Modulation of the Intestinal Flora of Mice by Parenteral Treatment with Broad-Spectrum Cephalosporins

Transcription

1 ANTIMICROBIAL AGNTS AND CHMOTHRAPY, My 1991, p /91/59767$2./ Copyright 1991, Americn Society for Microbiology Vol. 35, No. 5 Modultion of the Intestinl Flor of Mice by Prenterl Tretment with BrodSpectrum Cephlosporins M. L. VAN OGTROP,"12 H. F. L. GUIOT,12 H. MATTI,1* AND R. VAN FURTH' Deprtment of Infectious Diseses, University Hospitl,' nd J. A. Cohen Institute for Rdiopthology nd Rdition Protection,2 Leiden, The Netherlnds Received 1 August 199/Accepted 27 December 199 A study ws performed to determine the effect of prenterl tretment with four brodspectrum cephlosporins (cefoperzone, ceftrixone, ceftzidime, nd cefepime) on the number of erobic grmnegtive rods nd on the outgrowth of Cndid lbicns nd multiresistnt strin of Citrobcterfreundii in the feces of mice. The estimted frctions of prenterl dose tht were excreted into the gstrointestinl trct were.37 for cefoperzone,.11 for ceftrixone,.3 for ceftzidime, nd.2 for cefepime. All four cephlosporins significntly decresed the number of erobic grmnegtive rods in the feces, nd virtully ll grmnegtive rods were eliminted t high doses of cefoperzone, ceftzidime, nd ceftrixone. Furthermore, t high doses these three compounds led to significnt increse of the outgrowth of resistnt Citrobcter freundii. The outgrowth of Cndid lbicns ws incresed t high doses of cefoperzone nd ceftrixone, wheres ceftzidime nd cefepime did not hve this effect. The most profound chnges in the gstrointestinl ecology were observed during tretment with high doses of cefoperzone. The results suggest tht the coloniztion resistnce of the gstrointestinl trct cn be substntilly decresed by prenterl tretment with cefoperzone nd, to lesser extent, with ceftrixone nd ceftzidime. Brodspectrum cephlosporins re frequently used for empiricl tretment nd for the tretment of grmnegtive infections in grnulocytopenic ptients (15). These compounds possess powerful in vitro ctivity ginst grmnegtive bcteri, including Pseudomons eruginos. Most grmpositive bcteri re lso considered susceptible. Brodspectrum cephlosporins my lso disturb the nerobic nd erobic intestinl microflor (1, 2, 5), which cn led to decrese of coloniztion resistnce to opportunistic pthogens. This my led subsequently to overgrowth by fungi or multiresistnt bcteri, whichespecilly in grnulocytopenic ptientsincreses the risk of cquiring nosocomil infection. The degree of disturbnce of the intestinl microflor due to the prenterl dministrtion of ntibcteril gents is dependent on the intrinsic ctivity of the ntibcteril gent ginst the indigenous bcteri of the gstrointestinl trct s well s on the phrmcokinetic chrcteristics of the ntibcteril gent determining the concentrtion of the gent in the gstrointestinl trct. In the present study, the effect of prenterl dministrtion of four brodspectrum cephlosporins (cefoperzone, ceftrixone, ceftzidime, nd cefepime) on the erobic grmnegtive fecl flor nd the coloniztion resistnce of the intestinl trct ws determined. These four drugs were chosen for severl resons. Cefoperzone ws included becuse it hs brodspectrum ctivity ginst intestinl nerobic bcteri nd is excreted predominntly in the bile (6); ceftrixone ws included becuse it is lso excreted into the bile (23), but it lcks brod ntinerobe ctivity. Ceftzidime ws studied becuse it lcks brod ntinerobe ctivity nd, unlike cefoperzone nd ceftrixone, is minly excreted by the kidney (22); nd cefepime ws studied becuse, like ceftzidime, it is minly excreted by the kidney (9), but its ntibcteril spectrum includes grmpositive * Corresponding uthor. 976 cocci such s Stphylococcus ureus (13, 14, 25). Comprison of the effects of these four ntibcteril gents ws expected to provide informtion bout the reltionship mong the effect of brodspectrum cephlosporins on gstrointestinl flor, their in vitro ntibcteril spectrum, nd their phrmcokinetic chrcteristics. MATRIALS AND MTHODS Antibiotics. Cefoperzone (sodium slt, ctivity 96.4%) ws kindly donted by Pfizer Phrmceuticls, Rotterdm, The Netherlnds; ceftzidime (ctivity 84.2%) ws donted by Glxo, Nieuwegein, The Netherlnds; ceftrixone (ctivity 83.95%) ws donted by HoffmnnL Roche, Bsel, Switzerlnd; nd cefepime (ctivity 82.6%) ws donted by Bristol Myers, Brussels, Belgium. Just before use, solutions of the drugs were prepred freshly in phosphtebuffered sline (PBS), ccording to the mnufcturer's