Molecular epidemiology of VIM-1 producing Escherichia coli from Germany referred to the National Reference Laboratory

Accepted Manuscript Title: Molecular epidemiology of VIM-1 producing Escherichia coli from Germany referred to the National Reference Laboratory Author: Martin Kaase Niels Pfennigwerth Felix Lange Agnes Anders S¨oren G. Gatermann PII: DOI: Reference:

S1438-4221(15)00095-8 http://dx.doi.org/doi:10.1016/j.ijmm.2015.08.032 IJMM 50994

To appear in:

Please cite this article as: Kaase, M., Pfennigwerth, N., Lange, F., Anders, A., Gatermann, S.G.,Molecular epidemiology of VIM-1 producing Escherichia coli from Germany referred to the National Reference Laboratory, International Journal of Medical Microbiology (2015), http://dx.doi.org/10.1016/j.ijmm.2015.08.032 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ip t

Martin Kaase*[email protected], Niels Pfennigwerth, Felix Lange, Agnes Anders, Sören G. Gatermann

d

M

an

us

cr

Department of Medical Microbiology, National Reference Centre for multidrug-resistant gram-negative bacteria, Ruhr-University Bochum, Universitätsstraße 150, 44801 Bochum, Germany

te

2 3 4 5 6 7 8 9 10 11 12 13 14

Molecular epidemiology of VIM-1 producing Escherichia coli from Germany referred to the National Reference Laboratory

Ac ce p

1

1/1 Page 1 of 22

Abstract

16

The distribution of carbapenemase genes in Escherichia coli strains isolated between

17

September 2009 and May 2013 in Germany was investigated. Out of 192 isolates with

18

carbapenemase production OXA-48 was found in 44.8%, VIM-1 in 18.8%, NDM-1 in

19

11.5% and KPC-2 in 6.8%. Patients with VIM-1 producing E. coli (n = 36) differed from

20

patients with OXA-48 by an older age, less frequent mention of travel history and an

21

increased proportion of clinical over screening specimens. These data might indicate that

22

introduction from abroad is of minor importance for VIM-1 producing E. coli compared to

23

other carbapenemases. Multilocus sequence typing revealed that E. coli with VIM-1 were

24

mostly multiclonal, emphasizing the role of horizontal gene transfer in its spread.

25

Susceptibility testing of VIM-1 producing E. coli demonstrated aztreonam susceptibility in

26

55.6%. Among non-β-lactams susceptibility rates of >90% were observed for amikacin,

27

tigecycline, colistin, fosfomycin and nitrofurantoin.

28

Keywords

29

Enterobacteriaceae, Escherichia coli, carbapenemases, VIM-1, OXA-48, NDM-1, KPC-2

Ac ce p

te

d

M

an

us

cr

ip t

15

2/2 Page 2 of 22

Introduction

31

Carbapenem resistance in Enterobacteriaceae is of utmost clinical importance since

32

currently available treatment options are scarce. In Enterobacteriaceae carbapenem

33

resistance is usually caused either by carbapenemase production or by porin loss

34

combined with expression of ESBL or AmpC-β-lactamases (Doumith et al., 2009).

35

Carbapenemases in Enterobacteriaceae belong to Ambler class A enzymes such as KPC

36

or certain GES variants, Ambler class B enzymes which are synonymous with the term

37

metallo-β-lactamases such as VIM, IMP, NDM or GIM-1 and to class D enzymes also

38

known as oxacillinases such as OXA-48 and its variants (Patel and Bonomo, 2013).

39

Since 2009 the German National Reference Laboratory for multidrug-resistant gram-

40

negative bacteria investigates Enterobacteriaceae, Acinetobacter baumannii and

41

Pseudomonas aeruginosa with elevated MICs for carbapenems for the presence of

42

carbapenemases. Laboratories are encouraged to refer their isolates to the National

43

Reference Laboratory and the service is offered free of charge.

44

Escherichia coli has the potential to spread also outside the hospital environment

45

(Banerjee and Johnson, 2014) and therefore deserves special attention. In addition, a

46

recent report about the finding of VIM-1 producing E. coli isolated on a pig farm in

47

Germany (Fischer et al., 2012) prompted us to perform this study, whose aim is to

48

describe the molecular epidemiology of carbapenemase producing E. coli in Germany with

49

emphasis on isolates with VIM-1 production.

50

Material and methods

51

Strains

52

Laboratories in Germany were asked to send E. coli isolates to the National Reference

53

Laboratory if ertapenem, imipenem or meropenem were categorized as non-susceptible.

