Risk factors associated with the acquisition of linezolid-resistant Enterococcus faecalis

Background Linezolid is used to treat vancomycin-resistant enterococcal infections. Vancomycin-resistant and -susceptible Enterococcus faecalis can develop resistance to linezolid in clinical settings with high linezolid consumption. Linezolid-resistant E. faecalis (LREF) has emerged as a major nosocomial pathogen in our hospital, in which linezolid use is considerable. Aim To dene risk factors and outcomes associated with LREF infections. Methods A retrospective case-control study was designed to evaluate patients admitted to Hospital Civil de Guadalajara “Fray Antonio Alcalde” from January 2014 to October 2017. Fifty patients meeting case denitions for LREF infection and 100 controls hospitalized in the same rooms and dates as the cases were included. Clinical and demographic data were collected and analyzed.Results Risk factors for LREF, included hospitalization within the previous 6 months, intensive care unit admission, previous surgery, urinary catheterization, parenteral nutrition, and acute renal disease. Multivariate analysis identied prior exposures to linezolid [odds ratio (OR, 6.7] and clindamycin (OR, 6.7), previous hospitalization (OR, 2.8), previous surgery (OR, 5.7), and parenteral nutrition (OR, 3.4) as risk factors for LREF infection. The mortality rate for cases was 18% on the LREF versus 9% for controls. Conclusion Risk factors for LREF infections include antibiotic exposures to linezolid, clindamycin, third-generation cephalosporins, meropenem, and colistin; previous hospitalization; intensive care unit admission; previous surgery; and the use of parenteral nutrition.

Enterococci exhibit intrinsic resistance to multiple antibiotic classes, including aminoglycosides, cephalosporins, streptogramins, and lincosamides [2,3], with resistance noted more frequently in E. faecium than in E. faecalis. In addition, enterococci can develop resistance to linezolid, daptomycin, and vancomycin during treatment [4].
Although generally considered less capable of developing resistance than E. faecium, E. faecalis has emerged as a multidrug-resistant bacterium that may harbor linezolid resistance [5]. Linezolid-resistant E. faecalis (LREF) was rst described in a 2002 report from the UK describing two E. faecium and one E. faecalis isolates with linezolid minimum inhibitory concentrations of 64 mg/L obtained from patients with prior linezolid exposures [6]. In the following years, multiple reports appeared, including two cases of LREF infections in patients who received linezolid previously for the treatment of vancomycin-resistant E. faecium infections [7] and in patients that received prolonged (>30 days) linezolid courses for enterococcal [8], and mycobacterial infections [9,10] Here, we aimed to de ne risk factors and outcomes associated with LREF infections.

Study design
A retrospective case-control study of patients hospitalized from January 2014 to October 2017 at the Hospital Civil de Guadalajara "Fray Antonio Alcalde" was designed. We evaluated 50 patients de ned as cases (patients with a culture positive for LREF) and 100 de ned as controls (patients hospitalized in the same rooms and dates as cases), with two controls per case.
Control patients were selected from the same source population the case-patients to prevent biased estimates of relative risk that occurs when designating patients with positive cultures for susceptible bacteria as the control group [19,20]. We excluded patients hospitalized for <48 hours. Risk time was de ned as the number of days from admission to being diagnosed as having a positive culture; for controls, exposure data were collected from the date of admission until the date of discharge or death.
The US Centers for Disease Control and Prevention/National Healthcare Safety Network surveillance de nitions of healthcare-associated infections were used: the presence of purulent drainage in skin/soft tissue and surgical wound infections; bacterial isolates from more than one blood culture bottle for bloodstream infections; a positive culture from purulent material obtained during surgery for intraabdominal infections; fever and a positive urine culture for urinary tract infections; and for respiratory tract infections, a positive respiratory specimen in the presence of fever, leukocytosis, increased respiratory secretions, and tachypnea.

Data and analysis
Clinical and demographic data were collected for cases and controls from clinical records. These included previous hospitalizations (within 6 months), prior antibiotic exposures (within 30 days), and time of discharge. The Charlson comorbidity index was used to assess comorbidities. Linezolid use was quanti ed by calculating DDD per 100 bed-days.
Identi cation and drug susceptibility were determined using the automated VITEK-2 system (Biomérieux, Lyon, France). Linezolid resistance was de ned by a minimum inhibitory concentration >4 µg/mL. Data were analyzed using the T-test and 2 test in SPSS v. 24, and logistic regression analysis was conducted to calculate odds ratios (ORs).

Ethics statement
The local ethics and research committee Hospital Civil de Guadalajara. Fray Antonio Alcalde (121-17) and the ethics, research and biosafety committee (University of Guadalajara) approved the study (CI-058-18).

