comparative antiviral efficacies of cidofovir...

Comparative Antiviral Efficacies of Cidofovir,Trifluridine, and Acyclovir in the HSV-1 RabbitKeratitis Model

Eric G. Romanowski,1 Stephen P. Bartels,2 and Y.Jerold Gordon1

PURPOSE. TO determine the relative antiviral inhibitory activity of topical 1% and 0.5% cidofovir,topical trifluridine (Viroptic; Burroughs-Wellcome, Research Triangle Park, NC), and topicalacyclovir (Zovirax; The Wellcome Foundation, London, UK) during a 7-day period for the treatmentof herpes simplex virus type 1 (HSV-1) keratitis and HSV-1 replication in the New Zealand rabbitocular model.

METHODS. In a series of four experiments using a two-eye design, a total of 80 New Zealand rabbitswere inoculated in both eyes with HSV-1 McKrae after epithelial scarification. Forty-eight hoursafter inoculation, the rabbits were randomly assigned to a treatment group. Five treatment groups(16 rabbits/group) were evaluated: I, 1% cidofovir, twice daily for 7 days; II, 0.5% cidofovir, twicedaily for 7 days; III, 3% acyclovir ointment, five times daily for 7 days; IV, 1% trifluridine, nine timesdaily for 3 days, then 4 times daily for 4 days; and V, control vehicle twice daily for 7 days. HSV-1dendritic keratitis was graded in a masked fashion by slit-lamp examination on days 2, 3, 5, 7, 9, 11,and 14. Ocular viral cultures were obtained after slit-lamp examination on days 1, 3, 5, 7, 9, 11, and14.

RESULTS. Compared with the control group, all four treatment groups demonstrated significantlylower viral titers, fewer HSV-1-positive eyes/total during the treatment period, lower keratitisscores, fewer eyes with keratitis/total, and a shorter time to resolution of keratitis. Within thetreatment groups, the 1% and 0.5% cidofovir treatments were significantly more effective thanacyclovir and trifluridine as measured by the previous viral and keratitis parameters.

CONCLUSIONS. Topical 1% and 0.5% cidofovir both appeared to be significantly more efficacious thantopical trifluridine and acyclovir, during a 7-day course, in the treatment of experimental HSV-1ocular disease in the New Zealand rabbit keratitis model. (Invest Ophthalmol Vis Sci. 1999;40:378-384)

Herpes simplex virus type 1 (HSV-1) is the leading causeof infectious corneal blindness in industrialized na-tions. There are approximately 400,000 cases of her-

pes simplex ocular disease in the United States each year.1

There are several antiviral agents available today for the treat-ment of HSV-1 keratitis. These antivirals include trifluridine(F3T, Viroptic; Burroughs-Wellcome, Research Triangle Park,NC), acyclovir (acycloguanosine, Zovirax; The Wellcome Foun-dation, London, UK), idoxyuricline GDU; Herplex Liquifilm;Allergan Inc., Irvine, CA), and vidarabine (Ara-A, Vira-A; Parke-Davis, Morris Plains, NJ). IDU and trifluridine are nucleosideanalogues. These agents are incorporated into the viral DNAand disrupt viral DNA synthesis. Ara-A inhibits viral DNA poly-merase and acts as a viral DNA chain terminator.2 Acyclovir is

From the ' Department of Ophthalmology, University of PittsburghSchool of Medicine, Pennsylvania; and 2Bausch & Lomb Pharmaceuti-cals, Pearl River, New York.

Supported by Bausch & Lomb Pharmaceuticals, Tampa, Florida; byThe Eye and Ear Foundation of Pittsburgh, Pennsylvania; and by CoreGrant for Vision Research EY08098, National Institutes of Health,Bethesda, Maryland.

Submitted for publication June 1, 1998; accepted August 11, 1998.Proprietary interest category: C2 and C3 (ER, YJG), E (SPB).Reprint requests: Y. Jerold Gordon, The Eye & Ear Institute, 203

Lothrop Street, Pittsburgh, PA 15213.

specifically activated by viral thymidine kinase and then ismodified by cellular enzymes to form acyclovir triphosphate,which binds preferentially to HSV-1 DNA polymerase andblocks viral replication.