Ac ce p

te

d

M

an

us

cr

ip t

30

3/3 Page 3 of 22

Detection of carbapenemases

55

Several phenotypic tests were performed routinely on all isolates as previously described

56

(Kaase et al., 2014): the modified Hodge Test, the combined disk test with boronic acid

57

and the combined disk test with EDTA. In addition, the following PCRs followed by

58

sequencing were performed routinely: blaKPC, blaOXA-48, blaVIM, blaIMP and blaNDM. If the

59

routinely performed PCRs were negative but the phenotypic tests suggested the presence

60

of a carbapenemase further PCRs for rarely occurring carbapenemases such as IMI,

61

NMC-A, SME, GES and GIM-1 were performed. If negative, a microbiological bioassay

62

was performed similar as previously described (Marchiaro et al., 2005).

63

Susceptibility testing

64

Susceptibility testing was performed by microdilution in accordance with EUCAST

65

(www.eucast.org) recommendations using pre-configured microtiter plates (MERLIN

66

Diagnostika GmbH, Bornheim-Hersel, Germany).

67

Multilocus sequence typing (MLST)

68

For multilocus sequence typing (MLST), PCR and subsequent sequence analysis for

69

seven housekeeping genes (adk, fumC, gyrB, icd, mdh, purA, recA) was performed

70

according to Wirth et al. (Wirth et al., 2006) and the MLST web site

71

(http://mlst.warwick.ac.uk/mlst/dbs/Ecoli/). Genes for all seven allels were concatenated for

72

each MLST. An unrooted phylogenetic tree was constructed by DNA distance matrix

73

calculation and the Neighbor-Joining tree drawing method using the Dnadist, Neighbor and

74

Drawtree program of the PHYLIP software package. The strain NRZ-06045 with the allel

75

profile Ø-48-327-374-235-42-37 was excluded from this analysis, since the adk allel could

76

not be amplified.

Ac ce p

te

d

M

an

us

cr

ip t

54

4/4 Page 4 of 22

Statistical analyses

78

All analyses were performed using R (http://www.R-project.org). The Wilcoxon rank-sum

79

test was used for comparing age distributions, the Fisher test was used for comparison of

80

proportions. Statistical significance was established at P value of <0.05. Calculation of the

81

Simpson diversity index was done with the 'vegan' package.

ip t

77

cr

82

us

83

Results

85

During the time period from September 2009 till May 2014 693 E. coli isolates were

86

referred for carbapenemase detection to the National Reference Laboratory for multidrug-

87

resistant gram-negative bacteria. Fourty-one copy-strains were identified and excluded

88

from further analyses. Carbapenemase production was excluded in 460 (70.6%) isolates,

89

whereas in 192 isolates (29.4%) a carbapenemase was detected. All carbapenemase

90

genes were sequenced and fourteen different carbapenemases were found among the

91

strains (Table 1) with OXA-48 being most frequent (n = 86), followed by VIM-1 (n = 36) and

92

NDM-1 (n = 22). The 36 VIM-1 producing E. coli isolates were referred from eleven of the

93

sixteen federal states of Germany, mostly from Berlin (n = 12), Bavaria (n = 5), Rhineland-

94

Palatinate (N = 5) and Northrhine-Westphalia (n = 4).

95

The age of the patients with VIM-1 producing E. coli ranged from 0 to 86 years with a

96

median of 70.5 years (Table 2). The patients with VIM-1 producing E. coli were significantly

97

older than patients with OXA-48 producing E. coli (Wilcoxon test, P = 0.017). Although the

98

median age of patients with NDM-1 producing E. coli was 50 years and thus lower than in

99

those with VIM-1 producing E. coli (Table 2), this difference did not reach statistical

100

Ac ce p

te

d

M

an

84

significance (Wilcoxon test, P = 0.059). No statistically significant age difference between

5/5 Page 5 of 22

patients with VIM-1 producing E. coli was observed compared to patients with KPC-2

102

producing E. coli and E. coli isolates without any carbapenemase.

103

The referring laboratory did not mention any travel history for the patients with VIM-1

104

producing E. coli whereas a previous stay abroad was reported for E. coli with OXA-48 in

105

12.8%, with NDM-1 in 18.2%, with KPC-2 in 7.7% and without carbapenemase production

106

in 3.5%. The difference in reported travel history was statistically significant when

107

comparing VIM-1 and OXA-48 (Fisher test, P = 0.033) as well as VIM-1 and NDM-1

108

(Fisher test, P = 0.017), but not when comparing VIM-1 and KPC-2 (Fisher test, P = 0.265)

109

or isolates for which a carbapenemase had been excluded (Fisher test, P = 0.62).

110

The main sources of VIM-1 producing E. coli comprised urinary (41.7%) and rectal (13.9%)

111

specimens (Table 3). The proportion of urinary specimens was higher in E. coli producing

112

VIM-1 than in those with OXA-48, NDM-1, KPC-2 and without any carbapenemase

113

(Table 3) and this difference was statistically significant when comparing VIM-1 and OXA-

114

48 (Fisher test, P = 0.023). In contrast, the proportion of rectal specimens was lower in

115

E. coli harboring VIM-1 than in those with OXA-48, NDM-1, KPC-2 and without any

116

carbapenemase (Table 3) and again this difference was statistically different for the

117

comparison of VIM-1 and OXA-48 (Fisher test, P = 0.028).

