molecular typing of vancomycin-resistant enterococcus faecium with an automated repetitive...
TRANSCRIPT
Correspondence: M. Kard é n-Lilja, National Institute for Health and Welfare, THL, Department of Infectious Disease Surveillance and Control, Antimicrobial Resistance Unit, PO Box 30, 00271 Helsinki, Finland. Tel: � 358 29 524 8564. Fax: � 358 29 524 8238. E-mail: [email protected]
(Received 20 June 2012 ; accepted 2 October 2012 )
Scandinavian Journal of Infectious Diseases, 2013; 45: 350–356
ISSN 0036-5548 print/ISSN 1651-1980 online © 2013 Informa HealthcareDOI: 10.3109/00365548.2012.737475
ORIGINAL ARTICLE
Molecular typing of vancomycin-resistant Enterococcus faecium with an automated repetitive sequence-based PCR microbial typing system compared with pulsed-fi eld gel electrophoresis and multilocus sequence typing
MINNA KARD É N-LILJA 1 , JAANA VUOPIO 2 , MARKKU KOSKELA 3 , P Ä IVI TISSARI 4 & SAARA SALMENLINNA 5
From the 1 Department of Infectious Disease Surveillance and Control, Antimicrobial Resistance Unit, National Institute for Health and Welfare, THL, Helsinki, Finland, 2 Department of Infectious Disease Surveillance and Control, Antimicrobial Resistance Unit, National Institute for Health and Welfare, THL, Turku, Finland, 3 Clinical Microbiology Laboratory, Oulu University Hospital, Oulu, Finland, 4 Division of Clinical Microbiology, Department of Bacteriology, HUSLAB, Helsinki University Hospital, Helsinki, Finland, and 5 Department of Infectious Disease Surveillance and Control, Bacteriology Unit, National Institute for Health and Welfare, THL, Helsinki, Finland
Abstract Background: Pulsed-fi eld gel electrophoresis (PFGE) is the main typing method used for the molecular typing of vancomycin-resistant Enterococcus faecium (VREfm). However, more rapid and unambiguous typing methods are needed. DiversiLab, a repetitive sequence-based PCR (rep-PCR), offers an alternative method for strain typing. Methods: Thirty-nine VREfm isolates with known epidemiological relationships were characterized by semi-automated rep-PCR (DiversiLab), PFGE, and multilocus sequence typing (MLST). Results: The DiversiLab results were analysed in 2 ways: fi rst relying solely on the DiversiLab software, and second by DiversiLab analysis combined with manual interpretation. The analysis with interpretation yielded more DiversiLab profi les, correlated better with PFGE and MLST, and grouped the isolates better according to their relatedness in time and space. However, most of the DiversiLab groups also included isolates with different PFGE and MLST types. Conclusions: DiversiLab provides rapid information when investigating a potential hospital outbreak. However, the interpretation of E. faecium DiversiLab results cannot be fully automated and is not always straightforward. Other typing methods may be necessary to confi rm the analysis.
Keywords: Enterococcus faecium , rep-PCR , MLST , PFGE , typing
Introduction
Problems with vancomycin-resistant enterococci (VRE) emerged a few decades ago in European hospitals. The rising number of VRE infections is mainly due to the increase in vancomycin-resistant Enterococcus faecium (VREfm) [1]. In particular, one large group of VREfm strains, clonal complex 17 (CC17), has been associated with nosocomial out-breaks around the world [2]. Recent studies suggest that CC17 strains may have evolved from different ancestors and that CC17 actually consists of multiple different genetic lineages [3,4].
Reliable and fast strain typing is one of the cor-nerstones in trying to control the spread of resistant
bacteria. Pulsed-fi eld gel electrophoresis (PFGE) is the most common method used for typing of VRE for epidemiological purposes [5 – 7]. However, it is laborious and requires several days to obtain results. Also the interpretation of PFGE results requires experience and is not always straightforward [6,8,9].
Other typing methods used for E. faecium include the sequence-based typing method mul-tilocus sequence typing (MLST) and the PCR-based multiple-locus variable-number tandem repeat analysis (MLVA) [10]. MLST provides unambiguous results, but it is also time-consuming and the discriminatory power does not reach that of PFGE [11]. The MLVA protocol is faster than
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.
