Descemetic Versus Pre-Descemetic Lamellar Keratoplasty: Clinical and Confocal Study

Purpose: To compare the clinical findings and confocal microscopic features of the lamellar interface after 2 types of deep anterior lamellar keratoplasty (DALK): Descemetic with total stromal resection versus pre-Descemetic with deep stromal dissection. Methods: Twenty eyes of 20 patients who had corneal disease with healthy endothelium were treated by DALK using the air technique. Baring of Descemet membrane (DM) was achieved for 12 eyes (Descemetic group). A fine stromal layer was left in 8 eyes (pre-Descemetic group). Visual acuity, interface clarity, corneal topography, confocal microscopy, and endothelial cell count were analyzed. Results: DM microperforation occurred in 2 eyes (10%). No patient required conversion to penetrating keratoplasty (PKP) intraoperatively. Presumed stromal rejection occurred in 1 eye that was excluded from the study. In the Descemetic group, 90.9% achieved final best-corrected visual acuity (BCVA) of 20/30 or better; in the pre-Descemetic group, 75% achieved final BCVA of 20/30 or better. No statistically significant difference in mean visual outcomes was found between groups. The reflectivity of activated keratocytes at the interface was less in the Descemetic than that in the pre-Descemetic group. Ten to 12 weeks after pre-Descemetic DALK and 4 to 6 weeks after Descemetic DALK, keratocyte morphology and reflectivity had returned to normal. Mean combined topographic astigmatism was 2.17 +/- 0.75 diopters (D) at 6 months (sutures out). Conclusions: The depth of the lamellar bed, smoothness, and healing process at the interface are the keys to optimal visual acuity. Although dissection to bare DM is more difficult, keratocyte activation and interface haze were less, but differences in final visual acuity were not significant.