Antibacterial susceptibility testing in theclinical laboratory

Maria Joyce, MD, DTM&H,Christopher W. Woods, MD, MPH*

Division of Infectious Diseases, Duke University Medical Center,

Durham, NC 27710, USA

The clinical microbiology laboratory has a mandate to provide reliable,accurate susceptibility data in a time frame that is useful to the cliniciansrequesting the information to achieve good clinical outcome and, whenpossible, to reduce the emergence of resistance. This mandate is served byselective reporting of results to the ordering clinician from isolates obtainedfrom individual patients and by providing collective data on localprevalence of resistance to be used for empirical therapy. To meet thesechallenges and responsibilities, clinical microbiologists must continuouslyassess and update susceptibility testing and reporting strategies.

Indications for susceptibility testing

Routine susceptibility testing should be reserved for clinically signicantisolates retrieved from appropriately collected specimens that have a well-described capacity for resistance to primary therapeutic agents and forwhich standardized performance methodology and interpretive criteria areavailable. Testing is not routinely indicated for organisms with predictableresponses to antimicrobial agents. Patient allergy, intolerance, or epidemi-ologic studies, however, are alternative reasons for susceptibility testing forthese organisms. Susceptibility testing of isolates known or believed to becontaminants is actively discouraged, because provision of results may

Infect Dis Clin N Am 18 (2004) 401434encourage inappropriate antibiotic use [1]. Special consideration should alsobe given to isolates retrieved from patients with bacterial meningitis,

* Corresponding author. Medical Microbiology Section, Service 113, Durham VAMC,

Durham, NC 27705.

E-mail address: woods004@mc.duke.edu (C.W. Woods).

0891-5520/04/$ - see front matter 2004 Elsevier Inc. All rights reserved.doi:10.1016/j.idc.2004.04.001

infective endocarditis, osteomyelitis, infections of the eye or other protectedsites, and isolates with normally predictable susceptibility collected frompatients with compromised immune systems.

Methods: phenotypic (conventional)

Conventional inhibitory methods rely on the detection of in vitrophenotypic expression of resistance. The mean inhibitory concentration(MIC) is dened as the lowest concentration of an antimicrobial agent thatinhibits the growth of an organism over a dened interval and is typicallyexpressed in micrograms per milliliter. Measuring the MIC involvesexposing the organism to a series of twofold dilutions of the antimicrobialagent in a suitable culture media (broth or agar). This phenotypic resistancecan be quantitatively reported as the MIC for dilution methods orqualitatively (sensitive, intermediate, resistant) as provided by either di-lution or diusion methods. These qualitative interpretations are based onMIC distributions, pharmacokinetics and pharmacodynamics, and clinicaland bacteriologic response rates [2].

Regardless of the organism-antimicrobial combination and the methodused, the results of in vitro diagnostic tests can vary dramatically dependingon the type of growth media used (nutritional supplements, cation content,pH); the amount of organism tested (inoculum eect); and the conditions ofincubation (duration, temperature, atmosphere). For each test run, controlsfor viability (growth control), purity (purity plate), and methodology(quality control organisms) should be performed in parallel with the testorganism. The National Committee for Clinical Laboratory Standards(NCCLS) provides standards for these performance variables, for qualitycontrol practices, and for the interpretation of the results of referencemethods for antibacterial susceptibility testing. The organization publishesconsensus standards for the reference methods of antimicrobial susceptibil-ity testing. The M2 standard covers disk diusion, the M7 standard includesminimal inhibitory concentration techniques, and the M11 standard is foranaerobic methods. These standards are published every 3 years andsupplemental tables (M100) are updated annually. A new guideline for theanalysis and presentation of cumulative susceptibility test data (M-39-A) isalso now available. Although these guidelines are available globally,dierent methods and interpretive criteria may be used outside of NorthAmerica [3].

Broth dilution

For all broth methods, cation-adjusted Mueller-Hinton broth (CAMHB)

402 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434is recommended for the routine testing of commonly encountered non-fastidious organisms. Cation adjustment (calcium and magnesium) isrequired to ensure acceptable results when aminoglycosides are tested

against Pseudomonas isolates (and when tetracycline is tested against otherbacteria) [4,5]. Insucient cation concentrations result in increased amino-glycoside activity and vice versa [6]. A standardized inoculum may beprepared by growing organisms to log phase (exponential growth) or bydirect suspension of organisms to the turbidity of the 0.5 McFarlandstandard. Direct suspension is preferred for fastidious organisms (eg,Haemophilus inuenzae, Streptococcus pneumoniae, Neisseria meningitidis)and for detection of methicillin resistance in staphylococci. In brothmacrodilution, 13 mm 100 mm tubes are inoculated with a minimumvolume of 1 mL (usually 2 mL) and are evaluated macroscopically forgrowth. Microdilution methods typically use 96-well trays containing0.1 mL of broth and viewing devices may be used for reading and record-ing. For either method, the inoculum should be adjusted to achieve a con-centration of approximately 5 105 colony-forming units per milliliter.Incubation is typically in air at 35(C for 16 to 20 hours and increased CO2 isnot required. Detection of low-level or inducible resistance in certainorganisms, however, requires longer incubation times and additionalsupplementation (see later). Growth is best determined by comparison withthat in the growth control well and is generally indicated by turbiditythroughout the well or by buttons in the well bottom. To ensure that themethod is testing a single organism, a purity plate should also be planted.Trailing end points may be seen when trimethoprim or sulfonamides andcertain other bacteriostatic antibiotics are tested, and the concentration atwhich 80% of growth is inhibited compared with the growth control occursis recorded as the MIC [7].

Agar dilution

Agar dilution is a well-established technique for obtaining quantitativesusceptibility results and is the reference method commonly performed inEurope [8]. Mueller-Hinton agar (MHA) is the recommended medium forthe testing of most commonly encountered aerobic and facultative anaer-obic bacteria and is standardized such that calcium and magnesiumsupplementation is not indicated [9]. For predictable diusion, the depthof the agar should be between 3 mm and 4 mm. The recommended nalinoculum is 104 colony-forming units per spot [10]. The method is labor-intensive and costly, but is recommended for fastidious organisms that donot grow well in broth media (eg, Neisseria gonorrhoeae, anaerobes) [11,12].

Disk diusion

The disk diusion method has been standardized for the testing ofcommon, rapidly growing organisms and allows for a qualitative categori-

403M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434zation of isolates as susceptible, intermediate, or resistant [7]. An antibiotic-impregnated lter paper disk is placed on the surface of an agar plate(MHA) inoculated with a lawn of organism at a known turbidity (0.5

McFarland). The inoculum may be prepared by either log phase growth ordirect suspension from colonies on the agar plate. As with dilution methods,direct suspension is preferred for fastidious organisms and for detection ofmethicillin resistance in staphylococci. The antibiotic diuses through theagar almost instantaneously after placement on the agar and creates a zoneof inhibition that can be measured in millimeters (edge to edge, including thedisk) [13]. Interpretation of the test is based on the inverse correlation of thezone diameter with MIC for each combination of antimicrobial andorganism [7].

Because interpretation is dependent on the rate of antibiotic diusionversus bacterial growth, this method should not be used to evaluate theantimicrobial susceptibilities of bacteria that show marked variability ingrowth rates (eg, some glucose nonfermenting gram-negative bacilli andanaerobic bacteria). Agar dilution has been modied for some fastidiousorganisms, however, when special media and interpretive breakpoints areused [7,14].

Screening and breakpoint methods

In some instances, testing of a single drug concentration may be the mostreliable and convenient method for the detection of resistance (eg, screen forhigh-level aminoglycoside resistance in enterococci). In addition to singleconcentration screens, breakpoint susceptibility testing of antimicrobialagents only at the specic concentrations necessary for dierentiatingbetween the interpretive categories rather than the full range is often used.In particular, commercial methods with limited room on their panelsfrequently use breakpoint testing to permit a greater number of agents tobe tested [15].

Automated, short incubation, and rapid methods

Commercial microdilution accounts for two thirds of susceptibilitytesting in the United States [16]. When commercial systems are used, themanufacturers recommendations concerning storage, inoculation, incuba-tion, and interpretations should be followed precisely. An error rate forthese systems has to be determined by comparison with the referenceNCCLS broth dilution method using at least 100 clinical isolates of a singlegenus or species. The detection rate for very major errors (false suscepti-bility) should be less than 1.5% and major errors (false resistance) less than3% of isolates [17]. At present, there are ve automated broth dilutionantibiotic susceptibility testing (AST) systems currently approved for use inthe United States including the Vitek and Vitek 2 (BioMerieux, St. Louis,

404 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434MO), the Microscan WalkAway (Dade Behring, West Sacramento, CA), theSensititre ARIS (Trek Diagnostic Systems, Westlake, OH), and the PhoenixSystem (BD Diagnostic Systems, Sparks, MD). These systems vary in the

extent of automation; the method of detection (turbidity, uorometric); thespeed of detection; data management and interpretation packages to helpverify reporting of results; and provision of antibiograms [15]. The exibilityof commercially prepared broth microdilution trays is limited comparedwith in-house preparations. Special instrumentation is also available forreading, storing, and interpreting zone diameters from disk diusion testsand this may reduce interobserver variability [1820].

In addition to labor savings resulting from decreased media preparationtime and automated identication and susceptibility, most systems reducethe time to reporting. Two studies have demonstrated the clinical andeconomic benets derived from the use of rapid susceptibility testing andreporting [21,22]. Shortcomings of rapid methods, however, include di-culty in detecting some inducible or subtle resistance mechanisms [2325].

One of the simplest and fastest susceptibility tests uses a chromogeniccephalosporin (nitrocen) that changes color when hydrolyzed byb-lactamase. The test does not detect resistance to b-lactams from othermechanisms, but has been useful in evaluating H inuenzae, N gonorrhoeae,Moraxella catarrhalis, enterococci, and some anaerobes [4].

Conventional methods have also been adjusted to achieve more rapidreporting of results, including direct inoculation of positive blood culturesonto disk diusion plates [26,27] or microdilution panels [28]. Thesemethods need to be standardized and validated, however, with the emergingresistant organisms and fastidious organisms.

