Predicting The Amount Of Hyperopic Shift After Dmek

To evaluate changes of total corneal refractive power (TCRP) over a course of 60 months after uneventful Descemet membrane endothelial keratoplasty (DMEK), investigate possible correlations with changes of central corneal thickness (CCT) and present a predictive model for TCRP changes based on expected CCT changes after DMEK.

Department of Ophthalmology, Philipps University of Marburg, Germany

Sixty pseudophakic eyes, which underwent DMEK without intraoperative and postoperative complications, were included. The TCRP mean keratometry values (Km) at 2, 4, 6 and 8mm zones and the CCT were evaluated. The measurements were performed preoperatively and at 3, 6, 12, 24 and 60 months postoperatively.

The TCRP Km showed in all zones a significant decrease (hyperopic shift) at 3 months (P<0.001), followed by a continuous and significant increase (myopic shift) up to 60 months (P<0.001). A decrease of CCT was observed up to 6 months (P<0.001), followed by a continuous increase thereafter (P<0.001). A significant positive correlation between the decrease of CCT and the decrease of TCRP Km at 2, 4, 6 and 8mm zones (preoperative – 3-month values; P<0.001; r ranging from 0.553 to 0590) was found. For the optically relevant 4mm zone the equation of the linear regression model for the expected TCRP Km decrease (y) in relation with the decrease of central corneal thickness (preoperative – 3-month postop) (x) was y= 0.0115x + 0.1264.

Our study quantified the hyperopic shift observed shortly after DMEK. To date, in cases of cataract surgery prior to DMEK or in triple DMEK procedures the target refraction would be set to low myopia based on clinical experience rather than measurable data. Historical data of CCT prior to the development of a clinically significant corneal edema would enable preoperatively an estimation of the postoperative CCT of the dehydrated cornea and thus an estimation of the expected decrease of CCT after DMEK. The postoperative TCRP reduction suggested from our regression analysis model could enable a better prediction of the postoperative hyperopic shift and a more accurate selection of the intraocular lens power and target refraction.