Comparative antifungal activity of cilofungin ( LV 121 0 19 ) against Candida species / including evaluation of susceptibility testing method

AH C HAGLA, JH Hn, DJ HoBAN, et al . Comparative antifungal activity of cilofungin (LY121019) against Candida species, including evaluation of susceptibility testing method. Can J Infect Dis 1992;3(5): 231 -234. The in vitro activity of cilofungin against 100 Candida species was compared with 5-0ucytosine. amphotericin 8 and ketoconazole by two laboratories independenUy and in a blinded fash ion using a macrotitre dilution broth method in SAAM -F medium. Cilofungin howed good in vitro activity against Candida albicans. Candida tropicalis and Candida glabrata (90% minimal inhibitory concentration [MlC) 3.2 pg/mL) but was inactive against other Candida species . When testing the susceptibility of cilofungin. 5-0ucytosine and amphotericin 8 at the two centres. approximately 90% of the Candida strains had M!Cs differing by fourfo ld or less . However. when testing susceptibil ity of ketoconazole. on ly 51% of U1e Candida strains had MIC differences fourfo ld or less . MIC susceptibi lity testing \vith ci lofungi n , 5-0ucytosine and amphotericin 8 in SAAM -F medium is reproducible.

C JLOFU! GIN (N -P-OCTYLO:--.'YBENZOYLECHI NOCANDIN B: LY121019) is a sem isyn tl1e tic lipopeptide a n alogu e of th e polypeptide antibiotic ec hinocardin.It is believed to inhibit the biosyntl1esis of the beta-1-3 glu can compon ent of the Candida albicans cell wall in non sta tionary.m eta bolizing c lis .t hereby d is rupting ce ll wall integri ty (1.2).It is known to be active aga inst certain Candida species.In vitro susceptibility tes ts show that minimal inhibitory con centra tions (MI Cs) for ci lofungin against C albicans an d Candida tropicalis a re lower than those against Candida glabrata and other Candida species (3)(4)(5) .An imal stud ies show it to be 20-fold less toxic than amphote ricin B (6).
In vitro susceptibility testing for fungi and yeasts has been unsatisfactory beca u se of a lack of sta ndard test criteria.Variations in in oc ulum preparation.medium composition.pH.duration of in cubation.temperature a nd endpoint determina tion h ave a ll contributed to variable test results a m ong laboratories (7).Specifically.various laboratori es have reported a nearly eightfold difference in MICs for cilofungin aga inst C a lbicans.with results differing from 0.31 to 2 .5 ~tg/mL (3-6.8.9).Different media.inoculum s ize a nd temperature of incubation were used.Hall et al (8) have suggested that c il ofungin activity was affected by the growtl1 m edium and inoculum s ize .In contrast.Strippoli et a l (9) h ave noted Ulat compositional differences in tl1 e medium and/ or the presence of anim a l se rum did not adversely affect s u sceptibili ty tes ting witll cilofuno-in.The few stud ies which h ave analyzed t11 e antifungal activity of cilofungin have compared results based on differing metl1odologies.while none has evalu ated a method of cho ice to determine reproducibility between different laboratories.
The a u tho rs repo rt a s tudy on the in vitro activity of cilofungin again st 100 clinical isolates of Candida spec ies u ing a m acrolitre broth dilution m ethod with syntl1etic am in o ac id m edium for fungi (SAAM -F) (10.11).and compare test results meas ured by MIC and minimum fungicidal concentra tion (MFC) of th ese strains obtained between two laboratories in Canada.
