Abstract
Antibiotics have saved millions of human lives, and their use has contributed significantly to improving human and animal health and well-being. Use of antibiotics in food-producing animals has resulted in healthier, more productive animals; lower disease incidence and reduced morbidity and mortality in humans and animals; and production of abundant quantities of nutritious, high-quality, and low-cost food for human consumption. In spite of these benefits, there is considerable concern from public health, food safety, and regulatory perspectives about the use of antimicrobials in food-producing animals. Over the last two decades, development of antimicrobial resistance resulting from agricultural use of antibiotics that could impact treatment of diseases affecting the human population that require antibiotic intervention has become a significant global public health concern. In the present review, we focus on antibiotic use in lactating and nonlactating cows in U.S. dairy herds, and address four key questions: (1) Are science-based data available to demonstrate antimicrobial resistance in veterinary pathogens that cause disease in dairy cows associated with use of antibiotics in adult dairy cows? (2) Are science-based data available to demonstrate that antimicrobial resistance in veterinary pathogens that cause disease in adult dairy cows impacts pathogens that cause disease in humans? (3) Does antimicrobial resistance impact the outcome of therapy? (4) Are antibiotics used prudently in the dairy industry? On the basis of this review, we conclude that scientific evidence does not support widespread, emerging resistance among pathogens isolated from dairy cows to antibacterial drugs even though many of these antibiotics have been used in the dairy industry for treatment and prevention of disease for several decades. However, it is clear that use of antibiotics in adult dairy cows and other food-producing animals does contribute to increased antimicrobial resistance. Although antimicrobial resistance does occur, we are of the opinion that the advantages of using antibiotics in adult dairy cows far outweigh the disadvantages. Last, as this debate continues, we need to consider the consequences of “what would happen if antibiotics are banned for use in the dairy industry and in other food-producing animals?” The implications of this question are far reaching and include such aspects as animal welfare, health, and well-being, and impacts on food quantity, quality, and food costs, among others. This question should be an important aspect in this ongoing and controversial debate.
Introduction
Procedures mandated by the Pasteurized Milk Ordinance, which is a document that specifies safety standards of Grade A milk (Grade “A” Pasteurized Milk Ordinance, 2007 Revision), are in place to ensure that meat and milk of dairy cows for human consumption are free from antibiotic residues. However, the issue of development of antimicrobial resistance that could impact treatment of diseases affecting humans is not so clear. Some factors that contribute to antimicrobial resistance include the nature of disease-producing bacteria, the nature of diseases in which the antibiotics are used, environmental pressures, and widespread use of antibiotics in human medicine and animal agriculture (Hawkey, 2008). Over the last two decades, antimicrobial resistance associated with agricultural use of antibiotics and the impending propagation of antimicrobial-resistant bacteria from food-producing animals to humans has become a significant global public health concern (WHO, 2000; 2002; Heymann, 2006). This has led to important changes in perceptions and priorities of government agencies throughout the world with regard to antimicrobial usage—in particular, use of antimicrobials as growth promoters and as prophylactic agents (WHO, 2000, 2002; Nawaz et al., 2001; Erskine et al., 2002; Andersson, 2003; Angulo et al., 2004; Molbak, 2004; Phillips et al., 2004; Viola and DeVincent, 2006; Mathew et al., 2007).
There is little documented evidence on the amount of antimicrobials used in animal agriculture, and estimates vary considerably (Mellon et al., 2001; Nawaz et al., 2001; WHO, 2002; Viola and DeVincent, 2006). The Centers for Disease Control and Prevention (CDC) estimated that 50 million pounds of antibiotics are produced annually in the United States, and ∼20 million pounds or 40% of the total amount of antibiotics produced are used in agriculture (Nawaz et al., 2001). An often-cited report published by the Union of Concerned Scientists (Mellon et al., 2001) indicated that livestock producers in the United States use 24.6 million pounds of antibiotics annually for nontherapeutic purposes. Of this amount, ∼10.5 million pounds are used in poultry, 10.3 million pounds in hogs, and 3.7 million pounds in cattle. These estimates did not include antibiotics that are used therapeutically to treat sick animals. Thus, most of the antibiotics in animal agriculture in the report by Mellon et al. (2001) were used as growth promoters. Data from the Animal Health Institute representing U.S. manufacturers of animal pharmaceuticals summarized from 1998 to 2002 by Viola and DeVincent (2006) indicated that an average of 21.9 million pounds of antibiotics were produced annually and an average of 13.4% of the total was used for growth promotion.
Antimicrobial use in animal agriculture, especially at subtherapeutic levels, has met with considerable controversy and is at the center of the agriculture antibiotic use debate. A significant concern is that selection pressure from use of antimicrobials in food-producing animals could result in the emergence, maintenance, and horizontal transfer of antimicrobial-resistant determinants in bacteria (Witte, 1998; O'Brien, 2002; Molbak, 2004). Selection pressure through sustained use of antimicrobials at subtherapeutic concentrations in animal production systems could result in development of antimicrobial resistance in commensal and pathogenic bacteria, and favor genetic exchange of antimicrobial resistance determinants involving commensal bacteria (Summers, 2002). Bacteria exchange antimicrobial resistance genes, and these genes may ultimately enter bacteria pathogenic to human and/or opportunistic bacterial pathogens. Commensal bacteria from healthy animals can also serve as reservoirs for determinants of antimicrobial resistance and perhaps play a key role in their dissemination to other commensal and pathogenic bacteria (Levin et al., 1997; Guillemot and Courvalin, 2001; O'Brien, 2002). Use of antibiotics in food-producing animals for prophylaxis has also resulted in considerable controversy and is an important component of the agriculture antibiotic use debate.
Antimicrobial resistance resulting from agricultural use of antibiotics that could impact treatment of diseases affecting the human population that require antibiotic intervention has become a significant global public health concern. There are basically two opposing positions on this highly controversial polarizing topic. One position is that bacterial resistance to antimicrobials used in human medicine does result from agricultural use of antibiotics, and thus immediate action should be taken to prevent this from happening in the future. The alternative position is that resistance to antimicrobials used in human medicine does result from agricultural use of antibiotics; however, evidence of this having a major effect on human health and well-being is minimal or nonexistent, and therefore no action is required. As pointed out by Turnidge (2004), the real difference between these two positions is whether action should be taken, or should have been taken to effectively deal with bacterial antimicrobial resistance developed in food-producing animals. This on-going debate has led to important changes in perceptions and priorities of federal regulatory and public health agencies throughout the world with regard to antimicrobial usage—in particular, use of antimicrobials as growth promoters and as prophylactic agents.
In the present review, we focus on antibiotic use in lactating and nonlactating cows in U.S. dairy herds, and address four key questions: (1) Are science-based data available to demonstrate antimicrobial resistance in veterinary pathogens that cause disease in dairy cows associated with use of antibiotics in adult dairy cows? (2) Are science-based data available to demonstrate that antimicrobial resistance in veterinary pathogens that cause disease in adult dairy cows impacts pathogens that cause disease in humans? (3) Does antimicrobial resistance impact the outcome of therapy? (4) Are antibiotics used prudently in the dairy industry?
Antibiotic Use in Lactating and Nonlactating Cows in Dairy Herds in the United States
Antibiotics are used on dairy farms for treatment of a variety of diseases affecting dairy cows. Data from the U.S. Department of Agriculture (USDA) National Animal Health Monitoring System (NAHMS) Dairy 2007 study conducted in 17 major dairy states in the United States in 79.5% of dairy operations and 82.5% of dairy cows revealed that the percentage of farms that treated cows with any antibiotic was 85.4 for mastitis, 58.6 for lameness, 55.8 for respiratory, 52.9 for reproduction, 25.0 for diarrhea or other digestive problems, and 6.9 for all other categories (USDA APHIS, 2008a). The percentage of cows treated with antibiotics was 16.4 for mastitis, 7.4 for reproduction, 7.1 for lameness, 2.8 for respiratory, and 1.9 for diarrhea or other digestive problems (USDA APHIS, 2008a). Antibiotics are also used frequently on dairy farms for disease prevention. Over 90% of dairy farms practiced antibiotic dry cow therapy and used intramammary antibiotics following the last milking of lactation (USDA APHIS, 2008a, 2008b). Approximately 80% of farms that practiced antibiotic dry cow therapy treated all cows on the farm. Overall antibiotic use on dairy farms in the NAHMS Dairy 2007 study was similar to the NAHMS Dairy 2002 study (USDA APHIS, 2005a, 2008a). The NAHMS Dairy 2002 study was conducted in 21 major dairy states in the United States and represented 82.9% of dairy operations and 85.5% of dairy cows (USDA APHIS, 2005a).
Several antibiotics are used for treatment and prevention of diseases of dairy cows, and mastitis continues to be the most commonly treated disease. Cephalosporin was the most widely used antibiotic for treatment of mastitis; 53.2% of cows were treated with cephalosporin. Other antibiotics used frequently to treat cows with mastitis were lincosamide (19.4% of cows), and 19.1% of cows were treated with noncephalosprorin β-lactam antibiotics (USDA APHIS, 2008a). Penicillin G/dihydrostreptomycin and cephapirin were the two most commonly used antibiotics used for dry cow therapy (USDA APHIS, 2008a). Over 42% of cows treated for lameness received tetracycline, whereas 27.2% and 19.5% were treated with cephalosporin and noncehalosprorin β-lactam antibiotics, respectively (USDA APHIS, 2008a). Antibiotics were administered to dairy cows through two primary routes: intramuscular and intramammary (USDA APHIS, 2009a).
