S. C. Yiu

Fungal keratitis caused by Scedosporium apiospermum: report of two cases and review of treatment.

Purpose. We report our experience in treating two cases of Scedosporium apiospermum keratitis and provide a review of basic scientific and clinical data regarding the treatment of this visually devastating disease. Method. We present a case report and literature review. Results. A 35-year-old woman and a 73-year-old man both reported pain, redness, and a foreign body sensation in the eye after trauma. They were initially treated with antibacterials. When cultures were positive for S. apiospermum, the first patient was treated with fluconazole and amphotericin with good results. The second was treated with fluconazole and natamycin, but the infection persisted and the eye was eventually enucleated. Review of the literature showed variable responsiveness of S. apiospermum to antifungal treatment. Conclusions. It is difficult to predict how S. apiospermum keratitis will respond to treatment, but miconazole appears to be useful in conjunction with other antifungals. Voriconazole shows promise as an effective alternative.

Technique for preparation of the corneal endothelium-Descemet membrane complex for transplantation.

Purpose: Replacing diseased corneal endothelium with a preparation of Descemet membrane carrying functional endothelium and no stroma may be a feasible method for treating corneal endothelial decompensation. To obtain a viable donor of a Descemet membrane endothelium disc, we modified the Descemet membrane stripping technique and monitored the percentage of endothelial damage to the donor tissue preparation. Methods: Forty-eight human corneas were used. Cornea buttons were mounted on an artificial anterior chamber, endothelial side up. Endothelia were stained with alizarin red, examined under the microscope, and photographed at 5 different sites (microscope, ×100; digital magnification, ×2.83). A 6 × 7-mm rectangular piece of endothelium-Descemet membrane complex was obtained using a Grieshaber microsurgical knife and Kelman-McPherson forceps. Digital photographs of endothelia were analyzed with a computer, and the percentage of endothelial damage was calculated. Specimens were processed for hematoxylin-eosin staining. Results: Forty of 48 endothelium-Descemet membrane preparations (83.3%) were complete peels with minimal endothelial damage. Endothelial damage before and after the surgery was 1.57 ± 2.11% and 2.61 ± 1.77%, respectively. Eight preparations (16.7%) failed because of tearing. Multiple hematoxylin-eosin-stained sections showed the presence of endothelium with intact Descemet membrane and no stromal tissue. Conclusion: We modified the technique of Melles and obtained a sheet of Descemet membrane and endothelium with minimal endothelial damage and with no remaining stroma observed. This simple technique can be used to obtain the endothelium-Descemet membrane complex in minutes. It may be useful for corneal endothelium transplantation.

Investigation of femtosecond laser-Enabled keratoplasty wound geometry using optical coherence tomography

Purpose: To measure the wound geometry after femtosecond laser-enabled keratoplasty (FLEK) using optical coherence tomography (OCT). Design: Prospective nonrandomized clinical study and laboratory study. Participants: Patients who were candidates for penetrating keratoplasty at an academic referral center. Methods: Wound architecture was measured and analyzed by OCT in 8 eyes of 8 consecutive patients who underwent FLEK. Femtosecond laser lamellar cuts were performed on 3 eye bank corneas, and the wound diameters were measured by OCT. Results: Sutures were completely removed on average at 1 month per decade of age after surgery. No cases of wound dehiscence were noted, and graft-host tissue apposition appeared excellent on OCT. The mean spectacle-corrected visual acuity 1 month after suture removal was 20/35, and mean astigmatism was 5.0 diopters by manifest refraction and 8.7 diopters by computerized topography. The measured graft diameter was smaller than the laser setting by 2.5% (P = 0.007). The cut diameters of the eye bank corneas were also slightly smaller than the laser setting. Conclusions: FLEK offers benefits of rapid wound healing and predictable wound geometry. The measured wound geometry agreed well with laser setting other than a small systematic deviation that could be explained by the mechanics of corneal applanation during the laser cut. OCT is a useful tool to measure the actual graft diameter after FLEK.

Dual laser-assisted lamellar anterior keratoplasty with top hat graft: a laboratory study.

Objectives: To develop a dual laser-assisted lamellar anterior keratoplasty (LALAK) technique, using excimer and femtosecond lasers to perform surgery on eye bank eyes. Methods: First, we compared corneal stromal surfaces produced by (1) deep excimer ablation, (2) femtosecond lamellar cuts, and (3) manual dissection and evaluated the effect of excimer laser smoothing with fluid masking on each surface. Masked observers graded scanning electron microscopy images on a 5-point roughness scale. Then, we performed a 6-mm diameter excimer laser phototherapeutic keratectomy ablation to a residual bed thickness of 200 &mgr;m, followed by laser smoothing. We used the femtosecond laser to cut donors in a modified top hat design with a thin tapered brim, which fitted into a manually dissected circumferential pocket at the base of the recipient bed. Fourier domain optical coherence tomography was used to measure corneal pachymetry and evaluate graft fit. Results: Deep excimer ablation with smoothing (n = 4) produced a significantly (P < 0.05) smoother surface (grade = 3.5) than deep excimer alone (n = 4, grade = 3.8) or manual dissection with (n = 1, grade = 3.8) and without smoothing (n = 1, grade = 4.8). Deep femtosecond cuts (n = 2) produced macroscopic concentric ridges on the stromal surface. Experimental LALAK was performed on 4 recipients prepared by deep excimer ablation and 4 donors cut with the femtosecond laser. After suturing, good peripheral graft-host match was observed on Fourier domain optical coherence tomography imaging. Conclusion: These preliminary studies show that the LALAK technique permits improved interface smoothness and graft edge matching. Clinical trials are needed to determine whether these improvements can translate to better vision.