Low Proportion of Linezolid and Daptomycin Resistance Among Bloodborne Vancomycin-Resistant Enterococcus faecium and Methicillin-Resistant Staphylococcus aureus Infections in Europe

Vancomycin-resistant Enterococcus faecium (VREF) and methicillin-resistant Staphylococcus aureus (MRSA) are associated with significant health burden. We investigated linezolid and daptomycin resistance among VREF and MRSA in the EU/EEA between 2014 and 2018. Descriptive statistics and multivariable logistic regression were used to analyze 6,949 VREF and 35,131 MRSA blood isolates from patients with bloodstream infection. The population-weighted mean proportion of linezolid resistance in VREF and MRSA between 2014 and 2018 was 1.6% (95% CI 1.33–2.03%) and 0.28% (95% CI 0.32–0.38%), respectively. Daptomycin resistance in MRSA isolates was similarly low [1.1% (95% CI 0.75–1.6%)]. On the European level, there was no temporal change of daptomycin and linezolid resistance in MRSA and VREF. Multivariable regression analyses showed that there was a higher likelihood of linezolid and daptomycin resistance in MRSA (aOR: 2.74, p < 0.001; aOR: 2.25, p < 0.001) and linezolid in VREF (aOR: 1.99, p < 0.001) compared to their sensitive isolates. The low proportion of linezolid and daptomycin resistance in VREF and MRSA suggests that these last-resort antibiotics remain effective and will continue to play an important role in the clinical management of these infections in Europe. However, regional and national efforts to contain antimicrobial resistance should continue to monitor the trend through strengthened surveillance that includes genomic surveillance for early warning and action.


INTRODUCTION
Staphylococcus aureus and Enterococcus faecium are Grampositive pathogens that frequently colonize the skin, nostrils, and gut of humans with potential for invasive infections in humans (Fisher and Phillips, 2009). They are the most important Grampositive organisms causing nosocomial infections especially bloodstream and other invasive infections (Suetens et al., 2018). These pathogens have shown combined resistance to multiple antimicrobial classes including beta-lactams, cephalosporins, fluoroquinolones etc. making their treatment increasingly difficult (Hollenbeck and Rice, 2012;O'Driscoll and Crank, 2015;Lakhundi and Zhang, 2018). Consequently, they are responsible for significant health and economic burden on hospitalized patients, families and society (Zhen et al., 2019).
In Europe and many other countries across the world, Enterococcus spp. and S. aureus infections are commonly observed among hospital patients with high vancomycin and methicillin resistance proportions, respectively (Lakhundi and Zhang, 2018;García-Solache and Rice, 2019). Moreover, these multidrug-resistant pathogens have been associated with significant mortality and morbidity in several healthcare settings (Caballero-Granado et al., 2001;Pinholt et al., 2014;Boncagni et al., 2015;Brady et al., 2017;Nelson et al., 2017;Kramer et al., 2018;Cassini et al., 2019). Across Europe, the populationweighted proportion of methicillin resistance among S. aureus is 16.4%, while vancomycin resistance among E. faecium isolates was estimated to be 18.3% in 2019 (Ayobami et al., 2020a;European Centre for Disease Prevention Control, 2020). Recent studies have demonstrated the rising trend of vancomycin resistance in Enterococcus spp., especially in E. faecium, and its pervasiveness across Europe (Markwart et al., 2019;Ayobami et al., 2020a).
While a declining trend of MRSA is reported in Europe, the incidence is increasing among the extremes of age groups, and its resistance proportion is still high in many countries of the European Union (Cassini et al., 2019;European Centre for Disease Prevention Control, 2020). The diminishing antibiotic arsenal, widespread resistance to several antibiotics, and other pharmacological concerns, renders antimicrobials such as linezolid and daptomycin, the antibiotics of last-resort to manage VREF and MRSA infection especially among critically ill hospitalized patients who are at the greatest risk of mortality (Rodvold and McConeghy, 2014;Hashemian et al., 2018;Heidary et al., 2018). Despite the therapeutic utility of these drugs, there is an increasing report of linezolid, and daptomycin resistance among patients infected with VREF and MRSA worldwide (Miller et al., 2016;Bender et al., 2018a;Heidary et al., 2018).
Ensuring the effectiveness of these last resort antibiotics is a priority globally, considering the severity of multi-drug resistant Gram-positive bacterial infections, broken antibiotic market and the prevailing insufficient public health actions to address antimicrobial resistance. It is on this premise that this study sought to characterize the magnitude and trend of resistance of VREF and MRSA to linezolid and daptomycin using European-wide surveillance data (2014-2018) from patients with bloodstream infections.

