JOURNAL OF CLINICAL MICROBIOLOGY, Sept. 1984, p. 561-5670095-1137/84/090561-07$02.00/0Copyright © 1984, American Society for Microbiology

Influence of Growth Medium on the In Vitro Activities of Second-and Third-Generation Cephalosporins Against Streptococcus faecalis

DANIEL F. SAHM,lt* C. N. BAKER,2 R. N. JONES,3 AND C. THORNSBERRY2Laboratory Program Office, Centers for Disease Control,' and Center for Infectious Diseases, Hospital Infections

Program,2 Atlanta, Georgia 30333, and Kaiser Foundation Hospitals, Portland, Oregon 970153

Received 15 February 1984/Accepted 30 May 1984

The influence of culture medium of the MICs of eight cephalosporins for 45 strains of Streptococcusfaecaliswas investigated. The MICs of cephalothin, cefamandole, and cefoperazone were not substantially influencedby the type of culture medium used. In contrast, MICs of cefuroxime, ceftizoxime, cefotaxime, cefmenoxime,and ceftriaxone varied markedly with both the commercial brand and the blood content of the broth used. Theuse of Mueller-Hinton broths (from Oxoid Ltd., GIBCO Diagnostics, and Difco Laboratories) supplementedwith 5% lysed sheep blood frequently resulted in MICs that were -16 times lower than the MICs obtained withthese same broths without blood. Similar, but less marked, patterns were observed when supplemented andunsupplemented brain heart infusion and Sceptor broths were used. The influence of the broth on MICssuggests a complex interaction between some cephalosporins, medium components, and organisms. Thecephalosporins that were affected by media share an identical moiety at the 7-acyl position (cefuroxime isslightly different), but this structure is not shared by those cephalosporins that were not affected. Thiscommonality in structure at the 7-acyl position may be partially responsible for the observed results.

Numerous in vitro studies have demonstrated that entero-cocci are generally resistant to newer second- and third-generation cephalosporins such as cefamandole (25), cefur-oxime (20, 21), cefoperazone (14), ceftizoxime (7),cefnienoxime (1, 9), ceftriaxone (5, 28), and cefotaxime (8,13). However, an investigation done in our laboratory hasshown that the susceptibility of Streptococcus faecalis toone of these caphalosporins (cefotaxime) may be substantial-ly influenced by the commercial brand and the blood contentof the Mueller-Hinton (MH) agar medium used for suscepti-bility testing by the standard disk diffusion method (24). Thetesting of several S. faecalis strains against cefotaxime oftenresulted in zone sizes indicative of susceptibility whenblood-supplemented MH agar was used, whereas zone sizesindicative of resistance usually resulted when unsupplement-ed MH agar was employed. In addition, more S. faecalisstrains appeared susceptible to cefotaxime with the use ofblood-supplemented MH agars from Oxoid Ltd. (KC Biolog-ical Inc., Lenexa, Kans.) and BBL Microbiology Systems(Cockeysville, Md.) than was apparent with the use ofsupplemented MH agar from other sources. This medium-associated major discrepancy is of obvious clinical impor-tance as well as of major concern with respect to thestandardization of in vitro susceptibility testing.To aid in the understanding of the mechanism(s) responsi-

ble for this major discrepancy, we initiated further studiesthat might clarify and define the in vitro situations in whichthis phenomenon was observed. The purpose of this presentstudy was twofold. First, we sought to determine whetherthe medium-associated major discrepancy observed with S.faecalis and cefotaxinie by the disk diffusion method couldalso be demonstrated by the microdilution broth method;second, we sought to determine whether other second- andthird-generation cephalosporins would also demonstrate this

* Corresponding author.t Present address: Clinical Microbiology Laboratories, Universi-

ty of Chicago Medical Center, Chicago, IL 60637.

561

medium-associated discrepancy when tested against severalS. faecalis strains.

MATERIALS AND METHODS

For this investigation, 45 strains of S. faecalis were testedagainst various first-, second-, and third-generation cephalo-sporins by the microdilution broth method. These tests weredone with broth media of various types and from differentcommercial sources (see Table 1); each broth was testedwith and without 5% lysed sheep blood.

Organisms. Forty-five S. faecalis strains used for thisinvestigation were kindly supplied by Richard Facklarn,Reference Bacteriology Section, Respiratory and SpecialPathogens Branch, Centers for Disease Control, Atlanta,Ga. Control bacterial strains for susceptibility testing includ-ed Staphylococcus aureus ATCC 25923, Escherichia coliATCC 25922, and Pseudomonas aeruginosa ATCC 27853.

Media. MH broth media were supplied by Oxoid; GIBCODiagnostics, Madison, Wis.; and Difco Laboratories, De-troit, Mich. The brain heart infusion (BHI) broth used in thisstudy was also supplied by Difco. Sceptor gram-positivebroth was graciously donated by George Evans, BBL Micro-biology Systems. All broth media were prepared, with orwithout the addition of 5% lysed sheep blood, according tothe supplier's directions. Lysed sheep blood was preparedby previously described methods (27). In all, 10 differenttypes of broth media were used for the testing of the variouscephalosporins against the S. faecalis strains. These mediaincluded GIBCO MH broth with and without 5% lysed sheepblood, Oxoid MH broth with and without 5% lysed sheepblood, Difco MH broth with and without 5% lysed sheepblood, BHI broth with and without 5% lysed sheep blood,and Sceptor gram-positive broth (Sceptor broth) with andwithout 5% lysed sheep blood.

