suspected KPC and other carbapenemase producers among species of Nacional de Enfermedades Infecciosas (INEI)- ANLIS Dr. Carlos G.

Transcription

1 JCM Accepts, published online ahead of print on 1 December 0 J. Clin. Microbiol. doi:./jcm.0- Copyright 0, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved. 1 Can we use imipenem and meropenem VITEK MICs for the detection of suspected KPC and other carbapenemase producers among species of Enterobacteriaceae? Running Title: Detection of carbapenemases with VITEK Authors: Fernando Pasteran 1, Celeste Lucero 1, Rolando Soloaga, Melina Rapoport 1, and Alejandra Corso 1 Author s affiliation: 1 Servicio Antimicrobianos, Departamento Bacteriología, Instituto Nacional de Enfermedades Infecciosas (INEI)- ANLIS Dr. Carlos G. Malbrán, Ministerio de Salud y Ambiente de la Nación, Ciudad Autónoma de Buenos Aires, Argentina. Servicio de Microbiología, Hospital Naval Buenos Aires Cirujano Mayor Dr. Pedro Mallo, Ciudad Autónoma de Buenos Aires, Argentina. Address of the Institution at which the work was performed: Servicio Antimicrobianos, INEI-ANLIS Dr. Carlos G. Malbrán, Av. Velez Sarsfield (C1AFF), Buenos Aires, Argentina. Corresponding author: Alejandra Corso Address: Servicio Antimicrobianos, INEI - ANLIS Dr. Carlos G. Malbrán, Velez Sarsfield Ave. (C1AFF), Buenos Aires, Argentina. acorso@anlis.gov.ar Phone/FAX: Downloaded from on August 1, 01 by guest 1

2 ABSTRACT Imipenem and meropenem VITEK MICs were evaluated against a panel of Enterobacteriaceae for identification of carbapenemase producers. The sensitivity and specificity values for the new CLSI interpretative criteria (M0-S0-U) were: %/% for imipenem and %/% for meropenem, respectively. We propose an algorithm that is highly sensitive (%) and specific (%) for carbapenemase screening based on the combined use of imipenem and meropenem MICs. Downloaded from on August 1, 01 by guest

3 Carbapenems are increasingly utilized as drugs of last resort against a variety of infections due to the emergence of extended spectrum beta-lactamase-(esbl)- producing Enterobacteriaceae (). The emergence of carbapenem-resistant Enterobacteriaceae is therefore worrisome as the antimicrobial armamentarium is consequently restricted (, 1). Resistance to carbapenems in Enterobacteriaceae could be related to carbapenemases or to a dual mechanism associating an outer membrane permeability defect and β-lactamases such as AmpC cephalosporinase and ESBL, particularly the presence of CTX-M variants (-, 1, 1,, ). The vast majority of acquired carbapenemases belong to three of the four known classes of β-lactamases, namely Ambler class A (KPC, Sme, NMC-A, IMI and some allelic variants of the GES/IBC enzymes), Ambler class B (metallo- β-lactamases, MBLs), and Ambler class D (oxacillinases, OXAs) (). The location of carbapenemase genes on highly mobile genetic elements has contributed to their rapid spread and the frequent cotransfer of multiple other antibiotic resistance factors (, 1, 1). The ability to limit the spread of carbapenemase producers will require effective laboratory screening methods to rapidly identify patients infected with these organisms. Automated antimicrobial susceptibility testing systems, such as VITEK (biomérieux, Marcr LÈtoile, France), are commonly used in microbiology laboratories to decrease the laboratory turn-around-time. However, several reports have questioned the ability of VITEK to identify carbapenemase producers (1,, 1,, ). Previous reports suggest an ertapenem MIC.0 µg/ml (formerly, an ertapenem resistance result; ) as the most accurate way to detect KPC carbapenemase (, 1). In Argentina, among several other countries with endemic CTX-M, a large proportion of nosocomial Enterobacteriaceae display ertapenem resistance (about %, 1%, 0%, % of the Escherichia coli, Klebsiella pneumoniae, Citrobacter freundii and Enterobacter spp., respectively, WHONET- Downloaded from on August 1, 01 by guest

