Purpose
To determine whether deep anterior lamellar keratoplasty (DALK) using acellular glycerol-cryopreserved corneal tissue (GCCT) could prevent allograft rejection in high-risk corneas.
Design
Prospective, randomized, comparative study.
Methods
settings: The Eye Hospital, Wenzhou Medical College, Zhejiang, China. study population: All patients with herpes simplex virus keratitis, bacterial keratitis, fungal keratitis, or ocular burn, who were eligible as per study design, were invited to participate. observation procedures: According to randomized block design, all patients received either GCCT or fresh corneal tissue (FCT) during DALK. Best-corrected visual acuity (BCVA), slit-lamp microscopy, and in vivo confocal microscopy examinations at 1 week and 1, 3, 6, 12, and 24 months after surgery were analyzed. Kaplan-Meier survival analysis was used to evaluate graft survival rate. main outcome measures: Therapeutic success, 2-year rejection-free graft survival rate and 2-year graft survival rate, in vivo confocal microscopy results, BCVA, and endothelial cell density.
Results
Postoperative BCVA of 20/40 or better at the last follow-up visit was achieved in 57.6% (19/33) of eyes in the GCCT group and in 54.8% (17/31) of the FCT group. No graft rejection occurred in the GCCT group, while in the FCT group 10 episodes of stromal rejection developed in 7 eyes. Overall, the rejection-free graft survival rate at 2 years was significantly higher in the GCCT group as compared with the FCT group (100.0%, 78.8% respectively, P = .006).
Conclusions
Deep anterior lamellar keratoplasty using acellular glycerol-preserved cornea could prevent allograft rejection and promote graft survival rate in high-risk corneas.
Corneal transplantation is important since it is the ultimate cure for blinding corneal disease, and it represents one of the most frequent tissue allotransplantations performed all around the world. Advances have been achieved in corneal transplantation in the past decades. However, immune-mediated corneal allograft rejection remains the main cause of graft failure.
Deep anterior lamellar keratoplasty (DALK) is advantageous over penetrating keratoplasty (PK) because of the elimination of endothelium rejection and the ability to use lower-quality tissues. However, epithelial and stromal rejections can still occur after DALK. Cellular components of fresh corneal tissue (FCT) including the epithelium, keratocytes, and bone marrow–derived cells are sources of major histocompatibility complex (MHC) antigens that might be responsible for DALK rejection. One way to overcome this limitation may be through transplantation of corneal substrates without their cellular components. Glycerin cryopreservation could be a practical and effective technique to carry out decellularization. Acellular glycerol-preserved corneal tissue (GCCT), in contrast to FCT, cannot lead to activation of an indirect immune transplant rejection pathway. Because it lacks antigen-presenting cells, it cannot directly sensitize the recipient T cells. Our previous study has shown GCCT can be used safely and effectively in patients with low-risk corneal conditions. Confocal microscopy demonstrated glycerin-cryopreserved corneas were acellular at 2 weeks after DALK, which implied that all cells, including antigen-presenting cells, keratocytes, and other resident bone marrow–derived cells, are not viable after glycerin cryopreservation.
As a matter of fact, a large proportion of corneal transplantation in developing countries is considered to be at high risk of corneal graft failure. The current treatments to prevent or treat rejection are not very good; in fact, they have barely changed in over 50 years. The survival of corneal grafts is significantly lower in an inflamed eye than in the low-risk noninflamed eye. Accordingly, high-risk corneal transplantation remains a formidable challenge. Herein, a prospective randomized study was performed to determine whether acellular GCCT could prevent rejection in high-risk corneas, characterized by vascularized and/or infected host beds. To the best of our knowledge, no study has compared the clinical outcomes of GCCT with FCT used in high-risk corneal transplantation. In the present study, we hypothesized that DALK using GCCT could be a simple and practical technique to minimize allograft rejection in high-risk corneas.
Methods
Study Groups
All patients with herpes simplex virus keratitis (HSK), bacterial keratitis, fungal keratitis, and ocular burn who were eligible as per study design were invited to participate. These patients underwent DALK at the Cornea Service of the Eye Hospital of Wenzhou Medical College. The definition of high-risk cornea was preexisting corneal neovascularization in 2 or more quadrants or an infected recipient bed. The extent of corneal involvement had to be less than 8.5 mm. Surgical intervention was considered for: 1) active bacterial keratitis, for which ulceration progressed despite maximum antibacterial medication; 2) refractory fungal keratitis that did not respond to antifungal agents; 3) nonactive HSK, for which corneal opacities with or without new vessels involved the optical zone; and 4) ocular acid burn and thermal burn with partial limbal deficiency (50% or less) that, after more than half a year of preoperative treatment, showed re-epithelialization and less than 2 quadrants limbal neovascularization. Patients with keratolimbal allograft transplantation, total limbal stem cell deficiency secondary to ocular burns, and other ocular diseases (ie, amblyopia, age-related cataract, glaucoma, macular edema, and macular degeneration) were excluded.
