Graft Survival, Graft Rejection, and Visual Outcomes After Descemet Stripping Endothelial Keratoplasty for Previous Graft Failure.

PURPOSE: The objective of this study was to investigate outcomes of Descemet stripping endothelial keratoplasty (DSEK) in eyes regrafted for failed penetrating keratoplasty (PK) or failed DSEK.

METHODS

This was a retrospective consecutive series of all DSEK cases performed from April 2006 to November 2015 for failed PK or failed DSEK. Graft survival and graft rejection risk were estimated by Kaplan-Meier analysis.

RESULTS

Of 76 DSEK cases, 42 (55%) were regrafted for failed PK and 34 (45%) for failed DSEK. The mean age at regraft was 72 years; 49 participants (64%) were women. The median follow-up was 4.0 years. DSEK graft survival was 72% and 57% at 5 and 10 years, respectively (failed PK: 64% and 41%; failed DSEK: 81% and 76%, [P = 0.051]). The cumulative risk of graft rejection was 17% at 5 and 10 years (failed PK: 24% at 5 and 10 years; failed DSEK: 10% at 5 and 10 years [P = 0.21]). A rebubble procedure was needed in 16 cases (21%) (13 for failed PK cases and 3 for failed DSEK cases, P = 0.02). For eyes without vision-limiting comorbidities (n = 42), the mean preoperative best spectacle-corrected visual acuity (BSCVA) was 0.83 ± 0.36 logMAR (mean ± SD) in eyes with failed PK and 0.51 ± 0.32 logMAR for failed DSEK (P = 0.02). After adjusting for preoperative BSCVA, the BSCVA (least-square mean [95% confidence interval]) at one-year follow-up was 0.58 logMAR [0.44-0.72] for failed PK and 0.28 logMAR [0.18-0.30] for failed DSEK (P = 0.002).

CONCLUSIONS

Compared with DSEK regrafts for failed PK, DSEK regrafts for failed DSEK were associated with better visual outcomes and lower rebubble procedure rates.