Effects of Drug Therapy Ontoxoplasma Cysts in an Animal

Effects of Drug Therapy on Toxoplasma Cysts in an Animal Model of Acute and Chronic Disease

Paula D. Gormley, Carlos E. Pavesio, Darwin Minnasian, and Susan Lightman

PURPOSE. TO evaluate the effects of drug therapy on the clinical course of acute acquired Toxo- plasma retinochoroiditis and on the number of Toxoplasma cysts present in the brain and ocular tissues in the hamster animal model. METHODS. The Syrian golden hamster animal model of Toxoplasma retinochoroiditis was used. In acute disease, systemically administered atovaquone was compared with conventional therapies (pyrimethamine combined with sulfadiazine; clindamycin; and spiramycin). The clinical course of the ocular disease was determined with retinal examination and photography of the fundus. The number of Toxoplasma cysts remaining after treatment was evaluated in aliquots of brain homogenate and in retinal tissue. The effect of atovaquone on cerebral Toxoplasma cyst count was also studied in chronic disease. RESULTS. None of the drugs administered altered the course of the acute disease, judged by clinical examination. Atovaqvione alone significantly reduced the number of cerebral Toxoplasma cysts after acute disease. Atovaquone also significantly reduced the cerebral Toxoplasma cyst count in chronic disease. CONCLUSIONS. Tissue cysts are believed to be responsible for reactivation of Toxoplasma retinochoroiditis. Atovaquone has the potential to reduce the risk of recurrent disease. (Invest Ophthalmol Vis Set. 1998;39:1171-1175)

cular toxbplasmosis is the most common cause of Hydroxynaphthoquinones were first developed (The Well- human retinochoroiditis1"3 and accounts for as much come Foundation, Beckenham, Kent, UK) as antimalarial drugs Oas 55% of all cases worldwide.4 Treatment of the when shortage of quinine during World War II prompted disease presents a problem to ophthalmologists, and there is search for alternative preparations. The early drugs were inef- much debate about the most effective treatment regimen.5 fective in human disease because of poor absorption and rapid Clinical experience has shown that the current treatment does metabolism, but continued research resulted in the synthesis of not seem to alter the natural progression of the disease,6 and atovaquone in the 1980s. The drug proved more active against there is no evidence that the available drugs are effective strains of Plasmodium falciparum in vitro and in vivo.12 against the tissue cysts that are believed to be responsible for Unlike the earlier hydroxynaphthoquinones, atovaquone did the disease's recurrence.7 Evaluation of these drugs in the not undergo rapid hepatic metabolism. The drug was subse- treatment of cysts in patients is complicated by ethical consid- quently shown to be active against the related parasite T. erations and retinal biopsy cannot be justified in this condition. gondii. The hydroxynaphthoquinones are potent inhibitors of The Syrian golden hamster has recently been described as the mitochondrial electron transport chain, competing with a reliable animal model of Toxoplasma retinochoroiditis.8 This the electron carrier ubiquinone13 at the ubiquinone-cyto- model is used in the present study to evaluate and compare the chrome c reductase region (complex III). effect of currently available drugs on the clinical course of acute acquired Toxoplasma retinochoroiditis and on the number of Toxoplasma cysts remaining in the cerebral tissues after MATERIALS AND METHODS infection. Atovaquone, a new drug that is effective against the tachyzoite and the tissue cysts of Toxoplasma gondii in vivo Maintenance of Toxoplasma Stocks 9 11 and in vitro, " was evaluated in acute and chronic infections. The avirulent ME49 cyst-forming strain of T. gondii maintained Atovaquone is a hydroxynaphthoquinone, /rarcs-2[4-(4-chloro- in mouse brain was kindly supplied by Jack Remington, Palo phenyl)cyclohexyl]-3-hydroxy-l,4-naphthalenedione, with an Alto Medical Research Foundation, (Palo Alto, CA). For main- empiric formula C22H19O3C1. tenance of the stocks, a chronically infected mouse was killed, and its brain tissue was gently homogenized with 1 ml phosphate-buffered saline, by mortar and pestle. A 0.2-ml aliquot of From the Institute of Ophthalmology and Moorfields Eye Hospital, this suspension was inoculated intraperitoneally onto each of London, United Kingdom. the adult BALB/c mice (weight, 22-24 mg), which were used Supported by Frost TFC Charitable Trust, Research into Eye Dis- ease Fund. for maintenance of the stocks. Submitted for publication May 9, 1997; revised November 7, 1997; accepted January 28, 1998. Experimental Animals Proprietary interest category: N. Reprint requests: Susan Lightman, Institute of Ophthalmology and Adult female outbred Syrian golden hamsters (B and K Univer- Moorfields Eye Hospital, City Road, London ECIV 2PD, UK. sal; The Field Station, Grimston, Aldborough, Hull, UK), each

