Comparison of 12-month anatomic and functional results between Z6 femtosecond laser-assisted and manual trephination in deep anterior lamellar keratoplasty for advanced keratoconus.

The management of severe keratoconus requires corneal transplantation, for which the gold standard is deep anterior lamellar keratoplasty (DALK), preserving the healthy Descemet's membrane and endothelium. The safety and reproducibility of corneal cuts have been improved by the evolution of femtosecond lasers in refractive surgery, and femtosecond laser in DALK would seem to provide the same advantages over the manual method. In our retrospective study, we compare functional and anatomical results of femtosecond-assisted DALK versus manual trephination DALK in patients with keratoconus in stage 4 of the Krumeich classification. It is a retrospective study including all patients with stage 4 keratoconus who underwent femtosecond laser-assisted DALK between November 2012 and November 2015 in Nantes university medical center. We compared those patients to a group of patients who underwent manual DALK in the same period, paired by age and maximal keratometry. We assessed visual acuity, pachymetry, endothelial cell density (specular microscopy), and keratometry before surgery and at 4, 8 and 12 months of follow-up. Laser settings and intraoperative complications were recorded. Nineteen patients underwent surgery by femtosecond-assisted DALK, 6 women and 12 men with average age 30.2±10.8 years at transplantation. They were paired with a group of 17 patients who underwent manual DALK in order to compare results. Before surgery, mean visual acuity in the femtosecond group was 0.90 logMAR versus 0.89 logMAR in the manual group, showing no statistically significant difference (P=0.96). Both groups were similar in terms of preoperative age, mean keratometry, pachymetry and endothelial cell density. Average visual acuity post-surgery was 0.27; 0.26; and 0.14 logMAR for femtosecond DALK versus 0.27; 0.17 et 0.25 for manual DALK at 4, 8 and 12 months follow-up respectively, showing no statistically significant difference. After surgery, at 4, 8 and 12 months, mean pachymetry was similar in both groups, and average endothelial cell density was 2390 cells/mm2 for femto DALK versus 2531 cells/mm2 for manual DALK at 12 months of follow-up, showing no statistically significant difference (P=0.5726). The rate of Descemet's membrane microperforations during the procedure was low and similar for both groups. Our study allows for a 12-month follow-up, with assessment of visual recovery, anatomic result and endothelial safety in a series of 19 femtosecond laser-assisted DALK with no statistical significant difference versus the manual trephination group. Femtosecond laser allows for increased reproducibility of the DALK procedure without reducing adverse effects during surgery. Femtosecond laser seems to improve the technique of the DALK procedure, and future developments could improve the reproducibility of DALK even further. A medical economics study would be necessary to determine the cost-effectiveness of femtosecond laser-assisted DALK.