ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Jan. 1975, p. 15-21Copyright 0 1975 American Society for Microbiology

Vol. 7, No. 1Printed in U.S.A.

Broth-Dilution Method for Determining the AntibioticSusceptibility of Anaerobic BacteriaDONALD R. STALONS AND CLYDE THORNSBERRY*

Department of Microbiology, Health Sciences Center, Temple. University, Philadelphia, Pennsylvania 19140;and Antimicrobics Investigation Section, Center for Disease Control, Atlanta, Georgia 30333*

Received for publication 6 September 1974

A broth-dilution method for performing antimicrobial susceptibility tests on

anaerobic bacteria has been proposed. The medium used in the test wasSchaedler broth, with incubation in a glove box with an atmosphere of 5% CO2,10% H2, and 85% N2, or in the GasPak system. Minimal inhibitory concentrationsfor selected antibiotics were determined, under these conditions, by using a

conventional twofold dilution scheme for the antibiotics and a "categorizationthree-tube method" in which two or three clinically significant concentrations ofeach antibiotic were used. Minimal inhibitory concentrations obtained by bothmethods were very similar. The categorization method could be used routinely totest the antimicrobial susceptibility of anaerobic bacteria.

Because anaerobic bacteria are being isolatedfrom clinical material with increasing fre-quency, the diagnostic microbiology laboratoryis more often called upon to test the susceptibil-ity of these organisms to a variety of antimi-crobial agents. The determination of the drugsusceptibility of anaerobic organisms is gener-ally more complicated than for aerobic orga-nisms, because of the difficulty in isolating andidentifying them, the obligation to test them inthe anaerobic environment, and the lack of astandard method for testing their susceptibility.The approach of many investigators has been

to use conventional techniques for aerobic andfacultative organisms to test the drug suscepti-bility of anaerobes in an oxygenless environ-ment. Therefore, antimicrobial susceptibilitieshave been determined by disk-diffusion, agar-dilution, and broth-dilution techniques. Al-though there are several common sources ofvariation for these techniques, e.g., inoculumconcentrations, pH of the medium, and compo-nents of the medium, one major disadvantage ofthe disk-diffusion and agar-dilution techniquesfor testing anaerobic bacteria in comparisonwith the broth-dilution technique is that theorganisms are much more likely to be subjectedto lethal doses of oxygen when they are spreadonto the surface of agar. Furthermore, theaddition of agar to a broth medium adds to thecomplexity of the medium and to the possibilityof some effects on the susceptibility test. Forthese reasons we chose to use a broth-dilutiontechnique as our reference method for determin-ing minimal inhibitory concentrations (MICs)

of several antimicrobial agents for various an-aerobic bacteria.The initial consideration in attempts to stan-

dardize a method for susceptibility testingshould be the achievement of optimal growthconditions for the organisms being tested. Wehave previously shown that Schaedler broth(BBL) in an atmosphere of 5% CO2, 10% H2,and 85% N2 was superior to eight other mediaand other combinations of gas mixtures forgrowing anaerobic bacteria likely to be isolatedfrom clinical material (10). A broth-dilutionmethod, with this medium and atmosphere,were used in this study. Susceptibility testswere also performed with a modified broth-dilu-tion technique in which only two or threeconcentrations of each antimicrobial agent areused. We think that this test can be readilyadapted to routine use in a diagnostic labora-tory.

MATERIALS AND METHODSBacterial strains. The bacterial strains used in

this study were obtained from the stock-culture col-lection of the Anaerobe Unit, Center for DiseaseControl, and from the American Type Culture Collec-tion. Each strain was characterized by a battery ofmicroscopic, cultural, and biochemical tests, as de-scribed by Dowell and Hawkins (1) and Holdemanand Moore (5). These strains of anaerobes includedmany that are commonly encountered in humaninfections (1).

