Enterococci are gram positive, facultatively anaerobic cocci that form chains of various lengths. They are robust and versatile, with a remarkable propensity to survive under harsh conditions (1) . Enterococci have been known for more than a century as a common cause of endocarditis, a disease that is fatal without effective antimicrobial therapy. Since then, they have been shown to cause a range of infections including pelvic infections, wound infections, meningitis, intra-abdominal and pelvic infections, urinary tract infections (UTI) and neonatal infections (2). The genus Enterococcus is composed of 38 species, the most important of which are Enterococcus faecalis and Enterococcus faecium—both human gastro-intestinal (GI) colonisers. The clinical importance of the genus Enterococcus is directly related to its antibiotic resistance, which contributes to the risk of colonisation and infection. The tremendous progress of medical care in the last hundred years towards more intensive care and invasive medical procedures has undoubtedly contributed to the increased prevalence of these opportunistic pathogens (3).
Over the past twenty years infections due to multi-drug resistant organisms (MDRO) have escalated worldwide, affecting patient morbidity, mortality and healthcare costs. Among these bacteria Enterococcus faecium and Enterococcus faecalis represent opportunistic nosocomial pathogens that have been known to cause infections for decades (4). With ampicillin resistance in E. faecium being almost ubiquitous, treatment of infections caused by this isolate relies heavily on vancomycin (5-7).
First encountered in 1986, vancomycin resistance in Enterococcus species has increased in prevalence (8), with Ireland having the highest rate of vancomycin resistant Enterococcus (VRE) compared to any other European country. This has become a serious issue for nosocomial infection control, with increased pathogenicity seen in E. faecium due to an expansion of hospital-adapted genetic lineages showing more resistance and virulence traits. A staggering 44.4% of all E. faecium from bloodstream infections (BSI) in Ireland were resistant to vancomycin in 2016 (9). Resistance to glycopeptides in Enterococcus spp. is mediated by the vancomycin resistance (Van) operon, which may be carried chromosomally or on on a plasmid. This Van operon contains a variable ligase gene, which is central in determining the level of vancomycin resistance – the most commonly identified genes being vanA, vanB and vanC (8).
Typically, vanA-mediated vancomycin resistance is plasmid-borne and is found predominantly in E. faecium. Vancomycin resistance in vanA-bearing isolates is mediated by the group of genes vanR, vanS, vanH, vanA and vanX, which are usually carried on the Tn1546 transposon. The expression of these genes leads to replacement of the C-terminal D-Ala residue with D-Lac during cell wall synthesis, thus modifying the vancomycin-binding target. The transposon is often contained on plasmids, making it easy to transfer among enterococcal strains. Historically, there has been little sequence variation in this gene cluster (10). In recent years, vancomycin sensitive E. faecium with ‘silent’ vanA genotypes have been reported, further supporting the need for genotypic testing (11-15). These so-called vancomycin-variable vanA+ enterococci (VVE) have the ability to switch into vancomycin resistance during therapy through the constitutive expression of the vancomycin resistance cassette thus escaping phenotypic detection. By definition, the term VVE should be restricted to vancomycin-susceptible enterococci containing vanA and capable of reverting to a glycopeptide-resistant phenotype. Accordingly, enterococci containing remnants of the vanA cluster that are not able to revert to a resistant phenotype or enterococci with vanB showing an MIC below the clinical breakpoint are not VVE (11). Due to the majority of the detection of VRE being via phenotypic methods, the overall prevalence for VVE cannot be accounted for, with the possibility of unrestricted transmission of these strains occurring in health care facilities (13). In one case, these VVE isolates have been shown to grow (poorly) on chromogenic VRE agar, such as Brilliance Agar (Oxoid, Canada) – then when tested for vancomycin resistance, demonstrated a vancomycin MIC of <= 1mg/L (14). However, further research is required to determine the sensitivity of various