OBJECTIVES: To assess the relative efficacy of abaloparatide compared with osteoporosis treatment options (alendronate, denosumab, ibandronate, raloxifene, risedronate, romosozumab, strontium ranelate, teriparatide, zoledronic acid).

METHODS: PubMed^®, Embase^® and Cochrane Central Register of Controlled Trials were searched for all randomized controlled trials published prior to December 12, 2016 including postmenopausal osteoporotic women with and without prior fractures. Selection of trials for inclusion in the Network meta-analysis (NMA) were based on populations (inclusion/exclusion criteria), interventions (dose/frequency) and outcomes (fracture assessment). NMA were conducted by fracture sites with relative risk (RR) of fractures as the main clinical endpoint.

RESULTS: For vertebral fractures (VF) and nonvertebral fractures (NVF), 17 studies informed a network of 11 treatments, and 20 studies informed a network of 10 treatments, respectively. For VF, abaloparatide had the greatest effect relative to placebo (RR 0.13; 95% CrI: 0.04-0.34), with estimates ranging from 0.27 for romosozumab to 0.71 for strontium ranelate. For NVF, abaloparatide had a greater risk reduction versus placebo (RR 0.52; 95% CrI: 0.29-0.88) and was most effective (with a probability of 0.74) versus teriparatide (RR 0.67; 95% CrI: 0.48-0.93) and romosozumab (RR 0.74; 95% CrI: 0.52-1.05). In a further evaluation of specific fracture sites, 9 studies reporting wrist fractures informed a network of 7 treatments. Abaloparatide was associated with the greatest effect versus placebo (RR 0.35; CrI: 0.14-0.80) and reduced the risk of fractures versus teriparatide (RR 0.44; CrI: 0.17-1.02) and denosumab (RR 0.42; CrI: 0.16-1). The network meta-analyses illustrated a good level of agreement with the direct trial evidence and direct pairwise comparisons.

CONCLUSIONS: Based on the current NMA, abaloparatide treatment results in a reduction in RR of both vertebral and nonvertebral fractures in women with and without prior fractures versus placebo in comparison with other treatment options. Generalizability is limited to the trial populations included in the NMA.