Nonreference quantitative methods

The epsilometer test ([Etest] AB Biodisk, Solna, Sweden) uses a plastic-coated strip that releases an antimicrobial gradient into agar media. TheMIC is read where the ellipse of growth inhibition intercepts the scale on thestrip. Strips containing dierent antimicrobials can be placed in a radialfashion on the surface of a large MHA plate. The method is comparablewith reference dilution methods for staphylococci, enterococci, anaerobes,and various gram-negative organisms [2931]. Although very simple toperform and exible, this method is relatively expensive. It is mostcommonly used for infrequent drug requests or for fastidious or anaerobicorganisms because the strip may be placed onto enriched media to enhancegrowth. Strips may also be supplemented for detection of certain resistancemechanisms (ethylenediaminetetraacetic acid for metallo-b-lactamase) [32]or for eective use with certain agents (calcium for daptomycin) [33].

The spiral gradient end point method (Spiral Systems Instruments,Bethesda, MD) uses increasing concentrations of an antibiotic in a radial

405M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434fashion from the center of an agar plate [34]. The organism is streaked fromthe center and the end point is measured as the distance of growth to thecenter of the plate. This method is expensive, inecient, and not widely used.

Method selection

The choice of methodology to be used in individual laboratories is usuallybased on nancial and labor resources and the volume of tests to beperformed. Variables to be considered include relative ease of performance,cost of equipment and contract arrangements, cost of media and supple-mental materials, exibility in selection of drugs for testing, use of automatedor semiautomated devices to facilitate testing, and the perceived accuracy ofthe methodology [4]. Although the more traditional macrodilution pro-cedure is laborious and infrequently performed in modern clinical microbi-ology laboratories, microdilution (particularly commercial methodology) isstandard in many laboratories. Despite the growing popularity of commer-cial microdilution systems, the disk method is exible, technically simple,inexpensive, and its results are reproducible. With a few notable exceptions,clinicians usually prefer categorical results in most situations. There are fewsituations where MICs are used clinically (eg, endocarditis and meningitis)[35], and alternative methods that provide an MIC can be available whenquantitative results are indicated.

Selection of agents for routine testing and reporting

The battery of agents to be tested depends on the species, the formulary,and demography of patients in the institution, and the likelihood ofencountering highly resistant organisms [36]. Many compounds exhibitsimilar if not identical activities in vitro to other agents in their same class,so that in some cases one compound may be chosen as a surrogate for anantimicrobial class because of a greater ability to detect resistance.

To discourage the indiscriminate use of broad-spectrum antimicrobials,the laboratory should establish a reporting cascade in which a few, narrow-spectrum, less-expensive antimicrobials are reported, whereas other resultsare suppressed and released as needed. A restricted reporting policy canhave a strong inuence on prescribing patterns and ultimately on resistance[37]. Decisions regarding testing and reporting algorithms should be made inconsultation with local infectious disease practitioners; pharmacists; andrelevant committees (pharmacy and therapeutics, infection control) [36].

Bactericidal methods and combination therapy

The methods described previously are measures of the inhibitory capacityof an antibacterial. In the absence of a robust immune system or wheninfection occurs in relatively protected sites, however, cure often depends onthe killing capacity of a drug. Although, most experts agree that every new

406 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434drug should undergo testing to determine its bactericidal capacity, thebenet of testing isolates from individual patients with meningitis, endo-carditis, or osteomyelitis remains controversial [38].

Bactericidal activity can be measured by one of three methods: (1)calculation of the minimum bactericidal concentration, (2) performance oftime-kill studies, or (3) serum bactericidal assay (Schlichter test). Theminimum bactericidal concentration is obtained by subculturing the tubes(macrodilution) or wells (microdilution) that do not demonstrate growth at24 hours and is dened as the lowest concentration of antibiotic at whicha 99.9% reduction of viable organisms occurs [39]. A bactericidal drugachieves this within two dilutions of the MIC. A number of biologic andtechnical issues limit the clinical value of this information. Time-killbactericidal tests plot the proportion of bacteria killed over time whenexposed to a specic antimicrobial concentration. The concentrationschosen for testing are based on achievable serum levels. The serumbactericidal test is a modied broth dilution test, in which serial dilutionsof serum are tested rather than specic antimicrobial concentrations [40]. Acorrelation between specic bactericidal titers and outcome of antimicrobialtherapy has been dicult to establish.

Synergy is usually dened by a 2 log10 drop (or greater) in colony-forming units per milliliter between the combination of each of twoantimicrobials (at one fourth of their respective minimum bactericidalconcentrations) compared with the more active drug alone at a concentra-tion of one half its minimum bactericidal concentration [41]. The eects ofthese drug combinations may be additive, synergistic, or antagonistic. Theclinical importance of assessing synergy in the treatment of endocarditis iswell established, but is controversial for other infections. Other thanscreening enterococci for high-level aminoglycoside resistance (discussedsubsequently), the test most frequently used involves serial dilutions of twodrugs, alone and in combination (two-dimensional checkerboard titration).Results for individual drugs are reported as a fractional inhibitory con-centration (FIC). A sum of the FIC for each drug (FIC index) less than orequal to 0.5 suggests synergy. An FIC index greater than 4 is consistent withantagonism [42]. In many situations, the time-kill kinetic studies arepreferred for assessing synergy; however, these are labor-intensive and aretypically performed only on a small number of drugs.

Genotypic methods

In addition to rapid phenotypic methods, molecular methods targetingmutations responsible for antimicrobial resistance are increasingly availableto the clinical laboratory. The genotypic approach is primarily attractivebecause these assays may facilitate real-time detection directly from patientspecimens before culture results are available, they can arbitrate MIC results

407M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434at or near breakpoints, they make excellent epidemiologic tools, and theycan be used in the evaluation of new susceptibility tests [43]. A great numberof dierent probes and primers have been developed targeting genes

associated with resistance to most antimicrobial categories. Newer methods,including real-time polymerase chain reaction (PCR), ligase chain reaction,cleavase-based assays, and DNA sequence analysis, are often more rapidand more exible.

Development of molecular assays has focused on resistance mechanismsin common organisms that are limited to a few, well-characterized geneticmechanisms (eg, mecA in methicillin-resistant Staphylococcus aureus[MRSA], vanA in vancomycin-resistant enterococci). Additionally, sequenceanalysis methodology is useful for point mutations associated withuoroquinolone resistance [44] and extended spectrum b-lactamases (ESBL)in reference laboratories [45,46]. Although many laboratories use in-housemethods, only mecA methods have been made commercially available [47].

The expectation that molecular techniques would replace phenotypicsusceptibility testing has not yet been realized. DNA microarray technologyoers some hope for handling the large number of analyses required to pursuemultiple resistancemechanisms. Implementation of this technology awaits thesuccessful handling of several technical impediments, and routine suscepti-bility testing still needs to be performed to look for new mechanisms [48].

Pathogen-specic issues

Gram-positive organisms

Staphylococcus aureusAmong gram-positive pathogens, S aureus is the leading cause of

morbidity and mortality among nosocomial infections [49,50]. S aureus isinherently very susceptible to antibiotics [51] but is also adept at developingresistance.

Nearly all clinical isolates of S aureus produce a penicillinase [51];however, if the organism tests susceptible to penicillin, it is the drug ofchoice. If the penicillin disk zone is greater than or equal to 29 mm or theMIC is between 0.03 and 0.25 lg/mL, then a b-lactamase test is performedon growth from the margin around the oxacillin disk (induced b-lactamasetest) is indicated [7]. If the isolate is b-lactamase negative and oxacillinsusceptible, it can be reported as susceptible to all penicillins, cephalospor-ins, carbapenems, and monobactams [7].

Methicillin-resistant S aureus is responsible for over 50% of S aureusinfections and most are hospital-acquired S aureus infections [52,53].Methicillin resistance in staphylococci is caused by an altered penicillinbinding protein (PBP2a), which has a markedly low anity for all b-lactamantibiotics, including penicillins, cephalosporins, and carbapenems. PBP2a

408 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434is encoded by the mecA gene, which is carried by a large mobile geneticelement designated staphylococcal cassette chromosome mec (SCC mec) andintegrated into the chromosome of MRSA [54]. This large cassette

frequently carries additional genes encoding resistance to multipleantibiotics [5557]. A smaller SCCmec IV has been found in community-associated MRSA that characteristically retains susceptibility to non b-lactam antibiotics [58].

Phenotypic detection of methicillin resistance in S aureus is notstraightforward because not all subpopulations express methicillin re-sistance. When phenotypic expression of resistance is 1 in 104 to 1 in 108

of mecA-positive cells, the isolate is referred to as heterogeneous orheteroresistant and detection by standard methods is dicult [14]. Theexpression of heteroresistant populations is enhanced by higher saltconcentrations and lower temperatures [59]. For broth dilution referencemethods, CAMHB must be supplemented with 2% NaCl and incubationmust complete a full 24 hours [7]. This method is cumbersome and rarelyused in clinical practice. For routine disk diusion, MHA is not supple-mented with 2% NaCl because it may adversely aect the testing of otherantimicrobial agents [7]. This reduces the ability to detect heteroresistantcolonies around the oxacillin disk. The oxacillin-salt agar screen plate isa practical and reliable method for MRSA detection in the clinicallaboratory. MHA with oxacillin (6 lg/mL) and NaCl (4%) is spotinoculated with a direct growth suspension and incubated at 35(C for a full24 hours. The oxacillin Etest on MHA supplemented with 2% NaCl has alsobeen found to give reliable results [60].

Commercial panels are constantly being updated and their accuracy hasimproved with more discriminating tests. Rapid testing with certaincommercial panels has achieved a sensitivity for MRSA versus mecA PCRof 98% to 100% with a specicity of 100% with results available in 8 hours[15,61]. The oxacillin salt screen plate in the same study by Yamazumi et al[61] for MRSA was compared with PCR with a sensitivity of 98% andspecicity of 100%.

Recent studies have shown that a cefoxitin disk (30 lg) for diusiontesting is a better indicator of methicillin resistance than oxacillin [62,63],making the salt screen plate unnecessary. In addition to better correlationwith the presence of the mecA gene, the zone size is easier to read because ofcrisper zone margins than with oxacillin (Fig. 1). These ndings have beenincorporated into the most recent version of the NCCLS guidelines [7].

