Antimicrobial susceptibility of enteric bacteria recovered from feedlot cattle administered chlortetracycline in feed

Tammy M. Platt Feedlot Research Group, Department of Agricultural Sciences, College of Agriculture, Science and Engineering, West Texas A&M University, Canyon, TX 79016.

Search for other papers by Tammy M. Platt in
Current site
Google Scholar
PubMed
Close
 MS
,
Guy H. Loneragan Feedlot Research Group, Department of Agricultural Sciences, College of Agriculture, Science and Engineering, West Texas A&M University, Canyon, TX 79016.

Search for other papers by Guy H. Loneragan in
Current site
Google Scholar
PubMed
Close
 BVSc, PhD
,
H. Morgan Scott Department of Veterinary Integrative Biosciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX 77843.

Search for other papers by H. Morgan Scott in
Current site
Google Scholar
PubMed
Close
 DVM, PhD
,
Bo Norby Department of Veterinary Integrative Biosciences, College of Veterinary Medicine and Biomedical Sciences, Texas A&M University, College Station, TX 77843.

Search for other papers by Bo Norby in
Current site
Google Scholar
PubMed
Close
 CMV, MPVM, PhD
,
Daniel U. Thomson Department of Clinical Sciences, College of Veterinary Medicine, Kansas State University, Manhattan, KS 66506.

Search for other papers by Daniel U. Thomson in
Current site
Google Scholar
PubMed
Close
 PhD, DVM
,
Michel S. Brown Feedlot Research Group, Department of Agricultural Sciences, College of Agriculture, Science and Engineering, West Texas A&M University, Canyon, TX 79016.

Search for other papers by Michel S. Brown in
Current site
Google Scholar
PubMed
Close
 PhD
,
Samuel E. Ives Cactus Feeders, 2209 W 7th St, Amarillo, TX 79106.

Search for other papers by Samuel E. Ives in
Current site
Google Scholar
PubMed
Close
 DVM, PhD
, and
Mindy M. Brashears International Center for Food Industry Excellence, Department of Animal and Food Sciences, College of Agricultural Sciences and Natural Resources, Texas Tech University, Lubbock, TX 79409.

Search for other papers by Mindy M. Brashears in
Current site
Google Scholar
PubMed
Close
 PhD

Abstract

Objective—To evaluate administration of chlortetracycline in feed of cattle as a method to select for tetracycline resistance among enteric bacteria in feedlot settings.

Animals—20 steers.

Procedures—Steers were randomly assigned to an exposed cohort (n = 10) or an unexposed cohort (control cohort; 10). Chlortetracycline (22 mg/kg) in cottonseed meal was administered to the exposed cohort on days 0 through 4, 6 through 10, and 12 through 16. The control cohort was administered only cottonseed meal. Fecal samples were collected from 16 steers on days −7, 0, 2, 6, 8, 12, 14, 19, 22, 26, and 33, and Escherichia coli and Enterococcus spp were isolated. Minimum inhibitory concentration (MIC) of selected antimicrobials was estimated.

Results—Overall, 56.0% and 31.4% of E coli and Enterococcus isolates, respectively, were resistant to tetracycline. Exposure to chlortetracycline was associated with a significant temporary increase in log2 MIC for both genera but returned to preexposure values by day 33. Averaged across time, the proportion of tetracycline-resistant E coli and Enterococcus isolates was significantly greater in exposed than in unexposed steers. Although all ceftiofur-resistant E coli isolates were coresistant to tetracycline, exposure to chlortetracycline led to a significant decrease in the proportion of E coli resistant to ceftiofur during exposure.

Conclusions and Clinical Relevance—Exposure to chlortetracycline was associated with a temporary increase in the likelihood of recovering resistant bacteria. Exposure to chlortetracycline decreased the likelihood of recovering ceftiofur-resistant E coli isolates, even though isolates were coresistant to tetracycline. These findings warrant further investigation.

Abstract

Objective—To evaluate administration of chlortetracycline in feed of cattle as a method to select for tetracycline resistance among enteric bacteria in feedlot settings.

Animals—20 steers.

Procedures—Steers were randomly assigned to an exposed cohort (n = 10) or an unexposed cohort (control cohort; 10). Chlortetracycline (22 mg/kg) in cottonseed meal was administered to the exposed cohort on days 0 through 4, 6 through 10, and 12 through 16. The control cohort was administered only cottonseed meal. Fecal samples were collected from 16 steers on days −7, 0, 2, 6, 8, 12, 14, 19, 22, 26, and 33, and Escherichia coli and Enterococcus spp were isolated. Minimum inhibitory concentration (MIC) of selected antimicrobials was estimated.

Results—Overall, 56.0% and 31.4% of E coli and Enterococcus isolates, respectively, were resistant to tetracycline. Exposure to chlortetracycline was associated with a significant temporary increase in log2 MIC for both genera but returned to preexposure values by day 33. Averaged across time, the proportion of tetracycline-resistant E coli and Enterococcus isolates was significantly greater in exposed than in unexposed steers. Although all ceftiofur-resistant E coli isolates were coresistant to tetracycline, exposure to chlortetracycline led to a significant decrease in the proportion of E coli resistant to ceftiofur during exposure.

Conclusions and Clinical Relevance—Exposure to chlortetracycline was associated with a temporary increase in the likelihood of recovering resistant bacteria. Exposure to chlortetracycline decreased the likelihood of recovering ceftiofur-resistant E coli isolates, even though isolates were coresistant to tetracycline. These findings warrant further investigation.

