Methods: This study included 97 adults newly admitted to the ICU between February 2015 and June 2016. Rectal swabs were obtained at time of ICU admission and 72 hours later. VRE rectal colonization status was determined categorically for each sample by culture on selective media. Specimens were also cultured for methicillin-resistant Staphylococcus aureus (MRSA) and for MDR Gram negatives, defined as those with non-susceptibility to 3 or more antibiotic classes. 16S rRNA gene sequencing was performed and the relative abundance was calculated for B. producta. Differentially abundant bacteria taxa between VRE positive and VRE negative specimens were assessed using linear discriminant analysis effect size (LefSe) analysis.
Results: Among the 97 patients, 7 (7.2%) were colonized with VRE at time of ICU admission and 3 (3.3%) of the remaining patients became colonized 72 hours later. The microbiome composition differed significantly when accounting for VRE colonization status. The relative abundance of B. producta was 140-fold higher in VRE negative compared to VRE positive samples (0.0012% vs 8.48x10-6 %, p =.03). On LefSe analysis, there was also significantly lower differential abundance of B. producta when VRE was present (LDA score 4.65). The presence of VRE in culture was significantly associated with the co-presence of MRSA (23.5% co-colonized if VRE positive vs 8.4% if VRE negative, p= 0.046) but not with the co-presence of MDR Gram negative bacteria (29.4% if co-colonized if VRE positive vs. 34.3% if VRE negative, p=0.68).
Conclusion: In this ICU cohort, rectal colonization with VRE was inversely associated with the putatively protective organism B. producta. VRE was associated with rectal co-colonization with MRSA but not with MDR Gram negative bacteria. B. producta may have promise as a probiotic designed to prevent VRE colonization.
S. Stump, None
D. Moscoso, None
D. Chong, None
D. Freedberg, None