Laser Photocoagulation of Expe~entruCorneru Stromru Vascularization

Efficacy and Histopathology

V erinder S. Nirankari, MD, 1 Lalit Dandona, MD, MPH, 1

Merlyn M. Rodrigues, MD, PhD2

Background: Conventional treatment of corneal stromal vascularization is often inadequate. The authors developed a rabbit model of corneal stromal vascularization, treated it with laser photocoagulation, and then studied the histopathology.

Methods: A reproducible model of corneal stromal vascularization was developed in albino rabbits by injecting sodium hydroxide into the corneal stroma. Corneal stromal vascularization was produced in both eyes of 13 rabbits, and treated after stabilization at 5 weeks with 577 -nm yellow dye laser in 1 eye of each rabbit. Seven rabbits were followed for 6 months with corneal angiography and photography, and the corneal stromal vascularization quantified with a grid. The other 6 rabbits were killed at 1, 4, 8, 24, 48 hours, and 6 days after laser photocoagulation and examined by light and trans­mission electron microscopy.

Results: Stable corneal stromal vascularization was observed in the anterior and midstroma for at least 6 months in the model. Laser photocoagulation reduced corneal stromal vascularization significantly compared with the controls (P ~ 0.05), resulting in 40.7% ± 5.0%, 45.3% ± 3.3%, and 34.9% ± 5.2% (mean± standard error of the mean) reduction at 2, 4, and 6 months, respectively. Maximum inflammatory cell infiltrates were detected at 8 hours after laser photocoagulation, which diminished markedly at 6 days. The stroma of unlasered eyes showed no inflammatory cells and considerably more patent blood vessels than the lasered eyes. In the lasered eyes, transmission electron microscopy showed damaged vascular endothelial cells, extravasated eryth­rocytes, haphazardly arranged collagen fibrils, thrombus formation, and ghost vessels in the stroma. No damage was observed in the deep corneal stroma or endothelium in the lasered eyes.

Conclusion: Laser photocoagulation is effective in reducing corneal stromal vas­cularization in this model for at least 6 months. It does not damage the deeper stroma or endothelium. Ophthalmology 1 994; 100:111-118

Originally received: June 25, 1992. Revision accepted: September 4, 1992. 1 Cornea Service, Department of Ophthalmology, University of Maryland School of Medicine, Baltimore. 2 Laboratory of Eye Pathology & Cell Biology, Department of Ophthal­mology, University of Maryland School of Medicine, Baltimore.

It has been recognized for approximately 150 years that the human cornea is normally avascular but it can become vascular in certain pathologic situations. 1 Corneal vas­cularization is a "response to a call for help by a tissue in difficulty."2 In humans, this can occur in a variety of con-

Supported in part by a grant from Research to Prevent Blindness, New York, New York (Dr. Rodrigues).

This work was included in a thesis accepted as requirement for mem­bership to the American Ophthalmological Society (Dr. Nirankari).

Reprint requests to Verinder S. Nirankari, MD, Cornea Service, De­partment of Ophthalmology, University of Maryland Hospital, Baltimore, MD 21201.

111

Ophthalmology Volume 100, Number 1, January 1993

ditions, including corneal graft rejection, infections, con­tact lens wear, trauma, burns, vasculitides, metabolic dis­orders, toxins, and nutritional deficiency states. Although corneal vascularization represents a defense mechanism against disease or injury, its presence can lead to a decrease in vision due to the loss of corneal transparency.2 The presence of corneal vascularization increases the chance of corneal graft rejection. J-s Vascularized corneas also may be more susceptible to lipid deposition, resulting in corneal opacity.6 Vascularization may interfere with transdiffer­entiation of the migrating conjunctival epithelium during healing of the corneal epithelial defects, which may lead to impaired barrier function of the corneal epithelium.7

In this background, persistent corneal vascularization is an undesirable event. The conventional treatment of cor­neal vascularization, including suture removal if consid­ered to be the inciting agent and corticosteroid therapy,8

•9

often is inadequate. Corneal graft rejection is associated with stromal vas­

cularization, and the other sequelae of corneal vascular­ization such as edema and scarring are more extensive with stromal than with superficial vascularization. There­fore, we created a reproducible model of corneal stromal vascularization in New Zealand albino rabbits. We treated this corneal stromal vascularization with 577-nm yellow dye laser photocoagulation to evaluate its efficacy. We also studied the histopathology after laser photocoagula­tion to assess the changes that occurred in the cornea as a result of this treatment.

