HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by McDowell, A. Articles by Elborn, J. S. Search for Related Content PubMed PubMed Citation Articles by McDowell, A. Articles by Elborn, J. S. Agricola Articles by McDowell, A. Articles by Elborn, J. S.

Epidemiology of Burkholderia cepacia complex species recovered from cystic fibrosis patients: issues related to patient segregation

^1,3Department of Bacteriology, Northern Ireland Public Health Laboratory¹ and Northern Ireland Regional Adult Cystic Fibrosis Centre³, Belfast City Hospital, Belfast BT9 7AB, UK ²Cardiff School of Biosciences, Cardiff University, Cardiff CF1 3US, UK

Correspondence Andrew McDowell a.mcdowell{at}qub.ac.uk

Received December 2, 2003
Accepted March 16, 2004

Studies of the prevalence of Burkholderia cepacia complex species amongst cystic fibrosis (CF) patients in different geographical regions, and the association between cross-infection and putative transmissibility markers, will further our understanding of these organisms and help to address infection-control issues. In this study, B. cepacia complex isolates from CF patients in different regions of Europe were analysed. Isolates were examined for B. cepacia complex species and putative transmissibility markers [cable pilin subunit gene (cblA) and the B. cepacia epidemic strain marker (BCESM)]. Sporadic and cross-infective strains were identified by random amplification of polymorphic DNA (RAPD). In total, 79 % of patients were infected with Burkholderia cenocepacia (genomovar III), 18 % with Burkholderia multivorans (genomovar II) and less than 5 % of patients with B. cepacia (genomovar I), Burkholderia stabilis (genomovar IV) or Burkholderia vietnamiensis (genomovar V). The cblA and BCESM transmissibility markers were only detected in strains of B. cenocepacia. The BCESM was a more sensitive marker for transmissible B. cenocepacia strains than cblA, although sporadic B. cenocepacia strains containing the BCESM, but lacking cblA, were also observed. Furthermore, clusters of cross-infection with transmissibility marker-negative strains of B. multivorans were identified. In conclusion, B. cenocepacia was the greatest cause of cross-infection, and the most widely distributed B. cepacia complex species, within these CF populations. However, cross-infection was not exclusive to B. cenocepacia and cblA and the BCESM were not absolute markers for transmissible B. cenocepacia, or other B. cepacia complex strains. It is therefore suggested that CF centres cohort patients based on the presence or absence of B. cepacia complex infection and not on the basis of transmissibility marker-positive B. cenocepacia as previously suggested.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Burkholderia cepacia has emerged as a life-threatening cause of infection in patients with cystic fibrosis (CF). Strains of B. cepacia can be highly transmissible, resulting in patient-to-patient spread within and between CF centres (Govan et al., 1993). To date, the most effective strategy to combat B. cepacia infection has been the strict segregation of infected patients. Although successful, the introduction of such a policy has proved controversial and unpopular with many CF patients. Furthermore, segregation will not prevent sporadic acquisition of B. cepacia complex organisms from natural environments (LiPuma et al., 2002).

To complicate the clinical problems associated with B. cepacia infection, polyphasic taxonomic studies have now discovered that B. cepacia is a complex of nine genetically distinct species, all capable of causing CF-related infections (Mahenthiralingam et al., 2002). Putative transmissibility markers for B. cepacia complex bacteria have now also been identified. These markers include the cable pilin subunit gene (cblA), which encodes a giant cable-like pilus that facilitates adherence to respiratory mucins (Sajjan et al., 1995), and the B. cepacia epidemic strain marker (BCESM), a 1.4-kb open reading frame with homology to several negative transcriptional regulatory genes (Mahenthiralingam et al., 1997).

Initial studies by Vandamme et al. (1997) revealed that Burkholderia cenocepacia (formerly genomovar III) was the predominant species amongst their CF-related B. cepacia complex isolates. Further investigations by Clode et al. (2000) also found that B. cenocepacia was the prevalent species amongst 114 B. cepacia complex isolates recovered from CF patients in the UK. In addition, their investigations found that putative transmissibility markers, most specifically the cblA gene, were associated with epidemic spread of B. cenocepacia. This led to the suggestion that, in the UK at least, only patients colonized with transmissibility marker-positive B. cenocepacia strains should be strictly segregated from patients colonized with other transmissibility marker-negative B. cepacia complex strains, who themselves may not require segregation. Relaxation of segregation is an attractive proposal for CF centres since it would undoubtedly reduce the psychological impact upon the CF patient population, as well as the burden placed upon centres to provide and maintain this tedious and socially difficult policy. However, extensive multi-centre information on the distribution of B. cepacia complex species amongst CF patients, and continued data on the relationship between putative transmissibility markers and cross-infective strains, is required before such a recommendation could be seriously considered. Such studies are especially urgent since more recent data from the USA suggest that cblA and the BCESM may not be sufficient markers for transmissibility or indeed virulence of B. cenocepacia (LiPuma et al., 2001; Sajjan et al., 2002).

