Treatment of Acanthamoeba keratitis.

PURPOSE: To review the available treatment modalities for Acanthamoeba keratitis and to assess the most effective regimens. METHODS: The efficacy of therapeutic modalities and chemotherapeutic agents used in the treatment of Acanthamoeba keratitis was reviewed. Potential synergistic or additive drug interactions were documented both in vitro and in vivo. RESULTS: Early diagnosis of Acanthamoeba keratitis plays a crucial role in successful medical treatment. The cationic antiseptic agents, chlorhexidine and polyhexamethylene biguanide (PHMB) have the lowest minimal amoebicidal concentrations. Synergistic effects are seen when used with pentamidine, and additive effects are seen with propamidine or neomycin. Penetrating keratoplasty should be deferred if at all possible until a medical cure has been achieved. CONCLUSION: Early diagnosis and wide epithelial debridement are important elements in the successful treatment of Acanthamoeba keratitis. Recommended therapy would include the cationic antiseptic agents, chlorhexidine or PHMB in combination with propamidine isethionate and neomycin as part of triple therapy. Surgical intervention should be avoided until a medical cure has been achieved.

Ketorolac, prednisolone, and dexamethasone for postoperative inflammation.

This 6-week, partially masked, three-arm, multicenter study was conducted to evaluate the postoperative anti-inflammatory efficacy of ketorolac, a cyclooxygenase inhibitor. The study setting was the clinical practice of six ophthalmic surgeons. The study enrolled 157 candidates for routine extracapsular cataract extraction or phaco-emulsification and posterior-chamber intraocular lens implantation. Patients who received any glucocorticoid or cyclooxygenase inhibitor within 1 week of surgery were excluded. All patients were treated with solutions of 0.5% ketorolac, 1% prednisolone acetate, or 0.1% dexamethasone instilled into the operative eye three times daily from 1 day before surgery to 4 weeks after surgery. Efficacy variables included the signs of anterior-segment inflammation, primarily cells and flare in the anterior chamber, as observed by slit-lamp biomicroscopy; fluorescein leakage across the blood-aqueous barrier as measured by fluorophotometry; and the rating of efficacy by the investigator. No significant differences were seen between ketorolac and either glucocorticoid in cells and flare. No significant differences were found in other signs of inflammation, except conjunctival hyperemia and Descemet's folds at week 2. Ketorolac showed significantly greater efficacy than the glucocorticoids against blood-aqueous barrier breakdown at day 5 and week 2, as demonstrated by the difference in fluorescein concentration between the operated and nonoperated eyes. Investigators did not detect any significant difference in rating for overall effectiveness and acceptability. These findings support the use of ketorolac as an alternative to glucocorticoids for the treatment of postoperative inflammation.

The effect of vancomycin on the corneal endothelium.

Possible toxic effects of vancomycin on the corneal endothelium were assessed in rabbit eyes and subsequently in corneal transplant recipients. Fifteen New Zealand White rabbits were divided into five groups of three rabbits each. A paracentesis was performed on right eyes only. One hundred microliters of aqueous humor was removed followed by anterior chamber injection of 100 mu l of balanced salt solution (BSS) or varying concentrations of vancomycin (150, 750, 1,875, 7,500 mu g/ml). Left eyes served as untreated controls. Endothelial cell morphology and density were assessed by contact specular microscopy 48 h postinjection. There was no statistically significant difference in endothelial cell density between left (control) or right (treated) eyes receiving either BSS or varying concentrations of vancomycin. Transmission electron microscopy of rabbit corneal endothelium exposed to varying concentrations of vancomycin (150, 750, 1,875, 7,500 mu g/ml) showed no toxic effects. Six patients undergoing penetrating keratoplasty (PKP) received donor corneas stored in Optisol containing gentamicin to which vancomycin was added to make a final concentration of 150 mu g/ml. Eight patients undergoing PKP received corneas stored in Optisol alone. Specular microscopy was performed preoperatively and 3, 6, and 12 months postoperatively. No statistically significant difference in mean endothelial cell change was observed (Wilcoxon signed-rank test) at any time point postoperatively. Endothelial cell morphology and function are not adversely affected by therapeutic doses of vancomycin.

