Systematic review of carbapenem-resistant Enterobacteriaceae causing neonatal sepsis in China

Background Carbapenems are β-lactam antibiotics which are used to treat severe infections caused by multidrug resistant Enterobacteriacea. The recent emergence and rapid spread of Enterobacteriaceae resistant to carbapenems is a global concern. We undertook a systematic review of the antibiotic susceptibility and genotypic characteristics of carbapenem-resistant Enterobacteriaceae in Chinese neonates. Methods Systematic literature reviews were conducted (PubMed/Medline, Embase, Wanfang medical online databases, China National Knowledge Infrastructure (CNKI) database) regarding sepsis caused by carbapenem-resistant Enterobacteriaceae in Chinese neonates aged 0-30 days. Results 17 studies were identified. Eleven patients in the six studies reported the source of infection. Ten patients (10/11, 90.9%) were hospital-acquired infections. Genotypic data were available for 21 isolates in 11 studies (20 K. pneumoniae, 1 E. coli). NDM-1 was the most frequently reported carbapenem-resistant genotype (81.0%, 17/21). Carbapenem-resistant Klebsiella pneumoniae and Escherichia coli were resistant to many antibiotic classes with the exception of colistin and fosfomycin. Sequence type 105 (ST105) was the most commonly reported K. pneumoniae ST type (30.8%; 4/13), which was from the same hospital in Western China. ST17 and ST20 were the second and third most common K. pneumoniae ST type, 23.1% (3/13) and 15.4% (2/13) respectively. The three strains of ST17 are all from the same hospital in central China. The two strains of ST20, although not from the same hospital, belong to the eastern part of China. Conclusions Klebsiella pneumoniae with the NDM-1 genotype was the leading cause of neonatal carbapenem resistant sepsis in China. Hospital acquired infection is the main source of carbapenem resistant sepsis. There is currently no licenced antibiotic regimen available to treat such an infection in China. Improved surveillance, controlling nosocomial infection and the rational use of antibiotics are the key factors to prevent and reduce its spread.


Background
According to the global reports, in 2013, 51.8% of the 6.3 million children under the age of five died of infectious diseases, while 44% (276.1 million) died during the neonatal period. Neonatal sepsis is the third leading cause of neonatal death, killing 0.421 million neonates worldwide in 2013 [1]. The overall incidence of neonatal sepsis in four Asian centres (including mainland China, Thailand, Macau, and Malaysia) was 26.1 (95% CI 24.5 to 27.8) per 1000 admissions and Klebsiella spp. was the most common Gram negative organism causing most deaths [2]. Laxminarayan et al. [3] reported that 214 000 of 690, 000 annual neonatal deaths (31%) associated with sepsis are potentially attributable to antimicrobial resistance. Carbapenems are beta-lactam antibiotics which are used to treat severe infections caused by multidrug resistant Enterobacteriaceae, such as Klebsiella pneumoniae (K. pneumoniae) and Escherichia coli (E. coli). The recent
Recent studies suggest that carbapenem resistance is increasing in China. A national report using data from CHINET (a Chinese antimicrobial resistance surveillance network) has shown that the overall prevalence of imipenem-resistant K. pneumoniae increased from 3.0% to 20.9% and meropenem-resistance from 2.9% to 24.0% between 2005 and 2017. These data included both children and adults and most of the samples were from sputum and urine. Among the five children's hospitals, the resistance rate of K. pneumoniae isolated from one hospital to imipenem was 2.5%, while from the other four hospitals resistance rates ranged from 32.1% to 45.5%. Little information was available on age ranges and types of samples [6]. This systematic review aimed to summarize the current data from both English and Chinese language sources on the antibiotic susceptibility and genotypic characteristics of carbapenem-resistant Enterobacteriaceae (K. pneumoniae and E. coli) causing neonatal sepsis in China.

Definitions
Carbapenem-resistance was defined as resistance to any one of meropenem, imipenem, or ertapenem according to the US Central Laboratory Standards Institute (CLSI). In 2015 the breakpoint was changed from 2010. Laboratories using Enterobacteriaceae minimal inhibitory concentration (MIC) interpretive criteria for carbapenems described in M100-S20 (January 2010) performed the modified Hodge test (MHT), Carba NP test and/or a molecular assay when isolates of Enterobacteriaceae were suspicious for carbapenemase production based on impipenem or meropenem MICs of 2-4 ug/ml or ertapenem MIC of 2 ug/ml in 2015 [7]. Carbapenemresistant K. pneumoniae or E. coli sepsis was defined as a laboratory confirmed culture of K. pneumoniae or E. coli obtained from the blood accompanied with signs and symptoms of infection [8]. Neonates were defined as age 0-30 days [9].

