Safety of repeated intravitreous injections of antibiotics and dexamethasone.

PURPOSE: To determine the retinotoxicity of repeated intravitreous injections of vancomycin, ceftazidime, and dexamethasone in rabbit eyes. METHODS: Twenty pigmented New Zealand rabbits were divided into two groups. In Group 1, the right eyes received repeated intravitreous injections with vancomycin 0.3 mg, ceftazidime 0.7 mg, and dexamethasone sodium phosphate 0.13 mg at three consecutive 48-hour intervals. Group 2 right eyes received three times higher dose of the same intravitreous drugs as used in Group 1, repeated at the same frequency. All left eyes served as control eyes. Retinotoxicity was monitored by slit-lamp biomicroscopy, indirect ophthalmoscopy, electroretinography, and light and electron microscopy. RESULTS: No evidence of retinotoxicity was found in Group 1 eyes. Photopic A-waves were significantly elevated, and 30- and 50-Hz flicker fusion amplitudes were significantly depressed in Group 2 eyes. No changes were found by clinical or histopathologic examination in the retinas of either group. CONCLUSIONS: Three repeated intravitreous injections at 48-hour intervals of a combination of vancomycin, ceftazidime, and dexamethasone in rabbit eyes at dosages that approximate drug concentrations recommended for human endophthalmitis were nontoxic. Similar injections at three times higher doses resulted in mild electroretinogram changes.

High dose intramuscular methylprednisolone in experimental Staphylococcus aureus endophthalmitis.

We attempted to determine whether treatment using intramuscular methylprednisolone plus intravitreal vancomycin decreased ocular inflammation and preserved retinal function better in experimental Staphylococcus aureus (S. aureus) endophthalmitis than treatment with intravitreal vancomycin alone. Sixteen rabbits received intravitreal inoculations in both eyes with S. aureus and the rabbits were divided into two groups (group I and group II) of eight rabbits each. Group I rabbits were treated with one injection of intravitreal vancomycin in each eye at either 24, 36, 48 or 72 hours after bacterial inoculation followed by seven consecutive days of high dose intramuscular methylprednisolone (30 mg/kg per day). Group II rabbits were treated with only one intravitreal injection of vancomycin in each eye at equivalent time intervals as in Group I. Clinical evaluations of ocular inflammation were performed by slit-lamp biomicroscopy and indirect ophthalmoscopy. Electroretinography (ERG) was performed eight days after bacterial inoculation to assess retinal function in all eyes. The combination of intramuscular methylprednisolone and intravitreal vancomycin resulted in a degree of ocular inflammation equal to eyes treated with intravitreal vancomycin alone at all treatment intervals. ERG responses were not significantly different in either group. A single intravitreal injection of vancomycin plus daily intramuscular methylprednisolone for seven days were found neither to decrease ocular inflammation nor preserve retinal function better than a single intravitreal injection of vancomycin in our experimental model of S. aureus endophthalmitis.

Timing of dexamethasone treatment in experimental Staphylococcus aureus endophthalmitis.

PURPOSE: This study sought to determine whether intravitreal dexamethasone with vancomycin preserves retinal function in eyes with experimental Staphylococcus aureus endophthalmitis better than intravitreal vancomycin alone. METHODS: Twenty-four rabbits received intravitreal injections in both eyes with S. aureus. Right eyes were treated with intravitreal dexamethasone plus vancomycin and left eyes were treated with vancomycin alone at 24, 36, 48, or 72 hours after inoculation. Evaluation was performed by slit-lamp biomicroscopy, indirect ophthalmoscopy, and electroretinogram. Vitreous humor cultures and histopathologic examinations were performed on the eyes after the rabbits were killed. RESULTS: The combination of intravitreal dexamethasone and vancomycin resulted in significantly less inflammation than vancomycin alone at 24 and 36 hours after inoculation, but electroretinograms showed significantly better preservation only at 36 hours after bacterial inoculation. Viable bacteria were cultured from eyes treated 48 and 72 hours after inoculation. CONCLUSION: Intravitreal dexamethasone was found to be beneficial by electroretinography when administered 36 hours after infection. In the authors' model, a single intravitreal injection of vancomycin with or without the addition of dexamethasone was insufficient to sterilize eyes 48 and 72 hours after bacterial inoculation.

