PURPOSE: To describe and report outcomes of our single-pass microkeratome technique for preparation of ultrathin (UT, ≤100 μm) and nanothin (NT, ≤50 μm) Descemet stripping automated endothelial keratoplasty (DSAEK) grafts.
METHODS: To prepare NT-DSAEK grafts, a pachymetry nomogram specific to each technician and individual microkeratome head was developed based on accumulated precut and postcut pachymetry data from previous DSAEK grafts. Mean graft thickness as well as precut and postcut endothelial cell counts (ECCs) of NT-DSAEK, UT-DSAEK, and Descemet membrane endothelial keratoplasty (DMEK) grafts between July 2015 and July 2017 were calculated and compared statistically. Endothelial cell loss was evaluated using calcein AM stains and ImageJ analysis. Postcut graft thickness and rates of perforation/tissue loss for NT-DSAEK grafts between May and July 2017 were calculated to determine overall graft preparation success rates.
RESULTS: Mean postcut graft thickness for all grafts within the NT range was 41.0 ± 6.4 μm (range 26-50 μm). Mean ECC did not differ between NT-DSAEK, UT-DSAEK, and DMEK grafts (P = 0.759 and 0.633, respectively). The overall tissue loss rate from attempted NT-DSAEK was 4.8%. Excluding cases of perforation, the chance of achieving NT thickness was 60% and within the traditional UT range was 100%.
CONCLUSIONS: We propose the term "NT-DSAEK" for grafts ≤50 μm. The described nomogram allows for standardized creation of NT grafts with a low tissue loss rate. This technique is safe and does not result in significant ECC loss compared with UT-DSAEK and DMEK grafts. Further studies are necessary to corroborate the postsurgical results of NT grafts.