799
Antimicrobial Susceptibility Testing: Special Needs for Fastidious Organisms
and Difficult-to-Detect Resistance Mechanisms
James H. Jorgensen1 and Mary Jane Ferraro2
From the 1Department of Pathology, The University of Texas Health
Science Center, San Antonio, Texas; and 2Microbiology Laboratory,
Massachusetts General Hospital, Boston, Massachusetts
Clinical microbiology laboratories are faced with the challenge of accurately detecting
emerging antibiotic resistance among a number of bacterial pathogens. In recent years, vancomycin resistance among enterococci has become prevalent, as has penicillin resistance and
multidrug resistance in pneumococci. More recently, strains of methicillin-resistant Staphylococcus aureus with reduced susceptibility to vancomycin have been encountered. In addition,
molecular techniques have demonstrated that there are still problems detecting methicillin
resistance in staphylococci, especially in coagulase-negative species. Among members of the
family Enterobacteriaceae, mutated b-lactamase enzymes may confer difficult-to-detect resistance to later-generation penicillins and cephalosporins. Anaerobic bacteria are no longer
entirely predictable in their susceptibility to agents that might be selected for empiric therapy.
Therefore, clinical microbiology laboratories may not be able to rely on a single susceptibility
testing method or system to detect all those emerging resistant or fastidious organisms. For
reliable detection, laboratories may need to employ conventional, quantitative susceptibility
testing methods or use specially developed, single concentration agar screening tests for some
resistant species. Certain of these screening tests are highly specific, while others may require
additional confirmatory testing for definitive results. Therefore, laboratories must retain the
versatility to apply several different approaches to detect resistance in both common and
infrequently encountered bacterial pathogens.
Clinical microbiology laboratories are faced with the challenge of accurately detecting emerging antibiotic resistance
among several important bacterial pathogens. Certain of these
are fastidious organisms that require enriched media and modified growth conditions for reliable susceptibility testing (e.g.,
Streptococcus pneumoniae), whereas some other organisms have
subtle or inducible resistance mechanisms that may require special screening tests for reliable recognition (e.g., vancomycinresistant enterococci, staphylococci with reduced susceptibility
to vancomycin, methicillin-resistant staphylococci, and gramnegative bacilli that produce extended-spectrum b-lactamases
[ESBLs]). Clinical laboratories may not be able to rely on a
single susceptibility testing method or commercial system to
detect all of these emerging resistant organisms. For reliable
detection, it may be necessary to employ conventional broth
or agar dilution MIC procedures, special fixed–drug-concentration screening tests, and modified antibiotic interpretive
breakpoints that increase the opportunity for recognizing resistant strains. However, with several of the screening methods
it is important to confirm presumptive resistance by performing
Received 2 December 1999; electronically published 28 April 2000.
Reprints or correspondence: Dr. James H. Jorgensen, Department of Pathology, The University of Texas Health Science Center, 7703 Floyd Curl
Drive, San Antonio, Texas 78284-7750 ([email protected]).
Clinical Infectious Diseases 2000; 30:799–808
q 2000 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2000/3005-0010$03.00
more definitive phenotypic tests (e.g., tests designed to detect
vancomycin-resistance or ESBL-mediated resistance). Despite
the need for special or alternative testing methods for several
bacterial pathogens, practical test methods are available for use
in both large and small laboratories. In the near future, it may
become practical to use sensitive and specific molecular diagnostic methods capable of detecting the genes most often associated with resistance.
When Standard Methods Must be Modified
or Alternative Methods Applied
In a previous article [1], methods for susceptibility testing of
common, nonfastidious bacteria were reviewed. These included
broth and agar dilution, antimicrobial gradient, disk diffusion,
and automated instrument systems for performance of susceptibility tests in clinical laboratories. Those methods have been
developed and standardized for testing bacteria that grow rapidly (in !24 h) in unsupplemented Mueller-Hinton broth or agar
during incubation in ambient air. At least 90% of clinical isolates can be tested by these methods, including staphylococci,
enterococci, Enterobacteriaceae, and Pseudomonas and Acinetobacter species. However, clinically important resistance has
emerged in Streptococcus pneumoniae, certain other Streptococcus species, Haemophilus influenzae, Neisseria gonorrhoeae,
and some anaerobic bacteria [2]. These organisms require more
complex growth media, usually containing blood or blood
800
Jorgensen and Ferraro
CID 2000;30 (May)
Table 1.
bacteria.
Media and methods recommended for testing fastidious
products, and other growth additives, and depending on the
method used, they may require a CO2-enriched or strictly anaerobic atmosphere of incubation. However, despite the need
for enriched media and different incubation atmospheres, other
parameters of standardized susceptibility tests for nonfastidious
bacteria are maintained. Necessary modifications of standard
methods will be described in this review.
Some important mechanisms of antimicrobial resistance
among nonfastidious bacteria can be difficult to detect reliably
by standard dilution, diffusion, or instrument-based testing
methods. Molecular diagnostic methods that demonstrate wellcharacterized genes encoding for resistance mechanisms are
emerging as the new benchmarks defining resistance in individual bacterial isolates [3]. However, special resistance screens
that use a critical concentration of an antimicrobial agent are
a practical method to accurately recognize methicillin resistance
in staphylococci, vancomycin and aminoglycoside resistance in
enterococci, or the presence of an ESBL in certain common
gram-negative bacteria [4]. The phenotypic approaches to defining subtle or difficult-to-detect resistance properties in common bacterial pathogens are also described below.
Testing of Nutritionally Fastidious Organisms
Facultative bacteria.
The standard broth microdilution,
agar dilution, and disk diffusion tests can be adapted to test
several nutritionally fastidious bacterial species by appropriate
modification of the test medium and the incubation conditions.
The NCCLS has developed and standardized testing methods
for Haemophilus species, N. gonorrhoeae, Streptococcus species
(including S. pneumoniae), and, more recently, Helicobacter pylori [5–7]. H. influenzae can be tested with use of a supplemented
Mueller-Hinton medium referred to as haemophilus test medium (HTM; table 1). HTM includes hematin, nicotinamide
adenine dinucleotide, and yeast extract as supplements for
growth of Haemophilus species, and it requires incubation in a
5% CO2 atmosphere when used in the agar formulation [6].
Despite these additives, the broth and agar versions of HTM
are transparent and only slightly darker than Mueller-Hinton
medium.
Quality control and interpretive criteria have been published
for a large number of parenteral and oral agents useful in treating infections caused by these organisms [7], although routine
testing of agents other than ampicillin and trimethoprim/sulfamethoxazole may not be needed to manage most respiratory
infections due to Haemophilus species [7]. Indeed, H. influenzae
isolates are predictably susceptible to a number of combinations
of b-lactam/b-lactamase inhibitors, later-generation cephalosporins, and fluoroquinolones [8].
