Cover Image

HARDBACK
$59.95



View/Hide Left Panel

5
Categories of Scientific Evidence—Animal Data

GUIDING PRINCIPLE: Even in the absence of information on adverse events in humans, evidence of harm from animal studies is often indicative of potential harm to humans. This indication assumes greatest importance when the route of exposure is oral, the formulation tested is identical or highly similar to that consumed by humans, and more than one species show the same or similar toxicity.

Information about animal exposure to dietary supplement ingredients may be in the form of formal studies such as traditional toxicity studies, safety pharmacology data, or observations from clinical veterinary medicine. Because dietary supplements are not required to undergo formal animal toxicity testing before marketing, extensive toxicity studies common to drugs and other substances are not likely to exist, but limited amounts of animal data are available in the scientific literature for a number of dietary supplement ingredients. Despite the challenges of dealing with incomplete data, the animal data that are available warrant attention when assessing risk of dietary supplement ingredients.

The first section of this chapter describes types of animal data that may be available. Subsequent sections describe the rationale for using animal data, including its power and relevance to human health. Also described is the appropriate consideration of negative data, and how the seriousness of



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 156
Dietary Supplements: A Framework for Evaluating Safety 5 Categories of Scientific Evidence—Animal Data GUIDING PRINCIPLE: Even in the absence of information on adverse events in humans, evidence of harm from animal studies is often indicative of potential harm to humans. This indication assumes greatest importance when the route of exposure is oral, the formulation tested is identical or highly similar to that consumed by humans, and more than one species show the same or similar toxicity. Information about animal exposure to dietary supplement ingredients may be in the form of formal studies such as traditional toxicity studies, safety pharmacology data, or observations from clinical veterinary medicine. Because dietary supplements are not required to undergo formal animal toxicity testing before marketing, extensive toxicity studies common to drugs and other substances are not likely to exist, but limited amounts of animal data are available in the scientific literature for a number of dietary supplement ingredients. Despite the challenges of dealing with incomplete data, the animal data that are available warrant attention when assessing risk of dietary supplement ingredients. The first section of this chapter describes types of animal data that may be available. Subsequent sections describe the rationale for using animal data, including its power and relevance to human health. Also described is the appropriate consideration of negative data, and how the seriousness of

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety harm, strength of evidence, and dose administered to animals factor into assessing animal data, along with general guidelines for integrating these factors. Animal data that raise a higher level of concern warrant immediate attention to evaluate the potential of the ingredient to cause harm. For data classified as lower to moderate concern, it is important to consider whether other animal data or other types of data (e.g., human data, in vitro data, or data on related substances) add to the level of concern. POWER AND RELEVANCE OF ANIMAL DATA Animal testing provides invaluable information about the potential for ingested substances to cause harm in humans. Studies in animals are regularly used as an important step in attempting to predict untoward effects of substances in humans (see, for example, the Food and Drug Administration’s [FDA’s] Redbook [OFAS, 2001, 2003] or guidance documents for new drugs [CDER, 2002]). Animal studies are powerful because controlled studies can be conducted to predict effects that might not be detected from customary use by humans until they result in overt harmful effects. Animal studies are especially useful in detecting effects of chronic exposures and effects on reproductive and developmental processes because epidemiological methods of studying humans are especially problematic in these areas. The ability to administer agents to animals during their entire lifespan, if necessary, enables scientists to ascertain the potential toxic effects that may arise from long-term (chronic) exposure. Animal studies thus serve as important hypothesis generators and may be sufficient to indicate potentially unreasonable risk to human health, which justifies their use in evaluating the risks dietary supplement ingredients may pose to humans. In general, adverse effects observed in well-designed and well-conducted animal studies should be treated as if they would occur in at least some members of the human population, assuming humans receive a sufficiently high dose. With some notable and important exceptions, the biological factors affecting the capacity of chemical substances to cause toxicity are broadly similar across mammalian species. Unless there is scientific evidence that raises significant doubt regarding the relevance of specific toxicity findings to humans, it is prudent and scientifically appropriate to consider animal studies relevant in evaluating potential human toxicity, especially in the many cases of dietary supplement ingredients where sufficient human data are not available. Similar positions on the relevance of animal data to human health have been supported by other committees of the National Academies, as well as by other organizations in the United States and internationally (NRC, 1994, 2001; NTP, 2002; WHO, 1999).

