seattle, WA), with an interassay CV of 4.0% and are reported after  correction for creatinine [NTx/Cr, nanomoles of bone collagen equivalent  (BCE) per mmol creatinine].

BMD measurements and radiographic assessment

BMDs of the femoral neck, hip, lumbar spine, and total body were  measured by dual energy x-ray absorptiometry using Hologic, Inc., model  1000W, 2000, or 4500. BMD was determined twice (femoral neck and lumbar  spine) or once (total body) at baseline and after 3, 6, 9, 12, and 18  months of treatment. The primary comparison among treatments for BMD was  after 18 months of treatment to allow for the maximal duration of  treatment with MK-677/alendronate and alendronate alone. A common,  standardized procedure for patient positioning and utilization of  software was incorporated into the QA manual procedures provided by the  central QA center. The baseline scan was evaluated before follow-up hip  scan. Patient positioning was duplicated as closely as possible, and  identical scan parameters were used. As the scan was acquired, the  identical starting point and femur positioning used at baseline were  verified. If the match of baseline and follow-up acquisitions was not  optimal, then the patient was repositioned or rescanned. Internal dual  energy x-ray absorptiometry calibration was maintained at each center,  and calibration across centers was performed using Hologic, Inc., spine  and linearity phantoms. Hologic, Inc., Medical Data Management Services  was responsible for handling all aspects of quality assurance for BMD  measurements, including assessment of consistency of acquisition,  analysis, and data management at the study sites without knowledge of  treatment assignment. Lateral thoracic and lumbar spine radiographs were  evaluated at each center for the presence of prevalent or incident  vertebral fractures at baseline and after 12 and 18 months of treatment.  Radiographic fractures were defined as an x-ray report from an expert  reader noting one or more definite fractures or as a 20% or more  decrease in the height of a vertebral body and at least a 4-mm decrease  in vertebral height.

Assessment of treatment safety

Patients were questioned about intercurrent health problems at each  visit. Standard clinical evaluations and laboratory analyses, including  hematological and chemistry values, were performed at least every 6  weeks during the first 9 months of treatment and every 3 months  thereafter. Physical examinations were performed at baseline and after  12 and 18 months of treatment. Radiographs were obtained during the  study if needed to assess a clinical syndrome consistent with fracture.  All adverse events (including clinical reports of fracture) were  recorded by the physician investigator, who rated each event as to  whether it appeared causally related to the study drug. Study drug  referred collectively to any combination of MK-677/MK-677 placebo,  alendronate/alendronate placebo, and calcium supplement.

Statistical methods

The biochemical markers included osteocalcin (primary end point),  urine NTx (secondary end point), and BSAP. Data were transformed to ln  (fraction of baseline) for the analysis and backtransformed to percent  change from baseline for presentation. The analysis included the effects  of the four treatments on femoral neck BMD (prespecified key BMD end  point) as well as lumbar spine, total hip, and total body BMD. The  percent change from baseline was analyzed.

The percent change from  baseline was analyzed with ANOVA with factors for the effect of center,  treatment, and treatment by center interaction. If the P value  from the F test for the interaction effect was greater than 0.1, the  interaction term was dropped from the ANOVA model before assessment of  the treatment effect. Also before assessment of the treatment effect,  appropriate diagnostic tests were performed to ensure that the data  conformed to the statistical assumptions of common variance and  normality of distribution. Patients who completed 18 months of the study  and had valid BMD measurements at baseline and month 18 were included  in the analysis of the change from baseline BMD to month 18. A per  protocol approach was taken, which excluded data from patients with  serious protocol deviations and made no attempt to replace missing  values. The per protocol analysis was specified because it provided the  best evaluation of the scientific model underlying the protocol.  Comparisons were accomplished using the t test computed with  the least square means (LSMEANS) and root mean squared error provided by  SAS PROC GLM. Data are reported as the mean ± se.  There was 80% power (Ι = 0.05, by two-tailed test) with a sample size  of 60 patients in the MK-677/alendronate treatment group and 60 patients  in the alendronate alone treatment group to detect between group  differences from baseline in osteocalcin, NTx, femoral neck BMD, and  lumbar spine BMD of 17, 9, 2.4, and 2.0 percentage points, respectively.

Accounting for patients in the analysis

Four patients were excluded from the 18-month analysis of BMD data  due to new-onset concurrent therapy including thyroid, estrogen, and  steroid therapy. Patients who completed 12 months of the study and had  bone turnover marker measurements at baseline and month 12 were included  in the analysis of the change from baseline to month 12. Four patients  were excluded from the latter analysis due to extended periods off study  drug (failure to take >75% of doses, as prespecified in the data  analysis plan), and seven patients were excluded due to new-onset  concurrent therapy or missing biochemical marker data at the 12 month  point. All patients with available data were included in the safety  analysis.