Nomogram to Predict Graft Thickness in Descemet Stripping Automated Endothelial Keratoplasty: An Eye Bank Study

PURPOSE:The purpose of this study was to develop a nomogram to predict postcut thickness of corneal grafts prepared at an eye bank for Descemet stripping automated endothelial keratoplasty (DSAEK). METHODS:Retrospective chart review was performed of DSAEK graft preparations by 3 experienced technicians from April 2012 to May 2017 at the Eye Bank of Canada—Ontario Division. Variables collected included the followingdonor demographics, death-to-preservation time, death-to-processing time, precut tissue thickness, postcut tissue thickness, microkeratome head size, endothelial cell count, cut technician, and rate of perforation. Linear regression models were generated for each microkeratome head size (300 and 350 μm). RESULTS:A total of 780 grafts were processed during the study period. Twelve preparation attempts resulted in perforation (1.5%) and were excluded. Mean precut tissue thickness was 510 ± 49 μm (range363–670 μm). Mean postcut tissue thickness was 114 ± 22 μm (range57–193 μm). Seventy-nine percent (608/768) of grafts were ≤130 μm. The linear regression models included precut thickness and donor age, which were able to predict the thickness to within 25 μm 80% of the time. CONCLUSIONS:We report a nomogram to predict thickness of DSAEK corneal grafts prepared in an eye bank setting, which was accurate to within 25 μm 80% of the time. Other eye banks could consider performing similar analyses.