Vancomycin-resistant enterococci (VRE) are an increasingly recognized cause of infection worldwide. The vanD operon has been found to be exclusively chromosomally encoded, and it is characterized by constitutive resistance to moderate levels of vancomycin and teicoplanin.Reference Depardieu, Kolbert and Pruul ¹ Strains of Enterococcus faecium harboring the vanD gene have been sporadically reported in the United States, Canada, Brazil, and Europe,Reference Depardieu, Kolbert and Pruul ^{1 –} Reference Boyd, Lalancette, Lévesque and Golding ⁵ but nosocomial transmission of this genotype has not been described. We report the first nosocomial outbreak of E. faecium containing a vanD genotype in an orthopedic–neurosurgical ward at a Canadian tertiary-care center.

Sunnybrook Health Sciences Center is a 627-bed acute- and tertiary-care hospital located in Toronto, Ontario, Canada. Risk-factor–based screening for VRE using rectal swabs is performed for all patients on admission, and contact precautions and private room isolation are instituted for VRE-positive patients. Risk factors that prompt admission VRE screening include direct hospital transfer or prior admission within the last year, home healthcare services, hemodialysis, or patients unable to answer these questions. Patients with hospital stays >30 days receive repeat VRE screenings. Rectal swabs are plated on Brilliance VRE chromogenic agar (Oxoid, Nepean, Ontario), and all positive isolates are further characterized using pulsed-field gel electrophoresis (PFGE). Additionally, multiplex polymerase chain reaction (PCR) is performed for species identification (E. faecalis, E. faecium, E. gallinarum, and E. casseliflavus) and genotype (vanA, vanB, and vanC) detection. Over the past 4 years, ~40% of admissions have undergone VRE screening; among them, 0.5% were VRE positive on admission. Nosocomial transmission of VRE remained below 0.4 per 1,000 patient days over this period, with only 3 VRE bloodstream infections.

As part of an investigation of 2 nosocomially acquired cases of VRE (vanA subtype) in an inpatient unit, point prevalence screening performed by the Infection Prevention and Control program identified 2 nosocomial cases of VRE (E. faecium); both were negative by PCR for vanA, vanB, and vanC. In response, an outbreak was declared on May 7, 2017, and additional control measures were implemented: enhanced cleaning, hand hygiene audit-and-feedback, daily chlorhexidine gluconate bathing, dedicated equipment for isolated patients, and cohorting of healthcare providers to colonized patients.

Weekly point-prevalence screening identified 4 VRE (E. faecium) transmissions among patients with previously negative surveillance cultures, for a total of 6 cases in the unit between May 1, 2017, and May 15, 2017. Using PCR testing, the presence of vanD was confirmed in all 6 isolates.Reference Boyd, Lalancette, Lévesque and Golding ⁵ Based on PFGE analysis, 4 of the isolates were indistinguishable, and 2 isolates were closely related, with <3 band differences (Table 1). Multilocus sequence typing of 7 housekeeping genes (adk, atpA, ddI, gdh, gyd, purK, and pstD) demonstrated a complete match with sequence type 117 (ST117) in all isolates.Reference Homan, Tribe and Poznanski ⁶ Nucleic acid sequencing of the vanD gene indicated 100% shared identity with the vanD4 gene.Reference Camargo, Dalla Costa, Woodford, Gilmore and Darini ³

Table 1

Typing and Antimicrobial Susceptibility Testing of 6 vanD Enterococcus faecium Isolates Identified as Part of an Outbreak in a Canadian Neurosurgical Acute Care Ward and 2 Sporadic Isolates Identified 4 Months Prior

Note. AMP, ampicillin; NIT, nitrofurantoin; VAN, vancomycin; TEC, teicoplanin; CIP, ciprofloxacin; DAP, daptomycin; RIF, rifampin; TET, tetracycline; TGC, tigecycline; LZD, linezolid; S, susceptible; I, intermediate; R, resistant; NC, not completed; ST, sequence type; PFGE, pulsed-field gel electrophoresis.

