Risk factors for graft failure after penetrating keratoplasty: 5-year follow-up from the corneal transplant epidemiological study.

Purpose: To evaluate corneal graft survival over a 5-year period and to investigate whether factors related to the donor, eye bank practices, the recipient, surgery, and postoperative course influenced the outcome.

Methods: Nine hundred ninety-eight patients were randomly selected and monitored in the subsequent 3 years from a cohort of 4500 recipients who underwent penetrating keratoplasty between 2001 and 2004. Cox univariate regression analysis was used to select variables to be included in a multivariate Cox proportional hazards model with a backward selection procedure to identify potential risk factors for graft failure. Graft survival curves were obtained from Kaplan-Meier estimates.

Results: Ectasia/thinning was the most common indication (49.1%), followed by regraft (16.1%) and pseudophakic corneal edema (PCE) (9.4%). The overall rate of graft failure was 10.7% with 6 cases of primary graft failure. Adverse reactions and complications (other than graft failure) were reported in 2.7% of patients in the first postoperative week and in 22.8% during the full follow-up period. The probability of 5-year survival was 83.0%, best in eyes with ectasia/thinning (96.0%) and less favorable in PCE (67.0%) and regraft (64.0%). Multivariate analyses showed the following variables to be linked to an increased risk of graft failure: regraft for any reason, all clinical indication except PCE, history of ocular inflammation/infection, pseudophakic/aphakic eye, vitrectomy, graft Descemet folds, adverse reactions/complications, and surgeons' low caseload.

Conclusions: Penetrating keratoplasty shows an overall positive prognosis in the long-term. However, the probability of graft survival is largely dependent on the preoperative clinical condition and the lack of complications during follow-up.