Daptomycin versus Glycopeptides for the Treatment of Enterococcus faecium Bacteraemia: A Cohort Study

Background: Ampicillin resistant and glycopeptide susceptible Enterococcus faecium bloodstream infection (GSEF-BSI) incidence has risen. However, the treatment of choice remains unknown. Daptomycin use for the treatment of enterococcal infections has increased, despite effectiveness and safety concerns. The objective was to compare the effectiveness and safety of daptomycin and glycopeptides in the treatment of GSEF-BSI. Methods: This was a single-centre, retrospective observational cohort study performed at Hospital del Mar (Barcelona, Spain), from January 2006–May 2018. The primary outcome was clinical cure at the end of the therapy, and secondary outcomes included 14-day, 30-day, in-hospital mortality, and length of stay. Results: From a total of 192 patients with GSEF-BSI, 54 (28.1%) were treated with glycopeptides and 17 (8.9%) with daptomycin. Patients treated with daptomycin presented a lower clinical cure than patients treated with glycopeptides (58.8% vs. 83.3%, RR 0.416 (95% CI 0.189–0.915)). After controlling for confounding variables by means of multivariate analysis the significative difference was confirmed (aOR 4.313, 95% CI, 1.053–17.660). The need for treatment discontinuation due to adverse events was similar. Conclusions: Patients with GSEF-BSI treated with glycopeptides showed a higher clinical cure than those treated with daptomycin.


Introduction
Enterococci are ubiquitous Gram-positive bacteria that have been isolated from a wide variety of places, including the gut microbiota [1]. Although these microorganisms are considered commensal organisms of the gastrointestinal tract, they can cause serious infections such as bloodstream infections (BSI) or endocarditis, and have emerged as important nosocomial pathogens [1,2]. They are the second most common cause of healthcare associated infections in Europe and the United States of America (USA) [1].
Currently, there are two main species among enterococci that can cause infections: Enterococcus faecalis and Enterococcus faecium [1]. Although E. faecalis is more virulent and is the main enterococci involved in human infections, the incidence of E. faecium infections is rising [1]. Concerning enterococcal bacteraemia, its incidence is increasing, especially that

Results
A total of 192 patients presented a positive blood culture with GSEF during the study period (Table 1).
One hundred and twenty-one patients were excluded as they did not receive an active treatment (29,24.0%), received antibiotics less than 48 h (35, 28.9%), or were treated with other antibiotics (57, 47.1%). Finally, 71 (37.0%) patients were treated either with glycopeptides (54, 28.1%) or daptomycin (17,8.9%) and were included in the final analysis. From the 54 patients treated with glycopeptides, 6 (11.1%) received teicoplanin, while the rest were treated with vancomycin. Figure 1 summarised the study flow chart. From the 54 patients treated with glycopeptides, 6 (11.1%) received teicoplanin, while the rest were treated with vancomycin. Figure 1 summarised the study flow chart. Table 2 showed the baseline and clinical characteristics of analysed patients. The median age was 68.0 (56.0-78.0) years, most of the included patients were male and hospitalized in medical wards. Charlson Comorbidity Index was 2.0 (0-2.0), with both groups presenting similar comorbidities excepting for a higher incidence of chronic kidney disease in the daptomycin group (9.9% vs. 47.1%, p = 0.005). Concerning clinical presentation, 8 (11.3%) patients were diagnosed of septic shock and 10 (14.1%) were mechanically ventilated. SOFA and Pitt bacteraemia scores were 3.0  Table 2 showed the baseline and clinical characteristics of analysed patients. The median age was 68.0 (56.0-78.0) years, most of the included patients were male and hospitalized in medical wards. Charlson Comorbidity Index was 2.0 (0-2.0), with both groups presenting similar comorbidities excepting for a higher incidence of chronic kidney disease in the daptomycin group (9.9% vs. 47.1%, p = 0.005).
Concerning clinical presentation, 8 (11.3%) patients were diagnosed of septic shock and 10 (14.1%) were mechanically ventilated. SOFA and Pitt bacteraemia scores were 3.0 (1.0-5.0) and 1.0 (0.0-2.0), respectively. No differences were found in the clinical presentation, apart from a worse creatinine and glomerular filtration rate in the daptomycin group. Most of the bacteraemia were of high risk, being the main sources of infection intraabdominal, biliary and catheter related.

