Advisory Board Assessment of Antibiotic Use at Sanderson Farms, Inc. November 28, 2018

Advisory Board Assessment of Antibiotic Use at Sanderson Farms, Inc. November 28, 2018 Page 1. Executive summary 2 2. Introduction and definition of antibiotic stewardship 4 3. Advisory board assessment of Sanderson Farms, Inc. antibiotic use 5 Is there a legitimate need for antibiotics in the SFI system as it 5 currently stands? Is there evidence that antibiotic use at SFI is effective? 6 Is there a viable alternative to antibiotic use to prevent disease in 6 poultry at SFI? Is there a viable non-antibiotic alternative to treat disease in poultry 7 at SFI? What would happen if antibiotics were removed from the SFI system in terms of animal welfare, economics, preservation of antibiotics for use in animals and humans, and food safety? 4. Background information 15 Antibiotic stewardship 15 Medically important vs. non-medically important antibiotics 16 The special case of the ionophores 23 Reasons some antibiotics are not used in all or some food animals 24 Antibiotic resistance vs. residues 27 What are the highest risk uses of antibiotics in terms of perpetuating 30 resistance? What are the most serious resistant bacteria and where do they 32 come from? Are antibiotics used in human medicine to prevent and/or control 35 bacterial infections? The precautionary principle 36 References 38 Appendix - Members of the advisory board 41 8 1

1. Executive Summary of Advisory Board Assessment Potential human health risks related to the use of antibiotics in chicken management include the possibility of antimicrobial drugs persisting in chicken meat. Regular assessment of chicken meat in the U.S. indicates that violative concentrations of antibiotics are rarely found; only one positive sample was detected in the past three years. Another risk is the possibility that bacteria resistant to antibiotics will infect people that come into contact with treated chickens or their products. This is not just a risk related to chickens raised with antibiotics, antibiotic resistant bacteria can be found even on the surface of uncooked chicken raised without antibiotics or raised with no antibiotics ever (RWA/RAE). Given currently available research data, it is not possible to estimate with a high level of confidence the true risk to human health posed by antibiotic use practices in poultry production. Given that some producers market chicken as RWA/NAE, or raised without treatment with antibiotics classified as medically important, Sanderson Farms, Inc. (SFI) is vulnerable to criticism for their antibiotic use. The use at SFI of the antibiotics gentamicin and virginiamycin is particularly likely to draw critical attention, because these antibiotics have been designated as medically important. In spite of the RWA/NAE definition, poultry producers marketing such chicken must at times use antibiotics when flocks of chickens develop disease that can only be halted by antibiotics. This is necessary to prevent the death of large numbers of birds, in accord with attention to bird welfare and fiscal responsibility. Chickens in RWA/NAE systems that must receive antibiotics are then marketed as conventional chicken. Research indicates that RWA/NAE management represents a trade-off. While generation of antimicrobial resistant bacteria can be decreased in such systems, chickens in these systems are more likely to shed Salmonella or Campylobacter, bacteria that can cause illness in people who consume undercooked chicken. Moreover, rates of chicken death (mortality) in RWA/NAE operations are slightly higher than in traditional systems, according to Agristats data. Chicken mortality at SFI is below industry averages for chickens raised in traditional and RWA/NAE systems, and systems where medically important antibiotics are not used, indicating that chickens at SFI have better than average health. In 2017 1.04% of SFI chicken were treated with antibiotics for disease. Although data for comparison are not readily available, the impression of Advisory Board members with relevant experience is that this rate of treatment is low relative to conventional operations of comparable size, also indicating that chickens at SFI have better than average health. 2

Some of the antibiotics used at SFI are classified as not medically important. Ionophores are one example; these antibiotics have no applications in human medicine. Ionophores are used to decrease chicken sickness and death due to the disease coccidiosis. Currently available alternatives to ionophores for coccidiosis control are less effective. While it has been noted that 80% of all antibiotic sales in the United States are for use in animals, 40% of these sales are for antibiotics not classified as medically important, and the majority of this 40% are ionophores. Because ionophores are classified as antibiotics in the U.S., operations marketing RWA/NAE chicken cannot use ionophores. This can lead to relatively high rates of sickness and death of chickens due to coccidiosis and related diseases on RWA/NAE operations. Notably, ionophores are not classified as antibiotics in Europe; therefore, RWA/NAE poultry management in Europe can include the use of ionophores. This discrepancy puts U.S. poultry farmers marketing RWA/NAE chicken at a disadvantage relative to European producers. Although their use of the medically important antibiotics gentamicin and virginiamycin makes SFI vulnerable to criticism by individuals who advocate RWA/NAE management, currently available data indicate that movement to RWA/NAE management is expected to increase chicken death rates. The degree of increased mortality could be substantial in the near term, particularly if RWA/NAE management required exclusion of the use of ionophores for coccidiosis control. A move by SFI to a system where non-medically important antibiotics (i.e. ionophores) can be used for prevention, and medically important antibiotics can be used for treatment and control of disease, could represent a responsible compromise to better preserve efficacy of antibiotics important for human health, while also avoiding the adverse impacts of a RWA/NAE system on chicken health and welfare. Such a system would reflect principles of antibiotic stewardship described in this document. 3