Antibiotics. The cephalosporins tested in this study werereceived in powdered form and tested at concentrations of 2,4, 8, 16, 32, and 64 ,ug/ml in the various types of broth media.

Vol. 20, No. 3

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

562 SAHM ET AL.

The cephalosporins tested included the following: cepha-lothin and cefamandole (Eli Lilly & Co., Indianapolis, Ind.),cefuroxime (Glaxo Research Labs, Research Triangle Park,N.C.), cefoperazone (Pfizer Laboratories Division, NewYork, N.Y.), cefotaxime (Hoechst-Roussel PharmaceuticalsInc., Somerville, N.J.), ceftizoxime (SmithKline BeckmanCorp., Philadelphia, Pa.), cefmenoxime (Abbott Labora-tories, North Chicago, Ill.), and ceftriaxone (Hoffman-LaRoche Inc., Nutley, N.J.).

Susceptibility testing. Susceptibility testing of the S. faeca-lis and control strains with the various cephalosporins was

performed by the microdilution broth method recommendedby the National Committee for Clinical Laboratory Stan-dards (16). Because specific breakpoints have not beenestablished for some of these cephalosporins, the resultswere not tabulated in a categorical format (i.e., resistant,intermediate, susceptible), but rather were analyzed strictlyon the basis of the observed MICs.

RESULTS

We first analyzed the data by determining the number ofS. faecalis strains for which we obtained an MIC value of s8,ug/ml with each cephalosporin in each type of broth medium(Table 1). The number of strains with MICs of .8 Vig/ml forcephalothin, cefamandole, and cefoperazone was generallymuch less than for the other cephalosporins, and the numberof strains for which the MICs of cephalothin, cefamandole,and cefoperazone were .8 ,ug/ml did not vary extensivelywith the brand or blood content of the broth used. The use ofunsupplemented Oxoid MH resulted in none of the 45 S.faecalis strains having a MIC of .8 Fg/ml for any of thecephalosporins tested. In comparison to the results obtainedwith cephalothin, cefamandole, and cefoperazone, the MICsof cefuroxime, ceftizoxime, cefotaxime, ceftriaxone, andcefmenoxime were .8 pg/ml for a much larger number of S.faecalis strains; in addition, the number of strains for whichthe MIC values were .8 .g/ml was closely related to thebrand and to the blood content of the broth medium that wasemployed. Results obtained with the unsupplemented brothsrevealed some variations in the number of S. faecalis strainsfor which MIC values were .8 ,ug/ml. When unsupplement-ed Oxoid MH broth was used, the MICs of cefuroxime,ceftizoxime, cefotaxime, ceftriaxone, and cefmenoximewere never .8 ,ug/ml for any of the S. faecalis strains.However, the MICs of these five cephalosporins were .8,ug/ml for a much larger number of S. faecalis strains whenunsupplemented BHI broth was used. The number of strainsfor which the MICs were .8 ,g/ml did not vary appreciablywith the use of unsupplemented GIBCO MH, Difco MH, orSceptor broths.Comparisons of the MIC results obtained with the use of

blood-supplemented broths did reveal brand-to-brand varia-tions in the number of strains for which the MICs ofcefuroxime, ceftizoxime, cefotaxime, ceftriaxone, and cef-menoxime were .8 ,ug/ml (Table 1). However, the mostnotable pattern that was observed was the remarkableincrease, with the use of blood-supplemented broths, in thenumber of S. faecalis strains for which the MICs of thesefive particular cephalosporins were <8 p.g/ml (Table 1). Foreach brand of broth used, the number of strains for which theMICs of cefuroxime, ceftizoxime, cefotaxime, ceftriaxone,and cefmenoxime were .8 ,ug/ml was greater with the blood-supplemented broth than with the unsupplemented counter-part; a single exception was noted with cefuroxime tested in

TABLE 1. Number of S. faecalis strains with MICs of <8 jig/mlwhen tested in various broths with and without blood"

No. of strains with MICs of -8 pug/ml

AntibioticOxoid GIBCO Difco BHI SceptorMH MH MH

Bb NBC B NB B NB B NB B NB

Cephalothind 2 0 3 2 1 3 4 4 3 3Cefamandolee 3 0 3 1 2 2 4 4 2 2Cefoperazonef 5 0 2 1 2 1 2 2 5 3Cefuroximee 22 0 6 1 3 3 16 7 5 3Ceftizoximef 24 0 16 2 16 3 22 14 15 5Cefotaximef 24 0 15 3 13 4 21 15 14 8Ceftriaxonef 22 0 9 2 9 4 18 8 10 5Cefmenoximef 28 0 18 2 18 4 23 14 17 8

a A total of 45 strains were tested.b B, Presence of 5% lysed sheep blood.' NB, Absence of 5% lysed sheep blood.d First-generation cephalosporin.eSecond-generation cephalosporin.f Third-generation cephalosporin.

