Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Microarray-Based Genotyping and Clinical Outcomes of Staphylococcus aureus Bloodstream Infection: An Exploratory Study

  • Siegbert Rieg ,

    siegbert.rieg@uniklinik-freiburg.de

    Affiliation Department of Medicine, Center for Infectious Diseases and Travel Medicine and IFB-Center for Chronic Immunodeficiency, University Medical Center Freiburg, Freiburg, Germany

  • Daniel Jonas,

    Affiliation Department of Environmental Health Sciences, University Medical Center Freiburg, Freiburg, Germany

  • Achim J. Kaasch,

    Affiliation Institute for Medical Microbiology, Immunology and Hygiene, University of Cologne, Cologne, Germany

  • Christine Porzelius,

    Affiliations Institute of Medical Biometry and Medical Informatics, University Medical Center Freiburg, Freiburg, Germany, Institute for Medical Informatics, Statistics and Epidemiology, University of Leipzig, Leipzig, Germany

  • Gabriele Peyerl-Hoffmann,

    Affiliation Department of Medicine, Center for Infectious Diseases and Travel Medicine and IFB-Center for Chronic Immunodeficiency, University Medical Center Freiburg, Freiburg, Germany

  • Christian Theilacker,

    Affiliation Department of Medicine, Center for Infectious Diseases and Travel Medicine and IFB-Center for Chronic Immunodeficiency, University Medical Center Freiburg, Freiburg, Germany

  • Marc-Fabian Küpper,

    Affiliation Department of Medicine, Center for Infectious Diseases and Travel Medicine and IFB-Center for Chronic Immunodeficiency, University Medical Center Freiburg, Freiburg, Germany

  • Christian Schneider,

    Affiliation Institute of Medical Microbiology and Hygiene, University Medical Center Freiburg, Freiburg, Germany

  • Harald Seifert,

    Affiliation Institute for Medical Microbiology, Immunology and Hygiene, University of Cologne, Cologne, Germany

  • Winfried V. Kern

    Affiliation Department of Medicine, Center for Infectious Diseases and Travel Medicine and IFB-Center for Chronic Immunodeficiency, University Medical Center Freiburg, Freiburg, Germany

Microarray-Based Genotyping and Clinical Outcomes of Staphylococcus aureus Bloodstream Infection: An Exploratory Study

  • Siegbert Rieg, 
  • Daniel Jonas, 
  • Achim J. Kaasch, 
  • Christine Porzelius, 
  • Gabriele Peyerl-Hoffmann, 
  • Christian Theilacker, 
  • Marc-Fabian Küpper, 
  • Christian Schneider, 
  • Harald Seifert, 
  • Winfried V. Kern
PLOS
x

Abstract

The clinical course of Staphylococcus aureus bacteremia varies extensively. We sought to determine the relationship between genetic characteristics of the infecting pathogen and clinical outcomes in an exploratory study. In two study centers, 317 blood culture isolates were analyzed by DNA microarray and spa genotyping. By uni- and multivariate regression analyses associations of genotype data with 30-day all-cause mortality, severe sepsis/septic shock, disseminated disease, endocarditis, and osteoarticular infection were investigated. Univariate analysis showed significant association between S. aureus genes/gene-clusters or clonal complexes and clinical endpoints. For example CC15 was associated with 30-day mortality and CC22 with osteoarticular infection. In multivariate analysis methicillin resistance (mecA, OR 4.8 [1.43–16.06]) and the beta-lactamase-gene (bla, OR 3.12 [1.17–8.30]) remained independently associated with 30-day mortality. The presence of genes for enterotoxins (sed/sej/ser) was associated with endocarditis (OR 5.11 [1.14–18.62]). Host factors such as McCabe classification (OR 4.52 [2.09–9.79] for mortality), age (OR 1.06 [1.03–1.10] per year), and community-acquisition (OR 3.40 [1.31–8.81]) had a major influence on disease severity, dissemination and mortality. Individual genotypes and clonal complexes of S. aureus can only partially explain clinical features and outcomes of S. aureus bacteremia. Genotype-phenotype association studies need to include adjustments for host factors like age, comorbidity and community-acquisition.

Introduction

Staphylococcus aureus remains a major human pathogen that is able to cause a wide spectrum of clinical manifestations ranging from asymptomatic carriage to localized or disseminated infections, such as skin and soft tissue infections (SSTI), bone and joint infections, and infective endocarditis [1].

