Femto-second laser (FSL) techniques and consistency in corneal surgery: experimental study.

OBJECTIVE: To investigate the feasibility, effectiveness, and reproducibility of femtosecond laser (FSL) technology in different corneal cuts.

DESIGN: Experimental study at the Brandon Regional Health Centre.

PARTICIPANTS: Twenty two Human Cadaver eyes donated for research and supplied by the eye bank of Manitoba university.

METHODS: In this experimental study, the effectiveness and reproducibility of FSL were investigated in different corneal cuts. The corneas were divided into 5 groups: group I for penetrating keratoplasty; group II for anterior lamellar keratoplasty; group III for corneal pockets for inlays; group IV for intracorneal ring tunnels; and group V for corneal cross-linking pockets. Our primary endpoint was the repeatability of planned versus obtained thickness and diameter of the cuts with FSL. A secondary endpoint was the effectiveness and ease of dissection, rated from 0-4, with 4 being the easiest and 0 the most difficult to dissect.

RESULTS: There was a high correlation between planned and obtained thickness (r = 0.997, p < 0.001) and diameter (r = 0.998, p < 0.001). Both were statistically significant. Linear regression analyses showed that the obtained thickness of corneal cuts in micrometers was significantly predicted by the planned thickness of corneal cuts in micrometers (β = 0.996, t = 56.47, p < 0.001). Likewise, the obtained diameter of corneal cuts in millimeters was significantly predicted by the planned diameter of corneal cuts in millimeters (β = 0.971, t = 70.85, p < 0.001). The surgeons gave 15 out of a total of 22 corneas (68.2%) the maximum score for ease of dissection (Grade 4), where no adhesions were found and the dissection was smooth and steady. Five corneas out of a total of 22 (22.7%) were given a score of 3 of 4 where minor adhesions were found. Two corneas out of a total of 22 (9.1%) were given a score of 2 of 4 where >1 adhesion was found, with some difficulty in separating the cut cornea from the bed. No corneas were given grade 1 or grade 0. No significant differences were obtained for the ease of dissection among the 5 surgery types using Kruskal-Wallis H test (H [4] = 4.971, p > 0.05).

CONCLUSION: The measured corneal-cut geometry correlated well with laser settings in a variety of full- and partial-thickness FSL corneal patterns, including different depths and diameters. This reproducible efficacy and measurement accuracy of the planned versus obtained cuts could have a favourable result on a variety of corneal surgeries. It also yields a favourable ease of dissection of the cut lenticule from the stromal bed.