Donor cross-linking for keratoplasty: a laboratory evaluation.

This laboratory-based investigation compares the topographic outcomes of conventional penetrating keratoplasty with that of a novel procedure in which donor corneas are cross-linked prior to keratoplasty. Penetrating keratoplasty procedures with continuous running sutures were carried out in a porcine whole globe model. Sixty eyes were randomly paired as 'donor' and 'host' tissue before being assigned to one of two groups. In the cross-linked group, donor corneas underwent riboflavin/UVA cross-linking prior to being trephined and sutured to untreated hosts. In the conventional keratoplasty group, both host and donor corneas remained untreated prior to keratoplasty. Topographic and corneal wavefront measurements were performed following surgery, and technical aspects of the procedure evaluated. Mean keratometric astigmatism was significantly lower in the cross-linked donor group at 3.67D (SD 1.8 D), vs. 8.43 D (SD 2.4 D) in the conventional keratoplasty group (p < 0.005). Mean wavefront astigmatism was also significantly reduced in the cross-linked donor group 4.71 D (SD 2.1) vs. 8.29D (SD 3.6) in the conventional keratoplasty group (p < 0.005). Mean RMS higher order aberration was significantly lower in the cross-linked donor group at 1.79 um (SD 0.98), vs. 3.05 um (SD 1.9) in the conventional keratoplasty group (P = 0.02). Qualitative analysis revealed less tissue distortion at the graft-host junction in the cross-linked group. Cross-linking of donor corneas prior to keratoplasty reduces intraoperative induced astigmatism and aberrations in an animal model. Further studies are indicated to evaluate the implications of this potential modification of keratoplasty surgery.