Overview Page 1

Introduction
Despite important advances in recent years in prevention, diagnosis, and treatment, osteoporosis remains a major public health problem. The disease is extremely common: approximately 50% of all postmenopausal women will eventually experience fractures. Approximately one in three Caucasian women will experience fractures of the spine and one in six will have hip fractures.¹ Hip fractures in particular are associated with significant increases in morbidity and mortality.² Initial osteoporotic fractures often lead to a tragic downward spiral of increasing frailty, declines in physical activity and performance, and further increases in fracture risk.¹

Most of these fractures occur in women with low bone mineral density (BMD), a condition that is readily detectable with the use of dual energy radiograph absorptiometry (DXA), which is also useful for monitoring the effectiveness of therapy. Bone loss accelerates at menopause, independently of chronological age.³ Over her lifetime, a woman's BMD declines by 30-50%, and each standard deviation decrease of BMD is associated with a doubled risk of fracture.⁴ For this reason, the National Osteoporosis Foundation recommends BMD measurements for all women aged 65 or older and for postmenopausal women below age 65 who have one or more additional risk factors.

Osteoporosis is generally defined as having a BMD 2.5 standard deviations (SD) below the mean for young adults.⁵ A BMD between 1 and 2.5 SDs is defined as osteopenia, or low bone mass. The number of SDs above or below the mean is defined as the T-score (Figure 1). Treatment is currently recommended for women with BMD T-scores <-2 if no additional risk factors are present and < -1.5 if one or more additional risk factors (especially prior fractures) are present. Women with osteopenia who do not meet these criteria may also be candidates for treatment if they have several risk factors. For example, a patient with osteopenia who is on corticosteroid therapy should be considered at risk. However, these are recommendations only; the decision to treat should be a joint one between physician and patient.

Treatment Goals and Therapeutic Options
The goal of osteoporosis treatment is to increase BMD; the end point of treatment is a reduction in overall fracture risk. The ideal therapy should reduce risk not only for spinal fractures, but also for fractures of the hip, tibia, fibula, and wrist. Treatment should be effective across all age groups and different degrees of low bone density. Efficacy should be rapid and sustained for the duration of therapy. The treatment should also be safe and well tolerated. Current treatment options for osteoporosis and their dosage forms are listed in Table 1. Recombinant parathyroid hormone has recently been approved as a daily subcutaneous injection. The number of available treatment options for osteoporosis has complicated the decision-making process. The results of the Women's Health Initiative (WHI) trial have recently added a further complication: they suggest that long-term (> 4 years) hormone replacement therapy (HRT) may not be appropriate for most women.⁶ As the following discussion shows, the quality of the data supporting the antifracture efficacy of available therapies differs considerably from one therapy to another.



Calcium and Vitamin D
Calcium and vitamin D are essential as adjunctive therapies to the more potent antiresorptive therapies. They are also commonly used as controls in osteoporosis trials. Calcium is the simplest and least costly preventive therapy for osteoporosis, but its benefits on BMD are modest. Shea et al conducted a meta-analysis of controlled trials of the effects of calcium supplementation on bone density and fractures in postmenopausal women.⁷ They summarized the results of 15 trials that randomized postmenopausal women to calcium supplementation or usual calcium intake. These trials followed patients for at least one year and reported effects on BMD or on fracture incidence. They found that calcium supplementation had a small positive effect on BMD, and there was a trend toward reduction in vertebral fractures. However, conclusions could not be drawn about the effect of calcium on nonvertebral fractures.

In one of the studies summarized by Shea et al, Chapuy et al studied the effects of vitamin D and calcium supplementation on the frequency of hip and other nonvertebral fractures in 3270 healthy, ambulatory elderly women (mean age: 84 +/- 6 yrs).⁸ Each day for 18 months 1634 women received a dose of tricalcium phosphate containing 1.2 g of elemental calcium and 20 µg (800 IU) of vitamin D3. The control group (n=1636) received a double placebo. Among the women who received the supplements, the number of hip fractures was 43% lower (p=0.043), and the number of nonvertebral fractures was 32% lower (p=0.015) compared with those who received placebo. The bone density of the proximal femur increased 2.7% in the treatment group and decreased 4.6% in the placebo group (p < 0.001). This trial was conducted in institutionalized patients with marginal serum vitamin D levels; thus, this study population may not be representative of elderly postmenopausal women. The results of a study by Lips et al did not show a significant effect on hip fracture, but this study used a lower dose of vitamin D (400 IU).⁹