Media. Schaedler broth was used in this studybecause it supported the best growth of the majorityof anaerobic organisms (10). It was prepared accord-ing to the manufacturer's specifications and prere-

15

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

16 STALONS AND THORNSBERRY

duced in the anaerobic atmosphere that we havefound to best support the growth of most anaerobicstrains isolated from clinical material, i.e., 85% N,,10% H2, and 5% CO2 (10). Mueller-Hinton broth(BBL) was used in certain comparative studies, sinceit is the medium in which standard MICs of antibiot-ics are well documented. Stock cultures were main-tained on Schaedler agar (BBL) that had been en-riched with 5% defibrinated rabbit blood and 0.5%yeast extract (Difco); hemin, which stimulates thegrowth of various Bacteroides species, and menadi-one, a growth requirement of some Bacteroides mela-ninogenicus strains (4), were added to all media infinal concentrations of 5 sg/ml and 0.1 qg/ml, respec-tively.

Antibiotics. The antibiotics used in this study weresupplied by the following organizations: cephalothin,erythromycin, and vancomycin, Eli Lilly Co. (Indian-apolis); clindamycin and lincomycin, Upjohn Co.(Kalamzaoo, Mich.); chloramphenicol, Parke-DavisCo. (Detroit); gentamicin, Schering Laboratories(Bloomfield, N. J.): penicillin, Wyeth Laboratories(Philadelphia, Pa.); and tetracycline, Chas. Pfizerand Co., Inc. (New York). The antibiotics were di-luted according to the protocol described by Gavanet al. (3), so that the final concentration of eachstock solution was 2,560 ,tg of active antibiotic perml. In the tests on the effect of medium and atmos-phere on MICs, the effect of inoculum size on MICs,and in the determination of base line MICs, the stocksolutions were diluted so that the final concentra-tion in the first tube would be 128 ;g/ml and in sUc-ceeding tubes, serial twofold dilutions of this concen-tration. For the categorization test, the concentra-tions of antibiotics shown in Table 5 were achieved bydiluting the stock antibiotic solution in Schaedlerbroth.

Anaerobic systems. The anaerobic systems usedwere the GasPak (BBL) and an anaerobic glove box(Coy Manufacturing, Ann Arbor, Mich.). A palla-dium-coated aluminum catalyst (Englehard Indus-tries, Newark) was used with each of the anaerobicsystems.

Determination of effect of medium and atmos-phere on antibiotic activity. MICs were determinedon two standard strains of facultative organisms bythe microtiter broth-dilution technique described byGavan et al. (3), with Mueller-Hinton and Schaedlerbroths and aerobic and anaerobic incubation, todetermine whether the medium and atmosphere se-lected for optimal growth in earlier studies (10) had aneffect on antibiotic activity. Two standard referencestrains, Staphylococcus aureus (ATCC no. 25923) andEscherichia coli (ATCC no. 25922) that have beenrecommended as reference strains for the standard-ized agar diffusion test (7) but have been studiedextensively by dilution tests, were used in this phaseof the study. The antibiotics tested are listed in Table1. The cultures were incubated for 24 h at 35 C in theappropriate atmosphere.

Effect of inoculum size on MICs. The effect ofvariation in inoculum on the MIC of antibiotics was

determined. Three-tenths milliliter of each antibioticsuspension listed in Table 1 was transferred to screw-cap tubes (16 by 125 mm) containing 2.7 ml ofSchaedler broth; this gave an additional 1:10 dilutionof the antibiotic (the broth medium was concentrated10% to compensate for this dilution factor). Themedia containing the antibiotics were then prere-duced in the anaerobic glove box with a gaseousenvironment of 5% CO,, 10% H,, and 85% N, for 24 h.After prereduction, sterile, precalibrated droppers(Cooke Laboratory Products no. 220-65) were used toinoculate each tube and a series of controls (3.0 ml ofmedium without antibiotic) with 0.025 ml of theappropriate concentration of organism. Standard in-ocula of four representative anaerobic organisms(Bacteroides fragilis ssp. fragilis, PeptostreptococcusCenter for Disease Control group 2, Eubacteriumalactolyticum, and Clostridium perfringens) were di-luted serially to yield cultures containing 5 x 10', 5 x