N. gonorrhoeae can be tested by the agar dilution or disk
diffusion methods with use of a supplemented GC agar base
medium and incubation in 5% CO2 for 20–24 h [5, 6] (table 1).
Growth of N. gonorrhoeae relies on the “XV-like” supplement,
Organism, test method
Hemophilus species
Broth microdilution
E test
Disk diffusion
Neisseria gonorrhoeae
Agar dilution
Disk diffusion
Streptococcus pneumoniae
Broth microdilution
E test
Disk diffusion
Other Streptococcus species
Broth microdilution
E test
Disk diffusion
Anaerobic bacteria
Agar dilution
E test
Broth microdilution
Incubation
atmosphere
Recommended medium
Haemophilus test medium
Haemophilus test medium
Haemophilus test medium
Ambient air
5% CO2
5% CO2
a
GC agar with XV-like supplement
a
GC agar with XV-like supplement
5% CO2
5% CO2
MHB 1 3% lysed horse blood
MHA 1 5% sheep blood
MHA 1 5% sheep blood
Ambient air
5% CO2
5% CO2
MHB 1 3% lysed horse blood
MHA 1 5% sheep blood
MHA 1 5% sheep blood
Ambient air
5% CO2
5% CO2
Brucella agar with 5% sheep blood
Brucella agar with 5% sheep blood
Wilkins-Chalgren
Anaerobic
Anaerobic
Anaerobic
NOTE. GC, [author will define]; MHA, Mueller-Hinton agar; MHB, MuellerHinton broth.
a
A combination of vitamins and coenzymes needed for the growth of N.
gonorrhoeae.
which contains numerous vitamins and coenzymes required for
growth of this species. A cysteine-free version of the growth
supplement is required for agar dilution testing of carbapenem
antibiotics (e.g., imipenem and meropenem) or b-lactam combinations that include the b-lactamase-inhibitor clavulanic acid
[5]. However, cysteine does not significantly interfere with disk
diffusion testing of these antibiotics or with agar dilution testing
of any other drugs [5, 6].
Many clinical microbiology laboratories now use nonculture
methods (i.e., direct probe or nucleic acid amplification tests)
to detect N. gonorrhoeae directly in clinical samples and therefore do not recover viable gonococci for susceptibility testing.
For this reason and because gonococci continue to respond
effectively to the empirical therapeutic regimens currently recommended for treatment of gonorrhea [9], susceptibility testing
of N. gonorrhoeae isolates is performed primarily for surveillance purposes by public health reference laboratories.
S. pneumoniae and certain species of the viridans Streptococcus group have emerged with multiple antimicrobial resistance mechanisms within the past decade [10–12]. It is now
important for laboratories to test routinely pneumococcal isolates from most infections and, as a minimum, viridans streptococci from patients with bacteremia or endocarditis for susceptibility to penicillin [7]. Pneumococci should also be tested
against relevant extended-spectrum cephalosporins, macrolides,
and possibly other drug classes, depending on the type of infection [7].
The National Committee for Clinical Laboratory Standards
(NCCLS) has developed quality control and interpretive breakpoint criteria specific for pneumococci and other streptococci
CID 2000;30 (May)
Special Antimicrobial Susceptibility Testing
[5–7]. The NCCLS reference broth microdilution procedure for
these organisms is to use Mueller-Hinton broth supplemented
with 2%–5% lysed horse blood but with otherwise standard
inoculum and incubation conditions for 20–24 h [5] (table 1).
Commercial microdilution panels that closely resemble the
NCCLS reference method have received clearance from the
United States Food and Drug Administration (FDA) for use
in clinical laboratories [13]. In addition, the E test (AB BIODISK, Solna, Sweden) has become very popular for testing
pneumococci because of its simplicity and convenience [14].
Clinical laboratories may also employ the NCCLS disk diffusion test for most but not all drugs when testing pneumococci
and other streptococci [6]. The disk diffusion test uses sheep
blood supplemented with Mueller-Hinton agar and incubation
for 20–24 h in 5% CO2 [6] (table 1). For streptococcal isolates,
a CO2 atmosphere for incubation is important to prevent
growth failures. This requirement has been incorporated into
the standardized test and taken into account in the development
of quality control diameter ranges for zones of inhibition.
The pneumococcal strain ATCC 49619 has been designated
by the NCCLS for routine quality control of both MIC and
disk diffusion tests for all streptococci [5–7]. Despite its simplicity, the disk diffusion test for pneumococci has 2 shortcomings. The oxacillin disk can be used to “screen” for penicillin
susceptibility, which is indicated by a zone of inhibition >20
mm. Such strains are uniformly susceptible to other relevant
b-lactam antibiotics, including amoxicillin and most cephalosporins [7]. However, if the oxacillin zone of inhibition is !20
mm, it is necessary to determine a penicillin MIC to clarify
whether an isolate is resistant, intermediately resistant, or borderline susceptible to penicillin [7, 15]. Second, the disk test
does not provide acceptable accuracy with some of the most
important drugs (e.g., the extended-spectrum cephalosporins)
when tested against pneumococci [7, 15].
A convenient approach used by some laboratories is to apply
selected E test strips (e.g., penicillin and cefotaxime or ceftriaxone) along with disks for non–b-lactam drugs (e.g., erythromycin, trimethoprim/sulfamethoxazole, or a quinolone) on the
same large Mueller-Hinton sheep blood agar plate. Although
this is a very convenient test method, the laboratory personnel
should verify that the particular drug and specific organism
that they wish to test are a combination that has received FDA
clearance for clinical use.
The NCCLS reference methods that have been recently developed for H. pylori [7] are relevant at this time primarily for
research purposes, including for comparison of the in vitro
activities of new compounds under development for treatment
of H. pylori gastritis. Clinical laboratory testing of this species
is not recommended at present. In cases in which there appears
to be a clinical need for testing a particular isolate (such as
with apparent failure of empirical therapy), a laboratory might
consider using E test strips for ampicillin, clarithromycin, or
metronidazole with Mueller-Hinton 5% sheep blood agar plates
801
incubated in a microaerophilic environment at 357C for 3–5
days [16]. However, it should be noted that the E test has not
yet received FDA clearance for use with H. pylori.
Last, because of the predictable susceptibility of Moraxella
catarrhalis to agents other than penicillin or amoxicillin [8], the
NCCLS recommends direct b-lactamase testing of significant
isolates but not the use of growth-based susceptibility tests with
other antimicrobial agents [5–7].
