Nz vet journal october 2010 no covers.indd

New Zealand Veterinary Journal 58(5), 229-236, 2010 Scientifi c Article
Low levels of antibacterial drug resistance expressed by Gram-negative
bacteria isolated from poultry carcasses in New Zealand
EJ Pleydell*§, L Rogers*, E Kwan* and NP French*
antibacterial drugs. The use of cephalothin as a marker of resist-
Abstract
ance to fi rst-generation cephalosporins may not be appropriate
for non-type-specifi c E. coli of animal origin.
AIM: To provide baseline data on the levels and patterns of an-
tibacterial drug resistance expressed by Gram-negative bacteria
KEY WORDS: Antibacterial resistance, antibiotic resistance,
isolated from poultry carcasses in New Zealand.
antimicrobial resistance, Campylobacter jejuni, carcass rinse,
cephalothin, chicken, disc diffusion, Escherichia coli, poultry,
METHODS: Between July and December 2006, isolates of
Salmonella
Escherichia coli (n=407) and Salmonella spp. (n=3) originating
from carcass-rinse samples were submitted by testing laborato-
ries affi liated to fi ve major poultry processing plants. Isolates
of Campylobacter jejuni (n=193) originating from retail poultry
Introduction
carcasses in 2005–2006 were retrieved from the Massey Uni-
versity archives. All isolates underwent disc diffusion suscepti-
In 2005, an expert panel was convened by the New Zealand Food bility testing against panels of 12 (Enterobacteriaceae) and six
Safety Authority (NZFSA) to review the impact of the use of an- (Campylobacter spp.) antibacterial drugs. Cephalothin-resist-
tibacterial drugs in agricultural settings on the emergence of anti- ance in isolates of E. coli was confi rmed using ETest strips, and
bacterial drug-resistant human pathogens. One of the conclusions confi rmation of the resistance phenotypes for a subset of C. je-
of the panel was that the development of accurate within-country juni isolates used microbroth dilution assays. Patterns within
risk assessments was being impeded by a lack of data regarding the resistance phenotypes of the isolates were investigated using
resistance in animal-associated bacteria in New Zealand (Anony- hierarchical clustering, and logistic regression modelling.
RESULTS: The majority of isolates (71.5% E. coli, 99% C. je-
The rate of notifi cations of campylobacteriosis in New Zealand juni, and all three Salmonella spp. isolates) were fully suscep-
had been on an upward trajectory since the early 1980s. By 2006, tible to the drugs that were tested. Four (1%) E. coli isolates
the annual rate of notifi cation had reached a peak of 383.5 per showed resistance to three or more drugs. The proportions of
100,000 population, the highest of any developed country in the susceptible E. coli differed between the fi ve processing plants.
world (Baker et al. 2007). Poultry meat was estimated to be ac- Resistances were detected in E. coli isolates, using disc diffusion
countable for 80% of infections (Mullner et al. 2009a) and, with to cephalothin (18.2%), ampicillin (4.4%), tetracycline (4.4%)
969 people requiring hospitalisation for the disease in 2006, high and gentamicin (1.5%). There was an association between
levels of antibacterial drug resistance in Campylobacter spp. iso- cephalothin-resistant isolates of E. coli and decreased suscep-
lates entering the food chain would be of direct public health tibility to gentamicin. Using ETests to ascertain the minimum
inhibitory concentrations (MIC) of E. coli for cephalothin gave
The incidence of salmonellosis was comparatively lower within inconsistent results. One of 193 C. jejuni isolates was resistant
the country, with an average of 46.1 notifi cations per 100,000 to erythromycin, and microbroth dilution assays confi rmed that
population per year between 1995 and 2001 (Thornley et al. this panel of C. jejuni was generally susceptible to antibacterial
2003). However, this fi gure is still relatively high for a developed country, and source-attribution modelling of data from 2003 es- CONCLUSIONS: The levels of resistance shown by Gram-neg-
timated that 20% of human cases may have originated from poul- ative bacteria isolated from chicken carcasses in New Zealand
are among the lowest reported around the world. No resist-
Many antibacterial drug-resistance mechanisms are encoded by ance to extended-spectrum cephalosporin drugs was detected in
genes carried on extrachromosomal DNA such as plasmids and E. coli, suggesting that CTX-M and AmpC beta-lactamases are
transposons, and horizontal gene transfer between bacteria of the rare or absent. Salmonella spp. are rarely isolated from poultry
same and different species can act to spread these genes among carcasses during routine testing in New Zealand, and the iso-
bacterial populations (Salyers and Amabile-Cuevas 1997; Sunde lates identifi ed during this study were fully susceptible to the
and Sorum 2001). The conditions present within the intestinal drugs tested. A panel of C. jejuni isolates originating from retail
poultry carcasses were susceptible to fi rst-line and second-line
Clinical and Laboratory Standards Institute * Institute of Veterinary, Animal and Biomedical Sciences, Massey University, Private Bag 11222, Palmerston North 4442, New Zealand.
