Purchase this article with an account.
Georgios Kontadakis, Christopher Smith, Daniel Kaitis, Rocio Bentivegna, Jordan Winegar, Sonia H Yoo, Victor L Perez, Mohamed Abou Shousha; Diagnostic performance of the Endothelial/Descemet’s membrane thickness versus endothelial cell density in the diagnosis of corneal graft rejection.. Invest. Ophthalmol. Vis. Sci. 2016;57(12):1933.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess the diagnostic performance of the Endothelial/Descemet’s membrane thickness (En/DMT) as measured by the high resolution optical coherence tomography (HD-OCT) in the diagnosis of corneal graft rejection and compare it to that of endothelial cell density (ECD) and central corneal thickness (CCT).
This prospective study includes 42 corneal grafts. Assessment included detailed history, clinical examination, automated assessment of the corneal endothelial layer with corneal confocal microscopy, and corneal imaging with HD-OCT. Blinded observers calculated manually CCT and EnDMT on HD-OCT images. Eyes were divided into two groups, rejected and clear grafts. CCT, EnDMT and ECD were compared between groups. Diagnostic performance of EnDMT for the diagnosis of graft rejection was evaluated using receiver operating characteristic (ROC) curves.
CCT and EnDMT showed statistically significant differences among groups. Clear grafts’ mean CCT was 525 μm (±74) and mean EnDMT was 15μm (SD 3). Whereas, in the rejected grafts group, mean CCT was 809μm (±251) and mean EnDMT was 48 μm (SD 49, p<0.05 for corresponding comparisons, Mann-Whitney U test). On the other hand, no significant difference in ECD was found between the two groups. EnDMT showed better diagnostic accuracy than CCT (EnDMT area under the curve, AUC was 0.93, p<0.001 and for CCT AUC was 0.81, p<0.001).
Endothelial/Descemet’s membrane thickness demonstrated excellent accuracy in the diagnosis of corneal graft rejection, which is significantly better than that of central corneal thickness and endothelial cell density.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
This PDF is available to Subscribers Only