Anaerobic bacteria. Susceptibility testing of obligately anaerobic bacteria is not a standard procedure in many clinical
microbiology laboratories at this time. This is due in part to
the success of many empirical antibiotic regimens recommended
for treating anaerobic infections and also to the technical difficulty of performing anaerobic testing. Many times, cultures
of samples taken from anaerobic infections reveal the presence
of multiple facultative and anaerobic species that must be systematically isolated and tested separately, which can take several days [17]. In addition, the NCCLS reference MIC method
for anaerobes is an agar dilution procedure that requires preparation of antibiotic dilutions and special media at the site of
testing [18]. Indeed, there is a serious lack of practical, commercially available reagents for performing anaerobic susceptibility testing.
Some laboratories deal with these challenges by testing anaerobic bacterial isolates only selectively. Laboratories may test
isolates from only selected sites (e.g., blood or material from
brain abscesses), the anaerobic species most likely to harbor
resistance mechanisms (e.g., members of the Bacteroides fragilis
group or certain Clostridium species), or by specific requests of
physicians on a case-by-case basis. Some other laboratories may
choose to perform periodic surveillance testing of stored clinical
isolates in order to generate local data to guide clinicians with
empirical therapy. However, the need for some form of routine
testing may become more important, since resistant anaerobes
have been found at some centers [17, 19].
The NCCLS recommends an agar dilution method as the
reference procedure for anaerobes. The method employs brucella agar supplemented with sheep blood and other growth
additives [18]. It uses an inoculum 1 log2 higher (i.e., 105 cfu/
spot) than is used with aerobic organisms and requires incubation in an anaerobic gas atmosphere for 48 h before determination of MICs. Because interpretation of the MIC endpoints can be more difficult than with aerobic bacteria, the
NCCLS publication on anaerobic testing includes color photos
that depict various endpoints and guides to their appropriate
interpretation.
Interpretive breakpoints specific for anaerobic bacteria and
anaerobic quality control strains have been recommended by
the NCCLS [18]. Because agar dilution is not a practical method
for most clinical laboratories, the NCCLS has also described
a broth microdilution procedure that may represent a more
practical choice for routine use [18]. It includes use of WilkinsChalgren broth, an inoculum of 1 3 10 6 cfu/mL (rather than
802
Jorgensen and Ferraro
5 3 10 5 cfu/mL, as used for aerobes), and incubation for 48 h
in an anaerobic atmosphere. However, because not all anaerobes grow reliably in this broth medium, NCCLS studies are
under way to identify a more satisfactory broth for use in microdilution tests. In the meantime, use of E test strips on brucella blood agar (incubated in an anaerobic atmosphere) represents a convenient but somewhat expensive method for testing
clinical isolates of anaerobes [20]. It is important to note that
disk diffusion or disk elution testing is not recommended for
anaerobes because such tests have produced numerous errors
[18].
Difficult-to-Detect Resistance Mechanisms in
Nonfastidious Organisms
Methicillin/oxacillin resistance in staphylococci. Staphylococci continue to be among the most common of all nosocomial
and community-acquired bacterial pathogens. Both Staphylococcus aureus and the coagulase-negative species of Staphylococcus may be methicillin/oxacillin-resistant because of the production of a special, low-affinity penicillin-binding protein, PBP
2a [21]. Production of PBP 2a results in broad resistance to
semisynthetic penicillins, cephalosporins, and carbapenems.
Methicillin-resistant strains are often (but not always) multiply
resistant to several other drug classes, including macrolides,
clindamycin, aminoglycosides, chloramphenicol, fluoroquinolones, and trimethoprim/sulfamethoxazole [21].
Accurate methods for detection of methicillin resistance have
been a particular concern of many laboratories. Because of
lingering difficulties with standard phenotypic susceptibility
testing methods and concern over the possibility of missing
some highly heterogeneous strains, the most definitive method
is now recognized to be the detection of the gene (mecA) that
encodes production of PBP 2a [3]. This gene can be detected
by a PCR method that uses published primers or by a direct
DNA probe method [3]. However, such methods are beyond
the scope of many clinical laboratories, and FDA-cleared commercial kits for performing them are not yet available.
The most practical and reliable phenotypic test for detection
of methicillin-resistant S. aureus (MRSA) appears to be the
NCCLS oxacillin-salt agar screening procedure [5]. The test is
performed by spot-inoculating the surface of a Mueller-Hinton
agar plate containing oxacillin (6 mg/mL) and 4% NaCl with a
0.5-McFarland-density direct inoculum suspension [5]. After a
full 24-h incubation at a temperature no higher than 357C, any
growth on the plate, when examined with transmitted light,
indicates MRSA. FDA-cleared oxacillin screening plates are
available commercially from several sources. It is important to
note that although this procedure is referred to as a screening
test, it is sufficiently accurate that no further testing is required
to confirm that an isolate is indeed resistant to oxacillin [4, 5].
The second most reliable phenotypic method for detection
of MRSA appears to be the NCCLS broth microdilution test
CID 2000;30 (May)
procedure with incorporation of 2% NaCl in cation-adjusted
Mueller-Hinton broth for testing of oxacillin [5]. The NCCLS
disk diffusion procedure that employs a 1-mg oxacillin disk is
also generally reliable for detection of MRSA but cannot practically take advantage of salt-supplemented medium because of
adverse affects on drugs of other classes that might also be
tested on the same plate [6]. Last, the E test has been shown
to be reliable for detection of MRSA if 2% NaCl is included
in the Mueller-Hinton agar [22]. With any of the above methods, the direct inoculum-suspension procedure should be used,
and tests should be incubated for a full 24 h at 357C (not 377C)
[4–6].
Detection of methicillin-resistant coagulase-negative staphylococci (MR-CNS) by phenotypic methods is even more difficult than recognition of MRSA. Several studies have shown
that oxacillin MICs with MR-CNS may be lower and disk
diffusion zone-of-inhibition diameters larger than those with
MRSA [23, 24]. Therefore, the MIC and zone of inhibition
diameter breakpoints developed for detection of MRSA (i.e.,
oxacillin MIC of > mg/mL) may not detect a number of MRCNS isolates. Moreover, oxacillin-salt agar screening also fails
to detect a number of CNS strains that have the mecA gene
[24].
The NCCLS has recently proposed that the most sensitive
phenotypic method for detection of MR-CNS is either a broth
microdilution determination of the oxacillin MIC or the oxacillin disk diffusion test with specially derived breakpoints specific for CNS [7]. Oxacillin susceptibility of CNS is now defined
by an MIC <0.25 mg/mL or zone of inhibition >18 mm,
whereas resistance is indicated by an oxacillin MIC >0.5 mg/
mL or zone of inhibition <17 mm [7]. It is important to note
that these breakpoints differ substantially from those still recommended for S. aureus and that the breakpoints were derived
on the basis of the presence or absence of mecA rather than
the pharmacokinetic properties of oxacillin [7]. Clinical outcome data, however, are sparse for strains of CNS with MICs
of 0.5–1 mg/mL.
