To study the outcome of penetrating keratoplasty (PK) in eyes with a glaucoma drainage device (GDD).

Retrospective case-controlled study.

We reviewed all patients who underwent PK from December 1986 to September 2002 at the University of California, Davis (n = 1,974). We identified 33 patients (40 grafts) who were treated with a GDD and followed up for 6 months or more after PK. Graft survival and glaucoma control were compared with grafts in patients without glaucoma (n = 40) and patients with medically controlled glaucoma (n = 17). Kaplan-Meier survival analysis, log rank test, repeated-measures analysis of variance (ANOVA), Fisher exact test, and χ² were used in group comparisons. Multivariate analysis was performed using the Cox proportional hazards model.

The percentages of clear grafts in the glaucoma drainage device group were 58.5% and 25.8% at 1 and 2 years, respectively. At these time points, glaucoma was controlled in 74.0% and 63.1% of the eyes, respectively. Both medically controlled glaucoma patients and nonglaucomatous patients had higher graft survival percentages at comparable time points. The presence of a GDD was an important factor influencing graft survival (Hazard ratio = 6.8).

A GDD implant is an independent risk factor for graft failure. Although these devices are effective in controlling intraocular pressure (IOP) in the majority of eyes in the presence of PK, corneal graft clarity is often compromised.