Multiple vertebral fractures after antiosteoporotic medications discontinuation: A comparative study to evaluate the potential rebound effect of denosumab

Background

Discontinuation of anti-osteoporotic medications (AOM), except for bisphosphonates (BP), is not favorable for the bone, being especially negative for Prolia® (60 mg denosumab-DMAb). DMAb withdrawal leads to a rapid and significant increase in bone turnover markers, and to an important loss in bone mineral density, which has been associated with an increased risk of multiple vertebral fractures (MVF).

Objective

To assess the risk of MVF (≥2 VF) recorded at the same time) after discontinuation of different AOM.

Methods

A case-control analysis nested in a cohort of new users of DMAb, BP, Teriparatide (TPTD), Strontium Ranelate (SrRan), or selective estrogen receptor modulators (SERM), aged ≥18 years from 2011 to 2018 with ≥1 year of prior available data, was performed using the Pharmacoepidemiological Research Database for the Public Health System (BIFAP) in Spain. Cases were first MVF recorded after AOM initiation (index date). Up to 4 controls per case, matched on index date, age, sex, and location, were randomly selected among non-cases from the cohort. Adjusted conditional OR (AOR) and 95 % CI: between discontinuation of a given AOM (supply of the last prescription ended >90 days before the index date) and occurrence of MVF was assessed compared with their current use and alternatively, with discontinuation of BP, among individuals who did not switch therapy in the study.

Results

A total of 532 incident cases of MVF were identified and matched to 2121 controls (86 % women; median age 73 years). AOR of MVF after DMAb discontinuation was 2.82 (1.73–4.60) compared with DMAb current use. No risk was seen for the other AOM. The AOR was highest between 3 and 9 months after discontinuation of denosumab (8.58; 3.98–18.48) and after >1 year of cumulative use (5- and 11-times increased risk when discontinuing after 1–2 years and 2–5 years of use, respectively). Compared with BP discontinuers, discontinuation of DMAb (2.73, 1.66–4.50), TPTD (2.06, 1.09–3.88) and SrRan (1.93, 1.23–3.05) showed an increased risk of MVF; current use of DMAb showed no protective effect (1.44; 0.95–2.17).

Conclusions

Discontinuation of DMAb was associated with an immediate increased risk of MVF, especially after longest treatments compared with patients who continued therapy or discontinued BP. Although there were increased risks after discontinuation of other AOM in comparison with first- line therapy (BP), these were not found when the reference was current users. Confounding by indication cannot be discarded. Larger studies should investigate reasons for discontinuation and preventive retreatment options.