Osteoporosis: trials and tribulations

Abstract

A 30-50% reduction in fracture risk produced by a drug is biologically "worthwhile." The detection of this benefit, when truly present, is a challenge requiring large studies of 3-5 years' duration, because only a small number of women at risk actually sustain a fracture during this time. For example, in any year, fractures occur in 1-2 per 100 women approximately 65 years of age, 6-10 per 100 women approximately 75 years of age, and only 1-2 per 2,000 of the 15% of women < 60 years of age with osteoporosis. An appreciation of this low annual event rate is important because (1) it helps patients to understand their illness, (2) it determines the power of clinical trials, (3) it underscores the large numbers of patients that must be treated to prevent one fracture, and (4) it underscores the need for safety, particularly in groups at low absolute risk of fracture; all are exposed to drug side effects, and the vast majority derive no benefit from treatment because they would not have had a fracture without it, despite being at risk. Few studies have met the design requirements needed to identify the antifracture efficacy of a drug when it really exists, namely, (1) large patient samples randomized to treatment or placebo for 3-4 years, (2) blinding throughout follow-up, (3) statistical analyses of preplanned comparisons using intention to treat, and (4) avoidance of statistical analyses of associations discovered by post hoc analyses. Moreover, (5) few studies have assessed long-term safety and quality of life. Consequently, the uncertainty regarding efficacy and safety of available treatments may be more of a problem of the design, execution, and interpretation of the clinical studies than of the drugs themselves. In the reduction of vertebral fracture risk, the greatest optimism exists for hormone replacement therapy (HRT) and the bisphosphonates. HRT reduces bone turnover, increases bone mineral density (BMD), and decreases vertebral fracture rates by approximately 40%, even in women > 70 years of age. Reduction in hip fracture risk with HRT has been reported in observational studies. Two rigorously conducted studies provide credible evidence that the bisphosphonate alendronate reduces the risk of vertebral and hip fractures by approximately 40-50%. Etidronate, calcitonin, and 1,25-dihydroxyvitamin D3 may reduce risk of vertebral fracture; however, problems in study design leave uncertainty. Although 2 trials using fluoride suggest a reduction in fracture rates, the more rigorously conducted trials do not, despite having adequate power to do so. Calcium supplements are likely to slow bone loss, but reduction in fracture risk is uncertain. Vitamin D and calcium supplementation reduce risk of hip fracture in nursing home residents but not in community residents. There have been no studies of the efficacy of any treatment to prevent hip or vertebral fractures in men or in corticosteroid-related osteoporosis. The treatment of osteoporosis is becoming a reality. HRT and the bisphosphonates, particularly alendronate, appear to be the best options at present.