Materials and Methods

This study conformed to the ARVO Resolution on the Treatment of Animals, and was approved by the Animal Care and Use Committee ofthe University of Maryland School of Medicine.

Corneal stromal vascularization was produced in New Zealand albino rabbits by the following method. The rab­bit was anesthetized with an intramuscular injection of 35 mg/kg body weight of ketamine hydrochloride and 5 mg/kg body weight of xylazine hydrochloride. The eye was proptosed with a latex sleeve after giving a drop of proparacaine hydrochloride 0.5% topically as local an­esthetic. The corneal injection site was selected between the attachment of the extraocular muscles to avoid in­growth of vessels from the muscles to the injection site. While viewing through an operating microscope, the exact injection site was marked with calipers on the cornea 2 mm inside the limbus. A 30-gauge needle attached to a tuberculin syringe containing 0.1 N sodium hydroxide was introduced into the corneal stroma at the marked site with the bevel facing up. Then, 50 JLL of sodium hydroxide was gradually injected into the corneal stroma while care­fully observing the spreading haziness in the cornea. Spe­cial precaution was taken to avoid entering the anterior chamber of the eye and spillage of sodium hydroxide on the corneal surface. After the injection, the needle was immediately withdrawn, and the cornea was irrigated with

112

balanced salt solution. The eyelids were then brought to their normal position, and a drop of gentamicin sulfate 0.3% was given topically.

After the sodium hydroxide injection, the rabbit eyes were examined at least once a week. Corneal stromal vas­cularization was documented with photography and cor­neal angiography after injecting 1.5 ml of 10% fluorescein dye in the central ear vein of the rabbit. Fluorescein an­giography has been found to be useful in elucidating cor­neal vasculature. 10 Corneal stromal vascularization was observed to stabilize at 4 to 5 weeks after the injection of sodium hydroxide. The presence of blood vessels con­taining erythrocytes in the corneal stroma was confirmed histologically in our model.

Seven rabbits with corneal stromal vascularization produced in both eyes had one eye randomly chosen for treatment with laser photocoagulation after stabilization of corneal stromal vascularization at 5 weeks after the sodium hydroxide injection. The 577-nm yellow dye laser was chosen to treat corneal stromal vascularization be­cause of the relatively higher absorption of this wavelength by oxyhemoglobin and reduced hemoglobin. 11 Fluores­cein angiograms were studied before applying the laser treatment to better define the blood vessels in the cornea. The Abraham iridotomy laser lens was used for the laser treatment because it helps in stabilization of the globe, magnification of the corneal vessels, and focusing laser energy on the corneal vessels. The settings used for the laser treatment were 400-m W power, 50-J,Lm beam spot size, and 0.05- to 0.2-second pulse duration. The laser treatment was started paracentrally and continued pe­ripherally. If the afferent and efferent vessels could be dis­tinguished, the laser treatment was first applied to the afferent and then to the efferent vessels. Laser treatment to the central corneal region was avoided to prevent any possible damage to the visual axis. The central vessels regressed spontaneously because their blood supply was cut off when the paracentral and peripheral vessels were photocoagulated. Between 300 and 600 laser spots were applied, depending on the extent of corneal stromal vas­cularization.

The lasered and the unlasered control eyes in these 7 rabbits were followed for 6 months. Photographic docu­mentation of corneal stromal vascularization was done at 1- to 4-week intervals during this follow-up period. Corneal stromal vascularization was quantified from slides of the corneal photographs, using a grid consisting of 1 X 1 mm squares. This grid was superimposed on the slide, and the number of squares on the cornea that were at least half filled with blood vessels were counted. The area of these squares was added, and the magnification factor was taken into account to calculate the area of corneal stromal vascularization. The corneal stromal vasculariza­tion area in the seven rabbits and the percent change from baseline were expressed as mean ± standard error of the mean. The paired t test was used to determine the statis­tical significance of the differences between the lasered and control eyes just before laser treatment and at 2, 4, and 6 months of follow-up. P s 0.05 was considered sig­nificant.

