Background
Sepsis has been recognized as a global health priority by world health organization (WHO) in 2017. Better assessment of antimicrobial resistance (AMR) of pathogens causing bloodstream infections (BSIs) is one of the most significant resolutions to stem the tide of BSIs and thus alleviate the burden of sepsis [
1‐
3]. ESKAPEEc, a cluster of notorious pathogens including
Enterococcus faecium,
Staphylococcus aureus,
Klebsiella pneumoniae,
Acinetobacter baumannii,
Pseudomonas aeruginosa,
Enterobacter spp and
Escherichia coli, have been reported to contribute 50–70% of all causative pathogens to BSIs [
4‐
6]. The striking emergence and transmission of extensive beta-lactamases (ESBLs) or carbapenemase genes among Gram-negative robs have vigorously driven the prominence of Gram-negative ESKAPEEc in BSIs. In addition, due to the implement of vaccination campaigns and comprehensive antimicrobial stewardship activities, a shift in the distribution and AMR patterns of ESKAPEEc was ongoing worldwide [
7‐
11]. However, data concerning about these changes in BSIs were lacking in China.
Moreover, foregoing studies failed to discuss these alterations in pediatric populations. A longitudinal surveillance study in United States indicated that children witnessed an increased prevalence of carbapenem-resistant Enterobacteriaceae (CRE) in BSIs from 0.0% in 1999–2000 to 4.5% in 2011–2012 [
12]. Recently, national surveillance data from China in 2013 had reported that pediatric patients witnessed highest prevalence of carbapenem resistance in
Klebsiella pneumoniae (10.6%) [
4]. However, a paucity of data has referred to the shifting of this high prevalence among pediatric BSIs over time in China.
To fill this gap, we launched a 6-year retrospective multicenter study from 2012 to 2017 to elucidate the longitudinal alterations of pathogen distributions and antibiogram of ESKAPEEc among pediatric and adult BSIs in Chongqing, Southwest China. Age distributions and AMR patterns were further compared between 2012–2014 and 2015–2017 two periods.
Discussions
To our knowledge, this study presented the current available evidence of the time shifting of ESKAPEEc distributions and AMR patterns in children and adult BSIs from China. Our findings verified the increasing prevalence of ESKAPEEc in BSIs and their distinct AMR patterns in adult and pediatric BSIs.
ESKAPEEc contributed to 58.67% of BSIs, which was consistent with previous studies [
5,
16]. Our observation of an overall increasing trend of ESKAPEEc in BSIs was largely due to the striking upward trend of
E. coli and
K. pneumoniae in adult BSIs. Similar results were also concluded by a 9-year analysis of BSIs in Rome [
5]. However, a recent long-term surveillance study at an urban hospital in Malawi showed a significant declination in the incidence of Enterobacteriaceae in BSIs from 1998 to 2016 [
7]. This discrepancy was partly due to the heterogeneity of study design, since this present study was conducted in laboratories of more than 40 hospitals and had greater geographical representation.
Interestingly, in accordance with an 11-year retrospective study in US [
17], ESKAPEEc accounted for 27.2% of all causative pathogens in pediatric BSIs, but we firstly observed a sharp drop in the prevalence of ESKAPEEc in pediatric BSIs in China, mainly driven by the slight reduction of
A. baumannii and
S. aureus. This drop may suggest that other successful species are prevalent in pediatric BSIs, but we noticed that CoNS was predominant in this present study and its predominance may override the prevalence of ESKAPEEc. Added that, data from Typhoid Surveillance in Africa Program of Sub-Saharan Africa had verified that children had significantly high odds of having a contaminated blood culture compared with adults [
15]. Therefore, we deduced that blood contamination should be blamed for our high incidence of CoNS and decreased incidence of ESKAPEEc in pediatric BSIs, even though that previous studies in the US and Chongqing also reported CoNS was the most frequently isolated from pediatric BSIs [
17,
18].
