Journal of Antimicrobial Chemotherapy (2004) 54, 134 138 DOI: 10.1093/jac/dkh274 Advance Access publication 18 May 2004 Evaluation of a new cefepime clavulanate ESBL Etest to detect extended-spectrum b-lactamases in an Enterobacteriaceae strain collection JAC Enno Stürenburg*, Ingo Sobottka, Djahesh Noor, Rainer Laufs and Dietrich Mack Institut für Infektionsmedizin, Zentrum für Klinisch-Theoretische Medizin I, Universitätsklinikum Hamburg-Eppendorf, Martinistrasse 52, D-20246 Hamburg, Germany Received 20 February 2004; returned 31 March 2004; revised 13 April 2004; accepted 14 April 2004 Objectives: In this study, we evaluated the performance of a new ESBL Etest configuration based on clavulanate synergy with cefepime compared with cefotaxime clavulanate and ceftazidime clavulanate ESBL Etest strips for the detection of extended-spectrum b-lactamases (ESBL) in an Enterobacteriaceae strain collection, with special focus on Enterobacter spp. Methods: Overall, a total of 54 clinical isolates of ESBL-producing Enterobacteriaceae species were evaluated: Enterobacter aerogenes (n 5 3), Enterobacter cloacae (n 5 10), Escherichia coli (n 5 10), Klebsiella oxytoca (n 5 3), Klebsiella pneumoniae (n 5 25) and Proteus mirabilis (n 5 3). To check Etest behaviour with resistance phenotypes similar to ESBL, our panel was expanded by six clinical isolates of K. oxytoca that were identified as putative producers of their chromosomal K1 b-lactamase. Results: With this panel, ESBL Etest was 98% sensitive with cefepime clavulanate, 83% with cefotaxime clavulanate, and 74% with ceftazidime clavulanate strips. Concentrating on Enterobacter spp., reliable ESBL detection could only be achieved by the new cefepime clavulanate strip since it confirmed ESBL production in all strains (100% sensitivity) whereas only 4/13 (31%) of Enterobacter strains were positive using cefotaxime clavulanate or ceftazidime clavulanate strips. A limitation of using the new cefepime strip was less than optimal specificity with K1 phenotypes of K. oxytoca: among six strains, four isolates were scored false-positive by Etest strips containing cefepime clavulanate. Conclusion: The new Etest ESBL strip containing cefepime clavulanate is a valuable supplement to current methods for detection of ESBLs. In our study collection, the cefepime clavulanate strip was the best configuration for detection of ESBLs, particularly in Enterobacter spp. Keywords: ESBL detection, Etest methodology Introduction Resistance to third-generation cephalosporins through the acquisition and expression of extended-spectrum b-lactamases (ESBLs) among Enterobacteriaceae is increasing. The clinical implications of ESBLs are extremely serious, and sensitive diagnostic methods are urgently needed to guide therapy, monitor resistance development and implement intervention strategies. 1,2 Genotypic methods based on enzyme assays, PCR and others are not suitable for routine clinical testing. Two different Etest gradient formats were developed to detect ESBL based on reduction of ceftazidime or cefotaxime MICs by >_ 3 two-fold dilutions in the presence of clavulanic acid. Successful use of these strips for ESBL detection has been reported. However, evaluation studies have involved limited species coverage, including predominantly Escherichia coli and Klebsiella spp., 3,4 or the test strains were all transconjugants generated in vitro. 5,6 At the same time, ESBL phenotypes have become more complex due to the production of multiple enzymes including inhibitor-resistant TEM enzymes, AmpC, enzyme hyperproduction and porin loss. 7 10 In addition, ESBLs are now reported in a growing number of genera other than E. coli or Klebsiella spp., such as Enterobacter aerogenes, Enterobacter cloacae, Proteus mirabilis and Serratia marcescens. In these organisms, detection of ESBLs based on 11 13 reduction in oxyimino cephalosporin MICs by clavulanic acid is difficult. Clavulanic acid may induce high-level expression of AmpC chromosomal enzyme, and may then antagonize rather than protect the bacterial activity of the partner b-lactam, thus... *Corresponding author. Tel: +49-40-42803-3147; Fax: +49-40-42803-4881; E-mail: e.stuerenburg@uke.uni-hamburg.de... 134 JAC vol.54 no.1 q The British Society for Antimicrobial Chemotherapy 2004; all rights reserved.
