Characterization of Extensively Drug-Resistant (XDR) Carbapenemase-Producing Enterobacterales (CPE) in Canada from 2019 to 2020

ABSTRACT Data regarding the epidemiology of extensively drug-resistant (XDR) carbapenemase-producing Enterobacterales (CPE) in Canada are scarce. Among CPE patients identified by the Canadian Nosocomial Infection Surveillance Program, the following were each significantly associated with XDR status: international travel history; CPE acquisition from a health care exposure abroad; presence of the New Delhi metallo-β-lactamase (NDM) carbapenemase gene; E. coli sequence type (ST) 167, ST405, and ST648; E. cloaceae ST177; C. freundii ST22; and resistance to all antimicrobials except colistin, tigecycline, and ceftazidime-avibactam. IMPORTANCE Extensively drug-resistant (XDR) carbapenemase-producing Enterobacterales (CPE) are a global public health concern. XDR CPE are among the most drug-resistant and difficult-to-treat bacteria, and infected patients are likely to experience adverse outcomes. Because XDR status further reduces effective therapeutic options, it is critical for clinicians to consider resistance and therapeutic options not only in the context of a patient with CPE but also in the context of potential XDR status. Our study reports on patient characteristics associated with the acquisition of an XDR CPE. Our study also reports on the species and carbapenemases associated with XDR status among Enterobacterales identified in Canada. Among a panel of 22 antibiotics, including novel combination drugs, we showed which retained the highest activity against XDR CPE, which may help guide the selection of antibiotic treatments.

restricted our surveillance period to 2019 and 2020. Participating acute care facilities submitted isolates that exhibited nonsusceptibility to imipenem, meropenem, or ertapenem to the National Microbiology Laboratory (Winnipeg, Canada) for minimum inhibitory concentration confirmation using imipenem, meropenem, and/or ertapenem Etest strips (bioMérieux, St. Laurent, Quebec, Canada) (6). Broth microdilutions (Sensititre, panel CAN1MSTF) were performed to determine the antibiotic susceptibilities, following the Clinical and Laboratory Standards Institute interpretations (6). Isolates were sequenced using the Illumina NextSeq platform. Carbapenemase (github.com/phac-nml/staramr) and multilocus sequence typing profiles (pubmlst.org) were extracted from whole-genome sequencing data (7). Using the novel Canadian recommendation for the laboratory interpretation of XDR organisms, we defined CPE isolates as XDR if they were resistant to five or six of the following six antibiotic groups: tobramycin or gentamicin, piperacillin-tazobactam, imipenem or meropenem, cefotaxime or ceftriaxone or ceftazidime, ciprofloxacin, or trimethoprim-sulfamethoxazole (8). Through chart review, IPC professionals completed standardized questionnaires for patients harboring laboratory-confirmed CPE, and, using the standardized definitions detailed in our published surveillance protocol, determined the likely source of CPE acquisition (4,5).
Because a single CPE patient could harbor more than one unique CPE isolate (e.g., differing by organism or carbapenemase gene), we reported demographic and clinical data by the number of unique patients, and we reported molecular and microbiological data by the number of unique isolates. For analyses pertaining to newer combination drugs, we created CPE isolate subsets, each of which was restricted to isolates that harbored carbapenemase genes that the beta-lactamase inhibitor in question could inhibit (9). Using R version 4.1.1, we performed chi-square and Fisher's exact two-sided tests to compare XDR CPE to non-XDR CPE (10). We set the criterion for statistical significance at a = 0.05.
We identified 482 unique CPE isolates from 440 patients across 33 of 72 participating facilities. Over half of the patients (259/440, 59%) were colonized or infected with an XDR CPE. The incidence rate for XDR CPE significantly decreased from 0.13 per 1,000 admissions in 2019 to 0.10 per 1,000 admissions in 2020 (P = 0.048). Table 1 summarizes the demographic data available for the CPE patients, stratified by XDR status.
International travel in the year prior to positive culture was significantly more common in XDR patients than in non-XDR patients (80/183, 44% versus 32/122, 26%; P = 0.002). Specifically, we observed a significant association between XDR status and travel to South Asia (Nepal, India, Pakistan, Afghanistan, Sri Lanka, or Iran; P = 0.010). For those who had data available, over three quarters of the international travelers received medical care while abroad (79/100, 79%). Among patients with health care-associated CPE, acquisition from health care exposure outside Canada compared to within Canada was significantly more common among XDR patients than among non-XDR patients (49/183, 27% versus 17/126, 13%; P = 0.005). . Presence of the Ambler class B carbapenemase bla NDM was significantly more common in XDR CPEs than in non-XDR CPEs (130/281, 46% versus 23/201, 11%; P , 0.001), whereas the presence of the Ambler class A carbapenemases bla KPC and bla NMC was significantly less common. Results failed to show a significant association between any other carbapenemase and XDR status. The most common sequence types (present in .10 isolates; ST) were: ST167, ST38, ST405, ST410, and ST648 among E. coli, ST177 among Enterobacter spp., and ST22 among Citrobacter spp. The frequency of C. freundii ST22 was likely driven by a hospital outbreak and due to greater diversity among Klebsiella STs, none were classified as common. While results failed to show an association between species and XDR status, when we compared the occurrence of STs, we observed that E. coli ST167, ST405, and ST648, E. cloaceae ST177, and C. freundii ST22 were significantly more common among XDR isolates (Table 2). Specifically, bla NDM was the most common carbapenemase among CPE isolates harboring E. coli ST167 (16/18, 89%) and ST405 (15/16, 94%). Further, bla NDM was the only carbapenemase among the CPE isolates harboring E. coli ST648 (13/13, 100%) and E. cloaceae ST177 (13/13, 100%).
We present novel data regarding the epidemiology of XDR CPE, the most resistant and difficult-to-treat pathogens, in Canada. The travel and medical history risk factors that we identified are consistent with those described in the literature for CPE acquisition in general (11). Our findings align with surveillance from Greece and Spain, which found that the activity of ceftazidime-avibactam was higher than that of imipenem/relebactam against Enterobacterales harboring OXA-48-type carbapenemases (12,13). Further, this report applies novel Canadian recommendations for the laboratory interpretation of XDR clinical isolates to national surveillance data (8). These recommendations are the first of their kind in Canada, and they have been modified from those originally proposed by the European Centre for Disease Prevention and Control and the Centers for Disease Control and Prevention to consider the Canadian context and Canadian stakeholder input (14). Because relevant data were available for 2 years of surveillance and from acute care facilities only, and because a small number of patients were infected (62/406, 15%), the generalizability of our data and our ability to test for a difference in clinical outcome between the XDR and the non-XDR CPE patients were limited.
Given the increasing spread of XDR CPE, it is critical for clinicians to consider potential XDR status due to the fact that they reduce effective therapeutic options (1, 3). Our reporting of both the patient characteristics associated with XDR status and the resistance profiles associated with XDR CPE isolates may prompt further investigations into CPE diagnosis and may help guide IPC practices and empirical antibiotic selection. Ongoing national surveillance of XDR CPE is required to elucidate their epidemiological profiles in Canada.