Keratoprosthesis surgery for end-stage corneal blindness in asian eyes.

PURPOSE To establish a multidisciplinary surgical program for osteo-odonto-keratoprosthesis (OOKP) surgery in Asia and to evaluate efficacy and preliminary safety of this keratoprosthesis in end-stage corneal and ocular surface disease. DESIGN Prospective noncomparative case series. PARTICIPANTS Sixteen adults of Asian ethnic origin, bilaterally blind with end-stage corneal blindness from Stevens-Johnson syndrome, or severe chemical or thermal burns. METHODS Osteo-odonto-keratoprosthesis surgery involves 2 procedures-in stage 1, an autologous canine tooth is removed, modified to receive an optical polymethyl methacrylate cylinder, and implanted into the cheek. The ocular surface is denuded and replaced with full-thickness buccal mucosa. Stage 2 surgery, performed 2 to 4 months later, involves retrieval of the tooth-cylinder complex and implanting it into the cornea, after reflection of the buccal mucosal flap, corneal trephination, iris and lens removal, and anterior vitrectomy. Concurrent glaucoma and vitreoretinal procedures are also performed at either stage, as required. MAIN OUTCOME MEASURES Visual acuity (VA), field of vision, anatomical integrity and stability, and ocular and oral complications related or unrelated to the OOKP device. RESULTS Osteo-odonto-keratoprosthesis surgery was performed on 15 patients, with a mean follow-up of 19.1 months (range, 5-31). Intraoperative complications included expulsive hemorrhage (keratoprosthesis device not implanted), tooth fracture (n = 1), oronasal fistula (n = 1), and mild inferior optic tilt (n = 1). Anatomical stability and keratoprosthesis retention has been maintained in all eyes, with no dislocation, extrusion, retroprosthetic membrane formation, or keratoprosthesis-related infection. Other complications not directly related to device insertion included retinal detachment (RD) related to silicone oil removal (n = 1) and endophthalmitis related to endoscopic cyclophotocoagulation performed 1 year after OOKP surgery (n = 1). Eleven patients (73.3%) attained a stable best spectacle-corrected VA of at least 20/40 or better, whereas 9 (60%) attained stable 20/20 vision. Four patients achieved their best visual potential, ranging from 20/100 to counting fingers vision, related to preexisting glaucomatous optic neuropathy or previous RD. CONCLUSIONS Establishment of our OOKP program suggests that OOKP surgery has the potential to restore good vision to the most severe cases of corneal blindness in an Asian setting, with minimal device-related complications. Longer follow-up of these cases is currently underway.