Hormone Replacement Therapy (HRT)
Menopausal Hormone Therapy (MHT)
Estrogen has been used for the prevention or treatment of numerous health problems in postmenopausal women for approximately 50 years. Many health agencies and support groups endorsed its use, particularly for the prevention of heart disease, relief of menopausal symptoms, and the prevention and treatment of osteoporosis. The Framingham Health Study showed that women who took estrogen, both current and past users, had a lower incidence of hip fractures.¹⁰ However, the recent findings from the Women's Health Initiative (WHI) study led to a reexamination of the indications for HRT.⁶ The study was halted in 2001 when the risk-benefit ratio for treatment began to favor the risk. The study showed that women receiving HRT (conjugated equine estrogens, 0.625 mg/day plus medroxyprogesterone acetate 2.5 mg) had a 34% decrease in hip and vertebral fractures and a reduction in risk of all osteoporotic fractures of approximately 24%. However, there were also slight increases in the risk of heart attacks, strokes, blood clots, and breast cancer in the treatment group. The risk of invasive breast cancer was increased by approximately 25%, but HRT did not increase all-cause mortality. Overall, the WHI study population was generally healthy and representative, but it should be noted that two-thirds of the enrolled women were over age 60, which is rather high for a study of prevention in postmenopausal women.

Clearly, clinicians can no longer routinely recommend long-term therapy with HRT. Indeed, the concept of "hormone replacement therapy" is being reevaluated, because the term itself has tended to mislead both physicians and patients. The National Institutes of Health (NIH) now recommends use of the term "menopausal hormone therapy" (MHT) in recognition of the fact that treatment with hormones is not a "replacement" for premenopausal hormones and does not restore youthful physiology.

Treatment of vasomotor and urogenital menopausal symptoms remains the primary indication for hormone therapy. While many hormonal products are FDA-approved for the prevention of postmenopausal osteoporosis, the risks associated with these forms of therapy must be considered. How long should MHT be used? The WHI study suggests that two to four years may be a reasonable limit on therapy duration, because results started to diverge significantly between the two WHI groups during this interval. A general rule is to use MHT for the shortest possible time at the lowest possible dose. After stopping MHT, an antiresorptive therapy, such as a bisphosphonate, should be considered to prevent the rapid bone loss seen when estrogen levels decline or therapy is withdrawn.

Raloxifene HCl is the first SERM approved for the prevention and treatment of osteoporosis. Raloxifene is a nonhormonal agent that binds with high affinity to estrogen receptors, producing estrogen-agonist effects on bone and estrogen-antagonist effects on endometrial and breast tissue.¹¹ The rationale for using SERMs is to obtain the beneficial effects of estrogen on bone while avoiding or minimizing undesirable effects on the breast and uterus.

Cranney et al conducted a meta-analysis of seven published and unpublished randomized trials of raloxifene versus placebo.¹² Both treatment and placebo groups in these trials received calcium and vitamin D supplementation. BMD was measured for at least one year. The investigators found positive effects on bone density in the group receiving raloxifene; these increased over time and were independent of dose. The differences between raloxifene and placebo were statistically significant for total body, lumbar spine, combined forearm, and combined hip BMDs (p < 0.01). Compared with placebo, raloxifene resulted in a slightly increased rate of withdrawal from therapy due to adverse effects (p =0.05).

The results of the meta-analysis were rendered somewhat problematic by the inclusion of a very large trial, the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, which was a three-year trial of 7705 postmenopausal women (mean age 67 yrs; range: 31-80).¹³ Patients were randomized to 60 mg/day or 120 mg/day of raloxifene or to placebo. Vertebral fractures were determined radiographically at baseline and at 24 and 36 months. Nonvertebral fractures were determined at six-month interim visits. BMD was determined annually by DXA.

Evaluable radiographs were obtained from 6828 women. At 36 months, 7.4% of enrolled patients had at least one new vertebral fracture, including 10.1% of placebo patients, 6.6% of the 60 mg/day raloxifene patients, and 5.4% of the 120 mg/day raloxifene patients. The risk of vertebral fracture was reduced in both treatment groups: 30% for 60 mg/day and 50% for 120 mg/day. The overall reduction in risk for vertebral fractures was 40% (p < 0.01). The risk of vertebral fracture was reduced both in women who did and did not have existing fractures. The risk reduction for nonvertebral fractures was not statistically significant (p=0.27). The increase in BMD seen with raloxifene was modest and smaller than that seen with other therapies. This raises the possibility that factors other than BMD, such as bone quality, may account for the observed reduction in vertebral fractures. In the MORE study, women receiving raloxifene had a significantly increased risk of venous thromboembolism compared with placebo.