Downloaded from iovs.arvojournals.org on 09/28/2021 1172 Gormley et al. IOVS, June 1998, Vol. 39, No. 7

weighing 100 g to 120 g, were used in the experiment. Animals activity. The solubility of atovaquone was improved with 50 were kept in single cages at an average room temperature of ml/1 chemophore and 50 ml/1 acetone (provided by V. An- 20°C in natural light. They were fed pellets (HMB; Special Diet drews, Pharmacy Department, Moorfields Eye Hospital, Lon- Services, Witham, UK) and drank tap water ad libitum from a don, UK) and the final concentration was 160 /i-g/ml, measured drinking bottle. All hamsters had normal findings on physical by immunoassay (performed by Martin French, Wellcome Re- and ophthalmic examination and were free of serum antibod- search Laboratory, Beckenham, UK) The average daily number ies to T. gondii at the beginning of the experiment. of milliliters of water drunk by each animal was measured and All procedures used were in accordance with the Animal the drug supply changed weekly. After a 4-week treatment (Scientific Procedures) Act of 1986 and conformed to the course, all hamsters received fresh drinking water (no drug) for ARVO Statement for the Use of Animals in Ophthalmic and an additional 2 weeks. The hamsters were then killed in a Vision Research. carbon dioxide chamber, and their brains and eyes were removed. The fresh brain was used immediately for cerebral cyst Antimicrobial Agents counts, and the eyes were fixed in 10% formaldehyde for The drugs used were pyrimethamine (Sigma, St. Louis, MO), histologic examination. sulfadiazine sodium (May and Baker, Dagenham, UK), clindamycin phosphate (Upjohn, Crawley, UK), vancomycin (Eli Treatment Protocol for Chronic Infection Lilly, Basingstoke, UK), spiramycin (kindly supplied by Celso Reis, Gerencia de Pesquisas Medicas, Rhodiafarma, Sao Paulo, The study group consisted of 16 hamsters that had chronic, Brazil), and atovaquone (The Wellcome Foundation). inactive Toxoplasma retinochoroiditis (diagnosed by retinal examination) 12 weeks after inoculation. Eight hamsters were Induction of Toxoplasmosis in the Hamster treated with atovaquone at the same concentration as in the Animal Model previous experiment. The treatment period was 4 weeks, in A suspension of the ME49 T. gondii strain was prepared from keeping with the usual treatment period in human disease. the infected stock mouse 6 weeks after the inoculation of the Eight untreated animals served as control subjects. After 4 mouse with the parasite. According to the protocols described weeks, all the animals were killed, and their brains and eyes previously,8 each hamster was inoculated intraperitoneally were removed. The fresh brain was used immediately for with 0.2 ml mouse brain homogenate on day 0 of the experi- cerebral cyst counts, and the eyes were fixed in 10% formal- ment. dehyde for histologic examination.

Clinical Course Evaluation of the Effect of Treatment Retinal examination with an indirect ophthalmoscope and a Cerebral Cyst Counts. The brain of each hamster was 30-diopter lens was performed on all hamsters on days 0, 7, 10, cut along the sagittal plane, and one half was homogenized and 14 and weekly thereafter. Fundus photographs were taken in 1 ml phosphate-buffered saline, ground by mortar and with a fundus camera. Full pupillary dilation was achieved with pestle. A 0.025-ml aliquot of homogenate was placed on a one drop each of tropicamide 1% (Alcon Laboratories, Herts, microscope slide and mounted with a coverslip (24 mm X UK) and phenylephrine 2.5% (Smith and Nephew Medical, 24 mm). The number of Toxoplasma cysts were counted in Hull, UK) instilled 30 minutes before the examination. four O.O25-ml aliquots from each hamster by light microscopy (magnification, X100). To calculate the total number Treatment Protocol for Acute Infection of cysts in the whole brain, the sum of the number of cysts At the first clinical signs of retinal disease, the hamsters were counted in the four aliquots was multiplied by 20. Cyst randomly assigned to one of five groups of 10 animals, and counts in the brain after the different treatment protocols treatment was administered in the drinking water according to were compared using Student's f-test and one-way analysis the following dosage regimens: group 1, untreated control; of variance (ANOVA). group 2, treated with 15 mg/1 pyrimethamine and 1000 mg/1 Retinal Cyst Count. Retinal cysts were counted in sulfadiazine; group 3, treated with 400 mg/1 clindamycin; chronic infection only. Both eyes from each hamster were group 4, treated with 2000 mg/1 spiramycin; and group 5, fixed in 10% formal saline and embedded in paraffin. Thick treated with 476 mg/1 atovaquone. sections (0.3 ju,m) were cut from three levels. One section from The drug dose was calculated from the human treatment each level was stained with hematoxylin and eosin and exam- regimen correlating to milligrams per kilogram of body weight ined for Toxoplasma cysts by light microscopy (magnification, in the hamsters. The daily human treatment regimens were as follows: 50 nig pyrimethamine, 4 g sulfadiazine, 1200 mg X100). The number of cysts in 48 sections from each group clindamycin, 3 g spiramycin, and 3 g atovaquone. The 8-ml (one section from each of the three levels from both eyes of average daily fluid consumption of hamsters kept under labo- each animal) were compared. The relatively small sample did ratory conditions'4 was used to calculate the concentration of not provide sufficient evidence of normally distributed data. drug supplied in the drinking water. Therefore, nonparametric equivalents of the f-test and of In a previous small pilot study, diarrhea developed in ANOVA (Mann-Whitney and Kruskal-Wallis tests) were also three of five hamsters treated with clindamycin, and they died used to confirm the results, although an assumption of nor- within 24 hours. The addition of 170 mg/1 vancomycin (calcu- mally distributed cyst counts was not crucial to the validity of lated from the human treatment regimen of 1 g/day) to the the f-tests (the sampling distribution of the difference between drinking water in this treatment group prevented further side the means is expected to be normal, even when the data have effects. Vancomycin is not known to have anti-Toxoplasma a skewed distribution).

