Clin Exp Ophthalmol 2011_ p434

Original Article

Voriconazole versus natamycin as primarytreatment in fungal corneal ulcersRitu Arora MD,1 Deepa Gupta MS,1 Jawaharlal Goyal MD1 and Ravinder Kaur MD2

1Guru Nanak Eye Centre, and 2Department of Microbiology, Maulana Azad Medical College, New Delhi, India

ABSTRACT

Background: To evaluate the efficacy of topical 1%voriconazole versus 5% natamycin in treatment offungal corneal ulcers.

Design: A prospective, randomized pilot study in atertiary care hospital.

Participants: Thirty patients of microbiologicallyproven fungal keratitis divided randomly in twogroups of 15 patients each.

Methods: Two groups were treated with either 5%natamycin (group A) or 1% voriconazole (group B)topically as a primary treatment for fungal keratitis.The mean size, depth of infiltrate and LogMAR visualacuity at presentation were comparable in bothgroups (P > 0.05). Patients were followed up forminimum of 10 weeks or till complete resolution ofulcer, whichever was later. Cultures to identify thecausative organisms were performed.

Main Outcome Measure: Time of resolution of theulcer.

Results: Twenty-nine of the total 30 patients showedcomplete resolution. Average time of resolution andgain in LogMAR visual acuity was 24.3 days and1.12 in group A and 27.2 days and 0.77 in groupB. These were comparable in the two groups (P >0.05%). Aspergillus spp. (40%) and Curvularia spp.(30.0%) were found to be most common isolates.

Conclusion: Topical 1% voriconazole was found to besafe and effective drug in primary management offungal keratitis, its efficacy matching conventional

natamycin. There was no added advantage of usingtopical 1% voriconazole over topical natamycin asprimary treatment in fungal keratitis.

Key words: fungal keratitis, natamycin, voriconazole.

INTRODUCTION

Infectious keratitis is one of the predominant causesof corneal blindness being second only to cataract asper WHO bulletin.1 Fungal keratitis is more commonin tropical region as well as in developing countries,constituting over 50% of keratitis.2 Prevalence offungal keratitis, reported from tropical climates is17% in Nepal,3 36% in Bangladesh,4 38% in Ghana5

and 35% in south Florida in USA.6 It constitutesabout 44% of corneal ulcers in south India.7 In con-trast, fungal keratitis generally accounts for only1–5% of the keratitis in developed countries andtemperate regions such as Britan, northern USA andAustralia.1,8

Management of fungal keratitis requires timelydiagnosis of the infection and administration ofappropriate antifungal therapy. Antifungal agentsare still the major therapeutic options in fungalkeratitis, whereby success depends on the agent’sability to penetrate into the aqueous and achievetherapeutic levels.9 The choice of antifungal thera-pies available for fungal keratitis remains limited.9

Natamycin, the only FDA-approved agent for thetreatment of fungal keratitis, is a fungicidal tetraenepolyene antibiotic, derived from Streptomyces natalensisthat possesses in vitro activity against a variety ofyeast and filamentous fungi,10 including Candida,Aspergillus, Cephalosporium, Fusarium and Penicilliumspecies. Jones et al. reported 18 consecutive casesof Fusarium solani keratitis treated successfully

� Correspondence: Dr Deepa Gupta, Guru Nanak Eye Centre, Maulana Azad Medical College, 68 Kirpal Appartments, 44,I.P. Extension, Delhi-110092,

India. Email: [email protected]

Received 25 August 2010; accepted 25 October 2010.

Clinical and Experimental Ophthalmology 2011; 39: 434–440 doi: 10.1111/j.1442-9071.2010.02473.x

© 2011 The AuthorsClinical and Experimental Ophthalmology © 2011 Royal Australian and New Zealand College of Ophthalmologists

with natamycin.11 Natamycin is believed to be effec-tive only in superficial infections because of its poorpenetration. There have been reported 45.2% treat-ment successes with natamycin as a primary treat-ment for fungal keratitis.12

In the recent past, the use of newer triazoles (vori-conazole [VRC], posaconazole and ravuconazole) hasbeen suggested as better alternatives in the treatmentof fungal keratitis not responding to conventionalantifungals. They are fungistatic and act by inhibit-ing biosynthesis of ergosterol, an essential compo-nent in fungal cell wall.13 At present only VRC isavailable commercially (Vfend, Pfizer, New York,USA). VRC has been shown to have a broad spec-trum of activity against non-ocular isolates ofAspergillus spp., Paecilomyces lilacinus, Cryptococcusneoformans, Scedosporium spp., Curvularia spp. andothers.14–17 Its MIC was found to be the lowestamong all antifungal agents and in vitro suscep-tibility against a variety of fungi found to be 100%.17

