Abstract
Fluoroquinolones are a commonly used prophylaxis in transrectal ultrasound-guided prostate biopsy (TRUS-Bx), even though fluoroquinolone-resistant Escherichia coli has been associated with infectious complications after TRUS-Bx. The present study describes fluoroquinolone resistance mechanisms and antimicrobial susceptibility among intestinal E. coli, isolated from TRUS-Bx patients in a prospective study showing very few infectious prostate biopsy adverse events. This Multi-IMPROD sub-study included a total of 336 patients who received either ciprofloxacin, levofloxacin, or fosfomycin as prophylaxis before TRUS-Bx. E. coli could be cultured from 278 fecal swab samples, and 27 (9.7%) of these showed resistance to ciprofloxacin, and 14 (5.0%) were susceptible with increased exposure (I). Chromosomal and transferable fluoroquinolone resistance mechanisms were found among ciprofloxacin non-susceptible isolates, but both qnr genes and single gyrA mutations were found also among the ciprofloxacin-susceptible E. coli population. Low-level fluoroquinolone resistance is commonly associated with ESBL production in Enterobacterales. However, ESBL and qnr genes were not associated in our material, 14 isolates were ESBL producers and only 14.3% of them had the qnr gene, although 85.7% of the ESBL producers were ciprofloxacin non-susceptible. In the Multi-IMPROD substudy, only two mild urinary tract infections were reported, indicating that the antimicrobial susceptibility or resistance pattern of E. coli does not correlate with the onset of post-biopsy adverse events. We conclude that in our clinical settings, ciprofloxacin and levofloxacin prophylaxis is effective, and no severe post-biopsy infections were detected despite the intestinal colonization of genotypically and phenotypically fluoroquinolone-resistant E. coli.
Introduction
Prostate cancer is one of the most common cancer types in men, especially in Western countries. A prostate cancer diagnosis is based on transrectal ultrasound-guided prostate biopsy (TRUS-Bx) which is an invasive procedure, and antimicrobial prophylaxis is a common practice to lower the risk for infectious adverse events. Fluoroquinolones are widely used prophylaxis in prostate biopsy procedures, due to their good penetration to prostate tissue (1). Fluoroquinolone resistance has emerged globally during the last decades, and both fluoroquinolone-resistant Escherichia coli and isolates with reduced susceptibility have increasingly reported to cause sepsis or other infectious adverse events after prostate biopsy procedure (2-6). In Finland, fluoroquinolones are very common per oral treatment for urinary tract infections in men. Fluoroquinolone resistance in both urine and blood E. coli isolates has increased during the last decade, being 15.1% and 10.8%, respectively in Finres data of 2020 (7). The percentage of ESBL-producing E. coli has also increased during the last decade being 6.5% in men’s urine isolates and 6.6% in blood isolates in 2020 (7). In Enterobacterales, low-level fluoroquinolone resistance has been associated with ESBL production (8-11), and in Finland, 2.3% of the urine E. coli isolates were both ESBL producers and fluoroquinolone-resistant in 2020 (7). Low-level fluoroquinolone resistance is mainly caused by plasmid-mediated quinolone resistance (PMQR) determinants like qnr genes whereas high-level resistance is caused by several chromosomal mutations in the quinolone resistance determining region (QRDR) of DNA gyrase (gyrA/gyrB) and topoisomerase IV genes (parC/parE) (12-14). In E. coli, resistance to fluoroquinolones is highly associated with mutations in gyrA (12). In the present study, we determined the correlation between prostate biopsy adverse events, fluoroquinolone resistance mechanisms and antimicrobial susceptibility among intestinal E. coli isolated from fecal swab samples from men undergoing TRUS-Bx procedure in Finland.