Outcomes, Study Design, and the European Antimicrobial Resistance Surveillance Database
The primary outcome was the population-weighted proportion of linezolid resistance among vancomycin-resistant E. faecium, and resistance to linezolid and daptomycin among methicillinresistant S. aureus isolates. We conducted a retrospective observational study on E. faecium and S. aureus (2014-2018) using data retrieved from the European Antimicrobial Resistance Surveillance Network (EARS-Net) TESSy database with the approval of the European Centre for Disease Prevention and Control. EARS-Net is a network of European surveillance systems that collects routine clinical antimicrobial susceptibility (AST) data on invasive isolates [blood and cerebrospinal fluid (CSF)] from the 27 countries in the European Union as well as Norway, Iceland, and the United Kingdom (European Centre for Disease Prevention Control, 2018a).

Selection of Isolates
The TESSy database of EARS-Net only includes the first isolate from a given patient in the respective year. To identify unique isolates, we created a composite identifier for each isolate using "R" (v. 3.6.1). The composite identifier is composed of the (i) reporting country, (ii) unique laboratory identifier, (iii) hospital identifier, (iv) patient identifier, (v) date of sample collection, and (vi) the identified pathogen. These variables are part of the original dataset from EARS-Net and are provided by the participating countries. Isolates with duplicate composite identifiers, more than one AST against the same antibiotic, and those not assigned a hospital ID were excluded. Only E. faecium isolates that were tested for vancomycin susceptibility were included. For S. aureus, only isolates that were tested for "methicillin resistance" (see the section on Variables and Definitions below) were included for analysis.

Outcomes and Statistical Analyses
The population-weighted proportion of antibiotic-resistant E. faecium and S. aureus among all tested isolates was expressed as a percentage with its 95% confidence intervals (95% CI). To assess the potential association of last-resort antibiotic resistance with ward type (i.e., ICU vs. non-ICU) and year of sampling (temporal trend), respectively, multivariable logistic regression analyses were performed including the following independent variables: Year of sampling, gender, age group, European region, and ward type. These variables were selected before the analysis based on the availability of data and our prior hypotheses about variables that may be associated with linezolid and daptomycin resistance in VREF and MRSA. All variables were treated as categorical variables, except year of sampling which was treated as a continuous variable. All statistical analyses were performed using R version 3.6.1 (R Core Team, 2019) and the "survey" package (version 3.37) (Lumley, 2020). For all analyses in all strata (i.e., descriptive analyses of resistance proportions, logistic regression analyses, and Chi 2 -tests) we accounted for clustering at hospital level and applied European region population-based weighting (Ayobami et al., 2020b). The population of the four European regions was calculated from the individual populations of the region's countries. Population data of individual countries were obtained from the Eurostat database (Eurostat, 2020). European regions population weighting was used to ensure that the data from each major European region contributed proportionally (in relation to their respective population size) to the calculation of resistance proportions. This was done to minimize bias from significant differences in isolate numbers from various European regions.

Daptomycin Resistance in MRSA
Between 2014 and 2018, the daptomycin resistance proportion among MRSA isolates from patients with bloodstream infections was 1.11% (95% CI 0.75-1.63%) ( Figure 3A) (Figure 3C). This finding was also observed after adjusting for other potential predictors in a multivariable regression analysis, that showed that there was a higher likelihood of daptomycin resistance in MRSA blood isolates from ICUs compared to non-ICU wards [adjusted OR: 2.65 (95% CI 1.13-6.26), p = 0.0245] ( Table 4).