Nirankari et al · Corneal Stromal Vascularization

Six other rabbits with corneal stromal vascularization produced in both eyes had one eye randomly chosen for treatment with laser photocoagulation after stabilization of corneal stromal vascularization at 5 weeks after the sodium hydroxide injection. These rabbits were killed at 1, 4, 8, 24, 48 hours, and 6 days after laser photocoagu­lation, respectively, for study under light and transmission electron microscopy. The unlasered eyes of these rabbits served as controls. Paraffin sections of the cornea were stained with hematoxylin-eosin, periodic acid-Schiff, Masson trichrome, and Van Gieson stains for light mi­croscopy, in a masked fashion. Epoxy resin sections of the cornea were studied by transmission electron micros­copy.

Results

In our model, corneal stromal vascularization was pro­duced consistently in the peripheral and paracentral areas of the quadrant injected with sodium hydroxide (Figs 1 and 2). The central cornea was uninvolved, and none of the rabbits were blinded by these experiments. Corneal stromal vascularization was localized in the anterior and midstroma histologically (Fig 3).

Infiltration of the corneal stroma with leukocytes, mainly neutrophils, was noted at 8 hours after the sodium hydroxide injection. Increased vascularity of the anterior and midstroma was noted within a few days of the injec­tion. Leukocytes could not be detected in the corneal stroma by 2 to 3 weeks after the injection. Blood vessel growth stabilized at 4 to 5 weeks after the injection. His­topathology at this time showed moderate fibrosis and many patent blood vessels containing erythrocytes in the anterior and midstroma (Fig 3).

At 1 hour after laser treatment, the stroma showed edema and relatively fewer blood vessels (Fig 4) than the unlasered control (Fig 3). Scattered inflammatory cells were observed at 4 hours after laser treatment (Fig 5), and greater numbers were observed at 8 hours (Fig 6). His­topathology at 4 hours showed engorged and ruptured corneal blood vessels (Fig 5). This may be related to our technique of starting the laser treatment paracentrally and proceeding peripherally, which may lead to temporary engorgement of the central vessels and their subsequent rupture. Since this type of hemorrhage was noted in only one specimen, we think it is an uncommon occurrence. The inflammatory cells were less at 24 and 48 hours than at 8 hours. A few ghost vessels, devoid of erythrocytes and therefore believed to be nonfunctioning, were seen at 48 hours (Fig 7) and 6 days (Fig 8). The inflammatory cells had diminished further at 6 days. Blood vessels were fewer, and fibroblastic proliferation was pronounced in alllasered corneas compared with controls. No damage was detected in the deep stroma or endothelium in any of the lasered corneas (Figs 4 to 8).

The presence of fewer patent blood vessels, pronounced fibroblastic proliferation, and inflammatory cells in the anterior and midstroma oflasered corneas enabled all six

of these cases to be distinguished from controls by one masked observer and five of these six to be distinguished by another masked observer.

In the control corneas, transmission electron micros­copy showed smooth contoured stromal blood vessels lined with normal endothelial cells (Fig 9). These endo­thelial cells contained regular nuclei and cytoplasmic or­ganelles, and a uniform pattern of the collagen fibrils was observed (Fig 9). No inflammatory cells were seen in the corneal stroma of control eyes.

In the laser treated corneas, transmission electron mi­croscopy showed damaged endothelial cells, with cyto­plasmic edema and irregular nucleus, lining the vessels in the stroma (Fig 10). Granulocytes, other inflammatory cells, extravasated erythrocytes, and edema were observed in the stroma, and neutrophils were seen in the vessel lumen (Fig 10). A haphazard array of collagen fibrils was noticed in the stroma oflasered corneas (Figs 10 and 11 ). Distorted erythrocytes in the stroma were noticed in one section of a lasered cornea (Fig 11 ). Thin and empty ghost vessels were seen in the stroma (Fig 12). Granular fibrinous material was noted around closely packed erythrocytes in a vessel lumen, suggesting thrombus formation after laser photocoagulation (Fig 13 ).

