Company Publication

INTRODUCTION

Both osteoporosis and dyslipidemia are highly prevalent, clinically important, chronic medical problems. Currently the aminobisphosphonates constitute the most important drug class for the treatment of osteoporosis. These agents act predominantly to decrease bone resorption by inhibition of the farnesyl diphosphate synthase step in the mevalonic acid pathway (1, 2). 3-Hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors (statins) inhibit the same pathway at an earlier point (3) and may also antagonize osteoclasts by increasing expression of osteoprotegerin (4, 5). Statins may also enhance osteoblast activity by increasing synthesis of bone morphogenetic protein-2 (6, 7). In vitro and animal studies have demonstrated that statins, including atorvastatin, can both increase bone formation by osteoblasts and decrease bone resorption by osteoclasts (6, 8-13). However, some experiments in oophorectomized animals have not increased bone mass (14).

The publication of these findings led several groups to examine data from observational studies in which some of the patients took statins. Some of these analyses suggested a possible beneficial effect on osteoporosis or fracture risk (15-28). However, several other studies, including the Women’s Health Initiative, failed to demonstrate any such effect (29-32), and one report of early effects was not confirmed by observations extended to one year (33). The possibility of bias in non-randomized trials is considerable, and one group concluded that the apparent effect of statins was attributable to unmeasured confounding factors (34). Authors of observational studies have consistently called for prospective controlled trials. The authors of a recent meta-analysis found some evidence for reduced hip fracture risk, but not in prospective studies, and found little or no effect on biomarkers and bone mineral density (BMD) (35). This analysis was limited by the small size of the controlled trials available. Relatively small prospective studies of the effect of simvastatin have suggested either small effects on BMD or indices of bone turnover (36, 37), or no effect on BMD or bone remodeling biomarkers (38), and a small, short-term study demonstrated no effect of atorvastatin on bone turnover markers (39).

In view of the considerable therapeutic implications of even a modest beneficial effect of atorvastatin or similar drugs on bone metabolism in humans at clinical doses, we undertook a prospective, randomized, double-blind, placebo-controlled trial to determine whether atorvastatin, across the range of clinical doses, compared with placebo had any effects on bone mineral density and biochemical markers of bone turnover in postmenopausal women with modest elevations of low-density lipoprotein cholesterol (LDL-C).

SUBJECTS AND METHODS

This was a multicenter, randomized, double-blind, placebo-controlled study conducted at 62 sites in the United States. Institutional review board approval was obtained for each site and each participant gave written informed consent.

Subjects
The study enrolled 626 women, aged 40 to 75 yr, with LDL-C cholesterol levels of at least 130 mg/dL (3.4 mmol/L) and less than 190 mg/dL (4.9 mmol/L). At screening, lumbar spine BMD was required to be between 0.0 and 2.5 SD below the mean for young adult Caucasian women(>0.772 g/cm² and <1.047 g/cm² by Hologic densitometers). All individuals were postmenopausal, as demonstrated by serum estradiol levels of less than 110 pmol/L (30 pg/mL) and FSH levels of more than 30 IU/L.

Patients were excluded if they were treated with any of the following within 3 months (or as specified) before screening: any lipid-lowering medication, systemic hormone therapy (6 months), or other drugs affecting bone metabolism, including vitamin D (more than 1000 IU daily), calcitonin, hormone therapy or selective estrogen receptor modulators, bisphosphonates (within 12 months), sodium fluoride, or chronic systemic or inhaled glucocorticoids. Patients with a clinical history of diabetes mellitus, coronary heart disease, or any medical disease known to be associated with development of metabolic bone disease (eg, bone marrow disease, hereditary disorders of calcium or mineral metabolism, untreated or inadequately treated endocrine disorders, severe rheumatoid arthritis, or a history of malignancy within the past 5 years) were also excluded. After publication of the third National Cholesterol Education Program Adult Treatment Panel (NCEP ATP III) guidelines in 2001 (40), patients who presented with two or more cardiovascular risk factors and an LDL-C of at least 160 mg/dL (4.1 mmol/L) were not included.

Study Design
The study incorporated a two-week screening phase, followed by 52 weeks of randomized treatment. Eligible participants were randomly assigned to receive once-daily doses of atorvastatin 10 mg, 20 mg, 40 mg, 80 mg, or matching placebo according to a computer generated pseudo-random code using the method of random permuted blocks. Patients in each group received calcium and vitamin D supplementation in the form of tablets containing 500 mg elemental calcium and 200 IU vitamin D, taken twice daily. All patients were instructed to follow the NCEP ATP III Step I (40) or a comparable diet.

The primary endpoint was percent change in areal lumbar spine BMD (L1-L4, following International Society of Clinical Densitometry official positions for skeletal site selection (41), measured by dual-energy x-ray absorptiometry (DXA) at baseline and at wk 52 or completion. Secondary endpoints included percent change in femoral neck, trochanter and total proximal femoral areal BMD measured by DXA at baseline and wk 52. In addition, lumbar spine L1-L2 volumetric BMD was measured by quantitative computed tomography (QCT) in a subset of patients at selected investigative sites. Percent changes from baseline to wk 26 were also evaluated. All centers used Hologic densitometers (models QDR-1000, QDR-2000, 2000, 4500 or Delphi). Centralized bone density interpretation and quality control were carried out by Synarc Inc. (Portland, OR, USA).

