Click for next page ( 2


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
CHAPTER 1 THE USE OF ANT IMICROBIAL A(;ENTS1 The discovery of the first selective antimicrobial agent approximately four decades ago was a major milestone in the his- tory of medicine and human health. The subsequent development of antimicrobial therapy largely centered on the search for drugs with effectiveness against microbial species that were not susceptible to drugs then in use. These powerful new drugs have been shown to save lives when used to treat some severe infections and to reduce the burden of illness when used prophylactically in certain clinical situations (Utz, Appendix A). Since antibiotics] are isolated from microorga- nisms, strains of some microbial species have predictably evolved the capacity to inactivate them or became impermeable to them, i.e., these strains have developed resistance to these antibiotics. Re- sistance to synthetic antimicrobial agents arises from the variation normally displayed by individual microorganisms within species. Thus, the consequence of expanded use of antimicrobial drugs has been an increased prevalence of resistant organisms resulting from the selection process. In certain places, such as hospitals, contact between individuals has facilitated the spread of these resistant bacteria. Consequently, researchers sought agents that were active against strains in which resistance had become prevalent. The ex- panding array of antimicrobials, particularly antibiotics, provided alternatives in most cases. But control of infections is sometimes delayed if the resistance of the infecting organism is not recognized immediately, and physicians may need to use drugs that are more toxic, more expensive, or less effective than those that would be selected if the infecting organisms were not resistant (Utz, Appendix A). Trends in the antimicrobial resistance patterns of a number of clinically important pathogens have been reviewed by Finland (1979) and Stollenman (1978~. The prevalence of multiply ant~microbial- resistant Staphylococcus aureus increased until 1960 but subsequently declined in association with a change in phage types. Recently, strains of Streptococcus pneumonias with multiple resistance have been found in a number of countries. Strains of Haemophilus An antibiotic is a chemical substance produced by microorganisms that has the capacity at low concentrations to inhibit the growth of or to destroy bacteria and other microorganisms. An antimicro- bial is any agent that destroys microorganisms or suppresses their multiplication or growth. 1

OCR for page 1
2 influenzas producing f-lactamase and occasional strains with resistance to chloramphenicol have also been observed, as have strains of Neisseria gonorrhoeae producing plasmid-mediated S-lactamase. Other changes are noted by Finland (1979~. Differences in resistance encountered in certain pathogens of humans are often observed in separated geographical areas (Finland, 1979), in different hospitals (O'Brien, Appendix I), or in local outbreaks, e.g., of chloramphenicol-resistant strains of Salmonella typhi in Vietnam and Mexico (Finland, 1979~. The committee could find no comparable assessments of the trends in resistance to antimicrobials that might have occurred in the major pathogens of food animals over the last three decades. The prevalence of clinically significant antimicrobial- resistant bacterial strains correlates with the increasing use of antimicrobial agents in the course of clinical practice (Finland, 1955a,b,c; 1979~. The necessity for therapeutic use of antimicrobial agents in the treatment of overt disease in animals has not been questioned. However, the continuous use of subtherapeutic levels of antimicro- bials in animal feeds for growth promotion, improvement of feed efficiency, and disease prophylaxis has been criticized as posing dangers to human health by making an important quantitative con- tribution to the pool of antimicrobial-resistant bacteria that may be transferred to the human population. Possible "qualitative" effects of the selection pressure imposed by subtherapeutic usage on resistance profiles or transfer mechanisms are discussed below and by O'Brien (Appendix I). For regulatory purposes, the Food and Drug Administration (FDA) defines subtherapeutic use as the administration of doses less than or equal to 200 g of antimicrobial per ton of feed for 2 weeks or longer. However, the FDA has approved the marketing of some antimicrobial agents for use at levels below 200 g/ton to treat certain diseases (Animal Health Institute, 1979~. There- fore, the current definition of subtherapeutic use in animal feeds encompasses certain uses that are therapeutic in intent in addi- tion to those for prophylaxis and the improvement of growth and efficiency of feed conversion. In the hope of preserving the effectiveness of the antimicro- bial agents that are important in the therapy of human diseases, some governments (e.g., the United Kingdom) have regulated the use of these agents as growth promotants in animal feeds (Swann et al., 1969~. Antimicrobials used to treat humans are still approved in those countries for use in animals on veterinary prescription.

