Use of enzymes may facilitate DALK

This laboratory study explored the safety of a new technique of lamellar dissection using enzymatic digestion of the corneal stroma and extracellular matrix. The authors found the results encouraging and say that enzymolysis may represent an effective innovation in deep anterior lamellar keratoplasty (DALK) with an acceptable safety profile. They say further studies are required to refine the technique and application of the enzymes in vivo.

This research grew out of the search for an effective, safe and reproducible technique for DALK without a steep learning curve. In centers without high surgical volume, tissue accessibility and operating room access, the learning curve may be a deterrent to the performance of DALK. There is emerging interest in the role of enzymolysis of the corneal stroma in facilitating lamella keratoplasty. The differing profiles of collagen and proteoglycan subtypes between the cornea and Descemet membrane suggest that stromal components may be selectively softened or separated by enzymolysis.

They performed hyaluronidase and trypsin-assisted DALK in 17 cadaveric human corneal tissue samples. Rates of perforation and Descemet membrane exposure were recorded by clinical observation and by OCT in selected cases. Two tissue samples from the same donor were halved, with each half soaked in a different solution (Optisol, balanced salt solution, hyaluronidase, and trypsin) for 13.5 hours to observe maximal effect.

Successful exposure of Descemet membrane was achieved in eight specimens. In the remaining nine, manual dissection was possible to a residual depth of 25 to 90 mm where measured with OCT. Three tissues had perforation of Descemet membrane, all via manual maneuvers. No deleterious effects on residual host tissue were observed by light microscopy with no significant rates of endothelial cell loss in eight tissues in which a predissection cell count was obtainable. The two enzymes had differing effects on soaked specimens that were reflected intraoperatively.

The authors found several results from which they say further analysis can begin:

• At the tested concentrations, there is an observable clinical and histological effect of both enzymes on human corneal tissue.
• With prolonged exposure of trypsin far above the suggested surgical application, they observed cleavage of endothelial cells from Descemet membrane at a higher rate than in controls.
• At the concentrations and exposure times and with the delivery method used, early and limited analysis of dissection specimens suggests no adverse effect on Descemet membrane, endothelium or host margins.
• The enzymes facilitated surgery in different ways, as would be expected by their differing substrates. Hyaluronidase had an effect on the extracellular matrix but not the collagen lamellae. To achieve a bare Descemet membrane with the aid of hyaluronidase, they often repeated injections and layer-by-layer manual dissection was necessary. Trypsin had a lytic effect on planes below those accessed with the initial injection and this allowed a quicker and easier exposure of Descemet membrane .

The authors conclude that success may be improved with the use of more specific enzymes. Keratinases and collagenases would in theory be more specific for human corneal stroma. They say the results of this and other pilot studies indicate that enzymes show potential for facilitating DALK.