A cadaveric eye model for practicing anterior and posterior segment surgeries.

OBJECTIVE: To develop a simple and effective model for practicing anterior and posterior segment surgeries. DESIGN: A Lander wide-field keratoprosthesis was used as an artificial cornea in human cadaver eyes. RESULTS: The keratoprosthesis provided a tight seal which maintained the anterior chamber. Continuous curvilinear capsulorrhexis, hydrodissection, and phacoemulsification of the cadaveric lens followed by insertion of a foldable intraocular lens into the capsular bag were performed through a standard three-planed scleral tunnel incision. Pars plana vitrectomy, using the standard three-port approach, was performed. Panretinal photocoagulation was performed using a slit-lamp laser delivery system and a Rodenstock lens. CONCLUSIONS: This eye model very closely simulates the tactile behavior of a living eye and may be useful as a realistic practice and teaching platform for a wide variety of anterior and posterior segment surgical procedures, including laser photocoagulation.

Ocular surface flora in drowning victims.

Contaminated allograft donor tissue represents a potential source of infection unique to keratoplasty. We prospectively studied perilimbal cultures of drowning victims over 30 months to determine if the ocular surface flora of drowning victims was unique. Twenty-eight donor eyes were cultured from 14 drowning victims. Ninety-three percent of limbal cultures were positive for one or more organisms, compared to results of a previous study of surface contamination of donor globes in which 65% were found to be culture positive. Fifty-seven percent (16 of 28) of donor eyes grew at least one streptococcal species, while 46% (13 of 28) grew two or more streptococcal species. A previous study identified streptococcal species in 7.9% of perilimbal cultures from nondrowning donor eyes, which demonstrates the unique effect of drowning on the ocular flora. The incidence of gram-negative isolates was also markedly higher in drowning victims, compared to previous studies of donor globes. This prospective study has shown that the ocular surface flora of drowning victims harbors markedly increased numbers of streptococcal species as well as gram-negative organisms. These findings demand careful globe decontamination, and emphasize the need for appropriate antibiotic coverage in corneal storage media.

Open-sky phacoemulsification during corneal transplantation.

A continuous, circular capsulorhexis maintains the integrity of the anterior capsule. During corneal transplantation, a capsulorhexis may be combined with phacoemulsification, allowing the cataractous lens to be delivered in small fragments, thereby maintaining the integrity of the anterior capsular margin. This approach allows stable "in-the-bag" placement of a posterior chamber intraocular lens (PC-IOL) during combined cataract extraction and corneal transplantation. The technique also minimizes the risk of radial anterior capsular tears and subsequent decentration of the PC-IOL.

Cytotoxicity of viscoelastics on cultured corneal epithelial cells measured by plasminogen activator release.

