David G. Hwang

Identification of antiangiogenic and antiinflammatory proteins in human amniotic membrane.

Purpose. To identify the potential antiangiogenic and antiinflammatory proteins expressed in human amniotic membrane tissue. Methods. Human amniotic epithelial and mesenchymal cells were isolated from human amniotic membranes by sequential trypsin and collagenase digestion. Total RNAs were harvested from freshly obtained human amniotic epithelial and mesenchymal cells. Antiangiogenic and antiinflammatory proteins were detected by the reverse transcriptase-polymerase chain reaction (RT-PCR) technique and further confirmed by DNA sequencing of PCR-amplified transcripts. The distribution of tissue inhibitors of metalloproteinase (TIMPs) were studied further by immunohistochemistry performed on paraffin-embedded amniotic membrane tissue. Results. RT-PCR results showed that both human amniotic epithelial and mesenchymal cells express interleukin-1 receptor antagonist, all four TIMPs, collagen XVIII, and interleukin-10. Thrombospondin-1 was expressed in all of the epithelial cell specimens and in one out of five mesenchymal cell specimens. Furthermore, immunohistochemistry studies performed on freshly prepared amniotic membrane confirmed that all members of the TIMP family were present in epithelial and mesenchymal cells as well as in the compact layer of the amniotic stroma. In cryopreserved amniotic membranes, positive staining was seen in residual amniotic cells and stroma. Conclusions. Human amniotic membrane epithelial and mesenchymal cells express various antiangiogenic and antiinflammatory proteins. Some of those proteins also were found in amniotic membrane stroma. These findings may explain in part the antiangiogenic and antiinflammatory effects of amniotic membrane transplantation.

Effect of donor age on morphologic variation of cultured human corneal endothelial cells

Purpose. To examine the effect of donor age on the morphologic variation of cultured human corneal endothelial cells (HCEC). Methods. HCEC were obtained from the remaining corneoscleral rims of seven human corneas used for penetrating keratoplasty. The donor age ranged from 2 to 75 years. Primary cultures were established from explants of the endothelial cell layer, including the Descemets membrane, and were propagated on culture dishes coated with bovine corneal endothelial extracellular matrix. At the fourth passage, frequency distribution of cell area in the confluent monolayer was calculated and the effect of donor age on cell area was analyzed. Results. The percentage of HCEC with cell area over 2000 &mgr;m2significantly increased with donor age (r = 0.935, p = 0.0007). Conclusion. Cultured HCEC established from older donor tissue display greater heterogeneity. The use of HCEC from younger donors may be preferable to maximize the benefits of HCEC transplantation.

Late-onset Corneal Haze and Myopic Regression After Photorefractive Keratectomy (prk)

Objective: To report the incidence and clinical course of a series of patients who developed both delayed-onset, clinically significant progressive haze and myopic regression after photorefractive keratectomy (PRK). Methods: In this retrospective case series, the charts of 542 consecutive patients who had undergone PRK with the VISX Star Excimer or Nidek EC-5000 laser between July 1996 and October 1998 and who had a minimum of 6 months of follow-up were reviewed. Ten eyes of 8 patients developed progressive haze to greater than 1+ and myopic regression equal to or more than −1 D 3 months or more after PRK. The historical and clinical features were reviewed. Results: The incidence of combined progressive haze and myopic regression was 1.8%. The average age was 40.5 years. Three of the 8 patients were female. The median spherical equivalent (SE) attempted correction was −6.69 D (range −4.00 to −12.25 D). Five patients who underwent bilateral PRK had unilateral involvement. The mean SE regression was −2.01 ± 0.79 D (range −1.00 to −3.00 D). Regression plateaued at a mean of 9.8 months. Haze ranging up to 4+ peaked at a mean of 7.4 months. Topical steroid treatment and/or epithelial scraping was attempted in 3 eyes but was ineffective. Conclusions: Combined delayed-onset progressive haze and myopic regression can occur after PRK. In such cases, the amount of haze appears to correlate with the magnitude of attempted initial correction (r = 0.639, P = 0.046) although not with the magnitude of subsequent regression. Patients may need at least 10 months of follow-up to achieve a stable refraction and level of haze. These observations suggest a need for improved understanding of corneal wound healing following PRK and of biologic factors that may contribute to variability in outcomes.