Are Science-Based Data Available to Demonstrate Antimicrobial Resistance in Veterinary Pathogens That Cause Disease in Dairy Cows Associated with Use of Antibiotics in Adult Dairy Cows?
Although resistance to antimicrobial drugs among mastitis pathogens has been well documented for nearly four decades, evidence has not been published to suggest that this is either an emerging or progressing phenomenon. According to the National Mastitis Council's expert group (Erskine et al., 2004) that reviewed research on bovine mastitis pathogens and trends in resistance to antimicrobial drugs, currently available scientific evidence does not support widespread, emerging resistance among mastitis pathogens to antimicrobial drugs. There is a dearth of controlled studies that have determined the pharmacodynamic basis, including which drug therapeutic regimens may increase or decrease this risk. Although antimicrobial resistance can be induced in most pathogens, there is no evidence that antimicrobial resistance is an emerging process in mastitis pathogens in dairy cows. Further, if we are to infer developing resistance in relation to antimicrobial drug therapy, the history of drug administration, including dose, frequency, and duration, must be known. Science-based evidence would also entail use of comparative antibiograms before and after drug administration (Erskine et al., 2004).
Use of antibiotics, even in large doses such as in dry cow mastitis therapy programs, had minimal effects on bacterial resistance and minimal inhibitory concentration (MIC) of Escherichia coli in either the herd or environment (Rollins et al., 1974). More recently, Singer et al. (2008) tracked ceftiofur-treated and untreated dairy cattle to examine changes in genetic diversity and resistance to the antibiotic in E. coli. Ceftiofur-resistant E. coli were only isolated from treated cows during and immediately after cessation of treatment. E. coli counts dropped significantly in treated animals, reflecting a disappearance of sensitive strains; in addition, the resistant population did not increase in number within treated cows. In essence, the resistant population stayed low and was in fact replaced by antibiotic-susceptible E. coli populations. Conclusions were that antibiotic treatment did not appear to cause emergence or amplification of antibiotic resistance.
Have Resistance Patterns Changed in Mastitis Pathogens over Time, Presumably in Response to Antibiotic Use?
One way to assess the effect of antimicrobial use on antimicrobial resistance is to compare and contrast systems that employ different production strategies, for example, organic dairies that use little to no antibiotics, and conventional dairies where antibiotics are used in all categories of dairy animals (Call et al., 2008). To date, no studies have been published with the defined objective of comparing animal health on organic versus conventional dairy herds in the United States (Ruegg, 2009).
Use of antimicrobials has been suggested to be a selective force in determining the bacterial ecology of bovine mastitis. Despite these assumptions, very few studies (Mackie et al., 1988; Erskine et al., 2002; Rajala-Schultz et al., 2004) have demonstrated the long-term effects or trends over years on use of antibiotics on antimicrobial susceptibility of dairy cows. Mackie et al. (1988) found no change in proportions of Staphylococcus aureus isolates susceptible to antimicrobials. On the other hand, nonspeciated coliform bacteria demonstrated increased resistance to ampicillin and neomycin. In a 7-year study of Michigan dairy herds that included Gram-positive (Table 1) and Gram-negative (Table 2) mastitis pathogens (Erskine et al., 2002), the proportion (i.e., number of susceptible strains/number of strains evaluated) of bacterial isolates susceptible to antibiotics did not change for the majority of tests. Proportions of S. aureus that were susceptible to ampicillin, penicillin, and erythromycin increased with time. Streptococcus uberis demonstrated an increase in the proportion of isolates that were susceptible to gentamicin, oxacillin, sulfatrimethoprim, and pirlimycin, whereas there was increased resistance to penicillin. Streptococcus dysgalactiae showed increased susceptibility to erythromycin, gentamicin, sulfatrimethoprim, and tetracycline. There was an increased proportion of susceptible Streptococcus agalactiae with use of sulfatrimethoprim. Gram-negative isolates showed an increase in the proportion of E. coli that were susceptible to ampicillin and cephalothin, whereas Klebsiella pneumoniae demonstrated increased susceptibility to ceftiofur and apparent increased susceptibility to ampicillin. There were no changes in susceptibility profiles of Serratia marcesens and Pseudomonas aeruginosa.
There were no changes in susceptibility among resistance values that are not represented by a superscript.
The linear trend (not shown) from 1994 to 2000 was toward increased susceptibility.
The linear trend (not shown) from 1994 to 2000 was toward decreased susceptibility.
There were no changes in susceptibility among those not represented by a superscript.
The linear trend (not shown) from 1994 to 2000 was toward increased susceptibility.
Rajala-Schultz et al. (2004) conducted a 16-month study on antimicrobial susceptibility of mastitis pathogens isolated from first lactation and older dairy cows (Table 3). Their results appear to support the notion that use of antibiotics is a main factor in development of antimicrobial resistance. The study targeted coagulase-negative staphylococci (CNS), esculin-positive streptococci, and Gram-negative pathogens (E. coli, Serratia spp., Klebsiella spp., Citrobacter spp., and Enterobacter spp.). Based on visual assessment, there appeared to be no differences in MICs in esculin-positive streptococci and Gram-negative bacteria for the various antibiotics that were evaluated. CNS were isolated most frequently from intramammary infections (IMIs); 78% and 40%, respectively, were resistant to at least one antibiotic. Most resistance was observed against penicillin with 39% and 26% of CNS isolates from older cows and first lactation cows, demonstrating resistance to this antimicrobial. Although penicillin resistance was higher in older cows and tetracycline resistance was higher in first lactation cows, differences in proportions of resistant isolates between the two groups were not statistically significant. Resistance patterns of CNS were indicated to be concordant with usage of antibiotics in the dairy herd.
All isolates (Gram-positive and Gram-negative) in this study were tested against the same microdilution panel even though some organisms are intrinsically resistant to certain antimicrobials; for example, penicillin or pirlimycin do not have any activity against Gram-negative bacteria.
Pooled data for first lactation and older cows.
Gram-negative mastitis pathogens included E. coli, Klebsiella spp., Serratia spp., Citrobacter spp., Enterobacter spp., Pseudomonas spp., and others that were not speciated.
CNS are the most common pathogens associated with mastitis, and there is evidence that they may serve as a pool of antibiotic resistance genes that could be transmitted to S. aureus potentially resulting in the emergence of methicillin-resistant S. aureus (MRSA) strains, which demonstrate multiple resistance to penicillins, cephalosporins, and some fluoroquinolones (McAllister et al., 2001). Gillespie et al. (2009) isolated CNS from 11.4% (1407 of 12,412) of mammary quarter samples obtained from cows in three dairy research herds in 2005. Approximately 27% (383/1407) of CNS were identified to the species level. The dominant CNS species isolated were Staphylococcus chromogenes (48%), Staphylococcus hyicus (26%), and Staphylococcus epidermidis (10%). A majority of these CNS isolates were susceptible to ampicillin, oxacillin, cephalothin, ceftiofur, erythromycin, and pirlimycin (Sawant et al., 2009). The only exception was with S. epidermidis, where some strains exhibited efflux-based resistance to erythromycin encoded by msrA and one isolate that carried ermC encoding ribosomal methylase-based resistance to both erythromycin and pirlimycin. Methicillin-resistant S. epidermidis carried low affinity penicillin-binding protein encoded by mecA. Most multidrug-resistant (MDR) S. epidermidis (≥2 resistance genes) were resistant to ampicillin, erythromycin, and methicillin. Based on pulsed-field gel electrophoresis typing, MDR S. epidermidis were closely related genotypically, and were isolated from different cows on the same farm, suggesting clonal dissemination. Bovine S. epidermidis share antimicrobial resistance patterns and virulence determinants of strains observed in human infections (Sawant et al., 2009). Antimicrobial resistance of S. epidermidis may be important to those who consume raw milk because this organism is isolated frequently from bulk tank milk.
Pol and Ruegg (2007) analyzed relationships between antimicrobial usage at the farm level comparing organic versus conventional dairies and antimicrobial susceptibility of bacterial isolates collected from 1994 to 2000. Organic dairy herds received minimal or no exposure to antimicrobial drugs. More IMIs were present in organic than conventional herds. All isolates (CNS, Streptococcus spp., S. aureus, and S. agalactiae) except coliforms were more prevalent on organic herds. The MIC of selected antimicrobial drugs was determined for S. aureus, CNS, and Streptococcus spp. obtained from cows with subclinical mastitis. The majority of isolates were inhibited at the lowest dilution of antimicrobials tested. S. aureus and CNS demonstrated heterogeneity in MICs dependant on the amount of exposure to penicillin and pirlimycin. Farm type was associated with the MIC of ampicillin and tetracycline for CNS, and pirlimycin and tetracycline for Streptococcus spp. Further, the MIC for pirlimycin increased with increasing exposure to defined daily doses of pirlimycin. The level of exposure to most other antimicrobial drugs was not associated with MIC of mastitis pathogens. A dose–response effect between antimicrobial exposure and susceptibility was observed for some pathogen–antimicrobial combinations, for example, pirlimycin versus all isolates, whereas other antimicrobials commonly used for treating and preventing mastitis were not associated with resistance.