In the past, the NCCLS provided a cautionary statement
indicating that in vitro testing of various cephalosporins, penicillin–b-lactamase–inhibitor combinations, and carbapenems
could lead to false-susceptible results with methicillin-resistant
strains. In such cases, laboratories were urged to not report the
apparent susceptibility of these agents or to report them as
resistant on the basis of the recognized poor clinical response
of methicillin-resistant staphylococci to any of the b-lactam
antibiotics. Because this practice was sometimes confusing to
laboratory personnel and physicians and because it involved
unnecessary testing, the NCCLS now states that penicillin and
oxacillin are the only b-lactams that need be tested routinely
against staphylococci [5–7].
Susceptibility to all other currently available b-lactams can
be accurately deduced from the results of testing only penicillin
and oxacillin, as follows. Penicillin-susceptible strains can be
CID 2000;30 (May)
Special Antimicrobial Susceptibility Testing
safely assumed to be susceptible to all b-lactams with recognized activity against staphylococci. Penicillin-resistant, oxacillin-susceptible strains are susceptible to various b-lactams,
with the exception of b-lactamase–labile penicillins such as ampicillin, mezlocillin, piperacillin, and ticarcillin. Staphylococci
resistant to oxacillin are resistant to all currently available blactam antibiotics, including the penicillin–b-lactamase-inhibitor combinations. Therefore, while susceptibility breakpoints
for various cephalosporins or other b-lactam drugs have been
retained for research or epidemiological purposes, it is not necessary or useful for clinical laboratories to test those agents
against staphylococci on a routine basis [7].
Diminished vancomycin susceptibility in staphylococci.
There has been great concern that the genes encoding vancomycin resistance in enterococci could be transferred to staphylococci, especially to MRSA, thereby creating vancomycinresistant S. aureus. The vanA gene cluster has been transferred
in vitro on a plasmid to a recipient S. aureus, thereby creating
a perfectly functional vancomycin-resistant strain [25]. While
this has not yet occurred in nature, strains with reduced vancomycin susceptibility have been reported from Japan and several sites within the United States [26–28]. These strains are
now referred to as vancomycin-intermediate S. aureus (VISA)
or glycopeptide-intermediate S. aureus (GISA). These strains
have demonstrated a vancomycin MIC of 8 mg/mL and are
associated with a thickened cell wall and increased levels of
some cell wall precursors [28, 29]. They have not been shown
to contain the vanA, vanB, or vanC sequences from Enterococcus species [27, 28]. In addition, vancomycin-intermediate isolates of coagulase-negative staphylococci, particularly Staphylococcus epidermidis and Staphylococcus haemolyticus, have
been reported in the past [30, 31].
The NCCLS changed the vancomycin disk diffusion breakpoints in 1998 to facilitate the detection of vancomycin resistance in staphylococci, should it emerge. However, even the
modified disk breakpoints have not proven to be sufficient for
detection of the VISA strains reported to date [29]. Tenover et
al. [29] have examined several conventional and commercial
susceptibility testing methods with the limited collection of
VISA isolates . Their findings can be summarized by saying
that most MIC methods that incorporate at least an overnight
incubation period (>16 h) appear to offer reliable detection of
VISA. In addition, the vancomycin screening agar developed
for detection of vancomycin-resistant enterococci (VRE) seems
to offer a very simple and inexpensive way to screen for VISA.
Those instruments currently marketed in the United States
that provide rapid susceptibility results should be used with
caution [29]. With any of the procedures, the Centers for Disease Control and Prevention (CDC) recommends that S. aureus
isolates with vancomycin MICs >4 mg/mL should be regarded
as “possibly resistant” and forwarded to a public health reference laboratory for further characterization [32].
Vancomycin resistance in enterococci. The prevalence of
803
VRE has increased sharply in the United States in the past
decade [33]. A committee of the CDC has published recommendations for preventing the spread of VRE within health
care facilities, and these include prompt and accurate detection
of such strains by clinical microbiology laboratories [33]. However, there have been problems with accurate detection of VRE
with use of certain conventional and commercial antimicrobial
susceptibility testing methods [34].
For this reason, the NCCLS has recommended that the simplest and most sensitive test for recognition of VRE is the test
that uses vancomycin screening agar, originally described by
Willey et al. [5, 35]. This method incorporates use of 6 mg/mL
of vancomycin in brain-heart infusion (BHI) agar, spot inoculation with 105 or 106 cfu, and incubation for a full 24 h at
357C (table 2). Studies comparing use of this method with the
presence of the vanA, vanB, or vanC genes have confirmed the
sensitivity of this approach for detection of VRE [36]. However,
it is important to note that vancomycin screening agar is not
entirely specific for vanA- and vanB-type resistance.
Enterococcus casseliflavus and Enterococcus gallinarum contain the vanC1 and vanC2 genes that result in an intrinsic intermediate level of vancomycin susceptibility (e.g., usually a
vancomycin MIC of 8 mg/mL) that does not represent “true”
vancomycin resistance (usually indicated by a vancomycin MIC
>32 mg/mL) [37]. Therefore, it is important to perform several
confirmatory tests on isolates of presumed VRE found on vancomycin screening agar. These tests should include a gram stain
(to detect Lactobacillus species, also vancomycin-resistant), biochemical tests for species-level identification (most VRE in the
United States are Enterococcus faecium), and determination of
the vancomycin MIC by either broth microdilution or the E
test with 24-h incubation [5, 38, 39]. It may also be helpful to
perform a test for motility and determine the ability to acidify
methyl-a-D-glucopyranoside (MGP), which are properties of
E. casseliflavus and E. gallinarum [40].
As suggested above, vancomycin MIC determinations based
on a conventional overnight incubation period are generally
reliable for detection of vancomycin resistance [34]. In addition,
the rapid Vitek instrument (bioMerieux Vitek, Hazelwood,
MO) has received FDA clearance for its updated approach to
detection of VRE. The reliability of the disk diffusion method
can be improved by careful inspection of apparent zones of
inhibition around a standard vancomcyin disk after a full 24h incubation at 357C [6]. Some strains of enterococci with inducible vanB-type resistance may give rise to tiny, pinpoint
colonies within an otherwise apparent zone of inhibition. Therefore, all vancomycin disk diffusion tests should be read with
use of a strong transmitted light source to allow visualization
of subtle in-growth colonies.
High-level aminoglycoside resistance in enterococci. Enterococci may be resistant to high concentrations of aminoglycosides and thus be refractory to the synergistic interaction of
aminoglycosides and cell wall–active antibiotics. Resistance to
804
Jorgensen and Ferraro
Table 2.
Phenotypic screening tests for detection of some important resistance mechanisms in nonfastidious bacteria.