instructions. Animls. Femle specificpthogenfree Swiss mice (Broekmn Institutes, Someren, The Netherlnds), ged 5 to 6 weeks, were used throughout the study. The nimls were housed in groups of four to five in polycrbonte cges on sterile swdust nd were given cidified tp wter nd nonsterilized food pellets (type AMII; Hope Frms, Woerden, The Netherlnds) d libitum. Cges nd swdust were replced twice weekly. No specil mesures were tken to prevent coprophgi. The effect of prenterl tretment with oncedily doses of the cephlosporins under study ws ssessed over 9 to 11 dys, the exct durtion depending on the conditions of the individul experiments. ffect of cephlosporins on number of erobic grmnegtive rods in feces of conventionl mice. The mice were given one dily subcutneous injection of the drug under study t vrious dosges on nine consecutive dys. For ll experiments, the dose rnges studied were 1 to 16 mg/kg/dy for ceftzidime, 6.25 to 1 mg/kg/dy for cefepime, 25 to 4 mg/kg/dy for cefoperzone, nd 5 to 8 mg/kg/dy for ceftrixone. These experiments used five mice ech. Downloded from on October 21, 218 by guest

2 VOL. 35, 1991 FFCT OF BROADSPCTRUM CPHALOSPORINS ON GUT FLORA 977 On ech weekdy, fresh fecl pellets were collected, weighed, nd suspended in 1 ml of PBS. Tenfold seril dilutions of these suspensions were mde. Smples (.1 ml) of the dilutions were inoculted on both blood gr pltes nd McConkey III gr pltes (Oxoid, Bsingstoke, United Kingdom) which were incubted t 37 C. After incubtion, differentil counts were mde on the McConkey gr pltes. The lower limit of detection ws 5 CFU of erobic grmnegtive rods per g of feces. If no erobic grmnegtive rods were recovered from the feces, the number of CFU per grm of feces ws rbitrrily tken s 1 for further clcultions. Results re expressed s the geometric men of the number of erobic grmnegtive rods per grm of feces. Identifiction of members of the fmily nterobctericee ws bsed on Grm stining, bsence of oxidse production, nd biotyping by the API 2 system (API Lbortories, L Blme, Frnce). In ddition, the MICs of the cephlosporins for the isolted strins of erobic grmnegtive rods were determined by n gr dilution method (4), with MuellerHinton gr (Oxoid) s the test medium nd n inoculum density of between 5 x 13 nd 5 x 14 CFU per spot. ffect of cephlosporins on number of Cndid lbicns in feces of postntlly contminted mice. The effect of the cephlosporins on the outgrowth of Cndid lbicns (strin UC82) ws ssessed in murine model of gstrointestinl cndidisis in seprte group of mice, s described by Pope et l. (21). Sixdyold mice were orlly contminted with 2 x 16 to 5 x 16 CFU of Cndid lbicns. At the ge of 21 to 22 dys, the mice were wened nd housed in groups of four. ch tretment group consisted of eight mice. Tretment with the cephlosporins ws strted on dy 1 fter wening nd consisted of single subcutneous dose of the drug per dy. Control mice received PBS. Fresh feces were collected before nd vrious dys fter the strt of tretment. These smples were weighed, suspended, nd diluted in PBS, fter which the number of Cndid lbicns per grm of feces ws determined by quntittive culturing on Sbourud gr (Oxoid) supplemented with 2 mg of knmycin (GistBrocdes, Delft, The Netherlnds) per liter nd BIGGY (bismuth sulfite glucose glycine yest) gr (Oxoid). Pltes were incubted t 37 C for 48 h, nd the CFU were counted. The lower limit of detection of this ssy ws 5 CFU of Cndid lbicns per g of feces. Orl contmintion with Citrobcter freundii nd effect of cephlosporins on recovery of this microorgnism from feces of mice. The effect of ech of the cephlosporins on the outgrowth of resistnt strin of Citrobcterfreundii (MIC: ceftzidime, 128,ug/ml; ceftrixone, 64,ug/ml; cefoperzone, 64,ug/ml; nd cefepime,.25,ug/ml) ws ssessed fter dily orl dministrtion of n inoculum of 5 x 17 CFU of Citrobcter freundii through feeding needle for 7 dys. This strin of Citrobcter freundii, chosen from pnel of clinicl isoltes of grmnegtive bcteri resistnt to brodspectrum cephlosporins (kindly donted by J. vn der Klundert), displyed dequte colonizing properties in the gstrointestinl trct of mice combined with the highest level of resistnce to the cephlosporins under study. This experiment ws performed in seprte group of mice. ch tretment group consisted of six mice. These mice hrbored no cephlosporinresistnt erobic grmnegtive rods in their feces before the experiment. Tretment with cephlosporin nd orl contmintion