Abstract
Purpose:
Ultrathin Descemet stripping automated endothelial keratoplasty (UT-DSAEK) using a microkeratome-assisted double-pass technique has been shown to be a consistently reproducible technique for achieving thin grafts. Previous reports have noted improved visual acuity in thinner grafts. Here, we present clinical results of the double-pass technique highlighting graft thickness measured by anterior segment OCT (AS-OCT), best-corrected visual acuity (BCVA), and endothelial cell counts. We hypothesized that in UT-DSAEK, thinner grafts result in better visual outcomes.
Methods:
A retrospective case series of 36 eyes in 32 patients who underwent UT-DSAEK by two surgeons (RD and MT) with or without simultaneous phacoemulsification. Patients with macular pathology, advanced glaucoma, and pre-existing non-endothelial corneal pathology were excluded. Central graft thickness was measured postoperatively with AS-OCT. Graft endothelial cell count was measured by specular microscopy 6 months postoperatively. BCVA was recorded at 1, 3, and 6 months postoperatively. Patients with macular pathology, posterior capsule opacification, and pre-existing non-endothelial corneal pathology were excluded.
Results:
At six month follow up, 28 eyes in 25 patients were reviewed. A total of 8 patients were excluded. Two patients had age-related macular degeneration, 2 had advanced glaucoma, one had persistent postoperative cystoid macular edema, one had unrelated anterior corneal scarring, one had cortical visual impairment from prior trauma, and one had a history of Vogt-Koyanagi-Harada disease with involvement of the macula. The median graft thickness was 56 µm (range 28-158 µm) as measured by AS-OCT. Median BCVA at 6-month follow up (range 6 to 8 months) was 0.19 ± 0.16 logMAR. All but one eye reached BCVA better than 20/50, with 32% equal to or better than 20/25 and 68% 20/30 or better. Median graft endothelial cell count at 6 months was 1852 ± 362. A plot of central graft thickness against BCVA by logMAR did not show a statistically significant correlation (r2=0.02). Grafts thinner than the median (56 µm) did not have better BCVA at 6 months than grafts thicker than the median (p=0.91).
Conclusions:
Use of the double-pass technique results in excellent visual outcomes and preserved endothelial cell counts. In contrast to some previous reports, our results show that in-vivo graft thickness does not directly correlate to BCVA.