Comparison of E. faecium molecular typing methods 351
PFGE and divides E. faecium into distinct clonal complexes. However, for outbreak investigations PFGE is still recommended over the MLVA procedure [12].
The DiversiLab system is a commercially avail-able semi-automated method using repetitive sequence-based PCR (rep-PCR) for microbial strain typing. It has been used in VRE outbreak investiga-tions and its performance in comparison to PFGE for strain typing of enterococci has been evaluated previously. However, the criteria used to determine the relatedness of enterococci with DiversiLab have varied [13 – 16], and studies of different settings and enterococcal populations are needed. The purpose of our study was to compare DiversiLab with PFGE and MLST in the typing of 39 epidemiologically related and unrelated VREfm strains in Finland.
Materials and methods
Bacterial isolates and typing
All Finnish VRE cases are notifi ed to the National Infectious Disease Register and the corresponding strains are sent to the National Institute for Health and Welfare (THL). A multiplex PCR for verifi cation of E. faecium and Enterococcus faecalis species and van ( vanA , vanB and vanC1/vanC2-C3 ) genes is per-formed [17]. Typing is performed with PFGE and the results are interpreted according to the following criteria [5]: strains with an identical banding pattern are considered as indistinguishable, strains with � 3 band differences are considered as subtypes, and strains with � 3 but � 6 band differences are consid-ered as possible subtypes. Strains are classifi ed as epidemic or sporadic strain types based on PFGE and epidemiological data. A strain is considered as epidemic when a PFGE profi le with � 7 band differences is encountered from at least 5 persons. The Bionumerics 5.1 position tolerance settings for both optimization and for position tolerance were 1.00%.
For this study a total of 39 VREfm isolates with a known PFGE profi le were selected from the THL strain collection (Table I). The majority of the strains represented 2 different outbreak strains, VRE IV and VII and their subtypes. The epidemiological informa-tion related to the site (hospital district and health care facility) and time (by month and year) of isola-tion is shown in Table I. VRE IV is a strain originally linked to an outbreak in Northern Finland (North Ostrobothnia) during the period 2004 – 2008 and VRE VII to an outbreak in the Helsinki University Health District in 2006. In addition to these, repre-sentatives of 6 other Finnish epidemic VREfm strains (HKI II, VRE I, VRE II, VRE III, VRE VI, and VRE IX; 1 isolate each) since 1995 were studied. Also
13 sporadic strains from the period 2006 – 2009 were studied. The selection criteria for sporadic strains were: (1) all blood isolates ( n � 2), (2) at least 1 representative strain of VREfm showing the same PFGE profi le and isolated from 2 – 4 persons ( n � 4), (3) representatives of VREfm isolated from other hospitals ( n � 4), and (4) randomly selected strains from 2 health care districts (HUS and North Ostrobothnia) where most of the strains came from ( n � 3).
MLST
Multilocus sequence typing (MLST) was done as reported previously [11] (http://www.mlst.net). In addition, alternative primers were used for adk , atp , purK , and pstS genes according to the MLST website. MLST was performed for all but 6 isolates of PFGE type VRE IV strains from the North Ostrobothnia epidemic.
Rep-PCR
DNA extraction was done using the Ultra Clean TM Microbial Isolation Kit (Mo Bio Laboratories, Carlsbad, CA, USA) from VREfm that were grown overnight in 3 ml of brain – heart infusion broth at 37 ° C, in accordance with the kit instructions. The extracted DNA was diluted to 25 – 50 ng/ μ l, if neces-sary. The DNA was amplifi ed with the DiversiLab Enterococcus DNA Fingerprinting Kit (Bacterial Barcodes) in accordance with the manufacturer ’ s instructions, and the rep-PCR products were sepa-rated with electrophoresis performed in a Micro-fl uidics DNA LabChip (Caliper Life Sciences, Hopkinton, MA, USA). The rep-PCR patterns were analyzed by DiversiLab software (version 3.4) using the Pearson ’ s correlation for the calculation of the percentage of similarity. The data used for interpre-tation consisted of graphs of fl uorescence intensity (electropherograms, also visualized with virtual-gel images) which could be overlaid for the comparison of 2 strains, automatically generated dendrograms, similarity matrices, and scatter plot images. Two separate analyses with different criteria were used: in the fi rst analysis no interpretation was used and the strains with similarity � 97% were classifi ed to the same rep-PCR group. In the second analysis, strains with � 98% similarity and no differences in the presence or intensity of electropherogram peaks were considered as indistinguishable. Strains with � 96% similarity and � 1 difference in the electro-pherograms (peak presence or intensity difference) were considered as similar and as part of the same rep-PCR group. The isolates within a same rep-PCR group were assumed to be genetically related. Strains with � 96% similarity or with 2 or more
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.