§ Author for correspondence. Email: e.j.pleydell@massey.ac.nz New Zealand Veterinary Journal 58(5), 2010 tracts of mammals and birds are highly conducive for horizontal selected from each plate, and confi rmation of identity for E. coli transfer of genes (Blake et al. 2003), and enteric E. coli have been was undertaken using the indole test. The selected isolates were found to be adept at carrying and transferring plasmid-borne re- subcultured onto non-selective media, after which the purifi ed sistance genes (Sunde and Sorum 2001). For these reasons, ge- isolates were suspended in 2 ml 15% glycerol broth, and stored at neric E. coli are commonly included within resistance surveillance –80°C. When required for susceptibility testing, resuscitation was programmes both as markers of resistance within livestock popu- accomplished by spreading a small amount of the frozen glycerol lations, and also for their role as potential donors of resistance suspension onto a horse blood agar plate (Fort Richard Laborato- genes to more pathogenic bacteria (Hammerum et al. 2007; de ries Ltd, Auckland, NZ) prior to incubation at 37°C.
Susceptibility testing of the Enterobacteriaceae isolates was car- In light of the prominence of foodborne zoonoses in New Zea- ried out for 12 antibacterial drugs, using disc diffusion assays, land, and the recommendations of the NZFSA’s expert panel, a in accordance with the guidelines published by the Clinical and study was undertaken to collect baseline data on the occurrence Laboratory Standards Institute (CLSI), Baltimore, United States of antibacterial drug-resistant bacteria within the broiler indus- of America (Anonymous 2002). The details of the antibacterial try in New Zealand. The aim of this study was to assess poultry- drugs tested and the concentrations of discs used (Oxoid, Auck- associated antibacterial drug resistance in two zoonotic bacteria land, NZ) are shown in Table 1; the control strain used was E. coli of direct impact on human health, viz Campylobacter jejuni and Salmonella enterica, as well as non-type-specifi c E. coli.
Forty-four E. coli isolates showed unusual results to the 30-µg cephalothin disc; large isolated colonies grew within an obvious-ly demarcated clear zone. In accordance with the guidelines of Materials and methods
the CLSI (Anonymous 2002), colonies growing within the clear zones were subcultured onto plain media, their identities were Sampling strategies
confi rmed using biochemical test strips (API 20E; bioMérieux Panels of E. coli and Salmonella spp. isolates were assembled be- Inc, Marcy l’Étoile, France), and they were tested again using disc tween 17 July and 11 December 2006 from carcass-rinse samples diffusion. Having confi rmed that the colonies within the main collected in poultry processing plants in New Zealand. These sam- zone were E. coli and upon obtaining similar disc diffusion results ples were routinely collected using a randomised sampling pro- on the second run, the MIC of a randomly selected subset of 27 tocol as part of the National Microbiological Database (NMD) of these non-conforming isolates were ascertained using cepha- surveillance programme administered by the NZFSA (Anony- lothin ETest strips (AB Biodisk, Solna, Sweden). The ETests were mous 2008a). The aim of this study was to collect approximately run in duplicate for each isolate tested.
400 E. coli isolates across the fi ve largest processing plants in New The C. jejuni isolates were tested against six antibacterial drugs, Zealand, which between them accounted for approximately 90% using disc diffusion tests on Mueller-Hinton agar supplemented of the domestic broiler meat market. The sample size of 400 was with 5% defi brinated sheep blood (Fort Richard Laboratories chosen to provide 95% confi dence limits of ± 5% around an esti- Ltd). The antibacterial drugs tested are shown in Table 2; the con- mated prevalence of 50%, as well as acceptable precision around trol strain used was C. jejuni ATCC 35360. The inoculated plates estimates of lower and higher prevalence. The number of samples were incubated at 42°C for 48 h in a micro-aerobic atmosphere, that was requested from each processing plant was proportional to and the control strain was included with every batch of isolates that plant’s share of the domestic market.
Salmonella spp. are only isolated at low prevalence from all live- There are a number of reports of good correlation between disc stock species within the NMD programme, therefore all Salmon- diffusion and agar dilution assays for Campylobacter spp. (Gaud- ella spp. isolated from broiler carcasses at the NMD laboratories reau and Gilbert 1997; Luangtongkum et al. 2007; Gaudreau et during the sampling period were requested for inclusion within al. 2008), but due to diffi culties in standardising the interpreta- tion of results at the time of this study there was no standard The panel of C. jejuni isolates was assembled from the isolate ar- CLSI protocol for disc diffusion methods for Campylobacter spp. chive at Massey University, where they had been stored at –80°C. isolates (Fritsche et al. 2007). Full validation of the disc diffusion The archive constituted Campylobacter spp. that had been iso- results in this study was not possible within the project’s budget, lated from retail chickens purchased in the Manawatu region on but the results of a subset of 21 isolates selected randomly, as well a monthly basis since March 2005; isolates of C. jejuni had been as the isolate identifi ed as resistant to erythromycin, were re-test- identifi ed using PCR. The sampling strategy and confi rmatory ed in triplicate using the EUCAMP microbroth dilution assay testing related to that work has been published in more detail by (Trek Diagnostic Systems, East Grinstead, England).
