INTRODUCTION

The introduction of antimicrobial drugs into agriculture and veterinary medicine shortly after the Second World War caused a revolution in the treatment of many diseases of animals. In the “wonder drug era” of the late 1940s and early 1950s, the effective treatment of many infections that were previously considered incurable astonished veterinarians, such that some even feared for their livelihoods. Not all use of antimicrobial drugs in food animals is yet under veterinary prescription globally, despite repeated recommendations by the World Health Organization and other responsible organizations, so that the term “veterinary medicine” is used here rather generically to suggest use in animals rather than just use by veterinarians.

A broad overview of key features of the history of antimicrobial drug use in animals is given in Table 1, which traces developments from the preantibiotic era to the present day, where there are arguable fears that we are moving into the “postantibiotic” era but which may better be described as the antimicrobial stewardship era. Much of this overview will focus on antimicrobial use in food animals, the subject of an earlier review that partly focused on the public health aspects of the use of antimicrobials in food animals (1), but will include important aspects of use in companion animals.

Table 1 Historical time line of important events and trends in the use of antimicrobial drugs in animals, with emphasis on food animals

The table illustrates many important features in the history of the use of antimicrobials in animals: (i) The development of resistance to antimicrobial drugs followed soon after their introduction. (ii) Resistance was usually dealt with by the development of new classes of antimicrobials by the isolation from nature of novel antibiotics within a particular class or by development of synthetic analogs of an existing class. (iii) The antimicrobial drugs used in animals were the same as those in human medicine, although a number of them that were rejected by human medicine because of toxicity problems (e.g., bacitracin) became growth-promoting feed antimicrobials in food animals. At least one of these drugs (colistin) is now being reclaimed for systemic use in humans. (iv) Antimicrobials were used in agriculture in feed as growth promoters, or for subtherapeutic purposes, almost as soon as they were discovered. (v) The majority of antimicrobial drugs used in animals (and shared with human medicine) belong to a small number of major classes, and only one major new class of antimicrobial drugs (fluoroquinolones; pleuromutilins are an exception, but have very restricted use) has been introduced for food animal use in the past 30 years. (vi) Significant antimicrobial contamination in carcasses or selected tissues was detected in the 1970s and 1980s, leading either to the banning of potentially toxic (e.g., carcinogenic or idiosyncratic effects) drugs or to rigorous, ongoing programs of detection in carcasses after slaughter. This was the major focus of regulations of use in food animals in those decades, and there is still confusion in some quarters about the difference between residues and resistance. (vii) The public health impacts resulting from the development of resistance, and especially because of transmissible resistance, have been a major battleground between agriculture and medicine for nearly 50 years. (viii) The resistance crisis in human medicine has led to unprecedented concern at the highest political levels globally about the threat of resistance to humanity, to an unprecedented focus on stewardship, and to major ongoing reduction and ongoing changes in agricultural use of antibiotics, at least in the developed world. (ix) The science and practices supporting optimal antimicrobial drug use in animals and in humans has developed relatively slowly and is not complete.

DISCOVERY OF ANTIBIOTICS AND EARLY USAGE

Antimicrobial drugs were introduced for animal (and human) use with a minimum of controlled experimental studies, so that from the start of their use there were frequent calls to move from the wonder to the science. As in human medicine, much of the early dosage used was empirical and based on inadequately controlled small-scale trials (2, 3), so that there was a “confusing hodge-podge of widely divergent optimum dose-ranges for the many livestock diseases allegedly amenable to the activity of penicillin” (4). In the United States such empiricism led to a licensed dosage of penicillin G in cattle that was clearly inadequate. It took four or five decades before the licensed drug dosage was more scientifically determined, based on quantitative understanding of the interaction of drug with the target microorganism (dosage, pharmacokinetic and pharmacodynamic parameters, in vitro susceptibility) as well as clinical data (Table 1). Clinical evaluation is still an important component used in the licensing of antimicrobial drugs, in part because the predictive science is imperfect.