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety GENERAL TYPES OF ANIMAL DATA Different categories of animal studies (e.g., acute, subacute/subchronic [often used interchangeably], chronic, carcinogenicity) provide different information relevant to considering the potential toxic properties of a dietary supplement ingredient and can be classified as either traditional toxicology studies or as safety pharmacology studies. Traditional Toxicology Studies FDA’s Redbook describes several toxicology studies that are typically conducted in assessing the safety of food additives and other ingested substances (OFAS, 2001, 2003). These studies are applicable to evaluating most ingested substances, including dietary supplement ingredients, irrespective of what is known about their biological activities. It is highly probable that animal data from each type of toxicity study will not be available for every dietary supplement ingredient. However, consideration of the typical study protocols enables the animal data that are available on the dietary supplement ingredient in question to be placed in perspective regarding what type of information and conclusions about safety are appropriate to glean from the different study designs and endpoints. Perspective can also be gained by comparing the information available about a dietary supplement ingredient with the types of data that are often available about other ingested substances before they are considered safe. In acute (single dose), subacute/subchronic (repeated doses), and chronic toxicity testing, groups of animals are treated with increasing amounts of the test substance to determine the dose that induces overt toxic effects. The resulting toxicities might be within organs (detected by gross examination or by observing behavioral changes), cells (detected by histological examination, such as light or electron microscopic analysis of fixed tissue samples), or subcellular structures (detected in biochemical studies, such as enzyme assays or protein analysis). In chronic toxicity testing (and in subchronic toxicity testing, which is not as lengthy as chronic toxicity testing), the test substance is typically administered to animals on a daily basis for 3 to 24 months (depending on the species) to characterize possible longer-term toxicity. When conducting animal studies, blood concentrations of the test substance and its active metabolites are often determined. These blood levels are used to provide evidence that the test substance was absorbed, to describe the blood concentration–response curve, and to determine whether the metabolites formed in the test animal are qualitatively and quantitatively similar to those formed in humans. If the metabolites, especially active metabolites formed in the animal species studied, are not the same as

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety those formed in humans, the results are less meaningful and testing in species with metabolism similar to humans should be considered. Genetic, reproductive, developmental, immunological, neurobiological, and behavioral toxicity studies, as well as other types of studies, provide further information regarding the toxicity of the test substance. Safety Pharmacology Studies Safety pharmacology studies are conducted in various animal species to detect alterations in physiological functions at dosages lower than those used to elicit overt toxic effects detected in animal toxicity protocols. Guidance for conducting safety pharmacology studies for human pharmaceuticals is provided by FDA, which defines them as “those studies that investigate the potential undesirable pharmacodynamic effects of a substance on physiological functions in relation to exposure in the therapeutic range and above” (ICH/FDA, 2001). Safety pharmacology testing generally focuses on endpoints that differ from those examined in classic toxicity testing. The studies may be in vivo or in vitro and are designed to detect harmful effects in a core battery of vital organ systems, which include the cardiovascular, central nervous, and respiratory systems. Observations from Veterinary Medicine Veterinary toxicological observations may also prove useful in predicting the potential effect of dietary supplement ingredients on humans. The discipline of veterinary medicine/toxicology encompasses the entire spectrum of effects of natural and synthetic toxins, including drugs, pesticides, herbicides, and fungal and plant metabolites, on wildlife, livestock, and domestic animals (i.e., pets). The specific subdiscipline best described as plant-associated veterinary toxicology is likely to correlate most closely with adverse effects of botanical-derived dietary supplement ingredients. It is distinguished from toxicological studies in that it is primarily observational information or is based on studies not designed to predict effects on human health. Nevertheless, there are numerous examples of incidents of animal poisoning that have subsequently led to epidemiological studies and ultimately controlled experiments that resulted in identification of specific toxins and their mode of action. The well-known cases of aflatoxin-induced poultry toxicity led to the controlled animal and epidemiological studies that resulted in the classification of this important fungal metabolite as a human carcinogen (Mishra and Das, 2003). An advantage of considering plant-associated animal toxicity observations in livestock is that episodes of poisoning often occur on a large scale, affecting tens or even hundreds of animals, so that there is little doubt as to

