sign in sign up
<<
Home , Acanthamoeba, Acanthamoeba keratitis, Chlorhexidine, Keratitis, Propamidine

DRUG RESISTANCE AND : THE QUEST FOR ALTERNATIVE CHEMOTHERAPY

JOHN HAyl, COLIN M. KIRKNESSI, DAVID V. SEAL! and PETER WRIGHT" Glasgow and London

SUMMARY trate, dendritiform patterns and localised oedema, especially in a young person, as well as an association with Trophozoites and cysts of 20 isolates of Acanthamoeba from the and fivefrom related samples were tested previous contact wear, should suggest the possibility in vitro for sensitivity to ten drugs (three aromatic dia­ of Acanthamoeha keratitis. This can be confirmed by iso­ midines, two , two macrolides, a polyene lation and cultivation of the protozoan from corneal macrolide , an organoarsenical and an anti­ scrapes or biopsy material.7 Once a definitive diagnosis metabolite) and two cationic ( has been achieved, appropriate anti-acanthamoebal drug and polyhexamethylene , PHMB). Only chlor­ therapy will invariably be required. hexidine and PHMB showed uniform amoebacidal activ­ The literature attests to a wide variety of drugs provid­ ity. Aromatic diamidines (pentamidine isethionate, ing variable efficacy against different Acanthamoeha isethionate and diminazene aceturate) species or strains, both in riva and in vitro. In clinical prac­ generally proved effective against both forms of the tice, however, it is important to note that some of the ; only pentamidine gave synergy with the bigua­ agents, for example some of the aromatic diamidines, may nide while propamidine gave an additive effect. Other merely inhibit replication or induce encystment of the drugs tested proved erratic or ineffective against differ­ tropozoite form,x rendering it quiescent as a cyst, and thus ent isolates. Chlorhexidine alone, or together with pro­ often resistant to conventional drug therapy. In such cir­ pamidine, was subsequently used in two patients with cumstances the cyst retains the potential to exacerbate the proven keratitis; the causative isolates Acanthamoeba disease on discontinuation of the drugs or, if infected tis­ were sensitive to the individual compounds and to the sue is retained, following .9 combination The treatment provided resolution in vitro. In the United , empirical combination therapy of the clinical disease; amoebae were shown to be non­ of propamidine, dibromopropamidine and has viable by histology and culture. The combination of proved efficacious in some patients. \0 Drug resistance, II as chlorhexidine and propamidine is recommended for well as allergic or toxic reactions after prolonged therapy treatment of proven . with propamidine,12 has limited the use of this combi­ Keratitis associated with Acanthamoeha is a nation, and has prevented its widespread acceptance in relatively rare, sight-threatening condition occurring most clinical practice. often in wearers,I where there has been Novel approaches to chemotherapy of Acanthamoeha inappropriate or inadequate disinfection of contact lens keratitis continue to be forthcoming. For example, a for­ systems." The clinical presentation of the disease is often mulation containing the cationic polyhexameth­ mistakenly diagnosed as herpes simplex or fungal infec­ ylene biguanide at fairly low concentration, alone or in tion.3 This results in inappropriate anti-microbial agents combination with propamidine and/or neomycin, has being administered. Early features4-6 such as pain, photo­ proved very effective against both trophozoites and cysts phobia and recurrent epithelial breakdown with little infil- of Acanthamoeha derived from proven clinical cases of 1 1 From: 'Department of Bacteriology. The Royal Infirmary. Glasgow: the infection. 3. 4 Drug therapy, however, may be com­ 'Department of . Tennent Institute. University of plicated by a number of factors, mostly associated with Glasgow. Glasgow. UK. failure to attend to the natural history and metabolism of Correspondence to: Professor C. M. Kirkness. FRCS (Edinburgh and Glasgow). FRCOphth. Department of Ophthalmology. Tennent the protozoan within the diseased cornea and also to the Institute. University of Glasgow GIl 6NT. UK. pharmacology of the selected drug(s) within this location.

Eye (1994) 8, 555-563 © 1994 Royal College of Ophthalmologists 556 J. HAY ET AL.