Corneal stromal vascularization in this model was sta­ble for at least 6 months offollow-up (Fig 14). Just before laser photocoagulation, the area of corneal stromal vas­cularization in the eyes randomized for treatment was 24.6 ± 3.5 mm2 (range, 12 to 40 mm2

). The comparable area in the control eyes was 19.3 ± 4.1 mm2 (range, 10 to 40 mm2

). The difference between the groups was not statistically significant (P = 0.14). The percent decrease from baseline in the area of corneal stromal vasculariza­tion was more in the lasered eyes than in the control eyes at all time intervals (Fig 14). At 2 months, there was a 40.7% ± 5.0% decrease in lasered eyes and a 11.1% ± 6.2% increase in control eyes (P = 0.0002). At 4 months,

Figure 9. Electron micrograph of an unlasered rabbit cornea 5 weeks after the intrastromal injection of sodium hydroxide. The blood vessel shows a uniform lining of endothelial cells and the lumen contains eryth­rocytes. The surrounding collagen displays a uniform arrangement of fibrils. No inflammatory cells are seen in or around the vessel. (Original magnification, X3300).

113

Ophthalmology Volume 100, Number 1, January 1993

there was a 45.3% ± 3.3% decrease in lasered eyes and a 16.9% ± 8.2% decrease in control eyes (P = 0.04). At 6 months, there was a 34.9% ± 5.2% decrease in lasered eyes and a 3.0% ± 11.2% decrease in control eyes (P = 0.05).

Discussion

Persistent corneal stromal vascularization is undesirable because it can cause decreased vision due to edema and scarring, 2 increase the risk of graft rejection, J-s and result in lipid keratopathy.6 Topical, and sometimes periocular, corticosteroids are currently the mainstay of therapy for corneal stromal vascularization, but they may not always be completely effective. 8•

9 A major concern about corti­costeroids is that their long-term topical use in the eye may have side effects, including cataract, glaucoma, su­perinfection, and herpes simplex recurrence. Topical nonsteroidal anti-inflammatory drugs and cyclosporin-A also have been used for treating corneal stromal vascu­larization, but are not used clinically because of lack of consistent effectiveness or the possibility of serious side effects. 12

-14 Invasive treatments for corneal vascularization

that have been investigated but found to be clinically in­applicable include irradiation, cryotherapy, heat cautery, scar tissue barrier, and excision of vessels. 15

-18

The normal avascular nature of the cornea gives it rel­ative immune privilege from donor graft rejection. 19

•20

Corneal stromal vascularization provides the afferent and efferent limbs for the immune rejection of the graft to take place. Corneal stromal vascularization may reduce the success rate of corneal grafts to as low as 35% from 85% to 95% in avascular grafts.21

-23 A total of 41,393 cor­

neal grafts were performed in the United States during

1991 (data provided by Eye Bank Association of America, Washington, DC). Approximately 10% to 20% of all cor­neal grafts fall into the high-risk category for failure. 24 A large proportion of these high-risk grafts have persistent corneal stromal vascularization in the recipient cornea. Therefore, corneal stromal vascularization in a cornea being considered for grafting or corneal stromal vascu­larization into a donor corneal graft are both undesirable. The question whether matching for HLA antigens and lymphocytotoxic antibodies can reduce the incidence of corneal graft rejection in high-risk patients, and whether this would be cost-effective, is being investigated by the Collaborative Corneal Transplantation Studies.24 Another approach that may reduce the risk of corneal graft rejec­tion due to corneal stromal vascularization would be to ablate the vessels with laser photocoagulation.

Laser treatment of corneal vascularization after intra­venous injection of photosensitizing dyes rose bengal and dihematoporphyrin ether has been investigated.Z5-

28

However, this approach has not been applied to humans so far because of potential complications. Chorioretinal scarring has been seen in some rabbit eyes with the method using rose bengal probably because of absorption of laser energy by the choroidal and retinal vessels.27 Transient abnormalities of some liver enzymes and electrolytes also have been noted after intravenous injection of rose bengal in rabbits. 26 Corneal stromal vascularization consists of immature vessels that are leaky (Fig 2). Therefore, leakage of a photosensitizing dye into the corneal stroma could lead to a photochemical reaction and nonspecific damage on application of laser energy. Anterior corneal stromal damage has been noticed in some rabbit eyes with corneal vascularization treated with laser photocoagulation after intravenous injection of rose bengal. 27 A striking decrease in the number of stromal keratocytes, and basophilia and vacuoles in the lens anteriorly, have been noted in mouse

Top left, Figure 1. Stable corneal stromal vascularization in a rabbit eye 5 weeks after the intrastromal injection of sodium hydroxide.

Top right, Figure 2. Corneal fluorescein angiogram 34 seconds after dye injection showing corneal stromal vessels in the same rabbit eye as in Figure 1.