These antiviral agents are quite effective in the treatmentof herpetic keratitis, demonstrating cure rates of 76% to 95%.2

However, there are adverse effects associated with all thesedrugs. Long-term use of IDU, Ara-A, and trifluridine has beenassociated with ocular toxicity (irritation, superficial punctatekeratitis, lacrimal punctal occlusion, follicular conjunctivitis,and contact dermatitis).2 There are additional practical factorsto consider when using these drugs. The formulation of theantiviral is one such factor. Patients complain about ointmentssuch as acyclovir, Ara-A, or IDU because they blur vision.Frequency of administration is another important factor. Pa-tient compliance usually decreases as the frequency of dosesincreases. The recommended treatment regimen for trifluri-dine drops is nine times a day until dendrites heal, then fourtimes a day for 14 days. Similarly, IDU drops are administeredevery hour during the day and every other hour at night (18 -20doses/d) until the dendrites heal, then four times a day for 14days. The antiviral ointments are administered five times a dayfor 14 days. The development of a long-acting, effective, non-toxic antiviral drop that requires less frequent daily adminis-

378Investigative Ophthalmology & Visual Science, February 1999, Vol. 40, No. 2Copyright © Association for Research in Vision and Ophthalmology

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933430/ on 05/03/2018

10VS, February 1999, Vol. 40, No. 2 Efficacy of Cidofovir, Trifluridine, and Acyclovir 379

tration for fewer days would represent a significant improve-ment over the currently available topical antivirals.

Cidofovir (S-HPMPC) is a promising broad-spectrum long-lasting antiviral with significant inhibitory activity against anumber of DNA viruses (human cytomegalovirus, HSV-1,HSV-2, varicella-zoster virus, and adenoviruses).3 Cidofovir is anucleoside analogue of cytosine that is actively transportedacross cell membranes. Once inside the cell, cellular kinasesconvert cidofovir to its active form, cidofovir-diphosphate. Theantiviral effect of cidofovir-diphosphate is the result of a selec-tive interaction with the viral DNA polymerase after which itacts as both a competitive inhibitor and as a chain terminatorduring DNA synthesis. The long intracellular half-life of cido-fovir is probably attributed to the accumulation of the metab-olite cidofovir monophosphate-choline, which may act as areservoir for cidofovir-diphosphate."

We previously demonstrated in prevention and treatmentstudies that topical administration of cidofovir significantlyreduced ocular viral titers and the duration of viral shedding inthe Ad5 McEwen-New Zealand rabbit ocular model.5'6 Maud-gal and De Clercq7 first demonstrated that frequent daily dos-ing with topical 0.2% cidofovir was effective in reducing her-petic keratitis. Subsequently, our group demonstrated thatfrequent dosing with 0.2% cidofovir (nine times a day X 4 days,then four times a day X 6 days) was as effective as trifluridinein reducing HSV-1 dendritic keratitis, lowering ocular viraltiters, and shortening the duration of HSV-1 shedding in thetear film.8 Recently, Kaufman et al.9 also demonstrated thatcidofovir was as effective as trifluridine in the treatment ofexperimental herpetic keratitis in rabbits.

The goal of the present study was to determine whethertwice daily dosing with higher concentrations of cidofovir (1%and 0.5%) would be as effective as topical 1% trifluridine(Viroptic) and 3% acyclovir (Zovirax), the current standards ofcare worldwide, in the treatment of experimental HSV-1 oculardisease in the New Zealand rabbit keratitis model.

METHODS

HSV-1 Virus Strain and CellsHSV-1 McKrae strain is a well-characterized neurovirulent lab-oratory strain that produces typical dendritic keratitis wheninoculated ocularly. A stock of HSV-1 McKrae was grown inVero cell monolayers at 37°C in a 5% CO2-water vapor atmo-sphere, harvested, aliquoted, and frozen at —70°C. Before use,the stock HSV-1 McKrae was titered using a standard plaqueassay.

Vero cells, a cell line derived from African Green monkeykidneys (American Type Culture Collection [ATCC], Rockville,MD) and A549 cells, an epithelial-like cell derived from humanlung carcinoma (ATCC), were grown and maintained in Eagle'sminimum essential medium with Earle's salts (Sigma Cell Cul-ture Reagents, St. Louis, MO), supplemented with 6% fetalbovine serum (Harlan Bioproducts for Science, Indianapolis,IN), 2.5 /xg/ml amphotericin B, 100 U/ml penicillin G, and 0.1mg/ml streptomycin (Sigma).

Experimental DrugsThe synthesis of cidofovir (S-HPMPC [(S)-9-(3-hydroxy-2-phophonylmethoxypropyO-cytosine]) was originally describedby Holy et al.,8 initially evaluated as an antiviral by De Clercq et

al.,9 and was made available to us by Storz Ophthalmics (PearlRiver, NY). A topical ocular formulation was prepared as 1%and 0.5% solutions. Control eye drops for cidofovir consistedof the vehicle alone.

