J Antimicrob Chemother 2015; 70: 2038 – 2047
doi:10.1093/jac/dkv059 Advance Access publication 4 March 2015
Challenges to accurate susceptibility testing and interpretation
of quinolone resistance in Enterobacteriaceae: results of a
Spanish multicentre study
José-Manuel Rodriguez-Martinez1,2*, Jesús Machuca2,3, Jorge Calvo4,5, Paula Diaz-de-Alba1,
Cristina Rodrı́guez-Mirones4, Concha Gimeno6, Luis Martinez-Martinez4,5 and Álvaro Pascual1–3
1
Departamento de Microbiologı́a, Universidad de Sevilla, Sevilla, Spain; 2Spanish Network for Research in Infectious Diseases (REIPI
RD12/0015), Instituto de Salud Carlos III, Madrid, Spain; 3Unidad de Enfermedades Infecciosas y Microbiologı́a Clı́nica, Hospital
Universitario Virgen Macarena, Sevilla, Spain; 4Hospital Universitario Marques de Valdecilla and Valdecilla Biomedical Research Institute
(IDIVAL), Santander, Spain; 5Departamento de Biologı́a Molecular, Universidad de Cantabria, Santander, Spain; 6Servicio de Microbiologı́a,
Hospital General de Valencia, Valencia, Spain
*Corresponding author. Department of Microbiology, University of Seville, Avda Sanchez Pizjuan s/n, 41009, Spain. Tel: +34-954-55-28-63;
Fax: +34-954-37-74-13; E-mail: [email protected]
Received 15 December 2014; returned 16 January 2015; revised 6 February 2015; accepted 11 February 2015
Objectives: The objective of this study was to evaluate the proficiency of Spanish laboratories with respect to
accurate susceptibility testing and the detection and interpretation of quinolone resistance phenotypes in
Enterobacteriaceae.
Methods: Thirteen strains of Enterobacteriaceae were sent to 62 participating centres throughout Spain; strains
harboured GyrA/ParC modifications, reduced permeability and/or plasmid-mediated quinolone resistance genes.
The centres were requested to evaluate nalidixic acid and five quinolones, provide raw/interpreted clinical
categories and to detect/infer resistance mechanisms. Consensus results from reference centres were used to
assign minor, major and very major errors (mEs, MEs and VMEs, respectively).
Results: Susceptibility testing in the participating centres was frequently performed using the MicroScan
WalkAway, Vitek 2 and Wider systems (48%, 30% and 8%, respectively). CLSI/EUCAST breakpoints were used
in 71%/29% of the determinations. The percentage of VMEs for all quinolones was well below 2%. Only ofloxacin
and moxifloxacin showed higher values for raw VMEs (6.6%), which decreased to 0% and 2.9%, respectively, in
the interpreted VMEs. These errors were particularly associated with the CC-03 strain [qnrS2 + aac(6 ′ )-Ib-cr]. For
MEs, percentages were always ,10%, except in the case of ofloxacin and nalidixic acid. There was a significantly
higher percentage of all types of errors for strains whose MICs were at the border of clinical breakpoints.
Conclusions: The use of different breakpoints and methods, the complexity of mutation-driven and transferable
resistance mechanisms and the absence of specific tests for detecting low-level resistance lead to high variability
and represent a challenge to accuracy in susceptibility testing, particularly in strains with MICs on the border of
clinical breakpoints.
Keywords: QRDR, qnr, qepA, oqxAB, aac(6 ′ )-Ib-cr, efflux pumps, porins
Introduction
Over the past three decades, quinolone resistance has increased
in Enterobacteriaceae among human and veterinary isolates.
Fluoroquinolone resistance mainly occurs as a result of mutations
in chromosomal genes encoding the quinolone targets DNA gyrase and topoisomerase IV.1 Mutation-driven mechanisms also
include loss of porins and overexpression of one or more of several
efflux pumps (such as AcrAB-TolC), which can lead to resistance or
decreased susceptibility to b-lactams and fluoroquinolones.2
Three plasmid-mediated quinolone resistance (PMQR) mechanisms have also been described: (i) Qnr proteins (members of the
pentapeptide repeat protein family, which bind to DNA gyrase
and topoisomerase IV and protect the quinolone targets); (ii)
the Aac(6′ )-Ib-cr enzyme (which acetylates not only aminoglycosides but also ciprofloxacin and norfloxacin); and (iii) QepA and
OqxAB plasmid-mediated efflux pumps.3 Five types of qnr determinants encoded by the qnrA, qnrB, qnrS, qnrC and qnrD genes
# The Author 2015. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: [email protected]
2038
JAC
Challenges to accurate susceptibility testing and interpretation of quinolone resistance
Table 1. Panel of Enterobacteriaceae strains sent to the 62 participating centres
Strain
No.
Species and characteristics (known quinolone resistance mechanisms)
Source/reference
E. coli ATCC 25922
Ea-C2653
CC-00
CC-01
laboratory collection
27
Ecl-6087442
Kox-6059371
Cf-6067553
Ec-C1984
CC-02
CC-03
CC-04
CC-05
Ec-C1550
CC-06
Kpn-LB4
CC-07
Ec-25922 mut1
CC-08
Ec-25922 mut2
CC-09
Ec-25922 mut3
CC-10
Ec-25922 mut4
CC-11
Ec-38– 27
CC-12
CLSI control strain
Enterobacter aerogenes clinical isolate harbouring Asp86Tyr substitution in GyrA and the
aac(6 ′ )-Ib-cr gene
Enterobacter cloacae clinical isolate harbouring the qnrA1 gene
Klebsiella oxytoca clinical isolate harbouring qnrS2 and the aac(6 ′ )-Ib-cr gene
Citrobacter freundii clinical isolate harbouring the qnrB48 gene
E. coli clinical isolate harbouring Ser83Leu and Asp87Asn substitutions in GyrA, the Ser80Ile
substitution in ParC and the aac(6 ′ )-Ib-cr and oqxAB genes
E. coli clinical isolate harbouring Ser83Leu and Asp87Asn substitutions in GyrA, the Ser80Ile
substitution in ParC and the qepA1 gene
Klebsiella pneumoniae clinical isolate harbouring Ser83Tyr and Val112Ile substitutions in GyrA,
deficient in both OmpK35 and OmpK36, and expressing active efflux of fluoroquinolones
E. coli ATCC 25922 harbouring the Ser83Leu substitution in GyrA and the Ser80Arg substitution
in ParC (genetic replacement)
E. coli ATCC 25922 harbouring the Ser83Leu substitution in GyrA, the Ser80Arg substitution in
ParC (genetic replacement) and marR deficient (chromosomal inactivation)
E. coli ATCC 25922 harbouring Ser83Leu and Asp87Asn substitutions in GyrA and the Ser80Arg
substitution in ParC (genetic replacement)
E. coli ATCC 25922 harbouring Ser83Leu and Asp87Asn substitutions in GyrA, the Ser80Arg
substitution in ParC (genetic replacement) and marR deficient (chromosomal inactivation)
E. coli clinical isolate belonging to the ST131 clone harbouring the Ser83Leu substitution
in GyrA
have been described in Enterobacteriaceae.3,4 Several of these
mechanisms must be combined to achieve a clinical level of
resistance.
