The relationship of corneal epithelial defect size to drug penetration

OBJECTIVE

To determine the relationship between corneal epithelial defect size and corneal penetration of a triazole antifungal drug in an animal model.

METHODS

Corneas of adult rabbits were débrided of epithelium 25%, 50%, 75%, or 100% of surface area; the untreated fellow eye served as a control. Tritiated saperconazole was applied to each cornea every 5 minutes for 1 hour. The animals were killed and the cornea and aqueous of each eye were assayed for radiolabel activity.

RESULTS

Removal of 25% of the corneal epithelium produced an increase in corneal saperconazole concentration compared with eyes with intact epithelium. Increasing epithelial defect size from 25% to 50% produced a ninefold increase in mean corneal drug concentration (P = .0001). There was no further increase in corneal drug levels in eyes with 75% or 100% epithelial defects. A similar threshold effect was observed in aqueous drug concentration between 25% and 50% débridement (P = .0001).

CONCLUSION

In this experimental model, an apparent threshold was noted between 25% and 50% epithelial defect area, beyond which larger defects did not significantly increase drug penetration into the cornea or aqueous. This may be of clinical benefit in circumstances in which epithelial débridement is considered to enhance drug penetration.

Ocular pharmacokinetics of orally administered azithromycin in rabbits

Azithromycin was orally administered to Dutch-belted rabbits following extracapsular lens extraction in one eye. At various times the animals were sacrificed, and serum and ocular tissues were obtained for drug level determination by HPLC-EC. Following a single dose, peak levels of drug in ocular tissues were measured within 8 hours (cornea > 0.5 micrograms/g [15mg/kg]; > 1.5 micrograms/g [3Omg/kg]). Highest levels were obtained in iris and ciliary body ( > 15 micrograms). Measurable tissue levels persisted for at least 120 hours. Trough levels increased proportionately during drug multiple dose administration. Five days following five daily 15mg/kg doses, corneal levels exceeded 0.5 micrograms/g, and iris and ciliary levels were higher than 15 micrograms/g. Aqueous humor and serum levels were equivalent. Vitreous humor levels, though higher than aqueous humor, were consistently < 1 microgram/ml. Extracapsular cataract extraction did not significantly affect drug uptake.

The qualitative evaluation of the pharmacokinetics of subconjunctivally injected antifungal agents in rabbits

Dutch-belted rabbits with corneal epithelium either intact or debrided were injected subconjunctivally with 300 microliters of one of six antifungal agents: 10 mg/ml miconazole, 5 mg/ml fluconazole, 5 mg/ml ketoconazole, 2.5 mg/ml itraconazole, and 5 mg/ml amphotericin B. At intervals of 10 min to 96 h after injection, animals were killed and corneas removed at the limbus. Three vertical strips from the right cornea and four contiguous 3-mm disks trephined from the central vertical axis of the left cornea were placed on agar plates seeded with an appropriate indicator organism. After 24 h of incubation, the zones of inhibition were measured. For itraconazole, miconazole, fluconazole, saperconazole, and ketoconazole, central corneal levels peaked by 2 h in normal and debrided corneas. Little or no drug was detectable after 4-8 h, except for itraconazole, which persisted in the cornea for at least 24 h in both normal and debrided corneas. Peak levels of amphotericin B in the central cornea were achieved after 2 h in rabbits with debrided corneas, with no drug activity measured after 8 h. There were no detectable levels of drug found in the central corneas of rabbits with intact corneal epithelium. On the basis of this pilot study, the method offers a rapid approach to the screening of antifungal agents for possible use by subconjunctival injection.

Ocular pharmacokinetics of saperconazole in rabbits. A potential agent against keratomycoses

The ocular pharmacokinetics of saperconazole, an experimental lipophilic triazole with activity against filamentous fungi, including Aspergillus and Candida species, were evaluated in rabbits by radioassay. The drug was administered by topical, subconjunctival, and oral routes. Following a single 20-microL drop of 0.25% saperconazole in normal corneas, a mean (+/- SEM) peak level of 2.32 +/- 0.06 micrograms/g was achieved in 10 minutes. In débrided corneas, a peak level of 13.09 +/- 2.87 micrograms/g was achieved in 2 minutes. The drug was rapidly cleared from the cornea within 2 hours. The administration of 13 drops during 1 hour resulted in a threefold increase in normal corneal levels and in a sixfold increase in débrided cornea levels. Peak levels following subconjunctival injection in normal corneas (12.91 +/- 2.02 micrograms/g) were approximately twofold greater than those following sustained topical administration (6.19 +/- 0.16 micrograms/g) and, in débrided corneas, were a third higher than those following topical therapy in débrided corneas. Clearance was virtually complete by 8 hours. Levels following oral administration were low and probably subtherapeutic in all ocular tissues that were evaluated. Bioassay studies revealed that 44.17% of the drug in the cornea following topical administration was bioactive.