107, 5 x 106, and 5 x 10' organisms per ml. Therefore,the final concentrations of organisms used in thiscomparative study were 4 x 10, 4 x 10', 4 x 104, and4 x 103 per ml. After incubation at 35 C for anappropriate time, depending upon the growth re-

TABLE 1. MICs of selected antibiotics for two reference strains of bacteria obtained with a microtiterbroth-dilution method, with Mueller-Hinton and Schaedler broths, and aerobic and anaerobic atmospheres

Staphylococcus aureus Escherichia coli

Antibiotics Mueller-Hinton Schaedler Mueller-Hinton Schaedler

02a |Qb 02 02 02 02 02 02

Chloramphenicol 4-8c 8 8 8 4 4 8 8Clindamycin 0.5 0.25 0.25 0.25 64 32 128 128Erythromycin 0.25 0.5 0.5 0.5 32 64 32 32Gentamicin <S0.015 .<0.015 < 0.015 0.12 1 2 2 8Lincomycin 2 2 2 1 > 128 > 128 > 128 > 128Penicillin 0.06 0.03 0.12 0.03 32 32 64 64Tetracycline 0.06 0.12 1 0.25 1 0.5 2 2Vancomycin 2 1 2 1 > 128 > 128 > 128 > 128

a 02, aerobic incubation.b 2, anaerobic incubation.c MICs (ug/ml). Concentration of antimicrobial agents tested were serial twofold dilutions ranging from 128

to 0.015 gg/ml.

ANTimICROB. AGENTs CHEMOTHER.

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

BROTH-DILUTION METHOD 17

sponse of the organism, the end points and corre-sponding MIC values were determined through visualinterpretation of growth.

Reference MIC determinations. Base line MICswere determined on some commonly encounteredanaerobic bacteria using a broth dilution method. Theantibiotics were diluted in sterile distilled water tocontain 10 times the desired final concentration.Three-tenths milliliter of each antibiotic concentra-tion was added to 2.7 ml of supplemented Schaedlerbroth, thus achieving another 10-fold dilution ofantibiotic. The Schaedler broth had been concen-trated 10% to compensate for the dilution of the broth.The tests were performed in screw-capped tubes (16by 125 mm). This media was prereduced by placing itin the anaerobic glove box containing an atmosphereof 5% CO2, 10% H2, and 85% N2 for 24 h. The inoculawere prepared by growing the organisms in Schaedlerbroth until the desired turbidity (corresponding to 108colony-forming units/ml) was achieved or surpassed(generally overnight). The turbidity was determinedspectrophotometrically at 540 nm. Adjustment of theturbidity of the culture was made with Schaedlerbroth. The adjusted inoculum was added to each ofthe tubes with antibiotic-containing broth to achievea final concentration of approximately 106 col-ony-forming units/ml (0.025 ml into 3 ml of broth).The tubes were incubated at 35 C for 18 to 24 h (orlonger if growth not adequate) in the anaerobic glovebox containing 5% C02, 10% H2, and 85% N2. TheMIC was the least concentration of antibiotic thatprevented visible growth. To measure the reproduci-bility of the test, triplicate determinations wereperformed at each of three testing episodes, usingvarious lots of antibiotics, media, and anaerobic gas.

