Cite as: Portis E, Lindeman C, Johansen L, et al. Antimicrobial susceptibility of porcine Pasteurella multocida, Streptococcus suis, and Actinobacillus pleuropneumoniae from the United States and Canada, 2001 to 2010. J Swine Health Prod. 2013;21(1):30–41.
Also available as a PDF.

Keywords: swine, antimicrobial susceptibility, Actinobacillus, Pasteurella, Streptococcus
Search the AASV web site for pages with similar keywords.

Received: May 1, 2012
Accepted: June 29, 2012

Livestock veterinarians in the United States and Canada use antimicrobial drugs to treat sick animals and to control the spread of pathogenic bacteria to their healthy pen mates.1,2 Reduction of stress and suffering of animals is an important component of humane husbandry. Antimicrobial drugs are also used to promote growth in many production animals, including swine.1,3 Any use of antimicrobial drugs, however, does carry a risk that resistant bacteria will emerge,4,5 reducing the effectiveness of the drugs and resulting in prolonged illness and suffering, as well as increased numbers of sick animals. Monitoring antimicrobial susceptibility among significant pathogens is therefore an important activity in maintaining effective antimicrobial therapy.6,7 Swine respiratory disease is among the most frequently encountered bacterial infections in swine and can be caused by a number of bacteria, including Pasteurella multocida, Streptococcus suis, and Actinobacillus pleuropneumoniae.8 In this survey, the activities of ceftiofur, penicillin, enrofloxacin, florfenicol, tetracycline, tilmicosin, and tulathromycin against respiratory pathogens recovered from pigs across the United States and Canada between 2001 and 2010 were investigated as part of an on-going, long-term veterinary antimicrobial susceptibility surveillance program.

Materials and methods

Participating laboratories and characterization of isolates

Twenty-four veterinary diagnostic laboratories from the major pork-producing areas of the United States and Canada participated in this surveillance program. The regions from which isolates were obtained are shown in Table 1.

All A pleuropneumoniae, P multocida, and S suis were recovered from diseased or deceased pigs. The diagnostic laboratories selected the isolates on the basis of their own protocols, but were requested not to use susceptibility as a criterion for selection. In order to limit over-representation from any one geographic area, the participating laboratories were asked to submit no more than a maximum number of isolates each year. While this maximum number changed slightly during the 10-year period, the number was always ≤ 40 isolates of each bacterial species per laboratory per year. Starting in 2003, Pfizer Animal Health requested that the participating laboratories send no more than one isolate of each bacterial species from a herd each quarter-year to reduce the risk of over-representation of clones from local outbreaks. The total number of isolates recovered each year by each of the laboratories was not provided to Pfizer Animal Health.

Isolates were identified to the genus and species level by the submitting laboratory before shipment to the Pfizer Animal Health laboratory in Kalamazoo, Michigan. Standard biochemical tests and commercially available identification systems (API Microbial Identification Kits; bioMérieux, Durham, North Carolina, and Biolog Microbial Identification System; Biolog Systems, Hayward, California) were used to confirm or further characterize the isolates when necessary. All isolates were stored in 1.0 mL trypticase soy broth (BD Biosciences/Diagnostics, Sparks, Maryland) supplemented with 10% glycerol and were held at approximately -70°C until tested.

Minimal inhibitory concentration determinations

Over the 10 years of the survey, all minimal inhibitory concentration (MIC) determinations were conducted by two laboratories (Pfizer Animal Health, Kalamazoo, Michigan, and Microbial Research Inc, Fort Collins, Colorado) to minimize potential testing bias.9,10 Both laboratories adhered to Clinical and Laboratory Standards Institute (CLSI) standardized methods and quality control during susceptibility testing (Table 2). Minimal inhibitory concentrations (MICs) for all isolates were determined using a dehydrated broth microdilution system (Sensititre Division, Trek Diagnostic Systems, Inc, Cleveland, Ohio). This method conforms to the standards of the CLSI for testing veterinary pathogens.11 Direct colony suspensions were used when testing all organisms, and suspensions were prepared to yield a final bacterial concentration of approximately 5 × 105 colony forming units (CFU) per mL. The custom 96-well microtiter panels initially included serial doubling dilutions of the following antimicrobial agents: ceftiofur, enrofloxacin, florfenicol, penicillin, tetracycline, and tilmicosin. Enrofloxacin was not tested between 2004 and 2007, and when it was included in the panel again in 2008, the lowest concentration range was decreased from 0.03 µg per mL to 0.004 µg per mL to better accommodate quality-control ranges for this drug. Tulathromycin was added to the panel in 2004, prior to its approval by the US Food and Drug Administration (FDA) in 2005 for treatment of infections due to swine respiratory disease- (SRD-) associated pathogens, including A pleuropneumoniae and P multocida. Concentration ranges for each antimicrobial agent were chosen to encompass appropriate quality-control ranges and applicable clinical break points when available. In 2008, the range of tetracycline concentrations was altered to accommodate additional antimicrobial agents in the 96-well microtiter plates.

