BritishJournal ofOphthalmology, 1990,74: 40-42

Topical fluconazole for experimental candidakeratitis in rabbits

Wolfgang Behrens-Baumann, Bernd Klinge, Reinhard Ruchel

AbstractUsing a reproducible model of Candidaalbicans keratitis in rabbits we studied theeffect of topical fluconazole, a new triazole.Candida albicans DSM 70010 (2.5X10 cells)was injected into the corneal stroma of botheyes of21 rabbits. AU eyes developed a cornealulcer. Forty-eight hours after inoculation theanimals were divided into three groups: (1) 14eyes, received fluconazole (2 mg/ml) and theepithelium subsequently removed; (2) 14 eyes,received only fluconazole drops; (3) 14 eyes,received 0/9% NaCI: halfofthis group was alsodebrided. We applied one drop of either sub-stance 10 times a day for 24 days. A further sixrabbits were used to judge if the drug pene-trated into the cornea and aqueous humour.There was a highly significant differencebetween the fluconazole groups (1, 2) and thecontrol group (3) as to hypopyon and complica-tions (descemetocele, corneal perforation) aswell as recultivation of C. albicans fromcorneal tissue. The difference between thefluconazole groups with and without debride-ment was not significant. The drug penetratedinto the cornea and aqueous humour of bothuninflamed and inflamed eyes.

University of Gottingen,Federal Republic ofGermany, Department ofOphthalmologyW Behrens-BaumannB Klinge

Department of MedicalMicrobiologyR RuchelCorrespondence to: ProfessorW Behrens-Baumann,Department ofOphthalmology, University ofGottingen, Robert-Koch-Strasse 40, D-3400 Gottingen,FRG.

Accepted for publication4 September 1989

Fungal infections of the cornea have markedlyincreased during the last decades.' 2 Apart frompolyenes such as amphotericin-B, nystatin, andpimaricin (natamycin), azoles such as theimidazoles and ketoconazole have been appliedto fungal keratitis. The azoles proved effectiveand were non-toxic to the eye.34 However, thetreatment of clinical keratomycoses is stillinefficient.5 6

Fluconazole (2-(2,4,-difluorophenyl)-1,3-bis(1 H- 1,2,4-triazol-1-yl) propan-2-ol; UK 49*858)is a new triazole compound (molecular weight306&3) which has certain properties suggestingits use for the treatment of fungal keratitis. It hasbeen reported to be more efficient than keto-conazole in systemic candidosis of immuno-suppressed as well as normal mice.78 Moreover,fluconazole is the only antifungal azole derivativewhich penetrates into the cerebrospinal fluid.9 '0

Parenterally administered fluconazole underexperimental conditions was shown to penetratefreely into ocular tissues."

This is the first report on the use of topicalfluconazole in experimental Candida albicanskeratitis in rabbits. We also studied penetrationof the drug into the cornea and aqueous humour.

Material and methodsBioavailability. One drop of aqueous fluconazolesolution (2 mg/ml) was applied to both eyes of six

rabbits every hour for a period ofeight hours. Onthree animals the corneal epithelium had beengently removed with a hockey knife under amicroscope. Fifty minutes after the last treat-ment the eye was irrigated with saline.Aqueous humour samples (0-2 to 0 3 ml) were

obtained by puncture of the anterior chamberwith a 26 gauge needle. Corneal samples wereexcised at the limbus with scissors and ground ina tissue grinder (Ultra-Turrax, Janke andKunkel, Stauffen, Germany). These sampleswere centrifuged at 12 000 rpm.

Serial dilution tests of both the aqueoushumour samples and the supernatant of thecorneal samples were performed on microplates.A 100 1tl candida suspension (103 blastoconidia/ml) from a 24 h culture in glucose peptone brothcontaining 50 [ig/ml gentamicin sulphate wasadded to 50 [tl of the samples. The plates wereincubated for 24 hours at 370C. The minimalinhibition concentration (MIC) of fluconazolewas determined by serial dilution under compar-able conditions.

