Report: antimicrobial resistance in commensal E. coli from poultry, pigs, cows and veal calves. 2014

Transcription

1 CODA-CERVA Report: antimicrobial resistance in commensal E. coli from poultry, pigs, cows and veal calves Vicky Jasson and Pierre Wattiau Veterinary and Agrochemical Research Centre 1 Introduction Commensal E. coli are regarded as general indicators for resistance amongst Gram negative bacteria. They have the advantage of being present in nearly all animal species and in large numbers. As such they can be isolated from an animal at almost every sampling occasion. Because they are continuously present, they can be used to follow up resistance evolution in time. Moreover they have been studied frequently in other countries so they are useful in comparing geographic distribution of antimicrobial resistances. The genetic background or resistance in this species is also quite well known allowing a scientific interpretation of the resistance data. This report presents the fourth year of antimicrobial resistance in commensal E. coli from animals in Belgium. The surveillance program is conceived to follow up trends in prevalence of resistance at the national level. The data presented in this report and previous reported data ( will allow us to detect eventual trends in evolution of antimicrobial resistance. This is important seen measures are being prepared to reduce antimicrobial usage and the effect of the measures can then be compared to this baseline set. 2 Materials and Methods 2.1 Sampling Samples from fecal material were taken from 4 animal categories: broiler chickens, pigs, bovines (for meat production) and veal calves. Samples were taken by samplers of the Belgian Food Agency Poultry Caecal content of broiler chickens was taken at slaughter together with the samples in the framework of Salmonella control program. Caeca from 10 animals were collected and pooled. One sample originated from one farm Pig Pooled fresh fecal material of at least ten animals of approximately 6 months old was collected from slaughter pigs at the abattoir. One sample originated from one farm Bovines Pooled fresh fecal material was collected from the floor of barns harboring bovines for meat production of less than 7 months of age. One sample originated from one farm Veal calves Pooled fresh fecal material was collected at the abattoir from veal calves. One sample originated from one farm. 2.2 Isolation and identification Fecal material was inoculated and E. coli was identified at DGZ or ARSIA as described below. At DGZ fecal material was inoculated on McConkey agar and incubated at 37 C for 18 to 24 hours. Suspected colonies (pink, lactose positive) were inoculated on Kligler and indol medium and incubated at 37 C for 18 to 24 hours. When the test outcome was positive for E. coli a colony from the Kligler medium was inoculated on Mac Conckey agar, incubated at 37 C for hours and sent to CODA-CERVA. At ARSIA, fecal material was inoculated on Gassner medium and incubated at 37 C for 18 to 24 hours. Suspected colonies were purified on Columbia agar supplemented with 5% sheep blood. Identification was done by the OPNG test, Ureum test and indol test. Confirmed E. coli were sent to CODA-CERVA. 2.3 Susceptibility testing Strains were sent to the national reference laboratory (CODA-CERVA) for susceptibility testing. Upon arrival, the strain was purified on Columbia agar with 5% sheep blood, confirmed to be E. coli by MALDI-TOF and susceptibility was tested using a micro broth dilution method (Trek Diagnostics). To this end, 1 to 3 colonies were