selective agars in detecting VVE and whether this strategy in combination with molecular testing is cost-effective. Downing et al reported bacteraemia caused by VRE after intravenous vancomycin was used to treat a vancomycin sensitive E. faecium. Both initial and final isolates from blood cultures were identical by pulsed-field gel electrophoresis (PFGE). If VVE becomes an established nosocomial pathogen, it may be appropriate to recommend routine genotypic testing of E. faecium clinical isolates, particularly those from sterile sites, to determine the presence of vancomycin resistance genes (16). A number of studies have demonstrated the ability of vancomycin resistance determinants to be horizontally transferred within enterococci species (17). Although E. gallinarum are generally not considered significant pathogens, it has been shown in certain situations to cause serious infections, as well as several hospital-acquired outbreaks (18, 19). In addition to this, the low-level resistance usually seen in E. gallinarum is mediated by a non-transferrable chromosomally mediated vanC gene and therefore does not represent significant infection control challenges. However, an isolate of E.gallinarum carrying vanA & vanB genes conferring high-level resistance to vancomycin and teicoplanin has been reported (20). The ability of E. faecium to transfer the vanA gene cluster to E.gallinarum has been documented from additional sources (21). As the current algorithm in clinical microbiology laboratories does not routinely test for such resistant determinants in these species, this highlights the need for further investigation of any isolated E.gallinarum for the presence of vanA genes. Various selective media exist for the detection of VRE - some of which inhibit E. gallinarum and some of which do not (22- 24). Discovery of these isolates represents unique challenges to infection control practices as well as their diagnosis and management. These isolates have shown to be stable and capable of being transmitted among hospitalised patients. Despite considerable research, the reasons for the emergence & rapid spread of VRE remain ambiguous. If control of VRE in the Irish population is to be achieved a better understanding is required. Pathogenesis Although enterococci may not be considered as inherently virulent as other organisms, a number of virulence factors are involved in their success. Not only does their resistance to antibiotics provide a selective advantage, enterococci are armed with a number of genes encoding adhesion proteins that may mediate adherence to host tissues, consistent with their pathogenic role in infective endocarditis (1). There are a number of proteins secreted into the extracellular medium which have been implicated in enterococcal virulence; these include haemolysin–cytolysin (Cyl), gelatinase (GelE) and extracellular serine proteinase (SprE). Haemolysin–cytolysin (Cyl) is a toxin (encoded by plasmids or pathogenicity islands) which is produced by about 30% E. faecalis strains. It is secreted extracellularly and is responsible for lysing red blood cells and some white blood cells. Strains expressing cyl have been shown to be more virulent in animal models than those strains without (25). Gelatinase (GelE) is a protease produced by enterococci which has the ability to mediate virulence through degradation of host tissue as well as modulation of the host immune response (26). It plays a role in the activation of autolysin, which leads to the release of extracellular DNA and the formation of a biofilm (27). Studies of mutants lacking GelE have been shown to have a decreased biofilm formation and a decrease in translocation across intestinal cells (28). It is interesting to note, however, that although both Cyl and GelE are known virulence factors isolated from clinical isolates, they also are seen in equal amounts from stools of healthy individuals colonising the GI tract (29). There are also a number of cell surface determinants that have been involved in enterococcal virulence. These include aggregation substance proteins (AS proteins), enterococcal MSCRAMMs (microbial surface components recognizing adhesive matrix molecules), ElrA (enterococcal leucine-rich-repeat-containing protein), LPXTG-like motifs, immunoglobulin-like folds, enterococcal polysaccharide antigen (Epa), enterococcal cell membrane glycolipids and LTA93. Aggregation substance proteins (AS