ESCRS - FP18.07 - The Evaluation Of Corneal Tomography Using Swept-Source Optical Coherence Tomography And Scheimpflug Imaging For Ectasia Detection

The Evaluation Of Corneal Tomography Using Swept-Source Optical Coherence Tomography And Scheimpflug Imaging For Ectasia Detection

Published 2023 - 41st Congress of the ESCRS

Reference: FP18.07 | Type: Free paper | DOI: 10.82333/9wj6-wt83

Authors: Robert Herber* 1 , Lutz E. Pillunat 1 , Frederik Raiskup 1

1Department of Ophthalmology,University Hospital Carl Gustav Carus, TU Dresden,Dresden,Germany

Purpose

To determine the diagnostic ability of specific corneal tomography indices to detect corneal ectasia using a swept-source optical coherence tomography system (SS-OCT, ANTERION, Heidelberg Engineering, Germany) and a rotating Scheimpflug camera system (RSC, Pentacam HR, Oculus, Germany).

Setting

Prospective, observational, and cross-sectional study performed at the Department of Ophthalmology, University Hospital Carl Gustav Carus, TU Dresden, Germany

Methods

Subjects of the study cohort were divided into three groups according to specific inclusion criteria. Group 1 were healthy eyes with a K-max value < 47 D, inferior-superior (I-S) value < 1.5 D, and KISA% < 60. Group 2 included one eye of a bilateral clinical keratoconus (KC). Group 3 included eyes with normal topography (same criteria as group 1), where the contralateral eye showed a clinical keratoconus. The main outcome measure was the area under the curve (AUC) of the ROC analysis of the Belin/Ambrosio total deviation value (BAD D) and the Pentacam Random Forest Index (PRFI) provided by RSC compared to the SCORE parameter provided by the SS-OCT. Statistical analysis was performed using the Kruskal-Wallis test and DeLong test.

Results

A total of 354 eyes of 354 subjects were included (group 1: 90, group 2: 209, and group 3: 55). The median of BAD D, PRFI, and SCORE was 0.99, 0.05, and -1.69 for group 1; 7.82, 0.99, and 16.53 for group 2; and 1.46, 0.24, and 0.6 for group 3, respectively (all P < 0.05). When separating group 3 from group 1, the ROC analysis revealed a statistical significant higher AUC for the SCORE parameter (0.894) compared to PRFI (0.783, P=0.003) and BAD D (0.727, P<0.001). The sensitivity and specificity was 76% / 87% (SCORE) compared to 58% / 90% (PRFI), and 47% / 91% (BAD D). When distinguishing group 1 from both ectatic subgroups (group 2 and 3), the AUC was 0.981, 0.961, and 0.948 for SCORE, PRFI, and BAD D (for all comparisons P < 0.05).

Conclusions

The SCORE parameter provided by SS-OCT yielded the highest AUC and therefore the best diagnostic ability to detect corneal ectasia. A clinically acceptable sensitivity and specificity was found for SCORE to distinguish normal eyes from very early ectasia (group 3).