Rapid DSEK adoption

Data from the National Transplant Registry show there was rapid and steady adoption of Descemet’s stripping endothelial keratoplasty (DSEK) by surgeons in the UK after the procedure was introduced. And, in 2013, almost three-quarters of patients needing a corneal transplant for Fuchs’ endothelial dystrophy or pseudophakic bullous keratopathy (PBK) underwent the partial thickness procedure rather than penetrating keratoplasty (PK), reported Frank Larkin MD at the 5th EuCornea Congress in London.

Outcomes analyses for the two procedures using data from comparable six-year periods provide some evidence to support the current practice pattern, while findings from analyses of factors predicting surgical outcomes suggest targets for improving DSEK outcomes.

The comparative analyses included first PK procedures performed from January 2000 through to December 2005 (1,243 PBK, 1,373 Fuchs’) and first DSEKs performed from December 2006 through to January 2011 (799 PBK, 1,275 Fuchs’). All eyes had two years of follow-up.

Summarising the key findings, Dr Larkin reported that PK was associated with significantly better graft survival than DSEK. However, DSEK graft survival was improving as centres gained experience, and among patients with surviving grafts at two years, DSEK was associated with better visual and refractive outcomes compared with PK.

Multivariate analyses conducted to identify variables affecting graft survival identified centre experience and donor endothelial cell density among other independent predictive factors.

Speaking about the implications of the findings, Dr Larkin outlined several points to take away.

“Individual surgeons need to look at a new procedure and think about what advantages it gives over the procedure they are currently doing. Before transitioning to DSEK, surgeons getting good results with PK should consider how many procedures they will be doing and the length of their remaining surgical career,” he said.

“In addition, it is perfectly clear from these data that in the UK the current arrangements for training in evolving lamellar surgical techniques and for donor cornea preparation are inadequate.”

The UK National Transplant Registry provides an excellent source for comparing procedural outcomes because all grafts are registered centrally, and the time performing DSEK reached a plateau in 2010-2011. Data from Kaplan-Meier analysis of two-year graft survival rates showed outcomes varied by procedure and diagnosis, but the rate was significantly higher for PK vs DSEK for both eyes with Fuchs’ (94 per cent vs 81 per cent) and those with PBK (79 per cent vs 70 per cent).

“These findings are consistent with every published study in showing that graft survival in transplant procedures for PBK are worse than for Fuchs’. The reasons for that difference are not understood,” said Dr Larkin, consultant ophthalmic surgeon, Moorfields Eye Hospital, London.

“Considering the poorer graft survival with DSEK, one might wonder why patients in the UK are having endothelial keratoplasty. However, graft survival, which means a clear, functioning graft, is one thing, while data on visual outcomes for surviving grafts tells another interesting story.”

Visual outcomes

Analyses of the visual outcomes at two years among patients with surviving grafts showed that mean logMAR BCVA was significantly better for the DSEK group compared to the PK eyes both for patients with Fuchs’ (0.3 vs 0.4) and for those with PBK (0.6 vs 0.9).

In addition for the cohort with surviving grafts at two years, the proportion of eyes in which SE was within two dioptres of plano was significantly higher among DSEK patients than in the PK group, both in the Fuchs’ (75 per cent vs 51 per cent) and PBK (67 per cent vs 47 per cent) subgroups.

In the multivariate analyses to identify predictors of surgical outcomes for the EK procedures in patients with Fuchs', lower centre experience and lower donor ECD were associated with poorer graft survival, along with existing glaucoma, male donor, older donor (age 75+), and preoperative BCVA worse than 6/60.

Factors associated with poorer graft survival after DSEK in PBK eyes were lower centre experience, existing glaucoma and year of DSEK (2008-2009). Surgeon experience was not predictive of graft survival for either DSEK subgroup.

“Our interpretation of these data is that surgeons starting DSEK and operating in a centre with experienced colleagues should likely have a shorter learning curve and encounter fewer complications,” said Dr Larkin.

Focusing on some of the other predictors, Dr Larkin commented that, as with every type of corneal transplant, glaucoma at the time of DSEK is a significant predictor of graft survival. Likely patients with more advanced Fuch' had poorer outcomes because they have stromal disease rather than simply endothelial dysfunction.

He noted that five-year outcomes will be analysed as those data become available, and future analyses may also include eyes undergoing DMEK should that procedure become more widely adopted.

With that in mind and responding to an attendee’s comment about the potential for DMEK to provide patients with unaided driving vision, Dr Larkin noted that while the registry does not collect UCVA data, it is something to consider in the future.

Frank Larkin: f.larkin@ucl.ac.uk