Numerous case of successful treatment of fungalkeratitis resistant to conventional treatment by com-bination of topical as well as systemically adminis-tered VRC have been reported.18–23

The purpose of this pilot study was to investigatethe use of 1% VRC as a primary modality in thetreatment in fungal keratitis and compare its efficacywith 5% natamycin eye drops in a prospective, ran-domized controlled manner.

METHODS

This study was randomized, double-masked, inter-ventional, pilot study of patients with fungalkeratitis. Thirty patients of microbiologically provenfungal keratitis presenting at the outpatient depart-ment of Guru Nanak Eye Centre from September2007 to March 2009 were enrolled in the study. Allthese patients provided a written, informed consentfor their study participation. Only patients withcorneal scrapings positive for fungal hyphae on 10%KOH wet mount/Gram’s staining were includedin the study. Patients with any prior usage of antifun-gal drugs, history of herpetic keratitis or previouscorneal scars, impending perforation and no lightperception were excluded. All the patients in thestudy group were admitted in the hospital initially toensure that the treatment protocol was strictlyfollowed. Only with the signs of improvement werethey managed on outpatient basis. While in the hos-pital the treatment was delivered by nurse and byrelatives at home. They were randomly divided intotwo groups of 15 patients each using the lotterymethod. They received either 5% natamycin (groupA) or 1% VRC (group B) topically as the primarytreatment. Double masking of treatment assignmentwas achieved by dispensing the medications in iden-

tical opaque bottles and by having the ward nurseswipe any white residue from the patient’s eye priorto study assessment as natamycin is delivered viasuspension, whereas VRC is in solution.

A detailed history of any systemic disease or localpredisposing factors like trauma, contact lens useand usage of topical steroids was also taken. Detailedexamination of ocular adnexa and the corneal ulcerwas performed including measurements of the sizeof epithelial defect, stromal infiltrate and hypopyon.Infiltrate/scar size and epithelial defect size weremeasured according to the protocol adapted by theHerpetic Eye Disease Study.24 Intraocular pressurewas assessed digitally. Corneal scraping was per-formed for every patient after instillation of localanaesthetic (lignocaine 2%). Either a blunt Kimura’sspatula or a sterile no. 15 surgical blade was used toperform scraping and the material taken both fromthe base as well as the edges of the ulcer.

The material thus obtained from scraping wasused for direct microscopic examination usingGram’s stain and 10% KOH mount and also inocu-lated into Sabouraud’s dextrose agar and brain heartinfusion broth for transportation to our referral labo-ratory for culture and identification of species bystandard microbiological procedures.25,26 Routinemicroscopy was done at our laboratory and patientswho tested positive for fungal hyphae and negativefor bacteria were included in our study. The treat-ment comprising of either 5% natamycin or 1% VRCeye drops was instituted.

Among the study medications, 5% natamycintopical formulation available commercially wasused, and topical 1% VRC eye drops were preparedby reconstituting lyophilized powder available as200-mg vials with sterile deionized water to make1% (10 mg/mL) solution of VRC that was stored inrefrigerator for 48 h. The drug was reconstitutedevery 48–72 h for continued use. One drop of ran-domized medication was applied 1 hourly to theaffected eye at least till 2 weeks. The treatmentregimen was strictly monitored, and the patientswere admitted in the hospital. Further dosagetitrated according to the patient’s response. Thepatients were instructed to get VRC reconstitutedevery fourth day from the hospital pharmacy afterinitial stay in the hospital. Additional standard treat-ment protocol included topical 0.3% ofloxacinhydrochloride q.i.d., 2% homatropine bromide q.i.d.and 0.5% timolol maleate eye drops b.i.d. if needed.

Patients were scheduled for follow-up examina-tion everyday for 1 week/earliest sign of resolution,as determined by ulcer size being static or localiza-tion of infiltrate. Subsequently they were followedevery third day for 2 weeks, then every week for2 weeks, then every 2 weeks till 2 months or untilcomplete resolution of infiltrate whichever was later.

