Gut colonization with vancomicyn-resistant enterococci among patients with hematologic malignancies

The role of vancomycin-resistant enterococci (VRE) as medically important pathogens has significantly increased in the recent years. These bacteria are causative agents of hospital-acquired infections, usually involving the urinary tract, surgical sites and the bloodstream [1]. Also, it is well known that gastrointestinal tract carriage of VRE occurs in critically ill patients, especially those admitted to intensive care units, hematology-oncology and transplantation wards [2].

The frequency of VRE colonization in patients with hematologic malignancies is particularly high [3‐5] and geographical differences in VRE spread were described among such patients [6]. Many risk factors have been studied for VRE carriage and the most prominent are: advanced age, prolonged hospital stay, exposure to various antibiotics (e.g., cephalosporins, vancomycin and drugs with activity against anaerobic bacteria), underlying comorbid conditions, exposure to medical devices and contact with other VRE colonized patients [7‐12].

Intestinal colonization with VRE has been identified as a risk factor for subsequent blood stream infections (BSIs) [13]. A lot of studies have shown a high prevalence of VRE bacteremia in patients with hematologic malignancies [14, 15]. It is interesting to note that these patients are significantly more susceptible to VRE bacteremia than bacteremia caused by vancomycin-susceptible enterococci [13]. Other risks factors for VRE bacteremia are neutropenia and mucositis. The mucosal compromise due to chemotherapy allows VRE translocation into the bloodstream in patients with malignancies [16]. In addition, inefficient clearance of circulating pathogens due to neutropenia increases the risk of VRE BSIs among colonized patients [13].

The most common VR Enterococcus species associated with colonization/infection are E. faecium and E. faecalis, which carry the vanA or vanB genes [7]. Although, some authors have reported high prevalence of intestinal colonization with low-level vancomycin resistant E. gallinarum and E. casseliflavus [17], which are rare causative agents of nosocomial infections [18‐21].

Despite many reports for VRE acquisition worldwide, the data about the VRE carriage in hospitalized patients in our country are limited. Therefore, the purpose of our study was to determine the incidence of and risk factors for colonization with VRE among patients with hematologic malignancies.

A total of 119 patients with hematologic malignancies were enrolled in this investigation. The median age of the studied participants was 67 (range: 20–88) years, 69 (58%) were male and 50 (42%) were female. The mean length of hospital stay was 9 days (range: 6–46 days). The most frequent reason for admission was cytotoxic chemotherapy (n = 53, 44.5%). The most frequent underlying malignancies were myelodysplastic syndrome (n = 28), non-Hodgkin’s lymphoma (n = 26), multiple myeloma (n = 20) and chronic lymphocytic leukemia (n = 17). Overall, 256 fecal samples were collected with a mean of 2.2 samples per patient (range: 1 to 7). The same microorganism was detected in all patients with two or more positive cultures.

Intestinal colonization with VRE was observed in 18 (15.1%) out of 119 screened patients. One patient was a carrier of two strains (1 VR E. faecium and 1 VR E. faecalis), resulting in a total of 19 VRE isolates: 16 with VanC phenotype (12 E. gallinarum and 4 E. casseliflavus), 2 E. faecium and 1 E. faecalis.

Antimicrobial susceptibility profiles of the isolated VRE and their van genes are presented in Table 1. The results showed that only one E. faecium isolate demonstrated a VanA phenotype with high-level resistance to vancomycin (MIC ≥ 256 μg/ml) and teicoplanin (MIC = 96 μg/ml), as well as ampicillin (MIC ≥ 256 μg/ml), gentamicin (MIC ≥ 1024 μg/ml), ciprofloxacin (MIC ≥ 32 μg/ml) and quinupristin/dalfopristin (MIC ≥ 32 μg/ml). The VanA phenotype of this strain was verified by PCR detection of the vanA gene. The second E. faecium and E. faecalis expressed low-level resistance to vancomycin with MICs of 8 μg/ml and 12 μg/ml, respectively, and susceptibility to teicoplanin (MICs = 0.5 μg/ml). E. faecium was resistant to ampicillin (MIC ≥ 256 μg/ml) and ciprofloxacin (MIC ≥ 32 μg/ml), while E. faecalis was only highly resistant to gentamicin (MIC ≥ 1024 μg/ml). The vanB gene was detected in the two isolates. All E. gallinarum and E. casseliflavus isolates showed low-level resistance to vancomycin (MICs: 4–16 μg/ml), susceptibility to teicoplanin (MICs: 0.5–1.0 μg/ml) and to the rest of the tested antimicrobial agents. Furthermore, the E. gallinarum and E. casseliflavus strains carried vanC1 and vanC2 genes, respectively.

