Prevalence and time trend analysis of antimicrobial resistance in respiratory bacterial pathogens collected from diseased pigs in USA between 2006–2016

Journal Pre-proof Prevalence and time trend analysis of antimicrobial resistance in respiratory bacterial pathogens collected from diseased pigs in USA between 2006–2016

Shivdeep Singh Hayer, Albert Rovira, Karen Olsen, Timothy J. Johnson, Fabio Vannucci, Aaron Rendahl, Andres Perez, Julio Alvarez PII:

S0034-5288(19)30870-7

DOI:

https://doi.org/10.1016/j.rvsc.2019.11.010

Reference:

YRVSC 3923

To appear in:

Research in Veterinary Science

Received date:

25 August 2019

Revised date:

18 November 2019

Accepted date:

19 November 2019

Please cite this article as: S.S. Hayer, A. Rovira, K. Olsen, et al., Prevalence and time trend analysis of antimicrobial resistance in respiratory bacterial pathogens collected from diseased pigs in USA between 2006–2016, Research in Veterinary Science (2019), https://doi.org/10.1016/j.rvsc.2019.11.010

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier.

Journal Pre-proof Prevalence and time trend analysis of antimicrobial resistance in respiratory bacterial pathogens collected from diseased pigs in USA between 2006-2016

Shivdeep Singh Hayera, Albert Rovirab, Karen Olsenb, Timothy J. Johnsonc, Fabio Vannuccib, Aaron Rendahlc, Andres Pereza, Julio Alvarezade*

Department of Veterinary Population Medicine, College of Veterinary Medicine, University of

of

a

Veterinary Diagnostic Laboratory, College of Veterinary Medicine, University of Minnesota,

-p

b

ro

Minnesota, St. Paul, USA

c

re

St. Paul, USA

Department of Veterinary and Biomedical Sciences, College of Veterinary Medicine,

lP

University of Minnesota, St. Paul, USA

VISAVET Health Surveillance Center, Universidad Complutense, Madrid, Spain

e

Department of Animal Health, Facultad de Veterinaria, Universidad Complutense, Madrid,

ur Jo

Spain

na

d

* Corresponding author: Email- [email protected]

1

Journal Pre-proof Abstract Swine respiratory disease complex (SRDC) causes massive economic losses to the swine industry and is a major animal welfare concern. Antimicrobials are mainstay in treatment and control of SRDC. However, there is a lack of data on the prevalence and trends in resistance to antimicrobials in bacterial pathogens associated with SRDC. The objective of this study was to estimate the prevalence and changes in resistance to 13 antimicrobials in swine bacterial

of

pathogens (Streptococcus suis, Pasteurella multocida, Actinobacillus suis and Haemophilus

ro

parasuis) in the U.S.A using data collected at University of Minnesota Veterinary Diagnostic

-p

Laboratory between 2006-2016. For antimicrobials for which breakpoints were available,

re

prevalence of resistance remained below 10% except for tetracycline in S. suis and P. multocida isolates, and these prevalence estimates remained consistently low over the years despite

lP

statistical significance (p< 0.05) in trend analysis. For antimicrobial-bacterial combinations

na

without available breakpoints, the odds of isolates being resistant increased by >10% annually for 7 and 1 antimicrobials in H. parasuis and S. suis isolates respectively, and decreased >10%

ur

annually for 4 and 1 antimicrobials in A. suis and H. parasuis isolates, respectively, according to

Jo

the ordinal regression models. Clinical implications of changes in AMR for A. suis and H. parasuis should be interpreted cautiously due to the lack of interpretive criteria and challenges in antimicrobial susceptibility tests in the case of H. parasuis. Future studies should focus on surveillance of antimicrobial resistance and establishment of standardized susceptibility testing methodologies and interpretive criteria for these animal pathogens of critical importance. Keywords- Antimicrobial resistance, pigs, swine respiratory pathogens, surveillance

2

Journal Pre-proof Introduction Respiratory diseases are frequently associated with substantial financial losses in swine production systems worldwide due to high morbidity, decreased weight gain, increased culling rates, and additional medicine and labor costs (Fablet et al., 2012). Etiology of respiratory diseases is often complex, with a group of bacteria (such as Actinobacillus pleuropneumoniae) and viruses acting as primary pathogens with potential to cause the disease alone and other

of

microorganisms (such as Pasteurella multocida, Haemophilus parasuis and Streptococcus suis)

ro

acting as secondary pathogens that further aggravate the disease (Brockmeier et al., 2002). A.

-p

suis and H. parasuis are members of the Pasteurellaceae family and are associated with systemic

re

diseases including polyserositis, pleuritis, meningitis, arthritis, and respiratory diseases such as acute pneumonia (MacInnes and Desrosiers, 1999). P. multocida are commensal bacteria that

lP

can act as opportunistic secondary pathogens during respiratory diseases caused by other agents

na

such as Mycoplasma hyopneumoniae (Oliveira Filho et al., 2018). S. suis is a Gram-positive bacteria that primarily causes septicemia, but that can also contribute to the swine respiratory

ur

disease complex (SRDC) as a secondary pathogen (Opriessnig et al., 2011).

Jo

Management practices, vaccination and antimicrobial therapy are the main methods available to prevent, treat, and/or control infection caused by these pathogens. Bacteria such as H. parasuis and S. suis are commensals that colonize pigs early in their life, so that subclinically infected pigs can act as sources of infection for other susceptible pigs (Brean et al., 2016; MacInnes et al., 2008; Varela et al., 2013; Vötsch et al., 2018). Hence, management strategies aimed at preventing their circulation in a herd can be difficult to execute in practice due to their widespread presence. The effect of vaccination in preventing and controlling infections due to these bacterial pathogens has so far proven inconsistent because of the wide serovar diversity of 3

Journal Pre-proof the bacteria involved, poor cross-protection, low protective efficacy, and interference with maternal antibodies (Jiang et al., 2016; Lapointe et al., 2002; Lin et al., 2018; Macedo et al., 2015). Because of the challenges in preventing infection with ubiquitous pathogens (such as S. suis, P. multocida and H. parasuis) through management measures and the inefficacy of vaccination, use of antimicrobials plays a key role in the control of SRDC caused by these

of

bacteria. A number of antimicrobial compounds from different classes have been licensed for

ro

SRDC treatment in the U.S.A, including penicillins (ampicillin, penicillin), cephalosporin

-p

(ceftiofur), tetracyclines (oxytetracycline, chlortetracycline), macrolides (tulathromycin), and

re

pleuromutilins (tiamulin) (Food and Drug Administration, 2019a). Use of antimicrobials has been linked to the eventual development of resistance in bacteria (Magstadt et al., 2018) and

lP

there is a need for judicious antimicrobial use to preserve their efficacy (El Garch et al., 2016).

na

Surveillance of antimicrobial resistance (AMR) plays an important role in ensuring the long-term efficacy of antimicrobials but, unlike AMR in Escherichia coli and Salmonella enterica, which is

ur

routinely monitored as part of the National Antimicrobial Resistance Monitoring Systems

Jo

(NARMS), little information is available regarding AMR trends in SRDC pathogens in the United States of America (U.S.A) (Sweeney et al., 2017). Lack of clinical breakpoints for specific antimicrobial-bacterial combinations (Toutain et al., 2017) and difficulty in culturing certain bacteria such as H. parasuis (Brogden et al., 2018; Prüller et al., 2017) further contribute to an absence of harmonization in interpreting antimicrobial susceptibility testing results. The objectives of this study were to estimate the prevalence of, and changes in AMR over time (2006-2016) in swine bacterial isolates (A. suis, H. parasuis, P. multocida and S. suis) collected from diseased pigs in the U.S.A using antimicrobial susceptibility testing results from 4

Journal Pre-proof the University of Minnesota-Veterinary Diagnostic Laboratory (UMN-VDL). The antimicrobials tested for susceptibility are commonly used in swine medicine in the U.S.A. Minnesota is the third largest pork producing state and the UMN-VDL routinely receives diagnostic submissions from all the major pork producing states in the U.S.A, resulting in an alternative source of data for analyzing AMR in swine pathogens in the absence of formal systems to collect such information in the country. Results here will help to elucidate the trend of AMR in bacterial

of

pathogens causing SRDC in the U.S.A, which, will ultimately help to inform decisions related

ro

with the prevention and control of one of the disease syndromes with a huge financial impact on

