Ceftiofur Resistance in Salmonella Heidelberg | CDC EID

EID Journal Home > Volume 16, Number 1–January 2010

Volume 16, Number 1–January 2010
Research
Ceftiofur Resistance in Salmonella enterica Serovar Heidelberg from Chicken Meat and Humans, Canada
Lucie Dutil, Rebecca Irwin, Rita Finley, Lai King Ng, Brent Avery, Patrick Boerlin, Anne-Marie Bourgault, Linda Cole, Danielle Daignault, Andrea Desruisseau, Walter Demczuk, Linda Hoang, Greg B. Horsman, Johanne Ismail, Frances Jamieson, Anne Maki, Ana Pacagnella, and Dylan R. Pillai
Author affiliations: Public Health Agency of Canada, Saint-Hyacinthe, Québec, Canada (L. Dutil, D. Daignault); Public Health Agency of Canada, Guelph, Ontario, Canada (R. Irwin, R. Finley, B. Avery, L. Cole, A. Desruisseau); Public Health Agency of Canada, Winnipeg, Manitoba, Canada (L.K. Ng, W. Demczuk); Ontario Veterinary College, Guelph (P. Boerlin); Institut National de Santé Publique du Québec, Sainte-Anne-de-Bellevue, Québec, Canada (A.-M. Bourgault, J. Ismail); British Columbia Centre for Disease Control, Vancouver, British Columbia, Canada (L. Hoang, A. Pacagnella); Saskatchewan Disease Control Laboratory, Regina, Saskatchewan, Canada (G. Horsman); and Ontario Agency for Health Protection and Promotion, Toronto, Ontario, Canada (F. Jamieson, A. Maki, D.R. Pillai)

Suggested citation for this article

Abstract
The Canadian Integrated Program for Antimicrobial Resistance Surveillance describes a strong correlation (r = 0.9, p<0.0001) between ceftiofur-resistant Salmonella enterica serovar Heidelberg isolated from retail chicken and incidence of ceftiofur-resistant Salmonella serovar Heidelberg infections in humans across Canada. In Québec, changes of ceftiofur resistance in chicken Salmonella Heidelberg and Escherichia coli isolates appear related to changing levels of ceftiofur use in hatcheries during the study period, from highest to lowest levels before and after a voluntary withdrawal, to increasing levels after reintroduction of use (62% to 7% to 20%, and 34% to 6% to 19%, respectively). These events provide evidence that ceftiofur use in chickens results in extended-spectrum cephalosporin resistance in bacteria from chicken and humans. To ensure the continued effectiveness of extended-spectrum cephalosporins for treating serious infections in humans, multidisciplinary efforts are needed to scrutinize and, where appropriate, limit use of ceftiofur in chicken production in Canada.

Salmonella enterica serovar Heidelberg ranks among the top 3 serovars isolated from persons infected with Salmonella in Canada (1). It is more frequently reported in North America than in other regions of the world (2). Although many Salmonella Heidelberg infections result in mild to moderate illness, the bacterium also causes severe illness with complications such as septicemia, myocarditis, extraintestinal infections, and death (3,4). Salmonella Heidelberg appears more invasive than other gastroenteritis-causing serovars; ≈9% of isolates of this serovar received through the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) during 2003–2005 were recovered from blood samples (5). Treatment with antimicrobial agents may be life-saving in the case of invasive infections.

Sources of human Salmonella Heidelberg infection include consumption of poultry or eggs and egg-containing products (6–10). In Canada, Salmonella Heidelberg is commonly isolated from healthy chickens from farm, abattoir, and retail sources (11,12). It also has been isolated, although less frequently, from ground beef, pork, and turkey meat (13–15) and from clinical samples from various animal species (12).

Ceftiofur is an extended-spectrum cephalosporin drug approved in Canada for use with numerous label indications in cattle, swine, horses, sheep, turkeys, dogs, and cats. Ceftiofur is also injected in ovo to control Escherichia coli omphalitis in broiler chickens; this use is not an approved label indication.

A major public health concern is that use of third-generation cephalosporins, such as ceftiofur, in food animals is leading to resistance to other extended-spectrum cephalosporins, such as ceftriaxone and cephamycins (16–20), a group of antimicrobial agents used to treat a wide variety of human infections. Among other indications, ceftriaxone is the drug of choice for treating severe or invasive salmonellosis in children and pregnant women (16,17) where fluoroquinolones are not approved and treatment options are limited. Accordingly, third-generation cephalosporins have been classified as Critically Important Antimicrobials in Human Medicine by the World Health Organization (21) and as Class 1 Very High Importance in Human Medicine by the Canadian Veterinary Drugs Directorate, Health Canada (22).

In Canada, ceftiofur resistance in bacteria from healthy animals or food is mainly reported in chicken Salmonella Heidelberg isolates originating from farm, abattoir, and retail samples and in chicken abattoir and retail generic E. coli isolates (11,12). It also is occasionally reported in Salmonella isolates from sick animals or in bovine and porcine abattoir or retail E. coli isolates but at much lower frequency (12).

The objective of this study is to highlight the correlation between ceftiofur-resistant Salmonella Heidelberg isolated from retail chicken and the incidence of ceftiofur-resistant Salmonella Heidelberg infections in humans across Canada. Public health concerns raised by publication of the CIPARS 2003 annual report, specifically the higher rates of ceftiofur resistance in Salmonella Heidelberg isolates from chicken meat than from humans, prompted Québec broiler chicken hatcheries to voluntarily interrupt the extralabel in ovo use of ceftiofur during 2005–2006 (23). This study therefore also describes variations in ceftiofur resistance among chicken and human Salmonella Heidelberg and chicken E. coli strains in that province before, during, and after the voluntary withdrawal.