Only two patients were colonized with vanA or vanB enterococci. The first one, an 85-year-old woman with multiple myeloma, was a carrier of vanA E. faecium and vanB E. faecalis. The second patient, a 75-year-old man with chronic lymphocytic leukemia, was colonized with vanB E. faecium.

The results from the univariate statistical analysis of the risk factors for VRE colonization are shown in Table 2. The patients included in the analysis were divided in two groups – VRE colonized and non-VRE colonized. Of all 18 VRE-positive patients, 13 were included in the first group, because their VRE colonization was hospital-acquired. The remaining 5 patients had positive cultures for vanC-carring enterococci within the first 48 h of the hospital stay and they were excluded. None of the VRE colonized patients had acquired other infection during the hospital stay. Of all 101 non-VRE colonized patients, 5 developed an infection and were excluded from the statistical analysis. Comparison between 13 patients with hospital-acquired colonization and 5 patients with community-acquired colonization is presented in Table 3.

The data revealed that age of the patients in the range from 70 to 79 years (p = 0.025) and multiple myeloma (p = 0.001) were risk factors which were significantly associated with VRE intestinal acquisition. In the multivariate analysis, patient’s age (70–79 years) was confirmed as an independent risk factor for VRE colonization.

According to the Center for Disease Control and Prevention (CDC) and the Society for Healthcare Epidemiology of America (SHEA), active screening of hospitalized patients for VRE carriage is necessary [23‐25]. Thus, it has been suggested as an effective infection prevention and control measure.

In most of studies, the majority of VRE intestinal isolates are E. faecium and E. faecalis, but we detected a low ratio of these species among patients with hematologic malignancies in our hospital. Only two patients (1.7%) were colonized with VR E. faecium and/or VR E. faecalis, whereas vanC enterococci were found in 13.4% of the patients. The detailed epidemiological investigation of the colonized patients did not reveal any housing, environmental or work-related factors leading to higher likelihood of colonization with E. gallinarum and E. casseliflavus versus VR E. faecium and/or VR E. faecalis species. Our data correlate with the study of Batistao et al. [17], who demonstrated that 77 out of 78 VRE colonized patients were positive for vanC enterococci. Tresoldi et al. [26] also published similar results – 7 out of 9 colonized patients were carriers of E. gallinarum or E. casseliflavus strains. Maschieto et al. [27] did not isolate either VR E. faecium or VR E. faecalis, but they showed a high rate of vanC-carring enterococci – E. gallinarum (23.7%) and E. casseliflavus/E. flavescens (5.2%). In contrast, Gedik et al. [5] observed that a total of 50 (39.8%) patients carried VR E. faecium or VR E. faecalis. Other investigators [28] established that 72 (31.4%) children admitted in hematology-oncology wards were carrying VRE and the number of VR E. faecium isolates was significantly higher than vanC enterococci. Suzuki et al. [29] investigated the epidemiology of BSIs for a 10-year period in Japan and found that 9% and 2.7% of them were caused by E. casseliflavus/E. flavescens and E. gallinarum, respectively. All these data testify to the increasing number of patients infected/colonized with vanC enterococci.

