Genotypic characterisation and antimicrobial resistance of Pseudomonas aeruginosa strains isolated from patients of different hospitals and medical centres in Poland

Pseudomonas aeruginosa isolates (n = 742) were collected from 541 patients of 11 different hospitals and medical centres located in north-western Poland between December 2015 and March 2017. From the first isolate per patient, 202 were randomly selected for typing by Pulsed Field Gel Electrophoresis and antimicrobial drug resistance analysis.

Hospitals where the isolation was carried out were designated with letters H1 – H11. The largest number of selected isolates (n = 174) originated from four major general hospitals H1-H4, located in three different cities of north-western Poland. The isolation of P. aeruginosa strains was less frequent in other smaller specialised hospitals and care facilities in the chosen region of Poland and thus the number of strains from these medical units is lower. Detailed information regarding the hospitals and the amount of strains collected at different hospital wards is included in Table 1. The selected strains were isolated from blood (n = 35), the lower respiratory tract (n = 76), wounds (n = 65), urine (n = 12), and ear and eye swab samples (n = 14). Firstly, P. aeruginosa isolates were identified in the hospitals’ microbiology laboratories. Further genus identification was performed using the PCR method described by Spilker et al. [20].

The antipseudomonal drugs assessed in the study included carbapenems (meropenem, imipenem), cephalosporins (ceftazidime, cefepime), aminoglycosides gentamicin, amikacin, tobramycin, fluoroquinolone (ciprofloxacin), a penicillin + β-lactamase inhibitor (piperacillin-tazobactam) and polymyxin (colistin). To identify antimicrobial susceptibility, the disk diffusion method was performed. Pseudomonas aeruginosa ATCC 27853 was used as quality control. Minimal Inhibitory Concentration (MIC) of colistin was determined with a broth microdilution-based method (ComASP™ Colistin test, Liofilchem) as the disk diffusion method is no longer recommended to evaluate polymyxins susceptibility. Concentrations used varied between 0.25 μg/ml and 16 μg/ml. Tests were performed and interpreted according to the EUCAST guidelines [21].

The chromosomal P. aeruginosa DNA was isolated with the CHEF Bacterial Genomic DNA Plug Kit (Bio-Rad Laboratories) with some modifications of the manufacturer’s instructions. Pseudomonas aeruginosa isolates were cultured overnight at 37 °C on cetrimide agar. In the first stage of isolation, bacterial suspensions in PBS equivalent to 4–4.5 McFarland were prepared, then incubated in a water bath at 37 °C for 20 min. A 100 μl sample of each bacterial suspension was mixed with 100 μl of melted 2% agarose gel at 50 °C. These mixtures were allowed to solidify in plug moulds. The resulting agarose plugs were mixed by inversion with a proteinase buffer with 20 μl of proteinase (CHEF Bacterial Genomic DNA Plug Kit, Bio-Rad Laboratories), then incubated for 20 h at 50 °C. After incubation, the proteinase solution was removed and the plugs washed three times with 1 ml of 1x wash buffer from the commercial kit (1 wash/1 h). The plugs were either stored at 4 °C or used immediately in further analyses.

Before restriction digestion, plugs were washed once in a 0.1x wash buffer for 1 h, followed by a subsequent wash with a tango buffer (Thermofisher), again for 1 h. For the restriction endonuclease digestion, a SpeI restriction enzyme was used (Thermofisher). Each plug was resuspended in fresh tango buffer and 50 U of SpeI was added, followed by incubation for 20 h at 37 °C.

