Clinical profile, fundus fluorescein angiographic perfusion phenotypes, and visual outcomes in central retinal artery occlusion: a single-center retrospective cohort study

Central retinal artery occlusion (CRAO) is associated with poor visual outcomes, but the distribution and reliability of fundus fluorescein angiography (FFA) perfusion phenotypes in real-world CRAO cohorts remain incompletely defined. We described a contemporary CRAO cohort and evaluated exploratory associations between clinical/FFA features and visual acuity at 1 month or later. Consecutive CRAO affected-eye/clinical-event records treated between May 2025 and June 2026 were reviewed. Presenting and follow-up visual acuity were converted to approximate logarithm of the minimum angle of resolution (logMAR) values. Follow-up visual acuity was defined as outpatient visual acuity measured at least 1 month after symptom onset, because visual acuity after CRAO was considered clinically stable around 1 month. FFA-derived variables were analyzed only in records with matched FFA images. Key FFA perfusion phenotypes were independently reread by two masked ophthalmologists, with disagreements adjudicated by a third senior ophthalmologist. The primary outcome was poor follow-up visual acuity, defined as follow-up logMAR > = 1.3. Univariable odds ratios (ORs) with 95% confidence intervals (CIs) were estimated, with false discovery rate adjustment for exploratory comparisons. The final cohort included 132 affected-eye records from 130 unique patients; two patients contributed sequential bilateral events. Mean age was 65.3 years (SD, 12.1), and 33 records (25.0%) were female. Matched FFA data were available for 89 records. Final adjudicated rereading showed a high burden of perfusion disturbance: severe FFA ischemic phenotype was present in 81/89 records (91.0%), delayed or absent retinal arterial filling in 79/87 (90.8%), retinal nonperfusion or large/no-perfusion area in 81/89 (91.0%), and macular nonperfusion in 82/87 (94.3%). Pre-adjudication intergrader agreement ranged from 89.9% to 96.6%, with Cohen’s kappa values from 0.581 to 0.806. Median presenting and follow-up visual acuity were both 2.0 logMAR. Poor follow-up visual acuity occurred in 91 records (68.9%). Visual-acuity transition analysis showed persistent poor visual acuity in 91 records, transition from poor to non-poor visual acuity in 27, no transition from non-poor to poor visual acuity, and persistent non-poor visual acuity in 14. Final reread FFA perfusion phenotypes were highly prevalent and did not show clear discriminatory prognostic value in exploratory analyses. Visual outcome after CRAO remained poor at 1 month or later. Independent FFA rereading showed reproducible, highly prevalent perfusion disturbance, suggesting a ceiling effect for global ischemic phenotypes in this cohort. Presenting severe visual loss commonly persisted. Apart from this descriptive visual-acuity persistence pattern, no non-overlapping candidate clinical, laboratory, ocular, or FFA variable remained statistically robust after FDR correction; exploratory findings, including the documented cilioretinal macular supply sign and homocysteine, require validation in larger cohorts.