Abstract
BACKGROUND: Lactobacilli are gastrointestinal commensals and may represent skin contamination when isolated in blood cultures. Although uncommon, Lactobacillus bloodstream infections and endocarditis have been reported. METHODS: We conducted a retrospective cohort study of patients with Lactobacillus species bacteremia and/or endovascular infections over a 22-year period to identify potential predictors of possible or definite endocarditis using the modified Duke criteria. Lactobacillus growth in an initial blood culture set without growth in subsequent blood culture sets collected within 7 days was labeled as blood culture contamination (BCC), and these were excluded from the primary analysis. The primary outcome was the proportion of patients with possible or definite endocarditis. For all Lactobacillus isolates, we collected antimicrobial susceptibility data determined via broth microdilution methods according to Clinical and Laboratory Standards Institute guideline M45. RESULTS: We identified 331 patients with blood and/or endovascular cultures positive for Lactobacillus, of whom 100 were included. The primary outcome of possible or definite endocarditis was identified in 29% (29/100) of patients included in the primary analysis. Both the presence of an intracardiac device and/or non-native valve (relative risk [RR], 8.57; 95% CI, 1.89-38.83) and injection drug use (RR, 13.47; 95% CI, 3.18-57.03) were predictors of possible or definite endocarditis. All-cause mortality at ≤90 days was 22% (21/94). CONCLUSIONS: Nearly 1 in 3 patients with Lactobacillus bacteremia had possible or definite endocarditis, though polymicrobial infection was common. Growth of Lactobacillus spp. in blood cultures should be considered potentially pathogenic and interpreted in the clinical context before the isolate is labelled as inconsequential.