Enterobacteriaceae, particularly Escherichia coli, Klebsiella spp.
and Enterobacter spp., are among the most common organisms
causing community, nosocomial and opportunistic infections. The
emergence of a very close association between quinolone resistance and resistance to other antimicrobial agents, particularly
b-lactams and aminoglycosides, is a critical problem when managing such infections. One of the most striking features of these species is their remarkable capacity for developing antibiotic resistance.
The prevalence of infections caused by MDR Enterobacteriaceae
strains is increasing globally (with resistance rates at around 30%;
http://www.eurosurveillance.org/),5 often rendering antibiotic treatments useless and causing a significant rise in the morbidity and
mortality of affected patients.
The task of clinical microbiologists to properly perform antimicrobial susceptibility testing and interpretation and carry out
complementary phenotypic/genotypic tests proficient in detecting the resistance mechanisms involved is the necessary first
step for implementing appropriate treatments and antibiotic
use policies.6 In this regard, there are no reliable phenotypic
tests available for detecting PMQR mechanisms, either for strains
with borderline susceptibility to these determinants alone or for
resistant strains harbouring additional chromosomal mutations.4,7 One disturbing consequence of this situation is the underestimation of the prevalence of acquired and/or mutational
resistance mechanisms, which often leads to the application of
erroneous antibiotic treatments and subsequent clinical failure.
There is limited information about the proficiency of clinical microbiology laboratories in overcoming this challenge. Some previous
28
28
28
29
29
30,31
19
19
19
19
32
studies have shown the importance of correctly evaluating resistance to other antimicrobial agents, such as b-lactams, in wellcharacterized strains of Enterobacteriaceae or Pseudomonas
aeruginosa.8,9 The objective of this study was to assess the ability
of Spanish clinical microbiology laboratories to properly carry out
susceptibility testing and to infer underlying resistance mechanisms in a collection of well-characterized Enterobacteriaceae
strains, including those with complex combinations of mutational
and transferable quinolone resistance mechanisms causing lowlevel quinolone resistance (LLQR) phenotypes.
Methods
Bacterial strains, characterization of resistance
mechanisms and susceptibility testing
Thirteen Enterobacteriaceae strains, coded CC-00 to CC-12, were selected
for this study (Table 1). They included 12 clinical isolates or laboratory
strains (CC-01 to CC-12) with different chromosomal and/or horizontally
acquired resistance mechanisms and one reference strain: E. coli ATCC
25922 (CC-00). Identification of the strains, antimicrobial susceptibility testing and confirmation of the quinolone resistance mechanisms were verified
independently at the two reference laboratories used for this study: the
Department of Microbiology, School of Medicine, University of Seville,
Seville, Spain, and the Service of Microbiology, University Hospital Marqués
de Valdecilla, Santander, Spain. API 20-E (bioMérieux, Marcy-l’Étoile, France)
was used for species identification. Testing for susceptibility to ciprofloxacin,
levofloxacin, moxifloxacin, ofloxacin, norfloxacin and nalidixic acid was performed in duplicate at both the reference centres by disc diffusion, CLSI
broth microdilution and Etest (bioMérieux, Madrid, Spain). The 2013 CLSI
and EUCAST (http://www.eucast.org/clinical_breakpoints/) breakpoints
were used to interpret clinical categories.10,11
2039
Rodriguez-Martinez et al.
Study design
The work was designed as a nationwide proficiency study. In March 2013,
the 13 selected strains (Table 1) were sent to the 66 participating laboratories, together with detailed instructions (62 centres reported results).
Participating centres were requested to implement the methods they routinely used to test antimicrobial susceptibility and to consider the 13
strains as blood culture isolates. For each strain, participants were
requested to fill in an electronic form that included: (i) the results of the
antibiogram [quantitatively, in terms of inhibition zone diameters and
MIC values, and qualitatively, in terms of the derived ‘raw’ clinical categories (RCCs): susceptible (S), intermediate (I) and resistant (R)]; (ii) the breakpoints used (CLSI or EUCAST); (iii) the laboratory methods used (i.e. type of
automatic system or manual instrumentation); (iv) the interpreted clinical
categories (ICCs), when appropriate, according to internal laboratory criteria (also used in clinical practice); and (v) the inferred mechanism(s)
that might be responsible for the quinolone resistance phenotype.
Data analysis
Analysis of results focused on three aspects: (i) a descriptive analysis of
susceptibility testing methods, breakpoints applied, clinical categories
(CCs) assigned and discrepancies between centres deriving from these;
(ii) an analysis of errors detected in susceptibility testing results [minor
errors (mEs), major errors (MEs) and very major (VMEs), defined according
to standard criteria] compared with reference values;12 and (iii) an analysis
of the ability of participating laboratories to accurately infer the underlying
resistance mechanisms.
Results
Descriptive analysis of susceptibility testing methods,
breakpoints applied, CCs assigned and derived
discrepancies
Ninety percent of participating laboratories used automatic or
semi-automatic devices for routine susceptibility testing. These
included the MicroScan WalkAway (Dade MicroScan Inc., West
Sacramento, CA, USA; n ¼ 30, 48%), Vitek 2 (bioMérieux; n ¼ 19,
30%), the Wider (Francisco Soria Melguizo, Madrid, Spain; n ¼ 5,
8%) and Phoenix (BD Biosciences, Sparks, MD, USA; n ¼2, 3%) systems. The remaining 10% used manual methods. Nevertheless,
several automatic system users also performed complementary
manual tests (up to 15%) to check certain borderline susceptibility
values (mainly for CC-03, CC-07 and CC-09 strains). CLSI breakpoints were applied for 71% of determinations and EUCAST breakpoints for the remaining 29%.