Difco MH broth. This blood-associated phenomenon wasmost pronounced with Oxoid MH broth; approximately one-half of the S. faecalis strains tested demonstrated MICvalues that were .8 pg/ml in the presence of lysed sheepblood, whereas none of the strains demonstrated MIC values.8 ,g/ml in the absence of blood. Although not as drastic, asimilar pattern was observed when each of the other fourblood-supplemented brands was compared with its unsup-plemented counterpart. Therefore, the initial analysis of thedata provided evidence that both the presence of blood andthe brand of broth medium used may influence the in vitroactivity of cefuroxime, ceftizoxime, cefotaxime, ceftriax-one, and cefmenoxime against S. faecalis.To determine the extent to which lysed sheep blood and

the brand of broth medium used influenced the MICs of thevarious cephalosporins, the number of strains for whichthere was a 28-fold and a -16-fold decrease in MIC wastabulated for each type of broth used (Table 2). Cephalothin,cefamandole, and cefoperazone each had a >8-fold decreasein their MICs for a substantial number of S. faecalis strainswhen the use of blood-supplemented Oxoid MH was com-pared with the use of unsupplemented Oxoid MH; this dropin MICs was not observed with the use of any of the otherbrands of broth. The use of blood-supplemented Oxoid MHbroth resulted in only a few strains for which the MICs of

TABLE 2. Number of S. faecalis strains with -8-fold and -16-fold lower MIC when tested in blood-supplemented media than

when tested in same media without blood'No. of strains with lower MIC

AntibioticOxoid GIBCO Difco BHI SceptorMH MH MH

28 .16 >8 .16 .8 .16 .8 .16 .8 .16

Cephalothin 18 2 0 0 0 0 0 0 0 0Cefamandole 13 3 0 0 0 0 0 0 0 0Cefoperazone 23 4 0 0 0 0 0 0 0 0Cefuroxime 26 20 9 4 5 1 4 2 2 0Ceftizoxime 25 24 18 9 14 8 5 2 9 6Cefotaxime 27 24 14 12 11 9 4 3 8 5Ceftriaxone 26 21 20 6 10 2 7 3 1 0Cefmenoxime 28 24 20 15 11 7 7 3 4 0

a A total of 45 strains were tested.

J. CLIN. MICROBIOL.

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

GROWTH MEDIUM AND IN VITRO CEPHALOSPORIN ACTIVITY 563

cephalothin, cefamandole, and cefoperazone decreased.16-fold. In contrast, the use of each of the five brands ofbroth that were supplemented with blood resulted in a .8-fold drop in the MICs of cefuroxime, ceftizoxime, cefotax-ime, ceftriaxone, and cefmenoxime for several S. faecalisstrains. However, the number of strains for which -8-folddecreases in the MICs of these five cephalosporins wereobserved varied with the brand of broth. The use of blood-supplemented Oxoid MH broth resulted in the greatestnumber of S. faecalis strains for which 28-fold and .16-folddecreases in MICs were observed. Blood-supplementedGIBCO MH and Difco MH broths, respectively, demon-strated the second and third highest number of strains forwhich the MICs of cefuroxime, ceftizoxime, cefotaxime,ceftriaxone, and cefmenoxime decreased -8-fold and .16-fold. The use of blood-supplemented BHI and Sceptorbroths resulted in the lowest number of S. faecalis strains forwhich these substantial MIC decreases were observed.Because cefuroxime, ceftizoxime, cefotaxime, ceftriax-

one, and cefmenoxime were the cephalosporins with activi-ties that were most affected by the brand and blood contentof the broth used, the distribution of the number of strainsthat were inhibited by each cephalosporin concentrationtested was tabulated according to the broth type employed(Tables 3 through 7).The data in Table 3 aptly demonstrate the remarkable

influence that the brand and blood content of the broth usedhad on the MIC of cefotaxime for the S. faecalis strains. Theuse of unsupplemented Oxoid MH, GIBCO MH, and DifcoMH broths resulted in the MIC of cefotaxime being .64 ,ug/ml for the vast majority of S. faecalis strains. In strikingcontrast, the use of these same three broths supplementedwith blood resulted in the MIC of cefotaxime being .64 FLg/ml for only approximately one-half of the S. faecalis strainstested. The use of these three blood-supplemented brothsalso resulted in a more even distribution of the number ofstrains inhibited by each concentration of cefotaxime testedthan that which resulted with the use of their unsupplement-ed counterparts. For supplemented and unsupplementedBHI and Sceptor broths, the number of strains for which thecefotaxime MIC of .64 ,ug/ml was obtained was compara-ble, as was the distribution of the number of strains inhibitedby each cefotaxime concentration tested. However, thecefotaxime MIC of s2 ,ug/ml was observed for more strainswhen BHI broth was used than when Sceptor broth wasused.The MICs of ceftizoxime (Table 4), cefuroxime (Table 5),

TABLE 3. Number of S. faecalis strains that gave the designatedMIC value when tested against cefotaxime in various broths with

and without bloodaNo. of strains with designated MIC

MIC (,ug/ml)Oxoid GIBCO Difco BHI SceptorMH MH MH

Bb NBC B NB B NB B NB B NB

s2 17 0 3 1 4 2 16 11 7 34 3 0 7 1 3 0 2 3 3 18 4 0 5 1 6 2 3 1 4 4

16 1 0 3 0 2 0 1 3 4 132 2 0 7 1 5 0 4 3 2 3

.64 18 45 20 41 25 41 19 24 25 33a A total of 45 strains were tested.b B, Presence of 5% lysed sheep blood.c NB, Absence of 5% lysed sheep blood.