The purpose of this study was to suggest a more rational After several transfers, the amoebae were incubated in this approach to chemotherapy of Acanthamoeha keratitis medium for approximately 72 hours at either 25°C or based on in vitro drug sensitivity studies on cultures of 32 0c. Acanthamoeha isolated from patients and contact lens In order to determine the purity of Acanthameoha cul­ accoutrements (details of which have been published pre­ tures, Giemsa staining was performed. Viability of each viously). Representatives of selected drug classes and Acanthamoeha culture was assessed using a 0.2% trypan cationic antiseptics were used singly or, where considered blue. Cyst populations were obtained by incubating tro­ appropriate, in combination with each other, in order to phozoite cultures for about 7 days at 25°C or 32 0c. Puri­ determine whether a potent acanthamoebacidal action ty and viability were determined retrospectively. as could be identified ill vitro, where the mechanism of described above, on excysted cohorts of amoebae. Only action could be established, and the effect subsequently cultures with >98% purity and viability were used for exploited in vivo. Illustrative case reports are used to high­ drug sensitivity studies. light some important aspects. Drugs MATERIALS AND METHODS Aqueous solutions (100 Ilg/mi) of drug or Acanthamoeba and Cultivation were prepared immediately prior to use and filter-steril­ Eighteen Acanthamoeha corneal isolates, including three ised using Gelman filters with 0.22 11m pore size. Com­ from the same patient (TB) taken at different time inter­ pounds used for assessment of amoebacidal action are vals, were initially investigated. From one of these (AT), shown in Table L These agents were selected mainly as a isolates from the soft contact lens and its storage case were consequence of literature reports of their efficacy, or that also included. From another (MT), an isolate from the of related drugs or antiseptics, against Acanthamoeha, contact lens storage case and one from the water supply at either in vitro or in vim. work (the home was negative) were included. On completion of testing these isolates, another patient Drug and Antiseptic Screening (AB) presented with Acanthan10eha keratitis. Isolates Drug and antiseptic screening was performed using a from a corneal biopsy and scrape as well as a soft contact series of sterile 96-well microtitre plates containing a stan­ lens were examined. This afforded the opportunity of dardised concentration of 2 x 10" organisms per 100 IIIof assessing the in vitro findings from the 18 samples in a medium per well. One hundred microlitres of doubling clinical setting. dilutions of each compound (100-0.8 Ilg/ml) were pro­ All amoebae were maintained by routine passage on to duced vertically for each of the 12 compounds tested. Lids 1.5% high clarity bacteriological agar No. I (LAB M) were secured, then the contents of plates mixed gently for made up in amoebal saline.IS It was spread with heat-killed 10 minutes on a plate rotator prior to incubation at either Klehsiella aerogenes and moistened intermittently with 25°C or 32 0c. amoebal saline prior to incubation in air at either 25°C, Sensitivity of isolates was assessed after 48 hours of 32°C or 35°C. incubation, by recording either the lowest concentration of In order to obtain sufficient numbers of each of the 25 drug or antiseptic which resulted in complete lysis or isolates for in vitro drug screening, the surface of each degeneration of trophozoites and non-viability of result­ plate was flooded with amoebal saline and agitated in ing cysts (minimum trophozoite amoebacidal concentra­ order to permit transfer of amoebae to sterile plastic tion, MTAC) or, for cysts, the lowest concentration of test 75 cm" tissue-culture flasks (Sterlin, CelCult) containing compound that resulted in no excystment and trophozoite approximately 50-100 cm3 of a definedgrowth medium.16 replication (minimum cysticidal concentration, MCC),13

Table I. Drugs and antiseptics used in this study, their classification and proposed antimicrobial mechanism of action

Agent Class Inhihitor of

Chlorhexidine digluconate' (chlor)* Cationic antiseptic Membrane function Polyhexamethylene biguanide'" (phmb) Cationic antiseptic Membrane function Propamidine isethionate 1 (propam) Aromatic diamidine DNA synthesis Pentamidine isethionate" (penta) Aromatic diamidine DNA synthesis Diminazine aceturateb.5 (dim) Aromatic diamidine DNA synthesis Neomycin sulphate" (nco) antibiotic Protein synthesis sulphate7 (paro) Aminoglycoside antibiotic Protein synthesis ' Polyene macrolide antibiotic Membrane (ergosterol) biosynthesis Dirithromycin'" (dir) Macrolide antibiotic Protein synthesis Spiramycinu.1o (spir) Macrolide antibiotic Protein synthesis Cymelarsanb.1I (cymel) Organoarsenical Energy metabolism a-Diftuoromethylomithinebl2 (dfmo) Antimetabolite Substrate- reaction (inhibitor of ornithine decarboxylase)