Intraorbital air simulating an intraocular foreign body.

PURPOSE: To present an ultrasonographic finding that simulated an intraocular foreign body after repair of a ruptured globe. METHOD: Case report. An ultrasonogram of a post-trauma eye was correlated with a computed tomographic scan. RESULTS: B-scan ultrasonography was performed on an eye after repair of a corneoscleral laceration. The ultrasonogram showed a highly reflective echo source suggestive of a foreign body; however, an orbital computed tomographic scan demonstrated that the lesion was intraorbital air. CONCLUSION: Although a highly reflective echo source in the presence of a ruptured globe may suggest a foreign body, the presence of orbital air should also be considered when interpreting ultrasonograms used in the preoperative and postoperative management of globe trauma.

Determination of ocular toxicity in multiple applications of foscarnet iontophoresis.

This is the first study of multiple applications of drug iontophoresis in the eye. We repeated ocular foscarnet iontophoresis in 10 eyes of 10 rabbits every third day at the same paralimbal site for a total of seven applications over a period of 21 days to determine the efficacy and toxicity of multiple applications of ocular foscarnet iontophoresis. Mean vitreous human foscarnet concentration of 189 +/- 50.6 microM (SD) was achieved four hours after the seventh consecutive iontophoretic application over a period of twenty-one days. These levels were within the therapeutic range (25-800 microM) for the treatment of CMV retinitis and comparable to the intravitreal foscarnet concentrations achieved in eyes treated with a only a single application of ocular iontophoresis. Electroretinography (ERG) and Slit-lamp biomicroscopy responses revealed no evidence of ocular toxicity. Indirect ophthalmoscopy of the retinas and gross examinations of the calottes revealed a single, small burn in the retina and choroid corresponding to the application site of the iontophoresis probe similar to the lesion resulting from a single application of iontophoresis. Light and electron microscopy revealed local tissue injury and fibrosis at the iontophoresis site, but adjacent areas were unaffected.

Ocular iontophoretic supplementation of intravenous foscarnet therapy.

PURPOSE: Reactivation of cytomegalovirus retinopathy during intravenous antiviral therapy is usually treated with higher doses of drug. We sought to determine whether ocular iontophoresis increases the intravitreal foscarnet concentration attained by intravenous injection. METHODS: We injected foscarnet (120 mg/kg or 180 mg/kg) intravenously into 24 rabbits and determined the time of maximal concentrations in serum and vitreous humor. We injected the same doses into 24 additional rabbits and administered ocular foscarnet iontophoresis one hour later. Vitreous humor concentrations were assayed at one, four, eight, 24, 60, and 120 hours after iontophoresis and compared with those from injection alone. RESULTS: Maximum serum and vitreous humor concentrations were achieved one hour after each intravenous dose. Maximum vitreous humor concentrations were achieved four hours after 120 mg/kg intravenous doses plus iontophoresis and eight hours after 180-mg/kg intravenous doses plus iontophoresis. Vitreous humor levels were significantly higher in eyes receiving intravenous foscarnet (120 mg/kg, P < .0001; 180 mg/kg, P < .0001) plus ocular Foscarnet iontophoresis than in those receiving intravenous foscarnet alone. Vitreous humor foscarnet levels in eyes receiving 120 mg/kg intravenously did not differ significantly from those in the group receiving 180 mg/kg intravenously (P < .1). The intravenous dose did not significantly affect vitreous humor levels after iontophoresis (P < .1). Vitreous concentrations fell below therapeutic levels (25 microM) in all eyes 60 hours after intravenous foscarnet and ocular foscarnet iontophoresis. CONCLUSIONS: Ocular iontophoresis significantly increased intravitreous foscarnet concentrations above those attained by intravenous injection alone and may be an effective alternative to increasing the intravenous drug dose in patients with reactivated cytomegalovirus retinopathy.