Downloaded from iovs.arvojournals.org on 09/28/2021 IOVS, June 1998, Vol. 39, No. 7 Effects of Drug Therapy on Toxoplasma Cysts 1173

FIGURE 1. White chorioretinal lesions in animal treated with FIGURE 2. White chorioretinal lesions in animal treated with clindamycin. atovaquone.

RESULTS group had a markedly lower mean cyst count, which and was Drug Administration compared by Mest with the count in each of the other groups. There was no significant difference in the number of There was no significant difference in the volume of water cerebral Toxoplasma cysts between groups 1 through 4 drunk by each hamster in any of the different groups when (ANOVA test for equality of means, P — 0.48; Levene test for measured on a weekly basis. The average daily water consump- 14 equality of variance, P = 0.76). The atovaquone-treated group tion was 8 ml (range, 5-10 ml), as expected. had a significantly lower mean number of cerebral cysts than Clinical Course the control group or either of the other three treatment groups (Fig. 3). The ocular signs were in keeping with the course of the Chronic Disease. The mean number of cysts per sample 8 disease, as described previously. The first signs of ocular was 0.87 in the atovaquone-treated group (17.4 per whole disease were seen at 10 or 14 days after inoculation. Multiple brain) and 15.9 in the control group (318 per whole brain). small, white, well-demarcated lesions were noted, mainly at The difference was significant (P < 0.01). The nonparametric the posterior pole, without significant vitreitis. By 14 days after Mann-Whitney and Kruskal-Wallis significance tests con- inoculation, Toxoplasma lesions were observed in both eyes of firmed the results obtained from the f-test and ANOVA. The all hamsters. These lesions progressively enlarged and lost their highest significance wasP = 0.002 (Marin-Whitney test) and in sharp borders. At 6 weeks, the lesions reached their peak of the comparison of groups 1 through 4, P = 0.53 (Kruskal- inflammatory activity; thereafter, signs of regression were Wallis test). noted. At 12 weeks, all lesions were determined to be com- pletely inactive by clinical examination in all animals. Retinal Cyst Count There was no difference in the clinical course of active There were two cysts in 48 sections from treated eyes com- Toxoplasma retinal lesions between treatment groups or when pared with five cysts in 48 sections from control eyes. These compared with the course in control animals, when observed numbers are too small for meaningful statistical analysis. with indirect ophthalmoscopy or when assessed photographi- cally (Figs. 1, 2). No hamster exhibited signs of systemic Tox- oplasma infection during the course of the experiment. One DISCUSSION hamster in the control group died 6 weeks after inoculation. The cause of death was undetermined. The results of this study show that none of the drugs used altered the clinical course of acute, acquired Toxoplasma ret- Cerebral Cyst Counts inochoroiditis in the hamster animal model. The finding of no Acute Disease. The mean number of cerebral cysts per significant difference between groups 1 through 4 does not whole brain of each animal in the five groups is compared in necessarily indicate that there was no treatment effect. The Table 1, Groups 1 to 4 had similar mean cyst counts. To avoid study had little chance of detecting small or moderate differ- a large number of f-tests, these groups were compared using a ences as significant. In comparing any two means, the sample single significance test, one-way ANOVA. The atovaquone size provided a power of approximately 0.8 (80% chance) to