One of the more consistent uses of antibiotics in dairy herds is in dry cow mastitis therapy programs in which every cow is treated at the end of the lactation period. Over a 1-year study, intramammary administration of large doses of penicillin and dihydrostreptomycin had little or no effect on drug resistance in E. coli in the dairy herd and its immediate environment (Rollins et al., 1974). Peaks of resistant E. coli of 10% from both cow and environmental sources had a marked pattern of similarity, suggesting an enteric-environmental flora interaction. Prophylactic use of antibiotics, even in large doses, in dry cow mastitis therapy programs had minimal effects on bacterial resistance and MIC of E. coli in either the herd or the environment.
Have Resistance Patterns Changed over Time in Enteric Pathogens, Presumably in Response to Antibiotic Use?
E. coli and Salmonella
E. coli is the predominant bacterial species cultured from feces of dairy cattle (Sawant et al., 2007). In a study conducted to investigate the descriptive and molecular epidemiology of antimicrobial-resistant Gram-negative enteric bacteria in the feces of healthy lactating dairy cattle, nine bacterial species were isolated. The E. coli isolates exhibited resistance to ampicillin (48%), ceftiofur (11%), chloramphenicol (20%), florfenicol (78%), spectinomycin (18%), and tetracycline (93%). MDR (resistance to ≥3 to 6 antibiotics) was seen in 40% of isolates. Findings from the study by Sawant et al. (2007) suggest that commensal enteric E. coli from healthy lactating cattle could be an important reservoir for tetracycline and perhaps other antimicrobial resistance determinants.
Susceptibility to 26 antimicrobial agents used in veterinary and human medicine, and prevalence of antimicrobial resistance genes of 135 E. coli isolated from cows with mastitis were evaluated (Srinivasan et al., 2007). All isolates were resistant to two or more antimicrobials in different combinations. Most were resistant to antimicrobials used in veterinary medicine, including ampicillin (98.4%, ≥32 μg/mL), and many were resistant to streptomycin (40.3%, ≥64 μg/mL), sulfisoxazole (34.1%, ≥512 μg/mL), and tetracycline (24.8%, ≥16 μg/mL). Most isolates were resistant to antimicrobials used in human medicine, including aztreonam (97.7%, ≥32 μg/mL) and cefaclor (89.9%, ≥32 μg/mL). Some were resistant to nitrofurantoin (38%, ≥128 μg/mL), cefuroxime (22.5%, ≥32 μg/mL), and fosfomycin (17.8%, ≥256 μg/mL). All E. coli were susceptible to ciprofloxacin and cinoxacin. Almost 97% (123 of 127) of ampicillin-resistant isolates carried ampC. Eleven of 52 (21.2%) streptomycin-resistant isolates carried strA, strB, and aadA together, and 29 streptomycin-resistant isolates (55.8%) carried aadA alone. Among 44 sulfisoxazole-resistant E. coli, 1 isolate (2.3%) carried both sulI and sulII, 12 (27.3%) carried sulI, and 10 (22.7%) isolates carried sulII. Among 32 tetracycline-resistant isolates, 14 (43.8%) carried both tetA and tetC and 14 (43.8%) carried tetC. Results from this study demonstrated that E. coli from cows with mastitis were MDR and carried multiple resistance genes. These bacteria can be a reservoir for antimicrobial resistance genes and could play a role in dissemination of antimicrobial resistance genes to other pathogenic and commensal bacteria in the dairy farm environment.
Walk et al. (2007) provided data that support the hypothesis that the abundance and distribution of tetracycline resistance (Tetr) determinants are weakly influenced by antibiotic use. The purpose of the study was to assess the influence of antibiotic selection on genetic composition of E. coli populations from conventional and organic dairy farms. A significant association was established between low MDR, organic farms, and strains of E. coli phylogroup B1, which were the most preponderant. Data suggested that organic farming practices change the frequency of resistance strains as well as impact the overall population genetic composition of resident E. coli flora. Results also indicated that the current prevalence of tetr loci on dairy farms has little to do with the use of tetracycline.
Singer et al. (2008) conducted a study to follow ceftiofur-treated and untreated dairy cattle to examine changes in genetic diversity and resistance to the antibiotic, and particularly to resistance mediated by the β-lactamase gene (bla CMY-2) in E. coli. Ceftiofur-resistant E. coli were only isolated from treated cows during and immediately after cessation of treatment. E. coli counts dropped significantly in treated animals, reflecting a disappearance of sensitive strains. Further, the resistant population did not increase in number within treated cows; numbers stayed low and were replaced by antibiotic-susceptible E. coli populations. Conclusions were that antibiotic treatment enabled detection of ceftiofur-resistant E. coli but did not appear to cause its emergence or result in its amplification. Additionally, Singer et al. (2008) indicated that finding resistant strains after antibiotic treatment was not sufficient to estimate the strength of selection pressure nor was it sufficient to demonstrate a causal link between antibiotic use and emergence or amplification of antibiotic resistance.
Fey et al. (2000) indicated that Salmonella and E. coli had the same antimicrobial resistance determinants, suggesting interspecies transmissibility of antimicrobial resistance mechanisms. The wide distribution of the tetB efflux gene across Gram-negative genera, including Escherichia, Enterobacter, Proteus, Salmonella, Actinobacillus, Haemophilus, Moraxella, and Treponema spp., indicates that it is likely that horizontal transfer of tetracycline resistance occurs among these bacteria (Sawant et al., 2007).
Dairy cattle are considered to be reservoirs of Shiga-toxin-producing E. coli (STEC) O157, and contamination of raw milk or ground beef from dairy cattle poses a significant risk to humans. Many dairy STEC isolates are known to be of high virulence. Of 193 STEC serotypes of dairy cattle origin, 24 have been isolated from patients with hemolytic uremic syndrome. Culled dairy cows are used primarily for production of ground beef and therefore pose a significant health risk (Hussein and Sakuma, 2005). Fitzgerald et al. (2003) provided data on antimicrobial resistance of E. coli O157:H7 isolates from dairy cows (Table 4). Isolates demonstrated appreciable resistance (20%–60%) to a variety of antibiotics evaluated (10 of 13) but were universally sensitive to ceftiofur and trimethoprim/sulfamethoxazole. Multiple resistance to nine or more antibiotics was observed in five isolates. Further, lactating cows shed more E. coli O157:H7 than nonlactating cows (43% vs. 32%).
Antimicrobial susceptibility of STEC isolated from organic and conventional dairy farms was investigated (Cho et al., 2007). Resistance to at least one antimicrobial agent was observed in 62% and 48% of isolates from conventional and organic farms, respectively. The majority of STEC isolates from this study were similar to serotypes commonly associated with human infection (e.g., O2:NM, O2:H27, O26:H11, O103:H8, O111:NM, O113:H21, O119:HUT, O145:H28, and O157:H7). There were 29 STEC isolates from conventional farms, and 3, 3, and 2, respectively, were resistant to gentamicin, trimethoprim/sulfamethoxazole, and chloramphenicol. All 23 isolates from organic farms were sensitive to chloramphenicol, gentamicin, and trimethoprim/sulfamethoxazole. A significant difference was observed in the proportion of STEC that were resistant to spectinomycin (72.4% conventional vs. 39.1% organic farm sources; p < 0.05). A slightly greater proportion of isolates from conventional farms were resistant than those from organic farms; however, no statistically significant differences were observed in antimicrobial resistance of isolates from adult lactating dairy cows from the two management systems. Further, there were no significant differences in recovery of MDR (resistance to ≥4 antimicrobials).
Berge et al. (2007) described and compared antimicrobial susceptibility patterns of Salmonella enterica subspecies enterica and E. coli recovered from dairy bulk tank milk in California over a 29-month period. The most commonly recovered Salmonella serotypes were Montevideo (33%), Typhimurium (14%), Dublin (13%), and Give (11%). A total of 60% each of Salmonella and E. coli isolates were pansusceptible, and 23% of all isolates were classified as MDR. The MDR E. coli were 1.8 times more likely to be coisolated from dairies with MDR Salmonella in the same sampling period compared to herds without detectable Salmonella. However, resistance patterns of MDR pathogens differed. A longitudinal observational study conducted in Washington State indicated that acquisition of new MDR S. enterica strains was a common event; 33 of 59 (56%) herds had at least one strain introduced (Adhikari et al., 2009).
In 1985, an outbreak of MDR Salmonella serotype Newport in California in humans was genetically traced back to consumption of hamburger at a fast-food restaurant, and subsequently to a meat-processing plant and eventually to the dairy farms where the cattle were raised (Spika et al., 1987). Isolation of chloramphenicol-resistant Salmonella from dairies was associated with use of chloramphenicol at those dairies. In another study (Lyons et al. 1980), where evidence of direct linkage of antimicrobial resistance of bacteria from farm animals was directly associated with disease in humans, a Salmonella Heidelberg epidemic in a hospital nursery was traced to infected calves on a dairy farm where the mother of the infant lived. Isolates from all cases were resistant to chloramphenicol, sulfamethoxazole, and tetracycline.
