Hypothesis / aims of study
Mirabegron is a beta-3 agonist that is commonly used for the treatment of overactive bladder. There are beta-3 receptors in the eye, and mirabegron has been associated with adverse ocular effects[1]. Our objective was to evaluate if there is an increased risk of retinal disease in new users of mirabegron.
Study design, materials and methods
This was a retrospective cohort study. We used routinely collected data to identify adults ≥66 years of age from Ontario Canada who were newly started on an oral OAB medication between 2016-2020. We defined three exposure groups: oxybutynin users, newer OAB anticholinergic users (solifenacin, fesoterodine, tolterodine, darifenacin, and trospium), and mirabegron users.
The primary outcome was use of an intraocular anti-VEGF medication. These medications are indicated for the treatment of neovascular age-related macular degeneration, diabetic eye disease, and other neovascular retinal conditions (including choroidopathy), and represent a marker of chorioretinal disease. Secondary outcomes included vitrectomy, photocoagulation, and a more restrictive definition of retinal disease.
We incorporated a 90 day lag between OAB medication initiation and primary outcome assessment, and censored patients 90 days after they stopped continuous usage of their OAB medication, if they switched exposure groups, or on March 31 2025. We used matching weights to balance 82 relevant covariates, and standardized differences to ensure that additional ocular covariates were similar between groups. Weighted Cox proportional hazards models were used, and 95% CI obtained via bootstrapping are reported. A sensitivity analysis using a pre-OAB medication exposure time-period was prespecified to assess for potential confounding.
Results
We identified 103,024 new users of OAB medications (33,097 new anticholinergic users, 13,865 new oxybutynin users, and 56,062 new mirabegron users). Median age was 76 (IQR 70-82), approximately 70% of each group was female, and 18% had diabetes. All baseline covariates were similar in the weighted cohorts. The median duration of use of mirabegron was 6 (IQR 4-17) months. In the weighted cohort, there was a significantly increased hazard of chorioretinal disease among mirabegron users (HR 1.15, 95% CI 1.03-1.27), compared to newer OAB anticholinergic users (reference group); oxybutynin users were similar to newer OAB anticholinergic users (HR=0.99, 95% CI 0.80-1.18).
A sensitivity analysis was performed where we looked at the risk of chorioretinal disease in the 180 days prior to starting the OAB medications. Mirabegron users had a similarly elevated risk in this pre-exposure period (HR 1.15, 95% CI 1.03-1.28). Secondary outcomes of vitrectomy, photocoagulation, and a more restrictive definition of retinal disease were not significantly elevated in beta-3 agonist users.
Interpretation of results
Beta-3 receptors are found in retinal and choroidal endothelial cells, and influence nitrous oxide and VEGF signaling. Mirabegron may cause subtle reductions in optic nerve perfusion, however the clinical implications of this are unclear.[2] Conversely, activation of these receptors may also cause endothelial proliferation and thickening from pro-angiogenic changes; a recent case report described a mirabegron associated choroidopathy.[3] Taken together with the ocular adverse events that have been reported[1], it is important to assess this rare potential risk in a large population based dataset.
While we found that mirabegron users had an increased hazard of chorioretinal disease, this same magnitude of increased risk was present before they started using mirabegron, meaning this is likely due to residual confounding. Secondary outcomes related to chorioretinal disease were similar to OAB anticholinergic users. Limitations of our research include the use of an anticholinergic OAB comparison group, and the real-world utilization patterns of OAB medications.