Treatment discontinuation among users of GLP-1 receptor agonists and SGLT2 inhibitors in a national population of individuals with type 2 diabetes.

AIMS/HYPOTHESIS: Our aim was to assess treatment discontinuation, reinitiation and switching between drugs within the same drug class for glucagon-like peptide-1 (GLP-1) receptor agonists and sodium-glucose cotransporter 2 (SGLT2) inhibitors in individuals with type 2 diabetes.

METHODS

We used data from nationwide registers in Sweden to perform separate analyses for all patients with type 2 diabetes who filled a first prescription of a GLP-1 receptor agonist or an SGLT2 inhibitor between 2017 and 2021. Patients were considered to be on treatment for the period during which prescriptions were refilled before the estimated end date of the most recent prescription, including a 90-day grace period, i.e. the time allowed between and after prescriptions before treatment is considered as discontinued. We used the Aalen-Johansen estimator to estimate cumulative incidences of discontinuation and reinitiation, and Fine-Gray sub-distribution hazard models to assess the association of clinical variables with the risk of discontinuation.

RESULTS

Among 73,895 new users of GLP-1 receptor agonists, the cumulative incidence of treatment discontinuation was 23.6% at 1 year and 38.5% at 3 years. Among patients who discontinued, the cumulative incidence of treatment reinitiation was 41.1% at 1 year and 57.4% at 3 years after discontinuation. Among 113,207 new users of SGLT2 inhibitors, the cumulative incidence of treatment discontinuation was 27.9% at 1 year and 45.9% at 3 years, with a cumulative incidence of reinitiation of 40.4% at 1 year and 55.7% at 3 years after discontinuation. When varying the grace period between 60 days and 365 days, treatment discontinuation rates at 3 years ranged from 23.3% to 43.6% among GLP-1 receptor agonist users and from 28.8% to 50.6% among SGLT2 inhibitor users. The proportion of patients who had ongoing treatment, regardless of previous discontinuation episodes, ranged between approximately 70% and 80% for both drugs during a 1-5 year period after treatment initiation across analyses using various grace periods. In terms of switching, 22.9% of the GLP-1 receptor agonist users and 2.1% of the SGLT2 inhibitor users switched between drugs within the same drug class. Patient characteristics associated with treatment discontinuation were similar for GLP-1 receptor agonists and SGLT2 inhibitors, although the association between higher BMI and a lower likelihood of treatment discontinuation was stronger for GLP-1 receptor agonists.

CONCLUSIONS/INTERPRETATION: Approximately half of type 2 diabetes patients who had started using GLP-1 receptor agonists or SGLT2 inhibitors had discontinued treatment within 5 years of follow-up. However, more than half of those who discontinued treatment subsequently reinitiated treatment, such that the proportion with ongoing treatment was approximately 70-80% for both drugs during a 1-5 year period after treatment initiation. This suggests that the proportion of patients with long-term use of the medications is larger than indicated by analyses focusing on treatment discontinuation. Patient characteristics associated with treatment discontinuation were similar for GLP-1 receptor agonists and SGLT2 inhibitors.