Two antivirals used clinically for the treatment of HSV-1keratitis, 1% trifluridine (Viroptic Ophthalmic Solution) in theUnited States and 3% acyclovir (Zovirax Ophthalmic Ointment)worldwide were used as the standard of therapy controls.

AnimalsSix-week-old, 1.5-kg, female New Zealand albino rabbits ob-tained from either Green Meadows Rabbitry (Murrysville, PA)or Myrtle's Rabbitry (Thompson Station, TN). All animal studiesconformed to the ARVO Statement on the Use of Animals inOphthalmic and Vision Research. Institutional approval wasobtained, and institutional guidelines regarding animal experi-mentation were followed.

Experimental ProtocolIn a series of four experiments using a two-eye design, a totalof 80 New Zealand rabbits were inoculated in both eyes aftertopical anesthesia with 1% proparacaine eye drops and epithe-lial scarification (three interlocking circles of a 7.5-mm tre-phine) with 50 jLtl (32 X 105 plaque-forming units/eye) ofHSV-1 McKrae. Forty-eight hours after inoculation, the rabbitswere randomly assigned to a treatment group. Five treatmentgroups (16 rabbits/group) were evaluated: group I, 1% cidofo-vir, twice daily for 7 days; group II, 0.5% cidofovir, twice dailyfor 7 days; group III, 3% acyclovir ointment, five times daily for7 days; group IV, 1% trifluridine, nine times daily for 3 days,then four times daily for 4 days; and group V, control vehicletwice daily for 7 days. HSV-1 dendritic keratitis was graded ina masked fashion, using a scale of 0 to 4,'" by slit-lamp exam-ination with 0.2% Na fluorescein and Cobalt blue filter on days2, 3, 5, 7, 9, 11, and 14. Ocular viral cultures were obtainedafter slit-lamp examination by swabbing each eye in the upperand lower fornices with a cotton-tipped applicator on days 1,3, 5, 7, 9, 11, and 14. Each swab was then placed in 1.0 mloutgrowth media and frozen at — 70°C until titration.

Determination of HSV-1 Ocular TitersEach frozen HSV-1 ocular sample to be titered was thawed anddiluted serially (1:10) for three dilutions. Each dilution (0.1ml/well) was then inoculated onto A549 cells in duplicatewells of a 24-well plate. The virus was adsorbed for 1 hour at37°C in a 5% CO2-water vapor atmosphere. After adsorption, 1ml outgrowth media plus 0.5% methylcellulose was added toeach well, and the plates were incubated at 37°C in a 5%CO2-water vapor atmosphere. The wells were examined dailyfor progressive cytopathic effect, samples were stained with0.5% gentian violet after 4 days, and the number of plaques perwell was counted under a dissecting microscope (25 X). Theocular HSV-1 titer was then calculated and expressed as plaque-forming units per milliliter.

Statistical AnalysisAfter the completion of all experiments, the data from eachexperiment were analyzed statistically. Because comparableresults were obtained in each experiment, the data were thenpooled to obtain a larger subject number and analyzed usingthe appropriate statistical test (analysis of variance [ANOVA],


380 Romanowski et al. IOVS, February 1999, Vol. 40, No. 2

TABLE 1. Assessment of HSV-1 Replication in the HSV-1 New Zealand Rabbit Keratitis Model

1% Cidofovir 0.5% Cidofovir 3% Acyclovir 1% Trifluridine Control

Mean ocular titer(day 7), pfu/ml§

HSV-1-positive eyesper total||

Days 3-14(overall)

Days 3-7 (Rxperiod)

Days 9-14 (afterRx)

Mean duration ofHSV-1 shedding(days)§

0.5 ± 2.7 X 102* 2.5 ± 9.3 X 102 2.3 ± 6A X 102* 1.5 ± 4.0 X 103* 1.4 ± 2.3 X 105

48/174 (28)*t

48/91 (53)*t

0/83 (0)*t

4.1 ± 1.5*t

65/180 (36)*t

65/96 (68)**

0/84 (0)*f

5.1 ± 0.9*t

104/174 (60)**

69/92 (75)**

35/82 (43)

8.2 ± 4.3*

134/176 (77)

84/96 (88)*

50/80 (63)*

9.7 ± 3.9

127/178 (71)

91/91 (100)

36/87 (41)

9.6 ± 1.9

Rx, treatment.* P < 0.05 when compared with control.t P < 0.05 when compared with acyclovir and trifluridine.* P < 0.05 when compared with trifluridine.§ Analysis of variances, Duncan multiple comparison test.|| Chi-square analysis. Values in parentheses are percentages.