BACKGROUND: Plasminogen activator has been shown to be released by epithelial cells following corneal injury. The demonstration of the release of plasminogen activator from cultured corneal epithelial cells has been used for developing a cytotoxicity test, the Corneal Epithelial Plasminogen Activator test, which compares changes in the level of plasminogen activator in tissue culture media following chemical exposure as an index of chemical injury. METHODS: Cultured rabbit corneal epithelial cells were exposed to varying concentrations of several viscoelastics for 1 hour. Release of plasminogen activator into the tissue culture media following exposure to the viscoelastic agent was studied as an index of chemical injury. RESULTS: The least cytotoxicity to cultured rabbit epithelium was associated with those viscoelastic agents containing methylcellulose. A 1-hour exposure to most concentrations of methylcellulose and chondroitin sulfate (Phacote) and methylcellulose (Occucoat) demonstrated release of greater amounts of plasminogen activator than was seen following a similar exposure to balanced salt solution, suggesting the greatest protective effect of these two viscoelastics. In contrast, sodium hyaluronate and chondroitin sulfate (Viscoat) showed decreased amounts of plasminogen activator release after a 1-hour exposure to cultured corneal epithelial cells demonstrating cytotoxicity. Polyacrylamide (Orcolon) and most diluted preparations of sodium hyaluronate (Healon and Healon Yellow) showed only mild reductions in the release of plasminogen activator, whereas undiluted sodium hyaluronate preparations were nearly as cytotoxic as Viscoat. CONCLUSIONS: This study suggests that viscoelastic agents containing methylcellulose (Phacote and Occucoat) may be most protective of the corneal epithelium during ophthalmic surgery. The clinical success of several dilute viscoelastic solutions as tear substitutes was corroborated by the lack of cytotoxicity seen in this study. Viscoat and undiluted sodium hyaluronate preparations showed the greatest cytotoxicity to cultured rabbit corneal epithelium.

Occurrence of bacterial endosymbionts in Acanthamoeba spp. isolated from corneal and environmental specimens and contact lenses.

Free-living and parasitic protozoa are known to harbor a variety of endosymbiotic bacteria, although the roles such endosymbionts play in host survival, infectivity, and invasiveness are unclear. We have identified the presence of intracellular bacteria in 14 of 57 (24%) axenically grown Acanthamoeba isolates examined. These organisms are gram negative and non-acid fast, and they cannot be cultured by routine methodologies, although electron microscopy reveals evidence for multiplication within the amoebic cytoplasm. Examination for Legionella spp. with culture and nucleic acid probes has proven unsuccessful. We conclude that these bacteria are endosymbionts which have an obligate need to multiply within their amoebic hosts. Rod-shaped bacteria were identified in 5 of 23 clinical Acanthamoeba isolates (3 of 19 corneal isolates and 2 of 4 contact lens isolates), 4 of 25 environmental Acanthamoeba isolates, and 2 of 9 American Type Culture Collection Acanthamoeba isolates (ATCC 30868 and ATCC 30871) previously unrecognized as having endosymbionts. Coccus-shaped bacteria were present in one clinical (corneal) isolate and two environmental isolates. There was no statistical difference (P > 0.8) between the numbers of endosymbiont strains originating from clinical (26% positive) and environmental (24% positive) amoebic isolates, suggesting that the presence alone of these bacteria does not enhance amoebic infectivity. Rods and cocci were found in both clinical and environmental isolates from different geographical areas (Seattle, Wash., and Portland, Oreg.), demonstrating their widespread occurrence in nature. Our findings suggest that endosymbiosis occurs commonly among members of the family Acanthamoebidae and that the endosymbionts comprise a diverse taxonomic assemblage. The role such endosymbionts may play in pathogenesis remains unknown, although a variety of exogenous bacteria have been implicated in the development of amoebic keratitis, warranting further evaluation.

Stability and activity of vancomycin in corneal storage media.

Gentamicin is the only antibiotic currently added to commercially available corneal storage media. To reduce the potential for bacterial dissemination from donor corneal tissue to the recipient eye, we evaluated the addition of vancomycin to corneal storage media. When added to Dexsol at a concentration of 200 micrograms/ml, vancomycin levels were maintained, showing a 7% decrease in vancomycin concentration per month, measured < or = 90 days after its addition. Human corneas were stored in gentamicin-free Dexsol (Chiron Ophthalmics, Inc., Irvine, CA, U.S.A.) containing 150 micrograms/ml vancomycin. Corneal tissue levels of vancomycin determined by agar diffusion bioassay were 201, 226, 292 micrograms/ml at 1, 3, and 7 days of storage respectively, suggesting that corneal tissue concentrates vancomycin with time. No differences in endothelial cell count or cell death were seen in corneas stored in Dexsol (containing gentamicin) or Dexsol plus vancomycin when followed for < or = 14 days. Vancomycin added to corneal storage media should reduce the potential for endophthalmitis due to gentamicin-resistant organisms.

