Introduction
Descemet membrane endothelial keratoplasty (DMEK) has emerged as a transformative surgical technique for treating corneal endothelial dysfunction, offering rapid visual recovery and excellent graft survival compared to traditional penetrating keratoplasty. Understanding factors that influence long-term graft success is critical to optimizing patient outcomes and refining surgical protocols. The Diabetes Endothelial Keratoplasty Study (DEKS), a multicenter randomized clinical trial, systematically evaluated the impact of various donor, lenticule preparation, recipient, and operative variables on one-year postoperative success rates. This article critically examines DEKS findings to elucidate determinants of DMEK success and their clinical implications.
Background and Clinical Context
Corneal endothelial dysfunction, commonly caused by conditions such as Fuchs endothelial dystrophy or pseudophakic bullous keratopathy, leads to vision impairment secondary to corneal edema. DMEK selectively replaces the diseased Descemet membrane and endothelium with a healthy donor lenticule, minimizing stromal manipulation and associated complications. Despite excellent outcomes reported in controlled settings, various factors including donor tissue quality, recipient comorbidities like diabetes mellitus, and surgical technicalities may affect graft survival. Given the increasing application of DMEK worldwide, robust data identifying modifiable and non-modifiable risk factors is essential.
Study Design and Methods
The DEKS was a prospective, double-masked, multi-center clinical trial involving 28 sites across the United States. A total of 1,421 eyes undergoing DMEK for corneal endothelial dysfunction were randomized to receive donor corneas from diabetic or non-diabetic individuals in a 2:1 ratio. Eyes with previous penetrating keratoplasty or glaucoma tube shunts were excluded to reduce confounding. Donor, preparation, recipient, and intraoperative variables were prospectively recorded. Graft failures were classified as primary donor failure (persistent stromal opacity within 1 week), early failure (cloudiness associated with operative complications within 8 weeks), or late failure.
Primary outcome measured was graft clarity at one year postoperatively. Statistical analyses included proportional hazards and logistic regression models to estimate risk ratios and assess associations.
Key Findings
The overall one-year graft success rate was 97% (1,374/1,421 grafts remained clear), demonstrating high efficacy of DMEK across varied clinical settings. Graft failures were uncommon and categorized into 32 primary donor failures, 12 early failures related to operative issues, and 3 late failures.
Operative complications occurred in 5.5% (78 cases) and significantly reduced graft success to 88% compared to uncomplicated cases (hazard ratio for graft failure 4.0; 95% confidence interval 1.9–8.6, p<0.001). This underscores the critical impact of surgical dexterity and intraoperative factors on outcomes.
Three additional factors demonstrated associations with marginally improved graft survival, though these did not surpass the stringent statistical threshold (P<0.01) set by the investigators:
– Donor endothelial cell density ≥2500 cells/mm² was linked with a 97% success rate versus 95% in <2500 cells/mm² group (P=0.03).
– Lenticule preparation by the surgeon (versus eye bank) achieved a 99% success rate compared to 96% (P=0.02).
– Recipients without diabetes had a slightly higher success (97%) compared to diabetic recipients (94%) (P=0.02).
These findings suggest subtle benefits associated with higher-quality donor tissue, surgeon-prepared grafts, and absence of recipient diabetes, though clinical relevance may be limited.
Expert Commentary and Interpretation
The DEKS reinforces that DMEK is reliably successful with robust one-year graft survival even across diverse donor and recipient profiles. The significantly increased risk of graft failure associated with operative complications highlights the paramount importance of surgical expertise, meticulous technique, and complication prevention strategies.
The trend toward improved outcomes with higher donor endothelial cell density aligns with the biological premise that higher viable cell counts support endothelial function and graft longevity. Surgeon-prepared tissues may offer advantages in tissue handling and tailored preparation, though eye bank preparation remains a practical and safe alternative in many settings.
Diabetes mellitus as a recipient factor is intriguing, given its systemic effects on microvasculature and tissue healing. Slightly lower success rates in diabetic recipients could reflect metabolic influences on graft integration or endothelial cell health post-transplant.
Limitations of the study include exclusion of eyes with previous complex surgeries (penetrating keratoplasty, tube shunts), which may limit generalizability to more challenging clinical scenarios. Moreover, the imbalanced donor diabetes status and the 2:1 randomization ratio warrant cautious interpretation regarding donor diabetes effects.
Further longitudinal studies beyond one year are needed to assess the durability of these associations and long-term graft survival.
Conclusion
This comprehensive analysis from the Diabetes Endothelial Keratoplasty Study confirms that DMEK achieves excellent one-year graft survival rates across a broad spectrum of clinical variables. Operative complications remain the most potent risk factor for graft failure, emphasizing the critical role of surgical technique. While trends suggest benefits from higher donor endothelial cell counts, surgeon-prepared grafts, and non-diabetic recipients, these factors exert modest effects.
Overall, DMEK represents a highly successful therapeutic option for corneal endothelial disorders with opportunities for optimizing outcomes via improved surgical training and meticulous intraoperative care. This study provides valuable evidence to guide donor selection, surgical planning, and patient counseling in contemporary corneal transplantation.
Funding and Trial Registration
The Diabetes Endothelial Keratoplasty Study was supported by institutional grants and conducted at multiple US clinical sites. Registration details and additional administrative data are available in the original publication.
References
Price MO, Szczotka-Flynn LB, Reed ZW, et al. Factors Associated with Successful Descemet Membrane Endothelial Keratoplasty in the Diabetes Endothelial Keratoplasty Study. Ophthalmology. 2026 Jul 28. PMID: 42521031. https://pubmed.ncbi.nlm.nih.gov/42521031/
Additional relevant literature can be consulted for broader understanding of endothelial keratoplasty and factors influencing graft survival.