Search strategy and selection criteria
This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidance (PRISMA) [10]. We searched the published literature from PubMed/Medline, Embase, China National Knowledge Infrastructure (CNKI) and Wanfang med online databases between January 1, 2000, and June 28, 2018. We used the search terms ("beta-lactamases/or carbapenemase, or carbapenem resistance/ resistant or drug resistance or carbapenemase* or carbapenem adj1 resist* or MBL or metallo-b-lactamase or VIM or NDM or OXA or oxacillinase or IMP or KPC or Klebsiella pneumoniae carbapenemase or OmpK") AND ("Enterobacteriaceae/or enterobacteriaceae or Escherichia/or Escherichia or Escherichia coli or Klebsiella or Klebsiella/Klebsiella pneumoniae/Klebsiella oxytoca") AND ("China or Chinese") AND ("neonate or newborn or infant") for English databases. We used search terms ("Carbapenems or Carbapenem") AND ("antibiotic resistance") AND ("infant", OR "neonatal") for Chinese databases. We limited the searches to Chinese territories, including Taiwan, Hong Kong, and Macau. The full search strategy is available in Additional file 1: Table S1.

Inclusion and exclusion criteria
We include studies with original data on carbapenemresistant K. pneumoniae or E. coli sepsis in neonates, which contained any antimicrobial resistance (AMR) or genotype data, or showed the proportion of carbapenem resistant isolates of all Gram negative isolates, or clinical data (including patient demographics, underlying conditions, and antibiotic treatment). We only included blood stream infections. The full details of inclusion and exclusion criteria are presented in Additional file 2: Table S2.

Statistical analysis
Descriptive analysis was performed to investigate the distribution of genotype and MLST typing. Antimicrobial resistance rates were reported by median with interquartile interval (IQI).

Literature search and study selection
We identified 491 studies from Chinese and English database searches: 81 from CNKI, 214 from Wanfang med database, 96 from Pubmed/Medline and 100 from Embase (the flow chart is shown in Fig. 1). A total of 17 studies met the inclusion criteria and were included No full text (n = 1)  with lung disease, 2 with necrotizing enterocolitis (NEC) and 2 with recent surgery. Ten patients in these 9 studies reported antibiotic treatment: 5 received meropenem alone, 1 ceftazidime alone, 3 patients had received two antibiotics (piperacillin/sulbactam and ceftazidime; imipenem and amikacin; meropenem and ciprofloxacin) and 1 patient had received more than three antibiotics. Eight patients in these 4 studies had received antibiotics prior to the onset of the relevant infection: 5 received meropenem, 1 received panipenem and 2 didn't report the type the antibiotics. Clinical treatment outcomes were reported in 13 patients from 6 studies; 3 died and their deaths were attributed to the infection. Eleven patients of the six studies reported the source of infection. Ten patients (10/11, 90.9%) were hospital-acquired infections, while only one was considered to be a vertical transmission.

The proportion of carbapenem resistant strains of all Gram negative strains
Only 6 studies (35%; 6/17) reported the proportion of carbapenem resistant isolates relative to all Gram-negative isolates causing sepsis. Overall, 39 (5.3%) carbapenem resistant K. pneumoniae and E. coli isolates were reported of out of a total of 740 Gram-negative isolates.

Discussion
This is the first review of carbapenem-resistant Enterobacteriaceae (CRE) sepsis in Chinese neonates. Although the prevalence of adults and children with infections resistant to imipenem and meropenem reported by CHI-NET in 2017 increased significantly, the samples were mainly derived from non-sterile body fluids, and the data for children were not broken down by age. This review has demonstrated that there are very limited recent data on carbapenem resistant isolates in neonates in China. CRKP is reported more than CREC. NDM-1 was the most commonly reported carbapenemase genotype, consistent with previous reports from Asia [11], but different to reports from the United States, where KPC is the most common genotype identified in children [12]. It is worth noting that the CLSI breakpoints for carbapenem changed in 2010 and in 2015, the CDC revised the definition for CRE. In this review, 12 studies provided CLSI reference standards. Among the 12 studies, only one adopted the CLSI standard of 2015, and the others adopted the CLSI standard of before 2015.
In 2017, the World Health Organization published a list of priority pathogens in order to inform global AMR research. CRE is one of the highest priority pathogens for the development of new antibiotics [13], but there are few new antibiotics available. Cefiderocol, is a novel catecholsubstituted siderophore cephalosporin with potent activity against meropenem-non susceptible Enterobacteriaceae [14], including metallo-β-lactamases (NDM-1, VIM, IMP). This is the most clinically advanced drug active against NDM carbapenem resistant organisms (CROs) infections [15], but no paediatric studies have yet commenced recruitment. The current standard treatment for NDM CRE infections is polymyxin based combination therapy [16]. However, polymixin E has complex pharmacokinetics requiring hydrolysis of the prodrug colistimethate sodium to colistin, making this less suitable for neonates and infants, and there are no pharmacokinetics data for      polymixin B in neonates [17]. Other older, off patent drugs that have potential activity against CROs include fosfomycin and tigecycline, but again, these have no published PK data in neonates. In our study, we found that the currently reported carbapenem-resistant Enterobacteriaceae sepsis in neonate is mainly nosocomial infection. In view of the fact that there is no appropriate antibiotics to treat carbapenem resistant bacteria infection in neonates, it is very important to strengthen epidemiological surveillance, stringent standard infection control practices in healthcare settings, and to enhance the rational use of antibiotics.