In vitro susceptibilities of bacterial ocular isolates to fluoroquinolones.

Purpose. To evaluate and compare the in vitro antimicrobial activity of levofloxacin versus ciprofloxacin and ofloxacin against ocular isolates from patients with bacterial conjunctivitis. Methods. The in vitro antimicrobial susceptibilities of ocular isolates to levofloxacin, ofloxacin, and ciprofloxacin were determined using both the agar disk diffusion and broth dilution methods. Results. Disk diffusion susceptibility testing disclosed that 99% (100 of 101 isolates) of Gram-negative isolates and 98% (127 of 129 isolates) of Gram-positive isolates were susceptible to levofloxacin; 96% (97 of 101 isolates) of Gram-negative isolates and 78% (100 of 129 isolates) of Gram-positive isolates were susceptible to ofloxacin; and 94% (95 of 101 isolates) of Gram-negative isolates and 61% (79 of 129 isolates) of Gram-positive isolates were susceptible to ciprofloxacin. Broth dilution testing disclosed that 99% (72 of 73 isolates) of Gram-negative isolates and 98% (111 of 113 isolates) of Gram-positive isolates were susceptible to levofloxacin; 96% (70 of 73 isolates) of Gram-negative isolates and 92% (104 of 113 isolates) of Gram-positive isolates were susceptible to ofloxacin; and 95% (69 of 73 isolates) of Gram-negative isolates and 82% (93 of 113 isolates) of Gram-positive isolates were susceptible to ciprofloxacin. Conclusions. In this study, levofloxacin demonstrated superior in vitro activity against human bacterial conjunctival isolates compared with either ofloxacin or ciprofloxacin (levofloxacin > ofloxacin > ciprofloxacin).

Aspergillus fumigatus keratitis after laser in situ keratomileusis.

Purpose. To report a case of Aspergillus fumigatus keratitis after a laser in situ keratomileusis (LASIK) enhancement procedure. Method. Case report. Results. A 56-year-old woman developed an ulcer in the flap 13 days after LASIK enhancement. A 4-week course of fortified antibiotics for a presumed bacterial infection followed. The ulcer progressed, causing 60% thinning of the corneal stroma. A biopsy was performed 5 weeks after onset of symptoms, and antifungal agents were initiated. Cultures showed A. fumigatus. Her cornea perforated after the biopsy, requiring cyanoacrylate and lamellar overlay sutures, but the infiltrate resolved on antifungal agents. Conclusion. This report is the first description of Aspergillus keratitis after LASIK. We hypothesize that the infection became established on the stromal bed during surgery and led to melting, anteriorly through the flap and posteriorly through the stroma. Diagnosis was made by a corneal biopsy and inoculation of a wide array of media. This case demonstrates the need to consider atypical organisms, including fungi, in the differential diagnosis of post-LASIK infections when there is no response to therapy and highlights the role of corneal biopsy and flap lifting in the diagnosis of this condition.

Objective assessment of transient corneal haze and its relation to visual performance after photorefractive keratectomy

