bioRxiv preprint doi: https://doi.org/10.1101/2020.02.21.959742; this version posted February 25, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 1 Clinical and genomic epidemiology of carbapenem-non-susceptible Citrobacter spp. at a 2 tertiary healthcare center over two decades 3 4 Ahmed Babiker1,2†, Daniel R. Evans1, Marissa P. Griffith1,2, Christi L. McElheny1, Mohamed 5 Hassan3, Lloyd G. Clarke1, Roberta T. Mettus1, Lee H. Harrison1,2,4, Yohei Doi1, Ryan K. Shields1, 6 Daria Van Tyne1# 7 8 1Division of Infectious Diseases, Department of Medicine, University of Pittsburgh, Pittsburgh, 9 Pennsylvania, USA 10 2Microbial Genomic Epidemiology Laboratory, University of Pittsburgh School of Medicine, 11 Pittsburgh, Pennsylvania, USA 12 3Division of General Internal Medicine, Department of Medicine, University of Pittsburgh 13 Medical Center, Pittsburgh, Pennsylvania, USA 14 4Graduate School of Public Health, University of Pittsburgh, Pennsylvania, USA 15 16 Running Title: Genomic epidemiology of carbapenem-non-susceptible Citrobacter 17 18 † Present address: Department of Pathology and Laboratory Medicine, Emory University School 19 of Medicine, Atlanta, Georgia, USA 20 #Address correspondence to: Daria Van Tyne, [email protected] 21 1 bioRxiv preprint doi: https://doi.org/10.1101/2020.02.21.959742; this version posted February 25, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 22 Abstract 23 Carbapenem-non-susceptible Citrobacter spp. (CNSC) are increasingly recognized as healthcare- 24 associated pathogens. Information regarding their clinical epidemiology, genetic diversity, and 25 mechanisms of carbapenem resistance is lacking. We examined microbiology records of adult 26 patients at the University of Pittsburgh Medical Center (UMPC) Presbyterian Hospital (PUH) from 27 2000-2018 for CNSC, as defined by ertapenem non-susceptibility. Over this timeframe, the 28 proportion of CNSC increased from 4% to 10% (P=0.03), as did carbapenem daily defined 29 doses/1000 patient days (6.52 to 34.5, R2=0.831, P<0.001), which correlated with the observed 30 increase in CNSC (lag=0 years, R2=0.660). Twenty CNSC isolates from 19 patients at PUH and 31 other UPMC hospitals were available for further analysis, including whole-genome short-read 32 sequencing and additional antimicrobial susceptibility testing. Of the 19 patients, nearly all 33 acquired CNSC in the healthcare setting and over half had polymicrobial cultures containing at 34 least one other organism. Among the 20 CNSC isolates, C. freundii was the predominant species 35 identified (60%). CNSC genomes were compared with genomes of carbapenem-susceptible 36 Citrobacter spp. from UPMC, and with other publicly available CNSC genomes. Isolates encoding 37 carbapenemases (blaKPC-2, blaKPC-3, and blaNDM-1) were also long-read sequenced, and their 38 carbapenemase-encoding plasmid sequences were compared with one another and with publicly 39 available sequences. Phylogenetic analysis of 102 UPMC Citrobacter spp. genomes showed that 40 CNSC from our setting did not cluster together. Similarly, a global phylogeny of 64 CNSC 41 genomes showed a diverse population structure. Our findings suggest that both local and global 42 CNSC populations are genetically diverse, and that CNSC harbor carbapenemase-encoding 43 plasmids found in other Enterobacterales. 44 Keywords: Carbapenem, Citrobacter, Multidrug-resistance, Carbapenemase 2 bioRxiv preprint doi: https://doi.org/10.1101/2020.02.21.959742; this version posted February 25, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 45 Introduction 46 Carbapenem-resistant bacteria have become a major health concern worldwide (1). There are 47 limited therapeutic options for treating infections caused by these multidrug-resistant organisms, 48 resulting in greater morbidity and mortality compared to infections caused by susceptible 49 organisms (2). Furthermore, multidrug-resistant infections place an additional economic burden 50 on healthcare systems (3). In recognition of this threat, the treatment and control of carbapenem- 51 resistant organisms have been prioritized by both the Centers for Diseases Control and Prevention 52 and the World Health Organization (4, 5). 