The evaluation of therapeutic responses in experimental keratomycosis

Two different measures of response to therapy were evaluated in a model of keratitis caused by Aspergillus fumigatus in Dutch-belted rabbits. Combined pre and post-inoculation treatment with oral fluconazole 37.5 mg/kg bid or itraconazole 40 mg/kg bid was compared to post-inoculation treatment only and untreated controls using a standardized clinical disease severity score and quantitative isolate recovery techniques. For both drugs, there was no difference in isolate recovery rates among all three groups. However, a significant improvement in clinical disease was noted in the pre and post-inoculation treatment group compared to controls (p less than .01) and to the post-inoculation group (p less than .05) for fluconazole. A similar trend, though not statistically significant, was apparent with itraconazole treatment. This disparity highlights the difficulties associated with measuring responses to therapy in keratomycosis and emphasizes the need for more sensitive and specific measures.

Anomalous effect of subconjunctival miconazole on Candida albicans keratitis in rabbits

After intrastromal injection of a standardized inoculum of Candida albicans blastoconidia in Dutch-belted rabbits, the efficacy of subconjunctival therapy with amphotericin B and miconazole was evaluated using a quantitative isolate recovery technique. The subconjunctival injection of miconazole resulted in the recovery of significantly more viable organisms compared to the number recovered in controls (2,000 micrograms daily, P less than .02; 3,000 micrograms daily, P less than .01, respectively). Amphotericin B, in contrast, in dosages of 500 or 1,500 micrograms daily, had a significant therapeutic effect (P less than .03). The anomalous response to subconjunctival therapy with miconazole seen in this model suggests an effect on host defense mechanisms.

Influence of corticosteroid on experimentally induced keratomycosis

To assess the effect of corticosteroid on the establishment of experimentally induced keratomycoses, rabbits were injected subconjunctivally with triamcinolone acetonide on two successive days before inoculation with Candida albicans, Aspergillus fumigatus, or Fusarium solanae. Whereas isolate recovery rates declined steadily in normal control corneas, they remained stable over 15 days in corticosteroid-treated corneas. Clinically, inflammation was equivalent (A fumigatus and F solanae) or significantly less (C albicans; P = .001) until the 10th day. At 15 days, inflammation in corticosteroid-treated corneas was significantly worse in animals infected with A fumigatus (P = .003) or F solanae (P = .02). Inflammatory signs correlated inconsistently with isolate recovery. Pathogenicity of the infecting organism appears to be important in determining the degree to which corticosteroid is able to mask clinical signs of infection while enhancing fungal replication.

Toxicity and pharmacokinetics of subconjunctival amphotericin B. An experimental study

The pharmacokinetics and toxicity of subconjunctival (S/C) amphotericin B (AmB) were evaluated in Dutch-belted rabbits. Following the S/C injection of 1,500 micrograms of AmB, corneal and aqueous levels were determined by bioassay. The highest levels were present in the periphery of debrided corneas at 1 h (90.12 +/- 2.4 micrograms/g). The debrided central cornea contained 30.84 micrograms/g, almost double the amount present in the intact central cornea. These levels were transient; in the central intact cornea only 2.08 micrograms/g could be detected at 2 h. Peak aqueous levels were low (0.95 +/- .24 micrograms/mL in debrided corneas at 1 h). The S/C injection of 1,500 micrograms of AmB in sodium deoxycholate produced a severe inflammatory response in the conjunctiva, episclera, iris, anterior chamber, and superior rectus muscle that persisted 10 days. Injection of sodium deoxycholate alone produced a similar but less severe response.

Differences in response in vivo to amphotericin B among Candida albicans strains

A group of ten Candida albicans strains previously determined to be resistant or susceptible to topical amphotericin B in vivo and in vitro were exposed to treatment with different concentrations of the drug in a quantitative model of candidal keratitis in Dutch-belted rabbits. After 5 days of topical treatment with amphotericin B eye drops in concentrations of 0.3%, 0.03%, or 0.003%, quantitative isolate recovery in treated animals was compared with that of untreated controls. A dose response was observed for all five susceptible strains. The two strains that were most sensitive to amphotericin B in vitro also were the most susceptible in vivo. At each dose level there was a two- to eightfold reduction in isolate recovery among highly susceptible strains compared with less susceptible strains (P less than 0.05). The five resistant strains remained so even when the 0.3% concentration was used. Among strains of C. albicans susceptible to amphotericin B, there appeared to be a variation in degree of susceptibility in vivo that correlated with the minimum inhibitory concentration.