Categorization method for determining antibi-otic susceptibility of anaerobic bacteria. An abbre-viated version of the reference broth-dilution methodwas also evaluated. The antibiotics were tested in twoor three concentrations that permitted clinically use-ful categorization of susceptibility, similar to thecharacterization obtained by the method described byEricsson and Sherris (2). The concentrations of an-tibiotics used in these tests are shown in Table 5. Theantibiotics were diluted in sterile distilled water to 10times the desired final concentrations. Three-tenthsmilliliter of the antibiotic solution was added to 2.7 mlof Schaedler broth (contained in screw-capped intubes either 16 by 125 mm or 13 by 100 mm), andprereduced as in the reference method. The inoculumwas also prepared as in the reference method, that is,grown in Schaedler broth and adjusted spectro-photometrically to contain 108 colony-forming units/ml. Therefore, for each antibiotic, there were two orthree antibiotic concentrations (depending upon theantibiotic) and one control tube of antibiotic-freemedium. The tubes were incubated in the glove boxcontaining 5% CO2, 10% H2, and 85% N2, and/or theGasPak jar, at 35 C for 18 to 24 h (or longer to achievevisible turbidity). The end points were the lowestconcentrations of antibiotics that permitted visiblegrowth. The MICs obtained by this method couldthen be used to classify the organisms into one of fourcategories of susceptibility. We prefer the categoriessuggested by Ericsson and Sherris (2).

This test was evaluated by comparing resultsobtained with the category method to those obtainedwith the reference broth-dilution method, usingstrains of B. fragilis ssp. fragilis, PeptostreptococcusCDC group 2, E. alactolyticum, and C. perfringens.These evaluations were performed in the anaerobicglove box. In subsequent studies, the GasPak environ-ment was also evaluated for use with the category testby comparing results obtained in the glove box withthose obtained in the GasPak jar. In these studies, theinoculum preparation and subsequent inoculation ofbroth were done in the glove box in duplicate and oneset was removed to the GasPak jar. Therefore, theonly differences were the two atmospheres. Additionalstudies were also performed where the preparation ofthe inoculum, as well as the inoculation itself, wereconducted under aerobic conditions to determine theeffect on MIC values in the categorization method.

RESULTSDetermination of effect of medium and

atmosphere on antibiotic activities. The ef-fect of Schaedler broth and an anaerobic atmos-phere of 5% CO2, 10% H2, and 85% N2 onantimicrobical activity was determined by com-paring the minimal concentrations of certainantibiotics required to inhibit the growth ofreference facultative bacteria in this mediumand atmosphere to those obtained in standardtesting conditions. MICs of selected antibioticsfor the reference cultures, S. aureus (ATCC no.25923) and E. coli (ATCC no. 25922), obtainedin Schaedler broth and in Mueller-Hinton brothin both aerobic and anaerobic conditions areshown in Table 1. In general, the MICs obtainedin Schaedler broth, under both aerobic andanaerobic conditions, compared favorably withthose obtained with the standard medium indi-cating that the medium and the atmosphereproposed for this procedure had little adverseeffect on the action of the antibiotics. There-fore, any change in MIC from aerobic to an-aerobic conditions would most likely be due toa change in growth response of the facultativeorganisms.

Effect of inoculum size on MICs. When thenumber of organisms in the test inoculum wasvaried, different MIC values were obtained.Tenfold dilutions of a culture in the mid-log-lin-ear growth phase produced progessively smallerMIC values. These results are shown in Tables 2and 3. The most marked variations in MICpatterns produced by changes in inoculum con-centration were with the slower growing orga-nisms, e.g., E. alactolyticum (Table 3), ascompared to a rapid grower such as C. perfrin-gens (Table 2).Determination of baseline MICs of selected

antibiotics. Baseline MICs of antibiotics thathave been reported to be efficacious in the

VOL. 7, 1975

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

18 STALONS AND THORNSBERRY

TABLE 2. MICs of selected antibiotics for different.concentrations of Bacteroides fragilis ssp. fragilis andClostridium perfringens

B. fragilis C. perfringensAntibiotics_

106 a 101, 104 103 101, 105 104 103

Cephalothin 64° 32 16 8 0.25 0.25 <0.12 <0.12Chloramphenicol 2 1 0.5 0.5 2 2 1 1Clindamycin < 0.03 <0.03 < 0.03 < 0.03 0.25 0.12 0.12 0.06Erythromycin 0.25 0.12 0.06 <0.03 0.5 0.25 0.25 0.25Gentamicin > 128 > 128 128 128 32 32 32 16Lincomycin 4 2 1 0.5 2 1 0.5 0.5Penicillin 16 8 4 4 < 0.03 < 0.03 < 0.03 <0.03Tetracyclinr 0.25 0.12 0.06 0.06 0.06 0.06 0.06 0.06Vancomycin 8 8 4 4 0.12 0.12 0.06 0.06

a Numbers represent 4 x 106, 4 x 105, etc., organisms per ml.MICs (sgg/ml). Concentrations of antimicrobial agents tested were serial twofold dilutions ranging from 128

to 0.12 or 0.03 ug/ml.

