Persistence with brand versus generic oral bisphosphonates among newly treated osteoporosis patients: a retrospective cohort study in the French claims database

Abstract

Introduction The first line of osteoporosis (OP) treatment is oral bisphosphonate (BP) and aims to reduce the risk of osteoporotic fracture. Although BP demonstrated his efficacy in preventing fractures, persistence with BP was estimated very low in previous studies in France and other countries. It has been proposed that persistence with generic BP formulations may be lower compared to brand formulations, due to changes of excipients reported to result in differences in disintegration time increasing the risk of upper gastrointestinal irritation. We aimed to compare the persistence with brand versus generic oral BPs. Methods We conducted a retrospective, comparative cohort study in the EGB (“Echantillon Generaliste des Beneficiaires”), a representative sample of the French Health Insurance claims database. We included all patients aged 50 and older, newly treated with oral BP (alendronic acid, ibandronic acid, risedronic acid or etidronic acid, without association with vitamin D) for primary osteoporosis between 01/01/2009 and 31/12/2015. Patients were followed from the initiation date until: switch (change from generic to brand drug and conversely, or change of BP molecule); discontinuation of BP (no new dispensation during a period greater than twice the expected duration of supply if taken as prescribed); or 12 months. Persistence was evaluated as time to discontinuation and as proportion of patients still on treatment at 6 and 12 months. Time to discontinuation and 6- and 12-month persistence were compared between patients who initiated treatment with generic and patients who initiated with brand BP, excluding patients who switched. Results In total, 3903 patients initiated an oral BP treatment for primary OP during the 7-year inclusion period, 1710 (44.0%) with generic BP and 2193 (56.0%) with brand BP. Mean age was comparable between the 2 groups: 70.8 (± 10.6) in “generic BP” group and 70.0 (± 10.4) in “brand BP” group. There were fewer women in the “generic BP” group than in “brand BP” group (respectively 87.8% and 90.3%; P = 0.01. 236 (13.8%) patients in “generic BP” group and 306 (14.0%) patients in “brand BP” group switched to the other BP group during the 12 months after initiation; 56 (3.3%) patients in the “generic BP” group and 175 (8.0%) patients in “brand BP” group have changed to another BP molecule, respectively. After exclusion of patients who switched, 6- and 12-month persistence were not statistically different between the 2 groups: respectively 43.9% and 33.8% in “generic BP” group and 41.7% and 31.3% in “brand BP” group (P = 0.22 and 0.14 respectively for 6- and 12-month persistence). Mean time to discontinuation was 185.6 days (± 138.3) in “generic BP” group and 183.1 days (± 138.1) in “brand BP” group. Conclusions There were no differences in persistence with brand or generic BP in this population. Low persistence rates in our analysis are consistent with results of other studies, and confirm that discontinuation of BP treatment is currently a public health concern. The reasons for discontinuation require further investigation, to inform shared decision-making in treatment choice and strategies to support long-term BP use.