-p

the swine industry of the U.S.A.

re

Materials and methods

lP

Bacteriology

This study used the data on the antimicrobial susceptibility testing performed on bacteria

na

associated with the SRDC recovered from samples from diseased pigs submitted for routine

ur

diagnostic purposes to the UMN-VDL between January 1, 2006 to December 31, 2016 (A. suis, P. multocida and S. suis) and January 1, 2010 to December 31, 2016 (H. parasuis). Tissue

Jo

samples were selected for bacteriological evaluation if they had gross lesions and/or on the basis of clinical history. Organs selected for bacterial isolation included upper respiratory tract (nasal cavity, nasal turbinates), lower respiratory tract (trachea, bronchi, lungs) and other body sites (brain, meninges, spinal cord, heart, pericardium, pleura, abdominal cavity, intestines, stomach, spleen, liver, joints, tendons, kidney, vulva, urethra, umbilicus and uterus). Samples consisted of tissue sections or swabs and were cultured on tryptic soy agar with 5% sheep blood aerobically with and without 5% CO 2 for 18-24 hours for bacterial isolation except in the case of H. parasuis isolation, for which sheep blood agar plates were additionally supplemented with a streak of 5

Journal Pre-proof Staphylococcus aureus that acted as a source of a required factor (nicotinamide adenine dinucleotide) needed for H. parasuis growth. Bacterial colonies were characterized at the bacterial species level by standard biochemical testing and/or MALDI-TOF mass spectrometry. At this stage, bacteria were selected for antimicrobial susceptibility testing if they were suspected to be involved in the disease process. Bacteria were preferentially selected for antimicrobial susceptibility testing if they were isolated from organs suspected to be involved in a systemic

of

disease as these were suspected to be more pathogenic as compared to bacteria isolated from

ro

lungs. However, if no isolates were recovered from other body sites, those isolated from lungs

-p

or other parts of the respiratory tract were tested for antimicrobial susceptibility.

re

Minimum inhibitory concentrations (MIC) were determined by the broth microdilution method using the Sensititre automatized dilution system (Trek Diagnostic Systems, Cleveland,

lP

Ohio). Antimicrobial susceptibility testing of S. suis and P. multocida isolates was performed as

na

per the standardized Clinical and Laboratory Standards Institute (CLSI) methodology (Clinical and Laboratory Standards Institute, 2018) but the MIC values for A. suis and H. parasuis were

ur

determined by growing standardized quantities of bacterial suspensions on cation-adjusted

Jo

Mueller Hinton broth and Haemophilus test medium, respectively for 18-24 hours at 35-370 C aerobically with 5% CO 2 . The antimicrobials tested are presented in table1. Quality assurance/quality control (QA/QC) were performed based on CLSI guidelines for each run of testing (Clinical and Laboratory Standards Institute, 2018). In case of failure to comply with QA/QC, results were discarded, possible causes of failure were evaluated and QC was repeated until passed thrice in a row. Data extraction and statistical analysis

6

Journal Pre-proof Antimicrobial susceptibility data, along with information on the age of the pig, date of isolation, organs from which the bacteria were isolated, and geographical state of origin were extracted from the computerized database maintained at the UMN-VDL. Multiple isolates of a bacterial species were sometimes collected from samples sent by a singlesubmitting client (swine farm/ veterinary clinic). In these scenarios, isolates with identical MIC values for the complete panel of antimicrobials per submitting client were identified and only data from a single isolate was

of

retained for further analyses. Bacteria isolated from swine samples submitted for routine

ro

surveillance and from research trials were also excluded from the analyses.

-p

S. suis and P. multocida isolates were classified as “resistant” or “non-resistant” to

re

specific antimicrobials using CLSI breakpoints when available (table 1). For those antimicrobials for which swine specific CLSI breakpoints were not available, breakpoints applicable to other

lP

species (humans, cattle) or epidemiological cut-offs (gentamicin- P. multocida) were used, if

na

available (table 1). There were no breakpoints or epidemiological cut-offs available for A. suis and H. parasuis. MIC50 and MIC90, the 50th and 90th percentile of the MIC distributions

ur

respectively, were also estimated and presented in tables 2, 3, 4 and 5.

Jo

Changes in AMR levels over time were estimated for each antimicrobial-bacteria combination using logistic regression models. Specifically, the status of bacterial isolates (resistant or non-resistant) was used as a binary dependent variable and time (in years) as a continuous independent variable of interest as described elsewhere (Couto et al., 2016; FoxLewis et al., 2018). Models also accounted for the association with variables such as site of isolation (2 categories- respiratory tract or other body sites involved in systemic disease process), season (winter, fall, spring and summer), and state of origin (Minnesota or elsewhere in the U.S.), which were added as categorical fixed effects. Changes in AMR prevalence over time 7

Journal Pre-proof were quantified using the exponential of the coefficient for the time variable (in years), also referred to as “time associated odds ratio”. To simplify analysis of AMR for bacteria without any breakpoints available (A. suis and H. parasuis), a sensitivity analysis was conducted to study the impact of using breakpoints of related bacteria from family Pasteurellaceae or from previously published studies on estimation of AMR prevalence and changes over time in these bacteria. However, the AMR prevalence

of

estimates were highly sensitive to the choice of breakpoints used (supplementary material-1,

ro

tables S1 and S2). Hence, a statistical approach which required no prior specifications of

-p

breakpoints was adopted.

re

For those antimicrobial- bacterial species combinations for which dichotomization of antimicrobial susceptibility results was not possible due to the lack of available breakpoints,

lP

ordinal logistic regression models were built. In these models, the dependent variable was the

na

MIC value of the isolates and time (in years) was the independent variable of interest, while accounting for the association with the independent variables previously described. The

ur

exponential of the coefficient associated with years in the regression models was interpreted as

Jo

the odds of being at a MIC value or higher as compared to any MIC value below for a unit change in time (one year).

For these ordinal regression models, the proportional odds assumption was statistically tested (Brant, 1990) and if the assumption was not met, non-proportional odds ordinal logistic regression models were fitted. Due to convergence issues, MIC values ranging from 2-5 doubledilutions were binned together if needed. A primary difference between proportional and nonproportional odds logistic regression models is that the former provides a single odds ratio for all MIC values and it is more parsimonious as compared to the latter, which provides a separate 8

Journal Pre-proof time associated odds ratio for each MIC value. Essentially, these ordinal regression models utilize each MIC value as a potential breakpoint and provide changes over time in the odds of being resistant at each MIC value or higher as compared to respective lower values per antimicrobial-bacteria combination. For each antimicrobial-bacterial combination, the results of these ordinal regression models were categorized into 3 classes for the ease of interpretation- a) statistical evidence of

of

increase or decrease in AMR- if there was a significant increase or decrease (p<0.05) in time-

ro

associated odds at all MIC values or higher as compared to respective lower values, suggestive

-p

of changes in AMR over time regardless of MIC values chosen as breakpoints, b) no statistical

re

evidence of change in AMR- if there were no significant changes in time-associated odds (p>0.05) across all MIC values and c) inconclusive- if there were statistically significant changes

lP

in time-associated odds at some but not all MIC values as compared to the respective lower

na

values, or there were changes in direction of odds ratio at different MIC values, implying that the interpretation of changes in AMR over time for these antimicrobial-bacteria combinations are

ur

likely to be dependent heavily on MIC values used as breakpoints. The results of the

S6).