2. Introduction Much has been written and discussed in recent years regarding the production of meat from poultry and livestock that have not been treated with antibiotics. Some restaurant and retail grocery store chains have announced plans to move to some variation of a no antibiotic ever program with respect to chicken products. In light of investor questions regarding the SFI position on antibiotic use, and in order to provide additional information to outside board members to aid in the board s oversight responsibilities, SFI management convened a group of individuals with expertise in poultry production, livestock health and management, and the use of antibiotics in veterinary and human medicine to review and comment on the use of antibiotics in SFI poultry production. The members of this Advisory Board and their affiliations are listed in the Appendix. The Advisory Board members were provided with data describing current antibiotic use in SFI chicken production, including the antibiotics given, the reason the antibiotics are used, the amount or dose of antibiotics given, the stages of the production cycle when antibiotics are used, and the number of chickens receiving antibiotics. The Advisory Board members discussed the information and generated this document summarizing their assessment in both physical meetings and teleconferences. This document summarizes the Advisory Board review. The document is organized into four sections: 1) executive summary; 2) introduction and definition of antibiotic stewardship; 3) Advisory Board assessment of SFI antibiotic use, in the form of responses to questions that may be posed regarding this use; and 4) background information provided for reference, to clarify or expand on information in sections 1-3. Defining Antibiotic (Antimicrobial) Stewardship The concept of antibiotic stewardship underlies any assessment of antibiotic use, including the assessments in this document. Defining and practicing antibiotic stewardship will be critical to the ability of producers and veterinarians to ensure the health and welfare of animals in their care in the coming decades. This is because of expected constraints related to the impact of antibiotic resistance in bacteria that cause disease, and to regulatory oversight of various agencies. Perhaps most relevant to this document is the American Association of Avian Pathologists (AAAP) statement on Antimicrobial Stewardship for Poultry.(1) The excerpts below capture the core meaning of antimicrobial stewardship as defined by the AAAP. More detail regarding the defining principles is contained in the document. 4

Antimicrobial Stewardship for Poultry Veterinarians Defined Antimicrobial stewardship refers to the actions poultry veterinarians take individually and as a profession to preserve the effectiveness and availability of antimicrobial drugs through conscientious oversight and responsible medical decision-making while safeguarding poultry, public, and environmental health. Core Principles of Antimicrobial Stewardship in Poultry Veterinary Medicine Antimicrobial stewardship involves maintaining poultry health and welfare by implementing a variety of management strategies to prevent, control and treat common diseases; using an evidence-based approach in antimicrobial decisions; and then using antimicrobials judiciously, sparingly, and with continual evaluation of the outcomes of therapy; while protecting poultry health and ensuring safe, affordable food to the consumer. 3. Assessment of SFI antibiotic use Is there a legitimate need for antibiotics in the SFI system as it currently stands? Historically, many diseases of poultry were prevented through use of an antibiotic in the feed or water during the period when disease was most likely. The poultry industry has made significant strides toward improving the health and wellbeing of the birds being raised for human consumption. One of the major practices adopted over the past fifty years is to have veterinarians specializing in poultry medicine serving as integral company executives, working with the other company managers on day-to-day decisions regarding not just bird health but also many husbandry decisions. SFI employees include six veterinarians who advise the company on all decisions regarding treatment and prevention of disease in birds. Antibiotics are used to prevent, treat, or control bacterial infections in birds at SFI. The major diseases that require antibiotics for prevention, treatment, or control include necrotic enteritis, gangrenous dermatitis, and infections due to the bacteria E. coli. The prominent role of staff veterinarians in poultry health management at SFI improves the likelihood that antibiotics are used only for legitimate applications to prevent, treat, or control disease. One important disease due to E. coli is respiratory infection, including airsacculitis, which affects SFI broiler chicken houses following outbreaks of the viral disease infectious bronchitis. The infectious bronchitis virus is easily transmitted between chicken farms, and once it enters a farm, it spreads rapidly between birds in the same chicken house. Antibiotics are used to control secondary bacterial infections that can follow infectious bronchitis virus infection. In particular, this virus makes infected birds very susceptible to E. coli, which is found in all chicken houses.(6) Escherichia coli infection causes the birds to have more severe respiratory disease that often progresses to pneumonia and death. The treatment of E. coli infection of birds experiencing infectious bronchitis infection represents an example of an appropriate use of antibiotics in SFI management. Current position statements of the American 5

Association of Avian Pathologists (AAAP) and the American Veterinary Medical Association (AVMA) support this. Other diseases commonly treated with antibiotics at SFI are necrotic enteritis, which is due to the bacteria Clostridium perfringens, and gangrenous dermatitis, due to the bacteria Clostridium perfringens, Clostridium septicum, and Staphylococcus aureus. In 2017 lincomycin was the antibiotic most commonly used for treating sick birds at SFI, and penicillin was the second most commonly used. Tetracyclines and sulfonamides were also used. In 2017, 1.04% of birds received antibiotic treatment for disease. Although data for comparison are not readily available, this rate of treatment is low relative to conventional operations of comparable size. The antibiotics bacitracin methylene disalicylate (BMD) and virginiamycin are used at SFI for prevention of disease. This is an approved use of these antibiotics as determined by the FDA- CVM. Because virginiamycin is considered medically important (defined and discussed in Section 4), continued effort to use and confirm the benefit of disease-preventing strategies to minimize the need for antimicrobial prevention with virginiamycin is advised. Prevention can include means of excluding disease-causing agents, often referred to as biosecurity. In addition, vaccination is one of the major means of prevention, especially against primary viral disease agents to prevent secondary bacterial infections. SFI is currently using some nonantibiotic means to control disease. Birds are vaccinated against coccidiosis, and a probioticprebiotic (Poultry Star), which may help prevent birds from being infected (colonized) by disease causing bacteria, is used in breeder diets. Data from Agristats indicate that SFI mortality (death) of chicks up to 7 days of life is better than the industry average, indicating that management is effective to preserve health of young birds. Given this, it may be reasonable for SFI to phase out the use of the antibiotic gentamicin in-ovo. Most broiler companies have ceased use of in-ovo antibiotics. However, removal of in-ovo gentamicin will take time, as an increase in 7-day bird mortality which can be substantial is likely to follow. Management practices to limit infection in the brooding and hatchery phases will need to be tested to keep mortality rates below industry averages if inovo gentamicin is removed. Is there evidence that antibiotic use at SFI is effective? Data from Agristats indicate that SFI mortality (death) of chicks up to 7 days of life, and mortality over the entire life of birds, is better than the industry average, indicating that current SFI management practices, including their use of antibiotics, is effective to preserve health and life of chicks and birds. Is there a viable alternative to antibiotic use to prevent disease in poultry at SFI? In response to consumer concerns, poultry companies use nonantibiotic alternative products (probiotics, intestinal acidifiers, natural antibacterials, enzymes, etc.) in an effort to reduce the growth of the unfavorable intestinal bacteria such as Clostridium perfringens. (14, 25) All of 6