4 Argentina Network, n=.1), although most of them did not produce carbapenemases by molecular methods. The selection of CTX-M--producing mutants with porin loss was responsible for this ertapenem resistance. Using ertapenem and meropenem as indicators (MICs µg/ml and µg/ml, respectively), the VITEK was not able to differentiate true resistance mediated by carbapenemases from that mediated by AmpC cephalosporinases and ESBLs in combination with an outer membrane permeability defect, largely overestimating carbapenemase producers (). Therefore, in areas where these alternative mechanisms of carbapenem resistance are common, the use of indicators with such low specificity could significantly delay the identification of patients infected with true carbapenemases, affecting the appropriate infection control policies. In addition, the need to confirm a larger number of strains could also cause an increase in the lab costs. For these reasons, it is necessary to find a screening strategy based on the use of carbapenems other than ertapenem for a more accurate identification of suspicious isolates with true carbapenemases. In addition, many VITEK cards do not contain ertapenem, therefore, having an alternative strategy for carbapenemase screening also benefit users of these models. The Clinical Laboratory Standards Institute (CLSI) recently changed the susceptibility breakpoint for meropenem, imipenem, and doripenem to 1.0 µg/ml and the ertapenem susceptible cutoff was modified to 0. µg/ml (M-0S-0-U) (). While the resistant breakpoints were moved to µg/ml for imipenem, meropenem and doripenen and 1 µg/ml for ertapenem. Detection of carbapenemase producers by VITEK based on these new susceptibility breakpoints has not been evaluated so far. Herein, we evaluated the ability of VITEK to screen carbapenemase-producing Enterobacteriaceae using imipenem and meropenem MICs and the updated CLSI breakpoints. We also assessed the capability of these carbapenems to distinguish true Downloaded from on August 1, 01 by guest

5 carbapenemases from carbapenem resistance resulting from combinations of impermeability and AmpC or an ESBL. In addition, we propose an algorithm that will help clinical laboratories to detect true resistance mediated by carbapenemase producers in scenarios with high baseline resistance to ertapenem due to non-carbapenemase mechanisms (The findings of this study were partly presented at the 1th International Congress on Infectious Diseases, abstr. 1, 0) A panel of genotypically characterized Enterobacteriaceae (n=) composed of diverse bacterial genera with distinct carbapenem susceptibility patterns were included in this study. The genotypes of the isolates were characterized previously (1, ) by PCR for bla VIM, bla IMP, bla SPM, bla KPC, bla Sme, bla OXAs (subgroups I, II and III as defined in reference ), bla OXA-, bla IMI/NMC-A, bla CTX-M, bla TEM, bla PER, bla SHV, bla AmpC. In addition, to exclude the possible presence of other not yet described carbapenemases among the negative control panel, imipenemase activity of cell-free extracts from overnight broth culture was determined by spectrophotometric assays (1). Outer membrane protein detection was performed by SDS-PAGE among ertapenem resistant carbapenemase nonproducers. () AmpC hyperproduction was initially screened by isolectric focusing (IEF) using a substrate-base development method. Crude extracts with positive AmpC IEF bands and in situ inhibition with oxacillin (1 mm) were subjected to spectrophotometric analysis as describe previously (). The carbapenemases represented were: KPC ( K. pneumoniae, E. cloacae, C. freundii, 1 E. coli, 1 Serratia marcescens, Salmonella), Sme ( S. marcescens), IMI/NMC-A ( E. cloacae), GES ( K. pneumoniae, 1 E. agglomerans), OXA-1, a novel carbapenemase closely related to OXA-, with an extended activity to expanded- Downloaded from on August 1, 01 by guest