Second row left, Figure 3. Light micrograph of a rabbit cornea 5 weeks after the intrastromal injection of sodium hydroxide. Arrows indicate patent blood vessels containing erythrocytes in the anterior and mid-stroma. (Hematoxylin-eosin; original magnification, X200).

Second row right, Figure 4. Light micrograph of a rabbit cornea 1 hour after laser photocoagulation for corneal stromal vascularization. Notice the presence of fewer blood vessels than in the unlasered control eye in Figure 3. Arrows indicate inflammatory cells in the superficial stroma. The deep stroma and endothelium are intact. (Hematoxylin-eosin; original magnification, XlOO).

Third row left, Figure 5. Light micrograph of a rabbit cornea 4 hours after laser photocoagulation for corneal stromal vascularization. Inflammatory cells, dilated blood vessels, and hemorrhage are present in the anterior corneal stroma. (Hematoxylin-eosin; original magnification, XlOO).

Third row right, Figure 6. Light micrograph of a rabbit cornea 8 hours after laser photocoagulation for corneal stromal vascularization. Many inflammatory cells and irregular collagen are seen in the anterior and mid-corneal stroma. (Hematoxylin-eosin; original magnification, XlOO).

Bottom left, Figure 7. Light micrograph of a rabbit cornea 48 hours after laser photocoagulation for corneal stromal vascularization. Inflammatory cells still persist in the anterior and midcorneal stroma, but are less than at 8 hours in Figure 6. Arrows indicate empty ghost vessels, those without erythrocytes. (Hematoxylin-eosin; original magnification, X 100).

Bottom right, Figure 8. Light micrograph of a rabbit cornea 6 days after laser photocoagulation for corneal stromal vascularization. Inflammatory cells in the corneal stroma are less than at 48 hours in Figure 7. Arrows indicate empty ghost vessels interspersed in areas of stromal fibrosis. (Hematoxylin-eosin; original magnification, XlOO).

114

Nirankari et al · Corneal Stromal Vascularization

115

Ophthalmology Volume 100, Number 1, January 1993

Figure 10. Electron micrograph of a rabbit cornea 24 hours after laser photocoagulation for corneal stromal vascularization. Endothelial cells lining the blood vessel display cytoplasmic edema and irregular nucleus, indicating damage. Neutrophils occupy most of the vessel lumen. There are scattered stromal granulocytes (arrows) and an extravasated erythrocyte (arrowhead). Collagen fibrils in the stroma are haphazardly arranged. (Original magnification, X4300).

eyes that had laser treatment for corneal vascularization after intravenous injection ofdihematoporphyrin ether.28

In this background, we investigated the efficacy oflaser photocoagulation, without prior injection of any photo­sensitizing dye, in the treatment of corneal stromal vas­cularization and studied its effect on the cornea with light and electron microscopy in rabbits. We developed a re­producible model of corneal stromal vascularization in New Zealand albino rabbits by injecting 50 J.~,l of 0.1 N sodium hydroxide into the corneal stroma. This corneal stromal vascularization stabilized 4 to 5 weeks after the injection. The presence of stromal blood vessels in our model was established by corneal angiography and his-

Figure 11. Electron micrograph of a rabbit cornea 4 hours after laser photocoagulation for corneal stromal vascularization. The corneal stroma displays a marked disarray of collagen fibrils. Fibroblasts (curved arrow) and neutrophils (thick arrow) are scattered in the stroma. There is an accumulation of distorted extravasated erythrocytes (asterisk). (Original magnification, X3300).

116

Figure 12. Electron micrograph of a rabbit cornea 24 hours after laser photocoagulation for corneal stromal vascularization. The blood vessel lumen is empty. There are periluminal fibroblasts (curved arrow) and adjacent granulocytes (thick arrow). Collagen fibrils in the stroma are haphazardly arranged. (Original magnification, X3300).

tologic examination. This corneal stromal vascularization was found to remain stable for at least 6 months. In a randomized masked study, we found that photocoagu­lation of corneal stromal vascularization in our model with the 577-nm yellow dye laser caused significant re­duction of corneal stromal vascularization in the treated eyes compared with the control eyes over a follow-up pe­riod of 6 months.

