Sir,

Acinetobacter baumannii is recognized as one of the most problematic nosocomial pathogens, whose infections are often associated with epidemic spread and outbreaks of multidrug-resistant strains. The identification of resistant clones with epidemic potential is crucial to understanding the epidemiology of this microorganism, and to enable the review and implementation of adequate infection control guidelines. In Portugal, a sublineage of A. baumannii European clone II was identified by PFGE and amplified fragment length polymorphism in different Portuguese hospitals. The sublineage was responsible for the spread of the multidrug-resistant phenotype associated with this strain, including resistance to imipenem.1 Despite the good discrimination obtained by these typing methods, they are laborious and fingerprinting results are not always transferable between laboratories, as protocols and thresholds may differ. The development of multilocus sequence typing (MLST) using housekeeping genes for the characterization of A. baumannii isolates allows the comparison of results between laboratories, contributing to a global epidemiological understanding of the relationship between A. baumannii phenotypes and genotypes.2,3

The main objective of this study was, therefore, to identify the sequence types (STs) of carbapenem-resistant isolates of A. baumannii collected in Portugal between 1998 and 2009. Our carbapenem-resistant A. baumannii collection (540 isolates) showed three antimicrobial susceptibility patterns: A, resistant to all β-lactams, including carbapenems, but susceptible to quinolones and aminoglycosides (collected from 1998 to 1999); B, resistant to all antibiotics, except tobramycin and/or amikacin for some isolates and colistin (collected from 1999 to 2009); and C, resistant to β-lactam antibiotics and susceptible to gentamicin, netilmicin, tobramycin, amikacin and colistin (collected from 2004 to 2009). Fifteen isolates were selected from these groups according to the antimicrobial susceptibility profile, the date of isolation, the geographical location of the hospital and the genotype profile (Table 1). Antimicrobial susceptibility was confirmed using the disc diffusion method for β-lactams, quinolones and aminoglycosides. MICs of ceftazidime and imipenem were determined by Etest. MLST was performed using two schemes,2,3 and sequences for each gene were compared with those available at Acinetobacter MLST websites http://pubmlst.org/abaumannii/ and http://www.pasteur.fr/recherche/genopole/PF8/mlst/Abaumannii.html.2,3

Table 1.

Data for the carbapenem-resistant A. baumannii representative isolates collected from 1998 to 2009 in several Portuguese hospitals and results obtained by MLST

Susceptibility profile
Institute Pasteur MLST scheme3
Bartual et al.2 MLST scheme
Month/yearWardSampleHospitalIsolateAMXTICCAZFOXATMIPMGENNETTOBAMKCIPCSTPFGEacpn60fusAgltApyrGrecArplBrpoBSTgltAgyrBgdhBrecAcpn60gpirpoDST
122HUC3/98medical uniturineHUCRRRRRRSSSSSSA1122344321361436120
65HUC9/98neurotraumatologyurineHUCRRRRRRSSSSSSA1122344321361436120
140HUC1/99ICUbloodHUCRRRRRRSSSSSSA1122344321361436120
142HUC1/99burn unitwoundHUCRRRRRRRRSIRSB4222222221123223398
11HSM2/99orthopaedic unitbloodHSMRRRRRRRRRRRSB1222222221123223398
159HCR11/99medical uniturineHCRRRRRRRRRRRRSB1222222221123223398
4HSB8/00ICUpulmonary secretionHSBRRRRRRRRRRRSB1222222221123223398
189HUC10/00ICUpulmonary secretionHUCRRRRRRRRSIRSB1222222221123223398
40HSFX2/02ICUpusHSFXRRRRRRRRRRRSB2222222221123223398
130HSM6/02ICUpulmonary secretionHSMRRRRRRRRRRRSB3222222221123223398
2396HP2/04medical unitpulmonary secretionHPRRRRRRRRSIRSB1222222221123223398
A-2Ev1/09ICUpulmonary secretionHESRRRRRRRRSIRSB1222222221123223398
2HSAC4/04ICUpulmonary secretionHSACRRRRRRSSSSSSB5222222221123223398
437HUC12/07ICUbloodHUCRRRRRRSSSSSSB6222222221123223398
N-1Ev2/09ICUurineHESRRRRRRSSSSSSB5222222221123223398
Susceptibility profile
Institute Pasteur MLST scheme3
Bartual et al.2 MLST scheme
Month/yearWardSampleHospitalIsolateAMXTICCAZFOXATMIPMGENNETTOBAMKCIPCSTPFGEacpn60fusAgltApyrGrecArplBrpoBSTgltAgyrBgdhBrecAcpn60gpirpoDST
122HUC3/98medical uniturineHUCRRRRRRSSSSSSA1122344321361436120
65HUC9/98neurotraumatologyurineHUCRRRRRRSSSSSSA1122344321361436120
140HUC1/99ICUbloodHUCRRRRRRSSSSSSA1122344321361436120
142HUC1/99burn unitwoundHUCRRRRRRRRSIRSB4222222221123223398
11HSM2/99orthopaedic unitbloodHSMRRRRRRRRRRRSB1222222221123223398
159HCR11/99medical uniturineHCRRRRRRRRRRRRSB1222222221123223398
4HSB8/00ICUpulmonary secretionHSBRRRRRRRRRRRSB1222222221123223398
189HUC10/00ICUpulmonary secretionHUCRRRRRRRRSIRSB1222222221123223398
40HSFX2/02ICUpusHSFXRRRRRRRRRRRSB2222222221123223398
130HSM6/02ICUpulmonary secretionHSMRRRRRRRRRRRSB3222222221123223398
2396HP2/04medical unitpulmonary secretionHPRRRRRRRRSIRSB1222222221123223398
A-2Ev1/09ICUpulmonary secretionHESRRRRRRRRSIRSB1222222221123223398
2HSAC4/04ICUpulmonary secretionHSACRRRRRRSSSSSSB5222222221123223398
437HUC12/07ICUbloodHUCRRRRRRSSSSSSB6222222221123223398
N-1Ev2/09ICUurineHESRRRRRRSSSSSSB5222222221123223398