Effects of povidone-iodine chemical preparation and saline irrigation on the perilimbal flora.

PURPOSE: To analyze the effects of 5% povidone-iodine preparation and saline irrigation on the species composition of perilimbal flora. METHODS: Cultures were taken from the perilimbal conjunctiva in 100 eyes before preparation for ophthalmic surgery, after instillation of povidone-iodine solution, and after saline irrigation. RESULTS: Bacteria were isolated in 75% of eyes before preparation, in 28% after povidone-iodine instillation, and in 24% after saline irrigation. Fifty-one culture-positive eyes became negative with povidone-iodine, while only four culture-negative eyes became culture-positive (P < 0.001). The number of eyes yielding coagulase-negative staphylococci, Staphylococcus aureus, and Propionibacterium were significantly decreased after povidone-iodine instillation. Twenty-three culture-positive eyes became negative after saline irrigation, while 19 culture-negative eyes became culture-positive (P > 0.25). CONCLUSION: Povidone-iodine solution is effective in reducing bacterial recovery from the perilimbal conjunctiva, where most incisions for intraocular surgery occur. Saline irrigation after povidone-iodine preparation has no significant effect.

Capsulorhexis, phacoemulsification, and posterior chamber lens placement following iridocyclectomy.

Two patients developed cataract following iridocyclectomy for iris melanoma invading the angle. Although several clock hours of zonular fibers had been surgically resected, the use of a continuous, circular capsulorhexis followed by phacoemulsification maintained the integrity of the anterior capsular margin and allowed for stable "in-the-bag" placement of a posterior chamber intraocular lens (PC-IOL). Continuous, circular capsulorhexis is the optimal anterior capsulectomy procedure, minimizing the incidence of radial tears and the risk of subsequent PC-IOL decentration.

Bilateral angle-closure glaucoma associated with uveal effusion: presenting sign of HIV infection.

A 40-year-old homosexual man presented with acute myopia and bilateral angle-closure glaucoma. Recognition of an anterior chamber configuration of a modestly shallowed central chamber with marked peripheral shallowing clinically suggested uveal effusion. B-scan echography provided definitive, confirmatory evidence of diffuse choroidal thickening with ciliochoroidal effusion. Treatment with aqueous suppressants, cycloplegics, and topical steroids resulted in complete resolution of the angle closure and reversal of induced myopia. The patient, who was systemically well without signs of AIDS or AIDS-related complex, was later tested and found to be serologically-positive for the human immunodeficiency virus.

Pharmacokinetic study of recombinant human epidermal growth factor in the anterior eye.

This study investigated the pharmacokinetics of recombinant human epidermal growth factor (EGF) in the eyes of live rabbits and enucleated human eyes. After intracameral injection of 1.2 muCi (1.1 micrograms) of 125I-EGF into rabbit eyes, tissues in the anterior segment were analyzed for uptake of 125I-EGF after 1, 2, 4, and 24 hr. The half-life of EGF during the initial 4 hr was 1.3 +/- 0.6 hr (+/- the confidence limit at P = 0.05) in the cornea, 0.7 +/- 0.4 hr in the iris, 1.9 +/- 0.9 hr in the lens, and 0.6 +/- 0.5 hr in the aqueous humor (n = 18 for each tissue). Between 4-24 hr, the percent retention of EGF in the tissues (relative to the initial amount in respective tissue) is in the order of lens greater than cornea greater than iris greater than aqueous humor. The kinetics of EGF uptake by excised human corneas was determined by incubating the endothelial surface with 1.2 muCi/ml (100 ng/ml) of 125I-EGF for 4 hr at 37 degrees C or 4 degrees C. After 4 hr, the total amount of EGF accumulated in the corneas was 2.3 +/- 0.2 ng (+/- standard deviation; n = 6) at 37 degrees C, and 1.4 +/- 0.1 ng at 4 degrees C.(ABSTRACT TRUNCATED AT 250 WORDS)

An evaluation of saline irrigation and comparison of povidone-iodine and antibiotic in the surface decontamination of donor eyes.