ESBL detection by the new cefepime clavulanate ESBL Etest masking any synergy arising from inhibition of an ESBL. 14 One approach to overcoming this problem is to use cefepime as the indicator drug. High-level AmpC production has minimal effect on the activity of cefepime, making this drug a more reliable agent for ESBL detection in the presence of an AmpC enzyme. 15,16 Very recently, this idea has been translated into routine application as a new Etest strip based on clavulanate synergy with cefepime was launched to the market. In this study, we evaluated the performance of the new cefepime clavulanate Etest configuration with the other two existing strips in a collection of 54 clinical ESBL isolates, with special focus on Enterobacter spp. Furthermore, in order to gain an impression of how well the newly available Etest strip would perform with a resistant phenotype very similar to ESBL, we expanded our panel of test strains with six clinical K. oxytoca isolates that were identified as putative hyperproducers of their chromosomal K1 enzymes. Like TEM-, SHV- or CTX-M ESBLs, K1 b-lactamase is an Ambler class A enzyme too and is naturally susceptible to inhibition by clavulanate. Materials and methods Bacterial strains The bacterial isolates selected for this study included 54 clinical non-duplicate ESBL isolates, the majority of which were collected and characterized in a previous evaluation study. 17 A few other ESBL strains were sent from the laboratories acknowledged. The organisms were considered ESBL-positive when TEM-, SHV-, or CTX-M related ESBL genes had been identified by PCR and DNA sequencing and when a reduction of at least eight-fold in the MICs of either ceftazidime or cefotaxime in the presence of clavulanic acid had been detected, using broth microdilution according to the NCCLS. 18,19 The clinical isolates consisted of E. aerogenes (n = 3), E. cloacae (n = 10), E. coli (n = 10), K. oxytoca (n = 3), K. pneumoniae (n = 25) and P. mirabilis (n = 3). The E. aerogenes strains analysed harboured CTX-M-1 (n = 2) and SHV-5 (n = 1) ESBLs. The 10 non-repetitive isolates of E. cloacae, which were isolated over a period of 2 years, included TEM type (ABL: A184V) (n = 7) and SHV-12 (n = 3) ESBLs. The E. coli strains included SHV-2 (n = 1), SHV- 5 (n = 1), SHV-12 (n = 3), TEM-26 (n = 1), TEM-52 (n = 1), TEM-111 (n = 1), CTX-M-1 (n = 1) and CTX-M-23 (n = 1). The K. oxytoca isolates harboured CTX-M-1 (n = 2) and SHV-12 (n = 1). The P. mirabilis strains included CTX-M-1 (n = 1), CTX- M-22 (n = 1) and TEM-92 (n = 1). The K. pneumoniae strains harboured SHV-2 (n = 5), SHV-2a (n = 1), SHV-5 (n = 4), SHV- 12 (n = 11), SHV-19 (n = 1), LEN type (ABL: N53S, A201P, P218A) (n = 1), TEM-47 (n = 1) and TEM-110 (n = 1). The two NCCLS recommended ESBL test quality control strains ATCC 25922 E. coli (negative control) and ATCC 700603 K. pneumoniae (ESBL positive, SHV-18) were also included in all experiments. To check Etest behaviour with resistance phenotypes very similar to ESBL, our panel was expanded by clinical K. oxytoca isolates (six strains) that were identified as putative hyperproducers of their K1 chromosomal b-lactamase, based on susceptibility to ceftazidime (MIC<_ 8 mg/l) but resistance to cefuroxime (MIC>_ 32 mg/l) and negative results with the bla SHV, bla TEM and bla CTX-M PCR assays. Etest procedure, reading and interpretation The