Introduction

Multidrug-resistant Enterobacterales bacteria have become a serious global concern, with limited therapeutic options for their control [1]. To meet the current challenges, there is an urgent need to discover new antimicrobials, or to re-examine known compounds such as fosfomycin, polymyxins, and temocillin [2].

Temocillin is a β-lactamase-resistant penicillin. It is the 6-α-methoxy derivative of ticarcillin, and is resilient to all classical and extended-spectrum TEM, SHV, and CTX-M enzymes and AmpC β-lactamases. Temocillin is used in few Western Europe countries [3], and only limited data are available on temocillin susceptibility in Eastern Europe, where microbial resistance to many antimicrobials is prevalent. According to the most recent report from the European Centre for Disease Prevention and Control in Poland over 65% of Klebsiella pneumoniae isolates reported to the European Antimicrobial Resistance Surveillance Network (EARS-Net) in 2018 were resistant to extended-spectrum cephalosporins; this represents an increase compared to previous years [4].

The aim of the present study was to evaluate the in vitro activity of temocillin and compare it to the activities of commonly used antimicrobials in respect of a large collection of Enterobacterales bacteria, especially ESBL- and/or AmpC-producing strains isolated from Polish patients with various infections.

Materials and methods

The non-duplicate, well-characterized, clinical Enterobacterales isolates used in the present study were collected during laboratory surveillance conducted by the National Reference Centre for Susceptibility Testing between January 2000 and 2017 (Table 1). All the isolates were sub-cultured from storage (at -70 °C) and reidentified prior to testing. ESBL production and AmpC expression were verified using a double-disk synergy test, as previously described [5]. Klebsiella pneumoniae carbapenemase (KPC) production was verified using the disk test combined with phenylboronic acid [6] followed by polymerase chain reaction (PCR) [7]. β-lactamase production was controlled by PCR using specific primers for blaCTX-M-1-, blaSHV-, blaTEM-, and blaKPC-like genes [8]. The blaKPC gene amplicons were all digested using a restriction enzyme (RsaI; Thermo Scientific, Vilnius, Lithuania), which allowed blaKPC-2- and blaKPC-3-like variants to be distinguished [9].

Table 1 Characterisation of the Enterobacterales strains tested
Full size table

The minimal inhibitory concentrations (MICs) of temocillin and 16 other antibiotics listed in Table 2 were evaluated using the microdilution method according to standard ISO 20776-1. The MIC of ceftazidime/avibactam was determined only for KPC-producing strains using a MIC Test Strip (Liofilchem®, Roseto degli Abruzzi, Italy). The quality control strains used in the study were: Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853, and Escherichia coli mcr-1-producing strain. The data were interpreted using EUCAST guidelines, except for temocillin and cefoxitin, where BSAC and CLSI breakpoints were used, respectively [10,11,12].

Table 2 Results of antibiotic susceptibility evaluation of 400 Enterobacterales strains
Full size table

Results

The 400 isolates tested were recovered from urine (35%), blood (13%), and other clinical specimens (52%; from bronchial secretions, cerebrospinal fluid, peritoneal fluid, pleural fluid, pus, skin lesions, sputum, and wounds) (Table 1). With regard to the resistant phenotypes, isolates producing ESBLs (CTX-M, SHV, or TEM) and/or isolates with acquired or overexpressed AmpC-type β-lactamase were the most prevalent (n = 260, 65%), followed by KPC-producing isolates (n = 40, 10%), mostly represented by KPC-like-3-positive bacteria (n = 24, 60%) (Table 1).

The results of a detailed analysis of the susceptibility data are shown in Table 2. The MICs of temocillin ranged from 1 to 256 mg/L. Most of the isolates (n = 364, 91%) had an MIC value between 4 and 32 mg/L (MIC50 and MIC90 values of 8 and 32 mg/L, respectively) (Table 2). Overall, 91% of the isolates were susceptible to temocillin according to the BSAC urinary breakpoint (≤ 32 mg/L), and 61.7% were susceptible according to the systemic breakpoint. Temocillin was very effective against all the species tested when the urinary breakpoint was used (Table 3). According to the systemic breakpoint, temocillin was less effective against Klebsiella and CESP spp. (Citrobacter, Enterobacter, Serratia, Providencia, Morganella, and Hafnia that produced inducible chromosomally encoded AmpC-type) with resistance values of 53.1% and 44.6%, respectively. Among the CESP members, the most resistant species were Serratia marcescens (88.9%) and Enterobacter spp. (35.1%) (data not shown).