Downloaded from iovs.arvojournals.org on 09/28/2021 1174 Gormley et al. IOVS, June 1998, Vol. 39, No. 7 Mean number of toxoplasma cysts per animal

)

[ 1 ence 10- t o [ p o If) O) T3 CD C CO [ 1 0 10 9 10 10 10 Pyr.+Sulpha Control Atovaquone Spiramycin Clindamycin

Treatment Group FIGURE 3- Mean number and 95% confidence limits of Toxoplasma cerebral cysts per animal in the treatment groups for acute disease. Pyr, pyrimethamine; Sulfa, sulfadiazine.

detect a significant difference (P < 0.05), only if the true shown that 90% of total body cyst load is contained in the (expected) difference between the mean cyst counts was sub- cerebral tissue.l7 Toxoplasma cysts are only rarely noted in the stantial—that is, 9 or higher. ocular tissue,818"20 perhaps because of the smaller mass of Atovaquone was the only drug to affect the number of neural tissue in the eye.17 This was confirmed in our model in cerebral Toxoplasma cysts. It caused a significant reduction which only five cysts were counted in 48 sections from eyes (90%) in acute disease (Fig. 3)- Atovaquone also caused a with untreated Toxoplasma retinochoroiditis. Although atova- significant reduction in cerebral cyst count when used in quone caused a reduction in the number of cysts found in the chronic disease. In acute disease, the treatment was started at retinal tissue when compared with the number in control the first signs of retinal lesions (10-14 days after inoculation). animals, the numbers involved are very small. Much larger Because tissue cysts have been described as early as 7 days after 1516 groups of animals would be required for meaningful statistical inoculation in the animal model, it is uncertain whether analysis. For this reason, we elected to count the number of the drug was effective against the tachyzoites or the tissue cysts in the cerebral tissue. However, it is unknown whether cysts. The study of the effect of atovaquone in chronic, inactive there is a correlation between the number of Toxoplasma disease attempted to answer this question. In chronic infec- cysts in the brain and the number of cysts in the eye. Additional tion, the Toxoplasma organisms are expected to be in the study is required to resolve this question. latent, cystic form, and the effect of drug therapy on this stage of the life cycle may be determined. In this situation, atova- It is likely that reactivation of ocular toxoplasmosis is quone caused a significant reduction in the number of Toxo- caused by several factors. Rupture of the tissue cysts with plasma cysts in the cerebral tissue. Previous studies have infection of retinal cells by liberated organisms is believed to play a major role in the pathogenesis of recurrent disease,21'22 but immunologic factors may also be involved in the break- down of the cysts and the reactivation of the parasites.18'23 TABLE 1. Toxoplasma Cysts Regardless of the relative contribution by each of these mech- Treatment Number of Number of anisms, reduction in the number of cysts in the retinal tissue Group Animals Cysts should reduce the risk of disease recurrence. It is possible that atovaquone is effective against the Pyr. + Sulpha. 10 8.4 ± 1.93 tachyzoite and the encysted form of the Toxoplasma parasite. Control 9 12.89 ± 2.91 By destroying free tachyzoites, fewer viable organisms are left Clindamycin 10 11.3 ± 2.2 for subsequent cyst formation. Thus, atovaquone may become 10 Atovaquone 1.5 ± .52 the drug of choice for the treatment of acute episodes. New Spiramycin 10 8.9 ± 2 regimens with prolonged or repeated courses of treatment may Values are means ± SEM. be suggested, to achieve reduction of the cyst load and thereby

Downloaded from iovs.arvojournals.org on 09/28/2021 IOVS, June 1998, Vol. 39, No. 7 Effects of Drug Therapy on Toxoplasma Cysts 1175