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety the significance of the information. Livestock poisoning also tends to be worldwide in distribution, with numerous episodes reported in North America, Australia, and South Africa, where there have been significant efforts on the part of national and state governments to control losses. More recently, efforts to confront these problems, and consequently the amount and quality of information, have increased in areas such as China and parts of South America. With domestic animals, reports may only occur on an individual basis, but the close owner-pet relationship leads to more episodes being reported and carefully analyzed, cumulating in large numbers of documented cases in aggregate. However, the types of toxicological information obtained can differ significantly between livestock and domestic animals. Livestock producers, because of their economic interests, are often more likely to be aware of both acute toxicity and chronic effects, such as lack of weight gain, birth defects, infertility, and abortion. Owners of domestic animals, which are not only restricted to cats and dogs but may also include individual cattle, sheep, goats, and horses, frequently report only acute toxicity, typically resulting from poisoning by a house or garden plant. Serious adverse events reported in animals, such as livestock, may also provide helpful information. Reported effects of animal intoxications, similar to spontaneous human adverse event reports, tend to be scattered, with the only nationwide tracking system being the Animal Poison Control Center operated by the American Society for the Prevention of Cruelty to Animals, although this is generally more focused on pets than livestock (ASPCA, 2003). State veterinary diagnostic laboratories, usually located in close association with university veterinary schools, receive many reports of animal poisonings, but this information may not be routinely compiled for general use. However, a comprehensive database of poisonous plants with numerous links to other compilations is maintained by the Department of Animal Science at Cornell University (Cornell University, 2002). The U.S. Department of Agriculture (USDA)/Agricultural Research Service Poisonous Plant Research Laboratory (PPRL) in Logan, Utah, is the only laboratory in the world specifically conducting research entirely devoted to poisonous plants affecting livestock (USDA/ARS, 2003). This program was initiated over a century ago and data acquired since that time are voluminous. Staff at the PPRL, consisting of animal and range scientists, veterinarians, and natural products chemists, have access to most of this historical information and are aware of the most recent episodes of plant-livestock interactions. The veterinary literature provides anecdotal observational information, but there is no single source. There are compendia on effects of poisonous plants on livestock for North America, Australia, and Southern Africa (Cheeke, 1998; Everist, 1981; Keeler and Tu, 1983, 1991; Kellerman et al.,

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety 1988; Kingsbury, 1964). In addition, there are five volumes of proceedings of international symposia on poisonous plants, with particular emphasis on livestock poisonings (Colegate and Dorling, 1994; Garland and Barr, 1998; James et al., 1992; Keeler et al., 1978; Seawright et al., 1985). Most of the available veterinary toxicology reports are observational and not experimental, and the relevance of various species of livestock to human toxicity is not well established. Nevertheless, veterinary toxicology information may be quite useful when it corroborates concerns raised by other types of data. Independent of other types of data, evidence of harm in livestock and other veterinary toxicology information is appropriate to consider as a signal prompting an initial review of an ingredient. In addition, the veterinary toxicology literature is also useful for generating hypotheses in need of testing in well-established animal models. A careful mechanism for ensuring continuing awareness by FDA of this important data source is suggested. CONSIDERATIONS FOR ASSESSING ANIMAL DATA Human Versus Animal Dose The degree of potential human risk is a function of the nature and seriousness of the observed toxicity and the dose at which it occurs in animals relative to the expected human intake of the substance. One of the unique and powerful approaches of animal testing is the administration of high amounts of a substance over a short time period. This allows the detection of effects with small groups of animals, the prediction of possible effects following prolonged human exposure, and the prediction of possible effects on particularly sensitive human subpopulations. Many animal studies focused on toxicity evaluate increasing dosages until signs of toxicity are seen. While the amount administered may not appear relevant to the non-scientist, organ toxicities at elevated intake in acute or subchronic studies can be indicative of toxicities that may develop at lower doses during chronic use of the ingredient and should therefore not be disregarded simply because the dose administered is higher than that taken by humans. On the other hand, in certain instances, data will indicate that positive animal studies conducted at high doses may falsely predict human outcomes because the excessive doses used in animals overwhelm normal detoxification mechanisms that would protect against toxicity at actual levels of human exposure. While the assumption should be that any effects observed in animals are relevant to humans, under some circumstances known differences between humans and animals with respect to the pharmacokinetics and metabolism of a substance, interspecies differences in pharmacodynamics, or

