Endothelial Keratoplasty For Herpetic Endothelial Decompensation

Herpetic endothelial decompensation is defined as corneal endothelial dysfunction resulting from herpes simplex virus (HSV), characterised by  immune-mediated inflammation and/or viral replication leading to endothelial cell loss, stromal oedema, and visual impairment. This study assesses visual outcomes, corneal thickness changes, and graft success in eyes undergoing endothelial keratoplasty for corneal decompensation secondary to herpetic endothelial decompensation.

II Ophthalmology Department, Aristotle University of Thessaloniki

Tertiary ophthalmic referral center

Retrospective analysis of 20 eyes with herpetic endothelial decompensation (including keratouveitis, endotheliitis, disciform keratitis, and associated glaucoma) undergoing EK (Descemet membrane endothelial keratoplasty [DMEK] n=9; Descemet stripping endothelial keratoplasty [DSEK] n=11). All cases were diagnosed clinically and intraoperative Acqueous tap was taken. 

 Metrics included preoperative and postoperative logMAR best-corrected visual acuity (BCVA), graft detachment rates, and corneal thickness. Perioperative antiviral prophylaxis  for the first year post op (systemic Valacyclovir and topical agents) was administered. Failures were defined as persistent corneal oedema or decompensation requiring re-intervention. 

Mean follow-up was 12 months (range 17-145).  6 patients had a positive PCR for HSV, 1 for CMV and 2 had a Goldmann-Witmer coefficient (GWC) of <3.  Three grafts failed (15% overall; DMEK 22%, DSEK 9%), showing secondary graft failure. In successful cases (n=17), mean preoperative BCVA improved from 1.34 ± 0.38 logMAR to 0.57 ± 0.21 logMAR (mean gain 0.77 logMAR units). Mean postoperative total corneal thickness was 571 ± 39 μm. No viral recurrences or rejection episodes occurred under prophylaxis. Glaucoma comorbidity was noted in 20% of cases, potentially influencing outcomes.

EK yields an 85% success rate with substantial BCVA improvement in herpetic endothelial decompensation, aligning with or surpassing literature reports. For instance, a series of 17 herpetic eyes post-DMEK showed BCVA gains from 1.16 to 0.62 logMAR but with higher complications (70%, including 12% primary failure and 29% endotheliitis). Smaller cohorts report 100% success with intensive antivirals, mirroring our zero-recurrence rate for the first 12 months. Our findings support EK as a viable first-line therapy over penetrating keratoplasty, emphasising prolonged antiviral prophylaxis to minimise risks in this challenging aetiology. Larger prospective studies are warranted to optimize protocols for comorbid glaucoma cases