Essential thrombocythemia and central retinal vein occlusion with neovascular glaucoma.

PURPOSE: We report a case of unilateral central retinal vein occlusion resulting from essential thrombocythemia, a rare myeloproliferative disorder with abnormally increased platelet count. METHODS: A 59-year-old man had central retinal vein occlusion in the left eye as the initial sign of essential thrombocythemia. He later developed neovascular glaucoma and optic disk neovascularization. RESULTS: Laser panretinal photocoagulation, goniophotocoagulation, glaucoma medications, and control of the platelet count were effective treatment. CONCLUSIONS: Early thrombocythemia is associated with systemic and ocular thrombotic and embolic complications. Early diagnosis, recognition of ocular complications, and appropriate treatment were crucial in controlling central retinal vein occlusion and ocular neovascularization associated with essential thrombocythemia.

Pars plana incisions of four patients: histopathology and electron microscopy.

The pathology of pars plana incisions of four patients is described: three with light microscopy and one with light and electron microscopy. Two eyes were removed because of choroidal melanoma, immediately and 8 days after vitrectomy and transvitreous retinal biopsy. Considerable disruption of tissues surrounding the pars plana incisions was observed. Vitreous was incarcerated in the wounds, which healed with granulation tissue. One eye was examined 4 months after vitrectomy for diabetic retinopathy and a failed pars plana filtering operation. It contained fibrovascular ingrowth from all the incisions, infiltrating the vitreous base with granulation tissue and causing vitreous haemorrhage and retinal detachment. One eye was removed 1 year after vitrectomy for anterior hyaloidal fibrovascular proliferation and early phthisis. The wound had fibrous ingrowth histologically and evidence of active fibroplasia.

Pharmacodynamics of subconjunctival gentamicin after experimental cataract operations and traumatic lacerations.

We addressed the concern whether retinal toxicity could occur with subconjunctival gentamicin injections after cataract operations or traumatic corneoscleral lacerations. The lacerations were created along the superior limbus of rabbit eyes and the lenses removed. We sutured the right eyes to mimic a cataract operation and the left eyes remained unsutured to mimic a traumatic corneoscleral laceration. Vitreous humor samples were obtained at 1, 2, 4, 8 and 12 hours after the subconjunctival injection, but in no case did vitreal concentration approach retinotoxic levels (> 133 micrograms/ml).

Ocular toxicity of iontophoretic foscarnet in rabbits.

Transscleral iontophoresis of foscarnet is a noninvasive drug delivery system for the local treatment of cytomegalovirus (CMV) retinopathy. We determined the retinotoxic effects of transscleral iontophoresis of foscarnet. Slit-lamp biomicroscopy revealed no toxic effects for any of the treated eyes. Indirect ophthalmoscopy showed retinal and choroidal burns 1-3 mm in diameter at the site of iontophoresis in both foscarnet-treated eyes and saline-treated control eyes. Light and electron microscopy revealed focal retinal, retinal pigment epithelial, and choroidal necrosis at the site of iontophoresis but no abnormalities elsewhere. Ganzfeld electroretinographic studies revealed no response differences between foscarnet-treated eyes vs. controls.

Vitrectomy techniques in late-stage Coats'-like exudative retinal detachment.

Retinal telangiectasia is the hallmark of Coats' disease. In the late stages, leakage from these abnormal vessels can result in a total, bullous exudative retinal detachment with cholesterol-laden subretinal fluid. Secondary angle-closure glaucoma may result in a blind and painful eye which may require enucleation or evisceration. Surgical reattachment of the retina and destruction of the retinal telangiectasia may preserve these eyes. We have found that vitrectomy, internal drainage of subretinal fluid and cholesterol, direct treatment of the retinal telangiectasia with intraocular diathermy and intravitreal gas or silicone oil injection are effective surgical techniques for salvaging these severely damaged eyes.