As a result of their selective antibacterial properties, antimicrobials provide selection pressure that favors bacteria that survive therapeutic concentrations of specific drugs. This survival-of-the-resistant phenomenon contributes to the emergence and dissemination of drug-resistant bacteria and can pose therapeutic challenges for effective control of diseases and treatment of animals infected by bacteria.1–4 In terms of public health, resistant bacteria are associated with an increase in the duration of illness and risk of death, and higher health-care costs are associated with the illness.5–7 Resistant bacteria cause therapeutic challenges for human and animal health-care providers, but in addition, an increasing body of literature suggests that some resistant variants may be more virulent in that they are associated with an increased severity of illness, compared with the severity of illness resulting from pansusceptible variants.8,9

Animal agriculture is frequently blamed for the emergence of drug-resistant bacteria.3,6,10,11 For example, administration of subtherapeutic doses of specific antimicrobials has been prohibited in countries of the European Union.3,4,12,13 Ostensibly, if antimicrobial drug use were the sole factor that facilitated emergence, dissemination, and maintenance of resistant bacteria, then such a ban would appear logical. However, there is little evidence that prohibiting subtherapeutic administration of antimicrobials in animal agriculture in countries of the European Union has resulted in detectable improvements in human health or even of susceptibility among bacteria that cause diseases in humans.14 Of additional concern is the fact that evidence exists to indicate that as a consequence of the ban, animal health (and thereby well-being) has deteriorated because a larger number of therapeutic doses of antimicrobials have been administered since the prohibition.15,16 Although we are not arguing for or against the benefits of a prohibition of subtherapeutic administration of drugs, it appears that emergence, dissemination, and maintenance of the determinants of antimicrobial resistance are more complex than has been believed. To better evaluate these complexities, effective methods for examining emergence, dissemination, and maintenance of antimicrobial resistance in real-world settings are needed. Ideally, these methods would enable researchers to obtain a better understanding of factors associated with and potential mitigation strategies against antimicrobial resistance within the complex microbial consortia in the gastrointestinal tract of food-producing animals.

Tetracycline resistance among Escherichia coli in cattle is relatively common.17–19 In a study20 conducted by our laboratory group, 40.7% of E coli isolates were resistant to tetracycline. Chlortetracycline is approved for inclusion in feed for use in cattle at several doses and for various indications, including treatment of cattle with pneumonia caused by Pasteurella multocida. Because resistance determinants are already widely prevalent among bacteria in cattle populations, administration of chlortetracycline may facilitate expansion of tetracycline resistance within and among bacteria in animals. If chlortetracycline administration proves an effective method for evaluating resistance dissemination, investigators could examine how resistance is disseminated and the potential impact of various intervention strategies targeted at reducing the burden of resistance. Therefore, the objective of the study reported here was to evaluate administration of chlortetracycline in the feed as a method to select for tetracycline resistance among enteric bacteria in cattle in a real-world feedlot setting.

Materials and Methods

Animals—Twenty single-source crossbred steers were enrolled in a cohort study. Steers were allocated by use of a randomization method to 2 treatment cohorts (10 steers/cohort). Each cohort was housed in 2 pens (5 steers/pen). Steers were allowed to adapt to a typical steam-flaked, corn-based feedlot diet for approximately 3 weeks. During the period of adaptation, no injectable antimicrobials were administered. However, all steers were administered tylosin and monensin in the feed at labeled dosages, which was typical of feedlot production practices. The study was reviewed and approved by the Cooperative Research, Education and Extension Triangle Animal Care and Use Committee (protocol No. 2005-01A).

Procedures—After the 3-week adaptation period, steers in the exposed cohort were administered chlortetracyclinea (22 mg/kg) in feed (mixed with cottonseed meal) in accordance with label directions. Steers in the control group received only cottonseed meal. Chlortetracycline was added to approximately 0.4 kg of cottonseed meal in a large stainless-steel bowl and combined by stirring with a handheld mixer for 60 seconds. Approximately 5 minutes after the daily ration was delivered to the feed bunk, the chlortetracycline-cottonseed meal mixture (exposed cohort) or cottonseed meal alone (control cohort) was added on top of the feed remaining in the bunk.

Day of initial administration of chlortetracycline was designated as day 0. Chlortetracycline was administered during 3 separate 5-day periods, with 1 day of no administration of chlortetracycline or cottonseed meal between each 5-day period. Consequently, chlortetracycline was administered on days 0 through 4 (period 1), 6 through 10 (period 2), and 12 through 16 (period 3). To ensure accurate body weights for calculation of the amount of chlortetracycline to be administered, cattle were weighed on days −7, 3, and 10. Mean body weight of steers in each pen was used to determine the appropriate amount of chlortetracycline to be administered in the feed bunk for each pen.

Fecal samples were collected for analysis. Samples were collected by moving steers through an animal-handling facility on days −7, 0, 2, 6, 8, 12, 14, 19, 22, 26, and 33. Fecal samples were manually obtained directly from the rectum of 4 steers in each pen and placed in labeled plastic cups. Fecal samples were collected from the same 4 steers in each pen at each sample collection. Samples were stored in a cooler until transported to the laboratory; all samples were transported to the laboratory on the day of collection.