GijiFam: 'Moorfields Eye Hospital; "Prof. F. W. Jennings; 'Lilly Research Centre Ltd; "Rhone-Poulenc Ltd. Supplied as solulions: '0.050/,. Sterets, Unisept; 20.02'k contains hypromellose eye drops. 0.3%: 10.1 'k Brolene. May & Baker (contains ); "Minims, Smith & Nephew (contains ). Supplied as solids: "Pentacarinat, May & Baker; 7Sigma: 'Fungizone ilv. Squibb: '.: 12 donated compounds. * Abbreviations in parentheses are used in Figs. I and 2 and Table II. DRUG RESISTANCE AND AMOEBIC KERATITIS 557 after thoroughly washing cysts free of residual drug, and was administered for 9 months without a further adverse re-incubation in the medium described above.16 Obser­ reaction. With this course of treatment relapse of Acantha­ vations were performed in duplicate using an inverted moeba infection was not apparent. Isolates from a corneal microscope. biopsy, contact lens and its storage case grew conftuently For in vitro combination testing, a chequerboard at 35°C. MT had worn Acuvue disposable soft contact method was used. 17 With this procedure, four possible out­ lenses (FDA Group 4) and used Softab for contact lens comes of drug-drug or drug-antiseptic combinations disinfection.20 were possible: IS Patient AB I. Additivity, where the result with the two compounds A 48-year-old man with mild atopy and early was equivalent to their sum when used separately. had recently changed contact lenses from daily-wear soft 2. Autonomy (or indifference), where the result with the to rigid, gas-permeable lenses. Total (Allergan), an all-in­ two compounds was not different from the result with one sterilising solution containing , the more effective compound used alone. was used for lens hygiene. The patient attended an eye 3. Antagonism, where the result with the two compounds casualty department with a dendritiform , was less than the additive response. stromal oedema and a mild anterior chamber reaction. 4. Synergism, where the result with the two compounds This was considered to be herpetic kerato- and was greater than the additive response. treatment was commenced with acyclovir ointment and 0.5% drops. He was then referred to one of 9 Five replicates per determination were performedl using us (C.M.K.). Intensive antiviral and therapy did not the combinations shown in Table II; findings were based bring about improvement. The eye became increasingly on results obtained from 7 of the 18 corneal isolates. painful, injected, and a ring abscess developed. A corneal biopsy was performed 2 months after initial presentation. CASE REPORTS Acanthamoeha was observed histologically in the corneal Patient TB stroma and cultured from it. The protozoan was seen on Drug resistance occurred after treatment of Acantha­ the hydrogel contact lens and cultured from washings. moeba keratitis with propamidine alone; topical neomycin Treatment was commenced with 0.02% chlorhexidine had been withdrawn because of . I I The isolate col­ in isotonic saline drops every hour for a week then lected in early treatment was retested for drug and anti­ 2-hourly, combined with 0.1% propamidine at the same septic sensitivities, as were two later isolates. Both the frequency, and 3-hourly prednisolone I %. The eye latter have been recorded before as temperature-sensitive became white and comfortable within 7 days though the and propamidine-resistant. I I patient remained photophobic. The abscess began regress­ ing and the epithelium slowly healed. At 2 months there Patient AT was a central stromal opacity and 3 mm overlying epi­ Treatment was given with topical -neo­ thelial defect. mycin combination q.i.d. for 4 weeks, followed by pro­ The patient next presented as an emergency with a pamidine q.i.d. alone for 3 weeks, prior to laboratory 2-day history of sudden discomfort and further loss of confirmation of Acanthamoeba keratitis by cornealscrape vision. The cornea was found to be perforated centrally and biopsy. The isolated protozoan grew poorly at 35°C, with an intumescent lens. An emergency keratoplasty, compared with those isolated from the contact lens and extracapsular lens extraction and posterior chamber lens storage case. AT had worn Acuvue (Johnson & Johnson) Table II. In vitro combination testing of selected drugs against seven disposable soft contact lenses (FDA Group 4) and had Acanthamoeha corneal isolates used Softab (Alcon; a chlorine-based system) following Combination Effect on MTAC Effect on MCC the manufacturer's instructions for contact lens disinfection.20 Iso/ates(TBI2;1.2.IO.13;AB) AT responded satisfactorily over a 12 month period to a phmb + pentamidine Synergy (slight) Synergy (slight) phmb + neomycin Additivity Additivity combination of propamidine and neomycin, before under­ phmb + dirithromycin Autonomy Autonomy going a penetrating keratoplasty due to scarring in the propamidine + neomycin Additivity Additivity visual axis. Histology revealed only a few degenerate pentamidine + neomycin Additivity Additivity *diminazine + neomycin Additivity Additivity cysts. There was no recurrence of infection after I year. *diminiazine + dirithromycin Autonomy Autonomy *pentamidine + dfmo Autonomy Autonomy Patient MT *cymelarsan + dfmo Autonomy Autonomy *neomycin + dirithromycin Autonomy Antagonism Treatment of this patient's Acanthamoeba keratitis was initially with a topical propamidine-neomycin combi­ Iso/ate AB chlorhexidine + propamidine Additivity Additivity nation, which induced a toxic reaction in the cornea within chlorhexidine + pentamidine Additivity Synergy (slight) 2 months. At this stage trophozoites were still present in a chlorhexidine + neomycin Additivity Additivity corneal biopsy. MTAC, minimum trophozoite amoebacidal concentration; MCC, mini­ Guttae chlorhexidine (0.02% w/v in 0.9% saline) alone mum cysticidal concentration. *Not including AB. 558 J. HAY ET AL.

concentration minimum "I'stlcldal concentration minimum trophozoite arnoebac1dal (f'!IIml) (I'g/ml) 200 200

150 o [] 150 Jan 'IfFeb 'IfJuly

100 100

50 50

penta ReO paro dlr dim neo paro lr aplr cyme cIllDf phmb ptopam dimdrug d l drug trophozoite. cysts 1a :: concentration nlmum cysticldal (Jlg/ml) 2Q:;lnlmum 1rophozotte amoebac1dal ()'9'ml) 2 concentration

case comea [] D lena 150 150

100

50

ptopam ReO paro dim neo paro dlr aplr cIlIDf phmb penta dimdrug dlr chlDf phmb propam penta drug cymel trophozoite. cysts * 1b temperature-sensitive Isolate