352 M. Kard é n-Lilja et al.
Table I. The epidemiological information, resistance genotype, and typing results of 39 vancomycin-resistant Enterococcus faecium isolates analyzed.
Interpretation:
1 2
Sample ID
Month and y of isolation
Health care district a
Institute code b
Resistance gene PFGE c MLST
Rep-PCR group 97% d
Rep-PCR group e
Rep-PCR pattern f
77805 12/1996 HUS A vanA HKI II 17 I Unique P577857 01/1997 HUS B vanB VRE I 16 I A P2158048-6 02/2006 HUS C vanB VRE VII 343 I Unique P3160050-4 07/2007 HUS A vanB Sporadic 275 I A P6155307-6 09/2004 North Ostrobothnia D vanB VRE IV – I A P1155185-1 09/2004 North Ostrobothnia D vanB VRE IV subtype 496 I A P1155880-3 01/2005 North Ostrobothnia E vanB VRE IV subtype 18 I A P1156063-8 02/2005 North Ostrobothnia E vanB VRE IV – I A P1156566-9 05/2005 North Ostrobothnia D vanB VRE IV – I A P1158693-2 08/2006 North Ostrobothnia D vanB VRE IV possible
subtype18 I A P4
159456-2 02/2007 North Ostrobothnia F vanB VRE IV 18 I A P1159455-1 02/2007 North Ostrobothnia F vanB VRE IV 18 I A P4159542-7 03/2007 North Ostrobothnia F vanB VRE IV – I A P1159540-5 03/2007 North Ostrobothnia F vanB VRE IV 18 I A P4159541-6 03/2007 North Ostrobothnia F vanB VRE IV – I A P1160143-7 08/2007 North Ostrobothnia G vanB VRE IV possible
subtype18 I A P2
160527-5 12/2007 North Ostrobothnia H vanB VRE IV – I A P4157891-2 01/2006 HUS I vanB VRE VII
subtype343 II C P10
158050-7 02/2006 HUS I vanB VRE VII subtype
343 II C P12
158049-7 02/2006 HUS C vanB VRE VII 343 II C P13160545-5 12/2007 HUS C vanB VRE VII
subtype343 II C P9
159515-0 03/2007 HUS I vanA Sporadic 412 II C P12160385-9 10/2007 North Ostrobothnia D vanB Sporadic 78 II Unique P14161243-7 05/2008 North Ostrobothnia D vanB VRE VII
possible subtype
343 II C P12
161266-5 05/2008 HCD1 J vanB VRE IX 192 II Unique P11119577 01/2000 North Ostrobothnia K vanA VRE III 16 III D P18162144-6 02/2009 North Ostrobothnia L vanB Sporadic B 17 III E P22160516-8 12/2007 HCD2 M vanB Sporadic A 18 III D P17161099-8 04/2008 HCD2 M vanB Sporadic A 18 III Unique P20160366-7 10/2007 HCD3 N vanB Sporadic B 17 III Unique P19161198-8 04/2008 HCD3 N vanB Sporadic B 17 III E P22161596-2 08/2008 North Ostrobothnia D vanB VRE IV subtype 18 III Unique P21156882-1 07/2005 HCD4 O vanB VRE VI 17 IV B P7158350-9 04/2006 HCD4 O vanB Sporadic 275 IV B P8161218-4 05/2008 North Ostrobothnia D vanA Sporadic 498 4 Unique P15156149-8 03/2005 North Ostrobothnia D vanB VRE II 252 5 Unique P16162753-8 07/2009 HCD3 N vanB Sporadic 440 7 Unique P23162087-6 12/2008 HUS I vanA Sporadic 497 8 Unique P24159239-8 01/2007 HUS A vanB Sporadic 275 9 Unique P25
PFGE, pulsed-fi eld gel electrophoresis; MLST, multilocus sequence typing; rep-PCR, repetitive sequence-based PCR. a Health care districts (HCDs) other than HUS (Hospital District of Helsinki and Uusimaa) or North Ostrobothnia are coded from HCD1 to HCD4. b Health care facilities coded with letters. c Isolates marked with letter A are PFGE subtypes to each other; isolates marked with letter B are PFGE subtypes to each other (161198-8 and 162144-6 isolated from the same person). d Rep-PCR groups (I – IV) and unique patterns (4 – 9) analyzed as having � 97% similarity. e Rep-PCR groups (A – E) as interpreted according to � 96% similarity and a maximum of 1 difference in the electropherograms. Unique rep-PCR patterns without a group are marked as ‘ unique ’ . f Rep-PCR patterns (P1 – P25); identical patterns with � 98% similarity and no differences in electropherograms.