Discussion
The increase in the incidence of E. faecium bacteraemia as well as its high mortality make the study of this disease essential [3]. In this work we tried to elucidate the existing controversies about the use of daptomycin for GSEF-BSI. To the best of our knowledge, this is the first study assessing the treatment outcomes of daptomycin in the treatment of GSEF-BSI. In our study, the treatment with daptomycin showed a lower rate of clinical cure compared to glycopeptides, mainly due to the lack of clinical improvement. However, the mortality rate was similar between both groups.
Concerns about the effectiveness of daptomycin in the treatment of enterococcal bacteraemia have already been discussed based on available data with vancomycin-resistant strains [14]. In two meta-analyses of retrospective observational studies comparing daptomycin and linezolid for the treatment of vancomycin-resistant E. faecium bacteraemia, linezolid seemed to present a lower 30-day mortality, but these results were challenged in the highest retrospective observational study published where this oxazolidinone was associated with a higher risk of treatment failure and 30-day mortality [7,15]. Unfortunately, data on the outcomes of different treatments of GSEF-BSI are scarce, and the outcomes of daptomycin in this setting have never been studied.
In our study, although the small sample size, a higher risk of treatment failure was found in patients treated with daptomycin. Furthermore, a higher incidence of 14-day mortality was observed, which was not statistically significant probably due to the low sample size. Several reasons may explain these findings: daptomycin dosage, the breakpoints, and the risk of resistance development [14].
The optimal dosage of daptomycin for enterococcal bacteraemia remains unknown. It is likely that even with high doses (10-12 mg/kg/day) required pharmacokinetic/pharmacodynamic (PK/PD) targets are unachievable without safety issues with MICs > 2 mg/L [14]. From a PK/PD perspective, a free drug area under the plasma concentration-time curve to MIC ratio (fAUC/MIC) > 27.43 h.mg/L has been associated with 30-day survival in low acuity patients [18,19]. A dose of 6 mg/kg/day achieved a probability target attainment (PTA) > 90% only with MICs ≤ 1 mg/L, whereas an optimal PTA for a MIC of 2 mg/L could only be achieved with 12 mg/kg/day [18]. Higher rates of microbiological failure and mortality have been reported in BSIs caused by E. faecium strains with daptomycin MIC 3-4 mg/L when treated with daptomycin [20,21]. In fact, two recent studies found that no dose presented an acceptable PTA against a MIC of 4 mg/L in the Monte Carlo simulation [22,23]. When MICs of 2-4 mg/L were analysed, a fixed dose of 1500 mg daily was needed for ≥ 90% PTA, but this dosage carried an undue risk of toxicity as was associated with an increased 24.5-80.5% risk of achieving a Cmin ≥ 24.3 mg/L [24].
Based on these data, the CLSI revised the daptomycin breakpoints for Enterococcus spp. and established that all the MICs ≤ 4 mg/L should be reported as SDD (susceptible dose dependent), based on a dosage regimens of 8-12 mg/kg/day [25]. The EUCAST changed their breakpoints to insufficient evidence [14].
In our study the median daptomycin dose was near 8 mg/kg/day. Although this dose could play a role in the outcomes as doses > 9 mg/kg/day were related to a survival benefit and the risk of emergence of resistance [16,26,27], the CLSI recommends 8-12 mg/kg/day [25]. In addition, the median MIC was 1 mg/L, so this dosage may have been sufficient.
Another important drawback of daptomycin is the resistance development during the treatment, mainly due to changes in the LiaFSR regulatory system [28]. Eighty percent of E. faecium isolates with MICs between 3-4 mg/L harbour changes in this system, whereas these substitutions are rarely found in strains with lower MICs [28]. The size of the inoculum may also play a role [27].
To overcome these limitations, the use of combination therapy with β-lactams has been recommended based on in vitro data [29]. The treatment with these family of antibiotics may reduce the MIC to daptomycin, facilitating the achievement of optimal PK/PD indexes even in daptomycin-resistant strains, through the use of lower doses with an acceptable safety risk [28,29]. The addition of ampicillin to daptomycin resulted in bactericidal activity event at lower doses of daptomycin (8-10 mg/kg) in an in vitro biofilm model [28]. A murine model found that the addition of ampicillin in continuous infusion (but no ertapenem or ceftaroline) to daptomycin was effective, synergistic and prevented the development of resistance in the treatment of multidrug-resistant E. faecium infections [29]. Caution is nevertheless advised since the benefits of this synergy may be strain dependent and therefore not universal [28].
In our study, patients treated with vancomycin presented a higher incidence of side effects, mainly due to gastrointestinal side effects and nephrotoxicity. However, the need for treatment discontinuation was not different. The difference in the length of treatment may have played a role in these differences.
Based on the results of our study daptomycin was associated with limitations, and until evidence from randomised controlled clinical trials are available, vancomycin should be preferred over daptomycin for the treatment of GSEF-BSI. In cases of intolerance to glycopeptides, linezolid could be considered if a deep-seated infection (endocarditis, thrombophlebitis) is ruled out. Although some concerns have been raised due to its bacteriostatic activity, linezolid has proven to be similar to vancomycin in terms of effectiveness and safety in the treatment of GSEF-BSI or Staphylococcus aureus bacteraemia [6]. Dalbavancin and oritavancin are other potential options although clinical evidence on their use is still scarce [30][31][32]. Daptomycin should only be used in the treatment of GSEF-BSI when no treatment options (glycopeptides, linezolid) remain available or deep-seated infections occur. In these exceptional cases, doses of up to 12 mg/kg/daily, a close toxicity and therapeutic drug monitoring, an appropriate MIC determination and combination with continuous infusion of ampicillin could be considered to try to optimize treatment outcomes.
This study presents several limitations. It consists of a non-randomised single-centre retrospective observational study, which is subject to their typical drawbacks. The indication bias could be present, although no significant differences were found in baseline characteristics excepting from renal function. We acknowledge that a propensity-score analysis would have improved the quality of our work, although the low sample size prevented us from carrying out such analysis. The existence of other potential confounding factors not registered must be also considered. As expected, a significant incidence of polymicrobial bacteraemia was observed which may have influenced the outcomes. However, due to the difficulty of finding patients with monomicrobial bacteraemia a mixed analysis was performed, with all the concomitant microorganisms appropriately treated. We also acknowledge the small sample size. However, we were limited by the incidence of this infection in our centre. The number of included patients was low compared to the number of years studied, but only 192 patients had at least one positive culture during the studied period. Of this population, 121 had to be discarded based on the exclusion criteria. Finally, throughout the entire period only 17 patients were treated with daptomycin. Although the power of the study is limited by its small sample size, the finding of statistically significant results suggests that these findings could be maintained with larger samples.