^a Susceptibility testing was carried out using VITEK-2 antimicrobial susceptibility cards (bioMérieux, Marcy l’Etoile, France) and an in-house prepared microbroth dilution. Interpretative criteria provided in parentheses beside minimum inhibitory concentration values are based on CLSI M100-S27 performance standards for antimicrobial susceptibility testing.

^b Isolates 1, 2, 5, and 6 were indistinguishable based on PFGE analysis.

^c Isolate 8 was matched to ST117 with 5 of 7 matching housekeeping genes; all other isolates were matched to all housekeeping genes

Of the 6 cases, 5 were linked to shared rooms with overlapping time periods. None of the patients colonized with vanD E. faecium developed a clinical infection. The outbreak was declared over on June 8, 2017, after no further cases were identified in 3 consecutive weekly point-prevalence screens. Upon review, 2 nosocomially acquired cases of vanD-carrying E. faceium were identified during routine surveillance screening on geographically separated units 4 months before the outbreak. Both isolates belonged to ST117 and had a PFGE pattern closely related to the outbreak strain.

To our knowledge, this is the first outbreak of nosocomial VRE transmission of the vanD genotype to be documented. The temporal and geographic clustering of cases suggests that transmission occurred through direct contact between healthcare providers with patients and the environment. The persistence of vanD-carrying VRE on environmental surfaces has not been studied, but this report provides evidence that nosocomial transmission can occur.

Reports of vanD-carrying VRE have been sporadic in the literature. In addition to E. faecium, vanD has been identified in E. faecalis,Reference Depardieu, Foucault and Bell ⁷ E. avium,Reference Depardieu, Foucault and Bell ⁷ E. gallinarum,Reference Boyd, Miller and Mulvey ⁸ and E. raffinosus.Reference Tanimoto, Nomura and Maruyama ⁹ Non-enterococcal bacteria can also harbor the vanD gene and may serve as an important reservoir.Reference Domingo, Huletsky and Giroux ¹⁰ A case of vanD-carrying E. faecium septicemia in a hematology ward reported by Starlander et alReference Starlander, Tellgren-Roth and Melhus ² identified 8 other patients (30%) with enterococci carrying the vanD gene, but no specific species were identified nor did dissemination occur.Reference Starlander, Tellgren-Roth and Melhus ² The prevalence of vanD enterococcal carriage in the general population is unknown but may be underestimated because most microbiology laboratories do not routinely perform genotyping on VRE-positive patients.

To date, 5 distinct vanD alleles have been reported in enterococci (vanD1 to vanD5). While previous vanD-carrying VRE reported in Canada have included strains N97-330 (vanD3 allele) and N03-0072 (vanD5 allele),Reference Boyd, Lalancette, Lévesque and Golding ⁵ this outbreak represents the first Canadian cases of vanD4-carrying VRE. The vanD4 gene was initially identified in an E. faecium (isolate 10/96A, ST281) from a 9-year-old girl with aplastic anemia in Brazil.Reference Camargo, Dalla Costa, Woodford, Gilmore and Darini ³ The finding of 100% sequence homology with vanD4, yet different sequence types between the outbreak and the E. faecium 10/96A isolate, suggests acquisition of resistance through horizontal gene transfer rather than clonal expansion.

The index case in this outbreak was a patient hospitalized for more than a year who received multiple courses of antibiotics with no travel history to Brazil. We suspect that this patient acquired the organism from the hospital environment in the context of selective pressure. With the identification of a related vanD isolate 4 months prior, unrecognized nosocomial transmission of this strain likely took place months before the outbreak.

This report describes the first documented nosocomial outbreak of vanD-carrying E. faecium. Many institutions do not routinely perform genotyping on VRE isolates; therefore, vanD transmission in hospitalized patients may be underrecognized.