reduce the risk of recurrent disease in the future— especially 9. Hudson AT, Dickens M, Ginger CD, et al. 566C80: a potent broad important in the immunosuppressed patient. Further studies of spectrum anti-infective agent with activity against malaria and opportunistic infections in AIDS patients. Drugs Exp Clin Res the efficacy this drug are merited. 1991:17;427-435. 10. Araujo FG, Gutteridge WE, Remington JS. Remarkable activity of 566C80 against Toxoplasma gondii (abstract). Clin Res. 1990;38: CONCLUSIONS 779A. 11. Araujo FG, Huskinson J, Remington JS. Remarkable in vitro and in None of the drugs studied altered the clinical course of the vivo activities of 566C80 against tachyzoites and tissue cysts of ocular lesions. Atovaquone was the only drug to affect the fluorescein isothiocyanate Toxoplasma gondii. Antimicrob number of cerebral Toxoplasma cysts in acute disease. Atova- Agents Chemother. 1991;35:293-299. quone also caused a significant reduction of cerebral Toxo- 12. Hudson AT, Dickens M, Ginger CD, et al. 566C80: a potent broad plasma cysts in chronic disease. Tissue cysts are believed to be spectrum anti-infective agent with activity against malaria and responsible for disease reactivation, and this drug may reduce opportunistic infections in AIDS patients. Drugs Exp Clin Res. 191;17:427-435. the risk of disease recurrence. 13- Fry M, Pudney M. Site of action of the antimalarial hydroxynaphthoquinone 2-[trans 4(4 chlorophenyl) cyclohexyl] 3-hydroxy- Acknowledgments 1-4-aphoquinone (566C80). Biochem Pharmacol. 1992;43: 1545-1553. The authors thank V. Andrews for the preparation of the drugs, S. 14. Hornet TD, Watson GL, Bantin GC. Feeding and watering. In: Kelly Munford for animal photography, and W. Rudling for care of the PJ, Millican KG, Organ PJ, eds. Principles of Animal Technology. experimental animals. Vol. 1. London: Institute of Animal Technology; 1988. 15. Lainson R. Observations on the development and nature of References pseudocysts and cysts of Toxoplasma gondii. Trans R Soc Trop 1. O'Connor Gil. Experimental models of toxoplasmosis. In: Tabbara MedHyg. 1958;52:396-407. KF, Cello RM, eds. Animal Models of Ocular Diseases. Springfield, 16. Dubey JP, Frenkel JK. Cyst-induced toxoplasmosis in cats. / Pro- MA: William Thompson; 1982:97-104. tozool. 1972;19:155-177. 2. Tabbara KF. Ocular toxoplasmosis. In: Tabbara KF, Hyndiuk RA, 17. Dubey JP, Beattie CP. Toxoplasmosis of Animals and Man. Boca eds. Infections of the Eye. Boston: Little, Brown; 1986:635-652. Raton, FL: CRC Press. 1988:1-40. 3. Henderly DE, Genstler AJ, Smith RE, Rao NA. Changing patterns of 18. Frenkel JK. Ocular lesions in hamsters with chronic Toxoplasma uveitis. Am J Ophthalmol. 1987;103:131-136. and Besnoitia infection. Am J Ophthalmol. 1955;39:203-225. 4. Newman PE, Ghosheh R, Tabbara KF, O Connor GR, Stern W. The 19. Davidson MG, Lappin MR, English RV, Tompkins MB. A feline role of hypersensitivity reactions to Toxoplasma antigens in ex- model of ocular toxoplasmosis. Invest Ophthalmol Vis Sci. 1993; perimental ocular toxoplasmosis in non human primates. Am J 34:3653-3660. Ophthalmol. 1982;94:159-l64. 20. Brezin AP, Gazzinell R, Ci Q, Nussenblatt RB, Chow CC. Experi- 5. Engstrom RE, Holland GN, Nussenblatt RB, Jabs DA. Current prac- mental reactivation of chronic toxoplasmosis: stage specific in situ tices in the management of ocular toxoplasmosis. Am j Ophthal- hybridization, DNA amplification and immunohistochemistry mol. 1991 ;111:601- 610. [ARVO Abstracts]. Invest Ophthalmol Vis Sci. 1992;33:1320. 6. Rothova A, Meenken C, Buitenhuis HJ, et al. Therapy for ocular 21. Frenkel JK. Pathogenesis of toxoplasmosis with a consideration of toxoplasmosis. Am J Ophthalmol. 1993;115:517-523. cysts rupture in Besnoitia infection. Surv Ophthalmol. 1961 ;6: 7. Velimirovic B. Toxoplasmosis in immunosuppression and AIDS. 799-825. Infection. 1984;12:315-317. 22. Wilder HC. Toxoplasma chorioretinitis in adults. Arch Ophthal- 8. Pavesio CEN, Chiappino ML, Gormley P, Setzer PY, Nichols BA. mol. 1952;48:127-136. Acquired retinochoroiditis in hamsters inoculated with ME 49 23. Nussenblatt RB, Gery I, Ballintine EJ, Wacker WB. Cellular immune strain Toxoplasma. Invest Ophthalmol Vis Sci. 1995;36:2l66- responsiveness of uveitis patients to retinal S antigen. Am/ Oph- 2175. thalmol. 1980;89:173-179.

Downloaded from iovs.arvojournals.org on 09/28/2021