Interface quality of different corneal lamellar-cut depths for femtosecond laser-assisted lamellar anterior keratoplasty

Purpose To evaluate the interface quality of different corneal lamellar-cut depths with the femtosecond laser and determine a feasible range of depth for femtosecond laser-assisted lamellar anterior keratoplasty. Setting Casey Eye Institute, Portland, Oregon, USA. Design Experimental study. Methods Full lamellar cuts were made on 20 deepithelialized human cadaver corneas using the femtosecond laser. The cut depth was 17% to 21% (100 μm), 31%, 35%, 38% to 40%, and 45% to 48% of the central stromal thickness. Scanning electron microscopy images of the cap and bed surfaces were subjectively graded for ridge and roughness using a scale of 1 to 5 (1 = best). The graft-host match was evaluated by photography and optical coherence tomography in a simulated procedure. Results The ridge score was correlated with the cut depth (P =.0078, R = 0.58) and better correlated with the percentage cut depth (P =.00024, R = 0.73). The shallowest cuts had the fewest ridges (score 1.25). The 31% cut depth produced significantly fewer ridges (score 2.15) than deeper cuts. The roughness score ranged from 2.19 to 3.08 for various depths. A simulated procedure using a 100 μm host cut and a 177 μm (31%) graft had a smooth interface and flush anterior junction using an inverted side-cut design. Conclusions The femtosecond laser produced more ridges in deeper lamellar cuts. A depth setting of 31% stromal thickness might produce adequate surface quality for femtosecond laser-assisted lamellar anterior keratoplasty. The inverted side-cut design produced good edge apposition even when the graft was thicker than the host lamellar-cut depth. Financial Disclosure Proprietary or commercial disclosures are listed after the references.