The overall (all antibiotics/strain combinations) raw discrepancy
rate due exclusively to the differential use of breakpoints was
21.8% (Table 2). Ofloxacin and norfloxacin were often involved in
this kind of discrepancy, with particularly high percentages
(38.4% – 46.1%); CC-03 and CC-10 strains [qnrS2+aac(6 ′ )-Ib-cr
and GyrA Ser83Leu + GyrA Asp87Asn + ParC Ser80Arg +DmarR,
respectively] showed the highest discrepancy rates (50%). High
breakpoint-related discrepancies were also found for CC-07,
CC-08, CC-09 and CC-11 (33.3%), all of which had chromosomally
mediated mechanisms. On the other hand, CC-00, CC-04, CC-05
and CC-06 showed no breakpoint-related discrepancies (Table 2).
When the RCC was modified to ICC, the discrepancy rate due exclusively to the differential use of breakpoints decreased considerably
when some in the susceptible category were re-coded as I(r)
2040
(interpreted as intermediate susceptibility due to a less susceptible
phenotype, e.g. with MICs ≥0.125 mg/L for ciprofloxacin; Table 2).
ICCs provided by participating centres and the discrepancies
between them are shown in Figure 1. There was a high degree
of consensus among centres for WT strains (CC-00) and for
strains with high-level resistance patterns, such as CC-05, CC-06
and CC-11 [harbouring triple quinolone resistance-determining
region (QRDR) modification + aac(6 ′ )-Ib-cr+oqxAB, triple QRDR
modification + qepA1 and triple QRDR modification +DmarR,
respectively]. On the other hand, there was less consensus about
some antimicrobial/strain combinations, such as ciprofloxacin
and CC-03, CC-07 and CC-09 strains [harbouring qnrS2+aac(6 ′ )Ib-cr, GyrA Ser83Leu + OmpK352 + efflux and double QRDR
modification + DmarR, respectively, with MICs in the range of
0.5 – 2 mg/L], and levofloxacin and CC-01, CC-03, CC-07, CC-09
and CC-10 strains (MICs in the range of 0.5–2 mg/L).
Analysis of errors (mEs, MEs and VMEs) in susceptibility
testing results
Quinolone MICs, RCCs obtained after applying CLSI and EUCAST
criteria and consensus ICCs obtained at the two reference centres
are shown in Table 2. The distributions of discrepancies and
categorical error rates by antibiotic and strain are shown in
Tables 3 and 4, respectively. The percentage of VMEs (false susceptibility) for all quinolones was well below 2%, and only ofloxacin and moxifloxacin showed higher rates for raw VMEs (6.6%),
which reduced to 0% and 2.9%, respectively, in the interpreted
VMEs (Table 3) (these VMEs were associated with manual methods: disc diffusion for ofloxacin and agar dilution for moxifloxacin;
data not shown). These errors were particularly associated with
the CC-03 [qnrS2 +aac(6 ′ )-Ib-cr] strain and were almost exclusively observed among MicroScan WalkAway automated systems
users (50% of total raw VMEs with respect to this strain) (Table 4).
The percentages of MEs (false resistance) were always ,10%,
except for the raw MEs for ofloxacin and nalidixic acid (22.2%
and 19.1%, respectively) and the interpreted MEs for nalidixic
acid (27.1%). Eighty-five percent of raw MEs for ofloxacin occurred
in the CC-01 and CC-02 strains. The high percentage of mEs
observed was not surprising, given the MIC values exhibited by
several strains in this collection, which, in many cases, were on
the borderline between two CCs (Tables 2 and 3).
The analysis of categorical errors according to strain (Table 4)
revealed that those with the highest rates of raw VMEs were CC-03
[harbouring qnrS2 and aac(6 ′ )-Ib-cr, 5.8%] and CC-08 (single GyrA
and ParC mutation, 3.3%). These percentages dropped to 0%
when interpreted categories were applied. VMEs in the CC-03
strain were linked to ciprofloxacin and moxifloxacin and those in
CC-08 to norfloxacin. CC-03 was also a strain that gave a high rate
of raw MEs (13.8%) linked with nalidixic acid (73% of these).
Highest rates of raw MEs were observed for CC-02 strains
(qnrA1) (21.9%) in association with nalidixic acid; CC-07 (single
GyrA mutation with loss of porins and efflux) (22.6%) linked to
norfloxacin; and CC-10 (double GyrA and single ParC mutation)
(20.4%), linked to ciprofloxacin (71% of these) and levofloxacin
(18%). A significant increase in interpreted ME values was
observed for CC-02 and CC-03 strains (57.6% and 42.6%, respectively), due to the fact that some raw susceptible strains were
modified to the intermediate category using ICC criteria, which
obviously reduced the number of strains finally defined as
Ciprofloxacin
Strain
(characteristics)
CC-00 (WT)
CC-01 [GyrA 1,
aac(6 ′ )-Ib-cr]
CC-02 (qnrA1)
CC-03 [qnrS2,
aac(6 ′ )-Ib-cr]
CC-04 (qnrB48)
CC-05 [GyrA 2,
ParC 1,
aac(6 ′ )-Ib-cr,
oqxAB]
CC-06 (GyrA 2,
ParC 1, qepA1)
CC-07 (GyrA 1,
OmpK352,
efflux)
CC-08 (GyrA 1,
ParC 1)
CC-09 (GyrA 1,
ParC 1, MarR2)
CC-10 (GyrA 2,
ParC 1)
CC-11 (GyrA 2,
ParC 1, MarR2)
CC-12 (GyrA 1)
Levofloxacin
Moxifloxacin
RCCb
RCCb
CLSI/
CLSI/
MICa (mg/L) EUCAST ICCc MICa (mg/L) EUCAST ICCc
a
MIC
(mg/L)
RCCb
CLSI/
EUCAST
Ofloxacin
ICCc
Norfloxacin
RCCb
CLSI/
MICa (mg/L) EUCAST ICCc
MICa
(mg/
L)
Nalidixic acid
RCCb
RCCb
a
CLSI/
MIC
CLSI/
EUCAST ICCc (mg/L) EUCAST ICCc
0.004 –0.008
4
S
R
S
R
0.005 – 0.015
1-2
S
S
S
0.008
I(r) 1
NS/S
NS/I
NS/S
NS/I
0.016 –0.032
2–4
S
S/R
S
0.032
I(r)/R 128
S
R
S
R
0.125 –0.25
2
S
I/R
I(r)
I/R
0.5
0.5
S
S
I(r) 1
I(r) 2
NS/I
NS/R
NS/I
NS/R
1
1–2
S/I
S/I
I(r)/I 0.5– 1
I(r)/I 4
S
S/R
0.125
64
S
R
I(r)
R
0.25
8 –16
S
R
I(r) 1
R
16– 32
NS/I
NS/R
NS/I
NS/R
0.5– 1
32
S
R
I(r)
R
0.25
256
.256
R
R
32
R
R
64
NS/R
NS/R
64
R
R
1
S/I
I(r)/I 0.5
S
I(r) 0.5
NS/S
NS/I(r) 4 – 8
R
R
0.25
S
I(r)
0.25
S
I(r) 0.25 –0.5
NS/S
NS/I(r) 1
S/I
0.5
S
I(r)
0.5
S
I(r) 0.5
NS/S
NS/I(r) 1 – 2
1
S/I
I(r)/I 1 –2
S
I(r) 2
NS/R
NS/R
4
R
R
4
I/R
I/R 8
NS/R
NS/R
0.25
S
I(r)
0.25
S
I(r) 0.25
NS/S
NS/I(r) 1
1– 2
128
S/NS
R/NS
S/NS
R/NS
I(r)
16–32
I(r)/R 16
S/NS
S/NS
S/NS
S/NS
S
R
I(r)
R
4– 8
.256
S/NS
R/NS
S/NS
R/NS
.256
R
R
.256
R/NS
R/NS
4 –8
S/R
I(r)/R .256
R/NS
R/NS
I(r)/I 2
S/R
I(r)/R .256
R/NS
R/NS
S/I
I(r)/I 2 –4
S/R
I(r)/R .256
R/NS
R/NS
2–4
I/R
I/R
4
S/R
I(r)/R .256
R/NS
R/NS
8
R
R
8 –16
I/R
I/R
.256
R/NS
R/NS
S/I
I(r)/I 0.5
S
I(r)
.256
R/NS
R/NS
2041
JAC
NS, not specified by CLSI or EUCAST committees; I(r), interpreted as intermediate susceptibility due to a less susceptible phenotype (MIC ≥0.125 mg/L for ciprofloxacin, 0.25 mg/L for
levofloxacin, moxifloxacin and norfloxacin, and 0.5 mg/L for ofloxacin).