Voriconazole and fungal keratitis 435


All the parameters of corneal infiltrate, epithelialdefect, hypopyon were recorded during the followup. Standard follow-up visits were taken as after 1,2, 4 and 8 weeks for statistical analysis.

Primary outcome was defined by the time takenfor the complete resolution of the ulcer. Statisticalanalysis was performed using non-parametricMann–Whitney test for comparing data in betweenthe groups and Wilcoxon signed rank test for com-paring data within the groups during the follow up.

RESULTS

Two groups of 15 patients each of microbiologicallyproven fungal corneal ulcers were started on treat-ment with either 5% natamycin (group A) or 1%VRC (group B) drops.

The age distribution of patients in the studywas 37.93 � 15.14 years in group A and 48.47 �13.53 years in group B, P-value being 0.054. Fungalkeratitis was found to be more common in men, 67%in group A and 73.3% in group B. The difference wasstatistically not significant (P = 0.689). There were nodefinite accompanying systemic factors except diabe-tes mellitus that was found only in three patients. Oninquiring about the local predisposing factors,trauma was found in 17 (56.7%) patients and priorhistory of topical steroid use in two (6.6%) patients(one in each group). Trauma was due to prior injurywith vegetative matter in 12 (40%) patients, cattletail in two (6.6%) patients and unidentifiable objectin the rest.

Mean size of the ulcer in millimetres in terms oflongest diameter * longest perpendicular diameterwas 4.10*3.26 in group A and 3.51*2.75 in groupB. The distribution was comparable (P = 0.254).Figure 1 shows the change in infiltrate size atfollow-up visits. Based on the depth of infiltratenoted at presentation, patients were distributed intothree groups – <30%, 30–70% and >70%. Seventeenpatients (56.7%) had depth >70%, nine (30%) haddepth 30–70%, and four (13.3%) had depth <30%.However, the distribution was comparable in boththe groups (P = 0.465). Hypopyon was present in21 (70%) of patients ranging from 0.5 to 4 mm.The height and distribution was comparable in boththe groups (P = 0.160). Figure 2 shows change inhypopyon size in each group during the follow up.In group A, the change was significant at all thefollow ups, and in group B, it was significant afterthe first follow up (P = 0.293) itself. In fact, there wasan increase in hypopyon size in three patients ingroup B at first follow up, and they were continuedon VRC under close supervision. Then the hypopyonwas found to decrease subsequently.

All ulcers healed completely in group A, and therewas one patient in group B who failed to respond to

the treatment. The average time of complete resolu-tion of corneal infiltrate in 15 patients of group Awas 24.33 days, and in 14 patients in group B it was27.42 days. It ranged from minimum of 10 days to amaximum 60 days. The treatment failure patient ingroup B perforated and required therapeutic pen-etrating keratoplasty. The average scar size wasfound to be 3.26*3 � 1.13*1.18 mm in group A and3.6*3.4 � 1.84*1.86 mm in group B. P-value whenboth the dimensions were compared was not signifi-cant (>0.05%). The distribution of the depth of thescar was comparable (P = 0.114) and found to belesser than infiltrate depth in 11 patients in both thegroups (36.6%). There was also formation of compli-cated cataract in two patients after healing of theulcer. The visual acuity on Snellen was convertedinto LogMAR for statistical analysis. Mean LogMARvisual acuity at presentation was 2.48 � 0.791 in

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

Size(presentation)

Size 1(1 week)

Size 2(2 weeks)

Size 3(4 weeks)

Siz

e in

mm

Group A LD

Group A LPD

Group B LD

Group B LPD

Figure 1. Mean size of the ulcer in both the groups at followups in terms of longest diameter (LD) and longest perpendiculardiameter (LPD) at presentation, 1, 2 and 4 weeks, respectively.

0.955

0.545

0.273

0.045

1.3861.227

0.35

0.10

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Hypopyon Hypopyon 1 Hypopyon 2 Hypopyon 3

Group A

Group B

Figure 2. Mean size of hypopyon in both the groups on followups in millimetres at presentation, 1, 2 and 4 weeks, respectively.

436 Arora et al.


group A and 2.48 � 0.846 in group B. Figure 3shows the change in visual acuity at follow ups. Thebest-corrected visual acuity at the last follow up ineach group was 1.368 � 0.887 in group A and1.775 � 1.036 in group B. The difference being insig-nificant statistically (P = 0.227). No adverse reactionsto study medications were observed in either group.