According to randomized block design, 68 eyes of 68 patients who met the criteria were distributed into 2 treatment groups by blocking on etiology: the GCCT group and the FCT group. All patients received explanations about characteristics of both GCCT and FCT before surgery. All of the surgeries were performed by the same surgeon (C.W.).
Data Collection
Data from routine follow-up examinations at 1 week and 1, 3, 6, 12, and 24 months postoperatively included best-corrected visual acuity (BCVA), slit-lamp biomicroscopy, and in vivo confocal microscopy examinations. Visual acuity was determined using the standard Snellen chart and spectacle corrected. A laser scanning confocal microscope (Heidelberg Retina Tomograph II with Rostock Cornea Module; Heidelberg Engineering, Dossenheim, Germany) was used to scan the center of each graft as previously described. A caliper tool (Analysis 3.1; Soft Imaging System, Munster, Germany) was used to count the cells in each image (400 × 400 μm) manually. The results were expressed in cells per square millimeter.
Surgical Technique
All DALKs were performed under retrobulbar anesthesia. The host cornea was trephinated to a level of 60% to 70% of its thickness using the Hessburg-Barron disposable vacuum trephine (Katena Instruments, Denville, New Jersey, USA). The diameter of the trephine, ranging from 7.5 to 9.0 mm, was at least 0.5 mm bigger than that of the lesion to make sure that all the infected tissue would be removed. A sharp 15-degree tipped blade was used to deepen the trephination groove to a depth of 70% to 80% of corneal thickness. After a stromal pocket was made by a Sinskey hook, a blunt-tipped iris spatula was inserted gently into the pocket to the corneal periphery in every meridian by gliding and rotating movements. The anterior lamellar stroma (about 80% in thickness) was removed using Vannas scissors. Sterile water for injection with zero osmotic pressure was spread on the host bed to make the residual stromal fibers swell, and then another stromal pocket was formed with a Sinskey hook. The residual stroma was removed layer by layer with repeated dissections until Descemet membrane was exposed. A full-thickness graft without endothelium and Descemet membrane, with a diameter 0.25 mm larger than the host bed, was sutured in place using 16 interrupted stitches of 10–0 nylon.
Donor Corneal Preparation
Fresh donor corneas were preserved at 4 C in Optisol-GS (Bausch & Lomb, St. Louis, Missouri, USA). The preserved donor corneas were stored at −78 C in pure sterile glycerin. The average duration of cryopreservation before surgery was 9.6 ± 2.8 months.
Postoperative Management
For fungal keratitis patients, all received 0.5% natamycin eye drops (Natacyn; Alcon Laboratories Inc., Puurs, Belgium) and 0.25% amphotericin B eye drops (Oubo; North China Pharmaceutical Group Corporation, Shijiazhuang, China; compound was diluted with distilled water to arrive at a 0.25% concentration) 4 times daily for 2 weeks, and oral itraconazole capsules (Sporanox; Janssen Ltd, Xian, China) 200 mg daily for 3 weeks to reduce the risk of fungal recurrence. Topical 1% cyclosporin A (provided by the pharmacy of Eye Hospital, Wenzhou Medical College, Wenzhou, China) was administered in the early postoperative period as a prevention substitute for allograft rejection and for its potential antifungal effect. For bacterial keratitis patients, broad-spectrum antibiotic eye drops were commenced immediately, and topical steroids generally were commenced only 1 to 5 days after surgery and tapered gradually. For HSK patients, long-term therapy consisted of topical antibiotics and steroids, together with oral acyclovir, and tapered gradually. For ocular burn patients, postoperative medication consisted of topical antibiotics and steroids 4 times daily for 2 weeks and then tapered.
All patients were observed closely for signs of inflammation, graft rejection, infection, suture-related problems, and refractive outcome. Immunologic stromal rejection was diagnosed clinically in the presence of decreased vision and stromal edema with or without corneal neovascularization. The definition of graft failure was a regraft or an irreversible loss of central graft clarity for a minimum of 3 consecutive months. The definition was quite different with therapeutic failure, which was defined as recurrent corneal infection that progressed to endophthalmitis or evisceration despite medical and surgical intervention. Suture removal was performed at around 3 months after surgery. In patients with evidence of allograft rejection, topical steroid eye drops were given hourly for 5 to 7 days, and oral prednisolone was commenced and tapered according to the physician’s clinical judgment.
Statistical Analysis
The categorical data comparisons were assessed by χ 2 test with Bonferroni adjustment, and the quantitative data were assessed by the unpaired t test. Survival curves were analyzed by Kaplan-Meier survival method at 2 years. The statistical differences in the age and trephination size between the 2 groups were analyzed using Wilcoxon rank sum test. All tests were 2-tailed, and a P value < .05 was considered significant.