In this current study, we investigated the distribution of B. cepacia complex species, as well as the association between putative transmissibility markers and cross-infective strains, amongst B. cepacia complex organisms isolated from 131 patients receiving treatment in 14 CF centres in different regions of Europe.

	METHODS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
B. cepacia complex isolates.

For this study, B. cepacia complex sputum isolates were recovered from 131 CF patients receiving treatment in different regions of Europe between 1997 and 2001. Patients were from four centres in Northern and Southern Ireland (50 isolates), nine centres in Great Britain (46 isolates) and one centre in mainland Europe (35 isolates). Samples were received either as pure bacterial isolates or as expectorated sputum. B. cepacia complex organisms in sputum samples were immediately isolated by culture on MAST selective agar (MAST Diagnostics). Isolates were initially analysed using the API 20NE phenotypic identification system (bioMérieux) following the manufacturer's instructions. As in the study of LiPuma et al. (2001), when multiple isolates from one patient were received, only the first was included in the study so that the best distribution of B. cepacia complex species could be assessed.

Molecular identification of B. cepacia complex species.

Genomic template DNA for molecular analysis was prepared from all isolates as described previously (McDowell et al., 2001). B. cepacia complex species were identified by 16S and 16S–23S rRNA-based PCR, as well as recA-based methods (LiPuma et al., 1999; Mahenthiralingam et al., 2000). Our recA-based diagnostic scheme for B. cepacia complex species identification consisted of RFLP analysis (with HaeIII and MnlI) of the recA gene, followed by confirmation with species-specific recA primers. Analysis of PCR and RFLP products by agarose electrophoresis was as described before (McDowell et al., 2001). Nucleotide sequence analysis of recA amplicons was performed on isolates with novel RFLP patterns to confirm species identification. Sequences were obtained using an ALF Express II DNA sequencer (Amersham-Pharmacia). Multiple sequence alignments (CLUSTAL W) and phylogenetic analysis were performed as described previously (Mahenthiralingam et al., 2000). In addition to culture-based analyses, PCR-based detection and identification of B. cepacia complex species directly from patients’ sputum samples (where applicable) was also carried out in parallel, as described previously (McDowell et al., 2001).

Genotyping.

All B. cepacia complex isolates were genetically typed by RAPD, using the 10-base oligonucleotide primer 272, as described previously (Mahenthiralingam et al., 1996). RAPD fingerprints were compared visually and with the aid of computer analysis (GelCompar version 3.10 for Windows). Strains presumptively identified as ET-12 were confirmed by macrorestriction of whole genomic DNA with the restriction enzyme SpeI, followed by PFGE, as described previously (Vandamme et al., 2000). Profiles were compared with the J2315 Edinburgh index strain for the ET-12 lineage. PFGE was carried out with separation conditions of two cycles of 6 V cm^–1, with cycle 1 pulsing from 1 to 40 s over 8 h, and cycle 2 from 30 to 90 s over 16 h.

Detection of putative transmissibility markers.

PCR-based detection of the cblA gene and the BCESM were as described previously (Sajjan et al., 1995; Mahenthiralingam et al., 1997). To assess the presence and stability of the BCESM and cblA for cross-infective isolates identified by genotyping and epidemiological data, sensitivity and specificity values were determined. Sensitivity was calculated as the number of cross-infective isolates positive for each marker, expressed as a percentage of the total number of cross-infective isolates. Specificity was calculated as the number of sporadic isolates negative for each marker, expressed as a percentage of the total number of sporadic isolates.

	RESULTS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Prevalence of B. cepacia complex species

Table 1 summarizes the total distribution of B. cepacia complex species amongst isolates recovered from 131 CF patients attending a total of 14 CF centres in different regions of Europe. Using a combination of rRNA and recA- based PCR protocols, 79 % of isolates were identified as B. cenocepacia (genomovar III), while 18 % of samples were B. multivorans (genomovar II). The remaining isolates were identified as B. cepacia (genomovar I), B. stabilis (genomovar IV) or B. vietnamiensis (genomovar V). recA analysis of our B. cenocepacia isolates revealed that 91 % were B. cenocepacia recA lineage III-A, while the remaining 9 % were identified as B. cenocepacia recA lineage III-B. Within each geographical region, B. cenocepacia was also the predominant B. cepacia complex species responsible for infection amongst the patient population (Table 2).