Transscleral iontophoresis of foscarnet.

Current local treatments of cytomegalovirus retinopathy may result in serious intraocular complications. Using an animal model, we investigated transscleral iontophoresis as a technique for delivery of foscarnet to the vitreous. Using a probe tip surface area of 0.19 mm2, a current of 1 mA, and a duration of ten minutes, transscleral iontophoresis of 0.5 ml of a 24-mg/ml foscarnet solution was administered to 72 normal rabbits. Vitreous aspiration was performed at 12 intervals (15 minutes, 30 minutes, and one, two, four, eight, 16, 24, 32, 40, 48, and 60 hours) after iontophoresis, and samples were analyzed by high-performance liquid chromatography to determine the vitreous pharmacokinetics of foscarnet. A peak foscarnet concentration of 200 +/- 31 microM (mean +/- standard deviation) was attained four hours after iontophoresis and was well below the concentration reported to cause retinal toxicity. Therapeutic levels were maintained until 60 hours after iontophoresis. The elimination half-life was approximately 24 hours. No toxic effects to anterior chamber structures were observed by biomicroscopy. Transscleral iontophoresis of foscarnet may provide an effective and safe technique for local treatment of cytomegalovirus retinopathy in patients with acquired immunodeficiency syndrome.

Topical and intravenous gentamicin in traumatically lacerated eyes.

Intravenous or topical gentamicin may be the initial mode of treatment for lacerated or ruptured eyes by emergency room physicians while awaiting ophthalmic consultation and surgical repair. The purpose of this study was to determine the possibility of having retinotoxic intravitreal gentamicin concentrations in experimentally lacerated rabbit eyes treated with either intravenous or topical gentamicin separately or in combination with each other. Nontoxic concentrations of gentamicin were found in the vitreous bodies by all routes of drug administration. After 3 h intravitreal concentrations of gentamicin were: 0.20-0.30 microgram/ml when treated intravenously, 0-2.9 micrograms/ml when treated topically, and 0.20-0.51 microgram/ml when treated both intravenously and topically. While the upper range of topically applied gentamicin concentrations (2.9 micrograms/ml) is therapeutic for some pathogens, the wide range of intravitreal concentrations (0-2.9 micrograms/ml) achieved does not indicate that topically applied gentamicin with or without intravenously administered gentamicin can reliably achieve therapeutic concentrations.

Intravitreous granulation tissue and retinal detachment following pars plana injection for cytomegalovirus retinopathy.

Recently, the pars plana route of injection has been used to administer drugs for the treatment of severe intraocular infections. We observed a complication of this method in a patient with AIDS and cytomegalovirus (CMV) retinopathy.

Topical indomethacin in the treatment of chronic cystoid macular edema.

Eighty percent of 30 eyes with chronic cystoid macular edema (CME) after cataract extraction achieved improved visual acuity of three or more lines when treated with 1% indomethacin eye drops. Of these patients 53% demonstrated an "on/off" phenomenon induced by the initiation and cessation of treatment documented by visual acuity measurements and fluorescein angiography. This "on/off" phenomenon suggests that there is a direct relationship between the use of 1% indomethacin eye drops and the resolution of chronic CME after cataract extraction. Previous studies on the treatment of CME with indomethacin have not addressed the use of indomethacin eye drops for chronic CME.

Scleral buckling for rhegmatogenous retinal detachment associated with severe myopia.

