143 Cephalosporin and Carbapenem Inhibition Zone Breakpoints for Distinguishing between Extended- spectrum Beta-lactamase (ESBL)- and Non-ESBL- producing Enterobacteriaceae Keywords: Inhibition zone, breakpoints, extended-spectrum beta-lactamase, Enterobacteriaceae, CLSI, EUCAST Corresponding author: Chusana Suankratay, M.D., Ph.D., Division of Infectious Diseases, Department of Internal Medicine, King Chulalongkorn Memorial Hospital, Bangkok 10330, Thailand. ORIGINAL ARTICLE Palakorn Panarat, M.D. 1 , Malee Techapornroong, M.D. 2 , Chusana Suankratay, M.D., Ph.D. 3 1 Department of Internal Medicine, King Chulalongkorn Memorial Hospital, Bangkok, Thailand. 2 Division of Infectious Diseases, Department of Internal Medicine, Prapokklao Hospital, Chantaburi, Thailand. 3 Division of Infectious Diseases, Department of Internal Medicine, King Chulalongkorn Memorial Hospital, Bangkok, Thailand. ABSTRACT Background: In 2010, the United States Clinical Laboratory Standards Institute (CLSI) released the new inhibition zone and minimal inhibitory concentration (MIC) breakpoints of beta-lactam antibiotics for Enterobacteriaceae and recommended that there is no need to perform ESBL production tests. The reasons for lowering breakpoints are based on pharmacokinetic and pharmacodynamic data as well as the MIC distribution of clinical isolates of Enterobacteriaceae in several studies carried out in the United States. To our knowledge, there have been no studies to determine inhibition zone distribution in comparison with ESBL production among Enterobacteriaceae in Thailand. Patients and methods: A descriptive study was carried out to determine the inhibition zone distribution and breakpoints of beta-lactam antibiotics to distinguish between ESBL- and non-ESBL-producing clinical isolates of Enterobacteriaceae from sterile specimens collected from adult patients who were hospitalized at Prapokklao Hospital, Chantaburi, Thailand, from October 2011 to September 2013. We tested ESBL production by using the double disc and disc combination methods for cefotaxime. Moreover, we performed Receiver Operating Characteristic (ROC) curve to determine the most appropriate sensitivity and specificity of inhibition zone breakpoints of all antibiotics tested to distinguish between ESBL- and non-ESBL-producing Enterobacteriaceae. Results: Among all 930 specimens (879 (94.5%) blood and 51 (5.5%) other sterile specimens), there were 312 ESBL-producing isolates of Enterobacteriaceae including 239 (37.2%) E. coli, 65 (27.5%) K. pneumoniae, and 8 (16%) P. mirabilis. Five hundred and thirty-four (57.4%) specimens were from patients hospitalized at Internal Medicine ward. Based on ROC curve, the inhibition zone breakpoint of cefotaxime to distinguish between ESBL- and non-ESBL-producing Enterobacteriaceae was 26 mm, consistent with that recommended by CLSI 2010. However, the inhibition zone breakpoint of cefazolin to distinguish between ESBL- and non- ESBL-producing Enterobacteriaceae was 12 mm which was much different from that (23 mm) recommended
24
Embed
Cephalosporin and Carbapenem Inhibition Zone Breakpoints ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Vol. 31 No. 3 143Beta-lactam inhibition zone breakpoints for ESBL and non-ESBL Enterobacteriaceae:- Panarat P, et al.
143
Cephalosporin and Carbapenem Inhibition ZoneBreakpoints for Distinguishing between Extended-spectrum Beta-lactamase (ESBL)- and Non-ESBL-producing Enterobacteriaceae
Keywords: Inhibition zone, breakpoints, extended-spectrum beta-lactamase, Enterobacteriaceae, CLSI, EUCASTCorresponding author: Chusana Suankratay, M.D., Ph.D., Division of Infectious Diseases, Department of Internal Medicine, King Chulalongkorn Memorial
1Department of Internal Medicine, King Chulalongkorn Memorial Hospital, Bangkok, Thailand.2Division of Infectious Diseases, Department of Internal Medicine, Prapokklao Hospital, Chantaburi, Thailand.3Division of Infectious Diseases, Department of Internal Medicine, King Chulalongkorn Memorial Hospital, Bangkok, Thailand.