OCR for page 1
3 Similar actions have been under cons iteration in many other coun- tries, including the United States. Since penicillin and the tetracyclines are effective and widely used in the therapy of human disease, the FDA has proposed restriction of their subtherapeutic use in animal feeds. This committee has attempted to determine if human health is affected by the subtherapeutic use of antimicrobials in animal feeds and to ascertain what additional information is needed to make a more definitive determination. NATURE OF THE SELECTION PRESSURE IMPOSED BY ANTIMICROBIALS . . It is important to distinguish between the effects of an anti- microbial drug on a single ant~microbial-sensitive microbial strain and the effects on a heterogeneous mixture of species or strains. When antimicrobial drugs are brought into contact with multiplying susceptible microorganisms, the organisms are generally inhibited from multiplying further or are killed. When the susceptible orga- nisms constitute a portion of the total microbial flora that is exposed to the drugs, the elimination of the susceptible organisms is generally followed by some degree of compensatory multiplication of the more resistant or nonsusceptible strains. Such a shift in the composition of the enteric flora may facilitate infection by a pathogen (Seelig, 1966~. Another important consideration in the evaluation of possible effects on human health is the possibility of "qualitative" as well as quantitative changes i n resistance brought about by the continuing selection pressure exerted by subtherapeutic levels of antimicrobials in animal feeds. To date, most research has been focused on quan- titative changes, i.e., changes in the prevalence of resistance, primarily because techniques to study qualitative changes have only recently been developed. Qualitative changes could include the development of new combinations of resistance genes, combina- tion with genes for other characteristics, e.g., toxins, the spread of such plasmids to new hosts and, possibly most seriously, the evolution of more efficient, wider host-range transfer mechanisms. The evidence for such changes is discussed by Jacoby and Low (Appendix C). The implications of this "molecular epidemiology of plasmids" are considered by O'Brien (Appendix I). Another change that could be regarded as qualitative is the shift in the composition of the gastrointestinal flora under the selection pressure of sub- therapeutic dosages of antimicrobials. This shift will produce changes in the interactions of the flora components with each other and with the host. Little is known about the composition of the

OCR for page 1
4 gastrointestinal flora in either humans or animals, especially the anaerobic components (Savage, Appendix D). Therefore, because of the lack of data and methodology to carry out _ -viva experiments, it is not possible to assess the significance for human health of such shifts. Investigators who have studied resistance in the gastroin- testinal flora of humans or animals have generally observed changes in the prevalence of R factors in E. cold or its close relative Salmonella because these organisms are easy to culture and manipu- late and because they can be pervasive pathogens in both humans and animals. The significance of resistance in these two species may not be the sate since their "ecology" is different. This is dis- cussed briefly in the next section. A thorough assessment of the significance for human health of various uses of antimicrobials requires knowledge of the consequences of the selection pressures imposed by intermittent, therapeutic doses versus subtherapeutic continuous feeding and of different routes of administration. Unfortunately, there are insufficient data comparing These regimens. Such data would have been of invaluable assistance to the committee during its deliberations. MECHANISMS AND TRANSFER OF RESISTANCE In most instances bacterial resistance to antimicrobial agents is conferred by extrachromosomal genetic elements called plasmids. Those conferring resistance are called R factors or R plasmids (Jacoby and Low, Appendix C). These plasmids are widely distri- buted among bacterial species, including those that are pathogenic in humans and animals. Some plasmids transfer between species of different genera. Two general types of R plasmids have been identified on the basis of their transmissibility characteristics: Large R plasmids harbor approximately 100 genes and are trans- missible to other cells by a process called conjugation or bacterial mating. Approximately 25 plasmid genes code for functions that are required for transmissibility, several other genes code for repli- cation functions, and from four to six genes generally code for resistance to antimicrobials. The f unctions of other genes on the larger R plasmids have not been identified. Small R plasmids usually harbor approximately 10 genes. These plasmids do not carry the genes for transmissibility and are not transmissible when by themselves. However, a transmissible plasmid

