Assessing Molecular Epidemiology of Carbapenem-resistant Klebsiella pneumoniae (CR-KP) with MLST and MALDI-TOF in Central China

Carbapenem-resistant K. pneumoniae (CR-KP) posts significant public health challenge worldwide. The aim of this study is to assess clinical characteristics and molecular epidemiology of CR-KP infections with Multilocus sequence typing (MLST) and Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF) in Central China. A total of 71 CR-KP isolates were recovered in a teaching hospital from October 2014 to December 2015. Among all CR-KP isolates, 73.2% (52) produced K. pneumoniae carbapenemases-2 (KPC-2). Eighteen ST types were identified by MLST, among these ST types, forty-seven isolates belonged to ST11 type, which was the predominant outbreak strain in China, and most ST11 isolates produced KPC-2. Eleven mass spectrometry (MS) types were identified by MALDI-TOF MS analysis, 53.5% isolates were MS4 and MS6, which matched with ST11 in MLST analysis. CR-KP infection was associated with increased medical cost and longer hospitalization. Therefore, we found that KPC-2-producing ST11 (MS4 and MS6) CR-KP isolates were the predominant clone identified by MLST and MALDI-TOF, and CR-KP infection was associated with increased hospital costs and longer hospitalization.

The aim of the present study was to investigate the molecular epidemiology and clinical characteristics of 71 CR-KP isolates in a teaching hospital in Changsha, central China using MALDI-TOF MS and MLST. In addition, this study identified antimicrobial resistance genes of CR-KP strains, and investigated the financial burden of CR-KP infection.

Detection of Antimicrobial
(3), ST736 (2), and the following ST type each represented by one isolate: ST20, ST23, ST25, ST29, ST34, ST147, ST189, ST441, ST629, ST1224, ST1425, ST2236, ST2389, and ST2391. ST2389, ST2390, and ST2391 were identified as novel ST types in the MLST database. More information could be seen on the MLST website (http://bigsdb. pasteur.fr/klebsiella/). strain typing by MALDI-toF Ms. Even though the 71 clinical strains cultured on the blood agar plate showed diverse morphological characteristics, all of them were correctly identified as K. pneumoniae species by Clin-TOF II with score values (>25). The representative spectra of the CR-KP strains were shown in

Medical Costs of CR-Kp infection.
The CR-KP infected patients stayed longer in the hospital than the patients with CS-KP infection, meanwhile, the mortality of CR-KP infection patients was higher than that of CS-KP infection patients. Furthermore, the medical costs of CR-KP group (including total costs, medical test costs and total drug costs and anti-infective drug costs) was significantly higher than costs of the CS-KP group ( Table 2).

