To the Editor—Infections and gut colonization with vancomycin-resistant enterococci (VRE) have been increasingly reported in hospitalized patients from different regions of Brazil, where difficulties in controlling VRE colonization have been noted.^{Reference Campos, Batistão and Gontijo-Filho1-Reference de Almeida, de Araujo and Iwasaki4} Patient colonization with VRE is a major risk factor for developing subsequent infections with these strains.^{Reference Campos, Batistão and Gontijo-Filho1,Reference Alevizakos, Gaitanidis and Nasioudis5} The first VRE description dates from 2011 (M. Celeste Melo, personnel data), and VRE infections among hospitalized patients from Natal city (northeastern Brazil) remain low, contrasting with the high rates of VRE infection and colonization of the patients in the southern and southeastern regions, where they have occurred since 1998.^{Reference Campos, Batistão and Gontijo-Filho1-Reference Sacramento, Zanella and Esposito3} For early recognition of silent interhospital VRE transmission through colonized patients as in other parts of Brazil, we aimed to search and characterize VRE colonization strains from patients known to have a previously history of hospitalization.

From January 2015 to December 2016, 55 VRE isolates were recovered (chromID-VRE) from perianal swabs of 44 inpatients at 5 hospitals (range, 36–257 beds) in Natal (public hospitals A, B, and F; public–private hospital C, or private hospital E) during hospital admission and of 3 outpatients when receiving domiciliary nursing care at their own home. Most of these patients had previous hospitalization episodes. Six additional VRE were collected from different clinical specimens during the hospitalizations (in hospitals A, B, C, D, and F) of 6 patients (2 in common with those from the cohort of 44 patients). Species identification and search of antibiotic resistance, virulence, and plasmid replicase genes were conducted using polymerase chain reaction testing (PCR) in isolates representative of different clones and epidemiological contexts.^{Reference Freitas, Novais and Duarte6,Reference Freitas, Novais and Tedim7} Antibiotic susceptibility was assessed using disk diffusion (ie, for ampicillin, tetracycline, chloramphenicol, gentamicin, erythromycin, and streptomycin), Etest (vancomycin), agar dilution (teicoplanin), or broth microdilution (linezolid) following the European Committee on Antimicrobial Susceptibility Testing (EUCAST) clinical breakpoints (http://www.eucast.org/clinical_breakpoints/) or, when not possible, Clinical and Laboratory Standards Institute (CLSI) breakpoints. Clonal relationships were firstly evaluated by SmaI pulsed-field gel electrophoresis (PFGE) in the 61 VRE and by multilocus sequence typing (MLST) in 18 Enterococcus faecalis (Efs) (ie, different PFGE types, patients, and hospitals).^{Reference Freitas, Novais and Duarte6}

The VRE were identified as Efs (n = 60) and E. faecium (n = 1), all carrying vanA. In this study, we only proceeded with the Efs; all were multidrug resistant (MDR) and exhibited resistance to vancomycin (range, 16 to >256 mg/L), teicoplanin (range, 4–64 mg/L), ciprofloxacin, tetracycline, and erythromycin. Most were also resistant to gentamicin (93%) and streptomycin (42%). PFGE established 3 pulsotypes among the 60 VREfs: 30 for pulsotype A, 29 for pulsotype B, and 1 for pulsotype C. The dominant pulsotypes A and B were dispersed among all institutions and were detected throughout the study period. Representative VREfs from pulsotypes A and B were identified as sequence type 6 (ST6; n = 4: 2 of pulsotype A and 2 of pulsotype B) or ST525 (n = 13: 6 of pulsotype A and 7 of pulsotype B). Pulsotype-C was also an ST6 (Table 1). Examples of Efs strains showing the same or similar pulsotypes but clustering in different sequence types have previously been described.^{Reference Freitas, Coque and Novais8} ST6 is dispersed worldwide mostly in association with nosocomial infections, whereas ST525 has only been documented in Brazil southeastern regions (VREfs linezolid-resistant infections; https://pubmlst.org),^{Reference de Almeida, de Araujo and Iwasaki4,Reference Santos, Oliveira and Cardoso9} suggesting the dispersion of relevant strains of such lineage in this country. A patient was colonized with both ST525 A and B clones collected 14 months apart. The same PFGE-B/ST6 Efs was associated with 2 isolates from the same patient (bone fragment and colonization) obtained 7 days apart in different hospitals (Table 1). Moreover, vanA plasmid replicases from the 18 VREfs were identical (rep₉ from pheromone-responsive pTEF2/pAD1 plasmids) and known to transfer highly efficiently, which could have enhanced vancomycin-resistance spread also by horizontal gene transfer. Among the 18 VREfs, all carried vanA, erm(B), tet(M), aac(6')-aph(2”), whereas aph(3')-III, and ant(6)-Ia were variable. Putative virulence factors included adhesins, biofilm-associated genes, and survival genes (eg, ace, gelE, and agg/cyl) previously associated with dominant lineages of human infections (Table 1).^{Reference Raven, Reuter and Gouliouris10}

Table 1. Genetic Characterization of 18 Representative Enterococcus faecalis With Different Epidemiological Backgrounds

Note. PFGE, pulsed-field-gel-electrophoresis; MLST, multilocus sequence typing; OP, outpatient

^a When the same patient carried VRE in >1 occasion, the data are shown in bold.

Glycopetides    Aminoglycosides

Macrolides       Secreted factor for tissue damage

Tetracyclines      Adhesins and biofilm-associated genes

In summary, patients with a previous history of hospitalization and who had been admitted to different hospitals of Natal were colonized with 2 VREfs clonal lineages with potential to cause infection, one closely related to global ST6-VREfs and the other more region specific (ST525).^{Reference de Almeida, de Araujo and Iwasaki4,Reference Santos, Oliveira and Cardoso9} The detection of several Efs strains with the same PFGE type associated with different sequence types (via MLST) highlights both the limited accuracy of these methods to identify identical Efs strains and the need to use methods with a higher resolution to better follow transmission events between patients and institutions. The identification of 2 specific VREfs clones is reminiscent of the emergence of VRE in other regions of Brazil in previous years, where VREfs strains are now being replaced by dominant vancomycin-resistant hospital-associated E. faecium.^{Reference Resende, Caierão and Prates2,Reference Sacramento, Zanella and Esposito3} Our results highlight the need for better hygiene measures, systematic colonization survey of transferred patients, patient isolation, and antimicrobial stewardship to prevent future epidemic and endemic scenarios associated with infection in hospitals from Natal. Of special concern is the transfer of such VRE to the community level by colonized patients with domiciliary nursing, which represents another potential level of transmission requiring additional strategies to control their spread.