Analysis of a simple method for clinical and subclinical Corneal Endothelial Disfunction (CED) diagnosis with the association of Specular Microscopy (SM) and Pachymetry Maps (PM)

Abstract

Purpose :
Corneal transparency is maintained by the integrity of the endothelial barrier and the endothelial pump, both composed of endothelial cells.
CED is classified as reversible, partially reversible, or irreversible, depending on the degree of endothelial mosaic injury. It has acquired importance in recent years since it has come to be recognized as an indication of endothelial transplantation. The diagnosis and classification of CED may consider the symptoms, SM, and PM.
Purpose: to present a simple method and evaluate its reliability for diagnosing CED using the association of SM with PM.

Methods : Cross-sectional study of 178 eyes with ECD<1000 cells/mm². ECD was calculated with BioOptics LSM 2000C (Relative Error< 0.05); PM was calculated by the Pentacam HR (Oculus, Germany) using the Scheimpflug imaging to determine Pupil Center Pachymetry (PCP).
The ECD threshold value at which ECD=CCP was identified. The difference ECD-PCP defined two groups: Group ECD>PCP and Group ECD<PCP. In both cases, cases with and without CED were identified using ocular signs and symptoms and the MP.
The ROC curve was adjusted and the ECD accuracy was evaluated by estimating the Area Under the Curve (AUC). The cutoff point indicated by the adjustment was analyzed by estimating sensitivity, specificity, accuracy, positive and negative predictive values, and positive and negative likelihood values. Values of p<0.05 indicated statistical significance.

Results : The causes of ECD<1000 cells/mm^₂ are shown (Fig.1A). The value ECD=PCP was 630 (Fig.1B); 40 cases had ECD<PCP, of which 37 had CED; 138 ECD>PPC, of which 15 presented CED (Fig.1C).
The AUC of the ROC Curve and accuracy of Endothelial Cell Density in discriminating the studied groups was 0.98 with statistical significance (p<0.001) (Fig.2A).
In the test of the ECD-PCP rule, the estimated sensitivity was 89.1%, specified 92.5%, and the accuracy of 89.9%. The specified sensitivity, accuracy, and other analytical descriptions for the ECD-PCP rule are demonstrated (Fig.2B).

Conclusions : ECD-PCP rule: Δ- (ECD<PCP) has very good sensitivity for screening of CED and decompensation; Δ+ (ECD≥PCP) has very good sensitivity for normal endothelial function. SM and MP can be used as endothelial semiology if performed with the correct methodology. PM was decisive in the diagnosis of CED where ECD≥PCP (Δ+).

This abstract was presented at the 2024 ARVO Imaging in the Eye Conference, held in Seattle, WA, May 4, 2024.