Biochemical and Safety Measurements
Serum samples for biochemical markers of bone metabolism and lipid biomarkers and second-void urine specimens were collected after a 10-14 hour fast. To minimize variability due to diurnal variation, subjects were instructed to report between 0600 and 1000 h at all visits. Bone biomarkers included serum N-telopeptide of type I collagen (Ostex International, Seattle, WA; now Wampole, MA), serum C-telopeptide of type I collagen (Elecsys 2010, Roche Diagnostics, Indianapolis, IN, USA), osteocalcin (Elecsys 2010, Roche Diagnostics), bone-specific alkaline phosphatase (Quidel, San Diego, CA, USA), N-terminal propeptide of procollagen type I (Elecsys 2010, Roche Diagnostics), and urinary deoxypyridinoline (Quidel). Samples for bone biomarkers were stored at or below –70 C until they were analyzed in batches, including all samples from each subject. Serum lipid markers included total cholesterol, LDL-C [by Friedewald estimation unless triglycerides were >400 mg/dL (4.5 mmol/L), then determined directly by ultracentrifugation], triglycerides (glycerol blanked), and HDL cholesterol (chemical precipitation by dextran sulfate/Mg Cl2). Cholesterol and triglyceride measurements employed Centers for Disease Control and Prevention (CDC) standardized enzymatic methods on the Hitachi 911 analyzer (Roche Diagnostics). All bone marker and lipid measurements were performed at Pacific Biometrics, Inc. (Seattle, WA, USA). In addition, standard hematological and biochemical safety measurements were performed.

All treatment-emergent adverse events (AEs), including adverse drug reactions, illnesses with onset during the study, or exacerbation of pre-existing illnesses, were reported, regardless of treatment group or suspected causal relationship to study drug. Serious AEs, regardless of treatment group or suspected relationship to drug, were reported immediately and were defined as any adverse drug experience that was life-threatening, or resulted in death, inpatient hospitalization, or significant disability/incapacity. Events that required medical or surgical intervention to prevent one of these outcomes could also be classed as serious AEs. Safety was also assessed by clinical laboratory measurements (including hemograms, aspartate aminotransferase, alanine aminotransferase, and creatine phosphokinase [CPK]), in addition to physical examinations, vital signs, and electrocardiograms.

Statistical Methods
The a priori sample size projection of 575 participants (115 per treatment group) was estimated to provide approximately 90% power to detect a 2.0% treatment-related difference in lumbar spine BMD between atorvastatin and placebo, assuming a SD of 3.5% and a dropout rate of approximately 20%. Hypothesis testing was controlled for multiple comparisons (4 active treatments and a placebo) via Dunnett’s method, and all tests were two-sided and were conducted at a 5% significance level. All analyses, tables, listings, and plots were produced using the SAS® statistical software.

Three patient populations were identified for purposes of analysis: the efficacy-evaluable (“per-protocol”) population consisted of subjects who met all evaluability criteria determined before unblinding and database lock, for whom follow-up data were obtained within one month of the scheduled wk 52 visit, and who took at least 80% of their study drug. Because the purpose of this study was to optimize the ability to detect differences between groups, rather than evaluate overall therapeutic effect, efficacy results for the per-protocol population were considered to be of primary interest and are presented herein. The modified intent-to-treat (ITT) population comprised all individuals who received at least one dose of study medication and had both a baseline and at least one post-baseline BMD measurement. The results for all efficacy endpoints were similar for both the per-protocol and modified ITT populations. The safety population included all subjects that received at least one dose of study medication.

The principal analyses of efficacy were comparisons of the differences in percent changes from baseline to the end of the study (52 wk) between each atorvastatin treatment group and the placebo group; percent changes from baseline to wk 26 were also evaluated. For BMD measurements, an analysis of covariance model adjusted for treatment, center, and baseline value, was used to assess treatment differences from placebo at each visit, based on comparisons of least squares (LS) means. Percent responders (defined as those patients with any positive change in BMD from baseline) and dose response (linear regression) were also assessed. For bone markers, 10% trimmed means of percent changes and 95% confidence intervals (CI) are reported. Further comparisons were carried out on rank-transformed percent changes from baseline, using linear model methodology, including O’Brien’s procedure to determine the overall treatment effect on multiple markers combined, and separate analysis on each bone marker alone. Blood lipid parameters were analyzed using the same methods used for BMD on percent change data. The proportion of patients with LDL-C of 130 mg/dL (3.4 mmol/liter) was compared between each treatment versus placebo using Pearson’s Χ²chi-square test. Due to the exploratory nature of the study, no adjustments were made for multiple end points, except that multiple comparisons of atorvastatin vs placebo LS means in linear models were performed using Dunnett’s method.