these methods can help reduce the incidence or the severity of the bacterial infection, although they are more likely to be reliable for prevention than for treatment. Principles of antibiotic stewardship require that non-antibiotic preventive strategies be utilized when effective. SFI is currently using some non-antibiotic means to control disease. Birds are vaccinated against coccidiosis, and a probiotic-prebiotic (Poultry Star), which may help prevent birds from being infected (colonized) by disease causing bacteria, is used in breeder diets. A non-antibiotic medication (chemical), nicarbazin, is used along with the ionophores narasin and salinomycin, to control coccidiosis. Is there a viable non-antibiotic alternative to treat disease in poultry at SFI? There are currently no viable alternatives for antibiotic use when flocks of chickens become sick from a bacterial infection. Even broiler producers who are producers of RWA/NAE chickens have to use antibiotics to treat flocks that become sick due to bacterial infection, or the chickens die in numbers that can be substantial. These treated flocks are then marketed in the traditional channels that are not for RWA/NAE labels. It has been estimated that in the winter approximately 5% - 10% of the broilers placed by an RWA/NAE producer will have to be treated with antibiotics. Relevant to this, a recent report by the Council for Agricultural Science and Technology (CAST) on the Scientific, Ethical and Economic Aspects of Farm Animal Welfare states: Because of concerns related to antimicrobial resistance, the use of antimicrobials in food-producing animals is a topic of much discussion. Strategies to address antimicrobial resistance include discontinuing production uses (e.g., for growth promotant and feed efficiency uses); enhanced use of other means of infectious disease prevention (e.g., improved biosecurity measures, increased use of vaccination to prevent viral diseases that may often be followed by secondary bacterial infections); greater attention to how antimicrobials are selected and used in prevention and treatment protocols (i.e., targeted application, increased veterinary oversight); and the identification, development, and use of nonantibiotic alternatives for prevention, control, and treatment (e.g., organic acids in feed and water, gene-encoded natural antibiotics, prebiotics and probiotics, bacteriophages). A related emergent animal welfare problem is that increased consumer demand for meat from animals that have not been treated with antimicrobials for any purpose production or therapeutic has caused and may continue to cause producers and veterinarians to withhold treatment for animals intended for the consumer market. The negative impacts on animals welfare resulting from disease that could be prevented and/or that cannot be controlled and treated are significant and unacceptable (CAST, Task Force Report No. 143, April 2018, pages 24-25).(13) The need to retain the possibility of using antibiotics to treat disease while limiting their use in preventive management is highlighted by a 2018 study by Karavolias et al.,(30) which showed that broilers raised without antibiotics had a 3.6 times greater risk of eye injury (corneal burns), 1.3 times greater risk of foot infections leading to sore feet and reluctance to walk, and 1.6 times greater risk of respiratory infection (airsacculitis) leading to death, as compared to 7

birds raised conventionally. What would happen if antibiotics were removed from the SFI system in terms of animal welfare, economics, preservation of antibiotics for use in animals and humans, and food safety? Effects on animal welfare and economics In the U.S. broiler industry, many companies have eliminated the use of antibiotics in normal production, reserving antibiotics to treat and control disease in sick flocks. However, in these companies, a flock sometimes develops disease that can only be effectively treated with antibiotics. In such cases, to prevent excessive sickness and death, a flock may be treated with antibiotics. If this occurs, the birds in that flock cannot be sold with a no antibiotics ever (NAE) label and must be marketed through conventional distribution channels. This type of production is more generally referred to as raised without antibiotics or RWA/NAE. The birds marketed under the NAE or RWA/NAE claims do not receive antibiotics, but chickens that left the production system due to need for antibiotic treatment are typically not addressed in marketing information released by these producers. In the U.S. removal of antibiotics includes removal of in-ovo antibiotics and the polyether ionophore antibiotics (ionophore anticoccidials). The removal of medically important antibiotics for growth promotion was effective as of January 1, 2017. The removal of antibiotics for prevention, control, and treatment of disease in poultry production without viable alternatives may result in welfare concerns due to elevated mortality and loss of production efficiency due to illness and death loss. Additionally, more resources are used to raise more birds to replace those that die, resulting in greater environmental impacts from increased manure production and more use of grain per unit of meat produced. Higher rates of intestinal disease may also result in increased numbers of foodborne illness-causing bacteria such as Salmonella sp. or Campylobacter sp. on the carcass. The disease necrotic enteritis (NE) can in particular have a negative impact.(26) In a recent survey of consumers, 55% responded that they were extremely or very concerned about antibiotic use in chickens when they purchase chicken.(8) Unfortunately, this same survey demonstrated that the respondents generally have major misunderstandings about poultry production. For example, 60% of respondents considered themselves to be very knowledgeable or somewhat knowledgeable about the care of chickens, but 75% of respondents believed that there are added hormones or steroids in chicken meat and 71% of respondents believed that chickens raised to be eaten are raised in cages neither of which is true. The reality of raising broiler chickens without antibiotics is different than understood by most consumers. In general, the impact of raising broilers without antibiotics has been shown to have an overall negative effect on gut health and bird performance.(24, 35) Although these negative impacts can be minimized over time as producers adjust to this different style of 8