ICU, intensive care unit; HUC, Hospitals of the University of Coimbra; HSFX, Hospital S. Francisco Xavier, Lisboa; HSM, University Hospital of Santa Maria, Lisboa; HSB, Hospital S. Bernardo, Setúbal; HCR, Hospital das Caldas da Rainha; HP, Hospital de Pombal; HSAC, Hospital S. António dos Capuchos, Lisboa; HES, Hospital Espírito Santo, Évora; AMX, amoxicillin; TIC, ticarcillin; CAZ, ceftazidime; FOX, cefoxitin; ATM, aztreonam; IPM, imipenem; GEN, gentamicin; NET, netilmicin; TOB, tobramycin; AMK, amikacin; CIP, ciprofloxacin; CST, colistin; R, resistant; I, intermediate; S, susceptible.

aDNA fingerprinting of profiles B2 to B6 differ from B1 by one to three bands.

Table 1.

Data for the carbapenem-resistant A. baumannii representative isolates collected from 1998 to 2009 in several Portuguese hospitals and results obtained by MLST

Susceptibility profile
Institute Pasteur MLST scheme3
Bartual et al.2 MLST scheme
Month/yearWardSampleHospitalIsolateAMXTICCAZFOXATMIPMGENNETTOBAMKCIPCSTPFGEacpn60fusAgltApyrGrecArplBrpoBSTgltAgyrBgdhBrecAcpn60gpirpoDST
122HUC3/98medical uniturineHUCRRRRRRSSSSSSA1122344321361436120
65HUC9/98neurotraumatologyurineHUCRRRRRRSSSSSSA1122344321361436120
140HUC1/99ICUbloodHUCRRRRRRSSSSSSA1122344321361436120
142HUC1/99burn unitwoundHUCRRRRRRRRSIRSB4222222221123223398
11HSM2/99orthopaedic unitbloodHSMRRRRRRRRRRRSB1222222221123223398
159HCR11/99medical uniturineHCRRRRRRRRRRRRSB1222222221123223398
4HSB8/00ICUpulmonary secretionHSBRRRRRRRRRRRSB1222222221123223398
189HUC10/00ICUpulmonary secretionHUCRRRRRRRRSIRSB1222222221123223398
40HSFX2/02ICUpusHSFXRRRRRRRRRRRSB2222222221123223398
130HSM6/02ICUpulmonary secretionHSMRRRRRRRRRRRSB3222222221123223398
2396HP2/04medical unitpulmonary secretionHPRRRRRRRRSIRSB1222222221123223398
A-2Ev1/09ICUpulmonary secretionHESRRRRRRRRSIRSB1222222221123223398
2HSAC4/04ICUpulmonary secretionHSACRRRRRRSSSSSSB5222222221123223398
437HUC12/07ICUbloodHUCRRRRRRSSSSSSB6222222221123223398
N-1Ev2/09ICUurineHESRRRRRRSSSSSSB5222222221123223398
Susceptibility profile
Institute Pasteur MLST scheme3
Bartual et al.2 MLST scheme
Month/yearWardSampleHospitalIsolateAMXTICCAZFOXATMIPMGENNETTOBAMKCIPCSTPFGEacpn60fusAgltApyrGrecArplBrpoBSTgltAgyrBgdhBrecAcpn60gpirpoDST
122HUC3/98medical uniturineHUCRRRRRRSSSSSSA1122344321361436120
65HUC9/98neurotraumatologyurineHUCRRRRRRSSSSSSA1122344321361436120
140HUC1/99ICUbloodHUCRRRRRRSSSSSSA1122344321361436120
142HUC1/99burn unitwoundHUCRRRRRRRRSIRSB4222222221123223398
11HSM2/99orthopaedic unitbloodHSMRRRRRRRRRRRSB1222222221123223398
159HCR11/99medical uniturineHCRRRRRRRRRRRRSB1222222221123223398