with Citrobcter freundii were begun on the sme dy. The mice were given the drug subcutneously for 11 dys. The number of Citrobcter freundii per grm of feces ws determined by quntittive culturing on McConkey III gr pltes supplemented with 1,ug of ceftzidime per ml. The pltes were incubted t 37 C for 24 h, fter which the CFU were counted. The lower limit of detection of this ssy ws 5 CFU of Citrobcter freundii per g of feces. If no Citrobcter freundii ws recovered from the feces, the number of CFU per grm of feces ws rbitrrily tken s 1 for further clcultions. Results re expressed s the geometric men of the number of Citrobcterfreundii per grm of feces. ffect of cephlosporins on reltive cecl weight nd totl number of bcteri in the cecum. The reltive cecl weight (16) nd the totl number of bcteri per grm of cecum were determined fter 1 dys of prenterl tretment with cephlosporin. For these determintions, mice were killed by cervicl disloction, the cecum ws removed nd weighed, nd the totl wet weight ws expressed s milligrms per grm of body weight. For the totl count of bcteri per grm of cecl homogente, the excised cecum ws homogenized in 2 ml of sterile icecold PBS, the homogente ws diluted in PBS, nd the totl number of bcteri in the pproprite dilutions ws determined on Grmstined slides nd compred with n externl stndrd by the method of Holdemn nd Moore (11). This count reflects the number of nerobic bcteri in the cecum (12, 17) Ṗhrmcokinetics of cephlosporins in cecl contents of mice. Singledose phrmcokinetic studies were performed for the following subcutneous doses: cefoperzone, 1 mg/kg; ceftzidime, 16 mg/kg; ceftrixone, 5 mg/kg; nd cefepime, 1 mg/kg. At vrious time points fter dministrtion, mice were killed by exposure to 1% C2, the cecum ws excised, nd the cecl contents were removed, weighed, nd diluted in 2 ml of PBS. This mteril ws centrifuged t 1,5 x g for 1 min t room temperture, fter which the concentrtions in the superntnt were determined by highperformnce liquid chromtogrphy (HPLC). The re under the concentrtiontime curve (AUC) ws clculted ccording to the trpezoidl rule. HPLC nlysis of cephlosporin concentrtions. Acetonitrile, dichloromethne, nd sodium cette were of nlyticl grde nd were supplied by Merck (Drmstdt, Federl Republic of Germny). The liquid chromtogrphic system consisted of constntflow pump (model 1; Sykm, Anlytic BV, Rijswijk, The Netherlnds), Rheodyne model 7125 injection vlve equipped with 2,lI smple loop (Chrompck, Middelburg, The Netherlnds), stinlesssteel column (length, 1 cm, nd internl dimeter, 3 mm), nd Spectroflow 773 bsorbnce detector (Kipp & Zn., Delft, The Netherlnds). For monitoring, the detector ws set t 254 nm for ceftzidime, 274 nm for ceftrixone, 254 nm for cefoperzone, nd 28 nm for cefepime. The column ws homepcked with 5,um Hypersil ODS (Shndon SPL, Cheshire, United Kingdom) ccording to pressurizedslurry technique. The extrction procedure pplied to the smples contining cefoperzone, ceftzidime, or ceftrixone ws s follows. A 5,ul liquot of the smple ws vigorously mixed with n equl volume of cetonitrile in polypropylene tube on whirlmixer for 3 s. After centrifugtion for 5 min t 1,2 x g, 4 RI of the superntnt ws mixed with 3 ml of dichloromethne for 3 s nd centrifuged for 5 min t 1,2 x g. A 2pu liquot of the upper queous phse contining the nlyte ws injected onto the column nd chromtogrphed t flow rte of 1 ml/min with mobile phse consisting of.7% (vol/vol) cetonitrile in.5mol/liter Downloded from on October 21, 218 by guest

3 978 VAN OGTROP T AL. L lol lo, ; cefoperzons celfrixone 1mg/kg) Img/kgl * j! *. v ANTIMICROB. AGNTS CHMOTHR. 1r cefltzidime mg/kg) or ceftrixone. The results suggested tht complete erdic tion of erobic grmnegtive rods within 1 week ws o' chieved by cefoperzone t doses of 1 mg/kg/dy nd higher, by ceftzidime t 4 mg/kg/dy nd higher, nd by tceftrixone t 8 mg/kg/dy. ; 14 The grmnegtive species isolted from the feces before Nthe strt of tretment were scherichi coli, Proteus mirte 1 bilis, nd Morgnell morgnii, mong which. coli pre XOt \i ^ ^ ^ ^4 dominted IGO (usully ccounting for more thn 9%). The erobic grmnegtive rods still present in the feces fter exposure to low doses of ny of the cephlosporins were. I dys coli nd Proteus spp. The MICs of the four cephlosporins for the isolted strins of erobic grmnegtive rods re cele8pme Img/klg) given in Tble 1. No chnge in the ntimicrobil susceptibil ity of the isolted bcteri to ny of