Organism
Resistance trait
Screening method
Staphylococcus aureus
Methicillin/oxacillin
resistance
Oxacillin agar screen
CNS
Methicillin/oxacillin
resistance
Oxacillin MIC >0.5 mg/mL
S. aureus
Enterococcus species
Enterococcus species
Vancomycin intermediate
resistance
Vancomycin resistance
High-level aminoglycoside
resistance
Oxacillin zone of inhibition
<17 mm
Agar screen
Vancomycin agar screen
Agar screen
Broth screen (microdilution)
Broth screen (microdilution)
Escherichia coli and
Klebsiella species
NOTE.
CID 2000;30 (May)
Extended-spectrum
b-lactamase production
MIC >2 mg/mL (cefpodoxime,
ceftazidime, aztreonam,
cefotaxime, ceftriaxone)
Zone of inhibition
(cefpodoxime, <22 mm;
ceftazidime, <22 mm;
aztreonam, <27 mm;
cefotaxime, <27 mm;
ceftriaxone, <25 mm)
Medium
Confirmatory testing
Mueller-Hinton broth 1
2% NaCl 1 6 mg/mL
oxacillin
Mueller-Hinton broth
None
Mueller-Hinton agar
None
BHI agar 1 6 mg/mL
vancomycin
BHI agar 1 6 mg/mL
vancomycin
BHI agar 1 500 mg/mL
gentamicin
BHI agar 1 2000 mg/mL
streptomycin
BHI broth 1 500 mg/mL
gentamicin
BHI broth 1 1000 mg/mL
streptomycin
Mueller-Hinton broth
Vancomycin MIC
Mueller-Hinton agar
>5-mm increase in ceftazidime or cefotaxime zone
of inhibition with 10 mg
clavulanate
None
Vancomycin MIC; specieslevel identification
None
None
None
None
>3-fold decrease in MIC
with 4 mg/mL clavulanate
BHI, brain-heart infusion; CNS, coagulase-negative staphylococci.
streptomycin may be due to either alterations in the ribosomal
target or production of unique aminoglycoside modifying enzymes, particularly ANT(300/9) or ANT(9) [21]. Resistance to
gentamicin is most often mediated by production of the bifunctional enzyme AAC(60) 1 A PH(200) that inactivates all aminoglycosides except streptomycin [21].
Laboratories can test clinically important isolates of Enterococcus species for high-level gentamicin resistance with use of
a single 500-mg/mL concentration in either a BHI broth microdilution format or a BHI agar plate containing that concentration of gentamicin [5, 7]. Growth after 24 h in the presence of that concentration of the drug confirms high-level
gentamicin resistance, without the need for further testing. Gentamicin results predict those for tobramycin against Enterococcus faecalis but not against E. faecium, which is intrinsically
resistant to tobramycin [4].
Similarly, testing a 1000-mg/mL concentration of streptomycin by BHI broth microdilution or 2000 mg/mL with a BHI
agar dilution plate can establish the presence or absence of highlevel streptomycin resistance [5, 7]. Streptomycin results predict
resistance only to that agent and not to other aminoglycosides.
In the rare event that amikacin might be desired for combined
therapy, kanamycin testing would be required to predict susceptibility to amikacin [4].
Several commercial systems, including the Vitek and the
rapid version of the MicroScan WalkAway (Baxter Laboratories, West Sacramento, CA), now have FDA clearance for
high-level aminoglycoside resistance screening. In the majority
of cases, laboratories need only test gentamicin and streptomycin and should report the presence or absence of high-level
resistance to one or both of the aminoglycosides, without necessarily referencing the test concentrations of the drugs. Laboratories should also create concise comments that briefly explain the implications of high-level resistance for inclusion on
susceptibility reports. Testing for high-level aminoglycoside resistance is normally restricted to isolates from the blood or
CSF or isolates from bone or joint infections [5, 6]. There is
no evidence that other infections (e.g., urinary tract or wound
infections) require bactericidal therapy with an aminoglycoside
in combination with a cell wall–active agent.
Bush group I b-lactamases in some gram-negative bacteria.
Resistance to broad-spectrum penicillins and cephalosporins
may occur in some gram-negative bacteria that produce large
amounts of chromosomal Bush group I b-lactamase. This has
been described most often with regard to Enterobacter species,
Citrobacter freundii, Serratia marcescens, and Pseudomonas
aeruginosa [41]. The gene for production of Bush group I blactamase is usually under the control of a second, repressor
gene that allows production of only small amounts of enzyme
[42]. However, large amounts of the enzyme can be produced
by induction or derepression of the gene by certain highly enzyme-stable compounds, most notably cefoxitin and imipenem.
This can be demonstrated by performance of disk induction
tests that simply confirm that these organisms have the capa-
CID 2000;30 (May)
Special Antimicrobial Susceptibility Testing
bility of producing the enzyme in large amounts after induction
[42].
The phenomenon of derepression, however, is a reversible
induction associated with a return to production of low levels
of enzyme when the inducing drugs are removed. More significant is the fact that spontaneous mutations may occur in
the repressor gene, leading to irreversible constitutive production of Bush group I b-lactamase [42]. These spontaneous mutants may be selected during therapy for serious infections with
later-generation penicillins or cephalosporins, leading to therapeutic failures in some patients [43]. In such cases, susceptibility testing will indicate that the patient’s isolate is susceptible
to several later-generation b-lactams at the outset of therapy,
but resistance can be readily demonstrated in isolates 4–10 days
later, during therapy, if spontaneous mutants occur and are
“selected” by the presence of the drug [43].
There is no real value in performing the disk induction tests
on clinical isolates of Enterobacter, Citrobacter, or P. aeruginosa,
since virtually all isolates have the gene for producing the Bush
group I enzyme. However, laboratories may assist physicians
by including a brief comment on routine laboratory reports
that such resistance may occur during therapy with a penicillin
or cephalosporin [7]. In addition, laboratories should repeat
susceptibility tests on any isolates that appear to be the same
strains of these genera recovered from patients undergoing prolonged therapy for serious infections. Such repeated tests should
be performed every 3–4 days to detect the emergence of resistance as soon as possible [7].
Extended-spectrum b-lactamases in some Enterobacteriaceae.
Virtually every strain of Klebsiella pneumoniae produces a plasmid-mediated b-lactamase known as SHV-1, which normally
confers resistance to only ampicillin and ticarcillin [42]. A substantial percentage of contemporary Escherichia coli isolates
also produce a plasmid-mediated b-lactamase, known as TEM1, that confers ampicillin resistance [42]. However, spontaneous
mutations may occur in these very common enzymes that “extend” the spectrum to include hydrolysis of later-generation
penicillins, cephalosporins, and aztreonam.