TABLE 4. Number of S. faecalis strains that gave the designatedMIC value when tested against ceftizoxime in various broths with

and without blood'No. of strains with designated MIC

MIC (,ug/ml) Oxoid GIBCO Difco BHI SceptorMH mHi MH

Bb NBC B NB B NB B NB B NB

'2 15 0 4 2 1 0 11 9 5 24 6 0 5 0 9 2 8 4 6 18 3 0 7 0 6 1 3 1 4 2

16 1 0 5 0 2 0 1 4 3 332 2 0 4 1 2 1 1 1 1 1

.64 18 45 20 42 25 41 21 26 26 36

"A total of 45 strains were tested.bB, Presence of 5% lysed sheep blood.C NB, Absence of 5% lysed sheep blood.

cefmenoxime (Table 6), and ceftriaxone (Table 7), obtainedwith each of the different broths for the S. faecalis strains,were similar to those shown for cefotaxime in Table 3. Thatis, the testing of these cephalosporins in unsupplementedOxoid MH, GIBCO MH, and Difco MH broths resulted in asubstantially greater number of S. faectlis strains for whichthe MICs were -64 ,ug/ml than resulted when these threebroths supplemented with blood were used (Tables 4 through7). In addition, the use of these three brands of blood-supplemented broths resulted in a more even distribution ofthe number of S. faecalis strains that were inhibited by eachcephalosporin concentration tested than the distribution thatresulted from the use of these brands without blood supple-mentation. The use of blood-supplemented and unsupple-mented BHI and Sceptor broths for the testing of ceftizox-ime (Table 4), cefuroxime (Table 5), ceftriaxone (Table 6),and cefmenoxime (Table 7) did not result in an extensivedisparity in the distribution of the number of S. faecalisstrains inhibited by each cephalosporin concentration tested.However, the number of strains for which MICs of ceftizox-ime (Table 4) and cefmenoxime (Table 6) were s2 ,ug/ml wasgreater with BHI broths than with Sceptor broths.

DISCUSSION

The results of this investigation have demonstrated thatthe activity of cefotaxime against numerous strains of S.faecalis, as determined by the microdilution broth method,varied extensively with the blood content and the branid(Oxoid, GIBCO, Difco, BHI, or Sceptor) of broth mediumused for testing (Tables 1 through 3). Cefotaxime was leastactive against S. faecalis strains when Oxoid MH, GIBCOMH, and Difco MH broths without blood supplement wereused. In comparison, the use of BHI broth without bloodand, to a lesser extent, Sceptor broth without blood resultedin a greater activity of cefotaxime against the strains of S.faecalis than resulted with the use of the MH broths withoutblood. Although the activity of cefotaxime did vary with thebrand of broth used for susceptibility testing, the mostnotable variations in cefotaxime activity were associatedwith the content of 5% lysed sheep blood in the broth media.For each brand of broth used, cefotaxime was more activeagainst a greater number of strains when testing was done inbroth containing blood than when testing was done in theunsupplemented broths. These findings are in agreementwith our earlier study that demonstrated that the use ofblood-supplemented MH agar in the disk diffusion test

VOL. 20, 1984

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

564 SAHM ET AL.

TABLE 5. Number of S. faeccalis strains that give the designated MIC value when tested against cefuroxime in various broths with andwithout blood"

No. of strains with designated MICMIC Oxoid MH GIBCO MH Difco MH BHI Sceptor

(,ug/ml)B" NB' B NB B NB B NB B NB

.2 4 0 2 1 2 2 2 2 3 24 8 0 1 0 0 0 9 3 0 08 10 0 3 0 1 1 5 2 2 1

16 4 0 6 0 8 0 5 6 3 132 2 1 12 2 7 0 3 3 5 5

'64 17 44 21 42 27 42 21 29 32 36

A total of 45 strains were tested.B. Presence of 5% lysed sheep blood.NB, Absence of 5% lysed sheep blood.

resulted in a greater number of S. faecalis strains beingsusceptible to cefotaxime than were susceptible when agar

media without blood were used. In addition, these resultsconfirm our previous finding that blood-supplemented OxoidMH agar was the medium most frequently associated withthis phenomenon (24).Also of interest in this present study is the observation

that other cephalosporins were affected, in a manner similarto cefotaxime, by the type of broth medium used forsusceptibility testing (Tables 1, 2, and 4 through 7). Ceftizox-ime (Table 4), cefuroxime (Table 5), cefmenoxime (Table 6),and ceftriaxone (Table 7) were all least active against the S.faecalis strains when tested in Oxoid MH, GIBCO MH, andDifco MH broths without blood. As with cefotaxime, theactivities of these cephalosporins were greater when testedin BHI and Sceptor broths without blood than when tested inMH broths without blood. As was also noted with cefotax-ime, the activities of ceftizoxime, cefuroxime, cefmenoxime,and ceftriaxone were most markedly influenced by the pres-

ence of 5% lysed sheep blood in the broth media. Each ofthese cephalosporins was more active against a greaternumber of S. faecalis strains when blood-supplementedbroths were used than when unsupplemented broths were

used. In contrast, the activities of cephalothin, cefamandole,and cefoperazone were not extensively influenced by thetype of broth medium used for susceptibility testing (Tables1 and 2).Our observation that the in vitro activities of certain

cephalosporins against S. faecalis are markedly influencedby the growth medium used presents new and confoundinginformation with respect to the susceptibility testing of S.faecalis with these cephalosporins. Although earlier studiesby Washington and Yu (29) and Pursiano et al. (23) haveshown that the type of growth medium used may influence