production, data analyses still show overall negative effects on animal health, animal welfare, environmental sustainability and economic viability. A recent study compared three different types of production systems: conventional, NAE, and non-medically important only, wherein only antibiotics that are not considered important to human health are used to maintain bird health. The study analyzed data regarding three important health conditions of the birds in these three production systems: eye ammonia burns, footpad lesions, and airsacculitis. The presence of these conditions can indicate poor animal welfare, and birds with these conditions can have reduced weight gain caused by decreased feed intake because of the associated pain.(30) This study found that NAE production increased the risk and severity of all three of these health conditions. Using the nonmedically important antibiotics diminished this risk and severity somewhat, but the risk was still greater than for conventional production. Specifically, the odds of eye burns occurring in a bird given no antibiotics was found to be approximately 3.6 times higher than a bird given medically important antibiotics (conventional). Birds raised with non-medically important antibiotics had an approximate 1.3 times higher odds of eye burns than birds raised conventionally. The odds of footpad lesions were approximately 1.3 times higher for birds raised NAE compared to a bird raised conventionally, and this was the same difference between birds raised with nonmedically important antibiotics and birds raised conventionally. Finally, the odds of airsacculitis were approximately 1.6 times higher for birds raised NAE compared to a bird raised conventionally. Interestingly, the odds of airsacculitis were lower in birds raised with nonmedically important antibiotics compared to birds raised conventionally. All of the odds mentioned above were statistically significant. The authors highlight important limitations of this study. First, they emphasize that the analyses do not prove a cause and effect relationship; in other words, they are not stating that raising birds NAE causes these conditions to become worse. Second, they emphasize that they did not analyze management practices and other related on-farm variables. They state Transitioning from medically important antibiotics to no antibiotics ever generally requires changes be made to production including reduced stocking density, longer downtime between flock production cycles in a barn, providing an all-vegetarian feed, etc. (30) This point, as mentioned above, is key. Many of the negative impacts of NAE production can be diminished over time, but data suggest these negative impacts might never be completely eliminated. Other studies have also reported negative impacts of NAE production on animal health. In an early study by Smith (2011) (35), the author reports that in addition to being more expensive to produce, due to stricter and more expensive diet requirements, drug-free birds had a higher incidence of necrotic enteritis. In a more recent study by Gaucher et al. (2015) (24), the authors reported that the drug-free program was associated with an overall negative effect on key performance indicators and gut health, which is indicative of the potentially negative effects on the overall animal welfare. In particular, the drug-free program was associated with both an increased incidence of necrotic enteritis, as well as a significant increase in feed conversion, and a decrease in both daily weight gain and mean live weight at slaughter. 9

The paper by Karavolias et al. (30) states that Prior literature investigating the impact of removing access to antibiotics on poultry production focuses on subtherapeutic (e.g., growth promotion) uses only and/or focuses on productivity impact related to bird performance and grower financial outcomes. These authors conclude that Policies aimed at eliminating or restricting the use of antibiotics in broiler production may come with potentially negative consequences with respect to good animal welfare. A more effective policy approach should consider comprehensive animal care plans that incorporate good housing, management, and responsible antibiotic use, including the use of ionophores. Policies aimed at informing the consumer on the positive role of access to antibiotics in supporting good animal welfare while limiting risk of antibiotic resistance in humans are needed to address the current information gap. The impacts of raising animals without antibiotics are not restricted to animal health and welfare. There are also potential effects on environmental sustainability and economic viability. In a study conducted by Salois et al.,(33) a simulation model was used to evaluate the impacts of NAE production. The authors conclude: Compared to broilers produced in a conventional system, birds raised in a single broiler house under ABF conditions will have an annual reduction of between 50,000 100,000 lbs of edible meat (breast, legs, thighs, wings) equivalent to between 265,000 530,000 individual 3 oz. single servings. This loss represents enough to feed 600 1,000 people annually, based on average annual consumption of chicken in the United States in 2012. In order to maintain the same supply of meat under ABF conditions, a typical broiler house will require between 15,000 33,000 more marketed broilers per year. Due to the additional broilers needed, eliminating antibiotic use has an environmental impact. Compared to a conventional house, chickens raised in a single broiler house under ABF conditions will require between 185,000 390,000 additional lbs. of feed per a year; between forty-two and ninety additional acres a year to produce that feed; between 33,000 and 78,000 additional gallons of water consumed; and between 157,000 and 333,000 additional tons of manure produced. In addition, the cost to produce edible prime meat in a broiler house under ABF conditions is between $52,000 and $110,000 per year. The authors extended these house-level estimates to the entire U.S. broiler industry. If the entire broiler industry were to go NAE, the authors estimate that we would need 680-880 million more birds to maintain supply, and 5.4-7.6 million tons of additional feed would be needed which would require between 2.5-3.3 million additional acres of land to grow. The additional birds also would require between 1.9-3 billion gallons of additional water and would produce between 4.6-6.1 million tons of additional manure. The authors conclude that eliminating the use of antibiotics in the raising of broilers may have a negative effect on the conservation of natural resources as well as a negative economic effect via increased prices to the consumer. Results suggest the need to communicate to consumers 10