In Massachusetts, the prevalence of Salmonella Newport-MDRAmpC (with MDR to amoxicillin/clavulanic acid, ampicillin, cefoxitin, ceftiofur, cephalothin, chloramphenicol, streptomycin, sulfamethoxazole, tetracycline, and decreased susceptibility to ceftriaxone) among Salmonella serotype Newport isolates obtained from humans increased from 0% in 1998 to 53% in 2001 (Gupta et al., 2003). Infection with Salmonella Newport-MDRAmpC was associated with exposure to a dairy farm. Additionally, isolates from both humans and dairy cattle had closely related antibiograms and PFGE patterns. The emergence of Salmonella Newport-MDRAmpC strains in humans coincided with the emergence of Newport-MDRAmpC infections in dairy cattle.
Combined data for clinical and nonclinical Salmonella from dairy cattle from the National Antimicrobial Resistance Monitoring System indicated that the percentage of resistance for most of the antibiotics evaluated increased from 2001 to 2003. In the 2001 and 2003 reports, resistance to tested antibiotics was as follows: ampicillin, 28.8% versus 69.2%; ceftiofur, 12.5% versus 49.5%; chloramphenicol, 17.1% versus 56.4%; streptomycin, 37.2% versus 63.2%; and tetracycline, 29.6% versus 65.5%, respectively (Mathew et al., 2007). Based on three USDA NAHMS nationwide dairy studies (Dairy 1996, Dairy 2002, and Dairy 2007) conducted on U.S. dairy operations (USDA APHIS, 2009b), it was concluded that resistance of Salmonella to antimicrobial drugs continues to be relatively low and did not change from 1996 to 2007. Of all Salmonella isolates evaluated for antimicrobial susceptibility, 88.9%, 83.0%, and 92.8% of isolates were susceptible to all antimicrobials tested in Dairy 1996, Dairy 2002, and Dairy 2007, respectively. In each study, about 5% of Salmonella isolates were resistant to two or more antimicrobials.
In the NAHMS Dairy 2002 study on U.S. dairy operations (USDA APHIS, 2005b), fecal samples collected with rectal retrieval were cultured for Salmonella. Isolates were evaluated for resistance against a panel of 16 antibiotics (Table 5). Twenty-eight different serotypes were identified, and 62.9% of these were serotype Meleagridis (24.1%), Montevideo (11.9%), Typhimurium (9.9), Kentucky (9.5%), and Agona (7.5%). Salmonella isolates showed relatively little resistance to a number of antimicrobial agents, with 83% demonstrating susceptibility to all 16 antimicrobial agents tested. All isolates were sensitive to amikacin, ciprofloxacin, naladixic acid, and trimethoprim/sulfamethoxazole. A total of 17% of isolates were resistant to at least one drug. The most common resistances were to tetracycline (11.9%) and streptomycin (9.6%). MDR was observed in 4.8% of isolates among the Mbandaka, Newport, Reading, and Typhimurium serotypes. The conclusion from this nationwide study was that resistance of Salmonella to antimicrobial drugs continues to be relatively low. An analysis of distributions of two prevalent MDR Salmonella phenotypes in the United States, 2003–2005, indicated that Salmonella Newport isolates were MDR-AmpC in the 10 states that had the highest milk cow density. However, MDR-ACSSuT Typhimurium (resistant to at least ampicillin, chloramphenicol, streptomycin, sulfonamides, and tetracycline) was distributed evenly across regions (Greene et al., 2008).
Resistance to underlined antibiotics was common among all MDR Salmonella isolates.
Intermediate resistances were classified as susceptible.
Panel of 16 antimicrobials tested: Am, amikacin; Amo, amoxacillin-clavulinic acid; Amp, ampicillin; Cefo, cefoxitin; Ceft, ceftiofur; Ceftr, ceftriaxone; Ceph, cephalothin; Chlor, chloramphenicol; Cip, ciprofloxacin; Gen, gentamicin; Kan, kanamycin; Nal, naladixic acid; Str, streptomycin; Sulf, sulfamethoxazole; Tet, tetracycline; Tris, trimethoprim/sulfamethoxazole.
MDR, multidrug-resistant.
In the NAHMS Dairy 2007 study (USDA APHIS, 2009b), the percentage of dairy operations positive for Salmonella was almost double that observed in the NAHMS Dairy 1996 study (39.7% of dairies and 13. 7% of cows in 2007 vs. 21.1% of dairies and 5.4% of cows in 1996). Part of the differences observed could be attributed to differences in sampling methodology. However, it was suggested that Salmonella might be becoming more common on U.S. dairies, particularly on larger farms with >500 cows. Three serotypes, Meleagridis, Montevideo, and Mbandaka, ranked in the top six for each of the NAHMS dairy studies. Almost 93% of Salmonella isolated in the NAHMS Dairy 2007 study were susceptible to all antimicrobials evaluated. Only 1.7% were resistant to a single antimicrobial, and 5.5% of Salmonella were resistant to two or more antimicrobials.
Prevalence of antimicrobial resistance among Salmonella isolated from conventional and organic dairy herds in New York, Minnesota, Michigan, and Wisconsin from August 2000 to October 2001 was investigated (Ray et al., 2007). Over 81% of isolates were pansusceptible to the 14 antibiotics evaluated. For most herds (81.6%; n = 129), the predominant antimicrobial resistance pattern was pansusceptible. On 23.6% of herds, at least one Salmonella isolate resistant to five or more antibiotics was found. Resistant isolates most frequently exhibited insensitivity to tetracycline, streptomycin, and/or ampicillin. Previous analyses showed that isolation rates and antimicrobial susceptibility of Salmonella from organic and conventional farms were similar (Fossler et al., 2005a, 2005b); thus, data from these studies were combined for analysis. Within most farms, the dominant antimicrobial resistance pattern was pansusceptibility to all antimicrobial agents tested. MDR isolates were obtained from sick cows and calves.
A study was conducted to establish the prevalence of S. enterica serotype Cerro, a potentially emerging cattle pathogen, in New York dairy herds (Cummings et al., 2010). Reports of human disease due to this serotype are sparse; however, there has been an apparent increase in the prevalence of this serotype among cattle with salmonellosis. Salmonella was identified in 77% of the study herds and nearly half of the positive herds (20/44) yielded Salmonella Cerro. There was a low frequency of antimicrobial resistance among Cerro isolates with 95% being pansusceptible. Antimicrobial resistance was widespread among farms with isolates from clinically ill cattle demonstrating greater antimicrobial resistance. The New York State Department of Health isolated Salmonella Cerro from two human patients since 2003, and both isolates had PFGE banding patterns identical with predominant PFGE types identified in cattle in this study.
Ceftiofur, a broad-spectrum cephalosporin with activity against both Gram-positive and Gram-negative bacteria, is approved in the United States for treatment of dairy cattle (Hornish and Kotarski, 2002). Ceftiofur-resistant Salmonella strains isolated from dairy farms in New York have been reported and they represent multiple widely distributed subtypes that evolved by independent horizontal gene transfer (Alcaine et al., 2005). Resistance to ceftiofur and ceftriaxone was highly correlated with the presence of bla CMY-2, the β-lactamase gene. The increase of antimicrobial-resistant Salmonella subtypes, including those that are resistant to ceftiofur, represents a particular concern because ceftriaxone is used to treat invasive cases of Salmonella in children and is closely related to ceftiofur, an antibiotic used commonly to treat dairy cattle.
The incidence of salmonellosis and antimicrobial resistance of Salmonella isolates were studied in 831 dairy herds in New York (Cummings et al., 2009). Salmonella were isolated from 22.5% of herds, and Salmonella Newport was the most dominant serotype followed by Salmonella Typhimurium. Salmonella Newport accounted for 41% of salmonellosis cases. More than 68% of all isolates were resistant to one or more antibiotics, and only 19% of isolates were pansusceptible.
Campylobacter
The prevalence and antimicrobial susceptibility of Campylobacter has been examined in three USDA NAHMS nationwide dairy studies (Dairy 1996, Dairy 2002, and Dairy 2007) conducted on U.S. dairy operations (USDA APHIS, 2009b). In Dairy 1996, Campylobacter was detected in feces of at least one cow on all dairies sampled (n = 31) using multiplex polymerase chain reaction. In Dairy 2002, culture-based methods detected Campylobacter in feces of 97.9% of 97 dairy operations, and 51.3% of 1424 cows. In Dairy 2007, 92.6% of 121 dairy operations and 33.7% of 1885 cows tested were positive for Campylobacter. Antimicrobial resistance monitoring was conducted in Dairy 2002 and Dairy 2007 (USDA APHIS, 2009b). About 50% of 473 Campylobacter jejuni isolates were susceptible to all antimicrobials tested in Dairy 2002; whereas, 36.6% were susceptible to all antimicrobials tested in Dairy 2007. The percent of C. jejuni resistant to a single antimicrobial was 46.9 and 61.2, respectively, in Dairy 2002 and Dairy 2007 studies. Percent of C. jejuni resistant to two or more antimicrobials was 3.6 and 2.2, respectively, in Dairy 2002 and Dairy 2007 studies. The greatest resistance was to tetracycline: 47.4% and 62.9%, respectively. Little resistance of C. jejuni was observed against all of the other antimicrobials evaluated, including ciprofloxacin and erythromycin (USDA APHIS, 2009b).
