A Comparison of Lamellar and Penetrating Keratoplasty Outcomes A Registry Study

Objectives: To investigate changing patterns of practice of keratoplasty in Australia, graft survival, visual outcomes, the influence of experience, and the surgeon learning curve for endothelial keratoplasty. Design: Observational, prospective cohort study. Participants: From a long-standing national corneal transplantation register, 13 920 penetrating keratoplasties, 858 deep anterior lamellar keratoplasties (DALKs), and 2287 endokeratoplasties performed between January 1996 and February 2013 were identified. Methods: Kaplan-Meier functions were used to assess graft survival and surgeon experience, the Pearson chi-square test was used to compare visual acuities, and linear regression was used to examine learning curves. Main Outcome Measures: Graft survival. Results: The total number of corneal grafts performed annually is increasing steadily. More DALKs but fewer penetrating grafts are being performed for keratoconus, and more endokeratoplasties but fewer penetrating grafts are being performed for Fuchs' dystrophy and pseudophakic bullous keratopathy. In 2012, 1482 grafts were performed, compared with 955 in 2002, translating to a requirement for 264 extra corneal donors across the country in 2012. Comparing penetrating grafts and DALKs performed for keratoconus over the same era, both graft survival (P < 0.001) and visual outcomes (P < 0.001) were significantly better for penetrating grafts. Survival of endokeratoplasties performed for Fuchs' dystrophy or pseudophakic bullous keratopathy was poorer than survival of penetrating grafts for the same indications over the same era (P < 0.001). Visual outcomes were significantly better for penetrating grafts than for endokeratoplasties performed for Fuchs' dystrophy (P < 0.001), but endokeratoplasties achieved better visual outcomes than penetrating grafts for pseudophakic bullous keratopathy (P < 0.001). Experienced surgeons (>100 registered keratoplasties) achieved significantly better survival of endokeratoplasties (P < 0.001) than surgeons who had performed fewer grafts (< 100 registered keratoplasties). In the hands of experienced, high-volume surgeons, endokeratoplasty failures occurred even after 100 grafts had been performed. Conclusions: More corneal transplants, especially DALKs and endokeratoplasties, are being performed in Australia than ever before. Survival of DALKs and endokeratoplasties is worse than the survival of penetrating grafts performed for the same indications over the same timeframe. Many endokeratoplasties fail early, but the evidence for a surgeon learning curve is unconvincing. Ophthalmology 2014; 121: 979-987 (C) 2014 by the American Academy of Ophthalmology.