We evaluated the effect of saline irrigation on 38 pairs of donor eyes and determined the relative efficacy of antibiotic rinse/storage (neomycin-polymyxin B-gramicidin) and povidone-iodine immersion. Microbial growth was found in 49 (64.5%) of 76 eyes from which cultures were taken before irrigation, and only four (8.2%) became culture negative after irrigation. Of 27 eyes culture negative before irrigation, 15 (55.6%) became positive after irrigation. One eye of each pair was then assigned randomly to antibiotic rinse/storage and the other eye to 3-minute immersion in 1% povidone-iodine. Both antimicrobial treatments decreased the numbers of positive cultures, with povidone-iodine showing an advantage that did not reach significance. These results demonstrate that saline irrigation of donor eyes in situ does not reduce surface contamination and that povidone-iodine immersion should be considered as an alternative for donor eye decontamination.

Indications for penetrating keratoplasty: 1980-1988.

A retrospective analysis was undertaken of the clinical diagnoses of 1594 eyes that underwent penetrating keratoplasty performed in a private-referral corneal practice over a 9-year period, 1980-1988. The seven most common indications for surgery were keratoconus (24.0%), pseudophakic or aphakic bullous keratopathy (21.2%), corneal scarring (13.9%), Fuchs' endothelial dystrophy (12.5%), regraft (8.1%), and herpetic keratitis (5.3%). Keratoconus was the leading indication from 1980 to 1985. From 1985 to 1988, pseudophakic bullous keratopathy became the leading indication and correlates well with known complications associated with closed-loop anterior chamber lenses, which were widely used during the early 1980s. Less frequent indications for penetrating keratoplasty included the following: infectious (nonviral) keratitis (3.5%); acute or chronic ulcerative keratitis (2.7%); interstitial keratitis (1.8%); mechanical trauma (1.5%); other (non-Fuchs') corneal dystrophies (1.4%); congenital opacities (0.8%); and chemical burns (0.5%).

Surgical treatment of overcorrection following radial keratotomy: evaluation of clinical effectiveness.

Secondary hyperopia following radial keratotomy (RK) may result from an initial overcorrection or from a continued effect of the procedure with time. We retrospectively evaluated the effect of surgical intervention for the management of overcorrection following RK in six patients who had undergone reopening of all RK incisions, followed by irrigation and closure of the incisions with 10-0 mersilene sutures. Three patients with prior 4-incision RK, with follow-up of 4 to 26 months, were steepened a mean of 1.63 diopters keratometrically; three patients with prior 8-incision RK, with follow-up of 3 to 12 months, were steepened a mean of 1.46 D keratometrically. Comparison of intraoperative with postoperative keratometry demonstrated a substantial loss of effect with time, which stabilized by 2 months in 4-incision RK patients and by 3 months in 8-incision RK patients. Nevertheless, placement of interrupted sutures across reopened radial incisions consistently induced central corneal steepening in these six patients.

Scleral ectasia secondary to Acanthamoeba keratitis.

An unrecognized case of Acanthamoeba keratitis became quiescent after prolonged scleritis, resulting in a central corneal scar with extensive scleral ectasia. Twenty-one months after the onset of the sclerokeratitis, a penetrating keratoplasty was performed. Acanthamoeba cysts were found in the host corneal button. The corneal transplant has remained thin and clear for 2 1/2 years following surgery. Acanthamoeba keratitis extending to the limbus may become self-limited due to immunologic mechanisms available at the limbus, which do not appear to be active within the cornea itself. However, the prolonged inflammatory reaction manifesting as scleritis may result in extensive scleral ectasia.