MDR-2013-0029-ver9-Pereira_1P.3d
06/01/13
4:37pm
Page 1 MDR-2013-0029-ver9-Pereira_1P Type: research-article
EPIDEMIOLOGY
MICROBIAL DRUG RESISTANCE Volume 00, Number 0, 2013 ª Mary Ann Liebert, Inc. DOI: 10.1089/mdr.2013.0029
Prevalence and Molecular Epidemiology of Imipenem-Resistant Pseudomonas aeruginosa Carrying Metallo-Beta-Lactamases from Two Central Hospitals in Portugal
AU1 c
So´nia Gonc¸alves Pereira,1 Teresa Reis,1 Irene Perez Mendez,2 and Olga Cardoso1,*
AU2 c
AU3 c Metallo-beta-lactamases (MBLs) can confer broad-spectrum beta-lactam resistance, including carbapenems. The aim of this work was to document the occurrence of MBLs in 122 imipenem-resistant Pseudomonas aeruginosa isolates collected in two Portuguese central hospitals, to determine their antimicrobial susceptibility, and to observe if there were intra- and interhospital epidemic spread. About 20.5% of these isolates presented blaVIM2, which was found to be widespread in both hospitals. Clonal diversity was observed within hospitals, and no interhospital spread was observed. Ten of the blaVIM-2-positive isolates (44%), from both hospitals, presented one or two class 1 integrons. Two of those contained a VIM-2 gene, one from each hospital, which is indicative for the possibility of MBL gene transfer. No interhospital spread of integrons was observed. Regular screening and surveillance is needed to prevent spread of this worrisome resistance determinant.
tral Portugal, to understand if intra- or interhospital clonal spread has occurred and to optimize the therapeutics in patients carrying P. aeruginosa MBL producers.
Introduction
C
arbapenems, a class of beta-lactams, are the most potent antimicrobial agents for the treatment of Pseudomonas aeruginosa infections.4 Since their introduction in the clinical use, strains expressing metallo-beta-lactamases ( MBLs) have been reported from around the world and in the last decade have emerged particularly in P. aeruginosa. This pathogen is one of the most prevalent nosocomial infection agents, mainly associated to septicemias and respiratory infections in immunocompromised patients, and is sometimes responsible for hospital outbreaks associated with high mortality.10 MBLs hydrolyze all beta-lactams, but monobactams.4 Several MBLs have been reported in P. aeruginosa, including the IMP and VIM families, SPM-1, GIM-1, SIM-1, AIM-1, KHM-1, NDM-1, and DIM-1, where in particular, VIM-2 has emerged as a dominant MBL variant worldwide.4 A first description of VIM-2 (2002) and further studies dealing with its spread have demonstrated that VIM-2 is the only MBL in Portugal.8,11,12 VIM genes (blaVIM) are often encoded on mobile gene cassettes inserted in class 1 integrons, promoting their spread.2,4 The aim of this work was to document the occurrence of MBLs, as well as integrons, in imipenem (IP)-resistant P. aeruginosa isolates (IRPA) from two central hospitals in cen-
Materials and Methods Hospital setting H1 and H2 are 600 and 1,100 beds tertiary care hospitals serving, respectively, the Southern and Northern populations of Coimbra city in central Portugal. H1 is a cluster formed by a main central building hospital (CH), a pediatric hospital (PH), and a maternity hospital (MH) and also receives samples from Hospital de Pombal (HP) located 30 km far. H2 is a medical school hospital composed by a main building and a MH. Bacterial population A total of 122 clinical isolates of IRPA were obtained from clinical specimens from different wards of both hospitals: 91 nonduplicate isolates were collected from H1, during a 1year period, and 31 from H2, during 6 months, in 2008. Bacterial identification and antimicrobial susceptibility was performed by MicroScan WalkAway system (DadeBhering) in H1 and Vitek 2 (BioMerieux) in H2.
1
Faculty of Pharmacy, Center for Pharmaceutical Studies, University of Coimbra, Coimbra, Portugal. Faculty of Pharmacy, Complutense University, Madrid, Spain.