Genotyping and putative transmissibility markers

Cross-infected and sporadic B. cepacia complex strains were identified by RAPD, and their relationship with putative transmissibility markers was examined (Table 3). Analysis of the isolates identified the same B. cenocepacia recA lineage III-A strain as the cause of infection in 58 (44 %) patients. Of these isolates, a total of 57 (98 %) contained the cblA and BCESM putative transmissibility markers, a feature characteristic of the highly transmissible ET-12 (multilocus enzyme electrophoresis type 12) lineage (RAPD group 02). One clonal isolate (2 %) was PCR-negative for the cblA gene, but did contain the BCESM. Confirmation that isolates were indeed the ET-12 strain of B. cenocepacia recA lineage III-A was achieved by macrorestriction analysis of whole genomic DNA followed by PFGE (Fig. 1). The ET-12 clonal lineage was found amongst CF patients in Northern Ireland and Great Britain. RAPD analysis of isolates from a mainland European CF centre revealed 33 patients cross-infected with a B. cenocepacia recA lineage III-A strain highly distinct from ET-12. This strain contained the BCESM, but not cblA. All remaining B. cenocepacia recA lineage III-A isolates were genotypically distinct and PCR-positive for the BCESM, but negative for cblA. All B. cenocepacia recA lineage III-B isolates had unique genetic fingerprints. None of these sporadic strains had cblA, but five did contain the BCESM. In addition to the detection of transmissible strains of B. cenocepacia, RAPD analysis also identified two separate clusters of B. multivorans cross-infection within a CF centre. One cluster involved two patients and the other a total of four patients. Both strains of B. multivorans responsible for these clusters of cross-infection were PCR-negative for cblA and the BCESM. All remaining isolates of B. multivorans, as well B. vietnamiensis, had unique genotypes and were negative for both putative transmissibility markers.

View larger version (130K): [in this window] [in a new window]   Fig. 1. Macrorestriction fingerprinting of B. cenocepacia ET-12 strains recovered from CF patients during this study. Lanes 1–4, clinical isolates of ET-12 strains; 5, J2315 Edinburgh index strain for the ET-12 lineage (Govan et al., 1993); M, molecular size markers, size of relevant bands in kb is indicated.

For cross-infective B. cenocepacia isolates, the BCESM was found to have a sensitivity of 100 % compared with 63 % for the cblA gene. For all isolates implicated in patient-to-patient spread, the BCESM had a sensitivity of 94 % compared with 59 % for the cblA gene. The specificity of the BCESM for the detection of cross-infective B. cenocepacia was 33 % compared with 100 % for cblA. The overall specificity of the BCESM for the detection of cross-infective isolates was 76 % compared with 100 % for cblA.

	DISCUSSION

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
The emergence of B. cepacia complex organisms as life-threatening pathogens amongst patients with CF has proved a tremendous challenge for the clinician. For patients colonized with bacteria of the B. cepacia complex, management of infection under the same principles adopted for patients with chronic Pseudomonas aeruginosa is complicated by the highly transmissible nature of certain B. cepacia complex strains. In response to this problem, CF centres have now adopted strict infection-control policies that segregate B. cepacia complex-infected patients from non-infected individuals. Although highly effective, stringent segregation has proved unpopular with many patients and generated new psychosocial and economic issues for CF centres. PCR-based strategies to refine the segregation of patients based on the properties of the infecting B. cepacia complex strain rather than the presence or absence of the organism have been recommended (Clode et al., 2000). In particular, for CF centres in the UK it was suggested that only transmissibility marker-positive B. cenocepacia-infected patients need to be segregated. Before such a proposal could be implemented, especially within a broader geographical context, further multi-centre studies on the distribution of B. cepacia complex species amongst CF patients in different localities, and the relationship between putative transmissibility markers and cross-infective strains, are required.

In the present study, we examined B. cepacia complex isolates recovered from CF patients in different regions of Europe. Most isolates were identified as B. cenocepacia, with the majority belonging to the recA lineage III-A. Similar results have also been described in Canada (Speert et al., 2002) and Italy (Agodi et al., 2001), while in the USA the majority of B. cenocepacia isolates belong to the recA lineage III-B (LiPuma et al., 2001). Although the basis of variations between the USA and other geographical regions in terms of the prevalence of B. cenocepacia recA lineage is currently unclear, one possibility may be the nature of the transmissible B. cenocepacia strains found in these different locations. In our study, the highly transmissible ET-12 strain was identified amongst 56 % of B. cenocepacia isolates. This strain was only recovered from patients attending CF centres in Northern Ireland and Great Britain (UK). The ET-12 strain appears to have first infected patients in Toronto before spreading across Canada, where it is the predominant B. cenocepacia lineage (Speert et al., 2002). It is presumed that the ET-12 lineage was first introduced into the UK CF population as a consequence of contact at CF summer camps, with the Edinburgh index case of this strain first observed in 1989 (Govan et al., 1993). Infection with a strain closely related to ET-12 has now also been described amongst CF patients in Italy (Agodi et al., 2001). One characteristic feature of the ET-12 strain is the presence of both cblA and the BCESM. However, during our study, we identified one ET-12 isolate that was negative for the cblA gene. Although genetic instability of the BCESM has been described (Mahenthiralingam et al., 1997; Agodi et al., 2001), instability of the cblA gene has only recently been observed (Agodi et al., 2001). Organisms of the B. cepacia complex have large and complex genomes with up to three or four replicons, as well as multiple insertion sequences (Lessie et al., 1996). The observation that cblA and the BCESM possibly reside on unstable chromosomal regions of the genome raises further important questions regarding the sole use of these putative markers for clinical management of patients within CF centres.