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety other well-characterized biological differences may lessen or eliminate concerns about human toxicity. Clues to such differences may arise when findings in one species of animal are not observed, under similar dosing conditions, in a second species. Such an observation is, however, only a clue, and cannot be taken by itself as evidence of irrelevance to humans. Rather, data that can be used to explain species differences are necessary to draw strong inferences regarding relevance or lack thereof. Bioavailability, Pharmacokinetics and Knowledge of Absorption, Metabolism, Excretion, and Distribution Processes When comparing the quantified ingested dose resulting in animal adverse effects with information about a human ingested dose, it is useful to consider the relationship between the amount ingested and the amount of the substance or its metabolites that reach the active site1 (usually indicated by the concentration of unbound compound in the blood, and described in terms of bioavailability [see Chapter 3]). Pharmacokinetic processes, such as absorption, metabolism, excretion, and distribution, affect how much of the ingested substance actually reaches sites of action in the body. Differences in the pharmacokinetic processes of humans and experimental animals can lead to differences in the plasma concentration of active constituents that result from a given intake amount. Evaluating possible pharmacokinetic differences between experimental animals and humans requires some knowledge of the comparative absorption, distribution, metabolism, and excretion of the test substance in animals and in humans (Klaassen, 2001) and a judgment regarding the degree to which any observed differences in these measures are sufficient to discount animal test findings. The reality is that quantitative information about how these pharmacokinetic variables should appropriately impact the extrapolation of safety information from animals to humans is not available for many substances, especially dietary supplement ingredients. Thus this type of evaluation should be undertaken by experts on a case-by-case basis. When detailed understanding of absorption, distribution, biotransformation, or excretion in experimental animals or humans is not available to make a comparison possible, it is appropriate to assume the most sensitive experimental animal studies are relevant to humans. Linear Versus Nonlinear Dose-Response Assumption Mechanistic or mode-of-action information may be used to improve the risk assessment by providing information about the relationship be- 1   See also discussion in Chapter 8.

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety tween dose and response. The default assumption, in the absence of any mechanistic or mode-of-action information, should be that a threshold or low-dose nonlinear dose-response relationship exists for health effects other than cancer; that is, that for noncancer health effects, there is a dose below which concern is not warranted. The default assumption for cancer is a linear low-dose extrapolation. This principle is important in considering the relevance of the dose of dietary supplements causing harm, as a linear dose extrapolation leads to the conclusion that any amount of the substance is a risk. The rationale for a linear or nonlinear assumption is not without its detractors, but it is an established principle used in risk assessment of other ingested substances (Rodricks et al., 2001) that should be applied to dietary supplements as well. Pharmacodynamic Differences In addition to the pharmacokinetic differences described above, there are several well-described examples of pharmacodynamic differences between animals and humans, that is, differences in how a chemical affects the body (Klaassen, 2001). For example, while rodent carcinogenicity studies are often predictive for human carcinogenesis from chemicals (Huff, 1999; Rodricks et al., 2001; Tomatis, 2001), some target sites in rats and mice have been questioned as relevant endpoints for human risk assessment (Capen et al., 1999; Rodricks et al., 2001). Examples include kidney toxicity/carcinogenicity in male rats related to the production of alpha-2-globulin (Rodricks et al., 2001), liver toxicity/carcinogenicity in rodents related to peroxisome proliferation (Rodricks et al., 2001), thyroid toxicity/carcinogenicity in rats (Capen et al., 1999), and bladder tumors in rats caused by terephthalate acid or cyclamate (IARC, 1999). When these specific endpoints are observed, they raise significant questions regarding relevance to humans. Such findings, or others that suggest irrelevance of the particular animal study evidence to humans, should be used to reach conclusions about possible human toxicity only after careful review. In the absence of specific evidence that certain animal study findings are irrelevant to humans, animal evidence should be used to evaluate potential human risk. Variable Sensitivity of Humans to Adverse Effects When interpreting a substance’s effects or lack of effect in animal studies, it is important to remember the variability among humans in their sensitivity to toxic effects from ingested substances. Some members of the human population are more sensitive than the so-called average (Hayes, 2001), an issue best captured under the concept of “natural variability in response,” a well-documented phenomenon. Many of these differences are