Microbiologic evaluationEscherichia coli and Enterococcus spp were cultured and isolated from fecal samples by use of standard methods. In brief, 10 g of fecal material was placed into 90 mL of buffered peptone water, and the mixture was shaken for 1 minute. The solution was then streaked onto MacConkey agar (for E coli isolation) and agarb selective for Enterococcus spp and incubated at 37°C for 24 hours. Three separate and distinct colonies were selected from each plate for further characterization. Results of another study20 conducted by our laboratory group indicated that biochemical confirmation of non–type-specific E coli offered little advantage over simply selecting typical 2- to 4-mm, slightly convex, magenta-colored (ie, lactose-positive) colonies surrounded by an area of dark pink (precipitated bile salts) from MacConkey agar inoculated with bovine feces. In that study, 99.9% of typical colonies were confirmed as E coli via a validated, automated biochemical assay.c All Enterococcus colonies were speciated by use of an automated biochemical analyzer.c

Susceptibility testing was performed on all isolates with a panel of antimicrobialsd,e by use of microbroth dilution. Three to 5 clearly distinct colonies were selected from tryptic soy agar plates and placed into 4 mL of sterile deionized water; these solutions were then adjusted to a 0.5-McFarland standard. Ten microliters of the suspension was transferred into Mueller-Hinton broth and mixed by use of a vortexer. An automated inoculatorf was used to inoculate 96-well microbroth dilution plates. Plates were incubated at 37.3°C for 18 to 24 hours; results were then manually determined.

The MIC was reported as the lowest concentration of antimicrobial that inhibited visible growth, as determined in accordance with the manufacturer's instructions. When an isolate grew at the greatest concentration of an antimicrobial, the MIC was arbitrarily set at double the greatest concentration. Isolates were classified as susceptible or resistant on the basis of guidelines established by the Clinical and Laboratory Standards Institute.21 When breakpoints had not been established by the Clinical and Laboratory Standards Institute, breakpoints described by the National Antimicrobial Resistance Monitoring System for enteric bacteria22 were used.

Statistical analysis—Data were entered into an electronic databaseg and examined for errors. Data were then imported into a commercially available statistical analysis program.h Descriptive results were generated and provided graphicallyi or in a tabular format and analyzed by use of mixed-model methods. Linear and logistic regression models were used for the various outcomes (log2-transformed MIC or proportion resistant, respectively). Sample day was included as a main effect in linear regression models; however, because of convergence problems within the logistic regression models, sample periods that most closely aligned with the various exposure periods were constructed (preexposure period [days −7 and 0], exposure period 1 [days 2 and 6], exposure period 2 [days 8 and 12], exposure period 3 [days 14 and 19], and postexposure period [days 22, 26, and 33]) and included as a main effect in logistic regression models. Within-animal dependency was modeled by use of first-order autoregressive and compound symmetry covariance matrices.23 When a significant interaction between exposure and sample day (or sample period) was detected, an analysis of treatment effect within each sample day (or sample period) was performed, as was an analysis of the effect of day (or period) within each exposure. When no significant interaction was detected, main effects were evaluated.

Results

Escherichia coli—Five hundred twenty-five E coli isolates were recovered from 176 fecal samples collected from 16 steers (4 steers/pen). We recovered 3 isolates from 173 samples and 2 isolates from 3 samples. Of the 525 isolates recovered, 128 (24.4%) were pansusceptible to all antimicrobials tested, whereas 397 (75.6%) were resistant to 1 or more antimicrobials (Table 1). Isolates from the control cohort (129/184 [70.2%]) were significantly (P = 0.04) less likely to be resistant to 1 or more microbials, compared with the proportion of resistant isolates from the exposed cohort (173/213 [81.0%]). Of the 525 isolates recovered, 151 (28.8%) were resistant to 1 antimicrobial, 82 (15.6%) were resistant to 2 antimicrobials, 73 (13.9%) were resistant to 3 antimicrobials, and 91 (17.3%) were resistant to 4 or more antimicrobials. Included in the latter were 37 (7.1%) isolates resistant to 8 or more antimicrobials. Testing revealed that 294 (56.0%) isolates were resistant to tetracycline21,22 (Table 2). Resistance to ceftiofur was detected in 37 (7.1%) isolates, all of which were also resistant to tetracycline, and there was almost perfect agreement (κ = 0.83) between ceftiofur resistance and coresistance to ampicillin, chloramphenicol, streptomycin, sulfamethoxazole, and tetracycline. The ACSSuT phenotype was detected in 51 (9.7%) isolates, and 37 (72.6%) of these were also resistant to ceftiofur.

Table 1—

Percentage of antimicrobial-resistant phenotypes for 525 Escherichia coli isolates, on the basis of the number of antimicrobials to which the isolates were resistant.

No. of antimicrobialsPhenotypePercentage of all phenotypes
0Pansusceptible24.4
1Su17.9
T10.7
S0.19
2Su,T10.9
A,T2.1
T, Nal1.5
A,Su0.57
Other phenotypes0.57
3S,Su,T11.8
Nal,Su,T0.57
A,Su,T0.57
A,Aug,Su0.38
Other phenotypes0.57
4C,S,Su,T5.0
A, C,Su,T1.5
A,S,Su,T0.38
5A, C, S, Su, T1.0
A,S,Su,T,TrSu0.38
Other phenotypes0.38
6A, C,Kan,S,Su,T0.19
A, C, Fox, S, Su, T0.19
7A, C,Gen, Kan, S, Su,T0.77
A, Aug, C, Fox, S,Su,T0.57
8A, Aug, C, Cef, Fox, S, Su, T4.0
9A, Aug, C, Cef, Fox, Nal, S, Su, T0.76
A, Aug, Axo, C, Cef, Fox, S, Su, T0.76
A, C, Cef, Fox, Gen, Kan, S, Su, T0.19
11A, Aug, C, Cef, Fox, Gen, Kan, Nal, S, Su, T1.3

A = Ampicillin. Aug = Amoxicillin-clayulanic acid. Axo = Ceftriaxone. C = Chloramphenicol. Cef = Ceftiofur. Fox = Cefoxitin. Gen = Gentamicin. Kan = Kanamycin. Nal = Nalidixic acid. S = Streptomycin. Su = Sulfisoxazole. T = Tetracycline. TrSu = I Trimethoprim-sulfisoxazole.

