Glaucoma Structural and Functional Progression in American and Korean Cohorts

Abstract

 

Purpose

 

Numerous studies on glaucomatous structural and functional progression have been published but it is unclear whether these results can be generalized given different geographic locations, ethnicities and disease types. In this study, we compare the rate of glaucomatous progression in American and Korean cohorts.

 

Methods

 

Glaucoma suspect and glaucomatous eyes with ≥4 reliable visual fields (VF; Humphrey Field Analyzer; Zeiss, Dublin, CA) and optical coherence tomography (OCT; Cirrus HD-OCT, Zeiss) were included in the study. VF mean deviation (MD), visual field index (VFI), peripapillary retinal nerve fiber layer (RNFL) thickness, and cup to disc (C/D) ratio were used in linear mixed effect analysis while accounting for baseline age for estimation of progression rate.

 

Results

 

The American cohort included 103 glaucoma and 73 glaucoma suspect eyes and the Korean cohort included 100 and 103 eyes, respectively. The overall follow-up period was 36.2±9.2 months. The majority (96.6%) of the American glaucoma eyes were categorized as high-tension glaucoma, whereas 85.7% of the Korea glaucoma eyes were diagnosed as low-tension glaucoma. At baseline, the Korean cohort was statistically significantly younger than the American (56.8 vs. 61.8 yrs.) with thicker RNFL (81.3 vs. 77.4 μm) and a larger cupping (0.7 vs. 0.6). There was no statistically significant difference in baseline MD between cohorts (-2.5 vs. -2.8 dB). No significant difference was detected in the rate of change between the cohorts for any of the parameters in the glaucoma suspect group. However, Korean glaucoma eyes showed significantly steeper progression slopes by average and nasal RNFL thickness and significantly shallower slope for average C/D ratio (Table).

 

Conclusions

 

Patient ethnicity and disease type should be considered when comparing glaucoma structural progression studies.  

 

Table. Progression rate in glaucomatous eyes. Parameters are reported as mean and 95% confidence intervals (linear mixed effect modeling).<br /> <br />