Purpose: The aim of this study was to assess the presence of interface fluid at the end of Descemet membrane endothelial keratoplasty (DMEK) and to analyze the impact of this interface fluid on the postoperative graft detachment rate. Methods: At the end of DMEK, after clinically satisfactory graft adhesion, corneas were imaged intraoperatively using optical coherence tomography. The number of interface fluid zones was counted using a custom-written computer script. These zones were classified according to their size (small or large) and their location (graft center or periphery). The rebubbling rate for each patient was recorded. Results: A total of 69 eyes of 69 patients undergoing DMEK surgery were included in this analysis. There was a considerable difference in the number of graft detachments detected between the patients, with a mean number of 2.19 ± 1.94 small central, 2.97 ± 3.11 small peripheral, 0.51 ± 0.90 large central, and 0.81 ± 1.35 large peripheral interface fluid collection sites per eye. A total of 55 patients required no rebubbling, while 10 patients required one, three patients required two, and one patient required three rebubblings. A statistically significant correlation was observed between the number of large peripheral detachments and the rebubbling rate (r = 0.368, P = 0.002). Conclusions: Patients with incomplete graft attachment at the end of DMEK surgery demonstrated increased rebubbling rates. Corneal surgeons might detect these detachments using intraoperative optical coherence tomography and modify their surgical technique to reduce them. Furthermore, curvature matching between the host and the recipient might decrease the occurrence of interface fluid accumulation.
Mackenbrock et al. (2026) studied this question.