Descemet Membrane Endothelial Keratoplasty for Failed Penetrating Keratoplasty: Visual Outcomes and Graft Survival.

PURPOSE

To evaluate outcomes and graft survival of Descemet membrane endothelial keratoplasty (DMEK) for failed penetrating keratoplasty (PK).

METHODS

Ninety-three DMEK procedures performed in 84 eyes of 77 patients with failed PK were retrospectively reviewed. The main outcomes were corrected distance visual acuity and graft survival assessed with Kaplan-Meier survival analysis and proportional hazards modeling taking follow-up into consideration.

RESULTS

Sixty-nine eyes had 1 previous failed PK, 13 had 2, 1 had 3, and 1 had 4. Ten eyes had failed Descemet stripping endothelial keratoplasty (DSEK) performed under failed PK. Fourteen cases (15%) had previous glaucoma filtration surgery (9 trabeculectomy alone; 5 trabeculectomy and aqueous shunt). Median follow-up was 21 months (range, 1 month to 7 years). Median Snellen corrected distance visual acuity improved from 20/100 preoperatively (range, 20/30 to count fingers) to 20/30 at 6 months postoperatively (n = 73; range, 20/20-20/200). Rebubbling rates were 53% when the diameter of the DMEK graft was oversized, 27% when same sized, and 33% when undersized relative to that of the previous PK graft. Two grafts (2%) experienced an immunologic rejection episode, and 15 (16%) failed, including 5 primary/early failures and 10 late failures. Previous glaucoma surgery was the only significant risk factor for failure (relative risk, 7.1; 95% confidence interval, 2.1-37.0). The 1-, 2-, and 3-year graft survival rates were 96%, 89%, and 89% without versus 78%, 53%, and 39% with previous glaucoma surgery.

CONCLUSIONS

Treatment of failed PK with DMEK produced similar 4-year survival (76%) and better visual outcomes than previously reported with Descemet stripping endothelial keratoplasty or an initial PK regraft.