Discordance between structure and function in glaucoma: Possible anatomical explanations

Abstract

Purpose The aim of this study was to analyze glaucomatous eyes that show discordance between structure and function and hypothesize plausible anatomical explanations for discordance. Methods Fifty eyes from 50 consecutive subjects with glaucoma (which was diagnosed according to multicenter criteria) were studied. One eye from each subject was selected randomly for inclusion, counterbalancing right eye (O.D.) and left eye (O.S.). By comparing the available structural information (from clinical disc assessment as well as GDx retinal nerve fiber layer [RNFL] measurements) to the Swedish Interactive Threshold Algorithm (SITA) Standard 24-2 (SS24-2) visual field results, study eyes were assigned to the following categories: (1) positive structure-function correlation, (2) structural abnormalities with no functional deficits, and (3) functional abnormalities with no structural deficits. Structure and function also were compared on a statistical basis, utilizing the nerve fiber indicator (NFI) of the GDx and the mean deviation (MD) and pattern standard deviation (PSD) of SS24-2 visual fields. Results Forty-four eyes were classified in category 1, 5 eyes in category 2, and 1 eye in category 3. Scatter plots showing the relationship between the NFI and MD and between the NFI and PSD depict a positive correlation in the 44 category 1 eyes with R 2 values of 0.465 and 0.322, respectively. The remaining 6 eyes with discordant structural and functional findings were analyzed in detail. Conclusion The majority (88%) of glaucoma cases show concordance between structural loss and functional deficits. The primary proposed explanations for discordance include (1) visual field sampling and test selection limitations (i.e., the SS24-2 samples only axons that are anatomically connected to photoreceptors in the central retina, whereas the GDx samples virtually all axons) and (2) GDx measurement limitations in the papillo-macula bundle (i.e., the GDx has difficulty differentiating the normally thin RNFL from the pathologically even thinner RNFL). Tests of both structure and function are recommended in glaucoma suspects and patients, as neither mode is capable of identifying all glaucomatous deficits. Due to the apparent diagnostic benefit that the peripheral functional data may provide, we strongly recommend that future studies evaluating the structure–function relationship in glaucoma record peripheral (60-4) visual fields for all subjects at the outset of the study.