the susceptibility testing results obtained with cephalospo-rins, the results of these investigations cannot be accuratelycompared with those of this present investigation, for vari-ous reasons. In those studies the effect of blood supplemen-tation was not studied, the test organisms did not include S.faecalis, and, obviously, the newer second- and third-gener-ation cephalosporins were not available for investigation. Anearlier study by Brenner and Sherris (3) did report that thetype of growth medium used markedly affected the activityof cephalothin against enterococci, but cephalothin was theonly cephalosporin tested and only one strain of enterococ-cus (species not given) was tested. In more recent studies ofthe in vitro activities of newer cephalosporins such as

cefotaxime (19), ceftriaxone (2, 5), cefuroxime (20), andceftizoxime (7), no significant effects of media on the activi-ties of these cephalosporins were demonstrated. However,these studies were concerned with the activity of cephalo-sporins against various species of gram-negative bacilli anddid not report on the effect of media on susceptibility testingresults obtained with strains of S. faecalis. In studies with S.aureus, Fong et al. (6) and Peterson et al. (22) have reportedthat the in vitro activities of cefazolin, cephalothin, andcephaloridine were influenced by the growth medium usedfor susceptibility testing. These investigators thought thatthe differences were due to medium enhancement of I-

lactamase production (6); if this is true, the significance oftheir results with respect to ours obtained with S. faecalis,an organism that does not usually produce 1-lactamase, isdoubtful.By demonstrating that the in vitro activities of certain

cephalosporins (cefotaxime, ceftizoxime, cefuroxime, cef-menoxime, and ceftriaxone) against S. faecalis vary marked-ly with the type of growth medium used for susceptibilitytesting, we have raised obvious concerns about the medium

TABLE 6. Number of S. faecalis strains that gave the designated MIC value when tested against cefmenoxime in various broth with andwithout blood"

No. of strains with designated MICMIC Oxoid MH GIBCO MH Difco MH BHI Sceptor

(,ug/ml)Bb NB' B NB B NB B NB B NB

'2 14 0 5 2 4 2 13 11 4 24 1 0 8 0 5 1 3 2 7 58 13 0 5 0 9 1 7 3 6 1

16 0 0 5 0 2 1 4 2 3 432 3 1 6 0 7 5 3 3 2 7

.64 14 44 16 43 18 35 15 24 23 26

A total of 45 strains were used.B, Presence of 5% lysed sheep blood.NB, Absence of 5% lysed sheep blood.

J. CLIN. MICROBIOL.

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

GROWTH MEDIUM AND IN VITRO CEPHALOSPORIN ACTIVITY 565

TABLE 7. Number of S. faecalis strains that gave the designatedMIC value when tested against ceftriaxone in various broths with

and without bloodNo. of strains with designated MIC

MIC Oxoid GIBCO Difco BHI Sceptor(p.gIml) MH MH MH

Bb NB' B NB B NB B NB B NB

<2 3 0 3 1 2 2 2 3 3 34 10 0 4 0 2 1 6 1 4 18 9 0 2 1 5 1 10 4 3 116 4 0 13 0 11 1 4 6 9 532 4 0 6 0 3 3 5 4 3 5

.64 15 45 17 43 22 37 18 27 23 30

A total of 45 strains were tested.B, Presence of 5% lysed sheep blood.NB, Absence of 5% lysed sheep blood.

of choice for testing these particular cephalosporins againstS. faecalis and the true effectiveness of these antibiotics intreating S. faecalis infections. An understanding of mecha-nism(s) involved in this observed phenomenon would beuseful in resolving these concerns. Although this mecha-nism(s) cannot be delineated on the basis of our resultsalone, certain patterns have been observed that may beuseful for the understanding and the eventual resolution ofthis problem. Classically, the activity of antibiotics is depen-dent upon inherent characteristics of the organisms and thedrugs and is influenced by the environment (media or tissue)in which they interact. The explanation for our unusualresults will probably encompass all three of these factors.Inherent characteristics of the organism (S. faecalis), theantibiotics (cefotaxime, ceftizoxime, cefuroxime, cefmenox-ime, ceftriaxone), and the various environments (differentbrands of blood-supplemented and unsupplemented broth)all appear to be important factors that contributed to theresults obtained in this investigation.The possibility that certain singular inherent characteris-

tics of S. faecalis are at least partially responsible for theresults observed is supported by observations made both byus and by other investigators. As part of the present study,using the various types of broth media, we tested S. aureus,E. coli, and P. aeruginosa control strains against eachcephalosporin. The results demonstrated that the MICs ofcefotaxime, cefmenoxime, cefuroxime, ceftizoxime, andceftriaxone for these control strains did not vary frommedium to medium by more than one twofold dilution (datanot shown). Similarly, results reported by other investiga-tors have failed to demonstrate any effect of growth mediaon the in vitro effectiveness of these cephalosporins againstvarious gram-negative bacilli (2, 5, 7, 19, 20). Therefore,because the phenomenon observed in this present investiga-tion with S. faecalis has not been observed with otherbacterial species, it seems likely that inherent characteristicsof the S.faecalis strains are important contributing factors tothe mechanism(s) involved. Another indication that certainsingular characteristics of S. faecalis may be involved hasbeen provided by Muytjens and van der Ros-van de Repe(14). Although no comparative data were presented, theseinvestigators did discuss the potential for MICs for S.faecalis to vary up to 16-fold with different media used fortesting.