For reporting purposes, investigator terms describing AEs were coded to standard preferred terms based on the Coding Symbols for Thesaurus of Adverse Reaction Terms (COSTART) dictionary and were characterized by intensity and relationship to study drug. Consistent with usual practice, abnormal laboratory values of more than three times the upper limit of normal for hepatic transaminases (aspartate amino-transferase and alanine aminotransferase) and more than 10 times the upper limit of normal for CPK at two consecutive measurements obtained 4-10 d apart were predefined as clinically important.

RESULTS

Figure 1. Study Flow Chart

Baseline Characteristics
As shown in Figure 1, a total of 626 women were randomized to one of the five treatment groups, of whom 604 actually received at least one dose of study medication. Of these, 167 (27.6%) discontinued prematurely and 437 (72.4%) completed the study through the final visit. The most common reason for discontinuation was withdrawn consent or loss of the subject to follow up (n=87, 13.9%). The per-protocol, modified ITT, and overall safety populations comprised 318, 482, and 604 individuals, respectively. Patient demographics and baseline characteristics are presented in Table 1. There were no clinically important differences between groups in any of the baseline characteristics. Medications taken prior to study start and during the study were similar between treatment groups.

Table 1. Demographic and baseline characteristics
	Placebo, n = 119	Atorvastatin
		10 mg, n=118	20 mg,n=121	40 mg, n=124	80 mg, n=122
Age (yr)	58.8 (7.6)	58.6 (6.5)	59.2 (6.7)	59.4 (7.0)	57.8 (6.7)
Race
No. of white subjects (%)	107 (89.9)	108 (91.5)	98 (810)	110 (88.7)	105 (86.1)
No. of years menopausal	13.3 (8.3)	11.8 (8.2)	11.6 (*8.1)	12.8 (8.7)	11.3 (7.9)
Weight (kg)	73.8 (15.3)	72.7 (15.0)	72.6 (13.5)	74.6 (15.0)	73.0 (12.6)
Ex-smoker or current smoker, n (%)	55 (46.2)	56 (47.5)	50 (41.3)	62 (50)	62 (50.8)
Lumbar spine (L1-L4) BMD (g/cm²) by DXA a	0.91 (0.087)	0.92 (0.084)	0.92 (0.079)	0.93 (0..083)	0.91 (0.086)
Lumbar spine (L1-L2) BMD (g/cm³) by QCT b	111.88 (29.500)	127.52 (22.932)	124.13 (21.992)	120.64 (21.195)	125.21 (18.523)
Total hip BMD (g/cm²) by DXA a	0.84 (0.109)	0.87 (0.100)	0.85 (0.108)	0.87 (0.099)	0.87 (0.103)
Total cholesterol a
mg/dl	224.8 (21.7)	241.2 (19.8)	242.2 (21.0)	245.2 (21.8)	242.4 (23.0)
mmol/liter	6.3 (0.6)	6.2 (0.5)	6.3 (0.5)	6.3 (0.6)	6.3 (0.6)
LDL-C a
mg/dl	159.1 (16.9)	154.8 (16.5)	158.0 (17.9)	155.4 (17.0)	155.6 (15.5)
mmol/liter	4.1 (0.4)	4.0 (0.4)	4.1 (0.5)	4.0 (0.4)	4.0 (0.4)
Triglycerides a
mg/dl	141.2 (61.6)	124.7 (53.1)	139.9 (59.5)	143.2 (70.3)	152.6 (83.3)
mmol/liter	1.6 (0.7)	1.4 (0.6)	1.6 (0.7)	1.6 (0.8)	1.7 (0.9)
HDL-C a
mg/dl	57.5 (11.0)	61.6 (14.7)	56.2 (13.9)	61.1 (15.2)	56.4 (12.3)
mmol/liter	1.5 (0.3)	1.6 (0.4)	1.5 (0.4)	1.6 (0.4)	1.5 (0.3)
Numbers represent mean (SD) values unless specified otherwise.
a Mean (SD) values for the efficacy evaluable population.
b Mean (SD) values for the efficacy-evaluable population (from selected sites only) (total n= 47; placebo, n=11 10 mg, n=11; 20 mg, n=11; 40 mg, n= 6; 80 mg, n=8).

Effect on BMD
There were no significant differences between any dose of atorvastatin and placebo for percent changes in the lumbar spine (L1-L4) BMD measured by DXA from baseline to 52 weeks (Figure 2, Table 2). Furthermore, there was no significant difference between treatments at Week 26, and no significant change from baseline to Week 26 or 52 for any treatment group. Approximately 39% to 53% of patients in each treatment group were “responders” (positive change in lumbar spine [L1-L4] BMD). However, there was no significant difference between atorvastatin and placebo “responder” rates, and the overall dose-response relationship was not significant. Furthermore, there were no significant differences between groups with respect to changes in secondary BMD endpoints, including LS volumetric BMD (QCT) and the femoral DXA measurements from baseline to 26 or 52 weeks. There were no significant within-group changes, with the exception of a borderline decrease in total proximal femur BMD within the placebo group at week 52 (Table 2).

