Advancing Descemet Membrane Endothelial Keratoplasty: Efficient Tri-Folded Graft Delivery Protocol With A Hydrophobic Asymmetrical Injector

One of the recent advancements in preloaded Descemet membrane endothelial keratoplasty (DMEK) surgery involves delivering the graft with the endothelium inwards, enabling faster operations but necessitating a pull-through surgical technique. Despite the critical advantages of the tri-folded endo-in DMEK technique, the lack of suitable surgical instruments that can accommodate this method without the pull-through technique remains a barrier to its widespread adoption. This study presents an eye bank validation of an asymmetrical injector designed for the implantation of tri-folded DMEK grafts without requiring a pull-through technique. The device was also evaluated for Descemet Stripping Automated Endothelial Keratoplasty (DSAEK) surgery.   

 Four different internal compartment designs of the injectors were assessed using DMEK tissues. Donors' age in the study was 36-71 (57.9 average) with 2-23 days death to preparation days (14.3 average). Mates from two pairs were tested on each device type, totaling 16 grafts, all loaded with folded,endo-in grafts. Sixteen grafts, all loaded with tri-folded, endo-in orientation were tested. Two DSAEK grafts of 100-110 micron thickness were Tri-Folded, loaded into the injector with micro forceps.

Tissues were prepared, loaded into the injector, and ejected to simulate tissue manipulation during DMEK operations. The tissues were validated according to current standard operation protocols with Fiji viable/nonviable (Trainable Weka Segmentation v3.2.33 software) with a background connection. The injector, constructed from transparent, very hydrophobic plastic, permits direct microscopic tissue validation before injection. The factors affecting the outcomes of tri-folded DMEK include good/bad deployment and injection, tissue orientation during the injection, and double scrolled in the inserter were recorded.

Endothelium-in graft delivery was achieved through injection without the need for a pull-through technique. Only one graft (6.25%) experienced double-scrolling within the injector with a larger (1.5 mm) internal compartment. Cell loss ranged from 3-23% (average 13.98%), with no significant differences observed between injectors with different internal compartment sizes. Interestingly, DMEK grafts stored for >20 days demonstrated higher viability loss (17.3% +/- 5.7) compared to those stored for less than two weeks (10.9% +/-2.1). DSAEK could be easily loaded and moved within the device. However, the nozzle diameter of 0.9 mm worked well for DMEK grafts but prevented DSAEK from being injected.

The asymmetrical injector demonstrates promise as a tool for oriented, tri-folded DMEK injection without the need for a pull-through technique. e injector, asymmetrical design enables precise control and validation of graft orientation, aligning with best eye bank practices crucial for the successful clinical application of tri-folded DMEK grafts. The observed average cell loss after loading and ejection (13.98%) is comparable to or better than current best practices with the precut-preloaded technique of naturally folded DMEK.