Latex agglutination tests are available to detect the mecA gene product,PBP2a [64,65]. Test results are available in real-time, and methicillinsusceptibility is available up to 19 hours earlier than with standardsusceptibility testing [66]. The sensitivity and specicity are high whencompared with the mecA PCR.

There are three main reasons for borderline resistance (MIC 28 lg/mL):(1) borderline resistant S aureus, so-called hyper b-lactamase producers

409M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434(intermediate or resistant susceptibility to oxacillin, but susceptible toamoxicillin-clavulanic acid on disk diusion testing or salt screen plate);(2) modied intrinsic PBPs, normal PBPs with reduced anity for

b-lactams; (3) heteroresistant colonies that carry the mecA gene [67,68]. Theclinical laboratory must be able to dierentiate the various types ofresistance. The hyper b-lactamaseproducing isolates are not truly resistantto b-lactamb-lactamase inhibitor combinations or penicillinase-stablepenicillins and are treated eectively with penicillinase-stable b-lactam orb-lactamb-lactamase inhibitors [69]. Other susceptibility testing issues forS aureus include small colony variants (\10 times the size of the parentcolony) or CO2-dependant isolates. Small colony variants are fastidiousorganisms that are decient in electron transport and are auxotrophs forthiamine, hemin, and menadione [70]. The colonies grow to a normal sizewhen these nutrients are replaced. The CO2-dependent isolates requireenhanced CO2 conditions for growth.

The emergence of S aureus isolates with reduced susceptibility (MIC 4 lg/mL) and intermediate susceptibility (MIC 8 lg/mL) to vancomycin resultingfrom cell wall thickening iswell-documented [71,72].Detection of reduced andintermediate susceptibility in the clinical laboratory can be dicult withconventional methods. For laboratories primarily using disk diusion allS aureus isolates with a zone diameter of less than or equal to 14mm should beconrmed by an MIC method [7,73]. Isolates conrmed to have an MICgreater than or equal to 4 lg/mL should be reported to the local healthdepartment and the Centers for Disease Control and Prevention (CDC) [74].More recently, the rst fully vancomycin-resistant (MIC32 lg/mL)S aureusisolates containing the vanA resistance gene from enterococci were identiedin the United States [75,76]. Performance of automated susceptibility testingmethods for detection of these isolates was inconsistent, therefore, theCDC recommend inoculating a vancomycin-agar screening plate (6 lg/ml)routinely.

Fig. 1. Disk diusion testing of a hyper b-lactamaseproducing Staphylococcus aureus. Thezone of inhibition (clear halo) around the cefoxitin disk (FOX) is sharp. In contrast, the zone of

inhibition around the oxacillin disk (OX) is blurred by the presence of individual colonies.

410 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434Noncell-wall active agents and D diusion zone test. Most macrolide,lincosamide, and streptogramin B resistance in S aureus is coded for by

ermA and ermC genes [77,78]. Macrolide, lincosamide, and streptogramin Bresistance may be inducible (in vitro susceptible to clindamycin) orconstitutive (in vitro resistant to all macrolide, lincosamide, and streptog-ramin B) [77,79]. Treatment of inducible macrolide, lincosamide, andstreptogramin B resistant isolates with clindamycin may lead to clinicalfailure [8083]. For isolates that are erythromycin resistant and clindamycinsusceptible, the D diusion zone test can be used to detect the presence ofthe inducible phenotype. An erythromycin disk (15 lg) is placed between 15and 26 mm away from a clindamycin disk (2 lg) and observed for bluntingor formation of a D shape around the clindamycin zone resulting frominduced resistance (Fig. 2). The streptogramin combination quinupristin-dalfopristin is bactericidal for S aureus, but becomes bacteriostatic for thosewith constitutive expression of macrolide, lincosamide, and streptograminB. Clinical data suggest, however, that it does not impair the clinical ecacyso long as an adequate dose is given [84].

Other drugs reported selectively. Resistance to linezolid, the rst oxazolidi-none, is rare but has been documented [8587]. Accurate susceptibilitytesting requires an experienced technologist because interpreting results canbe dicult because the drug is bacteriostatic. At present, no NCCLSinterpretative criteria are available for a resistant category [7]. Fordaptomycin, MIC testing requires supplementation with 50 lg/mL ofcalcium to CAMHB and disk diusion testing should be performed oncation-adjusted MHA [7]. Trimethoprim-sulfamethoxazole is an alterna-tive to vancomycin in uncomplicated MRSA infections [8892]. Suscepti-bility testing can be done by MIC or disk diusion. To date, all threecases of VRSA reported by the CDC were susceptible to trimethoprim-sulfamethoxazole.

Several agents are used in various combinations to achieve synergy withcell wallactive agents or other bactericidal agents. Aminoglycosides are

411M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434Fig. 2. Disk diusion D test to detect the presence of the macrolide lincosamide

streptogramin b (MLSb) phenotype with inducible clindamycin resistance in a Staphylococcus

aureus isolate that was found to be resistant to erythromycin and sensitive to clindamycin on

routine susceptibility testing.

widely used in prosthetic valve endocarditis [9396]. In these situations,high-level aminoglycoside resistance implies that synergy will not beachieved [58]. As with aminoglycosides, uoroquinolone resistance is morecommon in MRSA than methicillin sensitive Staphylococcus aureus (MSSA)and resistance develops quickly. Rifampin should not routinely be reportedon S aureus isolates because resistance develops quickly when used asmonotherapy [97,98].

Coagulase-negative staphylococciThe mechanisms of resistance in coagulase-negative staphylococci are the

same as those of S aureus; however, resistance is usually expressed at a lowerlevel [51]. To enhance detection of methicillin-resistant strains, the MICbreakpoint for coagulase-negative staphylococci is less than or equal to 0.25lg/mL instead of 2 lg/mL, three dilutions lower than that for S aureus. Fordisk diusion, the susceptible breakpoint is greater than or equal to 18 mminstead of 13 mm. For more serious infections with coagulase-negativestaphylococci other than Staphylococcus epidermidis, NCCLS recommendstesting for the mecA gene or the gene product PBP2a when results are in theintermediate or resistant category [7]. Vancomycin resistance has beenrecognized and described in coagulase-negative staphylococci [99].

Enterococcus sppEnterococcus spp are part of the normal gastrointestinal ora; however,

they are an increasingly important cause of invasive disease includingbacteremia, meningitis, and endocarditis. Resistance in enterococci hasevolved over the past 30 years to become the prototype of multiresistantbacteria, carrying with it an independent risk of mortality [100]. Enterococciare intrinsically resistant to a large number of antibiotics includingcarboxypenicillins, cephalosporins, trimethoprim-sulfamethoxazole, andclindamycin [7]. These agents can appear susceptible in vitro, but do notwork in clinical situations. Ampicillin is the drug of choice for susceptibleenterococcal infections. Reduced anity for penicillin-binding proteinsmediates resistance to penicillin, ampicillin, and other b-lactams used intherapy including the ureidopenicillins and imipenem. Testing for PBP-mediated resistance can be done by either broth dilution or disk diusion.Another mechanism of resistance to penicillin in Enterococcus faecalis isb-lactamase production; this is much less common and the need for routinetesting is questionable [101,102]. NCCLS recommends routinely testingsterile site isolates. A positive b-lactamase test implies resistance to penicil-lin and ureidopenicillins, but susceptibility to imipenem and b-lactam

412 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434b-lactamase inhibitor combinations [103]. At present there are no NCCLScriteria for testing susceptibility to imipenem; however, a recent studyconrmed the widely held belief that ampicillin or penicillin may be used to

predict susceptibility to imipenem [104], although this practice has beenquestioned [105].

Vancomycin resistance was rst reported in Europe in 1988 [106] andsince that time has spread worldwide. The gene clusters responsible for high-level vancomycin resistance in Enterococcus faecium and E faecalis arepredominantly vanA and vanB. Other recently identied genes associatedwith vancomycin resistance include vanD (moderate level, MIC 16256 lg/mL); vanC; vanE; and vanG (intrinsic, low-level) [106110]. VanA and vanBphenotypes are highly transmissible, but vanC, found predominantly inEnterococcus gallinarum, Enterococcus avescens, or Enterococcus casseli-avus, is intrinsic and low level (MIC 816 lg/mL) and does not seem tocarry an infection control risk [111].

Most laboratories readily detect high-level vancomycin resistance. Sur-veys have shown, however, that laboratories that reported following NCCLSguidelines were, in practice, noncompliant [112,113]. The main area ofdiculty was with isolates that were intermediately resistant to vancomycin(MIC 816, vanC phenotype) [113]. NCCLS recommends testing by MICand further speciation of the isolate with a pigment and motility test to lookfor the vanC phenotype. Other problem areas noted in these studies werewith failure to use the correct medium (MHA or CAMHB) and failureto incubate for a complete 24 hours [112]. A comparison of agar dilution,broth microdilution, Etest, disk diusion, and automated Vitek methodsfor Enterococcus spp to vancomycin with the vancomycin-resistance geno-type found major errors primarily with the commercial method (Vitek) andwith disk diusion. Brain heart infusion (BHI) medium performed betterthan Mueller-Hinton [114].

Aminoglycosides are not clinically active for enterococci except whenused in conjunction with an active cell-wall agent (eg, ampicillin orvancomycin). For serious infections, particularly endocarditis, achievinga bactericidal eect is essential. Synergy between aminoglycosides and a cell-wall agent is predicted by using a high-level aminoglycoside screening test.For E faecium and E faecalis, resistance to gentamicin and streptomycin canbe tested by using high concentrations in a brain-heart infusion broth oragar for 24 hours. Any growth in broth is considered resistant. Streptomycinshould be incubated 48 hours if susceptible at 24 hours. A comparison wasmade of three NCCLS-approved screening methods for high-level amino-glycoside resistance: (1) Microscan broth microdilution, (2) synergy quadplate agar dilution, and (3) disk diusion. Screening methods showed anagreement of 99% for all three methods for high-level gentamicin and 96%for high-level streptomycin [115].