In addition to characteristics of the microorganisms (S.faecalis), certain broth-medium components must also signifi-cantly affect the in vitro activities of cefotaxime, ceftizox-

ime, cefuroxime, cefmenoxime, and ceftriaxone. The de-creased activities of these particular cephalosporins inunsupplemented Oxoid MH, GIBCO MH, and Difco MHbroths, as compared with activities in unsupplemented BHIand Sceptor broths, indicate that broth base constituentsmay be factors contributing to the observed results. Inaddition, the increase in activity associated with bloodsupplementation was more pronounced with Oxoid MH,GIBCO MH, and Difco MH media than with BHI or Sceptorbroths, further indicating that differences in broth baseconstituents are factors contributing to the observed phe-nomenon (Table 2). However, the complex and sometimesundefined nature of the components of these bases and theproprietary interests of the manufacturers concerning theseconstituents make drawing conclusions about the specificcomponent(s) responsible for these unusual results bothdifficult and highly speculative.Although differences in the MICs obtained with the use of

various unsupplemented broths did occur, more strikingdifferences in the MICs of the implicated cephalosporinswere observed when results obtained with the use of blood-supplemented broths were compared with the results ob-tained with unsupplemented broths (Table 1). The use ofblood-supplemented broth resulted in an increased activityof cefotaxime, cefuroxime, ceftizoxime, cefmenoxime, andceftriaxone against the S. faecalis strains. The mechanismby which blood or blood components may influence the invitro activity of these particular cephalosporins is not under-stood. Wright and Frogge (30) have demonstrated that in thepresence of lysed whole blood, cephalosporins having a 3-acetoxymethyl group may be hydrolyzed by acylases to theirdesacetyl derivative, resulting in various effects on the cepha-losporins' activities. Results obtained in this present studyas well as those of other studies offer contradictions to thisexplanation. In the present study, the activity of cepha-lothin, a cephalosporin containing a 3-acetoxymethyl group,was not nearly as affected by the presence of blood as werethe activities of cefuroxime, cefmenoxime, ceftizoxime, andceftriaxone, none of which contain a 3-acetoxymethyl group(Table 2, Fig. 1). In addition, the activity of ceftizoxime, acephalosporin that is naturally deacetylated at the 3 position,was affected by the presence of blood to an extent compara-ble to cefotaxime, a cephalosporin that contains a 3-acetoxy-methyl group (Fig. 1). Furthermore, Jones et al. (12) andNeu (18) have shown that the in vitro activities of desacetyl-cefotaxime and cefotaxime against S. faecalis are compara-ble. Deacetylation, therefore, does not appear to be themechanism by which lysed sheep blood affected the in vitroactivities of cefotaxime, cefuroxime, ceftizoxime, cefmen-oxime, and ceftriaxone.In addition to the factors associated with the S. faecalis

strains and the broth and blood components that contributeto the in vitro phenomenon observed in this investigation,certain characteristics of the particular cephalosporins arealso pertinent to the understanding of the underlying mecha-nism(s) involved. The activities of cephalothin, cefaman-dole, and cefoperazone were much less drastically influ-enced by the type of broth used than were the activities ofcefotaxime, ceftizoxime, cefuroxime, cefmenoxime, andceftriaxone (Table 1). Although no absolute explanation forthese results can be offered at this time, a closer analysis ofthe cephalosporins most affected by the broth medium usedindicates that certain structural characteristics of these par-ticular antibiotics may contribute significantly to the ob-served results. Cefuroxime, cefmenoxime, ceftizoxime, ce-fotaxime, and ceftriaxone all share common structures in

VOL. 20, 1984

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

566 SAHM ET AL.

HR2 , 'N s

coo e

Cephalosporin Rl R2

Cephalothin

Cefamandole

Cefoperazone

Cefuroxime*

Cefmenoxime*

Ceftizoxime

0

CH20CCH3

N-N

CH2-S

CH3

N-N

11/ \\- CH2-S N I

N

CH3

0

CH20CNH2

N-N

-CH2-S / \NN""CH3

-H

0

Cefotaxime* -CH20CCH3

CeftriaxoneCH3 eN

- CH2-S A

Qs-CH2

OH

HO Q< CH-

\/ NH

C=O

tNXO

C2H5

,0-CH3N

0 1 1

.O-CH3N

I'AS C

H2N s

N,-CH3

N -I'I7

N,O-CH3N3

H2N S

/O-CH3N

ATN

FIG. 1. Cephalosporins as structural configurations. The aster-isks indicate cephalosporins with activities most affected by the typeof broth medium used. RI is the 3-position moiety, and R2 is the 7-acyl moiety.

their R2 (7-acyl position) moieties (Fig. 1). Specifically, eachof these five cephalosporins contains a methoxyimino group

at the 7-acyl position that bridges the cephalosporin nucleuswith a heterocyclic ring. For cefmenoxime, ceftizoxime,cefotaxime, and ceftriaxone the heterocyclic structure is an

identical aminothiazolyl ring. Although cefuroxime also con-

tains the methoxyimino group at the 7-acyl position, itsheterocyclic ring (a furyl moiety) differs slightly from thoseof the other four implicated cephalosporins. Whereas thesefive cephalosporins share common structural features intheir R2 moieties, none of them share a common structure intheir R1 (3-position) moiety. Additionally, the cephalospo-rins least affected by the type of broth used (cephalothin,cefamandole, cefoperazone) are strikingly different in thestructure of their R2 moieties from the other five cephalospo-rins (Fig. 1). It is possible that the similarities in structure of