6 spectrum cephalosporins ( K. pneumoniae, 1 E. cloacae) (), IMP- (1 E. cloacae) and VIM- (K. pneumoniae, 1 E. cloacae, 1 Providencia retgeri). The carbapenemase nonproducers were (n=): K. pneumoniae, Proteus mirabilis, 1 E. coli, 1 K. oxytoca and 1 P. penneri isolates with diverse ESBLs (CTX-M-, CTX-M-, CTX-M- 1, PER-, SHV- or SHV-1), K. pneumoniae isolates with CTX-M- and porin loss, E. cloacae isolates with hyperproduction of AmpC plus porin loss, E. cloacae and 1 C. freundii isolates with hyperproduction of AmpC, S. marcescens isolates with inducible AmpC and ESBLs (CTX-M-) plus porin loss, E.cloacae, C. freundii, 1 S. marcescens, 1 Morganella morganii and 1 P. stuartii isolates with inducible AmpC with inducible expression, K. pneumoniae, 1 E. coli, 1 P. mirabilis and 1 Shigella flexneri with diverse plasmidic AmpCs (CMY, AAC-1, DHA-1 or FOX-), 1. E. coli and 1 C. koseri isolates with penicillinases, and, E. coli, 1 E. cloacae and 1 Samonella sp. quality control strains. By using agar dilution, we determined that % and % of the carbapenemase producers and nonproducers, respectively, were non-susceptible to ertapenem. Strains were subcultured twice before testing. The isolates were from clinical source (with the exception of the quality control strains and two Salmonella with KPC that were obtained by conjugation assays) and they were single isolates from each patient. All the isolates corresponded to different clonal types as revealed by PFGE, with the exception of K. pneumoniae possessing KPCs (strains were divided into six pulse types). Isolates of the major PFGE pattern of K. pneumoniae possessing KPC (0% of the strains) corresponded to the molecular type ST (1). Downloaded from on August 1, 01 by guest The MICs of carbapenems were determined in duplicate by VITEK using the AST- N0 card, which included imipenem (range 1-1 µg/ml) and meropenem (0.-1 µg/ml), according to the manufacturer's instructions. Inoculums were adjusted to the

7 required optical density using the VITEK Densi-Check. Plate counts of representative samples were analyzed to corroborate the compliance with the required inoculum size We evaluated two criteria for carbapenemase screening using VITEK : (i) we assessed the ability of the Advanced Expert System (AES), a software that uses the antimicrobial susceptibility results to suggest a presumed mechanism of resistance of the tested isolate; ii) we analyzed the carbapenem susceptibility results to identify isolates suspected of carbapenemase production. The VITEK AES compares each MIC result of the tested isolate with the modal MIC distribution of bacteria with known resistance mechanisms included in the database. Based on this comparison, the system indicates the potential presence of carbapenemases. (It is required that, at least, all but one of the MICs obtained for the tested strains fit the modal distribution of the presumed resistance mechanism). The AES software predicted the presence of carbapenemases in / (%) of carbapenemase producers (the highest efficiency for OXA-1 with 0% of correct detection, followed by KPCs with %, Sme %, other class A carbapenemases 0% and MBLs %). The AES offered no interpretation (isolates were erroneously reported with an inconsistent result) for carbapenemase producers ( S. marcescens isolates with Sme, E. cloacae with IMI/NMC-A, 1 E. agglomerans with GES, 1 K. pneumoniae and 1 P. retgeri with VIM and 1 E. coli and 1 S. marcescens with KPC), even though all of them were resistant to imipenem (MICs µg/ml). The remaining carbapenemase producers ( K. pneumoniae isolates with KPC, 1 K. pneumoniae with KPC plus PER-, 1 S. marcescens with Sme plus CTX-M- and 1 K. pneumoniae with VIM plus CTX-M-) were inferred by the AES to have ESBL production alone or in combination with impermeability. Nine out of (1%) isolates without true Downloaded from on August 1, 01 by guest

8 carbapenemases (all nine isolates with permeability defects, with CTX-M- and with AmpC) were also inferred to produce a carbapenemase giving % specificity To address the usefulness of the second criteria, the VITEK MIC s of the isolates were defined as susceptible, intermediate and resistant to imipenem and meropenem according to the updated 0 CLSI breakpoints (). Intermediate susceptibility or resistance (a carbapenem MIC µg/ml) was detected in % and % of the carbapenemase producers with imipenem (Figure 1a) and meropenem (Figure 1b), respectively. Strains that remained susceptible to at least one of the carbapenem tested (MIC 1 µg/ml) were: 1 K. pneumoniae isolate (VIM) with imipenem (Figure 1a) and K. pneumoniae ( KPC and 1 VIM), S. marcescens (Sme), 1 E. coli (KPC) and 1 E. cloacae (KPC) isolates with meropenem (Figure 1b). Only the K. pneumoniae isolate with VIM retained susceptibility to both carbapenems (Figure 1a and 1b). Intermediate susceptibility or resistance (a carbapenem MIC µg/ml) was also detected in 1% and 1% of the carbapenemase nonproducers with imipenem (Proteae members were excluded because these isolates have naturally elevated imipenem MICs as shown in Figure 1a) and meropenem (Figure 1b), respectively. Strains that showed MICs in the non-susceptible categories were: E. cloacae (with AmpC plus porin loss) isolates with imipenem (Figure 1a), K. pneumoniae (CTX-M- plus porin loss), S. marcescens (CTX-M- plus porin loss) and 1 E. cloacae (AmpC) isolates with meropenem (Figure 1b), and E. cloacae (AmpC plus porin loss) and 1 K. pneumoniae (CTX-M- plus porin loss) isolates with both carbapenems (Figure 1a and 1b). Thus, the recently approved CLSI breakpoints for meropenem and imipenem, which define nonsusceptibility by a MIC µg/ml, increased the capture of carbapenemase producers, especially with imipenem. However, these changes also enhanced the poor ability of the Downloaded from on August 1, 01 by guest