TABLE 3. MICs of selected antibiotics for different concentrations of Peptostreptococcus Center for DiseaseControl group 2 and Eubacterium alactolyricum

Peptostreptococcus E. alactolyticumAntibiotics

106 a 105 104 108 10 105 104 103

Cephalothin 0.03b <0.015 < 0.015 <0.015 0.06 <0.015 < 0.015 .50.015Chloramphenicol 2 1 0.03 < 0.015 1 0.25 < 0.015 <0.015Clindamycin <0.015 <0.015 < 0.015 < 0.015 <0.015 <0.015 < 0.015 <0.015Erythromycin < 0.015 < 0.015 <0.015 < 0.015 0.06 < 0.015 <0.015 <0.015Gentamicin 8 4 <0.015 <0.015 8 2 <0.015 <0.015Lincomycin 0.5 0.12 0.03 <0.015 <0.015 <0.015 <0.015 <0.015Penicillin 0.03 0.03 < 0.015 <0.015 0.03 <0.015 <0.015 < 0.015Tetracycline 0.25 0.12 0.06 0.06 0.12 0.03 < 0.015 . 0.015Vancomycin 8 8 4 4 0.5 0.5 <0.015 <0.015

a Numbers represent 4 x 106, 4 x 105, etc., organisms per ml.b MICs (Ag/ml). Concentrations of antimicrobial agents tested were serial twofold dilutions ranging from 128 to

0.015 sg/ml.

treatment of patients with anaerobic bacterialinfections were determined by a modified broth-dilution method in which Schaedler broth andan anaerobic glove box atmosphere of 5% CO2,10% H2, and 85% N2 were used.Each organism was tested in triplicate on

different days to measure the reproducibility ofthe test. The mean MICs of selected antibioticsfor several commonly encountered anaerobicorganisms are shown in Table 4. Results ob-tained by this test were very reproducible; therewas no variation in the MICs obtained at anyone testing episode and no greater than onedilution variation between episodes.Three-tube categorization method for de-

termining antibiotic susceptibility of anaer-obes. The susceptibility patterns of the fourrepresentative anaerobes used in this studywere also determined by a three-tube broth-dilution categorization method. The results ofsix replicate tests (Tables 5 and 6) are consist-

ent with the base line MIC values shown inTable 4. Moreover, the data shows that whenthe GasPak system was used as the mode ofanaerobic incubation, identical, or very similar,MIC values were obtained. When the inoculumpreparation and inoculation were done underaerobic conditions, the MIC values were alsoidentical, or very similar, to those obtainedwhen these manipulations were performedunder anaerobic conditions.

DISCUSSION

Much effort has been exerted in recent yearsto standardize the techniques for antibioticsusceptibility testing. For the most part, thesuccess of these efforts has been limited to theaerobic and facultative bacteria. The develop-ment of standard procedures for the anaerobeshas been more difficult. Consequently, one ofthe main purposes of this study and of an earlier

ANTimICROB. AGENTS CHEMOTHER.