the R2 moieties of cefuroxime, cefmenoxime, ceftizoxime,cefotaxime, and ceftriaxone may be merely coincidentalwith respect to being the cephalosporins most affected by thetype of broth medium used for susceptibility testing. Howev-er, another interesting, but speculative, explanation for thisin vitro phenomenon may also be proposed. Georgopapada-kou and Liu (10) have demonstrated that a strain of S.faecalis will contain at least six different penicillin-bindingproteins, and they have suggested that the resistance of S.faecalis to cephalosporins may be due to a lack of affinitybetween these penicillin-binding proteins and cephalospo-rins (11). This proposed mechanism of resistance to cephalo-sporins, specifically ceftizoxime, has also been put forth byNeu (17). In consideration of this possible mode of resist-ance, it is of interest that the 7-acyl moiety (R2 in Fig. 1) isthought to be a major determinant of cephalosporin potencyand spectrum of activity (4). More specifically, Nakano (15)has suggested that ceftizoxime's aminothiazolyl ring at the 7-acyl position (R2 moiety in Fig. 1), a structure shared bycefotaxime, cefmenoxime, and ceftriaxone, provides theaffinity for penicillin-binding proteins. In light of theseobservations, it is possible that certain broth and bloodcomponents interact with the heterocyclic ring at the 7-acylposition of the implicated cephalosporins and alter thepotency of these drugs by influencing their affinity for thepenicillin-binding protein of S. faecalis, thus resulting in theobserved medium-associated variabilities in the activities ofthese cephalosporins. Although speculative, our resultscould support such a hypothesis, and additional studies areplanned to further investigate the possibility that this mecha-nism may be the cause of the observed phenomenon.These results raise two major and closely related practical

questions. Which of these in vitro results is the true correctone for predicting therapeutic efficacy? Do these cephalo-sporins have any in vivo activity when used to treat patientswith S. faecalis infections? Obviously, the first questioncannot be answered until the second is answered. Althoughthere are some indications that these cephalosporins may beefficacious for treating certain enterococcal infections (26),data concerning the true efficacies of these drugs for entero-coccal infections are inconclusive. If such trials are done,the S. faecalis isolates should be tested for susceptibility inMH broths with and without blood to determine whichresults correlate with clinical efficacy. A third question thatmay be asked is, do other streptococci, enterococcal ornonenterococcal, show this phenomenon? In our previousstudy, we showed that the medium-associated discrepancieswere obtained most often with S.faecalis strain and also, butto a much lesser extent, with Streptococcus faecium strains(24). The medium-associated phenomenon was not observedwith strains of Streptococcus bovis.

In summary, we have demonstrated that the in vitroactivity (as measured by MICs) of certain second- and third-generation cephalosporins (cefuroxime, cefotaxime, cefti-zoxime, cefmenoxime, and ceftriaxone) on S. faiecalis ismarkedly affected by the medium in which the tests aredone, whereas other cephalosporins (cephalothin, cefaman-dole, and cefoperazone) are affected very little. The effectsof media could be correlated with the kind of broth (MH,BHI, Sceptor) used and, to a lesser degree, with the brand ofMH broth used. The most marked effects, however, wereassociated with the addition of blood to the base media. Wehave postulated that the reason for this phenomenon may beinteractions between medium components and the heterocy-clic ring at the 7-acyl position of the affected cephalosporinsthat alter the cephalosporins' activity by influencing their

J.CLIN. MICROBIOL.

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

GROWTH MEDIUM AND IN VITRO CEPHALOSPORIN ACTIVITY 567

affinity for penicillin-binding proteins. Correlations of thisphenomenon with therapy of S. faecalis infections mustawait the results of clinical studies.

LITERATURE CITED

1. Ahonkai, V. I., C. E. Cherubin, M. A. Shulman, and V.Bancroft. 1982. Comparative in vitro activities of cefenoxime(SCE-1365) and new cephalosporin derivatives of clinical utility.Antimicrob. Agents Chemother. 21:999-1002.

2. Angehrn, P., P. J. Probst, R. Reiner, and R. L. Then. 1980. Ro13-9904, a long-acting broad-spectrum cephalosporin: in vitroand in vivo studies. Antimicrob. Agents Chemother. 18:913-921.

3. Brenner, V. C., and J. C. Sherris. 1972. Influence of differentmedia and bloods on the results of diffusion antibiotic suscepti-bilty tests. Antimicrob. Agents Chemother. 1:116-122.

4. Dunn, G. L. 1982. Ceftizoxime and other third-generationcephalosporins: structure-activity relationships. J. Antimicrob.Chemother. 10(Suppl. C):1-10.

5. Eickoff, T. C., and J. Ehret. 1981. Comparative in vitro studiesof Ro 13-9904, a new cephalosporin derivative. Antimicrob.Agents Chemother. 19:435-442.

6. Fong, I. W., E. R. Engelking, and W. M. M. Kirby. 1976.Relative inactivation by Staphylococcus aureus of eight cepha-losporin antibiotics. Antimicrob. Agents Chemother. 9:939-944.

7. Fu, K. P., and H. C. Neu. 1980. Antibacterial activity ofceftizoxime, a P-lactamase stable cephalosporin. Antimicrob.Agents Chemother. 17:583-590.