In 2008, the CLSI formally issued a clarification on the methodology for susceptibility testing of veterinary streptococci.11 They recommended that the inoculation medium for MIC testing of streptococci be a cation-adjusted Mueller-Hinton broth (CAMHB) that contained 2.5% to 5% lysed horse blood. Previous testing of streptococci in our laboratory included use of CAMHB without lysed horse blood. To determine the effect of this change, the S suis strains isolated in 2007 were retested in 2008 using both types of broth, and the results were compared. The MIC50 and MIC90 values for florfenicol, penicillin, tetracycline, tilmicosin, and tulathromycin against S suis were the same using either method. However, with the addition of lysed horse blood in the medium, the ceftiofur MIC50 increased from ≤ 0.03 to 0.06 µg per mL, and the MIC90 increased from 0.12 to 1.0 µg per mL. The MIC results reported here for the S suis isolated from 2007 to 2010 were tested under the updated (2008) recommendations.

Results

Actinobacillus pleuropneumoniae

Table 3 shows the MIC distributions for the seven antimicrobial drugs tested against A pleuropneumoniae, along with the MIC50 and MIC90 values. All isolates tested each year showed MIC values that were less than the CLSI break points for susceptibility to ceftiofur (MIC ≤ 2 µg per mL) and to florfenicol (MIC ≤ 2 µg per mL). There was an increase in the ceftiofur MIC90 for A pleuropneumoniae in 2005, but this was not seen in subsequent years. Penicillin susceptibility did not change substantially over the 10 years. The penicillin MIC50 values were either 0.5 or 1 µg per mL for each year of the study, and the MIC90 values were all ≥ 32 µg per mL, except for 2010, when the MIC90 was 8 µg per mL. In 2003 and 2008, 0.6% and 3.5% of A pleuropneumoniae isolates, respectively, had enrofloxacin MICs that were higher than the CLSI-approved (but not yet published) susceptible break point (MIC ≤ 0.25 µg per mL), but no isolates with an MIC greater than the susceptible break point were detected in the other 4 years in which enrofloxacin was included in the testing panel. The susceptible break point for tetracycline is ≤ 0.5 µg per mL, and ≥ 92% of the A pleuropneumoniae isolates were above this, with MIC50 values > 8 µg per mL in each year of the survey. The number of A pleuropneumoniae isolates with tilmicosin MIC values greater than the susceptible break point (MIC ≤ 16 µg per mL) varied during the 10-year surveillance period, with all isolates having MIC values below the susceptible break point in 2001, and 10% to 15% testing above the susceptible break point the last 2 years of testing. All isolates had tulathromycin MIC values ≤ 64 µg per mL, the CLSI-approved (but not yet published) susceptible break point. The MIC50 and MIC90 values for both tilmicosin and tulathromycin increased over the surveillance period.

Pasteurella multocida

Table 4 shows the MIC distribution frequencies of P multocida isolates collected between 2001 and 2010 from across the United States and Canada. All P multocida isolates tested each year remained susceptible to ceftiofur. Penicillin was active against P multocida: > 95% of the isolates tested between 2001 and 2010 had MICs ≤ 0.25 µg per mL. The enrofloxacin MIC50 value changed from ≤ 0.03 µg per mL to 0.015 µg per mL between 2001 and 2008, but this might reflect the lower concentrations tested from 2008 to 2010. The enrofloxacin MIC90 values remained ≤ 0.03 µg per mL between the two testing periods, and the proportion of isolates that had MICs that were equal to or less than the CLSI break point for enrofloxacin susceptibility (MIC ≤ 0.25 µg per mL) also remained consistent, at or near 100%.