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

BROTH-DILUTION METHOD 19

TABLE 4. MICs (Agg/ml) of nine selected antibiotics against anaerobic organisms commonly encountered inclinical infections by using a broth-dilution technique with Schaedler broth and an anaerobic glove box

atmosphere of 5% CO2, 10% H2, and 85% N2

AntibioticsaOrganisms

Ceph Chlo Clin Ery Gen Lin Pen Tet Van

Bacteroides fragilis 32 4 0.12 0.5 > 32 4 8 0.25 168sp. fragilis no. 9053

B. fragilis sap. fragilis >32 4 0.12 0.5 >32 2 8 0.25 16no. 11239

B. fragilis sap. fragilis 32 4 0.06 0.5 >32 2 8 0.25 16no. 11732

B. fragilis sap. fragilis 32 4 0.25 0.5 >32 4 16 2 16no. 11736

B. fragilis asp. fragilis 32 2 0.06 0.5 >32 4 8 0.25 16no. 11737

B. fragilis sap. fragilis >32 4 <0.015 0.12 >32 0.5 16 0.25 8no. 23745

B. fragilis sap. 0.12 8 0.06 0.12 1 0.5 0.03 2 2thetaiotaomicron

Fusobacterium nu- 0.5 4 1 0.03 32 8 0.06 0.12 2cleatum no. 9052

F. nucleatum no. 0.03 2 0.06 4 16 0.25 0.06 0.12 >3214371

Eubacterium 0.12 2 <0.015 0.03 8 0.03 0.06 0.06 0.5alactolyticum

Propionibacterium 0.5 0.5 <0.015 <0.015 0.25 0.06 0.06 0.5 0.25granulosum

Clostridium 0.12 1 0.06 0.5 32 0.5 0.06 0.06 0.12perfringens

C. ramosum 0.5 4 2 0.25 8 4 1 0.5 2Peptostreptococcus 0.03 1 0.25 .0.015 8 0.25 < 0.015 0.25 0.25CDC group 2

a Antibiotics tested were cephalothin, chloramphenicol, clindamycin, erythromycin, gentamicin, lincomycin, penicillin,tetracycline, and vancomycin, respectively. They were tested in concentrations that were serial twofold dilutions ranging from32 to 0.015 .JVml.

TABLE 5. MICs of nine antibiotics against Bacteroides fragilis and Peptostreptococcus Center for DiseaseControl group 2a

Bacteroides fragilis PeptostreptococcusAntibiotics Concn tested sap. fragilis 9053 CDC group 2(Ag/ml)

Glove box GasPak Glove box GasPak

Cephalothin (2, 16,128) > 16, < 128 > 16, . 128 < 2 < 2Chloramphenicol (1, 8) > 1, .8 > 1, .8 < 1 < 1Clindamycin (1, 4, 64) . 1 < 1 . 1 . 1Erythromycin (1, 4, 64) < 1 < 1 < 1 < 1Gentamicin (0.5, 4, 64) > 64 > 64 > 4, .64 >4, S64Lincomycin (1, 4, 64) > 1, <4 > 1, .4 < 1 < 1Penicillin (0.25, 16, 128) > 0.25, . 16 > 0.25, . 16 .0.25 .0.25Tetracycline (1, 8, 32) < 1 < 1 . 1 < 1Vancomycin (4, 16) >4, .16 >4, .16 .4 .4

a The MICs were determined by a categorization method using Schaedler broth and two atmospheres(anaerobic glove box with 5% CO2, 10%H2, and 85% N2; GasPak).

study (10) was to consider the prerequisites thatmight lead to the recommendation of a refer-ence method and routine method that would beaccepted for use in a diagnostic laboratory.Of the three testing techniques commonly

used for determining antibiotic susceptibility,the broth-dilution method has fewer sources for

variation in testing anaerobic bacteria. Al-though contamination and the restriction to testone organism per series of tubes are drawbacksof the broth-dilution method, the use of brothmedia minimizes the amount of oxygen expo-sure and thus may increase the growth responseor reduce the time required to achieve eugonic

VOL. 7, 1975

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

20 STALONS AND THORNSBERRY

TABLE 6. MICs on nine antibiotics against Eubacterium alactolyticum and Clostridium perfringensa

MICs

Antibiotics Concn tested Eubacterium alactolyticum Clostridium p6erfringens(jig/mi)

