Diagnostic accuracy of Bruch's membrane opening: minimum rim width for detection of perimetric glaucoma and suspect glaucoma patients
Session Details
Session Title: Presented Poster Session: Glaucoma and Imaging
Venue: Poster Village: Pod 3
First Author: : Q.de Bosredon FRANCE
Co Author(s): : C. Dutheil D. Touboul C. Schweitzer
Abstract Details
Purpose:
Purpose: To evaluate diagnostic performances of Bruch Membrane Opening - Minimum Rim Width (BMO-MRW) and Retinal Nerve Fiber Layer (RNFL) parameters for detection of glaucoma and glaucoma suspects patients.
Setting:
Bordeaux University Hospital
Methods:
Methods: A total of 171 eyes from age- and sex-matched patients (n=93) with perimetric primaryopen-angle glaucoma (POAG) (n=31), glaucoma suspects (GS) (n=31) and healthy controls (n=31).
Patients were recruited between March and May 2016. Demographic and clinical factors wererecorded, including visual field parameters from the Humphrey Visual Field Analyzer (Carl Zeiss Meditec Inc, Dublin, CA). Spectral-Domain Optical Coherence Tomography (SD-OCT) (Spectralis, Heidelberg Engineering, Heidelberg, Germany) was performed to obtain BMO-MRW thickness and peripapillary RNFLs thicknesses (at 3,5 mm, 4,1 mm and 4,7 mm) on both eyes of 93 POAG, GS patients and controls.
Results:
Results: Globally, BMO-MRW yielded to better diagnostic performance than RNFL for identifyingglaucoma suspects. Global BMO-MRW AUROC was 0.86 (0.76;0.96), global RNFL was 0.72 (0.59;0.85) for glaucoma suspects. Global BMO-MRW AUROC was 0.92 (0.85;0.98), global RNFL was0.91 (0.83;0.99) for glaucoma patients. Sectorally, the AUROC ranged from 0.79 to 0.87 for BMO-MRW and from 0.58 to 0.73 for RNFL for glaucoma suspects, and from 0.85 to 0.93 for BMO-MRW and from 0.74 to 0.91 for RNFL for glaucoma patients.
Conclusions:
Conclusions: BMO-MRW is more discriminating than RNFL for identifying glaucoma suspects. A longitudinal study is ongoing to determine the role of BMO-MRW in the follow-up of glaucomasuspects and glaucoma patients.
Financial Disclosure:
NONE