Jo

abovementioned models are summarized in the supplementary material (tables S3, S4, S5 and

Results Description of data Overall, 19,498 isolates were analyzed in this study. Nearly 42.3, 34.6, 14.7 and 8.3 % of these isolates were P. multocida, S. suis, A. suis and H. parasuis, respectively. Nearly 74, 80 and 88% of the A. suis, H. parasuis and P. multocida isolates were obtained from the respiratory tract, respectively whereas 90% of the S. suis isolates with AMR data available were collected from 9

Journal Pre-proof organ systems other than respiratory system. The majority (45%) of the isolates were collected from diagnostic samples submitted by clients in Minnesota. Isolates from the rest of the top ten pork producing states in USA made up for 45% of the isolates. Age information was available for 84% of the isolates, of which 78% were recovered from pigs 1-6 months of age. The samples were submitted by 328 vertically integrated swine systems or veterinary clinics. Streptococcus suis

of

Briefly, breakpoints were available for 8 antimicrobials tested in >6,500 isolates, and prevalence

ro

of resistance to 6 of them was below 10% (table 2). In fact, the overall prevalence of resistance

-p

to five antimicrobials (ampicillin, ceftiofur, enrofloxacin, florfenicol and trimethoprim-

re

sulphamethoxazole) was low (<3%) (table 2) and ranged between 0-3% across the years. There were no statistically significant changes (p>0.05) in prevalence of resistance to ampicillin,

lP

enrofloxacin and trimethoprim-sulphamethoxazole (table S3). The overall prevalence of

na

penicillin resistance was 7.0% and varied from 5.3% in years 2006-09 to 9.7% in years 2013-16 (9% annual increase in odds, p<0.01) (table S3, figure 1). Prevalence of ceftiofur resistance also

ur

increased from 0.59% in 2006 to 2.54% in 2016 (annual increase in odds -19%, p<0.01) (table

Jo

S3, figure 1). Prevalence of florfenicol resistance decreased from 1.42% in 2006 to 0.51% in 2016 (annual decrease in odds-11%, p<0.01) (table S3, figure 1). Levels of oxytetracycline and chlortetracycline resistance were always high and ranged between 93-97% over the years without any statistically significant change (p>0.05) (table S3). For these antimicrobials with breakpoints available, there was no statistical difference in the proportion of resistant isolates depending on the body site (upper respiratory tract, lower respiratory tract or other body sites) (based on chisquare tests).

10

Journal Pre-proof Breakpoints were not available for gentamicin, spectinomycin, tiamulin, tulathromycin and sulphadimethoxine and MIC distributions for these antimicrobials are presented in table 2. There was a 11-17% annual increase in the odds of being resistant to gentamicin depending on the MIC value used as breakpoint (p<0.01) (figure 2a, table S3). For sulphadimethoxine, there was a modest (4%) decrease in annual odds of having MIC value of >256 µg/ml over the years (p<0.01) (table S3). The results were inconclusive for spectinomycin and tiamulin, with

of

statistically significant annual changes using some MIC values as breakpoints but decreases or

ro

no change in odds if other MIC values were used as breakpoints (table S3). For example, in case

-p

of spectinomycin, there was a 6% annual decrease (p<0.01) in odds of having MIC values 32

re

µg/ml but a 3% annual increase (p<0.01) in odds of having MIC values 128 µg/ml as compared to lower MIC values, respectively (table S3).

lP

Pasteurella multocida

na

Breakpoints were available for 10 antimicrobials tested in > 8000 isolates, with overall AMR prevalence being less than 10% and annual values ranging between 0 to 8% in 8 of them

ur

(ampicillin, ceftiofur, enrofloxacin, florfenicol, gentamicin, penicillin, spectinomycin and

Jo

tulathromycin) (table 3). There were no statistically significant changes (p>0.05) in prevalence of resistance to enrofloxacin, florfenicol, gentamicin and tulathromycin (table S4). However, there was 4-19% decrease in annual odds of resistance to ampicillin, ceftiofur, penicillin and spectinomycin (p<0.01) (table S4, figure 1). Resistance against the remaining two antimicrobials (oxytetracycline and chlortetracycline) varied widely and ranged between 28-56% (chlortetracycline) and 52-72% (oxytetracycline) during the study period, with logistic regression models suggesting a modest (4-5%) annual increase in the odds of resistance (p<0.01) (table S4). For these antimicrobials with breakpoints available, there was no statistical difference in the 11

Journal Pre-proof proportion of resistant isolates on the basis of body site (upper respiratory tract, lower respiratory tract or other body sites), with the exception of chlortetracycline resistance (based on chi-square test). Resistance to chlortetracycline varied significantly with body site of isolation (upper respiratory tract-62%, lower respiratory tract-61%, other body sites-56%) (chi-square test, p=0.01). Breakpoints were not available for tiamulin, sulphadimethoxine and trimethoprim-

of

sulphamethoxazole and MIC distributions for these are provide in table 3. For tiamulin, results

ro

were inconclusive (table S4). There was a modest annual increase (4%) in the odds of having

-p

sulphadimethoxine MIC values >256 µg/ml (p<0.01) (table S4). The odds of having MIC values

re

(>2 µg/ml) did not change significantly for trimethoprim-sulphamethoxazole (p=0.06) (table S4). Actinobacillus suis

lP

There were no breakpoints available for any of the antimicrobials for A. suis and detailed MIC

na

distributions are presented in table 4. There was no statistically significant change in the odds of resistance to enrofloxacin, trimethoprim-sulphamethoxazole and gentamicin (table S5). There

ur

was a statistically significant annual decrease of 15-29% in the odds of being resistant to

Jo

ceftiofur, spectinomycin, sulphadimethoxine and tiamulin depending on MIC value considered as breakpoint (figures 2b, c, d, table S5). For ampicillin, there was a statistically significant (p<0.05) yet modest decrease (4%) in the odds of being resistant (table S5). The change in odds of resistance were inconclusive for oxytetracycline, florfenicol, tulathromycin and penicillin, with either a statistically significant increase or decrease for some MIC values but nonsignificant changes for other MIC values as compared to respective lower MIC values (table S5). Similarly, the change in odds for chlortetracycline resistance were ambiguous and inconclusive,

12

Journal Pre-proof with a 4% annual increase at MIC value 1 µg/ml but a 5-7% annual decrease at MIC values 4 µg/ml as compared to respective lower MIC values (table S5). Haemophilus parasuis There were no breakpoints available for any of the antimicrobials for H. parasuis and detailed MIC distributions are presented in table 5. There was no significant change in the odds of resistance to chlortetracycline, tiamulin, trimethoprim-sulphamethoxazole and spectinomycin

of

(table S6). However, there was a consistent annual increase of 9-20% in the odds of ampicillin,

ro

florfenicol, gentamicin and penicillin resistance (p<0.01) (figures 2e, f, g, h, table S6). This

-p

annual increase in odds of resistance was even higher for ceftiofur, enrofloxacin and

re

tulathromycin, with the change in odds ranging between 23-34% depending on the MIC values used as breakpoint (p<0.01) (figures 2i, j, k, table S6). There was a 12% annual decrease in odds

lP

of being at MIC values >256 µg/ml for sulphadimethoxine (p<0.01) (table S6). The results of

na

annual change in odds of resistant to oxytetracycline were inconclusive, with a statistically significant decrease in annual odds for one MIC value but non-significant changes in odds for

Discussion

Jo

ur

other MIC values as compared to respective lower MIC values (table S6).

SRDC causes huge economic losses and increased mortality in swine herds worldwide. Antimicrobials are one of the main tools for control of SRDC, but there is limited available information on the prevalence of AMR in these bacteria in comparison with other zoonotic pathogens such as Salmonella or Campylobacter (Food and Drug Administration, 2019b), even though this information could help to guide therapy and detect changes over time. In this study, we estimated the prevalence of AMR against a panel of 13 antimicrobials and the changes in the 13

Journal Pre-proof levels of resistance over more than a decade in key bacterial pathogens involved in SRDC in the U.S.A using data collected at the UMN-VDL. For S. suis, prevalence of ampicillin, ceftiofur, enrofloxacin and florfenicol resistance determined in swine isolates from Europe were similar to the results reported here (<3%), whereas higher levels of trimethoprim-sulphamethoxazole resistance were observed for European isolates as compared to this study (3-12% compared with 2% here) (Callens et al.,

of

2013; El Garch et al., 2016; Hernandez-Garcia et al., 2017; van Hout et al., 2016; Vela et al.,

ro

2005; Wisselink et al., 2006). Reports on penicillin resistance levels in Europe were much more