proteins) are encoded by plasmids, which cause clumping of E. faecalis. They have been shown to increase E. faecalis binding to cultured renal epithelial cells, survival within polymorphonuclear neutrophils and internalization by intestinal cells (30). They are also an important virulence in endocarditis by aiding in the formation of large bacterial aggregates on the cardiac valve (31,32). Enterococcal MSCRAMMs (microbial surface components recognizing adhesive matrix molecules) are another well studied cell surface determinant responsible for enterococci virulence, which are considered to be important in the early stages of infection. A number of adhesins have been described, the most extensively studied being the collagen adhesins Ace in E. faecalis and Acm in E. faecium. These are cell wall-anchored adhesins, which bind collagen (33). Community-associated strains, unlike hospital-associated strains, often have an Acm pseudogene and rarely express Acm. The expression of these adhesins may therefore have a role in increased ability of hospital-associated Enterococcus spp. to cause disease (34). ElrA (entero-coccal leucine-rich-repeat-containing protein) is another cell surface protein which is from the WxL family, and is responsible for infecting macrophages (35). LPXTG-like motifs and immunoglobulin-like folds are other cell surface proteins which are bound into pili (36,37). Pili are filamentous structures which protrude from the cell surface, and have been shown to play an important role in infection and initial colonisation in a number of bacteria including enterococci. Ebp (endocarditis and biofilm-associated pili) pili are ubiquitous in E. faecalis (38) and have shown to be imperative for the formation of biofilm and the pathogenesis of endocarditis and UTI's (39). Polysaccharides are significant components of the Gram-positive cell surface and play an essential role in pathogenesis by mediating evasion of phagocytosis by polymorphonuclear neutrophils and promoting cytokine production (40). Enterococcal polysaccharide antigen (Epa), is a cell wall antigen which is seen in most patients with serious E. faecalis infections (41). Disruption of this epa cluster has been shown to result in a decrease in biofilm formation and translocation across an enterocyte monolayer (42), increase in susceptibility to polymorphonuclear neutrophil-mediated killing (43), and causes attenuation in mouse peritonitis and UTI models (44). Enterococcal cell membrane glycolipids and LTA93 have also been shown to be important in pathogenesis. An E. faecalis mutant that is deficient in one cell membrane glycolipid appeared to have reduced adherence to enterocytes and also reduced biofilm formation. As well as this, the mutant was cleared from the bloodstream quicker than the wild type (45). Other virulence factors have shown to play a role in enterococcal infections. Large transferable plasmids common among E. faecium isolates have been shown to increase the virulence of isolates and increase the ability of the organism to colonise the GI tract (46). Stress response proteins, for example Gls24, have also been shown to play a role in enterococcal virulence. This protein has been shown to mediate resistance to bile salts, and is important in endocarditis (47). Trends Regarding trends, Ireland has the highest rate of VRE than any other European country. Rates for bloodstream infection (BSI) with E. faecium VRE and E. faecalis VRE can be seen below in Tables 1 & 2, with rates peaking in 2014 for E. faecium at 45.9%. Since 2012, the rates have been relatively steady, and have not dropped below 40% since 2011. Figure 1 demonstrates this increasing trend seen in Ireland since 2004. Table 1: Summary of EARS-Net data by pathogen and year (data correct as of 10/03/2017). hpsc.ie Table 2: Summary of EARS-Net data by pathogen and year (data correct as of 10/03/2017). hpsc.ie Figure 1: Trends for E. faecium – total numbers of E. faecium and percentage resistance to high-level gentamicin (HLGR- Efm) and vancomycin (VREfm) with 95%CIs. ecdc.europa.eu. Rates of VRE in OLCHC however, are very low. There has only been one VRE isolated from a blood culture since 2012. Figure 2 below demonstrates the rates of VRE colonisation in the hospital since Q2 2008 to Q3 2017. Although the rates of colonisation increase and decrease between quarter/ years, there is an overall increase of the rate of colonisation since 2008. There is