Longitudinal changes in visual field sensitivity compared to estimates of retinal ganglion cell loss in glaucoma

Abstract

 

Purpose

 

To evaluate the relationship between rates of visual field loss in standard automated perimetry (SAP) and longitudinal changes in estimated numbers of retinal ganglion cells (RGCs) in glaucoma.

 

Methods

 

A longitudinal study of 202 eyes with glaucoma recruited from the Diagnostic Innovations in Glaucoma Study (DIGS) at the University of California, San Diego. All subjects underwent SAP and optical coherence tomography (OCT). The estimated number of RGCs was calculated using previously described formulas based on OCT retinal nerve fiber layer thickness. Glaucoma severity was defined by baseline SAP MD as early (better than -4dB), moderate (-4 to -8dB), and severe (worse than -8dB). The rates of change in MD and estimated RGC counts were calculated and the influence of disease severity on the relationship between rates of change in MD and estimated rates of ganglion cell loss was examined using a regression model.

 

Results

 

The mean (± standard deviation) age of patients was 66.3 ± 11.2 years. There was no significant difference in mean age among subjects with early, moderate and severe disease. The mean baseline MD was -3.4 ± 4.3 dB and the mean baseline estimated RGC count was 673,827 ± 221,949. The average follow up time was 6.3 years with a range of 2.1 to 11.0 years and an average of 8.6 ± 2.7 visits. Baseline disease severity had a significant influence on the relationship between rate of change in MD and the rate of change in estimated RGCs (P<0.001, Figure 1). A 1 dB/year rate of MD loss corresponded to an estimated RGC loss of 48,049 cells/year in early disease versus 32,833 cells/year in moderate and 16,707 cells/year in severe disease.

 

Conclusions

 

This longitudinal study, in agreement with previous cross-sectional analyses, confirms that the relationship between rates of change in SAP MD and rate of estimated neural loss is influenced by disease severity. In early disease, larger estimated neural losses are necessary for a 1dB change in MD than are necessary in more severe disease. Reliance on SAP to measure rates of disease progression is likely to underestimate RGC loss in early and moderate disease.

 

 

Figure 1. The relationship between rates of change in estimated RGC counts and SAP MD according to baseline disease severity.

 

Keywords: 758 visual fields • 531 ganglion cells • 642 perimetry