The prevalence and antimicrobial susceptibilities of Campylobacter spp. from bovine feces were compared between organic and conventional dairy herds (Sato et al., 2004). Thirty organic dairy herds were compared with 30 neighboring conventional dairy farms. The prevalence of Campylobacter spp. in fecal specimens was 26.7% and 29.1%, respectively. None of the isolates was resistant to gentamicin or erythromycin; 45% of isolates were resistant to tetracycline. There was no evidence that antimicrobial use on dairy farms correlated with prevalence of resistance to ciprofloxacin, gentamicin, erythromycin, and tetracycline in Campylobacter.
In another study (Table 6), Campylobacter isolates from dairy cattle and farms managed organically and conventionally had similar patterns of antimicrobial resistance (Halbert et al., 2006). However, the proportion of antimicrobial-resistant isolates was higher for conventional farms. Samples and data were collected every 2 months from 128 farms. Most isolates from both farming systems were susceptible to most antibiotics evaluated. On both farm types, low proportions of isolates were resistant to ampicillin (<10%) and moderate proportions (30%–60%) demonstrated resistance to kanamycin, sulfamethoxazole, and tetracycline. The majority of tetracycline-resistant isolates encoded the tetB determinant.
Greater resistance in organic than conventional farms for these antibiotics.
Other pathogens
Although the majority of antibiotic use in dairy cows is for treatment and prevention of mastitis, antibiotics are also used routinely for treatment of bovine respiratory disease (pneumonia), foot rot, metritis (uterine infection), and pink eye (ocular condition), all of which are caused by bacteria (McAllister et al., 2001). Unlike for mastitis, there is no substantive body of literature pertaining to development of antimicrobial resistance associated with administration of antibiotics for treatment of bovine diseases outside of the udder. It is, however, conceivable that the associated pathogens could develop antimicrobial resistance over time. On the other hand, the majority of these bovine pathogens (Table 7) have little or no clinical significance in humans.
Antimicrobial-Resistant Bacteria in the Dairy Environment
On most dairy farms, different combinations of antibiotics and routes of treatment are used for prevention and control of diseases affecting dairy cows. Antimicrobial resistance among bacteria isolated from food-producing animals (Aarestrup et al., 1998) and from the dairy farm environment (van Dijck and van de Voorde, 1976) has been reported. Food-producing animals have been suggested as a possible reservoir of both antimicrobial-resistant bacteria and antimicrobial resistance genes in the microbiome that could be transferred to humans either directly via the food chain or indirectly as a result of spread of animal waste on cropland. Data from CDC showed that the emergence of antimicrobial-resistant bacteria correlated with increased use of antimicrobial agents (Nwosu, 2001). Endogenous bacterial flora may play an important role as acceptors and/or donors of transmissible drug resistance genes. E. coli are commonly found in the intestinal tract of humans and animals and are also implicated in human and animal infectious diseases. Virtually, all research to date on antimicrobial-resistant bacteria in the environment has focused on bacteria of fecal origin and selective pressure of antimicrobials released through feces and urine on development of antimicrobial resistance in indigenous soil bacteria. Microorganisms move easily between multiple ecosystems: from human and animals to soil and water and vice versa. Thus, antimicrobial resistance genes acquired by organisms in one ecosystem can be transferred among organisms in other ecosystems. In addition, there is a greater global mobility of bacteria in humans and food, potentially facilitating the spread of microorganisms and their genes around the world.
Srinivasan et al. (2004) evaluated antimicrobial susceptibility, genotypic characterization of antimicrobial resistance, and presence of class 1 integrons in C. jejuni (n = 39) and Salmonella spp. (n = 12) isolated from the dairy farm environment. All these foodborne pathogens were sensitive to chloramphenicol. C. jejuni was also sensitive to erythromycin, gentamicin, and streptomycin; Salmonella spp. were sensitive to florfenicol. C. jejuni was resistant to trimethoprim and vancomycin; Salmonella spp. showed resistance to ampicillin, cephalothin, penicillin, rifamycin, streptomycin, and vancomycin. The intI integron was found in 56.4%, and 100% of C. jejuni, and Salmonella spp., respectively. C. jejuni (5.1%) and Salmonella spp. (100%) contained more than one antimicrobial resistance gene. tetA was found in 15.4% and 100% of C. jejuni and Salmonella spp., respectively; tetC was found only in Salmonella spp. In Salmonella spp., strA (100%), strB (83.3%), sulI (100%), ermB (58.3%), and penA (50%) were amplified and all Salmonella spp. were MDR. Results of this study indicated that a high prevalence of foodborne pathogens isolated from the dairy farm environment contain integrons and/or antimicrobial resistance genes. The potential exists for foodborne pathogens carrying antimicrobial resistance genes with intI to acquire additional resistance genes as well as to spread this genetic material to commensal and pathogenic bacteria in the dairy farm environment. These pathogens can act as reservoirs for antimicrobial resistance genes and transfer these genes to closely related species or to other bacteria. Bacteria not carrying antimicrobial resistance genes may have other physiological mechanisms, for example, decrease of outer membrane permeability, activation of efflux pump, or a mutation in a ribosomal protein gene (Yan and Taylor, 1991).
Antimicrobial resistance of Listeria monocytogenes (n = 38) isolated from four dairy farms to 15 antimicrobial agents was evaluated (Srinivasan et al., 2005). All L. monocytogenes isolates evaluated were resistant to cephalosporin (MIC ≥512), streptomycin (MIC ≥32), and trimethoprim (MIC ≥512). Most L. monocytogenes isolates were resistant to ampicillin (92%, MIC ≥2), rifampicin (84%, MIC ≥4), rifamycin (84%, MIC ≥4), and florfenicol (66%, MIC ≥32), and some were resistant to tetracycline (45%, MIC ≥16), penicillin G (40%, MIC ≥2), and chloramphenicol (32%, MIC ≥32). All L. monocytogenes isolates were susceptible to amoxicillin, erythromycin, gentamicin, kanamycin, and vancomycin. Susceptibility of L. monocytogenes to the antimicrobials evaluated was quite consistent among the dairy farms evaluated. Nineteen of 38 L. monocytogenes isolates contained more than one antimicrobial resistance gene sequence. A high frequency of floR (66%) was found in L. monocytogenes followed by penA (37%), strA (34%), tetA (32%), and sulI (16%). Results of this study demonstrated that L. monocytogenes isolated from the dairy farm environment were resistant to many antimicrobials and carried one or more antimicrobial resistance genes.
The prevalence of selected tetracycline and streptomycin resistance genes and class 1 integrons in Enterobacteriaceae (n = 80) isolated from dairy farm soil and nondairy soils was evaluated (Srinivasan et al., 2008). Among 56 bacteria isolated from dairy farm soils, 36 (64.3%) were resistant to tetracycline and 17 (30.4%) were resistant to streptomycin. Lower frequencies of tetracycline (9 of 24 or 37.5%) and streptomycin (1 of 24 or 4.2%) resistance were observed in bacteria isolated from nondairy soils. Bacteria (n = 56) isolated from dairy farm soil had a higher frequency of tetracycline resistance genes, including tetM (28.6%), tetA (21.4%), tetW (8.9%), tetB (5.4%), tetS (5.4%), tetG (3.6%), and tetO (1.8%). Among 24 bacteria isolated from nondairy soils, four isolates carried tetM, tetO, tetS, and tetW in different combinations, whereas tetA, tetB, and tetG were not detected. Similarly, a higher prevalence of streptomycin resistance genes, including strA (12.5%), strB (12.5%), ant(3′′) (12.5), aph(6)-1c (12.5%), aph(3′′) (10.8%), and addA (5.4%), was detected in bacteria isolated from dairy farm soils than in nondairy soils. A higher distribution of MDR genes was observed in bacteria isolated from dairy farm soil than in nondairy soil. Results of this small study suggest that the presence of multiple resistance genes and class 1 integrons in Enterobacteriaceae in dairy farm soil may act as a reservoir of antimicrobial resistance genes, and could play a role in the dissemination of these antimicrobial resistance genes to other commensal and indigenous microbial communities in soil. In addition, this could lead to an increase because of higher total bacterial populations bringing donor and recipient bacteria into closer physical proximity. The higher prevalence of antimicrobial-resistant bacteria in soil from dairy farms may be due to routine use of antimicrobials for the prevention and control of diseases affecting dairy cows, resulting in enhanced antimicrobial selection pressure. However, additional studies are required to support this hypothesis. This study also points out clearly that antimicrobial-resistant bacteria occur frequently in soils where antibiotic pressure is presumably very low.
Are Science-Based Data Available to Demonstrate That Antimicrobial Resistance in Veterinary Pathogens That Cause Disease in Dairy Cows Impacts Pathogens That Cause Disease in Humans?