A total of 100 clinical i ola tes of Candida s pec ies were examined.Fifty strain s recovered from blood cultures were collected at Health Sciences Centre in Winnipeg.Manitoba and the remaining 50 i olates.obtained from specimens of various body sites (50 isolates) .were coll ected at Moun t Sinai Hospital in Toronto.Ontario .Identi fication was based on microscop ic morphology.germ -tube formation.chl am ydospore development and b iochemical tests using tl1 e API 20C system (Analyt.abProducts.New York).These isolates in cluded 57 stra in s of C albicans.16 strains of C tropicalis.19 strain s ofCglabraLa.five Candidaparaps ilos is strain s. a nd one eac h of Candida quilliermondii.Candida lusiLaniae and Candida Jcrusei (co llectively referred to in the text as other Candida species).Stock cultures were preserved in skim milk at -70°C and regularly m a intain ed on 5% sheep blood agar or Sabouraud•s de>.. 't:rose agar.SAAM -F is comprised of 16 amino acids .glu cose .fumaric a nd pyruvic acid.a mmonium acetate.pota ss ium hydrogen phosphate.N-2-hydroxyetl1yl piperazine -N-2 -eth a n e sulphonic acid (H EPES) an d Tris buffe r (11).Vitamins a nd other salts were a dded separately.a nd the med ium adj u sted to pH 7.4.All reagents were purchased from the Sigma Ch emical Co (M issouri) a nd con stituted ind epend ently at each refe ren ce labora tory.
Cilofungin. provided by Eli Lilly Research Laboratories (Indiana) dissolved in 50% ethanol.Stock solutions of amphotericin B. provided by ER Squibb a nd Sons (New Jersey).and ketocon azo le.provided by Janssen Pharmaceutica Inc were dissolved in dimethyl s ulphoxide.The 5-flu cyt.osine.provided by Roche Laboratories (New Jersey).was dissolved in sterile distilled water.Th e stock solutions were stored at -72°C to retain potency.Se1ial twofold dilutions o f a ntifungal agen ts were m ade witl1 sterile distilled water to achi eve working solution ranging from 0 .05 to 200 pg/mL.A 0 .5 mL working solution was furt11er diluted by h alf witl1 a ddition of 0.5 mL 2x (doubly concentrated) SAAM -F medium to obtain fin a l drug concentration s ranging from 0.025 to 100 pg/mL.
Fungal cultures grown on blood agar for 48 h at 30°C were recovered a nd suspended in 5 mL sterile saline a nd standardized to 0 .5 MacFarland.The cell s u s pens ion was diluted 100-fold .and a n actu a l cell count was made on dilu ted s uspen s ion using a hemacytometer.A 50 pL aliquot was U1en used as the inoculum for s usceptibility testing to obtain a fin a l con centra tion of lx10 3 yeast cells/mL.Controls were used for medium a lon e and for medium with 2% ethanol.A sa mp le of inoculum was placed on blood agar to check for co ntamination.inocu lum quantitation and viability.The tubes were inc ubated unagitated for 48 h at 30°C.The M!Cs were determined after gen tle agitation of tl1e tubes a nd co mparison witl1 controls.The M! Cs were defined as t11e lowest concentratio n of antifu n gal agent tested which yielded no macroscopic turbidity compa red wit11 uninoculated control.To detem1 in e lhe MFC. 10 ~tL of agitated suspension .from one lube before a nd at least six tubes after Ml C endpoints .was s ubcultured on 5% sheep blood agar or Sabouraud•s dextrose agar.and incubated for 48 h at 30°C .Colonies were counted and the MFC was defined as tl1e lowest concentration at which 99% of t11e initia l inoculum was killed .An Eagle effec t was not produced using t11e present testing method .Controls were performed with each lest.using two reference str ains of C a lbicans ATCC 10231 and LY-A26 provided by Eli Lilly Research Laboratories.The study was blinded and carried out independently at the two refe rence labora tories.Cultured isolates were given numbers prior to the study and exchanged between Ule two laboratories without any accompa nying information concern ing t11e so urce or stra in iden tification.Each reference laboratory pe rformed susceptibility testing on the 100 clinical isolates using a common protocol.and without prior knowledge of olher laboratory resu lts.Based on NCCLS gui delin es. a difference of fourfold or less in MICs and MPCs was con idered acceptable (12) .An overall 95% confidence limit was computed.taking into consideration lhe comparison of MICs and MPCs for cilofungin to the other lhree drugs.The MIC5o.MJCgo and MPC results of in vitro susceptibility testing of U1e 100 isolates using an inocu lum of 10 3 yeast cells/mL are summarized in Tab le l.Cilofungin showed good in vitro activity against C albicans .C Lropicalis. and C g labrata (M! Cgo 3.2 pcr/mL).but was inactive against the oU1er Candida species.Amphotericin B showed significant in vitro activity against all lhe Candida strains tested.A significant TABLE 2 trail ing effect was seen when ketoconazole was tested.wiU1 considerable d isparity between partial an d complete end -point inhibition.Th is may be the reason fo r interlaboratory va riation in dete rmin ing antifungal activity of ketoconazole.and has been confirmed by other investigators (13).The determination of end-points by measurement of turbidity and calculation of IC5o (concentration of antifungal agent lhat.. causes 50% decrease in turbid ity) has been proposed as being more reliable than conventional (visual) scoring for absence of growth (14).In lhe presen t study trailing effect was not observed with other antifungal agents which were t sl.ed.and U1e resu lts were based on determination of clear cut.. end-points.