118

MLST revealed 24 different types (Figure 1), 17 of which were unique (47.2%). The

119

Simpson diversity index was 0.94. The unique types comprised ST69, ST88, ST95,

120

ST131, ST224, ST357, ST420, ST453, ST602, ST607, ST728, ST998, ST1266, ST1725,

121

ST4962, ST4963 and an allel profile Ø-48-327-374-235-42-37 which could not be

122

assigned a MLST designation because the adk allel could not be amplified. Among the

123

isolates with the non-unique MLST types ST10, ST12, ST393, ST410, ST624, ST648 and

124

ST1196 (Table 4), occurrence in the same geographical region within a related time period

125

of two months was observed only for five isolates (13.9% of all VIM-1 producing E. coli),

Ac ce p

te

d

M

an

us

cr

ip t

101

6/6 Page 6 of 22

namely two ST624 isolates referred from Rhineland-Palatinate and three ST1196 isolates

127

from Berlin.

128

Susceptibility testing of the VIM-1 producing E. coli isolates revealed resistance to all

129

penicillins and cephalosporins with high MIC50 values (Table 5). However, aztreonam was

130

resistant in only 44.4% of the isolates and presence of an ESBL could be inferred by a

131

positive synergy phenomenon in a double disk approximation test with aztreonam in 14

132

isolates. MIC50 values for imipenem, meropenem and doripenem were considerably lower

133

than for penicillins and cephalosporins and fell between 2 and 8 mg/l. Susceptibility rates

134

according to EUCAST were as high as 61.1% (Table 5). Regarding the non-β-lactams

135

susceptibility rates for aminoglycosides were 52.8% for tobramycin, 77.8% for gentamicin

136

and 100% for amikacin. Tigecycline demonstrated high susceptibility rates of 94.4%,

137

whereas susceptibility rates for sulfamethoxazole-trimethoprim were only 19.4%. Also for

138

fluoroquinolones susceptibility rates reached 37.1% at most; a similar low susceptibility

139

rate of 36.1% was shown for chloramphenicol. In contrast, colistin and fosfomycin were

140

susceptible in all isolates and nitrofurantoin in 97.2% (Table 5).

141

The EDTA combined disk test for meropenem demonstrated an increase of the inhibition

142

zone in the range between 3 mm and 18 mm (median 10 mm), for imipenem in the range

143

between 3 mm and 14 mm (median 9 mm).

Ac ce p

te

d

M

an

us

cr

ip t

126

144

145

Discussion

146

This study confirms that carbapenemase producing E. coli strains have arrived in

147

Germany. Several reports already mentioned E. coli from human sources with production

148

of certain carbapenemases such as NDM-1 (Pfeifer et al., 2011), NDM-7 (Göttig et al.,

149

2013), GIM-1 (Wendel et al., 2013) or OXA-244 in Germany (Valenza et al., 2014), but

7/7 Page 7 of 22

VIM-1 producing E. coli have not been described in Germany so far. Generally, studies on

151

VIM-1 producing E. coli are scarce and to the best of our knowledge the molecular

152

epidemiology of VIM-1 producing E. coli has only been published from Greece (Galani et

153

al., 2010; Scoulica et al., 2004) and Italy (Aschbacher et al., 2013, 2011), single E. coli

154

isolates with VIM-1 have been reported from Spain (Cendejas et al., 2010; Tato et al.,

155

2007) and Canada (Tijet et al., 2013).

156

Carbapenemase producing Enterobacteriaceae are still very rare in Germany. In a point

157

prevalence study performed on 98 intensive care units in the federal state of Saxony only

158

two KPC-2 producing K. pneumoniae were found among 1037 patients screened by rectal

159

swabs (Ehrhard et al., 2014). In the present study only 192 carbapenemase producing

160

E. coli were found among Enterobacteriaceae referred to the National Reference

161

Laboratory for multidrug-resistant gram-negative bacteria on a voluntary basis within a

162

time frame of three years and nine months. It can be assumed that the real number of

163

patients colonized or infected with carbapenemase producing E. coli is considerably larger

164

for several reasons: firstly, the awareness for the importance of carbapenemase detection

165

in hospitals and labs has increased only in recent years which might account for a lack of

166

rectal screening in patients at risk and for inappropriate methods for their detection in

167

microbiological laboratories. Secondly, it is not mandatory to refer strains suspicious for

168

carbapenemase production to the National Reference Laboratory. The yearly number of

