Donor Diabetes Mellitus Severity and Corneal Transplant Suitability in a US Eye Bank Donor Population.

PURPOSE: To determine whether donor diabetes mellitus (DM) severity is associated with differences in endothelial cell density (ECD) and surgically unsuitable tissue. METHODS: Raw data were obtained from Saving Sight Eye Bank (Kansas City, MO) including 10,454 donated eyes from 5346 eligible donors from July 2014 through May 2017. Donors were grouped into 5 categories by their insulin use and the presence of microvascular end-organ complications. The categories were non-DM (NDM), noninsulin-dependent DM without complications (NIDDMnc), noninsulin-dependent DM with complication (NIDDMc), insulin-dependent DM without complications, and insulin-dependent DM with complication. Outcome variables included ECD and tissue transplant suitability. Mixed effects models were used to adjust for the random effect of repeated measures and fixed effects of donor age, race, lens status, and death to refrigeration and death to preservation times. Interaction effects of DM severity group and donor age and DM severity group and lens status were included in the models. RESULTS: One thousand six hundred eighty-four (32.1%) donors had a diagnosis of DM. Six hundred fifty-eight donors were in the NIDDMnc group, 225 in the NIDDMc group, 404 in the insulin-dependent DM without complication group, and 397 in the insulin-dependent DM with complication group. Compared with non-DM, donors with DM were older (P < 0.001) and more likely to be pseudophakic (P < 0.001). DM severity groups did not affect adjusted ECD at mean donor age. There was no statistically significant ECD interaction between DM severity group and lens status. There was a statistically significant ECD crossover interaction with NIDDMnc and donor age (P < 0.001). In phakic eyes, NIDDMc was associated with a statistically significantly lower odds of transplant suitability (odds ratio 0.62, P = 0.006). CONCLUSIONS: DM severity does not affect lowering adjusted ECD at mean donor age. DM severity and pseudophakia were not associated with lower adjusted ECD. NIDDMnc was associated with an attenuation of the age-dependent decrease in ECD. NIDDMc was associated with decreased transplant suitability in phakic eyes. Future studies should include age, lens status, and interaction effects in their models of ECD and transplant suitability.

Effect of Diabetes Mellitus and Metformin on Central Corneal Endothelial Cell Density in Eye Bank Eyes.

PURPOSE: To determine whether metformin use and diabetes mellitus (DM) affect central corneal endothelial cell density (ECD) by examining an eye bank corneal donor database. METHODS: The Lions Eye Institute corneal donor database, which consists of 38,318 corneal samples, was examined. Associations of ECD with metformin use and DM were tested by mixed effects linear models that account for correlations of outcomes between eyes within subjects adjusting for age, intraocular lens status, and glaucoma. Subjects (N = 17,056) with observed ECD counts for both eyes are included for analysis. RESULTS: Average donor age was 56.3 (SD = 15.0). ECD was not associated with metformin use (mean ± SE = 2592 ± 11.9 (N = 1014) versus nonuse [2592 ± 3.0 (N = 16,042), P = 0.302]; further analysis showed that ECD was not significantly associated with metformin use in patients with diabetes. However, metformin use was significantly associated with lower ECD among patients with glaucoma: [2658 ± 50.7 (N = 27) for use versus 2789 ± 19.0 (N = 164) for nonuse, P = 0.018]. The presence of DM was significantly associated with lower ECD 2581 ± 5.6 (N = 4766) for DM versus 2595 ± 3.4 (N = 12,290) for non-DM, P = 0.031). CONCLUSIONS: Lower ECD was associated with DM. Lower ECD was not associated with metformin use except in a subgroup of patients with glaucoma, in which subgroup analysis showed lower ECD. The differences in ECD observed were small and unlikely to affect the suitability for transplantation of donor corneas.

Technician Consistency in Specular Microscopy Measurements: A "Real-World" Retrospective Analysis of a United States Eye Bank.

PURPOSE: To quantify consistency of endothelial cell density (ECD) measurements among technicians in a single US eye bank operating under typical operating conditions. METHODS: In this retrospective analysis of 51 microscopy technicians using a semiautomated counting method on 35,067 eyes from July 2007 to May 2015, technician- and date-related marginal ECD effects were calculated using linear regression models. ECD variance was correlated with the number of specular microscopy technicians. RESULTS: Technician mean ECDs ranged from 2386 ± 431 to 3005 ± 560 cells/mm. Nine technicians had statistically and clinically significant marginal effects. Annual mean ECDs adjusted for changes in technicians ranged from 2422 ± 433 to 2644 ± 430 cells/mm. The period of 2007 to 2009 had statistically and clinically significant marginal effects. There was a nonstatistically significant association between the number of technicians and ECD standard deviation. CONCLUSIONS: There was significant ECD variability associated with specular microscopy technicians and with the date of measurement. We recommend that eye banks collect data related to laboratory factors that have been shown to influence ECD variability.

Age-Stratified Analysis of Diabetes and Pseudophakia Effects on Corneal Endothelial Cell Density: A Retrospective Eye Bank Study.