9 commercial system to distinguish carbapenemase producers from isolates with an ESBL and/or AmpC combined with porin loss resulting in poor specificity (Figure 1a and 1b). Thus, we explored other screening cutoff points for a more specific detection of suspected carbapenemases by VITEK, while retaining the highest possible sensitivity. As the overlap in the MICs for isolates with these contrasting resistance mechanisms mostly involved only one of the two carbapenems tested, we explored a screening strategy based on the combined use of imipenem and meropenem MICs. As shown in the figure 1c, almost all of the carbapenemase producers (with the exception of 1 K. pneumoniae isolate with VIM-) showed simultaneously an imipenem MIC µg/ml and a meropenem MIC 1 µg/ml. Conversely, most of the strains with alternative carbapenem resistant mechanisms, including the proteae members, had at least one carbapenem MIC below these cutoff values (an imipenem MIC 1 µg/ml or a meropenem MIC 0. µg/ml). Only 1 K. pneumoniae isolate (CTX-M- plus porin loss) and E. cloacae isolates (AmpC plus porin loss) displayed MICs µg/ml for both carbapenem (Figure 1c). Based on these findings, we propose an algorithm for carbapenemase screening by VITEK (Figure ): when an isolate has both an imipenem MIC µg/ml and meropenem MIC 1 µg/ml (this means that both MICs have to be above their respective cutoff points), carbapenemase producer should be suspected. Subsequently it should be confirmed by the means of more specific methods (1, ). On the contrary, if the MIC to imipenem is 1 µg/ml (susceptible according to the updated CLSI interpretative criteria) or that of meropenem 0. µg/ml, the presence of carbapenemases can be excluded. The sensitivity, specificity, positive predictive value and negative predictive value for screening suspected carbapenemase producers ( imipenem MIC µg/ml and a meropenem MIC 1 µg/ml) were: %, %, % and %, respectively (Figure 1c). It should be mentioned that this algorithm is based Downloaded from on August 1, 01 by guest

10 on an epidemiological cutoff value of meropenem (MIC 1 µg/ml) within the redefined CLSI susceptible category (). It has been recently published that the VITEK system tends to produce meropenem MICs significantly lower than the broth microdilucion for the KPCs (). This tendency to give lower meropenem MICs by VITEK in carbapenemase producers could explain why we find that the meropenem epidemiological cutoff value is within the susceptible category. Based on our results, several recommendations can be offered to clinical microbiology laboratories to improve routine detection of carbapenemases by VITEK : i) considering that the AES software failed to infer carbapenemase as the resistance mechanism in several isolates, including those with KPCs, laboratories should avoid using the AES recommendations to detect carbapenemase production. A similar problem of lack of sensitivity of the AES was recently reported (), although in this report, several OXA- isolates (but not KPCs) were miss-detected. The differences observed between both studies (we observed that several class A enzymes, including KPCs, and MBLs, were the most undetected carbapenemases by the AES) could be due to the different carbapenems used in the cards. The discrepancy in the detection between OXA-1 and OXA- could also be due to differences in the hydrolytic profile between these enzymes (). ii) Ertapenem nonsusceptibility (MIC µg/ml as was defined with previous CLSI breakpoints; ) has been proposed as the carbapenem that most accurately detects the presence of KPC by several methods, including VITEK (,, 1). But for those areas or institutions where ertapenem-resistant Enterobacteriaceae by dual mechanisms have become more prevalent, routine laboratories using VITEK commercial system can identify strains with true carbapenemases using the imipenem and meropenem MICs (1). Isolates suspected to produce carbapenemases will have an Downloaded from on August 1, 01 by guest