the four cephlosporins ws detected during the study. Coloniztion with new bioo \ 625 types of grmnegtive species during this period ws not observed either. ffect of cephlosporins on number of Cndid lbicns in 14 feces of postntlly contminted mice. To determine whether 2 prenterl tretment with brodspectrum cephlosporins '. predisposes to intestinl coloniztion nd overgrowth with, U l Cndid lbicns, we dministered the drugs to mice U conii tminted postntlly with Cndid lbicns. In the control mice, which were treted with PBS from dy 21 on, the S O dys I dys number of Cndid lbicns tended to decrese slightly FIG. 1. ffect of prenterl tretment with ech of four ceph during the study period, i.e., from between 15 nd 16 losporins for 9 dys on the totl number of indigenous erobic CFU/g of feces t the strt to between 14 nd 15 1 dys grmnegtive rods (AGNR) in tlhe feces of mice. ch symbol lter (Fig. 2). A dose of 4 mg of cefoperzone per kg per represents the geometric men o If the number of erobic grm dy led to significnt increse in the numbers of Cndid negtive bcteri per grm of feces of five mice. The numbers t the lbicns within 2 dys fter the strt of tretment to between end of the curve give the dose of tthe drug dministered dily. 17 nd 18 CFU/g of feces. Tretment with 8 mg of ceftrixone per kg per dy hd smller, but still significnt positive effect on the outgrowth of Cndid lbicns. In cette buffer (ph 5.5). Smpli[es contining cefepime were mice given ll other dosges of cefoperzone, ceftrixone, extrcted by the method of Birbhiy et l. (3). Clibrtion ceftzidime, or cefepime, the outgrowth of Cndid lbicns plots were obtined by ssyinjg smples of pooled supern did not differ significntly from tht in controls. tnts of murine cecl contents, spiked in the rnge between Orl contmintion with Citrobcter freundii nd effect of.5 nd 1,ug/ml. cephlosporins on recovery of this microorgnism in feces of Sttisticl procedures. The retsults pertining to the qun mice. To determine whether tretment with the cephlospos the geometric men of the rins could led to coloniztion with resistnt grmnegtive tittive fecl cultures re given number of CFU per grm of feces. Other results re ex rods, we contminted dult mice orlly with resistnt pressed s men stndrd enror. Repetedmesures nl strin of Citrobcter freundii. In the control mice, the ysis of vrince ws used for st; tisticl comprisons (7). A P geometric men number of these bcteri during week 1 of level of.5 ws considered silgnificnt. the experiment ly between 15 nd 16 CFU/g of feces (Fig. 3). In week 2 of the experiment, the orl contmintion ws RSUILTS discontinued nd the number of CFU of Citrobcterfreundii tended to fll to levels below the detection limit (5 CFU/g ffect of cephlosporins on n umber of erobic grmneg of feces). For sttisticl purposes, these mice were rbitrr mice. The effect of 9 to 1 ily considered to hve 1 CFU of Citrobcterfreundii per g tive rods in feces of conventionl dys of tretment with one ( dily subcutneous dose of in their feces (Fig. 3). cefoperzone, ceftzidime, cefitrixone, or cefepime on the In cephlosporintreted mice, the dministrtion of the totl number of erobic grminegtive rods in the feces is drug nd orl contmintion with Citrobcter freundii were shown in Fig. 1. All four cephliosporins reduced the number begun simultneously. Tretment with 4 mg of cefoper the feces, but cefepime ws zone per kg per dy hd the strongest effect on the microbil of erobic grmnegtive rods iin less effective in this respect th: n cefoperzone, ceftzidime, ecology of the intestinl trct, resulting in n increse of the Downloded from on October 21, 218 by guest Orgnism TABL 1. MICs of the cephlosporins for cultured erobic grmnegtive rods from feces of mice MIC (mg/liter) Cefoperzone Ceftzidime Ceftrixone Cefepime scherichi coli Proteus mirbilis Morgnell morgnii