Calcitonin is a polypeptide hormone secreted by the C cells of the parafollicular tissue of the thyroid gland; it acts on bone by inhibiting bone resorption by osteoclasts. It also appears to have an analgesic effect in women with vertebral fractures.¹⁴

Most of the randomized trials used a synthetic version of salmon calcitonin, which is 40 to 50 times more potent than human calcitonin and available as a nasal spray. The recommended dose is one spray (200 IU) per day intranasally, alternating nostrils daily. However, there is some concern that prolonged exposure to calcitonin may result in down-regulation of calcitonin receptors on osteoclasts, which could reduce the therapeutic effect.¹⁵ Some commentators have recommended intermittent administration to prevent development of drug resistance. Currently, the National Osteoporosis Foundation (NOF) recommends calcitonin as an alternative to HRT/MHT or alendronate when patients cannot tolerate or fail therapeutically on these agents.¹⁶

Cranney et al conducted a meta-analysis of 30 studies that randomized women to calcitonin or placebo (or calcium and/or vitamin D) and followed patients for at least one year.¹⁷ The pooled analysis of the four studies reporting vertebral fractures showed that calcitonin reduced the risk of vertebral fractures by 54% (p = 0.02). For nonvertebral fractures the results were not significant.

The vertebral fracture results come from three small studies suggesting large treatment effects and one large study with more modest results. The last trial was the Prevent Recurrence of Osteoporotic Fractures (PROOF) study, which was a five-year, randomized trial of 1255 postmenopausal women, most of whom had severe osteoporosis.¹⁸ Subjects were randomly assigned to receive calcitonin nasal spray (100, 200, or 400 IU) or placebo daily. All subjects received daily supplements of calcium (1000 mg) and vitamin D (400 IU). The primary endpoint was the incidence of new vertebral fractures. Over a period of five years, 1108 subjects had at least one follow-up radiograph. Of these, 783 completed three years of treatment and 511 completed five years.

The risk of vertebral fractures was reduced by 33% (p =0.03) in women taking the 200-IU dose of calcitonin. In the 100-IU-dose and 400-IU-dose groups, the risk of vertebral fractures was reduced by 15% and 16%, respectively (not significant). Among the 817 women with one to five prevalent vertebral fractures at enrollment, the risk of new vertebral fractures was reduced by 36% (p = 0.03). Lumbar spine bone mineral density increased modestly from baseline in all treatment groups (1.0% to 1.5%; p < 0.01).

Results of the PROOF study are rendered problematic by the fact that 59% of the subjects were discontinued early from the trial, which casts some doubt on the validity of the statistical analysis. Furthermore, while there was a significant risk reduction with 200 IU of calcitonin, the risk reductions with 100 IU and 400 IU were not significant, a puzzling result. Furthermore, the PROOF study was only partly blinded, because both physicians and subjects were aware of the BMD results as the trial progressed.

Based on the quality of available data and the magnitude of the clinical benefit obtained with these drugs, this is truly the "era of the bisphosphonates." Bisphosphonates act by decreasing bone turnover, and therefore bone loss, and by increasing bone mineralization and density. These agents represent major additions to the therapeutic armamentarium for the prevention and treatment of osteoporosis. Alendronate sodium was the first drug of this class to receive FDA approval for the prevention and treatment of osteoporosis, with subsequent long-term studies yielding powerful data regarding efficacy and safety. Approximately four million patients worldwide are now taking this agent.

Cranney et al conducted a meta-analysis of 11 trials of alendronate that randomized women to alendronate or placebo and measured BMD for at least one year.¹⁹ The pooled risk reduction for vertebral fractures in patients given alendronate was 48%. For nonvertebral fractures in patients given 10 mg of alendronate or more, the risk reduction was 49%; this benefit was appreciably greater than that seen with the 5-mg dose. The investigators found similar risk reductions across all nonvertebral fracture types. Alendronate also demonstrated significant increases in BMD that were directly proportional with dose and length of treatment. After two to four years of treatment with 10 mg or more of alendronate, the pooled estimate of the difference in lumber spine BMD between alendronate and placebo was 7.45%; for the hip, 5.6%; for the forearm, 2.08%; and for the total body, 2.73%. The investigators concluded that alendronate increases BMD in both early postmenopausal women and those with established osteoporosis, while reducing the rate of vertebral fractures over two to three years of treatment. Alendronate also reduced nonvertebral fractures in women without prevalent fractures who have BMDs below the WHO threshold for osteoporosis.

The largest study included in the above meta-analysis was the Fracture Intervention Trial (FIT).²⁰ This trial had two treatment arms. The first treatment arm, the Vertebral Fracture Arm (VFA), included 2027 women with pre-existing vertebral fractures. The second treatment arm, the Clinical Fracture Arm (CFA), included 1631 women without vertebral fractures at baseline, but with a femoral neck T-score of less than -2.5. The magnitude of reduction of fracture incidence with alendronate was similar in both groups. In the VFA, the subjects (mean age: 71) were followed for three years. In the alendronate group, new vertebral fractures were reduced by 47%, multiple vertebral fractures were reduced by 90%, and hip fractures were reduced by 51% compared with placebo. In the CFA, the subjects (mean age: 69) were followed for four years. In the alendronate group, vertebral fractures were reduced by 49%, multiple vertebral fractures were reduced by 60%, and hip fractures were reduced by 56% compared with placebo.