Kruskal-Wallis ANOVA, Duncan multiple comparisons test forANOVA and Kruskal-Wallis ANOVA, Monte Carlo Randomiza-tion Test, Fisher exact test, and chi-square analysis). Signifi-cance was established at the P ^ 0.05 confidence level.

RESULTS

Efficacy of Cidofovir, Trifluridine, and AcyclovirTreatment on HSV-1 Replication: Serial HSV-1Ocular TitersThe mean HSV-1 ocular titers for all groups were determinedby calculating the mean of all HSV-1 cultures per group perday. These results are demonstrated in Table 1 (day 7) andgraphically in Figure 1. All antivirals tested demonstrated lower

Duration of Rx

10'

Days

FIGURE 1. Demonstration of the mean HSV-1 ocular titers over timefor 0.5% cidofovir (O), 1% cidofovir (D), trifluridine (A), acyclovir (V),and control (O) groups. Asterisks denote the days on which all anti-viral treatments demonstrated significantly lower mean HSV-1 oculartiters than the control group. The y axis is a log scale.

mean HSV-1 titers compared with the control group (ANOVA,Duncan multiple comparisons test, P < 0.0005), with no dif-ferences among the treatment groups, on days 3, 5, 7, and 9.The trifluridine (day 11) and acyclovir (day 14) groups demon-strated a significant increase in mean viral titers (rebound)compared with the control and 1%, and 0.5% cidofovir groups(P ^ 0.02), after the cessation of treatment. In contrast, eyestreated with either 1% or 0.5% cidofovir demonstrated norebound of viral titers after the cessation of treatment.

HSV-1-Positive Eyes per Total

The number of HSV-1 positive eyes per total was determinedfor each group by ascertaining the number of eye swabs thatdemonstrated a positive HSV-1 culture per total number ofcultures. Table 1 summarizes these results. Overall from days 3to 14, 1% and 0.5% cidofovir and acyclovir demonstrated fewerHSV-1-positive eyes than the control and trifluridine groups(chi-square analysis, Fisher exact test, P < 0.000001, P <0.000001, P < 0.03, respectively). Trifluridine demonstratedno difference compared with the control. There was no differ-ence between 1% and 0.5% cidofovir, and both had fewerHSV-1-positive eyes than the acyclovir group (P < 0.00002).

Breaking down these results to days 3 to 7 (during thetreatment period) and days 9 to 14 (after the cessation oftreatment), we find that all antiviral treatment groups demon-strated fewer HSV-1-positive eyes (P < 0.0004) from days 3 to7 than the control group. Within the treatment groups, 1%cidofovir demonstrated fewer HSV-1-positive eyes than thetrifluridine and acyclovir groups (P < 0.002), and 0.5% cido-fovir demonstrated fewer positive eyes than only the trifluri-dine group (P < 0.002). There was no difference between the1% and 0.5% cidofovir groups. The acyclovir group had fewerHSV-1-positive eyes than the trifluridine group (P < 0.04).

From days 9 to 14, after the cessation of treatment, 1% and0.5% cidofovir groups demonstrated complete clearance ofvirus from all eyes. In contrast, the acyclovir and trifluridinegroups maintained more HSV-1-positive eyes than both cido-fovir groups (P < 0.000001). Acyclovir demonstrated a com-


IOVS, February 1999, Vol. 40, No. 2 Efficacy of Cidofovir, Trifluridine, and Acyclovir 381

Duration of Rx

100

1 1 * 14

FIGURE 2. Demonstration of the percentage of HSV-1 -positive eyesover time for 0.5% cidofovir (O), 1% cidofovir (•) , trifluridine (A),acyclovir (V), and control (O) groups. *Days on which both 1% and0.5% cidofovir groups demonstrated significantly fewer HSV-1-positiveeyes than the trifluridine and acyclovir groups.

parable number of HSV-1-positive eyes from days 9 to 14 tothat of the control group but fewer than the trifluridine group(P < 0.02). The trifluridine group also demonstrated more(P = 0.01) than the control group. The number of HSV-1-positive eyes per group are different (Table 1) because theHSV-1 inoculation of several of the rabbit eyes was unsuccess-ful, and some of the rabbits died before the end of the studydue to viral encephalitis.