These enzymes are now referred to as ESBLs, and they have
been subjected to the latest enzymology and genetic sequencing
techniques to characterize the precise mutations and unique
hydrolytic activities of each enzyme [44]. The unique ESBLs
TEM-2 through TEM-69 and SHV-2 through SHV-23 have
now been described [45]. These enzymes all appear capable of
hydrolyzing penicillins, cephalosporins, and aztreonam at a
higher-than-normal inoculum density in standard MIC tests
and have been associated with adverse outcomes for patients
treated with those agents and in nosocomial infection outbreaks
[46, 47]. For these reasons, it is important that clinical laboratories be able to recognize ESBL-producing isolates and alert
physicians to their presence.
Because of differences in drug substrate affinities of some
ESBLs and the resulting inoculum effects with standard sus-
805
ceptibility tests, ESBL-producing isolates may give rise to inconsistent or illogical results when tested by standard MIC and
disk diffusion tests. For example, an isolate might appear to
be resistant or intermediate to ceftazidime but susceptible to
cefotaxime and aztreonam. Such differences in activities with
those drugs should not occur with Klebsiella species or E. coli.
The NCCLS has dedicated considerable effort in the past few
years to standardizing methods for reliable recognition of
ESBL. The current recommendations are that the drugs that
are most readily hydrolyzed by ESBL should be tested as indicators of possible ESBL [5–7]. These include (in order of
preference for greatest sensitivity of detection of North American strains) cefpodoxime, ceftazidime, aztreonam, cefotaxime,
and ceftriaxone.
It is further suggested that special “screening” breakpoints
be applied to interpretation of the results of tests with the “indicator” drugs [5–7]. For example, MICs of 2–8 mg/mL for
those compounds are very unlikely against Klebsiella or E. coli
strains that have only native TEM-1 or SHV-1 enzymes, but
those MICs do not reach the breakpoint that has been used
until now to separate the susceptible from the intermediate or
resistant categories of those drugs (except cefpodoxime). The
NCCLS suggests that MICs >2 mg/mL for those agents are
presumptive evidence of the presence of an ESBL [7]. The same
principle may be applied to disk testing by use of smaller,
“screening” zone-of-inhibition diameters [6, 7] (table 2).
Most recently, the NCCLS has described phenotypic confirmatory tests for possible ESBL-producing strains detected
by means of the screening procedures described above [5–7].
Either MIC or disk diffusion tests may be performed with a
substrate cephalosporin tested alone and with the addition of
clavulanic acid. Since ESBLs are plasmid-mediated enzymes,
they are inhibited by the b-lactamase inhibitors, clavulanate,
sulbactam, and tazobactam. The NCCLS recommends use of
clavulanate added in fixed concentration to both ceftazidime
and cefotaxime for optimal confirmation of most ESBLs [7]
(table 2).
With the MIC procedure, the MIC of the cephalosporin plus
clavulanate should be at least 3 log2 dilutions (or 8-fold) lower
than when the cephalosporin is tested alone [5, 7]. In the disk
test, the diameter of the zone of inhibition should increase by
at least 5 mm in the presence of the clavulanate [6, 7]. These
confirmatory tests are probably most important with E. coli
and Klebsiella oxytoca isolates, in which excess production of
native b-lactamase (chromosomal AmpC and K1, respectively)
may occasionally give rise to resistance to cefpodoxime. In contrast, cefpodoxime is both a sensitive and specific indicator of
an ESBL in K. pneumoniae.
Another potential clue to the presence of an ESBL is coresistance to gentamicin and trimethoprim/sulfamethoxazole
that may be associated with resistance genes carried on the
same plasmid as those encoding the ESBL [48].
Laboratories should make reasonable efforts to detect ESBLs
806
Jorgensen and Ferraro
in these 2 very common gram-negative genera. When they are
detected and confirmed by the methods described above, the
laboratory should change the result category to “resistance”
for all of the penicillins, aztreonam, and the true cephalosporins
(but not the cephamycins, i.e, cefotetan, cefmetazole, and cefoxitin) [5–7]. Alternatively, a laboratory could include a warning statement that an isolate produces an ESBL and should be
considered as clinically resistant to the compounds noted above.
Direct detection of some enzymes that mediate resistance.
Simple, rapid tests are available that allow direct detection of
2 bacterial resistance agents; specifically, some b-lactamases and
chloramphenicol acetyl transferase [4]. b-lactamase tests may
incorporate penicillin as a substrate along with an indicator
system that detects its destruction by bacterial enzymes or may
use a chromogenic cephalosporin (e.g., nitrocefin) that changes
color when it is hydrolyzed by a b-lactamase [4]. With either
of these reagents, direct testing of bacterial colonies taken from
an initial growth plate can provide evidence within a few
minutes that an enzyme is present that hydrolyzes penicillin as
well as ampicillin and amoxicillin.
Unfortunately, there are only a few bacterial species for which
this approach can be applied: H. influenzae, N. gonorrhoeae,
M. catarrhalis, staphylococci, enterococci, and some anaerobic
bacterial species [4, 5, 17]. These tests do not detect resistance
to penicillins due to other mechanisms (e.g., alteration of PBPs)
or resistance to extended-spectrum b-lactam antibiotics; for example, these tests cannot be used to detect ESBL.
Chloramphenicol acetyl transferase (CAT) can be detected
by a rapid tube or filter paper test within 1–2 h and provide
evidence of chloramphenicol resistance in H. influenzae, S.
pneumoniae, and some Enterobacteriaceae [4]. While not all
resistance to chloramphenicol is mediated by CAT, the majority
of resistance among H. influenzae and S. pneumoniae isolates
can be attributed to production of that enzyme. However, because the use of chloramphenicol has diminished dramatically
in developed countries, the CAT test is rarely performed in
clinical laboratories.
Other problem organisms. There are several other species
of bacteria that may harbor important intrinsic or acquired
resistance mechanisms that make therapy problematic. Unfortunately, little has been accomplished toward standardization
of susceptibility testing of Corynebacterium species, Bacillus
species, fastidious gram-negative bacilli of the Haemophilas-Actinobacillus-Cardiobacterium-Eikenella-Kingella
(HACEK)
group, or Pasteurella species. In addition, Stenotrophomonas
maltophilia and Burkholderia cepacia present problems because
the interpretive-category results of disk diffusion or automatedinstrument tests may not agree with categories of susceptibility
based on conventional MIC determinations. The NCCLS currently recommends that only MIC testing should be performed
against S. maltophilia [5, 6]. Indeed, with organisms for which
disk diffusion interpretive breakpoints have not been specifically derived, it may be most prudent to rely on empirical reg-
CID 2000;30 (May)
imens supported by literature citations. If it is deemed necessary
to do a susceptibility test on a particular isolate, the most reliable approach may be to determine MICs of relevant drugs
by a conventional MIC procedure (e.g., broth microdilution).