Study Design and Setting
This was a retrospective observational cohort study conducted from January 2006 to May 2018 at the Hospital del Mar, a 420-bed university tertiary care hospital located in Barcelona, Spain. The hospital includes two different intensive care units (ICU) with a program for renal transplantation. All patients were followed-up for 3 months by electronic records after hospital discharge or death. This study was reviewed and approved by the ethics committee of the institution and followed the Strengthening The Reporting of Observational Studies in Epidemiology (STROBE) statement guidelines [33].

Participants
We identified all adult patients (>18 years) with at least 1 positive blood culture for GSEF through computer-generated microbiological data. Thereafter, electronic records were reviewed, and patients were selected if they were treated more than 48 h with glycopeptides (vancomycin/teicoplanin) or daptomycin according to the responsible physician criteria. Patients were excluded if they did not receive antibiotic therapy or if they received another antibiotic against GSEF. In patients with multiple episodes only the first episode was included.
Daptomycin, vancomycin and teicoplanin dosages were adjusted by the responsible physician according to the weight and renal function, with the possible recommendation of the infectious diseases' pharmacist or physician. Vancomycin plasmatic levels were performed throughout the study period and were adjusted by clinical pharmacists.

Variables and Data Sources
Epidemiological data included age, sex, body mass index, hospitalization ward (medical or surgical), site of acquisition (community-acquired, healthcare-associated or nosocomial [34]), underlying diseases and their severity according to Charlson index score [35], immunosuppression, neutropenia and presence of devices (vascular or urinary catheter, drainages). Clinical data included the source of bacteraemia, severity of illness and antimicrobial therapy. Severity of illness was assessed through the Pitt bacteraemia score [36], Sepsis-related Organ Failure Assessment [37], diagnosis of sepsis and septic shock according to the 2016 definitions [38] and need for mechanical ventilation on the day of the episode. In terms of antimicrobial therapy, empirical therapy, definitive treatment, time to adequate therapy, length of definitive treatment and antibiotic-related adverse events were recorded. All the variables were obtained by reviewing the electronic chart records.