Percentage of HSV-1-positive eyes per day is demon-strated graphically in Figure 2. There were no differencesamong the treatment groups and the control group on days 1and 3. On day 5, the 1% cidofovir group demonstrated fewerHSV-1-positive eyes than all other groups (Monte Carlo Ran-domization Test, P < 0.001). By day 7, all treatment groups hadfewer HSV-1-positive eyes than the control group (P < 0.03).Both cidofovir groups and the acyclovir group demonstratedfewer positive eyes than the trifluridine group (P < 0.002). Ondays 9 and 11 the 1% and 0.5% cidofovir groups demonstratedfewer HSV-1-positive eyes than the control, trifluridine, and

acyclovir groups and on day 14 fewer than the trifluridine andacyclovir groups (P < 0.0008). The acyclovir group had fewerHSV-1-positive eyes than the control group on day 9 (JP <0.002) and fewer than the trifluridine group on day 11 (P =0.05). However, on days 11 and 14 for the trifluridine groupand day 14 for the acyclovir group, these groups had moreHSV-1-positive eyes than the control group (P < 0.03) due tothe rebound of viral replication.

Duration of Shedding

The duration of HSV-1 shedding was ascertained for each eyeby determining the final day on which a positive HSV-1 culturewas detected and calculating the mean for each treatmentgroup. These results are presented in Table 1. The 1% and 0.5%cidofovir groups demonstrated a significant reduction in themean duration of shedding (3 to 5 days) compared with thecontrol, trifluridine, and acyclovir groups (ANOVA, Duncanmultiple comparisons, P < 0.00004). There were no differ-ences between the cidofovir groups, between the trifluridineand control groups, and between the acyclovir and controlgroups. However, the acyclovir group did demonstrate a sig-nificant decrease in mean duration of shedding compared withthe trifluridine group (P < 0.04).

Efficacy of Cidofovir, Trifluridine, and AcyclovirTreatments on HSV-1 Keratitis: SerialKeratitis Scores

The mean epithelial keratitis scores were determined by cal-culating the mean keratitis scores of all eyes per group per day.These results are presented in Table 2 (day 7) and graphicallyin Figure 3- On days 3 to 11, all antiviral treatment groupsdemonstrated lower keratitis scores than the control group(Kruskal-Wallis ANOVA, Duncan multiple comparisons, P <0.01) except for trifluridine on day 11, which was no differentfrom the control group. From days 5 to 14, 1% and 0.5%cidofovir groups had lower keratitis scores than trifluridine andacyclovir groups (P < 0.01), whereas there were no differ-ences between the 1% cidofovir group and the 0.5% cidofovirgroup. During this period, acyclovir-treated eyes demonstratedlower keratitis scores than the trifluridine group (P < 0.01). Onday 14, a rebound in keratitis scores was demonstrated in theacyclovir and trifluridine groups that correlated to an increasein viral titers, with the result that their scores were greater

TABLE 2. Assessment of HSV-1 Keratitis in the HSV-1 New Zealand Rabbit Keratitis Model

1% Cidofovir 0.5% Cidofovir 3% Acyclovir 1% Trifluridine Control

Mean keratitis score (day 7)§Eyes with keratitis per total||

Days 3-14 (overall)Days 3-7 (Rx period)Days 9-14 (after Rx)

Mean time to resolution ofkeratitis (days)H

26/174(15)*t24 /91 (26)*t2 /83 (2)*t

4.5 ± 0.9*t

0.3 ±0.1*t

41/180 (23)*t36/96 (38)*t7/84 (8)*t

5.3 ± 1.1't

0.3 ± 0.3't

75/174(43)*$56/92 (61)**33/82 (40)**

7.0 ± 2.8*

0.7 ± 0.5*

134/176(76)*81/96 (84)*53/80 (66)

7.9 ± 4.1*

2.3 ± 0.5

116/178(65)91/91(100)54/87 (62)

13.2 ± 1.3

Rx, treatment; ANOVA, analysis of variance.* P < 0.05 when compared with control.t P < 0.05 when compared with acyclovir and trifluridine.t P < 0.05 when compared with trifluridine.§ Kruskal-Wallis ANOVA, Duncan multiple comparisons test.|| Chi-square analysis. Values in parentheses are percentages.H ANOVA, Duncan multiple comparison test.



3.0 - iDuration of Rx

0.5

0.0

FIGURE 3. Demonstration of the mean HSV-1 keratitis scores over timefor 0.5% cidofovir (O), 1% cidofovir (•) , trifluridine (A), acyclovir (V),and control (O) groups. *Days on which all antiviral groups demon-strated significantly lower keratitis scores than the control group;tdays on which the 1% and 0.5% cidofovir groups demonstratedsignificantly lower keratitis scores than the trifluridine and acyclovirgroups.

(P < 0.005) than the control scores. There was no significantrebound in keratitis scores demonstrated in the 1% and 0.5%cidofovir groups, nor was there any associated actively repli-cating HSV-1.

