3A-04 - Comparative Effectiveness of Denosumab and Oral Bisphosphonates in UK and Danish Real World Data

Background: Different pharmacotherapies were efficacious in reducing the risk of osteoporotic fractures in randomized trials, but head-to-head comparisons in clinical practice are lacking.

Objectives: To assess the comparative effectiveness of denosumab (DMAB) vs. oral bisphosphonates (OBP) on the risk of fractures in postmenopausal women.

Methods: We conducted an active comparator, new user cohort study, with new switcher as sensitivity analysis. Data were obtained from the UK Clinical Practice Research Datalink (CPRD) and the linked Danish National Registries (DNR). We included women aged ≥50 years initiating DMAB or OBP in 2011-2018, with ≥1 year of data pre-index date.
New users of DMAB or OBP were defined by lack of treatment in the previous year. The index date was the date of initiating DMAB or OBP.
The new switcher cohort included patients who had switched from OBP to DMAB or from one OBP to another OBP in the year before index date. The index date was the date of treatment switch.
Women were followed until the earliest of any fracture, treatment switch or discontinuation, or three years of follow-up. Cox regression estimated the hazard ratio (HR) and 95% confidence interval (CI) comparing the risk of fracture between treatment groups, adjusted for remaining imbalanced covariates after propensity score weighting (PSW).
Unmeasured confounding was assessed with up to 47 negative control outcomes (NCOs). Unmeasured confounding was present if more than 5% of NCOs had a 95% CI with an upper CI < 0.75 or lower CI >1.30. These analyses revealed better comparability in the new switcher cohort than the new user cohort.

Results: In the new user cohort we included 200,179 patients from CPRD (DMAB: 6,528; OBP: 194,191) and 79,569 patients from DNR (DMAB: 4,276; OBP: 79,523). In both new user and switcher analyses, the DMAB cohort was older, had greater occurrence of prior fracture and higher cumulative BP use than the OBP cohort. There was a greater difference in proportion of patients on DMAB with prior fracture than those on OBP in the new switcher vs new user cohorts. In CPRD, there were fewer imbalanced covariates after PSW in the new switcher cohort.
The HRs for new users of DMAB vs. OBP on fracture risk were HR 0.98 (95% CI 0.86, 1.11) in CPRD; 0.99 (0.85, 1.16)] in DNR. Among those switching to DMAB compared to those switching to another OBP, HR was 0.72 (0.54, 0.97), available only for CPRD.

Conclusions: Risk of any fracture was similar among new users of DMAB vs OBP, but lower among people switching to DMAB vs those switching to OBP. Given limited comparability and local treatment guidelines that recommend DMAB as a second line treatment, the new user results should be interpreted with caution, as residual confounding could account for some of these findings.