Purpose: To conduct a systematic review and meta-analysis on the efficacy of thin and ultrathin Descemet stripping automated endothelial keratoplasty (T-DSAEK and UT-DSAEK, with graft thickness <130 and <100 µm, respectively), depending on graft thickness.
Design: Systematic review and meta-analysis.
Method: PubMed, Cochrane Library, Embase, ClinicalTrials.gov, and ScienceDirect databases were searched until October 1, 2021. We computed random-effect meta-analysis on postoperative outcomes of T/UT-DSAEK, stratified by graft thickness (<80 μm, 80-100 μm, and 100-130 μm). The main postoperative outcome was visual acuity (logarithm of the minimum angle of resolution [logMAR]). Secondary outcomes were pachymetry (μm), endothelial cell count (cell/mm2), spherical equivalent (diopter [D]), rebubbling rate (%), and rejection rate (%). Meta-regressions compared postoperative outcomes depending on graft thickness and search for putative confusion factors.
Results: We included 47 articles for a total of 2141 eyes of 2040 patients. T/UT-DSAEK globally improved visual acuity (effect size = -0.38 logMAR [95% confidence interval {CI} -0.46 to -0.30 logMAR]), without difference depending on graft thickness. Overall, pachymetry improved (-60.6 µm [95% CI -101 to -19.7 µm]), endothelial cell count decreased (-1039 cells/mm2 [95% CI -1209 to -868 cells/mm2), spherical equivalent resulted in a hyperopic shift (0.74 D [95% CI -0.50 to 1.97 D), the graft rejection rate was 0.2% (95% CI -0.1% to 0.4%), and the rebubbling rate was 8.7% (95% CI 6.8%-10.5%). Grafts >100 μm induced a hyperopic shift. Metaregressions did not demonstrate differences between the 3 groups (<80 μm, 80-100 μm, or 100-130 μm) in any outcomes.
Conclusion: All T/UT-DSAEK thickness groups provided similar visual acuity, pachymetry, endothelial cell count, rejection rate, and rebubbling rate regardless of graft thickness. A hyperopic shift was induced by grafts >100 μm.
Copyright © 2022 Elsevier Inc. All rights reserved.