Eyes with Keratitis

The number of eyes with epithelial keratitis per total (Table 2)was calculated for each group by determining the number ofeyes that demonstrated at least a minimal keratitis score pertotal number of eyes examined. Overall from days 3 to 14, the1% cidofovir, 0.5% cidofovir, and acyclovir groups all demon-strated fewer eyes with keratitis than the control and trifluri-dine groups (chi-square analysis, Fisher exact test, P <0.000001). There was no significant difference demonstratedbetween the control and trifluridine groups or between the 1%and 0.5% cidofovir groups. The 1% and 0.5% cidofovir groupsboth demonstrated fewer eyes with keratitis than the acyclovirgroup (P < 0.00007).

Breaking down the overall number of eyes with keratitis todays 3 to 7 (during the treatment period) and days 9 to 14 (afterthe cessation of treatment), we find that all antiviral treatmentgroups demonstrated fewer eyes with keratitis (P < 0.00004)from days 3 to 7 than the control group. Within the antiviralgroups, 1% and 0.5% cidofovir groups demonstrated fewereyes with keratitis than the trifluridine and acyclovir groups(P < 0.003), and the acyclovir group had fewer than thetrifluridine group (P < 0.004).

Only 1% and 0.5% cidofovir and acyclovir groups demon-strated fewer eyes with keratitis after treatment was halted(days 9 to 14) than the control group (P < 0.006). There wasno difference between the trifluridine group and the controlgroup. Just as during days 3 to 7, from days 9 to 14 the 1% and0.5% cidofovir groups demonstrated fewer eyes with keratitisthan the trifluridine and acyclovir groups (P < 0.000002), andthe acyclovir group had fewer than the trifluridine group (P <0.002).

Serial percentage of eyes with keratitis is demonstrated inFigure 4. Day 3 was the first day that a significant differencewas demonstrated, when the 1% cidofovir group demonstratedfewer eyes with keratitis than the control and 0.5% cidofovirgrovips. From days 5 to 11, the 1% and 0.5% cidofovir groupsdemonstrated fewer eyes with keratitis than the control group(Monte Carlo Randomization Test, P < 0.000001) and fromdays 5 to 14 had fewer than the acyclovir and trifluridinegroups (P < 0.03). The 0.5% cidofovir group demonstratedseveral eyes with keratitis on days 9, 11, and 14, and the 1%cidofovir group had several eyes with keratitis on day 14, butthis keratitis was not associated with any viral replication. Thetrifluridine group demonstrated fewer eyes with keratitis thanthe control group on days 5 and 9 only (P < 0.005) and theacyclovir group on days 5, 7, and 9 only (P < 0.000006). Theacyclovir group had fewer eyes with keratitis than the trifluri-dine group on days 5, 7, and 14 (P < 0.03).

Time to Resolution of KeratitisThe time to resolution of keratitis was ascertained for each eyeby determining the first day on which a negative keratitisexamination was detected after a previously positive examina-tion and then calculating the mean time for each treatmentgroup. These results are presented in Table 2. All antiviraltreatment groups demonstrated a significant reduction in themean time to resolution of keratitis compared with the controlgroup (ANOVA, Duncan multiple comparisons test, P <0.000001). In addition, the 1% and 0.5% cidofovir groupsdemonstrated a shorter mean time to resolution of keratitisthan the acyclovir and trifluridine groups (P < 0.009). Therewere no differences between the two cidofovir groups orbetween the trifluridine and acyclovir groups.

DISCUSSION

Cidofovir is a promising broad-spectrum antiviral with signifi-cant in vitro inhibitory activity against a number of DNA viruses

Duration of Rx

100

Hi 40 -

20 -

FIGURE 4. Demonstration of the percentage of eyes with HSV-1 ker-atitis over time for 0.5% cidofovir (O), 1% cidofovir (•) , trifluridine(A), acyclovir (V), and control (O). 'Days on which the 1% and 0.5%cidofovir groups demonstrated significantly fewer eyes with keratitisthan the control group; fdays on which the 1% and 0.5% cidofovirgroups had significantly fewer eyes with keratitis than the trifluridineand acyclovir groups.