DISCUSSION
In this study, we examined linezolid and daptomycin resistance patterns among VREF and MRSA isolates isolated from patients with bloodstream infections in Europe. In sum, resistance proportions of the analyzed last line antibiotics in VREF and MRSA are relatively low in Europe. The population-weighted mean proportion of linezolid resistance among VREF isolates was 1.6% (95% CI 1.3-2.0%). While the in-country proportion varies across several local settings within Europe (Gawryszewska et al., 2016;Sassi et al., 2019;Xanthopoulou et al., 2020), this European mean proportion is comparably similar to what is seen in many within-country studies from other regions of the world such as South Korea, China, India, Iran, and United States where it remains less than 2% (Flamm et al., 2015;Houri et al., 2017;Yadav et al., 2017;Bi et al., 2018;Cho et al., 2018;Zhang et al., 2018;Huang et al., 2019). Also, the European mean proportion of linezolid resistance among MRSA blood isolates included in this study was found to be very 1ow [0.29% (95% CI 0.21-0.40%)], attesting to the overall susceptibility of MRSA isolates to linezolid in Europe. This is also similar to local studies in Germany and Spain (Sierra et al., 2013;Yayan et al., 2015), Russia (Gostev et al., 2015), China (Huang et al., 2019), and Latin America (Vega and Dowzicky, 2017), but remarkably lower compared to results from Pakistan (48.1%) (Azhar et al., 2017) and India (2.8-7.0%) (Kaur and Chate, 2015;Kumar, 2016). In comparison to other multi-country surveillance studies, such as ZAAPS and SENTRY, linezolid activity against VREF and MRSA remains similarly very high (>98%) in Europe (Mendes et al., 2016;Deshpande et al., 2018).
Our results suggested no significant change in the linezolid resistance proportions among MRSA and VREF isolates over the study period in Europe despite the reported increasing trend of linezolid consumption between 2009 and 2018 in Europe (European Centre for Disease Prevention Control, 2019a; Kramer et al., 2019). We also found that VREF and MRSA isolates are more likely to be linezolid-resistant than the vancomycin-sensitive and methicillin-sensitive isolates, respectively. While previous studies have shown linezolid resistance can be seen in both VREF and vancomycin-sensitive isolates (Allen and Bierman, 2009), the increased likelihood among VREF and MRSA isolates suggests the higher chance of co-resistance to linezolid among VREF and MRSA isolates under selective pressure such as linezolid exposure (Cantón and Ruiz-Garbajosa, 2011). Whether this is driven mainly by de novo mutations and/or horizontal transfer of resistance genes is for future studies to clarify. However, this observation reinforces the need to preserve the effectiveness of vancomycin and penicillinase-resistant β-lactams, to reduce the use of linezolid and safeguard its efficacy as an important lastline antibiotic. In line with the findings for linezolid resistance, the European population-weighted mean proportion of daptomycin resistance among MRSA isolates was similarly low [1.0% (95% CI 0.82-1.28%)]. The low proportion is comparable to the non-susceptibility proportion recorded in a multicenter ICU surveillance studies from Canada (Denisuik et al., 2018), China (Qin et al., 2016), and a global surveillance that reported daptomycin resistance in MRSA isolates of bone and joint infections . Similar to linezolid resistance, our results showed daptomycin resistance among MRSA isolates is substantially more likely compared to MSSA isolates. Our data showed MRSA blood isolates from the ICU are more likely to be linezolid and daptomycin resistant compared to isolates from non-ICU wards of the hospital. This might be due to the prominence of MRSA in the European ICUs (Vincent et al., 1995;Pujol et al., 1996;Boncagni et al., 2015;European Centre for Disease Prevention Control, 2018c) where linezolid and daptomycin are frequently prescribed antibiotic (Davis et al., 2005;Rodríguez et al., 2009;Hashemian et al., 2018;Kramer et al., 2019). These results can't overstate the need for early initiation of effective antibiotics to treat MRSA infections especially among vulnerable patients that usually populate ICU wards, as a matter of patient safety.
Various mechanisms underlie the resistance of gram-positive pathogens to linezolid including mutation of 23S rRNA, ribosomal proteins (L3, L4) but very few studies have reported the  transmissible optrA, cfr, and poxtA mediated linezolid resistance in Europe. The recent report of the highest prevalence (22.7%) of optrA, and poxtA genes ever in human linezolid-resistant enterococci isolates in Ireland (Egan et al., 2020), highlights its potential clinical and surveillance challenge in Europe (Feßler et al., 2013;Brenciani et al., 2015;Antonelli et al., 2016;de Dios Caballero et al., 2016;Gawryszewska et al., 2017;Lazaris et al., 2017;Bender et al., 2018a,b;Ruiz-Ripa et al., 2020, 2021. Experience with the evolution of antibiotic resistance warns that the present effectiveness comes with a price of triggered treatment failure among future patients. With the treatment difficulties associated with VREF and MRSA, an increasing reliance on linezolid and daptomycin is expected since both have been shown to be equally efficacious in the treatment of VREF and MRSA (Crank et al., 2010;Gonzalez-Ruiz et al., 2011;Twilla et al., 2012;Heidary et al., 2018). Prevention of resistance to linezolid and daptomycin will require multi-pronged approaches including genomic surveillance, screening of high-risk patients where beneficial, combination antibiotic therapy, and enhanced infection prevention and control strategies (Gonzales et al., 2015;Yim et al., 2017). Therefore, the low proportion of resistance to linezolid and daptomycin among VREF and MRSA isolates in Europe should be seen as a window of opportunity to strengthen existing AMR prevention and control measures, and tackle the difficulties plaguing the research and development of new antibiotics (McKenna, 2020).