Unexpectedly, decreased resistance rates to ceftriaxone and ceftazidime but increased resistant rates to imipenem were observed among
E. coli and
K. pneumoniae. Similar trends were also observed by Lei Tian et al. [
19]. The possible explanation of this alteration was that preferential antibiotic adoption of imipenem led to antibiotic selective pressure. Since ESBLs are notorious in nosocomial infection, most of empirical antibiotic therapies of BSIs are initiated with imipenem other than the third- or four- generation cephalosporins. Then, high consumption leads to high resistance. Recent studies in China and abroad have demonstrated that increased usage of carbapenems accelerates the production of carbapenemase, and then high proportion of CRKP [
20‐
22].
Of note, our study alarmed high prevalence of CRKP (12.79%) among pediatric BSIs, which was higher than data reported by CHINET [
4]. However, a comparative study of neonatal BSIs in Chongqing revealed that all of 49 strains were susceptible to imipenem [
18]. Our high prevalence may due to the heterogeneity of resistant element among strains and the dissemination of conservative mobile elements. Nationwide surveillance of CRE in China had confirmed carbapenemase production was the primary mechanism of carbapenem resistance in CRE. The mobile element of
ISKpn27-blaKPC-2-ISKpn2 played an important role in CRKP transmission [
23]. However, in pediatric patients, several studies have reported the predominance or outbreak of
blaNDM-1 among CRKP in Beijing, Shanghai and Shandong [
24‐
26]. Our previous studies also demonstrated a high prevalence of
blaNDM-1 among CRKP in Chongqing [
27,
28]. Therefore, more researches are ongoing to uncover the underlying molecular mechanisms of the high prevalence of CRKP in children BSIs and cautious adoption of carbapenems to fight against pediatric CRKP BSI is advocated.
The prevalence of carbapenem resistance among
E. coli was 0.76%, which was higher than results from an epicenter of CRE in the US with a ratio of 0.1% [
29]. Moreover, it was noteworthy that eight CRECO strains were isolated from pediatric BSIs, suggesting the emergence of carbapenem resistance phenotypes among pediatric inpatients in Chongqing. Despite this, carbapenems or piperacillin/tazobactam are still active in our settings; less than 3% strains were resistant to them.
Accompanied with declined proportion of
A. baumannii in BSIs, reduced resistance rates to most of antibiotics were also observed, especially to carbapenems (From 63.61 to 55.86%). This rate of carbapenem resistance was lower than our previous study suggesting 71.19% of CRAB in BSIs [
30], but similar decreasing trend was reported in Dhaka [
9]. This reduction may be owing to the implementation of CRASS strategies to fight against carbapenem-resistant robs, such as resistance surveillance, colonization clearance and environmental monitoring. However, it was noticed that in spite of this decreasing trend, resistance rates of
A. baumannii to all these antibiotics were 60% in adult BSIs and 25% in pediatric BSIs, respectively, hence, effective treatments to fight against
A. baumannii BSIs are still in urgent need. Recently, Tigecycline Evaluation and Surveillance Trial (TEST) concerning global blood-borne pathogens reported 70.8% of
A. baumannii were susceptible to minocycline and MIC
90 of tigecycline against it was 2 μg/mL, suggesting the potential efficiencies of these two antibiotics to overcome
A. baumannii BSIs [
13].
Moreover, TEST further reported high prevalence of MRSA and VREFM worldwide, with the proportions of 33.0 and 27.6%, respectively [
13], while our study observed relatively low prevalence, with 28.91 and 2.2%, respectively. In contradistinction to the sharp reduction of MRSA in US [
10], our finding suggested the proportion of MRSA remains constant. Despite penicillin resistance was dramatically increased and ampicillin resistance was consistently high among
E. faecium, vancomycin or linezolid is still efficient to fight against Gram-positive ESKAPEEc in our settings.
Several limitations should be considered in our study. Firstly, although this study included almost all the microbiology laboratories in the secondary or tertiary hospitals in Chongqing, Southwest China, it excluded some laboratories which failed to perform antimicrobial susceptibility testing. Careful interpretation of our findings was advocated in different hospital settings. Secondly, this multicenter study was focused on laboratory based, blood-culture proven BSIs attributable to ESKAPEEc, it, therefore, was not designed to analyze blood culture contamination and to discuss hospital-onset or community-onset without sufficient clinical patient data. Thirdly, our study excluded the isolates with antimicrobial susceptibility results reported by Kirby-Bauer, which might lead to an overestimation of the incidences of antimicrobial resistance.