Table 2—

Percentage of 525 E coli isolates recovered from fecal samples collected from feedlot steers, on the basis of MIC ratio.

AntimicrobialMIC ratio*Lowest concentration tested (μ/mL)
 012345678 
Amikacin0.9527.0061.709.100.950.19010.5
Gentamicin24.0062.308.601.700.380.760.951.300.25
Kanamycin96.000.760.570.382.308.0
Streptomycin71.2010.7018.1032.0
Ampicillin20.2048.009.500.575.500.9515.201 .0
Amoxicillin and clavulanic acid11.1037.1040.202.900.576.901.301.0 and 0.5
Cefoxitin1.105.3039.6037.308.000.760.767.100.5
Ceftiofur6.9040.2043.102.1000.761.505.500.12
Ceftriaxone91.600.3800.3801.504.001.300.760.25
Ciprofloxacin94.500.9500.571.300.382.30000.015
Nalidixic acid0.5710.3074.5010.3000.1904.200.5
Sulfisoxazole12.4010.9016.6060.200016.0
Trimethoprim and sulfamethoxazole79.8017.901.300.190.1900.12 and 2.38
Chloramphenicol2.7038.5040.801.300.1916.602.0
Tetracycline31.2012.801.5014.3040.204.0

For each antimicrobial, the MIC ratio was determined by dividing the MIC by the lowest concentration of antimicrobial tested. This ratio was then logarithmically (log2) transformed to create a comparable index. When an isolate grew at the highest concentration of antimicrobial tested, the MIC was arbitrarily assigned the value of the subsequent highest serial dilution.

Value represents concentration considered resistant.21,22

— = Value greater than the MIC of the assay.d

A significant (P = 0.03) interaction was detected between treatment and sample day for geometric mean MIC of tetracycline. A significant effect of treatment was detected on days −7 (P = 0.02), 2 (P = 0.03), 8 (P = 0.02), and 14 (P = 0.01) on the basis that isolates recovered from the control cohort had a lower MIC than those recovered from the exposed cohort (Figure 1). The interaction between treatment and sample period was not significant (P = 0.48) for the proportion of resistant isolates; however, an effect of treatment was detected in which isolates recovered from the exposed cohort were significantly (P = 0.03) more likely to be resistant than were those from the control cohort. Averaged across time, 42.3% (95% CI, 26.6% to 59.8%) and 68.8% (95% CI, 58.7% to 77.3%) of E coli isolates recovered from the control and exposed cohorts, respectively, were resistant to tetracycline. Sample period did not have a significant (P = 0.08) effect.

Figure 1—
Figure 1—

Geometric mean ± SEM (log2-transformed) MIC of tetracycline (A) and the proportion of isolates resistant to tetracycline (B) for Escherichia coli isolates recovered from fecal samples collected from steers administered chlortetracycline (22 mg/kg [white circles]) in cottonseed meal or nonmedicated cottonseed meal alone (black circles) on days 0 to 4, 6 to 10, and 12 to 16. Day 0 was the first day of chlortetracycline administration.

Citation: American Journal of Veterinary Research 69, 8; 10.2460/ajvr.69.8.988

A significant (P = 0.02) interaction was detected between treatment and day for geometric mean MIC of ceftiofur. On days 6, 8, and 12, isolates recovered from the control cohort had a significantly (P = 0.01) greater MIC than the MIC for isolates recovered from the exposed cohort (Figure 2). The interaction between treatment and sample day was not significant (P = 0.77) for the proportion of isolates resistant to ceftiofur. However, there was a significant (P = 0.01) effect of treatment in that the isolates recovered from the control cohort were more likely to be resistant than those recovered from the exposed cohort, when averaged across time. Averaged across time, 8.6% (95% CI, 3.1% to 21.9%) and 2.9% (95% CI, 0.7% to 10.1%) of E coli isolates recovered from the control and exposed cohorts, respectively, were resistant to ceftiofur.

Figure 2—
Figure 2—

Geometric mean ± SEM (log2-transformed) MIC of ceftiofur (A) and the proportion of isolates resistant to ceftiofur (B) for E coli isolates recovered from fecal samples collected from steers administered chlortetracycline in cottonseed meal or nonmedicated cottonseed meal alone on days 0 to 4, 6 to 10, and 12 to 16. Preexposure samples were collected before onset of chlortetracycline administration (days −7 and 0), exposure samples were collected for each of the 3 administration periods (period 1, days 2 and 6; period 2, days 8 and 12; and period 3, days 14 and 19), and postexposure samples were collected after cessation of chlortetracycline administration (days 22, 26, and 33). See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 69, 8; 10.2460/ajvr.69.8.988