GyrA 1 and ParC 1 mean one substitution in the QRDR and GyrA 2 means two substitutions in the QRDR (see Table 1).
a
Determined in duplicate by the two reference centres using the CLSI broth microdilution method.
b
According to CLSI/EUCAST criteria.
c
Interpreted by the reference centres according to pre-characterized resistance mechanisms and interpretive reading of the antibiogram.
Challenges to accurate susceptibility testing and interpretation of quinolone resistance
Table 2. Reference centres’ susceptibility testing results, derived RCCs and ICCs based on resistance mechanisms
%I
Rodriguez-Martinez et al.
2042
%S
%R
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
NAL
CIP
LVX
MXF
NOR
OFX
0%
CC-00
CC-01
CC-02
CC-03
CC-04
CC-05
CC-06
CC-07
CC-08
CC-09
CC-10
CC-11
CC-12
Figure 1. Overall ICC percentages provided by participating centres regarding quinolone antibiotics; ‘r’ values were re-coded as ‘I’. CIP, ciprofloxacin; LVX, levofloxacin; MXF, moxifloxacin;
NAL, nalidixic acid; NOR, norfloxacin; OFX, ofloxacin.
CIP, ciprofloxacin; LVX, levofloxacin; MXF, moxifloxacin; OFX, ofloxacin; NOR, norfloxacin; FQs, fluoroquinolones; NAL, nalidixic acid.
RCC and ICC results from participating laboratories were compared with reference centre values and discrepancies were classified as mE, ME and VME, following standard criteria.33
b
The most favourable CC for the evaluated centre was chosen when the MIC value (obtained by the reference centres) gave rise to two different CCs according to EUCAST/CLSI breakpoints.
In this case, the breakpoints chosen were those most in concordance with the CC issued by the evaluated centre.
c
The denominator is the number of susceptibility-testing determinations per antibiotic.
d
The denominator is the number of susceptible strains per antibiotic.
e
The denominator is the number of resistant strains per antibiotic.
f
The denominator is the total number of susceptibility determinations.
a
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
1.9/0
18.1/0
0/0
0/0
0/0
3.0/0
34.2/40.7
12.8/26.7
7.5/38.5
0/21.4
0/0
0/0
0/0
7.1/37.6
0/1.9
3.6/19.8
4.3/0
0/0
0/0
0/0
18.2/0
7.4/0
4.3/4.3
6.5/2.8
0.3/0.6
0/0
6.6/2.9
6.6/0
1.8/0
1.4/0.1
0/0.4
0.7/0.5
CIP (805)
LVX (243)
MXF (77)
OFX (354)
NOR (42)
All FQs (1521)
NAL (784)
All agents (2305)
19.1/42.0
13.5/53.3
25.9/36.7
28.5/37.8
27.4/43.1
20.7/43.2
7/10.8
16.1/31.5f
14.4/41.7
9.8/53.3
22.1/35.7
21.4/37.8
20/43.1
15.6/43.0
1.1/2.2
10.6/28.3f
7.4/0
4.7/0
3.4/0
22.2/0
8.8/0
7.3/0
19.1/27.1
9.6/18.1
17.9/43.1
3.7/68.5
15.4/28
0/0
23/46.3
17.7/45.6
0/0.6
11.3/29
12.4/0
0/0
0/0
0/0
9.2/0
8.3/0
9.2/20.4
9.7/13.8
0.7/1.4
0/0
0/0
0/0
1.2/0
0.6/0.4
0/0.8
0.4/1
8.3/48.5
0/0
0/35.3
0/0
31.5/57.8
13.2/48.1
0/1.4
7.4/29.3
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
MEd
MEd
mEc
VMEc
MEd
mEc
VMEe
MEd
Wider (162)
MicroScan WalkAway (919)
all methods and devices (2305)
mEc
Quinolone (n)
Percentage of
discrepancies in RCC/ICCb,c
Table 3. Distribution of discrepancies and categorical error rates by antimicrobial agent tested
Percentage of errors in RCC/ICC (n)a
VMEe
mEc
Vitek 2 (520)
VMEe
Challenges to accurate susceptibility testing and interpretation of quinolone resistance
JAC
susceptible. Additionally, interpretation of these results should
take into account the fact that some of the quinolones (e.g. ciprofloxacin and nalidixic acid) were tested by most of the centres
while others (e.g. moxifloxacin and norfloxacin) were tested
only by a lower number of centres.
VMEs were not found for the Vitek 2 and Wider systems compared with the MicroScan WalkAway (0.4% and 1% for raw and
interpreted data, respectively). Among all determinations performed using the MicroScan WalkAway, two raw VMEs and four
interpreted VMEs were found (affecting ciprofloxacin and norfloxacin in the CC-03, CC-07 and CC-08 strains). The Wider system
accounted for the lowest percentage of raw MEs overall, followed
by the MicroScan WalkAway and Vitek 2 systems; after interpretation, the percentage of MEs was still in the same order. The
fluoroquinolone giving the highest percentage of raw MEs was
norfloxacin among Wider users, levofloxacin among Vitek 2
users and ciprofloxacin among MicroScan WalkAway users. The
Wider system showed the lowest rate of raw mEs, followed by
the MicroScan WalkAway and the Vitek 2.