PURPOSE Photorefractive keratectomy has the potential to cause transient corneal haze. The purpose of this study was to evaluate the relationship between transient corneal haze as measured by an objective means and high and low contrast visual performance. METHODS In a prospective study, 44 eyes of 28 patients were examined preoperatively and at 1, 3, 6, and 12 months after photorefractive keratectomy. Five laser in situ keratomileusis and two intrastromal corneal ring segments (Intacs [KeraVision, Fremont, CA]) were included for comparison, because these procedures are not expected to cause haze. Haze was measured using a prototype objective hazemeter, TSPC-3, a modification of the Nidek EAS-1000. Visual performance was measured using high-contrast visual acuity and the Rabin Small Letter Contrast Test. RESULTS Corneal haze was greatest at the 1-month examination and was consistent with a decrease in visual performance on both tests. Corneal haze resolved in 82% of eyes by 10 +/- 4 months after photorefractive keratectomy. However, visual performance had not returned to preoperative levels in 65% and 81% of these eyes on the high-contrast visual acuity test and the Small Letter Contrast Test, respectively. Eyes that underwent laser in situ keratomileusis and Intacs did not develop corneal haze; however, visual decrements were measured. CONCLUSIONS As a clinical tool, the TSPC-3 hazemeter objectively measures very subtle changes in haze levels. Corneal haze appears to account for only approximately 50% of visual performance changes in the early healing period after photorefractive keratectomy. Other factors, namely topographic abnormalities, are more likely to be an important cause of persistent visual disturbances.

Complex limbal choristomas in linear nevus sebaceous syndrome

OBJECTIVE This study aimed to describe the clinical and histopathologic findings in four patients with complex limbal choristomas associated with linear nevus sebaceous syndrome (LNSS), a rare disorder including nevus sebaceous, seizures, and mental retardation, and often accompanied by ocular anomalies. DESIGN Small observational case series. METHODS A retrospective review of the clinical and histopathologic records of four patients. RESULTS Each of four patients had complex limbal choristomas in the setting of clinical and histopathologic LNSS. The limbal choristomas were multiple in three patients and bilateral in two patients. Most choristomas involved the superotemporal limbus (6 of 10), although nasal (3 of 10) and inferior (1 of 10) limbal tumors also were present. Three patients had significant corneal astigmatism or involvement of the central cornea requiring surgical removal of their choristomas, one accompanied by a lamellar keratoplasty and another accompanied by two consecutive penetrating keratoplasties. Each graft eventually vascularized, reducing vision. One patients vision was limited by amblyopia and another by occipital cortical dysgenesis with visual impairment. Histopathologic examination of the excised choristomas showed foci of lacrimal gland (3 of 4 patients), adipose tissue (3 of 4), neural tissue (1 of 4), cartilage (1 of 4), lymphoid follicles (1 of 4), skin adnexal tissue (1 of 4), and smooth muscle (1 of 4). Other associated ocular findings included an eyelid mass (1 of 4), colobomas of the eyelid (3 of 4), colobomas of the choroid and retina (2 of 4), nonparalytic strabismus (2 of 4), scleral ectasia (1 of 4), partial oculomotor palsy with ptosis and anisocoria (1 of 4), microphthalmia (1 of 4), hypertelorism (1 of 4), and cortical visual impairment (1 of 4). CONCLUSIONS Complex limbal choristomas, although rare, can occur in the setting of LNSS and can be associated with multiple ocular and systemic abnormalities. Visual prognosis appears poor in most cases despite aggressive management.

Expression of Matrix Metalloproteinases 2 and 9 and Tissue Inhibitors of Metalloproteinase 1 and 2 in Inflammation-Induced Corneal Neovascularization