Interpretation of the resulting MICs must then lie with the
infectious diseases specialist in light of the patient’s condition
and the previous experiences of the clinician.
Future Approaches to Detection of Antimicrobial
Resistance
The newest and perhaps most definitive approach to detection of antimicrobial resistance is the direct detection of genes
known to encode certain resistance mechanisms. It is possible
to detect mecA (the gene responsible for methicillin resistance
in Staphylococcus species), vanA and vanB (the genes responsible for vancomycin resistance in enterococci), and various blactamases and aminoglycoside-inactivating enzymes [3]. Methods that employ the use of genetic probes or nucleic acid
amplification techniques offer the promise of excellent sensitivity, specificity, and speed in the detection of resistance genes.
The detection of the genes for methicillin and vancomycin
resistance have become the benchmarks for evaluating new or
improved phenotypic tests for those resistance properties. Despite the accuracy of this approach with some resistance traits,
no molecular genetic tests are yet commercially available or
FDA-cleared for clinical laboratory use. Potential limitations
of this approach include the large number of different resistance
mechanisms that would need to be detected (e.g., many different
b-lactamases produced separately or together) and the possibility that certain genes that could be detected may not be
expressed or may not result in phenotypic resistance [3]. In some
instances multiple mutations result in remodeling of the molecular targets affected by antibiotics (e.g., PBPs in pneumococci) [21]. Therefore it might be necessary to sequence the
genes of interest in order to detect subtle mutations.
A recent technological advance that may facilitate both the
ability to probe for a large number of genes conveniently and
the ability to sequence other genes quickly and cheaply is the
application of computer chip technology to genetic analyses
[49]. Several companies are developing computer chip–based
products that will allow detection of as many as several hundred
resistance genes simultaneously or rapid sequencing of some
genes involved with resistance mediators. Therefore, the performance of highly definitive tests for antimicrobial resistance
may become a practical approach for routine but selective application in the future and could herald a new paradigm in
antimicrobial susceptibility testing.
References
1. Jorgensen JH, Ferraro MJ. Antimicrobial susceptibility testing: general principles and contemporary practices. Clin Infect Dis 1998; 26:973–80.
CID 2000;30 (May)
Special Antimicrobial Susceptibility Testing
2. Gold HS, Moellering RC Jr. Antimicrobial-drug resistance. N Engl J Med
1996; 335:1445–53.
3. Tenover FC, Rasheed JK. Genetic methods for detecting antibacterial and
antiviral resistance genes. In: Murray PR, Baron EJ, Pfaller MA, Tenover
FC, Yolken RH, eds. Manual of clinical microbiology. Washington, DC:
American Society for Microbiology, 1999:1578–92.
4. Swenson JM, Hindler JA, Peterson LR. Special phenotypic methods for
detecting antibacterial resistance. In: Murray PR, Baron EJ, Pfaller MA,
Tenover FC, Yolken RH, eds. Manual of clinical microbiology. Washington, DC: American Society for Microbiology, 1999:1563–77.
5. National Committee for Clinical Laboratory Standards. Methods for dilution
antimicrobial susceptibility tests for bacteria that grow aerobically. Approved standard M7-A5. Wayne, Pennsylvania: National Committee for
Clinical Laboratory Standards, 2000.
6. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial disk susceptibility tests. Approved standard M2A7. Wayne, Pennsylvania: National Committee for Clinical Laboratory
Standards, 2000.
7. National Committee for Clinical Laboratory Standards. Performance standards for antimicrobial susceptibility testing. Ninth informational supplement M100-S10. Wayne, Pennsylvania: National Committee for Clinical Laboratory Standards, 2000.
8. Doern GV, Jones RN, Pfaller MA, Kugler K, the SENTRY Participants
Group. Haemophilus influenzae and Moraxella catarrhalis from patients
with community-acquired respiratory tract infections: antimicrobial susceptibility patterns from the SENTRY antimicrobial surveillance program
(United States and Canada, 1997). Antimicrob Agents Chemother 1999;
43:385–9.
9. Centers for Disease Control and Prevention. 1998 Guidelines for treatment
of sexually transmitted diseases. MMWR Morb Mortal Wkly Rep 1998;
47(RR-1):59–69.
10. Butler JC, Hofmann J, Cetron MS, et al. The continued emergence of drugresistant Streptococcus pneumoniae in the United States: an update from
the Centers for Disease Control and Prevention’s pneumococcal sentinel
surveillance system. J Infect Dis 1996; 174:986–93.
11. Goldstein FW, Acar JF, the Alexander Project Collaborative Group. Antimicrobial resistance among lower respiratory tract isolates of Streptococcus pneumoniae: results of a 1992–1993 western Europe and USA collaborative surveillance study. J Antimicrob Chemother 1996; 38(Suppl A):
71–84.
12. Doern GV, Ferraro MJ, Brueggemann AB, Ruoff KL. Emergence of high
rates of antimicrobial resistance among viridans group streptococci in the
United States. Antimicrob Agents Chemother 1996; 40:891–4.
13. Jorgensen JH, McElmeel ML, Crawford SA. Evaluation of the Dade
MicroScan MICroSTREP antimicrobial susceptibility testing panel with
selected Streptococcus pneumoniae challenge strains and recent clinical
isolates. J Clin Microbiol 1998; 36:788–91.
14. Jorgensen JH, Ferraro MJ, McElmeel ML, Spargo J, Swenson JM, Tenover
FC. Detection of penicillin and extended-spectrum cephalosporin resistance among Streptococcus pneumoniae clinical isolates by the use of E
test. J Clin Microbiol 1994; 32:159–63.
15. Jorgensen JH, Swenson JM, Tenover FC, Ferraro MJ, Hindler JA, Murray
PR. Development of interpretive criteria and quality control limits for
broth microdilution and disk diffusion antimicrobial susceptibility testing
of Streptococcus pneumoniae. J Clin Microbiol 1994; 32:2448–59.
16. Hachem CY, Claridge JE, Reddy R, et al. Antimicrobial susceptibility testing
of Helicobacter pylori: comparison of E-test, broth microdilution, and disk
diffusion for ampicillin, clarithromycin, and metronidazole. Diagn Microbiol Infect Dis 1996; 24:37–41.
17. Hecht DW. Susceptibility testing of anaerobic bacteria. In: Murray PR,
Baron EJ, Pfaller MA, Tenover FC, Yolken RH, eds. Manual of clinical
microbiology. Washington, DC: American Society for Microbiology, 1999:
1555–62.
18. National Committee for Clinical Laboratory Standards. Methods for anti-
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
807
microbial susceptibility testing of anaerobic bacteria. Approved standard
M11-A4. Wayne, Pennsylvania: National Committee for Clinical Laboratory Standards, 1997.