Mullner et al. (2009b). For this study, using the criteria of only Statistical analysis
testing a single isolate from each chicken purchased, a panel of Hierarchical clustering of the resistance phenotypes of the 407 193 C. jejuni isolates was available for testing. The 95% confi - E. coli isolates was undertaken by designating isolates as re- dence limits around an estimated prevalence of 50% using a sam- sistant or susceptible according to CLSI breakpoints. Isolates of intermed iate susceptibility were grouped with the fully susceptible Laboratory methods
ones. The distance matrix was calculated using a simple matching The Enterobacteriaceae were submitted to the laboratory at Mas- algorithm, and the clustering algorithm used was the unweighted sey University on agar slopes or Petrifi lm plates (Thermo Fisher pair group method with arithmetic mean. These methods were Scientifi c, Auckland, NZ) by the NMD laboratories designated to chosen as they produced the clearest visual summary of the phe- each processing plant. A single isolate of typical morphology was New Zealand Veterinary Journal 58(5), 2010 Logistic regression modelling was used to estimate the OR of se- A dendrogram was produced using BioNumerics 5.1 software lecting a fully susceptible E. coli from each of the participating (Applied Maths NV, Sint Martens-Latem, Belgium). The multi- poultry processing plants in comparison with the plant with the nomial and logistic regression models were fi tted in the R statisti- highest percentage of fully susceptible E. coli.
cal environment v2.9.2 (R Foundation for Statistical Computing, Vienna, Austria).
Due to the apparent discordance between the E. coli disc diffu-sion results for ampicillin and cephalothin, unordered multinom-ial logistic regression models (Hosmer and Lemeshow 2000) were fi tted to ascertain whether cephalothin resistance, identifi ed using disc diffusion, was associated with consistent shifts in measure-ments of the size of the zones for other drugs. That is, to determine Enterobacteriaceae
whether these cephalothin-resistant E. coli showed similarities in The disc diffusion test results for the panel of 407 E. coli are dis- terms of their susceptibility to other drugs that demarcated them played in Table 1, with the CLSI breakpoints for interpreting the from the cephalothin-susceptible isolates. Each E. coli isolate was sizes of the zones superimposed upon the table in varying shades allocated to one of four groups, to form a discrete, nominally- of grey. The dendrogram in Figure 1 displays the resistance phen- scaled outcome variable corresponding to: cephalothin-suscep- otypes that were present within the E. coli panel, together with tible/ampicillin-susceptible, cephalothin-susceptible/ampicillin- the number of isolates that showed each phenotype, and the pro- resistant, cephalothin-resistant/ampicillin-susceptible, or cepha- cessing plant(s) of origin. The most commonly occurring phen- lothin-resistant/ampicillin-resistant. The reference category was otype was fully susceptible to all drugs tested, with 291 (71.5%) cephalothin-susceptible/ampicillin-susceptible, and four covari- isolates falling into this category. Sixteen resistance phenotypes ates were simultaneously fi tted that corresponded to the measure- were elucidated among the 116 resistant isolates, with 95/407 ments of the size of the mean-centred zone for tetracycline, strep- (23.3%) isolates showing resistance to one drug, 17 (4.2%) to tomycin, gentamicin and furazolidone. Drugs to which resistance two drugs, three (0.7%) to three drugs, and one (0.2%) showed was shown by only one isolate were excluded from the model.
resistance to fi ve of the drugs on the panel, which included all Table 1. Number of Escherichia coli isolates originating from fi ve poultry processing plants in New Zealand, that displayed various-sized diameters of
growth-free zones using disc diffusion for each of 12 antibacterial drugs. The resistance categories according to Clinical and Laboratory Standards
Institute (CLSI) breakpoint concentrations are superimposed upon the table in various shades of grey, with the exception of furazolidone, for which
no CLSI breakpoints were available.
Zone size (mm)
Disca (µg) Res (%)b 95% CIc
a Concentration of drug within the discb Percentage of isolates with zone sizes within the resistant category for that drugc 95% CI for the prevalence of resistance to that drug Table 2. Number of Campylobacter jejuni isolates originating from retail chicken meat, that displayed various-sized diameters of growth-free zones
using disc diffusion for each of six antibacterial drugs. A diameter of ≤6 mm of a zone (i.e. growth extending to the edge of the disc) was interpreted
as defi nitive resistance to that drug: this zone is highlighted in grey.
Zone size (mm)
Drug Disca (µg)
17 18 19 20 21 22 23 24 25 26 27 ≥28
a Concentration of drug within the discb Percentage of isolates with zone sizes within the resistant category for that drug New Zealand Veterinary Journal 58(5), 2010 Resistance phenotype Processing plant of origin (no. isolates)
Figure 1. A dendrogram showing the hierarchical clustering of the resistance phenotypes of 407 Escherichia coli isolates originating from fi ve
poultry processing plants (A to E) in New Zealand. The clustering algorithm used was the unweighted pair group method with arithmetic mean. AS
= susceptible to all drugs tested; Fur = furazolidone; Cep = cephalothin; Gen = gentamicin; Str = streptomycin; Tet = tetracycline; Amp = ampicillin;
Sul = sulfasoxazole; Kan = kanamycin; Fox = cefoxitin; Tax = cefotaxime; Chl = chloramphenicol; Cip = ciprofl oxacin
three aminoglycoside drugs tested (gentamicin, kanamycin and Of the 74 cephalothin-resistant E. coli detected using disc diffu- streptomycin) in addition to sulfasoxazole and tetracycline.