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety due to known genetic polymorphisms (i.e., differences in a gene’s DNA that occur in more than 1 percent of the population) (Hayes, 2001). In general, it can be said that the human population, because of its extremely diverse genetic, environmental, nutritional, and disease status, is far more variable in response to chemicals than are populations of experimental animals. Lack of Adverse Effects in Animals As with any type of scientific study in which an effect is not observed, it is important to remember that a lack of observed or reported detrimental effects in an animal study is not adequate evidence that a particular substance is “safe” to humans.2 The sensitivity of animal experiments in detecting particular effects is of utmost importance when extrapolating from animal studies to humans. Use of animal data to mitigate concerns raised by other data is appropriate only if animal studies are sensitive enough to detect adverse effects if they occur. Sensitivity depends on experimental design factors, such inclusion of positive controls, study power, and whether relevant endpoints were examined in the animals. For example, if an animal study only reported how many animals died or exhibited gross toxicity following short-term administration of an ingredient, it is not acceptable to conclude that this ingredient does not cause cancer following chronic intake by humans. Even if a lack of adverse effects in an appropriate model is reported, it is not scientifically valid to use such information to mitigate other types of information suggesting risk if the study does not have the statistical power necessary, is incompletely reported, does not include positive controls, or is otherwise inadequately designed to detect a risk. In summary, it is only negative data originating from well-designed studies or other credible sources that may mitigate or eliminate a concern raised by other data. Quality Issues While all animal experiments may be informative, the nature of the experimental design, the quality of the methodology, and the statistical significance of the results need to be taken into consideration in weighing the evidence of toxicity. As was mentioned earlier in this chapter, recommendations for well-designed safety tests using animals are described in FDA’s Redbook (OFAS, 2001, 2003) while general characteristics of ideal 2   See also Chapter 10.

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety BOX 5-1 Characteristics of High-Quality Animal Studies A high-quality animal study is one that: uses good laboratory practices; is specifically designed as a toxicity, safety pharmacology, or safety study and includes sufficiently large doses to detect toxicity; uses unanesthetized, unrestrained animals on a semipurified diet; includes adequate controls; uses fully characterized composition and formulation of the test substance; uses a species that has pharmacokinetics similar to humans (bioavailability, distribution, metabolism, excretion); tests multiple doses of test substance; estimates blood or other tissue levels to ensure absorption and adequate exposure to active components to increase the likelihood the response will occur in humans; conducts clinical chemistry, blood, and urine analysis; uses more than a single species; administers the test substance orally; and conducts animal necropsy and histopathology. animal studies are given in Box 5-1. Data from animal studies not meeting these criteria may be valuable as well and should be considered if they suggest a possible risk to human health. In summary, animal evidence should be used to evaluate potential concern for harm to human health unless the evidence indicates that the results are irrelevant. WHEN DO ANIMAL DATA WARRANT CONSIDERATION AS AN INDICATOR OF SERIOUS RISK TO HUMAN HEALTH? Pathophysiological Effects in Animals That Raise the Most Concern Animal toxicity outcomes with clearly definable pathological changes are more compelling with regard to their relevance to humans than are outcomes in which only physiological or biochemical abnormalities are found. Thus, the concerns about possible human toxicity rise in proportion to both the seriousness and the severity of the toxic effects observed in animals and, more closely, those effects suggest the presence of a disease or pathological condition development process. Clearly, animal studies that predict possible serious harm or death warrant more attention than those that predict mild, self-limiting effects on humans. Certain chronic animal toxicity or adverse biological effects data should be considered as immediate cause for higher or moderate concern,

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety TABLE 5-1 Guidelines for Relative Seriousness for Examples of Adverse Effects Obtained from Animal Studies Category A (most serious) Neoplasia (including genotoxic and nongenotoxic carcinogens), teratogenesis, mortality Severe target organ toxicity Necrosis, dysplasia Reproductive failure, fetotoxicity, severe developmental effects Severe neurobehavioral changes Category B (moderately serious) Moderate target organ toxicity Atrophy, hyperplasia Reduced reproductive capacity, moderate developmental effects Moderate neurobehavioral changes Clinical chemistry changes associated with histological lesions outside reference value ranges Category C (less serious) Reduced body weight gain Body weight/organ weight ratios Reduced food consumption Enzyme changes, other biochemical and toxicity biomarker alterations unaccompanied by histological changes Reversible degenerative changes regardless of the presence of high-quality human data suggesting no acute toxicity. This is because human exposure may need to be prolonged before such toxicities would be detected. Table 5-1 contains classifications of toxicity outcomes, ranked according to the nature of the effect, and provides a perspective on which effects are of greater concern. Three broad categories of effects are described. Those in Category A represent the clearest and most serious manifestations of toxicity, and if such effects are observed in well-conducted animal studies, there is a compelling basis for significant concern about comparable human toxicity (ignoring differences that may occur in dose and metabolism). Effects in Category B, while considered adverse, are of lesser concern, and those in Category C are of concern, but the concern is less than the other categories. Depending upon the ingested dosage at which the effect has been observed in animal studies relative to the level of human intake of the substance, and assuming there is no evidence that raises significant doubts about differences in toxic effects between animals and humans, effects in Category A should raise significant concerns about human toxicity even without data from other categories of evidence (e.g., human data). Effects in Category B may need to be but-tressed by other data, and effects in Category C are considered less useful in