Methods
Study population
This study is a substudy of IMPROD (Improved Prostate Cancer Diagnosis – Combination of Magnetic Resonance Imaging (MRI) and Biomarkers, NCT02241122) multi-center study (15). A prostate cancer screen with TRUS-Bx was performed for patients included in this study and a rectal swab sample was taken during the biopsy procedure. Between March 2015 and May 2017, a total of 336 rectal swab samples were collected from four hospitals in Finland: Helsinki University Hospital, 58 samples; Tampere University Hospital, 59 samples; Satakunta Central Hospital, 87 samples; and Turku University Hospital, 132 samples. Levofloxacin was used as antimicrobial prophylaxis in Turku, and ciprofloxacin in all the other study sites. In Helsinki, patients who had traveled abroad within three months before the TRUS-Bx received fosfomycin instead of ciprofloxacin.
Bacterial cultures
Swab samples were cultured on ChromAgar Orientation plate (Becton Dickinson, Heidelberg, Germany) and 5 µg-ciprofloxacin disk (OXOID, Thermo Scientific, Helsinki, Finland) was placed on top of the culture to select the patient’s most resistant E. coli strain. After the overnight incubation at 35 °C, two to three bacterial colonies with E. coli morphology (mauve to light purple colonies) were selected preferably near to the ciprofloxacin disk and pure cultures were made from these on CLED plates (Becton Dickinson, Heidelberg, Germany). Maldi-TOF (Bruker, Berlin, Germany) was used for species identification of the isolated strains. Only E. coli isolates were studied further.
Antimicrobial susceptibility testing
Antimicrobial susceptibility testing was performed with the disk diffusion method, according to EUCAST guidelines (16). Following antimicrobial disks were used: ciprofloxacin 5 μg, levofloxacin 5 μg, cefotaxime 5 μg, ceftazidime 10 μg, cefoxitin 30 µg, meropenem 10 µg, ampicillin 10 µg, amoxicillin-clavulanic acid 10/20 µg, mecillinam 10 µg, nitrofurantoin 300 µg, trimethoprim 5 µg, and trimethoprim-sulfamethoxazole 25 µg (OXOID, Thermo Scientific, Helsinki, Finland). Antimicrobial susceptibility profiles of each E. coli isolates were determined according to EUCAST clinical breakpoint table version 11.0 2021 (17). E. coli ATCC 25922 was used as a control strain in antimicrobial susceptibility testing.
PCR amplification and sequencing of chromosomal gyrA and parC gene mutations
Fluoroquinolone resistance mechanisms were studied in all E. coli isolates with a ciprofloxacin (CIP) disk inhibition zone ≤30mm. PCR amplification of gyrA and parC genes were performed with primers described in Table 1. The gyrA/parC PCR reaction (50 µL) consisted of 0.2 pmol/µL of each primer, 0.03 U/µL AmpliTaq Gold DNA polymerase, 5 µL AmpliTaq Gold buffer, 2 mM MgCl2 (Thermo Fisher Scientific Oy, Vantaa, Finland), and 0.2 mM dNTP mix (Life Technologies Europe, Espoo, Finland). The PCR program consisted of an initial denaturation at 94°C for 10 minutes, then 37 cycles of DNA denaturation at 94°C for 30 seconds, primer annealing at 55°C for 30 seconds, and extension at 72°C for 90 seconds. PCR-products were purified enzymatically with Exonuclease I- and FastAp Thermosensitive alkaline phosphatase -enzymes (Thermo Fisher Scientific Oy, Vantaa, Finland), and sequenced with BigDye v.3.1 sequencing using ABI3730xl DNA Analyzer at Institute for Molecular Medicine (FIMM, Helsinki, Finland). Detection of plasmid-mediated quinolone resistance genes. Transferable plasmid-mediated quinolone resistance (PMQR) qnr genes were screened for all the E. coli isolates with a ciprofloxacin disk inhibition zone ≤30mm. In addition, 34/90 randomly selected isolates with a ciprofloxacin inhibition zone >30mm were studied with previously reported primers and protocols (18).