In retrospect, for drugs whose use “has advanced the practice of medicine farther than any other single factor of any of the previous centuries” (5), the time taken to establish the science of the clinical use of antimicrobial drugs seems astonishing. As human medicine’s poor cousin, veterinary medicine lagged in the development of the science of optimal antimicrobial drug use, but the lag was only relative since the same delay was clearly visible in human medicine. In general, the science and practice of antimicrobial usage in animals has largely paralleled that in human medicine, in the same way that most antimicrobial drugs used were the same or were in the same drug class. There have been, however, a number of features unique to animal agriculture, as discussed below.

In most countries, approval by the appropriate regulatory authority must be obtained before an antimicrobial drug can be legally sold, and this depends on extensive testing to ensure safety and efficacy, as well as, in the case of food animals, studies of safety for people consuming their products. Registration requirements for veterinary medical products have been largely harmonized internationally under the International Cooperation on Harmonization of Technical Requirements of Veterinary Medicinal products (VICH) of the World Organization for Animal Health, membership of which includes the European Union and the United States. A harmonized VICH guideline, GL 27, defines the data requirements for risk of transfer of resistant bacteria or resistance determinants from foods of animal origin to humans. These data are assessed in terms of exposure of food-borne pathogens and commensal bacteria, and the “qualitative probability” that human exposure to resistant bacteria results in adverse human consequences. As part of this assessment, many countries are attempting to stratify the stringency of regulatory requirements by how important a drug is to public health, also discussed below. This is a highly contentious issue, since most antimicrobial drugs can, under various criteria, be claimed as “critically important.” A more useful and a more rational approach, which could be adopted in both human and veterinary medicine, is categorization into “lines” based in part on culture and susceptibility results (Table 2) (6). This approach has the advantage of simplicity (for example, labels on bottles could indicate the category) and would enhance the use of laboratory diagnosis. Research is needed into whether such a categorization scheme would be accepted in the animal (and human) health world, including the barriers to acceptance and what it would take to implement such a system so that it would be widely accepted.

Table 2 Suggested categorization of antimicrobial drugs for veterinary use (6)

The regulation of antimicrobial drug use in animals is a complex process that has jurisdictional differences. Regulation is more stringent for use of these drugs in food animals, although “off- or extra-label” use (use of the product in any manner not specified on the label) is often approved under specific circumstances and constraints. Use of antimicrobial drugs in companion animals is subject to less stringent regulation, and there is likely more off-label use in companion animals (as there is in human medicine).

Although the situation is changing, there has been historically no formal interest by regulatory authorities in postapproval use (or periodic relicencing) of antimicrobial drugs in animals. Some of the label claims for some antimicrobial drugs list approval for use in bacteria that have had their names changed several times since approval or for diseases that have subsequently been shown to be caused by other agents, so that reading the labels can be like reading a outdated veterinary microbiology textbook from the 1960s. The postapproval monitoring of resistance in Campylobacter jejuni following the introduction of fluoroquinolones for use in broiler chickens in the United States, and the subsequent withdrawal of fluoroquinolones from use in poultry in the United States is, however, a well-known example of postapproval monitoring of the approved use of a drug.

PRACTICES IN ANTIMICROBIAL DRUG USAGE UNIQUE TO ANIMAL HUSBANDRY

The greatest differences in usage of antimicrobial drugs between animal husbandry and human medicine were, and in many countries still are, in the use in agriculture of antimicrobials for growth promotion and for long-term disease prophylaxis, although the situation is changing relatively rapidly. This has occurred particularly in countries where livestock, notably chickens and pigs, are reared intensively.