Effects on Biochemical Markers of Bone Metabolism
Measurements of the biochemical markers of bone metabolism failed to demonstrate any significant differences between the atorvastatin and placebo groups for percent change from baseline to Week 52 (Table 2) or wk 26. Within individual groups, some reductions (95% CI <0) were observed, but there was no clear pattern of dose dependence. Multivariate analysis of the overall treatment effect of all markers compared with placebo was nonsignificant for all doses of atorvastatin.

In exploratory analyses, we found no evidence of differential treatment effects on bone markers between pateints younger and older than 62 yr (data not shown).

Table 2. Percent change from baseline to wk 52 (last observation carried forward a) in BMD and markers of bone metabolism (per-protocol population)
	% change from baseline
	Placebo n=67	Atorvastatin
		10 mg, n=59	20 mg, n=65	40 mg, n=67	80 mg, n=60
Lumbar spine (L1-L4) BMD by DXA, LS mean (95% CI)	0.16 (-0.51, 0.84)	-0.26 (-0.98, 0.45)	-0.38 (-1.05, 0.30)	-0.44 (-1.12, 0.23)	-0.03 (-0.75, 0.69)
Lumbar spine (L1-L2) BMD by QCT,b, LS mean (95% CI)	2.36 (-2.96, 7.68)	-3.95 (-8.97, 1.06)	-1.93 (-7.17, 3.30)	0.11 (-6.74, 6.96)	-3.43 (-9.41, 2.54)
Total femoral BMD by DXA, LS mean (95% CI)	-0.74 (-1.43, -0.04)	-0.58 (-1.31, 0.15)	-0.55 (-1.25, 0.15)	-0.51 (-1.20, 0.19)	-0.06 (-0.81, 0.68)
sNTX, 10% trimmed mean (95% CI)	-3.68 (-8.54, 1.18)	-3.50 (-8.24, 1.24)	0.02 (-5.86, 5.89)	-6.54 (-10.05,-3.03)	-5.04 (-9.98, -0.09)
sCTX, 10% trimmed mean (95% CI)	-4.18 (-14.77, 6.41)	-1.92 (-8.58, 4.74)	3.96 (-6.45, 14.36)	-4.47 (-12.25, 3.31)	-1.14 (-10.50, 8.22)
Osteocalcin, 10% trimmed mean (95% CI)	-4.38 (-10.39, 1.64)	-4.44 (-10.52, 1.64)	-2.88 (-8.78, 3.02)	-5.32 (-10.88, 0.24)	-10.56 (-15.70, -5.43)
BSAP, 10% trimmed mean (95% CI)	-2.74 (-6.44, 0.97)	0.55 (-3.25, 4.35)	-1.34 (-6.06, 3.38)	-0.80 (-4.76, 3.16)	-1.18 (-5.41, 3.05)
P1NP, 10% trimmed mean (95% CI)	-7.04 (-12.89, -1.19)	-0.25 (-7.15, 6.64)	-1.20 (-8.86, 6.45)	-1.63 (-9.53, 6.27)	-10.29 (-16.65, -3.92)
DPD, 10% trimmed mean (95% CI)	0.83 (-5.19, 6.85)	-3.44 (-9.37, 2.48)	1.88 (-3.62, 7.38)	-1.19 (-6.70, 4.32)	-5.81 (-10.62, -1.01)
For all measurements, the P value for the difference between each atorvastatin group and placebo was nonsignificant using Dunnett's method in linear models. BSAP, Bone-specific alkaline phosphatase; DPD, urinary deoxypyridinoline; P1NP, procollagen type I N propeptide; sCTX, serum C-telopeptide; sNTX, serum N-telopeptide.
a No values were imputed due to the definition of the per-protocal population, who required follow-up data within 1 month of the wk-52 visit.
b A subset of the efficacy evaluable population from selected sites only was analyzed for lumbar spine (L1-L2 ) BMD by QCT (total n=47; placebo, n=11; atorvastatin 10 mg, n=11; 20 mg, n=11; 40 mg, n=6; 80 mg, n=8).

Fig. 2. Lumbar spine (L1-L4) BMD by DXA: mean percent change from baseline (efficacy evaluable population).

Effect on Lipids
As anticipated from the well-characterized effects of atorvastatin, the mean total cholesterol, LDL-C, and triglyceride levels decreased from baseline to wk 52 for each of the atorvastatin groups in terms of percent change, and these differences were statistically significant compared with the placebo group (Figure 3). The mean HDL-C levels were not significantly different in terms of percent change between baseline and wk 52 within treatment groups or between atorvastatin and placebo groups (Fig. 3).

Fig. 3. Serum lipids: mean difference (atorvastatin vs. placebo) percent change from baseline at wk 52 (efficacy evaluable population).

Safety Evaluation
Overall, atorvastatin was well-tolerated, consistent with the product labeling and previous publications. Serious AEs were reported in 12 patients (placebo n=3; atorvastatin 10 mg n=1, 20 mg n=4, 40 mg n=2, 80 mg n=2), of which none was considered to be related to study treatment and only one led to permanent study drug discontinuation (atorvastatin 80 mg, hemorrhagic stroke). There were no deaths or cases of rhabdomyolysis during the study.