Fig. 1. (a) Minimum tropho::oiteamoebacidal concentration (MTAC) and minimum cysticidal concentration (MCC) often agents for three cornealisolates from patient TB. (b) MTAC and MCC often agents for corneal.storage case and contact lells isolates/i'ompatient AT. For abbreviations see Table I. implant was performed. Half the excised cornea was sent cymelarsan and a-diftuoromethylornithine (a-DFMO) for culture and half for histopathology. Streptococcus were ineffective against trophozoites and cysts for all acidominimus (weakly a-haemolytic) was cultured from three isolates. This patient, in whom medical treatment the corneal tissue but it remained persistently culture­ failed with propamidine and arsenicals,II would probably negative for Acanthamoeha. Degenerate cysts were seen have benefited from therapy with cationic antiseptics. in the corneal stroma but no trophozoites. The corneal epi­ thelium appeared healthy. Patient AT (Fig. IB) Post-operatively treatment was continued with Brolene Trophozoites and cysts of all three isolates (cornea, con­ (May & Baker) and chlorhexidine with prednisolone (l % tact lens and storage case) were highly sensitive to chior­ 2-hourly) and Polytrim (Burroughs Wellcome) drops hexidine and PHMB. Trophozoites and cysts from the 2-hourly. Three months after keratoplasty the graft corneal isolate, but not the others, were less sensitive to remained clear with no signs of further infection. The cor­ propamidine and were temperature-sensitive (confluent rected vision was 6/12. growth at 25°C, poor at 35 DC); this trend was also evi­ dent for the other two diamidines. Neomycin was more RESULTS effective than paromomycin but the corneal isolate was Patient TB (Fig. lA) more resistant than the others as regards both trophozoites Trophozoites and cysts of all three corneal isolates were and cysts. Amphotericin B, cymelarsan and a-DFMO highly sensitive to chlorhexidine and polyhexamethylene were ineffective. biguanide (PHMB). While trophozoites and cysts from 2A) the firstisolate were sensitive to propamidine, the two sub­ Patient MT (Fig. sequent isolates showed resistance as previously tested, a Trophozoites and cysts of all three isolates (cornea, stor­ phenomenon apparently related to temperature I I but not to age case, workplace water sample) were highly sensitive the action of the antiseptics. Both forms of the amoeba for to chlorhexidine and PHMB and grew well at 32°C. Pro­ all isolates were sensitive to neomycin. Amphotericin B, pamidine was more effective than either pentamidine or DRUG RESISTANCE AND AMOEBIC KERATITIS 559

minimum trophozoiteamoebacldal concentration (ug/mll minimum cystcldal concentration (ug/ml) 200 200

150 150 o cornea IT] case D waler source

100 100

so 50

O ��-d�-C�-L�-L����LU������L- chlor phmb propam penta dim neD paro dlr spiro cymel chlor phmb propam penla dim neo para dlr spiro cymel drug drug

trophozoite. cysls 2a

minimum trophozoldalamoebacldal concentration (ug/mll minimum cystlcidal concentration (ug/mll 200 200

150 150 o blopa aera� lena

NO: •• nIIU..,Uy not•• t t.d

100 100

50 50

NO NO NO NO NO NO NO NO NO

chlor phmbpropam penta dim neo para dlr apir cyme! chlor phmb propam penis dim neD paro dlr aplr cymel drug 2b drug

Fig. 2. (a) MTAC and MCC often agentsjiJr corneal, contact lens storage case and water-source isoiatesJi-ompatient MT. (h) MTAC and MCC (�ften agents for corneal hiopsy, corneal scrape alld contact lens washing isolates Ji-om patient AB. For ahhreviations see Tahle 1. diminazine against trophozoites of all three isolates, but described above, chlorhexidine and PHMB were most pentamidine was most effective against cysts.For all three efficacious.All isolates with the exception of nos. 2 and 10 isolates, the storage case isolate was less sensitive to dia­ for trophozoites and nos. 1 and 13 for cysts were sensitive midines than either the corneal or water sample isolates. to propamidine; there were no obvious trends with the Trophozoites of all three isolates were sensitive to neo­ other two diamidines. Aminoglycosides were relatively mycin but cysts were resistant. Amphotericin B, cyme­ ineffective against cysts while the trophozoites of four iso­ larsen and a-DFMO were ineffective. lates (nos. 1, 4, 7, 10) were insensitive also. Trophozoites and cysts from all isolates were insensitive to macrolides, Patient AB (Fig. 2B) amphotericin B, cymelarsan and a-DFMO. Fig. 3H shows the average values for MTAC and MCC Trophozoites and cysts of all three isolates (corneal for 10 of the 12 drugs tested. Chlorhexidine and PHMB biopsy, scrape and contact lens washings) were sensitive at were the most active compounds against both trophozoites the upper limit to chlorhexidine; the biopsy sample only and cysts. The diamidines were the next drug class in order was tested against PHMB and gave similar values. All of acanthamoebacidal activity. Neither the aminoglyco­ three isolates were sensitive to diamidines but again at the sides, macrolides nor the arsenical were effective against upper limit (MTAC 6.3-25 flg/ml, MCC 6.3-25 flg/ml). cysts and showed an increasing inability to destroy Trophozoites and cysts of the three isolates were resistant trophozoites. to acyclovir - which has been tested since it has been sug­ Table II gives findings obtained from in vitro combi­ gested that other antiviral agents maybe ineffective nation of selected drugs and PHMB against six corneal 9 against Acanthamoeha. isolates (i.e. TB 1, 2; 1, 2, 10, 13) and selected drugs and chlorhexidine against isolate AB. The only combinations Fig. 3A-G illustrates the findings for trophozoites and that gave a slight synergistic response were the cationic cysts obtained for the remaining 13 corneal isolates. As antiseptics (chlorhexidine or PHMB) and pentamidine; 560 J. HAY ET AL.

chIoftlealdlne poI,heumelh,lene bl9ulnide minimum lmoeblcldalconcentrilion (ug/mll minimum 1f1>OebIcId., concenlt_ (ug/ml) 100 100