-p

heterogeneous (ranging from <2% to >21%) (Callens et al., 2013; van Hout et al., 2016;

re

Wisselink et al., 2006) but in some cases agreed with values described here (7%) (HernandezGarcia et al., 2017; Vela et al., 2005). With a few exceptions, European studies also reported

lP

tetracycline resistance levels above 80% in S. suis isolates (Callens et al., 2013; El Garch et al.,

na

2016; Hernandez-Garcia et al., 2017; van Hout et al., 2016; Vela et al., 2005; Wisselink et al., 2006). For those antimicrobials with no available breakpoints (gentamicin, spectinomycin and

ur

tiamulin) the estimated MIC 50 and MIC90 values in determined in our study were within one

Jo

double-dilution value of those reported previously in Europe, (El Garch et al., 2016; HernandezGarcia et al., 2017; van Hout et al., 2016; Vela et al., 2005; Wisselink et al., 2006) with the exceptions of a lower MIC 90 value (16 µg/ml) for spectinomycin (Wisselink et al., 2006) and higher MIC50 (32 µg/ml) and MIC90 (128 µg/ml) values for tiamulin (Callens et al., 2013). Outside Europe, reports of the prevalence of AMR in S. suis were highly variable. Prevalence levels similar to those reported here were also reported for ampicillin, enrofloxacin, penicillin, tetracycline and trimethoprim-sulphamethoxazole resistance in Australian S. suis isolates, while higher levels of florfenicol resistance were observed in Australian S. suis isolates 14

Journal Pre-proof (15%) compared to our results (O’Dea et al., 2018). In stark contrast, Chinese S. suis were found to be highly resistant to clindamycin (98%) and ceftiofur (56%) (Li et al., 2012). Prevalence of resistance to ceftiofur, enrofloxacin, florfenicol, tulathromycin and trimethoprim-sulphamethoxazole and MIC 50 and MIC90 values for tiamulin and spectinomycin in European and Australian P. multocida isolates were also similar to values found here (Dayao et al., 2014a; El Garch et al., 2016). However, both Dayao et al. (2014a) and El Garch et al. (2016)

of

found that 20-28% of the P. multocida isolates were tetracycline resistant, whereas nearly 40-

ro

60% of P. multocida isolates in this study were chlortetracycline and/or oxytetracycline resistant.

-p

In contrast, the prevalence of resistance to antimicrobials (spectinomycin, chlortetracycline,

re

trimethoprim-sulphamethoxazole), MIC 50 , and MIC90 values (gentamicin) reported from clinical swine isolates in China were higher with the exception of ceftiofur and florfenicol resistance

lP

(Tang et al., 2009). Similarly, in isolates from diseased pigs in Taiwan, prevalence of

na

enrofloxacin (61%), florfenicol (92%), and tiamulin (MIC 50 – 128 µg/ml, MIC90 - >128 µg/ml) resistance was comparatively higher (Yeh et al., 2017).

ur

In absence of clinical breakpoints and epidemiological cut-offs, MIC distributions are an

Jo

alternative source of information on AMR levels (Schwarz et al., 2010). For H. parasuis, MIC50 and MIC90 values for ampicillin, ceftiofur, florfenicol, gentamicin, penicillin, tetracycline, tulathromycin and tiamulin reported here were at least three double-dilution levels higher compared with values reported in European studies (Brogden et al., 2018; El Garch et al., 2016). Dayao et al. (2014b) estimated MIC90 values similar (florfenicol, oxytetracycline) or 4-7 doubledilutions lower (ampicillin, ceftiofur, penicillin) in Australian H. parasuis isolates as compared to this study. In contrast, MIC 90 values estimated by Zhao et al. (2018) in Chinese H. parasuis isolates for ceftiofur, enrofloxacin, florfenicol, gentamicin and tetracycline were 2-7 double15

Journal Pre-proof dilution levels higher as compared to isolates in this study. To the best of our knowledge, this is the first study to publish MIC distributions for A. suis and hence, comparison of results to earlier studies was not possible. Significant changes in the MIC distributions of some antimicrobial-bacteria combinations over the years were observed in this study. Notable changes included increases (>10%) in annual odds of resistance to gentamicin (S. suis) and 7 antimicrobials in H. parasuis isolates. However,

of

decreases (>10%) in the annual odds of resistance to 4 antimicrobials in A. suis and

ro

sulphadimethoxine in H. parasuis isolates were also reported. Nevertheless, for bacterial-

-p

antimicrobial combinations for which breakpoints were available, the prevalence of resistance

re

never exceeded 12% and remained consistently low, even though some trends were statistically significant (tables S3 and S4). Similarly, for those antimicrobials with breakpoints available, El

lP

Garch et al. (2016) reported no changes in AMR in European S. suis and P. multocida isolates,

na

except for a significant increase to ceftiofur resistance and a non-significant increase to tetracycline resistance in S. suis isolates. A drastic increase in resistance to spectinomycin and

ur

tiamulin and non-significant changes in resistance to other antimicrobials in S. suis isolates were

Jo

also reported by Hernandez-Garcia et al. (2016). To the best of our knowledge, this is the first study to analyze changes in antimicrobial resistance in A. suis and H. parasuis over any time period. The interpretation of AMR data is severely hindered due to lack of host-bacteriaantimicrobial specific clinical breakpoints (El Garch et al., 2016). Although we estimated that the odds of resistance to several antimicrobials were changing significantly, the clinical or epidemiological interpretation of these results can be contentious due to lack of interpretive criteria. Despite using complex statistical models, the results were often inconclusive for several 16

Journal Pre-proof antimicrobial-bacterial combinations, indicating the complexity and inherent limitations of statistical modelling of AMR data. Moreover, the methodology of susceptibility testing of H. parasuis and A. suis has not been standardized by CLSI or EUCAST (Brogden et al., 2018; Prüller et al., 2017). Hence, there is an urgent need to establish methodology, epidemiological cut-offs and clinical breakpoints for bacteria relevant to animal health to make clear clinical and epidemiological inferences (Toutain et al., 2017).

of

There are some limitations that should be considered in the context of this study. First,

ro

information on the serovar or presence of virulence factors was not available for any bacteria

-p

which can be related with AMR prevalence. For instance, serovar distribution has been

re

associated with AMR in S. suis and H. parasuis (Yeh et al., 2017; Yongkiettrakul et al., 2019). Second, the information on antimicrobial use in the farm prior to collection of the analyzed

lP

samples was not available. Third, the isolates were mostly from Minnesota and might not be

na

representative of true prevalence of AMR in these swine pathogens in the whole U.S. However, incorporating geographical information in statistical models did not lead to significant changes in

ur

the estimates. Prevalence estimates and regression models built using data from states other than

Jo

Minnesota only also yielded results similar to those presented in the study, indicating that these results can be generalized to whole of the U.S.A. In conclusion, we described the prevalence and changes of AMR in some key bacterial pathogens of great importance to swine medicine in the U.S.A. Results described here will help in surveillance of AMR in these critical swine pathogens and aid in informed decision- making regarding antimicrobial use in swine medicine. Future research work should focus on continuing and strengthening AMR surveillance in these pathogens, as well as establishing standardized methods to test AMR in A. suis and H. parasuis and epidemiological cut-offs and clinical 17

Journal Pre-proof breakpoints for specific bacterial species so that results can be harmonized globally. Additionally, genomic characterization and correlating MIC values with presence/absence of AMR genes in these pathogens can aid in allowing microbiological/genomic inferences to these results in addition to statistical modelling.

Acknowledgments

of

This work was supported by the USDA National Institute of Food and Agriculture (Animal

ro

Health Formula Fund project MIN-62-091) and the Rapid Agricultural Response Fund (RARF)

Conflict of interests

na

Tables

lP

Authors declare no conflict of interests.

re

-p

at the University of Minnesota.

ur

Table 1. Breakpoints* (in g/ml) used for classifying isolates as resistant Caption- *- Isolates with MIC values equal to or more than these breakpoints were considered to

Jo

be “resistant”. All of these breakpoints are recommended by CLSI Vet08 4 th edition (Clinical and Laboratory Standards Institute, 2018), except for breakpoints for gentamicin which is based on epidemiological cut-off provided by EUCAST.

Table 2. MIC distribution frequencies of Streptococcus suis isolates collected at UMN-VDL from 2006-2016. Caption- Gray areas indicate concentrations not tested. Red lines demarcate resistant and notresistant isolates based on swine-specific breakpoints (Clinical and Laboratory Standards 18

Journal Pre-proof Institute, 2018) except trimethoprim-sulphamethoxazole (human-specific breakpoints). * number of isolates tested for susceptibility to this antimicrobial.

