GLP-1 receptor agonists show no overall increased risk of diabetic retinopathy progression in meta-analysis

Diabetic retinopathy (DR) is a leading cause of vision impairment in patients with diabetes. Despite the widespread use of glucagon-like peptide-1 receptor agonists (GLP-1RAs) for type 2 diabetes (T2D) management, their impact on DR progression has been a subject of controversy, particularly following signals from trials like SUSTAIN-6. Current standard-of-care for DR primarily focuses on glycemic control and direct retinal interventions, but the systemic effects of GLP-1RAs on microvascular complications like DR remain unclear, necessitating a comprehensive evaluation of their safety profile in this context.

This systematic review and meta-analysis identified 11 studies (randomized controlled trials and retrospective cohort studies) from a pool of 161 records, encompassing 137 to 9,463 participants across North America, Europe, and East Asia. The primary GLP-1RAs analyzed were albiglutide, liraglutide, semaglutide, exenatide, and dulaglutide, compared against placebo or SGLT-2 inhibitors. The primary outcome was DR progression, assessed via pooled relative risk (RR) estimates using a random-effects model. Subgroup analyses by GLP-1RA type and geographical location were performed, alongside sensitivity analyses and publication bias assessment using funnel plots, Egger's test, and Begg's test. Risk of bias was evaluated with the Cochrane tool.

The meta-analysis revealed no statistically significant association between overall GLP-1RA use and diabetic retinopathy progression, with a pooled RR of 1.07 (95% confidence interval [CI]: 0.90-1.28). Subgroup analyses, stratified by specific GLP-1RA type and geographical region, similarly showed no significant differences in DR progression risk. However, a notable exception emerged: two studies specifically evaluating semaglutide demonstrated a higher RR for DR progression. These RRs were 1.76 (1.11-2.79) and 6.41 (0.67-61.46), a unique finding that contrasted with the overall null result for the class. Sensitivity analyses confirmed the robustness of the primary findings, and no evidence of publication bias was detected across the included studies (P > 0.05 for both Egger's and Begg's tests). Risk of bias analysis indicated a low risk across most domains for the included studies.

The overall meta-analysis found no significant association between GLP-1RA use and diabetic retinopathy progression (pooled RR = 1.07, 95% CI: 0.90-1.28).