169

cases in German hospitals between 2009 and 2013 fell between 17,817,180 and

170

18,787,168 according to the Federal Statistical Office for Germany, Wiesbaden, Germany

171

(www.destatis.de, accessed 10-Oct-2014). Comparing these yearly data to the number of

172

192 carbapenemase producing E. coli found within the study period of three years and

173

nine months might nevertheless give an indication how rare carbapenemase production

174

still is in the species E. coli which can be assumed to be present in the gut flora of almost

Ac ce p

te

d

M

an

us

cr

ip t

150

8/8 Page 8 of 22

every patient. This study focuses on the metallo-β-lactamase VIM-1 which was found in

176

E. coli isolates from 36 patients. Of note, a wide geographical distribution of VIM-1

177

producing E. coli was observed with isolates referred from eleven of the sixteen federal

178

states of Germany. Interestingly the majority of the isolates originated in laboratories from

179

Berlin. Characteristics of patients with VIM-1 producing E. coli differed from patients

180

colonized or infected by E. coli with other resistance determinants in several ways: firstly,

181

the median age of patients was higher for VIM-1 than for NDM-1 and OXA-48, in the latter

182

case this difference reached statistical significance. Secondly, no travel history was

183

provided for any of the patients with VIM-1 in contrast to patients with other

184

carbapenemases and this difference was statistically significant for OXA-48 and NDM-1.

185

Thirdly, the proportion of urinary specimens was higher and the proportion of rectal

186

specimens was lower for VIM-1 compared to OXA-48, NDM-1 and KPC-2. This difference

187

was significant when comparing VIM-1 and OXA-48. It can be assumed that previous

188

hospitalization abroad occurs more frequently in younger patients, therefore the older age

189

of patients with VIM-1 producing E. coli does not point into the direction of acquisition of

190

this carbapenemase abroad. The same holds true for the lack of mention of travel history

191

in patients with E. coli harboring VIM-1. However, it has to be considered that data quality

192

for travel history can be assumed to be low, because laboratories referring the strains only

193

rarely have access to these data. Also the lower proportion of rectal specimens among

194

VIM-1 producing E. coli can be explained by a differing epidemiology of this

195

carbapenemase: rectal swabs are mostly performed if multidrug-resistant

196

Enterobacteriaceae are suspected in a patient and previous hospitalization abroad is one

197

risk factor mentioned in the German guidelines for prevention of the spread of multidrug-

198

resistant gram-negative bacteria (KRINKO, 2012). Thus, the significantly lower proportion

199

might indicate that VIM-1 producing E. coli occur less frequently than other

Ac ce p

te

d

M

an

us

cr

ip t

175

9/9 Page 9 of 22

carbapenemases in the patient population normally assumed to be at higher risk for

201

carbapenemases.

202

Typing by MLST revealed that VIM-1 producing E. coli were mostly multiclonal with a high

203

Simpson diversity index of 0.94. This indicates the importance of horizontal gene transfer,

204

probably by plasmids, in the spread of this resistance determinant. Hints for a clonal

205

spread were only found in five isolates (13.9%) with the same MLST type found in the

206

same geographical region, namely in two ST624 isolates referred from Rhineland-

207

Palatinate and in three ST1196 isolates from Berlin. Of the already assigned MLST types

208

five have been found only in human sources so far, two only in animal sources and the

209

remaining from animal and human sources. Six of the E. coli MLST types were described

210

in Germany for the first time. Apart from three MLST types none of the other types has

211

ever been associated with VIM-1 before: VIM-1 producing E. coli of type ST10 have been

212

found in Spain (Oteo et al., 2013) and of type ST131 in Italy (Aschbacher et al., 2013;

213

Mantengoli et al., 2011) and Canada (Tijet et al., 2013). As E. coli ST131 in association

214

with CTX-M-15 shows a worldwide spread (Banerjee and Johnson, 2014), the observation

215

of this clone together with VIM-1 deserves further attention. Of special interest is the

216

finding of one VIM-1 producing E. coli ST88 in our collection since this type has been also

217

found on a pig farm in Germany in 2011 as described in a recent report (Fischer et al.,

218

2012), which constitutes the first observation ever of carbapenemase producing

219

Enterobacteriaceae in livestock. The VIM-1 producing E. coli ST88 isolate from our study

220

was referred by a laboratory in the federal state of Mecklenburg-Western Pomerania in

221

January 2012. Without further investigations no statement can be made, however, if these

222

two isolates might be directly related.

223

Susceptibility testing demonstrated resistance with high MIC50 values for all

224

cephalosporins and penicillins including mecillinam and temocillin. As metallo-β-

Ac ce p

te

d

M

an

us

cr

ip t

200

10/10 Page 10 of 22

lactamases do not themselves exert any activity against aztreonam this antibiotic can be a

226

treatment option as long as no second resistance mechanism such as an ESBL or an

227

AmpC-β-lactamase is present. In our collection this was the case in 55.6% of the isolates.