2
1
MDR-2013-0029-ver9-Pereira_1P.3d
06/01/13
4:37pm
Page 2
2 Antimicrobial susceptibility testing, phenotypic screening of MBLs, and determination of minimal inhibitory concentration Susceptibility testing was performed by disk diffusion methodology, according to Clinical and Laboratory Standards Institute (CLSI) guidelines,3 using IP, meropenem (MP), ceftazidime (CAZ), cefepime (CEP), aztreonam (AZT), piperacillin (PIP), amikacin (AMK), and ciprofloxacin (CIP). Results were interpreted according to CLSI directives,3 with intermediate results considered as resistant. Screening of MBLs was evaluated using the combined double-disk test, considering an increase in the inhibition halo of more than 7 mm in the presence of EDTA as a presumptive predictor of a MBL-producing isolate.8 A minimal inhibitory concentration (MIC) of beta-lactams was determined in all MBL-positive isolates by the E-test (BioMe´rieux). Results were interpreted according to CLSI breakpoints.3 Genotypic screening of MBLs and integrons Phenotypic MBL-positive isolates were investigated for the presence of blaIMP, blaVIM, blaSPM, and blaGIM by PCR, as previously described.2 Class 1 integrons were screened by PCR using 5¢-CS (5¢-GGCATCCAAGCAGCAAG-3¢) and 3¢CS (5¢-AAGCAGACTTGACCTGA-3¢) primers, as previously described,2 in all isolates PCR-positive to MBLs. Sequencing of PCR products was performed and results were BLAST analyzed. When a class 1 integron was observed, a second PCR, of the related MBL, was performed in the integron PCR product. All integrons were compared by the megablast alignment tool to study their similarity. Random amplified polymorphic DNA analysis P. aeruginosa isolates with VIM-2 genes were submitted to Random Amplified Polymorphic DNA (RAPD) analysis using primer 272, as previously described.2 RAPD profiling was performed manually by two different observers and each pattern was assigned a letter. Patterns that differed by one or more major bands or by three or more minor bands were considered different, otherwise patterns were considered identical.2
PEREIRA ET AL. tivity (96.7%), but the fluoroquinolone CIP had a diminished activity (25.3%). The combined double-disk test screened for 50 MBLpositive results: 32 from H1 and 18 from H2. An 815 bp amplicon, corresponding to blaVIM, was detected in 25 of those 50 MBL: 12 from H1 and 13 from H2. Obtained amplicons were interpreted by Blast analysis and all regarded a VIM-2-type gene. No amplification was detected with other MBL primers. Overall, 20.5% of the isolates presented VIM-2. MICs of beta-lactams were determined on those isolates showing that 84% were resistant to MP, 72% to CAZ, 28% to b T1 PIP, and 16% to AZT (Table 1). Class 1 integrons were detected in 10 of the 25 VIM-2 isolates (40%): four from H1, where two had 3,000 bp integrons, one isolate presented two integrons, one with 3,000 bp and another with 1,000 bp, and the fourth isolate had a 2,000 bp integron. In H2 isolates, two of the six isolates presented a single integron of 3,000 bp. The other four isolates had two integrons, one with 3,000 bp and the other with 1,000 bp. Only two integrons, with 3,000 bp, contained the VIM-2 gene, one from H1 and another from H2. When their sequences were compared by the megablast alignment tool, we observed that the integrons were different. Other integrons, without VIM-2, from H2 were similar (100% identical) with each other, and contained aminoglycoside resistance genes (aacA4 and aacA7). H1 integrons were different from each other and from those of H2. These integrons also contained aacA4 and aacA7 genes. RAPD analysis of the isolates harboring VIM-2 showed 14 different patterns (table 1). A to G profiles were attributed to isolates from H2 and were spread through eight different wards. Pattern A was represented by six isolates, collected from three wards, and these isolates harbored all the class 1 integrons identified in H2. H to N profiles belonged to H1 and were distributed through six wards from CH, PH, and HP. Five of the 12 H1 VIM-2 producing isolates were collected from two wards of PH and presented the same profile H. Three of the four integrons observed in H1, belonged to this profile. L, M, and N profiles were obtained from different urine isolates of the CH Internal Medicine ward, showing the great variability between strains in the same ward. The M isolate contained the other integron detected in H1.
