Purpose:
To compare 5-year corneal endothelial safety of 3 micro-invasive glaucoma surgery devices (iStent inject, Hydrus Microstent, CyPass Micro-Stent).
Setting:
U.S. multicenter trials.
Design:
Post hoc 5-year analysis from prospective randomized single-masked pivotal trials.
Methods:
Mild to moderate open-angle glaucoma subjects received a MIGS implant with phacoemulsification (implant+phaco) or phaco alone (control). Five-year endpoints comparing the implant and control groups included proportion of eyes with significant endothelial cell loss (ECL) (>30% or ≥30% vs. baseline) and mean endothelial cell density (ECD).
Results:
Comparable proportions of eyes in the iStent inject+phaco and control groups had significant 60-month ECL (9.4% vs. 6.3%, respectively, diff: 3.2%, 95% CI: (-5.0%, 11.3%), P=0.77). Hydrus (20.8% vs. 10.6%, diff: 10.2%, 95% CI: (3.2%, 17.2%), P=0.01) and CyPass (27.2% vs. 10.0%, diff: 17.2%, 95% CI: (5.6%, 28.7%), P=0.02) had more eyes with ECL vs. controls (iStent inject 1.49X, Hydrus 1.96X, CyPass 2.72X vs. controls). Mean ECD over 60 months for iStent inject was indistinguishable vs. control, while greater ECL was observed primarily 3 months postoperative (Hydrus) or accelerated after 2 years (CyPass). No iStent inject or Hydrus subjects developed persistent corneal edema, while 7 CyPass-implanted eyes developed ECL-related complications.
Conclusions:
Through 5 years postoperative, there were no differences in proportion of eyes with significant ECL or mean ECD between the iStent inject and control groups. There was greater 5-year ECL and lower ECD in the Hydrus and CyPass groups versus controls. The Hydrus ECL rate mirrored control after 3 months; the CyPass ECL rate accelerated vs. control.