Swept-source OCT/OCTA quantitative biomarkers and early visual response after faricimab switch in type 1 macular neovascularization

Abstract Purpose To evaluate whether baseline swept-source optical coherence tomography angiography (SS-OCTA) parameters are associated with 3-month visual response after switching to faricimab in eyes with Type 1 macular neovascularization (MNV) and incomplete prior anti-VEGF response. Methods In this single-center, prospective exploratory cohort, 62 eyes of 62 patients with Type 1 MNV received three monthly faricimab injections. Baseline SS-OCTA parameters—including choroidal stromal-to-vascular ratio (CSVR) and layer-specific vessel densities—were quantified. Primary outcome was visual gain ( $$\geq 0.1\ \mathrm{logMAR}$$ improvement) at 3 months. Results Thirty eyes (48.4%) achieved visual gain at 3 months. In multivariable Firth’s logistic regression, lower baseline CSVR (adjusted odds ratio [OR] = 0.745 per 0.1-unit increase, 95% CI: 0.565–0.925; P = 0.006) and higher deep vascular complex (DVC) vessel density (adjusted OR = 1.096 per 1% increase, 95% CI: 1.010–1.212; P = 0.026) were independently associated with visual response. The combined predictive model demonstrated an area under the curve (AUC) of 0.759 (95% CI: 0.641–0.878; bootstrap-adjusted AUC: 0.743) with good apparent calibration (calibration slope: 1.134; Hosmer–Lemeshow P = 0.621). A data-driven baseline CSVR cut-off of 1.04 identified responders with a sensitivity of 86.7% and a specificity of 56.2%. Conclusions Lower baseline CSVR and higher DVC vessel density were associated with short-term visual gain after faricimab conversion in Type 1 MNV with incomplete prior anti-VEGF response. These preliminary predictive findings require external validation before clinical use.