sion, 6/407 (1.5%) grew right up to the edge of the disc (zone size ≤6 mm), while the remainder had zone sizes of between 10 and Using the CLSI guidelines for interpretation of measurements of the size of zones, 74/407 (18.2%) E. coli isolates were designated 14 (median 13) mm. However, scattered individual colonies grew as resistant to cephalothin, and these isolates originated from all within a mainly clear zone around the cephalothin disc for 44/74 fi ve of the participating processing plants (see Figure 1). Con- (59%) purportedly cephalothin-resistant isolates. Repeating the current cephalothin-ampicillin resistance was seen in just seven tests using the original stored isolates returned the same result, isolates, that had originated from Plants C, D and E. Ampicillin and the results of API 20E tests confi rmed that the scattered re- resistance was expressed by 18 (4.4%) isolates in total, and these sistant colonies were E. coli. Disc diffusion tests of purifi ed cul- originated from all fi ve plants. There were also 18 tetracycline- tures of colonies within the main clear zone also produced clearly resistant isolates, originating from all plants except Plant A. Re- demarcated zones containing individual scattered colonies, im- sistance to gentamicin was seen in six (1.5%) isolates; four mo- plying that this result was not due to a mix of two different strains no-resistant isolates from Plant C, and two multidrug-resistant within the original stored cultures. The ETest results determined isolates from Plant D, where multidrug resistance refers to an that only three of a subset of 21 of these non-conforming isolates isolate expressing resistance to three or more drugs. None of the showed a conclusive MIC ≥32 µg/ml in duplicate tests. However, E. coli isolates was resistant to ciprofl oxacin, cefoxitin, cefotaxime scattered colonies within a main zone of no growth were also seen, using ETest for a further six isolates. Two of the jointly cephaloth-in- and ampicillin-resistant isolates were also retested using ETest, The percentage of E. coli isolates that were susceptible to all 12 and the MIC of cephalothin for both isolates were ≥32 µg/ml.
antibacterial drugs tested varied between the individual process-ing plants, with the range extending from 50 to 84.7%. Logistic Multinomial modelling highlighted a number of associations regression modelling showed that the odds of picking an isolate between the cephalothin and ampicillin resistance status of the susceptible to all 12 drugs from a chicken carcass were signifi cant- E. coli isolates and measurements of the size of the mean-centred ly lower for Plants C, D and E compared with Plant B (Table 3).
zone obtained for other drugs (Table 4). An increase in size of Table 3. Results of a logistic regression model of the probability of picking an Escherichia coli isolate that was susceptible to all 12 drugs tested from
samples submitted from each of fi ve participating poultry processing plants (A–E).
Plant Sus
(%)a Log-odds
(β) SEb Z-statistic
P-valuec OR
[Exp(β)] 95%
a Percentage of isolates that were susceptible to all 12 drugsb SE of the log-oddsc Signifi cance of the difference in odds compared with Plant Bd 95% CI for the estimated OR New Zealand Veterinary Journal 58(5), 2010 the zone for tetracycline of 1 mm (increasing susceptibility) was of the erythromycin disc was ≥32 µg/ml, therefore confi rming associated with a decrease in the odds of an isolate being am- the resistance of this isolate. Based on microbroth dilution, 2/22 picillin-resistant but cephalothin-susceptible compared with the isolates demonstrated intermediate susceptibility to erythromy- cephalothin-susceptible and ampicillin-susceptible reference cat- cin. The sizes of the disc diffusion zones for these two isolates (18 egory, i.e. ampicillin-resistant isolates also had smaller zones of no and 21 mm) were smaller than the mean size of the zone of the growth (decreased susceptibility) to tetracycline. In contrast, the isolates, with MIC that corresponded to susceptibility to erythro- isolates that were designated as cephalothin-resistant (both those mycin (29.3 mm; 95% CI=27.9–30.6). In addition, 1/22 isolates that were susceptible to ampicillin and resistant to ampicillin) tested was also found to show resistance to streptomycin, with an were more likely to show reduced susceptibility to gentamicin, MIC of >16 µg/ml; however, this drug had not been included in A total of three Salmonella spp. were submitted during the sam-pling period. All three originated from a single processing plant, Plant C, and each isolate was fully susceptible to all of the drugs Discussion
Comparing the results of this study with data from other coun- Campylobacter jejuni
tries revealed that levels of resistance in Gram-negative bacteria For all drugs except nalidixic acid, >94% of isolates had zones recovered from broiler chickens at slaughter in New Zealand are with a diameter ≥24 mm, indicating that the isolates were suscep- among the lowest reported in the world. Only 4.4 % of the E. coli tible to these six drugs (Table 2). The median size of the zone for isolates in this study were resistant to ampicillin, with an equal nalidixic acid was 24 (range 16–34) mm. A single isolate showed number (although mainly different isolates) showing resistance to total resistance to 15 µg erythromycin, and growth extended to tetracycline. These two drugs have been used in human and vet- erinary medicine for many years, and in many parts of the world Ascertaining the MIC for a subset of 22 C. jejuni isolates, using indicator bacteria carrying resistance to them are commonly iso- microbroth dilution, confi rmed that these isolates were suscep- lated from both diseased and healthy people and animals, includ- tible to nalidixic acid, ciprofl oxacin, tetracycline and chloram- ing free-living wildlife species (Livermore et al. 2001; Dominguez phenicol (Table 5). The MIC of the isolate that grew to the edges et al. 2002; Nascimento et al. 2003; Nys et al. 2004; Dai et al. 2008; Hendriksen et al. 2008).