Figure 4 shows two of our patients; first patientpresented with pain and redness following traumawith grain husk for 2 days was started on natamy-cin after confirmation of fungal hyphae. The ulcerresolved in 4 weeks. Patient was found to haveimmature senile cataract for which cataract surgerywas performed 4 weeks after healing of the ulcer.The second patient had spontaneous onset of ulcerand presented with endothelial exudates. It wasmanaged with topical VRC, and the ulcer resolved inabout 5 weeks.

Fungal hyphae were present in 27 patients on10% KOH mount at presentation. Three patients ini-tially showed no hyphae on 10% KOH mount, butstaining from growth on Sabaraud’s dextrose agar(SDA) media on second/third day revealed fungalhyphae. They were then subsequently included inthe study. Antifungal treatment was instituted onlyafter visualizing fungal hyphae from growth. Table1 shows causative organisms found on culture.Overall, the culture was positive in 25 (83.3%)patients, Aspergillus spp. (40%) and Curvularia spp.(30%) being the most common isolates in eachgroup.

DISCUSSION

Corneal infections of fungal aetiology are commonand represent 30–40% of all cases of culture-positiveinfectious keratitis in India. The incidence of cornealinfections in India is almost 10 times of that reportedin USA.27

The major predisposing factor for fungal keratitishas been trauma and contact lens usage. Trauma has

been reported to be associated with 55–65% offungal corneal ulcers.28,29 The same was seen in ourstudy too where a large number of patients (56.7%)reported trauma prior to development of the ulcer.None of the patients had any history of contact lensusage.

Filamentous fungi have been reported as causativeagents in large proportions of mycotic corneal ulcersin tropical climates than in temperate climates1 asevident in the present study where filamentousorganisms were found in 100% of culture-positivecases, Aspergillus spp. and Curvularia spp. being themost common isolates. No case of yeast aetiologywas seen.

Fungal keratitis is difficult to treat and carries asignificant risk of intraocular involvement. Natamy-cin has been reported as the most effective medica-tion against Fusarium and Aspergillus.1,30 Primarytreatment failure has been reported in 31.3% of casesin a study including 115 patients. Large ulcer size,hypopyon and Aspergillus as causative organism havebeen reported as predictors of poor outcome withtopical 5% natamycin monotherapy.12 All the 15patients in the study healed well with topical nata-mycin alone. Average ulcer size being <5 mm andstrict compliance to treatment resulted in 100%success. The subset of patients was reasonablysmaller than earlier reported studies with largernumber of patients.

Of the newer antifungal agents VRC has beenreported as a highly potent triazole with 100% in vitrosusceptibility against common ocular fungal patho-gens compared with only 60–84% for fluconazole,itraconazole, amphotericin B and ketoconazole.18

VRC levels in aqueous following only topicaladministration have been found to be 6 mg/mL innon-inflamed human eyes that exceeds or meetsMIC90 for most pathogens.31 Its penetration topicallyis less clear and reports from human data areconflicting. Isolated reports of topical VRC alone orin combination with oral or intravenous route havegiven successful results for fungal keratitis notresponding to conventional treatment.32

There was present 93% success rate in healing ofthe fungal corneal ulcers with topical 1% VRCalone. On comparing the efficacy of topical 1% VRCwith that of topical 5% natamycin with comparabledrug frequencing, average time of resolution ofcorneal ulcer was more with VRC than natamycinbut the difference was statistically insignificant.Efficacy of both the drugs as a primary treatment infungal keratitis is being thereby comparable. Deependothelial abscess were present in 70% of patientsin the study. Complete healing in both the groupswith topical therapy alone highlights the penetra-tion of both the drugs through the cornea effectively.Penetration of VRC through intact corneal epithe-

VA VA 1 VA 2 VA 3 VA 4 VA fnNatamycin 2.48 2.233 2.037 1.784 1.384 1.368VRC 2.48 2.433 2.272 2.02 1.857 1.775

1

10

logM

AR

VA

Figure 3. Change in logMAR best-corrected visual acuity (VA)with natamycin (group A) and voriconazole (VRC) (group B) onfollow ups at 1, 2, 4, 8 weeks and follow up after healing.



lium effectively has been proven.33 Hence, systemicadjuvant therapy in the treatment of fungal keratitismust be weighed against the side-effects of systemicmedications.