With the exception of the first year of the survey (2001), all florfenicol MIC50 and MIC90 values remained at 0.5 µg per mL over the years of this study. The proportion of P multocida isolates that were susceptible to florfenicol by CLSI standards remained at or near 100% across all years. Less than 47% of the P multocida isolates were susceptible to tetracycline in each year of the survey (MIC ≤ 0.5 µg per mL). Pasteurella multocida showed high levels of susceptibility to the macrolides tilmicosin and tulathromycin, although against both drugs, MIC50 and MIC90 values increased with time. The number of P multocida isolates with tilmicosin MIC values greater than the susceptible break point (MIC ≤ 16 µg per mL) was very low from 2001 to 2010, ranging from 0% in 2003 to approximately 6% above the susceptible break point the last 2 years of testing. The proportion of isolates that were susceptible to tulathromycin, according to CLSI break points, did not change substantially over time: 100% of isolates in most years were categorized as susceptible.

Streptococcus suis

As a consequence of the change in MIC testing methodology for S suis in 2007, an increase in ceftiofur MIC50 and MIC90 values between 2006 and 2007 is evident (Table 5). For each method of testing, ceftiofur MIC50 values remained consistent and MIC90 values indicate small fluctuations. The penicillin MIC50 values did not change over the 10 years of the survey, although the MIC90 values increased from 0.25 to 1 µg per mL after the change in testing method. The proportion of enrofloxacin MIC values among the S suis that were susceptible (MIC ≤ 0.5 µg per mL) during 2008 to 2010 declined slightly from the 2001 to 2003 testing period. Susceptibility to florfenicol remained high, with > 97% of isolates each year susceptible to this drug. Conversely, < 4% of S suis isolates each year were susceptible to tetracycline. Tilmicosin and tulathromycin are not indicated for S suis, and therefore there are no CLSI-approved break points for these two macrolides against S suis. The data indicate that there was very little antimicrobial activity of these drugs against this organism, with MIC50 and MIC90 values > 64 µg per mL during all years of this study.

Discussion

The relatively high prevalence of the three potentially serious pathogens in swine herds in the United States and Canada8,12-14 and the need to treat and control further infection in litter and pen mates indicates the importance of high levels of susceptibility to the antimicrobial drugs that are available to veterinarians. However, as the FDA and the American Veterinary Medical Association have noted, there is no publicly funded nationwide monitoring of antimicrobial susceptibility among swine pathogens in the United States.15 Only a few countries conduct nationwide surveys of swine pathogens, and most of these surveys focus upon zoonotic bacteria such as Salmonella and Escherichia coli. Systematic surveillance of porcine respiratory pathogens conducted annually or at regular time intervals are even scarcer. Germany has a national program, GERM-Vet,7,16 which examines the antimicrobial susceptibility of A pleuropneumoniae, P multocida, and S suis isolated from swine in Germany, and results of recent testing have been published.17

This report seeks to provide a picture of the antimicrobial susceptibility of a convenience sample of SRD pathogens isolated from swine across the United States and Canada over the period 2001 to 2010. This program was not designed to estimate the prevalence of resistant porcine pathogens and indeed, denominator data, which would be required for estimating the size of the population being sampled, was not available. Instead, the program monitors changes in in vitro susceptibility among representative samples of identified pathogens and provides a warning system for the emergence of resistance, a feature of other antimicrobial susceptibility surveillance programs.7 The program is ongoing and continues to collect SRD bacteria from across North America and test them for antimicrobial susceptibility.

As Schwarz et al18 have stated, there are many important reasons for presenting MIC frequencies in surveillance data in the format that is used in this report. These include permitting comparisons between MIC datasets in which different break points or epidemiological cut-off points are used, or even where no break points have been established. Publishing this data also allows for observation of MIC shifts that are not reflected in calculated values such as MIC90 or percentages susceptible and resistant. While it added substantially to the length of this report, we believe that there is value in including the MIC frequency distributions and that these provide more details of the dynamics of antimicrobial susceptibility changes among swine respiratory pathogens than would be available if just the summarized values were included.