Glove box GasPak Glove box GasPak

Cephalothin (2, 16, 128) < 2 . 2 < 2 < 2Chloramphenicol (1, 8) > 1, .8 >1, .8 > 1, .8 > 1, .8Clindamycin (1, 4, 64) < 1 . 1 . 1 < 1Erythromycin (1, 4, 64) < 1 < 1 < 1 < 1Gentamicin (0.5, 4, 64) > 4, . 64 >4, .64 >4, .64 >4, .64Lincomycin (1, 4, 64) <1 <1 <1 <1Penicillin (0.25, 16, 128) .0.25 .0.25 .0.25 .0.25Tetracycline (1, 8,32) <1 .1 <1 <1Vancomycin (4, 16) <4 .4 <4 <4

aThe MICs were determined by a categorization method using Schaedler broth and two atmospheres(anaerobic glove box with 5% Co2, 10% H2, and 85% N2; GasPak).

growth in culture. Other advantages of thebroth-dilution technique are that it can be usedfor simultaneous determination of minimal in-hibitory and bactericidal concentrations; it canbe used to test several antibiotics in synergismstudies; and it is more easily adapted to auto-mated procedures that have been used forsusceptibility testing. (3, 6; C. Thornsberry, J.C. Sherris, L. D. Thrupp, J. M. Matsen, J.Washington, T. Gavan, L. D Sabbath, A. Ba-lows, and F. Schoenknecht. Prog. Abstr. Inter-sci. Conf. Antimicrob. Ag. Chemother., 13th,Washington, D.C., Abstr. 204, 1973).Although the agar dilution method is well

suited for the simultaneous testing of largenumbers of organisms, the broth-dilutionmethod is a more efficient method for routineuse in most laboratories where only a fewanaerobic isolates would be tested at any onetime.The medium and atmosphere used in these

broth-dilution tests had little deleterious effecton MIC values as compared to values achievedthrough standard methods with reference facul-tative organisms. A higher cation concentra-tion in the Schaedler medium may contribute tovariations in the MICs obtained with gentami-cin and tetracycline, since effectiveness of ani-moglycosides and tetracyclines may be associ-ated with ion concentration (2), but anaerobio-sis and increased CO2 could have played a role(8).Even though the data obtained with this

reference method are reliable, it is not anoptimal method for the-routine diagnostic labo-ratory because of the antibiotic dilution proto-col and the physical requirements for the han-dling and processing of a large number of tubes.Consequently, our approach to overcoming this

cumbersome problem was to abbreviate thenumber of antibiotic dilutions used in the test.Similar systems have been used by Schneirsonand Amsterdam in a broth-dilution test (9) andby Washington in an agar-dilution test (11).Ericsson and Sherris and the World HealthOrganization-International CollaborativeStudy Group (2) also advocate such an ap-proach to in vitro susceptibility testing. Theyrecommend an extended dilution scheme inwhich two or three concentrations of each anti-biotic are used so that susceptibility can becategorized according to levels attainable inblood, urine, or other body fluids or tissues, andaccording to efficacy in clinical experience. Byusing this method, there would generally be fourcategories of susceptibilities. Ericsson andSherris (2) suggested that these be called groups1, 2, 3 and 4. Group 1 would include the verysusceptible organisms that would probably re-spond to the usual dosage of the antimicrobialagent being tested. Group 2 would include thoseorganisms that cause systemic infections thatwould be likely to respond to the antimicrobialagent if it is given in high doses. Group 3 wouldinclude organisms causing localized infectionsthat could probably be treated successfullybecause the antimicrobial agent is concen-trated by physiological processes, as in theurine, or by direct application. Group 4 orga-nisms would be those with degrees of resistancethat make clinical response unlikely and, there-fore, would be considered resistant withoutqualifaction. If these group definitions were tobe used in the clinical laboratory, the reportshould contain a definition of each group. Ex-amples of the application of this system arediscussed in the International CollaborativeStudy report of Ericsson and Sherris (2).

ANTimICROB. AGENTS CHEMOTHER.