We used the 577-nm yellow dye laser because this wavelength has been reported to be absorbed relatively more by oxyhemoglobin and reduced hemoglobin than is the argon blue-green 514.5-nm wavelength. 11 Therefore, theoretically, it appears that less total energy would be needed for ablation of corneal stromal vascularization with the 577-nm laser than the 514.5-nm laser, and

Figure 13. Electron micrograph of a rabbit cornea 4 hours after laser photocoagulation for corneal stromal vascularization. The endothelial cell lining the blood vessel has cytoplasmic edema and irregular nucleus. The vessel lumen is packed with erythrocytes and a granular fibrinous material (asterisk), suggesting thrombus formation. Two fibroblasts are seen touching the vessel. (Original magnification, X9000).

Nirankari et al · Corneal Stromal Vascularization

0

1 120

a _ 100 ~ '#. • 0

~ ~ 80 iii ~

" iii ~ ~ 60 8 ~

~ J 40 .i-

20

p: 0.0002 P"'0.04

Time after Laser Treatment (Months)

p: 0.05

-o-- Control eyes

~ Lasered eyes

Figure 14. Percentage change in the area of corneal stromal vascularization in 7 control and 7 laser treated rabbit eyes over 6 months follow-up. The vertical bars represent 1 standard error of the mean on each side of the mean. The P values are obtained by using the paired t test.

thereby damage to the surrounding tissue could be min­imized.

Histopathologic study of the rabbit corneas by light and electron microscopy within 6 days after laser pho­tocoagulation of corneal stromal vascularization showed damage to the endothelial cells lining the stromal vessels and the presence of thrombi in blood vessels. Alllasered corneas had fewer patent stromal vessels than the control corneas. Lasered corneas also showed presence of edema, inflammatory cells, fibroblastic proliferation, extravasated erythrocytes, and a disarray of collagen fibrils in the stroma. The stromal edema and inflammatory cells had diminished markedly 6 days after laser treatment. None of the lasered corneas showed any damage to the deep corneal stroma or the endothelium.

Argon laser photocoagulation of corneal vasculariza­tion has been reported to be effective under certain cir­cumstances in rabbits and humans.29

-33 Recently, success

in some categories of patients has been reported in treating corneal vascularization with the 577-nm yellow dye laser photocoagulation. 34 In this study, we have described a reproducible model of corneal stromal vascularization in rabbits. We found the 577-nm yellow dye laser to beef­fective in significantly reducing the extent of corneal stromal vascularization confirmed by clinical and histo­pathologic study. We also have described the effect oflaser treatment on corneal stromal vascularization by light and electron microscopy.

References

I. Toynbee J. Researchers, tending to prove the non-vascularity and the peculiar uniform mode of organization and nutrition of certain animal tissues, viz. articular cartilage, and the cartilage of the different classes of fibro-cartilage; the cornea, the crystalline lens, and the vitreous humour; and the epi­dermoid appendage. Philos Trans R Soc Lond 1841; Pt.1: 159-92.

2. Duke-Elder S, ed. System of Ophthalmology. Vol. VIII: Diseases of the Outer Eye. Pt.2: Cornea and Sclera. St. Louis: CV Mosby, 1965;676-91.

3. Khodadoust AA, Silverstein AM. Transplantation andre-

jection of individual cell layers of the cornea. Invest Ophthalmol 1969;8: 180-95.

4. Langston RHS, Pavan-Langston D. Penetrating keratoplasty for herpetic keratitis: decision-making and management. Int Ophthalmol Clin 1975;15(4):125-40.

5. Herbort CP, Matsubara M, Nishi M, Mochizuki M. Pene­trating keratoplasty in the rat: a model for the study of im­munosuppressive treatment of graft rejection. Jap J Ophthalmol 1989;33:212-20.

6. Cogan DG, Kuwahara T. Lipogenesis of cells of cornea. Arch Pathol 1955;59:453-6.

7. Huang AJW, Tseng SCG, Kenyon KR. Alteration of epi­thelial paracellular permeability during corneal epithelial wound healing. Invest Ophthalmol Vis Sci 1990;31 :429-35.

8. Ashton N, Cook C, Langham M. Effect of cortisone on vas­cularization and opacification of the cornea induced by al­loxan. Br J Ophthalmoll951;35:718-24.

9. Michaelson IC. Effect of cortisone upon corneal vascular­ization produced experimentally. Arch Ophthalmol1952;47: 459-64.

10. Billings EE, Nirankari VS. Corneal angiography of ghost vessels useful in evaluating risks of corneal surgery. J Ophthalmic Photogr 1981 ;4( 1 ):6-1 0.