ABSTRACT
Background: In 2010, the United States Clinical Laboratory Standards Institute (CLSI) released the new
inhibition zone and minimal inhibitory concentration (MIC) breakpoints of beta-lactam antibiotics for
Enterobacteriaceae and recommended that there is no need to perform ESBL production tests. The reasons
for lowering breakpoints are based on pharmacokinetic and pharmacodynamic data as well as the MIC
distribution of clinical isolates of Enterobacteriaceae in several studies carried out in the United States. To
our knowledge, there have been no studies to determine inhibition zone distribution in comparison with
ESBL production among Enterobacteriaceae in Thailand.
Patients and methods: A descriptive study was carried out to determine the inhibition zone distribution
and breakpoints of beta-lactam antibiotics to distinguish between ESBL- and non-ESBL-producing clinical
isolates of Enterobacteriaceae from sterile specimens collected from adult patients who were hospitalized at
Prapokklao Hospital, Chantaburi, Thailand, from October 2011 to September 2013. We tested ESBL production
by using the double disc and disc combination methods for cefotaxime. Moreover, we performed Receiver
Operating Characteristic (ROC) curve to determine the most appropriate sensitivity and specificity of
inhibition zone breakpoints of all antibiotics tested to distinguish between ESBL- and non-ESBL-producing
Enterobacteriaceae.
Results: Among all 930 specimens (879 (94.5%) blood and 51 (5.5%) other sterile specimens), there were
312 ESBL-producing isolates of Enterobacteriaceae including 239 (37.2%) E. coli, 65 (27.5%) K. pneumoniae,
and 8 (16%) P. mirabilis. Five hundred and thirty-four (57.4%) specimens were from patients hospitalized
at Internal Medicine ward. Based on ROC curve, the inhibition zone breakpoint of cefotaxime to distinguish
between ESBL- and non-ESBL-producing Enterobacteriaceae was 26 mm, consistent with that recommended
by CLSI 2010. However, the inhibition zone breakpoint of cefazolin to distinguish between ESBL- and non-
ESBL-producing Enterobacteriaceae was 12 mm which was much different from that (23 mm) recommended
144 J INFECT DIS ANTIMICROB AGENTS Sep.-Dec. 2014
by CLSI 2010. In addition, based on ROC curves we could not determine the inhibition breakpoints of
ceftazidime, piperacillin/tazobactam, and all carbapenems.
Conclusions: To our knowledge, the present study is the first to determine the inhibition zone distribution
and breakpoints of beta-lactam antibiotics to distinguish between ESBL- and non-ESBL-producing
Enterobacteriaceae in Thailand. Only cefotaxime inhibition zone breakpoint is consistent with that
recommended by CLSI 2010. However, we cannot calculate the inhibition zone breakpoints due to no
correlations between the inhibition zone distributions of ceftazidime, piperacillin/tazobactam, and all
carbapenems with ESBL production in Enterobacteriaceae. (J Infect Dis Antimicrob Agents
2014;31:143-66.)
ROC curve for the inhibition zone breakpoint of cefotaxime ROC curve for the inhibition zone breakpoint of cefazolin
INTRODUCTION
Widespread use of broad-spectrum antibiotics
leads to development of drug-resistant bacteria.
One of the important mechanisms that mediate
development of antibiotic resistance is antibiotic
destruction by the enzymes. The beta-lactamases are
the most common enzymes that cause worldwide
medical problems. ESBL (Extended spectrum beta-
lactamase), one of the most common beta-lactamase,
has a property for hydrolyzing many betalactams
such as penicillins, cephalosporins, and monobactams.
However, it has no effect on cephamycins and
carbapenems. Beta-lactamase inhibitors can inhibit
this enzyme. This problem results in increasing use
of carbapenem group and other classes of antibiotic,
leading to increased hospitalization, hospital cost,
and adverse reactions from the antibiotics.1-3
ESBL was originally found in 1983.4 Bacteria
Vol. 31 No. 3 145Beta-lactam inhibition zone breakpoints for ESBL and non-ESBL Enterobacteriaceae:- Panarat P, et al.