For clinically signicant isolates that are ampicillin and vancomycin

413M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434resistant, additional testing is required. Linezolid is bacteriostatic for both Efaecalis and E faecium. Quinupristin-dalfopristin is bactericidal (erm genenegative), but testing and reporting are recommended for E faecium only(less than 2% of E faecalis are susceptible). The drug daptomycin, newly

approved by the Food and Drug Administration, seems to be equallyeective against vancomycin-resistant enterococci and vancomycin-suscep-tible enterococci; however, these were interpreted with a tentative MIC ofless than or equal to 2 lg/mL as susceptible. NCCLS has since establishedan MIC of less than or equal to 1 as susceptible. There are no resistantcriteria established [116,117]. Other drugs that should be tested and areeective for all enterococci if susceptible are tetracycline and rifampin.

Streptococcus pneumoniaeStreptococcus pneumoniae is the leading bacterial cause of community-

acquired pneumonia and bacterial meningitis [118]. Subsequently, theemergence of penicillin resistance in the pneumococcus has becomea worldwide problem [119122]. Resistance results from complex alterationsin penicillin binding proteins caused by formation of mosaic genes thataect binding anity for all b-lactam antibiotics [123]. The extended-spectrum cephalosporins and carbapenems retain greater activity thanpenicillin [124,125]. To complicate therapeutic issues further, isolates haveoften acquired the transposon Tn1546 that possesses ermB, tetM, andaphA3 resulting in resistance to macrolides, tetracyclines, and aminoglyco-sides [126]. Other therapeutic options for multidrug-resistant isolates includethe newer uoroquinolones (levooxacin, moxioxacin, gemioxacin) andvancomycin. Resistance has begun to emerge among the uoroquinolones,currently at 1% in the United States. Of particular concern is a small groupof highly resistant clones that dominate multidrug-resistant S pneumoniae(eg, Spain 23F), which has disseminated to many parts of the world [127131]. Resistance to trimethoprim-sulfamethoxazole and macrolides is alsoincreasing, particularly among penicillin-resistant isolates [121,123].

Accurate susceptibility testing of S pneumoniae requires special con-ditions. Mueller-Hinton broth or agar is supplemented with 5% sheep bloodand incubated at 35(C for 20 to 24 hours in ambient air for broth dilutionor 5% CO2 for agar dilution. Broth or agar dilution may be used to test allantimicrobial agents recommended by NCCLS. No reliable interpretativecriteria exist, however, for disk diusion testing of b-lactam antibiotics on5% sheep blood MHA except for oxacillin [7]. Nonb-lactam antibioticsmay be tested by disk diusion. The Etest for penicillin, ceftriaxone, andvancomycin was found to be more than 98% accurate when compared withNCCLS methods on 5% sheep blood MHA, but for other antimicrobials isunreliable [15,132]. A small study looked at the D diusion zone test (seepreviously) compared with PCR for phenotypic detection of the mefA andermB in S pneumoniae and concluded that they could be determined by diskdiusion, but PCR was more reliable [133]. S pneumoniae poses a challenge

414 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434to commercial methods because of special nutritional and incubationrequirements. Commercial microdilution methods have been studied exten-sively and most current panels (frozen and dried) were found to be

comparable with NCCLS reference broth dilution in 2% to 5% lysed horseblood [15,134,135].

The MIC values are reported based on source using either meningitiscriteria or nonmeningitis criteria. For cerebrospinal uid isolates, onlypenicillin, ceftriaxone or cefotaxime, meropenem, and vancomycin arereported. The b-lactam antimicrobials are tested by MIC for cerebrospinaluid isolates. Interpretative criteria are dierent for ceftriaxone andcefotaxime based on source site: a cerebrospinal uid isolate is interpretedas susceptible onefold dilution lower that those from noncerebrospinaluid site. Vancomycin may be tested by either MIC (broth, agar, or Etest)or disk diusion. Laboratories may choose to report only the MIC values orthe MIC values with interpretations specic to the source of the isolate.

For noncerebrospinal uid source sites, b-lactams, macrolides, anduoroquinolones are reported. Disk diusion using a 1-lg oxacillin disk hasbeen advocated as a screen for penicillin susceptibility in pneumococcirecovered from nonsterile sites, such as the respiratory tract. A zonediameter of greater than or equal to 20 mm is considered susceptible toall b-lactam antibiotics including cephalosporins and carbapenems and nofurther testing is required. Isolates with a zone diameter of less than or equalto 19 mm to oxacillin, however, do not reliably predict resistance topenicillin. For these isolates, MICs should be used to determine suscepti-bility to penicillin, ceftriaxone or cefotaxime, and meropenem whenclinically indicated [7,136]. The laboratory should bypass the oxacillinscreen test on sterile-site isolates (eg, cerebrospinal uid and blood) and anMIC determination should be made available for penicillin, ceftriaxone,meropenem, or ertapenem on these isolates. Vancomycin susceptibility maybe determined by MIC or disk diusion. Although isolates recovered fromcerebrospinal uid and other sterile sites should receive anMIC test, a surveyby the Centers for Disease Control and Prevention showed that 53% oflaboratories performed the oxacillin-screening test inappropriately onisolates from sterile sites, delaying MIC results by more than 24 hours [118].

Viridans group and other streptococciViridans group streptococci refer to a group of streptococcal species that

are part of the normal ora in healthy humans: they are found on themucosa, gastrointestinal tract, and upper respiratory tract and as transientcommensals of skin. Most viridans streptococci are divided into ve speciesgroups: (1) mitis, (2) anginosus, (3) mutans, (4) salivarius, and (5) bovis. Thesmall (\0.5 mm) colony-forming b-hemolytic strains of Lanceeld group A,C, F, and G are included in the viridans anginosus group [137]. Viridansspecies are frequent contaminants and only clinically signicant isolatesfrom a normally sterile site should be routinely tested for susceptibility.

415M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434When performing susceptibility testing for viridans group streptococci,either disk diusion or MIC methods may be used for nonb-lactams. Aswith S pneumoniae, viridans group streptococci require sheep blood

supplemented MHA, direct colony inoculum preparation, and incubation in5% CO2 for 20 to 24 hours. Penicillin, ampicillin, and oxacillin diskdiusion testing is considered unreliable and should be tested by an MICmethod. For broth dilution, the CAMHB must be supplemented with lysedhorse blood (2%5%). MHA with sheep blood is used for agar dilution anddisk diusion. The sheep blood should be replaced with lysed horse bloodwhen testing a sulfonamide [7]. The Etest for b-lactam testing is convenientand reliable when compared with the reference agar dilution method, whichcan be time consuming and cumbersome [138].

For nutritionally decient streptococci (Abiotrophia defectiva or adaciens)susceptibility testing requires sheep blood Mueller Hinton (SBMH) agarsupplemented with pyridoxine (Vit B6) or cysteine. Etest is comparable withthe agar dilution reference method [139].

Other than the small colony variants, b-hemolytic streptococci (groups A,B, C, and G) are not fastidious and should be considered separately fromviridans group streptococci. Penicillin is considered uniformly active againstStreptococcus pyogenes (group A) and Streptococcus agalactiae (group B)and no susceptibility testing to penicillin is recommended. For the penicillin-allergic patient, testing is recommended for the nonb-lactam antibiotics.

Other fastidious organismsListeria monocytogenes is an opportunistic pathogen that is empirically

treated with ampicillin and gentamicin. Listeria spp have remained stable insusceptibility to antimicrobials over the past 50 years and most of the drugsremain active [140]. Susceptibility to ampicillin may be determined by MICin broth microdilution, CAMHB with lysed horse blood (2%5%) at 35(Cfor 16 to 20 hours [7]. Aminoglycosides may be added for synergy [141].Cephems may test susceptible, but are not eective clinically [7].Corynebacterium spp are frequent laboratory contaminants and no stan-dardized method for susceptibility testing is available. Vancomycin istraditionally the drug of choice for serious Corynebacterium jeikeiuminfections [142]. Like corynebacterium, Bacillus spp (not B anthracis) arefrequent environmental contaminants and have no standardized methodfor susceptibility testing. Bacillus cereus produces a broad-spectrum b-lactamase and is frequently resistant to penicillin, ampicillin, and cepha-losporins. It is also frequently resistant to trimethoprim-sulfamethoxazole,but usually susceptible to carbapenems, clindamycin, gentamicin, chloram-phenicol, and vancomycin [143].

Gram-negative organisms

Enterobacteriaceae

416 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434The family Enterobacteriaceae includes many species of aerobic orfacultatively anaerobic, gram-negative, nonspore-forming bacilli. Al-though these organisms are endogenous ora, several cause disease

including outpatient urinary tract infections and outbreaks of multidrug-resistant nosocomial infections. In fact, 20 species are responsible forapproximately 50% of all organisms isolated in the clinical microbiologylaboratory. Resistant phenotypes for most antibiotic classes have beendescribed and are increasingly common. Susceptibility testing should beperformed on all clinically signicant isolates in this group of organisms.NCCLS has specic standards for all Enterobacteriaceae. Either diskdiusion or MIC obtaining dilution methods may be performed usingreference procedures or comparable commercial products [7].

Resistance to b-lactams in Enterobacteriaceae is primarily mediated byan ever-increasing number of b-lactamases [144,145]. The presence of classA (Bush 2b) b-lactamases that confer resistance to penicillin and amino-penicillins is commonly detected among enteric gram-negative bacilli inclinical laboratories. Detection of Bush 2be ESBLs, AmpC (Bush 1,chromosomal or plasmid-mediated), metallo-b-lactamases, and other re-sistance mechanisms, however, creates a vexing problem for laboratories.

For the most part, ESBLs result from simple point mutations in the genesgenerally responsible for ampicillin resistance in Escherichia coli (TEM-1,TEM-2) and Klebsiella spp (SHV-1). Over 100 dierent ESBLs have beendetected in dierent species and present with a variety of susceptibilityproles [144]. The in vitro results for these isolates often do not follow thestandard hierarchy rules for cephalosporins. In particular, the cephamy-cins (cefoxitin and cefotetan) are often susceptible. Traditional breakpointsof extended-spectrum cephalosporins for Enterobacteriaceae do not reliablydetect the presence of these b-lactamases [24,146]. NCCLS recommends theuse of lower screening breakpoints for certain extended-spectrum cepha-losporins or aztreonam for E coli and Klebsiella spp [7]. Conrmation relieson the susceptibility of ESBLs to clavulanic acid. For broth microdilution,cefotaxime and ceftazidime are tested with and without clavulanic acid (4 lg/mL in broth, 10 lg in disk). A decrease of three dilutions or more in the MICor a 5 mm or greater increase in the zone size with clavulanic acid comparedwith the cephalosporin alone is indicative of ESBL production (Fig. 3).