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety raising questions of significant human toxicity without significant additional data. In general, it would not be advisable to solely use Category B or C effects to specify the seriousness of the adverse effects expected in humans. The level of concern appropriate for different adverse findings increases when effects have been documented in well-designed and well-conducted animal studies, when the observed effects increase in severity or incidence with increasing dose, and/or when the observed effects are otherwise clearly related to the substance. An ideal study would be appropriately controlled, define the composition of the test material, administer the test material in measured quantities by the oral route, and use standardized and validated methods to measure toxicity accompanied by appropriate statistical analysis, interpretation, and reporting (see Box 5-1). While many studies will not meet this ideal, they will nonetheless provide useful information and should be used if they suggest possible risk to human health.3 The strength of the evidence for toxicity is substantially increased if the effects were observed in more than one animal species, and even more so if supported by additional experimental data (e.g., in vitro data) or human data. RISK ASSESSMENT STRATEGY FOR CONSIDERING ANIMAL DATA Evaluating Risk with Animal Data Under current law, FDA has the burden of providing evidence that one or more uses of a dietary supplement poses some identifiable significant or unreasonable risk to human health. Issues confronting FDA in regulating dietary supplements are not exactly analogous to those that arise in the premarketing approvals of other substances such as food additives, for which protocols for using animal data to establish safety have been developed. The traditional use of animal toxicity data to establish acceptable exposures has imbedded within it an element of caution—animal toxicity findings are used without significant question regarding the predictive power of specific findings for humans, and uncertainty factors are used to ensure safety. Use of protocols for setting safe levels (as opposed to evaluating risk) are outlined in Box 5-2. In assembling evidence regarding risks to health, 3   For example, animal data resulting from non-oral exposure may be available and indicate adverse effects. Concentration of the ingredient (or its active constituents or metabolites) in animal blood that results in adverse effect can be compared with blood levels likely to result from human ingestion, with consideration of additional uncertainty factors as discussed in the text.

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety BOX 5-2 Setting Safe Levels and Tolerances in Foods Over the past half century, a large effort has been devoted to the development and validation of a wide variety of protocols to guide the study of chemical toxicity in animals. At present, such protocols are available to study a broad range of adverse health effects, including the effects of acute, subchronic, and chronic dosing, effects on reproduction and development, and effects on the nervous and immune systems (Klaassen, 2001). The premarketing approval of food additives, both direct and indirect, food-use pesticides, and the evaluation of generally recognized as safe substances all depend heavily upon results from such animal studies. Those who propose to market such substances are required to conduct such studies and to ensure compliance with Good Laboratory Practice Regulations (21 C.F.R. § 58 [1978]). Adverse effects elucidated in animal toxicology studies are used to evaluate the safety of food ingredients and pesticide residues. The following assumptions have long been applied in making those safety evaluations: The most sensitive indicator of adverse effects is selected from the entire body of reported animal data, relying on quality of the data and/or weight of the evidence.a A lowest-observed-adverse-effect level (LOAEL—the minimum toxic dose) and a no-observed-adverse-effect level (NOAEL) for that effect are identified. The NOAEL is divided by a series of uncertainty factors that are designed to accommodate variability in response between animals and humans and among humans (typically, factors of 10 for each). Additional factors may sometimes be introduced to deal with uncertainties in the database or to estimate a NOAEL from a LOAEL if the former is not available from the study. The dosage (or intake) resulting from the above is taken as a safe level of daily intake for the human population; it is assumed to satisfy the “reasonable certainty of no harm” requirements of law. a   In the case of ingredients that are carcinogenic in animals, direct addition to food is prohibited for substances coming within the purview of the Delaney Clause Amendment to the Food, Drug, and Cosmetic Act (and regulated by FDA/Center for Food Safety and Applied Nutrition); for other substances (regulated by Environmental Protection Agency, Consumer Product Safety Commission, FDA/Center for Drug Evaluation and Research), a quantitative estimate of risk is derived from the data and a “safe” level is established at a very low level of risk. the questions of the predictive power of animal studies, the dosages to which humans might be expected to be exposed, and the various types of toxicity observed in animals can become meaningful and significant. The purpose of the section that follows is to offer guidance on issues of risk. Risk is defined as the probability that a substance or situation will produce harm under specified conditions and is a combination of probabil-