S. aureus is equipped with a broad range of virulence factors including exoproteins and proteases, superantigens and toxins, microbial surface components recognizing adhesive matrix molecules (MSCRAMMs) and adhesins. Moreover, S. aureus is capable to circumvent innate and adaptive host defences by multiple immune evasion strategies and to withstand antibiotics due to acquisition of diverse resistance mechanisms [2]. The question whether particular virulence genes, resistance determinants, or specific clonal lineages of S. aureus predispose to certain clinical manifestations and are associated with poor outcome has raised substantial interest [3].

A range of studies identified genotypic S. aureus factors likely to contribute to invasive disease by comparing S. aureus isolates recovered from the anterior nares of healthy carriers with isolates obtained from patients with different clinical disease manifestations [4][10]. In contrast, studies comparing the genetic determinants of isolates from patients with different clinical disease manifestations and/or poor versus good outcomes have been rare and hampered by small isolate numbers [11], non-inclusion of MSSA [12], paucity of clinical data [13] and few genes investigated [14]. Using microarray techniques, the presence of several hundred genes within an isolate can now be readily determined, thus enabling a comprehensive study of putative S. aureus virulence and resistance determinants [15].

The aim of the current study was to explore whether the genetic makeup of S. aureus can be linked with mortality, infection severity, metastatic disease, and tropism for endovascular and osteoarticular sites of infection. For this purpose we used clinical data and isolates from a large prospective cohort study of SAB at two German university hospitals.

Patients and Methods

Patients

The study was a non-interventional observational study conducted at the University Medical Centers Freiburg (1.500-bed tertiary care center) and Cologne (1.300 bed tertiary care center) within the framework of an ongoing quality assurance programme at both institutions. Blood cultures growing S. aureus were reported daily by the microbiology laboratory, and adult patients were assessed by infectious disease physicians onsite who recommended intensified diagnostic studies and optimized treatment if needed. The patients were usually followed until discharge from the hospital. For each case relevant clinical and microbiological data were entered into a database that included age, sex, underlying medical conditions, clinical signs and symptoms at SAB onset, diagnostic and therapeutic procedures, antimicrobial therapy, complications and outcomes. Outcomes after discharge were routinely assessed by active case finding, by contact with the primary care physician, or by assessing the patient in the infectious diseases outpatient clinics.

For the present analysis we retrospectively chose every second case seen between January 2006 and May 2010 for which clinical data was complete and the bacterial isolate was available for further study. The subgroup of SAB cases included in the study was representative of the complete cohort with respect to underlying diseases, proportions of MRSA isolates, endocarditis, disseminated disease as well as 30-day all-cause mortality (File S1). The study and data collection were approved by the Institutional Review Boards of the University Medical Centers Freiburg and Cologne. Written informed consent was obtained from the patients at the University Medical Center Cologne. The Institutional Review Board of the University Medical Center Freiburg considered the investigation as evaluation of service within a quality assurance programme and waived the need for written informed consent.

Clinical definitions

Mode of acquisition (community-acquired, community-onset healthcare-associated or hospital-acquired SAB) was defined as previously described [16]. Severity of illness was assessed by the McCabe-Jackson classification. Intravascular catheter/device-related SAB was considered when clinical signs of catheter/device-infection and/or positive culture results of catheter material were present without evidence of an alternative source of SAB. Infective endocarditis was diagnosed according to the modified Duke criteria. Metastatic lesions were actively sought by clinical examination and imaging studies. Severe sepsis and septic shock were defined according to definitions of the ACCP/SCCM consensus conference and was considered an SAB-related outcome when present within 48 h before and 72 h after the first positive blood culture was drawn. Osteoarticular infection (defined as SAB with vertebral/non-vertebral osteomyelitis or septic arthritis) was considered as documented focus if there were suggestive clinical and pathological findings, evidence from imaging studies and if S. aureus was grown from a joint aspirate or osteoarticular biopsy sample. Disseminated disease was defined as SAB with at least one other hematogenous metastatic manifestation distant from the primary focus. 30-day all-cause mortality was included as outcome.

spa typing

spa typing was performed according to standard protocols. Cycle sequencing products were separated on an ABI310 Genetic Analyser. Sequence data were analyzed and assigned to spa-types by use of the Ridom Staph Type (version 2.0.3) software (Ridom Bioinformatics Münster, Germany).

