Effectiveness of Osteoporosis Medications on the Risk of Fracture: A French Population-based Cohort Study

Background: Although medications indicated for osteoporosis (OP) have proven to be effective in clinical trials, data on their effectiveness in real-world settings are currently limited.

Objectives: To assess the effectiveness of OP medications on fracture risk in real-world settings.

Methods: A cohort study was conducted using data of the French nationwide claims database, the SNDS, including all women ≥ 55 years, initiating an OP medication (i.e., denosumab, zoledronic acid (ZA), oral bisphosphonates (OBis), estrogens, teriparatide and raloxifene) between 2014 and 2016 according to a hierarchical approach that follows the relative order of drugs commercialization in France. Included patients were treated for ≥ 6 months, and followed up to 2 years. Fracture risk was estimated within treatment cohorts, comparing incidence rates between the 3 months following index date (baseline fracture risk) and the subsequent 12, 18 and 24 months. Results of the "as treated" own-control analysis are presented in terms of Incidence Rate Ratio (IRR) and 95% Confidence Interval (95%CI).

Results: Among the 200,496 patients of the cohort, 33% initiated denosumab, 26% OBis, 21% ZA, 10% estrogens, 6% raloxifene and 4% teriparatide. Patients were on average between 63 years old (SD ± 8 years) in the estrogen group and 76 years old (± 9 years) in the teriparatide group. Prior OP treatment was highest in the denosumab group (55%). Baseline fracture risk was highest in the ZA (hip) and teriparatide (clinical vertebral) cohorts, and lowest in the estrogen cohort. In comparison with the 3-month baseline period, the vertebral fractures were significantly reduced within the subsequent 12, 18 and 24 months with denosumab (IRRs: 0.7 [0.5; 0.8], 0.6 [0.5; 0.7] and 0.6 [0.5; 0.7], respectively), ZA (IRR: 0.4 [0.3; 0.4] for all 3 exposure periods), teriparatide (IRRs: 0.4 [0.3; 0.6], 0.4 [0.2; 0.5] and 0.3 [0.2; 0.5]) and OBis (IRRs: 0.6 [0.4; 0.8] for all 3 exposure periods). Compared to the 3-month baseline period, the incidence of hip fracture was significantly lower with denosumab within the subsequent 18 and 24 months (IRRs: 0.8 [0.6 ; 0.9]) but significantly higher for the OBis for all 3 exposure periods (IRRs: 1.6 [1.1; 2.2] in the next 12 months and 1.7 [1.2; 2.3] in the next 18 and 24 months); no significant difference was observed with ZA and teriparatide. A significant decrease in the incidence of peripheral fractures was observed only in the denosumab group for the 3 periods (IRRs: 0.85 [0.75; 0.96] in the next 12 months, 0.83 [0.73; 0.94] in the next 18 months and 0.82 [ 0.73; 0.92] in the next 24 months).

Conclusions: Overall, vertebral fracture incidence was reduced with treatment by OP therapies. Significant decreases in hip fracture and non-hip/non-vertebral fractures were only observed with denosumab treatment.