4 VOL. 35, 1991 FFCT OF BROADSPCTRUM CPHALOSPORINS ON GUT FLORA 979 t) ctfoperzone s 1dys 1 1o ;. lo o :D Ly 14 U. 1 ceftzidime 16 5 lodys O 1 dysj * * Ws 1 ol! U.3 lm ; cefoperzone (mg/kgl A==~~/ 4 1 _v>u2s 25 C o II dys g 6 log o l og 1 dl (L. t IIIttt ceftzidime oi.l s 6 7 I dys we. 1Iu,14 O ceftrixone s 1 dys lo,. l3 C) o c*efpime 1mg/kg) l225~~~~~~2 o s 1 dys FIG. 2. ffect of 9 to 1 dys of prenterl tretment with the four cephlosporins under study on the recovery of Cndid lbicns in the feces of mice. ch symbol represents the geometric men of the number of Cndid lbicns per grm of feces of eight mice. The numbers t the end of the curve give the dose of the drug dministered dily. number of Citrobcter freundii to between 19 nd 11 CFU/g of feces (Fig. 3). Cefoperzone t lower dosges, 4 mg or more of ceftzidime per kg, or 2 mg or more of ceftrixone per kg dily hd less pronotunced effects on the numbers of Citrobcter freundii in the feces, which incresed to between 16 nd 18 CFU/g of feces. Tretment with cefepime (up to 1 mg/kg/dy) did not hve ny significnt effect on the recovery of this strin (Fig. 3), but this strin ws susceptible to cefepime (MIC,.25,ug/ml). ffect of cephlosporins on reltive cecl weight nd totl number of bcteri in the cecum. The reltive cecl weight is nother prmeter of coloniztion resistnce, nd the totl number of bcteri in the cecum reflects the number of nerobes in tht orgn. Administrtion of 4 mg of cefoperzone per kg per dy or 2 or 8 mg of ceftrixone per kg per dy led to significnt increse in the reltive cecl weight (Tble 2). Qnly 4 mg of cefoperzone per kg per dy gve significnt decrese in the totl number of bcteri in the cecum (Tble 2). Other tretments hd no effect on either the reltive cecl weight or the totl number of bcteri in the cecum. Phrmcokinetics of cephlosporins in cecl contents in mice. The concentrtions of the cephlosporins in the contents of the cecum re shown in Fig. 4. The AUC from to 5 h ws 4,548,ug. h/g for dose of 1mg/kg cefoperzone, 583 jig h/g for 16mg/kg ceftzidime, 68.8 jig h/g for 5mg/kg ceftrixone, nd 2.5,ug h/g for 1mg/kg cefepime. The rtios between the AUC nd the dose, s reltive prmeters of the proportion of the dose excreted into the gstrointestinl trct, were 45.5 for cefoperzone, 3.64 for ceftzidime, 13.8 for ceftrixone, nd.16 for cefepime. 1".6 we' * Go fi C1 l 1 1t1tt t t ceftrixone (mg/kgl t 11 dys 1o * * It e ; _,! I (.3, ti ti t cefepime (mg/kgl ttt o_1 t t t O o * 1lo dys FIG. 3. ffect of prenterl tretment with ech of the four cephlosporins for 11 dys on the recovery of Citrobcterfreundii in the feces of mice during dily orl contmintion of the mice with 5 x 17 CFU for the first 7 dys. ch symbol represents the geometric men of the number of Citrobcter freundii per grm of feces of six mice. The numbers t the end of the curve give the dose of the drug dministered dily. The rrows indicte the times of contmintion of the mice with Citrobcter freundii. To estimte the percentge of prenterl dose excreted into the gstrointestinl trct, we lso mesured concentrtions of ceftzidime in cecl contents fter orl dministrtion. The AUC from to 5 h (AUCO5) for cecl contents fter n orl dose of 16 mg of ceftzidime per kg ws 19,42 p,g. h/g. If it is correct to ssume tht orlly dministered ceftzidime is not bsorbed or metbolized in the gstroin TABL 2. ffect of 1 dys of tretment with cephlosporins on the reltive cecl weight nd the totl number of bcteri per grm of cecl homogente Reltive cecl Log no. of Drug (mg/kd ) weight (mg/g bcteri/g of of body wt) cecum Control 16.3 (3.) 1.9 (.2) Cefoperzone (12.2)b 8.8 (1.1) (5.5) 1.7 (.6) (4.6) 11. (.2) Ceftzidime (3.) 1.9 (.2) (2.9) 11. (.2) (2.8) 11.1 (.1) Ceftrixone (3.4)b 1.8 (.2) (5.5)b 1.8 (.3) (2.4) 11.1 (.2) Cefepime (2.1) 11. (.1) (2.7) 11. (.1) (2.3) 11.1 (.2) Vlues in prentheses give the stndrd devition. bsignificntly different from the control mice (P <.5). Downloded from on October 21, 218 by guest

5 98 VAN OGTROP T AL. I l Is ".1 : so u Ic 8.. * &, c u us 1. I I &A AL LA A A dos: 1 mg/kg s.c. I hours A,M.s dose: 5 mg/kg s.e hours.u 1 1. o FIG. 4. Drug concentrtions in cecl contei single subcutneous (s.c.) dose of cefoperzon trixone, or cefepime. ch symbol represents in single mouse. ANTIMICROB. AGNTS CHMOTHR. dose: 16 mg/kg s.c. tminted with resistnt strin of Citrobcterfreundii, the number of these orgnisms in the feces incresed during prenterl dministrtion of cefoperzone, ceftzidime, nd * * ceftrixone but not during dministrtion of cefepime. It is &. noteworthy tht this outgrowth of Citrobcter freundii occurred despite the fct tht the concentrtions of the cephlosporins in the cecl contents were much higher thn the * MIC. For exmple, the outgrowth of the Citrobcter strin & in the feces incresed with the dose of cefoperzone up to 4 mg/kg/dy. At this dose of cefoperzone, the concentrtions in cecl contents cn be expected to be s high s hours 4, pg/g, wheres the MIC of cefoperzone for this strin ws only 64,ug/ml. Therefore, this strin of Citrobcter freundii cn be considered resistnt in vivo (except for cefepime), nd therefore it is suitble indictor strin for doss: loo mg/kg 5.c. the detection of decrese of coloniztion resistnce. Prenterl dministrtion of high doses of cefoperzone nd ceftrixone lso led to n increse of the outgrowth of Cndid lbicns nd n increse of the reltive cecl weight of mice, but only high doses of cefoperzone gve mrked decrese of the totl number of bcteri in the cecum. An increse of the reltive cecl weight is generlly observed in & :. & &.