Table 3. MIC distribution frequencies of Pasteurella multocida isolates collected at UMN-VDL from 2006-2016. Caption- Gray areas indicate concentrations not tested. Red lines demarcate resistant and not-

of

resistant isolates based on swine-specific breakpoints (Clinical and Laboratory Standards

ro

Institute, 2018) except spectinomycin (cattle-specific breakpoints) and gentamicin (EUCAST

re

-p

epidemiological cut-offs). * - number of isolates tested for susceptibility to this antimicrobial.

Table 4. MIC distribution frequencies of Actinobacillus suis isolates collected at UMN-VDL

lP

from 2006-2016.

na

Captions- Gray areas indicate concentrations not tested. * - number of isolates tested for

ur

susceptibility to this antimicrobial.

from 2010-2016.

Jo

Table 5. MIC distribution frequencies of Haemophilus parasuis isolates collected at UMN-VDL

Captions- Gray areas indicate concentrations not tested. * - number of isolates tested for susceptibility to this antimicrobial.

Figures

19

Journal Pre-proof Figure 1. Changes in prevalence of resistance to select antimicrobials in Streptococcus suis and Pasteurella multocida isolates collected between 2006-2016 Caption- Vertical axis represents percentage of resistant isolates, horizontal axis represents the year of isolation, legend on the right represents antimicrobials-bacterial combinations with statistically significant change (p<0.05) in odds annually (tables S3 and S4). Abbreviations used-

ro

suis (Streptococcus suis) and P. multocida (P. multocida).

of

XNL (ceftiofur), FFN (florfenicol), PEN (penicillin), AMP (ampicillin), SPT (spectinomycin), S.

-p

Figure 2. Changes in MIC distributions over years for select antimicrobial-bacteria combinations.

lP

na

represents the years of isolation.

re

Caption- “MIC” represents minimum inhibitory concentrations in g/ml. Horizontal axis

References

ur

Brant, R., 1990. Assessing Proportionality in the Proportional Odds Model for Ordinal Logistic

Jo

Regression. Biometrics 46, 1171–1178. https://doi.org/10.2307/2532457 Brean, M., Abraham, S., Hebart, M., Kirkwood, R.N., 2016. Influence of parity of birth and suckled sows on piglet nasal mucosal colonization with Haemophilus parasuis. Can. Vet. J. 57, 1281–1283. Brockmeier, S.L., Halbur, P.G., Thacker, E.L., 2002. Porcine Respiratory Disease Complex, in: Brogden, K.A and Guthmiller, J.M., Polymicrobial diseases. ASM Press, Washington D.C, pp.231-258. Brogden, S., Pavlović, A., Tegeler, R., Kaspar, H., De Vaan, N., Kehrenberg, C., 2018. 20

Journal Pre-proof Antimicrobial susceptibility of Haemophilus parasuis isolates from Germany by use of a proposed standard method for harmonized testing. Vet. Microbiol. 217, 32–35. https://doi.org/10.1016/j.vetmic.2018.02.017 Byrt, T., Bishop, J., Carlin, J.B., 1993. Bias, prevalence and kappa. J. Clin. Epidemiol. 46, 423– 429. https://doi.org/10.1016/0895-4356(93)90018-V Callens, B.F., Haesebrouck, F., Maes, D., Butaye, P., Dewulf, J., Boyen, F., 2013. Clinical

of

Resistance and Decreased Susceptibility in Streptococcus suis Isolates from Clinically

ro

Healthy Fattening Pigs. Microb. Drug Resist. 19, 146–151.

-p

https://doi.org/10.1089/mdr.2012.0131

Clinical and Laboratory Standards Institute, 2018. Performance standards for antimicrobial disk

re

and dilution susceptibility tests for bacteria isolated from animals VET08, 4th ed. Clinical

na

Pennsylvania-19087, USA.

lP

and Laboratory Standards Institute, 950 West Valley Road, Suite 2500, Wayne,

Couto, N., Monchique, C., Belas, A., Marques, C., Gama, L.T., Pomba, C., 2016. Trends and

ur

molecular mechanisms of antimicrobial resistance in clinical Staphylococci isolated from

Jo

companion animals over a 16 year period. J. Antimicrob. Chemother. 71, 1479–1487. https://doi.org/10.1093/jac/dkw029 Cusack, T.P., Ashley, E.A., Ling, C.L., Rattanavong, S., Roberts, T., Turner, P., Wangrangsimakul, T., Dance, D.A.B., 2019. Impact of CLSI and EUCAST breakpoint discrepancies on reporting of antimicrobial susceptibility and AMR surveillance. Clin. Microbiol. Infect. 25, 910–911. https://doi.org/10.1016/j.cmi.2019.03.007 Dayao, D.A.E., Gibson, J.S., Blackall, P.J., Turni, C., 2014a. Antimicrobial resistance in bacteria associated with porcine respiratory disease in Australia. Vet. Microbiol. 171, 232–235. 21

Journal Pre-proof https://doi.org/10.1016/j.vetmic.2014.03.014 Dayao, D.A.E., Kienzle, M., Gibson, J.S., Blackall, P.J., Turni, C., 2014b. Use of a proposed antimicrobial susceptibility testing method for Haemophilus parasuis. Vet. Microbiol. 172, 586–589. https://doi.org/10.1016/j.vetmic.2014.06.010 de la Fuente, A.J.M., Tucker, A.W., Navas, J., Blanco, M., Morris, S.J., Gutiérrez-Martín, C.B., 2007. Antimicrobial susceptibility patterns of Haemophilus parasuis from pigs in the

of

United Kingdom and Spain. Vet. Microbiol. 120, 184–191.

ro

https://doi.org/10.1016/j.vetmic.2006.10.014

-p

El Garch, F., de Jong, A., Simjee, S., Moyaert, H., Klein, U., Ludwig, C., Marion, H., HaagDiergarten, S., Richard-Mazet, A., Thomas, V., Siegwart, E., 2016. Monitoring of

re

antimicrobial susceptibility of respiratory tract pathogens isolated from diseased cattle

lP

and pigs across Europe, 2009–2012: VetPath results. Vet. Microbiol. 194, 11–22.

na

https://doi.org/10.1016/j.vetmic.2016.04.009 Fablet, C., Marois-Créhan, C., Simon, G., Grasland, B., Jestin, A., Kobisch, M., Madec, F., Rose,

ur

N., 2012. Infectious agents associated with respiratory diseases in 125 farrow-to-finish

Jo

pig herds: A cross-sectional study. Vet. Microbiol. 157, 152–163. https://doi.org/10.1016/j.vetmic.2011.12.015 Food and Drug Administration, 2019a. Code of Federal Regulations - Title 21 - Food and Drugs [WWW Document]. URL https://www.fda.gov/medical-devices/medical-devicedatabases/code-federal-regulations-title-21-food-and-drugs (accessed 10.19.19). Food and Drug Administration, 2019b. NARMS now, https://www.fda.gov/animalveterinary/national-antimicrobial-resistance- monitoring-system/narms-now- integrateddata (accessed 10.19.19) 22

Journal Pre-proof Fox-Lewis, A., Takata, J., Miliya, T., Lubell, Y., Soeng, S., Sar, P., Rith, K., McKellar, G., Wuthiekanun, V., McGonagle, E., Stoesser, N., Moore, C.E., Parry, C.M., Turner, C., Day, N.P.J., Cooper, B.S., Turner, P., 2018. Antimicrobial Resistance in Invasive Bacterial Infections in Hospitalized Children, Cambodia, 2007–2016. Emerg. Infect. Dis. 24, 841– 851. https://doi.org/10.3201/eid2405.171830 Hernandez-Garcia, J., Wang, J., Restif, O., Holmes, M.A., Mather, A.E., Weinert, L.A., Wileman,

of

T.M., Thomson, J.R., Langford, P.R., Wren, B.W., Rycroft, A., Maskell, D.J., Tucker,

ro

A.W., 2017. Patterns of antimicrobial resistance in Streptococcus suis isolates from pigs