228

Despite the carbapenemase activity of VIM-1 MICs for carbapenems were only moderately

229

increased and susceptibility rates according to EUCAST were as high as 61.1% for

230

imipenem. This observation is in accordance with previous observations (Aschbacher et

231

al., 2013, 2011) and emphasizes how challenging the detection of carbapenemases can

232

be in Enterobacteriaceae. For this reason EUCAST has recommended the screening cut-

233

off >0.12 mg/l for the detection of carbapenemases and applying this cut-off all VIM-1

234

producing E. coli in our study would have been found.

235

Regarding treatment alternatives among non-β-lactams good susceptibility rates of more

236

than 90% were only found for amikacin, tigecycline, colistin, fosfomycin and nitrofurantoin.

237

There are some limitations of our study. Firstly, a full coverage of all carbapenemase

238

producing E. coli observed in microbiological laboratories cannot be assumed in our study,

239

since it is voluntary to refer strains to the reference lab. In addition, it might not have been

240

known in all laboratories that the National Reference Laboratory offers the free service of

241

carbapenemase detection at the beginning of the study period. Also the awareness that

242

carbapenemases might also be present in isolates with increased MIC values for

243

carbapenems which are still in the susceptible range has probably been only risen in

244

recent years. Secondly, the number of isolates suspicious for carbapenemase production

245

found in a microbiological lab is dependent on the screening policies for such bacteria

246

which can considerably vary between institutions. Thirdly, only an internal 801-bp amplicon

247

of blaVIM-1 was sequenced. Although this fragmentary sequence showed complete

248

homology to blaVIM-1 and allowed the exclusion of all other hitherto known blaVIM variants, a

249

novel variant characterized by substitutions in the extremes of the gene cannot be ruled

Ac ce p

te

d

M

an

us

cr

ip t

225

11/11 Page 11 of 22

out. A further limitation of our study is the scarcity of patient data available for the isolates.

251

Because the microbiological laboratory only rarely has direct access to patient data it can

252

only occasionally provide data regarding the role of the isolate in infection, the previous

253

use of antibiotics, the travel history or even the outcome of the infection.

254

In conclusion, this is the first nationwide report on isolates of VIM-1 producing E. coli from

255

human specimens in Germany. Patients with VIM-1 producing E. coli differ from patients

256

with OXA-48 and other carbapenemases by an older age, less frequent mention of travel

257

history and an increased proportion of clinical over screening specimens. These

258

characteristics might indicate that import of VIM-1 producing E. coli from abroad might be

259

of minor importance compared to other carbapenemases. A high diversity of different

260

clones indicates that horizontal gene transmission plays a major role in the spread of VIM-

261

1 within E. coli.

M

an

us

cr

ip t

250

d

262

te

263

Ac ce p

264

265

Funding

266

This work was supported by the Robert Koch-Institute with funds provided by the German

267

Ministry of Health (grant no. 1369-402).

268

Transparency declaration

269

MK and SGG have accepted speaking invitations from various pharmaceutical and

270

diagnostical companies though none poses a conflict of interest with the work presented

271

here.

272

NP, FL and AA declare that they have no competing interests.

12/12 Page 12 of 22

273

Acknowledgements

275

The authors thank all laboratories referring isolates to the German National Reference

276

Laboratory for multidrug-resistant gram-negative bacteria. The authors would like to thank

277

Anja Kaminski, Susanne Friedrich, Doris Jaromin, Michaela Stieglitz-Rumberg, Anke

278

Albrecht, Brigitte Hemmerle, Marion Schmidt, Svenja Hirle for excellent technical

279

assistance.

Ac ce p

te

d

M

an

us

cr

ip t

274

13/13 Page 13 of 22

References Aschbacher, R., Giani, T., Corda, D., Conte, V., Arena, F., Pasquetto, V., Scalzo, K., Nicoletti, M., Rossolini, G.M., Pagani, E., 2013. Carbapenemase-producing Enterobacteriaceae during 2011-12 in the Bolzano area (Northern Italy): increasing

ip t

diversity in a low-endemicity setting. Diagn. Microbiol. Infect. Dis. 77, 354–356. doi:10.1016/j.diagmicrobio.2013.08.029

Aschbacher, R., Pagani, L., Doumith, M., Pike, R., Woodford, N., Spoladore, G., Larcher,

cr

C., Livermore, D.M., 2011. Metallo-β-lactamases among Enterobacteriaceae from routine samples in an Italian tertiary-care hospital and long-term care facilities

us

during 2008. Clin. Microbiol. Infect. Off. Publ. Eur. Soc. Clin. Microbiol. Infect. Dis. 17, 181–189. doi:10.1111/j.1198-743X.2010.03225.x

an

Banerjee, R., Johnson, J.R., 2014. A New Clone Sweeps Clean: the Enigmatic Emergence of Escherichia coli Sequence Type 131. Antimicrob. Agents Chemother. 58, 4997– 5004. doi:10.1128/AAC.02824-14