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.
Comparison of E. faecium molecular typing methods 353
differences (peak presence or intensity difference) were considered as different strains. The analysis criteria were modifi ed from the general criteria found in the DiversiLab Analysis Guide.
Results
PFGE divided the 39 E. faecium isolates analyzed into 8 epidemic types (26 isolates) and 10 sporadic PFGE types (13 isolates) (Table I). Among the 33 isolates analyzed with MLST, 13 types were encoun-tered (Table I). Thirteen of the isolates belonged to sequence type (ST) 18 (9 isolates) and its single locus variants ST275 (3 isolates) and ST496 (1 isolate). Nine of the isolates belonged to ST17 (5 isolates) and its single locus variants ST16 (2 isolates), ST252 (1 isolate), and ST78 (1 isolate). ST343 was encountered from 6 isolates and its single locus variant ST412 from 1 isolate. Also 4 other STs (ST192, ST440, ST497, and ST498) were encountered among the isolates investigated.
The fi rst rep-PCR analysis of the 39 isolates clustered the isolates into 4 rep-PCR groups and 5 unique rep-PCR patterns (Table I) when grouping according to � 97% similarity was used. All but 1 of the PFGE type VRE IV strains (including 13 ST18 isolates and 1 ST496 isolate) belonged to rep-PCR group I. Four other strains belonging to the same rep-PCR group I were of PFGE types VRE VII (ST343), VRE I (ST16), HKI II (ST17), and a sporadic isolate (ST275). The rep-PCR group II consisted of 5 PFGE type VRE VII (ST343) isolates, 1 VRE IX isolate (ST192), and 2 sporadic (ST412 and ST78) isolates. The isolates in rep-PCR group III were of PFGE types VRE IV (ST18), VRE III (ST16), and 5 sporadic isolates (ST18 and ST17). Rep-PCR group IV consisted of 2 isolates, PFGE type VRE VI (ST17) and 1 sporadic isolate (ST275). Five isolates had a unique rep-PCR pattern (ST275, ST497, ST440, ST252, and ST498).
The second rep-PCR analysis was performed using interpretation criteria � 96% similarity and a maximum of 1 difference in the electropherograms (peak presence or intensity difference) for group defi nition. For a rep-PCR pattern, � 98% similarity and no differences in the electropherograms were required. In this analysis, the 39 E. faecium isolates clustered into 5 rep-PCR groups (A – E, 14 different rep-PCR patterns) and 11 unique rep-PCR patterns (Table I, Figure 1). Thirteen isolates with PFGE pro-fi le VRE IV (ST18, ST496) were clustered into rep-PCR group A and these isolates could be divided into 3 different rep-PCR patterns (Figure 1). One VRE IV isolate had a unique rep-PCR pattern P21. How-ever, by visual inspection of this pattern, a peak at datapoint ~ 450 could be detected. In this study mate-rial this peak was typical only for isolates with PFGE
profi le VRE IV (all VRE IV isolates except 160143-7). For 2 isolates – VRE I epidemic isolate (ST16) with rep-PCR pattern P2 and 1 sporadic isolate with rep-PCR pattern P6 (160050-4, ST275) – an unambig-uous grouping was not possible (Figure 1). These isolates could be included in rep-PCR group A or alternatively a new rep-PCR group could have been formed with HKI II epidemic strain with rep-PCR pattern P5 (ST17). Isolates were included in rep-PCR group A as the difference between these isolates and HKI II was a peak (datapoint ~ 400) unique to only HKI II isolate within all isolates. The rep-PCR group C included 6 isolates: 5 with PFGE profi le VRE VII (ST343) and 1 with a sporadic PFGE pro-fi le (ST412). These 6 isolates could be divided into 4 rep-PCR patterns. One isolate with PFGE profi le VRE VII had a unique rep-PCR pattern P3. Rep-PCR groups B, D, and E consisted of 2 strains each and the group E strains were isolated from the same person in different years.