2 suspended in sterile physiological water to an optical density of 0.5 McFarland. Ten microliter of this suspension is inoculated in 11ml cation adjusted Mueller Hinton broth with TES buffer. Fifty microliter of the Mueller-Hinton broth with bacteria was brought on a micro-titer plate with the antimicrobials lyophilised, the EUVSEC plate as produced by Trek Diagnostics, using the auto-inoculating system of Trek Diagnostics. Plates were incubated hours at 35 C and read. The Minimal Inhibitory Concentration (MIC) was defined as the lowest concentration by which no visible growth could be detected. MICs were semi-automatically recorded by the Trek Vision system using the SWIN software. Results were automatically exported to an Excel file. Table 1 shows the antimicrobials tested and their abbreviations. 2.4 Analysis of data Data from the Excel file generated by the software of the semi-automated susceptibility equipment (sensivision, Trek Diagnostics) and merged to the administrative date from the LIMS system at CODA. These files were validated for consistency. The excel file was then imported into an Access file in which the number of strains having an MIC for a certain antibiotic were calculated. These data were set in a table that was subsequently exported to an Excel file. The data were interpreted for susceptibility using breakpoints based on the EUCAST ECOFFs or as defined by the EU reference laboratory on antimicrobial resistance (DTU). The number of resistant strains was counted and resistance percentages were calculated. Exact confidence intervals for the binomial distribution were calculated using a visual basic application in Excel. A 95% symmetrical two-sided confidence interval was used with p= The lower and upper bound of confidence interval for the population proportion was calculated. Based on the Pearsons chi-square test, and where appropriate the Fischer exact test, significance of the differences were calculated. Multi-resistance was determined by transforming the MIC data into resistant (R) and susceptible (S). Number of antimicrobials to which a strain was resistant to was counted and cumulative percentages were calculated. The modal number of antimicrobials to which 50% of the strains was resistant was calculated. Graphical representations were prepared in Excel. 3 Results Results are shown in tables 2 to 5 and figures 1 to Poultry Susceptibility of commensal E. coli originated from poultry, towards the different antimicrobials tested is shown in table 2. A total of 158 strains were tested. Highest resistance was seen against ampicillin (72.8%) followed by ciprofloxacin (69.6%), nalidixic acid (63.3%), closely followed by sulphamethoxazolone (58.2%), trimethoprim (49.4%) and tetracycline (45.6%). Less than 10% of the strains were resistant to extended spectrum cephalosporins and none of them were resistant towards meropenem. All ceftazidime resistant strains were resistant to cefotaxime, as expected, while one strain was cefotaxime resistant and showed not ceftazidime resistance. All strains were further analyses with EUVSEC2 plates as they were suspected to be ESBL carrying strains. Sevens strains showed to be presumptive ESBL, four strains were presumptive ESBL+pAmpC, one strain was presumptive pampc and one strain had an unusual phenotype. The presumptive ESBL strains are clearly multi-resistant with all strains resistant to at least 7 antimicrobials tested (data not shown). Ciprofloxacin and nalidixic acid are both antimicrobials from a same class and likewise, resistance is frequently cross-resistance. The extreme high prevalence of fluoroquinolone resistance is worrisome. Slightly more resistance against ciprofloxacin can be noted, which indicates the presence of mobile fluoroquinolone resistance, without mutational resistance. This resistance can be mediated by qnr (DNA protection), qepa (efflux) or aac(6 )- Ib-cr genes (aminoglycoside acetyl transferase showing cross resistance with several aminoglycosides)(redgrave et al., 2014). Resistance to sulphamethoxazolone (58.2%) and trimethoprim (49.4%) was substantial while resistance to the combination (as determined by resistance to both components) was evident in 48.1% of the strains (CI , data not shown), which is substantial as well. It should be noted that chloramphenicol resistance remains high with 20.9% while the resistance against meropenem, gentamicin, colistin and tigecycline remained below 6%. Considering multiresistance (Figure 1 and 2) it should be noted that 11.4% of the strains tested remained fully susceptible. 54.4% of the strains were resistant to at least 4 different antimicrobials. However, cross-resistances should be taken into account. All nalidixic acid resistant strains are also resistant to ciprofloxacin, and strains resistant to cephalosporins are also resistant to ampicillin. One strain was resistant to 9 antimicrobials tested, remaining only susceptible to colistin, gentamicin, tigecycline and meropenem. Eight strains (5%) were resistant