Microarray Typing

Microarray-based DNA genotyping (StaphType, Alere, Jena, Germany) was performed as previously described [17]. The array covers 334 target sequences corresponding to 185 distinct genes and their allelic variants and detects a wide range of genetic determinants including virulence and resistance genes, genes encoding exotoxins, superantigens, and MSCRAMMs as well as SCCmec, capsule and agr group typing makers [18]. Data interpretation, threshold definition and categorization was performed as described [Monecke FEMS Immunol Med Microbiol [15]. Results categorized as ambiguous were counted as negative. On the basis of hybridisation patterns and by comparing to a database of reference strains isolates were assigned to clonal complexes (CC) using MLST-based nomenclature [15].

Statistical Methods

All statistical analyses were performed using R version 2.13.1 [R foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/]. In univariate analyses, Fisheŕs exact test was used for binary variables (including demographic and clinical data, individual genotypes, clonal complex assignments), and a χ2-test was used for agr types.

In multivariate analyses, a two-step procedure was applied for each endpoint. In the first step, logistic regression models were fitted for each covariate separately including agr type, the most prevalent clonal complexes, and the individual multiarray-defined genetic resistance determinants and pathogenicity factors excluding low (<5%) and high (>95%) prevalent genes. For this, agr type and clonal complex were taken as binary covariate each. Furthermore, age, study center, sex, methicillin resistance, McCabe-Jackson classification (non-fatal vs. ultimately/rapidly fatal) and mode of acquisition (dummy coded) were included as mandatory covariates. In the second step, covariates with p values≤0.05 were included in a multivariate logistic regression model including the same mandatory covariates as indicated above, allowing to assess the additional value of the covariates of interest beyond readily available clinical variables. For covariates highly correlated with each other (such as individual genes linked in gene clusters), representatives to be included in the analysis were selected by hand.

P values<0.05 were considered to be statistically significant. As this investigation had an explorative character, no adjustment for multiple testing was done.

Results

Patient Characteristics

A total of 317 isolates from SAB cohort patients were included in the study. The patient characteristics are summarized in table 1. The median age was 65 years, 69% were male, and 63% had an underlying disease classified as non-fatal. Almost half of the SAB episodes were hospital-acquired, 24% were community-onset healthcare-associated, and 27% were community-acquired. In 45% the primary focus was an intravascular catheter/device. An osteoarticular focus was identified (vertebral/non-vertebral osteomyelitis or septic arthritis) in 16%. Other primary foci such as respiratory tract or cardiovascular infection or skin and soft tissue infection were rare (<10% each). In 11% a primary focus could not be identified. The patients of the two centers had similar age, gender ratio, McCabe classification, SAB primary focus distribution, and outcomes. Endstage renal disease was more prevalent in one center (4% vs. 16%, p<0.01).

thumbnail
Table 1. Patient characteristics and demographic data of 317 patients with SAB.

https://doi.org/10.1371/journal.pone.0071259.t001

Microarray and Spa Typing Results

Forty-nine of the genes detected in the DNA microarray were highly prevalent (≥95% of isolates, 55 genes in ≥90% of isolates) whereas 59 genes were detected in ≤5% of isolates (80 genes in ≤10% of isolates). Thirty-two isolates (10%) were methicillin-resistant (mecA positive), 29 of them originated from community-onset healthcare-associated or hospital-acquired SAB. 163 isolates (51%) belonged to agr group I, 94 isolates (30%) to agr group II, 49 isolates (15%) were agr group III and 7 isolates (2%) were agr group IV. Four isolates could not be classified. The prevalence of other resistance and major virulence genes is summarized in table 2. The complete hybridisation profiles for the individual strains are provided in File S2.

thumbnail
Table 2. Prevalence of major virulence factors and resistance determinants in 317 SAB isolates.

https://doi.org/10.1371/journal.pone.0071259.t002

Based on the microarray genotyping 307 isolates were assigned to 21 different clonal complexes. Seven clonal complexes comprised 74% of all isolates. The most prevalent clonal complex was CC5 (50 isolates, 16%), followed by CC45 (46 isolates, 15%), CC30 (39 isolates, 12%), CC15 (28 isolates, 9%), CC7 (27 isolates, 9%), CC8 (24 isolates, 8%) and CC22 (22 isolates, 7%) (table 2).

spa typing revealed 157 different spa types among 316 typeable isolates and showed a substantial biodiversity within each of the clonal lineages and no overlap between microarray-derived clonal lineages (File S3). Only seven spa types were detected 10 times or more often, and 121 isolates represented single spa types.