In addition to the detection of the highly transmissible ET-12 strain, we also identified a strain of B. cenocepacia recA lineage III-A which was responsible for multiple cross-infections amongst patients attending a CF centre in mainland Europe. This strain, which was genotypically distinct from ET-12, lacked the cblA gene but did contain the BCESM. Cross-infection with BCESM-positive B. cenocepacia RAPD types other than ET-12 has been observed before, such as the epidemic strain of B. cenocepacia responsible for transmission between CF patients in Manchester, UK (Haworth et al., 1997). Studies have now also identified cross-infection amongst CF patients due to cblA- and BCESM-negative B. cenocepacia strains, such as the epidemic PHDC strain which infects CF patients in the mid-Atlantic region of the USA (Chen et al., 2001). Such studies further highlight the caveats associated with sole reliance on these markers for identification, and infection-control management, of CF patients infected with potentially cross-infective B. cenocepacia strains.

Within our CF study population we also observed clusters of cross-infection amongst patients due to B. multivorans. None of these strains carried the cblA and BCESM putative transmissibility markers. As our understanding of B. cepacia complex infection has continued to grow, it has become clear that serious episodes of cross-infection due to B. cepacia complex species other than B. cenocepacia can arise. Cross-infection due to B. multivorans has been reported amongst CF patients in the UK and France (Whiteford et al., 1995; Segonds et al., 1999). Also, the first epidemic strain identified as the cause of patient-to-patient spread of B. cepacia has now been identified as Burkholderia dolosa (genomovar VI) (LiPuma et al., 2001). Very recently, reports of potential cross-infection with B. stabilis amongst 13 CF patients in the Slovak Republic, and the demonstration of transmission of B. cepacia (genomovar I), B. multivorans and B. dolosa, provides striking evidence of the danger in confining segregation to B. cenocepacia-infected patients only (Biddick et al., 2003; Drevinek et al., 2003).

Overall, our sensitivity and specificity calculations revealed that neither the BCESM nor cblA were absolute markers for spread amongst CF patients. The BCESM was present in sporadic isolates of B. cenocepacia, and absent from cross-infective B. multivorans isolates, while cblA was only found in the transmissible ET-12 strain of B. cenocepacia. However, previous studies have identified the presence of cblA gene sequences in a small number of B. cepacia (genomovar I) isolates recovered from CF patients in the USA (LiPuma et al., 2001). The majority of these isolates were genotypically distinct and had sequences with high identity (98.2–100 %) to each other, but lower identity (87.8–88.4 %) to the ET-12 cblA sequence (Sajjan et al., 2002). Interestingly, electron microscopy revealed the complete absence of pilus expression by these isolates. Therefore, while variant cblA-containing sequences have been found amongst a small number of sporadic non-ET-12 strains, the gene appears to be relatively specific for the ET-12 strain of B. cenocepacia and not for transmissible B. cepacia complex strains in general.

In conclusion, our data demonstrate the prevalence of B. cenocepacia amongst B. cepacia complex-infected CF patients. Cross-infection was not solely confined to B. cenocepacia, and cblA and the BCESM were not absolute markers for transmissible strains. We would therefore recommend that CF centres cohort patients based on the presence or absence of B. cepacia complex infection, and not on the basis of transmissibility marker-positive B. cenocepacia. This is in agreement with the UK CF Trust's guidelines on cross-infection control (http://www.cftrust.org.uk/index.jsp). Further segregation of patients infected with B. multivorans (and other B. cepacia complex species) from those infected with B. cenocepacia is also prudent to reduce the risk of cross-infection and superinfection (Mahenthiralingam et al., 2001).

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
This work was supported by grants (PJ470 and PJ472) from the Cystic Fibrosis Trust, Bromley, UK. We would also like to acknowledge the various CF centres involved in this study for their cooperation and submission of B. cepacia complex isolates to our laboratory.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES