Diabetic macular edema (DME) is a leading cause of vision loss in diabetic retinopathy and involves complex neurovascular dysfunction beyond simple vascular permeability. Although anti-VEGF agents have improved outcomes, a subset of eyes remain refractory with persistent fluid and limited visual gain. Faricimab, a bispecific antibody targeting VEGF-A and angiopoietin-2, aims to reduce leakage and stabilize vasculature via Tie2 pathway effects. This 12-month real-world retrospective study evaluated anatomical and functional outcomes after faricimab in both naïve and refractory DME eyes and examined baseline spectral-domain OCT biomarkers associated with visual outcomes.
This single-center retrospective cohort included eyes treated with intravitreal faricimab between December 2024 and March 2026. Eyes were classified as naïve or refractory based on prior intravitreal injection history. The refractory definition required at least four anti-VEGF injections in the prior six months with persistent edema (CMT > 350 μm and ≤15% reduction in that period). Inclusion criteria were age >18 years, type 2 diabetes, nonproliferative diabetic retinopathy (NPDR) on fluorescein angiography, CMT > 350 μm, and completion of three consecutive faricimab loading doses. Standard exclusion criteria applied for other ocular/systemic confounders and recent ocular procedures.
All eyes received three monthly loading injections of faricimab (baseline, month 1, month 2), then entered a treat-and-extend regimen from month 3 with 4-week interval adjustments based on anatomical and functional responses. Clinical and OCT evaluations occurred every 4 weeks during follow-up.
Two experienced ophthalmologists independently extracted data including demographics, DR severity, lens status, HbA1c, prior ocular treatments, and prior anti-VEGF history in the refractory group. SD-OCT (Cirrus 5000) macular cube scans were used to measure central macular thickness (CMT) within the central 1-mm zone and to grade predefined OCT biomarkers as present/absent: intraretinal cysts (IRCs), subretinal fluid (SRF), hyperreflective foci (HRFs), disorganization of the retinal inner layers (DRIL), and external limiting membrane/ellipsoid zone (ELM/EZ) disruption. Two masked graders performed assessments with adjudication by a senior specialist for disagreements.
Primary outcomes were changes in best-corrected visual acuity (BCVA, converted to logMAR) and CMT from baseline to month 12. Secondary outcomes included changes in proportions of eyes with the specified OCT biomarkers. Repeated measures and within-subject correlation (including bilateral cases) were analyzed using generalized estimating equations (GEEs). A multivariable GEE identified baseline factors independently associated with 12-month BCVA. Statistical significance was defined as two-sided P < 0.05.
The analysis included 110 eyes from 104 patients (54 eyes naïve, 56 eyes refractory). The refractory group had significantly longer DME duration (mean 19.06 ± 5.41 vs 4.53 ± 1.07 months; P < 0.001) and a higher baseline prevalence of DRIL and ELM/EZ disruption (both P < 0.001). Age, HbA1c, baseline BCVA, baseline CMT, and NPDR severity distribution were comparable between groups. In the refractory cohort, prior therapy averaged 10.28 ± 2.13 anti-VEGF injections over 18.14 ± 4.59 months, with a mean prior injection interval of 6.72 ± 2.41 weeks; aflibercept was the most frequently used prior agent.
Baseline mean BCVA was similar between groups (naïve 0.54 ± 0.22 vs refractory 0.57 ± 0.27 logMAR; adjusted P = 0.660). The naïve group demonstrated progressive, statistically significant BCVA improvement at months 3, 6, and 12, reaching 0.37 ± 0.27 logMAR at month 12 (all adjusted P < 0.05 vs baseline). The refractory group showed modest, non-significant improvements (month 12 mean 0.49 ± 0.21 logMAR; adjusted P > 0.05). Between-group comparisons favored naïve eyes at months 3, 6, and 12 (all adjusted P < 0.05).
Baseline mean CMT did not differ significantly between groups (naïve 418.17 ± 46.25 μm vs refractory 433.22 ± 58.93 μm; adjusted P = 0.132). Both groups experienced significant CMT reductions from baseline at months 3, 6, and 12 (all adjusted P < 0.001). At month 12 mean CMT values were 283.48 ± 53.57 μm (naïve) and 294.61 ± 58.84 μm (refractory) with no significant between-group difference (adjusted P = 0.295).
Both naïve and refractory groups showed significant reductions in the proportions of eyes with fluid-related OCT findings: IRCs, SRF, and HRFs (all P < 0.001), indicating effective retinal fluid control after faricimab. However, the refractory cohort had a higher baseline frequency of microstructural damage markers (DRIL and ELM/EZ disruption), suggesting greater chronic structural compromise.
In the multivariable GEE model, baseline DRIL (B = 0.044; 95% CI: 0.004 to 0.084; P = 0.031) and ELM/EZ disruption (B = 0.147; 95% CI: 0.069 to 0.225; P < 0.001) were independently associated with poorer 12-month BCVA. The treatment group (naïve vs refractory) was not independently associated with 12-month BCVA (P = 0.969), indicating that baseline retinal microstructural integrity had greater prognostic value for visual outcomes than prior treatment history alone.
The study protocol included active monitoring for ocular complications (elevated intraocular pressure, endophthalmitis, retinal detachment, vitreous hemorrhage) and systemic events (cardiovascular or cerebrovascular). The provided source excerpt did not report specific adverse event rates or detailed safety outcomes; those details were not available in the supplied text.
In this real-world cohort, intravitreal faricimab produced meaningful anatomical improvement with significant reductions in CMT and fluid-related OCT biomarkers in both treatment-naïve and previously refractory DME eyes. Functional gains in BCVA occurred in naïve eyes but were limited in refractory eyes despite comparable fluid resolution. Baseline microstructural damage—DRIL and ELM/EZ disruption—was more prevalent in refractory eyes and independently predicted poorer 12-month visual outcomes. These results emphasize that anatomical success (fluid control and reduced CMT) does not necessarily equate to visual recovery when photoreceptor and inner retinal architecture are compromised.
For eyes with chronic refractory DME and structural disruption, clinical benefits of switching to faricimab may include sustained fluid control and reduced treatment burden even when substantial visual improvement is unlikely. OCT biomarkers such as DRIL and ELM/EZ disruption offer clinically relevant prognostic information and may aid in counseling and management decisions.
This analysis is retrospective, single-center, and observational, which can introduce selection bias and limit generalizability. The provided excerpt did not include detailed safety outcome data or full biomarker time-course tables; those details were not reported in the supplied source text. The operational definition of refractory DME was study-specific and may differ from other definitions used in the literature.
Faricimab achieved effective retinal fluid control and reduced CMT in both naïve and refractory DME eyes over 12 months in real-world practice. However, visual recovery depended on baseline retinal microstructural integrity: DRIL and ELM/EZ disruption were independently associated with poorer 12-month BCVA, whereas prior treatment status was not. In refractory eyes with structural damage, sustained anatomical control and lower treatment burden may represent meaningful clinical outcomes despite limited visual gains.