An evaluation of intrastromal injection of antifungal agents

The effect of the intrastromal injection of a mixture of amphotericin B, miconazole, natamycin, and nystatin and the component agents alone was studied in Dutch-belted rabbits by standard bioassay. Twenty-four hours following injection, the inhibiting effect of the mixture was equivalent to that of nystatin alone. All other agents exhibited inferior activity. The intrastromal injection of this mixture appears to offer no therapeutic advantage.

Ocular uptake of fluconazole following oral administration

The ocular penetration and distribution of oral fluconazole was studied in Dutch-belted rabbits. Measured by high-pressure liquid chromatography, fluconazole readily penetrated all ocular tissues and fluids. No difference was observed between the levels obtained in phakic and aphakic eyes. Four hours after a single oral dose of 20 mg/kg, the mean levels and SEs were as follows: cornea, 13.3 +/- 1.4 micrograms/g; aqueous, 7.4 +/- 0.3 mg/L; vitreous, 9.8 +/- 0.9 mg/L; and choroid/retina, 5.2 +/- 0.4 micrograms/g. These levels were approximately twice those obtained with a 10-mg/kg dose. The corneal concentrations correlated highly with serum levels (r = .89). A steady accumulation in both normal corneas and corneas infected with Candida albicans was noted when 17.5 mg/kg of fluconazole was administered twice daily over a 5-day period. Drug levels did not increase in the cornea when fluconazole was administered as a single daily dose of 35 mg/kg. In view of its excellent ocular pharmacokinetic profile, fluconazole merits further attention as an orally administered agent for ocular fungal infections.

Measurements of antifungal levels in corneal tissue: a simplified bioassay for amphotericin B

Measuring precise antifungal levels in the cornea with broth-dilution bioassays is difficult, as standard techniques involving visual determination of endpoints are hindered by corneal debris. To increase the precision of the measurement, we modified the sample preparation for bioassay of rabbit corneas treated with subconjunctival amphotericin B. Endpoint determination and variance were compared for a freshly thawed corneal suspension and the supernatant after 24 h equilibration; bioassay of the corneal suspension after 24 h equilibration served as an additional control. All endpoints were read visually in a masked fashion and were verified by culture. The three methods gave comparable endpoint values with equivalent degrees of variance. Amphotericin B levels were consistent by both visual and culture determination; however, endpoints were clearly visible and easier to read for the supernatant. Visual determination of the endpoints for the supernatant following 24 h equilibration simplified and ensured the precision of the bioassay technique.

Corneal toxicity of propamidine

Two patients with Acanthamoeba keratitis developed a corneal abnormality following prolonged treatment with topical 0.1% [corrected] propamidine isethionate. In both instances, withdrawal of drug therapy resulted in a gradual clearing of the keratopathy, with no permanent sequelae. The changes we observed may be confused with those of active Acanthamoeba infection.

In vitro and in vivo susceptibility of Candida keratitis to topical polyenes

The susceptibility of Candida albicans to topical amphotericin B and natamycin was evaluated in a model of stromal keratitis in Dutch-belted rabbits and compared with minimal inhibitory concentrations in vitro. Treatment was delayed 24 hr to allow invasive disease to occur and was then continued for 5 days. Ten strains of Candida albicans comprised the test panel. For amphotericin B, the minimal inhibitory concentration (MIC) by tube dilution classified the same strains as resistant or susceptible as did the in vivo response. A dose-response was observed with different concentrations of the drug. For natamycin, the MIC misclassified two strains. The rate of administration of natamycin required in this model was much higher than for amphotericin B, a therapeutic effect being observed with natamycin only when the drug was administered every 30 min during the in vivo efficacy and in vitro susceptibility with these strains is in agreement with that observed in the authors' previous studies using a model of immediate treatment.

Correlation of in vitro and in vivo susceptibility of Candida albicans to amphotericin B and natamycin

The efficacy of topical 0.15% amphotericin B and 5% natamycin was examined in a model of Candida keratitis in rabbits and correlated with three tests of in vitro susceptibility: tube dilution minimal inhibitory concentration (MIC), minimal fungicidal concentration (MFC) and agar diffusion zones of inhibition. For a panel of 17 strains, the MIC classified precisely the same strains as resistant or susceptible to amphotericin B as did the in vivo response. Several strains were misclassified using the MFC and the zone of inhibition. For natamycin the MIC misclassified two strains but it was still superior to the other two tests. For all strains, amphotericin B was equal or superior in efficacy to natamycin in vivo. The tube dilution MIC for amphotericin B was a reliable indicator for natamycin efficacy in vivo.

