Purpose: We compared the hand-guided Moria Carriazo-Barraquer (CBm) microkeratome with the fully automatic SLc microkeratome for Descemet's stripping automated endothelial keratoplasty (DSAEK)-lamella preparation and storage, vis-à-vis accuracy, endothelial cell loss (ECL), and lamellar surface roughness (LSR).
Methods: A total of 18 human corneas were dissected with both the 300 μm CBm multi-use (n = 9) and the 300 µm SLc (n = 9) single-use heads, after which they were incubated for 6 d in a 5% dextran medium. Before preparation (0 h) and 1, 24, and 144 h after dissection, ECL and corneal thickness (CT) were measured by ultrasound pachymetry (USP) and optical coherence tomography (OCT). LSR was assessed by scanning electron microscopy (SEM) and evaluated by three masked observers.
Results: Prior to cutting, CTs did not differ significantly between OCT or USP measurements, with a high correlation between the two modalities (r(2)= 0.94, p < 0.0001). One hour after preparation the anterior lamella showed a significantly higher dissection depth with the CBm (429.4 ± 21.8 µm) than the SLc (311.7 ± 54.8 µm, p = 0.0006), with the variance of the SLc system showing a trend towards higher values (p = 0.07). Anterior and posterior lamellae swelled significantly in the subsequent culture period. Both groups showed a significant ECL 1 h after preparation (p < 0.0001) with no significant difference between the systems (1 h: p = 0.44; CBm: - 9.4%, SLc: -11.7%), which stabilized over 144 h (144 h CBm: -13.9%, 144 h SLc: -10.3%). LSR did not differ significantly between both systems (p = 0.60).
Conclusions: The SLc system agrees more with the designated cutting depth than the CBm. The dissection produced a comparable LSR and a ∼10% ECL independently of the system. Further incubation of the prepared lamellae led to a swelling, but no further ECL.
Keywords: Corneal transplantation; Descemet’s stripping automated endothelial keratoplasty; lamellar graft; organ culture.