(C-095) Trajectories of Osteoporosis Pharmacotherapy among Older Treatment-Naïve Females

Background: Denosumab has been listed on the Ontario (Canada) Drug Benefit formulary since 2012/02 in females aged ≥65 years and who failed, did not tolerate, or were contraindicated to receive oral bisphosphonates. Abrupt discontinuation or delays in denosumab therapy may lead to increased bone turnover and multiple vertebral fractures. Therefore, continued therapy or transition to another osteoporosis pharmacotherapy (e.g. oral bisphosphonates) is an important consideration with denosumab.

Objectives: To identify trajectories of denosumab and oral bisphosphonate therapy among female initiators of denosumab and describe their characteristics.

Methods: We identified all community-dwelling female residents aged ≥67 years who initiated denosumab between 2013 and 2021 in Ontario, Canada. Patients with history of any osteoporosis pharmacotherapy within 2 years prior to denosumab initiation were excluded. We assessed whether patients were on denosumab, oral bisphosphonate or no therapy over 90-day observation windows up to 7 years or until 2022/12. Windows within 2020/03/15 to 2020/09/30 were excluded to account for healthcare disruptions from COVID-19; and observation windows were censored upon patient death, entry into long-term care or end of follow-up. A group-based multi-trajectory model was used to identify distinct patterns of osteoporosis pharmacotherapy. Medical and pharmacy claims within the year prior to denosumab initiation were used to measure baseline characteristics.

Results: Among 34,079 eligible female denosumab initiators, we identified 5 distinct patterns of denosumab therapy: 1) persistent use with ≥80% probability to be on denosumab throughout follow-up (56%); 2) one-time use and < 10% transition to oral bisphosphonate (14%); 3) early discontinuation and < 10% transition to oral bisphosphonate (12%); 4) early delays and return to persistent use with ≥75% probability to be on denosumab and < 3% transition to oral bisphosphonate (10%); and 5) late discontinuation after 2.5 years of denosumab with < 8% transition to oral bisphosphonate (8%). One-time users and those discontinuing early were older and had more comorbidities.

Conclusions: Two-thirds of female denosumab initiators had persistent or one-time use of denosumab, and another 12% persisted for 7 years after early delays in therapy. Notably, 1 in 12 patients were at high risk of rebound fractures with abrupt discontinuation after 2.5 years of therapy.