M

Cendejas, E., Gómez-Gil, R., Gómez-Sánchez, P., Mingorance, J., 2010. Detection and characterization of Enterobacteriaceae producing metallo-beta-lactamases in a tertiary-care hospital in Spain. Clin. Microbiol. Infect. Off. Publ. Eur. Soc. Clin.

d

Microbiol. Infect. Dis. 16, 181–183. doi:10.1111/j.1469-0691.2009.02888.x

te

Doumith, M., Ellington, M.J., Livermore, D.M., Woodford, N., 2009. Molecular mechanisms disrupting porin expression in ertapenem-resistant Klebsiella and Enterobacter spp. clinical isolates from the UK. J. Antimicrob. Chemother. 63, 659–667.

Ac ce p

280

doi:10.1093/jac/dkp029

Ehrhard, I., Karaalp, A.-K., Hackel, T., Höll, G., Rodewald, N., Reif, U., Kaase, M., Eckmanns, T., Sydow, W., 2014. [Prevalence of carbapenemase-producing bacteria in hospitals in Saxony, Germany]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 57, 406–413. doi:10.1007/s00103-013-1914-z

Fischer, J., Rodríguez, I., Schmoger, S., Friese, A., Roesler, U., Helmuth, R., Guerra, B., 2012. Escherichia coli producing VIM-1 carbapenemase isolated on a pig farm. J.

Antimicrob. Chemother. 67, 1793–1795. doi:10.1093/jac/dks108 Galani, I., Souli, M., Mitchell, N., Chryssouli, Z., Giamarellou, H., 2010. Presence of plasmid-mediated quinolone resistance in Klebsiella pneumoniae and Escherichia coli isolates possessing blaVIM-1 in Greece. Int. J. Antimicrob. Agents 36, 252–254. doi:10.1016/j.ijantimicag.2010.05.004 14/14 Page 14 of 22

Göttig, S., Hamprecht, A.G., Christ, S., Kempf, V.A.J., Wichelhaus, T.A., 2013. Detection of NDM-7 in Germany, a new variant of the New Delhi metallo-β-lactamase with increased carbapenemase activity. J. Antimicrob. Chemother. 68, 1737–1740. doi:10.1093/jac/dkt088 Kaase, M., Szabados, F., Anders, A., Gatermann, S.G., 2014. Fosfomycin susceptibility in

ip t

carbapenem-resistant Enterobacteriaceae from Germany. J. Clin. Microbiol. 52, 1893–1897. doi:10.1128/JCM.03484-13

KRINKO, 2012. [Hygiene measures for infection or colonization with multidrug-resistant

cr

gram-negative bacilli. Commission recommendation for hospital hygiene and infection prevention (KRINKO) at the Robert Koch Institute (RKI)], 2012. .

us

Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 55, 1311–1354. doi:10.1007/s00103-012-1549-5

an

Mantengoli, E., Luzzaro, F., Pecile, P., Cecconi, D., Cavallo, A., Attala, L., Bartoloni, A., Rossolini, G.M., 2011. Escherichia coli ST131 producing extended-spectrum βlactamases plus VIM-1 carbapenemase: further narrowing of treatment options.

M

Clin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am. 52, 690–691. doi:10.1093/cid/ciq194 Marchiaro, P., Mussi, M.A., Ballerini, V., Pasteran, F., Viale, A.M., Vila, A.J., Limansky, A.S.,

d

2005. Sensitive EDTA-based microbiological assays for detection of metallo-βlactamases in nonfermentative gram-negative bacteria. J. Clin. Microbiol. 43, 5648–

te

5652. doi:10.1128/JCM.43.11.5648-5652.2005 Oteo, J., Saez, D., Bautista, V., Fernández-Romero, S., Hernández-Molina, J.M., Pérez-

Ac ce p

Vázquez, M., Aracil, B., Campos, J., 2013. Carbapenemase-Producing Enterobacteriaceae in Spain in 2012. Antimicrob. Agents Chemother. 57, 6344–

6347. doi:10.1128/AAC.01513-13

Patel, G., Bonomo, R.A., 2013. “Stormy waters ahead”: global emergence of carbapenemases. Front. Microbiol. 4, 48. doi:10.3389/fmicb.2013.00048

Pfeifer, Y., Witte, W., Holfelder, M., Busch, J., Nordmann, P., Poirel, L., 2011. NDM-1producing Escherichia coli in Germany. Antimicrob. Agents Chemother. 55, 1318–