Results The 91 P. aeruginosa isolates collected from H1 were distributed in HP, PH, and in eight different wards from CH: HP (5.8%), PH (11.3%), internal medicine (16.8%), surgery (14.6%), neurology (13.5%), pneumology (12.5%), neurosurgery (7.9%), infectious diseases (6.9%), emergency room (5.9%), and intensive care unit (ICU) (4.8%). In H2, the 31 isolates were collected from 9 wards: surgery (19.3%), liver transplant unit (16.1%), internal medicine (9.6%), neurotraumatology (12.9%), ICU (12.9%), orthopedic (9.7%), cardiology (6.5%), infectious diseases (6.5%), and emergency room (6.5%). Regarding clinical specimens, 44.3% isolates were from bronchial secretions, 27% from urine, 17.2% from wounds, 4.1% from blood, and 7.4% from other origins. Susceptibility to beta-lactams of all IRPA was 79.1% for CAZ, 76.9% for CEP, 70.3% for PIP, 34.1% for AZT, and 23.1% for MP. The aminoglycoside AMK showed good ac-
Discussion P. aeruginosa has clearly emerged as a leading nosocomial pathogen, because of its ubiquitous nature and its ability to acquire resistance to several antimicrobial agents under drug exposure. Carbapenems, mainly IP and MP, are good agents to treat infections caused by multidrug resistance (MDR) P. aeruginosa, however, the prevalence of carbapenem-resistant P. aeruginosa has been increasing over the years due to the loss of OprD porin, AmpC overexpression, efflux pumps, and also by MBL production.18 In the present study, the overall MBL prevalence in IRPA was relatively higher (20.5%) than in those reported in Italy (12.7%), Germany (11.7%), and Spain (6.8%),13,14,16 which could be associated to isolates coming from patients with invasive devices, complicated infections, and long-term hospitalization. A recent study from Netherlands stated higher values of VIM-2 producers (33%),17 and an older
MDR-2013-0029-ver9-Pereira_1P.3d
VIM-2 IN P.
AERUGINOSA
06/01/13
4:37pm
Page 3
FROM TWO PORTUGUESE HOSPITALS
3
Table 1. Minimal Inhibitory Concentrations (MICs) of Beta-Lactams of VIM-2-Producing Pseudomonas aeruginosa with Respective Integrons, Wards, Products, and RAPD Patterns MIC (mg/l) Isolate N 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
IP
MP
AZT
CAZ
PIP
Integron (bp)
> 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32
> 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 > 32 32 > 32 > 32 8 3 1 3 3 32 32 > 32 > 32 12 > 32 > 32
0.35 0.35 0.75 0.047 1 0.75 24 32 > 256 > 256 3 3 2 2 1 0.5 1 1.5 0.75 3 3 8 2 6 8
48 48 256 16 256 48 96 > 256 > 256 > 256 64 24 6 24 8 4 8 3 12 12 3 2 24 256 256
16 16 32 2 24 96 > 256 > 256 6 6 > 256 64 8 12 12 12 48 16 12 8 8 32 256 256 > 256
3000 3000/1000 3000/1000 3000/1000 3000/1000 3000
3000/1000 3000 3000
2000
Ward/ Hospital
Sample product
RAPD pattern
NT/2 NT/2 S/2 S/2 NT/2 LTU/2 LTU/2 LTU/2 M/2 C/2 ER/2 ID/2 O/2 PH/1 PH/1 PH/1 PH/1 PH/1 P/1 NS/1 HP/1 ID/1 M/1 M/1 M/1
Urine Urine Bronchial secretions Blood Bronchial secretions Catheter Bile Blood Wound Urine Urine Wound Urine Urine Urine Catheter Catheter Wound Bronchial secretions Bronchial secretions Urine Urine Urine Urine Urine
A A A A A A B C D D E F G H H H H H I I J K L M N
RAPD, Random Amplified Polymorphic DNA; IP, Imipenem; MP, meropenem; CAZ, ceftazidime; AZT, aztreonam; PIP, piperacillin; NT, Neurotraumatology ward; S, Surgery ward; M, Internal Medicine ward; C, Cardiology ward; LTU, Liver Transplant unit; ER, Emergency room; ID, Infectious Diseases ward; O, Orthopedic ward; PH, Pediatric Hospital; P, Pneumology ward; NS, Neurosurgery ward; HP, Hospital de Pombal.