3 to 8 antibiotics remaining only susceptible to meropenem, colistin, tigecycline and one other varying antimicrobial. 3.2 Pigs Susceptibility of commensal E. coli with pigs origin towards the different antimicrobials tested is shown in table 3. A total of 184 strains were tested. Highest resistance was seen against sulphamethoxazolone (52.2%) and trimethoprim (50.0%), followed by tetracycline (44.0%) and ampicillin (41.3%). Resistance towards extended spectrum cephalosporins ceftazidime and cefotaxime was low (0.5%). The strain was resistant towards cefotaxime and ceftazidime (presumptive ESBL; data not shown). All strains tested remained susceptible to meropenem and tigecycline. The resistances towards quinolones were low (1.1% for nalidixic acid and 2.2% for ciprofloxacin). The combination sulphamethoxazolone -trimethoprim was determined to be 47.8% (CI ) (data not shown), which is only marginally lower than for the separate components (52.2% and 50.0% respectively), indicating that most strains are resistant to both compared to the single components separately. Resistance against 8 of the antibiotics tested remained below 2.5%, of which notably colistin, an antibiotic largely used in the treatment of diarrhoea. Resistance to chloramphenicol remains with 28.8% relatively high. Figure 3 shows that 28.8% of all strains tested remained fully susceptible. 51.6% of the commensal E. coli from pigs was resistant to at least 2 antimicrobials (mostly the combination sulphamethoxazolone-trimethoprim). One stain was resistant towards 8 antimicrobials remaining only susceptibly towards meropenem, colistin, tigicycline and the cephalosporins tested. 3.3 Veal calves (PRI 036) 188 strains isolated from veal calves were analysed (table 4). The highest prevalence of resistance in bovine commensal E. coli is against tetracycline (68.6%), sulphamethoxazolone (58%), ampicillin (55.3%) and trimethoprim (51.6%) followed by resistance towards chloramphenicol (25.5%), ciprofloxacin (22.3%) and nalidixid acid (20.7%). Low resistance can be noticed for gentamycin, colistin and the cefalosporines cefotaxime and ceftazidime. One strain isolated was resistant towards cefotaxime and ceftazidime and showed to be a presumptive ESBL (data not shown). All strains tested remained fully susceptible towards meropenem and tigecycline. The combined resistance against sulphamethoxazolone-trimethoprim, as calculated from the combine resistance against the two components is 49.5% (CI: data not shown). This is nearly the same as trimethoprim resistance. As for E. coli isolated from poultry and pigs, ciprofloxacin resistance in E. coli isolated from veal calves was higher than nalidixic acid resistance, indicating the presence of plasmid mediated quinolone resistance. Figures 5 and 6 show that 26.1% of the strains are fully susceptible to all antimicrobials tested while two strains are resistant to 9 different antimicrobials. 3.4 Bovines (PRI 515) From the 164 strains isolated from bovine, resistance in commensal E. coli (table 5) was highest against sulphamethoxazole (23.2%), ampicillin (20.1%) and tetracycline (17.7%) followed by chloramphenicol (15.9%) and trimethoprim (15.2%). Resistance towards the other antimicrobials tested remained under 10% and all strains tested remained susceptible towards meropenem and tigecycline. The combined resistance against sulphamethoxazole-trimethoprim, as calculated from the combined resistance against the two components is 15.2% (CI: data not shown). Four strains were resistant to cephalosporines cefotaxime and ceftazidime, three strains were presumptive ESBL and one strain was presumptive pampc (data not shown). One strain showed resistance towards 10 antimicrobials remaining only susceptible towards meropenen, tigecycline and colistin. One stain was colistin resistant and was also associated with multi-resistance, (9 antimicrobials). 3.5 Beta-lactam resistance 19 strains were subjected to a confirmatory test for ESBL production (data not shown). 12 strains (63.2%) were determined as presumptive ESBL, 21.1% of the strains were had the phenotype presumptive ESBL + pampc and 2 stains (10.6%) were presumptive pampc. Only one strains had an unusual phenotype. 3.6 Comparison between animal species Clearly, commensal E. coli from poultry are the most resistant E. coli strains tested with only 11.4% of the strains reaming fully susceptible. Next in row are the strains from pigs and veal calves, which have less than 30% of the

4 strains remaining fully susceptible (28.8% and 26.1% respectively). Commensal E. coli from bovine for meat production are the less resistant, 71.3% remaining fully susceptible. Multi-resistance is mostly seen in commensal E. coli from poultry (Figure 1) and strains from poultry are most resistant to cephalosporins of the third generation. Chloramphenicol resistance varies from % for the different origins. There is a large difference in ciprofloxacin resistance (2.2% in pigs and 69.6% in poultry). The resistance against ciprofloxacin in bovine for meat production is slightly larger than for veal calves. Colistin resistance remains low for E. coli from all animal species (>2.7%). All strains tested remained susceptible towards meropenem and tigecycline. 3.7 Conclusions Resistance in commensal E. coli from different animal species remains problematic. Resistance against quite some antibiotics is so high that their efficacy may be questioned whether they would be pathogenic. However, the extrapolation to pathogenic strains is difficult due to the scarcity of the data and the fact that different bacteria have different capabilities to take up resistance genes or have different rates of mutations leading to antimicrobial resistance. However, the resistance in commensal E. coli from veal calves is quite low compared to what has found in Belgium in pathogenic (neonatal) strains. Since last year, the sample is now only at the slaughterhouse, thus from the older animals, the age effect should be taken into account seen that in general, in the younger animals, resistance is higher. Poultry strains remain the most resistant strains and show also the highest level of multi-resistance. In cattle, resistance is lowest, however showing a large variety in multi resistance with some strains highly multiresistant. Resistance against the older antibiotics remains highest while against the more newly introduced antibiotics resistance remains quite low. An exception to this is colostin for which resistance remains low. At the other hand, chloramphenicol resistance remains at remarkably high levels seen it is not used anymore for over 20 years. 3.7 References Redgrave L.S., Sutton S.B., Webber M.A., and Piddock L.J.V. (2014). Fluoroquinolone resistance: mechanisms, impact on bacteria, and role in evolutionary success. Trends in Microbiology, (22),