-p

with or without streptococcal disease in England between 2009 and 2014. Vet. Microbiol. 207, 117–124. https://doi.org/10.1016/j.vetmic.2017.06.002

re

Jiang, X., Yang, Y., Zhu, L., Gu, Y., Shen, H., Shan, Y., Li, X., Wu, J., Fang, W., 2016. Live

lP

Streptococcus suis type 5 strain XS045 provides cross-protection against infection by

na

strains of types 2 and 9. Vaccine 34, 6529–6538. https://doi.org/10.1016/j.vaccine.2016.05.003

ur

Kucerova, Z., Hradecka, H., Nechvatalova, K., Nedbalcova, K., 2011. Antimicrobial

Jo

susceptibility of Actinobacillus pleuropneumoniae isolates from clinical outbreaks of porcine respiratory diseases. Vet. Microbiol. 150, 203–206. https://doi.org/10.1016/j.vetmic.2011.01.016 Lapointe, L., D’Allaire, S., Lebrun, A., Lacouture, S., Gottschalk, M., 2002. Antibody response to an autogenous vaccine and serologic profile for Streptococcus suis capsular type 1/2. Can. J. Vet. Res. 66, 8–14. Li, L.-L., Liao, X.-P., Sun, J., Yang, Y.-R., Liu, B.-T., Yang, S.-S., Zhao, D.-H., Liu, Y.-H., 2012. Antimicrobial Resistance, Serotypes, and Virulence Factors of Streptococcus suis Isolates 23

Journal Pre-proof from Diseased Pigs. Foodborne Path. Dis. 9, 583–588. https://doi.org/10.1089/fpd.2011.1106 Lin, W.-H., Shih, H.-C., Lin, C.-F., Yang, C.-Y., Chang, Y.-F., Lin, C.-N., Chiou, M.-T., 2018. Molecular serotyping of Haemophilus parasuis isolated from diseased pigs and the relationship between serovars and pathological patterns in Taiwan. PeerJ 6, e6017. https://doi.org/10.7717/peerj.6017

of

Macedo, N., Rovira, A., Torremorell, M., 2015. Haemophilus parasuis: infection, immunity and

ro

enrofloxacin. Vet. Res. 46, 128. https://doi.org/10.1186/s13567-015-0263-3

-p

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

re

MacInnes, J.I., Gottschalk, M., Lone, A.G., Metcalf, D.S., Ojha, S., Rosendal, T., Watson, S.B.,

lP

Friendship, R.M., 2008. Prevalence of Actinobacillus pleuropneumoniae, Actinobacillus

na

suis, Haemophilus parasuis, Pasteurella multocida, and Streptococcus suis in representative Ontario swine herds. Can. J. Vet. Res. 72, 242–248.

ur

Magstadt, D.R., Schuler, A.M., Coetzee, J.F., Krull, A.C., OC., Or, A.M., Coetzee, J.F.,

Jo

Engelken, T.J., 2018. Treatment history and antimicrobial susceptibility results for Mannheimia haemolytica, Pasteurella multocida, and Histophilus somni isolates from bovine respiratory disease cases submitted to the Iowa State University Veterinary Diagnostic Laboratory from 2013 to 2015. J. Vet. Diagn. Investig. Off. Publ. Am. Assoc. Vet. Lab. Diagn. Inc 30, 99–104. https://doi.org/10.1177/1040638717737589 O’Dea, M.A., Laird, T., Abraham, R., Jordan, D., Lugsomya, K., Fitt, L., Gottschalk, M., Truswell, A., Abraham, S., 2018. Examination of Australian Streptococcus suis Isolates From Clinically Affected Pigs in a Global Context and the Genomic Characterisation of 24

Journal Pre-proof ST1 as a Predictor of Virulence (preprint). Microbiology. https://doi.org/10.1101/375543 Oliveira Filho, J.X. de, Morés, M.A.Z., Rebellato, R., Kich, J.D., Cantão, M.E., Klein, C.S., Guedes, R.M.C., Coldebella, A., Barcellos, D.E.S.N. de, Morés, N., 2018. Pathogenic variability among Pasteurella multocida type A isolates from Brazilian pig farms. BMC Vet. Res. 14, 244. https://doi.org/10.1186/s12917-018-1565-2 Opriessnig, T., Giménez- Lirola, L.G., Halbur, P.G., 2011. Polymicrobial respiratory disease in

of

pigs. Anim. Health. Res. Rev. 12, 133–148. https://doi.org/10.1017/S1466252311000120

ro

Prüller, S., Turni, C., Blackall, P.J., Beyerbach, M., Klein, G., Kreienbrock, L., Strutzberg-

-p

Minder, K., Kaspar, H., Meemken, D., Kehrenberg, C., 2017. Towards a Standardized Method for Broth Microdilution Susceptibility Testing of Haemophilus parasuis. J. Clin.

re

Microbiol. 55, 264–273. https://doi.org/10.1128/JCM.01403-16

lP

Schwarz, S., Silley, P., Simjee, S., Woodford, N., van Duijkeren, E., Johnson, A.P., Gaastra, W.,

na

2010. Editorial: Assessing the antimicrobial susceptibility of bacteria obtained from animals. J. Antimicrob. Chemother. 65, 601–604. https://doi.org/10.1093/jac/dkq037

ur

Sweeney, M.T., Lindeman, C., Johansen, L., Mullins, L., Murray, R., Senn, M.K., Bade, D.,

Jo

Machin, C., Kotarski, S.F., Tiwari, R., Watts, J.L., 2017. Antimicrobial susceptibility of Actinobacillus pleuropneumoniae, Pasteurella multocida, Streptococcus suis, and Bordetella bronchiseptica isolated from pigs in the United States and Canada, 2011 to 2015. J. Swine Health Prod. 25, 15. Tang, X., Zhao, Z., Hu, J., Wu, B., Cai, X., He, Q., Chen, H., 2009. Isolation, Antimicrobial Resistance, and Virulence Genes of Pasteurella multocida Strains from Swine in China. J. Clin. Microbiol. 47, 951–958. https://doi.org/10.1128/JCM.02029-08 Toutain, P.-L., Bousquet-Mélou, A., Damborg, P., Ferran, A.A., Mevius, D., Pelligand, L., 25

Journal Pre-proof Veldman, K.T., Lees, P., 2017. En Route towards European Clinical Breakpoints for Veterinary Antimicrobial Susceptibility Testing: A Position Paper Explaining the VetCAST Approach. Front. Microbiol. 8. https://doi.org/10.3389/fmicb.2017.02344 van Hout, J., Heuvelink, A., Gonggrijp, M., 2016. Monitoring of antimicrobial susceptibility of Streptococcus suis in the Netherlands, 2013–2015. Vet. Microbiol. 194, 5–10. https://doi.org/10.1016/j.vetmic.2016.03.014

of

Varela, N.P., Gadbois, P., Thibault, C., Gottschalk, M., Dick, P., Wilson, J., 2013. Antimicrobial

ro

resistance and prudent drug use for Streptococcus suis. Anim. Health. Res. Rev. 14, 68–

-p

77. https://doi.org/10.1017/S1466252313000029

Vela, A.I., Moreno, Miguel.A., Cebolla, J.A., González, S., Latre, M.V., Domínguez, L.,

re

Fernández-Garayzábal, J.F., 2005. Antimicrobial susceptibility of clinical strains of

lP

Streptococcus suis isolated from pigs in Spain. Vet. Microbiol. 105, 143–147.

na

https://doi.org/10.1016/j.vetmic.2004.10.009 Vötsch, D., Willenborg, M., Weldearegay, Y.B., Valentin-Weigand, P., 2018. Streptococcus suis –

ur

The “Two Faces” of a Pathobiont in the Porcine Respiratory Tract. Front. Microbiol. 9.