Materials and Methods
CIPARS is a national program led by the Public Health Agency of Canada (PHAC) dedicated to the preservation of effective antimicrobial drugs for humans and animals through the collection, integration, analysis, and communication of trends in antimicrobial resistance in selected bacterial organisms. Data presented here were collected during 2003–2008 from CIPARS' surveillance of human clinical Salmonella isolates and E. coli and Salmonella isolates from retail chicken. Detailed methods for sample collection, bacterial isolation, antimicrobial resistance testing, and data analysis are described in CIPARS's reports (12).

Sample Collection
Human Salmonella Isolates

Hospital-based and private clinical laboratories isolated and forwarded human Salmonella isolates to their Provincial Public Health Laboratory (PPHL). PPHLs forwarded Salmonella isolates to the Enteric Diseases Program, National Microbiology Laboratory (NML), PHAC, for phage type characterization and antimicrobial resistance testing. All isolates (outbreak and nonoutbreak) received passively by the Saskatchewan PPHL were forwarded; the more populated provinces (British Columbia, Ontario, and Québec) forwarded isolates received from days 1–15 of each month. Only 1 isolate per patient was kept for the analysis.

Retail Meat Samples

To use a similar geographic scale as CIPARS surveillance of human clinical Salmonella isolates and because we expected a certain level of provincial clustering in food distribution, we designed the study of CIPARS retail surveillance to provide a representative measurement of what consumers from each province were exposed to through ingestion of improperly cooked raw meat or cross-contamination. Randomization and weighted allocation of samples according to demography of the human population ensured that the data generated through retail sampling were representative and reliable at the provincial level. Retail raw chicken samples (most often chicken thigh with skin on) were collected as part of CIPARS retail program that purchases samples weekly (Ontario and Québec) or biweekly (Saskatchewan, British Columbia) from chain, independent, and butcher stores in 15–18 randomly selected census divisions in each participating province. Retail surveillance was initiated in Ontario and Québec in mid-2003 and at the beginning of 2005 in Saskatchewan. Surveillance also was conducted during part of 2007 and all of 2008 in British Columbia.

Microbiologic Analysis
Recovery of Isolates from Retail Chicken Meat

Primary isolations of E. coli and Salmonella spp. were conducted at the Laboratory for Foodborne Zoonoses, PHAC. Every retail chicken meat sample received was cultivated for Salmonella, but only 1 of every 2 samples was systematically selected to be tested for generic E. coli isolation. Incubated peptone rinses of chicken meat samples were streaked on eosin-methylene blue agar (Becton Dickinson, Sparks, MD, USA). Presumptive E. coli colonies were identified by using the Simmons' citrate and indole tests. Colonies showing negative indole results were identified by using the API 20E (bioMérieux Clinical Diagnostics, Marcy l'Étoile, France). All chicken samples were tested for Salmonella with a modified MFLP-75 method of the Compendium of Analytical Methods (24). Incubated peptone rinses were injected into modified semisolid Rappaport-Vassiliadis media. Presumptive E. coli colonies were injected into triple sugar iron and urea agar slants and subjected to the indole test. Presumptive Salmonella isolates were verified by slide agglutination using PolyA-I and Vi Salmonella antiserum (Difco, Becton Dickinson). Salmonella isolates were shipped between laboratories on a tryptic soy agar slant by priority courier. No selective media were used to isolate ceftiofur-resistant bacteria.

Serotyping, Phage Typing, and Susceptibility Testing

Human and chicken Salmonella isolates were serotyped and phage typed by using published methods (25–28). MICs were determined by the NML (human isolates) and the Laboratory for Foodborne Zoonoses, PHAC (chicken isolates) by the broth microdilution method (Sensititre Automated Microbiology System; Trek Diagnostic Systems Ltd., Westlake, OH, USA). Salmonella and E. coli isolates were tested by using the National Antimicrobial Resistance Monitoring System custom susceptibility plate for gram-negative bacteria. The breakpoint used to determine ceftiofur resistance was >4 μg/mL (29).

Data Analysis
We analyzed data using SAS version 9.1 (SAS Institute Inc., Cary, NC, USA). The yearly proportion of retail chicken samples contaminated with ceftiofur-resistant Salmonella Heidelberg (or E. coli) and the incidence rate of human infection with ceftiofur-resistant Salmonella Heidelberg was calculated as described in CIPARS 2006 annual report (12). The Pearson product-moment correlation was used to verify the correlation between ceftiofur-resistant Salmonella Heidelberg isolated from retail chicken and human incidence estimates by using the Pearson option in the PROC CORR procedure in SAS. We computed the overall correlation coefficient using data across all provinces under study and computed a specific coefficient for provinces with >5 observations (30)

To describe ceftiofur resistance changes by quarter and reduce the noise around the estimate caused by the small number of observations per quarter, we computed a nonweighted rolling average of the prevalence of ceftiofur resistance using data from the current quarter and the previous 2 quarters for chicken E. coli, chicken Salmonella Heidelberg, and human Salmonella Heidelberg isolates from the province of Québec. We tested differences in ceftiofur resistance between years with SAS using χ2 or Fisher exact tests when appropriate.

abrir aquí para acceder al documento CDC EID completo (extenso):
Ceftiofur Resistance in Salmonella Heidelberg | CDC EID