Snydman DR, McDermott L, Cuchural GJ, et al. Analysis of trends in
antimicrobial resistance patterns among clinical isolates of Bacteroides
fragilis group species from 1990 to 1994. Clin Infect Dis 1996; 23(Suppl):
S54–65.
Citron DM, Ostavari A, Karlsson A, Goldstein EJC. Evaluation of the epsilometer (E test) for susceptibility testing of anaerobic bacteria. J Clin
Microbiol 1991; 29:2197–203.
Quintiliani R Jr, Sahm DF, Courvalin P. Mechanisms of resistance to antimicrobial agents. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC,
Yolken RH, eds. Manual of clinical microbiology. Washington, DC: American Society for Microbiology, 1999:1505–25.
Huang MB, Gay ET, 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; 31:2683–8.
McDonald CL, Maher WE, Fass RJ. Revised interpretation of oxacillin
MICs for Staphylococcus epidermidis based on mecA detection. Antimicrob Agents Chemother 1995; 39:982–4.
Marshal SA, Wilke WW, Pfaller MA, Jones RN. Staphylococcus aureus and
coagulase-negative staphylococci from bloodstream infections: frequency
of occurrence, antimicrobial susceptibility, and molecular (mecA) characterization of oxacillin resistance in the SCOPE program. Diagn Microbiol Infect Dis 1998; 30:205–14.
Noble WC, Virani Z, Cree RG. Co-transfer of vancomycin and other resistance genes from Enterococcus faecalis NCTC 12201 to Staphylococcus
aureus. FEMS Microbiol Lett 1992; 93:195–8.
Hiramatsu K, Hanaki H, Ino T, Yabuta K, Oguri T, Tenover FC. Methicillinresistant Staphylococcus aureus clinical strain with reduced vancomycin
susceptibility. J Antimicrob Chemother 1997; 40:135–40.
Smith TL, Pearson ML, Wilcox KR, et al. Emergence of vancomycin resistance in Staphylococcus aureus. N Engl J Med 1999; 340:493–501.
Sieradzki K, Roberts RB, Haber SW, Tomasz A. The development of vancomycin resistance in a patient with methicillin-resistant Staphylococcus
aureus infection. N Engl J Med 1999; 340:517–23.
Tenover FC, Lancaster MV, Hill BC, et al. Characterization of staphylococci
with reduced susceptibilities to vancomycin and other glycopeptides. J
Clin Microbiol 1998; 36:1020–7.
Schwalbe RS, Stapleton JT, Gilligan PH. Emergence of vancomycin resistance
in coagulase-negative staphylococci. N Engl J Med 1987; 316:927–31.
Sanyal D, Johnson AP, George RC, Cookson BD, Williams AJ. Peritonitis due to vancomycin-resistant Staphylococcus epidermidis [letter]. Lancet
1991; 337:54.
Centers for Disease Control and Prevention. Interim guidelines for prevention and control of staphylococcal infections associated with reduced susceptibility to vancomycin. MMWR Morb Mortal Wkly Rep 1997; 46:
626–8, 635.
Recommendations for preventing the spread of vancomycin resistance: recommendations of the Hospital Infection Control Practices Advisory Committee (HICPAC). Am J Infect Control 1995; 23:87–94.
Tenover FC, Tokars J, Swenson J, Paul S, Spitalney K, Jarvis W. Ability of
clinical laboratories to detect antimicrobial agent–resistant enterococci. J
Clin Microbiol 1993; 31:1695–9.
Willey BM, Kreiswirth BN, Simor AE, et al. Detection of vancomycin resistance in Enterococcus species. J Clin Microbiol 1992; 30:1621–4.
Swenson JM, Clark NC, Ferraro MJ, et al. Development of a standardized
screening method for detection of vancomycin-resistant enterococci. J Clin
Microbiol 1994; 32:1700–4.
Leclercq R, Corvalin P. Resistance to glycopeptides in enterococci. Clin Infect
Dis 1997; 24:545–56.
Facklam R, Hollis D, Collins MD. Identification of gram-positive coccal and
808
Jorgensen and Ferraro
coccobacillary vancomycin-resistant bacteria. J Clin Microbiol 1989; 27:
724–30.
39. Schulz JE, Sahm DF. Reliability of the E test for detection of ampicillin,
vancomycin, and high-level aminoglycoside resistance in Enterococcus species. J Clin Microbiol 1993; 31:3336–9.
40. Corvalho MDGS, Teixeira LM, Facklam RR. Use of tests for acidification
of methyl-a-D-glucopyranoside and susceptibility to efrotomycin for differentiation of strains of Enterococcus and some related genera. J Clin
Microbiol 1998; 36:1584–7.
41. Pfaller MA, Jones RN, Marshal SA, et al. Inducible ampC b-lactamase producing gram-negative bacilli from blood stream infections: frequency, antimicrobial susceptibility, and molecular epidemiology in a national surveillance program (SCOPE). Diagn Microbiol Infect Dis 1997; 28:211–9.
44.
45.
46.
47.
48.
42. Sanders CC, Sanders WE Jr. b-lactam resistance in gram-negative bacteria:
global trends and clinical impact. Clin Infect Dis 1992; 15:824–39.
43. Chow JW, Fine MJ, Shlaes DM, et al. Enterobacter bacteremia: clinical
49.
CID 2000;30 (May)
features and emergence of antibiotic resistance during therapy. Ann Intern
Med 1991; 115:585–90.
Jacoby GA. Extended-spectrum b-lactamases and other enzymes providing
resistance to oxyimino-b-lactams. Infect Dis Clin N Am 1997; 11:875–87.
Jacoby GA, Bush K. Amino acid sequences for TEM, SHV, and OXA
extended-spectrum b-lactamases. 1999; http://www.lahey.org/studies/webt
.htm. Accessed November, 1999.
Meyer KS, Urban C, Eagan JA, Berger BJ, Rahal JJ. Nosocomial outbreak
of klebsiella infection resistant to late-generation cephalosporins. Ann
Intern Med 1993; 119:353–8.
Rice LB, Eckstein EC, DeVene J, Shlaes DM. Ceftazidime-resistant Klebsiella
pneumoniae isolates recovered at the Cleveland Department of Veterans
Affairs Medical Center. Clin Infect Dis 1996; 23:118–24.
Katsanis GP, Spargo J, Ferraro MJ, Sutton L, Jacoby GA. Detection of
Klebsiella pneumoniae and Escherichia coli strains producing extendedspectrum b-lactamases. J Clin Microbiol 1994; 32:691–6.
Fodor SP, Rava RP, Huang XC, Pease AC, Homes CP, Adams CL. Multiplexed biological assays with biological chips. Nature 1993; 364:555–6.