Teriparatide reduces the fracture risk associated with increasing number and severity of osteoporotic fractures.

The relationship between prior fractures and risk of new fractures was evaluated in 931 postmenopausal women with prevalent vertebral fractures randomized to daily placebo or teriparatide (20 mug) in the Fracture Prevention Trial. The median observation time was 21 months. Among placebo patients with one, two, or three or more prevalent vertebral fractures, 7%, 16%, and 23%, respectively, developed vertebral fractures (by Cochran-Armitage trend test, P < 0.001), and 3%, 9%, and 17% developed moderate or severe vertebral fractures (P < 0.001). Among placebo patients with mild, moderate, or severe prevalent vertebral fractures, 10%, 13%, and 28%, respectively, developed vertebral fractures (P < 0.001), and 4%, 8%, and 23% developed moderate or severe vertebral fractures (P < 0.001). Among placebo patients with zero, one, or two or more prior nonvertebral fragility fractures, 4%, 8%, and 18%, respectively, developed nonvertebral fragility fractures (P < 0.001). In the teriparatide-treated group, there was no significant increase in vertebral or nonvertebral fracture risk in these subgroups. In summary, the number and severity of prevalent vertebral fractures independently predicted the risk for new vertebral fractures, and the number of prior nonvertebral fractures predicted the risk for new nonvertebral fractures in placebo patients. However, in teriparatide-treated patients, the increased fracture risk associated with prior number and severity of fracture was not observed.

Prognostic utility of a semiquantitative spinal deformity index.

The semiquantitative spinal deformity index (SDI) is a summary measure of the vertebral fracture status of the spine incorporating both the number and severity of vertebral fractures. For each vertebra, a visual semiquantitative grade of 0, 1, 2, or 3 is assigned for no fracture or mild, moderate, or severe fracture, respectively, and the SDI is calculated by summing the fracture grades of all vertebrae (T4 to L4). We investigated the effect of prevalent vertebral fracture number and severity, as integrated by the SDI, on 3-year vertebral fracture risk by performing logistic regression modeling with data from the MORE trial. There was a striking linear relationship between baseline SDI and the model-based vertebral fracture risk estimates, with a near-perfect correlation (r = 0.98, P < 0.001). However, the SDI may be overly simplistic, as a given SDI value can be attained through differing vertebral fracture scenarios (i.e., an SDI of 3 can be realized three ways), each corresponding to potentially different vertebral fracture risk. To address this issue, a second, more complex model was constructed that included individual predictor variables for number of mild, number of moderate, and number of severe prevalent vertebral fractures. The model-based risk estimates for vertebral fracture using the SDI and the more complex model were highly correlated (r = 0.91, P < 0.001), giving almost identical values up to an SDI of 5. Thus, for most clinical scenarios, it is not necessary to consider the particular fracture configuration that led to a given SDI score for predicting a patient's future vertebral fracture risk. These results validate the SDI as an accurate tool for assessing future vertebral fracture risk; patients with greater baseline SDI had greater future risk for vertebral fractures.

Reduction in vertebral fracture risk in teriparatide-treated postmenopausal women as assessed by spinal deformity index.

Teriparatide is the first bone-building drug available for the treatment of osteoporosis. We investigated the efficacy of this compound as assessed by spinal deformity index (SDI) using data from the Fracture Prevention Trial (FPT). The FPT was a randomized, double blind trial of placebo versus teriparatide 20 microg (TPTD20) versus teriparatide 40 microg (TPTD40) administered by daily self-injection. Patients included in the current analyses were those patients from the placebo (n = 398) and TPTD20 (the approved dose, n = 403) groups with baseline and follow-up radiographs and at least one vertebral fracture at baseline. For each vertebra, a visual semiquantitative grade of 0, 1, 2, or 3 was assigned for no fracture or mild, moderate, or severe fracture, respectively; the SDI was calculated by summing the fracture grades of all T4 to L4 vertebrae. The mean SDI increased in the placebo and TPTD20 groups by 0.485 and 0.134, respectively (P < 0.001). The proportions of patients with SDI increases >1, >2, and >3 were reduced by 85%, 80%, and 80%, respectively. In the placebo group, increasing baseline SDI was correlated with the mean increase in SDI during the trial (r = 0.080, P = 0.01), consistent with the progressive natural history of osteoporosis. However, in the TPTD20 group, increasing baseline SDI was not correlated with the mean increase in SDI during the trial (P = 0.297) indicating that teriparatide mitigated or eliminated the increased risk associated with increasing fracture burden. Increases in SDI during the trial were associated with increasing proportions of patients with new or worsening back pain and greater mean height loss (P < 0.0001), demonstrating an association with important clinical consequences. The results indicate that teriparatide greatly reduced the increase in fracture burden in the FPT and mitigated or eliminated the risk for future fractures imparted by increasing baseline fracture burden.

Relationship between changes in biochemical markers of bone turnover and BMD to predict vertebral fracture risk.

