Comparison of Two OCT Device RNFL and GCIPL Measurements in Mild, Moderate, and Severe Glaucoma

Overview

This research study was designed to evaluate how consistently two different imaging devices measure the thickness of retinal layers in patients with glaucoma. The two machines, the Topcon Maestro2 and the Zeiss Cirrus 5000, are commonly used in eye clinics to take high-resolution scans of the retina and optic nerve. These scans help doctors monitor glaucoma and detect disease progression. The goal of this study was to compare the results from both machines when scanning the same patients, and to determine how closely their measurements align. Researchers also developed mathematical models that allow doctors to convert results from one machine to the other, which could help ensure consistency when a patient receives scans on different devices over time or at different clinics. A total of 141 eyes from patients with varying levels of glaucoma severity were included in the study. Scans were performed using standard imaging protocols on both machines. Researchers analyzed the average thickness of two important retinal layers: the retinal nerve fiber layer (RNFL) and the ganglion cell-inner plexiform layer (GCIPL). This study helps improve our understanding of how different imaging devices compare and provides tools for standardizing results across platforms. The findings may help eye care providers make more confident decisions when reviewing images from different machines, improving the care and monitoring of patients with glaucoma. This was an investigator-initiated study sponsored by Twin Cities Eye Consultants with funding support from Topcon Healthcare. No new medications, treatments, or procedures were tested as part of this research. The study only involved non-invasive eye imaging scans that are already widely used in clinical practice.

This investigator-initiated, cross-sectional study was designed to assess the level of agreement between optical coherence tomography (OCT) measurements captured by two commercially available imaging platforms: the Topcon Maestro2 and the Zeiss Cirrus 5000. These devices are widely used in the clinical management of glaucoma to quantify structural biomarkers such as retinal nerve fiber layer (RNFL) thickness and ganglion cell-inner plexiform layer (GCIPL) thickness. The motivation behind this study stems from real-world variability in imaging hardware across clinical settings. Patients may undergo imaging on different machines due to equipment availability, clinic referral patterns, or evolving device infrastructure. As each device uses proprietary software algorithms and scan acquisition protocols, there can be small but clinically significant differences in reported measurements-even when scanning the same eye under similar conditions. Understanding this variability is critical for ensuring continuity of care and accuracy in longitudinal monitoring of glaucoma. The study enrolled 141 eyes from patients with established glaucoma, screened from a single clinical practice. The first 72 eyes enrolled were designated as the derivation cohort, while the subsequent 69 eyes comprised the validation cohort. Each subject underwent back-to-back imaging on both devices during the same session using manufacturer-recommended scan protocols: the Topcon Wide Scan, RNFL Cube, and Macular Cube protocols; and the Zeiss Optic Disc Cube 200×200 and Macular Cube 512×128 protocols. All imaging was performed by a single, trained technician using factory-default acquisition settings and standard dilation (phenylephrine 2.5% and tropicamide 1%), with eye-tracking (FastTrac) enabled where available. Average RNFL thickness and GCIPL thickness were extracted from each scan and compared between devices. Measurements were further analyzed by quadrant (RNFL) and sector (GCIPL) to evaluate regional agreement. Generalized linear models and Deming regression were used to evaluate the level of agreement and construct conversion equations between devices. Results were evaluated based on clinical thresholds of ±3 microns-a commonly accepted tolerance for scan-to-scan variability on a single device. Repeatability of Topcon scans was also assessed using three consecutive scans on the same visit. Intra-device variability was analyzed using mean differences and 95% limits of agreement to determine reliability across protocols. Exploratory subgroup analyses were performed by glaucoma severity, gender, race/ethnicity, and visual acuity to explore potential drivers of scan variability or poor agreement. The study found that while mean RNFL and GCIPL measurements from the two devices were highly correlated, statistically significant differences in average values were observed. This reinforces the importance of accounting for device-specific variability in both research and clinical practice. Derived conversion equations demonstrated strong performance in the validation cohort, particularly for global RNFL and GCIPL metrics. Sector-level agreement was more variable, particularly in the temporal RNFL and superotemporal GCIPL sectors, indicating that clinicians should use caution when interpreting regional values across platforms. Importantly, scan repeatability on the Maestro2 was excellent, particularly for macular GCIPL metrics, highlighting its reliability for longitudinal monitoring. This supports the growing role of widefield and integrated OCT imaging in clinical glaucoma care. The study was approved by Sterling IRB and conducted at Twin Cities Eye Consultants (Minnesota). It involved no investigational treatments or interventions beyond standard OCT imaging. Funding support was provided by Topcon Healthcare, which had no role in study design, data analysis, or manuscript preparation. Results from this study are intended to inform clinicians, researchers, and device manufacturers regarding the interoperability and clinical translation of OCT imaging biomarkers in glaucoma.