Inoculum. We used Candida albicans DSM(German Collection of Micro-organisms) No70010, which shows filamentous growth inaqueous humour. 12 Yeast cells from a 24 hculture in Sabouraud agar were suspended inpeptone broth, counted with a haemacytometer,and adjusted to a concentration of 2 5x10'cells/l.

Animals. Twenty-one male, pigmented,inbred rabbits, weighing approximately 2 5 kgeach were used. No pretreatment was per-formed.

Inoculum procedure. This method has beendescribed previously. 13 The animals wereanaesthetised with intramuscular xylazin andketamine, 5 mg/kg body weight. To avoid pre-servatives no local anaesthetic was applied.Under an operating microscope a 27 gaugeneedle was inserted into the central cornealstroma to a depth ofabout one-halfof the cornealthickness. A 10 [d inoculum, containing2-5xI10 cells, was injected in both eyes. Noreflux of the inoculum was observed. If penetra-tion into the anterior chamber occurred, theanimal was removed from the study.

Therapy. Forty-eight hours after inoculationthe animals were randomised into three groups(14 eyes of 7 animals each). In groups 1 and 2aqueous fluconazole solution (2 mg/ml in 0 9%NaCl) and in group 3 0-9% NaCl was adminis-tered. In addition, the epithelium of the eyes ingroup 1 was gently removed. To exclude anymechanical effect of debridement the same wasdone in half of those in the control group 3. Thisprocedure was done every third day with ahockey knife under a microscope. All eyesreceived a daily treatment of 0-5% aqueous

40

on May 17, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.74.1.40 on 1 January 1990. D

ownloaded from

Topicalfluconazolefor experimental candida keratitis in rabbits

gentamicin sulphate (without preservatives) toavoid a bacterial infection.

Evaluation of ocular infections. All eyes wereexamined every day and a standardised photo-graphic record was made. The occurrence of adescemetocele, corneal perforation, andhypopyon was noted.Fungal recultivation. After 24 days the animals

were killed by intravenous thiopentone sulphate.Aqueous humour samples were obtained from alleyes to study the penetration of fluconazole intoinflamed eyes. Serial dilution tests for antifungalactivity were performed as described above. Thecornea was excised at the limbus and dividedthrough the centre of the ulcer. The cut surfacesof both sections were pressed on to Sabouraudagar containing 50 [tg/g gentamicin sulphate andthe plates incubated for one week at 21C.Positive cultures were plated on blood agar andincubated for 48 hours at 37°C for furtherdifferentiation. These cultures were subjected toan auxanogram'4 (API-20-C) to establish thepresence of C. albicans.

ResultsBioavailability. Penetration of fluconazole intothe cornea and aqueous humour of uninflamedeyes is shown in Table I. There is a moderatedifference between eyes with and withoutcorneal epithelium. In inflamed eyes (24 daysafter infection) the results of aqueous humoursamples are the same as in non-inflamed eyes.

Course of infection and complications. All 42eyes inoculated with Candida albicans DSM70010 developed corneal infiltrates on day 2 withan average size of 3x3 mm. These infiltratesprogressed to an ulcer between days 5 and 8. Theincidence of hypopyon and complications(descemetocele or corneal perforation) is shownin Table II for each group. There was a markeddifference between the therapy groups (1 and2) and the control group (3a and 3b) in theoccurrence of descemetoceles or perforations(p<001; X2 test). Figures 1-4 show typicalcorneal infections of each group on day 21.The drug was well tolerated by all animals. No

adverse effect could be observed.Recultivation. Table III lists the results of

Candida albicans recultivation. In group 1 noviable yeasts were recovered. In group 2, two of14 samples showed candida growth. These

TABLE I Penetration offluconazole into the rabbit corneaand aqueous humour ofuninflamed eyes. Growth ofC.albicans was inhibited by the indicated dilutions ofcornealhomogenate or aqueous humour

With debridement Without debridement

Corneal homogenate 1:4 1:2Aqueous humour 1:8 1:2

TABLE II Influence oftopicalfluconazole on hypopyon, descemetocele, or perforation inexperimental candida keratitis

DescemetoceleEyes Hypopyon or perforation

Group 1 (fluconazole+debridement) 14 3 0Group 2 (fluconazole without debridement) 14 6 1Group 3a (0 9% NaCl+debridement) 6 6 5Group 3b (0 9% NaCI without debridement) 8 7 8

rtgure 1: Ganadaa keratttos oj group I (puconazole+debridement) on day 21.

results are highly significant compared with thatof group 3a+b (p<001); x2 test).