Jo

https://doi.org/10.3389/fmicb.2018.00480 Wisselink, H.J., Veldman, K.T., Van den Eede, C., Salmon, S.A., Mevius, D.J., 2006. Quantitative susceptibility of Streptococcus suis strains isolated from diseased pigs in seven European countries to antimicrobial agents licenced in veterinary medicine. Vet. Microbiol. 113, 73–82. https://doi.org/10.1016/j.vetmic.2005.10.035 Yeh, J.-C., Lo, D.-Y., Chang, S.-K., Chou, C.-C., Kuo, H.-C., 2017. Antimicrobial susceptibility, serotypes and genotypes of Pasteurella multocida isolates associated with swine pneumonia in Taiwan. Vet. Rec. 181, 323–323. https://doi.org/10.1136/vr.104023 26

Journal Pre-proof Yongkiettrakul, S., Maneerat, K., Arechanajan, B., Malila, Y., Srimanote, P., Gottschalk, M., Visessanguan, W., 2019. Antimicrobial susceptibility of Streptococcus suis isolated from diseased pigs, asymptomatic pigs, and human patients in Thailand. BMC Vet. Res. 15, 5. https://doi.org/10.1186/s12917-018-1732-5 Zhao, Y., Wang, Q., Li, J., Lin, X., Huang, X., Fang, B., 2018. Epidemiology of Haemophilus parasuis isolates from pigs in China using serotyping, antimicrobial susceptibility,

of

biofilm formation and ERIC-PCR genotyping. PeerJ 6, e5040.

ro

https://doi.org/10.7717/peerj.5040

-p

Zhou, X., Xu, X., Zhao, Y., Chen, P., Zhang, X., Chen, H., Cai, X., 2010. Distribution of antimicrobial resistance among different serovars of Haemophilus parasuis isolates. Vet.

Jo

ur

na

lP

re

Microbiol. 141, 168–173. https://doi.org/10.1016/j.vetmic.2009.05.012

27

Journal Pre-proof Table 1. Breakpoints* (in g/ml) used for classifying isolates as resistant Streptococcus Pasteurella Actinobacillus Haemophilus multocida

suis

parasuis

Ampicillin

2

2

NA

NA

Ceftiofur

8

8

NA

NA

Chlortetracycline

2

2

NA

NA

Enrofloxacin

2

1

NA

NA

Florfenicol

8

8

NA

Gentamicin

NA

16

NA

NA

Oxytetracycline

2

2

Penicillin

1

1

Spectinomycin

NA

Sulphadimethoxine

NA

Tiamulin

NA

Trimethoprim-

4

re

NA

>64

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

NA

64

NA

NA

ur

NA

-p NA

lP

NA

na

Tulathromycin

NA

NA

Jo

Sulphamethoxazole

ro

suis

of

Antimicrobial

*- Isolates with MIC values equal to or more than these breakpoints were considered to be “resistant”. All of these breakpoints are recommended by CLSI Vet08 4 th edition (CLSI, 2018), except for breakpoints for gentamicin which is based on epidemiological cut-off provided by EUCAST.

28

Journal Pre-proof Table 2. MIC distribution frequencies of Streptococcus suis isolates collected at UMN-VDL from 20062016. Antimicrobia

0.

0.

0.

l (n)*

12

25

5

1

2

4

8

16

32

6

12

25

51

%

MI

MI

4

8

6

2

resista

C50

C90

≤0.

≤0.

25

25

≤0.

≤0.

5

5

95.0

>8

>8

1.4

0.5

0.5

1.0

2

2

-

4

8

95.4

>8

>8

6.9

≤0.

0.5

98

1.

0.

0.

0.

0.

0.

(6701)

.2

0

2

1

2

1

2

Ceftiofur

93

3.

1.

0.

0.

0.

(6702)

.9

1

4

6

4

5

Chlortetracy

5.

0.

2.

13

78

cline (6702)

0

7

0

.4

35

51

5.

0.

0.

(6074)

2

.4

.0

9

8

7

Florfenicol

0.

3.

20

68

(6693)

9

4

.0

.4

8.

ro

3

3

28

46

14

1.

1.

ur

4

na

0.

7

1

.0

.4

.6

5

(7025)

ine (6702)

6.

Jo

0.

Gentamicin

Oxytetracycl

re

6.

9.3

.6

lP

Enrofloxacin

0.6

of

Ampicillin

-p

nce

2.

1.

0.

1.

4.

89

9

7

7

1

3

.3

Penicillin

86

3.

3.

2.

2.

0.

0.

0.

(6702)

.6

4

2

5

7

9

4

3

Spectinomyc

6.

51

29

2.

10

in (6702)

7

.0

.4

8

.0

12

Sulphadimet

41

58

hoxine

.2

.8

29

-

16

>64

-

-

-

Journal Pre-proof (6694) Tiamulin

20

37

13

13

1.

2.

4.

6.

(6690)

.6

.2

.8

.4

7

7

4

2

Trimethopri

98

2.

m-

.0

0

-

1

2

Sulphametho

of

xazole (6694)

ro

Gray areas indicate concentrations not tested. Red lines demarcate resistant and not-resistant isolates

-p

based on swine-specific breakpoints (CLSI, 2018) except trimethoprim-sulphamethoxazole (human-

Jo

ur

na

lP

re

specific breakpoints). * - number of isolates tested for susceptibility to this antimicrobial.

30

32

Journal Pre-proof Table 3. MIC distribution frequencies of Pasteurella multocida isolates collected at UMN-VDL from 2006-2016. Antimicrobia

0.

0.

0.

l (n)*

12

25

5

1

2

4

8

16

32

64

12 25

51

%

MI

MI

8

2

resista

C50

C90

≤0.

0.5

6

Ampicillin

84

8.

2.

1.

0.

0.

0.

1.

(8237)

.8

5

4

2

6

4

4

6

Ceftiofur

96

1.

0.

0.

0.

0.

(8240)

.8

4

5

3

4

7

Chlortetracy

33

26

23

11

3.

1.

cline (8240)

.8

.6

.0

.7

5

-p

nce

3.

2.

0.

0.

0.

(7264)

.9

2

1

5

2

2

Florfenicol

21

59

11

5.

(8240)

.5

.6

.6

7

37

Oxytetracycl ine (8241)

of

0.

6

na

2

8

48

11

1.

0.

1.

.0

.1

ur

.3

7

9

0

Jo

(8241)

ro

0.

Gentamicin

0.

≤0.

≤0.

5

5

39.6

1

4

0.9

≤0.

≤0.

12

12

1.0

0.5

1

1.9

2

4

61.8

2

>8

5.0

≤0.

0.25

1.1

re

93

25

4

lP

Enrofloxacin

4.3

24

13

26

8.

2.

24

.3

.9

.1

9

8

.1

Penicillin

77

14

2.

1.

1.

0.

0.

1.

(8241)

.4

.8

8

4

0

4

5

8

Spectinomyc

14

32

43

7.

2.

in (8241)

.6

.3

.9

0

2

12 2.2

Sulphadimet

29

70

hoxine

.3

.7

31

-

32

32

Journal Pre-proof (8236) Tiamulin

2.

8.

31

43

14

(8237)

3

4

.2

.6

.6

Trimethopri

95

4.

m-

.8

2

-

32

>32

≤1

4

-

Sulphametho

(8236)

of

xazole

54

23

12

3.

0.

0.

0.

1.

in (6897)

.2

.0

.6

7

8

7

7

4

2.1

-p

ro

Tulathromyc

re

Gray areas indicate concentrations not tested. Red lines demarcate resistant and not-resistant isolates based on swine-specific breakpoints (CLSI, 2018) except spectinomycin (cattle-specific breakpoints) and

lP

gentamicin (EUCAST epidemiological cut-offs). * - number of isolates tested for susceptibility to this

Jo

ur

na

antimicrobial.

32

Journal Pre-proof Table 4. MIC distribution frequencies of Actinobacillus suis isolates collected at UMN-VDL from 20062016. Antimicrobial

0.1

0.2

(n)*

2

5

Ampicillin

83.

(2846)

5

Ceftiofur

93.

(2551)

8

0.5

9.2

1

2

1.5

0.7

7

2

3.0

1.6

0.5

0.3

8.6

72.

15.

3.1

0

3

Penicillin (2850)

49.

Jo

e (2849)

61.

33.

3

1

ur

(2850) Oxytetracyclin

6.1

12

25

51

MIC

MIC

4

8

6

2

50

90

≤0.2

0.5

2.9

20.

1

1

10.

52.

22.

7

6

9

13.

7.7

0.9

0.2

2.0

5

1

4

0.12

1

≤1

2

1

>8

0.5

1

32

32

16

16

0.1

0.5

0.5

3.7

0.7

0.6

0.5

1.7

27.