Enterococcus spp—Three hundred sixty-six Enterococcus isolates were recovered from 176 fecal samples. Three isolates were recovered from 78 samples, 2 isolates were recovered from 55 samples, and 1 isolate was recovered from 22 samples. Of the 366 isolates recovered, 330 (90.2%) were resistant to at least 1 antimicrobial or grew at the greatest concentration tested. Overall resistance did not differ significantly (P = 0.98) between exposed and control cohorts, with 170 (90.9%) resistant organisms recovered from the control group and 160 (89.4%) resistant organisms recovered from the exposed group. Of the 366 isolates recovered, 66 (18.0%) were resistant to 1 antimicrobial, 79 (21.6%) were resistant to 2 antimicrobials, 112 (30.6%) were resistant to 3 antimicrobials, 42 (11.5%) were resistant to 4 antimicrobials, and 31 (8.5%) were resistant to 5 or more antimicrobials (Tables 3 and 4). Three isolates (2 Enterococcus gallinarum and 1 Enterococcus casseliflavus) were recovered that were resistant to vancomycin. Of the Enterococcus isolates, 115 (31.4%) were resistant to tetracycline (Table 5). Resistance to tetracycline varied significantly (P = 0.01) by species of Enterococcus organism (Figure 3). For the various species of enterococci, 8 of 79 (10.1%) isolates for E casseliflavus, 0 of 1 (0%) isolate for Enterococcus columbae, 9 of 9 (100%) isolates for Enterococcus durans, 25 of 97 (25.8%) isolates for Enterococcus faecium, 39 of 91 (42.8%) isolates for E gallinarum, and 35 of 89 (39.3%) isolates for Enterococcus hirae were resistant to tetracycline.

Table 3—

Percentages of antimicrobial-resistant phenotypes for 366 Enterococcus isolates, on the basis of the number of antimicrobials to which the isolates were resistant for isolates resistant to 0 to 3 antimicrobials.

No. of antimicrobialsPhenotypePercentage of all phenotypesEnterococcus spp*
0Pansusceptible9.61,4,5,6
1T6.93, 4, 5, 6
N4.65,6
F2.21,4,5
L1.94,5,6
Bac1.41,4
Ery0.824,5
Other phenotypes0.544,5
2L,T4.45,6
L Tylt3.66
F, N2.24,5
Ery,F1.44,5
F,T1.45,6
Ery, N1.44
Bac, L1.11,5
Ery,T0.824,5
Bac, Ery0.821
Bac,T0.824
N,T0.825,6
Syn,T0.556
Other phenotypes1.621,4,5,6
3Bac, Ery, L12.01,5
Ery, L, Tylt7.95,6
F,T, L1.91,4,5,6
T, L, Tylt1.61,4,5,6
Ery, F, T0.824,5
Ery, F, L0.551,4
Dap, F, L0.555,6
Ery, L, N0.554,5
L, N,T0.553,6
Other phenotypes4.321, 2, 4, 5

Enterococcus spp were as follows: 1, Enterococcus casselifla-vus; 2, Enterococcus columbae; 3, Enterococcus durans; 4, Enterococcus faecium; 5, Enterococcus gallinarum; and 6, Enterococcus hirae.

Bac = Bacitracin. Dap = Daptomycin. Ery = Erythromycin. F = Flavomycin. L = Lincomycin. N = Nitrofurantoin. Syn = Quinupristin and dalfopristin. Tylt = Tylosin tartrate.

Table 4—

Percentages of antimicrobial-resistant phenotypes for 366 Enterococcus isolates, on the basis of the number of antimicrobials to which the isolates were resistant for isolates resistant to ≥4 antimicrobials.

No. of antimicrobialsPhenotypePercentage of all phenotypesEnterococcus spp*
4Bac, Ery, F, L1.91
Ery, L, N, Tylt1.16
Ery, F, L, Tylt0.825, 6
Ery, L, Syn, Tylt0.825, 6
F, L, T, Tylt0.821, 3, 5
Ery, F, N, T4, 5 
Ery, L, T, Tylt0.556
Dap, F, L, T0.555, 6
Ery, F, L, N0.554, 5
Ery, L, N, T0.555, 6
F, L, N, T0.555, 6
Other phenotypes2.161, 4, 5, 6
5Ery, F, L, Syn, Tylt1.11
Dap, Ery, F, L, Tylt0.826
C, Ery, L, Syn, Tylt0.556
Ery, F, L, N, T0.554
Ery, L, N, T, Tylt0.555
Other phenotypes1.351, 4, 5, 6
6Bac, Ery, L Syn, T, Tylt0.554, 5
Cip, Ery, F, L, Syn, Tylt0.551
Other phenotypes0.811, 4
7Cip, Dap, Ery, F, L, Syn, Tytl0.275
Bac, Ery, F, Kan, L, T, Tylt0.271
Ery, F, Kan, L, Syn, T, Tylt0.271
Cip, Ery, F, L, N, T, Tylt0.274
Cip, Ery, L, N, Syn, T, Tylt0.275
Ery, Kan, L, N, S, T, Tylt0.274
11Bac, Dap, Ery, F, L, Lzd, P, Syn, T, Tylt, V0.275
12Dap, Ery, F, Gen, L, Lzd, P, S, Syn, T, Tylt, V0.271
14Bac, C, Dap, Ery, F, Kan, L, Lzd, P, S, Syn, T, Tylt, V0.275

Cip = Ciprofloxacin. Lzd = Linezolid. P = Penicillin. V = Vanco-mycin.

See Tables 2 and 4 for remainder of key.

Table 5—

Percentage of 366 Enterococcus isolates recovered from fecal samples collected from feedlot steers, on the basis of MIC ratio.

AntimicrobialMIC ratio*Lowest concentration tested (μ/mL)
 012345678 
Bacitracin54.1010.909.302.50†0.27†23.00†8
Chloramphenicol38.5014.5038.307.400.27†0.27†0.27†2
Erythromycin32.000.829.309.809.00†39.10†0.5
Flavomycin44.006.807.707.706.80†4.40t21.60†1
Penicillin61.8016.9015.604.400.550†0.82†0.5
Daptomycin44.3022.7022.406.600.551.60†1.90†0.5
Quinupristin and dalfopristin68.3024.304.60†1.90†0†0†0.82†1
Tetracycline67.201.404.404dR12.00†15.00†4.0
Vancomycin93.401.901.901.90000†0.82†0.5
Lincomycin5.100.554.609.6022.101.80†43.20†1.00
Tylosin tartrate9.9012.6019.7023.800.82000.820.55†26.50†0.25
Ciprofloxacin29.8014.8018.0025.109.901.60†0.82†0.12
Linezolid35.500.9947.805.200.27†1.40†0.5
Nitrofurantoin10.7014.009.8014.8030.6018.00†2.20†2
Kanamycin74.6017.50†6.60†0.27†1.10†128
Gentamicin99.200.550†0.27†128
Streptomycin98.900.55†0.27†0.27†512

— = Value greater than the MIC of the assay.e

See Table 2 for remainder of key.

