Abstract
Purpose:
To develop and study femtosecond laser techniques and settings that facilitate deep stromal resections for DALK.
Methods:
Three approches in human eye bank eyes were used to create smooth beds for DALK using a 150 kHz FS laser: Procedure 1) Debulking (N=12) removing 300 µm @ 9.0 mm (3x3 spot/line SL, 0.6 µJ) followed by 100 µm resection (2x2 or 3x3 SL, 0.6 µJ); Procedure 2) Deep single or double raster pass (N=3) @ 7.0 mm diameter, 8x8 SL, 0.3 µJ); Procedure 3) Multiple raster passes using two spiral and then 2 raster patterns using the same patient interface, followed by a raster pattern with a side cut (N= 8) forcing the FS-produced gases to dissect Descemet’s membrane (DM) away from the stroma (7.0 mm diameter, 8x8 SL, 900 side cut 1.6 µJ, 2.1 µJ raster). The corneas were then fixed in glutaraldehye and prepped for LM, SEM, and TEM.
Results:
Procedure 1 created a bed as smooth as more superficial FS laser dissections, but left a significant thickness of posterior stroma in place. Procedure 2 removed more stroma than #1, but it was not as smooth and still left significant stroma behind. Procedure 3 removed the most stroma and in two cases left an almost bare DM behind, but the edges of the resection were rougher than the other procedures. One case of perforation of DM occurred. There was no morphologic damage to the endothelium.
Conclusions:
FS laser-created deep lamellar beds are not normally as smooth as < 120µm flaps, but the debulking and single or double raster passes can create smooth beds. The multiple raster pass technique, dubbed the “Little Bubble Technique” was able to bare DM and remove the most stroma, but has the risk of perforation of DM similar to the “Big Bubble Technique”. It appears to have a more predictable depth resection and does not require mechanical air injection. Modifications of SL, energy settings, side cut angles and side cut energy may be more predictable in baring DM.
Keywords: 578 laser •
479 cornea: clinical science •
678 refractive surgery