work from Greece pointed 62% VIM producers in nine hospitals.6 In our study, VIM-2 IRPA isolates were highly resistant to MP (84%), but AZT and PIP displayed some activity (16% and 28%, respectively). Spanish, Greek, and Italian studies evidenced higher rates of antibiotic resistance, showing a MDR pattern, including CAZ, CEP, PIP/tazobactam, gentamicin, tobramycin, and quinolone resistance.9,13–15 In Europe, blaVIM-type determinants are more prevalent and widespread in P. aeruginosa than blaIMP type. In the present study, molecular analysis only detected blaVIM2, the only MBL observed until now in Portugal, as well as the studies from Germany, Netherlands, and Greece.6,16,17 IMP has only been detected in P. aeruginosa in Italy, France, and Spain.7,13,14 These great differences in MBL prevalence, drug susceptibility patterns, and MBL content between different European countries are probably due to different antibiotherapy policies. The MBL phenotypic screening yielded 50% of falsepositive results since from 50 phenotypic-positive isolates, only 25 demonstrated to harbor a MBL gene, VIM-2. In Valenza’s study, the EDTA disk test was able to discriminate between all MBL-positive and MBL-negative by using a breakpoint > = 14 mm.16 This could be a remark to be applied in future studies to diminish false-positive results. However, PCR failure in enzyme variants or newer MBLs may also have occurred in this study, for which, the observed differences between genotypic and phenotypic MBL detection might be
due, in part, to PCR constraints and not only to EDTA-testing failure. The psresence of class 1 integrons in VIM-2 P. aeruginosa isolates was observed in 40% of all VIM-2 isolates, which suggests that horizontal gene transfer may play a role in the dissemination of MBL producing P. aeruginosa in these settings. Indeed, two integrons contained the MBL gene, one from each hospital. Association of MBL genes with class 1 integrons was also demonstrated in Germany and Italy.14,16 RAPD analysis of VIM-2 isolates showed 14 patterns scattered in both hospitals. Although the number of studied isolates was limited, the results suggest that there is no interhospital dissemination, but intrahospital spread was observed. This clonal diversity was also seen in a French survey, where 137 isolates were clustered in 113 PFGE patterns and in most cases, the clonally related iso- b AU4 lates were recovered from the same hospital.7 A Belgium study, however, documented an interhospital epidemic spread,5 as well as in Italy, Spain, Greece, and Netherlands, where the same cluster was observed in different hospitals wards.9,14,15,17 Pattern A was the most prevalent in the 25 detected VIM-2 P. aeruginosa isolates, represented by six isolates collected from three different wards of H2. The other patterns were from different wards and the majority was represented by a single isolate, suggesting their possible source in the patients
MDR-2013-0029-ver9-Pereira_1P.3d
06/01/13
4:37pm
Page 4
4
PEREIRA ET AL.
