Effect of Faricimab on Optical Coherence Tomography Angiography and Artificial Intelligence-Based Analysis in Resistant Choroidal Neovascularization

Ophthalmologica. 2025;248(5-6):346-355. doi: 10.1159/000548690. Epub 2025 Oct 24.

Abstract

Introduction: Treatment-resistant choroidal neovascularization (CNV) remains a major challenge despite anti-VEGF therapy. Faricimab, with dual inhibition of VEGF-A and Ang-2, offers a novel approach. The purpose of this study was to evaluate its effect using advanced OCTA-based vascular analysis and AI-assisted anatomical assessment, providing new insights into therapeutic response in resistant CNV.

Methods: This retrospective study analyzed OCTA and OCT biomarkers before and after treatment switch in patients with CNV that was resistant to prior anti-VEGF therapy. Quantitative vascular analysis of CNV lesions was performed using AngioTool 2.0 software. Anatomical response was evaluated by retinal fluid biomarkers, including central retinal thickness (CRT), intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (PED) volume using the RetinAI software.

Results: Following the switch to Faricimab, OCTA analysis revealed reductions in vessel area (p = 0.011), vessel percentage area (p = 0.002), total vessel length (p = 0.011), and total number of junctions (p = 0.006). Lacunarity significantly increased (p = 0.009), indicating notable vascular remodeling. Moreover, AI-assisted imaging analysis showed that the CRT decreased by 9.94% (p = 0.0001), and PED volume decreased by 20.94% (p = 0.011). Total retinal fluid, including IRF and IRF, was reduced by 84.06% (p = 0.007), reflecting substantial anatomical improvement. Additionally, the mean injection interval increased by 3.19 ± 2.47 weeks, indicating enhanced disease stability and a reduced treatment burden.

Conclusion: This study highlights the role of OCTA vascular analysis in assessing CNV treatment response. The significant vascular and AI-assisted anatomical improvements observed after switching to faricimab suggest its potential efficacy for treating resistant CNV cases. Given the scarcity of literature on OCTA vascular changes in this context, our findings provide novel insights into treatment response assessment. Further studies with larger cohorts are needed to validate these observations.

Keywords: Artificial intelligence; Faricimab; OCTA vascular changes; Resistant choroidal neovascularization; RetinAI software.

MeSH terms

  • Aged
  • Angiogenesis Inhibitors / administration & dosage
  • Antibodies, Bispecific
  • Artificial Intelligence*
  • Choroid* / blood supply
  • Choroid* / pathology
  • Choroidal Neovascularization* / diagnosis
  • Choroidal Neovascularization* / drug therapy
  • Female
  • Fluorescein Angiography* / methods
  • Follow-Up Studies
  • Fundus Oculi
  • Humans
  • Intravitreal Injections
  • Male
  • Middle Aged
  • Retrospective Studies
  • Tomography, Optical Coherence* / methods
  • Treatment Outcome
  • Vascular Endothelial Growth Factor A / antagonists & inhibitors
  • Visual Acuity

Substances

  • Angiogenesis Inhibitors
  • Vascular Endothelial Growth Factor A
  • faricimab
  • Antibodies, Bispecific