145
that produce ESBL mostly are a member of the
Enterobacteriaceae family. The most common
species are Escherichia coli, Klebsiella pneumoniae,
Klebsiella oxytoca, and Proteus mirabilis, but ESBLs
are also found in other groups of bacteria. ESBL-
producing organisms usually cause nosocomial
infection, but in the past recent years, many
institutes have reported those organisms causing
community-acquired infection.5,6
The screening methods that are used for
identification of ESBL-producing organism are
disc diffusion method or dilution antimicrobial
susceptibility and cephalosporin/clavulanate
combination discs. Broth microdilution and E-
test are used as confirmatory tests.7 Those
phenotypic tests have a sensitivity of 98% and
specificity of 96.3% compared with detection of
ESBL gene by PCR amplifications.8
Before 2010, carbapenems are the treatment
of choice for those patients with infection
caused by ESBL-producing organisms regardless
of drug-susceptibility test by disc diffusion
method. Later in 2010, the United States Clinical
Laboratory Standards Institute (CLSI) released
the new inhibition zone and minimal inhibitory
concentration (MIC) breakpoints of beta-lactam
antibiotics for Enterobacteriaceae and recommended
that testing for ESBL production is not necessary.9
The reasons for lowering breakpoints are based on
pharmacokinetic and pharmacodynamic data as
well as the MIC distribution of clinical isolates of
Enterobacteriaceae in several studies carried out
in the United States. To our knowledge, there
has been no study to determine inhibition zone
distribution in comparison with ESBL production
among Enterobacteriaceae in Thailand. Thus,
our study aim was to determine the optimal
susceptibility breakpoint criteria of cefotaxime
and other beta-lactams to distinguish between ESBL-
and non-ESBL-producing Enterobacteriaceae,
and compare it with the susceptibility breakpoint
criteria that was launched by CLSI both before 2010
and in 2010, as well as the European Committee
on Antimicrobial Susceptibility Testing 2013
(EUCAST), and the British Society for Antimicrobial
Chemotherapy 2012 (BSAC).
METHODS
Study population
This study was conducted in blood or other
sterile specimens with Escherichia coli, Klebsiella
pneumoniae, Klebsiella oxytoca , and Proteus
mirabilis grown at Prapokklao Hospital, Chantaburi,
both out-patient and in-patient department, from
October 2011 to September 2013. The study included
patients 18 years of age or older. We excluded
patients who had blood cultures or other sterile
specimens that grew other organisms, and also the
above organism but in non-sterile specimens. The
institutional review board approved the protocol.
Design
A descriptive study was carried out. The data was
obtained from Microbiology laboratory, Prapokklao
Hospital, Chantaburi, which still perform phenotypic
test for ESBL production. The collected data included
organisms, specimens, ESBL-producing property,
inhibition zone size (mm) for the interesting
antibiotics (cefazolin, cefotaxime, ceftazidime,
piperacillin/tazobactam, ertapenem, imipenem,
meropenem). We determined the ESBL-producing
property by double disc and disc combination test.
In our study we used only cefotaxime for both
methods. The cefotaxime inhibition zone breakpoint
is much more reliable than ceftazidime inhibition
zone breakpoint in prediction of ESBL production.
146 J INFECT DIS ANTIMICROB AGENTS Sep.-Dec. 2014
Statistical analysis
Our study was design to determine the suscep-
tibility breakpoint of inhibition zone to distinguish
between ESBL- and non-ESBL-producing Entero-
bacteriaceae for each antibiotic by using ROC
curve (Receiver Operating Characteristic). We
calculated the sample size with the assumption
that the ratio of ESBL-producing organisms would
be 0.3.10 Thus, 461 specimens would be required
to detect a statistical difference (type 1 error 5%,
sensitivity 90%, and range of confident interval
for sensitivity 5%). All data were analyzed by using
microsoft excel 2010 and the SPSS 13.0 software
program. The categorical variables were compared
using the chisquare test, and the continuous
variables were compared using the student two-
tailed t-test. P < 0.05 was considered statistical
significant.
RESULTS
A total of 930 blood and sterile specimens were
obtained from many subspecialty wards. Those
specimens had grown 643 E. coli, 236 K. pneumoniae,
1 K. oxytoca, and 50 P. mirabilis. The proportion of
organisms that produce ESBL was 0.33. The
proportions of ESBL-producing strain for each
organism are shown in Table 1. The baseline
characteristics between ESBL and non-ESBL group
are shown in Table 2.
