Effectiveness And Safety Of Endothelial Corneal Transplant To Treat Corneal Diseases: An Overview Of Systematic Reviews And Meta-Analyses

Evaluate the best available evidence from systematic literature reviews (SLRs) regarding the effectiveness and safety of descemet membrane endothelial keratoplasty (DMEK) and descemet stripping automated endothelial keratoplasty (DSAEK) for the treatment of endothelial diseases.

We conducted a systematic search in the Cochrane Library, MEDLINE, EMBASE, and Epistemonikos. The last search was performed on July 14, 2024. We also examined the reference lists of the included SLRs, the PROSPERO repository, and selected high-quality, full-text electronic journals related to corneal and endothelial transplantation. The search was limited to studies published from 2006 onward, as that is when the initial publications on DMEK were released.

We conducted an overview following Cochrane and PRIO-harms guidelines. Our systematic search identified SLRs comparing DMEK and DSAEK in endothelial disease, such Fuchs' endothelial dystrophy (FED), pseudophakic bullous keratopathy (PBK), and iridocorneal syndrome, assessing corrected-distance visual acuity (CDVA), endothelial cells, spherical equivalent (SE), and safety at 6- and 12-month follow-ups. Paired reviewers extracted data and assessed the methodological quality of SLRs with AMSTAR 2.0. We estimated overlap using the corrected covered area (CCA) and performed a meta-analysis on primary and unique studies from the SLRs using a random-effects model to prevent double-counting. We registered the protocol in PROSPERO (CRD42024493403).

We analyzed 11 SLRs encompassing 31 primary studies and 6,416 eyes with endothelial disease. Eight SLRs were low quality and three were high. Due to high overlap (CCA= 0.18), we meta-analyzed 17 primary and unique studies, excluding 14 for issues with outcomes (N= 6), interventions (N= 4 ), or study type (N= 4).

DMEK improved CDVA in LogMAR versus DSAEK at 6 months (N= 920, MD= 0.1, 95% CI: 0.04 to 0.16) and 12 months (N= 443, MD= 0.1, 95% CI: 0.05 to 0.16), but increased complication risk (N= 1,013, RR= 0.43, 95% CI: 0.30 to 0.61). In addition, postoperative endothelial cell counts were similar between the two techniques (N= 355, MD= 27.32, 95% CI: –148.51 to 203.15) as was the SE (N= 195, MD= –0.00, 95% CI: –0.61 to 0.61).

This is the first overview comparing DMEK and DSAEK for endothelial diseases. The best available evidence from SLRs suggests that DMEK offers superior visual outcomes at 6 and 12 months compared to DSAEK, although with a higher relative risk of complications. Notably, only 0.7% (N= 4 eyes) of DMEK complications were vision-threatening due to cystoid macular edema, with the remainder being transient issues such as graft dislocations requiring re-bubbling. Furthermore, both techniques resulted in comparable postoperative endothelial cell counts and SE outcomes.