0.6

8 2.5

0.3

0.5

4.4

2.4

14.

75.

5.

2.

3

9

2

2

Sulphadimetho

79.

20.

xine (2847)

7

3

0.1

5

0.4

(2849)

0.3

≤0.2

0.12

0.2

Spectinomycin

Tiamulin

≤0.2

-p

Gentamicin

0.1

6

na

(2849)

8.7

0.5

lP

Florfenicol

0.9

6

of

1.3

ne (2849)

3

0.7

32

ro

1.5

30.

(2594)

0.6

16

re

2.7

37.

94.

8

5

Chlortetracycli

Enrofloxacin

4

0.2

1.2

42. 33

50.

4.6

0.

Journal Pre-proof (2843) Trimethoprim-

98.

Sulphamethoxa

6

1

6

2.6

0.5

8

1.4

zole (2848) Tulathromycin

20.

61.

11.

(2417)

8

4

3

0.3

0.

2.

7

4

ur

na

lP

re

-p

ro

- number of isolates tested for susceptibility to this antimicrobial.

Jo

*

of

Gray areas indicate concentrations not tested.

34

2

4

Journal Pre-proof Table 5. MIC distribution frequencies of Haemophilus parasuis isolates collected at UMN-VDL from 2010-2016.

5

Ampicillin

46.

(1615)

9

Ceftiofur

82.

(1615)

9

9.5

5.6

Chlortetracycli

82.

ne (1615)

4

Enrofloxacin

94.

(1615)

9

2.7

1.6

58.

34.

(1615)

5

4

11.

10.

3

3

3.7

1.2

Gentamicin

7.4

2.1

0.7

0.4

0.9

32

0.6

0.4

0.2

0.3

1.3

0.1

0.1

8

2

4

9

15.

13.

5.1

4.5

3

7

0 5.9

1.1

1.2

ur 38.

17.

14.

10.

(1615)

6

5

8

8

25

51

MIC

MIC

8

6

2

50

90

0.5

>16

≤0.2

1

0.3

5

0.1

3.0

≤0.1

2

2

≤0.2

0.5

8.4

≤0.5

8

10.

0.25

>8

0.8

≤8.0

32.0

4

8

2 12.

(1615)

8

8

7.4

3.

2.

9

1

Sulphadimetho

45.

55.

xine (1615)

0

0

12.

≤0.1

8

73.

7.7

1

4

Spectinomycin

Tiamulin

≤0.5

5 15.

Penicillin

12

5.6

1.2

5.6

64

11.

6.6

33.

Jo

16

5

24.

53.

e (1658)

8

22.

(1615) Oxytetracyclin

4

9.0

na

Florfenicol

2

of

2

1

ro

(n)*

0.5

-p

0.2

re

0.1

lP

Antimicrobial

21.

27.

22. 35

7.5

1.0

0.

Journal Pre-proof (1615)

2

5

7

Trimethoprim-

86.

13.

Sulphamethoxa

1

9

0

5

zole (1615) Tulathromycin

21.

16.

21.

24.

(1615)

3

7

0

0

9.4

4.2

1.

1.

5

9

ur

na

lP

re

-p

ro

- number of isolates tested for susceptibility to this antimicrobial.

Jo

*

of

Gray areas indicate concentrations not tested.

36

4

16

Journal Pre-proof Figure 1. Changes in prevalence of resistance to select antimicrobials in Streptococcus suis and Pasteurella multocida isolates collected between 2006-2016

10

8

XNL-S. suis FFN-S. suis

PEN-S. suis

of

6

AMP-P. multocida XNL-P. multocida

ro

4

PEN-P. multocida SPT-P. multocida

-p

2

0 2006

2007

2008

2009

2010

re

Percentage of resistance to different antimicrobials

12

2011

2012

2013

2014

2015

2016

lP

Year

na

Vertical axis represents percentage of resistant isolates, horizontal axis represents the year of isolation, legend on the right represents antimicrobials -bacterial combinations with statistically significant change (p<0.05) in odds annually

ur

(tables S3 and S4). Abbreviations used-XNL (ceftiofur), FFN (florfenicol), PEN (penicillin), AMP (ampicillin), SPT

Jo

(spectinomycin), S. suis (Streptococcus suis) and P. multocida (P. multocida).

37

Journal Pre-proof Figure 2. Changes in MIC distributions over years for select bacterial-antimicrobial combinations. a) Gentamicin- S. suis

90 80

MIC

70

>16

60

4

30

2 ≤1

20 10

60 50 40 30 20

ur

70

na

90

Jo

Distribution of MIC values per 100 isolates

100

lP

b) Ceftiofur- A. suis

re

0

80

ro

8

40

of

16

50

-p

Distribution of MIC values per 100 isolates

100

MIC >8 8

4 2 1

0.5 ≤0.25

10 0

38

Journal Pre-proof c) Spectinomycin- A. suis

90 80

MIC

70 >64

60

64

50

32

40

16 30

≤8

20 10

of

Distribution of MIC values per 100 isolates

100

-p

ro

0

d) Tiamulin- A. suis

re

90 80

lP

MIC

70

>32

60

30 20 10 0

na

40

32

ur

50

16 8

≤4

Jo

Distribution of MIC values per 100 isolates

100

39

Journal Pre-proof e) Ampicillin- H. parasuis

90

MIC

80

>16

70

16

60

8

50

4

2

40

1

30

0.5 20

≤0.25

10

ro

0

-p

2010 2011 2012 2013 2014 2015 2016

f) Florfenicol- H. parasuis

re

100 90

MIC

lP

80

>8

70

8

60

30 20

10 0

na

40

4

ur

50

2 1

0.5 ≤0.25

Jo

Distribution of MIC values per 100 isolates

of

Distribution of MIC values per 100 isolates

100

2010 2011 2012 2013 2014 2015 2016

40

Journal Pre-proof g) Gentamicin- H. parasuis

90

MIC

80 70

>16

60

16

50

8

4

40

2

30

≤1 20

of

Distribution of MIC values per 100 isolates

100

10

ro

0

-p

2010 2011 2012 2013 2014 2015 2016

h) Penicillin- H. parasuis

re

90

MIC

80

lP

>8

70

8 4

40 30 20

10 0

ur

50

na

60

2

1 0.5 0.25

≤0.12

Jo

Distribution of MIC values per 100 isolates

100

2010 2011 2012 2013 2014 2015 2016

41

Journal Pre-proof i) Ceftiofur-H. parasuis

90

MIC

80 >8

70

8 60

4

50

2

40

1

30

0.5

20

≤0.25

of

Distribution of MIC values per 100 isolates

100

10

ro

0

-p

2010 2011 2012 2013 2014 2015 2016

j) Enrofloxacin-H. parasuis

re

90

MIC

lP

80 70

>2 2

40 30 20

10 0

ur

50

na

60

1

0.5 0.25 ≤0.12

Jo

Distribution of MIC values per 100 isolates

100

2010 2011 2012 2013 2014 2015 2016

42

Journal Pre-proof k) Tulathromycin-H. parasuis

90

MIC

80

>64

70

64

60

32

50

16

8

40

4

30

2 20

≤1

of

Distribution of MIC values per 100 isolates

100

10

ro

0 2010 2011 2012 2013 2014 2015 2016

-p

“MIC” represents minimum inhibitory concentrations in g/ml.

Jo

ur

na

lP

re

Horizontal axis represents the years of isolation.

43

Jo

ur

na

lP

re

-p

ro

of

Journal Pre-proof

44

Journal Pre-proof Highlights AMR data for bacterial swine pathogens associated with SDRC are lacking



Review of data from a 11-year dataset revealed low levels of AMR for most pathogens



AMR in S. suis and P. multocida isolates mostly remained low.



There were statistically significant changes in AMR in A. suis and H. parasuis



Analysis of MIC data for antimicrobial-bacteria with no breakpoints revealed significant changes (decreases and increases) in AMR over time

of



Use of surrogate breakpoints can lead to different AMR estimates for certain bacteria



Monitoring of AMR in animal pathogens is important to assess the impact of policy

-p

ro



Jo

ur

na

lP

re

changes regulating antimicrobial use in food animals.

45

Figure 1

Figure 2A

Figure 2B