Table 4. The estimated coeffi cients (β), SE and OR [Exp(β)] obtained
With respect to E. coli isolated specifi cally from poultry, data col- from an unordered multinomial logistic regression model of cephalothin
lected from fi ve central European countries in 2002 and 2003 and ampicillin resistance status (as determined by disc diffusion) of
demonstrated between-country differences in the proportion of 407 Escherichia coli isolates against the size of the mean-centred disc
diffusion zone of four other antibacterial drugs (measured in mm). For
E. coli collected from chickens at slaughter that were resistant to each cephalothin:ampicillin category the parameters with log-odds
ampicillin, with a range of 38.6–71.1%, and tetracycline resist- that are greater than twice the SE and with OR with 95% CI that do
ance ranged from 49 to 81.2% of isolates (de Jong et al. 2009). not include 1 were considered to be signifi cantly different from the
reference category, and are highlighted in bold.
National surveillance data also allow for comparisons; fi gures from Canada in 2005 showed similar magnitudes to those from Cephalothin:ampicillin
the study in central Europe (Anonymous 2007ab), whilst data resistance status
Log-odds (β) SE OR
[Exp(β)] 95%
from the Scandinavian countries were more similar to the data from New Zealand, with ampicillin resistance ranging between 4 and 17.1%, and tetracycline 3.7 and 6.5% (Anonymous 2005b, Tetracycline –0.12
0.83–0.94
Although resistance was generally low in E. coli, signifi cant differ- ences were seen in the percentages of E. coli, from the fi ve partici- pating processing plants, that were susceptible to all drugs tested (Table 2). There was a lower probability of obtaining a susceptible Table 5. A comparison of the results of two susceptibility tests for
22 Campylobacter jejuni isolates, using microbroth dilution to obtain
minimum inhibitory concentrations (MIC), and sizes of zones, using
Gentamicin
0.69–0.87
disc diffusion.
Furazolidone
0.88–0.98
MIC-based
Mean zone
resistance
Drug category
isolates
Gentamicin
0.54–0.79
CSAS = cephalothin-susceptible/ampicillin-susceptible (n=322); CSAR = cephalothin-susceptible/ampicillin-resistant (n=11); CRAS = cephalothin- resistant/ampicillin-susceptible (n=67); CRAR = cephalothin-resistant/ampicillin- New Zealand Veterinary Journal 58(5), 2010 isolate from three plants (C, D and E) compared with the plant strains producing acylesterases are able to grow in broths contain- (B) with the highest percentage of fully susceptible isolates. All ing cephalothin in concentrations of up to four times the MIC the farms supplying an individual plant belonged to the same (Nishiura et al. 1978). Nonetheless, regardless of the actual mech- poultry company and therefore were using the same husbandry anism behind these observations, it would appear that cephaloth- protocols. However, there were differences between the poultry in may not be a suitable marker of resistance to fi rst-generation companies in the numbers of different growth-promoting anti- cephalosporins in generic E. coli of animal origin.
bacterial drugs that had been used on the company farms during In this study, the Enterobacteriaceae isolates were obtained from the 6 years preceding the study. Zinc bacitracin had been used carcass-rinse samples that had been collected after the carcasses exclusively on the farms supplying Plants A and B, however the had passed through an immersion chiller. This type of sample farms supplying Plants C, D and E had also used two other drugs was chosen for reasons of economics and logistics as these samples (tylosin and avilamycin) as routine, in-feed prophylactics at vari- were already being collected routinely for the NZFSA’s NMD. ous times over the same period (Pleydell et al. 2010). Therefore, However, many of the surveillance programmes in other coun- it is possible that these company-level differences in the use of tries utilise caecal contents or cloacal swabs taken from broiler antibacterial drugs in feed could have been the cause of the differ- chickens at slaughter. Therefore, there could be potential prob- ences seen between plants in the patterns of resistance in E. coli. lems with the direct comparison of results obtained from these Similar differences between companies have also been reported two types of sample. However, E. coli isolated from retail chicken in the resistance patterns of Enterococcus faecium isolated from meat samples in Scandinavia showed similar levels of resistance as chicken carcasses at those plants (Pleydell et al. 2010).
caecal samples from slaughtered birds within the same country, None of the E. coli isolates in this study was resistant to cefo- implying that both techniques were sampling the same popula- taxime or cefoxitin. Sampling from the probability mass func- tions of E. coli (Anonymous 2005b, 2006ab, 2007c).
tion of the binomial distribution, it can be estimated that if With respect to Salmonella spp., this work demonstrated that they the true prevalence of cefotaxime or cefoxitin resistance within are rarely isolated from freshly dressed broilers in New Zealand E. coli isolated from poultry carcasses was either 1% or 0.1%, then the probabilities of 0/407 isolates showing resistance to that within the NMD programme, and that the three isolates that drug would be 0.02 or 0.67, respectively. Therefore, although no were tested did not express antibacterial drug resistance.