The growth-promotional benefits of adding low concentrations of many antibacterial drugs to feed was recognized almost as soon as antibiotics were introduced. The enhancement of growth rates and improved efficiency of use of feed were noted when pigs and poultry were fed the fungal waste derived from antibiotic production, originally intended as a source of vitamins and protein, but mostly as an efficient way to use the waste. The effect was originally attributed to the presence of vitamin B12 (“animal protein factor”) in the mycelial mass, but with time it was recognized to be a direct effect of residual antibiotic. Interestingly, how antimicrobial drugs improve growth rate and efficiency of utilization is still unknown, although it is thought to be through an inhibitory or metabolic effect of some kind on the Gram-positive intestinal microflora. Curiously, until about the mid-1950s, low prolonged oral dosage of tetracycline was even used to improve the growth of underweight human infants and children, but this practice was dropped because of both resistance and discoloration of teeth. In animals, the growth-promotional and disease-prophylactic benefits appear to have remained constant over the years (7), supporting the idea that these effects result from metabolic rather than antibacterial activities. As the use of antibiotics as growth promoters in intensively reared livestock becomes illegal in much of the world, alternative approaches to manipulation of the microbiome may replace their growth-promoting effects.

Not only have antimicrobial drugs been used for growth promotion, but some drugs were and in some countries still are administered in feed for prolonged periods at somewhat higher concentrations, the “subtherapeutic levels” (defined in the United States as less than 200 g per ton of feed), which are lower concentrations than those approved for therapeutic purposes. The historical origin of subtherapeutic usage and indeed the meaning of this term are obscure, but it seems to have both beneficial growth-promotional and disease-prophylactic effects, particularly against pathogens that do not readily develop or acquire resistance. Drugs such as tetracyclines are administered “subtherapeutically” for many defined, licensed purposes at a range of concentrations varying with the drug, the food-animal species, and the purpose. Such usage, which can often be prolonged and thus inconsistent with important general principles of antimicrobial drug dosage (6), has been particularly widespread in the swine industry in countries in which the drugs are still allowed for this purpose (8). The practice is coming under increased scrutiny and will likely also be banned and replaced by short-term antimicrobial prophylaxis or short-term treatment targeted to specific pathogens. As noted earlier, a number of antimicrobial drugs (such as the streptogramins) that were too toxic for parenteral use in humans were relegated to growth-promotional and disease-prophylactic use in food animals.

Another practice that has historically been far more common in food animal use than in human medicine has been short-term mass medication with therapeutic concentrations of drugs immediately before an outbreak of disease can be anticipated, or immediately at the onset of disease in a population (9). This type of prophylaxis has been commonly practiced in beef feedlot and swine medicine and is most akin to the prophylactic use of antimicrobial drugs to prevent Haemophilus or Neisseria meningitis in humans. Prophylactic use of intramammary antimicrobial drugs to prevent development of new infections and to treat existing infections has become a standard practice in dairy cows in the two months before calving and re-entering the milking herd (“dry cow treatment”), with no apparent adverse effect on resistance development, perhaps in part because the very high concentrations of drugs achieved in the udder result in rapid killing of the target bacteria. Blanket use of dry cow treatment is also coming under scrutiny and is being replaced by use only when udders are known to be infected and likely to carry an infection over to the next lactation.

USE OF ANTIMICROBIAL DRUGS IN ANIMALS

PUBLIC HEALTH ASPECTS OF ANTIMICROBIAL DRUG USE IN ANIMALS

Food Animals

The effect of antimicrobial drug use in food animals on the development of resistance in bacteria that can cause disease in humans has been the subject of prolonged, acrimonious, and ongoing debate. The major and most accessible reviews of this issue are summarized in Table 3, which shows that the intensity of the criticism of agricultural usage of antimicrobial drugs intensified from the mid-1990s and has now reached a crescendo, paralleling the antimicrobial resistance crisis in human medicine (Table 1).