Results
Patient Data
Sixty-eight consecutive patients who underwent DALK for corneal keratitis or ocular burn between November 7, 2006 and July 10, 2008 (53 male and 15 female; mean age, 48.2 ± 12.6 years; range, 17–77 years) were included in our study, which included HSK (n = 30), bacterial keratitis (n = 15), fungal keratitis (n = 14), and ocular burns (n = 9). The study only analyzed 2 years of follow-up data. Demographic data and operative details of these patients are summarized in Table 1 . Our results show that there was no significant difference in the age, sex, eyes, and trephination size between 2 groups.
| GCCT (n = 34) | FCT (n = 34) | P | |
|---|---|---|---|
| Male: female (n) | 26:8 | 27:7 | .77 |
| Left: right (n) | 17:17 | 18:16 | .81 |
| Age (mean ± SD, years) | 50.7 ± 13.5 | 45.9 ± 11.5 | .12 |
| Primary pathogenesis | .98 | ||
| Herpes simplex virus keratitis (n) | 15 | 15 | |
| Bacterial keratitis (n) | 8 | 7 | |
| Fungal keratitis (n) | 7 | 7 | |
| Ocular acid burn & thermal burn (n) | 4 | 5 | |
| Host diameter [median (IQR), mm] | 7.75 (7.50, 8.00) | 7.75 (7.50, 8.00) | .84 |
| Donor diameter [median (IQR), mm] | 8.00 (7.75, 8.25) | 8.00 (7.75, 8.25) | .96 |
Therapeutic Success
Of the infectious keratitis cases (n = 59), 2 patients had recurrence of infection. The primary etiology was 1 fungal from the GCCT group and 1 bacterial from the FCT group. Therapeutic success was defined as complete eradication of primary infection without signs of recurrent infection after DALK with appropriate adjunctive medical therapy. In this study, the therapeutic success rate achieved was 96.6%. The recurrence time for both patients was within 2 weeks after surgery. The patient with recurrent fungal infection underwent a repeat lamellar graft. Another patient with recurrent bacterial infection refused a repeat lamellar graft and underwent instead a PK procedure. Neither patient showed further recurrence. No recurrence occurred in any HSK patient. In ocular burn cases (n = 9), corneal opacifications were all successfully removed by the surgery.
Intraoperative Descemet membrane microperforation occurred in 5 eyes (2 eyes in the GCCT group and 3 eyes in the FCT group) undergoing the modified manual dissection technique, but the procedure was completed in all of them without the need to convert to PK. Two of them developed into a second anterior chamber, which was resolved spontaneously within 3 days postoperatively.
Allograft Rejection and Graft Survival
Ten episodes of stromal rejection developed in 7 of 68 eyes (10.3%), including fungal keratitis (n = 3), HSK (n = 3), and ocular thermal burn (n = 1) ( Table 2 , Figure 1 ), all from the FCT group. No graft rejection occurred in the GCCT group; the preoperative and postoperative slit-lamp photographs of 4 typical patients are shown in Figure 2 . Four cases developed peripheral stromal vessels at the interface between recipient and graft in the GCCT group but without stromal edema ( Figure 2 , Botttom right). Seven eyes with long-standing keratitis developed pseudopterygium postoperatively ( Figure 1 , Top right). Rejections in the FCT group occurred between 2 and 17 months postoperatively. At the time of rejection, all 7 patients were given topical steroid therapy. This treatment led to reversal of rejection in 5 eyes, but 1 eye with fungal keratitis and 1 eye with HSK developed into irreversible stromal rejection ( Figure 1 ). Above all, the rejection-free graft survival rate at 2 years was 100.0% in the GCCT group and 78.8% in the FCT group (Kaplan-Meier test, P = .006). Meanwhile, the graft survival rate at 2 years was 100.0% in the GCCT group, and in the FCT group it was 93.9% (Kaplan-Meier test, P = .154) ( Figure 3 ). At 2-year follow-up, the difference in the rejection-free graft survival rates between the groups was statistically significant; however, the difference was not that significant in graft survival rate.
| Patient | Group | Age (years), Gender | Eye | Disease | Rejection Time (months) | Final BCVA/Graft Status | Figure |
|---|---|---|---|---|---|---|---|
| 4 | FCT | 44, male | R | Fungal keratitis | 10 | Graft failure | 1, Top |
| 7 | FCT | 51, male | R | Fungal keratitis | 5, 17 | 20/30 | |
| 27 | FCT | 53, female | R | HSK | 10, 18, 20 | Graft failure | |
| 33 | FCT | 46, male | R | Fungal keratitis | 3 | 20/70 | |
| 52 | FCT | 60, male | R | HSK | 2 | 20/50 | |
| 61 | FCT | 48, male | L | Thermal burn | 6 | < 20/200 | |
| 66 | FCT | 56, male | L | HSK | 15 | 20/40 | 1, Bottom |
Visual Acuity Outcomes
Eyes with therapeutic failure and irreversible rejection were excluded from the visual acuity evaluation. Preoperative BCVA worse than 20/200 was seen in 78.8% (26/33) of the GCCT group and 74.2% (23/31) of the FCT group respectively, and it was improved in all cases after DALK. Postoperative BCVA of 20/40 or better at the last follow-up visit was achieved in 57.6% (19/33) of eyes in the GCCT group and in 54.8% (17/31) of the FCT group. The qualitative classification of preoperative and postoperative BCVAs for each group was shown in Table 3 . No significant difference was found for BCVAs in different groups preoperatively and postoperatively ( P > .05).