& mice with disturbed intestinl microflor (17) nd hs been shown to correlte with decrese in coloniztion resistnce of the gstrointestinl trct (16). Coloniztion resistnce of the gstrointestinl trct hs hours clssiclly been linked to n intct nerobic microflor (26). Until now, the question of which nerobic bcteri re nts of mice given importnt in mintining coloniztion resistnce nd e, ceftzidime, cef whether erobic bcteri cn ply role in this process s the mesured vlue well is unsolved. The min problem in ddressing these questions is the culturing nd identifiction of murine intestinl nerobes. Despite lborious nerobic culture methods, no one hs yet succeeded in culturing more thn 6 to testinl trct of mice, this vlue cn be us ed to estimte the 8% of the nerobic intestinl bcteri of mice (12, 18). In frction of the subcutneous (s.c.) dose of drug tht is the present study, we decided to count the nerobic bcteccording to the ri indirectly vi microscopicl method (11, 17). Since excreted into the gstrointestinl trct, following eqution: F = AUCS.C./AUCOr, where F is the nerobic bcteri ccount for more thn 9% of the flor of excreted frction. For subcutneous dose of 16 mg of the mouse cecum (19), the decrese in the totl number of ceftzidime per ml, this vlue ws.3, iwhich mens tht bcteri implies decrese in the number of nerobic 3% of subcutneously dministered dose of ceftzidime IS bcteri in the cecum. excreted into the gstrointestinl trct. Ursing this estimte Vrious prmeters of disturbed coloniztion resistnce in for the other three drugs, we clculted the proportion ofthe mice treted with brodspectrum cephlosporins cn be dose excreted into the gstrointestinl 36.5% for cefoperzone, 11.1% for ceftriztrct, which ws compred on the bsis of the present results. A stndrd xone, nd.16% method used to estblish decrese of coloniztion resisconsidered rough tnce is mesurement of the outgrowth of resistnt erobic for cefepime. These percentges should be estimtes of the true percentge of the dosse tht is excreted grmnegtive rod in the feces of mice during tretment with into the gstrointestinl trct of mice. ntimicrobil gents. In this study, Citrobcterfreundii ws used s the test strin becuse it combines two chrcteristics: low sensitivity to three of the four cephlosporins under DISCUSSION study nd dequte recovery from the feces of mice. In tht prenterl cresed outgrowth of Citrobcter freundii in the feces oc The results of the present study showtzd dministrtion of brodspectrum cephlossporins cn hve curred t low doses of ceftzidime nd cefoperzone, mjor ecologicl impct on the intestinl flilor of mice, since wheres t these dosges ll other prmeters of decresed the investigted drugs (cefoperzone, ceftzidime, cefepime, coloniztion resistnce, i.e., outgrowth of Cndid lbicns nd ceftrixone) cused mrked dosedependent decrese in feces, reltive cecl weight, nd totl number of bcteri of the number of erobic grmnegtive rods. It my be in the cecum, showed no devition from the controls (Tble rgued tht this decrese might even be m( ore if coprophgi 3). This led us to conclude tht the ltter three prmeters would hve been prevented. However, thlis does not men re less sensitive for the detection of reduced coloniztion tht the effects of the vrious dosge iregimens re not resistnce thn the outgrowth of Citrobcterfreundii in the comprble, becuse the mice were lh vys housed in feces of mice treted with brodspectrum cephlosporins is. closed system under identicl conditions. Nevertheless, the The ecologicl chnges of the intestinl flor of mice by coprophgic behvior of mice hs to be bcrne in mind if the prenterl tretment with cephlosporins described in the results in mice re to be extrpolted to the humn sitution. present study cn lso be expected to occur in humns. The Other ecologicl effects of prenterl tiretment with the concentrtions of cefoperzone in the feces of group of cephlosporins were observed s well. In mice orlly con four ptients treted with 4 g of cefoperzone per dy vried Downloded from on October 21, 218 by guest