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

BROTH-DILUTION METHOD 21

The test concentrations of each antibioticthat we have used in this study were notadopted just for testing anaerobic bacteria, butfor testing all bacteria. General agreement onwhat the most clinically useful test concentra-tions should be may not occur, but with thismethod each organization could set its own testlevels.Although we must test more strains, our

results lead us to believe that the methodsproposed here can be used routinely in theclinical bacteriology laboratory to test the an-timicrobial susceptiblility of anaerobic bacte-ria. The three-tube categorization method, per-formed with the GasPak system, can be used asthe routine method. If more definitive MICs aredesired, the more conventional, twofold dilutionscheme in the same broth and atmosphere canbe used.The categorization would be even more prac-

tical for the clinical laboratory, if paper diskswere used to deliver the antibiotic to the broth.This method has been used by Isenberg et al. (6)and Thornsberry et al. (C. Thornsberry et al.Prog. Abstr. Intersci. Conf. Antimicrob. Ag.Chemother., 13th, Washington, D.C., Abstr.204, 1973) for automated or mechanized suscep-tibility tests on aerobic and facultative orga-nisms, and by Wilkins and Thiel (12) foranaerobic bacteria. This modification of thecategorization method is being studied.

LITERATURE CITED

1. Dowell, V. R., Jr., and T. M. Hawkins. 1973. Laboratorymethods in anaerobic bacteriology. U.S. Departmentof Health, Education, and Welfare Publication no.(HSM) 73-8222.

2. Ericsson, H. M., and J. C. Sherris. 1971. Antibioticsensitivity testing: Report of an international col-laborative study. Acta Pathol. Microbiol. Scand. SectB. Supplement no. 217.

3. Gavan, T. L., H. W. McFadden, Jr., and E. L. Cheatle.1971. Antimicrobial susceptibility testing. AmericanSociety of Clinical Pathologists, Chicago, Ill.

4. Gibbons, R. J., and J. B. MacDonald. 1959. Hemin andvitamin K compounds as required growth factors forthe cultivation of certain strains of Bacteroides melani-nogenicus. J. Bacteriol. 80:164-170.

5. Holdeman, L. V., and W. E. C. Moore. 1972. Anaerobelaboratory manual. Virginia Polytechnic Institute andState University, Blacksburg, Va.

6. Isenberg, H. D., A. Reichler, and D. Wiseman. 1971.Prototype of a fully automated device for determina-tion of bacterial antibiotic susceptibility in the clinicallaboratory. Appl. Microbiol. 22:980-986.

7. National Committee for Clinical Laboratory StandardsSubcommittee on Antimicrobial Susceptibility Test-ing. 1974. Performance standards for antimicrobialsusceptibility tests as used in clinical laboratories, p.138-155. In A. Balows (ed.), Current techniques forantibiotic susceptibility testing. Charles C Thomas,Springfield, Ill.

8. Rosenblatt, J. E., and F. Schoenknecht. 1972. Effect ofseveral components of anaerobic incubation on antibi-otic susceptibility test results. Antimicrob. Ag. Chemo-ther. 1:433-440.

9. Schneirson, S. S., and D. Amsterdam. 1959. A simplifiedtube procedure for the routine determination of bacte-rial sensitivity to antibiotics. Amer. J. Clin. Pathol.31:81-86.

10. Stalons, D. R., C. Thornsberry, and V. R. Dowell. 1974.The effect of culture medium and carbon dioxideconcentrations on the growth of anaerobic bacteriafrom clinical specimens. Appl. Microbiol.27:1098-1104.

11. Washington, J. A. The agar dilution method, p. 127-141.In T. L. Gavan, H. W. McFadden, and E. L. Cheatle(ed.), Antimicrobial susceptibility testing. AmericanSociety of Clinical Pathology, Chicago, Ill.

12. Wilkins, T. D., and T. Thiel. 1973. Modified broth-diskmethod for testing the antibiotic susceptibility ofanaerobic bacteria. Antimicrob. Ag. Chemother.3:350-356.

VOL. 7, 1975

on May 23, 2018 by guest

http://aac.asm.org/

Dow

nloaded from