Safety of Concurrent Boston Type I Keratoprosthesis and Glaucoma Drainage Device Implantation.

ABSTRACT

PURPOSE: To evaluate the safety of concurrent Boston type I keratoprosthesis (KPro) and glaucoma drainage device (GDD) implantation. DESIGN: Retrospective comparative study of a consecutive cohort of patients. SUBJECTS: Patients who underwent KPro implantation by a single surgeon (A.J.A.) with or without 1 concurrent Ahmed GDD (New World Medical, Inc., Rancho Cucamonga, CA) implantation between January 1, 2005, and January 31, 2015, were included. Patients with fewer than 3 months of follow-up or a history of previous KPro implantation were excluded. METHODS: Preoperative, operative, and postoperative data were collected for each procedure. All comparisons were made between KPro procedures performed with or without concurrent GDD implantation. The Fisher exact test (2-tailed) was used to compare proportions, Student t test and Wilcoxon rank-sum test were used to compare means, and the log-rank test was used to compare time-to-outcome events. MAIN OUTCOME MEASURES: The primary outcome was frequency of the composite variable, that is, any serious vision-threatening postoperative complication, which included sterile vitreitis, endophthalmitis, hypotony maculopathy, suprachoroidal hemorrhage, retinal detachment, stromal necrosis, and infectious keratitis. Secondary outcomes included intraocular pressure control, worsening of visual acuity, cystoid macular edema, retroprosthetic membrane formation, persistent epithelial defect formation, GDD exposure, and KPro removal. RESULTS: One hundred thirty-seven KPro procedures were performed in 129 patients 91 (66.4%) KPro alone and 46 (33.6%) KPro plus GDD. There were no statistically significant differences between the 2 groups in terms of the incidence of vision-threatening postoperative complications. None of the 46 GDDs placed at the same time as the KPro became exposed during an average follow-up of 44 months. CONCLUSIONS: Compared with KPro alone, GDD placement combined with KPro was not associated with increased postoperative complications.