phen otypic evidence of CTX-M or AmpC extended-spectrum Disc diffusion methods demonstrated that 192/193 C. jejuni iso- beta-lactamases among E. coli of poultry origin in New Zealand lates were fully susceptible to the drugs tested, with the exception was found in this study, this sample size could not exclude the being a single isolate that grew to the edge of the erythromycin possibility that CTX-M or AmpC could have been present at very disc (Table 2). The use of microbroth dilution assays for a subset low levels. This situation is worthy of future monitoring, due to of isolates confi rmed that isolates with wide zone sizes were sus- both the recent detection of CTX-M-15 enzymes in E. coli caus- ceptible to the drugs on the panel, and that one isolate (0.5% of ing urinary tract infections in humans in New Zealand (Moor et the total tested) was resistant to erythromycin. In comparison, al. 2008), and because other countries are starting to report the 11% of C. jejuni isolated from retail poultry meat reared in Den- presence of genes encoding for the CTX-M family within E. coli mark in 2007 were resistant to nalixidic acid and ciprofl oxacin isolated from poultry (Dai et al. 2008; Smet et al. 2008; Costa et (Anonymous 2008b). Elsewhere in the world, levels of resistance in Campylobacter spp. can be even higher, e.g. in fi ve central Eu- Despite a low level of ampicillin resistance and no evidence of ropean countries, 10.6–83.3% of C. jejuni isolated from slaugh- the presence of extended-spectrum beta-lactamases, evidence tered poultry were resistant to ciprofl oxacin, and 23.5–58.3% to of cephalothin resistance was seen for 18.2% of E. coli isolates, tetracycline, but no resistance was seen to erythromycin (de Jong and these purportedly cephalothin-resistant isolates also showed et al. 2009). Furthermore, in a study in Korea, 87.9% of 594 iso- decreased susceptibility to gentamicin. This level of cephalothin lates of Campylobacter spp. isolated from raw retail chicken meat resistance was surprising, because beta-lactamase enzymes, with a were resistant to ciprofl oxacin, 87.2% to tetracycline, and 19.4% spectrum of activity including cephalothin, also degrade ampicil- to erythromycin (Kang et al. 2006).
lin (Livermore 1995; Li et al. 2007). However, these observations When interpreting the results for Campylobacter spp. in the study are not unprecedented, as another disc diffusion study of E. coli presented here, it should be borne in mind that the C. jejuni iso- isolated from human sewage, animals and the environment found lates originated from chickens bought within the Manawatu re- that 20–30% of isolates originating from animal faeces and farm gion of the lower North Island. However, surveillance work com- environments were resistant to cephalothin, with just 4–6% of missioned by the NZFSA has shown that Campylobacter spp. iso- the same isolates showing resistance to ampicillin (Sayah et al. lated from retail chickens purchased in the Manawatu are broadly representative of those isolated from chickens in other regions of Possible explanations for the appearance of scattered resistant the North Island (Anonymous 2008c). Furthermore, the poultry colonies on a plate of pure growth of a single strain could in- companies operating in the South Island are the same as those clude the following. Hypermutator strains due to defects within in the North Island. Studies of enterococci, as well as the E. coli the methyl-directed mismatch repair system (LeClerc et al. 1996). results shown here, suggested that the company of origin had a greater effect upon the resistance phenotype than did the region of Epigenetic-phase regulation, whereby the expression of a gene is origin (Pleydell et al. 2010), so it is not anticipated that the resist- switched on and off under differing environmental conditions ance phenotypes of C. jejuni isolated in the South Island would be (Holden et al. 2007; Deitsch et al. 2009; Srikhanta et al. 2009). substantially different from those isolated in the Manawatu.
Or, the production of acylesterase enzymes, i.e. narrow-spectrum cephalosporinases which act on the sidechain of the cephalospor- It should also be noted that prior to susceptibility testing all iso- in ring (Nishida et al. 1968). With increasing incubation periods, lates were resuscitated from storage at –80°C. It is hypothetically New Zealand Veterinary Journal 58(5), 2010 possible that storage could have affected the resistances expressed *Anonymous. DANMAP 2006: Use of Antimicrobial Agents and Occurrence of
by the isolates, however we are not aware of any published evi- Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in dence that careful storage at –80°C and subsequent resuscitation Denmark. Statens Serum Institut, Danish Veterinary and Food Administration, causes alterations in the resistances expressed by those isolates.
Danish Medicines Agency, and Danish Institute for Food and Veterinary Research, Copenhagen Denmark, 2007 c In conclusion, this study provided baseline data on antibacterial *Anonymous. Schedule 1 National Microbiological Database Programme. New
drug resistance in Gram-negative bacteria present on chicken car- Zealand Food Safety Authority, Wellington, NZ, 2008 a casses in poultry-processing plants in New Zealand. It has shown *Anonymous. DANMAP 2007: Use of Antimicrobial Agents and Occurrence of
that the majority of E. coli and C. jejuni isolates were fully suscep- Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in tible to fi rst-line and second-line antibacterial drugs. However, Denmark. Statens Serum Institut, Danish Veterinary and Food Administration, there were between-plant differences, with increased levels of re- Danish Medicines Agency, and Danish Institute for Food and Veterinary sistance detected in E. coli obtained from some plants. The study also provided evidence that the use of cephalothin as a marker of *Anonymous. Enhancing Surveillance of Potentially Foodborne Enteric Diseases
resistance to fi rst-generation cephalosporins may produce incon- in New Zealand: Human Campylobacteriosis in the Manawatu. New Zealand sistent results if applied to generic E. coli of animal origin. The Food Safety Authority, Wellington, NZ, 2008 c data provided by this study have contributed to the scope and Baker M, Wilson N, Edwards R. Campylobacter infection and chicken: an update
design of a national surveillance programme of antibacterial drug on New Zealand’s largest ‘common source outbreak’. New Zealand Medical resistance in bacteria of animal origin in New Zealand, and pro- Journal 120 (1261), No. 2717, September 2007 vide a statistically valid set of results for comparison with future Blake DP, Hillman K, Fenlon DR, Low JC. Transfer of antibiotic resistance
data regarding E. coli and C. jejuni.
between commensal and pathogenic members of the Enterobacteriaceae under ileal conditions. Journal of Applied Microbiology 95, 428–36, 2003 Costa D, Vinue L, Poeta P, Coelho AC, Matos M, Saenz Y, Somalo S, Zarazaga
M, Rodrigues J, Torres C. Prevalence of extended-spectrum beta-lactamase-
Acknowledgements
producing Escherichia coli isolates in faecal samples of broilers. Veterinary Microbiology 138, 339–44, 2009 The authors would like to thank Vanessa Wintle and Natalie Dai L, Lu LM, Wu CM, Li BB, Huang SY, Wang SC, Qi YH, Shen JZ.
