Use of Collagen Shields Containing Amphotericin B in the Treatment of Experimental Candida albicans-induced Keratomycosis in Rabbits U w e Pleyer, M.D., A n n e Legmann, M.D., Bartly J. M o n d i n o , M.D., and David A. Lee, M.D. We evaluated the effect of collagen shields presoaked with amphotericin B on the treat ment of experimental Candida albicans-in duced keratitis. Treatment results were com pared to those of amphotericin B eyedrops instilled hourly. Forty-eight albino rabbits received intrastromal injections of 108 C. albicans organisms. Twenty-four hours later, eyes were treated for eight hours each day with hourly instillation of 0.15% amphotericin B drops, hourly instillation of saline drops, or application of a collagen shield presoaked in 0.5% amphotericin B for one hour. The rabbits were killed after one, three, or five days of treatment. Quantitation of fungi in the cornea was achieved by culturing homogenates and counting colony-forming units. Treatment with amphotericin B applied either as hourly instilled drops or absorbed in collagen shields significantly (P < .05) reduced corneal fungal counts at all time points when compared to saline-treated control eyes. Rabbit eyes treat ed with amphotericin B-soaked collagen shields had significantly lower fungal counts compared with hourly instilled amphotericin B drops at Days 1 (P = .02) and 3 (P = .04), but
Accepted for publication Dec. 30, 1991. From the Jules Stein Eye Institute and the Department of Ophthalmology, UCLA School of Medicine, Los An geles, California. This study was supported in part by National Eye Institute grants EY4606 (Dr. Mondino) and EY07701 (Dr. Lee); t h e Wasserman Fund, Los Angeles, California (Dr. Mondino); the Lucille E. Simon Research Fund, Los Angeles, California (Dr. Lee); a n d t h e Deutsche Forschungsgemeinschaft, PL 1 5 0 / 1 , Tubing en, Germany (Dr. Pleyer). This study was presented at the International Symposium of the German O p h t h a l mological Society, Miinster, Germany, Sept. 2 1 , 1991. Reprint requests to Bartly J. Mondino, M.D., Jules Stein Eye Institute, 100 Stein Plaza, Los Angeles, CA 90024-7003.
not at Day 5. The collagen shields were as effective in reducing the number of colonyforming units as were amphotericin B drops at Day 5. These data suggest that collagen shields soaked in amphotericin B could be a useful and convenient treatment device in keratomycosis such as that caused by C. albicans. JXERATOMYCOSIS remains a treatment chal lenge. Although many antifungal agents are available, treatment with these agents is com plicated by a narrow spectrum of activity, lack of effective penetration into the eye, and toxicity. Polyenes, which bind to the sterol moiety in the cell membrane of the organism and induce an osmotic pressure change of the cell constitu ents, possess the greatest antifungal activity in vitro. 1 Amphotericin B is the most commonly used polyene, and its efficacy against most yeasts is well documented. 2,3 The toxicity of this agent after topical use in keratomycosis is now circumvented by preparations with low concen trations. 3 Because penetration into ocular tissue by systemic or periocular injection is limited, frequent topical applications are necessary to reach adequate intracorneal concentrations. The recommended initial treatment of kerato mycosis is one drop of 0.15% amphotericin B every 30 minutes for the first three days. 3 Frequent topical applications are time-con suming and inconvenient. To alleviate these problems and provide reliable delivery of medi cation, drug delivery by collagen shields is being explored. Our previous study suggested that amphotericin B delivery to the cornea by collagen shields is comparable to frequent drop therapy over a six-hour period. 4 In this study, we determined the effectiveness of collagen shields containing amphotericin B in the treat ment of experimental C. albicans-induced kera titis.
©AMERICAN JOURNAL OF OPHTHALMOLOGY 113:303-307, M A R C H ,
1992
303
304
AMERICAN JOURNAL OF OPHTHALMOLOGY
Material and Methods We used a C. albicans strain that was suscepti ble to amphotericin B in vitro and was obtained from a patient at the UCLA Medical Center. The minimal inhibitory concentration (evaluated by the tube dilution method) was 0.08 ^.g/ml. The isolated fungi were kept at - 7 0 C; to avoid mutation and drug resistance, only prepara tions of the original strain without further pas sages were used. One day before inoculation, the fungi were plated onto Sabouraud dextrose agar and incubated at 37 C. Blastoconidia were harvested and suspended in normal (0.9%) sa line solution to yield a concentration of 5 X 109 colony-forming units/ml, and were stored at 4 C overnight. To verify the concentration of organisms in the inoculum, the number of colo ny-forming units/ml was determined on the next day no more than two hours before injec tion. A total of 48 healthy adult New Zealand (3- to 4-kg) rabbits were used. The rabbits were housed in wire-bottomed cages with controlled light/dark cycles, fed by standard laboratory diet, and given free access to tap water. All experiments were performed in accordance with current animal care guidelines and heeded protocols approved by the UCLA Animal Re search Committee. Before intracorneal injec tion, general anesthesia was induced in the rabbits by an intramuscular injection of 2 ml of chlorpromazine hydrochloride followed by in travenous injection of 3 ml of Equi-Thesin (UCLA Pharmacy, Los Angeles, California). Af ter topical anesthesia was achieved in the rab bits with proparacaine hydrochloride 0.5%, the eye was gently proptosed. An 8.5-mm trephine blade was used to create a circular corneal incision approximately 0.1 mm in depth. With in this area, the epithelium was completely removed using a sterile cotton-tip applicator to enhance drug penetration. A 30-gauge needle attached to a 1-ml tuberculin syringe was in serted into the corneal stroma at the edge of the trephine line and advanced toward the central cornea before injection of the suspension. Each cornea of each rabbit was inoculated with 20 (il of C. albicans in normal (0.9%) saline solution at a concentration of 5 X 109 organisms per milliliter. If penetration into the anterior cham ber occurred, the rabbit was removed from the study and replaced. Treatment was delayed for 24 hours after inoculation to allow transforma
March, 1992