the supportive role that prudent, responsible use of antibiotics for animal disease treatment, control, and prevention plays in the sustainable production of broilers. The special case of ionophores: Ionophores are medications used to prevent coccidiosis, a protozoal (parasite) infection. In the U.S., the FDA defines ionophores as antibiotics, although they are not medically important (see Section 4). Coccidiosis is a common cause of intestinal disease in poultry. Moreover, coccidiosis increases the likelihood that birds will be infected by Clostridium perfringens, the bacteria that causes necrotic enteritis, a disease which can kill birds if they are not treated with antibiotics. If SFI stops using all antibiotics they will have to stop using the ionophore anticoccidial drugs. The ionophores have no use in humans and have no documented impact on human bacterial infections or antibiotic resistance. Exclusion of ionophore antibiotics in a RWA/NAE management system will likely result in more chickens having intestinal disease, and more chickens dying because of necrotic enteritis. There are non-ionophore, non-antibiotic coccidiostats available as discussed in section 4. These are collectively referred to as chemical coccidiostats, and are often used in RWA/NAE programs. The protozoal parasites that cause coccidiosis have developed significant resistance to chemical coccidiostats, leading to efficacy that is inferior to ionophores. When an RWA/NAE producer makes the decision to exclude all antibiotics, not just medically-important antibiotics, they are making the decision to remove the ionophores from the production system and to rely solely on the chemical anticoccidials. In summary, if SFI were to move to RWA/NAE production, this could improve opportunities to market certain products through certain channels. However, current assessments indicate that the percent of their chickens that get sick and that die will increase. This increase in the number of chickens expected to die means that more chickens will have to be hatched and raised in order to produce the same number of pounds of chicken as a traditional system. Hatching and raising more birds will require increased resources such as feed and energy, and will result in some increase in waste production, over that of the current traditional system. Effects on preservation of antibiotics for animals and people Effective, relatively non-toxic antibiotics have been available since 1935, with the introduction of sulfonamides. Following the first large scale availability of penicillin in 1942, there was an approximately 40-year flurry of new antibiotic class discovery, ending in the mid 1980 s. The last new antibiotic group for which a member of that group was eventually available for use in food animals was the 1978 release of ciprofloxacin (a fluoroquinolone) for human use. The history of this group in chickens and turkeys is discussed in Section 4; fluoroquinolone antibiotics are no longer legal for use in chicken production. While some new molecules in existing antibiotic groups have been made available for use in some food animal species since 1978, no new antibiotic groups have become available for use in food animals, and the probability that more will become available is extremely low. 11

While preserving antibiotics for use in humans should be at the forefront of antimicrobial stewardship efforts, preserving the ability to treat the animals under our care is also necessary. If resistance builds to the remaining effective antibiotics we are left with little for intervention in disease outbreaks when preventive practices are overwhelmed. The cost of using antibiotics as routine preventive practices is that they may not be effective in the future when they are needed for therapeutic interventions. Specifically, in relation to the antibiotic use practices of SFI, the question is whether the use of the following antibiotics in preventive programs have the potential to introduce resistant pathogens into the food chain or environment where there is a detrimental effect on human health, or to contribute to the erosion of antibiotics as therapeutic interventions in chicken disease outbreaks in the future. Medically important o Gentamicin in-ovo o Virginiamycin in feed Non-medically important o Ionophores o Bacitracin Methylene Disalicylate (BMD) in feed (while classified as nonmedically important in the U.S., the cyclic polypeptides [the group which includes bacitracin] are classified as medically important by the World Health Organization) Although used on a much smaller scale, the same questions could be asked for the antibiotics used only for treatment of disease at SFI; the FDA considers all of these medically important in human medicine. They are used only in response to disease outbreaks and are administered through the water. Lincomycin Penicillin Sulfadimethoxine Oxytetracycline The definite establishment of links between poultry antibiotic use and human health, and the magnitude of such links, are beyond the scope of this document. However, it is clear that bacterial populations in chickens, humans, and the environment can overlap. Moreover, the potential for erosion of therapeutic efficacy of our few remaining effective antibiotics because of the development of antibiotic resistance is an issue that is quite real. Therefore producers and veterinarians should focus on the preventive use of antibiotics as an interim practice while management practices that allow continued reduction of such antibiotic use are pursued. Separating the precautionary principle (as discussed in Section 4) from legitimate, sciencedriven concerns is a challenge in both this report, and in addressing consumer concerns about poultry production practices. 12