In the current study, we observed high diversity of VRE genes among the intestinal enterococcal isolates with a higher proportion of vanC1 gene (63.2%), followed by vanC2 gene (21%), vanB gene (10.5%), and vanA gene (5.3%). Although all vanC-carrying enterococci (n = 16, 84.2%) displayed a classical VanC phenotype, they can be reservoirs of resistance genes. Corso et al. [30] reported 14 intestinal E. gallinarum isolates, which carried dual vanC1 and vanA genes. Shirano et al. described 98 E. gallinarum intestinal isolates, from which 88 were positive for vanA and 10—for vanB. Also, there are evidence for carriage of vanC1, vanA and vanB cassette by E. gallinarum [31].

Many risk factors, which can contribute to acquisition of VRE in hematologic malignancy patients, were evaluated in the current study. Analysis of our results showed an association between advanced patient’s age (70–79 years; p = 0.025) and fecal carriage of VRE. A correlation between the age of patients and VRE colonization was also found by Metallidis et al. [10], whereas some other studies found no association between the two variables [4, 11, 17, 32]. We also found that patients who suffer from multiple myeloma were most commonly colonized with VRE (p = 0.001). The multivariate analysis identified only the advanced age of the patient as an independent risk factor for VRE acquisition. This result was expected, because the prevalence of multiple myeloma increases with age (in patients over the age of 60 years), while most of the other discussed malignancies are more prevalent in younger/middle age. The studies by Mioljevic et al. [12] and Worth et al. [33] demonstrated significantly high prevalence of VRE carriage in patients with acute myeloid leukemia (AML), who usually have a longer length of hospital stay. In our study the evaluated number of patients with AML was small and no one acquired fecal VRE.

A variety of variables were identified as significant factors for VRE acquisition among different groups of immunocompromised patients [11, 17, 34]. Suntharam et al. [11] found that prolonged length of stay, prior hospitalization, previous admission in intensive-care units and amikacin administration played a role in VRE colonization among hematology-oncology patients. In other study, nephropathy, prior antibiotic use and carbapenem administration were identified as risk factors for VRE acquisition among critically ill patients [17]. Papadimitriou-Olivgeris et al. [34] reported the quinolone use, the number of positive VRE patients in nearby beds, the chronic renal failure and the chronic obstructive pulmonary disease as risk factors for VRE colonization during hospital stay in intensive care units. In our study, there was not significant association between gender; mean length of hospital stay; underlying disease; antibiotic use; other types of medications and VRE acquisition.

Intestinal colonization with VRE precedes and plays a key role in the pathogenesis of BSIs in patients with malignancies. The data in the literature showed that the individuals who carried VRE had 24 times higher risk for developing bacteremia with these microorganisms in comparison to the non-colonized patients [6]. Study by Kara et al. [3] founded that among 193 colonized patients, 3 developed bacteremia caused by VRE and 1 of these patients have been under intensive chemotherapy. In our study, 8 of the colonized patients (6 with vanC enterococci and 2 with E. faecium and/or E. faecalis) were admitted to the hospital due to the need of a chemotherapy course. Although that the high dose chemotherapy promotes VRE-related BSI, none of the patients developed an infection. During the study period in two other hospital wards (Surgery and Intensive Care Unit) there were 3 cases of vanC enterococcal infections (2 wound infections and 1 urinary tract infection) without bacteremia.

It is interesting to note that only 13 (11.93%) out of 109 patients included in the study had used antibiotics in the last 30 days and during the current hospital stay. In contrast to our findings was a study by Mioljevic et al. [12] in which only 11 (15.4%) patients did not receive any antibiotic during the hospitalization. In the present study, the patient caring vanA E. faecium and vanB E. faecalis was treated with third-generation cephalosporin during the hospital stay, whereas the patient with E. gallinarum was on therapy with ciprofloxacin prior to hospitalization. In many studies, treatment with these antibiotics was determined as a risk factor for colonization with VRE [7, 34]. Due to this reason, we could suggest that consumption of third-generation cephalosporins and fluoroquinolones might have contributed for the acquisition of intestinal VRE in these patients.

To our knowledge, this is the first report from Bulgaria which investigated the prevalence of VRE colonization among patients with hematologic malignancies. Intestinal VRE carriage was identified in 15.1% of the patients with a significant prevalence of vanC enterococcal isolates. Among the analyzed risk factors, advanced age and multiple myeloma contributed to VRE acquisition.