Agreement of VCDR measurements: ultra-widefield imaging versus standard fundus photography compared to BMO-OCTA

This study evaluates the consistency of vertical cup-to-disc ratio (VCDR) measurements using standard fundus photography (Method F) and ultra-widefield (UWF) imaging (Method U), against three-dimensional (3D) Bruch’s membrane opening optical coherence tomography angiography (BMO-OCTA) as an anatomical comparator. Measurement disparities between primary open-angle glaucoma (POAG) and glaucoma suspect (GS) eyes were also assessed. This cross-sectional study included 157 eyes (80 POAG, 77 GS) evaluated via standard fundus photography, UWF imaging, and OCTA. VCDR was manually delineated in two-dimensional (2D) images by a masked observer and automatically segmented in BMO-OCTA. Consistency and measurement bias were analyzed using Bland-Altman plots and multivariable Generalized Estimating Equations (GEE) adjusting for patient age to account for inter-eye correlation. UWF and standard fundus photography were highly correlated (Spearman ρ = 0.951) but exhibited systematic bias (mean bias − 0.024) and wide 95% limits of agreement (-0.112 to 0.065). Compared to the 3D BMO-OCTA, both 2D methods demonstrated a significant overestimation bias. The estimated marginal mean bias of standard fundus photography was significantly lower than that of UWF imaging (0.035 vs. 0.059, P adj < 0.001). 2D measurements demonstrated wider inter-individual measurement dispersion (broader limits of agreement) in GS eyes compared to POAG eyes, reflecting greater measurement variability in this subgroup. Although UWF imaging correlates well with standard photography, the two modalities are not interchangeable for quantitative optic disc assessment. Both monoscopic 2D modalities systematically overestimate VCDR, exhibiting wider measurement dispersion in GS eyes, a phenomenon potentially related to their atypical optic cup morphology. While 2D imaging remains a fundamental tool for initial screening, relying solely on longitudinal 2D monitoring may introduce clinical uncertainties. To mitigate potential risks associated with the inherent limitations of 2D imaging, we suggest integrating 3D BMO-OCT/OCTA as a complementary structural confirmation tool when a borderline or elevated VCDR is flagged in a glaucoma suspect.