To date, there are no studies that show that use of antimicrobials to treat mastitis in dairy cows has resulted in the emergence and establishment of dominant antimicrobial-resistant clonal types in both human and dairy cattle populations. However, it is well documented that use of antimicrobials in both human and veterinary medicine will inevitably over time result in the emergence of antimicrobial-resistant bacteria. Antimicrobial-resistant bacteria become a significant public health issue when a resistant determinant becomes established as the dominant clonal type in humans (e.g., MRSA) or animals (ceftiofur-resistant Salmonella).
Raw food and undercooked food products of plant and animal origin can be a source of antimicrobial-resistant pathogenic bacteria such as Salmonella, enterotoxigenic E. coli, and C. jejuni (Aarestrup et al., 2008; Hammerum and Heuer, 2009; Presi et al., 2009; Sammarco et al., 2010). With respect to foodborne pathogens originating from dairy cows and the dairy environment, raw milk has received the most attention (Oliver et al., 2009). The dairy farm environment and animals on the farm serve as important reservoirs of pathogenic and commensal bacteria that could potentially gain access to milk in the bulk tank via several pathways: from infected mammary glands, contaminated udders and milking machines, fecal contamination, and/or from the dairy farm environment. Contaminated raw milk when consumed by humans or fed to animals on the farm can result in gastroenteric infections, and this scenario becomes much more complicated when pathogenic bacteria such as Salmonella, STEC, and commensal Gram-negative enteric bacteria encode for antimicrobial resistance determinants. Commensal bacteria, especially Gram-negative bacteria, can be a significant reservoir of antimicrobial determinants (Straley et al., 2006).
There are several well-documented reports on the prevalence of foodborne pathogens in raw milk and their involvement in disease outbreaks after consumption of raw (unpasteurized) milk and milk products made from raw milk (CDC, 1983, 2001, 2003, 2007; Jayarao and Henning, 2001; Mazurek et al., 2004; Jayarao et al. 2006). For a recent review of this topic, see Oliver et al. (2009). Salmonella isolated from an outbreak after consumption of cheese made from raw milk in Washington State (Vilar et al., 1999) and Northern California (Cody et al., 1999) and from raw milk in Arizona (Tacket et al., 1985) were observed to be MDR. As compared to raw milk, far fewer related foodborne outbreaks have been associated with pasteurized milk; in almost all instances, contamination was caused by postpasteurization contamination (Ackers et al. 2000; Olsen et al., 2004; CDC, 2008; Oliver et al., 2009).
To safeguard public health from exposure to antimicrobials through milk, stringent regulatory practices are implemented to ensure that milk for human consumption is free of antibiotic residues (Grade “A” Pasteurized Milk Ordinance, 2007 Revision). This practice has resulted in milk as one of only a few products that is free of antibiotic residues.
Use of antimicrobials to treat other bacterial infections of dairy cattle has resulted in emergence of antimicrobial-resistant pathogenic (e.g., bla CMY-2-resistant Salmonella) and commensal (e.g., bla CMY-2-resistant E. coli) bacteria (Sawant et al., 2005; Donaldson et al., 2006). There are reports on the occurrence of MDR Salmonella in dairy cattle and their link to human illnesses (Holmberg et al., 1984; CDC, 2002; Zansky et al., 2002; Edrington et al., 2008).
Does Antimicrobial Resistance Impact the Outcome of Therapy?
Antimicrobial resistance decreases success of bacterial pathogen eradication or control. There are several types of evidence (Swartz, 2002) that link the risks of humans becoming infected with antimicrobial-resistant pathogens that emerged through use of drugs in food-producing animals (Table 8). Antimicrobial resistance is a concern in dairy cattle production systems, where antibiotic-resistant pathogens can contribute to increased morbidity and mortality of livestock with commensurate increases in production expenses for livestock producers (Mathew et al., 2007). From a public health perspective, there is potential for antimicrobial-resistant pathogens and commensal organisms to disseminate to humans via direct contact with animals or via the food chain (Call et al., 2008). Many of the antibiotics that could potentially lead to antimicrobial-resistant human pathogens have not been approved for use in North America (McAllister et al., 2001). A direct linkage between the outcome of therapy in humans and association of antimicrobial resistance from pathogens acquired from dairy animals is rare.
Each year in the United States, Salmonella species cause an estimated 1.4 million infections, 16,000 hospitalizations, and nearly 600 deaths (Mead et al., 1999). MDR Salmonella are associated with increased bloodstream infections, hospitalizations, and deaths compared to pansusceptible strains (Greene et al., 2008). Emergence of antimicrobial-resistant Salmonella has been traced to dairy farms in a few prominent studies (Lyons et al., 1980; Spika et al., 1987). Spika et al. (1987) established a dairy animal-to-human transmission of antimicrobial-resistant Salmonella Newport and concluded that food-producing animals are a major source of antimicrobial-resistant Salmonella infections in humans. Lyons et al. (1980) also reported animal-to-human spread of MDR (resistant to chloramphenicol, sulfamethoxazole, and tetracycline) Salmonella Heidelberg to hospital patients. The epidemic was remarkable in that it had its origins in infected dairy calves.
Mulvey et al. (2009) established a potential linkage of MDR E. coli in food-producing animals and humans in Canadian hospitals. They concluded that plasmids that carried MDR could limit options for treatment of human infections caused by these strains. A similar phenomenon has been observed in humans whereby infections with antimicrobial-resistant Salmonella tend to be more severe and have higher hospitalization and/or mortality rates than those caused by pansusceptible strains (Helms et al., 2002, 2004; Varma et al., 2005). Antimicrobial resistance was widespread among dairy farms with isolates from clinically ill cattle demonstrating greater antimicrobial resistance (Cummings et al., 2010).
Phillips et al. (2004) provided an alternative viewpoint that emphasized that although some antibiotics are used in both animals and humans, most resistance problems in humans have arisen from human use. There has been no correlation between the carriage of antimicrobial resistance in farm animals and carriage of resistant enterococci of animal origin and human infection presumably because animal enterococci do not establish themselves in the human intestine. Commensal E. coli also exhibit host animal preferences. Antimicrobial resistance is a microbiological phenomenon that may or may not have clinical implications depending on pharmacokinetic and pharmacodynamic parameters as they apply to specific antibiotics. Phillips et al. (2004) concluded that resistance acquired in animals may add, although very little, to the burden of human disease even when infection occurs. However, in the case of Salmonella and Campylobacter spp., risk analyses suggest that resistance possibly acquired in animals may, to a small degree, add to the burden of human disease.
In a study by Chow et al. (1991), MDR Enterobacter spp. in patients initial blood culture were associated with a higher mortality rate (12 of 37, 32%) than was isolation of a more sensitive Enterobacter spp. (14 of 92, 15%; p = 0.03). In a study involving Enterobacter spp., S. marcesens, Citrobacter freundii, and Morganella morganii resistance to broad-spectrum cephalosporins during antimicrobial therapy, the emergence of resistance was associated with a low risk of mortality and was more prevalent in Enterobacter spp. (Choi et al., 2008). Deal et al. (2007) identified predictors of in-hospital mortality among patients with bacteremia caused by Enterobacter cloacae, Enterobacter aerogenes, or C. freundii. Their data indicated that among patients with Enterobacter spp. or C. freundii bloodstream infections, those with trimethoprim-sulfamethoxazole-resistance or second or third-generation cephalosporin-resistant strains of these bacteria had an increased risk of mortality.
In pediatrics, important recent developments in the area of antimicrobial resistance include the emergence of clinically relevant resistance to β-lactam antibiotics (including cefotaxime and ceftriaxone) in Streptococcus pneumoniae and the emergence and spread of infections caused by community-acquired strains of MRSA (English and Gaur, 2010). In both pediatric and adult medicine, development of resistance to multiple classes of antibiotics has transformed enterococci from a second-rate pathogen to a first-rate problem. Antibiotic use in dairy cattle has not been linked to development of antimicrobial-resistant Staphylococcus or other pathogens that are a primary concern to human health (McAllister et al., 2001).
Prevention and Control of Mastitis
Bovine mastitis is one of the most important bacterial diseases in dairy cattle throughout the world. Antimicrobials are used frequently in dairy cows for mastitis therapy and prophylaxis. To control mastitis and to avoid potential problems associated with bacterial antimicrobial resistance and treatment failure, it is important to be aware of antimicrobial resistance characteristics of mastitis pathogens. Resistance surveillance programs are usually aimed at assessing resistance phenotypes. These resistance phenotypes may arise from many different genetic determinants and each determinant may present specific epidemiological features. Therefore, assessment of antimicrobial resistance at the genetic level would be an important asset in the understanding and control of antimicrobial resistance in general. Antibiotic therapy of clinical mastitis involves (1) detection of the infected quarter, (2) prompt initiation of treatment, (3) administration of the full series of recommended treatments, (4) maintaining a set of treatment records, (5) identification of treated cows, and (6) making sure milk is free of antibiotic residues before adding to the bulk tank.