no data for rates of colonisation in Ireland, only rates of bloodstream infection. Table 3 demonstrates the number of (new) colonised patients in OLCHC from 2008 to 2017. The rates are very low compared to the national average. Figure 2: Summary of VRE colonisation rates in OLCHC 2008-2017. Table 3: Summary of OLCHC colonisation 2008-2017. In comparison to Europe, Ireland has the highest rate of VRE compared to any other country (with the exception of Cyprus in 2016) (6). Figure 3 below demonstrates the percentage of invasive isolates which are resistant to VRE, with the majority of European countries being at 10% or below. A difference of 30% is significant when comparing Irish rates to European average rates. Figure 3: Enterococcus faecium. Percentage (%) of invasive isolates with resistance to vancomycin, by country, EU/EAA countries, 2016. ecdc.europa.eu. Table 4 below demonstrates the comparison of rates of VRE BSI between different European countries from countries with no reported VRE to those that have higher rates similar to Ireland. It can be seen that Estonia has reported 0% VRE. Finland has very low rates, with 0.3% BSI reported as being vancomycin resistant. It is important to note that although Cyprus had a higher rate of VRE in 2016 than Ireland, there were only 41 VRE invasive infections in the year compared to 422 in Ireland. With population differences between the two countries taken into account, Ireland still had double amount of BSI with VRE than Cyprus. Table 4: Enterococcus faecium. Total number of invasive isolates tested (N) and percentage with resistance to vancomycin, including 95 % confidence intervals (95 % CI), lowest to highest incident countries, 2013–2016. ecdc.europa.eu. Current Diagnostics EUCAST have recently published recommended methods for the detection of glycopeptide resistance in E. faecium and E. faecalis. According to EUCAST, resistance may be detected by three methods - minimum inhibitory concentration (MIC) determination, disk diffusion and breakpoint agar method, with a strong emphasis on incubation of plates for a full 24 hours to detect isolates with inducible resistance. As vanA- mediated resistance results in high-level resistance to vancomycin (MIC 64-1024 mg/L) and teicoplanin (MIC 8-512 mg/L), these three methods should readily detect this resistance. Detection of vanB-mediated resistance however, may be more challenging due to the lower levels of resistance they confer to vancomycin (MIC 4-1024 mg/L) and teicoplanin ( MIC 0.06-1 mg/L). Genotypic testing may also be performed targeting vanA and vanB genes. Disk diffusion is recommended using a 5ug vancomycin disk with a full 24 hours of incubation. Any isolates with a 'fuzzy zone' or colonies inside the zone must be interpreted as resistant regardless of the zone size, with confirmation determined by MIC. See below Figure 4 demonstrating the interpretation of vancomycin discs (48). Figure 4: Reading of vancomycin disk diffusion tests on Enterococcus spp. www.eucast.org a) Sharp zone edges and zone diameter ?12 mm. Report as susceptible. b-d) Fuzzy zone edges and/or colonies within the zone. Report as resistant regardless of zone diameter. Breakpoint agar tests with Brain Heart Infusion agar and 6 mg/L vancomycin may also be used for detection of VRE. Plates must be incubation for 24 h at 35 ± 1°C in air is required to detect resistance in isolates with inducible resistance (49). A number of commercial chromogenic plates are available (22-24). The present algorithm for the isolation and identification of VRE in OLCHC uses faecal samples, which are cultured into VRE broth (containing 6ug/ml vancomycin, 2ug/ml amphotericin B, 50ug/ml colistin and 8ug/ml clindamycin). This is incubated for 24hrs at 37C. The broth is then subcultured to bioMerieux's CHROMID®? VRE agar and incubated for 48hrs at 37C. If growth is observed on the VRE agar, colonies are identified using the Vitek-MS. If the isolate is identified as E. faecalis or E. faecium, this is subbed to blood agar. Susceptibility testing is performed after 18-24hrs at 35C using the Vitek®? 