Etest ESBL strip (AB Biodisk, Solna, Sweden) is a plastic drug-impregnated strip, one end of which generates a stable concentration gradient of ceftazidime (MIC test range, 0.5 32 mg/l) and the remaining end of which generates a gradient of ceftazidime (MIC test range, 0.064 4 mg/l) plus 4 mg/l clavulanic acid. An Etest ESBL strip containing cefotaxime (MIC test range, 0.25 16 mg/l) and cefotaxime (MIC test range, 0.016 1 mg/l) plus 4 mg/l clavulanic acid is also available. Very recently, a third strip for ESBL detection containing cefepime (MIC test range, 0.25 16 mg/l) and cefepime (MIC test range, 0.064 4 mg/l) plus 4 mg/l clavulanic acid has been produced. The Etest procedure, reading and interpretation were carried out according to the manufacturer s instructions. Isolated colonies from an overnight agar plate were suspended in saline (0.85% NaCl) to achieve an inoculum equivalent to 0.5 McFarland Standard. This suspension was swabbed on a Mueller Hinton agar plate (Oxoid, UK) and allowed to dry completely. An ESBL Etest strip was then applied to the agar surface with sterile forceps and the plate was incubated at 368C for 18 h. ESBL results were read either as MIC values or observation of phantom zones or deformation of inhibition ellipses. Reduction of MIC by >_ 3 two-fold dilutions in the presence of clavulanic acid is indicative of ESBL production. Deformation of ellipses or the presence of a phantom zone is also indicative of ESBL production even if the MIC ratio is <8 or cannot be read. Results Cefotaxime, ceftazidime, or cefepime MIC reduction by >_ 3 twofold dilutions with clavulanic acid is indicative of ESBL activity (Figure 1a). Applying this rule, for the majority of our ESBL test strains, it was easy and straightforward to assess their ESBL status (Table 1). Nevertheless, varying degrees of clavulanatemediated MIC reduction were noted. For the cefotaxime clavulanate and ceftazidime clavulanate strips, the MIC reductions were consistently more pronounced than for the cefepime clavulanate strips (Table 1). Moreover, we have shown that the results of the ESBL Etests were sometimes difficult to read because the isolates exhibited the effects demonstrated in Figure 1(b and c). According to the Etest interpretation guidelines, these interferences are phantom zones (P) or ellipse deformations (D), which are directly indicative of ESBL production. In our set of ESBL strains, P and D effects were particularly frequent when using cefepime clavulanate or cefotaxime clavulanate strips. The frequencies of observing P or D effects by cefotaxime clavulanate, ceftazidime clavulanate and cefepime clavulanate strips were 61%, 20% and 85%, respectively. The Etest guidelines state that even if the MIC reductions with clavulanate fall below the threshold of 3 two-fold dilution steps, the isolate should be considered as ESBL-positive when a P or D effect is present. By these criteria, the overall sensitivities for cefotaxime clavulanate, ceftazidime clavulanate and cefepime clavulanate strips were 83% (45/54 strains), 74% (39/54 strains) and 98% (53/54 strains), respectively (Table 1). The cefepime clavulanate strips were found to be particularly accurate for detecting ESBLs among Enterobacter isolates (100% detection sensitivity), whereas cefotaxime clavulanate or ceftazidime clavulanate strips yielded an unacceptably high number of nondeterminable (Figure 1d) or negative results and thus showed marked inability to detect ESBL production (31% detection sensitivity) within this group (Table 1). 135