tion of the organisms to the hyphal phase of the fungi and invasion of the corneal stroma as previously reported by O'Day and associates. 6 Gentamicin ophthalmic ointment was applied to each eye after inoculation and once daily in the evening for the duration of the experiment. Amphotericin B was reconstituted in sterile water to yield a concentration of 0.5 mg/ml. Bio-Cor 24-hour corneal collagen shields (Bausch and Lomb, Tampa, Florida) were im mersed in the 0.5% amphotericin B solution for 60 minutes before use. Amphotericin B drops were prepared at the 0.15% (1.5 mg/ml) con centration recommended for clinical use. 5 The same amphotericin B concentrations used for drops and collagen shields were used in our previous study, which showed that equivalent amounts of drug could be delivered to rabbit corneas using these concentrations. 4 Each morning a fresh vial of amphotericin B was used to prepare the treatment solutions. Rabbits were randomly assigned to treatment with a single collagen shield presoaked in 0.5% amphotericin B, 0.15% amphotericin B drops, or 0.9% saline drops in both eyes. The rabbits were treated for eight hours a day for one, three, or five days. In 18 rabbits assigned to the collagen shield group, one collagen shield was placed on the center of each cornea for a period of eight hours each day, after which all rem nants of the collagen shield were removed. Eighteen rabbits were assigned to amphotericin B drop treatment. A five-drop (50 |xl/drop) loading dose of amphotericin B was instilled in each eye of each rabbit with a one-minute interval between drops. After the loading dose, one drop was applied to each eye hourly over an eight-hour period each day. Twelve control rab bits were treated with a loading dose of 0.9% saline drops and hourly saline drops over eight hours each day. Rabbits were killed after one, three, or five days of treatment. At each time point, the rabbits were killed by means of intravenously administered Equi-Thesin 16 hours after treat ment regimens were completed. Corneas for isolate recovery were placed in 1 ml of sterile 0.9% saline solution, minced within two hours with a scalpel, and thoroughly homogenized with a mortar and pestle. Serial dilutions of these corneal solutions from 10" 1 to 10" 9 were made in 0.9% sterile saline solution, and 100|xl aliquots were plated onto trypticase soy agar with 5% sheep blood. Each dilution was plated in duplicate. Plates were incubated for 48 hours
Amphotericin B-Collagen Shield and Candida Keratitis
Vol. 113, No. 3
at 37 C and the number of colony-forming units per cornea was counted using the last four countable plates. At each time point, two corne as from one randomly assigned rabbit were embedded in paraffin, sectioned into 4-ji.mthick sections, and stained with hematoxylin and eosin and Giemsa. The resulting counts in the right and left corneas were highly correlated (r = .80, P < .001). The counts/cornea from right and left corneas were averaged to obtain a single count/ cornea for each rabbit. Therefore, all analyses and sample sizes were based on rabbits, not corneas. Mean counts among the treatment groups were compared via analysis of variance and post hoc f-tests. Analysis was performed on the log (base 10) scale because log count/ cornea is better modeled by a normal distribu tion. Statistical significance was accepted at P < .05. The Statistical Analysis System software (SAS Institute, Inc., Cary, North Carolina) was used to perform all computations.
Results Immediately after inoculation of C. albicans, a circumscribed opacity was visible in the cornea. Control rabbits treated with saline drops devel oped corneal infiltrates that enlarged over the next three days and gradually decreased over the last two days of observation. All control eyes developed a hypopyon after three to five days of saline-drop instillations. The corneas of rabbits treated with amphotericin B-soaked shields or amphotericin B drops did not differ from the corneas of control eyes after only one day of treatment. After three days of treatment, a hypopyon developed in one of 12 eyes (6%) treated with amphotericin B-soaked shields and four of 12 eyes (33%) treated with ampho tericin B drops. After five days of treatment with amphotericin B-soaked collagen shields, four of 12 eyes (33%) developed hypopyon as compared to six of 12 eyes (50%) that were treated with amphotericin B drops. Central epithelial defects were present on Day 1 in all corneas administered amphotericin B-soaked shields, amphotericin B drops, or saline drops. At Day 3, the epithelial defects were healed in all three groups. Histologic studies of control eyes one day after inoculation demonstrated blastospores with pseudohyphae and true hyphae invading the corneal stroma
305
Fig. 1 (Pleyer and associates). Candida albicansinduced keratomycosis in control rabbit on Day 1 after inoculation. Blastospores are present at the injection site. Active transformation to hyphae is occurring with invasion into the corneal stroma (he matoxylin and eosin, original magnification x 330). (Fig. 1). Active stromal infiltration with mycelial growth accompanied by polymorphonuclear leukocytes was observed in control eyes five days after saline-drop application. Blastospores predominated in eyes treated for one day with amphotericin B-soaked shields or amphotericin B drops. Compared to control corneas, infiltration of the stroma and transfor mation to true hyphae were reduced. After three to five days of treatment with amphoteric in B-soaked shields or drops, hyphal growth was present but appeared less severe compared to that in control rabbits (Fig. 2). Quantification of C. albicans growth in con trol rabbits showed a high isolate recovery rate. Even after five days, an average of 417,000 colony-forming units/cornea were isolated from control eyes. At all time points, treatment with either am photericin B-presoaked collagen shields or am photericin B drops significantly reduced the number of colony-forming units/cornea (P < .05) when compared to control rabbits (Fig. 3). At Days 1 and 3, rabbits treated with amphoter icin B-soaked collagen shields had significantly lower fungal counts (P = .02 at Day 1, P = .04 at Day 3) compared to rabbits treated with hourly instilled amphotericin B drops. At Day 5, the number of colony-forming units/cornea was still lower in rabbits treated with collagen shields, but the difference was not statistically significant compared to rabbits treated with amphotericin B drops (P = .2).
306
March, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
10000
—•— —■—
c E
o u.