UNLABELLED: The change in BMD is a poor predictor of vertebral fracture risk after raloxifene treatment. One-year percent change in bone turnover and BMD was used to predict vertebral fracture risk. The percent change in osteocalcin was determined to be a better predictor of vertebral fracture risk than BMD. INTRODUCTION: The association between baseline BMD and fracture risk is well understood. However, the relationship between changes in BMD and fracture risk is not well defined. It has previously been demonstrated that BMD change was a poor predictor of vertebral fracture risk in raloxifene-treated women, whereas bone turnover markers were significantly associated with fracture risk. In the current analysis, we explore the prediction of vertebral fracture risk using changes in both BMD and bone turnover. MATERIALS AND METHODS: The Multiple Outcomes of Raloxifene Evaluation (MORE) trial was a randomized, placebo-controlled trial of 7705 women with osteoporosis treated with raloxifene 60 or 120 mg/day for 3 years. Markers of bone turnover were measured in one-third of the study population (n = 2503), and the present analyses include these women. Logistic regression models were constructed using one-year percent changes in BMD and bone turnover and relevant baseline demographics to predict the risk of vertebral fracture with pooled raloxifene therapy at 3 years. All covariates were standardized before modeling to facilitate direct comparisons between changes in BMD and bone turnover. RESULTS AND CONCLUSION: Prevalent vertebral fracture status (p < 0.0001), baseline lumbar spine BMD (p < 0.0001), and number of years postmenopausal (p = 0.0005) were independent predictors of fracture risk in raloxifene-treated patients. Therapy-by-change in femoral neck BMD (p = 0.02) and therapy-by-change in osteocalcin (OC; p = 0.01) were also significant for all treatment groups, indicating that changes in BMD and OC have different effects on fracture risk for the placebo and pooled raloxifene groups. The final model included significant baseline variables and change in OC (p = 0.01), whereas change in femoral neck BMD was not significant. After adjustment of each significant baseline variable, the percent change in OC was better able to predict the reduction in vertebral fracture risk than the percent change in femoral neck BMD in patients treated with raloxifene.

Efficacy and safety of raloxifene 60 milligrams/day in postmenopausal Asian women.

In healthy Caucasian postmenopausal women, raloxifene increases bone mineral density (BMD), decreases biochemical markers of bone turnover, and lowers low-density lipoprotein (LDL) cholesterol, without effects on high-density lipoprotein (HDL) cholesterol and triglycerides. This randomized, double-blind study examines the effects of raloxifene 60 mg/d (n = 483) or placebo (n = 485) in healthy postmenopausal Asian women (mean age 57 yr) from Australia, Hong Kong, India, Indonesia, Malaysia, Pakistan, Philippines, Singapore, Taiwan, and Thailand. Serum osteocalcin, serum N-telopeptide, total cholesterol, HDL cholesterol, LDL cholesterol, and triglycerides were assessed at baseline and 6 months. Lumbar spine BMD was measured at baseline and 1 yr in 309 women from 4 countries. Clinical adverse events were recorded at each interim visit. At 6 months, raloxifene 60 mg/d significantly decreased osteocalcin, N-telopeptide, total cholesterol, and LDL cholesterol by medians of 15.9%, 14.6%, 5.3%, and 7.7%, respectively, from placebo. Changes in HDL cholesterol and triglycerides were similar between raloxifene and placebo. Raloxifene 60 mg/d increased mean lumbar spine BMD (1.9%) from placebo at 1 yr (P = 0.0003). The incidences of hot flashes (placebo 3.5%, raloxifene 5.6%, P = 0.12), and leg cramps (placebo 2.7%, raloxifene 4.3%, P = 0.16) were not different between groups. No case of venous thromboembolism was reported. The effects of raloxifene 60 mg/d on bone turnover, BMD, and serum lipids in healthy postmenopausal Asian women were similar to that previously reported in Caucasian women.

How conservative is Fisher's exact test? A quantitative evaluation of the two-sample comparative binomial trial.

The debate as to which statistical methodology is most appropriate for the analysis of the two-sample comparative binomial trial has persisted for decades. Practitioners who favor the conditional methods of Fisher, Fisher's exact test (FET), claim that only experimental outcomes containing the same amount of information should be considered when performing analyses. Hence, the total number of successes should be fixed at its observed level in hypothetical repetitions of the experiment. Using conditional methods in clinical settings can pose interpretation difficulties, since results are derived using conditional sample spaces rather than the set of all possible outcomes. Perhaps more importantly from a clinical trial design perspective, this test can be too conservative, resulting in greater resource requirements and more subjects exposed to an experimental treatment. The actual significance level attained by FET (the size of the test) has not been reported in the statistical literature. Berger (J. R. Statist. Soc. D (The Statistician) 2001; 50:79-85) proposed assessing the conservativeness of conditional methods using p-value confidence intervals. In this paper we develop a numerical algorithm that calculates the size of FET for sample sizes, n, up to 125 per group at the two-sided significance level, alpha = 0.05. Additionally, this numerical method is used to define new significance levels alpha(*) = alpha+epsilon, where epsilon is a small positive number, for each n, such that the size of the test is as close as possible to the pre-specified alpha (0.05 for the current work) without exceeding it. Lastly, a sample size and power calculation example are presented, which demonstrates the statistical advantages of implementing the adjustment to FET (using alpha(*) instead of alpha) in the two-sample comparative binomial trial.

Anti-vertebral fracture efficacy of raloxifene: a meta-analysis.

In the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, raloxifene reduced the risk of vertebral fracture. However, a systematic analysis of the anti-vertebral fracture efficacy of raloxifene, which includes the results of newly reported studies, has not been performed. A meta-analysis was carried out using all randomized, double-blind, placebo-controlled trials to determine whether the reduction in the risk for vertebral fracture, reported with raloxifene, was consistent among studies, and to define more accurately the point estimate of the odds ratio. Three prevention studies, two arms of the MORE trial, and three additional treatment studies in which fracture data were available from prospectively scheduled spinal radiographs were included in the analysis. A systematic review of the literature (MedLine, EMBASE) confirmed that no studies with raloxifene had been excluded from this analysis. The effects of raloxifene 60 mg/day (RLX60) and 120 mg/day pooled with 150 mg/day (RLX120/150) were analyzed by intention to treat. There was no significant heterogeneity among the studies included in the meta-analysis. Odds ratio estimates (95% CI) were 0.60 (0.49, 0.74) for RLX60 and 0.51 (0.41, 0.64) for RLX120/150. From these data we infer that raloxifene consistently reduces the risk of vertebral fracture in postmenopausal women.