The enterococcus is ever with us, a commensal and a low-level pathogen, except for its ability to cause morbidity and occasionally mortality in bloodstream infections. Vancomycin-resistant enterococci (VRE), primarily species specific Enterococcus faecium, emerged 20 years ago, and VRE still poses a threat. During the past two decades, VRE has been the subject of widespread publication, with the latest flurry of publications coming between 2007 and 2012. Many of the risk factors have been determined in various settings, but a proven strategy for prevention has not been devised.

From the LDS Hospital in Salt Lake now comes a study by Ford et al of VRE bloodstream infections in patients with newly diagnosed acute leukemia. They studied 214 consecutive patients with acute leukemia for colonization and related bloodstream infections with VRE. The study covered the years 2006 to 2014. Patients during the time of their leukemia were housed in a laminar-flow room in a ward dedicated to patients with hematologic malignancies. All patients had central catheters and all patients had weekly stool cultures for VRE. One hundred seventy of the patients had AML. The mean hospital admission lasted 29 days, with a range of 1-69 days.

Fifteen (7%) of the patients acquired a bloodstream infection with VRE. Only 3% had stool colonization with VRE on admission, and 35% acquired VRE after admission. Almost 10% developed toxin-positive stool for C. difficile and 80% of the patients with bloodstream infection had VRE stool colonization. Variables associated with stool colonization by both univariate and multivariate analysis included exposure to corticosteroids and carbapenems, and an increased number of stools per day. When antimicrobials were analyzed separately, only carbapenems were a significant risk factor (P = 0.002).

The molecular analysis done by repetitive element PCR was very interesting considering the patients were all housed on the same ward over many years. Of 24 strains analyzed, 21 had highly similar or identical patterns (my reading) with 80% genetic similarity. Two of the three remaining had about 10% less genetic similarity, leaving one isolate with a highly divergent type. The authors report nine of the isolates as an identical molecular group. Of nine bloodstream infection/stool pairs, 78% had identical pattern types. Hospital costs were doubled in patients with bloodstream infections due to VRE, primarily due to prolonged length of stay (LOS).


What have we learned about VRE in 20 years? We know that prolonged hospitalization, hematologic malignancies, neutropenia, exposure to antibiotics, stool colonization, length of stay, and probably central lines are risk factors for VRE bloodstream infection. Even though the numbers are small, this current study adds evidence that most bloodstream infections are related to prior colonization with VRE. There is a constant “colonization pressure” on units such as the one in this study. We also know the intestinal microbiome is a very complex entity, with hundreds of species of bacteria colonizing the normal gut. Yet we do not know which microbiomes are more protective than others from high-grade VRE colonization. If we can solve that problem, perhaps using probiotics, find nonpathogenic strains of VRE or vancomycin-susceptible enterococci to colonize the at-risk gut, or invent a useful immunization strategy, we can reduce the gastrointestinal colonization that is linked to bloodstream infection.1

Additionally, there may be new strategies based on altering the microbial burden in the environments of high-risk wards. Perhaps the use of sterilizing tough surfaces like those containing a high percentage of copper, periodic disinfection of the environment with UV light or hydrogen peroxide, or newer forms of surface decontamination will lessen the risk of GI colonization with VRE.

The authors also are concerned with the persistence of similar strains during the six years of the study. Is there a way to rid a unit like this of the predominating strains of VRE? Some of the strategies described above may be used in conjunction with newer infection control methods of preventing transmission and colonization, such as very early detection of colonization using PCR or protein-based rapid diagnosis.

For now, the outlook is not good for those patients who do become colonized with VRE. Their risk of bloodstream infection after colonization increases steadily with exposure to steroids, broad-spectrum antibiotics, and length of stay. Survival itself in a VRE group compared to controls is reduced by 20% at 12 months into their leukemia. Clearly there is a lot more work to be done to understand how best to protect these very vulnerable patients from infection with VRE. 


  1. Crouzet L, Rigottier-Gois L, Serror P. Potential use of probiotic and commensal bacteria as non-antibiotic strategy against vancomycin-resistant enterococci. FEMS Microbiol Lett. 2015 Feb 8. pii: fnv012. [Epub ahead of print]