DiscussionMost antifungal drugs are effective againstsuperficial mycoses but are less successful incounteracting deep mycotic infiltrations. In par-ticular, the treatment of keratomycosis is oftenfrustrating owing to limited tissue penetration,narrow antimicrobial spectrum, and toxicity ofthe antifungal agents currently available.The new triazole antifungal fluconazole, by

virtue of its pharmacological properties, offers afresh opportunity for the topical treatment ofkeratomycosis. The drug acts against all patho-genic Candida species except C. krusei'5 andhence encompasses the majority of fungi whichare notorious as causative agents of keratomyco-sis. 1-18 Fluconazole is exceptionally hydrophilic,which is reflected by low protein binding andhigh penetration into cerebrospinal fluid, as wellas by intestinal absorption and renal excre-tion.'5 1920 In rabbits fluconazole was shown toreach ocular tissues after intravenous adminis-tration." In addition fluconazole was shown to benon-mutagenic and less toxic than the otherazoles.'9

In the present study we used a rabbit model toinvestigate the potential of fluconazole in thetherapy of deep keratitis due to C. albicans. Themodel is reproducible without the need ofimmunosuppressive pretreatment. ' Therapy is

TABLE iII Recultivation ofCandida albicans DSM 70010after 24 days oftreatment with topicalfluconazole

Eyes Positive

Group 1 (fluconazole+debridement) 14 0Group 2 (fluconazole without debridement) 14 2Group 3a (0-9% NaCl+debridement) 6 5Group 3b (0 9% NaCi without debridement) 8 7

Figure 2: Candida keratitis ofgroup 2 (fluconazole withoutdebridement) on day 21.

41

on May 17, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.74.1.40 on 1 January 1990. D

ownloaded from

Behrens-Baumann, Klinge, Riichel

Figure 3: (Candida keratitis of group 3a (0Udebridement) on day 21.

started only on day 2, when stromal keratitis ismanifest; hence the model more closely parallelshuman corneal candidiasis than a regimen whichbegins treatment one hour after infection.2'Our results demonstrate the efficacy of

fluconazole in the topical therapy of experi-mental candida keratitis. As to complicationssuch as descemetocele and corneal perforation,there was a highly significant difference betweenthe treated eyes and the controls. The beneficialeffect was confirmed because almost no reculti-vation of C. albicans was possible after 24 days oftreatment.However, no significant difference between

the debridement and no-debridement groupswas found. This finding contrasts with resultsfrom similar experiments using amphotericin B,which revealed a highly significant difference infavour of debrided eyes.22 The effect of debride-ment on the penetration of amphotericin Breflects its higher molecular weight (924- 11),since the limiting size of molecules for diffusinginto the normal cornea is about 500.00.23We conclude that fluconazole proved highly

effective in the therapy of experimental candidakeratitis. If debridement of the corneal epithe-lium is undesirable, this drug seems to besuperior even to amphotericin B.22

1 Jones BR. Principles in the management of oculomycosis.XXXI Edward Jackson memorial lecture. AmJI Ophthalmol1975; 79: 719-51.

2 Forster RK. Fungal diseases. In: Smolin G, Thoft R, eds. Thecornea. Boston: Little, Brown, 1983: 168-77.

3 Ishibashi Y, Kaufman HE. The effects of subconjunctivalmiconazole in the treatment of experimental candidakeratitis in rabbits. Arch Ophthalmol 1985; 103: 1570-3.