The analysis also showed that electropherograms with clear differences can have an extremely high percentage of similarity, as shown for example iso-lates 119577 and 160366 (Figure 2a and b). These isolates also had different PFGE, MLST, and van gene profi les.
Discussion
We analyzed 39 E. faecium isolates in order to assess the possibility of using the DiversiLab system for the recognition of epidemic VRE strains mostly belonging to MLST sequence types associated with hospital-adapted VREfm genetic lineages. The study material constituted a broad coverage of the Finnish VREfm isolates, as it included all Finnish vancomy-cin-resistant E. faecium epidemic strains, several representatives of 2 epidemics, and a number of spo-radic isolates. VRE has been a notifi able fi nding in Finland since 1995 and all strains are stored and typed at a single reference laboratory.
In our study, when relying only on a percentage threshold as a grouping criterion, 4 rep-PCR groups were formed, which all included isolates with differ-ent PFGE and MLST types. For example the largest rep-PCR group I included 5 different PFGE and 6 different MLST types. Therefore, an interpretation relying only on percentages can be misleading. In previous studies evaluating rep-PCR for E. faecium the analysis criteria have varied [13 – 15,18]. A report investigating a VRE outbreak with the DiversiLab system used a criterion where strains with 2 bands difference (95 – 97% similarity) were still classifi ed as similar, and the report was done without comparison to PFGE [15]. In a recent report comparing rep-PCR and PFGE the samples were classifi ed into rep-PCR groups based only on a percentage of similarity
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.
354 M. Kard é n-Lilja et al.
� 97%. In this report, rep-PCR was concluded to be less discriminatory than PFGE but the results cor-related with PFGE typing [13]. A new report by Fluit et al. analyzed the usefulness of rep-PCR for a wide
range of bacteria. In that report, for Gram-positive species the strains with a similarity � 96% were con-sidered different, while isolates with a similarity � 99% were considered as indistinguishable. The
Figure 1. Rep-PCR, PFGE, and MLST results of the 39 vancomycin-resistant Enterococcus faecium isolates. Rep-PCR groups (A – E) as interpreted according to � 96% similarity and a maximum of 1 difference in the electropherograms. Rep-PCR patterns (1 – 25): identical patterns with � 98% similarity and no differences in electropherograms. The 97% similarity line is marked in the dendrogram.
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.
Comparison of E. faecium molecular typing methods 355
unique rep-PCR patterns were identifi ed, resulting in an interpretation more similar to that of PFGE than the fi rst analysis. The previous reports analyzing both E. faecium or/and MRSA rep-PCR results have con-cluded recommending PFGE over rep-PCR for con-fi rming strain relatedness [18,19]. In contrast to other reports, one of the latest reports published concluded that rep-PCR was more discriminative than PFGE. The report still recommended PFGE as the reference method for typing of epidemic isolates [16].
Even though the DiversiLab rep-PCR protocol is strictly standardized to achieve high reproducibility, the manual analysis is still subjective. The majority of the interpretation problems come with very small peaks or minor variations in intensity, as they can easily be regarded as non-signifi cant. The same kind of variations could also be seen in the electropherograms of a repeated analysis of the same strain (data not shown). The analysis program also allows the use of different similarity methods for the analysis with different weight put on the presence of the peaks or peak intensity. The Kullback – Leibler method rather than Pearson ’ s corre-lation is recommended for strains with typically only a few large peaks (typically Gram-positives), as the pres-ence of peaks is weighted higher than the intensity of peaks. In our study, the Kullback – Leibler method clustered the strains even more closely to each other, and clustering by Pearson ’ s correlation resulted in a
strains with similarity between these values were analyzed manually. The report concluded that rep-PCR could not be recommended for methicillin-resistant Staphylococcus aureus (MRSA) or E. faecium strain typing due to its poor performance with these species for the detection of hospital outbreaks, with the currently used typing techniques being pre-ferred [18]. In a recent report, rep-PCR data were compared to MLST and PFGE data with 2 sets of VRE E. faecium isolates. The rep-PCR clusters were specifi ed as the discriminant threshold calculated by the software. The report concluded that rep-PCR was useful for rapid VRE screening during an out-break, but that PFGE should still be used for the confi rmation of genetic relatedness of epidemic iso-lates. It was also pointed out that an analysis without taking account of both the intensity and position of the bands may be impossible [16].
