The rapid emergence of bacterial pathogens presenting resistance to multiple antimicrobial agents (MDR) together with the decreasing trend in the development of new antimicrobial compounds constitute an extremely serious threat to public health (Roca et al., 2015). Resistance rates are particularly alarming among Gram-negative bacteria (GNB) as available treatment options are severely impaired. In 2016 a report commissioned by the UK Prime minister, known as the “O’Neill report,” attributed 700.000 annual human deaths worldwide to microbial infections caused by antimicrobial resistant (AMR) pathogens (O’neill, 2016). The same report estimated that this figure would rise to 10 million annual deaths by 2050 unless there is a global effort to tackle AMR. Likewise, organizations such as the US Centers for Disease Control and Prevention (CDC), the European Centre for Disease Prevention and Control (ECDC) and the World Health Organization (WHO) are considering infections caused by MDR bacteria as an emergent global disease and a major public health problem (Roca et al., 2015). According to their mortality, health-care and community burden, prevalence of resistance and treatability, carbapenem-resistant Enterobacterales have been included into the 2017 WHO priority list of antibiotic-resistant bacterial pathogens and currently constitute one of the major public health threats worldwide (Tacconelli et al., 2017).

Among carbapenem-resistant Enterobacterales, Klebsiella pneumoniae stands out as a formidable nosocomial pathogen causing several infections associated with high mortality rates (Tumbarello et al., 2012). In Europe, carbapenem-resistant K. pneumoniae accounted for almost 16,000 infections and more than 2,000 deaths in 2015 (Cassini et al., 2018). Carbapenem-resistance is usually associated with the carriage of genes encoding carbapenem-hydrolyzing enzymes, most of which are commonly located within conjugative plasmids and, therefore, are easily disseminated among nosocomial pathogens (Bonomo et al., 2017). Predominant carbapenemases in Europe include OXA-48 oxacillinase, New Delhi metallo-β-lactamase (NDM) and Klebsiella pneumoniae carbapenemase (KPC), among others. In Spain, OXA-48 is currently the predominant mechanism associated with the dissemination of carbapenem-resistant Enterobacterales (Grundmann et al., 2017). NDM-producing bacteria are reported less frequently in Spain but their presence is increasing and associated with a few clonal lineages (Marí-Almirall et al., 2021) while KPC-producing Enterobacterales are also sporadic in Spain, and only a few outbreaks have been reported in Madrid and Andalusia (Oteo et al., 2016).

Klebsiella pneumoniae carbapenemase is likely one of the most common mechanisms of resistance to carbapenems in K. pneumoniae, and the worldwide spread of the bla_KPC gene was initially associated with the dissemination of a few K. pneumoniae clonal lineages within the ST258 clonal group (CG258), a particular plasmid backbone (pKpQIL) and a highly conserved Tn4401 transposon (Chen et al., 2014). More recently bla_KPC has been found on a variety of clonal lineages and plasmids from different incompatibility groups (Wyres et al., 2020), and although it is usually associated with a composite Tn4401 transposon, some structural variations have also been described (Shen et al., 2009; Naas et al., 2012; Chmelnitsky et al., 2014). In this retrospective study we have examined the rapid dissemination of KPC-producing Enterobacterales within a tertiary hospital in Barcelona and we have investigated the inter- and intra-species spread of carbapenem resistance.

This study included 125 carbapenem-resistant isolates of Enterobacterales collected from July 2018 to February 2019 at one tertiary hospital in Barcelona, Spain. Strains were recovered from surveillance and clinical samples (Supplementary Table 2). Identification at species level was performed by matrix-assisted laser desorption ionization-time time of flight mass spectrometry (MALDI-TOF MS) in a Microflex LT benchtop instrument (Bruker Daltonics) operated in linear positive mode.

The emergence of KPC-producing Enterobacterales has been increasing worldwide during the last decades (Lee et al., 2019). Since its first description in 1996 in the United States, KPC has spread through the globe, becoming endemic in areas such as the United States, Israel, Greece, and Italy, and sporadic spread being reported in additional European countries as well as in the Asia-Pacific region (Lee et al., 2019). Many of the descriptions from European countries have been related to patients traveling to endemic areas.