Chrystal of the Poultry Industry Association New Zealand for Characterization of antimicrobial resistance among Escherichia coli isolates their logistic support with collecting the samples and isolates; Da- from chickens in China between 2001 and 2006. FEMS Microbiology Letters vid Marks, Neil Christensen and the Poultry Industry Association New Zealand (PIANZ) Vet Tech Committee for sharing their Deitsch KW, Lukehart SA, Stringer JR. Common strategies for antigenic
variation by bacterial, fungal and protozoan pathogens. Nature Reviews, knowledge and making helpful comments about the manuscript; Anne Midwinter for identifying and retrieving the C. jejuni iso- de Jong A, Bywater R, Butty P, Deroover E, Godinho K, Klein U, Marion
lates from storage; Kathryn Goodwin for assisting with data en- H, Simjee S, Smets K, Thomas V, Valle M, Wheadon A. A pan-European
try; Simon Spencer for statistical support; and Peter Cripps and survey of antimicrobial susceptibility towards human-use antimicrobial drugs Tom Besser for reading and commenting on the manuscript. This among zoonotic and commensal enteric bacteria isolated from healthy food- study was supported and funded by the PIANZ.
producing animals. Journal of Antimicrobial Chemotherapy 63, 733–44, 2009 Dominguez E, Zarazaga M, Saenz Y, Brinas L, Torres C. Mechanisms of
antibiotic resistance in Escherichia coli isolates obtained from healthy children in Spain. Microbial Drug Resistance 8, 321–7, 2002 References
Fritsche TR, McDermott PF, Shryock TR, Walker RD, Morishita TY. Agar
dilution and disk diffusion susceptibility testing of Campylobacter spp. Journal *Anonymous. Performance Standards for Antimicrobial Disk and Dilution
of Clinical Microbiology 45, 2758–9, 2007 Susceptibility Tests for Bacteria Isolated from Animals; Approved Standard. 2nd Gaudreau C, Gilbert H. Comparison of disc diffusion and agar dilution methods
Edtn. CLSI Document M31-A2. Clinical and Laboratory Standards Institute, for antibiotic susceptibility testing of Campylobacter jejuni subsp. jejuni and Campylobacter coli. Journal of Antimicrobial Chemotherapy 39, 707–12, 1997 *Anonymous. A Review of the Impact of the Use of Antimicrobials in Animals and
Gaudreau C, Girouard Y, Gilbert H, Gagnon J, Bekal S. Comparison of disk
Plants on the Development of Antimicrobial Resistance in Human Bacterial diffusion and agar dilution methods for erythromycin, ciprofl oxacin, and Pathogens. New Zealand Food Safety Authority, Wellington, NZ, 2005 a tetracycline susceptibility testing of Campylobacter coli and for tetracycline *Anonymous. SVARM 2004: Swedish Veterinary Antimicrobial Resistance
susceptibility testing of Campylobacter jejuni subsp. jejuni. Antimicrobial Monitoring. National Veterinary Institute (SVA), Uppsala, Sweden, 2005 b Agents and Chemotherapy 52, 4475–7, 2008 *Anonymous. DANMAP 2005: Use of Antimicrobial Agents and Occurrence of
Hammerum AM, Heuer OE, Emborg HD, Bagger-Skjot L, Jensen VF, Rogues
Antimicrobial Resistance in Bacteria from Food Animals, Foods and Humans in AM, et al. Danish integrated antimicrobial resistance monitoring and research
Denmark. Statens Serum Institut, Danish Veterinary and Food Administration, program. Emerging Infectious Diseases 13, 1632–9, 2007 Danish Medicines Agency, and Danish Institute for Food and Veterinary Hendriksen RS, Mevius DJ, Schroeter A, Teale C, Jouy E, Butaye P, et al.
Occurrence of antimicrobial resistance among bacterial pathogens and indicator *Anonymous. NORM/NORM-VET 2006: Usage of Antimicrobial Agents and
bacteria in pigs in different European countries from year 2002–2004: the Occurrence of Antimicrobial Resistance in Norway. National Veterinary ARBAO-II study. Acta Veterinaria Scandinavica 50, doi:10.1186/1751-0147- Insititute, and Norwegian Insititute of Public Health, Oslo, Norway, 2006 b *Anonymous. CIPARS 2005: Canadian Integrated Program for Antimicrobial
Holden N, Totsika M, Dixon L, Catherwood K, Gally DL. Regulation of
Resistance Surveillance. Public Health Agency of Canada, Guelph ON, P-fi mbrial phase variation frequencies in Escherichia coli CFT073. Infection *Anonymous. CIPARS PR 2006: Canadian Integrated Program for Antimicrobial
Resistance Surveillance: Preliminary Results 2006. Public Health Agency of Canada, Guelph ON, Canada, 2007 b New Zealand Veterinary Journal 58(5), 2010 *Hosmer DW, Lemeshow S. Special topics. Applied Logistic Regression. 2nd Edtn.