Conrmed ESBL producers should be reported as resistant to penicillins(not including b-lactamb-lactamase inhibitor combinations); cephalospor-ins (excluding the cephamycins); and aztreonam [147,148]. Reporting thecephamycins and b-lactamb-lactamase inhibitor combinations is contro-versial, because no clinical data exist to support their use in these patients.For this reason, some laboratories choose to report these agents as resistant.Although resistance factors for other antibiotic classes are often found inthese organisms, the results of in vitro tests should not be altered outside ofroutine NCCLS reporting guidelines. Those isolates that are not conrmedto have ESBLs despite positive screening tests highlight the limitations of

417M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434phenotypic testing. In addition to an ESBL, this heterogeneous group oforganisms may contain alterations in porin production; hyperproductionof clavulanic acidsusceptible b-lactamase (including ESBLs); production of

a class I b-lactamase (AmpC) that is not clavulanic acid susceptible; ora combination of these resulting in the clavulanic acidresistant phenotype[149]. Although not formally addressed in the most recent guidelines, use ofthe NCCLS criteria for detecting ESBL production among Enterobacter-iaceae other than E coli and Klebsiella spp as routine practice in the UnitedStates is discouraged as a result of poor specicity [150].

Chromosomally mediated inducible AmpC is present in essentially allEnterobacter, Serratia, Providencia, Morganella morganii, Citrobacterfreundii, Hafnia alvei, and Aeromonas spp (in addition to Pseudomonasaeruginosa). NCCLS guidelines encourage reporting in vitro susceptibilitypatterns obtained in routine testing with a comment that these organismsmay develop resistance during prolonged treatment with third-generationcephalosporins and that repeat testing may be warranted 3 to 4 days afterinitiation of therapy. Detection of ESBLs is particularly dicult amongspecies or strains that co-produce an inhibitor-resistant b-lactamase, such asAmpC. For organisms that possess chromosomal AmpC and E coli orKlebsiella spp that contain plasmid-mediated AmpC, clavulanate may act asan inducer of high-level AmpC expression resulting in a false-negative ESBLtest. Methods that use either tazobactam or sulbactam as inhibitors havebeen considered for these organisms [151]. Also, high-level AmpC is muchless active on cefepime, so inclusion of cefepime as a screening agent or aspart of conrmatory testing coupled with an inhibitor has improvedsensitivity for ESBLs, but is not widely accepted as a standard method [152].

Carbapenem resistance in Enterobacteriaceae should be veried byadditional testing in the laboratory. Similarly, Klebsiella spp, Providenciaspp, Proteus vulgaris, C freundii, Enterobacter spp, or Serratia marcescenssusceptible to ampicillin or rst-generation cephalosporins require verica-

Fig. 3. Extended spectrum b-lactamase conrmation test on Klebsiella pneumoniae ATCCstrain 700603, a known ESBL producer (SHV-18). The zone of inhibition surrounding the disk

impregnated with both ceftazidime and clavulanic acid is more than 5 mm larger than the zone

around ceftazidime alone, demonstrating susceptibility to the b-lactamase inhibitor.

418 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434tion. Enterobacteriaceae may also possess numerous resistance factors foraminoglycosides, uoroquinolones, tetracyclines, chloramphenicol, andtrimethoprim-sulfamethoxazole. Fortunately, the reference methods for

susceptibility testing are generally accurate for the detection of thesephenotypes.

Special consideration should be given to the enteric pathogens Salmonellaand Shigella. For these pathogens, susceptibility results for rst- andsecond-generation cephalosporins and aminoglycosides can be misleadingand should not be reported. The 2004 NCCLS guidelines recommend testingnalidixic acid against invasive Salmonella isolates, acknowledging growingevidence that nalidixic acid resistance in Salmonella spp is associated withhigher ciprooxacin MICs and possibly an association with worse outcome[7,153,154].

Non-EnterobacteriaceaeLike the Enterobacteriaceae, most of the fast-growing, nonfermenting,

gram-negative bacilli can be tested using the published reference methodsor a wide variety of approved commercial methods. In addition toP aeruginosa, NCCLS provides specic interpretive guidelines for non-Enterobacteriaceae including Stenotrophomonas maltophilia, Burkholderiacepacia genomovars, and Acinetobacter spp. Other nonfastidious, glucosenonfermenting, gram-negative bacilli are usually interpreted using thePseudomonas guidelines.

Pseudomonas aeruginosa. Pseudomonas aeruginosa is infrequently part ofthe normal human ora. Hospitalization, however, may greatly increaserates of colonization. This colonization often precedes nosocomial infection.P aeruginosa possesses a number of resistance mechanisms for all majorantibiotic groups requiring that most severe infections have to be treated bymore than one eective antibiotic [155]. P aeruginosa isolates are intrin-sically resistant to narrow-spectrum penicillins, rst- and second-generationcephalosporins, and trimethoprim-sulfamethoxazole. P aeruginosa is usuallyfast growing, so susceptibility testing may be accomplished by any of thereference methods (agar dilution, broth dilution, or disk diusion) orapproved commercial systems. Because of the large number of inducibleresistance characteristics, repeat isolates after 3 or 4 days should beconsidered for retesting. For multidrug-resistant specimens, colistin andpolymyxin B are often considered. For these agents disk diusion is notrecommended and MIC interpretive criteria are not yet available [156].

Isolates of P aeruginosa obtained from patients with cystic brosis posea special problem for susceptibility testing and usually require separatetesting protocols. Multiple isolates are often obtained from single speci-mens, but mixed morphotype susceptibility should be discouraged [157,158].These isolates often grow slowly, produce a mucoid exopolysaccharide, andare multiply resistant after many years of antibiotic exposure [159].

419M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434Commercial microdilution systems have not performed well with theseisolates and their use is discouraged in favor of the reference disk diusion,broth microdilution, or Etest [160,161]. Incubation should be increased to

24 hours for the slow-growing isolates, but little is gained from extendedincubation to 48 hours [162]. Although not generally available in clinicallaboratories, synergy testing may be performed in a reference laboratory ona variety of antimicrobial combinations for particularly resistant isolates.The clinical value of this testing has not been established, but in vitro testinghas identied occasional synergistic and additive eects [163,164].

Stenotrophomonas maltophilia. Stenotrophomonas maltophilia is emergingas a leading cause of nosocomial infection, often second in frequency only toP aeruginosa [165]. The organism is not part of the normal ora and usuallyaects immunocompromised or debilitated patients [166]. Trimethoprim-sulfamethoxazole remains the treatment of choice despite increasing re-sistance [167]. Treatment is a signicant challenge secondary to inherentmultidrug resistance that aects many b-lactams, aminoglycosides, andother drug classes through b-lactamase, decreased outer membrane perme-ability, or multidrug resistance eux pumps [168]. Notably, most of theseisolates produce L1 (a zinc-dependent metallo-b-lactamase that doesnot respond to clavulanic acid), or L2 (an extended-spectrum cephalos-porinase that is inhibited by clavulanic acid). Independent studies havedemonstrated methodologic problems associated with susceptibility testingof S maltophilia [169]. Disk diusion particularly was problematic withciprooxacin and trimethoprim-sulfamethoxazole. Specic MIC break-points and new disk diusion recommendations were recently added,however, to the NCCLS guidelines [7]. Etest has been shown comparablewith broth microdilution. Many recent studies have demonstrated in vitrosusceptibility of the organism to new uoroquinolones, including levoox-acin, gatioxacin, and moxioxacin, but few clinical data are available.Testing a newer uoroquinolone in addition to minocycline should beconsidered for primary reporting. In vitro evidence of synergy has beendemonstrated with certain combinations including ticarcillin-clavulanatewith aztreonam [170]. Any carbapenem-susceptible isolate should be veriedbefore reporting.

Burkholderia cepacia. Organisms belonging to the B cepacia complex areproblematic for patients with cystic brosis and are also emerging asa nosocomial problem. The organism is very resistant to many agentscommonly used in patients with cystic brosis. Commercial identicationmethods have not performed well with this group of organisms. Brothmicrodilution or Etest are the methods most frequently used, although newbreakpoints for disk diusion are available in 2004 [7]. As with the mucoidPseudomonas, tests for synergy are often performed, but the usefulness ofthe data is controversial. Although tests of combinations of two and three

420 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434agents have been performed using dierent methodologies, consistent invitro synergy has not been observed and the patient response to therapywhen synergy is observed has not been conrmed [171].

Acinetobacter spp. This genus consists of strictly aerobic, gram-negativecoccobacillary rods that can be dicult to decolorize and may be mistakenfor gram-positive cocci in blood culture media. Susceptibility testing isrequired for all signicant strains and is performed using reference methods,including dilutional MIC or disk diusion.

Fastidious gram-negative organismsHaemophilus inuenzae. Globally, 5% to 40% of H inuenzae isolatesproduce b-lactamase (plasmid-mediated TEM-1) resulting in resistance toampicillin and amoxicillin [121,172]. Occasionally, resistant strains do notproduce b-lactamase, but have altered PBP. These isolates are referred to asb-lactamase negative, ampicillin resistant. Resistance to extended-spectrumcephalosporins has not been reported and resistance to new macrolides anduoroquinolones is rare; any of these phenotypes should be veried.Trimethoprim-sulfamethoxazole testing may be warranted because ofresistance rates of 10% to 45% [172,173]. Because resistance is predomi-nantly mediated by a predictable b-lactamase, most facilities continue onlyto screen. When indicated and requested, broth dilution or disk diusioncan be performed using a hematin, NAD (thymidine phosphorylase-fordilution tests only), and yeast extract supplemented MHB or MHA(Haemophilus test medium), inoculated directly from a chocolate plate.Growth on solid media requires 5% CO2. Agar dilution techniques have notbeen studied. Alternatively, commercial methods approved by the Food andDrug Administration, including Etest, are available.