There continues to be concern over the low efficacy of antibiotic mastitis therapy against certain mastitis pathogens. Efficacy of mastitis therapy is extremely low for chronic S. aureus infections, and β-lactamase production may be partly responsible for the low cure rate. However, even with antibiotics to which the bacteria were sensitive in vitro, the cure rate was still low (Owens et al., 1997). This suggests the presence of other environmental or physiological mechanisms that interfere with therapy such as formation of microabscesses in mammary tissues and internalization into phagocytic and epithelial cells (Almeida et al., 1996), including formation of biofilms (McAllister et al., 2001). Most antibiotics used in mastitis therapy do not penetrate into the infected area and have poor intracellular penetration. Pirlimycin has been studied extensively to treat cows with chronic S. aureus IMIs because of its lower MIC and its tissue-penetrating property.
Results of research (Owens et al., 1997; Deluyker et al., 2000; Gillespie et al., 2000; Oliver et al., 2004a, 2004b) support the concept that extended intramammary antimicrobial therapy is significantly more effective at eliminating natural and experimentally induced mastitis than standard intramammary treatment regimens. It would appear that lengthening the duration of antibiotic therapy increases treatment efficacy. This has been demonstrated for ceftiofur and pirlimycin against a variety of mastitis pathogens, including S. uberis, and other environmental Streptococcus species, S. aureus, Corynebacterium bovis, and CNS. Effectiveness of extended antimicrobial therapy must be weighed against several factors, including the price of the antibiotic, loss of milk due to withholding time, marketability of the milk, potential of infecting the cow through repeated infusions via the teat canal, increased milk production following elimination of the chronic infection, reduced spread of contagious mastitis pathogens, and reduced culling because of a greater emphasis on milk quality. Caution should be taken to avoid extended antibiotic therapy during stressful situations for the animals such as heat stress. Studies to evaluate economic benefits of extended antimicrobial therapy need to be conducted to fully evaluate costs and benefits associated with this type of therapy.
The importance of the nonlactating (dry) period in the control of mastitis in dairy cows has been recognized for 60 years. A classic study by Neave et al. (1950) demonstrated that udders were markedly susceptible to new IMIs during the early dry period. Studies have also shown that udders are highly susceptible to new IMIs during the periparturient period (Oliver and Mitchell, 1983, 1984; Smith et al., 1985a, 1985b; Oliver, 1988a, 1988b; Oliver and Sordillo, 1988). Increased susceptibility to new IMIs is likely associated with physiological transitions of the mammary gland either from or to a state of active milk production. Many IMIs that occur at this time persist throughout the dry period and are often associated with clinical mastitis after calving. Thus, the dry period was identified as an extremely important time for the control of mastitis in dairy cows.
Since the early work by Neave et al. (1950), effective preventative procedures were developed to control infections during the dry period. Most dairy advisors recommend that all mammary quarters of all cows be infused with antibiotics approved for use in dry cows following the last milking of lactation. Objectives of dry cow therapy are twofold: (1) to eliminate infections present during late lactation, and (2) to prevent new infections during the early dry period when mammary glands are highly susceptible. Dry cow therapy is particularly effective against streptococci and to a lesser extent against S. aureus. Smith et al. (1985a, 1985b) demonstrated that antibiotic therapy at drying off reduced the rate of new environmental streptococcal infection during the early dry period only and that the rate of new coliform IMIs was not affected at all. Thus, two significant limitations of present antibiotic formulations used for dry cow therapy are (1) ineffectiveness against coliform bacteria, which can cause a high proportion of IMIs during the early dry period and near calving, and (2) ineffectiveness in preventing new IMIs by a broad spectrum of mastitis pathogens during the period near calving when mammary glands are highly susceptible to new infection (Oliver, 1988a, 1988b; Oliver and Sordillo, 1988, 1989).
Research has also shown that mastitis in breeding age and pregnant heifers is much higher than previously thought. A review on this topic was published recently (Oliver et al., 2005). Many IMIs in heifers can persist for long periods, are associated with elevated somatic cell counts, and may impair mammary development during gestation and affect milk production after calving. Presence of mastitis before calving increased the risk of infection during lactation, increased the risk of clinical mastitis in the first week after calving, and increased the risk of further cases of mastitis and culling during the first 45 days of lactation. In some studies, prepartum intramammary antibiotic infusion of heifer mammary glands was shown to be an effective procedure for eliminating many infections in heifers during late gestation and for reducing the prevalence of mastitis in heifers both during early lactation and throughout lactation. Prepartum antibiotic-treated heifers also produced significantly more milk than control heifers and had significantly lower somatic cell counts scores than untreated control heifers. These observations are likely due to the lower prevalence of mastitis pathogen isolation in prepartum antibiotic-treated heifers throughout lactation (Oliver et al., 2005).
Are Antibiotics Used Prudently in the Dairy Industry?
Over the last 70 years, antibiotics have saved millions of lives and their use has contributed significantly to improving human and animal health. Ironically, this very achievement now threatens the future usefulness of antibiotics, as some bacteria are becoming resistant to antibiotics (IFT Expert Panel, 2006). Antimicrobial resistance is now a global issue. Internationally, and more so in developed countries, public health officials have increased communication efforts to reduce indiscriminate use of antibiotics. International agencies, including the Food and Agriculture Organization (FAO) of the United Nations, the World Health Organization (WHO), and the World Organization for Animal Health (OIE), have stressed the need to find alternative approaches to address the issue of antimicrobial use in food-producing animals and its effect on the emergence of antimicrobial resistance of human pathogens (FAO/WHO/OIE, 2007). In the United States, growing demand from human health organizations, nongovernmental agencies, and consumers on curtailing use of antimicrobials in food-producing animals has resulted in development of programs and policies that favor prudent use of antimicrobials (Torrence, 2001). To address the concern of antimicrobial use in food-producing animals, including dairy cows, the American Association of Bovine Practitioners (AABP) and other groups have taken pro-active measures through education and outreach programs for veterinarians such that they can work effectively with their clientele on judicious or prudent use of antimicrobials. The AABP Prudent Drug Usage Guidelines for Cattle (
There are several guidelines that have been developed and are employed in prudent use of antimicrobials in bovine medicine, of which the best documented evidence is that of use of antimicrobials for dry cow therapy. Dry cow therapy has been in practice for almost four decades. Studies have shown clearly that β-lactam antimicrobials and other classes of antimicrobials used for dry cow therapy are as effective today as they were when first used for treating dry cows. This clearly demonstrates that guidelines for prudent use of antimicrobials for dry cow therapy that were developed are effective at preventing the emergence of antimicrobial resistance of bovine mastitis pathogens.
Guidelines for antimicrobial use serves as the first effective approach toward judicious use of antimicrobials. A major purpose of the guideline should be to minimize inappropriate use that includes the type of drug used, dosage, and duration of treatment. Guidelines should be developed in consultation with veterinarians, microbiologists, epidemiologists, the pharmaceutical industry, state and federal regulatory agencies, and nonprofit organizations such as the NMC (formerly referred to as the National Mastitis Council), American Veterinary Medical Association (AVMA), and AABP.
Morley et al. (2005) suggested that antimicrobials used for treatment be categorized such as primary, secondary, and tertiary, or as first, second, and third line. The following categorization was suggested with the view that the higher class of antimicrobials (e.g., such as third-generation cephalosporins) should only be used when lower classes of antibiotics are not appropriate. The decision to use/select the appropriate class should generally be on culture-based antibiotic sensitivity testing. On dairies, use of a particular antimicrobial should also take into consideration the relative importance of the antimicrobial in human medicine (WHO, 2005), and regulations that are placed for a class of antimicrobials being considered for therapy. Categorization of antimicrobials can be practice specific or disease condition specific, and the approach to categorization should include the species of food animal and antimicrobial resistance patterns (Weese, 2006).
Guidelines for prudent use of antimicrobials have been developed by several veterinary groups; most are relatively generic and address only a few key points (Weese, 2006). Some veterinary practices and veterinary teaching hospitals have developed specific policies and guidelines for antimicrobials. However, widespread use and implementation of these guidelines are yet to become a reality. One major constraint is lack of proper methods to document outcomes and success of using a guideline for prudent use of antimicrobials. This is critical as widespread use of guidelines developed for prudent use antimicrobials will only become a reality once outcomes can be measured objectively (e.g., emergence of antimicrobial resistance, continued effectiveness of an antimicrobial of concern).
Use of antimicrobials in humans and animals is governed by policies that are placed primarily to protect public health. Thus, use of antimicrobials in food-producing animals has always been under close scrutiny, and for nearly 30 years legislators have introduced bills to enforce stricter bans on use of antimicrobials in food animal practice. Several organizations such as the Union of Concerned Scientists, The Pew Trust, and the American Medical Association have called for stricter regulation or “phasing out” use of antimicrobials in food animal practice, starting with antimicrobials in feed for prophylactic purposes. The primary contention of these organizations is that use of antimicrobials is not regulated and antibiotics are used extensively in food animal practice, which has resulted in increased antimicrobial-resistant bacteria in the human population. The AVMA is of the opinion that science-based data and decisions should be considered before any legislative action is called for. Further, AVMA stated that decisions on use of antimicrobials should be entrusted to veterinarians who have the knowledge, skill, and training on prudent use of antimicrobials. Veterinarians are considered as the primary source of information to dairy producers on use of antibiotics, more so with educating producers on appropriate use of antibiotics and consequences that could result following misuse. To address the issue of antimicrobial resistance, prudent drug use guidelines developed by the AVMA state that “Prophylactic or metaphylactic use of antimicrobials should be based on a group, source, or production unit evaluation rather than being utilized as standard practice” (DHHS:FDA:CVM, 2000).