E. Stürenburg et al. Figure 1. (a) Clear cut ESBL positive: MIC ratio of ceftazidime/ceftazidime + clavulanate >_ 32/0.064 >_ 512. (b) A rounded phantom inhibition zone below cefepime end indicative of ESBL. (c) Deformation of the cefepime inhibition ellipse indicative of ESBL. (d) When MIC values were above the test device range, interpretation was non-determinable. Antibiotic codes: TZ, ceftazidime; TZL, ceftazidime plus clavulanate; PM, cefepime; PML, cefepime plus clavulanate; CT, cefotaxime; CTL, cefotaxime plus clavulanate. Table 1. ESBL Etest results for the Enterobacteriaceae isolates studied + CTX-CLA ESBL Etest CAZ-CLA ESBL Etest FEP-CLA ESBL Etest Log 2 reduction in MIC a ND/ + Log 2 reduction in MIC a ND/ + Log 2 reduction in MIC a Performance by organisms E. aerogenes (n = 3) 3 3 3 5.7 (4.0 7.0) E. cloacae (n = 10) 4 6.0 (3.6 10) 6 4 7.5 (6.4 9.0) 6 10 4.1 (3.0 6.0) E. coli (n = 10) 10 7.8 (4.0 10) 10 7.3 (3.0 9.0) 10 5.7 (3.0 8.0) K. oxytoca (n = 3) 3 7.7 (5.0 9.0) 2 6.1 (6.4 9.0) 1 3 6.8 (5.6 8.0) K. pneumoniae (n = 25) 25 8.7 (6.0 10) 20 8.4 (6.0 9.0) 5 24 5.4 (3.0 8.0) 1 P. mirabilis (n = 3) 3 8.6 (7.4 10) 3 4.0 (4.0) 3 4.7 (3.6 6.4) Total (n = 54) 45 8.0 (3.4 10) 9 b 39 7.7 (3.0 9.0) 15 c 53 5.3 (3.0 8.0) 1 d Performance by ESBL types TEM type (n = 12) 6 7.4 (3.4 10) 6 5 6.9 (4.0 9.0) 7 11 4.8 (3.0 8.0) 1 SHV type (n = 34) 32 8.1 (3.6 10) 2 28 8.3 (6.0 9.0) 6 33 5.3 (3.0 8.0) 1 CTX-M type (n = 9) 7 9.0 (7.0 10) 2 6 6.6 (5.0 8.4) 3 9 5.8 (3.6 8.0) K1 hyperproducers of K. oxytoca K. oxytoca K1 hyperproducer (n = 6) 3 4.7 (4.4 5.0) 3 6 4 4.0 (4.0) 2 CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; CLA, clavulanate; ND, not determinable. a Geometric mean and range of results. b Results from cefotaxime clavulanate strips were not determinable or negative in two E. aerogenes (CTX-M-1), one E. aerogenes (SHV-5), and six E. cloacae (TEM type, ABL:A184V) strains, respectively. c Results from ceftazidime clavulanate strips were not determinable or negative in one E. aerogenes (SHV-5) strain, two E. aerogenes (CTX-M-1), two E. cloacae (TEM type, ABL:A184V), one E. cloacae (SHV-12), one K. oxytoca (CTX-M-1), one K. pneumoniae (SHV-12), one K. pneumoniae (SHV-2; TEM-110), one K. pneumoniae (SHV-5), and two K. pneumoniae (SHV-2) strains, respectively. d Results from cefepime clavulanate strips were not determinable or negative in one K. pneumoniae (SHV-2; TEM-110) strain. ND/ Some difficulties were encountered when using the new cefepime clavulanate strip with resistant phenotypes of K. oxytoca. Among six strains hyperproducing their chromosomal K1 b-lactamase but lacking ESBL, four isolates were scored false-positive by cefepime clavulanate strips (Table 1). Similar behaviour was seen with cefotaxime clavulanate strips: 3/6 gave a false-positive result by this ESBL Etest. None did so by ceftazidime clavulanate ESBL Etest. However, with those K1 strains that were tested false-positive by Etest, the reduction in MICs was significantly less than that for the three K. oxytoca strains producing ESBL (Table 1). Discussion Using our study collection, the new Etest ESBL strip containing cefepime clavulanate appeared the most sensitive test for ESBL detection. Even as a stand-alone test, this strip had an excellent detection sensitivity of 98%. It is important to observe that ESBL 136
ESBL detection by the new cefepime clavulanate ESBL Etest results involved reading the MIC values to calculate the ratio of MIC reduction by clavulanic acid as well as observing phantom zones or ellipse deformations, which are directly indicative of ESBL activity. This was particularly important when using cefepime clavulanate or cefotaxime clavulanate strips. In line with previous reports, 20 our study confirmed that neither ceftazidime clavulanate nor cefotaxime clavulanate strips were suited for detecting ESBL in Enterobacter species. The cefepime clavulanate strip was the only configuration that enabled accurate detection (100%) within this group of organisms, where inducible chromosomal AmpC b-lactamase can interfere with clavulanate synergy. Detection of ESBL production among such strains is known to be challenging, 14,21 and false-negative results using cefotaxime or ceftazidime have also been described in other studies. 