The incidences of the most frequently reported treatment-emergent AEs are summarized in Table 3. The most frequently reported treatment-related AE was myalgia (four, five, 10, five, and three cases reported for 10, 20, 40, 80 mg atorvastatin, and placebo, respectively). We did not observe CPK elevations above the predefined threshold in any patient. Hepatic transaminase elevations exceeding the predefined thresholds were observed in two patients, both of whom were receiving atorvastatin (40 and 80 mg), and both discontinued treatment. All laboratory abnormalities were followed up and resolved by the end of the study.

Table 3. All-cause, treatment-emergent AEs occurring in at least 10% of patients in any treatment group
	%No. of patients experiencing AE (%)
	Placebo n=67	Atorvastatin
		10 mg, n=59	20 mg, n=65	40 mg, n=67	80 mg, n=60
Any AE	102 (86)	98 (83)	109 (90)	104 (84)	102 (84)
Respiratory infection	20 (17)	20 (17)	13 (11)	27 (22)	25 (20)
Arthralgia	14 (12)	16 (14)	21 (17)	20 (16)	14 (11)
Myalgia	8 (7)	10 (8)	8 (7)	19 (15)	16 (13)
Urinary tract infection	12 (10)	5 (4)	12 (10)	14 (11)	13 (11)
Headache	10 (8)	10 (8)	8 (7)	14 (11)	11 (9)
Dyspepsia	10 (8)	10 (8)	11 (9)	9 (7)	12 (10 )
Constipation	12 (10)	8 (7)	11 (9)	10 (8)	8 (7)
Pain	9 (8)	7 (6)	9 (7)	15 (12)	8 (7)
Accidental injury	9 (8)	5 (4)	4 (3)	14 (11)	8 (7)
Asthenia	6 (5)	4 (3)	13 (11)	9 (7)	6 (5)
Sinusitis	8 (7)	12 (10 )	5 (4)	5 (4)	5 (4)

 

DISCUSSION

We found no evidence that systemic atorvastatin administration produced any significant effect on bone mass or markers, despite highly provocative reports that statins, when used in experimental systems, affected both bone formation and bone resorption in the direction of a positive remodeling balance (6, 8-12). Although the orally administered doses employed in humans result in much lower exposure than the relatively high concentrations that were used in those experiments, and orally administered statins undergo a first-pass effect, the reports of a possible clinical benefit from such medications, although not consistent, warranted a prospective controlled trial. We conducted such a study using the entire clinical dosage range of atorvastatin (10-80 mg), the most widely prescribed of these medications, in order to determine whether an effect on any measurement of BMD or an effect on bone metabolism could be detected and, if so, to determine whether a dose-response relationship existed. As far as we are aware, this prospective, randomized controlled clinical trial is unique in employing a range of doses to determine whether there is an effect of HMG-CoA reductase inhibition on bone metabolism. We found no such effects over 52 wk of treatment on lumbar (L1-L4) spine BMD by DXA, lumbar (L1-L2) spine BMD by QCT, total hip BMD by DXA, or biochemical markers of bone metabolism, although the expected lipid-lowering effect was apparent. While we could not exclude the possibility of subtle effects over many years, all medications that have been shown to be beneficial for the prevention or treatment of osteoporosis have produced measurable effects on several of these endpoints.

Case-control analyses have suggested potential benefits of statins in reducing the risk of fracture (17-19, 23, 28). However, the association between fracture risk and the level of statin exposure differed across these studies and, in two of the studies, potentially confounding factors such as body mass index were inadequately controlled (17, 18). In addition, the effects of statins on BMD or bone turnover were either not assessed in these studies, or changes in BMD were not associated with the risk of fracture (23). Subsequent studies have not supported this association between fracture risk reduction and statin use (29-31, 42).

Several studies have investigated the effects of statins on BMD in hyperlipidemic patients, but were limited by the absence of a randomized, controlled design and have yielded inconsistent findings (29, 30, 36, 43-45). The data from the current trial are consistent with those of a smaller randomized, controlled trial in which simvastatin showed no benefit on BMD and biochemical bone markers in a similar population of women (38). Both studies enrolled postmenopausal women with low BMD and moderately elevated LDL-C and randomized them in a double-blind fashion to either a statin or placebo. The potential effects of statins on bone turnover were thereby assessed prospectively in a well-characterized target population using a well-matched control group.

At the time of study initiation, no clinical data were available on the effects of atorvastatin (or statins in general) on BMD, therefore statistical considerations were based on information available from selective estrogen receptor modulators. After controlling for multiple comparisons (four active treatments vs placebo), approximately 90 patients per treatment group were calculated to have 90% power to detect an assumed difference of 2.0% between active and placebo, with a standard deviation of 3.5% and a dropout rate of approximately 20%. However, this was a conservative assumption regarding the standard deviation for the distribution of percent change from baseline values and, based on the results obtained, the actual SD was in the range 2.5-2.8%. Recalculating the power assuming a standard deviation of 3% with 60 subjects in each treatment group (as in the per protocol analysis), yields a power of more than 85% for each comparison. Therefore, the power to detect a treatment difference of 2.0% in the per protocol analysis with a SD of 2.8% would be consistent with the original power estimate.