Table 3 Historical time line of major reports and their conclusions or recommendations relating to the public health aspects of antimicrobial drug use in food animals

The first major review of the effect of antimicrobial drug use on resistance in human and animal pathogens was carried out in the United Kingdom under the chairmanship of M.M. Swann. The impetus for the review was a combination of recognition of the increasing importance of (i) the phenomenon of “infectious,” transferable, drug resistance associated in part with the pioneering work of the distinguished British veterinary microbiologist H. Williams Smith, (ii) the emergence and dissemination in calves in Britain of multidrug-resistant Salmonella enterica serotype Typhimurium and its spread to humans, and (iii) experiences around this time of a difficult-to-control epidemic of chloramphenicol-resistant S. enterica serovar Typhi in Central America. Chloramphenicol was then the drug of last resort for typhoid fever in humans (25). The 1969 Swann Report to the British government gave a careful analysis of how different usage of antimicrobial drugs in animals might lead to selection of resistant bacteria and resistance plasmids and how such resistant bacteria, or their transmissible resistance traits, could lead to difficult to treat infections in humans. The major recommendations of the committee were as follows: (i) “Feed” antimicrobial drugs could only be used for growth promotion without prescription if they had little or no implication as therapeutic agents in humans, would not impair the value of prescribed drugs, and produced an economic benefit. Since penicillin and tetracyclines did not meet these criteria, they were withdrawn from growth-promotional use and could only be used therapeutically by veterinary prescription. (ii) The “therapeutic” antimicrobials (those other than growth-promotional antimicrobials) tylosin, sulfonamides, and nitrofurans should no longer be used without veterinary prescription. The spirit of the Swann report was to restrict the use of therapeutically effective antibiotics to only therapeutic use on a veterinary prescription basis. Withdrawal of penicillin and tetracyclines for growth-promotional and subtherapeutic purposes was, however, soon followed by their substitution by bacitracin, flavomycin, nitrovin, and virginiamycin for similar purposes.

It was perhaps unfortunate that little effort was made in Britain following the Swann report to improve the scientific base of understanding of the effect of antimicrobial drug use in animals on human health, or to document the effect of implementation of the report. Nevertheless, the sustained work of A.H. Linton (26) and that of his colleagues led to important conceptual understanding of the routes of movement of resistant bacteria between animals and humans, and the factors which enhanced the movement, although the scale of the movement still has considerable uncertainty (Fig. 1).

Figure 1: Routes of exchange of Escherichia coli between animals and humans. Note the areas where antimicrobial drug selection for resistance is most likely. The size of the circles or boxes does not indicate the extent of the scale of the movement. After Linton (26), modified by R. Irwin; reproduced with permission.

In the United States, the response to the issues raised in the Swann report was largely unenthusiastic and critical (Table 3). Resistance to Swann’s recommendations was based on the estimates of the considerable economic contribution that growth-promoting and subtherapeutic (feed) antimicrobial drugs made to agriculture in comparison to what was criticized as the inadequate evidence, the dubious and slender risk, and the “special case pleading” on which the recommendations of the Swann report were regarded as based. The strong lobby of antimicrobial drug manufacturers and the absence in the United States of a national health system (i.e., the patient pays for illness, whereas in Europe it is the nation that bears the cost) may have helped to fuel the criticism. The data were regarded as inadequate to make clear judgements, but the scale of the problem was also thought to be minor. For example, the 1989 Institute of Medicine study (Table 3) suggested that use of subtherapeutic or growth-promoting drugs might contribute to perhaps 26 human deaths a year from antimicrobial-resistant Salmonella. For perspective, these numbers would have compared to about 40,000 automobile accident and 10,000 gunshot fatalities in the United States in the same year.

Despite the inconclusive nature of many of the reports in the United States in the period between 1972 and 1995, the issue refused to die. There were periodic highly publicized reports throughout this period of serious human illness caused by Salmonella carrying resistance genes thought to be acquired from subtherapeutic, or even therapeutic, use of antimicrobials in animals. One of several examples was that of Spika and others (27) of chloramphenicol-resistant S. enterica serotype Newport traced from hamburger meat to dairy farms. Such reports led to apparently carefully orchestrated media and even major science journal frenzies about the discovery of the “smoking gun,” with consequent fervent denials by the animal antimicrobial drug industry. Given the existing well-established understanding of the epidemiology of the movement of resistant intestinal bacteria (Fig. 1), these periodic frenzies seemed at the time both astonishing and somehow hysterical. The periodic surges in public interest, however, produced no political will in the United States to re-examine the problem.