Antimicrobial resistance surveillance programs are subject to a number of limitations, including sampling bias.9,10,19 In a study of published antimicrobial resistance surveillance studies,20 it was concluded that sampling bias and failure to address the potential bias introduced by isolates from a common outbreak are frequent, but that case definition and laboratory practices and procedures may also influence the validity of the results. In the current study, the sampling strategy changed in 2003, when the number of isolates of a target species from any herd was restricted to one isolate during any quarter. The impact of this change has not been determined, but the data from 2003 and all years following were from isolates collected using this restriction. The number of isolates submitted by each laboratory was different each year, and not all of the participating laboratories submitted isolates every year. Additionally, testing methods have changed for streptococci.

Bias due to laboratory testing practices were minimized by using only two laboratories to conduct MIC testing, and both adhered to standard microbiological methods for susceptibility testing and quality-control standards. The data in this program came from over 6000 clinical swine isolates, and while this is a substantial number, it is only a small representative sample of the SRD pathogen population in the United States and Canada. To increase the likelihood that our sample was representative, veterinary diagnostic laboratories from across the major pork-producing areas of the United States and Canada provided isolates for this program. The isolates may have been collected and sent to state or provincial laboratories only after treatment with antimicrobial drugs had failed, and so the isolates in this study may reflect a more resistant bacterial population compared to isolates collected from animals without previous antimicrobial treatment or where treatments were successful, resulting in additional selection bias.

This report provides the first extensive survey of the antimicrobial susceptibility of major SRD pathogens isolated from swine across the United States and Canada during the years 2001 to 2010. The data show that, over those 10 years, A pleuropneumoniae and P multocida remained susceptible to ceftiofur, enrofloxacin, florfenicol, tilmicosin, and tulathromycin. Streptococcus suis remained susceptible to ceftiofur, enrofloxacin, and florfenicol. While the data show consistently low penicillin MIC values for P multocida and S suis, along with higher MIC values for A pleuropneumoniae, the CLSI has not approved interpretive criteria for penicillin against swine pathogens. The inoculating broth used for susceptibility testing of S suis was modified in 2007, during this study, and it is unknown if the increases in penicillin MIC90 values between 2008 and 2010 are related to this change. Most isolates of the three organisms were resistant to tetracycline, with little change in the MIC distributions over the 10 years of the survey.

The data presented in this report, especially those data that show that there has been some increase in MICs of important antimicrobial agents, should serve to underscore the importance of prudent use of these drugs when treating SRD (and other infections). Careful stewardship may allow for effective use of these drugs for many years. On-going surveillance of the in vitro susceptibility of these SRD pathogens will continue to be an important component in antimicrobial stewardship.

Implications

•  Monitoring antimicrobial susceptibility among swine pathogens over time provides valuable information about changes which may be occurring in the antimicrobial susceptibility of these organisms. Having current susceptibility information is an important function in the maintenance of effective antimicrobial therapy.

•  Surveillance of the in vitro susceptibility of SRD pathogens should continue as an important component in antimicrobial stewardship.

Acknowledgements

The authors would like to thank the veterinary diagnostic laboratories for their generous assistance by providing the bacterial isolates for this study: Cornell University, Iowa State University, Kansas State University, Manitoba Agriculture Services, Michigan State University, North Carolina Department of Agriculture, Ohio Department of Agriculture, Oklahoma State University, Pennsylvania State University, Purdue University, South Dakota State University, Texas A&M (Amarillo), Texas A&M (College Station), University of California Davis (Davis), University of California Davis (Tulare), University of Guelph, University of Illinois, University of Minnesota, University of Nebraska, University of Saskatchewan, University of Wisconsin, Washington State University, and two additional veterinary diagnostic laboratories that wish to not be acknowledged.

References

1. Cromwell GL. Why and how antibiotics are used in swine production. Anim Biotechnol. 2002;13:7–27.

2. Mackinnon JD. The proper use and benefits of veterinary antimicrobial agents in swine practice. Vet Microbiol. 1993;35:357–367.

3. Hughes P, Heritage J. Antibiotic Growth-Promoters in Food Animals. 2002. Available at http://www.fao.org/DOCREP/ARTICLE/AGRIPPA/555_EN.HTM. Accessed 15 October 2012.