Primers used in gyrA and parC PCR and Sanger sequencing
Detection of ESBL genes
Possible ESBL producers were screened according to EUCAST guidelines on the detection of Resistance mechanisms (19). ESBL + AmpC confirmation Kit (Rosco Diagnostics) was used to confirm ESBL production from E. coli isolates with reduced cefotaxime and ceftazidime susceptibility, inhibition zone <21 and <22mm, respectively. ESBL genes (blaCTX-M, blaSHV and blaTEM) were screened from these same isolates using previously described primers and PCR protocols (20).
Results
Antimicrobial susceptibility profiles
E. coli could be isolated from 278 patient samples (out of 336 fecal samples). The highest resistance rates were detected against ampicillin and amoxicillin-clavulanate (27.7% and 24.1%, respectively), also trimethoprim resistance was quite common (14.0%) (Figure 1). Totally 27 (9.7 %) E. coli strains were ciprofloxacin-resistant i.e. disk inhibition zone was <22mm and 14 (5.0 %) strains were ciprofloxacin-susceptible with increased exposure (I, CIP inhibition zone 22-24mm), according to the 2021 EUCAST breakpoints (17). Furthermore 22 (7.9%) E. coli strains were levofloxacin-resistant (R) (Figure 1). With correlation to given prophylaxis, 11 (9.6%) isolates from the levofloxacin group were resistant to both ciprofloxacin and levofloxacin, and six (5.3%) and four (3.5%) were ciprofloxacin- and levofloxacin-susceptible with increased exposure (I), respectively. Among the ciprofloxacin group, 16 (9.8%) and 11 (6.7%) isolates were ciprofloxacin and levofloxacin R, respectively, and eight (4.9%) and 11 (6.7%) isolates were ciprofloxacin and levofloxacin I, respectively.
Antimicrobial susceptibility profiles of 278 E. coli, isolated from fecal swab samples taken from men undergoing TRUS-Bx
Resistance against cefotaxime and ceftazidime was detected in 14 (5.0%) and 13 (4.7%) E. coli isolates. Of the ciprofloxacin-resistant E. coli isolates, nine (3.2%) were cefotaxime-resistant, and eleven (4.0%) showed co-resistance to trimethoprim. These eleven strains were also resistant to ampicillin and sulfamethoxazole-trimethoprim. Four of these were also gentamicin-resistant and three of these were ESBL producing strains. Ten ciprofloxacin-resistant E. coli (3.6%) were also ESBL producers and five (1.8%) isolates were both trimethoprim-resistant and ESBL producers. Antimicrobial susceptibility results from all tested antimicrobials are presented in Figure 1.
Fluoroquinolone resistance mechanisms and ESBL producers
Fluoroquinolone resistance mechanisms were found in all fluoroquinolone-non-susceptible isolates, i.e. CIP disk inhibition zone <25mm (17). Of the ciprofloxacin-resistant isolates, 23 had point mutations in the QRDR of both qyrA and parC, and four had point mutations only in gyrA. The most common gyrA mutation was S83L + D87N –double mutation found in 20 isolates, nine of these had also S80I + E84G –double mutation, and ten had S80I –single mutation in parC (Figure 2). Of the 14 E. coli isolates being ciprofloxacin I (susceptible with increased dose, according to EUCAST (17), disk inhibition zone 22-24mm), ten had S83L –single mutation and one had D87Y –single mutation in the QRDR of gyrA, and three isolates had only qnr gene (qnrA/B) as their only fluoroquinolone-resistance mechanism detected.
Fluoroquinolone (FQ) resistance mechanisms and ESBL genes detected among 276 E. coli isolates with ciprofloxacin (CIP) disk inhibition zone 6-30 mm. A dashed black line with R represents EUCAST and CLSI breakpoint (16, 23) for CIP-resistant isolates. A dashed line with black S represents the EUCAST breakpoint and a grey dashed line with S represents the CLSI breakpoint for CIP susceptible isolates.