For clonal population analysis, such as MRSA, the manufacturer recommends that strains with 2 differ-ences in the electropherogram should be considered as different. The same criterion was also considered suit-able for E. faecium isolates even though these infectious isolates may not be strictly clonally related [4]. The second analysis using percentages of similarity com-bined with manual interpretation with electrophero-gram overlays to detect the peak and intensity differences resulted in 5 rep-PCR groups. In this analysis more
Figure 2. (a) PFGE images and MLST results of 160366-7 and 119577; (b) overlaid rep-PCR electropherograms of the same strains with similarity of 98.6%. Differences in datapoint ~ 210 (peak) and datapoint ~ 390 (intensity) are marked with arrows.
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.
356 M. Kard é n-Lilja et al.
isolates using restriction endonucleases with infrequent recognition sites . J Clin Microbiol 1990 ; 28 : 2059 – 63 . Suppola JP , Kolho E , Salmenlinna S , Tarkka E , [6] Vuopio-Varkila J , Vaara M . vanA and vanB incorporate into an endemic ampicillin-resistant vancomycin-sensitive Ente-rococcus faecium strain: effect on interpretation of clonality . J Clin Microbiol 1999 ; 37 : 3934 – 9 . Turabelidze D , Kotetishvili M , Kreger A , Morris JG Jr, [7] Sulakvelidze A . Improved pulsed-fi eld gel electrophoresis for typing vancomycin-resistant enterococci . J Clin Microbiol 2000 ; 38 : 4242 – 5 . Tenover FC , Arbeit RD , Goering RV , Mickelsen PA , [8] Murray BE , Persing DH , et al . Interpreting chromosomal DNA restriction patterns produced by pulsed-fi eld gel electrophoresis: criteria for bacterial strain typing . J Clin Microbiol 1995 ; 33 : 2233 – 9 . van Belkum A , Tassios PT , Dijkshoorn L , Haeggman S , [9] Cookson B , Fry NK , et al . Guidelines for the validation and application of typing methods for use in bacterial epidemiol-ogy . Clin Microbiol Infect 2007 ; 13(Suppl 3) : 1 – 46 . Top J , Schouls LM , Bonten MJ , Willems RJ . Multiple-locus [10] variable-number tandem repeat analysis, a novel typing scheme to study the genetic relatedness and epidemiology of Enterococcus faecium isolates . J Clin Microbiol 2004 ; 42 : 4503 – 11 . Homan WL , Tribe D , Poznanski S , Li M , Hogg G , [11] Spalburg E , et al . Multilocus sequence typing scheme for Enterococcus faecium . J Clin Microbiol 2002 ; 40 : 1963 – 71 . Werner G , Klare I , Witte W . The current MLVA typing [12] scheme for Enterococcus faecium is less discriminatory than MLST and PFGE for epidemic-virulent, hospital-adapted clonal types . BMC Microbiol 2007 ; 7 : 28 . Chuang YC , Wang JT , Chen ML , Chen YC . Comparison of [13] an automated repetitive-sequence-based PCR microbial typ-ing system with pulsed-fi eld gel electrophoresis for molecular typing of vancomycin-resistant Enterococcus faecium . J Clin Microbiol 2010 ; 48 : 2897 – 901 . Pounder JI , Shutt CK , Schaecher BJ , Woods GL . Clinical [14] evaluation of repetitive sequence-based polymerase chain reaction using the Diversi-Lab System for strain typing of vancomycin-resistant enterococci . Diagn Microbiol Infect Dis 2006 ; 54 : 183 – 7 . Sherer CR , Sprague BM , Campos JM , Nambiar S , [15] Temple R , Short B , et al . Characterizing vancomycin-resistant enterococci in neonatal intensive care . Emerg Infect Dis 2005 ; 11 : 1470 – 2 . Bourdon N , Lemire A , Fines-Guyon M , Auzou M , [16] Perichon B , Courvalin P , et al . Comparison of four methods, including semi-automated rep-PCR, for the typing of vancomycin-resistant Enterococcus faecium . J Microbiol Methods 2010 ; 84 : 74 – 80 . Dutka-Malen S , Evers S , Courvalin P . Detection of [17] glycopeptide resistance genotypes and identifi cation to the species level of clinically