Streptococcus bovis bacteraemia revisited: Clinical and microbiological correlates in a contemporary series of 59 patients

Summary

Objective

To characterise the clinical features, associations and outcome in a contemporary series of patients with Streptococcus bovis bacteraemia (SBB).

Methods

Retrospective analysis of all episodes of SBB at the University Hospital 12 de Octubre (Madrid, Spain) between January 1997 and November 2008 was performed. Patient data were reviewed, focusing on clinical and microbiological associations with the different biotypes of S. bovis.

Results

Fifty-nine episodes of SBB were documented in 59 adult patients (30 males; mean age: 70.9 ± 15.0 years). Chronic liver disease was identified in 20 patients (33.9%). Sixteen patients (27.1%) presented infective endocarditis (IE) and 14 (23.7%) had a biliary source of bacteraemia. Thirty-three patients (55.9%) underwent colonic evaluation, adenomatous polyps being the most common finding (21 patients). Malignancy was diagnosed following SBB in 9 cases, including 6 patients with colorectal carcinoma (18.2% of those who underwent colonic evaluation). Of 22 isolates biotyped, 12 were S. bovis biotype I and 10 were S. bovis biotype II. IE was more frequent among patients with S. bovis biotype I (P = 0.010), whereas bacteraemia due to biotype II species was more likely to be of biliary origin (P = 0.078).

Conclusions

S. bovis biotyping identifies some clinically relevant associations.

Introduction

Streptococcus bovis, a non-enterococcal group D streptococcus, is a commensal inhabitant of the human digestive tract that may be isolated in up to 16% of faecal samples from normal individuals.1, 2 Since the first case report in 1951 by McCoy and Mason³ and the pivotal control-case study by Klein et al,⁴ the association between S. bovis bacteraemia and gastrointestinal diseases, specifically colorectal carcinoma, has been extensively established in the literature.5, 6 Based on such evidence, some authors have recommended a complete colonic evaluation in all patients with documentation of S. bovis bacteraemia, although the reliability and cost-effectiveness of this strategy remain to be assessed.7, 8 More recent studies have stressed the frequent association between the infection caused by this agent and the presence of chronic liver and biliary tract disorders as predisposing conditions.9, 10, 11S. bovis is also responsible for approximately 6% of all cases of infective endocarditis (IE) worldwide, as recently reported by the International Collaboration on Endocarditis (ICE) group, with particular relevance in southern Europe.¹² In addition, this prevalence rises up to 8.3% in individuals aged 65 years or older.¹³ As demonstrated by Tripodi et al¹⁴ in a large series from Italy, S. bovis IE is frequently correlated with advanced liver disease.

Traditionally, S. bovis strains have been physiologically divided into biotypes I (also called “classic”) and II or “variant” (further separated into II/1 and II/2 strains), according to the ability of the former to hydrolyse starch and ferment mannitol.¹⁵ It has been suggested that bacteraemia caused by S. bovis biotype I has a much stronger association with the presence of an underlying gastrointestinal cancer or the occurrence of IE, as opposed to infection caused by biotype II.⁶ To date the clinical correlates of S. bovis biotype II have been poorly studied. On the other hand, some extensive taxonomic changes have been proposed in the S. bovis/S. equinus group in the last few years.¹⁶ According to current terminology, based on DNA–DNA reassociation and whole-cell protein analysis, strains formerly termed S. bovis (biotype I) and S. bovis variant (biotype II/2) are now denominated S. gallolyticus and S. pasteurianus, respectively, while the former S. bovis variant (biotype II/1) is now named S. infantarius.¹⁷ Nevertheless, most clinicians still remain unfamiliar with biotypes or newer S. bovis species, and clinical correlates are uncommon.¹⁸

Most of the studies regarding the clinical and microbiological correlates of patients with S. bovis bacteraemia were performed several years ago;7, 19, 20 they included relatively small samples7, 9, 14, 15 or suffered from considerable heterogeneity in the diagnostic and methodological criteria.⁶ Thus, this study was aimed to review our experience with S. bovis bacteraemia in a large, contemporary series of patients from a single tertiary-care centre, paying special attention to demographic and clinical associations in relation to the different biotypes and microbiological findings.