R= resistant; S= susceptible
*Randomly selected 34 isolates with disk inhibition zone 31-36mm were tested for qnr genes
Fluoroquinolone resistance mechanisms were also found among ciprofloxacin-susceptible isolates. A total of seven isolates with ciprofloxacin inhibition zone between 25-29mm had a single gyrA mutation, and 31 isolates had qnr gene. The distribution of fluoroquinolone resistance mechanisms among the tested E. coli population is presented in Figure 2.
According to EUCAST Detection of resistance mechanisms guideline (19), 15 E. coli isolates (5.4%) were ESBL producers, nine of these were positive for blaCTX-M gene, four for blaCTX-M+blaTEM, and one strain for blaSHV +blaTEM. One screening positive isolate did not have any of these three ESBL genes analyzed. Of the ESBL E. coli strains, ten were ciprofloxacin-resistant and had chromosomal mutations, two were ciprofloxacin susceptible with increased dose, one of them had qnrS gene and the other had a single gyrA mutation. Two of the ESBL strains were ciprofloxacin-susceptible, but one of these had a single gyrA mutation. The distribution of ESBL genes and fluoroquinolone resistance mechanisms is presented in Figure 2.
Fluoroquinolone susceptibility and resistance mechanisms in correlation to post-biopsy adverse events
In the Multi-IMPROD study, only 12 minor post-biopsy adverse events were reported.21 The list of complications, prophylaxis, and microbiological findings, including ciprofloxacin susceptibility and possible resistance mechanisms of these 12 patients’ are presented in Table 2.
Patients with post-biopsy adverse events, used prophylaxis, ciprofloxacin susceptibility testing results and resistance mechanisms detected.
Discussion
In the present study, the antimicrobial susceptibility and fluoroquinolone resistance mechanisms of intestinal E. coli, isolated from faecal swab samples from men undergoing prostate biopsy procedures were analysed. Post-biopsy infections are a common side effect of prostate biopsy procedures, and especially fluoroquinolone-resistant and multidrug-resistant E. coli has been reported of being the main factor for infectious adverse events (2-5, 22). Our results showed that 14.7% of the tested E. coli strains were ciprofloxacin non-susceptible, and 85.3% were susceptible according to EUCAST breakpoints (17). If CLSI breakpoints would have been used the number of ciprofloxacin-resistant strains would have been the same (CIP ≤21mm), whereas the ciprofloxacin I isolates would have increased from 14 to 19 isolates (CIP 22-25mm vs. CIP 22-24mm) (23). Resistance to levofloxacin was more common in the study population who got levofloxacin prophylaxis than in ciprofloxacin prophylaxis group (9.6% vs. 6.7%, respectively), whereas ciprofloxacin resistance was practically equal in both groups (9.6% vs. 9.8%, respectively). Of the ciprofloxacin-resistant E. coli, 4.0% showed co-resistance to trimethoprim, 3.6% of were ciprofloxacin-resistant and ESBL producers, and 1.8% were ciprofloxacin- and trimethoprim-resistant and ESBL producers. Compared to overall susceptibility levels of E. coli isolates in Finland in 2020, fluoroquinolone resistance among the study population was higher compared to invasive isolates (10.8%), but equal to resistance levels in urine isolates (15.1%) (7). ESBL percentage among the intestinal E. coli isolates was somewhat higher, 5.4%, compared to the earlier reported ESBL carriage rate (4.7%) in Finland (24), but lower compared to ESBL-positivity rate in urine E. coli – isolates (6.5%) (7). In addition, resistance to trimethoprim was lower, whereas multidrug resistance was more common compared to that reported in Finres 2020 (7).
