The relative immunogenicity of corneal epithelium, stroma, and endothelium. The role of major histocompatibility complex antigens.

Corneal allografts have been shown to give rise to immune responses, but the role and relative importance of individual corneal cell populations in evoking such responses remain unclear. We dissected ACI (RT1a) rat corneas into separate epithelial, stromal, and endothelial components by a method that yields pure cell populations in tissue culture, and grafted these components separately to groups of fully allogeneic PVG (RT1c) recipients). Grafts of corneal stroma elicited strong cellular cytotoxic immune responses in a cell-mediated lymphocytotoxic assay, but corneal epithelium failed to generate any detectable response. Grafts of corneal endothelium alone, however, evoked a potent cellular cytotoxic response. Using congenic rats, it was found that grafts from PVG.1A (RT1a) donors to PVG (RT1c) recipients (which differ at both the RT1.A and B loci) yielded identical results. However, no corneal component graft from PVG.R1 (RT1rl) donors to PVG recipients (which differ only at RT1.A) generated a detectable immune response. Use of target lymphoblasts from congenic strains established that at least part of all responses detected were directed against class I (RT1.A) major histocompatibility complex antigens. These findings indicate that there is differential immunogenicity of specific corneal tissue components in the rat that may be further influenced by the degree of MHC disparity between donor and recipient.

Cellular immunity to heterotopic corneal allografts in the rat.

Although the immune nature of corneal allograft rejection has been recognized for over thirty years, the specific mechanisms involved in such reactions remain obscure. We investigated the cellular immune responses of PVG (RT1c) rats that were grafted with fully allogeneic ACI (RT1a) skin, ACI cornea, or PVG cornea to the chest wall or given sham grafts. Cell-mediated lymphocytotoxicity (CML) was tested at 10 days posttransplant by placing recipient spleen cells in culture with irradiated ACI stimulator cells, and six days later measuring specific lysis of 51Cr-labeled target lymphoblasts at several effector-to-target ratios. Effector cells from animals receiving allogeneic skin or cornea grafts lysed targets from the donor (ACI) strain at levels significantly (P less than 0.01) above those obtained using effectors from control (sham grafted or syngeneic corneal graft recipient) animals. Significant lysis was also seen using target cells from PVG.1A (RT1a) or PVG.R1 (RT1r1) congenic rats, which differ from recipients only at the RT1 complex and the RT1.A (class I antigen) region, respectively. Stimulator cells from PVG.1A and PVG.R1 animals also permitted detection of specific responses in secondary CML, but syngeneic PVG stimulators did not, indicating that in vitro restimulation of effector cells can be met by using stimulator cells bearing only allogeneic class I major histocompatibility complex (MHC) antigens. These results indicate that corneal allografts evoke specific cellular immune responses in the rat, and that class I MHC antigens act as effective targets for these responses.

The association of lymphocytotoxic antibodies with corneal allograft rejection in high risk patients. The Collaborative Corneal Transplantation Studies Research Group.

The Collaborative Corneal Transplantation Studies are a pair of multicenter prospective clinical trials evaluating the effectiveness of histocompatibility matching in high risk keratoplasty patients. The antigen matching study (AMS) evaluated HLA matching in patients without circulating lymphocytotoxic antibody to HLA antigens and the cross-match study (CS) evaluated the effect of using cross-match-negative donors in patients with identified circulating lymphocytotoxic antibodies to HLA antigens. Sera from 510 patients considered for enrollment in the studies were screened preoperatively for the presence of anti-class I lymphocytotoxic antibodies (LA). The 42 patients (8%) found to have detectable LA entered the CS. The 468 patients found not to have detectable LA preoperatively entered the AMS. Fifteen of the 37 transplanted CS patients were found to have donor-specific anti-class I antibody (before or after surgery). These patients were also screened for anti-class II LA and 25 had anti-class II panel reactive antibody > or = 5%. Forty-nine of the 419 transplanted AMS patients (12%) were found to have produced anti-class I LA after surgery, and in 19 patients, antibody specificities were those of donor HLA antigens. There was a significant association between the number of mismatched class I antigens and the number of donor-specific LA produced. The production of LA by AMS patients was significantly associated with reaction episodes; eighty-two percent of patients (40 of 49) with LA had reaction, compared with 63% of patients (230 of 365) without LA (P = 0.02). Likewise, production of donor-specific LA was significantly associated with immune-mediated graft failure (P = 0.025). For CS patients, there was no correlation between the production of donor-specific anti-class I or nonspecific anti-class II antibodies and graft outcome. However, the CS patients had poorer graft survival than did AMS patients at 3 years (57% vs. 66%, P = 0.01). These data demonstrate that LA, especially directed against donor class I HLA antigens following corneal transplantation in high risk patients, are associated with immune graft rejection and can be an indicator of allograft rejection.

