Development and validation of the epithelial thickness ratio for detecting subclinical keratoconus
Authors: Mary B. Holdsworth 1, Davin Johnson2. 1Queen's University School of Medicine, 2Department of Ophthalmology, Queen's University.
Author Disclosures: M.B. Holdsworth: None. D. Johnson: None.
Abstract Body:
Purpose: The aim of this study was to develop a quantitative method of describing corneal epithelial changes associated with corneal ectasia. We described and validated an epithelial thickness ratio (ETR) derived from OCT corneal epithelial thickness maps.
Study Design: This retrospective chart review consisted of patients with known keratoconus presenting from a hospital-based practice as well as consecutive patients presenting for laser refractive surgery.
Methods: The study sample included 93 eyes from 47 patients. We developed an ETA to quantify corneal epithelial changes seen on OCT. The ETR was defined as the ratio of the mean epithelial thickness at 4 points 90 degrees apart at a defined radius from the thinnest point of the cornea to the epithelial thickness at the thinnest point of the cornea. To validate the ETR at radius 1 mm (ETR1), 1.5 mm (ETR1.5) and 2 mm (ETR2), we generated receiver operating characteristic (ROC) curves using the Pentacam D score at cut-offs of 1.5 and 2.
Results: he median age of the participants was 35 (IQR 15). We report the sensitivity and specificity of all three ETR distances that were considered (1 mm, 1.5 mm, and 2 mm). For a D score cut-off of >2 and ETR cut-off of >1, the sensitivities and specificities for ETR1 were 77% and 76%, for ETR1.5 91% and 78%, and for ETR2 91% and 83%. Comparatively, with a more conservative D-score cut off of >1.5, we found ETR 1 to be 56% sensitive and 78% specific, ETR1.5 was 64% sensitive and 80% specific, and ETR2 was 59% sensitive and 83% specific.
Conclusions: In conclusion, the ETR derived from corneal epithelial thickness mapping shows promise for quantitatively differentiating subclinical keratoconus. Given the increased availability and affordability of OCT versus Pentacam technology, methods for detecting early keratoconus that rely primarily on epithelial mapping may be more accessible.