The scar size was found to be comparable in twogroups; in patients with dense infiltrate it was morethan the size of infiltrate because of scarring resultingfrom surrounding corneal oedema. There was reduc-tion in depth of scar when compared with depth ofinfiltrate in both the groups after treatment in accor-dance with general criteria for the healing of the ulcer.

Both the drugs were found to be efficacious agentsin primary fungal keratitis management. There wasno additional beneficial effect of 1% topical VRCover 5% natamycin in the study. Considering the

a b

e f

c d

Figure 4. Day 1 of presentationof ulcer treated with 5% natamycin(a), at 2 weeks (b) and at 10 weeksafter cataract surgery (c). Day 1 ofpresentation of ulcer treated with1% voriconazole (d), at 3 weeks (e)and at 5 weeks (f).

Table 1. Identification of fungi isolated from corneal scrapingsof patients in both the groups

Fungi Natamycin(group A)

Voriconazole(group B)

No. % No. %

Aspergillus spp. 6 40 6 40A. flavus 2 13.2 2 13.2A. fumigatus 2 13.2 2 13.2A. niger 2 13.2 2 13.2

Curvularia spp. 4 26.6 5 33.3Fusarium spp. 1 6.6 2 13.2Aureobasidium pullans 1 6.6 0 0Culture negative 3 20 2 13.2Totals 15 100 15 100

438 Arora et al.


cost, the shelf life and the variable bioavailability oftopical 1% VRC, it may be maintained as second lineof treatment in the management of fungal keratitis,refractory to topical natamycin or other conventionalantifungal agents. Although in vitro studies and anec-dotal reports support VRC, a larger trial with morenumber of patients in comparable situations mayprove or disprove the efficacy of topical VRC in clini-cal settings.

REFERENCES

1. Srinivasan M. Fungal keratitis. Curr Opin Ophthalmol2004; 321–7.

2. Galaretta D, Tuft S, Ramsay A, Dart J. Fungal keratitisin London: microbiological and clinical evalution.Cornea 2007; 26: 1082–6.

3. Upadhyay MP, Karamcharya PC, Koirala S et al. Epi-demiologic characteristics, predisposing factors, andetiologic diagnosis of corneal ulceration in Nepal. Am JOphthalmol 1991; 111: 92–9.

4. Dunlop AA, Wright ED, Howlader SA et al. Suppura-tive corneal ulceration in Bangladesh: a study of 142cases examining the microbiological diagnosis, clini-cal and epidemiological features of bacterial andfungal keratitis. Aust N Z J Ophthalmol 1994; 22: 105–10.

5. Hagan M, Wright E, Newman M et al. Causes of sup-purative keratitis in Ghana. Br J Ophthalmol 1995; 79:1024–8.

6. Rosa RH Jr, Miller D, Alfonso EC. The changing spec-trum of fungal keratitis in south Florida. Ophthalmology1994; 101: 1005–13.

7. Srinivasan M, Gonzales CA, George C et al. Epidemi-ology and aetiological diagnosis of corneal ulcerationin Madurai, south India. Br J Ophthalmol 1997; 81: 965–71.

8. Bhartiya P, Daniell M, Constantinou M, Islam F,Taylor H. Fungal keratitis in Melbourne. Clin Experi-ment Ophthalmol 2007; 35: 124–30.

9. Manzouri B, Vafidis G, Wyse R. Pharmacotherapy offungal eye infections. Expert Opin Pharmacother 2001; 2:1849–57.

10. O’Day DM. Selection of appropriate antifungaltherapy. Cornea 1987; 6: 238–45.

11. Jones DB, Forster RK, Rebell G. Fusarium solani kerati-tis treated with natamycin. Arch Ophthalmol 1972; 88:147–54.

12. Lalitha P, Prajna N, Kabra A, Mahadevan K,Srinivasan M. Risk Factors for treatment outcome inFungal Keratitis. Ophthalmology 2001; 113: 526–30.

13. Espinel-Ingrofff A, Boyle K, Sheehan DJ. In vitro anti-fungal activities of Voriconazole and reference agentsas determined by NCCLS methods; review of theliterature. Mycopathologia 2001; 150: 101–15.

14. Ghannoum MA, Kuhn DM. Voriconazole: betterchances for patients with invasive mycoses. Eur J MedRes 2002; 7: 242–56.

15. Marco F, Pfaller MA, Messer SA et al. Antifungal activ-ity of a new triazole, voriconazole (UK-109,496), com-pared with three other antifungal agents tested againstclinical isolates of filamentous fungi. Med Mycol 1998;36: 433–6.