4. McEwen SA, Fedorka-Cray PJ. Antimicrobial use and resistance in animals. Clin Infect Dis. 2002;34(suppl 3):S93-S106.

5. Prescott JF. Antimicrobial use in food and companion animals. Animal Health Research Reviews / Conference of Research Workers in Animal Diseases. 2008;92:127–133.

6. Aarestrup FM. Monitoring of antimicrobial resistance among food animals: principles and limitations. J Vet Med B Infect Dis Vet Public Health. 2004;51:380–388.

7. Wallmann J. Monitoring of antimicrobial resistance in pathogenic bacteria from livestock animals. Int J Med Microbiol. 2006;296(suppl 2):81–86.

8. MacInnes JI, Gottschalk M, Lone AG, Metcalf DS, Ojha S, Rosendal T, Watson SB, Friendship RM. Prevalence of Actinobacillus pleuropneumoniae, Actinobacillus suis, Haemophilus parasuis, Pasteurella multocida and Streptococcus suis in representative Ontario swine herds. Can J Vet Res. 2008;72:242–248.

9. Kahlmeter G, Brown DF. Resistance surveillance studies—comparability of results and quality assurance of methods. J Antimicrob Chemother. 2002;50:775–777.

10. Bax R, Bywater R, Cornaglia G, Goossens H, Hunter P, Isham V, Jarlier V, Jones R, Phillips I, Sahm D, Senn S, Struelens M, Taylor D, White A. Surveillance of antimicrobial resistance–what, how and whither? Clin Microbiol Infect. 2001;7:316–325.

11. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial disk and dilution susceptibility tests for bacteria isolated from animals; Approved standard – Third Edition. M31-A3. Volume 28, Number 8. Wayne, Pennsylvania: Clinical and Laboratory Standards Institute; 2008. Available at: http://www.clsi.org/source/orders/free/m31-a3.pdf. Accessed 18 October 2012.

12. Marsteller TA, Fenwick B. Actinobacillus pleuropneumoniae disease and serology. Swine Health Prod. 1999;7:161–165.

13. USDA/APHIS. Swine 2006. Part II: Reference of swine health and health management practices in the United States, 2006. 2007. Available at: http://www.aphis.usda.gov/animal_health/nahms/swine/downloads/swine2006/Swine2006_dr_PartII.pdf. Accessed 15 October 2012.

14. MacInnes JI, Desrosiers R. Agents of the “suis-ide diseases” of swine: Actinobacillus suis, Haemophilus parasuis, and Streptococcus suis. Can J Vet Res. 1999;63:83–89.

15. US FDA/CVM and AVMA. Judicious use of antimicrobials for swine veterinarians. 2002. Available at: http://www.fda.gov/downloads/AnimalVeterinary/SafetyHealth/AntimicrobialResistance/JudiciousUseofAntimicrobials/UCM095578.pdf. Accessed 15 October 2012.

16. Wallmann J, Schroter K, Wieler LH, Kroker R. National antibiotic resistance monitoring in veterinary pathogens from sick food-producing animals: the German programme and results from the 2001 pilot study. Int J Antimicrob Agents. 2003;22:420–428.

17. Bundesamt für Verbraucherschutz und Lebensmittelsicherheit. GERMAP 2008. Antibiotika-Resistenz und -Verbrauch [Antibiotic resistance and consumption]. 2008. Available at: http://www.bvl.bund.de/SharedDocs/Downloads/08_PresseInfothek/Germap_2008.pdf?__blob=publicationFile&v=2. Accessed 15 October 2012.

18. Schwarz S, Silley P, Simjee S, Woodford N, van Duijkeren E, Johnson AP, Gaastra W. Assessing the antimicrobial susceptibility of bacteria obtained from animals. Vet Microbiol. 2010;141:1–4.

19. Rempel OR, Laupland KB. Surveillance for antimicrobial resistant organisms: potential sources and magnitude of bias. Epidemiol Infect. 2009;137:1–9.

20. Rempel O, Pitout JDD, Laupland KB. Antimicrobial resistance surveillance systems: Are potential biases taken into account? Can J Infect Dis Med Microbiol. 2011;22:e24–e28.