1\ B 75 o 75 D

50 50

25

I_number laoa.te number ptopatnldlne pen'-mldlne

minimum _blcldoJ concentrollon (ug/mll minimum lmoeblcldal concentrollon (ug/mll 100 100

75 () 75 c

50 50

25

lao&.le number lao&.ate numbel

dlmtn.aDne ac;:eturat. �ycln mlnlm .....�-*' (ug/mll mln_ �bIddaI ooncon1ra.1on (ug/m') 100 100

-

50 50

25 25

o n r r r r r r IlOIal.numbw -ydn m ___ Iug/ml) 100 A_ IITAC &IICC r... III comuI... ,..... (ug/ml) 100

75 80 G II

60 50

40 -

25 20

o r- I I r- ·Ii r r i r

Fig. 3. MTAC and MCC for individual drugs: A. chlorhexidine; B. PHMB; C. propamidine; D. pentamidine; E. dimina:ene. F. neomycin; G. paramomycin; H. average MTAC and MCC for all corneal isolates. DRUG RESISTANCE AND AMOEBIC KERATITIS 561 additivity was found between chlorhexidine and propam­ and cyst drug-sensitivity studies which may cause con­ idine for the AB isolate, a successful combination in prac­ fusion with drugs that merely induce encystment and are tice. All three diamidines showed additivity with not acanthamoebacidal in action. neomycin as did the combination of PHMB and neo­ Several compounds have been used with varying effect mycin. Other combinations showed autonomy, except that in the clinical setting. These include: itraconazole plus of neomycin and dirithromycin, where there was ;36 ;'7 ;38 dibromo­ antagonism. propamidine plus propamidine and neomycin; 10 propam­ idine isethionate as Brolene;39 propamidine in DISCUSSION combination with neomycin--gramidicin as Failure of drug therapy in Acanthamoeha keratitis has Neosporin (Calmic );40 Neosporin with or without micona­ been recognised for some time but the reasons are not zole or ketoconazole;41 pimaricin plus Neodecadron (dex­ always understood in the absence of drug sensitivity test­ amethasone phosphate, neomycin sulphate) ing. This study suggests that some of the commonly used hydroxyuracil, and ; 2B PHMB solution drugs such as neomycin and paromomycin are not particu­ (which contained 0.3% hypromellose, 0.45% NaCl, larly effective amoebacides - a finding in accordance 0.37% KCl, 0.19% borax and 0.19% ) alone or with that of other workers.1.1 Moreover the case reports in combination with propamidine,1.1 or PHMB in a solu­ (TB, AT) suggest that resistance may develop as a result of tion of artificial tears combined with propamidine and low-dose, single-drug anti-amoebic therapy. Sensitivity neomycin. 14 testing can therefore be useful on all occasions. It is not As with ill I'itro testing, clinical reports suggest that sufficient to rely on isolation of the organism from the drug selection has been relatively haphazard. As a result, it contact lens or its storage case since amoebae from these was decided to investigate a cohort of 18 corneal isolates sources may have different sensitivities from that isolated from cases of keratitis, three from contact-lens-associated from the cornea, particularly when the diagnosis has been materials and one from a water supply used to clean the late and there has been pretreatment with a variety of storage case. Most of the drugs and antiseptics used in the drugs. Since an animal model is not yet available for the present study have been previously assessed for potential sensitivity testing of Acanthamoeha isolates, in I'itro anti-acanthamoebal activity. Two macrolides, however, assessment is a necessary but relatively unsatisfactory were included (dirithromycin and spiramycin), since alternative. erythromycin is known to induce encystment of Acantha­ 42 Agents tested in I'itro and found to have an effect on dif­ moeha. The organoarsenical cymelarsen was included as ferent species and strains of Acanthamoeha include: clo­ a novel representative of this group with less inherent tox­ trimazole21 and ketoconazole,2:' although like other azoles icity than earlier arsenical compounds which, in general, I . the effectis likely to be amoebastatic rather than amoeba­ exhibit relatively poor activity against Acanthamoeha. 1 25 cidal,23 and with some drugs in this group the organisms Chlorhexidine, a cationic antiseptic, was selected for may be highly resistant;24 5-ftuorocytosine (a nucleotide inclusion since it has been shown to have good anti­ 43 analogue),2'i.26 although this drug has been found inef­ acanthamoebal activity il1 vitro and had previously been fective by other workers;27.2X the diamidines pentamidine suggested anecodotally for therapy by one of us (D.Y.S.). 2 isethionate, 10. 9 although others have reported insensitivity PHMB, a related compound but as yet unlicenced for top­ to this drug,22 hydroxystilbamidine isethionate,25.3o dimin­ ical use in humans, has been shown to have considerable . izene aceturatelO and propamidine isethionate, 10 29 activity against Acanthamoeha, both in vitro and in vivo. although again this drug has been identified as insensitive The cationic antiseptics showed outstanding efficacy in other studies, except if combined with dimethyl sul­ against all isolates, with chlorhexidine giving the lowest phoxide (DMSO);31 membrane-active peptides, the MTAC and MCC (Fig. 3H). The aromatic diamidines as a magainins,32 the effect being enhanced when in combi­ group were second in order of efficacy, although in keep­ nation with nitrate or propamidine; pimaricin;2x ing with previous literature reports the effect varied con­ amphotericin B or AB methyl ester.22 although found to be siderably between the isolates. Interestingly, diminazine ineffective in other studies; 27 certain inhibitors of folate (an encystment-enhancing agent)/ showed satisfactory biosynthesis; ,3 triftuoroperazine; 34 the aminoglycosides activity against some isolates, a finding in keeping with 1O paramomycin,24.30 which was ineffectivein other studies,25 that of Wright et al. but at odds with the results of other and neomycin,'ll also without effectin other studies except workers.30 The aminoglycosides again showed variability if combined with propamidine; " polymyxin E (colistin);22 amongst the strains, and had no effect against cysts, both acriftavine30 (although some workers have reported resist­ findings confirming previous literature reports.1.1 Macro­ ance,28) and other .3'i lides showed similar behaviour to aminoglycosides but There is considerable disparity regarding in vitro effi­ were less effective. A note of caution should be intro­ cacy of drugs which are active against Acanthamoeha; duced, however, since the macrolides used were prodrugs, none demonstrate uniform activity against all isolates, and the metabolites being more effective, at least against bac­ there is differential sensitivity between trophozoites and teria.44 The organoarsenical showed poor activity against cysts, the former being more sensitive than the latter. both forms of the protozoan. Others have reported similar . Furthermore, some reports are based on mixed trophozoite resistance to arsenicals. 11 25 In keeping with the findingsof 562 J. HAY ET AL.