To become a high-level fluoroquinolone-resistant, bacterial strain needs to have mutations in both QRDR of gyrA and parC, however mutation even in one of these gene targets, S83L in gyrA, S80I in parC, and D87N in gyrA can cause a clinically relevant fluoroquinolone resistance in E. coli (25). In our study, double gyrA+parC mutations were detected only in high-level resistant isolates whereas single gyrA mutations were detected even phenotypically ciprofloxacin-susceptible isolates. EUCAST has started using the term ATU (Area of Technical Uncertainty) to warn laboratories that, there can be interpretative difficulties with susceptibility test results in this area, and susceptibility interpretation should be carefully evaluated (26). According to EUCAST breakpoints for ciprofloxacin (17), the ATU area is situated between the resistant and susceptible populations (22-24mm), i.e. I area, and includes the isolates which can be considered susceptible with increased exposure. As we show in Figure 2, isolates within the area of technical uncertainty, have chromosomal gyrase mutations, indicating that it is justified to consider these isolates rather resistant than susceptible. PMQR determinants like qnr genes are linked to low-level fluoroquinolone resistance but these genes also enhance the selection of high-level resistance, furthermore, the qnr genes are also easily missed until further mechanisms are acquired and detected (11-12,14, 27-28). We show that the qnr genes were detected among ciprofloxacin-resistant, ciprofloxacin I, and ciprofloxacin-susceptible strains, indicating that qnr genes do not necessarily have a clinical relevance without additional resistance mechanisms (25). PMQR genes are associated with the same plasmids as ESBL genes (9-11, 27) and there are reports on post-biopsy infections linked to ESBL E. coli isolates with co-resistance to fluoroquinolones (29). In our study, 85.7% of the ESBL strains had chromosomal mutations in QRDR and only one ESBL strain had a PMQR gene as the only fluoroquinolone resistance mechanism, indicating that ESBL-producing strains and strains with qnr genes were not associated in our material. In addition, among the tested ciprofloxacin-susceptible E. coli isolates with disk inhibition zone 31-36mm, no fluoroquinolone resistance determinants were detected, as expected. Despite the ciprofloxacin and multidrug resistance detected in our study population, no severe post-biopsy infections were reported in the Multi-IMPROD study (21). It is known, that the use of fluoroquinolones increases the risk for intestinal colonization of fluoroquinolone-resistant E. coli, and further post-biopsy infectious complications (6,30). In the present study, there were both phenotypically ciprofloxacin-resistant E. coli and isolates with susceptible phenotype but having fluoroquinolone resistance mechanisms, and despite this, only two mild UTI cases were reported. One of them had ciprofloxacin-susceptible E. coli in a swab sample, and the other had ciprofloxacin I E. coli with a single gyrA mutation, indicating that fluoroquinolone resistance alone does not explain the possible adverse events after TRUS-Bx. There are reports on certain E. coli clones, such as ST131 that have spread widely and have caused severe infections (3), and thus further studies are needed to evaluate the role of virulence factors and other resistance mechanisms such as efflux pumps, in post-biopsy infections.
Conclusion
We show in this study that fluoroquinolone resistance mechanisms, both the qnr genes and gyrA mutations, were also found among the ciprofloxacin-susceptible E. coli population and that ESBL and transferable qnr genes were not associated in our material. We conclude that the onset of post-biopsy adverse events did not correlate with antimicrobial susceptibility, since no severe post-biopsy infections were detected despite the intestinal colonization of genotypically and phenotypically fluoroquinolone-resistant E. coli.
Data Availability
All data produced in the present study are available upon reasonable request to the authors
Funding
This study was funded by the Sigrid Juselius Foundation (for Peter J. Boström), and Finnish Governmental Special Funding, The Cancer Foundation Finland, University of Turku Combined Research Funding, and the Turku University Hospital Foundation (for Juha Knaapila).
Conflict of Interest
All authors declare no conflict of interest regarding this publication.
Transparency declarations
None to declare
Acknowledgments
We thank Minna Lamppu for her skillful technical assistance with cultured E. coli isolates
References
