Heads-up Descemet Membrane Endothelial Keratoplasty Surgery: Feasibility, Surgical Duration, Complication Rates, and Comparison With a Conventional Microscope.

Purpose: To assess the feasibility of Descemet membrane endothelial keratoplasty (DMEK) surgery using the glasses-assisted 3-dimensional (3D) display system NGENUITY (Alcon Laboratories, Fort Worth, TX) and to compare with standard DMEK surgery by using a conventional operating microscope.

Methods: Twelve consecutive cases of DMEK surgery were performed using the glasses-assisted 3D display system NGENUITY (Alcon) and matched with similar cases performed by using the OPMI-Lumera 700 surgical microscope (Carl Zeiss Meditec, Jena, Germany) in this prospective cross-sectional study realized at the Rothschild Foundation, Paris, France. DMEK graft preparation time, graft unfolding time, time to perform the descemetorhexis (DM), and overall surgical time were recorded. Best corrected visual acuity, endothelial cell density of the donor tissue measured by specular microscopy, and the recipient's central corneal thickness were recorded preoperatively and again at 1 and 3 months postoperatively.

Results: In the 3D group, time to perform the graft preparation, DM, and global surgical time were significantly higher in the 3D group than in the conventional groups (17 minutes 36 seconds ± 3 minutes 13 seconds and 14 minutes 50 seconds ± 1 minutes 28 seconds, P = 0.04; 6 minutes 18 seconds ± 2 minutes 30 seconds vs. 3 minutes 50 seconds ± 2 minutes 8 seconds, P = 0.03; and 31 minutes 30 seconds ± 6 minutes 10 seconds vs. 24 minutes 10 seconds ± 3 minutes 26 seconds, respectively, P = 0.05). The time to unfold the graft was higher, although not significantly so, in the 3D group compared with the conventional group (5 minutes 15 seconds ± 2 minutes 40 seconds vs. 4 minutes 40 seconds ± 2 minutes 25 seconds, P = 0.89). Mean central corneal thickness decreased about 168 ± 82 μm at 3 months after surgery in the 3D group and about 228 ± 123 μm in the conventional group without significant difference (P = 0.60). Best corrected visual acuity in the 3D group was 0.33 ± 0.32 LogMAR at 3 months after surgery with no significant differences found when compared with conventional surgery. The percent endothelial cell loss from preoperative at 3 months was 35% in the 3D group and 26% in the conventional group, respectively, with no significant difference (P > 0.05).

Conclusions: Performing DMEK surgery using a 3D display system is feasible; however, it is more challenging and the total surgical time is longer. This method would certainly be useful for instructional courses.