From Jan. 1, 1980, to Dec. 31, 1989, we performed scleral buckling surgery on 48 eyes of 46 patients for rhegmatogenous retinal detachments associated with severe myopia (greater than 5.00 diopters). Forty eyes of 38 patients were observed for at least six months, and the mean follow-up period was 46 months. Intraoperative complications occurred in four of 48 eyes (8%) and included retinal incarceration (two eyes), choroidal hemorrhage (one eye), and choroidal detachment (one eye). Three of the 40 eyes (7.5%) followed up for more than six months developed a recurrent retinal detachment and underwent a revision of the scleral buckle. At the last follow-up examination, the retinas of all 40 eyes were totally reattached. Final visual acuity of 20/40 or better was attained in 26 of 40 eyes (65%). Because of the low rate of intraoperative complications and the high rate of success, scleral buckling is recommended for most patients with rhegmatogenous retinal detachments associated with severe myopia.

The ocular effects of gases when injected into the anterior chamber of rabbit eyes.

We studied the toxic effects of sulfur hexafluoride and perfluoropropane in comparison with air, balanced salt solution, hyaluronate sodium, and aqueous humor in a rabbit model. Sixty normal pigmented rabbits were studied during a period of 4 weeks. The variables studied were slit-lamp biomicroscopic examination of the anterior segment, intraocular pressure as measured by pneumotonometry, corneal thickness and endothelial cell count as measured by specular microscopy, lens opacity by Scheimpflug photography, and light and transmission electron microscopy. All three gases were more toxic to the cornea and lens than were balanced salt solution, hyaluronate, and aqueous humor. However, 15% perfluoropropane and 50% sulfur hexafluoride were no more toxic to the eye than air was. Therefore, this study suggests that nonexpansile mixtures of perfluoropropane and sulfur hexafluoride may be beneficial and relatively safe in re-forming persistently flat anterior chambers in situations where the use of air is being considered.

Spontaneous dural carotid-cavernous fistula with central retinal vein occlusion and iris neovascularization.

Spontaneous dural carotid-cavernous fistulas are dural vascular malformations that usually run a benign course. We present a case of a spontaneously occurring dural carotid-cavernous fistula complicated by central retinal vein occlusion and iris neovascularization that led to progressive visual failure.

Experimental transscleral iontophoresis of ciprofloxacin.

Ciprofloxacin was administered into the aqueous humor and vitreous body of the rabbit eye by transscleral iontophoresis. Positively and negatively charged forms of the drug molecule were tested. Therapeutic concentrations of ciprofloxacin were achieved in the aqueous body only when the negatively charged drug molecule was used. Ciprofloxacin did not reach the vitreous body in therapeutic concentrations in either the positively or negatively charged form, but higher concentrations were achieved when the drug was negatively charged. Peak levels were obtained in the aqueous and vitreous bodies (0.62 micrograms/ml and 0.19 micrograms/ml, respectively) one hour after transscleral iontophoresis of negatively charged ciprofloxacin at 5 mA for 15 minutes.

Taxol treatment of experimental proliferative vitreoretinopathy.

Taxol is a potent stabilizer of microtubules, and inhibitor of in vitro replication, migration, and contraction of fibroblasts. It has been found to limit the development of experimental tractional retinal detachments in nonvitrectomized rabbit eyes. We used taxol in vitrectomized, phakic rabbit eyes with experimentally induced proliferative vitreoretinopathy and tractional retinal detachments. Taxol was dissolved in 30% DMSO because of poor aqueous solubility. A single 0.1 ml intravitreal dose of 2 x 10(-4) M taxol in 30% DMSO was injected immediately after 250,000 heterologous corneal fibroblasts had been injected; 0.1 ml of 30% DMSO was injected into control eyes. Taxol reduced the incidence of tractional retinal detachments seen 3-4 weeks later. When taxol injection was delayed for 3 days after the initial intravitreal injection of fibroblasts into nonvitrectomized eyes, the extent of retinal detachments was reduced, but the incidence of retinal detachment was unchanged from the untreated eyes at the end of 4 weeks. These data indicate that taxol may be most useful when given early in the course of proliferative vitreoretinopathy.