The effect of corneal contact lenses on the oxygen tension in the anterior chamber of the rabbit eye.

The oxygen tension in the aqueous humor in the anterior chamber of rabbit eyes was measured continuously with a polarographic electrode. The normal oxygen tension in the anterior chamber was 23 +/- 2 mm Hg (mean +/- SD, n = 4). A contact lens was then placed on the cornea for at least 10 min and the drop in oxygen tension recorded. A hard polymethylmethacrylate lens reduced the oxygen tension by 16 +/- 4 mm Hg, and a larger hydroxyethylmethacrylate soft lens (Soflens) decreased oxygen tension by 17 +/- 4 mm Hg (mean +/- SD, n = 4). Comparable statistically significant decreases were seen with the Permalens, Polycon II, and Silcon lenses. Only the elastofilcon A lens (Silsoft) did not decrease the oxygen tension in the anterior chamber significantly.

Effects of ultraviolet radiation on corneal epithelial metabolism.

Acute exposure to ultraviolet light to 257 nm wavelength produced a photokeratitis associated with characteristic metabolic alterations of corneal epithelial metabolism in rabbits. Significant increases in corneal hydration occurred simultaneously with decreased corneal epithelial glycogen content, but adenosine triphosphate content and enzyme activity of epithelial extracts were not affected despite clinical and histological damage to the corneal epithelium. The pattern and time course of ultraviolet damage to the cornea are distinctly different from those of other forms of trauma.

Inhibition of human subconjunctival fibroblast proliferation by immunotoxin.

The ability to target proliferating cells is important for agents used to modulate wound healing by decreasing the growth of fibroblasts. Proliferating cells are known to express increased numbers of transferrin receptors and have increased receptor turnover. 454A12 Mab-rRA, an immunotoxin containing anti-human transferrin receptor monoclonal antibody conjugated to recombinant ricin A chain, was shown to inhibit the proliferation of human subconjunctival fibroblasts in vitro. A dose-related reduction of cell counts was observed in proliferating cells. More than 90% inhibition was achieved with an immunotoxin concentration of 10 ng/ml per 20,000 cells plated. In contrast, confluent fibroblasts were markedly less sensitive to the immunotoxin at equivalent concentrations. Comparative experiments demonstrated that 5-fluorouracil has less specificity for proliferating cells, with significant death of confluent fibroblasts at high drug concentrations.

Results of penetrating keratoplasty by residents.

There is little information regarding the results of penetrating keratoplasty for the novice transplant surgeon. We reviewed 63 consecutive cases of penetrating keratoplasty performed by senior residents at our institution. Of 59 cases with a 1-year follow-up, 76% remained clear; the mean astigmatism was 4.15 diopters, and the median visual acuity was 20/50. The two major postoperative complications were graft rejection and increased intraocular pressure, which occurred in 22% and 16% of cases, respectively. In a well-supervised setting, residents can be introduced to the basic principles of keratoplasty and achieve results comparable with those reported in the literature.

Ocular cicatricial pemphigoid with granular IgA and complement deposition.

A 65-year-old woman developed chronic redness of both eyes, and, over the ensuing 2 1/2 years, she had progressive conjunctival scarring with symblepharon formation. Other mucosal surfaces were not involved. A conjunctival biopsy specimen 12 months following onset of her disease showed areas of epithelial separation from the basement membrane zone as well as subepithelial chronic inflammation and scarring. Two years later, another conjunctival biopsy specimen showed granular deposition of IgA and C3 along the epithelial basement membrane zone using direct immunofluorescent staining. Electron microscopy confirmed the presence of deposits that were morphologically consistent with antigen-antibody complexes. These findings suggest that antigen-antibody (IgA) immune-complex deposition may provide an alternative pathogenetic mechanism to basement membrane zone autoantibody formation for development of progressive conjunctival scarring.

The effect of silicone oil on the cornea.

Intraocular silicone oil tamponade is being used with increasing frequency as an adjunct to conventional vitrectomy techniques in the treatment of proliferative vitreoretinopathy. When used in patients with aphakia, corneal decompensation is a common long-term complication. In this study, we evaluated the clinical and morphologic changes that occurred in the corneas of 14 rabbits and seven cats when the anterior chamber was filled with silicone oil. Within six days, wide-field specular microscopy showed a 40% reduction in endothelial density in the area of the silicone oil bubble in both groups. Progressive stromal thinning occurred in the rabbit cornea, with gradual development of a retrocorneal membrane at the junction of silicone-endothelial cell contact. In contrast, persistent stromal edema, peripheral vascularization, irregular plaques on the endothelium, and eventual epithelial ulceration and corneal thinning occurred in cat eyes.