Amphotericin BCollagen shields Amphotericin 8Drops Saline Drops
100:
c
O O
Days of
Fig. 2 (Pleyer and associates). Candida albicansinduced keratomycosis in a rabbit treated for eight hours with a collagen shield presoaked in 0.5% amphotericin B. Treatment was begun 24 hours after inoculation. Invasion of the corneal stroma is re duced, and transformation of blastospores to hyphae diminished (hematoxylin and eosin, original magnifi cation x 300).
Discussion The m o d e l of k e r a t o m y c o s i s u s e d in this study is c o m p a r a b l e to that p r e v i o u s l y r e p o r t e d b y O ' D a y a n d associates. 6 In t h a t m o d e l of d e e p corneal infection, b l a s t o s p o r e t r a n s f o r m a t i o n to a h y p h a l p h a s e o c c u r r e d w h e n t r e a t m e n t was delayed for 24 h o u r s . T h u s , the m o d e l of infec tion parallels h u m a n k e r a t o m y c o s i s , w h i c h is c h a r a c t e r i z e d by corneal i n v a s i o n b y p s e u d o h y p h a e a n d t r u e s e p t a t e h y p h a e . U n l i k e fungal infections in h u m a n s , k e r a t o m y c o s i s in r a b b i t s resolves w i t h o u t t r e a t m e n t . In s t u d i e s u s i n g a similar infection m o d e l , k e r a t o m y c o s i s w a s still active even n i n e a n d 14 days after fungal infec tion. 6 ' 7 A m p h o t e r i c i n B a n d n a t a m y c i n are the m o s t effective a g e n t s for the t r e a t m e n t of k e r a t o m y cosis. A l t h o u g h n a t a m y c i n h a s g a i n e d favor as an antifungal d r u g b e c a u s e of its w i d e a n t i f u n gal effect a n d relatively low toxicity, a m p h o t e r icin B r e m a i n s a p o t e n t a g e n t in the t r e a t m e n t of keratomycosis. Amphotericin B has a wider s p e c t r u m of activity against yeast-like fungi t h a n any other c u r r e n t l y available a n t i f u n g a l d r u g . A m p h o t e r i c i n B is preferred for t r e a t m e n t of infection caused by Coccidioides immitis, Histoplasma capsulatum, Blastomyces dermatitidis, C. albicans species, a n d o t h e r less c o m m o n fungi. 8 In a d d i t i o n to its a n t i f u n g a l effect, a m photericin B has immunoadjuvant properties a n d may have an i m m u n o p o t e n t i a t i n g effect. 9
Treatment
Fig. 3 (Pleyer and associates). Effect of treatment on isolate recovery of C. albicans in experimental kerato mycosis. Twenty-four hours after intracorneal inocu lation, treatment with 0.15% amphotericin B drops applied hourly, saline drops applied hourly, or a collagen shield presoaked in 0.5% amphotericin B was begun. Treatment was given for eight hours per day for one, three, or five days. Results are expressed as mean ± standard error of the mean colony-form ing units/cornea. Sample size equals five rabbits for each time point of the experimental groups and three rabbits for each time point of the saline-treated control group. O c u l a r p e n e t r a t i o n of n a t a m y c i n as well as a m p h o t e r i c i n B is limited. 1 0 In d e e p i t h e l i a l i z e d corneas, higher drug concentrations were found with natamycin than with amphotericin B. This finding, r a t h e r t h a n a s u p e r i o r antifun gal effect, m a y favor t h e use of n a t a m y c i n in k e r a t o m y c o s i s . I m p r o v e m e n t of the efficacy of a m p h o t e r i c i n B may, t h e r e f o r e , lie in its formu lation a n d m o d e of a p p l i c a t i o n . A t t e m p t s to i n c o r p o r a t e n a t a m y c i n into collagen s h i e l d s have b e e n unsuccessful. 4 Use of collagen s h i e l d s i m m e r s e d in a m p h o t e r i c i n B p r o d u c e d corneal c o n c e n t r a t i o n s t h a t w e r e h i g h e r or equivalent to h o u r l y instilled d r o p t r e a t m e n t in a p r e v i o u s study. 4 In t h e s i x - h o u r trial, corneal a m p h o t e r i c i n B c o n c e n t r a t i o n s o b t a i n e d b y col l a g e n shields w e r e h i g h e s t o n e h o u r after a p p l i cation a n d d e c l i n e d a f t e r w a r d s . Even after six h o u r s , the corneal d r u g c o n c e n t r a t i o n r e m a i n e d at least four t i m e s greater (2.3 fxg/g) t h a n t h a t typically c o n s i d e r e d as the m i n i m a l i n h i b i t o r y c o n c e n t r a t i o n for C. albicans (0.5 (Ag/ml). 6 The m i n i m a l i n h i b i t o r y c o n c e n t r a t i o n for the yeast in our e x p e r i m e n t was 0.08 u , g / m l ; t h u s , o n e collagen shield p r e s o a k e d in 0 . 5 % a m p h o t e r i c i n B s h o u l d h a v e p r o v i d e d effective a n t i f u n g a l d r u g c o n c e n t r a t i o n s for at least six h o u r s . O u r s t u d y did n o t a d d r e s s the use of m o r e t h a n o n e collagen shield d u r i n g a 2 4 - h o u r p e r i o d or t r e a t m e n t b e y o n d five days.