Neisseria spp. Like H inuenzae, N gonorrhoeae can either producea plasmid-mediated b-lactamase or develop b-lactam resistance by alteredPBP encoded on the chromosome. Tetracycline resistance is also commonand uoroquinolone resistance is emerging in the United States. Resistanceto an extended-spectrum cephalosporin has not been reported in the UnitedStates and should be conrmed if identied in the laboratory. Agar dilutionor disk diusion on a supplemented gonococcal (GC) agar base inoculatedfrom direct growth and incubated in 5% CO2 for 20 to 24 hours is thereference method [7]. When testing carbapenems or clavulanate containingcompounds, cysteine should be left out of the GC supplement. Etest is theonly nonreference method to demonstrate comparable results [174]. Withthe introduction of molecular diagnostics, however, N gonorrhoeae suscep-tibility testing is not recommended and not practical, and is usually onlyperformed as part of surveillance programs.

Provisional breakpoints exist for N meningitidis tested by either brothdilution using CAMHB supplemented with 2% to 5% lysed horse blood or

421M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434agar dilution on MHA with 5% sheep blood incubated in 5% CO2. Etestresults with meningococci have demonstrated variable performance toreference methods [175177].

Other testing issuesMore than 90% of M catarrhalis produce b-lactamase. Reporting of the

presence of b-lactamase may be useful by reinforcing that ampicillin is notappropriate therapy.

For research purposes, the NCCLS has provided reference agar dilutionmethod for testing Helicobacter pylori [7]. In clinical situations, however, anEtest on SHMHA incubated in a microaerophilic atmosphere for 3 to 5 daysis a simpler approach [178] but may sacrice accuracy with metronidazole[179]. A similar approach is appropriate with most Campylobacter spp [180].

Bordetella, Legionella, Pasteurella, and HACEK organisms (Haemophi-lus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella) usually re-spond to medications of choice; are infrequently encountered; are hard togrow; and little has been accomplished toward standardization of testingmethods [181]. Additionally, most clinicians do not appreciate the com-plexity of the testing for these organisms and the potential diculties withinterpretation.

Anaerobes

Anaerobes, both strict and aerotolerant, are part of the normal humanora. Over 100 dierent species have been identied in clinical specimens.Approximately one third of those isolates are of the Bacteroides fragilisgroup; another third are peptostreptococcus; and the last third are usuallyPrevotella, Fusobacterium, or Clostridia spp Resistance, primarily becauseof b-lactamases, is increasing [182]. Although appropriate therapy foranaerobic infections has been associated with signicant reductions inmortality [183], most clinical laboratories still do not perform routineanaerobic susceptibility testing [184]. Reference methods are labor-in-tensive and interpretation may be complicated. NCCLS recommends thatisolates obtained from patients with brain abscess, endocarditis, osteomy-elitis, joint infection, grafts and prosthetic material, and bacteremia shouldbe considered for testing [185]. Virulent organisms with unpredictablesusceptibility include Bacteroides, Prevotella, Fusobacterium, Clostridium,Bilophila, and Sutterella. Agar dilution using Brucella blood agar supple-mented with hemin and vitamin K1 is the reference method [186].Microdilution (Brucella broth), however, is now recognized as a referencemethod for B fragilis [185]. Gradient methods, including both Etest [187]and spiral gradient, have been shown to be eective for certain anaerobes.Disk diusion is not an alternative. Testing for b-lactamase activity isa reasonable approach for nonB fragilis group organisms beforesusceptibility testing. It may not always predict hydrolysis of imipenemand cephamycins, however, and other mechanisms for b-lactam resistance

422 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434may be present. Most facilities either send signicant isolates to a referencelaboratory on request or perform annual reviews to assess the prevalenceof resistance.

Agents associated with bioterrorism

Public health authorities should be notied immediately on the pre-liminary identication of any pathogen potentially related to a deliberaterelease. The three Centers for Disease Control and Prevention class Abacterial agents (Bacillus anthracis, Yersinia pestis, Francisella tularensis),although infrequent, are endemic to the United States and may be routinelyisolated in a clinical laboratory. Biosafety level 2 with biosafety level 3personnel precautions is appropriate for diagnostic quantities of infectiouscultures, but ideally any subsequent conrmation or susceptibility workshould take place in an approved reference or public health facility. Despiteprecautions, recent exposures of laboratory workers to agents associatedwith biological terrorism have been documented, including some resulting inclinical disease [188,189].

The 2004 NCCLS guidelines contain interpretive standards and recom-mendations for performance variables and quality control organisms forbroth microdilution with B anthracis and Y pestis [7]. A comparison of 50historical and 15 recent B anthracis isolates demonstrated that although theEtest was comparable with broth microdilution, the MICs were one to ninedilutions lower and reading the results through biosafety cabinets wasdicult [190]. One of these isolates was b-lactamase positive and resistant topenicillin, but resistance to uoroquinolones, tetracyclines, clindamycin,and vancomycin was not identied. In broth microdilution Y pestis requires24 hours incubation and potentially 48 hours if growth is inadequate [7].Although Y pestis may seem susceptible to b-lactam agents in vitro, theseagents lack ecacy in animal models and should not be reported [191].

Summary

Despite ongoing eorts to curtail antibiotic use and implement aggressiveinfection control eorts, emergence of new resistant pathogens continues tobring more challenges to the clinician and the clinical microbiologylaboratory. Clinicians and laboratory personnel must understand thelimitations of current susceptibility testing methods and work together toimprove clinical outcome and reduce emergence of resistance when possible.Although genotypic methods have an increasing role in many laboratories,conventional phenotypic susceptibility testing will maintain its central rolefor the foreseeable future.

References

423M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434[1] Bates DW, Goldman L, Lee TH. Contaminant blood cultures and resource utilization:

the true consequences of false-positive results. JAMA 1991;265:3659.

[2] Ferraro MJ. Should we reevaluate antibiotic breakpoints? Clin Infect Dis 2001;33 (Suppl

3):S2279.

[3] Gould IM, MacKenzie FM, Struelens MJ, van der Meer JM. Towards a European

strategy for controlling antibiotic resistance Nijmegen, Holland August 2931, 1999. Clin

Microbiol Infect 2000;6:6704.

[4] Jorgensen JH, Ferraro MJ. Antimicrobial susceptibility testing: general principles and

contemporary practices. Clin Infect Dis 1998;26:97380.

[5] Reller LB, Schoenknecht FD, Kenny MA, Sherris JC. Antibiotic susceptibility testing of

Pseudomonas aeruginosa: selection of a control strain and criteria for magnesium and

calcium content in media. J Infect Dis 1974;130:45463.

[6] DAmato RF, Thornsberry C, Baker CN, Kirven LA. Eect of calcium and magnesium

ions on the susceptibility of Pseudomonas species to tetracycline, gentamicin polymyxin B,

and carbenicillin. Antimicrob Agents Chemother 1975;7:596600.

[7] NCCLS. Performance standards for antimicrobial susceptibility testing. Fourteenth

Informational Supplement, vol. 4, no. 1. Wayne, PA: NCCLS.

[8] European Committee for Antimicrobial Susceptibility Testing (EUCAST) of the

European Society of Clinical Microbiology and Infectious Diseases (ESCMID).

EUCAST Denitive Document E.DEF 2.1, August 2000: Determination of antimicrobial

susceptibility test breakpoints. Clin Microbiol Infect 2000;6:5702.

[9] NCCLS. Quality assurance for commercially prepared microbiological culture media:

approved standard. 2nd edition. Wayne (PA): NCCLS; 1996.

[10] NCCLS. Methods for dilution antimicrobial susceptibility tests for bacteria that grow

aerobically; approved standard. 6th edition. Wayne (PA): NCCLS; 2003.

[11] Jorgensen JH, Turnidge JD. Susceptibility test methods: dilution and disk diusion

methods. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH, editors.

Manual of clinical microbiology. Washington: American Society of Microbiology; 2003.

p. 110827.

[12] Labbe AC, Bourgault AM, Vincelette J, Turgeon PL, Lamothe F. Trends in antimicrobial

resistance among clinical isolates of the Bacteroides fragilis group from 1992 to 1997 in

Montreal, Canada. Antimicrob Agents Chemother 1999;43:25179.

[13] Bauer AW, Kirby WM, Sherris JC, Turck M. Antibiotic susceptibility testing by

a standardized single disk method. Am J Clin Pathol 1966;45:4936.

[14] Jorgensen JH, Ferraro MJ. Antimicrobial susceptibility testing: special needs for

fastidious organisms and dicult-to-detect resistance mechanisms. Clin Infect Dis 2000;

30:799808.

[15] Evangelista AT, Truant AL. Rapid systems and instruments for antimicrobial

susceptibility testing of bacteria. In: Truant AL, editor. Manual of commercial methods

in medical microbiology. Washington: American Society of Microbiology; 2002.

p. 41329.

[16] Ferraro MJ, Jorgensen JH. Susceptibility testing instrumentation and computerized

expert systems for data analysis and interpretation. In: Murray PR, Baron EJ, Jorgensen

JH, Pfaller MA, Yolken RH, editors. Manual of clinical microbiology. Washington:

American Society of Microbiology; 1999. p. 1593600.

[17] Food and Drug Administration. Review criteria for assessment of antimicrobial

susceptibility. Rockville (MD): Food and Drug Administration; 1991.

[18] Korgenski EK, Daly JA. Evaluation of the BIOMIC video reader system for determining

interpretive categories of isolates on the basis of disk diusion susceptibility results. J Clin

Microbiol 1998;36:3024.

[19] Medeiros AA, Crellin J. Evaluation of the Sirscan automated zone reader in a clinical

microbiology laboratory. J Clin Microbiol 2000;38:168893.

[20] Nijs A, Cartuyvels R, Mewis A, Peeters V, Rummens JL, Magerman K. Comparison and

evaluation of Osiris and Sirscan 2000 antimicrobial susceptibility systems in the clinical

424 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434microbiology laboratory. J Clin Microbiol 2003;41:362730.

[21] Barenfanger J, Drake C, Kacich G. Clinical and nancial benets of rapid bacterial

identication and antimicrobial susceptibility testing. J Clin Microbiol 1999;37:14158.

[22] Doern GV, Vautour R, Gaudet M, Levy B. Clinical impact of rapid in vitro susceptibility

testing and bacterial identication. J Clin Microbiol 1994;32:175762.

[23] Jett B, Free L, Sahm DF. Factors inuencing the Vitek gram-positive susceptibility

systems detection of vanB-encoded vancomycin resistance among enterococci. J Clin

Microbiol 1996;34:7016.

