Effectiveness of curvilinear approach in dissection of Descemet's membrane: first 500 cases - factors influencing graft preparation

Purpose To confirm the reproducibility of manual graft preparation using curvilinear forceps and evaluate the incidence and type of structural abnormalities of Descemet's membrane (DM) preventing successful grafts preparation. Methods Five hundred corneo-scleral buttons were prepared. Factors such as endothelial cell number before preparation, donor age, post-mortem time, time in culture, pigmentation of the trabecular meshwork and preparation characteristics of the fellow eye were analysed. According to the preparation characteristics, three groups were formed: A, uncomplicated; B, complicated preparation with stripping from the contralateral side; and C, failure of preparation. Three failed grafts were examined by transmission electron microscopy (TEM). Results Using curvilinear forceps, manual separation of DM was achieved without any adverse effects in 457 of 500 corneas (91.4%). In 32 corneas (6.4%) with micro-tears during preparation, stripping from the opposite side was possible. However, 11 of the 500 corneas (2.2%) showed extremely strong adhesion leading to multiple tears of DM and preventing successful preparation of the graft. Endothelial cell number, donor age, post-mortem time, time in culture and pigmentation of the trabecular meshwork showed no significant correlation with failure to successfully obtain a DM graft. Complicated graft preparations of one eye showed a highly significant correlation with complicated graft preparations in the fellow eye. TEM analysis of failed grafts showed abnormal invasion of stromal parts into the DM, cell accumulation and pigmentation in the DM plane. Conclusion Using curvilinear forceps for dissecting of the graft shows valid and reproducible results in the vast majority (97.8%) of donor corneas.