AMERICAN JOURNAL OF OPHTHALMOLOGY
mires by keratornetry were distorted. Three weeks after removal of the epithelium, visual acuity in the left eye was correctable to 20/20. A 61-vear-old man with decreased vision in his right eye had some symptoms of itching, redness, and tearing in the right eye in the past although recurrent erosions were never diagnosed. Best corrected visual acuity was R.E.: 20/60 and L.E.: 20/20. Both corneas showed central map-dot dystrophic changes with the right cornea having larger white intraepithelial deposits in the visual axis. After removal of the epithelium of the right cornea, corrected visual acuity improved to 20/20. That in the left eye remained 20/20 for the next five years. At that time visual acuity in the left eye decreased to 20/50 as large white intraepithelial cysts developed in the visual axis. Three weeks after the removal of the epithelium in the left eye, best corrected visual acuity had improved to 20/20. Trobe and Laibson ' described 35 patients with anterior membrane dystrophy of whom only five had significant visual loss, although most had symptoms of recurrent erosion. In onlv one of these five cases was the visual loss as much as four Snellen lines. Both eyes of my two patients demonstrated the typical, albeit in one eye subtle, findings of anterior membrane dystrophy. The initial complaint was decreased vision. Anterior membrane changes of the cornea are best seen at the slit lamp under direct visualization with a broad beam, or with retroillumination against the red reflex with the pupil dilated. The involved epithelium is easily removed at the slit lamp with a sterile cotton swab while the patient is under topical anesthesia. Although it is only necessary to remove the central epithelium to clear the visual axis, the abnormal epithelium tends to be loosely adherent and often comes off in large sheets. After removal the eyes are patched and antibiotic eyed rops and cycloplegics are administered for several days until the epithelial defect has healed. Sodium chloride 5% ointment used at bedtime for one to two months after surgery may prevent problems with recurrent erosions. Within three weeks the patients showed maximal visual improvement. With healing further anterior membrane dystrophic changes appeared, but these were not visually significant. On the basis of my experience with cases such as these, anterior membrane dystrophy must be included in the differential diagnosis of unexplained visual loss. It is best detected by careful slit-lamp examination and by keratometry. Once the condition is diagnosed, the
visual loss can be easily corrected by removal of the epithelium.
References 1. Trobe, J. D., and Laibson, P. R.: Dystrophic changes in the anterior cornea. Arch. Ophthalmol.
2. Brown, N., and Bron, A.: Recurrent erosion of the cornea. Br. J. Ophthalmol. 60:84, 1976. 3. Shoch, D. E., Stock, E. 1., and Schwartz, A. E.: Stromal keratitis complicating anterior membrane dystrophy. Am. J. Ophthalmol. 100:199, 1985.
Optic Disk and Iris Neovascularization After Surgical Interruption of the Retinal Circulation Robert K. Hutchins, M.D., John W. Gittinger, Jr., M.D., and John J. Weiler, M.D. Department of Ophthalmology, New England Medical Center Hospitals, and Eye Research Institute.
Inquiries to John f. Weiter, M.D., Retina Associates, 100 Charles River Plaza, Cambridge St., Boston, MA 02114.
Retinal neovascularization has not been associated with central retinal artery occlusion. 1,2 Although it is generally accepted that iris neovascularization follows 1%to 2% of central retinal artery occlusions," Hayreh and Podhajsky'' argued that central retinal artery occlusion and iris neovascularization are not related in a cause-and-effect manner. We studied a patient in whom optic disk and iris neovascularization followed the acute interruption of retrobulbar blood vessels. After ten months of gradual visual deterioration and progressive ophthalmoplegia of the right eye, an otherwise healthy, 31-year-old woman underwent a right craniotomy and exploration of the right optic nerve with partial excision of tumor in the right orbit and probable interruption of the ophthalmic artery. The histopathologic findings were consistent with a sclerosing variant of inflammatory pseudotumor. Six weeks before surgery, her visual acuity was R.E.: no light perception and L.B.: 20/15. A
Vol. 101, No.5
Letters to the Journal
Goldmann visual field of the left eye was normal. She had 4 mm of proptosis in the right eye and restriction of gaze in all but the inferior field. Findings of an anterior segment examination were normal. The intraocular pressure was 13 mm Hg in both eyes. The right optic disk was swollen, with dilated veins. Several retinal petechial hemorrhages were observed in all four quadrants. The left fundus was normal except for a drusen-like lesion on the optic disk. Ophthalmologic evaluation the day before surgery disclosed a swollen right optic disk but no abnormal optic disk vessels. One month after surgery the right optic disk was pale with a tuft of new vessels. The arterial blood columns were segmented and stationary; and the veins were dilated. There was a cherryred spot in the macula and dot hemorrhages in the periphery. New vessels involving the pupillary margin and the angle of the right eye were first seen three months after surgery. The intraocular pressure was 11 mm Hg in the right eye. The neovasctilar tuft on the right optic disk had grown into a large membrane extending into the vitreous cavity (Figure). The retinal vessels were sheathed and contained no blood. A vitreous and subhyaloid hemorrhage was present. A fundus fluorescein angiogram showed normal filling of the choroid, a large optic disk neovascular membrane, and no filling of the retinal vessels (Figure). Panretinal photocoagulation (500 urn, 2,270 spots) resulted in complete regression of the
optic disk and iris neovascularization. Followup at 40 months showed aberrant right third nerve regeneration, an atrophic fibrotic optic disk membrane, and no iris neovascularization, Both retinal and iris neovascularization can develop in cases of chronic ocular ischemia.r" Although our patient had signs suggestive of a central retinal vein obstruction before craniotomy, ocular neovascularization did not develop until after surgical disruption of the posterior orbital vessels. Optic disk and iris neovascularization began and flourished in an eye that had the ophthalmoscopic and fluorescein angiegraphic characteristics of a central retinal artery occlusion. The complete regression of the optic disk and iris new vessels after panretinal photocoagulation suggested that ischemic retinal tissue contributed to the neovascularization. The rarity of these findings may be explained by a scarcity of vasoproliferative substance released from infarcted retinal tissue." Although the angiogram performed one month after surgery showed no filling of the retinal vessels and normal filling of the choroid, we suspect that there was an element of ocular ischemia in this case secondary to the interruption of the ophthalmic artery. Multiple and important potential anastomoses occur between the ophthalmic and external carotid arteries in humans." The rate and degree of development of these potential or existing anastomoses could easily account for a degree of ocular ischemia leading to the
Figure (Hutchins, Gittinger, and Weiter). Left, The right eye shows the optic disk neovascular membrane and retinal vessels replaced by yellow lines. Right, Angiogram of the right fundus shows the optic disk neovascular membrane and no filling of the retinal vessels.
