Potential Associations between Hematogenous Complications and Bacterial Genotype in Staphylococcus aureus Infection

Vance G. Fowler 2 Jr. 0 2 3 Charlotte L. Nelson 0 2 Lauren M. McIntyre 2 7 Barry N. Kreiswirth 2 6 Alastair Monk 2 5 Gordon L. Archer 2 5 Jerome Federspiel 2 3 Steven Naidich 2 4 Brian Remortel 2 8 Thomas Rude 2 3 Pamela Brown 2 3 L. Barth Reller 1 2 3 G. Ralph Corey 0 2 3 Steven R. Gill 2 9 0 Duke Clinical Research Institute, Duke University Medical Center , Durham, North Carolina 1 Clinical Microbiology Laboratory 2 Received 3 October 2006; accepted 21 March 2007; electronically published 20 July 2007. Potential conflicts of interest: V.G.F. has received grant funding from Cubist, Inhibitex, Merck, Nabi, and Theravance; is a consultant for Biosynexus, Cerexa, Cubist, Inhibitex, and Merck; and is on the speakers' bureau of Pfizer and Cubist. G.R.C. has received grant funding from Inhibitex, Merck, and Theravance and is a consultant for Cubist, Cerexa, and Inhibitex. All other authors report no potential conflicts. Financial support: National Institutes of Health (grant R01AI059111 to V.G.F.). Diseases, Duke University Medical Center , Durham, NC 27710 3 Division of Infectious Diseases, Department of Medicine 4 eGenomics , Inc., New York 5 Virginia Commonwealth University Medical Center , Richmond 6 Public Health Research Institute , Newark, New Jersey 7 University of Florida , Gainesville 8 Institute for Genomic Research , Bethesda, Maryland 9 State University of New York , Buffalo Background. The impact of bacterial clonality on infections caused by Staphylococcus aureus is unclear. Methods. Three hundred seventy-nine S. aureus isolates (125 methicillin-resistant S. aureus [MRSA] and 254 methicillin-susceptible S. aureus [MSSA]) were genotyped by spa typing and multilocus sequence typing. For MRSA isolates, the staphylococcal chromosomal cassette mec (SCCmec) element was also typed. Three clinical categories were identified: nasal carriage only (n p 118), uncomplicated infection (n p 104 ), and bacteremia with hematogenous complications (n p 157). Results. By use of eBURST, 18 clonal complexes (CCs) were found in 371 isolates. Eight CCs accounted for 89% of isolates and occurred in all clinical categories. CC5 (P p .0025) and CC30 (P p .0308) exhibited a significant trend toward more frequent hematogenous complications. Isolates within spa types 2 and 16 showed the same significant trend and grouped within CC5 and CC30, respectively. SCCmec II isolates also showed the same significant trend compared with SCCmec IV; 96% were CC5 or CC30. Conclusions. Although most S. aureus genotypes exhibited the capacity to cause invasive disease, strains within CC5 and CC30 exhibited a significant trend toward increasing levels of hematogenous complications. Isolates within these CCs were also implicated by use of spa and SCCmec typing. The genetic determinants underlying these findings remain to be demonstrated. The clinical spectrum of Staphylococcus aureus infection ranges from asymptomatic nasal carriage to infective endocarditis (IE). Although bacterial characteristics may be associated with distinct clinical manifestations of S. aureus [1-5], previous studies evaluating potential - associations between bacterial genotype and virulence have yielded conflicting results [69]. Previous studies have been limited by the heterogeneity of S. aureus infections [10, 11] and by the absence of a large, wellcharacterized collection of isolates from differing forms of infection. Thus, the association between S. aureus genotypes and severity of illness is incompletely understood. Multilocus sequence typing (MLST) uses nucleotide sequences to identify alleles at loci that diversify slowly by random accumulation of neutral or nearly neutral variation [12]. MLST effectively differentiates closely related genotypes or clonal complexes (CCs) of S. aureus [8, 13, 14]. spa typing is another sequence-based S. aureus genotyping method that is complementary to MLST. It is based on the genetic variation in the polymorphic 24-bp variable number tandem repeat within the 3 coding region of the S. aureusspecific gene spa (staphylococcal protein A) [15]. The sequence variation within the 24-bp repeats and the variation in the number of the repeats each contribute to the overall genetic diversity. In the current investigation, we evaluated potential associations between the severity of infection and the genotype of the infecting S. aureus strain. To perform this analysis, we employed isolates from a large collection of prospectively characterized S. aureus isolates, stringent clinical grouping definitions, and 3 genotyping strategies. Our specific goals for this investigation were to (1) develop reproducible clinical definitions for an ordinal spectrum of S. aureus infections; (2) characterize the genetic diversity of a large cohort of clinically wellcharacterized MSSA and MRSA isolates; and (3) evaluate potential associations between the presence of hematogenous complications in patients and the bacterial genotype of the infecting S. aureus strains, as determined by MLST, spa, and staphylococcal chromosomal cassette mec (SCCmec) typing. METHODS Patient characteristics. The current study was approved by the Duke University Medical Center (DUMC) Institutional Review Board. Between September 1994 and October 2003, all hospitalized adult patients with 1 blood culture positive for S. aureus at our institution were identified. Clinical data and corresponding bloodstream isolates from consenting patients were prospectively collected [11]. Patients were excluded for any of the following: outpatient status, age !18 years, polymicrobial infection, neutropenia (absolute neutrophil count !1.0 109 cells/L), and death before evaluation by the investigators. Only the initial episode of S. aureus bacteremia was included in the current investigation. Heterogenous patient characteristics can obscure potential associations. Accordingly, we used a narrow case definition for each clinical group. All patients with bacteremia were hospitalized with monomicrobial S. aureus bacteremia and had no history of injection drug use, neutropenia, surgery within the previous 30 days, or indwelling prostheses (e.g., orthopedic prostheses, cardiac valves, defibrillators, or pacemakers). These exclusion criteria eliminated important nonpathogen characteristics that can influence the severity of staphylococcal disease. Next, we used strict definitions to identify 3 clinical groups that represent a progression from healthy individuals to those who are severely affected: (1) nasal carriage only, (2) uncomplicated infection, and (3) bacteremia with hematogenous complications. Nasal carriage isolates were obtained from the nares of noninfected subjects in 2 distinct epidemiologic settings within the DUMC referral area: (1) healthy undergraduate students with no current infection or health care contact (community carriage isolates) and hospitalized or (2) hemodialysisdependent subjects with no active infection (health care associated carria (...truncated)