Ten-year outcomes of repeat keratoplasty for optical indications

To analyze the clinical outcomes of repeat keratoplasties following a failed primary optical keratoplasty in an Asian population.

In this prospective cohort study, clinical data of consecutive patients who had repeat keratoplasty at Singapore National Eye Centre from 2007 to 2020 were recorded from our ongoing Singapore Corneal Transplant Registry.

A total of 284 first regrafts (181 Descemet stripping automated endothelial keratoplasty (EK), 63 penetrating keratoplasty (PK), 21 Descemet membrane endothelial keratoplasty, and 19 deep anterior lamellar keratoplasty (DALK)) were analyzed. Graft rejection (38.4%) and late endothelial failure (15.8%) were the main causes of initial graft failure. PK/EK and EK/EK had better one-year postoperative best corrected visual acuity compared to PK/PK (p = 0.006 and p < 0.001 respectively). Kaplan-Meier 10-year cumulative regraft survival probabilities were 69.2% for PK/EK, 52.8% for EK/EK, and 43.1% for PK/PK. Regraft survival was 100% for DALK/PK, DALK/DALK, and DALK/EK at three to 5 years. Log-rank test showed higher survival of PK/EK compared to PK/PK (p = 0.002) and EK/PK (p = 0.009), and of EK/EK compared to PK/PK (p = 0.003) and EK/PK (p = 0.005). High-risk regrafts had significantly lower 10-year survival probabilities compared to non-high-risk regrafts (p = 0.045). Cox multiple regression analysis showed male gender (p = 0.023), PK regraft (p = 0.003), regraft rejection (p = 0.003), and initial graft indications of pseudophakic bullous keratopathy (p = 0.005) and aphakic bullous keratopathy (p = 0.004) to be risk factors for regraft failure, while longer time to regraft was associated with decreased risk of failure (p = 0.013).

Performing EK for failed optical PK or EK significantly improved regraft survival compared to repeat PK. Regrafts performed for failed initial DALK grafts did well regardless of type.