Results
The characteristics of the 50 cases and 100 controls are summarized in Table 1. Case-patients had a higher length of stay on average, 35.0 days compared to 11.1 days for controls (p < 0.001), with a risk time of 0.044. LREF was recovered from skin and soft tissues in 32% of case-patients, followed by blood (19%), intra-abdominal (16%), urine (14%), surgical wound (13%), and respiratory specimens (6%).
Risk factors for LREF, indicated by a signi cantly higher rate among cases than controls, included hospitalization within the previous 6 months, intensive care unit admission, previous surgery, urinary catheterization, parenteral nutrition, and acute renal disease (Table 1). Antibiotic exposures within the previous 30 days, particularly to third-generation cephalosporins, meropenem, linezolid, clindamycin, and colistin, were signi cantly increased among case-patients (Table 1).
Multivariate risk factor analysis indicated that LREF was more frequent in patients with previous hospitalization (OR 2.8), previous surgery (OR 5.7), parenteral nutrition (OR 3.4), exposure to clindamycin (OR 6.7), or exposure to linezolid (OR 6.7) ( Table 2).

Discussion
Infections due to drug-resistant bacteria including LREF are emerging as an important challenge in healthcare settings [21,22]. The need for effective control policies to prevent drug-resistant nosocomial infections became apparent during increasingly frequent hospital outbreaks [23].
The predominant mechanism of enterococcal linezolid resistance is conferred by the G2576T mutation in the 23S rRNA gene. Other mechanisms include mutations in genes encoding the L3 and L4 ribosomal proteins, and in two plasmid-borne genes, cfr and optrA [1]. The cfr gene was rst reported in 2000 in Staphylococcus sciuri [24]. Linezolid-resistant enterococcal infections were rst identi ed during nosocomial outbreaks of E. faecium resistant to both vancomycin and linezolid [25][26][27][28].
Prior linezolid use was one the most signi cant risk factors for LREF infection identi ed in this study, similar to the ndings of previous reports. Intensive care unit outbreaks of linezolid-resistant enterococci, including LREF, have occurred in the setting of prolonged linezolid treatment courses [12,15,17,29,30]. Other reported risk factors for linezolid-resistant enterococcal infections include immunosuppression, neutropenia, and invasive procedures [31]. In our cases, leading risk factors included the previous use of antibiotics (chie y linezolid), previous hospitalization, and surgery.
Linezolid resistance can be acquired by multiple bacterial species through horizontal transmission of genetic elements [32][33][34]. The cfr gene was rst documented in our hospital in 2009 in three LREF isolates [35]. All isolates were susceptible to tetracycline, tigecycline, daptomycin, and vancomycin. We detected linezolid-resistance in Staphylococcus epidermidis, and Staphylococcus cohnii [35], and E. faecalis in our hospital at approximately the same time, which suggests that cfr in staphylococci may act as a reservoir for this resistance factor.
Linezolid was approved for clinical use in the United States in 2000, two years after its approval in Mexico. Linezolid consumption in our hospital has increased progressively over the past 17 years owing to its use in tuberculosis treatment as a parenteral drug when oral administration is not available. Usually, the duration of the initial stabilization period is 15-30 days, within the time range identi ed in a 2018 report as risk factor for LVEF acquisition; the mean linezolid treatment duration was 29.8 ± 48.8 days [36]. Prolonged linezolid exposure was also documented as a risk factor for LVEF infections in 2004 and 2017 [8,13].
In hospitals where vancomycin-resistant enterococci and linezolid-resistant enterococci co-circulate, extreme care should be taken to optimize empiric and directed therapies [37]. Delays in initiating appropriate therapy of enterococcal bloodstream infections can increase 30-day mortality signi cantly [37]. In our hospital, linezolid use is substantial, particularly during the initial, aggressive treatment of severe tuberculosis. Prolonged treatment courses for tuberculosis have likely facilitated the emergence of linezolid resistance in Gram-positive bacteria [18,36]. A dedicated antimicrobial stewardship intervention intended to reduce linezolid use in our hospital is needed. A Spanish hospital reduced its linezolid consumption by 76%, while seeing a reduction in LREF isolation, after initiating a focused antimicrobial stewardship program [38].
Our study has several limitations. Firstly, stool cultures were not subjected to assess fecal carriage of LREF. Secondly, the mechanisms of resistance in our LREF isolates have not yet been de ned.

Conclusions
LREF has emerged as a signi cant nosocomial pathogen in our hospital, in association with a high linezolid consumption load. Patients with previous hospitalization, surgery, parenteral nutrition, previous use of meropenem, prior use of clindamycin, and past linezolid exposure appear to be at increased risk of acquiring LREF infections.

Declarations Data availability
The data used to support the ndings of this study are included within the article. These data and further information are available on request.

Con icts of interest
The authors declare that they have no con icts of interest.  Tables   Table 1. Characteristics of cases of patients infected with LVEF and controls.