11 imipenem MIC µg/ml (a non-susceptible result) and at the same time, a meropenem MIC 1 µg/ml (Figure ) by VITEK. Our data indicates that laboratories using this VITEK algorithm will detect >% of Enterobacteriaceae with true carbapenemase production. iii) The shortcomings of using a screening strategy that is different than current CLSI interpretative criteria might be overcome by customizable rules that can be added to the VITEK system by routine laboratories. iv) It is important to mention that the algorithm proposed here is not intended to replace the use of ertapenem in those areas where ertapenem has already demonstrated optimal performance as screening test for KPCs,(, 1). However, key ertapenem issues could arise in the near future for two main reasons: a) The recently approved CLSI breakpoints, which define ertapenem non-susceptibility by a MIC 0. µg/ml (), could be associated with a large number of interfering isolates and may even affect those areas with low prevalence of dual mechanisms; b) In addition, the worldwide emergence of E. coli ST producing bla CTX-M-1 as a major cause of serious multidrug-resistant infections could produce a dramatic epidemiological change (1, 0), as this ESBL has a high potential to contribute to the selection of ertapenem resistant mutants by binding with high affinity to this molecule (, ). Thus, in our opinion, laboratories should be prepared for a new evolutionary scenario with rising levels of resistance to ertapenem due to the mobilization of bla CTX-M-1 by this dominant strain. v) Finally, laboratories that cannot test ertapenem on their current VITEK systems, as it is not available in many cards, may consider using the approach proposed in this study. Downloaded from on August 1, 01 by guest In conclusion, we reviewed the strategies for identification of KPC and other carbapenemases by VITEK for scenarios with high base-line ertapenem resistance. The use of a strategy based on the combined use of imipenem and meropenen MICs

12 (cutoff values of µg/ml and 1 µg/ml, respectively) will enable routine labs to identify, with high confidence levels those isolates suspected of producing carbapenemases. This includes one of the most important epidemiological challenges of recent times, the KPCs. Downloaded from on August 1, 01 by guest 1

13 ACKNOWLEDGEMENTS. We are indebted to the valuable cooperation of WHONET-Argentina Network for their proficiency in detecting bacteria with unusual resistance mechanism and to Dr. R. Melano, Dr. A. Medeiros, Dr. D. Marcano, Dr. L. Martinez-Martinez, Dr. J. Smayevsky and Dr. L. Fernandez Canigia, for providing reference strains for this study. Downloaded from on August 1, 01 by guest 1

14 REFERENCES Anderson, K., D. Lonsway, J. Rasheed, J. Biddle, B. Jensen, L. McDougal, R. Carey, A. Thompson, S. Stocker, B. Limbago, and J. Patel. 00. Evaluation of methods to identify the Klebsiella pneumoniae carbapenemase in Enterobacteriaceae. J. Clin. Microbiol. :-.. Boyd, N. K., R. Mercier, D. Burgess, P. Schreckenberger, K. Singh, and S. Young. 00. Abstr. th Intersci. Conf. Antimicrob. Agents Chemother., abstr. D-1/1.. Bratu, S., M. Mooty, S. Nichani, D. Landman, C. Gullans, B. Pettinato, U. Karumudi, P. Tolaney, and J. Quale. 00. Emergence of KPC-possessing Klebsiella pneumoniae in Brooklyn, New York: epidemiology and recommendations for detection. Antimicrob. Agents Chemother. : Bulik, C., K. Fauntleroy, S. Jenkins, M. Abuali, V. Labombardi, D. Nicolau, and J. Kuti. 0. Comparison of meropenem MICs and susceptibility for carbapenemase producing Klebsiella pneumoniae by various testing methods. J. Clin. Microbiol. doi:./jcm Bush, K., and G. Jacoby. 0. Updated functional classification of betalactamases. Antimicrob. Agents Chemother. :-.. Carmeli,Y., M. Akova, G. Cornaglia, G. Daikos, J. Garau, S. Harbarth, G. Rossolini, M. Souli, and H. Giamarellou. 0. Controlling the spread of carbapenemase-producing Gram-negatives: therapeutic approach and infection control. Clin. Microbiol. Infect. 1:-. Downloaded from on August 1, 01 by guest 1