Ability of centres to detect acquired PMQRs and to infer
resistance mechanisms
An analysis of errors of inference is shown in Table 5, in which
strains CC-01, CC-02, CC-03 and CC-04 (all with LLQR and harbouring PMQR mechanisms) are highlighted as those giving the highest
percentage of errors, documented in 11.3% – 54.8% of centres.
One of the most important aspects was that acquired PMQR
genes were not detected in certain strains. In such cases, laboratories attributed resistance exclusively to mutational mechanisms, with percentages of .40% in the strains specified. By
contrast, PMQR genes were inferred in strains harbouring only
chromosomal mechanisms (CC-07, CC-08, CC-09, CC-10, CC-11
and CC-12).
Discussion
The growing prevalence of quinolone resistance in
Enterobacteriaceae severely compromises the therapeutic
arsenal available.1,13,14 The complexity of the mechanisms
involved also affects the ability of the clinical microbiology laboratory to perform proper standardized antimicrobial susceptibility
tests. This situation may cause greater concern with borderline
resistance phenotypes. The detection of specific resistance
mechanisms is particularly important for guiding therapy and
infection control strategies.8,9 It was this clinical challenge that
motivated this large multicentre study, in which we evaluated
the proficiency of 62 Spanish laboratories with respect to accurate
susceptibility testing and detection of quinolone resistance phenotypes in a collection of well-characterized Enterobacteriaceae
strains showing different combinations of mutational and transferable quinolone resistance mechanisms.
Of the various aspects that had a significant impact on discrepancies in the susceptibility testing results provided by participating
centres, the differential application of the currently nonharmonized CLSI or EUCAST breakpoints should be mentioned.
EUCAST breakpoints were used in only 29% of the participating
centres, although this represents a significant increase when
compared with similar Spanish studies performed in 2001 and
2043
Rodriguez-Martinez et al.
2044
Table 4. Distribution of categorical error rates by tested strain
Percentage of errors in RCC/ICC (n)a,b
all methods and devices (2305)
Strain (characteristics) (n)
CC-00 (WT) (175)
CC-01 [GyrA 1, aac(6 ′ )-Ib-cr] (178)
CC-02 (qnrA1) (183)
CC-03 [qnrS2, aac(6 ′ )-Ib-cr] (181)
CC-04 (qnrB48) (174)
CC-05 [GyrA 2, ParC 1, aac(6 ′ )-Ib-cr, oqxAB] (176)
CC-06 (GyrA 2, ParC 1, qepA1) (177)
CC-07 (GyrA 1, OmpK352, efflux) (175)
CC-08 (GyrA 1, ParC 1) (181)
CC-09 (GyrA 1, ParC 1, MarR2) (178)
CC-10 (GyrA 2, ParC 1) (176)
CC-11 (GyrA 2, ParC 1, MarR2) (175)
CC-12 (GyrA 1) (176)
MicroScan WalkAway (919)
Wider (162)
Vitek 2 (520)
mEc
MEd
VMEe
mEc
MEd
VMEe
mEc
MEd
VMEe
mEc
MEd
VMEe
0/0
3.9/10.5
11.4/47.9
24.3/36.4
5.1/60.9
0.5/0.6
0/0
20/35
8.8/40.2
23/35.1
23.8/41.2
14.8/13.4
2.2/45.7
0/0
4.3/0
21.9/57.6
13.8/42.6
1.2/8.2
0/0
0/0
22.6/0
0/0
14/0
20.4/0
0/0
0.9/0
0/0
0/0
0/0
5.8/0
0/0
0/0
0/0
1.1/0
3.3/0
1.1/0
1/0
0/0.7
0/1.6
0/0
0/8.4
11.6/52.2
36.7/44.1
0/60.8
0/0
0/0
29.2/35.6
10.9/37.1
25/30.8
19.1/50.7
13.8/13.8
2.7/47.2
0/0
0/0
16.4/43.4
0/41.6
0/0
0/0
0/0
35.3/0
0/0
17.7/0
12/0
0/0
2.1/0
0/0
0/0
0/0
8.3/0
0/0
0/0
0/0
0/0
2.6/0
0/0
0/0
0/1.7
0/3.8
0/0
0/20
8.3/53.3
14.2/18.7
0/50
0/0
0/0
7.6/50
0/58.3
25/33.3
25/41.6
16.6/16.6
0/54.5
0/0
0/0
0/0
12.5/0
7.7/16.6
0/0
0/0
50/0
0/0
28/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
2.5/2.5
0/31.7
5.1/26.3
0/52.6
0/0
0/0
5/23.7
2.5/31.5
7.7/24.3
16.6/30.7
7.3/4.8
0/32.5
0/0
0/0
48.7/48.7
30/47.3
0/11.1
0/0
0/0
0/0
0/0
0/0
10.1/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
0/0
GyrA 1 and ParC 1 mean one substitution in the QRDR and GyrA 2 means two substitutions in the QRDR (see Table 1).
RCC and ICC results from participating laboratories were compared with reference centre values and discrepancies were classified as mE, ME and VME, following standard criteria.33
b
The most favourable CC of the evaluating centre was chosen when an MIC value (obtained by reference centres) gave rise to two different CCs according to EUCAST/CLSI breakpoints. In
this case, the breakpoints chosen were those showing most concordance with the CC issued by the evaluating centre.
c
The denominator is the number of susceptibility-testing determinations for each strain.
d
The denominator is the number of susceptibility-testing determinations for susceptible strains.
e
The denominator is the number of susceptibility-testing determinations for resistant strains.