Vol. 113, No. 3
Amphotericin B-Collagen Shield and Candida Keratitis
The collagen shield has been successfully used in the treatment of experimental bacterial keratitis. Rabbit corneas intrastromally inject ed with Pseudomonas aeruginosa were treated with tobramycin. 11 In that experiment, one col lagen shield rehydrated in 4% tobramycin was as effective in reducing the number of colonyforming units as were drops with an identical drug concentration applied every 30 minutes over a four-hour trial. In another similar infec tion model, rabbit corneas injected with P. aeruginosa were treated with one collagen shield with tobramycin drops or with tobramy cin drops alone. 12 Treatment every 30 minutes with 3% tobramycin drops was continued for 12 hours in two separate groups of rabbits. The first group of eyes was instilled with topical drops alone, whereas the second group of eyes" was administered a collagen shield in addition to drops. In this trial, 12 cultures of excised corneas from eyes treated with collagen shields demonstrated a significantly lower number of colony-forming units. Collagen shields are effective in delivering various drugs to the cornea and aqueous hu mor. Shields soaked in water-soluble drugs produce corneal and aqueous concentrations that are comparable to frequent topical treat ment. 13 The prolonged exposure time provided by a presoaked collagen shield may produce higher and more consistent tissue concentra tions and may result in increased drug penetra tion into the cornea and the anterior chamber. Prolonged high drug concentrations are impor tant especially in the treatment of keratomycosis.6 The results of our study confirm the effective ness of amphotericin B delivery by collagen shields. Our study demonstrated the use of collagen shields not only significantly reduced the isolate recovery rate at each time point compared to the use of saline drops, but also was superior to the hourly application of am photericin B drops after one and three days of treatment. Superiority of treatment with colla gen shields was greatest one day after treatment was started. Thereafter, colony-forming units declined at a similar rate in all three treatment groups. The effectiveness of an intact epitheli um as a barrier to stromal drug penetration was previously shown and may affect treatment re sults. 710 To enhance corneal penetration of am photericin B, we removed the corneal epitheli um. But this barrier effect is not absolute, because a 0.15% concentration of amphotericin B was still effective even in the presence of an
307
intact epithelium. 8 Histopathologic examina tion of our rabbits killed at Day 3 showed that all corneas were reepithelialized. Extrapolation of data from studies in rabbits to humans is limited by differences in blink rate, tear volume, and corneal thickness. 14 Nevertheless, comparisons between different treatment regimens may be possible when a reproducible model of disease is used. Our study demonstrates that collagen shields im pregnated with amphotericin B are effective in the treatment of experimental C. albicans-induced keratitis in rabbits. Collagen shields may have a role in enhancing antifungal drug deliv ery, providing high initial drug concentrations and subsequent successful eradication of fun gal pathogens in keratomycosis. Additionally, delivery of drugs by collagen shields is more convenient than frequent topical applications of drops. Because epithelial defects may be associated with keratomycosis and cause pa tient discomfort, treatment with collagen shields may also reduce the irritating effects of the eyelids on the cornea. The additional cost of collagen shields must be considered when they are used for drug delivery especially in chronic diseases.
ACKNOWLEDGMENT
C o l l a g e n s h i e l d s w e r e p r o v i d e d b y Bausch a n d L o m b , T a m p a , Florida.
References 1. Kotler-Brajtburg, J., Medoff, G., and Kobayashi, G. S.: Classification of polyene antibiotics according to chemical structure and biological effects. Antimicrob. Agents Chemother. 15:716, 1979. 2. O'Day, D. M., Robinson, R. D„ and Head, W. S.: Efficacy of antifungal agents in the cornea. A comparative study. Invest. Ophthalmol. Vis. Sci. 24:1098, 1983. 3. Wood, T. O., and Williford, W.: Treatment of keratomycosis with amphotericin B 0.15%. Am. J. Ophthalmol. 81:847, 1976. 4. Schwartz, S. D., Harrison, S. A., Engstrom, R. E., Jr., Bawdon, R. E., Lee, D. A., and Mondino, B. J.: Collagen shield delivery of amphotericin B. Am. J. Ophthalmol. 109:701, 1990. 5. O'Day, D. M., Ray, W. A., Robinson, R. D., Head, W. S., and Savage, A. M.: In vitro and in vivo susceptibility of Candida keratitis to topical polyenes. Invest. Ophthalmol. Vis. Sci. 28:874, 1987. 6. Behrens-Baumann, W., Uter, W., Vogel, M., and Ansorg, R.: Tierexperimentelles Modell einer Ker-
308
AMERICAN JOURNAL OF OPHTHALMOLOGY
atomykose. Klin. Monatsbl. Augenheilkd. 190:496, 1987. 7. O'Day, D. M., Ray, W. A., Head, W. S., and Robinson, R. D.: Influence of the corneal epithelium on the efficacy of topical antifungal agents. Invest. Ophthalmol. Vis. Sci. 25:855, 1984. 8. Havener, W. H.: Antibiotics. In Ocular Pharma cology. St. Louis, C. V. Mosby Company, 1983, p. 132. 9. Shirley, S. F., and Little, J. R.: Immunopotentiating effects of amphotericin B. J. Immunol. 123:2878, 1979. 10. O'Day, D. M., Head, W. S., Robinson, R. D., and Clanton, J. A.: Corneal penetration of topical amphotericin B and natamycin. Curr. Eye Res. 5:877, 1986.
March, 1992
11. Hobden, J. A., Reidy, J. J., O'Callaghan, R. J., Hill, J. M., Insler, M. S., and Rootman, D. S.: Treat ment of experimental Pseudomonas keratitis using collagen shields containing tobramycin. Arch. Oph thalmol. 106:1605, 1988. 12. Sawusch, M. R., O'Brien, T. P., Dick, J. D., and Gottsch, J. D.: Use of collagen corneal shields in the treatment of bacterial keratitis. Am. J. Ophthalmol. 106:279, 1988. 13. Friedberg, M. L., Pleyer, U., and Mondino, B. J.: Device drug delivery to the eye. Ophthalmology 98:725, 1991. 14. Chrai, S. S„ Patton, T. F., Metha, A., and Rob inson, J. R.: Lacrimal and instilled fluid dynamics in rabbit eyes. J. Pharm. Sci. 62:1112, 1973.
