A new donor cornea harvesting technique for posterior lamellar keratoplasty

For cases of pure endothelial damage, posterior lamellar keratoplasty (PLK) involves replacement of only the diseased endothelium along with some underlying posterior stroma, most often through a deep lamellar limbal incision, as an alternative to a full thickness penetrating graft.^1–3 We describe a simplified technique, based on the approach proposed by Ignacio et al,⁴ for preparing the donor cornea for PLK, and evaluate it with regard to lamellar planarity and endothelial cell loss.

MATERIALS AND METHODS

Eight corneoscleral buttons stored in Optisol GS at 4°C and not suitable for clinical transplantation were obtained (Central Florida Lions Eye and Tissue Bank, Tampa, Florida; Donor Network of Arizona, Phoenix, AZ, USA) and divided into two groups. Ultrasonic pachymetry was used to measure the preoperative central corneal thickness after mechanical de-epithelialisation as well as after posterior lenticule removal.

Three buttons were mounted endothelial side down onto an artificial anterior chamber (Moria, Inc, France) and pressurised by air injection. Deep limbus to limbus lamellar dissections were carried out to approximately 80–90% depth using the three blade Melles lamellar keratoplasty set (Dorc, Netherlands) (group 1, n = 3).

Three buttons were mounted endothelial side up on the same chamber, pressurised with air and dissected limbus to limbus to 10–20% depth using a 2 mm wide Culler iris spatula (Katena, Denville, NJ, USA) after initial entry with a 3.0 mm steel keratome (Sharpoint, Reading, PA, USA) (group 2, n = 3). The endothelium was lubricated with Optisol throughout the procedure. After removal of the corneas from the artificial anterior chamber, all six posterior lenticules were then removed from the endothelial side using a 6.5 mm manual trephine (Surgistar, Carlsbad, CA, USA).

Endothelial cell damage after graft harvest was assessed by staining with trypan blue 0.25% (Sigma-Aldrich, Inc, St Louis, MO, USA) and alizarin red S 0.2% (Sigma-Aldrich, Inc).⁵

Two corneoscleral buttons underwent stromal pocket dissection as described above but without trephination and were submitted for routine histological processing and staining with periodic acid Schiff (PAS).

RESULTS

The average harvested lenticule thickness was 118 (SD 32) μm (group 1) and 92 (23) μm (group 2) (Student’s t test, p = 0.324). Percentage of devitalised endothelial cells was 0.86% (1.48%) (group 1) and 3.9% (2.9%) (group 2) (p = 0.185). The dissections using each of the two harvesting techniques remained in single deep lamellar planes and were equally smooth (fig 1).

View larger version:

• In a new window

• Download as PowerPoint Slide

Figure 1

 Comparison of two different donor cornea harvesting techniques. The left column shows a cornea mounted endothelial side down. The right column shows a cornea mounted endothelial side up. The initial entry incision is made using either the first of three Melles blades (A) or a steel keratome (B). The dissection is then performed with the second and third Melles blades (C) or a Culler iris spatula (D). Histological analysis demonstrates that the lamellar dissections are equally planar (E) and (F). Vital staining of the endothelium reveals low cell loss using both techniques (G) and (H).

DISCUSSION

Here we describe a technique for harvesting donor tissue for posterior lamellar keratoplasty in which a corneoscleral button is mounted onto an artificial anterior chamber endothelial side up and stromal dissection is accomplished with a simple blunt spatula.

The planarity and endothelial cell loss associated with the harvest appear comparable to previously described techniques. This modified method is advantageous in that a standard re-usable blunt corneal dissector is employed without the added complexity and cost associated with sharp, disposable three knife sets. Furthermore, the orientation of the endothelium up on the anterior chamber allows greater surgeon control during the harvest.