4HSB8/00ICUpulmonary secretionHSBRRRRRRRRRRRSB1222222221123223398
189HUC10/00ICUpulmonary secretionHUCRRRRRRRRSIRSB1222222221123223398
40HSFX2/02ICUpusHSFXRRRRRRRRRRRSB2222222221123223398
130HSM6/02ICUpulmonary secretionHSMRRRRRRRRRRRSB3222222221123223398
2396HP2/04medical unitpulmonary secretionHPRRRRRRRRSIRSB1222222221123223398
A-2Ev1/09ICUpulmonary secretionHESRRRRRRRRSIRSB1222222221123223398
2HSAC4/04ICUpulmonary secretionHSACRRRRRRSSSSSSB5222222221123223398
437HUC12/07ICUbloodHUCRRRRRRSSSSSSB6222222221123223398
N-1Ev2/09ICUurineHESRRRRRRSSSSSSB5222222221123223398

ICU, intensive care unit; HUC, Hospitals of the University of Coimbra; HSFX, Hospital S. Francisco Xavier, Lisboa; HSM, University Hospital of Santa Maria, Lisboa; HSB, Hospital S. Bernardo, Setúbal; HCR, Hospital das Caldas da Rainha; HP, Hospital de Pombal; HSAC, Hospital S. António dos Capuchos, Lisboa; HES, Hospital Espírito Santo, Évora; AMX, amoxicillin; TIC, ticarcillin; CAZ, ceftazidime; FOX, cefoxitin; ATM, aztreonam; IPM, imipenem; GEN, gentamicin; NET, netilmicin; TOB, tobramycin; AMK, amikacin; CIP, ciprofloxacin; CST, colistin; R, resistant; I, intermediate; S, susceptible.

aDNA fingerprinting of profiles B2 to B6 differ from B1 by one to three bands.

The three isolates of antimicrobial susceptibility group A, producing IMP-5 metallo-β-lactamase,4 were recovered from only one hospital and were not detected in further years. Nine isolates were selected from group B and three from group C according to the date of isolation, hospital, ward and type of sample. Isolates of groups B and C, producing OXA-40 enzyme, were recovered from seven hospitals located in the centre of Portugal and Lisbon region (HUC, HSM, HCR, HSB, HSFX, HP and HSAC) and, recently (2008–09), from a hospital located in the south of Portugal (HES). All isolates showed MICs of imipenem and ceftazidime of >32 and >256 mg/L, respectively.

According to the http://pubmlst.org/abaumannii/ database,2 isolates of groups B and C (the sublineage of European clone II)1 are ST98 (gltA-1, gyrB-12, gdhB-3, recA-2, cpn60-2, gpi-3 and rpoD-3), described for the first time in Portugal. ST98 shared six alleles with ST4, ST6, ST27, ST76, ST118 and ST130 genotypes, isolates recovered between 1982 and 2009 in Italy, Spain, Germany, Japan and Australia. The IMP-5 producer revealed a new ST, deposited in this database as ST120 (gltA-1, gyrB-3, gdhB-6, recA-1, cpn60-4, gpi-3 and rpoD-6). ST120 was not clonally related to ST98, but was clonally related to ST40, a Danish isolate from 1990, sharing five alleles.