As expected, consistent with data from previous studies (46), atorvastatin therapy decreased total cholesterol, LDL-C, and triglycerides in a dose-dependent fashion in this population of postmenopausal women with dyslipidemia. No changes in HDL-C were observed in any treatment group, although HDL-C levels were relatively high in all groups at baseline. Atorvastatin was well tolerated at all doses, and overall safety was comparable to placebo. The reported incidence of serious AEs did not increase significantly with higher doses, and only two patients were reported to have clinically meaningful abnormal laboratory liver function tests with atorvastatin treatment, which were resolved by the end of the study. Thus, our findings were generally consistent with labeled product-safety information.

Despite the in vitro data suggesting that statins inhibit bone-resorption and enhance osteoblast activity (6), no effects on bone metabolism were observed in our study. Following absorption, statins are biotransformed in the liver with a high hepatic first-pass clearance, resulting in a substantial lipid-lowering effect, but a low systemic exposure to unbound pharmacologically active drug (47). The observed lack of effect of atorvastatin on BMD and bone biomarkers may therefore have been the result of low statin uptake into bone tissue. The maximum doses of the currently available medications in this class are limited by tolerability and safety considerations. Therefore, it is unlikely that substantially higher systemic doses of oral HMG-CoA reductase inhibitors could be employed clinically. Thus, our findings do not support a role for conventional doses of atorvastatin in the prevention or treatment of postmenopausal osteoporosis.

ACKNOWLEDGMENTS

The following investigators participated in this study: Ronald D Emkey, MD, James M McKenney, PharmD, Alfred H Moffett Jr., MD, Douglas H Moffett, MD, Julio Rosenstock, MD, Joel Charles Silverfield, MD, Elliott N Schwartz, MD, Susan B Broy, MD, Charles H Chesnut III, MD, E Michael Lewiecki, MD, Raymond L Malamet, MD, R Walter Powell, MD, Angelo Licata, MD, PhD, Barbara A Soltes, MD, Barry C Lubin, MD, Charles L Goldsmith, MD, David J Portman, MD, Dean J Kereiakes, MD, Douglas P Kiel, MD, Grattan C Woodson, MD, Harold E Bays, MD, Henry G Bone, MD, Lydia G Corn, MD, Avi A Ostrowsky, MD, Samuel N Lederman, MD, Michael H Davidson, MD, Michael Roy McClung, MD, Patrick H Bowen, MD, Peter N Weissman, MD, Ralph M Vicari, MD, Robert Lindsay, MD, Stephen Gordon, MD, William D Koltun, MD, William Insull, MD, Wulf H Utian, MD, PhD, Stephen W Halpern, MD, Davis Walsh Baldwin, MD, Cesar Roberto Libanati, MD, Michael Anthony Bolognese, MD, Susan M Nattrass, PhD, Maria W Greenwald, MD, Robert M Griffin Jr., DO, Michael I Keller, MD, Robert Lang, MD, Louis L Shane, MD, Stuart L Silverman, MD, Judith A Hsia, MD, Barbara Troupin, MD, David R Hassman, DO, Paul D Miller, MD, Vance Bray, MD, Alan Jan Kivitz, MD, Patrick H Peters Jr., MD, Steven L Duckor, MD, Henryk Cioczek, MD, Grazyna Szenkiel, MD, Anu Rao, MD, Luis Mario Marquez, MD
The authors would like to acknowledge the contributions made by Rachel Laskey, PhD and Attila Kursun, MD (both full-time employees of Pfizer, Inc), and Steve Dobson of Envision Pharma, a medical writer funded by Pfizer Inc, in the development of this paper. Funding was provided by Pfizer Inc.

REFERENCES

1. van Beek E, Pieterman E, Cohen L, Lowik C, Papapoulos S 1999 Farnesyl pyrophosphate synthase is the molecular target of nitrogen-containing bisphosphonates. Biochem Biophys Res Commun 264:108-11
   
2. Fisher JE, Rogers MJ, Halasy JM, Luckman SP, Hughes DE, Masarachia PJ, Wesolowski G, Russell RG, Rodan GA, Reszka AA 1999 Alendronate mechanism of action: geranylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Proc Natl Acad Sci U S A 96:133-8
   
3. Halcox JP, Deanfield JE 2004 Beyond the laboratory: clinical implications for statin pleiotropy. Circulation 109:II42-8
   