6 VOL. 35, 1991 FFCT OF BROADSPCTRUM CPHALOSPORINS ON GUT FLORA 981 TABL 3. Comprison of the vrious effects of the cephlosporins on the intestinl ecology of mice Drug Reltive Totl count Outgrowth of Outgrowth of Dose cecl of bcteou Cndid Citrobcter wt in cecum lbicns in freundii feces feces in Cefoperzone 4 l t t 1 = = = 25 = = = Ceftzidime 16 = = = t 4 = = = 1 = = = Ceftrixone 8 t = t t 2 1 = 5 = = = = Cefepime 1 = = =b 25 = = =b 6.25 = = =b =, Not significntly different from control mice;, significntly higher thn in control mice;, significntly lower thn in control mice. b The strin of Citrobcterfreundii ws susceptible to cefepime (MIC,.25 p.g/ml), nd therefore these results re difficult to interpret. from 1,7 to 21,7,ug/g of feces (2). In nother study, Silv et l. (24) found lower concentrtions of cefoperzone (<2 to 15,43,ug/g) in the feces of volunteers treted three times with dose of 1 or 2 g of cefoperzone, nd in yet nother study, Giulino et l. (1) found concentrtions of cefoperzone in the feces rnging from 2,727 to 8,84 p,g/g in group of four volunteers receiving 4 g of cefoperzone per dy. Since only 19 to 36% of prenterl dose of cefoperzone is excreted renlly nd considerble biliry excretion of cefoperzone hs been documented (6), high concentrtions of cefoperzone in the feces during prenterl tretment with this drug re not surprising. Arvidsson et l. (2) mesured biliry excretion of ceftrixone during the stedy stte in five helthy mle volunteers. They estimted tht 11 to 65% of the prenterl dose of ceftrixone ws excreted into the biliry trct. In these individuls, ceftrixone concentrtions were mesured in the feces 1 dy fter they hd received totl dose of pproximtely 5 mg of ceftrixone. In two of these subjects, no ceftrixone concentrtions could be detected in the feces, wheres in the other three subjects, the ceftrixone concentrtions rnged from 28 to 75,ug/g. To our knowledge, there is no informtion on the concentrtions of ceftzidime nd cefepime in the feces of humns, but concentrtions of these two drugs in the feces cn be expected to be much lower thn those of cefoperzone nd ceftrixone, becuse these ntibiotics re minly renlly clered. The present study on the effects of cephlosporins on the gstrointestinl ecology in mice shows some similrities with those in humns. For exmple, Giulino et l. (1) found tht tretment of helthy volunteers with cefoperzone led to drmtic increse in the number of yests in the feces, wheres tretment with cefoxitin hd much less effect. The ntibcteril spectr of the two drugs re similr, but the concentrtions of cefoxitin in the feces were much lower thn those of cefoperzone. Tht the coloniztion resistncedisturbing potentil of ntibiotics is cliniclly relevnt hs never been proved conclusively, but decrese of coloniztion resistnce is believed to increse the chnce of overgrowth of the gstrointestinl trct by multiresistnt bcteri or yests. In ptients with incresed susceptibility to infection, intestinl overgrowth with potentilly pthogenic microorgnisms increses the chnce of n infection with those orgnisms, especilly if the mucos of the gstrointestinl trct hs been dmged by cytosttic gents or irrdition. This sitution underscores the importnce of knowing the therpeutic efficcy s well s the coloniztion resistncedisturbing effect of given ntibiotic. A study on both these effects could be expected to yield n indiction of the reltionship between the dosge t which coloniztion resistnce is decresed nd the dosge giving mximum therpeutic efficcy. In vitro cefepime nd ceftrixone hve much lower MICs ginst members of the fmily nterobctericee thn ceftzidime nd cefoperzone (8, 13, 25). We hve observed tht the in vitro order of potency of these four cephlosporins is similr to the in vivo order of potency in n experimentl thigh muscle infection with. coli in irrdited mice (27), indicting tht in therpeutic doses cefepime nd ceftrixone hve less effect on coloniztion resistnce thn ceftzidime nd cefoperzone. Although direct extrpoltion to humns should be done cutiously, our results suggest tht the mrgin of sfety with respect to coloniztion resistnce is greter for cefepime nd ceftrixone thn for ceftzidime nd cefoperzone. ACKNOWLDGMNTS We thnk A. M. Hzekmpvn Dokkum nd. vn Strijen for their technicl ssistnce nd A. Zwindermn from the Deprtment of Medicl Sttistics for sttisticl dvice. This work ws supported by the J. A. Cohen Institute for Rdiopthology nd Rdition Protection. RFRNCS 1. Alestig, K., H. Crlberg, C.. Nord, nd B. Trollfors ffect of cefoperzone on fecl flor. J. Antimicrob. Chemother. 