22,23 A limitation using the new cefepime clavulanate strip was less than optimal specificity with resistant phenotypes of K. oxytoca. As demonstrated, cefepime clavulanate strips as well as cefotaxime clavulanate strips were considerably susceptible to false-positive results with strains hyperproducing their chromosomal K1 b-lactamase but lacking ESBL. Careless use of the new strip may thus result in an unacceptably high number of falsepositive results. In order to assess the ESBL status in K. oxytoca, it might be more prudent to consider the overall susceptibility profile rather than depending solely on singular Etest results. K1 hyperproducer organisms have a typical pattern of resistance, being consistently resistant to cefuroxime and aztreonam, having borderline resistance to cefotaxime and cefepime, but remaining fully susceptible to ceftazidime (as shown here with the ceftazidime ESBL Etest). 21 In conclusion, today s most commonly used practice to confirm ESBL enzymes by carrying out clavulanate synergy tests with ceftazidime, cefotaxime or cefpodoxime may no longer be sufficient in populations with a high prevalence of ESBLproducing Enterobacter species. For such situations, where inducible chromosomal AmpC b-lactamase can interfere with clavulanate synergy from a co-existing ESBL, the new cefepime clavulanate strip could be a more sensitive alternative. Finally, facing the growing complexity of ESBLs, it becomes more and more apparent that regular species identification forms the base on which accurate ESBL detection should be built. Acknowledgements We are grateful to the following for sharing their ESBL strains with us for this research project: Mrs Aline Wenger, Centre Hospitalier Universitaire Vaudois, Lausanne, Switzerland (10 Enterobacter cloacae isolates); Dr Elke Halle, Institut für Mikrobiologie und Hygiene, Campus Charité Mitte, Berlin, Germany (one Proteus mirabilis strain); Dr Karin Schwegmann, Niedersächsisches Landesgesundheitsamt, Hannover, Germany (one Proteus mirabilis strain). This work was presented in part at the 55th annual convention of the Deutsche Gesellschaft für Hygiene und Mikrobiologie (DGHM), 28 September 1 October 2003, Dresden, Germany. This study was partly supported by Viva Diagnostika, Köln, Germany. References 1. Bradford, P. A. (2001). Extended-spectrum beta-lactamases in the 21st century: characterization, epidemiology, and detection of this important resistance threat. Clinical Microbiology Reviews 14, 933 51. 2. Stürenburg, E. & Mack, D. (2003). Extended-spectrum b-lactamases: implications for the clinical microbiology laboratory. Journal of Infection 47, 273 95. 3. Vercauteren, E., Descheemaeker, P., Ieven, M. et al. (1997). Comparison of screening methods for detection of extended-spectrum beta-lactamases and their prevalence among blood isolates of Escherichia coli and Klebsiella spp. in a Belgian teaching hospital. Journal of Clinical Microbiology 35, 2191 7. 4. Leverstein-van Hall, M. A., Fluit, A. C., Paauw, A. et al. (2002). Evaluation of the Etest ESBL and the BD Phoenix, VITEK 1, and VITEK 2 automated instruments for detection of extended-spectrum beta-lactamases in multiresistant Escherichia coli and Klebsiella spp. Journal of Clinical Microbiology 40, 3703 11. 5. Cormican, M. G., Marshall, S. A. & Jones, R. N. (1996). Detection of extended-spectrum beta-lactamase (ESBL)-producing strains by the Etest ESBL screen. Journal of Clinical Microbiology 34, 1880 4. 6. MacKenzie, F. M., Miller, C. A. & Gould, I. M. (2002). Comparison of screening methods for TEM- and SHV-derived extended-spectrum beta-lactamase detection. Clinical Microbiology and Infection 8, 715 24. 