An outbreak of Candida parapsilosis endophthalmitis: analysis of strains by enzyme profile and antifungal susceptibility

Twenty-two isolates from patients with postsurgical endophthalmitis due to infection with Candida parapsilosis as a result of exposure to a contaminated ocular irrigating solution were classified by enzyme profile analysis and antifungal susceptibility. These isolates were identical to a single isolate obtained from a contaminated vial but could be differentiated, on the basis of enzyme profile and antifungal susceptibility, from randomly selected stock isolates. The combination of these tests appears to have value in discriminating epidemic from non-epidemic strains.

The influence of yeast growth phase in vivo on the efficacy of topical polyenes

We compared the efficacy of two polyenes, amphotericin B and natamycin, in two models of yeast infection. In one, treatment was begun immediately after inoculation, in the other it was delayed 24 hours. In each model infection with Candida albicans was established in the corneal stroma of dutch-belted rabbits and treated topically with 5% natamycin or amphotericin B 0.15% and 0.075%. Quantitative isolate recovery techniques were used to assess response after 5 days of treatment. A significant therapeutic effect was present for amphotericin B in both models. However, delayed treatment with natamycin was ineffective using treatment schedules efficacious when begun 1 hour after inoculation. A therapeutic effect was present only with administration of the drug every 1/2 hr. This altered response may reflect a difference in susceptibility between different growth phases in yeasts.

Herpes simplex masquerade syndrome: acanthamoeba keratitis

Acanthamoeba is a free-living, fresh-water protozoan that can cause severe corneal disease. Acanthamoeba keratitis can closely mimic epithelial and stromal Herpes simplex keratitis. Three cases of severe keratitis, were referred for treatment. One patient presented with a pseudodendritic epithelial lesion that gradually progressed to stromal involvement. A second patient presented with central stromal infiltrate and necrosis, while a third exhibited features of a disciform lesion with the later development of an immune ring. Acanthamoeba was recovered from the cornea in each case. The distinctive characteristics of the history and clinical findings in Acanthamoeba keratitis can aid the clinician in distinguishing between these two clinical entities. Cytopathology and special staining and culture techniques can confirm the diagnosis.

Corneal penetration of topical amphotericin B and natamycin

The corneal uptake and penetration of 14C-labelled 0.15% amphotericin B and 5% natamycin were studied in Dutch-belted rabbits. Corneal levels of natamycin were substantially higher than amphotericin B. For both drugs, these levels were influenced by corneal contact time. In corneas debrided of epithelium, both agents entered the corneal stroma and levels were detected in aqueous in the therapeutic range. However, in corneas with intact epithelium, penetration was negligible for amphotericin B (0.23 microgram/gm at 2 mins). By contrast, although penetration of natamycin was greatly reduced, 7.0 micrograms/gm were present at 2 mins.

Bioavailability and penetration of topical amphotericin B in the anterior segment of the rabbit eye

The penetration of topical amphotericin B was studied in Dutch-belted rabbits using bioassay and radioassay techniques. Seven percent of total amphotericin B in the corneas, as measured by recovered 14-C, and 5% in the aqueous was in a bioactive form. Drug was not detectable by bioassay after single drop administration but after a loading dose approach, drug was present in therapeutic amounts. In inflamed corneas, a pass-through effect was noted with higher levels initially but rapid fall-off subsequently.

Statistical analysis for experimental models of ocular disease: continuous response measures

Experimental designs in ophthalmologic research frequently treat both eyes of a subject in the same fashion: e.g., therapy with a specific drug or control. In these two-eye designs, observations from the same subject are often positively correlated. Failure to account for this correlation is a serious error which overstates the precision of studies, resulting in falsely significant results. This paper reviews the statistical methods appropriate for studies where endpoints are quantitative. We present: (1) the use of analysis of variance (t-test when there are 2 treatment conditions) to estimate differences between all experimental treatments, (2) the use of contrasts to estimate differences between specific treatments, and (3) methods for analysis of data from multiple experiments. Because of the ubiquity of incorrect analysis of data from two-eye designs in the ophthalmologic research literature and the serious consequences of this error, we propose a limited statistical review of manuscripts to ascertain if the statistical analysis matched the experimental design.

Intraocular penetration of systemically administered antifungal agents

Amphotericin B, 5-flucytosine (5-FC), and ketoconazole levels were estimated in vitreous and aqueous samples taken from four patients undergoing therapeutic vitrectomy for fungal endophthalmitis. The levels of amphotericin B in the vitreous of three patients were low (.04 - .17 microgram/ml). However, 5-FC was present in a concentration of 22.2 micrograms/ml in one patient. In another case the aqueous level of ketoconazole was 0.35 microgram/ml. The vitreous in the same patient contained 0.71 microgram/ml of the drug.