Eye bank-prepared femtosecond laser-assisted automated descemet membrane endothelial grafts

Purpose: The aim of this study was to investigate the use of a femtosecond laser (FL) in the eye bank preparation of corneas for Descemet membrane (DM) automated endothelial keratoplasty (fDMAEK) and to compare endothelial cell death in graft preparations between fDMAEK, Descemet stripping endothelial keratoplasty (DSEK), and DM endothelial keratoplasty (DMEK). Methods: Twenty cadaveric tissues were used to test the fDMAEK method. A 9.0-mm-diameter lamellar incision was made using the FL with a 6.0-mm perpendicular anterior ring cut that enabled a stromal rim by acting as a venting incision for bubble expansion. DM was pneumodissected off the central 6.0 mm of the tissue. The fDMAEK grafts were trephined and stained with a viability dye, calcein AM. The entire stained endothelial surface was digitally captured and the endothelial cell loss (ECL) was calculated using trainable segmentation software. For comparison, a series of 6 DSEK grafts and 8 DMEK grafts were created and analyzed. Results: Six of 20 tissues (30%) were lost during fDMAEK preparation. In the 14 successful tissues, the average ECL was 30.4% [95% confidence interval (CI), 25.3-35.6] compared with 21.1% (95% CI, 13.2-28.9, P = 0.09) in the 6 DSEK grafts and 22.5% (95% CI, 18.0-27.0, P = 0.04) in the 8 DMEK grafts. Conclusions: FLs are useful in preparing DMAEK tissue at the eye bank and may promote predictable and precise big bubbles and stromal rims. The fDMAEK preparation success improved with experience and laser adjustments. In fDMAEK, the ECL is higher than was previously reported in DMEK and DSEK, likely due to greater tissue manipulation, although not significantly higher than DSEK controls.