Table III. In I'itro combination testing of drugs against A('(Jlltila­ in a way such as occurs in human neutrophilic granu­ mocha corneal biopsy isolate from patient AB locytes by inhibition of co-factors5" or cytoplasmic 5 Effect on mean , 3 or may on their own exert an inhibitory effect trophozoite on multiplication of Acanthamocha. Analogues of the dia­ amoebacidal Effect on mean midine series,4 may considerably enhance this effect. Combination concentration concentration The combination of chlorhexidine and propamidine phmb + pentamidine Synergy (slight) Synergy (slight) seems to have had effective amoebacidal action within the phmb + neomycin Additivity Additivity cornea. This could shorten the time during which anti­ phmb + propamidine Additivity Additivity propamidine + neomycin Additivity Additivity acanthamoebic drugs are required and their frequency of pentamidine + neomycin Additivity Additivity application. This, in turn, may reduce the likelihood of chlorhexidine + propamidine Additivity Additivity toxic reaction. Ie and obviate effects of inherent29 or chlorhexidine + pentamidine Additivity Synergy (slight) II chlorhexidine + neomycin Additivity Additivity acquired resistance to the diamidine.

We thank Drs D. Warhurst and De 10nckheere for preservation other workers,45 the inhibitor of ornithine decarboxylase, 1. of the 'London' strains of Acanthamocba and Mr V. Andrews, a-DFMO, had no discernible effect on the growth of Principal Pharmacist. Moorfields Eye Hospital, for supplying Acanthamoeha. There is, however, reason to believe that the formulation of PHMB. Mrs R. Connor rendered invaluable other components of polyamine metabolism in Acantha­ technical assistance. moeha may yet be found suitable as the basis for develop­ Key words: A('(JllthamociJa keratitis. Antiprotozoal chemotherapy. ment of more active chemotherapy against the Chlorhexidine. Diamidines. Drug resistance. protozoan.46.47 Single drug therapy of Acanthamocha keratitis with REFERENCES currently used compounds appears inadequate, and may I. Stehr-Green lK. Bailey TM. Visvesvara GS. The epidemi­ lead to emergence of drug resistance. Combination ther­ ology of Acanthamocba keratitis in the United States. Am 1 apy must be considered. Notable in this context ill vil'o is Ophthalmol 1989; 107:331-6. 2. Moore MB. Acanthamocba keratitis and contact lens wear: the combination of PHMB and propamidine, i1 or neo­ the patient is at fault. Cornea 1990;9 (Suppl J:S33-5. 1 I mycin and propamidine; 4 or neomycin, dibromopropami­ 3. Auran 10. Starr MB. lakobiec FA. Acanthamocba keratitis: III dine and propamidine; and, ill I'itro, DMSO and a review of the literature. Cornea 1987;6:2-26. propamidine isethionate.31 In the present ill I'itro study, 4. Lindquist TO. Sher NA. Doughman 01. Clinical signs and additive effects were observed with cationic antiseptics medical therapy of early Acanthamocba keratitis. Arch Ophthalmol 1988; 106: 1202-6. plus propamidine or neomycin, and slight synergy 5. Moore MB, M Culley lP. Kaufman HE, Robin lB. Radial between the antiseptics and pentamidine (Ta bles II, III). keratoneuritis as a presenting sign in Acanthamocba kera­ Following the demonstration of ill vitro efficacy of titis. Ophthalmology 1986;93:1310-5. chlorhexidine two patients have been treated with this 6. Bacon AS. Dart lKG. Ficker LA. Matheson MM, Wright P. drug. One patient (MT), who had developed an idiosyn­ Acanthamocba keratitis: the value of early diagnosis. Oph­ thalmology 1993: 100: 1238--43. cratic reaction to both neomycin and propamidine, was 7. Ficker LA. Kirkness e. Wright P. Prognosis for keratoplasty treated satisfactorily with monotherapy chlorhexidine. in Acamhamocba keratitis. Ophthalmology 1993;100: The other patient (AB) received combination therapy of 105-10. chlorhexidine with propamidine with rapid control of the 8. Byers n, Kim BG. King LE. Hugo ER. Molecular aspects Acanthamoeha infection. It is evident, however, that use­ of the cell cycle and encystment of Acanthamocba. Rev Infect Dis 1991; 13 (Suppl 5):S373-84. ful anti-acanthamoebic drugs may not have universal 9. 10nes DB. Visvesvara GS. Robinson NM. Acanthamocba activity against all amoebae. We believe that, in general, po/yp!Jaga keratitis and Acanthamocba uveitis associated combination therapy should always be employed. firstly with fatal meningoencephalitis. Trans Ophthalmol Soc UK because of the possibility of an additive anti-amoebal 1975:95:221-32. effect and secondly to prevent the emergence of resist­ 10. Wright P. Warhurst D. 10nes B1. Acallthamocba keratitis ance. On the basis of these two patients, plus one other successfully treated medically. Br 1 Ophthalmol 1985;69: 778-'1',2. now treated successfully for 3 months with a similar com­ II. Ficker L. Seal 0, Warhurst D. Wright P. Acanthamocba bination to AB, and on anecdotal evidence from several keratitis: resistance to medical therapy. Eye 1990;4:'1',35-8. personal communications, chlorhexidine seems to be well 12. 10hns Kl, Head WS, O'Day OM. Corneal toxicity of pro­ tolerated in the eye. pamidine. Arch Ophthalmol 1988;106:68-9. The findings from the present study are suggestive of 13. Larkin DFP, Kilvington S. Dart lKG. Treatment of Acantha­ mocba keratitis with polyhexamethylene biguanide. Oph­ membrane effects, which permit easier access of drug into thalmology 1992;99:185-91. the amoebae. Cationic antiseptics such as chlorhexidine4x 14. Varga lH. Wolf Te. lensen HG. Parmley Ve. Rowsey 11. and, to a lesser extent, neomycin perturb the plasma­ Combined treatment of Acallthamocha keratitis with pro­ lemma; this may facilitate the entry of an effective drug pamidine. neomycin and polyhexamethylene biguanide. such as an aromatic diamidine. Diamidines are either Am 1 Ophthalmol 1993;115:466-70. 15. Page Fe. A new key to freshwater and soil gymnamoebae. inhibitors of S-adenosylmethionine decarboxylase in Ambleside. Cumbria: Freshwater Biological Association, 49 Acanthamoeha, or drugs which interact directly with the 1988. nucleic acid of the organism.50.51 In addition they may act 16. Byers n. Akins RA. Maynard Bl. Lefken RA. Martin SM. DRUG RESISTANCE AND AMOEBIC KERATITIS 563

