Study Overview
Objective. To determine if romosuzumab, an antisclerostin antibody, is superior to alendronate in reducing the incidence of fracture in postmenopausal women with osteoporosis at high-risk for fracture.
Design. Multicenter, international, double-blind, randomized clinical trial.
Setting and participants. 4093 postmenopausal women with osteoporosis and a previous fragility fracture were enrolled from over 40 countries worldwide. Patients were eligible for the study if they were 55 to 90 years old and were deemed at high risk for future fracture based on bone mineral density (BMD) T score at the total hip or femoral neck and fracture history. This included T score ≤ –2.5 and ≥ 1 moderate or severe vertebral fractures or ≥ 2 mild vertebral fractures; T score ≤ –2.0 and either ≥ 2 moderate or severe vertebral fractures or proximal femur fracture within 3 to 24 months before randomization. Subjects with a history of prior use of medications that affect bone metabolism were excluded, as were those with other metabolic bone disease, vitamin D deficiency, uncontrolled metabolic disease, malabsorption syndromes, history of transplant, severe renal insufficiency, malignancy or severe illness.
Intervention. Patients were randomized to either subcutaneous romosuzumab 210 mg monthly or oral alendronate 70 mg weekly for 12 months. Following the 12-month double-blind period, all patients received open-label weekly alendronate until the end of the trial, with maintenance of blinding to the initial treatment assignment. Primary analysis occurred when all subjects had completed the 24-month visit and clinical fractures had been confirmed in at least 330 patients. All patients received daily calcium and vitamin D. Lateral radiographs of the thoracic and lumbar spine were obtained at screening and months 12 and 24. The BMD at the lumbar spine and proximal femur was evaluated by dual-energy x-ray absorptiometry at baseline and every 12 months thereafter. Serum concentrations of bone-turnover markers were measured in a subgroup of patients.
Main outcome measures. The primary outcomes were the incidence of new vertebral fracture and the incidence of clinical fracture at 24 months. Clinical fractures included symptomatic vertebral fracture and nonvertebral fractures. The secondary outcomes were the BMD at the lumbar spine, total hip, and femoral neck at 12 and 24 months, the incidence of nonvertebral fracture, and fracture category. Safety outcomes included the incidence of adjudicated clinical events, including serious cardiovascular adverse events, osteonecrosis of the jaw, and atypical femoral fracture. Serious cardiovascular events were defined as cardiac ischemic event, cerebrovascular event, heart failure, death, non-coronary revascularization and peripheral vascular ischemic event not requiring revascularization.
Analysis. An intention to treat approach was used for data analysis. For the incidence of fractures, the treatment groups were compared using a Cox proportional-hazards model and the Mantel-Haenszel method with adjustment for age (< 75 vs ≥ 75 years), the presence or absence of severe vertebral fracture at baseline, and baseline BMD T score at the total hip. Between-group comparisons of the percentage change in BMD from baseline were analyzed by means of a repeated-measures model with adjustment for treatment, age category, baseline severe vertebral fracture, visit, treatment-by-visit interaction, and baseline BMD. Percentage changes from baseline in bone turnover were assessed using a Wilcoxon rank-sum test. The safety analysis included cumulated incidence rates of adverse outcomes. Odds ratios and confidence intervals were estimated for serious cardiovascular adverse events with the use of a logistic regression model.
Main results. 2046 participants were randomized to the romosozumab group and 2047 to the alendronate group. A total of 3654 participants from both groups (89.3%) completed 12 months of the trial, and 3150 (77.0%) completed the primary analysis period. The treatment groups were similar in baseline age, ethnicity, and fracture history. The majority of patients in both groups were non-Hispanic (> 60%) and ≥ 75 years old (> 50%). The mean age of the patients was 74.3 years. Baseline mean bone mineral density T scores were –2.96 at the lumbar spine, –2.8 at the total hip, and –2.9 at the femoral neck.
After 24 months of treatment, 6.2% of patients in the romosozumab-alendronate group had a new vertebral fracture as compared to 11.9% in the alendronate-alendronate group. This represents a 48% lower risk (risk ratio 0.52, 95% confidence interval [CI] 0.4–0.66; P < 0.001) of new vertebral fractures with romosozumab. At the time of the primary analysis, romosozumab followed by alendronate resulted in a 27% lower risk of clinical fracture than alendronate alone (hazard ratio 0.73, 95% CI 0.61–0.88; P < 0.001). 8.7% of the romosozumab-alendronate group had a nonvertebral fracture versus 10.6% in the alendronate-alendronate group, representing a 19% lower risk with romosozumab (hazard ratio 0.81, 95% CI 0.66–0.99; P = 0.04). Hip fractures occurred in 2.0% of the romosozumab-alendronate group as compared with 3.2% in the alendronate-alendronate group, representing a 38% lower risk with romosozumab (hazard ratio 0.62, 95% CI 0.42–0.92; P = 0.02).
Patients in the romosozumab-alendronate group had greater gains in BMD from baseline at the lumbar spine (14.9% vs 8.5%) and total hip (7% vs 3.6%) compared to the alendronate-alendronate group. (P < 0.001 for all comparisons). At 12 months, romosozumab treatment resulted in decreased levels of bone resorption marker β-CTX and increased levels of bone formation marker P1NP. β-CTX and P1NP decreased and remained below baseline levels after transitioning to alendronate. In the alendronate-alendronate group, P1NP and β-CTX decreased within 1 month and remained below baseline levels at 36 months.
Overall, the adverse events and serious event rates were similar between the 2 treatment groups during the double-blind period with 2 exceptions. In the first 12 months, injection-site reactions were reported in 4.4% of patients receiving romosozumab compared to 2.6% in those receiving alendronate. Patients in the romosozumab group had an increased incidence of adjudicated serious cardiovascular outcomes during the double-blind period, 2.5% (50 of 2040 patients) compared to 1.9% (38 of 2014 patients) in the alendronate group. During the open-label period, osteonecrosis of the jaw occurred in one patient in each group. Two atypical femoral fractures occurred in the romosozumab-alendronate group, compared to 4 in the alendronate-alendronate group. During the first 18 months of the study, binding anti-romosozumab antibodies were observed in 15.3% of the romosozumab group, with neutralizing antibodies in 0.6%.
Conclusion. In postmenopausal woman with osteoporosis and high fracture risk, 12 months of romosozumab treatment followed by alendronate resulted in significantly lower risk of fracture than use of alendronate alone.