JOVS, February 1999, Vol. 40, No. 2 Efficacy of Cidofovir, Trifluridine, and Acyclovir 383

(human cytomegalovirus, HSV-1, HSV-2, varicella-zoster virus,and adenoviruses).3 Cidofovir (Vistide; Gilead Sciences, FosterCity, CA) is now available in the United States and Europe as anapproved systemic therapy for cytomegalovirus retinitis inAIDS patients. It has been used experimentally both topically11

and systemically12 to treat AIDS patients with mucocutaneousinfections due to acyclovir-resistant HSV. Maudgal and DeClercq first demonstrated that frequent daily dosing with top-ical 0.2% cidofovir was effective in reducing herpetic keratitisin rabbits.7 Subsequently, our group demonstrated that fre-quent daily dosing with 0.2% cidofovir was as effective astrifluridine in reducing HSV-1 keratitis and viral replication inthe HSV-1-New Zealand rabbit keratitis model.8

In the present study, all the antiviral agents tested wereeffective in reducing HSV-1 ocular replication and keratitiscompared with the control group. However, treatment witheither 1% or 0.5% cidofovir twice daily was significantly moreeffective than trifluridine and acyclovir in their multidose reg-imens during the 7-day treatment period. After the cessation oftreatment, there was total clearance of virus from all eyes in the1% and 0.5% cidofovir groups, and there was no rebound ofHSV-1 replication and keratitis such as that demonstrated in thetrifluridine and acyclovir groups. This sttidy confirms and ex-tends those previous reports by demonstrating that twice dailydosing with higher concentrations of cidofovir (1% and 0.5%)was more effective in reducing HSV-1 replication and keratitisthan the current standards of care, over a 7-day treatmentperiod, in the New Zealand rabbit keratitis model.

The present study also demonstrated that 3% acyclovirointment was superior to 1% trifluridine drops in the treatmentof experimental herpetic keratitis during the 7-day treatmentperiod. These results differ from those obtained from twoprevious clinical trials that demonstrated no significant differ-ence between trifluridine and acyclovir in the treatment ofherpetic keratitis.1314 However, those clinical studies com-pared 3% acyclovir ointment to 2% trifluridine ointment usingthe same treatment regimen (5 times per day until the den-drites healed or a maximum of 2 weeks). The differencesbetween the vehicles (liquid drops versus ointment), concen-trations of trifluridine (1% versus 0.2%), and the duration oftreatment (7 days versus therapy until the dendrites healed ora maximum of 2 weeks) may explain the differences in resultsbetween the present study and previous studies.

This study was specifically designed to assess whether thelong intracellular half-life of cidofovir observed in vitro3 wouldreflect as prolonged antiviral inhibitory activity compared withtrifluridine and acyclovir in an in vivo herpetic eye model.Recommended clinical treatment for both antivirals calls for 14days of continuous therapy. Consequently, the experimentaldesign of this study does not reflect the antiviral efficacy oftrifluridine and acyclovir therapy after 2 weeks of treatment. Inthe present study, if the therapy had not been abruptly termi-nated at 7 days, there would not have been a rebound of HSV-1ocular titers and keratitis.

Treatment of patients with HSV-1 ocular disease usingtopical 1% or 0.5% cidofovir offers several potential advantagesover the current standards of care. The advantages of cidofovirover trifluridine would include decreased frequency of dosing(twice daily for cidofovir versus nine times daily for trifluri-dine), decreased duration of treatment (7 days versus 14 days),decreased toxicity (topical 0.5% cidofovir twice daily for 7 dayshas been shown to be nontoxic to the corneas of healthy

volunteers15 and topical administrations of as much as 2%cidofovir twice daily for 10 days have been demonstrated to benontoxic to corneas of uninfected New Zealand rabbits5), and,possibly, better antiviral efficacy. The advantages of cidofovirover acyclovir would include decreased frequency of dosing(twice daily for cidofovir versus five times daily for acyclovir),decreased duration of treatment (7 days versus 14 days), for-mulation (liquid drop versus ointment), and, possibly, betterantiviral efficacy.

This study also demonstrated that 1% cidofovir wasslightly more effective than 0.5% cidofovir in the rapid reduc-tion of viral replication and keratitis. This is most likely becauseof a quicker achievement of therapeutic tissue levels of cido-fovir with the more concentrated solution. This slight earlyadvantage was quickly overcome by continued treatment with0.5% cidofovir. Despite these findings, there were no majordifferences between 0.5% cidofovir and 1% cidofovir in thetreatment of HSV-1 ocular disease in the New Zealand rabbitkeratitis model. We therefore advocate topical 0.5% cidofovirtwice daily for 7 days as the treatment for HSV-1 keratitis inpatients. Phase 2 clinical trials are currently under way todetermine the efficacy of this treatment regimen.

Topical trifluridine and acyclovir have been shown to bevery effective in the treatment of HSV-1 ocular disease. In orderfor a new topical antiviral agent to be introduced for patientuse, it must be as effective as or more effective than currentantivirals. It should be formulated as a liquid, require infre-quent daily dosing for fewer days, and be nontoxic to the eye.Cidofovir meets all these criteria in experimental and phase 1clinical studies. If phase 2 and 3 clinical studies confirm thesefindings, then topical cidofovir would represent a significantimprovement over current topical antiviral agents for the treat-ment of ocular herpes simplex infections.