Discussion
Carbapenem resistance in Enterobacteriaceae, especially K. pneumoniae, has become a significant public health challenge in china. Due to the limited efficacy of antimicrobials in treating carbapenem-resistant Enterobacteriaceae (CRE) infection, the mortality of patients infected with CRE is higher than that of patients infected with carbapenem-susceptible Enterobacteriaceae(CSE) [16][17][18] . In a case-control study at a New York City hospital, patients infected with CR-KP showed 48% in-hospital mortality and 38% infection-specific mortality 19 .
In this study, patients with CR-KP infection suffered from significant higher mortality, longer hospital stay and higher financial burden compared to patients with CS-KP infection as previously reported 20 . CR-KP isolates were most frequently isolated from patients from ICU in this study. The first in vivo isolation of CR-KP strain was reported in 2000 in an ICU in North Carolina 21 . In fact, ICU was the breeding ground that produced, spread, and amplified antimicrobial resistance because of the presence of extremely vulnerable patients, the use of invasive procedures and the frequent use of antimicrobial agents 22,23 . Published literature reported that In China, Klebsiella pneumoniae carbapenemase (KPC) is the most clinically significant serine carbapenemase. KPC-2-producing K. pneumoniae isolates spread widely and rapidly across the country, after the first KPC-2-producing K. pneumoniae(KPC-KP) was isolated in China 26 . In our study, most of the CR-KP strains were KPC-2-producing K. pneumoniae, which were major hospital pathogens 27,28 . Further, 18 strains harbored the NDM-1 carbapenemase gene, which made the strain confer resistance to almost all β-lactams, except aztreonam 29 . NDM-producing isolates were usually resistant to multiple antimicrobials, leaving few or no therapeutic options 30  Our study presented that 26.8% (19/21) CR-KP harbored ≥3 different resistance-associated genes. This was consistent with the finding of Li B, et al. 33 . Of note, multi-carbapenemase production is associated with multi-drug resistance, which leads to limited anti-infection treatment options. We have surveillance systems for multiple drug resistant (MDR) bacteria in the hospital. Patients with MDR bacteria would be isolated in hospital, and workers who had contacted the patient would enhance hand hygiene. Once an outbreak was detected, we should notify the infection departments to isolate hospitalization with MDR infection, and strengthen disinfection.
Our study provided information on clinical characteristics and molecular epidemiology of CR-KP infection in central China by typing isolates from an outbreak using MLST and MALDI-TOF MS. In this study, ST11was the predominant strain attributed to the outbreak. ST11 is the epidemic ST type of KPC-producing K.pneumoniae in China 34 , and almost all ST11 isolates were KPC-2-producing K.pneumoniae, contributing to the spread of antibiotic resistance in the hospitals. In addition, almost all ST11 isolates were matched with MS4 and MS6 in MS typing, which covered 53.5% of all the MADI-TOF MS types. During the outbreak described in this study, over half of the ST11 isolates were distributed from May 2015 to August 2015, most of which were MS4 in MALDI-TOF MS typing. Hierarchical cluster analysis of strains by MALDI-TOF MS was acquired and analyzed simultaneously with the antimicrobial sensitivity results. The quick identification of an outbreak was critical for infection control.
In summary, an outbreak of KPC-2-producing CR-KP isolates was reported in our hospital, which was associated with higher financial burden and longer hospital stay. ST11 isolates were the predominant ST type attributed to the outbreak, and most of these isolates were matched with MS4 and MS6 in MS typing. MALDI-TOF MS typing can rapidly identify and type the CR-KP isolates. This is the first report of the utilization of MALDI-TOF MS in understanding the lineage of isolates contributing to CR-KP outbreak in china. This is also the first study that evaluates the financial burden of CR-KP infection in china. Findings of this study will help establish Antimicrobial Stewardship Program (ASP) and develop HAI outbreak surveillance, prevention and control programs on CR-KP infection in China.
This study presents several limitations. First, because of the restrictions on technology and funds, PFGE and wzi gene sequencing was not conducted. Moreover, this study was conducted in a single medical center, the sample size of the CR-KP group and CS-KP group was small for the risk factor assessment. Study on a larger population may produce more comprehensive results, and patient-to-patient transmission of HAI caused by CRKP was not assessed. Instead, MALDI-TOF MS typing was utilized for its short turnaround time which may have produced a relatively rough description of outbreak 35 . Material and Methods study setting and Bacterial Isolates. This was a retrospective study carried out in Xiangya Hospital, a 3,500-bed university teaching hospital in Changsha, Hunan Province, Central South China. This hospital provides medical and surgical care for all patients including adults and children.
All non-duplicate bacterial isolates were collected from clinical samples from October 2014 to December 2015. Isolates recovered from the same patients were counted only once. Patients' medical records were retrospectively reviewed and all data collected were de-identified.
Detection of Antibiotic Resistance Genes. Antibiotic resistance genes were detected by polymerase chain reaction (PCR) using primers and conditions as previously described 33,36 . PCR was performed for all the CR-KP strains to detect the carbapenemase genes (bla NDM-1 , bla KPC-2 , bla IMP-1 , and bla OXA-48 ) and β-lactamases genes (bla CTX-M-15 and bla SHV ). PCR products were purified and sequenced using the amplification primers at Sangon Biotech (Shanghai, China).
MLst analysis. PCR amplification of seven housekeeping genes (gapA, infB, mdh, pgi, phoE, rpoB, and tonB) were performed on all CR-KP isolates as previously described 37 . The allele number and sequence type (ST) were assigned by MLST website (http://bigsdb.pasteur.fr/klebsiella/). strain typing by MALDI-toF Ms. For each isolate, three separate spectra were obtained using the measurements performed on Clin-ToF II (Bioyong Technologies Inc, Beijing). Escherichia coli ATCC8739 was used for the calibration of the instrument. All spectra were recorded in the linear positive-ion mode at a laser frequency of 40 Hz and mass ranging from 2,000 to 20,000 kDa. For each sample spot, one spectrum was acquired as a sum of 500 shots across a spot. Species were confirmed by comparison with the mass-spectrum library, using the BioExplorer V2.1 DataBase (Bioyong Technologies Inc) under standard conditions. The basal MALDI-TOF MS classification data were obtained from three points with the highest degree. The data with the highest degree were selected for cluster analysis. MALDI-TOF results were analyzed using MALDI MS software (Bioyong Technologies Inc, Beijing). The differences of K. pneumoniae MS were analyzed using the Launch pad software (Bioyong Technologies Inc, Beijing). All peak spectra and positions were compared with 1/1000 m/z offset value. Matrix was constructed with m/z peak position, and hierarchical clustering was analyzed through Flashclust software (Bioyong Technologies Inc, Beijing). The differences in homology were set using the perimeter distance of a dendrogram. According to the type assignment, we defined a cut-off value was >70% similarity.
Clinical Data Collection and Definitions. CR-KP strains were defined as isolates with resistance or intermediate sensitivity to at least one type of carbapenem (imipenem, ertapenem, or meropene). For each CR-KP infected patient, one patient with CS-KP infection was randomly selected. The two groups were admitted within the same period (within 30 days) and matched for age and sex. The clinical data of these patients were retrospectively reviewed.
Total cost was defined as all costs of the patients in hospital; medical examination cost was defined as the cost associated with examining the patients, including imaging and other laboratory tests; medical testing cost was defined as the cost of laboratory testing; drug cost was defined as the cost of all medications; and anti-infective drug cost was defined as the cost of antimicrobial agents. statistical Analysis. All the statistical analyses were performed using SPSS 20.0 (IBM). The categorical variables were expressed by rates and tested by Chi-square. The Wilcoxon Rank SumTest was used to compare the continuous variables which were shown as median. Two-tailed P value of less than 0.05 was considered significant. ethics statement. All procedures performed in this study involving human participants were in accordance with the Ethics Committee of the Xiangya Hospital of Central South University (No. 201701017). The study was conducted in accordance with the Declaration of Helsinki. Oral informed consent was obtained from all individual participants included in the study. This article does not contain any studies with animals performed by any of the authors.

Data Availability
The datasets supporting the conclusions of this article are included within the article. The raw data can be made available to the interested researchers by the authors of this article if requested.