Exploring the precision of femtosecond laser-assisted descemetorhexis in Descemet membrane endothelial keratoplasty

OBJECTIVE: Descemet membrane endothelial keratoplasty (DMEK) remains a challenging technique. We compare the precision of femtosecond laser-assisted DMEK to manual DMEK. METHODS AND ANALYSIS: A manual descemetorhexis (DR) of 8 mm diameter was compared with a femtosecond laser-assisted DR of the same diameter (femto-DR) in 22 pseudophakic patients requiring DMEK. We used OCT images with a centred xy-diagram to measure the postoperative precision of the DR and the amount of endothelial denuded area. Endothelial cell loss (ECL) and best corrected visual acuity were measured 3 months after surgery. RESULTS: In the manual group, the median error of the DR was 7% (range 3%–16%) in the x-diameter and 8% (range 2%–17%) in the y-diameter. In the femto group, the median error in the respective x and y-diameters was 1% (range 0.4%–3%) and 1% (range 0.006%–2.5%), smaller than in the manual group (p=0.001). Endothelial denuded areas were larger in the manual group (11.6 mm(2), range 7.6–18 mm(2)) than in the femto group (2.5 mm(2), range 1.2–5.9 mm(2)) (p<0.001). The ECL was 21% (range 5%–78%) in the manual DR and 17% (range 6%–38%) in the femto-DR group (p=0.351). The median visual acuity increased from 0.4 logMAR (range 0.6–0.4 logMAR) in both groups to 0.1 logMAR (range 0.4–0 logMAR) in the manual group and to 0.1 logMAR (range 0.3–0 logMAR) in the femto group (p=0.461). Three rebubblings were required in the manual group, whereas the femto group required only one. CONCLUSION: The higher precision of the femto-DR bears the potential to improve DMEK surgery.