Figure 3—
Figure 3—

Mean ± SEM proportion of Enterococcus casseliflavus (A), Enterococcus faecium (B), Enterococcus galinarum (C), and Enterococcus hirae (D) isolates recovered from fecal samples collected from steers administered chlortetracycline in cottonseed meal or nonmedicated cottonseed meal alone on days 0 to 4, 6 to 10, and 12 to 16 on the basis of the number of antimicrobials to which the isolates were resistant.

Citation: American Journal of Veterinary Research 69, 8; 10.2460/ajvr.69.8.988

A significant (P = 0.02) interaction was detected between treatment and sample day for mean MIC of tetracycline (Figure 4). A significant (P = 0.01) effect of treatment was detected on sample days 2, 6, 8, 12, 14, and 22 on the basis that isolates recovered from the control cohort had a lower geometric mean MIC of tetracycline than the values for isolates recovered from the exposed cohort. The interaction between treatment and sample period was not significant (P = 0.16) for the proportion of isolates resistant to tetracycline; however, there was a significant (P = 0.01) effect of treatment in that isolates recovered from the exposed cohort were more likely to be resistant than were isolates recovered from the control cohort, when averaged across time. Averaged across time, 12.3% (95% CI, 6.5% to 22.1%) and 53.3% (95% CI, 39.0% to 67.0%) of Enterococcus isolates recovered from the control and exposed cohorts, respectively, were resistant to tetracycline. Sample period did not have a significant (P = 0.36) effect.

Figure 4—
Figure 4—

Geometric mean ± SEM (log2-transformed) MIC of tetracycline (A) and the proportion of isolates resistant to tetracycline (B) for Enterococcus isolates recovered from fecal samples collected from steers administered chlortetracycline in cottonseed meal or nonmedicated cottonseed meal alone on days 0 to 4, 6 to 10, and 12 to 16. See Figure 1 for remainder of key.

Citation: American Journal of Veterinary Research 69, 8; 10.2460/ajvr.69.8.988

Discussion

Chlortetracycline administration increased our ability to recover E coli and Enterococcus spp with reduced susceptibility to tetracycline. Although we detected a significant difference between cohorts on day −7, the reason for this is uncertain. When the mean MIC for tetracycline for days −7 and 0 was determined, no preexposure effect was detected. The association between chlortetracycline administration and reduced susceptibility was transitory, and the MIC of tetracycline for isolates returned to or approached preexposure values by the final day of sample collection (ie, 17 days after discontinuation of chlortetracycline). For the proportion of resistant isolates, an interaction between time and exposure was not detected for E coli or Enterococcus spp, although graphically the data appeared consistent with an interaction (Figures 1 and 4). Therefore, it was likely that the binomial data (ie, proportion resistant) lacked the statistical power to detect an interaction even though there was sufficient power within the continuous data (ie, log2 MIC). When all results (both descriptive and analytic) were considered, the data are consistent with a transitory effect of chlortetracycline administration on tetracycline resistance in E coli and Enterococcus organisms. Although there was not a significant effect, there appeared to be an increase in the proportion of resistant E coli (and a corresponding increase in MIC) after exposure, compared with preexposure estimates for both the exposed and control cohorts. The reason for this is unknown.

Another important consideration of our analysis is that for isolates that grew at the greatest concentration of an antimicrobial included in the assay, we forced the MIC to be the next highest serial dilution. This assumption was used for analytic and reporting purposes. We believe it likely resulted in some right censoring of the data, which would have biased the results toward the null hypothesis. This bias, if present, would not have affected the analysis of the proportion resistant because the breakpoints were within the dilution range of the plates.

It is not possible to determine from our data whether chlortetracycline administration was associated with an increased population of resistant organisms or the proportional observations were merely a consequence of a decrease in the numbers of susceptible organisms. We did not quantify bacterial load in this study. Future research in which bacterial loads are quantified is needed because if the latter were true (ie, similar numbers of resistant organisms but reduced numbers of susceptible bacteria), public health consequences would be limited despite the fact that as a proportion of bacteria recovered, resistance was greater in the chlortetracycline-exposed cohort.