In Spain, the first KPC report dates back to 2009 and it was associated with the rapid spread of a ST384 K. pneumoniae strain carrying bla_KPC–3 within a Tn4401a structure (Curiao et al., 2010). KPC-2 and KPC-3-producing Enterobacterales were later reported in the central region of Spain and Andalusia mainly associated with K. pneumoniae isolates belonging to the CG258 or ST101 (López-Cerero et al., 2014; Porres-Osante et al., 2014; Oteo et al., 2016). In Catalonia, a single KPC-3-producing K. pneumoniae isolate belonging to ST258 was recovered from a human patient in 2015 (Piedra-Carrasco et al., 2018) but, to our knowledge, our study constitutes the first hospital outbreak caused by KPC-producing Enterobacterales in this region.

Here we report an outbreak caused by KPC-producing K. pneumoniae and E. coli in a tertiary hospital in Barcelona that initiated during the summer of 2018. It involved 125 Enterobacterales isolates recovered from surveillance (74.4%) or diagnosis (25.6%) samples. Outbreak identification was possible upon the implementation of an active surveillance program in ICU patients, but it is likely that KPC-producing K. pneumoniae isolates had already been circulating in asymptomatic fecal carriers for some time. The outbreak was caused by the rapid spread of a carbapenem-resistant ST15 K. pneumoniae strain carrying bla_KPC–2 in an IncFIIk plasmid of approximately 106 Kb and located within a non-Tn4401 genetic element (NTE_KPC-pMC-2-1).

Nevertheless, upon the initial stage of clonal spread, we speculate that the IncFIIk plasmid was successfully transferred first to E. coli isolates in patients either infected or colonized with the original ST15 strain, but later to other strains of K. pneumoniae or even Enterobacter spp. (Figure 1 and Supplementary Figure 4). This is supported by the fact that in at least 14 instances the same patient co-carried a bla_KPC–2-ST15 K. pneumoniae strain together with either a bla_KPC–2-K. pneumoniae belonging to a different sequence type or a bla_KPC–2-E. coli or Enterobacter sp. strain, but also because all plasmids carrying bla_KPC–2 in K. pneumoniae or E. coli strains shared high similarity with the IncFIIk plasmids from the ST15 K. pneumoniae strains, including the NTE_KPC-pMC-2-1 structure surrounding bla_KPC–2. Some genetic rearrangements, however, seem to have occurred, mainly in non-K. pneumoniae strains where the 106 Kb IncFIIk plasmid either co-integrated with other plasmids or suffered extensive genetic reduction. Therefore, intra- and inter-species dissemination of resistance also contributed to the spread of the outbreak. KPC transfer to other K. pneumoniae strains was not as heavily associated with co-carriage within the same patient, as opposite to inter-species dissemination, but such finding may also reflect a selection bias in the microbiology laboratory. Interestingly though, the clonal dissemination of non-ST15 strains was not detected. Recently, San Millán and co-workers proposed a similar hospital transmission dynamic for pOXA-48, where patient-to-patient transmission was tightly associated with the dissemination of a particular high-risk clone, while intra- and inter-species transmission of the plasmid was linked to concurrent gut colonization (León-Sampedro et al., 2021). As in the case of OXA-48, this genetic exchange represents an opportunity for the resistance gene to rearrange and shuffle into new plasmids and/or hosts, some of which may become more successful (Millan, 2018).

The genetic structures associated with the core structure of NTE_KPC-pMC-2-1 identified in this work (ISKpn27-bla_KPC–2-ΔISKpn6-korC) had already been reported by other studies and seem to constitute a common rearrangement associated with bla_KPC–2 mostly in isolates from China (Wang et al., 2015; Zhang et al., 2017). In Spain though, this structure has also been identified in IncP-6 plasmids from Citrobacter freundii, Enterobacter cloacae, and Klebsiella oxytoca but, to our knowledge, have never been associated with K. pneumoniae isolates of human origin, most likely suggesting a recent acquisition (Yao et al., 2017; Pérez-Vazquez et al., 2019).

Two sporadic K. pneumoniae strains belonging to the clonal group CG258 and carrying the bla_KPC–3 gene within a canonical Tn4401 in two different non-related plasmids were also detected in the study and, most likely, constituted an independent event unrelated to the spread of the outbreak. Nevertheless, one of such strains also managed to acquire a partial plasmid sequence containing the entire NTE_KPC-pMC-2-1 element.