Association between Microarray Genotypes and Mortality and Disease Severity

Univariate analyses were performed to evaluate a potential association of the seven most prevalent clonal lineages or agr types with different clinical outcomes. CC 15 was found to be associated with a higher mortality at day 30 (36% vs. 17%, p = 0.017), a tendency for more frequent septic shock or severe sepsis/septic shock (not significant), and a lower prevalence of disseminated disease (7% vs. 27%, p = 0.022) compared to non-CC15 isolates. Next, a univariate analysis of possible associations of 185 prevalent virulence and resistance genes was performed. An association with day-30 mortality (p<0.05) was found for mecA, bla, ermA, ermC, aadD, sed/sej/ser, and ssl11. Genes identified to be associated with severe sepsis or septic shock in univariate analysis were aadD, fosB, sed/sej/ser. The genes lukE, sak, chp, fib were found to be associated with disseminated disease.

Association between Microarray Genotypes and Infection Site (Endovascular or Osteoarticular)

Univariate analyses revealed an association between CC22 versus non-CC22 isolates and osteoarticular infection (36% vs. 14%, p = 0.015) but no other significant associations between clonal complex or agr type and infection site as investigated here. Individual genes associated with osteoarticular infections were egc gene cluster, lukD, splA/B, ssl3, ssl8, cna, ebh. There was no significant association in univariate analysis between individual genes and gene clusters with endocarditis. Complete data of univariate analyses are provided in File S4.

Association of Genotypes versus Host Factors with Clinical Outcomes

The following determinants were included in multivariate regression analyses: S. aureus clonal complex (7 most prevalent clonal complexes), agr type, microarray-defined virulence and resistance genes or gene clusters, mode of acquisition of SAB, study center, and age, sex and McCabe classification as host factors (table 3).

thumbnail
Table 3. Multivariate logistic regression analyses for endpoints 30-day mortality, severe sepsis or septic shock and disseminated disease.

https://doi.org/10.1371/journal.pone.0071259.t003

30-day all-cause mortality.

age (OR 1.06, 95% CI 1.03–1.1), McCabe classification (OR 4.5, 95% CI 2.1–9.8) and mode of acquisition (OR 3.4, 95% CI 1.3–8.8, for community-acquired SAB and OR 3.7, 95% CI 1.6–8.5, for community-onset healthcare-associated SAB) were factors found to significantly contribute to day-30 mortality, in addition to the pathogen factors mecA (OR 4.8, 95% CI 1.4–16.1), bla encoding for beta-lactamase (OR 3.1, 95% CI 1.2–8.3) and ermC encoding macrolide, lincosamide and streptogramin resistance (OR 4.6, 95% CI 1.3–16.4). A trend towards a higher day-30 mortality was noted for CC15 isolates (OR 2.7, 95% CI 0.96–7.4), whereas CC5 tended to be associated with lower day-30 mortality (OR 0.4, 95% CI 0.1–1.4).

Disease severity.

The development of severe sepsis/septic shock was associated with age (OR 1.02, 95% CI 1.0–1.04) and mode of acquisition (OR 5.0, 95% CI 2.6–9.6, for community-acquired SAB, and OR 3.0, 95% CI 1.6–5.7, for community-onset healthcare-associated SAB).

Disseminated disease.

Community acquisition (OR 5.2, 95% CI 2.6–10.6) and community-onset healthcare-associated SAB (OR 2.7, 95% CI 1.3–5.6), and study center (OR 0.5, 95% CI 0.3–0.9) but none of the pathogen factors were found to be associated with disseminated disease.

Infective endocarditis.

Mode of acquisition (OR 7.1, 95% CI 2.5–19.7, for community-acquired SAB and OR 3.6, 95% CI 1.2–10.7, for community-onset healthcare-associated SAB) was also associated with endocarditis (table 4). Independent pathogen factors identified here were the genes sed, sej, ser (located on the same mobile genetic element) coding for enterotoxin D, J and R (OR 5.1, 95% CI 1.4–18.6) which had not been identified by univariate analysis.

thumbnail
Table 4. Multivariate logistic regression analysis for endpoint infective endocarditis.

https://doi.org/10.1371/journal.pone.0071259.t004

Osteoarticular infection.