themselves and not in the hospital environment.6 The five VIM-2 isolates obtained in PH, from different wards, presented the same profile H indicating the selection of a unique clone in this hospital, as stated in an Italian sudy.14 Three of these isolates presented a 3,000 bp integron, therefore, horizontal gene transfer may be important in this setting and the hospital personnel should be aware to avoid cross contamination.4,14 L, M, and N profiles were obtained from urine isolates from the Internal Medicine ward, reflecting the variability between isolates, and reducing the possibility of cluster dissemination in this ward.6 Emergence of MBLs producing bacilli resistant to carbapenems is an increasing clinical problem. Moreover, they are associated to a MDR phenotype that not only includes carbapenems, but also last generation cephalosporins and ureidopenicilins.4 In this study, we demonstrated that the presence of VIM-2 is a reality in our hospitals, with P. aeruginosa carrying blaVIM-2 typed into 14 different patterns detected in different wards of nearby hospitals, suggesting that the prevalence of carbapenemase-encoding genes was mainly due to gene spread than to clonal dissemination. These results reinforce the need of screening all carbapenem-resistant isolates for MBL production to better address the prevalence and dissemination of these enzymes. We suggest the use of enhanced phenotypic tools, as proposed by Valenza’s work regarding the combined double-disk test interpretation.16 This might be useful to improve antibiotherapy regimens and policies of infection control in Portuguese and other European hospitals, to prevent the wider spread of this worrisome resistance determinant. Acknowledgments The authors wish to thank the CEF-Centre for Pharmaceutical Studies and the Portuguese Foundation for Science and Technology (POCTI-SFA-8-177) for supporting this research. The authors also thank MD Grac¸a Ribeiro, from the Laboratory of Clinical Pathology of Hospitais da Universidade de Coimbra, for providing the P. aeruginosa isolates used in this study, and Trindade Marques and Jorge Marques, from Centro Hospitalar de Coimbra, for assistance with data management. Author Disclosure Statement No competing financial interests exist. References 1. Cardoso, O., R. Leita˜o, A. Figueiredo, J.C. Sousa, A. Duarte, and L. Peixe. 2002. Metallo-beta-lactamase VIM-2 in clinical isolates of Pseudomonas aeruginosa from Portugal. Microb. Drug Resist. 8:93–97. 2. Cardoso, O., A.F. Alves, and R. Leita˜o. 2008. Metallo-betalactamase VIM-2 Pseudomonas aeruginosa isolates from a cystic fibrosis patient. Int. J. Antimicrob. Agents 31:375–379. 3. Clinical and Laboratory Standards Institute. 2011. Performance Standards for Antimicrobial Susceptibility Testing; Nineteenth Informational Supplement. M100-S21, CLSI, Wayne, PA, USA.
4. Cornaglia, G., H. Giamarellou, and G.M. Rossolini. 2011. Metallo-beta-lactamases: a last frontier for beta-lactams? Lancet Infect. Dis. 11:381–393. 5. Deplano, A., H. Rodriguez-Villalobos, Y. Glupczynski, P. Bogaerts, P. Allemeersch, A. Grimmelprez, G. Mascart, L. Berge`s, C. Laurent, B. Byl, and M.J. Struelens. 2007. Emergence and dissemination of multidrug resistant clones of Pseudomonas aeruginosa producing VIM-2 metallo-betalactamase in Belgium. Eurosurveillance 12:18. 6. Giakkoupi, P., G. Petrikkos, L.S. Tsouvelekis, S. Tsonas, The Whonet Greece Study Group, N.J. Legakis, and A.C. Vatapoulos. 2003. Spread of integron-associated VIM-type metallo-beta-lactamase genes among imipenem-nonsusceptible Pseudomonas aeruginosa strains in Greek hospitals. J. Clin. Microbiol. 41:822–825. 7. Hocquet, D., P. Ple´siat, B. Dehecq, P. Mariotte, D. Talon, and X. Bertrand, on behalf of ONERBA, 2010. 