15 Clinical and Laboratory Standards Institute. 0. Performance Standards for Antimicrobial Susceptibility Testing: 0th Informational Supplement, CLSI publication M0-S0. Clinical and Laboratory Standard Institute, Wayne, PA.. Clinical Laboratory and Standards Institute. 0. Performance Standards for Antimicrobial Susceptibility Testing: Twentieth Informational Supplement (June 0 Update), CLSI publication M0-S0-U. Clinical and Laboratory Standard Institute, Wayne, PA. Doumith, M., M. J. Ellington, D. M. Livermore, and N. Woodford. 00. Molecular mechanisms disrupting porin expression in ertapenem-resistant Klebsiella and Enterobacter spp. clinical isolates from the UK. J. Antimicrob. Chemother. :.. Girlich, D., L. Poirel, and P. Nordmann. 00. Do CTX-M beta-lactamases hydrolyse ertapenem? J. Antimicrob. Chemother. :-.. Girlich, D., L. Poirel, and P. Nordmann. 00. CTX-M expression and selection of ertapenem resistance in Klebsiella pneumoniae and Escherichia coli. Antimicrob. Agents Chemother. :-. 1. Giske, C., L. Gezelius, Ø. Samuelsen, M. Warner, A. Sundsfjord, and N. Woodford. 0. A sensitive and specific phenotypic assay for detection of metallo-β-lactamases and KPC in Klebsiella pneumoniae with the use of meropenem disks supplemented with aminophenylboronic acid, dipicolinic acid and cloxacillin. Clin. Microbiol. Infect. doi:./j Downloaded from on August 1, 01 by guest 1. Gomez, S., D. Faccone, M. Rapoport, R. Melano, M. Galas, Whonet Argentina Network, A. Corso, F. Pasteran, and A. Petroni. 0. Abstr. 0th Intersci. Conf. Antimicrob. Agents Chemother., abstr. C-. 1

16 Johnson, J.R., B. Johnston, C. Clabots, M. Kuskowski, and M. Castanheira. 0. Escherichia coli sequence type ST as the major cause of serious multidrug-resistant E. coli infections in the United States. Clin. Infect. Dis. 1:-. 1. Lartigue, M.F., L. Poirel, C. Poyart, H. Réglier-Poupet, and P. Nordmann. 00. Ertapenem resistance of Escherichia coli. Emerg. Infect. Dis. 1: Leavitt, A., I. Chmelnitsky, R. Colodner, I. Ofek, Y. Carmeli, and S. Navon- Venezia. 00. Ertapenem resistance among extended-spectrum-beta-lactamaseproducing Klebsiella pneumoniae isolates. J. Clin. Microbiol. :-. 1. McGettigan, S., K. Andreacchio, and P. Edelstein. 00. Specificity of ertapenem susceptibility screening for detection of Klebsiella pneumoniae carbapenemases. J. Clin Microbiol. :-. 1. Miriagou, V., G. Cornaglia, M. Edelstein, I. Galani, C. G. Giske, M. Gniadkowski, E. Malamou-Lada, L. Martinez-Martinez, F. Navarro, P. Nordmann, L. Peixe, S. Pournaras, G. M. Rossolini, A. Tsakris, A. Vatopoulos, and R. Cantón. 0. Acquired carbapenemases in Gram-negative bacterial pathogens: detection and surveillance issues. Clin. Microbiol. Infect. 1:-. 1. Nordmann, P., G. Cuzon, and T. Naas. 00. The real threat of Klebsiella pneumoniae carbapenemase-producing bacteria. Lancet Infect. Dis. :-. 0. Oteo, J., M. Pérez-Vázquez, and J. Campos. 0. Extended-spectrum β- lactamase producing Escherichia coli: changing epidemiology and clinical impact. Curr. Opin. Infect. Dis. :0-. Downloaded from on August 1, 01 by guest 1