Corneal complications after closed vitrectomy through the pars plana.

A retrospective study of 150 patients who underwent closed vitrectomy through the pars plana evaluated the incidence and character of postoperative corneal complications. Corneal changes developed in 43% of patients. Thirty patients had significant corneal complications, half of which resolved with treatment. Fourteen patients had complications that necessitated corneal surgery or resulted in loss of useful vision. Diabetic patients were at greatest risk for corneal problems, particularly postoperative epithelial defects. Intraoperative lensectomy was the only other factor that increased the corneal complication rate. The importance of early recognition and therapy for corneal changes after vitrectomy becomes apparent.

The effect of perfluoropropane on the cornea in rabbits and cats.

The effects of perfluoropropane gas on the cornea were compared with those of sulfur hexafluoride gas following injection into the anterior chamber of rabbits and cats. Injection of 0.15 mL of gas produced a perfluoropropane bubble that lasted 22 days compared with a sulfur hexafluoride bubble that lasted seven days. The sulfur hexafluoride bubble produced corneal edema for as long as the gas was present. With perfluoropropane, corneal edema persisted even after the gas bubble disappeared. In another group of cats, multiple sequential injections of sulfur hexafluoride to maintain a gas bubble of duration comparable with the perfluoropropane bubble also produced edema as long as gas was present. Clinical corneal edema, endothelial fibrin deposition, endothelial opacities, and retrocorneal membrane were observed by slit-lamp and light, scanning, and transmission electron microscopic examinations of the corneas exposed to perfluoropropane but not of those exposed to sulfur hexafluoride. Prolonged corneal endothelial contact by perfluoropropane results in corneal edema due to endothelial dysfunction, which persists in cats probably because feline endothelium is less capable of regeneration than that of rabbit.

Immunohistochemical analysis of the pathogenesis of posterior polymorphous dystrophy.

The pathogenesis of posterior polymorphous dystrophy was analyzed by immunohistologic methods. Sections of corneal buttons from two patients undergoing transplantation owing to posterior polymorphous dystrophy were stained with 2B4.14.1, a monoclonal antibody that reacts with human corneal endothelium, and with a cocktail of antihuman cytokeratin monoclonal antibodies that do not react with normal corneal endothelium. Single-stained sections revealed a variegated, intermittent staining pattern of antibody reactive and nonreactive cells. Double-stained sections revealed some cells that stained with only one of the antibodies and many cells that stained with both antibodies. The presence of cells staining positively for both 2B4.14.1 antigen and cytokeratins supports the hypothesis that the cytokeratin-expressing epithelial-like cells found in corneas with posterior polymorphous dystrophy arise via a metaplastic process in which the phenotype of endothelial cells becomes progressively abnormal.

Factors related to corneal epithelial complications after closed vitrectomy in diabetics.

In a series of 30 vitrectomy patients, postoperative corneal complications developed in 14 (47%). Thirteen (92%) of the 14 patients with complications were diabetic. Other factors that were positively correlated with the development of corneal complications included decreased corneal sensitivity, intraoperative lensectomy, and intraoperative epithelial debridement. The corneal complications involved the epithelium in 11 (78%) of the 14 complicated cases, with involvement of the endothelium (as manifested by stromal edema) in five (35%) cases. The diabetic human corneal epithelium contains considerable levels of sorbitol and fructose, which indicates the probable presence of the sorbitol pathway. As in the lens, this pathway may lead to osmotic changes that make the diabetic corneal epithelium more vulnerable to damage intraoperatively. Postoperatively, reepithelialization is delayed because of a probable failure of adhesion of cells to the underlying basement membrane.

Beneficial effects of histocompatibility in high-risk corneal transplantation.

Forty-six consecutive patients at high risk for corneal graft rejection because of significant vascularization or previous graft rejection received transplanted corneas from donors who were selected on the basis of a negative lymphocyte crossmatch, ABO antigen blood group compatibility, and maximized HLA-A and HLA-B antigen matching. Graft survival in this study group was compared to that in a retrospective control group of 71 consecutive high-risk transplants performed in patients receiving corneas obtained randomly. We found no differences between groups in terms of demographic factors, cause of primary corneal disease, or the incidence of previous graft rejection. However, graft survival in the control group was only 49% (35 of 71) vs 83% (38 of 46) for the study group (P less than .001) with mean follow-up times of 29.6 and 15.5 months, respectively. All graft failures in the study group were in recipients with poorly matched HLA antigens. Patients with evidence of humoral presensitization to HLA antigen had twice the rejection rate of unsensitized recipients (three of 11 vs five of 35).