OPHTHALMIC MINIATURE
" O h , I know all about him," continued Charousek venomously, as if he h a d only been waiting for me to ask him. " A n d I knew his son, Dr. Wassory. Never heard of him? Wassory the famous ophthalmologist? The whole town was mad about him a year ago. A great specialist, they used to call him. They never knew that not so long ago his name h a d been Wassertrum. He loved to play the part of a m a n of science w h o had renounced the world. Any awkward questions about his origin h e ' d turn aside with modest chat about his father in the Ghetto, his own humble beginnings, and how he had kept the lamp of learning alight despite sorrows and hardships. Sorrows and hardships! He was right there, but he never revealed those sorrows and hardships, nor the means he had used. But I k n e w . " Gustav Meyrink, The Golem New York, Dover Publications, 1976, p. 17
Accepted for publication Dec. 30, 1991. From the Jules Stein Eye Institute and the Department of Ophthalmology, UCLA School of Medicine, Los An geles, California. This study was supported in part by National Eye Institute grants EY4606 (Dr. Mondino) and EY07701 (Dr. Lee); t h e Wasserman Fund, Los Angeles, California (Dr. Mondino); the Lucille E. Simon Research Fund, Los Angeles, California (Dr. Lee); a n d t h e Deutsche Forschungsgemeinschaft, PL 1 5 0 / 1 , Tubing en, Germany (Dr. Pleyer). This study was presented at the International Symposium of the German O p h t h a l mological Society, Miinster, Germany, Sept. 2 1 , 1991. Reprint requests to Bartly J. Mondino, M.D., Jules Stein Eye Institute, 100 Stein Plaza, Los Angeles, CA 90024-7003.
not at Day 5. The collagen shields were as effective in reducing the number of colonyforming units as were amphotericin B drops at Day 5. These data suggest that collagen shields soaked in amphotericin B could be a useful and convenient treatment device in keratomycosis such as that caused by C. albicans. JXERATOMYCOSIS remains a treatment chal lenge. Although many antifungal agents are available, treatment with these agents is com plicated by a narrow spectrum of activity, lack of effective penetration into the eye, and toxicity. Polyenes, which bind to the sterol moiety in the cell membrane of the organism and induce an osmotic pressure change of the cell constitu ents, possess the greatest antifungal activity in vitro. 1 Amphotericin B is the most commonly used polyene, and its efficacy against most yeasts is well documented. 2,3 The toxicity of this agent after topical use in keratomycosis is now circumvented by preparations with low concen trations. 3 Because penetration into ocular tissue by systemic or periocular injection is limited, frequent topical applications are necessary to reach adequate intracorneal concentrations. The recommended initial treatment of kerato mycosis is one drop of 0.15% amphotericin B every 30 minutes for the first three days. 3 Frequent topical applications are time-con suming and inconvenient. To alleviate these problems and provide reliable delivery of medi cation, drug delivery by collagen shields is being explored. Our previous study suggested that amphotericin B delivery to the cornea by collagen shields is comparable to frequent drop therapy over a six-hour period. 4 In this study, we determined the effectiveness of collagen shields containing amphotericin B in the treat ment of experimental C. albicans-induced kera titis.
©AMERICAN JOURNAL OF OPHTHALMOLOGY 113:303-307, M A R C H ,
1992
303
304
AMERICAN JOURNAL OF OPHTHALMOLOGY
Material and Methods We used a C. albicans strain that was suscepti ble to amphotericin B in vitro and was obtained from a patient at the UCLA Medical Center. The minimal inhibitory concentration (evaluated by the tube dilution method) was 0.08 ^.g/ml. The isolated fungi were kept at - 7 0 C; to avoid mutation and drug resistance, only prepara tions of the original strain without further pas sages were used. One day before inoculation, the fungi were plated onto Sabouraud dextrose agar and incubated at 37 C. Blastoconidia were harvested and suspended in normal (0.9%) sa line solution to yield a concentration of 5 X 109 colony-forming units/ml, and were stored at 4 C overnight. To verify the concentration of organisms in the inoculum, the number of colo ny-forming units/ml was determined on the next day no more than two hours before injec tion. A total of 48 healthy adult New Zealand (3- to 4-kg) rabbits were used. The rabbits were housed in wire-bottomed cages with controlled light/dark cycles, fed by standard laboratory diet, and given free access to tap water. All experiments were performed in accordance with current animal care guidelines and heeded protocols approved by the UCLA Animal Re search Committee. Before intracorneal injec tion, general anesthesia was induced in the rabbits by an intramuscular injection of 2 ml of chlorpromazine hydrochloride followed by in travenous injection of 3 ml of Equi-Thesin (UCLA Pharmacy, Los Angeles, California). Af ter topical anesthesia was achieved in the rab bits with proparacaine hydrochloride 0.5%, the eye was gently proptosed. An 8.5-mm trephine blade was used to create a circular corneal incision approximately 0.1 mm in depth. With in this area, the epithelium was completely removed using a sterile cotton-tip applicator to enhance drug penetration. A 30-gauge needle attached to a 1-ml tuberculin syringe was in serted into the corneal stroma at the edge of the trephine line and advanced toward the central cornea before injection of the suspension. Each cornea of each rabbit was inoculated with 20 (il of C. albicans in normal (0.9%) saline solution at a concentration of 5 X 109 organisms per milliliter. If penetration into the anterior cham ber occurred, the rabbit was removed from the study and replaced. Treatment was delayed for 24 hours after inoculation to allow transforma
March, 1992