Purpose: Matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinase (TIMPs) have been linked to the angiogenic process in general. In order to understand the potential roles of MMP-2, MMP-9 and TIMPs in the corneal neovascularization process, we examined the expression and activities of MMP-2, MMP-9 and TIMPs during the course of cauterization-induced corneal neovascularization in a rat model. Methods: Neovascularization of rat corneas was induced by silver nitrate cauterization. The expression of MMP-2, MMP-9, TIMP-1 and TIMP-2 was examined by immunohistochemistry and RT-PCR. The protein activities of MMPs and TIMPs were compared in pre- and postcauterization corneas by gelatin zymography and reverse zymography, respectively. Results: MMP-2, TIMP-1 and TIMP-2 immunoreactivities were expressed in normal corneas, predominantly in the corneal epithelium. After injury, immunoreactivities of both MMPs and TIMPs were increased, notably in the healing corneal epithelium, infiltrating inflammatory cells, stromal fibroblasts and ingrowing vascular endothelial cells. The increase in gross MMP-2 enzymatic activity paralleled the maximal vascular ingrowth on day 4, while the gross MMP-9 enzymatic activity rose immediately on day 1, then decreased steadily, which paralleled the magnitude of inflammatory cell infiltration. The immunoreactivity of MMPs/TIMPs decreased significantly 2 weeks after cauterization. On day 35, MMP-2, TIMP-1 and TIMP-2 staining was seen only in corneal epithelium and vascular endothelial cells. Both the RT-PCR and reverse zymography results revealed a more constant expression of TIMP-2, while the TIMP-1 expression appeared to be more inducible. Conclusion: MMPs as well as TIMPs were upregulated in cauterization-induced corneal neovascularization, suggesting that both may participate in extracellular matrix remodeling in the corneal wound healing, inflammation and neovascularization processes.

Pharmacokinetics of Azithromycin and Moxifloxacin in Human Conjunctiva and Aqueous Humor During and After the Approved Dosing Regimens

PURPOSE To evaluate the ocular pharmacokinetics of azithromycin and moxifloxacin in human conjunctiva and aqueous humor in subjects undergoing cataract surgery. DESIGN Multicenter, open-label, randomized study. METHODS Subjects scheduled for routine cataract surgery and with normal-appearing conjunctiva were eligible. One conjunctival biopsy sample and 1 aqueous humor sample were obtained from subjects randomly assigned to 1 of 10 prespecified time points (1 to 312 hours) after treatment initiation of azithromycin ophthalmic solution 1% or moxifloxacin ophthalmic solution 0.5%. Samples were assayed using liquid chromatography tandem mass spectrometry. RESULTS Azithromycin 1% provided high concentrations (peak level, 559.7 μg/g) in human conjunctiva that were sustained at levels 1 to 2 orders of magnitude higher than those of moxifloxacin 0.5% throughout the 7-day dosing period and for at least 7 days thereafter. Azithromycin also showed an extended half-life (65.7 hours) in conjunctiva relative to that of moxifloxacin (28.6 hours). Accordingly, the concentration of azithromycin was maintained well above the minimum inhibitory concentration required for inhibition of growth of 90% of tested bacterial isolates for at least 7 days, whereas moxifloxacin conjunctival levels fell to levels at or less than the minimum inhibitory concentration required for inhibition of growth of 90% of tested bacterial isolates approximately 24 hours after the last dose. Peak aqueous humor concentration of moxifloxacin was higher (0.77 μg/mL) than that of azithromycin (0.053 μg/mL). No clinically relevant safety findings were observed. CONCLUSIONS Azithromycin 1% demonstrated high, therapeutic levels in the conjunctiva that were maintained up to 7 days after completion of a 1-week dosing regimen. Aqueous humor levels, however, were subtherapeutic with this dosing regimen. In comparison, moxifloxacin achieved lower conjunctival tissue levels, but higher aqueous humor levels.

Flap haze after epithelial debridement and flap hydration for treatment of post-laser in situ keratomileusis striae.

Purpose. To report a case of corneal haze after attempts to eliminate post–laser in situ keratomileusis (LASIK) lamellar striae. Methods. Case report. Results. A 24-year-old woman with visually significant flap striae 2 months after LASIK underwent manual epithelial debridement and flap hydration, refloating, and stretching to eliminate the striae. Three weeks after this intervention, she developed visually significant haze that was confined to the stroma of the flap. The haze slowly improved with use of a topical steroid. Conclusion. Stromal haze can develop after treatment of flap striae with epithelial debridement and hypo-osmolar irrigation. We speculate that these maneuvers may have induced cell death of anterior keratocytes and led to haze formation, as can occur after simple epithelial debridement and epithelial scrape-photorefractive keratectomy.