Tab le 2 shows lhe variations in in vitro susceptibili ty tes ting of U1e four antifungal agents between the 1:\vo  r ference laboratories.Approximately 90% of U1 e strains tested against c il ofungin, 5 -0ucyt.osineand amphotericin B showed a d ifference in M!C and MF'C by fourfold or less.Ninety per cent of ci lofungin MlCs had less U1an or equal to two well dilution differences.but on ly 74% had 1:\vo or fewer well d ilution differences.This variability cou ld not be explained by a trailing effect.since end-po ints were clearcul.A discordance in boU1 M!C and MF'C was seen wiU1 ketoco nazole between U1e 1:\vo centres .Antifungal testing wiU1 cilofungin has been performed variably using microtilre and macrotitre broU1 dilution m eU1ods.various m edia includin<J SAAM -F. and various inoculum sizes.In U1ese reports (3-6.8.12) U1ere was a fourfold or greater variation in MIC e ndpoints bel:\veen studies.In U1e present study.U1e auU1ors carried out interlaboratory comparison of antifungal suscepUbiUty res ults using a standardized meU1od which included an inoculum of 10 3 yeast cells/ mL and a constitutionally defined medium.SAAM -F.recommended for anUfun<Jal susceptibility testing (10).
SAAM -F has only been used in 1:\vo studies using a similar inoculum (3.15).and was U1ought to account for U1e loss of susceptibility to c ilofungin in some isolates of C albicans and C trop icalis: an Eagle effect was also described wiU1 testing several isolates of C iropicaLis (3).This problem was not observed in U1e current study.
An optimal inoculum of 10 3 yeast ce lls/mLwas used in U1e present study.comparable to U1e inoculum used by oU1ers (4.6).Th is was determined by testing 27 representative strains (data not shown) .An in oculum greater Ulan 10 5 yeast cells/mL resulted in decreased activity of cilofungin.nucyt.osineand ketoconazole.but not of amphotericin B. in agreement with the finding of oU1ers (8.16).The disparity in MIC end-points by using high inocula may renect an absolute increase in inocu lum to drug concentration or. which is less likely.resistance to c ilofungin (16).
Cilofungin demonstrated less activity against C albicans.C tropicalis a nd C glabrata U1an reported e lsewhere (3 -6.8).This may have resulted from testing in SAAM • F (3) .In contrast to the present findings.Hobbs et a l (3) r eport a MlCgo threefold higher for cilofungin against C g labrata.using a microtitre assay system in SAAM -F.In agreement wiU1 oilier reports (3 -6.8).c ilofungin has demonstrated litUe or no in vitro activity against U1e oilier Candida species tested.
Cilofungin (LY121019) demonstrates good in vitro activity against C albicans.C tropicali.sand C glabrala wiU1 MlCs well wiU1in achievable anim al serum concentrations.A common problem with antifungal suceptibility testing has been an unacceptable discordance of results bel:\veen various laboratories due to differing test condilions (7).However. in U1e present study U1 e auU1ors were able to provide reproducible inte rlaboratory MIC results when testing cilofungin.5-0ucytosine and amphotericin using a standardized meU1odology. auainst.

C
A N J INFEC T DIS V OL 3 No 5 SEPTEMBER/OCTOBER 1992