2 AST-P607 card. If the isolate is vancomycin resistant, this is reported as a VRE, and MIC's to vancomycin / teicoplanin provide the "vanA-like" or "vanB-like" genotype. This algorithm has changed since 2015 when an EQA failure resulted after a VRE was reported as a vancomycin sensitive E. faecium. Previously, any E. faecium or E. faecalis isolated from a stool sample had vancomycin and teicoplanin disc diffusion performed. In this case, vancoymcin was reported as sensitive, when in fact it was resistant due to the guidelines provided by EUCAST not being followed meticulously (see Figure 4). Since then, MIC's to vancomycin/teicoplanin are used for the diagnosis of VRE in OLCHC using the Vitek®? 2 AST-P607 card. An increase in VRE isolated from stool samples has been seen since this change in the algorithm for the isolation and identification of VRE in OLCHC. Discussion Why are the rates of VRE in Ireland the highest in Europe? Despite considerable research, the reason(s) remain largely unknown. There are a number of possibilities as to why this may be the case. It could be postulated that there is a high proportion of vancomycin-variable vanA+ enterococci (VVE) in Ireland (according to EUCAST, the proportion and geographical location are unknown) which is spreading through the community. Upon hospitalisation and treatment with many antibiotics, the resistance may be induced. In one study of VRE among hospitalised patients in Ireland, 72% of inpatients had been treated with broad-spectrum antibiotic regimens, most commonly piperacillin/tazobactam, third-generation cephalosporins and carbapenems, prior to developing VRE bloodstream infections (50). It is also possible that commercial chromogenic media commonly used for VRE detection in laboratories are not detecting all VRE. EUCAST recommend breakpoint agar testing using 6mg/L vancomycin (48), however Chrom ID® Agar, for example, contains vancomycin at a concentration of 8mg/l (22). Although this technically may only pose a problem for vanB VRE, which have low MIC's to vancomycin, it could also inhibit the growth of these VVE isolates, which have lower MIC's than the 'classic' vanA VRE. Due to the majority of the detection of VRE being via phenotypic methods, the overall prevalence for VVE cannot be accounted for, with the possibility of unrestricted transmission of these strains occurring in health care facilities. It is also a possibility that EUCAST guidelines are not being followed meticulously when disc diffusion methods are employed. This requires a full 24 hours of incubation as well as inspecting zones carefully for fuzzy edges and/or microcolonies with transmitted light. It is possible that under-diagnosis of VRE is occurring in healthcare facilities that employ this method of detection. The ability of E. faecium to transfer the vanA gene cluster to E.gallinarum (21) could potentially be another factor leading to the spread of VRE. As discussed, the current algorithm in clinical microbiology laboratories does not routinely test for such resistant determinants in these species, therefore the presence of vanA in these isolates is currently unknown. It is possible that E. gallinarum are serving as a reservoir for transferable vanA genes. It is important to note also that lack of facilities to isolate VRE-colonized patients in single rooms (a common issue in many Irish hospitals) may lead to our rising colonisation rates despite other optimal infection control strategies. There is a strong emphasis on environmental screening in OLCHC. In 2017, 12,000 environmental screens were performed in the hospital. It is protocol for infection control to perform a deep clean on any area in the hospital after a patient with a MDRO carriage status has been present. Before this area is re-opened to another patient, 10 swabs are taken of various areas of the room, for example - the bed rail, door handle etc. These are then screened for VRE, and the room is not re-opened until the screening has come back negative. It is feasible that this level of screening is keeping colonisation rates well below the national average. Conclusion Enterococci are possibly the most marked examples of organisms that, historically, were regarded as second-rate pathogens but which have become one of the most challenging nosocomial problems in Ireland. This is most likely due to the increased use of antibiotics in our hospitals which can change the gut flora of patients. Enterococci take advantage of this phenomenon and 'conquer' the prized niche of the GI tract, where these organisms then cause infections. Despite considerable research, the reasons for the emergence & rapid spread of VRE remain ambiguous. If control of VRE in the Irish population is to be achieved a better understanding is required.