Prognostic biomarkers of chronic diabetic macular edema treated with a fluocinolone acetonide intravitreal implant

Abstract Background: This study aimed to investigate retinal imaging biomarkers, such as disorganization of the retinal inner layers (DRIL) and/or ellipsoid zone (EZ) disruption by spectral domain optical coherence tomography (SD-OCT), and functional outcomes in eyes treated with ۰.۲ µg/day of a fluocinolone acetonide intravitreal implant (FAc) after an insufficient response to previous treatments. Methods: This was a retrospective comparative study of ۱۸ eyes (۱۵ patients) with persistent and/or recurrent diabetic macular edema (DME) treated with FAc. Eyes were divided according to the number of prior intravitreal treatments: group ۱ (n = ۸) with less than or equal to ۶ injections (early switch) and group ۲ (n = ۱۰) with > ۶ injections (late switch). Outcomes included percentage of eyes with DRIL and/or EZ disruption at baseline and analysis of the best corrected visual acuity (BCVA) using ETDRS letters, central macular thickness (CMT), DRIL, and EZ disruption at the last observation. Results: Group ۲ revealed a significantly higher percentage of DRIL and/or EZ disruption than group ۱ (P < ۰.۰۵). At the last observation, group ۱ revealed a higher percentage of eyes achieving vision stability/improvement, gaining greater than or equal to۱۵ letters, and achieving greater than or equal to ۷۰ letters (P > ۰.۰۵ for all comparisons). The mean BCVA gain was ۸.۸ and ۰.۷ letters for groups ۱ and ۲ (P = ۰.۳۹۷). Both groups revealed a significant mean CMT reduction (>۲۰% reduction from the baseline value), without a significant statistical difference between them (P = ۰.۷۴۹). After treatment, most eyes from both groups showed resolution of DRIL and EZ disruption. Conclusions: Patients with DME presenting with a lower percentage of DRIL and/or EZ disruption at baseline had better functional outcomes, supporting the possible benefit of an early switch to FAc after insufficient response to previous treatments. Future randomized studies with a larger patient cohort are warranted to confirm our conclusions.