a
Challenges to accurate susceptibility testing and interpretation of quinolone resistance
JAC
Table 5. Description of errors in the inferred resistance mechanisms
Wrong inferred mechanisms
Strain (characteristics)
no. (%)
description of mistakes
CC-00 (WT)
CC-01 [GyrA 1, aac(6 ′ )-Ib-cr]
CC-02 (qnrA1)
CC-03 [qnrS2, aac(6 ′ )-Ib-cr]
0 (0)
34 (54.8)
15 (24.2)
7 (11.3)
CC-04 (qnrB48)
10 (16.1)
no mistakes
multiple QRDR modifications (12); PMQR not detected (22)
single QRDR modification at GyrA and PMQR not detected (15)
single QRDR modification at GyrA and PMQR not detected (4); chromosomal
efflux pump phenotype and PMQR not detected (3)
single QRDR modification at GyrA and PMQR not detected (6); chromosomal
efflux pump phenotype and PMQR not detected (4)
no mistakes
no mistakes
PMQR (6)
PMQR (4)
PMQR (5)
PMQR (3)
PMQR (2)
PMQR (1)
CC-05 [GyrA 2, ParC 1, aac(6 ′ )-Ib-cr, oqxAB]
CC-06 (GyrA 2, ParC 1, qepA1)
CC-07 (GyrA 1, OmpK352, efflux)
CC-08 (GyrA 1, ParC 1)
CC-09 (GyrA 1, ParC 1, MarR2)
CC-10 (GyrA 2, ParC 1)
CC-11 (GyrA 2, ParC 1, MarR2)
CC-12 (GyrA 1)
0 (0)
0 (0)
6 (9.6)
4 (6.5)
5 (8)
3 (4.8)
2 (3.2)
1 (1.6)
2011, in which 100% and 86% of the centres, respectively, applied
CLSI breakpoints.8,9 Furthermore, this value is expected to increase
significantly in the near future once the current institutional recommendations in Europe are followed. The application of European
breakpoints has increased significantly in Spain, due particularly
to the 2012 constitution of the Spanish Antibiogram Committee
(Coesant), whose main commitment is to endorse EUCAST
breakpoints.15
Regarding the discrepancies that derived from breakpoints,
ofloxacin and norfloxacin were the antibiotics most affected.
These discrepancies may be particularly important in PMQRproducing strains (such as CC-03) or those with combinations of
two or three mutational modifications (such as CC-07, CC-08,
CC-09 and CC-10), since several strains that showed mutational
resistance mechanisms or relevant acquired PMQR genes, such
as qnrS2 and aac(6 ′ )-Ib-cr, were reported as susceptible according
to CLSI breakpoints but resistant according to EUCAST. Less relevant are discrepancies deriving from different ciprofloxacin
or levofloxacin breakpoints, where discrepancies were mostly
reported as intermediate according to CLSI breakpoints and
resistant according to EUCAST breakpoints.10,11
The method used to carry out susceptibility testing was
another source of discrepancy in the results between centres.
Most centres (90%) used semi-automatic or automatic devices,
mainly the MicroScan WalkAway, Vitek 2 and Wider systems.
Vitek 2 and Wider devices did not show VMEs, although they
were used considerably less than the MicroScan WalkAway.
Regarding MEs, both raw and after interpretation, the lowest percentage overall corresponded to the Wider, followed by the
MicroScan WalkAway and then the Vitek 2 systems. Among the
antibiotics tested, moxifloxacin and ofloxacin had the highest
recorded VME rates. Among the studied strains, CC-03 [a combination of qnrS2 plus aac(6 ′ )-Ib-cr] and CC-08 (single mutation in
GyrA and ParC) were those associated with a higher rate of
VMEs overall. These data indicate that borderline phenotypes
may imply a source of discrepancy for susceptibility testing in
clinical microbiology and to have consequences for treatment
outcome.16,17
A further source of divergent results concerns the interpretive
reading of susceptibility testing data, and hence the reporting of
ICCs. This ICC parameter is important because the in vivo activity
of quinolones in the clinical setting is related to MIC values and
AUC/MIC or Cmax/MIC, which are good indicators of a favourable
outcome for quinolone treatment.18 Consequently, it may be
assumed from these indicators that an increase in the MIC of a
quinolone (of even just a few dilution steps) will have a negative
impact on its therapeutic efficacy. In our study, 15% of all susceptibility determinations involved a change in the final CC, with percentages of .40% for some antibiotic and strain combinations. A
high percentage of discrepancies in ICCs involved a change to
intermediate susceptibility [I(r)] when interpreted at the reference
centres; e.g. an MIC of ≥0.125 mg/L for ciprofloxacin was modified
to intermediate susceptibility, depending on the associated resistant molecular mechanism (although it should be noted that
PMQR mechanisms could be present in isolates with lower MICs
values, as reported).19 We suggest that the current breakpoints
defined by EUCAST/CLSI be re-evaluated, and that the already
agreed general principles for the clinical categorization of antibiogram results for Salmonella could also be applied to other enterobacteria when considering severe infections.10,11
Various well-known factors that limit the efficiency of interpretive
readings include a lack of knowledge about the basis of resistance in
a significant proportion of medical microbiologists; unknown factors
affecting resistance in clinical strains; low-level resistance mechanisms; and certain mechanisms that may be camouflaged as a result
of the complex interplay of intrinsic and acquired resistance
mechanisms. For example, it is not possible to infer PMQR mechanisms directly when they are associated with QRDR modifications
that confer a high level of nalidixic acid resistance; meanwhile the
PMQR phenotype is generally characterized by reduced susceptibility
to fluoroquinolones, with susceptibility to nalidixic acid in the
absence of QRDR modifications.20 – 22 The implementation of expert
2045
Rodriguez-Martinez et al.
systems in most automatic/semi-automatic systems may help in
this task.23,24 Recommendations for interpretive reading, inferring
and detecting resistance mechanisms have never been issued by
CLSI or EUCAST, despite the fact that the breakpoints are certainly
not adapted to overcoming PMQR mechanisms. As observed in our
study, a combination of one to three relevant mechanisms of mutational resistance or PMQR determinants frequently yields a borderline MIC. This, therefore, is a major source of conflicting results in
the absence of an interpretive reading of the antibiogram. The pharmacokinetic parameter for the clinical efficacy of quinolones is the
AUC/MIC. Small variations in MIC values may decrease this parameter and affect the treatment outcome. In this way, these low-level
resistant phenotypes (breakpoint borderlines) could have clinical
relevance leading to in vivo resistance. In particular, it has been
reported that PMQR mechanisms can reduce fluoroquinolone activity in vivo.25,26 In this sense, the epidemiological cut-off for ciprofloxacin in E. coli, established as 0.032 mg/L according to EUCAST
(www.eucast.org), seems to be useful for LLQR detection.
Taken together, the data obtained in this work clearly suggest
that the use of different breakpoints and systems, as well as the
complexity of the mechanisms that lead to reduced susceptibility
and produce borderline phenotypes, may have important consequences for the treatment and control of infections caused by
these microorganisms.
