Morphometry of corneal endothelium in patients with corneal guttata

Abstract

Objective: To study the morphology of the corneal endothelium in patients diagnosed with corneal guttata using an image processing algorithm based on a contour detection method.

Design: Retrospective observational case series.

Participants: Twenty-four subjects with known corneal guttata.

Methods: Two hundred eight images of corneal endothelium, captured with a noncontact specular microscope were analyzed using the Contour method, which demonstrates endothelial cell density (ECD), coefficient of variation of cell area, percentage of 4- to 8-sided cells as well as the number, area, and coefficient of variation of corneal guttata.

Main outcome measures: The number, surface area, and coefficient of variation of corneal guttata.

Results: Corneal position had no significant effect on ECD or on the percentage of endothelial cells with 4, 5, 7, or 8 sides. However, the coefficient of variation of images taken from the central cornea was significantly larger than those taken at the 2- and 6-o'clock positions. In addition, the percentage of hexagonal cells was significantly lower in pictures of the central position compared to those located in the upper paracentral position. The numbers and surface areas of guttata were significantly larger in pictures of the central compared to some paracentral positions. Subjects who had previously undergone cataract surgery with intraocular lens (IOL) implantation did not show different areas of corneal guttata, but exhibited a significantly lower cell density (1825+/-582) compared with unoperated patients (2400+/-457/mm2). Analogously, the only significant change observed in paired comparisons between the operated eye of patients with unilateral cataract extraction with IOL implantation and their unoperated fellow eye was a lower cell density obtained in operated eyes. Compared with normal subjects, subjects with corneal guttata were shown to have a significantly lower ECD, a lower proportion of hexagonal cells, and a higher coefficient of variation of cell area in the central cornea.

Conclusion: This study supports the finding that corneal guttata mainly affect the central corneal area. A future prospective study using the described Contour detection method would be helpful to evaluate more accurately the risks associated with the evolution of corneal guttata into Fuchs' dystrophy.