Ophthalmology

Common design and operational causes of unsuccessful phase 2–3 retinal trials

Retrospective analysis of unsuccessful US phase 2–3 retinal trials (2015–2025) attributes many failures to endpoint choice, comparator selection, and population heterogeneity. Later-phase trials had larger size, wider geographic dispersion, longer follow-up, and greater assessment burden, which may dilute treatment effects. Sponsor-level operational and strategic factors, including feasibility and financial considerations, contributed to discontinuations. Recommendation: align biological rationale, feasible endpoints, and operational plans before progressing to pivotal trials. 1

Lower AMD incidence associated with NSAID prescriptions in a large multi-center cohort

In a TriNetX retrospective cohort (634,794 NSAID users matched to equal controls), NSAID prescription was associated with lower incident AMD over follow-up (overall HR 0.58). Protective associations persisted across timepoints (eg, HR 0.31 at 6 months) and for aspirin and non-selective COX inhibitors, and for nonexudative and exudative AMD. This is observational data from electronic records and susceptible to residual confounding, indication bias, and exposure misclassification. 2

Atropine dose–response for myopia: greater efficacy with higher doses but more side effects

Dose-response meta-analysis of 33 RCTs (6,301 children, mean 19.5 months) found increasing myopia control with higher atropine concentrations up to a plateau. Estimated annual spherical equivalent reductions ranged from 0.21 D at 0.01% to 0.99 D at 1%, with diminishing incremental benefit beyond intermediate doses. Higher concentrations increased accommodation loss, pupil dilation, and photophobia, and high-dose estimates are less precise due to small trial numbers. 3