7. Bonomo, R. A. & Rice, L. B. (1999). Inhibitor resistant class A beta-lactamases. Frontiers in Bioscience 4, e34 41. 8. Philippon, A., Arlet, G. & Jacoby, G. A. (2002). Plasmiddetermined AmpC-type beta-lactamases. Antimicrobial Agents and Chemotherapy 46, 1 11. 9. Xiang, X., Shannon, K. & French, G. (1997). Mechanism and stability of hyperproduction of the extended-spectrum beta-lactamase SHV-5 in Klebsiella pneumoniae. Journal of Antimicrobial Chemotherapy 40, 525 32. 10 Martinez-Martinez, L., Pascual, A., Hernandez-Alles, S. et al. (1999). Roles of beta-lactamases and porins in activities of carbapenems and cephalosporins against Klebsiella pneumoniae. Antimicrobial Agents and Chemotherapy 43, 1669 73. 11. Albertini, M. T., Benoit, C., Berardi, L. et al. (2002). Surveillance of methicillin-resistant Staphylococcus aureus (MRSA) and Enterobacteriaceae producing extended-spectrum beta-lactamase (ESBLE) in Northern France: a five-year multicentre incidence study. Journal of Hospital Infection 52, 107 13. 12. Spanu, T., Luzzaro, F., Perilli, M. et al. (2002). Occurrence of extended-spectrum beta-lactamases in members of the family Enterobacteriaceae in Italy: implications for resistance to beta-lactams and other antimicrobial drugs. Antimicrobial Agents and Chemotherapy 46, 196 202. 13. Coudron, P. E., Moland, E. S. & Sanders, C. C. (1997). Occurrence and detection of extended-spectrum beta-lactamases in members of the family Enterobacteriaceae at a veterans medical center: seek and you may find. Journal of Clinical Microbiology 35, 2593 7. 14. Thomson, K. S. (2001). Controversies about extended-spectrum and AmpC beta-lactamases. Emerging Infectious Diseases 7, 333 6. 15. Hancock, R. E. W. & Bellido, W. (1992). Factors involved in the enhanced efficacy against gram-negative bacteria of fourth-generation cephalosporins. Journal of Antimicrobial Chemotherapy 29, Suppl. A, 1 6. 16. Gottlieb, T. & Wolfson, C. (2000). Comparison of the MICs of cefepime for extended-spectrum beta-lactamase-producing and non-extended-spectrum beta-lactamase-producing strains of Enterobacter cloacae. Journal of Antimicrobial Chemotherapy 46, 330 1. 17. Stürenburg, E., Sobottka, I., Feucht, H. H. et al. (2003). Comparison of BD Phoenix and VITEK2 automated antimicrobial susceptibility test systems for extended-spectrum beta-lactamase detection in Escherichia coli and Klebsiella species clinical isolates. Diagnostic Microbiology and Infectious Disease 45, 29 34. 137
E. Stürenburg et al. 18. National Committee for Clinical Laboratory Standards (2003). Performance Standards for Antimicrobial Susceptibility Testing; Thirteenth Informational Supplement. In MIC Testing Supplemental Tables. NCCLS document M100-S13. NCCLS, Wayne, PA, USA. 19. National Committee for Clinical Laboratory Standards (2003). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically Sixth Edition: Approved Standard M7-A6. NCCLS, Wayne, PA, USA. 20. Laffineur, K., De Gheldre, Y., Berhin, V., et al. (2003). Evaluation of three different Etest strips for the detection of extended-spectrum b-lactamases (ESBL) in various Enterobacteriaceae species. In Program and Abstracts of the 43rd Interscience Conference on Antimicrobial Agents and Chemotherapy, Chicago, IL, 2003. Abstract D- 203, p. 167. American Society for Microbiology, Washington, DC, USA. 21. Livermore, D. M. & Brown, D. F. (2001). Detection of betalactamase-mediated resistance. Journal of Antimicrobial Chemotherapy 48, Suppl.1, 59 64. 22. Tzelepi, E., Giakkoupi, P., Sofianou, D. et al. (2000). Detection of extended-spectrum beta-lactamases in clinical isolates of Enterobacter cloacae and Enterobacter aerogenes. Journal of Clinical Microbiology 38, 542 6. 23. De Gheldre, Y., Avesani, V., Berhin, C. et al. (2003). Evaluation of the Oxoid combination disks for detection of extended-spectrum b-lactamases. Journal of Antimicrobial Chemotherapy 52, 591 7. 138