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety TABLE 5-2 Relative Spectrum of Concern: Guidelines for Types of Evidence from Animal Studiesa Increasing Concern At least one acceptable, quality study showing effects of Category A at Dose > 1,000× Human Intake At least one acceptable, quality study showing effects of Category A at Dose > 100 to < 1,000× Human Intake At least one acceptable, quality study showing effects of Category A at Dose < 100× Human Intake OR OR OR At least one acceptable, quality study showing effects of Category B at Dose > 100× Human Intake At least one acceptable, quality study showing effects of Category B at Dose > 10 to < 100× Human Intake At least one acceptable, quality study showing effects of Category B at Dose < 10× Human Intake OR OR OR At least one acceptable, quality study showing effects of Category C at Dose > 10× Human Intake OR Studies showing adverse effects, but which cannot be interpreted because of deficiencies in design, conduct, or reporting OR Acceptable, quality non-oral studies indicating adverse effect from Category A, B, or C At least one acceptable, quality study showing effects of Category C at Dose > 1 to < 10× Human Intake At least one acceptable, quality study showing effects of Category C at Dose ≤ 1× Human Intake a Categories A, B, and C refer to relative seriousness of a variety of adverse effects identified in animal studies, ranging from reproductive failure (A) to reduced food consumption (C). See Table 5-1 for further examples. ity and consequences. Risk assessment is an organized process used to describe and estimate the likelihood of adverse health outcomes from exposure to chemicals. The four steps in risk assessment are hazard identification, dose-response assessment, exposure assessment, and risk characterization (NRC, 1996). The risk assessment model proposed in this chapter for using animal data is to consider that the data are a means to integrate information about the seriousness of the observed animal toxicity (Table 5-1) with information about the human dose and the animal dose at which the toxicity occurs. The result is incorporated in the relative spectrum of concern figure for animal data (Table 5-2), providing a practical and gen-

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety eral mechanism to estimate the relevance of animal dose when setting priorities for further evaluation. Rationale for the Risk Assessment Strategy Box 5-2 outlines an approach for risk assessment based on animal data. This approach starts with a dose known to cause a no-observed-adverse-effect level (NOAEL) and/or the lowest dose known to cause any effect (lowest-observed-adverse-effect level, or LOAEL) to derive a dosage that is considered safe. These concepts are generally accepted by the toxicology community to provide some general guidance when determining how much of a substance can safely be consumed. In the case of the postmarketing situation that currently exists for dietary supplements where limited resources may necessitate a focus primarily on serious adverse effects, it is appropriate to apply some of the scientifically accepted concepts to determine which serious adverse effects observed in animals warrant further investigation or suggest an unreasonable risk may exist. In developing safe limits, uncertainty factors have been applied to animal toxicity threshold values (NOAELs) to reach estimates of human dosages that are likely to represent thresholds for the most sensitive members of the human population (see Box 5-2). These uncertainty factors are a scientifically accepted framework for setting priorities when complete data are not available. Uncertainty “default” values of 10 are used for each significant source of variability, such as cross-species differences and interindividual differences between and among humans. A series of studies provides evidence that the factors of 10 are generally adequate to deal with these sources of variability and, in most cases, are more than is necessary (Dourson and Stara, 1983; Dourson et al., 1996; NRC, 1994). The factors of 10 are widely used as default values in the United States and internationally. There is wide recognition that, in specific cases, pharmacokinetic data, if available, provide better estimates of variability. As comparative pharmacokinetic data that allow the development of models for quantitative interspecies extrapolation (physiologically based pharmacokinetic models) become available, they may be used to replace at least a fraction of the interspecies default uncertainty factor. Guidelines for Considering Seriousness of Effect and Dose Using a Risk Assessment Model The following guidelines relate the human intake level of the dietary supplement ingredient under review to the minimum experimental dose required to cause toxicity (LOAEL). For toxic effects that fall into the most serious category, Category A (see Table 5-1), human intakes that exceed