16. Sabo JA, Abdel- Rahman SM. Voriconazole: a newtriazole antifungal. Ann Pharmacother 2000; 34: 1032–43.

17. Maragon FB, Miller D, Giaconi JA, Alfonso EC. In vitroinvestigation of voriconazole susceptibility for kerati-tis and endophthalmitis fungal pathogens. Am J Oph-thalmol 2004; 137: 820–5.

18. Bunya VY, Hammersmith KM, Rapuano CJ, Ayres BD,Cohen EJ. Topical and oral voriconazole in treatmentof fungal keratitis. Am J Ophthalmol 2007; 143: 151–3.

19. Shah KB, Wu TG, Wilhelmus KR, Jones DB. Activityof voriconazole against corneal isolates of Scedospo-rium apiospermum. Cornea 2003; 22: 33–6.

20. Horstkotte MA, Weibmann J, Engelmann K, GarnerJC, Sobottka I. Sucessful treatment of ulceratingkeratitis due to Acremonium recifei with voriconazole.ESCMID 2007: Abstract no.1134_01_345.

21. Stewart RMK, Quah SA, Neal TJ, Kaye SB. The use ofVoriconazole in treatment of Aspergillus fumigateskeratitis. Br J Ophthalmol 2007; 1095–6.

22. Deng SX, Kamal MK, Hollander DA. The use of Vori-conazole in treatment of post- penetrating keratoplastyPaecilomyces keratitis. J Ocul Pharmacol Ther 2009; 25:175–8.

23. Baron DD, Ares R, Peralba T, Sanchez P, Jung R, BarciaM. Use of topical Voriconazole in fungal keratitis. ArcSof Esp Oftalmol 2008; 83: 493–6.

24. Barron BA, Gee L, Hauck WW et al. Herpetic EyeDisease Study: a controlled trial of oral acyclovir forherpes simplex stromal keratitis. Ophthalmology 1994;101: 1871–82.

25. Forbes BA, Sahm DF, Weissfeild AS. Laboratorymethods in basic mycology. Bailey and Scoff’s DiagnosticMicrobiology, 11th edn. St. Louis: Mosby, 2002; 711–98.

26. Koneman EW, Allen SD, Janda WM, SchreckenbergerPC, Wim WC. Colour Atlas and Textbook of Diagnostic Micro-biology, 5th edn. Philadelphia, PA: Lippincott, Williams& Wilkins, 1997; 983–1057.

27. Gonzales CA, Srinivasan M, Witcher JP, Smolin G.Incidence of corneal ulceration in Madurai district,south India. Ophthalmic Epidemiol 1996; 3: 159–66.

28. Panda A, Sharma N, Das G, Kumar N, Satpathy G.Mycotic keratitis in children: epidemiologic andmicrobiological evaluation. Cornea 1997; 16: 295–99.

29. Bharathi MJ, Ramakrishnan R, Vasu S et al. Epidemio-logical characteristics and laboratory diagnosis offungal keratitis: a three-year study. Indian J Ophthalmol2003; 51: 315–21.

30. Thomas PA. Current perspectives on ophthalmicmycoses. Clin Microbiol Rev 2003; 16: 730–97.

31. Vemulaconda GA, Hariprasad SM, Mieler WF, PrinceRA, Shah GK, Van Gelder RN. Aqueous and vitreous



concentration following topical administration of 1%voriconazole in humans. Arch Ophthalmol 2008; 126:18–22.

32. Hariprasad SM, Mieler WF, Lin TK, Sponsel WE,Graybill JR. Voriconazole in treatment of fungal eyeinfections: a review. Br J Ophthalmol 2008; 98: 871–8.

33. Thiel MA, Zinkernagel AS, Burhenne J, Kaufmann C,Haefeli WE. Voriconazole concentration in humanaqueous and plasma during topical or combinedtopical and systemic administration for fungalkeratitis. Antimicrob Agents Chemother 2007; 51: 239–44.

440 Arora et al.


Copyright of Clinical & Experimental Ophthalmology is the property of Wiley-Blackwell and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

Clin Exp Ophthalmol 2011_ p434

Documents

Transcript of Clin Exp Ophthalmol 2011_ p434