A clonal analysis of P. aeruginosa strains was carried out for 202 strains isolated from 11 different hospitals located in north-western Poland. Pseudomonas aeruginosa isolates were distributed among 116 different pulsotype groups. Thirty clusters consisted of two or more strains. The clusters were assigned with different letter designations from A to AH. A total of 86 clusters contained only one strain. Strains clonally related with at least one other isolate in this study were found in all hospitals except hospital H6. Five groups of clonally related strains (D, E, J, T, AC) included more than 5 strains (Fig. 1). The most prevalent D cluster included 17 strains isolated from two hospitals H2 (n = 16) and H1 (n = 1). Strains of those groups were isolated at the ICU (n = 7), BU (n = 8), and surgical unit (SU, n = 1), from December 2015 to March 2017. All isolates of group D were resistant to imipenem. Group E included 13 isolates collected between December 2015 and December 2016. Isolates of this cluster were isolated from different patients of an ICU (n = 12) and orthopaedic unit (n = 1) of hospital H3. Isolates of group D and E represented 8.6 and 6.4% of all isolates collected, respectively. Strains of cluster E were isolated 12 times from patients at ICUs of hospital H3, which stands for 25% (12/48) of all isolates collected in this hospital. Strains of cluster D were collected 16 times from patients of a burn unit and the ICU of H2, representing 25.4% of all isolates from this medical centre. Most of the isolates of cluster D were resistant to carbapenems (14/17), aminoglycosides (12/17), and other antimicrobials tested except piperacillin with tazobactam and colistin. Similar high rates of resistance to various antimicrobial classes were noted for strains from pulsotype E. Both clusters D and E share resistance to imipenem and aminoglycosides.

Strains of pulsotypes J, T, and AC were isolated 8, 7 and 6 times from different patients. Restriction patterns, hospitals, wards, isolation dates, and antimicrobial susceptibility patterns of P. aeruginosa strains isolates belonging to the major pulsotype groups are shown in Fig. 1. In this study, we found 13 clusters of strains: A, D, J, K, M, N, Ó, P, T, X, AC, AD, and AH that were isolated from patients of two or more hospitals. Strains of group T were collected in 3 different hospitals H1 (n = 2), H3 (n = 4), and H7 (n = 1). Isolates of the T cluster were susceptible to most of the antimicrobials tested, except for imipenem, for which 4 strains were resistant.

Most of the P. aeruginosa strains from this study were isolated in ICUs (n = 81). All remaining isolates were isolated in BU (n = 35), SU (n = 35), and other units (n = 51). Strains from clusters containing two and more strains were most frequently isolated in ICUs. Among strains isolated from these units, 56/81 (69.1%) were clonally related to at least one other strain in this study. In BU, 21/35 (60%) isolates have similar restriction patterns shared by two or more strains. In Fig. 2, we demonstrated a distribution of PFGE clusters of various sizes among medical units of different types.

Overall, our findings show the presence of multiple P. aeruginosa pulsotypes and a constant exchange of strains inside a single hospital ward, between different hospital wards and between different hospitals. Part of the clonally related strains were constantly isolated from different patients of the same hospital ward. Results presented in this work suggest a non-clonal structure of P. aeruginosa population of strains isolated at hospital centres located in north-western Poland.

Resistance to tested antipseudomonal drugs was as follows (202 strains): imipenem 67.8% (n = 137), meropenem 29.2% (n = 59), ceftazidime 33.2% (n = 67), cefepime 42.6% (n = 86), piperacillin-tazobactam 39.6% (n = 80), gentamicin 37.6% (n = 76), amikacin 30.2% (n = 61), tobramycin 38.1% (n = 77), ciprofloxacin 39.6% (n = 80), and colistin 3.0% (n = 6). The prevalence of resistance to both carbapenems was 29.2%, and for all aminoglycosides 22.8%. Combined resistance to carbapenems and aminoglycosides was 14.4% (n = 29). Out of 202 isolates in this study, 48% (n = 97) were multidrug-resistant (MDR). Multidrug-resistance was defined as not susceptible to 3 or more antibiotic classes (n = 20) [22].

Among MDR strains, 71.1% (n = 69) of isolates were clonally related with at least one more isolate in the study. However non-MDR susceptible strains (n = 105) less frequently exhibited clonal relatedness with other strains 44.8% (n = 47).

We observed a variation in antimicrobial resistance rates between different hospital wards. Strains isolated from BU were most often resistant to the tested antimicrobials except piperacillin with tazobactam, ciprofloxacin and colistin. Data regarding antimicrobial resistance of strains isolated at different medical units are shown in Table 2.