According to the Institute Pasteur database,3 the isolates of groups B and C showed the same allelic profile (cpn60-2, fusA-2, gltA-2, pyrG-2, recA-2, rpIB-2 and rpoB-2). This allelic profile corresponds to European clone II, recommended to be designated by ST2 or CC2 (where CC stands for clonal complex) for uniform nomenclature, and is widely disseminated in the world.3 The IMP-5 producer was shown to be ST32 (cpn60-1, fusA-1, gltA-2, pyrG-2, recA-3, rpIB-4 and rpoB-4), corresponding to CC32 isolated in Denmark in 1990.3 The closest matches were ST28 and ST53 collected in Sweden in 1980–81, with only one different allele (recA and rpoB, respectively).3

Recently, it has been suggested that the gyrB and gpi loci, genes used in the Bartual et al.2 scheme, may be affected by horizontal gene transfer, and are thus not good candidates for phylogeny studies.5 In this work, the two databases showed identical results.

The successful spread of ST2 (CC2)/ST98 may be related to the ability to develop multidrug resistance. Differences in resistance profiles may be due to local antibiotic selective pressures. Furthermore, the high dissemination of CC2 might be related to virulence factors. Recently, the presence of putative alien islands (pAICU20 and pAICU32) that may contribute to the virulence of this clone was shown.6 In contrast, the spread of the ST32 (CC32)/ST120 clone is limited. We describe for the first time that the IMP-5 producer strain belongs to ST32, according to the Institute Pasteur database, and apparently does not have epidemic potential.

Overall, not all of the carbapenem-resistant A. baumannii strains have the same ability to spread. However, the surveillance of carbapenem-resistant A. baumannii strains and better knowledge of well-identified clones with defined characteristics are crucial to enable the tracking of potential epidemic strains and to improve infection control measures, thus preventing further dissemination.

Funding

This work was supported financially by iMed.UL, University of Lisbon and by the Center of Pharmaceutical Sciences from Fundação para a Ciência e a Tecnologia, Lisbon, Portugal. N. M. is supported by grant SFRH/BPD/45815/2008 from Fundação para a Ciência e a Tecnologia, Lisbon, Portugal.

Transparency declarations

None to declare.

Acknowledgements

Part of this work was presented at the Twentieth European Congress of Clinical Microbiology and Infectious Diseases, Vienna, Austria, 2010 (Poster 1299).

References

1
Da Silva
G
Dijkshoorn
L
van der Reijden
T
, et al. 
Identification of widespread, closely related Acinetobacter baumannii isolates in Portugal as a subgroup of European clone II
Clin Microbiol Infect
2007
, vol. 
13
 (pg. 
190
-
5
)
2
Bartual
SG
Seifert
H
Hippler
C
, et al. 
Development of a multilocus sequence typing scheme for characterization of clinical isolates of Acinetobacter baumannii
J Clin Microbiol
2005
, vol. 
43
 (pg. 
4382
-
90
)
3
Diancourt
L
Passet
V
Nemec
A
, et al. 
The population structure of Acinetobacter baumannii: expanding multiresistant clones from an ancestral susceptible genetic pool
PLoS One
2010
, vol. 
5
 pg. 
e10034
 
4
Da Silva
G
Correia
M
Vital
C
, et al. 
Molecular characterization of blaIMP-5, a new integron-borne metallo-β-lactamase gene from an Acinetobacter baumannii nosocomial isolate in Portugal
FEMS Microbiol Lett
2002
, vol. 
215
 (pg. 
33
-
9
)
5
Hamouda
A
Evans
BA
Towner
KJ
, et al. 
Characterization of epidemiologically unrelated Acinetobacter baumannii isolates from four continents by use of multilocus sequence typing, pulsed-filed gel electrophoresis, and sequence-based typing of blaOXA-51-like genes
J Clin Microbiol
2010
, vol. 
48
 (pg. 
2476
-
83
)
6
Iacono
M
Villa
L
Fortini
D
, et al. 
Whole-genome pyrosequencing of an epidemic multidrug-resistant Acinetobacter baumannii strain belonging to the European clone II group
Antimicrob Agents Chemother
2008
, vol. 
52
 (pg. 
2616
-
25
)