Effects on food safety Human health risks related to the use of antibiotics in chicken management include the possibility of antimicrobial drugs persisting in chicken meat, and the possibility that bacteria resistant to antibiotics will infect people that come into contact with treated chickens or their products. Regular assessment of chicken meat in the United States indicates that violative concentrations of antibiotics are rarely found (discussed below). In the last three years of testing, only one violative level of antibiotic residue was found in chicken meat, and it was an antibiotic, nitrofurazone, which is not approved for use in animals raised for food. The two most common bacterial foodborne illnesses in the U.S. are non-typhoidal Salmonella and Campylobacter, accounting for a Centers for Disease Control and Prevention (CDC) estimated 1.0 million and 0.8 million cases per year respectively.(9) The most common source of these infections is contamination of meat by material from the intestines of animals raised for food; these bacteria are found in other food-producing animals and are not restricted to poultry. It is possible that exposure of animals to antibiotics in food or water may contribute to antibiotic resistance in these two bacterial pathogens. The latest National Antibiotic Resistance Monitoring System (NARMS) report states that no antibiotic resistance was found in 76% of non-typhoidal Salmonella isolates. Multidrug antibiotic resistant bacteria were found but the majority of the antibiotics to which Campylobacter and Salmonella were resistant were ampicillin, chloramphenicol, streptomycin, sulfonamides and tetracyclines, all antibiotics that would not be used to treat human illness due to Salmonella or Campylobacter. Overall there was a decline in multi-antibiotic resistant Salmonella from 17% to 12% over ten years. Ciprofloxacin is a critical antibiotic to treat Salmonella in adults and was banned from use in poultry by the FDA in 2005. Only 0.7% of Salmonella isolates from chicken meat samples showed resistance to ciprofloxacin, whereas 6% of human isolates were resistant. Another consideration regarding a move to an NAE system is that, as discussed above, more birds are expected to become sick. When chickens are sick, they eat more bedding material (litter), which can result in higher numbers of Salmonella sp. and Campylobacter sp. in their intestinal tract.(12) Also, it has been shown that birds from flocks having higher rates of carcass condemnation due to airsacculitis had higher levels of E. coli and Campylobacter on their carcasses.(32) There are food safety implications specifically related to any decision to remove ionophore antibiotics, which would be required in a RWA/NAE system as currently practiced in the United States. Baba, et al.,(5) found that coccidiosis resulted in more Salmonella typhimurium in the livers and spleens of chickens, and Volkova, et al.,(36) noted that the method of coccidiosis control may influence the prevalence of Salmonella at the processing plant. The disease necrotic enteritis, which is exacerbated by coccidiosis, may not directly result in an increased colonization by Salmonella, but it can result in greater variation in bird size due to the 13

subclinical form of the disease.(27) As Russell found, poorer body weight uniformity can result in greater intestinal tract tears and greater risk of Campylobacter contamination on chicken carcasses.(32) Thus, reducing animal illness likely plays a critical role in reducing the chances of contamination during processing. When assessing the potential risks of antibiotic use in poultry, we must begin to take a more holistic view of health into consideration. Specifically, we should be assessing the potential risks and the potential benefits associated with antibiotic use. Phrased another way, are there potential unintended consequences of removing antibiotics from use in food animals? Recent models have predicted that there might be significant negative human health consequences associated with the removal of certain antibiotics from animal production. This is an instance in which the precautionary principle would lead to an action of banning antibiotics in animal agriculture, but that action could have even worse unintended consequences. It might not be intuitive, however, how an antibiotic that is used in animal agriculture can actually benefit human health. Mathematical models have been developed to relate animal illness to human illness.(34) These models demonstrate that there can be large increases in human illness associated with small increases in animal illness, suggesting that agricultural management strategies may have significant impacts on human health. Antibiotics that are administered to poultry via the feed for disease prevention raise concern about their potential to increase rates of antibiotic resistance, posing a risk to human health. However, these applications also improve animal health and can promote size uniformity among animals in the flock. Antibiotic uses in animals can therefore have potential human health risks, but also benefits. Models such as the one cited previously, are able to evaluate simultaneously the human health risks and benefits associated with antibiotic use in animal agriculture. The cited model addressed the relationship between the negative human health impact of increased antibiotic resistance and the positive human health impact of fewer foodborne infections, both of which are due to the use of the antibiotic in animal agriculture. The model showed that the potential benefits to human health associated with the use of antibiotics in animal agriculture can far outweigh the potential risks. This finding has been validated by additional studies.(7, 28) In summary, if SFI moves to a RWA/NAE system of chicken production, they will have more opportunity to market certain chicken through certain channels. Moreover, rates of identification of antibiotic resistant bacteria are expected to be lower in a RWA/NAE system. However, it is important to remember that even in a RWA/NAE system, flocks of chickens will occasionally need antibiotic treatment for bacterial disease, or chickens will die in numbers that would be considered unacceptable in terms of animal welfare and fiscal responsibility. Also, removal of ionophores antibiotics, which would be required in a RWA/NAE system as currently practiced in the United States, would increase sickness and death due to coccidiosis and to necrotic enteritis, which is exacerbated by coccidiosis. Increased numbers of Salmonella and Campylobacter would be expected to be identified in and on chickens in a RWA/NAE system; this represents a food safety concern because these bacteria can cause disease in people who consume undercooked chicken or who come into contact with chickens carrying these bacteria. 14