The identification of a carbapenem resistant ST307 K. pneumoniae isolate in this study is also worth mentioning, as strains from this epidemic clonal group have recently been responsible for a hospital outbreak in Germany associated with multidrug resistance but also with the hypervirulent Klebsiella pneumoniae phenotype (hvKp). The hvKp phenotype was identified among ST307 isolates upon the acquisition of several plasmid-mediated virulence genes that merged with a resistance plasmid, hence creating a mosaic plasmid carrying both resistance and virulence genes (Heiden et al., 2020). In our study the ST307 strain (MC-2-230) carried bla_KPC–2 in a plasmid of 280 Kb also likely resulting from the merging of two different plasmids and, hence, the potential carriage of plasmid-mediated virulence genes was investigated. Fortunately, none of the plasmid-mediated virulence factors associated with the hvKp phenotype were present in plasmid pMC-2-230 nor in any other plasmid sequenced in this study. Likewise, none of the K. pneumoniae strains showed a hypermucoid phenotype, which is characteristic (although not exclusive) of hvKp (Lan et al., 2021).

We acknowledge several limitations in our study. First, WGS was performed under a single long-read sequencing approach, and we acknowledge that the use of a hybrid approach would have allowed for additional and more accurate comparisons. In addition, only a selected group of isolates were sequenced and sequence similarity was, therefore, assumed for the remaining isolates on the basis of PFGE, MLST and conventional PCR data. Unfortunately, further WGS analyses were beyond our possibilities but we expect that results from this study will contribute to a better implementation of WGS pipelines in our institution.

We report a hospital outbreak caused by the clonal dissemination of KPC-producing ST15 K. pneumoniae mainly among colonized carriers but also by the intra- and inter-species transmission of the bla_KPC–2 gene associated with plasmid conjugation and/or transposon dissemination. The ST15 clonal lineage is considered a high-risk clone and has been associated with KPC-2-producing isolates in Bulgaria and Vietnam (Markovska et al., 2015; Berglund et al., 2019), and it has also been reported in Portugal and Italy (Rodrigues et al., 2016; Fasciana et al., 2019). In Spain, ST15 has been reported but only associated with the production of OXA-48 (Madueño et al., 2017). Nevertheless, ST15 has been attributed with a high potential for horizontal gene acquisition and dissemination and it is critical that active surveillance strategies are prolonged over time to allow for the rapid detection and eradication of these highly resistant and virulent clones (Berglund et al., 2019).

Pepa Pérez Jove, Emma Padilla, and Mónica Ballestero-Téllez (Catlab, Centre Analítiques Terrassa AIE), Yuliya Zboromyrska, Miguel Ángel Benítez, Raquel Clivillé, Sabina González, and Iolanda Calvet (Consorci del Laboratori Intercomarcal de l’Alt Pendès, l’Anoia i el Garraf), Carmen Gallés (Corporació de Salut del Maresme i la Selva), Goretti Sauca (Hospital de Mataró), Carmina Martí-Sala and M^a Angeles Pulido (Hospital General de Granollers), Anna Vilamala (Hospital General de Vic), Araceli González-Cuevas (Hospital General del Parc Sanitari Sant Joan de Déu), Amadeu Gené (Hospital Sant Joan de Déu de Barcelona), Gloria Trujillo and Joan Lopez Madueño (Hospital Sant Joan de Déu de Manresa), Xavier Raga (Hospital Sant PauI Santa Tecla), Frederic Gómez, Ester Picó, and Carolina Sarvisé (Hospital Universitari Joan XXIII de Tarragona), Isabel Pujol and Xesca Font (Hospital Universitari Sant Joan de Reus).

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

MM-A, NF, and MF contributed to the conception, design, and implementation of the study, acquisition of laboratory and clinical data, analysis of the results, drafting the manuscript, and approval of the final version of the manuscript. CC, JViñ, ER, AC, and LM contributed to the acquisition of laboratory data, analysis of the results, and review and approval of the final version of the manuscript. MP, AV, IC, LR-S, GS, and AD contributed to the design of the study, acquisition of clinical data, analysis of the results, and review and approval of the final version of the manuscript. OF, PC, FB, FM, JM, and ÁS contributed to the design of the study, analysis of the results, and review and approval of the final version of the manuscript. CP, JVil, and IR contributed to the conception, design, and implementation of the study, analysis of the results, drafting the manuscript, and approval of the final version of the manuscript. All authors critically revised the manuscript for intellectual content and read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.