1319. doi:10.1128/AAC.01585-10 Scoulica, E.V., Neonakis, I.K., Gikas, A.I., Tselentis, Y.J., 2004. Spread of blaVIM-1producing E. coli in a university hospital in Greece. Genetic analysis of the integron carrying the blaVIM-1 metallo-beta-lactamase gene. Diagn. Microbiol. Infect. Dis. 48, 167–172. doi:10.1016/j.diagmicrobio.2003.09.012 Tato, M., Coque, T.M., Ruíz-Garbajosa, P., Pintado, V., Cobo, J., Sader, H.S., Jones, R.N., 15/15 Page 15 of 22

Baquero, F., Cantón, R., 2007. Complex clonal and plasmid epidemiology in the first outbreak of Enterobacteriaceae infection involving VIM-1 metallo-beta-lactamase in Spain: toward endemicity? Clin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am. 45, 1171–1178. doi:10.1086/522288 Tijet, N., Macmullin, G., Lastovetska, O., Vermeiren, C., Wenzel, P., Stacey-Works, T., Low,

ip t

D.E., Patel, S.N., Melano, R.G., 2013. Verona integron-encoded metallo-βlactamase 1 in Enterobacteria, Ontario, Canada. Emerg. Infect. Dis. 19, 1156–1158. doi:10.3201/eid1907.121294

cr

Valenza, G., Nickel, S., Pfeifer, Y., Eller, C., Krupa, E., Lehner-Reindl, V., Höller, C., 2014. Extended-spectrum-β-lactamase-producing Escherichia coli as intestinal colonizers

us

in the German community. Antimicrob. Agents Chemother. 58, 1228–1230. doi:10.1128/AAC.01993-13

an

Wendel, A.F., Brodner, A.H.B., Wydra, S., Ressina, S., Henrich, B., Pfeffer, K., Toleman, M.A., Mackenzie, C.R., 2013. Genetic characterization and emergence of the metallo-β-lactamase GIM-1 in Pseudomonas spp. and Enterobacteriaceae during a

M

long-term outbreak. Antimicrob. Agents Chemother. 57, 5162–5165. doi:10.1128/AAC.00118-13

d

Wirth, T., Falush, D., Lan, R., Colles, F., Mensa, P., Wieler, L.H., Karch, H., Reeves, P.R., Maiden, M.C.J., Ochman, H., Achtman, M., 2006. Sex and virulence in Escherichia

2958.2006.05172.x

Ac ce p

281

te

coli: an evolutionary perspective. Mol. Microbiol. 60, 1136–1151. doi:10.1111/j.1365-

282 283 284

Table 1 Carbapenemase genes among 192 E. coli with carbapenemase production carbapenemase gene

number

percentage of carbapenemase producers

blaGES-5

3

1.6

blaKPC-2

13

6.8

blaKPC-3

8

4.2

16/16 Page 16 of 22

blaNDM-3

1

0.5

blaNDM-5

5

2.6

blaVIM-1

36

18.8

blaVIM-4

2

1.0

blaVIM-19

1

0.5

blaOXA-48

86

44.8

blaOXA-162

2

1.0

blaOXA-181

5

2.6

blaOXA-232

2

1.0

blaOXA-244

4

2.1

ip t

11.5

cr

22

us

blaNDM-1

285

Ac ce p

te

d

M

an

286

17/17 Page 17 of 22

Table 2: Demographic characteristics of the four most frequent carbapenemases in E. coli and isolates without any carbapenemase

age (range)

OXA-48

NDM-1

KPC-2

carbapenemase excluded

36

86

22

13

460

a

0, 86

b

0, 100

age (median) 70.5

11, 82

54

70 f

4 (18.2%)

70

g

1 (7.7%)

16 (3.5%)h

cr

travel history 0 (0%) 11 (12.8%) a no age data available for three isolates

0, 103d

37, 80

50 e

287

c

ip t

number

VIM-1

b

no age data available for two isolates

289

c

no age data available for one isolate

290

d

no age data available for 16 isolates

291

e

Irak, Libya (n = 6), Russia, Saudi Arabia, Turkey, United Arab Emirates

292

f

Kosovo, Eastern Europe, Oman, Philippines

293

g

Greece

294

e

Afghanistan (n = 3), Belarus (n = 2), Bulgaria, Libya, Russia (n = 2), Syria (n = 2), Turkey

295

(n = 2), Bahrain, Iraq, Ucraine

Ac ce p

te

d

M

an

us

288

296

18/18 Page 18 of 22

NDM-1 (n = 22)