A move by SFI to a system where non-medically important antibiotics can be used for prevention, and medically important antibiotics can be used for treatment and control of disease, could represent a responsible compromise to better preserve efficacy of antibiotics important for human health, while also avoiding the adverse impacts of a RWA/NAE system on chicken health and welfare. Such a system would reflect principles of antibiotic stewardship described in this document. Section 4: Background Information Antibiotic stewardship Stewardship has been defined differently by physicians and veterinarians. The Infectious Disease Society of America has a stewardship definition applied to human medicine which does not include infection prevention.(29) In human medicine, infection prevention and antibiotic stewardship are considered two separate but overlapping processes. However, in veterinary medicine, these two aspects of stewardship are addressed together. Therefore, the concept of antibiotic stewardship in veterinary medicine is all encompassing, including disease prevention and the judicious use of antibiotics when they are needed (Figure 1). Figure 1: Components of an Antibiotic Stewardship Program Within this stewardship cycle, there are aspects of antibiotic use which may be benchmarked to make the veterinarian and their clients aware of the primary disease challenges being experienced by others, and how they are being addressed. The nature of different production 15

systems creates a situation in which the optimal antibiotic stewardship benchmarking metrics are often specific to the species, and even the production system type within species. The American Veterinary Medical Association (AVMA) recently defined antimicrobial stewardship and core principles.(4) Antimicrobial stewardship refers to the actions veterinarians take individually and as a profession to preserve the effectiveness and availability of antimicrobial drugs through conscientious oversight and responsible medical decisionmaking while safeguarding animal, public, and environmental health. Core principles as defined by the AVMA are Antimicrobial stewardship involves maintaining animal health and welfare by implementing a variety of preventive and management strategies to prevent common diseases; using an evidence-based approach in making decisions to use antimicrobial drugs; and then using antimicrobials judiciously, sparingly, and with continual evaluation of the outcomes of therapy, respecting the client s available resources. More details on the principles are provided on the AVMA website. Medically important vs. non-medically important antibiotics Medically important antibiotics are those antibiotics which are deemed important for treating diseases in humans. This classification has no bearing on the likelihood or magnitude of resistance development due to use in food animals. In the United States the regulatory status of antibiotics as medically important, as listed below, is defined in Appendix A of Guidance for Industry (GFI) 152.(16) The FDA Center for Veterinary Medicine (CVM) is currently in the process of working with the FDA Center for Drug Evaluation and Research (CDER) to revise this document. The groups with superscript 1 are those that the World Health Organization cites as being the Highest Priority of the Critically Important classification.(37) The groups with superscript 2 have at least one member of the group with a food animal label. Beta-lactams a. Penicillins 2 natural, penase resistant, antipseudomonal and amino-penicillins b. Cephalosporins 1,2 1 st through 4 th generation in the document, a revision would most likely also include 5 th and subsequent generations. The example with a poultry label is ceftiofur sodium (Naxcel ) c. Carbapenems d. Monobactams Quinolones and fluoroquinolones 1,2 (no products in the U.S. for poultry) Aminoglycosides 2 Macrolides 1,2 (includes ketolides, triamalides, and azolides) Lincosamides 2 (clindamycin is the class representative) Tetracyclines 2 Glycopeptides 1,2 Streptogramins 2 (Virginiamycin in food animals, Synercid in humans) Oxazolidinones 16

Pyrazinamide Isoniazid Rifamycins Chloramphenicol 2 (florfenicol in food animals) Metronidazole Trimethoprim/sulfa 2 (food animal labels in the U.S. are not potentiated with trimethoprim or other potentiators) Polymyxin B 1 Of these drugs, the medically important antibiotics used in preventive programs at SFI, Inc., include virginiamycin (Stafac, a streptogramin). Antibiotics used in preventive programs at SFI, Inc., which are not medically important antibiotics are Bacitracin methylene disalicylate (BMD ), and the narasin component of the combination drug Maxiban. It should be noted that the cyclic glycopeptides, which include bacitracin, are not classified as medically important in the U.S., but they are included as a medically important antibiotic in the WHO list of medically important antibiotics. While not used systemically in humans, it is a component of topical products in which bacitracin serves as a component aimed at Gram + bacteria. Other non-medically important antibiotics which are available for use in poultry feed that are not used at SFI, Inc., are Avilamycin (Inteprity ), and Bambermycins (Flavomycin ). In-feed medically important antibiotics available for treatment of E. coli infection include sulfadimethoxine/ormetoprim (RofenAid ) and neomycin/oxytetracycline (Neo-Terramycin ). In 2017 at SFI, Inc., 90,500 broilers received Neo-Terramycin in a broiler study, and 90,300 broilers received RofenAid in the feed. RofenAid has been unavailable for approximately a year and a half at the time of this writing. The other uses of medically important antibiotics at SFI, Inc. in 2017 included the following. In-ovo o Gentamicin (an aminoglycoside) to all eggs at 18 days of incubation at a dose of 0.05 to 0.1 mg/embryo for the control of post-hatch mortality related to E. coli. Broilers administered medically-important antibiotics through the water o Lincomycin (a lincosamide) in the water if water acidification is unsuccessful for control of necrotic enteritis or gangrenous dermatitis. The threshold for moving to lincomycin is more than 50 dead birds/house per day. Records indicate administration to 4,217,720 birds in 2017 (0.75% of birds) o Penicillin (a natural penicillin of the beta-lactam group) in the water as a rescue drug if lincomycin was ineffective. Records indicate administration to 1,016,000 birds in 2017 (0.18% of birds) o Sulfadimethoxine (a sulfa) in the water to 16,500 birds (0.003% of birds in 2017) o Oxytetracycline (a tetracycline) to 446,100 birds (0.08%) 17