Osteoarticular infection was associated with 24 covariates at a p-value<0.05. Due to this large number and high intercorrelations, a multivariate regression model could not be built. In the basic model with methicillin resistance, age, study center, sex, comorbidity and mode of acquisition, only the latter showed a significant effect (community-acquired SAB, OR 16.4, 95% CI 5.8–46.2; and community-onset healthcare-associated SAB, OR 7.9, 95% CI 2.7–22.7). Table 5 lists all covariates with a p value<0.05 when included separately in this basic model (step 1 of the above described procedure). CC22 (OR 4.6, 95% CI 1.5–14.1), presence of enterotoxins G/I/M/N/O/U (egc gene cluster, OR 2.4, 95% CI 1.2–5.0) and collagen-binding adhesin (cna, OR 2.6, 95% CI 1.3–5.2) were factors associated with osteoarticular infections, whereas a range of other factors was negatively associated with osteoarticular infections, notably splA and splB encoding serine proteases A and B, ssl3 and ssl8 encoding superantigen-like proteins, and cell wall-associated fibronectin-binding protein (ebh) among others (table 5).

thumbnail
Table 5. Logistic regression analysis for endpoint osteoarticular infection adjusted for age, study center, sex, comorbidity, mode of acquisition and methicillin resistance.

https://doi.org/10.1371/journal.pone.0071259.t005

Discussion

Studies investigating associations between genetic traits of S. aureus and clinical manifestations and outcomes are challenging and have yielded conflicting results for several reasons. First, S. aureus isolates within a clonal lineage possess a consensus repertoire of virulence genes with limited variation due to mobile genetic elements. Thus, studies on the association of virulence genes with invasiveness or other endpoints can be biased by an uneven distribution of clonal complexes between the two groups investigated (hitchhiker effect) [6], [7], [13]. Second, grouping of invasive isolates from a variety of infections such as skin infection, pneumonia, joint infection and bacteremia may have hampered some studies [4][7], [9], [10], [19]. Finally, host factors such as underlying disease [20] and age [21], have a major impact on clinical manifestation and outcome of S. aureus disease and need to be considered as possible confounders. The present study aimed to account for these issues and found few associations between clonal complex or specific pathogen genes and outcome.

Similar to other studies from other areas (UK, Netherlands, USA) we found isolates of CC5, CC45, CC30, CC15 and CC8 to be among the most prevalent to cause invasive disease [4], [22], [23]. Several groups have linked CC30 isolates to more severe disease, i.e. hematogenous complications [22], infective endocarditis [24], persistent bacteremia [12], or invasive disease per se [4], [10]. In contrast to these studies we did not observe a greater pathogenic potential of CC30 isolates.

In addition, there was no evidence in our cohort of an association of CC5 isolates with disseminated disease as previously described in an investigation that compared carriage and uncomplicated infections [22]. Interestingly, in our study CC5 was associated with lower 30-day all-cause mortality while for CC15, there was a tendency for higher mortality, independent of antimicrobial resistance determinants.

The most significant finding regarding clonal lineages was the association between CC22 and osteoarticular infection. However, it proved difficult to confirm this to be an independent association since a meaningful multivariate analysis could not be performed due to too many intercorrelating factors. An association between CC22 and osteoarticular infection has not been described previously [8], [25]. Also, cna as a potential factor for osteoarticular infection has not been identified in earlier studies.

Several of the genes found associated with selected clinical endpoints in this study (e.g. cna, egc gene cluster, sea) have been described by others to be associated with invasive, disseminated or severe disease [9], [13], [26]. Some of them remained significant in the current study if tested in a multivariate analysis that included further genetic determinants and patient factors. Antimicrobial resistance determinants rather than virulence factors were linked to poor survival which in the case of mecA corroborates previous findings with phenotypic methicillin resistance [16], [26] but which has not been described so far for bla (prevalence 71%) and ermC (prevalence 6%). The significance of these two associations remains unclear. In line with findings of a French study we could not determine any of the S. aureus superantigens to be associated with development of severe sepsis or septic shock [19].