KPC-2 (n = 13)

carbapenemase excluded (n = 460)

n

%

n

%

n

%

n

%

n

%

blood

1

2.8

3

3.5

0

0

0

0

25

5.4

catheter tip

0

0

0

0

0

0

0

0

3

0.7

abdominal

1

2.8

8

9.3

1

4.5

0

0

13

2.8

respiratory

3

8.3

5

5.8

1

4.5

1

7.7

41

8.9

urinary

15

41.7

17

19.8

5

22.7

2

15.3

134

29.1

wound

3

8.3

8

9.3

1

4.5

1

7.7

56

12.2

skin

1

2.8

2

2.3

0

0

1

7.7

3

0.7

rectal

5

13.9

29

33.7

9

40.9

5

38.4

102

22.2

other screening locations

3

8.3

2

2.3

1

other

3

8.3

11

12.8

1

no information given

1

2.8

1

1.2

3

us

cr

ip t

OXA-48 (n = 86)

4.5

2

15.4

28

6.1

4.5

0

0

27

5.9

13.6

1

7.7

28

6.1

d

specimen source

M

VIM-1 (n = 36)

an

Table 3: Specimen Sources of the four most frequent carbapenemases in E. coli and isolates without any carbapenemase

Ac ce p

te

297

19/19 Page 19 of 22

isolate number

isolation date geographical region

ST1196

NRZ-03157

02/2012

Lower Saxony

NRZ-03236

02/2012

Berlin

NRZ-05285

08/2012

Northrhine-Westphalia

NRZ-06940

01/2013

Berlin

NRZ-06941

01/2013

Berlin

NRZ-06974

01/2013

Berlin

NRZ-02827

01/2012

Berlin

NRZ-04638

07/2012

Rhineland-Palatinate

NRZ-07920

05/2013

Rhineland-Palatinate

NRZ-06565

12/2012

Northrhine-Westphalia

NRZ-11185

12/2013

Bavaria

NRZ-00900

01/2011

Northrhine-Westphalia

NRZ-08927

07/2013

Saxony-Anhalt

NRZ-04404

06/2012

NRZ-05915

10/2012

NRZ-07028

01/2013

Rhineland-Palatinate

NRZ-08180

04/2013

Berlin

ST393 ST410 ST648

NRZ-00591

us

an

Berlin

Rhineland-Palatinate

08/2010

Berlin

05/2013

Saxony

Ac ce p

NRZ-08243

M

ST12

d

ST10

te

ST624

cr

MLST

ip t

Table 4: Geographical distribution of VIM-1 producing E. coli isolates with non-unique MLST types

298

20/20 Page 20 of 22

Table 5: Susceptibility test results of VIM-1 producing E. coli isolates (n = 36) range (mg/l)

MIC50 (mg/l)

MIC90 S I (mg/l) (%) (%)

R (%)

amoxicillin

>128

>128

>128

0

0

100

amoxicillin/clavulanic acid

>128

>128

>128

0

0

100

piperacillin

32, >128

>128

>128

0

0

100

piperacillin/tazobactam

32, >128

>128

>128

0

0

100

mecillinam

128, >128

>128

>128

0

0

100

temocillin

16, >128

128

>128

na

na

na

cefuroxime

>64

>64

>64

0

0

100

cefotaxime

32, >256

128

>256

0

0

100

ceftazidime

32, >256

cefepime

8, >256

cefoxitine

16, >128

aztreonam

>256

us

0

0

100

64

>256

0

0

100

64

>128

na

na

na

≤0.25, >32

0.5

>32

55.6 0

ertapenem

≤0.00625, 4

0.25

2

64.7 14.7 20.6

imipenem

1, 16

2

4

61.1 36.1 2.8

meropenem

0.5, 64

8

32

13.9 50.0 36.1

doripenem

2, 64

8

16

0

gentamicin

≤0.5, 32

1

4

77.8 13.9 5.7

≤0.5, 16

2

8

52.8 36.1 11.1

1, 8

2

4

100 0

0

1, >64

64

>64

na

na

na

minocycline

≤0.5, 32

4

16

na

na

na

tigecycline

0.25, 2

0.5

1

94.4 5.6

0

trimethoprim

≤0.125, >16

>16

>16

16.7 0

83.3

sulfamethoxazole/trimethoprim

≤0.125, >16

>16

>16

19.4 0

80.6

ciprofloxacin

≤0.0156, >32

16

>32

37.1 0

62.9

levofloxacin

≤0.0625, >8

8

>8

36.1 2.8

61.1

chloramphenicol

4, >64

16

>64

36.1 0

63.9

colistin

≤0.25, 2

0.5

1

100 0

0

fosfomycin

≤1, 32

4

16

100 0

0

nitrofurantoin

8, 128

32

64

97.2 0

2.8

amikacin

M

Ac ce p

tetracycline

te

tobramycin

an

256

d

cr

ip t

antibiotic

44.4

27.8 72.2

299 300

21/21 Page 21 of 22

301 Figure 1: Neighbor-joining tree of MLST of the studied VIM-1 producing Escherichia coli.

Ac ce p

te

d

M

an

us

cr

ip t

302

303 304

22/22 Page 22 of 22