Reproducibility and Screening Capability of Corneal Epithelial Thickness Measurement for Keratoconus Using Anterior Segment Optical Coherence Tomography

Taro Oshika; Ayako Sawaki; Tomoya Nishida; Tomoaki Nakamura; Takashi Kojima

doi:10.2147/OPTH.S528630

Reproducibility and Screening Capability of Corneal Epithelial Thickness Measurement for Keratoconus Using Anterior Segment Optical Coherence Tomography

Clin Ophthalmol. 2025 Jun 30:19:2057-2065. doi: 10.2147/OPTH.S528630. eCollection 2025.

Authors

Taro Oshika¹, Ayako Sawaki^{2

3}, Tomoya Nishida³, Tomoaki Nakamura³, Takashi Kojima^{2

3}

Affiliations

¹ Department of Ophthalmology, Japanese Red Cross Gifu Hospital, Gifu, Japan.
² Department of Ophthalmology, Japanese Community Healthcare Organization Chukyo Hospital, Nagoya, Japan.
³ Department of Ophthalmology, Nagoya Eye Clinic, Nagoya, Japan.

Abstract

Purpose: To investigate the reproducibility of corneal epithelial thickness measurements using anterior segment optical coherence tomography (AS-OCT, CASIA2, TOMEY) and evaluate its potential for the diagnosis of keratoconus.

Patients and methods: This study included patients diagnosed with stage 1-2 keratoconus (Amsler-Krumeich classification) at the Nagoya Eye Clinic and healthy controls. Corneal epithelial thickness was measured with the CASIA2 AS-OCT. Sixteen patients with keratoconus (16 eyes) and 11 healthy controls (11 eyes) underwent three consecutive measurements to determine the intraclass correlation coefficient (ICC). Measurements were taken for eight corneal sections in 40 keratoconus eyes (mean age: 25.5 ± 8.0 years) and 44 normal eyes (mean age: 24.2 ± 5.1 years). The mean corneal epithelial thickness, the upper-to-lower corneal epithelial thickness ratio, and the standard deviation of thickness across segments were compared between the two groups. Logistic regression analysis was used to evaluate corneal epithelial thickness-related parameters for the diagnosis of keratoconus. The cutoff value was determined using the Youden index derived from the receiver operating characteristic (ROC) curve.

Results: The ICC was ≥0.95 in all areas for both keratoconus and normal eyes. The superior paracentral corneal epithelium was significantly thicker in the keratoconus eyes, whereas the inferior paracentral area was thinner compared to normal eyes (p<0.001). The standard deviation of corneal epithelial thickness was significantly higher in keratoconus eyes across all regions. Logistic regression analysis identified the standard deviation of corneal epithelial thickness (2-5 mm) and the mean thickness ratio (0-5 mm) as explanatory variables. The area under the ROC curve was 0.955, with sensitivity and specificity of 92.1% and 100%, respectively.

Conclusion: Corneal epithelial thickness measurements using the AS-OCT CASIA2 were highly reproducible and showed significant differences between early-stage keratoconus and normal corneas. These findings suggest its potential utility as a screening tool for keratoconus.

Keywords: anterior segment optical coherence tomography; corneal epithelial thickness; diagnosis; keratoconus.