OCR for page 1
5 may sometimes mediate the transfer of an otherwise untransmiss- ible one when present simultaneously in the same cell. Some pairs of R plasmids cannot coexist stably in the same host cell, a phenomenon referred to as incompatibility. This simple criterion has enabled investigators to differentiate R plas- mids into more than 40 incompatibility groups. The mechanism of incompatibility is not yet understood. It may be related to the control of plasmid replication. Different R plasmids may represent a group of genetic elements of diverse origin. However, even if located on plasmids in different incompatibility groups, the genes that confer resistance to specific antimicrobials are very similar. This may be explained by recent observations that certain antimicrobial resistance genes are located on segments of DNA (transposons) that may spontaneously translocate from one plasmid to another, thereby disseminating the resistance genera) among various plasmids (Kleckner, 1977~. Factors that pro- mote or inhibit the transfer of R factors In viva are not completely understood (Jacoby and Low, Appendix C). The biochemical mechanisms of resistance are known in most instances. R plasmids confer resistance to antimicrobials either by encoding for enzymes that chemically modify and thus inactivate the agent, by specifying a substitute metabolic enzyme that is insensitive to the agent, or by specifying a decrease in cell permeability to the agent. All three types of resistance mecha- nisms may be determined by the same plasmid, whether transmissible or nontransmissible. Thus, R plasmids are endowed with genes that increase the probability of survival of host cells in the presence of combinations of antimicrobials. The continued spread among bac- teria of resistance to more than one antimicrobial and the further acquisition of additional resistance genes by individual R plasmids results from the selection pressure imposed by the use of antimicro- bials. A number of factors affect the transfer rate+ of plasmids between species, e.g., the frequency with which R enteric bacteria come into contact with other bacteria and the environment in which they meet. The logistics of transfer of bacteria from animals to humans and of interbacterial transfer of plasmids may be different for different organisms. Salmonellae are generally infrequent, abnormal components of the flora of animals and humans, but when present they occur in enormous numbers that increase the potential for transfer at such times. _ coli, compared to the anaerobic flora, usually constitute a numerically small proportion of the gastrointestinal flora of animals and humans, but its continuous

OCR for page 1
6 presence (Savage, Appendix D) offers a different potential for transfer. Bacterial genetic aspects of drug resistance have been reviewed by Jacoby and Low (Appendix C). The dissemination of drug-resistance genes among diverse bacterial genera is also dis- cussed by these authors as well as by O'Brien (Appendix I). The epidemiology of plasmid transfer has not been studied in suffi- cient detail. Whatever motivation lies behind the use of antimicrobial agents in any specific situation, the consequences will be the same. Administration of an antimicrobial will result in a selec- tion pressure favoring an increase in the prevalence of resistant organisms. THE USE OF ANTIMICROBIALS IN HUMAN MEDICINE The use of antimicrobials in both hospitalized and ambulatory patients is extensive. Kunin (1979) reported that between 23% to 37.8% of hospitalized individuals receive them. Finkel (1978) estimated from dispensed prescription data and FDA certification records that approximately 190 million prescriptions for the major ant~microbials were filled for ambulatory patients in the United States in 1977. This is nearly one course of treatment per year for each person in the United States and includes approximately 43 million prescriptions for tetracycline. The consequences resulting from the administration of anti- microbials to humans have been examined by Finland (1979~. Hartley and Richmond (1975) reported that oral intake of tetracycline leads to the emergence of a predominantly tetracycline-resistant coliform gastrointestinal flora within 48 hours in those treated. The excre- tion of resistant organisms continues at least 10 days after the treatment is terminated (Richmond, 1975~. A complete evaluation of the increased prevalence of resistance to antimicrobials would require consideration of not only the contri- butions from subtherapeutic and therapeutic use in animals but also the extent to which these agents are administered to humans. Richmond and Linton ( 1980) studied the use of tetracyclines in the County of Avon in England and its possible relation to the excreti on of tetracycline-resistant bacteria. They estimated that one in 130 individuals in the county carried a large proportion of tetracycline- resistant organisms in their alimentary tracts. Examination of swabs from sewers in predominantly residential areas of Bristol (in Avon) (Lipton _ al., 1974) indicated that approximately 3% of the isolated

