Identifying the Best Structural and Functional Parameters to Detect Glaucoma Progression

Abstract

Purpose: : To identify the best parameters to detect glaucoma progression among visual field (VF) and optic nerve head (ONH) parameters.

Methods: : 110 eyes of 60 subjects (24 healthy, 48 glaucoma suspects and 38 glaucomatous eyes) had comprehensive ocular examination and ≥4 visits with reliable VF testing (Carl Zeiss Meditec, Dublin, CA) and good quality spectral-domain optical coherence tomography (SD-OCT; RTVue-100, Optovue, Freemont, CA) and scanning laser ophthalmoscopy (SLO; HRT III, Heidelberg Engineering, Heidelberg, Germany). Structural equation models (SEM) were constructed to compute the common latent factor measurement error for each device using standardized global parameters as provided by the devices along with clinical diagnosis and baseline age.

Results: : The median follow-up period was 3.93 years, and the median number of visits was 6.5. For VF, the Visual Field Index (VFI) showed statistically significantly better correlation with the common latent progression (ρ=0.93; confidence interval (CI): 0.76-1.00) than both mean deviation (MD) and pattern standard deviation (PSD). For SD-OCT the parameters with the highest correlation with latent progression were cup area (ρ=0.92; CI: 0.80-1.00) and rim area (ρ=-0.92; CI: -1.00--0.80) and statistically significantly better than other measured parameters. SLO also showed cup area (ρ=-0.85; CI: -1.00--0.73) and rim area (ρ=-0.85; CI: 0.74-1.00) to be the parameters most correlated with latent progression. For all three devices the diagnosis had similar effect and was statistically significant. Using latent progression scores for each eye derived from the SEM, there was poor correlation between VF and SD-OCT progression scores (r=0.17), SD-OCT and HRT (r=0.36), and HRT and VF (r=0.03).

Conclusions: : Cup and rim area were the most useful structural measurements of progression, while VFI was the most useful functional parameter. Even though SD-OCT and HRT quantify ONH structure there was a poor correlation between the devices in detecting progression.

Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • visual fields