relevant enterococci by PCR . J Clin Microbiol 1995 ; 33 : 24 – 7 . Fluit AC , Terlingen AM , Andriessen L , Ikawaty R , [18] van Mansfeld R , Top J , et al . Evaluation of the DiversiLab system for detection of hospital outbreaks of infections by dif-ferent bacterial species . J Clin Microbiol 2010 ; 48 : 3979 – 89 . Tenover FC , Gay EA , Frye S , Eells SJ , Healy M , [19] McGowan JE Jr . Comparison of typing results obtained for methicillin-resistant Staphylococcus aureus isolates with the DiversiLab system and pulsed-fi eld gel electrophoresis . J Clin Microbiol 2009 ; 47 : 2452 – 7 . Hunter PR , Gaston MA . Numerical index of the discrimina-[20] tory ability of typing systems: an application of Simpson ’ s index of diversity . J Clin Microbiol 1988 ; 26 : 2465 – 6 .
differentiation pattern resembling more closely that of other typing methods (results not shown).
The DiversiLab system may be a potential typing method for analyzing a large number of samples to obtain a fast result for epidemiological investigation purposes, when the epidemiological background data of the samples are available. A part of the PFGE type VRE IV isolates with a known epidemiological linkage (time and place clustering) differed in manual analysis by 1 peak difference. Similar variation could be seen in repeated rep-PCR runs (results not shown). How-ever, these isolates clustered into the same rep-PCR group as did most of the other PFGE type IV strains.
This study was limited by the small number of isolates analyzed and the scarce background data. There was no information about possible contacts between patients. Also, the rep-PCR interpretation was not blinded, which could have affected to some degree the objectivity of the fi nal results. The dis-criminatory power describes the ability to distinguish between unrelated strains and was not calculated between rep-PCR and PFGE and MLST, as the iso-lates from the outbreaks were not unrelated [20].
In our opinion, rep-PCR can be considered as a rapid typing method that gives results in 1 working day. However, consideration should be given to how the results are interpreted. Interpretation relying only on a percentage threshold may overestimate the possibility of a clonal relationship between isolates, with PFGE still being recommended to verify the typing result.
Declaration of interest: All the authors declare that they have no confl ict of interest. This study was funded by the European Commissions Seventh Framework Programme (FP7) project TROCAR – Translation Research On Combating Antimicrobial Resistance.
References
Werner G , Coque TM , Hammerum AM , Hope R , [1] Hryniewicz W , Johnson A , et al . Emergence and spread of vancomycin resistance among enterococci in Europe . Euro Surveill 2008 ; 13 . pii: 19046 . Willems RJ , Top J , van Santen M , Robinson DA , Coque TM , [2] Baquero F , et al . Global spread of vancomycin-resistant Enterococcus faecium from distinct nosocomial genetic complex . Emerg Infect Dis 2005 ; 11 : 821 – 8 . Willems RJ , Hanage WP , Bessen DE , Feil EJ . Population [3] biology of Gram-positive pathogens: high-risk clones for dis-semination of antibiotic resistance . FEMS Microbiol Rev 2011 ; 35 : 872 – 900 . van Schaik W , Top J , Riley DR , Boekhorst J , Vrijenhoek JE , [4] Schapendonk CM , et al . Pyrosequencing-based comparative genome analysis of the nosocomial pathogen Enterococcus faecium and identifi cation of a large transferable pathogenic-ity island . BMC Genomics 2010 ; 11 : 239 . Murray BE , Singh KV , Heath JD , Sharma BR , Weinstock GM . [5] Comparison of genomic DNAs of different enterococcal
Scan
d J
Infe
ct D
is D
ownl
oade
d fr
om in
form
ahea
lthca
re.c
om b
y M
ichi
gan
Uni
vers
ity o
n 11
/01/
14Fo
r pe
rson
al u
se o
nly.