OCR for page 1
7 coliforms were resistant to tetracycline. Richmond and Linton (1980) concluded that, . . . there hardly seems a need to postulate a veterinary source for the resistant coliforms encoun- tered in the human population. This is not to say that resistant _ cold of animal and poultry origin cannot reach the human popu- lation: clearly they can and do (Lipton et al., 1977~. And some resistant salmonellae of animal origin certainly seem to have caused serious human epidemic disease (Anderson, 1968[b]~. But whether the use of antibiotics in the animal and poultry rearing industries has a major quantitative impact must be question- able; . . ." The approach adopted by Linton and Richmond is necessarily indirect and requires a number of approximations and assumptions. Although no similar studies have been conducted in the United States, the data reported by Kunin (1979) and Finkel (1978) on prescriptions in this country indicate that the administration of antimicrobials to humans is widespread. THE USE OF ANTIMICROBIALS IN AGRICULTURE The livestock and poultry industry has undergone dramatic changes since 1950. Operations that were extensive became more intensive. There were increases in the size of facilities, the number of animals reared, and a move toward centralization. Socioeconomic changes, as well as advances in biomedical sciences, nutrition, engineering, and management, have all contributed to this evolution. Shortly after anthmicrobials had been discovered and their therapeutic use in humans and animals had begun, investigators J earned that the addition of antimicrobials to animal feed was effective in growth promotion, improvement of feed conversion, prophylaxis, and treatment of certain diseases. These effects of antimicrobials are especially useful when animals are stressed either by intensive husbandry practices or shipment. The use of antibiotics (and most probably sulfonamides) in animal husbandry has steadily increased since 1950 as has animal production (Table 1~. In 1978 approximately 48% of the antibiotics produced were designated for addition to animal feeds or for other (minor) uses (U.S. International Trade Commission, 1979~. The motivation for such use and the economic consequences of restrict- ing subtherapeutic concentrations of antimicrobials in feed have been dealt with in reports by the National Academy of Sciences Board on Agriculture and Renewable Resources (BARR, Appendix K) and the U.S. Department of Agriculture (1978~.

OCR for page 1
8 TABLE 1 Antibiotic Production from 1950 to 1978 (millions of kg~a~b Medicinal Use in Added to Animal Feed and Year Total Humans and Animals Other Applications 1978 11.66 6.08 5.58 1977 10.48 6.35 4.58 1976 9.30 4. 72 4.54 1975 8. 30 4. 26 4.04 19 74 9.30 5.99 3. 36 1973 9.43 5.72 3.72 1972 7.53 4.45 3.08 1971 8.12 4.90 3.22 1970 7.67 4.35 3.31 1969 5. 99 3. 36 2.63 1968 4.67 2. 72 1.95 1967 4.29 2.36 1.91 1966 4.40 2.45 1. 91 1965 3.40 2.13 1.27 1964 2. 95 1.77 1. 18 1963 3.04 1.91 1.13 19 62 2. 86 1.81 1.04 1961 2.31 1.50 0.82 1960 2.13 1.36 0.77 1959 1.68 1.04 0.64 1958 1.59 1.18 0.41 1957 1.47 1.08 0.39 1956 1.24 0.89 0.35 1955 0.95 0.71 0.24 1954 1.05 0.83 0.22 1953 0.94 0.74 0. 2C 1952 0.79 0.67 0. 12 1951 O. 69 0.58 0.11 1950 0.39 0. 39 Mentioned, but no figure Data extracted from reports of the U. S. ~ 1951-1979) . bValues exclude production of sulfonamides. In te rna t tonal Tr ade Commi s sio n

OCR for page 1
9 The mechanisms by which antimicrobials improve growth and the efficiency of feed conversion are not fully understood. Some suggested effects include modification of host metabolism, nutrient sparing or alteration of nutrient absorption, and selec- tive activity against microorganisms (BARR, Appendix K). Total amounts used and patterns of usage, which vary with species and geographic location, are described in reports by BARR (Appendix K) and the Animal Health Institute (1979~. Information on the total amount of specific antimicrobials administered for each application listed above does not appear to be available. Certain antimicrobials (streptomycin, tetracyclines, peni- cillin) are also used to control plant pathogens. Although this application might also have consequences for human health, the amounts used are much smaller (Goodman, Appendix B). The addition of subtherapeutic amounts of antimicrobials to animal feeds continues to be of concern because of its implica- tions for human health and because some believe that this use is unessential. In the United States therapeutic concentrations of drugs are given to livestock extensively, with or without veteri- nary prescription. Treated animals are not always isolated from untreated ones, and animal-to-animal transfer of Rob organisms is known to occur (Levy, 1978~. Since most herds and flocks receive antimicrobials somewhere in the production chain either for growth promotion, prophylaxis, or therapy, it is difficult to identify slaughtered livestock that have not been given antimicrobials or have not been exposed to animals that had. DEF INITION OF HUMAN HEALTH HAZARD The difficulty of determining whether human health is affected by the subtherapeutic use of antimicrobials in feeds is compounded by the diversity of opinions concerning the definition of a hazard to human health. Some view an increase in the pool of antimicrobial-resistant bacteria or an increase in Salmonella shedding by food animals as a source of danger. Others maintain that a significant hazard to health exists only if antimicrobial- resistant organisms can be shown to be transmitted to animal hand- lers or to meat processors. Others attach importance only to the passage of resistant microorganisms to meat consumers. Still others view these circumstances as unrealized, potential hazards. These scientists insist that incremental morbidity and, perhaps, excess mortality that can reasonably be attributed to resistant

