Pharmacoepidemiology and Pharmacovigilance Division. Spanish Agency of Medicines and Medical Devices (AEMPS). Madrid, Spain, Spain
Background: Persistence with antiosteoporotic medications (AOM) is essential for favorable outcomes, but it is known to be poor in real practice. Few long-term follow-up studies have assessed persistence with denosumab (Prolia®), a long-acting injectable therapy administered every 6 months, which is supposed to improve compliance compared to oral AOM.
Objectives: To estimate the incidence of denosumab discontinuation and time on continuous treatment (i.e persistence) between 2011 and 2018 in Spain.
Methods: A cohort study of new users of denosumab from October 2011 (first availability in Spain) to December 2018, aged ≥18 years and with ≥ 1 year of available data, was performed using the Spanish Database for Pharmacoepidemiological Research in Public Health Systems (BIFAP). Patients were followed from therapy initiation to the earliest of discontinuation (> 90-day refill gap after 6 months of the last denosumab prescription), loss to follow-up, death, or end of the study period. Cumulative incidences of discontinuation were estimated by time after initiation of treatment using Kaplan-Meier survival analyses. Patients who filled only one denosumab prescription were classified as “early discontinuers”. We also estimated the proportion of patients that reinitiated therapy (those with at least one denosumab prescription after the 90-day permissible refill gap) and the number of extended gaps between treatment episodes.
Results: Out of 62,445 Prolia new users (97.1% women, mean age of 72 years), 34,723 discontinued denosumab during a median follow-up of 2.79 years. Early discontinuers accounted for a total of 13,770 new users (23.1%). Discontinuation was 31.2% at the end of the 1st year, 50.4% at 2nd, 79.1% at 5th, and 87.6% at 7th year of treatment. The overall median [IQR] duration of the first treatment episode was 419 days [180-872], and among discontinuers 361 days [IQR 180-711]. Of total discontinuers, 41.4% (N= 14,370) reinitiated denosumab within a median of 190 days [IQR 155-321] and remained on therapy over a median of 346 days [IQR 180-651] more. The majority of these patients (80.6%) had only one extended refill gap. Women were more likely to return to therapy than men (58.3% vs 52.6%).
Conclusions: The proportion of patients that persisted with therapy declined significantly over time, from 68.8% at 1 year to 20.9% at 5 years of treatment. A pattern of interrupted treatment is suggested for a considerable number of patients, as new denosumab prescriptions were identified soon after a first permissible extended gap, mostly within the following 6 months. Recent evidence has identified potentially rapid and harmful effects on bone following discontinuation of denosumab, and since its clinical impact is not yet fully known, further research on this topic is warranted.
