Abstract
Purpose :
This study aims to determine if anterior segment-optical coherence tomography (AS-OCT) epithelial thickness mapping may provide a novel diagnostic tool to identify early-stage neurotrophic keratopathy (NK) from other ocular surface diseases.
Methods :
A retrospective comparative pilot study was conducted on 20 patients, with 14 diagnosed with stage 1 NK (NK-1) and 6 diagnosed with dry eye disease (DED). All patients underwent AS-OCT with corneal epithelial thickness mapping in a central 7 mm diameter, divided into 3 rings (0-2mm, 2-5mm, and 5-7mm) and measured for minimum and maximum values and the difference between minimum and maximum. Patients with NK-1 were diagnosed based on criteria including 3+ central corneal staining and Cochet-Bonnet esthesiometry of less than 4 mm. DED patients were diagnosed based on signs and symptoms of DED and Cochet-Bonnet esthesiometry of 4mm or greater.
Results :
There was no significant gender difference between groups (p=0.573). However, there was a greater age in NK-1, with a mean age of 62.5±15.4 in contrast to 46.6±10.5 in DED (p=0.035). Epithelial thickness was globally reduced in the NK-1 compared to DED. The distribution within the corneal rings were (in mm): 0-2mm (46.7±7.1 in NK-1 and 52.5±3.0 in DED, p=0.021), 2-5mm (46.7±5.4 in NK-1 and 53.0±3.3 in DED p=0.007), 5-7mm (46.8±4.9 in NK-1 and 52.1±2.0 in DED, p=0.024), 0-7mm (46.9±5.2 in NK-1 and 52.6±2.7 IN DED, p=0.005). The minimum and maximum values also showed statistical significance, with a minimum of 32.2±9.3 mm in the NK-1 and 47.1±1.4 mm in DED (p<0.001), and a maximum of 57.3±6.5 mm in NK-1 and 58.3±4.3 mm in DED (p=0.736 the only measurement that did not show statistical significance). The difference between minimum and maximum values was also statistically significant, with 11.1±3.1 mm in NK-1 and 25.0±10.7 mm in DED (p<0.001).
Conclusions :
AS-OCT epithelial thickness mapping may aid in the differentiation of early-stage NK from DED. Larger studies are needed to confirm this pilot data.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.