Acknowledgements
This study was performed under the auspices of the Study Group of
Antimicrobial Mechanisms of Action and Resistance (GEMARA) from the
Spanish Society of Clinical Microbiology and Infectious Diseases (SEIMC)
and the Spanish Network for Research in Infectious Diseases (REIPI). We
are grateful to the 62 participating centres and the SEIMC Quality
Control Program for their essential contribution to making this study
possible. The complete list of the 62 participating hospitals is as
follows: Hospital Universitario Virgen de la Macarena (Sevilla), Hospital
Universitario de Valme (Sevilla), Hospital Universitario de Tarragona Joan
XXIII (Tarragona), Hospital General de Gran Canaria Dr Negrı́n (Las
Palmas Gran Canaria, Las Palmas), Hospital Costa del Sol (Marbella,
Málaga), Hospital Universitario Miguel Servet (Zaragoza), Hospital
Sierrallana (Torrelavega, Cantabria), Hospital Provincial de Castellón
(Castellón), Hospital Santa Barbara (Complejo Hosp. de Soria, Soria),
Hospital Clı́nico Universitario de Salamanca (Salamanca), Hospital Clı́nico
Universitario de Valladolid (Valladolid), Hospital Universitario Rio Hortega
(Valladolid), Complejo Asistencial de Avila (Ávila), Hospital General
Universitario de Albacete (Albacete), Hospital General de Ciudad
Real (Ciudad Real), Hospital Virgen de la Salud (Toledo), Hospital
Universitario de Girona Dr Josep Trueta (Girona), Hospital Clı́nic
(Barcelona), Hospital Virgen de Puerto (Plasencia, Cáceres), Complejo
Hospitalario Universitario de Vigo (Vigo, Pontevedra), Complejo
Hospitalario Universitario A Coruña (A Coruña), Hospital Infantil Niño
Jesús (Madrid), Hospital General U. Gregorio Marañón (Madrid), Hospital
Universitario Puerta de Hierro Majadahonda (Majadahonda, Madrid),
Hospital Carlos III (Madrid), Complejo Hospitalario de Navarra
(Pamplona, Navarra), Hospital de Txagorritxu (Vitoria, Álava), Complejo
Hospitalario Donostia (Donosti-San Sebastián, Gipuzkoa), Hospital de
Cruces (Barakaldo, Vizcaya), Hospital de Galdakao (Galdakao Vizcaya),
Consorcio Hospital General Universitario de Valencia (Valencia), Hospital
Universitario La Fe (Valencia), Hospital General Universitario de Elche
(Elche, Alicante), Complejo Asistencial Universitario de Burgos (Burgos),
Hospital Universitario Puerta del Mar (Cádiz), Hospital Nª Sra de la
Candelaria (Santa Cruz de Tenerife, Tenerife), Hospital Severo Ochoa
(Leganés, Madrid), Complejo Hospitalario de Pontevedra (Pontevedra),
2046
Hospital Universitario Reina Sofı́a (Córdoba), Hospital Universitario La Paz
(Madrid), Hospital Universitario Son Espases (Palma de Mallorca,
Baleares), Hospital Ramón y Cajal (Madrid), Hospital Universitario Insular
de Gran Canaria (Las Palmas de Gran Canaria, Las Palmas), Hospital
Virgen de Altagracia (Manzanares, Ciudad Real), Hospital Son Llatzer
(Palma de Mallorca, Baleares), Hospital de Jerez (Jerez de la Frontera,
Cádiz), Hospital de la Ribera (Alzira, Valencia), Hospital Universitari
Germans Trias i Pujol (Badalona, Barcelona), Hospital San Pedro de
Alcantara (Cáceres), Hospital Universitario Vall d′ Hebron (Barcelona),
Hospital General Reina Sofia (Murcia), Hospital Universitario Virgen del
Rocı́o (Sevilla), Hospital Universitario San Cecilio (Granada), Hospital
Universitario Santa Cristina (Madrid), Hospital de Cabueñes (Gijón,
Asturias), Hospital Universitario Marqués de Valdecilla (Santander,
Cantabria), Hospital Universitario de Bellvitge (L’Hospitalet de Llobregat,
Barcelona), Hospital de la Princesa Madrid (Madrid), Hospital Clı́nico
Universitario San Carlos (Madrid), Hospital Clı́nico Universitario Lozano
Blesa (Zaragoza), Hospital Universitario Fundación Alcorcón (Alcorcón,
Madrid), Hospital Universitario Virgen de la Arrixaca (El Palmar, Murcia).
Funding
This work was supported by the Ministerio de Sanidad y Consumo, Instituto
de Salud Carlos III (projects PI11-00934 and PI11/01117) and the
Consejerı́a de Innovación Ciencia y Empresa, Junta de Andalucı́a
(P11-CTS-7730), Spain, by the Plan Nacional de I+D+i 2008-2011 and
the Instituto de Salud Carlos III, Subdirección General de Redes y
Centros de Investigación Cooperativa, Ministerio de Economı́a y
Competitividad, the Spanish Network for Research in Infectious Diseases
(REIPI RD12/0015)—co-financed by European Development Regional
Fund ‘A way to achieve Europe’ ERDF.
Transparency declarations
None to declare.
References
1 Drlica K, Hiasa H, Kerns R et al. Quinolones: action and resistance
updated. Curr Top Med Chem 2009; 9: 981–98.
2 Singh R, Swick MC, Ledesma KR et al. Temporal interplay between efflux
pumps and target mutations in development of antibiotic resistance in
Escherichia coli. Antimicrob Agents Chemother 2012; 56: 1680–5.
3 Strahilevitz J, Jacoby GA, Hooper DC et al. Plasmid-mediated quinolone resistance: a multifaceted threat. Clin Microbiol Rev 2009; 22:
664 – 89.
4 Rodriguez-Martinez JM, Cano ME, Velasco C et al. Plasmid-mediated
quinolone resistance: an update. J Infect Chemother 2011; 17: 149–82.
5 Fraimow H, Nahra R. Resistant Gram-negative infections. Crit Care Clin
2013; 29: 895–921.
6 Paterson DL. Impact of antibiotic resistance in Gram-negative bacilli on
empirical and definitive antibiotic therapy. Clin Infect Dis 2008; 47 Suppl 1:
S14– 20.
7 Woodford N, Eastaway AT, Ford M et al. Comparison of BD Phoenix, Vitek 2,
and MicroScan automated systems for detection and inference of mechanisms responsible for carbapenem resistance in Enterobacteriaceae. J Clin
Microbiol 2010; 48: 2999–3002.
8 Canton R, Loza E, Del Carmen Conejo M et al. Quality control for b-lactam
susceptibility testing with a well-defined collection of Enterobacteriaceae
and Pseudomonas aeruginosa strains in Spain. J Clin Microbiol 2003; 41:
1912– 8.
Challenges to accurate susceptibility testing and interpretation of quinolone resistance
9 Juan C, Conejo MC, Tormo N et al. Challenges for accurate susceptibility
testing, detection and interpretation of b-lactam resistance phenotypes in
Pseudomonas aeruginosa: results from a Spanish multicentre study.