tion of the organisms to the hyphal phase of the fungi and invasion of the corneal stroma as previously reported by O'Day and associates. 6 Gentamicin ophthalmic ointment was applied to each eye after inoculation and once daily in the evening for the duration of the experiment. Amphotericin B was reconstituted in sterile water to yield a concentration of 0.5 mg/ml. Bio-Cor 24-hour corneal collagen shields (Bausch and Lomb, Tampa, Florida) were im mersed in the 0.5% amphotericin B solution for 60 minutes before use. Amphotericin B drops were prepared at the 0.15% (1.5 mg/ml) con centration recommended for clinical use. 5 The same amphotericin B concentrations used for drops and collagen shields were used in our previous study, which showed that equivalent amounts of drug could be delivered to rabbit corneas using these concentrations. 4 Each morning a fresh vial of amphotericin B was used to prepare the treatment solutions. Rabbits were randomly assigned to treatment with a single collagen shield presoaked in 0.5% amphotericin B, 0.15% amphotericin B drops, or 0.9% saline drops in both eyes. The rabbits were treated for eight hours a day for one, three, or five days. In 18 rabbits assigned to the collagen shield group, one collagen shield was placed on the center of each cornea for a period of eight hours each day, after which all rem nants of the collagen shield were removed. Eighteen rabbits were assigned to amphotericin B drop treatment. A five-drop (50 |xl/drop) loading dose of amphotericin B was instilled in each eye of each rabbit with a one-minute interval between drops. After the loading dose, one drop was applied to each eye hourly over an eight-hour period each day. Twelve control rab bits were treated with a loading dose of 0.9% saline drops and hourly saline drops over eight hours each day. Rabbits were killed after one, three, or five days of treatment. At each time point, the rabbits were killed by means of intravenously administered Equi-Thesin 16 hours after treat ment regimens were completed. Corneas for isolate recovery were placed in 1 ml of sterile 0.9% saline solution, minced within two hours with a scalpel, and thoroughly homogenized with a mortar and pestle. Serial dilutions of these corneal solutions from 10" 1 to 10" 9 were made in 0.9% sterile saline solution, and 100|xl aliquots were plated onto trypticase soy agar with 5% sheep blood. Each dilution was plated in duplicate. Plates were incubated for 48 hours
Amphotericin B-Collagen Shield and Candida Keratitis
Vol. 113, No. 3
at 37 C and the number of colony-forming units per cornea was counted using the last four countable plates. At each time point, two corne as from one randomly assigned rabbit were embedded in paraffin, sectioned into 4-ji.mthick sections, and stained with hematoxylin and eosin and Giemsa. The resulting counts in the right and left corneas were highly correlated (r = .80, P < .001). The counts/cornea from right and left corneas were averaged to obtain a single count/ cornea for each rabbit. Therefore, all analyses and sample sizes were based on rabbits, not corneas. Mean counts among the treatment groups were compared via analysis of variance and post hoc f-tests. Analysis was performed on the log (base 10) scale because log count/ cornea is better modeled by a normal distribu tion. Statistical significance was accepted at P < .05. The Statistical Analysis System software (SAS Institute, Inc., Cary, North Carolina) was used to perform all computations.
Results Immediately after inoculation of C. albicans, a circumscribed opacity was visible in the cornea. Control rabbits treated with saline drops devel oped corneal infiltrates that enlarged over the next three days and gradually decreased over the last two days of observation. All control eyes developed a hypopyon after three to five days of saline-drop instillations. The corneas of rabbits treated with amphotericin B-soaked shields or amphotericin B drops did not differ from the corneas of control eyes after only one day of treatment. After three days of treatment, a hypopyon developed in one of 12 eyes (6%) treated with amphotericin B-soaked shields and four of 12 eyes (33%) treated with ampho tericin B drops. After five days of treatment with amphotericin B-soaked collagen shields, four of 12 eyes (33%) developed hypopyon as compared to six of 12 eyes (50%) that were treated with amphotericin B drops. Central epithelial defects were present on Day 1 in all corneas administered amphotericin B-soaked shields, amphotericin B drops, or saline drops. At Day 3, the epithelial defects were healed in all three groups. Histologic studies of control eyes one day after inoculation demonstrated blastospores with pseudohyphae and true hyphae invading the corneal stroma
305
Fig. 1 (Pleyer and associates). Candida albicansinduced keratomycosis in control rabbit on Day 1 after inoculation. Blastospores are present at the injection site. Active transformation to hyphae is occurring with invasion into the corneal stroma (he matoxylin and eosin, original magnification x 330). (Fig. 1). Active stromal infiltration with mycelial growth accompanied by polymorphonuclear leukocytes was observed in control eyes five days after saline-drop application. Blastospores predominated in eyes treated for one day with amphotericin B-soaked shields or amphotericin B drops. Compared to control corneas, infiltration of the stroma and transfor mation to true hyphae were reduced. After three to five days of treatment with amphoteric in B-soaked shields or drops, hyphal growth was present but appeared less severe compared to that in control rabbits (Fig. 2). Quantification of C. albicans growth in con trol rabbits showed a high isolate recovery rate. Even after five days, an average of 417,000 colony-forming units/cornea were isolated from control eyes. At all time points, treatment with either am photericin B-presoaked collagen shields or am photericin B drops significantly reduced the number of colony-forming units/cornea (P < .05) when compared to control rabbits (Fig. 3). At Days 1 and 3, rabbits treated with amphoter icin B-soaked collagen shields had significantly lower fungal counts (P = .02 at Day 1, P = .04 at Day 3) compared to rabbits treated with hourly instilled amphotericin B drops. At Day 5, the number of colony-forming units/cornea was still lower in rabbits treated with collagen shields, but the difference was not statistically significant compared to rabbits treated with amphotericin B drops (P = .2).
306
March, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
10000
—•— —■—
c E
o u.
Amphotericin BCollagen shields Amphotericin 8Drops Saline Drops
100:
c
O O
Days of
Fig. 2 (Pleyer and associates). Candida albicansinduced keratomycosis in a rabbit treated for eight hours with a collagen shield presoaked in 0.5% amphotericin B. Treatment was begun 24 hours after inoculation. Invasion of the corneal stroma is re duced, and transformation of blastospores to hyphae diminished (hematoxylin and eosin, original magnifi cation x 300).