Definitions
Bacteraemia was defined according to the Centre for Disease Control and Prevention guidance [39]. Immunosuppression was considered in patients receiving chemotherapy, radiotherapy or other immunosuppressive drugs including corticosteroids during the previous month. The source of bacteraemia was defined following both clinical and microbiological criteria [39,40], and was classified into two groups: low risk (urinary tract, vascular catheter and biliary tract) and high risk (all the others) [41].
Empiric therapy was deemed appropriate when at least one of the antibiotics was active in vitro against E. faecium [41]. Time to adequate therapy was defined as the days until receiving an appropriate treatment.
Clinical cure was considered when patients presented a resolution of signs and symptoms of infection [13]. For this purpose, we also evaluated laboratory parameters as C-reactive protein, procalcitonin, leucocytes and neutrophils. Otherwise, treatment failure was defined as the need for treatment change due to the persistence of infection, death or toxicity [13]. Microbiological eradication was recorded when the last blood culture drawn after initiation of therapy was negative [13] or, if not available, when clinical cure/improvement were achieved [6]. Persistent bacteraemia was defined as the isolation of GSEF in blood cultures after > 72 h of adequate treatment and relapse was defined as the isolation of positive blood culture after documented clearance within 30 days of index culture [6]. Readmissions were recorded up to 3 months after the discharge. Side events were defined as the development of an adverse event proven or suspected to be related to the agent used after the initiation of the therapy [13]. Side effects were recorded according to the information available in the medical records or analytics. Renal toxicity was assessed through the RIFLE criteria [42], whereas thrombocytopenia was defined as the decrease in the platelet count to < 75% from the baseline [6]. Anaemia was considered when a reduction ≥ 2 g/dL in haemoglobin concentration was developed [6]. Creatine phosphokinases (CPT) were measured at the discretion of the responsible physician. CPK elevation was considered as an elevated value ≥ 3 times the upper limit of normal in those with normal baseline CPK and ≥ 5 times the in those with elevated baseline CPK [24].
The primary outcome was the assessment of clinical cure at the end of the treatment. Secondary outcomes included 14-day, 30-day and in-hospital mortality, time to defervescence (time until fever returns to normal temperature), microbiological eradication, relapse, and superinfection. Finally, hospital length of stay, readmission rates and treatment-related side effects were also recorded.

Microbiology
Polymicrobial bacteraemia was considered when at least 1 non E. faecium species were isolated from the same blood culture than the enterococci [43]. The appropriateness of antibiotic therapy against these concomitant microorganisms was recorded, considering adequate when was active in vitro.
The EUCAST clinical breakpoints were applicated to determine the susceptibility to the different antimicrobials [17]. The full susceptibility test was performed by microdilution (MicroScan©, Mahwah, NJ, USA). The MIC of vancomycin, teicoplanin and daptomycin were confirmed by gradient diffusion (E-test, bioMérieux©, Marcy-l'Etoile, France).

Statistical Analysis
The sample size was limited by the low number of patients that received daptomycin over the study period. After the normality test of Kolmogorov-Smirnov, all the statistical analyses were performed with non-parametrical tests. In the case of categorical variables, these were presented as absolute numbers (percentages) and were compared using Fisher exact tests. Continuous variables were described as median (Q1-Q3) and were analysed by Mann-Whitney U test. Multivariate analysis was performed using a forward stepwise logistic regression model.
A p < 0.05 was considered statistically significant in all the analyses. Statistical analyses were performed with IBM SPSS Statistics 22.0 (IBM Co., Armonk, NY, USA).

Conclusions
Treatment with daptomycin compared to glycopeptides resulted in a lower clinical cure for GSEF-BSI, results that persisted after adjusting for confounding factors in the multivariate analysis. Treatment with this lipopeptide was also associated with a lower microbiological eradication. Both treatments showed a similar safety profile. Randomised controlled trials are needed to determine the role of daptomycin in the treatment of GSEF-BSI. Informed Consent Statement: Patient consent was waived due to the retrospective observational nature of the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy and ethical reasons.