8. Fuchs, P. C., A. L. Barry, C. Thornsberry, R. N. Jones, T. L.Gavan, E. H. Gerlach, and H. M. Sommers. 1980. Cefotaxime:in vitro activity and tentative interpretive standards for disksusceptibility testing. Antimicrob. Agents Chemother. 18:88-93.

9. Fuchs, P. C., R. N. Jones, C. Thornsberry, A. L. Barry, E. H.Gerlach, and H. M. Sommers. 1981. Cefmenoxime (SCE-1365),a new cephalosporin: in vitro activity, comparison with otherantimicrobial agents, beta-lactamase stability, and disk diffusiontesting with tentative interpretive criteria. Antimicrob. AgentsChemother. 20:747-759.

10. Georgopapadakou, N. H., and F. Y. Liu. 1980. Penicillin-bindingproteins in bacteria. Antimicrob. Agents Chemother. 18:148-157.

11. Georgopapadakou, N. H., and F. Y. Liu. 1980. Binding of -lactam antibiotics to penicillin-binding proteins of Staphylcoc-cus aureus and Streptococcusfaecalis; relation to antibacterialactivity. Antimicrob. Agents Chemother. 18:834-836.

12. Jones, R. N., A. L. Barry, and C. Thornsberry. 1982. Antimicro-bial activity of desacetylcefotaxime alone and in combinationwith cefotaxime: evidence of synergy. Rev. Infect. Dis.4(Suppl.):S366-S373.

13. Jones, R. N., and C. Thornsberry. 1982. Cefotaxime: a review ofin vitro antimicrobial properties and spectrum of activity. Rev.Infect. Dis. 4(Suppl.):S300-S315.

14. Muytjens, H. L., and J. van der Ros-van de Repe. 1982.

Comparative activities of 13 ,-lactam antibiotics. Antimicrob.Agents Chemother. 21:925-934.

15. Nakano, H. 1981. Structure-activity relationships related toceftizoxime (FJ 749). Med. Res. Rev. 1:127-157.

16. National Committee for Clinical Laboratory Standards. 1983.Standard methods for dilution antimicrobial susceptibility testsfor bacteria which grow aerobically: M7-T. National Committeefor Clinical Laboratory Standards, Villanova, Pa.

17. Neu, H. C. 1982. Factors that affect the in vitro activity ofcephalosporin antibiotics. J. Antimicrob. Chemother. 10(Suppl.C):11-23.

18. Neu, H. C. 1982. Antibacterial activity of desacetylcefotaximealone and in combination with cefotaxime. Rev. Infect. Dis.4(Suppl.):S374-S378.

19. Neu, H. C., N. Aswapokee, P. Aswapokee, and K. P. Fu. 1979.HR 756, a new cephalosporin active against gram-positive andgram-negative aerobic and anaerobic bacteria. Antimicrob.Agents Chemother. 15:273-281.

20. Neu, H. C., and K. P. Fu. 1978. Cefuroxime, beta-lactamase-resistant cephalosporin with a broad spectrum of gram-positiveand -negative activity. Antimicrob. Agents Chemother. 13:657-664.

21. O'Callaghan, C. H., R. B. Sykes, A. Griffiths, and J. E.Thornton. 1976. Cefuroxime, a new cephalosporin antibiotic:activity in vitro. Antimicrob. Agents Chemother. 9:511-519.

22. Peterson, L. R., D. N. Gerding, W. H. Hall, and E. A. Schierl.1978. Medium-dependent variation in bactericidal activity ofantibiotics against susceptible Staphylococcus aureus. Antimi-crob. Agents Chemother. 13:665-668.

23. Pursiano, T. A., M. Misiek, F. Leitner, and K. E. Price. 1973.Effect of assay medium on the antibacterial activity of certainpenicillins and cephalosporins. Antimicrob. Agents Chemother.3:33-39.

24. Sahm, D. F., C. N. Baker, R. N. Jones, and C. Thornsberry.1983. Medium-dependent zone size discrepancies associatedwith susceptibility testing of group D streptococci againstvarious cephalosporins. J. Clin. Microbiol. 18:858-865.

25. Sosna, J. P., P. R. Murray, and G. Medoff. 1978. Comparison ofthe in vitro activities of HR 756 with cephalothin, cefoxitin, andcefamandole. Antimicrob. Agents Chemother. 13:665-668.

26. Thornsberrry, C., R. N. Jones, A. L. Barry, and P. C. Fuchs.1982. Antimicrobial susceptibility tests with cefotaxime andcorrelation with clinical bacteriologic response. Rev. Infect.Dis. 4(Suppl.):S316-324.

27. Thornsberry, C., and J. M. Swenson. 1980. Antimicrobial sus-ceptibility tests for Streptococcus pneumoniae. Lab. Med.11:83-86.

28. Verbist, L., and J. Verhaegen. 1981. In vitro activity of Ro 13-9904, a new P-lactamase-stable cephalosporin. Antimicrob.Agents Chemother. 19:222-225.

29. Washington, J. A., II, and P. K. W. Yu. 1970. Regression curveanalysis of cephalosporin activity. Appl. Microbiol. 19:589-593.

30. Wright, W. E., and J. A. Frogge. 1980. Hydrolysis of 3-acetoxymethyl cephalosporins by lysed whole blood. Antimi-crob. Agents Chemother. 17:99-100.

VOL. 20, 1984

on April 3, 2020 by guest

http://jcm.asm

.org/D

ownloaded from