Table 3 Results of temocillin susceptibility evaluation of 400 Enterobacterales strains
Full size table

According to the urinary breakpoint, among the ESBL- and/or AmpC-producing isolates, the temocillin susceptibility rates were between 94% and 97%. According to the systemic breakpoint, the susceptibility rates were 22.7% for ESBL- and AmpC-producers, 61.6% for ESBL-producers, and 74.4% for AmpC-producers (Table 2).

The KPC producers were resistant to several antimicrobials. These isolates were all resistant to temocillin according to the systemic infection breakpoint, but only 42.5% were susceptible according to the urinary tract infection breakpoint. Ceftazidime/avibactam (100% susceptibility) and colistin (95% susceptibility) were the most effective against KPC-producing isolates.

Discussion

Temocillin has been used for several years in some European countries, and is approved for the treatment of septicaemia, and urinary tract and lower respiratory tract infections [13,14,15,16]. However, it is still not available in Poland. Therefore, in the present analysis we assessed the prevalence of susceptibility to temocillin in Enterobacterales bacteria before it becomes available in our country.

There are currently no EUCAST breakpoints for temocillin [10]. Therefore, Enterobacterales bacteria are categorized as susceptible at MIC values of 8, 16, or 32 mg/L, depending on the country [14]. In the present study, we used BSAC [11] clinical breakpoints with criteria established separately for systemic (8 mg/L) and urinary tract (32 mg/L) infections. The temocillin MIC50 and MIC90 values for the total set of isolates calculated according to the urinary breakpoint were higher than those reported by Alexandre et al. [14] for urinary tract infection (UTI) isolates (8 and 32 mg/L versus 3 and 6 mg/L, respectively). However, our set of isolates comprised both UTI cases and isolates from other sites of infection. When MICs were calculated only for isolates from urine, the MIC50 and MIC90 values were similar (data not shown), suggesting a possible role for temocillin in the treatment of urinary tract infections in Poland, regardless of the clinical breakpoint used. As expected, based on the literature non-ESBL/AmpC/KPC-producing isolates were very susceptible (100% and 99%), regardless of the clinical breakpoint used [15,16,17,18].

With regard to bacterial species tested and according to the urinary breakpoint, temocillin retained a high level of activity against all bacterial species producing ESBL and/or AmpC enzymes. However, when the systemic infection breakpoint was applied, only E. coli and Proteus spp. species remained susceptible, suggesting a possible role for temocillin in the treatment of urinary tract infections due to those two species [19]. More generally, temocillin was highly effective against most Enterobacterales bacteria, especially when the urinary tract infection breakpoint was used, and was least active against K. pneumoniae. The second most temocillin-resistant species was S. marcescens, with 77.8% of resistant isolates harbouring CTX-M enzymes (data not shown).

The majority of ESBL- or ESBL/AmpC-producers were susceptible using the urinary breakpoint but the resistance rate increased significantly using the systemic breakpoint. Previously, Rodriguez-Villalobos et al. [18] and Kresken et al. [20] pointed out that CTX-M-15 producers were less frequently susceptible to temocillin than other CTX-M-type-producing isolates. In Poland, as in other European countries, the population of Enterobacterales bacteria has been dominated by CTX-M-15 producers [21].

With regard to K. pneumoniae, the temocillin-resistant strains were dominated by CTX-M- and KPC-producing isolates, regardless of the sample origin (data not shown). Contrary to the previous reports by Adams-Haduch et al. [22] and Woodford et al. [23], we were unable to confirm the susceptibility to temocillin among KPC-producing isolates. In the present study, 100% and over 50% of the KPC-producers were resistant to temocillin according to the systemic and urinary breakpoints, respectively. Our data are however in line with a report from Greece describing low temocillin activity against KPC-producers: 97.3% and 42% were resistant according to systemic and urinary breakpoints, respectively [24]. The low number of KPC-producing isolates is a limitation of the study, but the set of clinical isolates tested was representative of the population of KPC-producing isolates cultured from clinical specimens between 2010 and 2017 in Poland. However, among our KPC-producers, 20% had CTX-M enzymes and this could already be a reason why they are more resistant to temocillin. It would be of interest to further investigate the presence of other resistance mechanisms such as upregulated efflux or permeability for instance.

The present study is the first Polish evaluation of the in vitro susceptibility of Enterobacterales isolates to temocillin. It demonstrated the effectiveness of the antibiotic against a collection of tested microbes, especially when the urinary breakpoint was used. In contrast with previously published data, our study did not confirm the susceptibility of KPC-producing isolates to temocillin.