Real-world effectiveness of intravitreal dexamethasone implants - Comparison between eyes eligible and ineligible for clinical trials and their associated outcomes

Background: Concerns about the generalizability of pivotal randomized controlled trials (pRCTs) findings have been raised. We aimed to compare intravitreal dexamethasone implants' (IDIs) effectiveness for diabetic macular edema (DME) and central retinal vein occlusion (CRVO), between eyes eligible and ineligible for pRCTs. Methods: This retrospective cohort study analyzed Taiwan's Chang Gung Research Database, including DME or CRVO eyes initiating IDIs during 2015-2020. We classified all treated eyes as eligible or ineligible for pRCTs following major selection criteria of the MEAD and GENEVA trials, and evaluated three-, six-, and twelve-month changes in central retinal thickness (CRT) and visual acuity (VA) after initiating IDIs. Results: We included 177 IDI-treated eyes (DME: 72.3%; CRVO: 27.7%), of which 39.8% and 55.1% were ineligible for DME and CRVO pRCTs, respectively. LogMAR-VA and CRT changes at different times were comparable in DME eyes eligible (LogMAR-VA difference: 0.11 to 0.16; CRT difference: -32.7 to -96.9 mu m) and ineligible (LogMAR-VA difference: -0.01 to 0.15; CRT difference: -54.5 to -109.3 mu m) for the MEAD trial. By contrast, CRVO eyes ineligible for the GENEVA trial had greater LogMAR-VA changes (0.37 - 0.50) than those eligible (0.05 - 0.13), with comparable CRT reductions (eligible eyes: -72.3 to -106.4 mu m; ineligible eyes: -61.8 to -110.7 mu m) (all p-values <0.05 of the mean differences between eligible and ineligible CRVO eyes for all followups). Conclusions: IDIs had similar VA and CRT outcomes among DME eyes, regardless of pRCT-eligibility. However, among CRVO eyes, those ineligible for pRCTs showed greater deterioration in VA than those eligible.