The use of in-ovo gentamicin may contribute to the relatively high rates of gentamicin resistance found in the bacteria E. coli isolated from U.S. poultry products or poultry processing environments, as compared to samples from U.S. beef or pork (Figure 2). Figure 2: Gentamicin resistance in E. coli isolated from various commodities. These data were retrieved from the National Antibiotic Resistance Monitoring System (NARMS) Now: Integrated Data Website hosted by the Food and Drug Administration Center for Veterinary Medicine. These data are available at https://www.fda.gov/animalveterinary/safetyhealth/ antimicrobialresistance/nationalantimicrobialresistancemonitoringsystem/ucm416741.htm Table 1 illustrates antibiotics approved in the U.S. with food animal labels by species, indication, and route.(3) Classes of antibiotics sold for use in food animals are reported in Table 2 from the FDA Center for Veterinary Medicine report on 2016 sales of antibiotics with a food animal label (excerpted as Table 3b in the FDA/CVM report).(23) Species values are based on estimates by drug sponsors as to species distributions of their product sales. 18

Table 1: Antibiotics labeled for food animal use in the United States classified by medical importance status with reported sales proportion in 2013, by species and labeled routes of administration. These routes may be as part of a combination product, are approved specific to disease indications, and may also include only specific use classes and age restrictions within the species. This table contains label approvals as represented at "Animal Drugs@FDA", the electronic version of the "Green Book" containing FDA CVM approvals. Medically important antimicrobials as defined in Guidance 152 Appendix A Beef % of antibiotic sales for food animals in U.S. for this category in 2013 Class Drug Swine Cattle, Non- Lactating Dairy Cattle Lactating Dairy Cattle Goats Sheep Chickens Turkeys Dihydrostreptomycin IMM Streptomycin O O IMM O 2.9% Aminoglycosides Gentamicin W, I, O T I I Neomycin W,M,F,O W,M,F,O W,M,F,O W,M,F,O W,F W,F Spectinomycin O W, I I 0.3% Cephalosporins Ceftiofur I I I, IMM I I I I Cephapirin IMM 0.2% Fluoroquinolones Enrofloxacin I I I Danofloxacin I 2.6% Lincosamides Lincomycin I,F,W W,F Pirlimycin IMM Tulathromycin I I Erythromycin F,I IMM W,F F Gamithromycin I 6.1% Macrolides Tildipirosin I Tilmicosin O O,I Tylosin F,I,W F,I F,W W Tylvalosin W 19

9.0% Penicillins Amoxicillin IMM Ampicillin O,W,I O,I I Cloxacillin IMM Hetacillin IMM Penicillin G I I IMM W Sulfadimethoxine O,W,I I W W Sulfadimethoxine/ Ormetoprim F F 4.2% Sulfonamides Sulfamethazine W,F W,F,O W W Sulfachlorpyridazine W,O W,O,I Sulfaethoxypyridazine W W,O,I Sulfamerazine W W Sulfaquinoxaline W F,W F,W Chlortetracycline W,F,O W,F,O F W,F W,F 70.8% Tetracyclines Oxytetracycline W,I,F W,I,F,O I W,F W,F W,F Tetracycline W W W W Not Individually Amphenicols Florfenicol W,F I Reported = 3.9% Streptogramins Virginiamycin F F F F Antimicrobials not categorized as medically important as defined in Guidance 152 Appendix A Beef % of antibiotic sales for food animals in U.S. for this category in 2013 Class Drug Swine Cattle, Non- Lactating Dairy Cattle Lactating Dairy Cattle Goats Sheep Chickens Turkeys Monensin F F F Lasalocid F F F F 79.3% Polyether Ionophores Laidlomycin F Salinomycin F Narasin F F F 20.7% Aminocoumarins Novobiocin IMM Glycolipids Bambermycins F F F 20

Bacitracin zinc F F F F Polypeptides Bacitracin Methylene Disalicylate F F F F Pleuromutilins Tiamulin W,F Quinoxaline derivatives Carbadox F F IMM I IU M O S T V W Feed Intramammary Injectable Intrauterine Milk Oral Solid dose implant Topical Intravaginal Water 21

Advisory Board Assessment of Antibiotic Use at Sanderson Farms, Inc. November 2018 Table 3 from the same report illustrates the high proportion of medically important antibiotics administered to food animals through the feed (72% of medically important antibiotics administered to food animals) and water (23%) (excerpted as Table 4 from the FDA/CVM report.(23) These values are based on reported percent of total weight of active ingredient and do not reflect potency of the products. At the writing of this report the 2017 sales data have not yet been released; it is expected that feed antibiotic administration in food animals will decrease in the 2017 data due to the new product labels with Veterinary Feed Directive (VFD) requirements for use which became effective January 1, 2018. Table 2: Table 3b from FDA/CVM report on 2016 sales of antibiotics labeled for food animals. 22

Table 3: Table 4 from FDA/CVM report on 2016 sales of antibiotics labeled for food animals. The special case of the ionophores In the United States, the ionophores are considered to be antibiotics. This is because they meet the definition of an antibiotic, which is a substance produced by one organism which impedes the growth of, or kills another organism. In Europe, the ionophores are classified as anticoccidials rather than antibiotics, allowing their use in antibiotic-free systems. Because coccidiosis is one of the most common diseases of poultry, this difference in classification of ionophores makes it more difficult for U.S. RWA/NAE poultry systems to keep birds healthy, as compared to European RWA/NAE systems. It is important to recognize that ionophores are considered to be non-medically important because they are not used in human medicine. There is no credible evidence that use of ionophores in any food animal species is selecting for resistance to medically-important antibiotics for human or animal therapy. There are multiple ionophores available for use in poultry feed in the United States. Rotation of ionophores in an attempt to minimize selection for resistant coccidia is a common practice in the industry, and is practiced at SFI. This list includes commentary on use in SFI production, 23