4. Kaji H, Kanatani M, Sugimoto T, Chihara K 2005 Statins modulate the levels of osteoprotegerin/receptor activator of NFkappaB ligand mRNA in mouse bone-cell cultures. Horm Metab Res 37:589-92
5. Viereck V, Grundker C, Blaschke S, Frosch KH, Schoppet M, Emons G, Hofbauer LC 2005 Atorvastatin stimulates the production of osteoprotegerin by human osteoblasts. J Cell Biochem 96:1244-53
6. Mundy G, Garrett R, Harris S, Chan J, Chen D, Rossini G, Boyce B, Zhao M, Gutierrez G 1999 Stimulation of bone formation in vitro and in rodents by statins. Science 286:1946-9
7. Ghosh-Choudhury N, Mandal CC, Choudhury GG 2007 Statin-induced Ras activation integrates the phosphatidylinositol 3-kinase signal to Akt and MAPK for bone morphogenetic protein-2 expression in osteoblast differentiation. J Biol Chem 282:4983-93
8. Garrett IR, Chen D, Gutierrez G, Zhao M, Escobedo A, Rossini G, Harris SE, Gallwitz W, Kim KB, Hu S, Crews CM, Mundy GR 2003 Selective inhibitors of the osteoblast proteasome stimulate bone formation in vivo and in vitro. J Clin Invest 111:1771-82
9. Staal A, Frith JC, French MH, Swartz J, Gungor T, Harrity TW, Tamasi J, Rogers MJ, Feyen JH 2003 The ability of statins to inhibit bone resorption is directly related to their inhibitory effect on HMG-CoA reductase activity. J Bone Miner Res 18:88-96
10. Maeda T, Matsunuma A, Kawane T, Horiuchi N 2001 Simvastatin promotes osteoblast differentiation and mineralization in MC3T3-E1 cells. Biochem Biophys Res Commun 280:874-7
11. Murray SS, Tu KN, Young KL, Murray EJ 2002 The effects of lovastatin on proteasome activities in highly purified rabbit 20 S proteasome preparations and mouse MC3T3-E1 osteoblastic cells. Metabolism 51:1153-60
12. Maritz FJ, Conradie MM, Hulley PA, Gopal R, Hough S 2001 Effect of statins on bone mineral density and bone histomorphometry in rodents. Arterioscler Thromb Vasc Biol 21:1636-41
13. Ruiz-Gaspa S, Nogues X, Enjuanes A, Monllau JC, Blanch J, Carreras R, Mellibovsky L, Grinberg D, Balcells S, Diez-Perez A, Pedro-Botet J 2007 Simvastatin and atorvastatin enhance gene expression of collagen type 1 and osteocalcin in primary human osteoblasts and MG-63 cultures. J Cell Biochem 101:1430-8
14. Yao W, Farmer R, Cooper R, Chmielewski PA, Tian XY, Setterberg RB, Jee WS, Lundy MW 2006 Simvastatin did not prevent nor restore ovariectomy-induced bone loss in adult rats. J Musculoskelet Neuronal Interact 6:277-83
15. Bauer DC, Mundy GR, Jamal SA, Black DM, Cauley JA, Ensrud KE, van der Klift M, Pols HA 2004 Use of statins and fracture: results of 4 prospective studies and cumulative meta-analysis of observational studies and controlled trials. Arch Intern Med 164:146-52
16. Edwards CJ 2002 Statins and bone morphogenetic proteins: new pathways in bone formation. Ann Acad Med Singapore 31:245-7
17. Chan KA, Andrade SE, Boles M, Buist DS, Chase GA, Donahue JG, Goodman MJ, Gurwitz JH, LaCroix AZ, Platt R 2000 Inhibitors of hydroxymethylglutaryl-coenzyme A reductase and risk of fracture among older women. Lancet 355:2185-8
18. Wang PS, Solomon DH, Mogun H, Avorn J 2000 HMG-CoA reductase inhibitors and the risk of hip fractures in elderly patients. JAMA 283:3211-6
19. Meier CR, Schlienger RG, Kraenzlin ME, Schlegel B, Jick H 2000 HMG-CoA reductase inhibitors and the risk of fractures. JAMA 283:3205-10
20. Solomon DH, Finkelstein JS, Wang PS, Avorn J 2005 Statin lipid-lowering drugs and bone mineral density. Pharmacoepidemiol Drug Saf 14:219-26
21. Funkhouser HL, Adera T, Adler RA 2002 Effect of HMG-CoA reductase inhibitors (statins) on bone mineral density. J Clin Densitom 5:151-8
22. Schoofs MW, Sturkenboom MC, van der Klift M, Hofman A, Pols HA, Stricker BH 2004 HMG-CoA reductase inhibitors and the risk of vertebral fracture. J Bone Miner Res 19:1525-30
23. Pasco JA, Kotowicz MA, Henry MJ, Sanders KM, Nicholson GC 2002 Statin use, bone mineral density, and fracture risk: Geelong Osteoporosis Study. Arch Intern Med 162:537-40
24. Rejnmark L, Olsen ML, Johnsen SP, Vestergaard P, Sorensen HT, Mosekilde L 2004 Hip fracture risk in statin users--a population-based Danish case-control study. Osteoporos Int 15:452-8
25. Tikiz C, Unlu Z, Tikiz H, Ay K, Angin A, Onur E, Var A, Tuzun C 2004 The effect of simvastatin on serum cytokine levels and bone metabolism in postmenopausal subjects: negative correlation between TNF-alpha and anabolic bone parameters. J Bone Miner Metab 22:365-71
