Abstract
Purpose:
To evaluate the endothelial cell damage caused by two different injectors for Descemet’s membrane endothelial keratoplasty (DMEK) grafts.
Methods:
Twelve DMEK grafts from 6 pairs of donor eyes not suitable for transplantation were prepared by the Lions VisionGift Eyebank (Portland, OR) with the standard partial pre-peel technique and placement of an ‘S’ stamp for orientation. One eye from each pair was assigned to injection with the glass modified Jones tube injector (Gunther Weiss Scientific Glass, Portland, OR) and one eye was assigned to injection with the Viscoject 2.2 injector (Medicel, Wolfhalden, Switzerland). The grafts were prepared and loaded into the injectors using standard surgical technique, injected onto a bed of viscoelastic on a glass slide and unscrolled using gentle viscoelastic-mediated manipulation. The grafts were stained with the vital fluorescent dye Calcein AM and digitally imaged and analyzed using the open source software FIJI (http://fiji.sc/Fiji). The percentage of cell loss was calculated by measuring the area of non-fluorescent pixels and dividing it by the total graft area pixels. Statistical comparison was performed using the Mann-Whitney U test.
Results:
Grafts injected using the modified Jones tube injector had overall 24±5% (average ± standard deviation, 95% confidence interval (CI) 19 to 30%) cell loss. Grafts injected using the Viscoject had 46±13% (95% CI 32-59%) cell loss. This difference was statistically significant (p=0.01). An area of endothelial damage consistent with the ‘S’ stamp mark was visible on two grafts, one each from the two groups. All grafts demonstrated curvilinear zones of damage. Jones tube grafts had more discrete patterns of cell loss. Viscoject grafts tended to have a higher density of linear damage as well as broader non-linear zones of cell dropout. There was significant edge loss in both groups, thought to be due to trephination damage.
Conclusions:
In this in vitro study, graft injection with the modified Jones tube resulted in significantly less endothelial cell loss than with the Viscoject system. The pattern of cell loss was specific to the injection method, and is suggestive of the types of stresses that grafts are put under during passage through the injector. Trephination resulted in significant peripheral cell loss in both groups, and improved trephination techniques may decrease endothelial loss.