AMERICAN JOURNAL OF OPHTHALMOLOGY
observed iris and optic disk neovascularization.
References 1. Wise, G. N., Dollery, C. T., and Henkind, P.: The Retinal Circulation. New York, Harper and Row, 1971, p. 299. 2. Hayreh, S. 5., and Podhajsky, P.: Ocular neovascularization with retinal vascular occlusion. II. Occurrence in central and branch retinal artery occlusion. Arch. Ophthalmol. 100:1585, 1982. 3. Gartner,S., and Henkind, P.: Neovascularization of the iris (rubeosis iridis). Surv. Ophthalmol. 22:291, 1978. 4. Brown, G. c.. Magargal, L. E., Simeone, F. A., Goldberg, R. E., Federman, J. L., and Benson, W. E.: Arterial obstruction and ocular neovascularization. Ophthalmology 89:139, 1982. 5. Sturrock, G. D., and Mueller, B. R.: Chronic ocular ischaemia. Br. J. Ophthalmol, 88:716, 1984. 6. Henkind, P., and Wise, G. N.: Retinal neovascularization, collaterals, and vascular shunts. Br. J. Ophthalmol. 58:413; 1974. 7. Hayreh, S. 5.: The ophthalmic artery. In Cant, J. S. (ed.): Vision and Circulation. St. Louis, C. V. Mosby, 1976, pp. 171-179.
Gelatin Implants in Glaucoma Filtering Surgery Andrew Antoszyk, M.D., David Robinson, M.D., Alan D. Proia, M.D., and M. Bruce Shields, M.D.
Eight eyes of four albinotic rabbits were studied. After anesthetizing the animals with intramuscular ketamine, we prepared 3-mm peritomies and subconjunctival pockets in three separate quadrants. Thin 3 x 3-mm portions of Gelfilm or Gelfoam were then inserted into two of the pockets, leaving the third as a control, and each peritomy was closed with an interrupted 10-0 nylon suture. The eyes were inspected weekly for the first two weeks and monthly thereafter. Two eyes were enucleated two weeks after surgery, four at two months, and two at six months. All surgical sites were studied by light microscopy. By gross inspection, the implants appeared to be well tolerated in all animals with no evidence of severe ocular inflammation. Histologic evaluation, however, disclosed a marked reaction to the implant material at two weeks, especially to the Gelfoam, with many foreignbody giant cells and fibrous tissue production (Fig. 1). By two months, the implant material had been digested and there was no discernible subconjunctival pocket. In two patients with neovascular glaucoma, two with glaucoma secondary to trauma, one with congenital glaucoma, and one with glaucoma in aphakia, filtering surgery was combined with subconjunctival implantation of Gelfilm (one eye) or Gelfoam (five eyes). The patients were 15 to 80 years old. Each eye received subconjunctival injections of gentamicin and dexamethasone at the end of the procedure and topical prednisolone 1 % four to eight times daily and atropine 1 % twice a day were prescribed for a minimum of one month.
Department of Ophthalmology, Duke University Medical Center. This study was supported by Core Grant 1-P30 EYO 5722 from the National Institutes of Health.
Inquiries to M. Bruce Shields, M. D., Duke University Eye Center, Durham, NC 27710.
In 1955, Laval! reported encouraging results in rabbit and human eyes using gelatin implants to promote successful glaucoma filtering blebs, although subsequent studies failed to confirm his findings.P We evaluated the subconjunctival implantation of absorbable gelatin film (Gelfilm) or sponge (Gelfoam) in rabbit eyes and in high-risk patients undergoing glaucoma filtering surgery.
Fig. 1 (Antoszyk and associates). Spicules of gelatin material surrounded by foreign-body reaction in subconjunctival tissue of rabbit eye two weeks after implantation of Gelfoam (hematoxylin and eosin, x 250).