In a recent multinational cohort study that compared infective endocarditis (IE) with SSTI isolates, IE isolates were more likely to be positive for adhesin genes sdrC, cna, map/eap and for genes tst, sea, sed, see and sei that code for toxins [24]. Using a different study design by comparing SAB isolates with vs. without IE we also found the presence of the enterotoxin gene sed (accompanied by sej and ser) to be independently associated with IE. As a plausible pathophysiological hypothesis is not imminent so far, a linkage of sed to yet unrecognized virulence genes deserves further consideration [24].

The relative paucity of individual pathogen factors that were associated with the investigated endpoints is contrasted by the fact that several host factors like age, comorbidity and mode of acquisition substantially impacted on SAB outcome and manifestation. The identification of mode of acquisition (particularly community-acquisition) as strong independent factor for each of the five endpoints is consistent with results of several previous studies [27], [28]. The observed influence may partly be a consequence of a longer “incubation period” with the longer presence of S. aureus in the bloodstream leading to different infective foci, such as endocarditis. Apart from host and pathogen factors, clinical management was shown to impact SAB outcome [16]. In our study, time from initial symptoms to initiation of adequate antimicrobial therapy could not be included as an additional variable in the multivariate model, as it was highly correlated with community-acquisition of SAB. Excluding the 22 cases with initiation of adequate antibiotic therapy of ≥3 days after SAB onset revealed no changes of factors contributing to the endpoints endocarditis, severe sepsis/septic shock, disseminated disease and osteoarticular infections and only very minor changes to the endpoint 30-day all-cause mortality.

Several conclusions may be drawn from our investigation. Studies on associations of genetic determinants with S. aureus manifestations and outcome need to be adjusted for host characteristics. In fact, our study revealed that most of the genetic determinants identified by univariate analysis were no longer significant after adjustment for host characteristics. Notably, age and mode of acquisition exhibited the strongest impact on clinical outcomes. In view of the present and previous reports [4], [7], [14] it appears that single virulence factors as currently determined by DNA microarray tests cannot sufficiently explain the invasive and pathogenic potential of S. aureus nor the clinical phenotype of S. aureus infections. Unknown genes [5], gene expression, and gene regulation may be more relevant for SAB outcome [3], [29].

There are limitations of the present study. As our investigation had an exploratory character, our results have to be validated in other cohorts. Only genes present on the microarray were investigated, yet other genes may be of major importance for the outcome of SAB. The study was conducted at only two tertiary care centers in a single country. Although the distribution of clonal complexes was similar to that described in a recent multinational European study of invasive S. aureus infections [30], our results can only be generalized after confirmation in other large SAB cohorts in geographically distant regions. Though one of the largest studies in this field, a larger sample might be needed to be able to demonstrate relatively modest effects (odds ratio of <3) of bacterial genotypes, therefore collaborative cohorts or multinational consortia may be of importance to address these issues in future.

In summary, we found only a limited number of microbial genetic determinants to be independently associated with the investigated clinical endpoints of SAB. Conversely, host factors such as age, comorbidity and community-acquisition of SAB revealed a substantial independent impact on the course and outcome of SAB and need to be accounted for as effect modifiers in genotypic association studies. Clinical awareness for host factors that predict a poor SAB outcome is warranted. Finally, the study indicates the need to intensify research on immunological host factors underlying intermittent and persistent S. aureus colonization and the susceptibility to severe, invasive and systemic S. aureus infections.

Supporting Information

File S1.

Patient characteristics and demographic data of all SAB patients within the INvasive STapylococcus aureus INfections CohorT (INSTINCT) study and the subgroup of SAB cases included within this study.

https://doi.org/10.1371/journal.pone.0071259.s001

(PDF)

File S2.

Complete hybridisation results for S. aureus strains examined in this study. Data are ordered by the assigned clonal complex and the strain designation.

https://doi.org/10.1371/journal.pone.0071259.s002

(PDF)

File S3.

Distribution of clonal complexes and spa types in 317 patients with SAB.

https://doi.org/10.1371/journal.pone.0071259.s003

(PDF)

File S4.