References

1. Pepose JS, Leib DA, Stuart PM, Easty DL. Herpes simplex virusdiseases: anterior segment of the eye. In: Pepose JS, Holland GN,Willhelmus KR, eds. Ocular Infection and Immunity. St. Louis,MO: Mosby; 1996:905-932.

2. Pavan-Langston D. Clinical disease herpetic infections. In: SmolinG, Thoft RA, eds. The Cornea. 3rd eel. Boston: Little, Brown & Co;1994:183-215.

3- De Clercq E. Therapeutic potential of HPMPC as an antiviral drug.Rev Meet Virol. 1993;3:85-96.

4. Naesens L, Snoeck R, Andrei G, et al. HPMPC (cidofovir), PMEA(adefovir) and related acyclic nucleoside phosphonate analogues:a review of their pharmacology and clinical potential in the treat-ment of viral infections. Antiviral Cbetn Chemother. 1997;8:1-23.

5. Gordon YJ, Romanowski EG, Araullo-Cruz T. Topical HPMPCinhibits adenovirus type 5 in the New Zealand rabbit ocular rep-lication model. Invest Ophthalmol Vis Set. 1994;35:4l35-4l43.

6. Gordon YJ, Romanowski E, Araullo-Cruz T, De Clercq E. Pretreat-ment with topical 0.1% (s)-l<3-hydroxy-2-phosphonylmethoxypro-pyl)cytosine inhibits adenovirus type 5 replication in the NewZealand rabbit ocular replication model. Cornea. 1992; 11:529 -533.

7. Maudgal PC, De Clercq E. (S>l-(3-hydroxy-2-phosphonylme-thoxypropyOcytosine in the therapy of thymidine kinase-positiveand -deficient herpes simplex virus experimental keratitis. InvestOphthalmol Vis Sci. 1991;32:18l6-1820.

8. Gordon YJ, Romanowski EG, Araullo-Cruz T. HPMPC, a broad-spectrum topical antiviral, inhibits HSV-1 replication and promoteshealing of dendritic keratitis in the NZ rabbit ocular model. Cor-nea, 1994;13:5l6-52O.



9. Kaufman HE, Varnell ED, Thompson HW. Trifluridine, cidofovir,and penciclovir in the treatment of experimental herpetic keratitis.Arch Ophthalmol. 1998;116:777-780.

8. Holy A, Rosenberg I. Synthesis of 9-(2-phosphonylmethoxyethy-l)adenine and related compounds. Coll Czech Chem Commun.1987;52:2801-2909.

9. De Clercq E, Sakuma T, Baba M, et al. Antiviral activity of phos-phonylmethoxyalkyl derivatives of purines and pyrimidines. Anti-viral Res. 1987;8:26l-272.

10. Gordon YJ, Armstrong JA, Brown SI, Becker Y. Herpesvirus type 1thymidine kinase. Am J Ophthalmol. 1983:95:175-181.

11. Snoeck R, Andrei G, Gerard M, et. al. Successful treatment ofprogressive mucocutaneous infection due to acyclovir and foscar-net-resistant herpes simplex virus with (S)-l<3-hydroxy-2-phos-phonylmethoxypropyOcytosine (HPMPC). Clin Infect Dis. 1994;18:570-578.

12. Lalezari JP, Drew WL, Glutzer E, et al. Treatment with intrave-nous (S)-l-[3-hydroxy-2-(phosphonylmethoxy)propyl]cytosineof acyclovir-resistant mucocutaneous infection with herpes sim-plex virus in a patient with AIDS. / Infect Dis. 1994;170:570-572.

13. La Lau C. Oosterhuis JA, Versteeg J, et al. Acyclovir and trifluoro-thymidine in herpetic keratitis: a multicentre trial. BrJ Ophthal-mol. 1982;66:506-508.

14. Hovding G. A comparison between acyclovir and trifluorothymi-dine ophthalmic ointment in the treatment of epithelial dendritickeratitis: a double blind, randomized parallel group trial. AdaOphthalmol. 1989:67:51-54.

15. Romanowski EG, Gordon YJ. The effect of topical NSAIDS on theantiviral activity of topical cidofovir in the Ad5/NZ rabbit ocularmodel [ARVO Abstract]. Invest Ophthalmol Vis Sci. 1998;39(4):SI48. Abstract nr 691.


comparative antiviral efficacies of cidofovir...

Documents