Précis: GDD implantation in the anterior chamber is associated with progressive corneal endothelial cell loss over time, with evidence suggesting a more pronounced effect in Asian compared to Caucasian eyes. Purpose: Corneal endothelial cell damage occurs with various surgical procedures. This systematic review and meta-analysis aims to investigate the impact of Ahmed or Baerveldt glaucoma drainage device (GDD) implantation on corneal endothelial cell density (ECD). Methods: Database search involved MEDLINE, EMBASE, Scopus, NIH Clinical Trials (http://www.clinicaltrials.gov/), and Google Scholar (PROSPERO ID:CRD42024521140). The included studies focused on standalone GDD implantation in eyes without pre-existing corneal endothelial pathologies and with a minimum follow-up of 12 months. The quality of the studies was assessed using the Cochrane Risk of Bias Tool (ROBINS-I). Results: Fourteen studies were included in the quantitative synthesis, comprising 572 eyes. ECD after anterior chamber GDD implantation decreased by 204.7 cells/mm 2 (95%CI:149.2,260.1, I 2 =61.04%) at 12 months. Subgroup analysis revealed a greater ECD loss in predominantly Asian (MD,225.5cells/mm 2 ; 95% CI:151.1,300.0, I 2 =67.4%) compared to predominantly Caucasian population (MD,155.3cells/mm 2 ; 95%CI:94.5,216.2, I 2 =0%) and no difference between Ahmed and Baerveldt GDD. Sensitivity analysis showed ECD loss of 227.1 cells/mm 2 (95%CI:182.6,271.6, I 2 =32.6%). At 24 months after anterior chamber GDD implantation, ECD decreased by 287.8 cells/mm 2 (95%CI: 233.0, 342.7, I 2 =53.4%). ECD loss at 12 months after sulcus or pars plana placement varied between 12-629.6cells/mm 2 , but significant heterogeneity in studies precluded meta-analysis of available data. Conclusion: Overall, implantation of a GDD leads to clinically relevant corneal endothelial cell loss. ECD loss may be more pronounced in Asians compared to Caucasians. No difference in ECD loss was observed between the two GDD types GDD. The existing evidence is characterized by heterogeneity, and further high-quality randomized controlled trials are necessary to establish definitive estimates.
Dastiridou et al. (Tue,) studied this question.