Nishiura T, Kawada Y, Shiomi Y, O’Hara K, Kono M. Microbial degradation of
cephalothin by cephalothin-susceptible Escherichia coli. Antimicrobial Agents Kang YS, Cho YS, Yoon SK, Yu MA, Kim CM, Lee JO, Pyun YR. Prevalence
and Chemotherapy 13, 1036–9, 1978 and antimicrobial resistance of Campylobacter jejuni and Campylobacter coli Nys S, Okeke IN, Kariuki S, Dinant GJ, Driessen C, Stobberingh EE.
isolated from raw chicken meat and human stools in Korea. Journal of Food Antibiotic resistance of faecal Escherichia coli from healthy volunteers from eight developing countries. Journal of Antimicrobial Chemotherapy 54, 952–5, LeClerc JE, Li BG, Payne WL, Cebula TA. High mutation frequencies among
Escherichia coli and Salmonella pathogens. Science 274, 1208–11, 1996 Pleydell E, Rogers L, Kwan E, French N. Evidence for the clustering of
Li XZ, Mehrotra M, Ghimire S, Adewoye L. Beta-lactam resistance and beta-
antibacterial resistance phenotypes of Enterococci within integrated poultry lactamases in bacteria of animal origin. Veterinary Microbiology 121, 197–214, companies. Microbial Ecology 59, 678–88, 2010 Salyers AA, Amabile-Cuevas CF. Why are antibiotic resistance genes so resistant
Livermore DM. Beta-lactamases in laboratory and clinical resistance. Clinical
to elimination? Antimicrobial Agents and Chemotherapy 41, 2321–5, 1997 Microbiological Reviews 8, 557–84, 1995 Sayah RS, Kaneene JB, Johnson Y, Miller R. Patterns of antimicrobial resistance
Livermore DM, Warner M, Hall LM, Enne VI, Projan SJ, Dunman PM,
observed in Escherichia coli isolates obtained from domestic- and wild-animal Wooster SL, Harrison G. Antibiotic resistance in bacteria from magpies
fecal samples, human septage, and surface water. Applied and Environmental (Pica pica) and rabbits (Oryctolagus cuniculus) from west Wales. Environmental Smet A, Martel A, Persoons D, Dewulf J, Heyndrickx M, Catry B, Herman
Luangtongkum T, Morishita TY, El-Tayeb AB, Ison AJ, Zhang QJ. Comparison
L, Haesebrouck F, Butaye P. Diversity of extended-spectrum beta-lactamases
of antimicrobial susceptibility testing of Campylobacter spp. by the agar and class C beta-lactamases among cloacal Escherichia coli isolates in Belgian dilution and the agar disk diffusion methods. Journal of Clinical Microbiology broiler farms. Antimicrobial Agents and Chemotherapy 52, 1238–43, 2008 Srikhanta YN, Dowideit SJ, Edwards JL, Falsetta ML, Wu HJ, Harrison OB et
Moor CT, Roberts SA, Simmons G, Briggs S, Morris AJ, Smith J, Heffernan
al. Phasevarions mediate random switching of gene expression in pathogenic
H. Extended-spectrum beta-lactamase (ESBL)-producing enterobacteria:
Neisseria. Plos Pathogens 5, doi: 10.1371/journal.ppat.1000400, 2009 factors associated with infection in the community setting, Auckland, New Sunde M, Sorum H. Self-transmissible multidrug resistance plasmids in
Zealand. Journal of Hospital Infection 68, 355–62, 2008 Escherichia coli of the normal intestinal fl ora of healthy swine. Microbial Drug Mullner P, Jones G, Noble A, Spencer SEF, Hathaway S, French NP. Source
attribution of food-borne zoonoses in New Zealand: A modifi ed Hald model. Thornley CN, Simmons GC, Callaghan ML, Nicol CM, Baker MG,
Risk Analysis 29, 970–84, 2009 a Gilmore KS, Garrett NKG. First incursion of Salmonella enterica serotype
Mullner P, Spencer SEF, Wilson DJ, Jones G, Noble AD, Midwinter AC,
Typhimurium DT160 into New Zealand. Emerging Infectious Diseases 9, Collins-Emerson JM, Carter P, Hathaway S, French NP. Assigning the
source of human campylobacteriosis in New Zealand: A comparative genetic and epidemiological approach. Infection, Genetics, and Evolution 9, 1311–9, Nascimento AMA, Cursino L, GoncalvesDornelas H, Reis A, ChartoneSouza
E, Marini MA. Antibiotic-resistant gram-negative bacteria in birds from the
Brazilian Atlantic Forest. Condor 105, 358–61, 2003
Nishida M, Yokota Y, Okui M, Mine Y, Matsubar T. Studies on microbial
degradation of cephalosporin C deriviatives: 1. Role of beta-lactamase and acylesterase in enzymatic degradation of cephalosporins. Journal of Antibiotics 21, 165–9, 1968

Source: http://www.pianz.org.nz/pianz/wp-content/uploads/2011/06/Pleydell_LowLevelResGramNegNZVet-journal_2010.pdf