The expression of HLA antigens by cells in the human cornea.

Although antigens of the human major histocompatibility complex, HLA, appear to be involved in the rejection of corneal allografts, the expression of these antigens by cells of the human cornea remains controversial. Using sensitive immunoperoxidase techniques, we readily demonstrated class I HLA antigens on corneal epithelial, stromal, and endothelial cells, regardless of donor age. Moreover, the expression of class I antigens by epithelium increased markedly from the central to the peripheral cornea. Finally, class II HLA antigens were found on cells scattered throughout both central and peripheral epithelium as well as the stroma. A cornea obtained from a donor of known HLA type expressed all expected and no unexpected polymorphic HLA determinants in the cornea, in a pattern similar to that seen for monomorphic HLA determinants.

Lens capsular cyst.

A 33-year-old man had a unique lens capsular cyst that became clinically evident eight years after a penetrating ocular injury in which the lens capsule was disrupted. Microscopic examination indicated that the cyst may have arisen from displaced lens capsule and epithelial cells that became separated from the lens.

Temporary keratoprosthesis for use during pars plana vitrectomy.

We have developed a series of temporary keratoprostheses that permit closed pars plana vitrectomy in eyes with abnormal corneas. These devices are placed into a trephined opening in the cornea, and provide a clear, stable view of the intraocular contents. Immediately after vitrectomy, the keratoprosthesis is replaced with a corneal graft. We have used this technique successfully in four eyes with opaque or severely traumatized corneas.

Epithelial cyst of the iris following penetrating keratoplasty.

In a 35-year-old patient with keratoconus a cyst of the iris was first observed 22 months after routine penetrating keratoplasty. The lesion was completely excised and shown by light and transmission electron microscopy to be composed of conjunctival epithelium organised as a single large cyst containing serous material and occasional free epithelial cells. The management of epithelial implantation cysts is discussed. When indicated, complete surgical excision without ruputure is preferred to avoid possible intraocular dissemination of viable epithelial cells that might result in further complications.

Inhibition of proliferating lens epithelium with antitransferrin receptor immunotoxin.

We investigated the effect of an antitransferrin receptor immunotoxin (454A12-rRA) on proliferating human and baboon lens epithelium in vitro. Human and baboon lens epithelial cells grown in modified TC-199 medium at 35 degrees Celsius in 7% CO2 were seeded in 24 well plates at a density of 17,500 cells/ml to 40,000 cells/ml. The cells were exposed to various concentrations of 454A12-rRA for seven days. The sensitivity of proliferating human lens epithelium to 454A12-rRA was dependent on the dose, with a 60% to 70% reduction in cell counts at immunotoxin concentrations of 100 ng/ml and above. The immunotoxin had no significant effect on baboon lens epithelium in vitro, which suggests that it is specific for human tissue. By preventing the proliferation of human lens epithelial cells, immunotoxin 454A12-rRA may be useful in the management of posterior capsule opacification after planned extracapsular cataract surgery.

Evaluation of the effect of lissamine green and rose bengal on human corneal epithelial cells.

PURPOSE: To examine the effects of lissamine green and rose bengal on proliferating human corneal epithelial (HCE) cells in vitro. METHODS: HCE cells derived from explants of discarded corneoscleral rims were cultured by the standard technique. Experimental cells were exposed to 1, 0.5, or 0.1% of either lissamine green or rose bengal for 10 min while control cells were exposed to a phosphate buffer solution (PBS). RESULT: Cell viability was 92% greater for 1% lissamine green (p = 0.013) and 81.2% greater for 0.5% lissamine green (p = 0.006) compared to 1 and 0.5% rose bengal, respectively. The difference between the effect of 0.1% rose bengal and 0.1% lissamine green on cell viability was not statistically significant (p = 0.83). Rose bengal staining of HCE cells was immediate and readily detectable with unaided eyes at all three concentrations, whereas no observable staining of healthy HCE cells was noted with lissamine green. CONCLUSION: Rose bengal adversely affects HCE cell viability and stains normal proliferating HCE cells in contrast to lissamine green, which exhibited neither of these characteristics. Therefore, we recommend the use of lissamine green over rose bengal in evaluating ocular surface disorders.