Purpose
The long-term outcome of corneal grafts (> 10 years) is poorly understood because of a lack of subjects with a sufficient follow-up time. A retrospective cohort study of corneal graft recipients was set up in a single French center. Our aim was to describe local care practices, patient demographics, as well as keratoplasty outcome and graft survival.<br /> <br />
Methods
Patients who underwent at least one corneal graft between 1983 and 2013 in the Besançon University hospital department of Ophthalmology were included. Data relating to all aspects of usual care (recipient, graft, donor, transplant surgical technique) have been prospectively collected since 1983 into single database. Postoperative evaluations at 1, 3, 6, 9, and 12 months and then yearly, were also included.<br />
Results
The 1133 included patients received 1563 grafts. A single graft was performed for 813 subjects (72%), while 123 (11%) received multiple unilateral grafts (up to 4). There were 148 (13%) a single bilateral graft, and 4% of the patients had bilateral multiple grafts with at least one graft per eye. Main graft indications were postoperative bullous keratopathy (35%), keratoconus (26.5%), Fuchs’ dystrophy (8%), non-herpetic corneal infections (5.5%), and graft failure (5%). The surgical procedure was a penetrating keratoplasty in 96,6% of the cases. As of the end of 2013, the mean hindsight was 190 months (range 2 - 363). The number of patients who were still followed-up with a clear graft after ten years was 246, 128 after 15 years, 21 after 20 years, and 3 after 25 years. Among the first grafts, 406 patients (36%) experienced a graft failure.<br />
Conclusions
The distribution of graft indications is representative of the French data (source : Agence of Biomédecine) but quite different when compared to other cohorts from country to country. The creation of this cohort will allow to perform further analyses about factors associated with corneal graft long-term survival. Good knowledge of the distribution will lead to an improved interpretation of the data.<br />