[24] Katsanis GP, Spargo J, Ferraro MJ, Sutton L, Jacoby GA. Detection of Klebsiella

pneumoniae and Escherichia coli strains producing extended-spectrum beta-lactamases.

J Clin Microbiol 1994;32:6916.

[25] Tenover FC, Swenson JM, OHara CM, Stocker SA. Ability of commercial and reference

antimicrobial susceptibility testing methods to detect vancomycin resistance in

enterococci. J Clin Microbiol 1995;33:15247.

[26] Doern GV, Scott DR, Rashad AL, Kim KS. Evaluation of a direct blood culture disk

diusion antimicrobial susceptibility test. Antimicrob Agents Chemother 1981;20:6968.

[27] Mirrett S, Reller LB. Comparison of direct and standard antimicrobial disk susceptibility

testing for bacteria isolated from blood. J Clin Microbiol 1979;10:4827.

[28] Ling TK, Liu ZK, Cheng AF. Evaluation of the VITEK 2 system for rapid direct

identication and susceptibility testing of gram-negative bacilli from positive blood

cultures. J Clin Microbiol 2003;41:47057.

[29] Baker CN, Stocker SA, Culver DH, Thornsberry C. Comparison of the E Test to agar

dilution, broth microdilution, and agar diusion susceptibility testing techniques by using

a special challenge set of bacteria. J Clin Microbiol 1991;29:5338.

[30] Huang MB, Baker CN, Banerjee S, Tenover FC. Accuracy of the E test for determining

antimicrobial susceptibilities of staphylococci, enterococci, Campylobacter jejuni, and

gram-negative bacteria resistant to antimicrobial agents. J Clin Microbiol 1992;30:

32438.

[31] Schulz JE, Sahm DF. Reliability of the E test for detection of ampicillin, vancomycin, and

high-level aminoglycoside resistance in Enterococcus spp. J Clin Microbiol 1993;31:

33369.

[32] Walsh TR, Bolmstrom A, Qwarnstrom A, Gales A. Evaluation of a new Etest for

detecting metallo-beta-lactamases in routine clinical testing. J Clin Microbiol 2002;40:

27559.

[33] Fuchs PC, Barry AL, Brown SD. Evaluation of daptomycin susceptibility testing by Etest

and the eect of dierent batches of media. J Antimicrob Chemother 2001;48:55761.

[34] Hill GB. Spiral gradient endpoint: a new method of susceptibility testing. Hosp Pract (O

Ed) 1990;25(Suppl 4):317.

[35] Craig WA. Qualitative susceptibility tests versus quantitative MIC tests. Diagn Microbiol

Infect Dis 1993;16:2316.

[36] Jorgensen JH. Selection of antimicrobial agents for routine testing in a clinical

microbiology laboratory. Diagn Microbiol Infect Dis 1993;16:2459.

[37] Greenwood D. Detection of antibiotic resistance in vitro. Int J Antimicrob Agents 2000;

14:3036.

[38] Peterson LR, Shanholtzer CJ. Tests for bactericidal eects of antimicrobial agents:

technical performance and clinical relevance. Clin Microbiol Rev 1992;5:42032.

[39] Shanholtzer CJ, Peterson LR, Mohn ML, Moody JA, Gerding DN. MBCs for

Staphylococcus aureus as determined by macrodilution and microdilution techniques.

Antimicrob Agents Chemother 1984;26:2149.

[40] Wolfson JS, Swartz MN. Drug therapy: serum bactericidal activity as a monitor of

antibiotic therapy. N Engl J Med 1985;312:96875.

[41] Hallander HO, Dornbusch K, Gezelius L, Jacobson K, Karlsson I. Synergism between

aminoglycosides and cephalosporins with antipseudomonal activity: interaction index and

425M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434killing curve method. Antimicrob Agents Chemother 1982;22:74352.

[42] Woods GL. Antibacterial therapy: In vitro testing, pharmacodynamics, pharmacology,

new agents. Infect Dis Clin N Am 1995;3:46281.

[43] Rasheed JK, Tenover FC. Detection and characterization of antimicrobial resistance

genes in bacteria. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH,

editors. Manual of clinical microbiology. Washington: American Society of Microbiol-

ogy; 2003. p. 1196212.

[44] Deguchi T, Yasuda M, Asano M, Tada K, Iwata H, Komeda H, et al. DNA gyrase

mutations in quinolone-resistant clinical isolates of Neisseria gonorrhoeae. Antimicrob

Agents Chemother 1995;39:5613.

[45] Rasheed JK, Jay C, Metchock B, Berkowitz F, Weigel L, Crellin J, et al. Evolution of

extended-spectrum beta-lactam resistance (SHV-8) in a strain of Escherichia coli during

multiple episodes of bacteremia. Antimicrob Agents Chemother 1997;41:64753.

[46] Rasheed JK, Anderson GJ, Yigit H, Queenan AM, Domenech-Sanchez A, Swenson JM,

et al. Characterization of the extended-spectrum beta-lactamase reference strain,

Klebsiella pneumoniae K6 (ATCC 700603), which produces the novel enzyme SHV-18.

Antimicrob Agents Chemother 2000;44:23828.

[47] Arbique J, Forward K, Haldane D, Davidson R. Comparison of the Velogene rapid

MRSA identication assay, Denka MRSA-screen assay, and BBL crystal MRSA ID

system for rapid identication of methicillin-resistant Staphylococcus aureus. Diagn

Microbiol Infect Dis 2001;40:510.

[48] Marshall A, Hodgson J. DNA chips: an array of possibilities. Nat Biotechnol 1998;16:

2731.

[49] Archer GL. Staphylococcus aureus: a well-armed pathogen. Clin Infect Dis 1998;26:

117981.

[50] Chang FY, Peacock JE Jr, Musher DM, Triplett P, MacDonald BB, Mylotte JM, et al.

Staphylococcus aureus bacteremia: recurrence and the impact of antibiotic treatment in

a prospective multicenter study. Medicine (Baltimore) 2003;82:3339.

[51] Livermore DM. Antibiotic resistance in staphylococci. Int J Antimicrob Agents 2000;

16(Suppl 1):S310.

[52] Scriven JM, Silva P, Swann RA, Thompson MM, Naylor AR, Bell PR, et al. The

acquisition of methicillin-resistant Staphylococcus aureus (MRSA) in vascular patients.

Eur J Vasc Endovasc Surg 2003;25:14751.

[53] Simor AE, Ofner-Agostini M, Bryce E, Green K, McGeer A, Mulvey M, et al. The

evolution of methicillin-resistant Staphylococcus aureus in Canadian hospitals: 5 years of

national surveillance. CMAJ 2001;165:216.

[54] Katayama Y, Ito T, Hiramatsu K. A new class of genetic element, staphylococcus cassette

chromosome mec, encodes methicillin resistance in Staphylococcus aureus. Antimicrob

Agents Chemother 2000;44:154955.

[55] Crisostomo MI, Westh H, Tomasz A, Chung M, Oliveira DC, de Lencastre H. The

evolution of methicillin resistance in Staphylococcus aureus: similarity of genetic

backgrounds in historically early methicillin-susceptible and -resistant isolates and

contemporary epidemic clones. Proc Natl Acad Sci U S A 2001;98:986570.

[56] Schmitz FJ, Fluit AC, Gondolf M, Beyrau R, Lindenlauf E, Verhoef J, et al. The

prevalence of aminoglycoside resistance and corresponding resistance genes in clinical

isolates of staphylococci from 19 European hospitals. J Antimicrob Chemother 1999;43:

2539.

[57] Ito T, Katayama Y, Asada K, Mori N, Tsutsumimoto K, Tiensasitorn C, et al. Structural

comparison of three types of staphylococcal cassette chromosome mec integrated in the

chromosome in methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemo-

ther 2001;45:132336.

[58] Ma XX, Ito T, Tiensasitorn C, Jamklang M, Chongtrakool P, Boyle-Vavra S, et al. Novel

type of staphylococcal cassette chromosome mec identied in community-acquired

methicillin-resistant Staphylococcus aureus strains. Antimicrob Agents Chemother 2002;

426 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 40143446:114752.

[59] Finan JE, Rosato AE, Dickinson TM, Ko D, Archer GL. Conversion of oxacillin-

resistant staphylococci from heterotypic to homotypic resistance expression. Antimicrob

Agents Chemother 2002;46:2430.

[60] Huang MB, Gay TE, Baker CN, Banerjee SN, Tenover FC. Two percent sodium chloride

is required for susceptibility testing of staphylococci with oxacillin when using agar-based

dilution methods. J Clin Microbiol 1993;31:26838.

[61] Yamazumi T, Marshall SA, Wilke WW, Diekema DJ, Pfaller MA, Jones RN.

Comparison of the Vitek Gram-Positive Susceptibility 106 card and the MRSA-

screen latex agglutination test for determining oxacillin resistance in clinical blood-

stream isolates of Staphylococcus aureus. J Clin Microbiol 2001;39:536.

[62] Felten A, Grandry B, Lagrange PH, Casin I. Evaluation of three techniques for detection

of low-level methicillin-resistant Staphylococcus aureus (MRSA): a disk diusion method

with cefoxitin and moxalactam, the Vitek 2 system, and the MRSA-screen latex aggluti-

nation test. J Clin Microbiol 2002;40:276671.

[63] Skov R, Smyth R, Clausen M, Larsen AR, Frimodt-Moller N, Olsson-Liljequist B, et al.

Evaluation of a cefoxitin 30 microg disc on Iso-Sensitest agar for detection of methicillin-

resistant Staphylococcus aureus. J Antimicrob Chemother 2003;52(2):2047.

[64] Cavassini M, Wenger A, Jaton K, Blanc DS, Bille J. Evaluation of MRSA-Screen,

a simple anti-PBP 2a slide latex agglutination kit, for rapid detection of methicillin

resistance in Staphylococcus aureus. J Clin Microbiol 1999;37:15914.

[65] Chediac-Tannoury R, Araj GF. Rapid MRSA detection by a latex kit. Clin Lab Sci 2003;

16:198202.

[66] Marlowe EM, Linscott AJ, Kanatani M, Bruckner DA. Practical therapeutic application

of the oxoid PBP2 latex agglutination test for the rapid identication of methicillin-

resistant Staphylococcus aureus in blood cultures. Am J Clin Pathol 2002;118:28791.