Discussion The m o d e l of k e r a t o m y c o s i s u s e d in this study is c o m p a r a b l e to that p r e v i o u s l y r e p o r t e d b y O ' D a y a n d associates. 6 In t h a t m o d e l of d e e p corneal infection, b l a s t o s p o r e t r a n s f o r m a t i o n to a h y p h a l p h a s e o c c u r r e d w h e n t r e a t m e n t was delayed for 24 h o u r s . T h u s , the m o d e l of infec tion parallels h u m a n k e r a t o m y c o s i s , w h i c h is c h a r a c t e r i z e d by corneal i n v a s i o n b y p s e u d o h y p h a e a n d t r u e s e p t a t e h y p h a e . U n l i k e fungal infections in h u m a n s , k e r a t o m y c o s i s in r a b b i t s resolves w i t h o u t t r e a t m e n t . In s t u d i e s u s i n g a similar infection m o d e l , k e r a t o m y c o s i s w a s still active even n i n e a n d 14 days after fungal infec tion. 6 ' 7 A m p h o t e r i c i n B a n d n a t a m y c i n are the m o s t effective a g e n t s for the t r e a t m e n t of k e r a t o m y cosis. A l t h o u g h n a t a m y c i n h a s g a i n e d favor as an antifungal d r u g b e c a u s e of its w i d e a n t i f u n gal effect a n d relatively low toxicity, a m p h o t e r icin B r e m a i n s a p o t e n t a g e n t in the t r e a t m e n t of keratomycosis. Amphotericin B has a wider s p e c t r u m of activity against yeast-like fungi t h a n any other c u r r e n t l y available a n t i f u n g a l d r u g . A m p h o t e r i c i n B is preferred for t r e a t m e n t of infection caused by Coccidioides immitis, Histoplasma capsulatum, Blastomyces dermatitidis, C. albicans species, a n d o t h e r less c o m m o n fungi. 8 In a d d i t i o n to its a n t i f u n g a l effect, a m photericin B has immunoadjuvant properties a n d may have an i m m u n o p o t e n t i a t i n g effect. 9
Treatment
Fig. 3 (Pleyer and associates). Effect of treatment on isolate recovery of C. albicans in experimental kerato mycosis. Twenty-four hours after intracorneal inocu lation, treatment with 0.15% amphotericin B drops applied hourly, saline drops applied hourly, or a collagen shield presoaked in 0.5% amphotericin B was begun. Treatment was given for eight hours per day for one, three, or five days. Results are expressed as mean ± standard error of the mean colony-form ing units/cornea. Sample size equals five rabbits for each time point of the experimental groups and three rabbits for each time point of the saline-treated control group. O c u l a r p e n e t r a t i o n of n a t a m y c i n as well as a m p h o t e r i c i n B is limited. 1 0 In d e e p i t h e l i a l i z e d corneas, higher drug concentrations were found with natamycin than with amphotericin B. This finding, r a t h e r t h a n a s u p e r i o r antifun gal effect, m a y favor t h e use of n a t a m y c i n in k e r a t o m y c o s i s . I m p r o v e m e n t of the efficacy of a m p h o t e r i c i n B may, t h e r e f o r e , lie in its formu lation a n d m o d e of a p p l i c a t i o n . A t t e m p t s to i n c o r p o r a t e n a t a m y c i n into collagen s h i e l d s have b e e n unsuccessful. 4 Use of collagen s h i e l d s i m m e r s e d in a m p h o t e r i c i n B p r o d u c e d corneal c o n c e n t r a t i o n s t h a t w e r e h i g h e r or equivalent to h o u r l y instilled d r o p t r e a t m e n t in a p r e v i o u s study. 4 In t h e s i x - h o u r trial, corneal a m p h o t e r i c i n B c o n c e n t r a t i o n s o b t a i n e d b y col l a g e n shields w e r e h i g h e s t o n e h o u r after a p p l i cation a n d d e c l i n e d a f t e r w a r d s . Even after six h o u r s , the corneal d r u g c o n c e n t r a t i o n r e m a i n e d at least four t i m e s greater (2.3 fxg/g) t h a n t h a t typically c o n s i d e r e d as the m i n i m a l i n h i b i t o r y c o n c e n t r a t i o n for C. albicans (0.5 (Ag/ml). 6 The m i n i m a l i n h i b i t o r y c o n c e n t r a t i o n for the yeast in our e x p e r i m e n t was 0.08 u , g / m l ; t h u s , o n e collagen shield p r e s o a k e d in 0 . 5 % a m p h o t e r i c i n B s h o u l d h a v e p r o v i d e d effective a n t i f u n g a l d r u g c o n c e n t r a t i o n s for at least six h o u r s . O u r s t u d y did n o t a d d r e s s the use of m o r e t h a n o n e collagen shield d u r i n g a 2 4 - h o u r p e r i o d or t r e a t m e n t b e y o n d five days.