J Antimicrob Chemother 2013; 68: 619–30.
10 Clinical and Laboratory Standards Institute. Performance Standards
for Antimicrobial Susceptibility Testing: Twenty-fourth Informational
Supplement M100-S24. CLSI, Wayne, PA, USA, 2014.
11 EUCAST. Clinical Breakpoints and Epidemiological Cut-off Values. 2014.
http://www.eucast.org/clinical_breakpoints/.
12 Camaro-Sala ML, Martinez-Garcia R, Olmos-Martinez P et al. Validation
and verification of microbiology methods. Enferm Infecc Microbiol Clin
2014; doi:10.1016/j.eimc.2013.11.010.
13 Hooper DC. Emerging mechanisms of fluoroquinolone resistance.
Emerg Infect Dis 2001; 7: 337– 41.
14 Jacoby GA. Mechanisms of resistance to quinolones. Clin Infect Dis
2005; 41 Suppl 2: S120–6.
15 Martinez-Martinez L, Pascual A, Canton R. The Spanish antibiogram
committee (COESANT), in step with EUCAST. Enferm Infecc Microbiol Clin
2013; 31: 639–40.
16 Rodriguez-Martinez JM, Briales A, Velasco C et al. Discrepancies in
fluoroquinolone clinical categories between the European Committee on
Antimicrobial Susceptibility Testing (EUCAST) and CLSI for Escherichia coli
harbouring qnr genes and mutations in gyrA and parC. J Antimicrob
Chemother 2011; 66: 1405– 7.
17 Poirel L, Pitout JD, Calvo L et al. In vivo selection of fluoroquinoloneresistant Escherichia coli isolates expressing plasmid-mediated quinolone
resistance and expanded-spectrum b-lactamase. Antimicrob Agents
Chemother 2006; 50: 1525– 7.
18 Schentag JJ, Meagher AK, Forrest A. Fluoroquinolone AUIC break
points and the link to bacterial killing rates. Part 2: human trials. Ann
Pharmacother 2003; 37: 1478– 88.
19 Machuca J, Briales A, Labrador G et al. Interplay between plasmidmediated and chromosomal-mediated fluoroquinolone resistance and
bacterial fitness in Escherichia coli. J Antimicrob Chemother 2014; 69:
3203– 15.
20 Cavaco LM, Aarestrup FM. Evaluation of quinolones for use in detection
of determinants of acquired quinolone resistance, including the new transmissible resistance mechanisms qnrA, qnrB, qnrS, and aac(6 ′ )Ib-cr, in
Escherichia coli and Salmonella enterica and determinations of wild-type
distributions. J Clin Microbiol 2009; 47: 2751 –8.
21 Corvec S, Cremet L, Reynaud A et al. Vitek2 system: a reliable tool to
detect qnr determinants in Enterobacteriaceae without quinolone
resistance-determining region modifications. Diagn Microbiol Infect Dis
2009; 64: 455–7.
JAC
22 Lupo A, Papp-Wallace KM, Sendi P et al. Non-phenotypic tests to detect
and characterize antibiotic resistance mechanisms in Enterobacteriaceae.
Diagn Microbiol Infect Dis 2013; 77: 179–94.
23 Calvo J, Cano ME, Pitart C et al. Evaluation of three automated systems
for susceptibility testing of Enterobacteria containing qnrB, qnrS, and/or
aac(6 ′ )-Ib-cr. J Clin Microbiol 2011; 49: 3343 –5.
24 Rodriguez-Martinez JM, Conejo MC, Martinez-Martinez L et al.
Evaluation of antimicrobial susceptibility of bacteria containing the qnr
gene and FOX-5 b-lactamase by four automated systems. Clin Microbiol
Infect 2005; 11: 402– 4.
25 Rodriguez-Martinez JM, Pichardo C, Garcia I et al. Activity of ciprofloxacin and levofloxacin in experimental pneumonia caused by Klebsiella
pneumoniae deficient in porins, expressing active efflux and producing
QnrA1. Clin Microbiol Infect 2008; 14: 691–7.
26 Guillard T, Cambau E, Chau F et al. Ciprofloxacin treatment failure in a
murine model of pyelonephritis due to an AAC(6 ′ )-Ib-cr-producing
Escherichia coli strain susceptible to ciprofloxacin in vitro. Antimicrob
Agents Chemother 2013; 57: 5830 –5.
27 Ruiz E, Ocampo-Sosa AA, Alcoba-Florez J et al. Changes in ciprofloxacin
resistance levels in Enterobacter aerogenes isolates associated with variable expression of the aac(6 ′ )-Ib-cr gene. Antimicrob Agents Chemother
2012; 56: 1097– 100.
28 Cano ME, Calvo J, Agüero J et al. Plasmid-mediated quinolone resistance among enterobacteria with reduced susceptibility or resistance
to ciprofloxacin but susceptible to nalidixic acid. In: Abstracts of the
Forty-seventh Interscience Conference on Antimicrobial Agents and
Chemotherapy, Chicago, IL, 2007. Abstract C2 – 152, p. 103. American
Society for Microbiology, Washington, DC, USA.
29 Ruiz E, Saenz Y, Zarazaga M et al. qnr, aac(6 ′ )-Ib-cr and qepA genes in
Escherichia coli and Klebsiella spp.: Genetic environments and plasmid and
chromosomal location. J Antimicrob Chemother 2012; 67: 886–97.
30 Martinez-Martinez L, Hernandez-Alles S, Alberti S et al. In vivo selection
of porin-deficient mutants of Klebsiella pneumoniae with increased resistance to cefoxitin and expanded-spectrum-cephalosporins. Antimicrob
Agents Chemother 1996; 40: 342–8.
31 Martinez-Martinez L, Garcia I, Ballesta S et al. Energy-dependent accumulation of fluoroquinolones in quinolone-resistant Klebsiella pneumoniae
strains. Antimicrob Agents Chemother 1998; 42: 1850– 2.
32 Lopez-Cerero L, Bellido MM, Serrano L et al. Escherichia coli O25b:H4/
ST131 are prevalent in Spain and are often not associated with ESBL or
quinolone resistance. Enferm Infecc Microbiol Clin 2013; 31: 385–8.
33 Elder BL, Hansen SA, Kellogg JA et al. Verification and Validation of
Procedures in the Clinical Microbiology Laboratory. In: McCurdy BW, ed.
CUMITECH: Cumulative Techniques and Procedures in Clinical Microbiology.
Washington, DC: American Society for Microbiology, 1997.
2047