FS-DSEK versus PK

[caption id="attachment_8827" align="alignnone" width="201"] Yanny Cheng HS[/caption]   Femtosecond laser-assisted Descemet’s stripping endothelial keratoplasty (FS-DSEK) provides better refractive outcomes with less postoperative astigmatism than penetrating keratoplasty (PK) but yields poorer visual outcomes, according to Yanny Cheng MD, PhD, of Leiden University Medical Centre, Leiden, The Netherlands. In a keynote lecture at a Cornea Day session during the 21st ESCRS Winter Meeting in Maastricht, The Netherlands, she noted that the last decade has seen a revolution in corneal transplantation, with endothelial keratoplasty increasingly replacing PK as the treatment of choice for eyes with 
endothelial dysfunction. She pointed out, for example, that in the USA in 2006 PK accounted for 85% of keratoplasty procedures and endothelial keratoplasty only accounted for 12%. By contrast, in 2016 endothelial keratoplasty accounted for 52% of keratoplasty procedures and PK only accounted for 40%. That experience was mirrored in The Netherlands. The proportion of keratoplasty accounted for by PK fell from around two-thirds in 2006 to only one-third in 2016. Conversely, the proportion accounted for by endothelial keratoplasty rose from less than 10% in 2006 to 60 % in 2016, Dr Cheng said. The reason for endothelial keratoplasty’s increasing popularity is that it overcomes many of the limitations of PK, she noted. Those limitations include postoperative astigmatism of more than 5.0D, affecting over a third of patients, suture events, including infection, long-term tectonic instability, and ocular surface disease. However, posterior lamellar keratoplasty also has limitations, including low proportions of patients achieving a best corrected visual acuity (BCVA) of 20/20 and high early endothelial cell loss, she pointed out. TESTING THE FEMTOSECOND LASER Dr Cheng noted that, compared to conventional microkeratome dissection, femtosecond laser graft dissection has several potential advantages over mechanical microkeratomes, including better cut safety and creation of planar grafts. To examine whether these advantages carry over into clinical experience, Dr Cheng and her associates carried out a prospective randomised multicentre study, the Dutch Lamellar Corneal Transplantation Study, comparing 12-month outcomes in 36 eyes undergoing FS-DSEK and 40 eyes undergoing PK. The patients had a mean age of 69 years. Their main indication for surgery was Fuchs’ endothelial keratopathy, accounting for 21 patients in each group, followed by pseudophakic bullous keratopathy, which accounted for 14 and 18 in the FS-DSEK and PK groups, respectively. One eye in each group had aphakic bullous keratopathy. The FS-DSEK group had a higher proportion of patients with vision limiting retinal macular diseases - 39% versus 20%. To prepare the lamellar discs, Dr Cheng and her associates used whole donor globes. They began by removing the epithelium. Then they created a horizontal lamellar cut using a flat applanation lens and an Intralase 30KHz femtosecond laser set to a depth of 400 microns and a diameter of 9.5mm, and using a mean energy level of 1.07mJ and a raster pattern. Following the femtosecond laser dissection, the globe was returned to the eye bank for endothelial evaluation and microbiological and serological examination. The corneoscleral button was then stored in organ culture and the surgeon performed trephination and transplantation. VISUAL OUTCOMES Throughout the 12 months follow-up, mean BCVA was significantly better after PK than after FS-DSEK (0.5 logMAR versus 0.28 logMAR). The difference was statistically significant even after accounting for pre-existing vision-limiting comorbidities, she said. However, among patients followed for up to three years, those undergoing FS-DSEK continued to improve for up to 
two years. The continuing improvement may result from corneal remodelling. In terms of refraction, FS-DSEK performed significantly better than PK. At 12-months follow-up the mean spherical equivalent was +0.78D in the FS-DSEK group and -1.23D in the PK group (p<0.001). In addition, refractive astigmatism was -1.22D in the FS-DSEK group compared to -2.98D in the PK group. Furthermore, topographic astigmatism was only +1.58D in the FS-DSEK group, compared to +3.67D in the PK group (p<0.001). HIGHER ENDOTHELIAL CELL LOSS Corneal endothelial cell loss was much higher in the FS-DSEK group. At one year, endothelial cell density fell by a mean of 65% to 1,067 cells/mm2 in the FS-DSEK group, compared to a 22% reduction to 2,064 cells/mm2 in the PK group. However, in both groups endothelial cell density remained stable thereafter. The poor endothelial cell density results might be partly attributable to the learning curve involved in this new surgical procedure, she noted. Complications of FS-DSEK included donor tissue location in 10 eyes, primary graft failure in three eyes, folds in the posterior lamellar discs in one eye and elevated intraocular pressure in two eyes. Dr Cheng noted that there are several factors that limited visual acuity for FS-DSEK in the study. They include the use of a thick lamellar graft, which could explain the high incidence of graft folds visible by confocal microscopy. Other factors include applanation strain during the lamellar dissection and poor matching of the host donor interface leading to high-order posterior cornea aberrations. In addition, keratocyte activation in the donor and host corneal tissue resulted in fibrosis and corneal haze, she added. Yanny Cheng: y.y.cheng@lumc.nl