Complete data of univariate analyses of pathogen factors (clonal complex, agr type and microarray-derived virulence and resistance genes) associated with investigated clinical endpoints.

https://doi.org/10.1371/journal.pone.0071259.s004

(PDF)

Acknowledgments

We thank Hanna Birkholz, Andreas Langhorst, Katharina Achilles, Georg Peppinghaus, and Stephan Neumann for collection of clinical data and Raffaele de Luca and Christa Hauser for expert technical assistance.

Author Contributions

Conceived and designed the experiments: SR DJ AJK CP HS WVK. Analyzed the data: SR DJ AJK CP HS WVK. Wrote the paper: SR DJ AJK CP HS WVK. Clinical care and infectious diseases consultation: SR AJK CT MFK HS WVK. Microbiological and array procedures: DJ AJK GPH CS HS. Data management: SR DJ AJK CP GPH.

References

  1. 1. Kern WV (2010) Management of Staphylococcus aureus bacteremia and endocarditis: progresses and challenges. Curr Opin Infect Dis 23: 346–358.
  2. 2. Chambers HF, DeLeo FR (2009) Waves of resistance: Staphylococcus aureus in the antibiotic era. Nat Rev Microbiol 7: 629–641.
  3. 3. Priest NK, Rudkin JK, Feil EJ, van den Elsen JM, Cheung A, et al. (2012) From genotype to phenotype: can systems biology be used to predict Staphylococcus aureus virulence? Nat Rev Microbiol 10: 791–797.
  4. 4. Feil EJ, Cooper JE, Grundmann H, Robinson DA, Enright MC, et al. (2003) How clonal is Staphylococcus aureus? J Bacteriol 185: 3307–3316.
  5. 5. Gill SR, McIntyre LM, Nelson CL, Remortel B, Rude T, et al. (2011) Potential associations between severity of infection and the presence of virulence-associated genes in clinical strains of Staphylococcus aureus. PLoS One 6: e18673.
  6. 6. Holtfreter S, Grumann D, Schmudde M, Nguyen HT, Eichler P, et al. (2007) Clonal distribution of superantigen genes in clinical Staphylococcus aureus isolates. J Clin Microbiol 45: 2669–2680.
  7. 7. Lindsay JA, Moore CE, Day NP, Peacock SJ, Witney AA, et al. (2006) Microarrays reveal that each of the ten dominant lineages of Staphylococcus aureus has a unique combination of surface-associated and regulatory genes. J Bacteriol 188: 669–676.
  8. 8. Luedicke C, Slickers P, Ehricht R, Monecke S (2010) Molecular fingerprinting of Staphylococcus aureus from bone and joint infections. Eur J Clin Microbiol Infect Dis 29: 457–463.
  9. 9. Wehrhahn MC, Robinson JO, Pascoe EM, Coombs GW, Pearson JC, et al. (2012) Illness severity in community-onset invasive Staphylococcus aureus infection and the presence of virulence genes. J Infect Dis 205: 1840–1848.
  10. 10. Wertheim HF, van Leeuwen WB, Snijders S, Vos MC, Voss A, et al. (2005) Associations between Staphylococcus aureus Genotype, Infection, and In-Hospital Mortality: A Nested Case-Control Study. J Infect Dis 192: 1196–1200.
  11. 11. Desachy A, Lina G, Vignon P, Hashemzadeh A, Denis F, et al. (2007) Role of superantigenic strains in the prognosis of community-acquired methicillin-susceptible Staphylococcus aureus bacteraemia. Clin Microbiol Infect 13: 1131–1133 CLM1810 [pii];10.1111/j.1469-0691.2007.01810.x [doi].
  12. 12. Xiong YQ, Fowler VG, Yeaman MR, Perdreau-Remington F, Kreiswirth BN, et al. (2009) Phenotypic and genotypic characteristics of persistent methicillin-resistant Staphylococcus aureus bacteremia in vitro and in an experimental endocarditis model. J Infect Dis 199: 201–208.
  13. 13. Peacock SJ, Moore CE, Justice A, Kantzanou M, Story L, et al. (2002) Virulent combinations of adhesin and toxin genes in natural populations of Staphylococcus aureus. Infect Immun 70: 4987–4996.
  14. 14. Chi CY, Wang SM, Lin CC, Liu CC (2010) Microbiological characteristics of community-associated Staphylococcus aureus causing uncomplicated bacteremia and infective endocarditis. J Clin Microbiol 48: 292–294.