12: Arvidsson, A., G. Alvn, B. Angelin,. Borg, nd C.. Nord Ceftrixone: renl nd biliry excretion nd effect on the colon microflor. J. Antimicrob. Chemother. 1: Brbhiy, R. H., S. T. Forgue, W. C. Shyu,. A. Ppp, nd K. A. Pittmn Highpressure liquid chromtogrphic nlysis of BMY28142 in plsm nd urine. Antimicrob. Agents Chemother. 31: Brry, A. L Procedure for testing ntimicrobil gents in gr medi: theoreticl considertions, p In V. Lorin (ed.), Antibiotics in lbortory medicine, 2nd ed. Willims & Wilkins, Bltimore. 5. Bodey, G. P., V. Finstein, I. Grci, B. Rosenbum, nd Y. Wong ffect of brodspectrum cephlosporins on the microbil flor of recipients. J. Infect. Dis. 148: Brogden, R. N., A. Crmine, R. C. Heel, P. A. Morley, T. M. Speight, nd G. S. Avery Cefoperzone: review of its in vitro ntimicrobil ctivity, phrmcologicl properties nd therpeutic efficcy. Drugs 22: dwrds, A. L Multiple regression nd the nlysis of vrince nd covrince, 2nd ed., p W. H. Freemn & Co., New York. 8. Fss, R. J Comprtive in vitro ctivities of thirdgenertion cephlosporins. Arch. Intern. Med. 143: Forgue, S. T., W. C. Shyu, C. R. Gleson, K. A. Pittmn, nd R. H. Brbhiy Phrmcokinetics of the novel cephlosporin cefepime (BMY28142) in rts nd monkeys. Antimicrob. Agents Chemother. 31: Giulino, M., M. Brz, N. V. Jcobus, nd S. L. Gorbch ffect of brodspectrum prenterl ntibiotics on the composition of intestinl microflor of humns. Antimicrob. Agents Chemother. 31: Holdemn, L. V., nd W.. C. Moore (ed.) Anerobe Downloded from on October 21, 218 by guest

7 982 VAN OGTROP T AL. lbortory mnul, 2nd ed., p. 17. Virgini Polytechnic Institute nd Stte University, Blcksburg. 12. Itoh, K., nd T. Mitsuok Comprison of medi for isoltion of mouse nerobic fecl bcteri. Lb. Anim. 19: Kessler, R.., M. Bies, R.. Buck, D. R. Chisholm, T. A. Pursino, Y. H. Tsi, M. Misiek, K.. Price, nd F. Leitner Comprison of new cephlosporin, BMY 28142, with other brodspectrum,lctm ntibiotics. Antimicrob. Agents Chemother. 27: Khn, N. J., J. A. Bihi, R. F. Schell, J. L. LeFrock, nd S. J. Weber Antimicrobil ctivities of BMY28142, cefbuperzone, nd cefpirmide compred with those of other cephlosporins. Antimicrob. Agents Chemother. 26: Klstersky, J mpiric tretment of infection during grnulocytopeni: comprehensive pproch. Infection 17: Koopmn, J. P., F. G. J. Jnssen, nd J. A. M. Vn Druten The reltion between the intestinl microflor nd intestinl prmeters in mice. Z. Versuchstierkd. 19: Koopmn, J. P., H. M. Kennis, A. M. Stdhouders, nd H. De Boer Selective elimintion of nterobctericee species from the digestive trct in mice nd rts. Z. Versuchstierkd. 26: Koopmn, J. P., G. W. Welling, A. W. M. Huybregts, J. W. M. A. Mullink, nd R. A. Prins Assocition of germfree mice with intestinl microflor. Z. Versuchstierkd. 23: Lee, A., J. Gordon, C.J. Lee, nd R. Dubos The mouse intestinl microflor with emphsis on the strict nerobes. J. xp. Med. 133: ANTIMICROB. AGNTS CHMOTHR. 2. Mullign, M.., D. M. Citron, B. T. McNmr, nd S. M. Finegold Impct of cefoperzone therpy on fecl flor. Antimicrob. Agents Chemother. 22: Pope, L. M., G. T. Cole, M. N. Guentzel, nd L. J. Berry Systemic nd gstrointestinl cndidisis of infnt mice fter intrgstric chllenge. Infect. Immun. 25: Richrds, D. M., nd R. N. Brogden Ceftzidime: review of its ntibcteril ctivity, phrmcokinetic properties nd therpeutic use. Drugs 29: Richrds, D. M., R. C. Heel, R. N. Brogden, T. M. Speight, nd G. S. Avery Ceftrixone: review of its ntibcteril ctivity, phrmcologicl properties nd therpeutic use. Drugs 27: Silv, M., N. A. Cornick, nd S. L. Gorbch Suppression of colonic microflor by cefoperzone nd evlution of the drug s potentil prophylxis in bowel surgery. Antimicrob. Agents Chemother. 33: Tsuji, A., A. Mnitis, M. A. Bertrm, nd L. S. Young In vitro ctivity of BMY28142 in comprison with those of other 1lctm ntimicrobil gents. Antimicrob. Agents Chemother. 27: Vn der WiJ, D., J. M. Berhguisde Vries, nd J.. C. Lekkerkerkvn der Wees Coloniztion resistnce of the digestive trct in conventionl nd ntibiotictreted mice. J. Hyg. 69: Vn Ogtrop, M. L., H. Mttie, H. F. L. Guiot,. vn Strien, A. M. Hzekmpvn Dokkum, nd R. vn Furth Comprtive study of the effects of four cephlosporins ginst scherichi coli in vitro nd in vivo. Antimicrob. Agents Chemother. 34: Downloded from on October 21, 218 by guest