The proportion of E coli isolates that had coresistance to ACSSuT was relatively small (9.7%) and similar to the value reported in another study.20 The majority (72.6%) of these isolates were also coresistant to ceftiofur. Because ceftiofur-resistant organisms were always coresistant to tetracycline and chlortetracycline exposure was associated with an increased likelihood of recovering tetracycline-resistant E coli, we expected that chlortetracycline administration would have increased the likelihood of recovering ceftiofur-resistant E coli. However, this was not evident in that chlortetracycline administration was associated with a decrease in ceftiofur resistance (and the ACSSuT phenotype); consequently, chlortetracycline administration was associated with a decrease in the proportion of isolates that were coresistant to ceftiofur and tetracycline. The majority (92.7%) of isolates resistant to > 4 drugs were of the ACSSuT phenotype. However, chlortetracycline administration in the study reported here appeared to select preferentially for tetracycline-resistant E coli isolates harboring only a limited number of other resistance determinants. This transient effect also returned to preexposure values by the completion of sample collection. Although this unexpected finding may simply be spurious,24 it may also indicate that subpopulations of E coli are associated with varying fitness (or metabolic) costs, and this is determined, in part, by the resistance determinants the bacteria harbor. If so, selective pressures (eg, chlortetracycline administration) may provide a subpopulation (ie, those singly resistant to tetracycline) with more of a fitness advantage over another subpopulation (ie, those coresistant to ceftiofur and tetracycline). Whether this advantage simply results in clonal expansion favoring singly tetracycline-resistant strains over multiple-resistant strains (eg, ACSSuT plus ceftiofur) or whether there are more complex actions such as the cellular expulsion of mobile elements (eg, plasmids)25 containing genes coding for multiple resistance remains unknown at this juncture. Further research is warranted to better understand the results obtained in the study reported here and to determine whether novel selection pressures that favor specific subpopulations of bacteria may be exploited to reduce the burden of resistance to certain antimicrobials.

Chlortetracycline was administered for three 5-day periods with 1 day between each successive period. Fifteen consecutive days of administration may have been a more intuitive approach; however, the formulation of chlortetracyclinea used had label directions for no more than 5 consecutive days of administration. On the basis of the experiences of 2 of the authors (DUT, SEI), 3 consecutive regimens, which were used in this study, are not an uncommon industry practice to aid in the control of bovine respiratory tract disease and treatment of affected cohorts of feedlot cattle. Because an objective of the study was to evaluate chlortetracycline administration as a method for examining tetracycline resistance in real-world feedlot settings, an industry-used technique that includes antimicrobials within label restrictions was adopted. The method developed may provide utility in examining emergence and dissemination of resistant bacteria and resistance determinants. If this method is to be used, we recommend incorporation of microbiologic methods to quantify the outcomes of interest, regardless of whether they are bacterial populations or genetic elements conferring resistance.

In the study described here, we selected colonies from MacConkey agar plates that were morphologically typical of E coli and classified them as such without further biochemical assays. We did this because in another study20 conducted by our laboratory group, we found that further biochemical assay of suspect E coli isolates of bovine origin offered little advantage because 952 of 953 (99.9%) suspect colonies were confirmed as E coli. When the pretest likelihood (or probability) that suspect colonies are indeed E coli is so great (probability ≥ 99%, as determined on the basis of our data20), there is little opportunity for improvement in positive predictive value (ie, probability that a test-positive colony is a truly positive result) because it can only increase from 99% to 100%. However, the negative predictive value (probability that a test-negative isolate is truly a negative result) will be low given an imperfect confirmatory test, and in most scenarios, the negative predictive value is < 50% (sometimes substantially so). Therefore, test-negative isolates are more likely to be truly positive (ie, they are in fact E coli) than truly a negative result. Thus, researchers should weigh the burden of pretest misclassification versus posttest misclassification when judging a value to place on confirmatory assays. When the pretest probabilities approach 100% (as indicated in our data20 for E coli recovered from feces of cattle), confirmatory tests frequently and paradoxically result in a greater degree of misclassification than if they were never performed. Therefore, we do not recommend biochemical confirmatory steps in this specific animalsubstrate-bacterium (ie, bovine–feces–E coli) scenario.

On the other hand, of the suspect Enterococcus organisms recovered from Enterococcus-selective agar,b only 73.9% (n = 366) were biochemically confirmed as being Enterococcus spp. Substantial improvements in positive (and negative) predictive values are possible, and depending on the sensitivity and specificity of assays, additional biochemical confirmatory assays are typically warranted to satisfactorily differentiate Enterococcus organisms from organisms of other genera recovered from Enterococcus-selective agarb inoculated with bovine feces. In this animal-substrate-bacterium (ie, bovine–feces–Enterococcus spp) scenario, misclassification is typically (and substantially) reduced with additional biochemical confirmatory steps.

Inevitably, error resulted from our microbial methods. We selected typical colonies for further characterization. Atypical colonies (even though potentially of the genera being investigated) were not selected, and they would not have been selected regardless of the use of additional confirmatory steps. This sampling error (or bias) ultimately results in decreased sensitivity in identifying bacterial variants. On the other hand, because the pretest probability that E coli isolates were indeed E coli was < 100% (albeit only slightly less), a small number of non–E coli isolates were probably included in our analysis. However, the number of incorrectly classified isolates (as determined on the basis of data generated by our laboratory group20) would be expected to be limited. Furthermore, because no assay is perfect for sensitivity and specificity, we almost certainly excluded some enterococci from the analyses and also included some nonenterococci in our analyses.

Such errors of bacterial misclassification and their implication for measures of burden or effect are rarely discussed in the scientific literature. If the misclassification is nondifferential with regard to exposure, it will always bias estimates toward the null hypothesis or increase the likelihood of type II error. In the study described here, there were inevitably misclassification errors, but on the basis of data in another study20 conducted by our laboratory group, we believed their magnitude was negligible. Furthermore, because we detected effects of chlortetracycline exposure, these misclassification errors did not result in type II error.

ABBREVIATIONS

ACSSuT

Ampicillin, chloramphenicol, streptomycin, sulfamethoxazole, and tetracycline

CI

Confidence interval

MIC

Minimum inhibitory concentration

a.

Aureomycin, chlortetracycline calcium complex equivalent to 100 g of chlortetracycline/lb, Alpharma, Bridgewater, NJ.

b.

Enterococcosel agar, Becton, Dickenson & Co, Sparks, Md.

c.

Vitek 2, bioMérieux Inc, Durham, NC.

d.