The reason for the extensive re-examination of the issue from the mid-1990s was related to several factors. The most important of these was the antimicrobial resistance crisis in medicine, in which for the first time resistant bacteria moved “out of the hospital and into the community.” The very serious nature of the crisis led to a re-examination within the human medicine community of all aspects of antimicrobial use and even to the apparent rediscovery of the importance of basic infection control procedures such as hand-washing. The antimicrobial resistance crisis in medicine again focused the medical establishment on agricultural usage of antimicrobials, in some cases almost to the extent of using it as the scapegoat for the crisis in medicine.

Improvements in understanding of the microbiology of infectious diseases acquired from animals were less important, but also critical, forces in the re-examination of antimicrobial usage in agriculture. For example, at the time of the Swann report, C. jejuni was not recognized as a human pathogen, although it subsequently became identified as the most common cause of bacterial gastroenteritis in humans. The emergence of fluoroquinolone resistance in C. jejuni of poultry origin in the United States because of the use of fluoroquinolones to treat Escherichia coli infections in chickens (28) subsequently led to the ban of all use of this class of drug in chickens in the United States (Table 3). A subsequent risk analysis in the United States suggested that 8,678 citizens treated for this illness with fluoroquinolones had fluoroquinolone-resistant C. jejuni illness acquired from chickens (Table 3), a huge number compared to the “26 possible deaths because of resistant Salmonella” identified in the 1989 Institute of Medicine report.

Similarly, at the time of the Swann report, vancomycin-resistant enterococci (VREs) were also unknown, although subsequently, enterococcal infections emerged as major nosocomial, largely hospital-acquired, infections in humans, with vancomycin as the “drug of last resort” in such infections. Acquisition of transmissible vancomycin-resistance genes by these hospital-associated bacteria made them essentially untreatable, again raising the specter of the postantibiotic era (Table 1). Work by Aarestrup and his colleagues in Denmark was important in identifying avoparcin, a glycopeptide antimicrobial related to vancomycin, as selecting for the massive presence of VREs in the intestine of poultry and swine fed this drug as a growth promoter (29). For the first time, there was convincing large-scale evidence that eliminating the use of antimicrobial drugs in food animals could dramatically reverse the rise of resistant bacteria in these animals (30). Convincing molecular genetic typing evidence showed that VREs from animals colonized humans (31) and, most dramatically, the marked decline in human intestinal colonization by VREs in Europe following the withdrawal of avoparcin as a growth promoter (32) suggested that the scale of the movement of resistant intestinal bacteria from animals to humans, which had always been a matter of great uncertainty, was far larger than generally suspected previously. Molecular genetic typing and whole-genome sequencing of resistance genes and gene regions were unavailable at the time of the Swann Report but were subsequently used extensively to characterize the relatedness (and therefore sometimes the source) of both bacteria and their resistance genes obtained from animals and humans.

More recent application of whole-genome sequencing to antimicrobial-resistant extrapathogenic E. coli from human urinary tract infections and comparison to isolates from healthy chickens clearly indicates that some resistant human isolates derive from chickens (33). Recognition of the likelihood of such a previously unsuspected “insidious epidemic” suggests that the scale and importance of the movement of resistant bacteria and their genes into the human population may be far greater than suspected. One group suggested that cephalosporin use in poultry was responsible for about 1,500 human deaths annually in Europe (34). A voluntary ban on ceftiofur use in Danish swine production was shown to effectively reduce extended-spectrum cephalosporinase-producing E. coli in slaughter pigs (35). It seems likely that similar bans, voluntary or involuntary, may be adopted in different jurisdictions in agriculture. The World Health Organization ranking of antimicrobials according to their importance in human medicine (36) remains an important issue for animal use, since the World Health Organization classification of “critically important” includes drugs such as penicillin, and essentially all antibiotics are classified as important, highly important, or critically important. Further discussion, which will likely focus on “highest-priority critically-important antimicrobials” (36), is clearly required as one of the steps in addressing the global resistance crisis.