  15. 15. Monecke S, Slickers P, Ehricht R (2008) Assignment of Staphylococcus aureus isolates to clonal complexes based on microarray analysis and pattern recognition. FEMS Immunol Med Microbiol 53: 237–251.
  16. 16. Rieg S, Peyerl-Hoffmann G, de With K, Theilacker C, Wagner D, et al. (2009) Mortality of S. aureus bacteremia and infectious diseases specialist consultation–a study of 521 patients in Germany. J Infect 59: 232–239.
  17. 17. Monecke S, Jatzwauk L, Weber S, Slickers P, Ehricht R (2008) DNA microarray-based genotyping of methicillin-resistant Staphylococcus aureus strains from Eastern Saxony. Clin Microbiol Infect 14: 534–545.
  18. 18. Monecke S, Luedicke C, Slickers P, Ehricht R (2009) Molecular epidemiology of Staphylococcus aureus in asymptomatic carriers. Eur J Clin Microbiol Infect Dis 28: 1159–1165.
  19. 19. Ferry T, Thomas D, Genestier AL, Bes M, Lina G, et al. (2005) Comparative prevalence of superantigen genes in Staphylococcus aureus isolates causing sepsis with and without septic shock. Clin Infect Dis 41: 771–777.
  20. 20. Kaasch AJ, Rieg S, Neumann S, Seifert H, Kern WV (2011) Measuring mortality in Staphylococcus aureus bloodstream infections: are 3 months of follow-up enough? Infection 39: 281–282 10.1007/s15010-011-0104-x [doi].
  21. 21. van Hal SJ, Jensen SO, Vaska VL, Espedido BA, Paterson DL, et al. (2012) Predictors of mortality in Staphylococcus aureus Bacteremia. Clin Microbiol Rev 25: 362–386.
  22. 22. Fowler VG Jr, Nelson CL, McIntyre LM, Kreiswirth BN, Monk A, et al. (2007) Potential associations between hematogenous complications and bacterial genotype in Staphylococcus aureus infection. J Infect Dis 196: 738–747.
  23. 23. Melles DC, Gorkink RF, Boelens HA, Snijders SV, Peeters JK, et al. (2004) Natural population dynamics and expansion of pathogenic clones of Staphylococcus aureus. J Clin Invest 114: 1732–1740.
  24. 24. Nienaber JJ, Sharma Kuinkel BK, Clarke-Pearson M, Lamlertthon S, Park L, et al. (2011) Methicillin-susceptible Staphylococcus aureus endocarditis isolates are associated with clonal complex 30 genotype and a distinct repertoire of enterotoxins and adhesins. J Infect Dis 204: 704–713.
  25. 25. Kechrid A, Perez-Vazquez M, Smaoui H, Hariga D, Rodriguez-Banos M, et al. (2011) Molecular analysis of community-acquired methicillin-susceptible and resistant Staphylococcus aureus isolates recovered from bacteraemic and osteomyelitis infections in children from Tunisia. Clin Microbiol Infect 17: 1020–1026.
  26. 26. Fowler VG Jr, Justice A, Moore C, Benjamin DK Jr, Woods CW, et al. (2005) Risk factors for hematogenous complications of intravascular catheter-associated Staphylococcus aureus bacteremia. Clin Infect Dis 40: 695–703.
  27. 27. Fowler VG Jr, Olsen MK, Corey GR, Woods CW, Cabell CH, et al. (2003) Clinical identifiers of complicated Staphylococcus aureus bacteremia. Arch Intern Med 163: 2066–2072.
  28. 28. Kaech C, Elzi L, Sendi P, Frei R, Laifer G, et al. (2006) Course and outcome of Staphylococcus aureus bacteraemia: a retrospective analysis of 308 episodes in a Swiss tertiary-care centre. Clin Microbiol Infect 12: 345–352.
  29. 29. Ziebandt AK, Kusch H, Degner M, Jaglitz S, Sibbald MJ, et al. (2010) Proteomics uncovers extreme heterogeneity in the Staphylococcus aureus exoproteome due to genomic plasticity and variant gene regulation. Proteomics 10: 1634–1644.
  30. 30. Grundmann H, Aanensen DM, van den Wijngaard CC, Spratt BG, Harmsen D, et al. (2010) Geographic distribution of Staphylococcus aureus causing invasive infections in Europe: a molecular-epidemiological analysis. PLoS Med 7: e1000215.