OCR for page 1
10 organisms resulting from subtherapeutic use of antimicrobials must be documented before risk to human health from this cause can be substantiated. This divergence of opinion was reflected in the committee itself. There are no data from which to predict quantitatively the change in the morbidity and mortality of humans that might result from an increased prevalence of resistant bacteria in animals or from the transfer of these organisms to humans. Although a measurable risk to human health cannot be ascribed to these phe- nonema, they remain plausible potential hazards since some of the individual steps in the transmission chain to humans have been independently demonstrated but not quantified (see Chapter 3~. An increased prevalence of resistant bacteria may result from the administration of both therapeutic and subtherapeutic levels of antimicrobials to animals and human beings. Thus, the question is raised whether antimicrobials used subtherapeutically in the meat industry add measurably to the carriage of resistant organisms, the incidence of clinical illness, or the number of complications resulting from antimicrobial resistance in the treatment of diseases. The committee could find little relevant information on the relative selection pressures for antimicrobial resistance exerted by continu- ous low-dose feeding versus intermittent higher, therapeutic doses. This subject is examined later in this report. Other possible hazards to health (also discussed below) re- ceived committee attention but were eliminated from further deliber- ations because they were not central to the major issue or because there was insufficient information. Katz (Appendix E) prepared for committee consideration an assessment of the potential hazards to humans presented by resi- dues of antimicrobials in livestock and poultry meat products. Surveys have shown that slaughtered animals may contain residues of penicillin or tetracycline that probably resulted from inade- quate withdrawal times or large dosages (Huber, 1971~. More recent surveys (Katz, Appendix E; Mussman, 1975; USDA, 1979) indicate that the residues resulting from penicillin or tetracycline used as feed additives were generally below the limits currently permitted by the FDA. Residues of tetracyclines were undetectable in animals slaughtered 1 to 5 days after withdrawal of antimicrobial-containing feed. The small amounts of these residues in the muscle tissues of animals do not survive normal food preparation because of heat in- activation during cooking (Katz, Appendix E).

OCR for page 1
11 Katz (Appendix E) reported that there has been concern that the therapeutic use of penicillin has resulted in significant residues. However, he noted that current regulations governing the subtherapeutic use of penicillin as a feed additive appear to result in very low or infrequent residues in meat. Katz concluded, "It is doubtful that antibiotic residues or their degradation products will provide any selective pressure on enteric bacteria contaminating the carcasses of animals." The committee concurs with this assessment. It believes that further studies of the effect on human health resulting from penicillin and tetracycline residues in meat would not elucidate the hazard from subtherapeutic levels of antimicrobials in animal feeds. The committee also viewed an assessment of the immunological consequences to humans resulting from penicillin and tetracycline residues in livestock and poultry meat products (Adkinson, Appendix J). Although pertinent information is limited, the committee con- curs with Adkinson's conclusion that "there is little reason to believe that foodstuffs obtained from animals fattened with anti- biotic-supplemented feed impose a significant risk to human health by contributing to antib~otic-induced allergic reactions." Adkinson indicated that further investigations in several areas could provide information that would be useful in clinical situations. However, the committee believes that immunological problems arising from the use of antimicrobials in animal feeds are not a serious health risk for the general population. In attempting to define the possible hazards to human health, the committee wished to know how the acquisition of antimicrobial resistance affected the virulence of pathogens infecting humans and animals. Since limited conclusive information appeared to be available (Jacoby and Low, Appendix C), the committee decided not to review this topic in depth.