Comparison of inhibition zone among Entero-
bacteriaceae (E. coli, K. pneumonia, K. oxytoca
and P. mirabilis) with criteria from CLSI 2009,
CLSI 2010, EUCAST 2013, and BSAC 2012 are
shown in Table 3. Inhibition zone distribution of
Enterobacteriaceae to cefotaxime, ceftazidime, and
cefazolin were shown in Figure 1-3. The inhibition
zones distribution for other antibiotics are shown
in supplementary index.
To distinguish ESBL- and non-ESBL producing
Enterobacteriaceae using inhibition zone, we
performed ROC curve and found that only in
cefotaxime and cefazolin, the inhibition zones can
be used with inhibition zone of 26 mm and 12 mm
respectively. With this susceptibility breakpoint,
both sensitivity and specificity are greater than 95%.
ROC curve and table of sensitivity and specificity of
cefotaxime, cefazolin, and ceftazidime are shown
in Figure 4-6 and Table 4-5, respectively. ROC curve
and table of sensitivity and specificity for other
antibiotics are shown in supplementary index.
DISCUSSION
From all 930 sterile specimens, 312 specimens
(33.55%) had ESBL-producing organisms, this
proportion is similar to the previous study from
Siriraj Hospital. The study was the cross-sectional
study that collected 346 specimens with Gram-
negative bacilli grown from 249 admitted patients
from August to September 2003. One hundred and
four specimens (30.1%) had ESBL-producing
organisms.10
Several studies demonstrated that the outcome
of treatment for infection caused by Entero-
bacteriaceae with cephalosporin was associated
with minimal inhibitory concentration (MIC) and
pharmacokinetic/pharmacodynamics more than
ESBL producing property.15 This rationale motivated
reference institutes; CLSI and EUCAST, to adjust the
recommended susceptibility breakpoint criteria in
2010. Physicians can use MIC as a guidance to select
cephalosporin for treatment.15,16
From our study, the susceptibility breakpoint
that has high sensitivity (95%) and specificity (99%)
is inhibition zone of 26 mm for cefotaxime, which is
equal to that of CLSI 2010. Also, the susceptibility
breakpoint for cefazolin can also distinguish ESBL-
Vol. 31 No. 3 147Beta-lactam inhibition zone breakpoints for ESBL and non-ESBL Enterobacteriaceae:- Panarat P, et al.
147
Table 2. Comparison of baseline characteristics between organisms that produce and not produce ESBL.
MICU - Medicine intensive care unit, CCU - Coronary care unit, SICU - Surgery care unit
IPDs - In-patient departments, OPD - Out-patient department
scitsiretcarahC puorgLBSE puorgLBSE-noN eulavP
DSegaegarevA:egA
)sraey(
49.519.26 19.6176.26 2048.0
)%(elaM:redneG )%31.55(271 )51.54(972 4400.0
draW
enicidemlanretnI )%81.26(491 )%20.55(043 8100.0
UCCdnaUCIM )%65.2(8 )%38.5(63
scidepohtrOdnayregruS )%38.02(56 )%8.71(011
UCIS )%12.3(01 )%26.1(01
sDPIrehtO )%69.0(3 )%31.1(7
slatipsohytinummoCdnaDPO )%62.01(23 )%16.81(511
snemicepS
worramenobdnadoolB )%13.29(882 )%36.59(195 8641.0
diulflanipsorbereC )%29.1(6 )%92.1(8
diulflaivonyS )%29.1(6 )%56.0(4
diulflaruelP )%35.3(11 )%34.2(51
diulflaidracireP )%23.0(1 )%0(0
± ± ±
Table 1. Proportion of Enterobacteriaceae that produce and non- ESBL isolates.
)n(smsinagrO segatnecrepdnasetalosifo.oN
iloc.E )346( LBSE 932 %71.73
LBSE-noN 404 %38.26
eainomuenp.K )632( LBSE 56 %45.72
LBSE-noN 171 %64.27
acotyxo.K )1( LBSE 0 %0
LBSE-noN 1 %001
silibarim.P )05( LBSE 8 %61
LBSE-noN 24 %48
)039(latoT LBSE 213 %55.33
LBSE-noN 816 %54.66
148 J INFECT DIS ANTIMICROB AGENTS Sep.-Dec. 2014
Table 3. Inhibition zone distribution of ESBL-producing Enterobacteriaceae for various types of antibiotics
separated by criteria of 3 institutes.
Pip/tazo – piperacillin/tazobactam, R – Resistant, I – Intermediate resistant, S – Susceptible