5 Table 1. List of abbreviations Abbreviation AMP AZI CHL CIP COL FOT GEN MERO NAL SMX TAZ TET TGC TMP Ampicillin Azithromycin Chloramphenicol Ciprofloxacin Colistin Cefotaxime Gentamicin Meropenem Nalidixic acid Sulphamethoxazole Ceftazidime Tetracycline Tigecycline Trimethoprim

6 Table 2. Antibiotic resistance in commensal E. coli from poultry Concentration AMP FOT TAZ MERO NAL CIP TET TGC SMX TMP CHL GEN COL AZI <= <= <= <= <= <= <= <= <= <= <= N NR ND %R 72,8 8,2 7,6 0,0 63,3 69,6 45,6 0,0 58,2 49,4 20,9 5,7 0,0 ND CI 65,1-80 4, , , , , , ,8-28 2, ND Line: breakpoint, N: Number, NR: Number resistant, %R percent resistant, CI: confidence interval, ND: not determined

7 Table 3. Antibiotic resistance in commensal E. coli from pigs Concentration AMP FOT TAZ MERO NAL CIP TET TGC SMX TMP CHL GEN COL AZI <= <= <= <= <= <= <= <= <= <= <= N NR ND %R 41,3 0,5 0,5 0,0 1,1 2,2 44,0 0,0 52,2 50,0 28,8 2,2 0,5 ND CI 34, ,1-4 0,6-5 34, , , ,4-36 0, ND Line: breakpoint, N: Number, NR: Number resistant, %R percent resistant, CI: confidence interval, ND: not determined

8 Table 4. Antibiotic resistance in commensal E. coli from veal calves Concentration AMP FOT TAZ MERO NAL CIP TET TGC SMX TMP CHL GEN COL AZI <= <= <= <= <= <= <= <= <= <= <= N NR ND %R 55,3 0,5 0,5 0,0 20,7 22,3 68,6 0,0 60,0 51,6 25,5 5,9 2,7 ND CI 47, , , , , , , ,9-6 ND Line: breakpoint, N: Number, NR: Number resistant, %R percent resistant, CI: confidence interval, ND: not determined

9 Table 5. Antibiotic resistance in commensal E. coli from bovines for meat production (<7 months) Concentration AMP FOT TAZ MERO NAL CIP TET TGC SMX TMP CHL GEN COL AZI <= <= <= <= <= <= <= <= <= <= <= N NR ND %R 20,1 2,4 2,4 0,0 7,3 8,5 17,7 0,0 23,2 15,2 15,9 4,9 0,6 ND CI 14,3-27 0,7-6 0, ,8-12 4, , , , ,6-22 2, ND Line: breakpoint, N: Number, NR: Number resistant, %R percent resistant, CI: confidence interval, ND: not determined

10 Figure 1. Percentage of strains isolated from poultry with multi-resistance.

11 Figure 2. Cumulative percentage of strains isolated from poultry with multi-resistance.

12 Figure 3. Percentage of strains isolated from pigs with multi-resistance.

13 Figure 4. Cumulative percentage of strains isolated from pigs with multi-resistance.

14 Figure 5. Percentage of strains isolated from veal calves with multi-resistance.

15 Figure 6. Cumulative percentage of strains isolated from veal calves with multi-resistance.

16 Figure 7. Percentage of strains from bovines for meat production (<7 months) with multi-resistance.

17 Figure 8. Cumulative percentage of strains from bovines for meat production (<7 months) with multi-resistance.