2010. Nationwide investigation of extended-spectrum-betalactamases, metallo-beta-lactamases and extended-spectrumoxacillinases produced by ceftazidime-resistant Pseudomonas aeruginosa strains in France. Antimicrob. Agents Chemother. 54:3512–3515. 8. Pena, A., A.M. Donato, A.F. Alves, R. Leita˜o, and O. Cardoso. 2008. Detection of Pseudomonas aeruginosa producing metallo-beta-lactamase VIM-2 in a central hospital from Portugal. Eur. J. Clin. Microbiol. Infect. Dis. 27:1269– 1271. 9. Pen˜a, C., C. Suarez, F. Tubau, O. Gutierrez, A. Domı´nguez, A. Oliver, M. Pujol, F. Gudiol, and J. Ariza. 2007. Nosocomial spread of Pseudomonas aeruginosa producing the metallo-beta-lactamase VIM-2 in a Spanish hospital: clinical and epidemiological implications. Clin. Microbiol. Infect. 13:1026–1029. 10. Queenen, A.M., and K. Bush. 2007. Carbapenemases: the versatile beta-lactamases. Clin. Microbiol. Rev. 20:440– 458. 11. Quinteira, S., H. Ferreira, and L. Peixe. 2005. First isolation of blaVIM-2 in an environmental isolate of Pseudomonas pseudoalcaligenes. Antimicrob. Agents Chemother. 49:2140– 2141. 12. Quinteira, S., and L. Peixe. 2006. Multiniche screening reveals the clinically relevant metallo-beta-lactamase VIM-2 in Pseudomonas aeruginosa far from the hospital setting: an ongoing dispersion process? Appl. Environ. Microbiol. 72:3743–3745. 13. Riera, E., G. Cabot, X. Mulet, M. Garcia-Castillo, R. Campo, C. Juan, R. Canton, and A. Oliver. 2011. Pseudomonas aeruginosa carbapenem resistance mechanisms in Spain: impact on the activity of imipenem, meropenem and doripenem. J. Antimicrob. Chemother. 66:2022–2025. 14. Rossolini, G., F. Luzzaro, R. Migliavacca, C. Mugnaioli, B, Pini, F. Luca, M. Perilli, S. Pollini, M. Spalla, G. Amicosante, A. Toniolo, and L. Pagani. 2008. First countrywide survey of acquired metallo-beta-lactamases in Gram-negative pathogens in Italy. Antimicrob. Agents Chemother. 52:4023– 4029. 15. Tsakris, A., A. Poulou, I. Kristo, T. Pittaras, N. Spanakis, P. Pournaras, and F. Markou. 2009. Large dissemination of VIM-2-metallo-lactamase–producing Pseudomonas aeruginosa strains causing health care-associated community-onset Infections. J. Clin. Microbiol. 47:3524–3529. 16. Valenza, G., B. Joseph, J. Elias, H. Claus, A. Oesterlein, K. Engelhardt, D. Turnwald, M. Frosch, M. Abele-Horn, and C. Schoen. 2010. First survey of metallo-beta-lactamases in clinical
MDR-2013-0029-ver9-Pereira_1P.3d
VIM-2 IN P.
AERUGINOSA
06/01/13
4:37pm
Page 5
FROM TWO PORTUGUESE HOSPITALS
isolates of Pseudomonas aeruginosa in a German university hospital. Antimicrob. Agents Chemother. 54:3493–3497. 17. Van der Bij, A.K., R. Van Mansfeld, G. Peirano, W.H.F. Goessens, J.A. Severin, J.D.D. Pitout, R. Willems, and M. Van Westreenen. 2011. First outbreak of VIM-2 metallobeta-lactamase-producing Pseudomonas aeruginosa in The Netherlands: microbiology, epidemiology and clinical outcomes. Int. J. Antimicrob. Agents 37:513–518. 18. Walsh, T.R., M.A. Toleman, L. Poirel, and P. Nordmann. 2005. Metallo-beta-lactamases: the quiet before the storm? Clin. Microbiol. Rev. 18:306–325.
5 Address correspondence to: Olga Cardoso, PhD Faculty of Pharmacy Center for Pharmaceutical Studies University of Coimbra Azinhaga de Santa Comba 3000-548 Coimbra Portugal E-mail:
[email protected]
MDR-2013-0029-ver9-Pereira_1P.3d
06/01/13
4:37pm
Page 6
AUTHOR QUERY FOR MDR-2013-0029-VER9-PEREIRA_1P AU1: Please note that gene symbols in any article should be formatted per the gene nomenclature. Thus, please make sure that gene symbols, if present in this article, are italicized. AU2: Please review all authors’ surnames for accurate indexing citations. AU3: Abstract has been taken from the metadata file. Please check. AU4: Please expand PFGE.