[67] Jorgensen JH. Mechanisms of methicillin resistance in Staphylococcus aureus and

methods for laboratory detection. Infect Control Hosp Epidemiol 1991;12:149.

[68] Liu H, Buescher G, Lewis N, Snyder S, Jungkind D. Detection of borderline oxacillin-

resistant Staphylococcus aureus and dierentiation from methicillin-resistant strains. Eur

J Clin Microbiol Infect Dis 1990;9:71724.

[69] Massanari RM, Pfaller MA, Wakeeld DS, Hammons GT, McNutt LA, Woolson RF,

et al. Implications of acquired oxacillin resistance in the management and control of

Staphylococcus aureus infections. J Infect Dis 1988;158:7029.

[70] Proctor RA, Kahl B, Von Ei C, Vaudaux PE, Lew DP, Peters G. Staphylococcal small

colony variants have novel mechanisms for antibiotic resistance. Clin Infect Dis 1998;

27(Suppl 1):S6874.

[71] Hiramatsu K, Hanaki H, Ino T, Yabuta K, Oguri T, Tenover FC. Methicillin-resistant

Staphylococcus aureus clinical strain with reduced vancomycin susceptibility. J Antimicrob

Chemother 1997;40:1356.

[72] Tenover FC, Biddle JW, Lancaster MV. Increasing resistance to vancomycin and other

glycopeptides in Staphylococcus aureus. Emerg Infect Dis 2001;7:32732.

[73] Marlowe E, Cohen M, Hindler JF, Ward K, Bruckner DA. Practical strategies for

detecting and conrming vancomycin-intermediate Staphylococcus aureus: a tertiary-care

hospital laboratorys experience. J Clin Microbiol 2001;39:26379.

[74] Fridkin SK. Vancomycin-intermediate and -resistant Staphylococcus aureus: what the

infectious disease specialist needs to know. Clin Infect Dis 2001;32:10815.

[75] Chang S, Sievert DM, Hageman JC, Boulton ML, Tenover FC, Downes FP, et al.

Infection with vancomycin-resistant Staphylococcus aureus containing the vanA resistance

gene. N Engl J Med 2003;348:13427.

[76] Centers for Disease Control and Prevention. Brief report: vancomycin-resistant

staphylococcus aureusNew York, 2004. MMWR Morb Mortal Wkly Rep 2004;53:

427M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 4014343223.

[77] Lina G, Quaglia A, Reverdy ME, Leclercq R, Vandenesch F, Etienne J. Distribution of

genes encoding resistance to macrolides, lincosamides, and streptogramins among

staphylococci. Antimicrob Agents Chemother 1999;43:10626.

[78] Roberts MC, Sutclie J, Courvalin P, Jensen LB, Rood J, Seppala H. Nomenclature for

macrolide and macrolide-lincosamide-streptogramin B resistance determinants. Anti-

microb Agents Chemother 1999;43:282330.

[79] Volokhov D, Chizhikov V, Chumakov K, Rasooly A. Microarray analysis of

erythromycin resistance determinants. J Appl Microbiol 2003;95:78798.

[80] Frank AL, Marcinak JF, Mangat PD, Tjhio JT, Kelkar S, Schreckenberger PC, et al.

Clindamycin treatment of methicillin-resistant Staphylococcus aureus infections in child-

ren. Pediatr Infect Dis J 2002;21:5304.

[81] Panagea S, Perry JD, Gould FK. Should clindamycin be used as treatment of patients

with infections caused by erythromycin-resistant staphylococci? J Antimicrob Chemother

1999;44:5812.

[82] Rao GG. Should clindamycin be used in treatment of patients with infections caused by

erythromycin-resistant staphylococci? J Antimicrob Chemother 2000;45:715.

[83] Siberry GK, Tekle T, Carroll K, Dick J. Failure of clindamycin treatment of methicillin-

resistant Staphylococcus aureus expressing inducible clindamycin resistance in vitro. Clin

Infect Dis 2003;37:125760.

[84] Johnson AP, Livermore DM. Quinupristin/dalfopristin, a new addition to the anti-

microbial arsenal. Lancet 1999;354:20123.

[85] Gonzales RD, Schreckenberger PC, Graham MB, Kelkar S, DenBesten K, Quinn JP.

Infections due to vancomycin-resistant Enterococcus faecium resistant to linezolid. Lancet

2001;357:1179.

[86] Pillai SK, Sakoulas G, Wennersten C, Eliopoulos GM, Moellering RC Jr, Ferraro MJ,

et al. Linezolid resistance in Staphylococcus aureus: characterization and stability of

resistant phenotype. J Infect Dis 2002;186:16037.

[87] Tsiodras S, Gold HS, Sakoulas G, Eliopoulos GM, Wennersten C, Venkataraman L,

et al. Linezolid resistance in a clinical isolate of Staphylococcus aureus. Lancet 2001;

358:2078.

[88] Bishara J, Pitlik S, Samra Z, Levy I, Paul M, Leibovici L. Co-trimoxazole-sensitive,

methicillin-resistant Staphylococcus aureus, Israel, 19881997. Emerg Infect Dis 2003;9:

11689.

[89] Johnson JR. Linezolid versus vancomycin for methicillin-resistant Staphylococcus aureus

infections. Clin Infect Dis 2003;36:2367.

[90] Markowitz N, Quinn EL, Saravolatz LD. Trimethoprim-sulfamethoxazole compared

with vancomycin for the treatment of Staphylococcus aureus infection. Ann Intern Med

1992;117:3908.

[91] Schmitz FJ, Verhoef J, Fluit A, Heinz HP, Jones ME. Stability of the MICs of various

antibiotics in dierent clonal populations of methicillin-resistant Staphylococcus aureus.

J Antimicrob Chemother 1998;41:3113.

[92] Yeldandi V, Strodtman R, Lentino JR. In-vitro and in-vivo studies of trimethoprim-

sulphamethoxazole against multiple resistant Staphylococcus aureus. J Antimicrob

Chemother 1988;22:87380.

[93] Graham JC, Gould FK. Role of aminoglycosides in the treatment of bacterial

endocarditis. J Antimicrob Chemother 2002;49:43744.

[94] Mulazimoglu L, Drenning SD, Muder RR. Vancomycin-gentamicin synergism revisited:

eect of gentamicin susceptibility of methicillin-resistant Staphylococcus aureus. Antimi-

crob Agents Chemother 1996;40:15345.

[95] Mulligan ME, Murray-Leisure KA, Ribner BS, Standiford HC, John JF, Korvick JA,

428 M. Joyce, C.W. Woods / Infect Dis Clin N Am 18 (2004) 401434et al. Methicillin-resistant Staphylococcus aureus: a consensus review of the micro-

biology, pathogenesis, and epidemiology with implications for prevention and manage-

ment. Am J Med 1993;94:31328.

[96] Watanakunakorn C, Tisone JC. Synergism between vancomycin and gentamicin or

tobramycin for methicillin-susceptible and methicillin-resistant Staphylococcus aureus

strains. Antimicrob Agents Chemother 1982;22:9035.

[97] ONeill AJ, Cove JH, Chopra I. Mutation frequencies for resistance to fusidic acid and

rifampicin in Staphylococcus aureus. J Antimicrob Chemother 2001;47:64750.

[98] Wichelhaus TA, Boddinghaus B, Besier S, Schafer V, Brade V, Ludwig A. Biological cost

of rifampin resistance from the perspective of Staphylococcus aureus. Antimicrob Agents

Chemother 2002;46:33815.

[99] Schwalbe RS, Stapleton JT, Gilligan PH. Emergence of vancomycin resistance in

coagulase-negative staphylococci. N Engl J Med 1987;316:92731.

[100] Vergis EN, Hayden MK, Chow JW, Snydman DR, Zervos MJ, Linden PK, et al.

Determinants of vancomycin resistance and mortality rates in enterococcal bacteremia.

a prospective multicenter study. Ann Intern Med 2001;135:48492.

[101] Fontana R, Canepari P, Lleo MM, Satta G. Mechanisms of resistance of enterococci to

beta-lactam antibiotics. Eur J Clin Microbiol Infect Dis 1990;9:1035.

[102] Murdoch DR, Mirrett S, Harrell LJ, Monahan JS, Reller LB. Sequential emergence of

antibiotic resistance in enterococcal bloodstream isolates over 25 years. Antimicrob

Agents Chemother 2002;46:36768.

[103] Okhuysen PC, Singh KV, Murray BE. Susceptibility of beta-lactamase-producing

enterococci to piperacillin with tazobactam. Diagn Microbiol Infect Dis 1993;17:21924.

[104] Weinstein MP. Comparative evaluation of penicillin, ampicillin, and imipenem MICs

and susceptibility breakpoints for vancomycin-susceptible and vancomycin-resistant

Enterococcus faecalis and Enterococcus faecium. J Clin Microbiol 2001;39:272931.

[105] El Amin N, Wretlind B, Wenger A, Brandt V, Bille J. Ampicillin-sensitive, imipenem-

resistant strains of Enterococcus faecium. J Clin Microbiol 2002;40:738.

[106] Leclercq R, Derlot E, Duval J, Courvalin P. Plasmid-mediated resistance to vancomycin

and teicoplanin in Enterococcus faecium. N Engl J Med 1988;319:15761.

[107] Boyd DA, Cabral T, Van Caeseele P, Wylie J, Mulvey MR. Molecular characterization of

the vanE gene cluster in vancomycin-resistant Enterococcus faecalis N00410 isolated in

Canada. Antimicrob Agents Chemother 2002;46:19779.

[108] Depardieu F, Reynolds PE, Courvalin P. VanD-type vancomycin-resistant Enterococcus

faecium. Antimicrob Agents Chemother 2003;47:718.

[109] Dutta I, Reynolds PE. Biochemical and genetic characterization of the vanC-2 vanco-

mycin resistance gene cluster of Enterococcus casseliavus ATCC 25788. Antimicrob

Agents Chemother 2002;46:312532.

[110] Malani PN, Thal L, Donabedian SM, Robinson-Dunn B, Kauman CA, Chow JW, et al.

Molecular analysis of vancomycin-resistant Enterococcus faecalis from Michigan

hospitals during a 10 year period. J Antimicrob Chemother