Background
Pseudomonas aeruginosa is a leading nosocomial pathogen and infections can be difficult to treat because of rapid resistance development. The emergence of multidrug-resistant (MDR) isolates is a serious public health threat and often affects immunocompromised patients within special units (intensive care units (ICU), haematology-oncology wards or burn units) [
1‐
4]. Resistance to carbapenems is mediated either by intrinsic resistant mechanisms (a combination of efflux pumps, AmpC overexpression and porin loss) or acquisition of a carbapenemase, especially a metallo-β-lactamase (MBL) [
5]. Carbapenemase-producing
P. aeruginosa (CPPA) isolates harbour antimicrobial resistance genes located on mobile genetic elements (mainly integrons, transposons or plasmids) that can spread to other bacteria [
6‐
8], so microbiological monitoring and infection control surveillance is of utmost importance. Prevalence of CPPA among MDR
P. aeruginosa differs greatly between regions, with VIM- and IMP-family carbapenemases being the most widespread [
9,
10]. Additionally, CPPA are known to cause protracted outbreaks, e.g. IMP-8 or GIM-1-producing types [
11,
12]. However, there is little surveillance data available combining molecular and epidemiological information. The aim of this study was to analyse the prevalence and epidemiology of CPPA in three German medical centres isolated from 2015 to 2017.
Discussion
In contrast to carbapenem-resistant
Acinetobacter baumannii complex or carbapenem-resistant Enterobacterales
, carbapenemases are detected less frequently in carbapenem-resistant
P. aeruginosa in which carbapenem-non-susceptibility is predominantly mediated by other mechanisms (a combination of efflux pumps, AmpC overexpression and porin loss) [
5,
25]. However, early detection of these mobile broad-spectrum β-lactamases is necessary to prevent the propagation mainly of metallo- β-lactamases, across other Gram-negative organisms in the healthcare-setting [
25,
26].
In our study, carbapenemases, mainly VIM-2, were detected in one third of the MDR/XDR
P. aeruginosa isolates. The rate of CPPA and proportions of the different carbapenemase gene families in this study are in line with other observations. In 2017 approximately 27.7% of the
P. aeruginosa isolates referred to the German reference centre carried a carbapenemase, VIM-2 being by far the most prevalent one [
27]. In a German multicentre study, 32% of the carbapenem-resistant
P. aeruginosa isolates were carbapenemase producers, with VIM-2 being the most prevalent enzyme [
28]. Studies combining molecular surveillance and prevalence data at two German tertiary care centres detected a CPPA proportion of 40% in MDR isolates (all
blaVIM) and 23% in XDR isolates (mostly
blaVIM-1 and
blaVIM-2) [
29,
30]. Nevertheless, the local epidemiology can differ greatly between medical centres, e.g. in a tertiary care centre 40 km from Cologne the most prevalent carbapenemase gene in
P. aeruginosa was
blaGIM-1 [
6]. In another hospital in southern Germany
blaIMP was widespread [
12]. Overall, it is difficult to compare prevalence studies as bacterial isolate selection, inclusion and screening criteria, as well as test algorithms differ greatly. Until now there are no official recommendations by EUCAST addressing carbapenemase screening cut-off values in
P. aeruginosa comparable to those existing for Enterobacterales [
31]. Official screening recommendations are based on the three antibiotics imipenem, meropenem and ceftazidime (German National Reference Centre) or on imipenem, meropenem and piperacillin-tazobactam (British standards) [
32,
33]. Overall, we chose a well-defined significant subgroup of MDR
P. aeruginosa since all isolates non-susceptible to piperacillin, ceftazidime, cefepime, imipenem, meropenem and ciprofloxacin (4MRGN) directly result in infection prevention and control (IPC) measures [
14].
Molecular surveillance of bacterial isolates combined with epidemiological and infection data can lead to direct implementation of targeted IPC measures. Surveillance of
P. aeruginosa is of utmost importance as it can reside in the inanimate patient environment and subsequently lead to transmission and to colonization or infection.
P. aeruginosa can reside in the sink drains in the patient room for long periods. The spreading and distribution of MDR
P. aeruginosa in the shower and sink drains, and sewage system of the ward is quite complex as several studies have shown [
11,
34]. We found direct and indirect evidence for both modes of transmission (patient-to-patient and room-to-patient). Although, most
blaVIM-2-carrying
P. aeruginosa isolates clustered in the PFGE analysis, we were only able to confirm a few transmission events. Interestingly, transmission happened exclusively on the intensive care units of the tertiary care centre. Therefore IPC measures should focus on the ICU, where the relevant patients at risk for colonization/infection with CPPA are found (e. g. antimicrobial therapy, prolonged hospitalization, medical devices, and severe underlying disease) [
2,
12,
35]. Moreover, two out of the 13 patients who carried a related (cluster 1) CPPA at admission were referred from another hospital in the region. Thus, genetically related strains may be endemic in the region.
There are a few limitations in this study. We were not able to provide full prevalence data, as only two third non-duplicate 4MRGN isolates detected during this period were available. However, our prevalence data is in line with other studies. Secondly, we were able to detect a dominant
blaVIM-2-carrying strain using PFGE; for further discrimination whole genome sequencing is needed and further studies will address this. Thirdly, our inclusion criteria were probably not sensitive enough to detect all CPPA. On the other hand, CPPA is often associated with MDR- or XDR-phenotypes, corresponding to our inclusion criteria [
36]. Extending the screening inclusion criteria would lead to more negative results and clinical microbiology laboratories may not have the resources.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.