Predicting Perimetric Glaucoma Development in Suspects Using Widefield OCT-Based Risk Scores.
Huiyuan Hou, Takashi Nishida, Evan Walker, Mary K Durbin, Anya Guzman, Alireza Kamalipour, Natchada Tansuebchueasai, Mohsen Adelpour, Linda M Zangwill, Robert N Weinreb, Sasan Moghimi
Summary
Baseline OCT-based risk scores from a single wide scan, incorporating both peripapillary RNFL and macular thickness measurements, are associated with the subsequent development of perimetric glaucoma development in glaucoma suspect eyes, offering valuable insights for…
Abstract
PURPOSE
To evaluate the predictive ability of baseline widefield OCT-based risk scores for the development of perimetric glaucoma in glaucoma suspect eyes.
DESIGN
A prospective cohort study.
PARTICIPANTS
Two hundred fifty-seven glaucoma suspect eyes with normal visual fields (VFs) at baseline from 180 patients.
METHODS
Baseline macular and peripapillary inner retinal layer thickness measurements were obtained using swept-source OCT widefield scans (12 mm × 9 mm) that included both the optic disc and macula. Three OCT-based risk scores were calculated for each eye using measurement segments and grids from the OCT reports. Glaucoma progression was defined as the development of repeatable abnormal VF results. Marginal Cox proportional hazards model was applied to evaluate baseline predictors, including risk scores and individual OCT thickness metrics, with adjustment for confounding factors.
MAIN OUTCOME MEASURES
Time-dependent receiver operating characteristic (ROC) curves and decision curve analysis evaluated model performance and clinical utility.
RESULTS
The mean follow-up time for glaucoma suspects was 2.8 years. During this period, 80 suspect eyes progressed to perimetric glaucoma. Circumpapillary retinal nerve fiber layer (RNFL) thickness, ganglion cell complex thickness, ganglion cell-inner plexiform layer thickness, mean deviation, and the 3 OCT-based risk scores were predictors of progression in univariable models, showing significant hazard ratios, whereas only the OCT-based risk scores remained statistically significant in multivariable models. Each 10-point increase in OCT-based risk scores across the 3 models corresponded to a 1.15-fold to 1.21-fold increase in risk. The risk scores demonstrated higher area under ROCs, compared to individual OCT parameters, across all time points. Decision curve analysis showed that individual OCT parameters offered no greater net benefit than treating all patients regardless of risk, whereas the Fukai risk scores consistently yielded higher net benefit.
CONCLUSIONS
Baseline OCT-based risk scores from a single wide scan, incorporating both peripapillary RNFL and macular thickness measurements, are associated with the subsequent development of perimetric glaucoma development in glaucoma suspect eyes, offering valuable insights for risk stratification in a single integrated metric and may aid in early intervention and clinical management in glaucoma suspects. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
Keywords
More by Huiyuan Hou
View full profile →Development and Validation of a Deep Learning System to Detect Glaucomatous Optic Neuropathy Using Fundus Photographs.
Measurement Floors and Dynamic Ranges of OCT and OCT Angiography in Glaucoma.
Macular and Optic Nerve Head Vessel Density and Progressive Retinal Nerve Fiber Layer Loss in Glaucoma.
Top Research in Optic Nerve & Disc
Browse all →Efficacy of a Deep Learning System for Detecting Glaucomatous Optic Neuropathy Based on Color Fundus Photographs.
Relationship between Optical Coherence Tomography Angiography Vessel Density and Severity of Visual Field Loss in Glaucoma.
Inflammation in Glaucoma: From the back to the front of the eye, and beyond.
In the Knowledge Library
Discussion
Comments and discussion will appear here in a future update.