Rapid growth of Acanthamoeha in defined media: induction mura S. Yasuraoka K, Ishii K.Oral itraconazole and topical of encystment by glucose-acetate starvation. J Protozool miconazole with for Acanthamoeba keratitis. 1980;27:216-9. Am J OphthalmoI1990;109:121-6. 17. Krogstad DL Moellering RC.Antimicrobial combinations. 37. Driebe WT, Stem GA, Epstein Rl, Visvesvara GS, Adi M, In: Lorian V, editor. in laboratory medicine, 2nd Komadina T. Acanthamoeba keratitis. Arch Ophthalmol ed.Baltimore: Williams & Wilkins, 1986:537-95. 1988; 106: 1196-201. 18. King TC, Schlessinger D, Krogstad DJ.Th e assessment of 38. Cohen El, Parlato Cl, Arentsen 11, et al. Medical and sur­ drug combinations.Rev Infect Dis 1981 ;3:627-33. gical treatment of Acanthamoeba keratitis. Am 1 Ophthal­ 19.Rand KH, Houck HJ, Brown p, Bennett D.Reproducibility mol 1987;103:615-25. of the microdilution method for antibiotic synergy. Anti­ 39. Yeoh R, Warhurst DC, Falcon MG.Acal1thamoeha keratitis. microb Agents Chemother 1993;37:613-5. Br 1 Ophthalmol 1987;71 :500-3. 20. Seal DV, Hay 1, Devonshire P, Kirkness CM. Acantha­ 40. Moore MB, MCCulley lP. Acanthamocba keratitis associ­ mocha and contact lens disinfection: should chlorine be dis­ ated with contact lenses: six consecutive cases of successful continued? Br J Ophthalmol 1982;77:128. management. Br 1 Ophthalmol 1989;73:271-5. 21. Stevens AR. Willaert E. Drug sensitivity and resistance of 41. Sharma S, Srinivasan M, George C.Acanthamoeha keratitis Acallthamocha four species. Trans R Soc Trop Med Hyg in non-contact lens wearers. Arch Ophthalmol 1990:108: 1980;74:806-8. 676-8. 22. Ferrante A, l'itro Rowan-Kelly B, Thong YH.III sensitivity 42. Akins RA, Byers Tl. Differentiation promoting factors of Acallthalllocba clIlbertsoni to a variety of drugs and anti­ induced in Acanthamocba by inhibitors of mitochondrial biotics.Int J Parasitol 1984;14:53-6. macromolecule synthesis.Dev Bioi 1980;78:126-40. 23. Schuster FL. Comparative effects of selected azole com­ 43. Anthony Y, Davies D1G, Meakin Bl, Halliday 1, Kumar R, pounds on trophic and cystic stages of Acanthamocba poly­ MacDonald L Ritchie M.A chlorhexidine contact lens disin­ phaga. J Euk Microbiol 1993;40:563-9. fection tablet: design criteria and antimicrobial efficacy in 24. Osato MS, Robinson NM, Wilhelmus KR, 10nes DB. l'it­ III potable tap water.J Br Contact Lens Assoc 1991:14:99-108. ro evaluation of antimicrobial compounds for cysticidal 44. Cairns 0, Hay L Seal DV The new macrolides: expanding activity against Acallthallloeha. Rev Infect Dis 1991;13 the frontiers of antimicrobial chemotherapy.Pharrn J 1993; (Suppl 5):S431-4. 251:3 17-20. 25. Casemore DP. Sensitivity of Hartlllallella (Acallthallloeba) 45. Ferrante A, Abell TJ, Robinson B, Lederer E. Effects of to 5-ftuorocytosine, hydroxystilbamidine and other sub­ singefungin and diftuoromethylornithine on pathogenic stances.1 Clin Pathol 1970:23:649-52. free-living amoebae in vitro. FEMS Microbiol Lett 1987;40: 26. Stevens AR. O'Dell WD. III l'itro and ill l'il'O activity of 67-70. 