Search Title by author or title

Outcomes of repeat corneal transplantation performed in a tertiary centre: indications, graft survival, and visual outcomes

Poster Details

First Author: V.Dimacali SINGAPORE

Co Author(s):    H. Ong   S. Lang   H. Cajucom-Uy   J. Mehta           

Abstract Details


To determine the indications and methods of repeat corneal transplantation, and to compare the outcomes among the different types of regrafts.


Tertiary eye centre


This was a retrospective, consecutive, non-comparative case series conducted at the Singapore National Eye Centre. Data were gathered from the charts of 463 patients who underwent repeat corneal transplantations from 2007 to 2018, of which 376 were for optical indications.


The most frequent indications for repeat transplants were previous graft rejection (36.1%) and late endothelial failure (12.5%). The methods of regrafting were Descemet membrane endothelial keratoplasty (DSAEK) 47.5%, penetrating keratoplasty (PK) 40.0%, anterior lamellar keratoplasty (ALK) 11.0%, and Descemet membrane endothelial keratoplasty (DMEK) 1.5%. The types of regrafts performed were PK/PK 31.1%, DSAEK/DSAEK 25.9%, PK/EK 20.5%, DSAEK/DMEK 1.1%, ALK/ALK 11.0%, ALK/PK 5.8%, DSAEK/PK 3.0%, DSAEK/DMEK 1.1%, DMEK/DSAEK 1.1%, and DMEK/DMEK 0.4%. Looking at regrafts for optical indications, 77.8% of cases had a best corrected visual acuity of at least 6/12 at one year, and were greatest in EK/EK group (83.7%). Among all PK regrafts, at one year, repeat EK showed a significantly greater success rate (87.5%) compared to repeat PK (70.2%) (p<0.001), defined as an optically clear graft recorded in the charts. Kaplan-Meier survival analysis showed that PK/EK and EK/EK had the highest graft survival rates at 3 years (respectively 75.9%, 75.7%) compared to PK/PK (p=0.001) and EK/PK (p<0.001).


In patients with previous PK, performing repeat EK improved graft survival rates. The EK/EK group had the highest success rate and best visual outcomes among the different types of regrafts.

Financial Disclosure:


Back to Poster listing