Development of standardized use of antimicrobials for common situations has been advocated (Morley et al., 2005) with the understanding that the goals do not dictate the practice of medicine, but rather to supplement clinical judgment and ensure that the art of medicine remains science based. In some veterinary practices, standardized protocols for treatment of clinical mastitis are followed. Standardized protocols are developed and used for common situations, more so to complement clinical decisions such that treatment decisions are closer to evidence-based medicine. Like any other field, veterinary medicine is continuously evolving, and veterinarians are at the forefront of utilizing cutting edge technologies and therapies. These efforts could also perhaps include pressure to use new antimicrobials used in human medicine but rarely used in animals. This may be more relevant in small animal or companion animal practices but highly unlikely in food animal practice, particularly in dairy cows. Food animal veterinarians and producers understand that their actions have a direct affect on contaminating the human food supply affecting human health, and therefore they refrain from attempting new and undocumented therapies.
Friedman et al. (2007) conducted in-person interviews with 20 dairy farmers in rural counties of South Carolina to determine farmers' knowledge and attitudes about prudent antibiotic use among livestock. Results showed that participants (100%) typically determined a need for antibiotic treatment using symptom assessment and reported following some form of operating procedures regarding administration of antibiotics. Few farmers (32%) had actual written antibiotic protocols. Preferred information sources about antibiotics were veterinarians (100%) and other dairy farmers (50%). Most farmers (86%) were not concerned that overuse of antibiotics in animals could result in antimicrobial resistance among farm workers. Qualitative analysis of focus groups revealed significant barriers to following proper antibiotic procedures, including limited finances and lack of time. The need for bilingual educational resources for Hispanic/Latino dairy workers was expressed. Desired formats for educational materials were posters, flowcharts, videos, and seminars. Authors concluded that education of South Carolina dairy farmers by veterinarians and public health professionals on the appropriate use of antibiotics in dairy cattle is needed to ensure antibiotic effectiveness in both animals and humans.
A comprehensive study by Cattaneo et al. (2009) surveyed Ohio bovine practitioners to determine the knowledge, beliefs, and practices regarding antibiotic resistance. Knowledge of veterinarians who participated in the study regarding selection pressures that cause bacteria to acquire resistance was positively correlated with their knowledge of transmission routes of antimicrobial-resistant bacteria to dairy cows (r = 0.60) and potential consequences of antimicrobial-resistant bacteria with respect to animal health (r = 0.52). Veterinarians believed that antibiotics are used by producers to treat a variety of illnesses and that specific antibiotics (e.g., gentamicin) are used without veterinary consultation. More than 75% of veterinarians thought that one-on-one meetings and handouts containing good management practices, diagnosis descriptions, and appropriate dosages for antibiotics would be effective ways to educate their clients about antimicrobial-resistant bacteria. Only 23% of veterinarians consistently provided treatment protocols for antibiotic use. They concluded that rather than expending resources to develop educational materials directed at improving bovine practitioner knowledge of the subject, communication and outreach efforts that encourage and facilitate information flow from veterinarians to dairy producers may be more effective tools to affect prudent use of antibiotics on dairy farms.
The study by Cattaneo et al. (2009) also revealed that veterinarians preferred to obtain information from trusted print and personal sources. One-on-one meetings between veterinarians and producers ranked as the most effective way to educate producers about antibiotic resistance. This approach was not always feasible because of the time commitment required by veterinarians. Veterinarians felt that distribution of handouts containing management practices, diagnosis descriptions, and dosage guidelines provided a practical and realistic method for veterinarians to communicate information to dairy producers. Raymond et al. (2006) developed similar materials in English and Spanish. A key finding of the study by Raymond et al. (2006) revealed that veterinarians felt that use of antibiotics without prior veterinary consultation signified a serious communication gap. Thus, veterinarians should be encouraged to share their knowledge with dairy producers regarding the importance of using the appropriate antibiotic for the corresponding illness. Further, bovine practitioners should encourage dairy producers to seek veterinary consultation or a treatment protocol approved by the herd veterinarian before administering antibiotics. A survey of 113 dairy farms in Pennsylvania conducted between 2001 and 2002 revealed that only 21% of dairy producers had written protocols for treating sick cows, and only 32% contacted a veterinarian before administering antibiotics (Sawant et al., 2005). A 2003 study of 381 dairy producers in Washington State yielded very similar results with 27% of dairy producers having written treatment protocols (Raymond et al., 2006).
The issue of prudent use of antimicrobials has become complicated due to issues such as (1) access and availability of antimicrobials, (2) need for prescription, (3) policies that dictate and encourage the sale and use of antimicrobials, and (4) how individuals who administer antimicrobials come into play in a given country or region. For example, in several countries (e.g., in the European Union) and in the United States, antibiotics can be obtained from sources other than veterinarians. Expert reviews and committees in many countries have highlighted the need for better control of licensing of antibiotics, and codes for prudent use of antibiotics by veterinary practitioners and farmers (Cattaneo et al., 2009; US FDA, 2008). Communicating effective risk management solutions in agriculture requires a comprehensive understanding of how antibiotics are used within and across each food-animal industry. Improving the appropriate use of antibiotics requires participation of all stakeholders administering antibiotics, including veterinarians, producers, and animal handlers (Salisbury et al., 2002; Cattaneo et al., 2009).
Sischo (2006) provided a unique viewpoint on public health concerns and issues as they relate to bacterial antimicrobial resistance. The dairy industry, contrasted to other major animal commodities, is focused not on meat production but on milk production. Milk production is constrained by disease and antimicrobial treatment is a common management tool. Unlike many other animal agricultural systems where the value and safety of the product is measured in the future, the value of milk is zero when an antimicrobial is used in a lactating cow and milk must be discarded because of potential antibiotic residues. He further stated that the main food product, milk, is mainly pasteurized and all shipments of milk from the farm to the processing plant are tested for the presence of antimicrobials. This makes the likelihood of farm-origin antimicrobials or bacteria appearing in the finished product very low. This suggests that the use and quantity of antimicrobials in the dairy system has little direct impact on public health.
Conclusions
In the present review, we focused on antibiotic use in lactating and nonlactating cows in U.S. dairy herds, and addressed four key questions/issues: (1) Are science-based data available to demonstrate antimicrobial resistance in veterinary pathogens that cause disease in dairy cows associated with use of antibiotics in adult dairy cows? (2) Are science-based data available to demonstrate that antimicrobial resistance in veterinary pathogens that cause disease in adult dairy cows impacts pathogens that cause disease in humans? (3) Does antimicrobial resistance impact the outcome of therapy? (4) Are antibiotics used prudently in the dairy industry? Based on this review, we conclude that antibiotic use in adult dairy cows has not increased antimicrobial resistance of veterinary pathogens to antibiotics used routinely in the dairy industry. Scientific evidence does not support widespread, emerging resistance among mastitis pathogens to antibacterial drugs even though many of these antibiotics have been used in the dairy industry for treatment and prevention of disease for several decades. However, it is clear that use of antibiotics in food-producing animals does contribute to increased antimicrobial resistance. While antimicrobial resistance does occur, we are of the opinion that advantages of using antibiotics in adult dairy cows far outweigh the disadvantages. Clinical consequences of antimicrobial resistance of dairy pathogens affecting humans appear small. Antimicrobial resistance among dairy pathogens, particularly those found in milk, is likely not a human health concern as long as the milk is pasteurized. However, there are an increasing number of people who choose to consume raw milk. Transmission of an antimicrobial-resistant mastitis pathogen and/or foodborne pathogen to humans could occur if contaminated unpasteurized milk is consumed; this is another very important reason why people should not consume raw milk. Likewise, antibiotic-resistant bacteria contaminating meat from dairy cows should not be a significant human health concern if the meat is cooked properly.
We emphasize and advocate prudent use of antibiotics in the dairy industry; it is important, worthwhile, and necessary. Use of antibiotics at times when animals are susceptible to new infection is a sound management decision and a prudent use of antibiotics on the farm. Strategies involving prudent use of antibiotics for treatment encompass identification of the pathogen causing the infection, determining the susceptibility/resistance of the pathogen to assess the most appropriate antibiotic to use for treatment, and a long enough treatment duration to ensure effective concentrations of the antibiotic to eliminate the pathogen. Last, as this debate continues, we need to consider the consequences of “What would happen if antibiotics are banned for use in the dairy industry and in other food-producing animals?” The implications of this question are far reaching and include such aspects as animal welfare, health, and well-being, and impacts on food quantity, quality, and food costs. This question should be an important aspect in this ongoing and controversial debate.
Footnotes
Acknowledgment
This work was supported, in part, by a grant from the Midwest Dairy Association.
Disclosure Statement
No competing financial interests exist.