5-ftuorocytosine on Acanthallloeha. Antimicrob Agents 46. Shukla OP, Kishore P. Gupta S, Srivastava DK. Potential Chemother 1974:6:282-9. metabolic targets in chemotherapy for Acanthamoeha 27. Duma RJ, Finley R. In l'itro susceptibility of pathogenic : polyamine metabolism.Rev Infect Dis 1991:13 Naegleria and Acanthamoeba species to a variety of thera­ peutic agents. Antimicrob Agents Chemother 1976;10: (Suppl 5):S438. 370-6. 47. Shukla OP, Muller S, Walter RD.Polyamine oxidase from 28. Ma P, Willaert E, leuchter KB, Stevens AR.A case of kera­ Acanthamocba clIlhertsoni specific for N� -acetylsperrni­ titis due to Acanthamoeha in New York, New York, and fea­ dine.Mol Biochem Parasitol 1992;51:91-8. tures of 10 cases. 1 Infect Dis 1981: 143:662-7. 48. Kuyyakanond T, Quesnel LB.Th e mechanism of action of 29. Kilvington S, Larkin DFP, White DG, Beeching JR.Labora­ chlorhexidine. FEMS Microbiol Lett 1992; 100:211-6. tory investigation of Acallthalllocba keratitis.J Clin Micro­ 49. Hugo ER. Byers TJ. S-adenosyl-L-methionine decarboxy­ bioi 1990;28:2722-5. lase of Acanthamocba castellanii (Neff): purification and 30.Nagington 1, Richards lE. Chemotherapeutic compounds properties. Biochem 1 1993;295:203-9. and Acallfhamoehae from eye infections. 1 Clin Pathol 50. Greenidge PA, 1cnkins TC, Neidle S. DNA minor groove 1976;29:648-51. recognition properties of pentamidine and its analogs: a 31. Saunders PPR. Proctor EM, Rollins DE Richards JSF. molecular modelling study.Mol Pharmacol 1993;43:982-8. Enhanced killing of Acanthamoeba cysts ill l'itro using 51. lenkins TC, Lane AN, Neidle S, Brown DG. NMR and dimethylsulfoxide. Ophthalmology 1992;99: 1197-200. molecular modelling studies of the interaction of berenil and 32. Schuster FL, lacob LS.Effects of magainins on ameba and pentamidine with d(CGCAAATTTGCGh. Eur 1 Biochem cyst stages of Acanthalllocba polvphaga. Antimicrob 1993:213:1175-84. Agents Chemother 1992;36: 1263-71. 52. Amott MA, Hay 1. The effect of pentamidine salts on the 33.Mehlotra RK, Shukla OP. In 1'itro susceptibility of Acallfha­ NADPH-oxidase system of stimulated neutrophilic granu­ mocba clllbertsoni to inhibitors of folate biosynthesis.1 Euk locytes.l Antimicrob Chemother 1990;25:247-53. Microbiol 1993;40:14-7. 53. Arnott MA, Bennett ND, Cairns D, Hay 1.Selective effects 34. Schuster FL, Mandel N. Phenothiazine compounds inhibit of pentamidine on cytosolic and granule-associated enzyme in l'itl'o growth of pathogenic free-living amoebae. Anti­ release from zymosan-activated human neutrophilic granu­ microb Agents Chemother 1984;25: 109-12. locytes.1 Pharm Pharmacol 1994;46:394-6. 35. Osuna A, Rodriguez-Santiago n, Ruiz-Perez L-M, Gamarro 54. Perrine D, Barbier D, Chenu P, Georges P. Comparative F, Castanys S, Giovannangeli G, et al. Antiamebic activity study of cysticidal effects of three diamidines on Acantha­ of new acridinic derivatives against Naegleria and Acantha­ mocba strains isolated from keratitis.In: VIth internatiol1al moeha species in l'itro. Chemotherapy 1987;33: 18-21. Conference on the Biology and Pathogenicity of Free-Liv­ 36.Ishibashi y, Matsumoto Y, Kabata T, Watanabe R, Hom- ing Amoebae 1992, abstract 288.