53 54 The recent increase in infections caused by carbapenem-resistant organisms in the United States 55 has been largely driven by the dissemination of plasmid-encoded carbapenemases, which are often 56 carried by members of the Enterobacterales, particularly Klebsiella pneumoniae (6, 7). However, 57 rates of other carbapenem-resistant bacterial species have also increased (8). Among them, 58 carbapenem-non-susceptible Citrobacter spp. (CNSC) have become increasingly recognized as a 59 healthcare-associated pathogen (9-13). CNSC isolates have been found to be both genotypically 60 and phenotypically diverse (14, 15), and their resistance to carbapenems is frequently caused by 61 plasmid-encoded carbapenemases, which can be readily acquired through horizontal gene transfer 62 (12, 16). 63 64 Information regarding the clinical epidemiology, genetic diversity, and mechanisms of 65 carbapenem resistance among CNSC in the United States are currently limited to a small number 66 of studies and very few isolates (9, 17-19). Here we aimed to investigate the emergence of CNSC 67 within our healthcare system using epidemiology and genomics approaches. We conducted a 3 bioRxiv preprint doi: https://doi.org/10.1101/2020.02.21.959742; this version posted February 25, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 68 retrospective analysis of CNSC prevalence and carbapenem use over the last two decades, and 69 compared the genomes of CNSC isolates from our center with other local Citrobacter isolates, as 70 well as with CNSC genomes sampled from around the globe. We found that while the CNSC 71 sampled from our center are highly genetically diverse, their diversity is consistent with the local 72 carbapenem-susceptible Citrobacter population, as well as with CNSC sampled elsewhere. 73 74 Methods 75 Study design and isolate collection 76 This study was conducted at the University of Pittsburgh Medical Center (UPMC) Presbyterian 77 Hospital (PUH), an adult medical/surgical tertiary care hospital with 762 total beds, 150 critical 78 care unit beds, more than 32,000 yearly inpatient admissions, and over 400 solid organ transplants 79 per year. CNSC isolates were collected from both UPMC-PUH as well as other UPMC hospitals. 80 Ethics approval for this study was obtained from the Institutional Review Board of the University 81 of Pittsburgh. 82 83 To investigate CNSC epidemiology at UPMC-PUH, microbiology records of adult patients with a 84 positive clinical culture for CNSC were evaluated from January 1, 2000 to December 31, 2018. 85 Cases were excluded from re-inclusion within 90 days of any CNSC culture. Carbapenem non- 86 susceptibility was defined as non-susceptibility to any carbapenem according to the 2017 Clinical 87 Laboratory Standards Institute (CLSI) interpretative criteria (20). Antibiotic consumption was 88 measured by daily defined doses (DDDs) of any carbapenem (21). To further phenotype and 89 genotype CNSC at UPMC, 20 available CNSC isolates from 19 patients collected between 2013 90 and 2019 from were included. Isolates were considered community-associated if the organism was 4 bioRxiv preprint doi: https://doi.org/10.1101/2020.02.21.959742; this version posted February 25, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. 91 isolated from a specimen collected within 72 hours following hospital admission; isolates collected 92 after 72 hours were considered healthcare-associated (22). Clinical characteristics and outcomes 93 of patients with CNSC isolates that underwent further characterization were collected through 94 retrospective chart review. The primary clinical outcome was in-hospital mortality and/or transfer 95 to hospice. 96 97 CNSC isolate characterization 98 Initial species assignment was performed using standard clinical microbiology laboratory 99 methods, and was confirmed or modified after whole-genome sequencing. Carbapenem non- 100 susceptibility was initially determined by standard clinical microbiology laboratory methods, and 101 was confirmed by the Kirby-Bauer disk diffusion method as per the 2017 Clinical Laboratory 102 Standards Institute (CLSI) interpretative criteria (20). Susceptibility to additional agents was 103 determined by the broth microdilution method (20). Presence of carbapenemase enzyme activity 104 was assessed by modified carbapenem inactivation (mCIM) test (23). 105 106 Genome sequencing and analysis 107 Genomic DNA was extracted from pure overnight cultures of single bacterial colonies using a 108 Qiagen DNeasy Blood & Tissue Kit according to the manufacturer’s instructions (Qiagen, 109 Germantown, MD). Library construction and sequencing were conducted using the Illumina 110 Nextera NGS Library Prep Kit or the Illumina Nextera XT DNA Library Prep Kit (Illumina,