Abstract
<title>Abstract</title> <p> <bold>Purpose</bold> To compare the anatomical and functional efficacy of an early switch to intravitreal ranibizumab (IVR) or dexamethasone implant (IVD) in eyes with persistent diabetic macular edema (DME) after bevacizumab, and to explore OCT and systemic predictors of treatment response. <bold>Methods</bold> This prospective, observational, non-randomized, open-label clinical study included 115 eyes of 80 patients with treatment-naïve, center-involved DME that persisted despite three consecutive monthly intravitreal bevacizumab injections. At month 4, eyes were switched either to IVR (0.5 mg/0.05 mL; 68 eyes) or IVD (0.7 mg; 47 eyes) at the discretion of the treating ophthalmologist. Patients were followed monthly for 12 months on a pro re nata regimen. Best-corrected visual acuity (BCVA, logMAR), central macular thickness (CMT, SD-OCT), intraocular pressure (IOP), OCT biomarkers, and systemic laboratory parameters (including systemic immune-inflammation index, SII) were recorded. Treatment outcomes were systematically evaluated. Predictors of CMT gain and BCVA outcome were analyzed. <bold>Results</bold> Baseline CMT was similar between the groups (IVR: 435.80 ± 132.27 µm; IVD: 449.51 ± 104.19 µm). After switching, both groups showed a significant reduction in CMT over time (p < 0.001), with comparable final CMT at month 12 (339.14 ± 90.90 µm vs. 369.66 ± 116.57 µm; p = 0.131). Dexamethasone produced a significantly greater CMT reduction at month 2 (p < 0.001). Baseline BCVA was 0.48 ± 0.38 logMAR in the IVR group and 0.52 ± 0.42 logMAR in the IVD group (p = 0.453). Visual acuity remained comparable through months 1, 2, 3, and 6 (p = 0.088, 0.087, 0.109, and 0.070, respectively. However, at month 12, a statistically significant difference emerged in favor of IVR, with better visual acuity compared to IVD (0.35 ± 0.37 vs. 0.66 ± 0.49 logMAR; p = 0.001). IOP increase was more frequent in the IVD group, with 17% of eyes requiring topical antiglaucoma therapy, but all were controlled medically. Baseline CMT was the only OCT parameter significantly associated with treatment efficacy (p < 0.01). Among systemic markers, higher SII was significantly associated with greater early CMT reduction at month 2 in the dexamethasone arm (p = 0.028). <bold>Conclusions</bold> Both intravitreal ranibizumab and dexamethasone implants resulted in meaningful anatomical improvement in refractory DME. Dexamethasone implants induced a faster early reduction in CMT, whereas ranibizumab achieved more sustained visual gains over 12 months. Elevated SII may help identify patients who are more likely to benefit from an early anatomical response to dexamethasone. These prospective, real-world data support individualized treatment decisions integrating anatomical, functional, and inflammatory profiles. </p>