Strengths and Limitations
With 7,000 and 35,000 VREF and MRSA isolates from patients with bloodstream infections, respectively, this study is the largest and most comprehensive analysis of last line antibiotic resistance profiles among these difficult to treat multidrug resistant pathogens in the EU/EEA. The analyzed dataset shows a high level of representativeness for the EU/EEA region, since the isolates were derived from routine clinical microbiological sampling and antimicrobial susceptibility testing and were transmitted to EARS-Net from 29 EU/EEA countries and the United Kingdom. Notably, population, hospital, and isolate sample representativeness was assessed as "high" in 25 countries (European Centre for Disease Prevention Control, 2019b). In addition, the validity of these AMR data is regularly assessed in external quality assessments of participating laboratories (European Centre for Disease Prevention Control, 2018b) ensuring high data quality.
A potential limitation is the variation in population coverage among reporting countries, although half of the participating countries reported a national coverage greater than 80% (European Centre for Disease Prevention Control, 2019b). To minimize possible bias from differences in population size and isolate numbers from various countries, EU/EEA-wide resistance proportions and multivariable regression analyses were population-weighted based on the population sizes of the individual regions. Moreover, differences in sampling routines, admission characteristics (e.g., stay duration, bed space density) and healthcare resources across EU/EEA countries can result in biased estimates of resistance proportions. However, to reduce the extent of this potential bias, in EARS-Net only isolates from bloodstream infections are collected, since microbiological sampling routines for these infections are generally similar across countries of the EU/EEA, even though some variations (such as frequency of sampling) between hospitals and countries cannot fully excluded.
Also, despite the off-label use of linezolid as a therapy option in some settings, it is questionable whether blood isolates are the most appropriate to determine linezolid resistance among VREF isolates from patient with nosocomial bloodstream infections because of its non-bactericidal effects. Therefore, our results should be interpreted with these limitations in mind.

CONCLUSION
Proportions of linezolid and daptomycin resistance among VREF and MRSA blood isolates remain low in the EU/EEA throughout the study period. However, VREF and MRSA were consistently more resistant to linezolid and daptomycin compared to their sensitive isolates. MRSA blood isolates from the ICU were more likely to be linezolid and daptomycin resistant compared to isolates from non-ICU wards of the hospital. In addition to existing antibiotics stewardship programs, it is necessary to strengthen surveillance in Europe, especially the genomic characterization of the resistance genes that compromise the efficacy of these last line antibiotics in VREF and MRSA.

DATA AVAILABILITY STATEMENT
The data analyzed in this study is subject to the following licenses/restrictions: The datasets will be made available with the express permission of the European Centre for Disease Control (ECDC) Stockholm Sweden. The ECDC on behalf of the EU/EEA countries have the exclusive control over the sharing of the data. Requests to access these datasets should be directed to DATA ACCESS REQUEST European Centre for Disease Prevention and Control (ECDC) Gustav III:s boulevard, 40 16973 Solna Sweden. Further inquiries should be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
OA, RM, and TE: conceptualization. RM and OA: original draft. NW, RM, and OA: statistical analysis. RM, OA, GW, and TE: critical review, revision, and Editing. RM, OA, NW, GW, and TE: final approval. All authors contributed to the article and approved the submitted version.

FUNDING
The study was funded with the internal funds of the Robert Koch Institute, a federal health agency of the German Ministry of Health.

ACKNOWLEDGMENTS
We thank our colleague Angelina Taylor at the Robert Koch Institute for her contribution to this study. We would like to also thank all EARS-Net participating laboratories and hospitals in the EU/EEA countries for providing data. We wish to thank the European Centre for Disease Prevention and Control for providing the dataset for our paper. "The views and opinions of the authors expressed herein do not necessarily state or reflect those of ECDC. The accuracy of the authors' statistical analysis and the findings they report are not the responsibility of ECDC. ECDC is not responsible for conclusions or opinions drawn from the data provided. ECDC is not responsible for the correctness of the data and for data management, data merging and data collation after provision of the data. ECDC shall not be held liable for improper or incorrect use of the data".