Vol. 113, No. 3
Amphotericin B-Collagen Shield and Candida Keratitis
The collagen shield has been successfully used in the treatment of experimental bacterial keratitis. Rabbit corneas intrastromally inject ed with Pseudomonas aeruginosa were treated with tobramycin. 11 In that experiment, one col lagen shield rehydrated in 4% tobramycin was as effective in reducing the number of colonyforming units as were drops with an identical drug concentration applied every 30 minutes over a four-hour trial. In another similar infec tion model, rabbit corneas injected with P. aeruginosa were treated with one collagen shield with tobramycin drops or with tobramy cin drops alone. 12 Treatment every 30 minutes with 3% tobramycin drops was continued for 12 hours in two separate groups of rabbits. The first group of eyes was instilled with topical drops alone, whereas the second group of eyes" was administered a collagen shield in addition to drops. In this trial, 12 cultures of excised corneas from eyes treated with collagen shields demonstrated a significantly lower number of colony-forming units. Collagen shields are effective in delivering various drugs to the cornea and aqueous hu mor. Shields soaked in water-soluble drugs produce corneal and aqueous concentrations that are comparable to frequent topical treat ment. 13 The prolonged exposure time provided by a presoaked collagen shield may produce higher and more consistent tissue concentra tions and may result in increased drug penetra tion into the cornea and the anterior chamber. Prolonged high drug concentrations are impor tant especially in the treatment of keratomycosis.6 The results of our study confirm the effective ness of amphotericin B delivery by collagen shields. Our study demonstrated the use of collagen shields not only significantly reduced the isolate recovery rate at each time point compared to the use of saline drops, but also was superior to the hourly application of am photericin B drops after one and three days of treatment. Superiority of treatment with colla gen shields was greatest one day after treatment was started. Thereafter, colony-forming units declined at a similar rate in all three treatment groups. The effectiveness of an intact epitheli um as a barrier to stromal drug penetration was previously shown and may affect treatment re sults. 710 To enhance corneal penetration of am photericin B, we removed the corneal epitheli um. But this barrier effect is not absolute, because a 0.15% concentration of amphotericin B was still effective even in the presence of an
307
intact epithelium. 8 Histopathologic examina tion of our rabbits killed at Day 3 showed that all corneas were reepithelialized. Extrapolation of data from studies in rabbits to humans is limited by differences in blink rate, tear volume, and corneal thickness. 14 Nevertheless, comparisons between different treatment regimens may be possible when a reproducible model of disease is used. Our study demonstrates that collagen shields im pregnated with amphotericin B are effective in the treatment of experimental C. albicans-induced keratitis in rabbits. Collagen shields may have a role in enhancing antifungal drug deliv ery, providing high initial drug concentrations and subsequent successful eradication of fun gal pathogens in keratomycosis. Additionally, delivery of drugs by collagen shields is more convenient than frequent topical applications of drops. Because epithelial defects may be associated with keratomycosis and cause pa tient discomfort, treatment with collagen shields may also reduce the irritating effects of the eyelids on the cornea. The additional cost of collagen shields must be considered when they are used for drug delivery especially in chronic diseases.
ACKNOWLEDGMENT
C o l l a g e n s h i e l d s w e r e p r o v i d e d b y Bausch a n d L o m b , T a m p a , Florida.
References 1. Kotler-Brajtburg, J., Medoff, G., and Kobayashi, G. S.: Classification of polyene antibiotics according to chemical structure and biological effects. Antimicrob. Agents Chemother. 15:716, 1979. 2. O'Day, D. M., Robinson, R. D„ and Head, W. S.: Efficacy of antifungal agents in the cornea. A comparative study. Invest. Ophthalmol. Vis. Sci. 24:1098, 1983. 3. Wood, T. O., and Williford, W.: Treatment of keratomycosis with amphotericin B 0.15%. Am. J. Ophthalmol. 81:847, 1976. 4. Schwartz, S. D., Harrison, S. A., Engstrom, R. E., Jr., Bawdon, R. E., Lee, D. A., and Mondino, B. J.: Collagen shield delivery of amphotericin B. Am. J. Ophthalmol. 109:701, 1990. 5. O'Day, D. M., Ray, W. A., Robinson, R. D., Head, W. S., and Savage, A. M.: In vitro and in vivo susceptibility of Candida keratitis to topical polyenes. Invest. Ophthalmol. Vis. Sci. 28:874, 1987. 6. Behrens-Baumann, W., Uter, W., Vogel, M., and Ansorg, R.: Tierexperimentelles Modell einer Ker-
308
AMERICAN JOURNAL OF OPHTHALMOLOGY
atomykose. Klin. Monatsbl. Augenheilkd. 190:496, 1987. 7. O'Day, D. M., Ray, W. A., Head, W. S., and Robinson, R. D.: Influence of the corneal epithelium on the efficacy of topical antifungal agents. Invest. Ophthalmol. Vis. Sci. 25:855, 1984. 8. Havener, W. H.: Antibiotics. In Ocular Pharma cology. St. Louis, C. V. Mosby Company, 1983, p. 132. 9. Shirley, S. F., and Little, J. R.: Immunopotentiating effects of amphotericin B. J. Immunol. 123:2878, 1979. 10. O'Day, D. M., Head, W. S., Robinson, R. D., and Clanton, J. A.: Corneal penetration of topical amphotericin B and natamycin. Curr. Eye Res. 5:877, 1986.
March, 1992
11. Hobden, J. A., Reidy, J. J., O'Callaghan, R. J., Hill, J. M., Insler, M. S., and Rootman, D. S.: Treat ment of experimental Pseudomonas keratitis using collagen shields containing tobramycin. Arch. Oph thalmol. 106:1605, 1988. 12. Sawusch, M. R., O'Brien, T. P., Dick, J. D., and Gottsch, J. D.: Use of collagen corneal shields in the treatment of bacterial keratitis. Am. J. Ophthalmol. 106:279, 1988. 13. Friedberg, M. L., Pleyer, U., and Mondino, B. J.: Device drug delivery to the eye. Ophthalmology 98:725, 1991. 14. Chrai, S. S„ Patton, T. F., Metha, A., and Rob inson, J. R.: Lacrimal and instilled fluid dynamics in rabbit eyes. J. Pharm. Sci. 62:1112, 1973.
OPHTHALMIC MINIATURE
" O h , I know all about him," continued Charousek venomously, as if he h a d only been waiting for me to ask him. " A n d I knew his son, Dr. Wassory. Never heard of him? Wassory the famous ophthalmologist? The whole town was mad about him a year ago. A great specialist, they used to call him. They never knew that not so long ago his name h a d been Wassertrum. He loved to play the part of a m a n of science w h o had renounced the world. Any awkward questions about his origin h e ' d turn aside with modest chat about his father in the Ghetto, his own humble beginnings, and how he had kept the lamp of learning alight despite sorrows and hardships. Sorrows and hardships! He was right there, but he never revealed those sorrows and hardships, nor the means he had used. But I k n e w . " Gustav Meyrink, The Golem New York, Dover Publications, 1976, p. 17