Sensititre gram-negative NARMS plate, catalogue No. CMV1AGNF, TREK Diagnostic Systems, Cleveland, Ohio.

e.

Sensititre gram-positive NARMS plate, catalogue No. CMV1AGPF, TREK Diagnostic Systems, Cleveland, Ohio.

f.

Sensititre autoinoculator, TREK Diagnostic Systems, Cleveland, Ohio.

g.

Access, Microsoft Corp, Redmond, Wash.

h.

SAS, version 9.1.3, SAS Institute Inc, Cary, NC.

i.

SigmaPlot release 10.0, Systat, San Jose, Calif.

References

  • 1.

    Levy SB. Antibiotic availability and use: consequences to man and his environment. J Clin Epidemiol 1991;44(suppl 2):83S87S.

  • 2.

    Lipsitch M. The rise and fall of antimicrobial resistance. Trends Microbiol 2001;9:438444.

  • 3.

    Threlfall EJ, Ward LR, Frost JA, et al. The emergence and spread of antibiotic resistance in food-borne bacteria. Int J Food Microbiol 2000;62:15.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4.

    van den Bogaard AE, Stobberingh EE. Epidemiology of resistance to antibiotics. Links between animals and humans. Int J Antimicrob Agents 2000;14:327335.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5.

    Barza M, Travers K. Excess infections due to antimicrobial resistance: the “attributable fraction.” Clin Infect Dis 2002;34(suppl 3):S126S130.

  • 6.

    Khachatourians GG. Agricultural use of antibiotics and the evolution and transfer of antibiotic-resistant bacteria. Can Med Assoc J 1998;159:11291136.

    • Search Google Scholar
    • Export Citation
  • 7.

    Travers K, Barza M. Morbidity of infections caused by antimicrobial-resistant bacteria. Clin Infect Dis 2002;34(suppl 3):S131S134.

  • 8.

    Helms M, Vastrup P, Gerner-Smidt P, et al. Excess mortality associated with antimicrobial drug-resistant Salmonella typhimurium. Emerg Infect Dis 2002;8:490495.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9.

    Varma JK, Molbak K, Barrett TJ, et al. Antimicrobial-resistant nontyphoidal Salmonella is associated with excess bloodstream infections and hospitalizations. J Infect Dis 2005;191:554561.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10.

    Finch RG. Antibiotic resistance. J Antimicrob Chemother 1998;42:125128.

  • 11.

    Lupski JR. Molecular mechanisms for transposition of drug-resistance genes and other movable genetic elements. Rev Infect Dis 1987;9:357368.

  • 12.

    Levy SB. Multidrug resistance—a sign of the times. N Engl J Med 1998;338:13761378.

  • 13.

    Singer RS, Finch R, Wegener HC, et al. Antibiotic resistance—the interplay between antibiotic use in animals and human beings. Lancet Infect Dis 2003;3:4751.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14.

    Phillips I. Withdrawal of growth-promoting antibiotics in Europe and its effects in relation to human health. Int J Antimicrob Agents 2007;30:101107.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15.

    Casewell M, Friis C, Marco E, et al. The European ban on growth-promoting antibiotics and emerging consequences for human and animal health. J Antimicrob Chemother 2003;52:159161.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16.

    Anonymous. DANMAP 2005. Use of antimicrobial agents and occurrence of antimicrobial resistance in bacteria from food animals, foods and humans in Denmark. Søborg, Denmark: Danish Institute for Food and Veterinary Research, 2006.

    • Search Google Scholar
    • Export Citation
  • 17.

    Aslam M, Service C. Antimicrobial resistance and genetic profiling of Escherichia coli from a commercial beef packing plant. J Food Prot 2006;69:15081513.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18.

    Sawant AA, Hegde NV, Straley BA, et al. Antimicrobial-resistant enteric bacteria from dairy cattle. Appl Environ Microbiol 2007;73:156163.

  • 19.

    Wagner BA, Dargatz DA, Salman MD, et al. Comparison of sampling techniques for measuring the antimicrobial susceptibility of enteric Escherichia coli recovered from feedlot cattle. Am J Vet Res 2002;63:16621670.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20.

    Lowrance TC, Loneragan GH, Kunze DJ, et al. Changes in antimicrobial susceptibility in a population of Escherichia coli isolated from feedlot cattle administered ceftiofur crystalline-free acid. Am J Vet Res 2007;68:501507.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21.

    Clinical and Laboratory Standards Institute. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically: approved standard. CLSI document M7–A7. Wayne, Pa: Clinical and Laboratory Standards Institute, 2006.

    • Search Google Scholar
    • Export Citation
  • 22.

    CDC. National Antimicrobial Resistance Monitoring System for Enteric Bacteria (NARMS): 2003 human isolates final report. Atlanta: US Department of Health and Human Services CDC, 2006.

    • Search Google Scholar
    • Export Citation
  • 23.

    Littell RC, Henry PR, Ammerman CB. Statistical analysis of repeated measures data using SAS procedures. J Anim Sci 1998;76:12161231.

  • 24.

    Funk JA, Lejeune JT, Wittum TE. The effect of subtherapeutic chlortetracycline on antimicrobial resistance in the fecal flora of swine. Microb Drug Resist 2006;12:210218.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25.

    DeNap JC, Hergenrother PJ. Bacterial death comes full circle: targeting plasmid replication in drug-resistant bacteria. Org Biomol Chem 2005;3:959966.

    • Crossref
    • Search Google Scholar
    • Export Citation
All Time Past Year Past 30 Days
Abstract Views 76 0 0
Full Text Views 1518 1222 368
PDF Downloads 268 135 12
Advertisement