Authors: Claudio I. Perez Valenzuela, Mahadev Bhalla, Kulbir Gill, Priya Gupta, Frederick S. Mikelberg, Steven Schendel. University of British Columbia.

Author Disclosure Block:C.I. Perez Valenzuela: None.M. Bhalla: None.K. Gill: Any direct financial payments including receipt of honoraria; Name of for-profit or not-for-profit organization(s); Allergan, Thea. Membership on advisory boards or speakers’ bureaus; Name of for-profit or not-for-profit organization(s); Allergan.P. Gupta: Any direct financial payments including receipt of honoraria; Name of for-profit or not-for-profit organization(s); Allergan, IDx. Membership on advisory boards or speakers’ bureaus; Name of for-profit or not-for-profit organization(s); Bausch and Lomb, Allergan, Johnson & Johnson, Santen.F.S. Mikelberg: None.S. Schendel: Any direct financial payments including receipt of honoraria; Name of for-profit or not-for-profit organization(s); Allergan, Alcon, Bausch and Lomb, Glaukos, Novartis, Thea, Santen. Membership on advisory boards or speakers’ bureaus; Name of for-profit or not-for-profit organization(s); Allergan.

Abstract Body:

Purpose: To evaluate if the structure-function relationship between the visual field sensitivity and the retinal nerve fiber layer (RNFL) thickness changed with different peripapillary circle scan diameters, axial length, and Bruch’s membrane opening (BMO) area values.
Study Design: Cross-sectional study
Methods: Eyes of adult subjects with or without glaucoma that had a 24-2 SITA standard visual field test and global peripapillary RNFL thickness value measured with a spectral domain optical coherence tomography (OCT) the same day were included in the study. All eyes had axial length measurement. Exclusion criteria were history of or current retinal disease, nonglaucomatous optic neuropathy, unreliable visual field and OCT segmentation artifact. Mixed-effects nonparametric regression analysis were performed to calculate the r-squared values to evaluate the relationship between the mean deviation (MD) and the global RNFL thickness measured with 3.5-mm, 4.1-mm, and 4.7-mm peripapillary scans. RNFL dynamic range and number of steps to RNFL floor were calculated.
Results: 346 eyes from 291 patients were enrolled after applying inclusion and exclusion criteria. The median age of the patients was 69.4 (IQR 57.9, 77.1) years and 64.45% of them had glaucoma diagnosis, while 35.55% did not. The median of the visual field test MD was -3.7 (IQR -11.3, -1.1) dB and for the global RNFL was 73 (IQR 54, 90) µm. Overall, the best R-squared value was obtained when the RNFL thickness was calculated in the 3.5-mm scan (0.71). When using the 4.1-mm and 4.7-mm scans the value decreased to 0.65 and 0.59, respectively. In addition, the R-squared value improved to 0.79 in the subgroup of eyes with axial length <23.5 mm and BMO area >2.0 mm². When measuring with the 3.5-mm scan, the dynamic range of the RNFL thickness was 47.1 µm, which decreased to 44.4 µm and improved to 50.9 µm when this parameter was analyzed in eyes with BMO area <2.0 mm² and >2.0 mm², respectively. In addition, the number of steps to RNFL floor improved from 9.2 to 10.3 when comparing eyes with axial length >23.5 versus <23.5 mm, respectively, and increased from 8.9 to 10.3 when comparing eyes with BMO area <2.0 versus >2.0 mm², respectively.
Conclusions: RNFL thickness measurement should be evaluated using the 3.5-mm peripapillary scan, unless there is a segmentation artifact. A greater dynamic range and number of steps to RNFL floor among eyes with shorter axial length and bigger BMO area might improve the structure-function relationship and the ability to detect OCT progression in these eyes.