It is highly ironic that the recent emergence and global dissemination of MRSA in livestock, particularly of clonal complex 398 associated with swine, and the emergence and dissemination of livestock-associated MRSA infections in humans (37, 38) has been linked to the use of zinc oxide in the feed of intensively reared livestock (39). Following the European ban on antimicrobial growth promoters, zinc oxide was introduced as an alternative to to prevent enteric infections in young animals.

THE EMERGING CONCEPT AND PRACTICE OF ANTIMICROBIAL STEWARDSHIP IN VETERINARY MEDICINE

As the science of antimicrobial resistance moves onto the political stage, the past 2 years have seen dramatic changes in the global response to the antimicrobial resistance crisis, culminating most recently in the September 2016 United Nations High-Level Meeting and the commitment of members to address the issue in a multifaceted way. This follows earlier similar commitment by members of the G7 and numerous important analyses of how to address the crisis (45, 46) (Table 1). In the United States, there has been game-changing commitment (Guidance 213, Guidance for Industry 233) to remove antibiotics from use as growth promoters and to bring all antibiotics used in feed or water of food animals under veterinary oversight. Canada has followed suit. Another major change has been adoption of a “One Health” approach to resistance (reviewed in chapter 26), an approach that involves multidisciplinary and multi jurisdictional approaches to very complex problems involving people, animals, and the environment (47). An evolving concept and practice, a One Health approach may reduce some of the conflict between the use of antimicrobials in human and veterinary medicine by focusing efforts and energy on resolving resistance issues in a collaborative rather than blaming manner.

Antimicrobial resistance is a multifaceted problem with all the complexities of climate change, to which it is highly analogous. It has multiple causes, with no single actor or factor that can be blamed, has the well-established ability to be self-sustaining, and has the potential to be catastrophic. No single intervention will address the problem, but the combination of multiple interventions and approaches has the potential to have a cumulative impact that will help in its control. A stewardship approach which integrates so much of what we now know about effective antimicrobial use (6), and about infection generally, is the best approach for first-line veterinary practitioners to address the resistance crisis. “Antimicrobial stewardship,” reviewed in chapter 32, is the term increasingly used in medicine to describe the multifaceted approaches required to sustain the efficacy of antibiotics and minimize the emergence of resistance. The concept and practice of antimicrobial stewardship continues to evolve in human and veterinary medicine, but it is an approach that takes an active, dynamic process of continuous improvement encapsulated in the idea of good stewardship practice (GSP) (6, 48). Only a GSP mind-set will ensure the long-term sustainability of antimicrobial drugs. Antimicrobial stewardship and GSP involve coordinated approaches and interventions designed to promote, improve, monitor, and evaluate the judicious use of antimicrobials to preserve their future effectiveness and promote and protect human and animal health. This involves a “5R” approach of responsibility, reduction, refinement, replacement, and review (49). Critically, a GSP approach to stewardship also could be evaluated quantitatively as a standard of veterinary practice.

The question for the future is how to preserve existing and develop new drugs in the face of bacterial pathogens, some of which appear to have become particularly adept at developing or acquiring resistance over the past 60 years. Using some of the tools available as we enter the “golden age of microbiology” to improve the way we diagnose infections and develop new, likely targeted, antimicrobials is promising (50). However, as noted by one writer in respect to resistant bacterial infections in companion animal practice (40), resolving the issue of multidrug-resistant endemic bacterial infections will not be through development of new antibiotics if current hygiene practices remain and if we don’t undertake good stewardship practices to preserve our existing drugs.

With respect to the changing relationship to antimicrobial drugs in human and veterinary medicine that the resistance crisis has produced, there’s a sense that humanity is perhaps in the intermission after the first act of a three-act play, and we’re still trying to determine if the play is a comedy or a tragedy. It currently feels like both. There’s a huge amount to be done.