4 Ishibashi Y, Kaufman HE. Topical ketoconazole for experi-mental candida keratitis in rabbits. Amj Ophthalmol 1986;102:522-6.

5 O'Day DM, Robinson R, Head WS. Efficacy of antifungal

Figure 4: Candida keratitis ofgroup 3b (0debridement) on day 21.

agents in the cornea. 1. A comparative study. InvestOphthalmol Vis Sci 1983; 24: 1098-102.

6 Komadina TG, Wilkes TDI, Shock JP, Ulmer WC, Jackson J,Bradsher RW. Treatment of Aspergillusfunigatus keratitis inrabbits with oral and topical ketoconazole. AmJ7 Ophthalmol1985; 99: 476-9.

7 Richardson K, Brammer KW, Marriott MS, Troke PF.Activity of UK-49,858, a bis-triazole derivate, againstexperimental infections with Candida albicans andTrichophyton mentagrophytes. Antimicrob Agents Chemother1985;27:832-5.

8 Troke PF, Andrews RJ, Brammer KW, Marriott MS,Richardson K. Efficacy of UK-49,858 (fluconazole) againstCandida albicans experimental infections in mice. AntimicrobAgents Chemother 1985; 28: 815-8.

9 Palou de Fernandez E, Patino MM, Graybill JR, Tarbit MH.Treatment of cryptococcal meningitis in mice withfluconazole. J Antimicrob Chemother 1986; 18: 261-70.

10 Arndt CAS, Walsh TJ, Lester McCully C, Balis FM, PizzoPA, Poplack DG. Fluconazole penetration into cerebro-spinal fluid: implications for treating fungal infections of thecentral nervous system. J Infect Dis 1988; 157: 178-80.

11 Savani DV, Perfect JR, Cobo LM, Durack DT. Penetration ofnew azole compounds into the eye and efficacy in experi-mental candida endophthalmitis. Antimicrob AgentsChemother 1987; 31: 6-10.

12 Uter W, Behrens-Baumann W, Ansorg R. Stimulation offilamentous growth of Candida albicans by aqueous humor.Mycoses 1988; 31: 313-5.

13 Behrens-Baumann W, Uter W, Vogel M, Ansorg R. Animalmodel of reproducible candida keratomycosis. KlinMonatsbl Augenheilkd 1987; 190: 496-500.

14 De Louvois J, Mulhall A, Hurley R. Biochemical identifica-tion of clinically important yeasts. J Cln Pathol 1979; 32:715-8.

15 Stevens DA. The new generation of antifungal drugs. Eur rClin Microbiol 1988; 7: 732-5.

16 Thygeson P, Okumoto M. Keratomycosis: a preventabledisease. Ophthalmology 1974; 78: 433-9.

17 Romano A, Segal E, Stein R, Eylan E. Yeasts in banal externalocular inflammations. Ophthalmologica 1975; 170: 13-21.

18 O'Day DM. Antifungal agents. In: Leibowitz HA. Cornealdisorders. Clinical diagnosis and management. Philadelphia:Saunders, 1984: 345-51.

19 Fromtling RA. Overview of medically important antifungalazole derivatives. Clin Microbiol Rev 1988; 1: 187-217.

20 Walsh TJ, Pizzo A. Amphotericin B is the first-line drug ofchoice for treatment of most systemic fungal infections.Eurj Clin Microbiol 1988; 7: 460-75.

21 O'Day DM, Ray WA, Head WS, Robinson RD. Influence ofthe corneal epithelium on the efficacy of topical antifungalagents. Invest Ophthalmol Vis Sci 1984; 25: 855-9.

22 Behrens-Baumann W, Uter W, Ansorg R. Experimentalstudies relating to local amphotericin B treatment of candidakeratitis. Klin Monatsbl Augenheilkd 1987; 191: 125-8.

23 Maurice DM. The permeability of the cornea. In: Duke-ElderS, Perkins ES. The transparency of the cornea. Oxford:Blackwell, 1960: 67-71.

42

on May 17, 2020 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.74.1.40 on 1 January 1990. D

ownloaded from