PURPOSE: To perform an age-stratified analysis of the effect of diabetes and pseudophakia on corneal endothelial cell density (ECD). METHODS: This is a comparative analysis of donor characteristics from data supplied by the Lions Eye Institute for Transplant and Research on tissue harvested from July 1, 2007, through May 23, 2014. The mixed-effects model was used to compare age-adjusted mean corneal ECD between donors with and without diabetes. RESULTS: A total of 20,026 nondiabetic donor eyes and 13,617 diabetic donor eyes were included in this study. ECD was 2604 cells per square millimeter in nondiabetic corneas and 2576 cells per square millimeter in diabetic corneas (P < 0.001). Among phakic patients, diabetic ECD was significantly less in the middle-age subgroups: -33 cells per square millimeter in the 21-to-40-year-old subgroup (P = 0.048) and -25 cells per square millimeter in the 41-to-60-year-old subgroup (P = 0.009). Among pseudophakic patients, diabetic ECD was significantly less only in the subgroup 61 years or older: -56 cells per square millimeter (P = 0.026). The magnitude of difference in ECD between phakic and pseudophakic donors was greater in patients with diabetes in the subgroup 61 years or older (P < 0.001). CONCLUSIONS: Donor eyes with a history of diabetes had a slightly lower ECD (-29 cells/mm) than eyes without a history of diabetes. Although this statistical relationship is consistent with our pathophysiologic understanding of diabetes and the corneal endothelium, such a minor difference in ECD would be expected to have minimal clinical impact on overall corneal endothelial function.

Association Between Corneal Endothelial Cell Density and Topical Glaucoma Medication Use in an Eye Bank Donor Population.

PURPOSE: To evaluate the association between corneal endothelial cell density (ECD) and donor topical glaucoma medication use in an eye bank database. METHODS: Raw eye bank data included 19,159 donors over the period July 2007 to May 2015. Free-text, donor medication lists were retrospectively searched for glaucoma medication. Exclusion criteria were age less than 40 years, history of eye surgery, endothelial trauma, guttae, and cell densities <1000 or >3300/mm. Analysis of covariance was used to test differences in cell density between groups while adjusting for age. Linear regression was used to test the correlation of independent interval variables while adjusting for age. RESULTS: Twelve thousand one hundred fifty-seven donors were included in the final analysis; 134 were on topical glaucoma medication. The mean ECD for donors not on glaucoma medication and pooled donors on glaucoma medication was 2561 ± 348 and 2516 ± 320 cells/mm, respectively (P = 0.42). Subgroup analysis by medication class resulted in nonstatistically significant differences between ECDs of nonmedicated donors and donors on alpha agonists (P = 0.76), beta blockers (P = 0.90), carbonic anhydrase inhibitors (P = 0.13), cholinergics (P = 0.37), and prostaglandin analogs (P = 0.62). The number of glaucoma medication classes used by donors was not a statistically significant predictor of endothelial density (P = 0.298). CONCLUSIONS: Donors on topical glaucoma medication do not have ECDs statistically significantly lower than donors not on medication.

Analyses of Factors Affecting Endothelial Cell Density in an Eye Bank Corneal Donor Database.

PURPOSE: To analyze the factors affecting central corneal endothelial cell density (ECD) in an eye bank corneal donor database. METHODS: The Lion's Eye Institute corneal donor database consisting of 18,665 donors (34,234 corneas) aged 20 years or older was analyzed. In particular, differences in the ECD based on age, sex, race, prior ocular surgery, a history of systemic diseases, and smoking were investigated. Furthermore, risk factors for donor cell count inadequacy (defined here as ECD less than 2000/mm) were identified. RESULTS: ECD decreased with age. Regarding race, the average ECD of African American donors was higher than those of white or Hispanic donors. A history of diabetes mellitus (DM) and ocular surgery were associated with a lower ECD. Donor medical history of hypertension, glaucoma, depression, dementia, Parkinson disease, hyper- or hypothyroidism, or smoking did not seem to affect the ECD. The risk factors for donor cell count inadequacy, based on binary logistic regression analyses were advanced age [65-74 years yielded an odds ratio of 17.8; confidence interval (CI), 10.6-29.8; P < 0.001; and 75-99 years yielded an odds ratio of 24.6 (CI, 14.5-41.61; P < 0.001) when compared with 20-34 years], cataract surgery (odds ratio, 4.3; CI, 4.0-4.8; P < 0.001), and DM (odds ratio, 1.2; CI, 1.1-1.3; P = 0.001). CONCLUSIONS: Age, race, ocular surgery (cataract and refractive), and DM seem to significantly affect donor corneal ECD. Of these variables, age, a history of cataract surgery, and DM were found to be the greatest risk factors for inadequate donor cell density (less than 2000/mm).

Keratoplasty in the United States: A 10-Year Review from 2005 through 2014.

PURPOSE: To report evolving indications and preferred techniques of corneal transplantation in the United States. DESIGN: Retrospective review. METHODS: Annual reports from the Eye Bank Association of America on corneal graft distribution in the United States from 2005 through 2014 were reviewed. MAIN OUTCOME MEASURES: Number and percentage of corneal grafts distributed for various types of keratoplasty and their surgical indications in the United States. RESULTS: The total number of corneal transplants increased from 44 277 in 2005 to 46 513 in 2014. In the past decade, penetrating keratoplasty dramatically decreased (from 95% to 42%) and largely has been replaced by various lamellar keratoplasty (LK) techniques (from 5% to 58%). Descemet stripping (automated) endothelial keratoplasty was the most common (50%) type of corneal transplantation performed in the United Stated in 2014. The volume of Descemet membrane endothelial keratoplasty (DMEK) has been doubling every year since 2011 and accounted for 11% of total endothelial keratoplasties in 2014. There was a significant shift in indication for corneal transplantation, with Fuchs' endothelial dystrophy (22%) being the most common, followed by corneal edema occurring after cataract surgery (12%) in 2014. Eye banks supplied precut corneal grafts for 68% of LK techniques in 2014. CONCLUSIONS: In the United States, there has been a major shift in preferred keratoplasty techniques over the past decade, with a wide adoption of new LK techniques.