17 Pasteran, F., T. Mendez, L. Guerriero, M. Rapoport, and A. Corso. 00. Sensitive screening tests for suspected class A carbapenemase production in species of Enterobacteriaceae. J. Clin. Microbiol. :-.. Pasteran, F., T. Mendez, L. Guerriero, M. Rapoport, and A. Corso. 0. Controlling false-positive results obtained with the Hodge and Masuda assays for detection of class A carbapenemase in species of enterobacteriaceae by incorporating boronic acid. J. Clin. Microbiol. :1-.. Pitout, J.D. 0. Infections with extended-spectrum beta-lactamase-producing enterobacteriaceae: changing epidemiology and drug treatment choices. Drugs. 0:1-.. Poirel, L., M. Castanheira, A. Carrer, R. N. Jones, C. A. Parada Rodriguez, J. Smayevsky, and P. Nordmann. 0. Abstr. 0th Intersci. Conf. Antimicrob. Agents Chemother., abstr. C1-.. Tenover, F., R. Kalsi, P. Williams, R. Carey, S. Stocker, D. Lonsway, J. Rasheed, J. Biddle, J. McGowan, and B. Hanna. 00. Carbapenem resistance in Klebsiella pneumoniae not detected by automated susceptibility testing. Emerg. Infect. Dis. 1:-1.. Walther-Rasmussen, J., and N. Høiby. 00. OXA-type carbapenemases. J Antimicrob. Chemother. :-.. Wang, X.D., J. Cai, H. Zhou, R. Zhang, G. Chen. 00. Reduced susceptibility to carbapenems in Klebsiella pneumoniae clinical isolates associated with plasmid-mediated beta-lactamase production and OmpK porin deficiency. J Med Microbiol. :-0. Downloaded from on August 1, 01 by guest 1

18 . Woodford, N., A. Eastaway, M. Ford, A. Leanord, C. Keane, R. Quayle, J. Steer, J. Zhang and D. Livermore. 0. Comparison of BD Phoenix, Vitek, and MicroScan Automated Systems for Detection and Inference of Mechanisms Responsible for Carbapenem Resistance in Enterobacteriaceae. Microbiol. :-00. J. Clin.. Woodford, N., J. W. T. Dallow, R. L. R. Hill, M. F. Palepou, R. Pike, M. E. Ward, M. Warner, and D. M. Livermore. 00. Ertapenem resistance among Klebsiella and Enterobacter submitted in the UK to a reference laboratory. Int. J. Antimicrob. Agents :. Downloaded from on August 1, 01 by guest 1

19 FIGURES LEGENDS Figure Scattergrams showing number of isolates at the indicated carbapenem MICs (average of duplicates values, in µg/ml, of imipenem-ipm- and meropenem-mem-) obtained by VITEK for carbapenemase producers and carbapenemase nonproducers. The sensitivity (SN), specificity (SP), positive predictive value (PPV) and negative predictive value (NPV) were calculated for carbapenemase screening using: a) the imipenem M0 S0-U CLSI non-susceptible breakpoint (a MIC µg/ml). (Members of the Proteae tribe were excluded from the carbapenemase nonproducers group for the calculation of SP, PPV and NPV); b) the meropenem M0 S0-U CLSI nonsusceptible breakpoint (a MIC µg/ml, dash lines); and c) the cutoff values recommended in this work, a combination of imipenem MIC µg/ml and a meropenem MIC 1 µg/ml. The mechanisms of resistance were indicated for undetected carbapenemases or isolates with interfering mechanisms (Prot: Proteae; CTX-M: production of CTX-M- ESBL plus porin loss; AmpC: derrepressed expression of chromosomal cephalosporinase plus porin loss) Downloaded from on August 1, 01 by guest 1

20 FIGURE 1 Downloaded from on August 1, 01 by guest 0

21 Figure. Proposed flowchart for screening suspected carbapenemase production in Enterobacteriaceae with VITEK. IPM, imipenem; MEM, meropenem; R, resistant; I, intermediate; S, susceptible (as defined by CLSI M0 S-0-U, reference ); APB, aminophenylboronic acid; DPA, dipicolinic acid. a Phenotypic confirmation by more specific methods such as the double modified Hodge test (as indicated in reference ), or by the use of meropenem disks supplemented with APB, DPA and cloxacillin (as indicated in reference 1) is recommended for these isolates. Downloaded from on August 1, 01 by guest 1

22 FIGURE Downloaded from on August 1, 01 by guest