II. L'Esperance FA Jr. Clinical photocoagulation with the or­ganic dye laser. A preliminary communication. Arch Ophthalmoll985;103:1312-6.

12. Deutsch TA, Hughes WF. Suppressive effects of indometh­acin on thermally induced neovascularization of rabbit cor­neas. Am J Ophthalmol 1979;87:536-40.

13. Haynes WL, Proia AD, Klintworth GK. Effect of inhibitors of arachidonic acid metabolism on corneal neovasculari­zation in the rat. Invest Ophthalmol Vis Sci 1989;30: 1588-93.

14. Belin MW, Bouchard CS, Frantz S, Chmielinska J. Topical cyclosporine in high-risk corneal transplants. Ophthalmol­ogy 1989;96: 1144-50.

15. Ainslie D, Snelling MD, Ellis RE. Treatment of corneal vascularization by Strontium 90 beta plaque. Clin Radio! 1962;13:29.

16. Lavergne G, Calmant lA. Comparative study of the action of thiotepa and triamcinolone on corneal vascularization in rabbits. Br J Ophthalmol 1964;48:416-22.

17. Mayer W. Cryotherapy in corneal vascularization. Arch Ophthalmol 1967;77:637-41.

18. Ey RC, Hughes WF, Bloome MA, Tallman CB. Prevention of corneal vascularization. Am J Ophthalmol 1968;66: 1118-31.

19. Billingham RE, Boswell T. Studies on the problem of corneal homografts. Proc R Soc Lond [Bioi] 1953;141:392-406.

20. Khodadoust AA, Silverstein AM. Studies on the nature of the privilege enjoyed by corneal allografts. Invest Ophthal­mol 1972;11:137-48.

21. Polack FM. Corneal transplantation [editorial]. Invest Ophthalmol 1973;12:85-6.

22. Khodadoust AA. The allograft rejection reaction: the leading cause of late failure of clinical corneal grafts. In: Corneal Graft Failure. Amsterdam: Elsevier, 1973;151-67. (Ciba Foundation Symposium: 15 new series.)

23. Batchelor JR, Casey T A, Werb A, eta!. HLA matching and corneal grafting. Lancet 1976;1:551-4.

24. Klein PE, Stark WJ, Maguire MG, eta!. Donor-recipient crossmatching and typing to avoid corneal allograft rejection. In: Cavanaugh HD, ed. The Cornea: The World Congress on the Cornea III. New York: Raven Press, 1988;395-8.

117

Ophthalmology Volume 100, Number 1, January 1993

25. Mendelsohn AD, Watson BD, Alfonso EC, et al. Amelio­ration of experimental lipid keratopathy by photochemically induced thrombosis of feeder vessels. Arch Ophthalmol 1987; 105:983-8.

26. Huang AJW, Watson BD, Hernandez E, Tseng SCG. Pho­tothrombosis of corneal neovascularization by intravenous rose bengal and argon laser irradiation. Arch Ophthalmol 1988; 106:680-5.

27. Corrent G, Roussel TJ, Tseng SCG, Watson BD. Promotion of graft survival by photothrombotic occlusion of corneal neovascularization. Arch Ophthalmol 1989; 107:1501-6.

28. Epstein RJ, Hendricks RL, Harris DM. Photodynamic therapy for corneal neovascularization. Cornea 1991; 10: 424-32.

29. Cherry PMH, Faulkner JD, Shaver RP, et al. Argon laser

118

treatment of corneal neovascularization. Ann Ophthalmol 1973;5:911-20.

30. Reed JW, Fromer C, Klintworth GK. Induced corneal vas­cularization remission with argon laser therapy. Arch Ophthalmol1975;93:1017-9.

31. Cherry PMH, Garner A. Corneal neovascularization treated with argon laser. Br J Ophthalmol 1976;60:464-72.

32. Marsh RJ, Marshall J. Treatment oflipid keratopathy with the argon laser. Br J Ophthalmol 1982;66:127-35.

33. Nirankari VS, Baer JC. Corneal argon laser photocoagula­tion for neovascularization in penetrating keratoplasty. Ophthalmology 1986;93: 1304-9.

34. Baer JC, Foster CS. Corneal laser photocoagulation for treatment of neovascularization. Efficacy of 577 nm yellow dye laser. Ophthalmology 1992;99:173-9.