26. Nakashima A, Nakashima R, Ito T, Masaki T, Yorioka N 2004 HMG-CoA reductase inhibitors prevent bone loss in patients with Type 2 diabetes mellitus. Diabet Med 21:1020-4
27. Bjarnason NH, Riis BJ, Christiansen C 2001 The effect of fluvastatin on parameters of bone remodeling. Osteoporos Int 12:380-4
28. Scranton RE, Young M, Lawler E, Solomon D, Gagnon D, Gaziano JM 2005 Statin use and fracture risk: study of a US veterans population. Arch Intern Med 165:2007-12
29. LaCroix AZ, Cauley JA, Pettinger M, Hsia J, Bauer DC, McGowan J, Chen Z, Lewis CE, McNeeley SG, Passaro MD, Jackson RD 2003 Statin use, clinical fracture, and bone density in postmenopausal women: results from the Women's Health Initiative Observational Study. Ann Intern Med 139:97-104
30. Sirola J, Honkanen R, Kroger H, Jurvelin JS, Maenpaa P, Saarikoski S 2002 Relation of statin use and bone loss: a prospective population-based cohort study in early postmenopausal women. Osteoporos Int 13:537-41
31. van Staa TP, Wegman S, de Vries F, Leufkens B, Cooper C 2001 Use of statins and risk of fractures. JAMA 285:1850-5
32. De Leo V, Morgante G, la Marca A, Lanzetta D, Cobellis L, Petraglia F 2003 Combination of statins and hormone replacement therapy in postmenopausal women is associated with increased bone mineral density. Gynecol Endocrinol 17:329-32
33. Tikiz C, Tikiz H, Taneli F, Gumuser G, Tuzun C 2005 Effects of simvastatin on bone mineral density and remodeling parameters in postmenopausal osteopenic subjects: 1-year follow-up study. Clin Rheumatol 24:447-52
34. Ray WA, Daugherty JR, Griffin MR 2002 Lipid-lowering agents and the risk of hip fracture in a Medicaid population. Inj Prev 8:276-9
35. Hatzigeorgiou C, Jackson JL 2005 Hydroxymethylglutaryl-coenzyme A reductase inhibitors and osteoporosis: a meta-analysis. Osteoporos Int 16:990-8
36. Lupattelli G, Scarponi AM, Vaudo G, Siepi D, Roscini AR, Gemelli F, Pirro M, Latini RA, Sinzinger H, Marchesi S, Mannarino E 2004 Simvastatin increases bone mineral density in hypercholesterolemic postmenopausal women. Metabolism 53:744-8
37. Rosenson RS, Tangney CC, Langman CB, Parker TS, Levine DM, Gordon BR 2005 Short-term reduction in bone markers with high-dose simvastatin. Osteoporos Int 16:1272-6
38. Rejnmark L, Buus NH, Vestergaard P, Heickendorff L, Andreasen F, Larsen ML, Mosekilde L 2004 Effects of simvastatin on bone turnover and BMD: a 1-year randomized controlled trial in postmenopausal osteopenic women. J Bone Miner Res 19:737-44
39. Braatvedt GD, Bagg W, Gamble G, Davidson J, Reid IR 2004 The effect of atorvastatin on markers of bone turnover in patients with type 2 diabetes. Bone 35:766-70
40. Expert Panel on Detection Evaluation And Treatment of High Blood Cholesterol In Adults 2001 Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 285:2486-97
41. Lewiecki EM, Watts NB, McClung MR, Petak SM, Bachrach LK, Shepherd JA, Downs RW, Jr. 2004 Official positions of the international society for clinical densitometry. J Clin Endocrinol Metab 89:3651-5
42. Toh S, Hernandez-Diaz S 2007 Statins and fracture risk. A systematic review. Pharmacoepidemiol Drug Saf 16:627-40
43. Edwards CJ, Hart DJ, Spector TD 2000 Oral statins and increased bone-mineral density in postmenopausal women. Lancet 355:2218-9
44. Chung YS, Lee MD, Lee SK, Kim HM, Fitzpatrick LA 2000 HMG-CoA reductase inhibitors increase BMD in type 2 diabetes mellitus patients. J Clin Endocrinol Metab 85:1137-42
45. Montagnani A, Gonnelli S, Cepollaro C, Pacini S, Campagna MS, Franci MB, Lucani B, Gennari C 2003 Effect of simvastatin treatment on bone mineral density and bone turnover in hypercholesterolemic postmenopausal women: a 1-year longitudinal study. Bone 32:427-33
46. Nawrocki JW, Weiss SR, Davidson MH, Sprecher DL, Schwartz SL, Lupien PJ, Jones PH, Haber HE, Black DM 1995 Reduction of LDL cholesterol by 25% to 60% in patients with primary hypercholesterolemia by atorvastatin, a new HMG-CoA reductase inhibitor. Arterioscler Thromb Vasc Biol 15:678-82
47. Corsini A, Bellosta S, Baetta R, Fumagalli R, Paoletti R, Bernini F 1999 New insights into the pharmacodynamic and pharmacokinetic properties of statins. Pharmacol Ther 84:413-28