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety one-one hundredth (0.01) of the test animal dose for that effect (and not for any lower dose effects that may fall into Categories B or C) should be considered to represent a significant risk to human health. Such a recommendation, which does not include several of the cautious (public health protective) assumptions that are associated with a safety assessment, nevertheless represents a balance between overinterpretation of animal findings and the need to consider population variability in response. For toxic effects falling into Category B, human intakes that exceed one-tenth (0.10) of the test animal dose for that effect (and not for any lower dose that may fall into Category C) should be considered to represent a significant risk to human health. These are offered as general guidelines, but they should not be interpreted as inflexible rules. The general guidelines offered here should be seen as useful for relatively rapid decision-making to set priorities for utilization of limited resources, with more thorough evaluation of all relevant data necessary to confirm the strength of the evidence. Thus, FDA should evaluate evidence of toxicity and make some general findings regarding the strength of the evidence. For example, for effects falling into Category A and exhibiting clear dose-response relationships, similar findings in multiple species/strains/sexes of animals, supporting in vitro data or information about related substances, and/or evidence from human studies, there can be justification for considering human intakes at levels less than 0.01 of the animal effect dose as representing a potentially serious health risk. While studies of acceptable quality are most useful, combined evidence from other studies may also be useful depending upon the limitations of the studies. Carcinogenicity findings, particularly those that are accompanied by evidence of genotoxicity4 and within the 100× expected human exposure, are of particular concern. Any dietary supplement ingredient having such activity presents the highest degree of potential seriousness (Category A). The guidelines described above for relating the type of effect observed to the dose are summarized in Table 5-2. Situations described in the righthand column of Table 5-2 signal the highest degree of concern for human risks and suggest a significant risk to human health, even in the absence of any human information regarding adverse effects. Situations described in the middle column are of less concern, and ingredients with this level of evidence may not represent a significant health risk unless such risk is confirmed with human or other types of data. Situations described in the left-hand column are of lower concern and thus by themselves present a relatively minor public health concern. 4   Genotoxicity is discussed in Chapter 7.

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety SUMMARY The guiding principle related to animal data is as follows: “Even in the absence of information on adverse events in humans, evidence of harm from animal studies is often indicative of potential harm to humans. This indication assumes greatest importance when the route of exposure is oral, the formulation tested is identical or highly similar to that consumed by humans in an ingredient, and more than one species shows the same or similar toxicity.” The rationale and importance of this principle have been presented, and the following corollaries, along with their rationales and limitations: In the absence of specific evidence that certain animal study findings are irrelevant to humans, animal evidence should be used to evaluate potential human risk. A lack of observed or reported detrimental effects in an animal study is not evidence that a particular substance is “safe.” Veterinary toxicology information may be useful when it corroborates concerns raised by other types of data. Independent of other types of data, evidence of harm in livestock and other veterinary toxicology information is appropriate to consider as a signal prompting an initial review of an ingredient. In addition, the veterinary toxicology literature is also useful for generating hypotheses in need of testing in well-established animal models. When there is no detailed understanding of pharmacokinetics to make a comparison between animals or humans possible, it is appropriate to assume that the most sensitive experimental animal studies are relevant to human health. Much of the animal study data available for dietary supplement ingredients will not have the characteristics of ideal studies, but these studies should nonetheless be considered if they suggest possible human health risk. Animal studies that predict possible serious harm or death warrant more attention than those that predict mild, self-limiting effects in humans. Certain chronic animal toxicity or adverse biological effects data should be considered as immediate cause for higher or moderate concern, regardless of the presence of high-quality human data suggesting no acute toxicity (see Category A in Table 5-1). The default assumption for cancer is a linear low-dose extrapolation. Carcinogenicity findings, particularly those that are accompanied by evidence of genotoxicity and observed in animals at ingested amounts within 100× of expected human exposure, are of particular concern.

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety As a general guideline, human intakes that exceed one-one hundredth (0.01) of animal doses that produce Category A effects (see Table 5-1) should be considered to represent a significant risk to human health. Human intakes that exceed one-tenth (0.1) of the animal doses that produce Category B effects should be considered to represent a significant risk to human health. REFERENCES ASPCA (American Society for the Prevention of Cruelty to Animals). 2003. Animal Poison Control Center. Online. Available at http://www.napcc.aspca.org. Accessed July 8, 2003. Capen CC, Dybing E, Rice JM, Wilbourn JD, eds. 1999. Species Differences in Thyroid, Kidney and Urinary Bladder Carcinogenesis. International Agency for Research on Cancer (IARC) Scientific Publication No. 147. Lyon, France: IARC. CDER (Center for Drug Evaluation and Research). 2002. Guidance Documents. Online. Food and Drug Administration. Available at http://www.fda.gov/cder/guidance/index.htm. Accessed May 2, 2002. Cheeke PR. 1998. Natural Toxicants in Feeds, Forages, and Poisonous Plants. 2nd ed. Danville, IL: Interstate Publishers. Colegate SM, Dorling PR, eds. 1994. Plant-Associated Toxins: Agricultural, Phytochemical and Ecological Aspects. Wallingford, UK: CAB International. Cornell University. 2002. Poisonous Plants Informational Database. Online. Available at http://www.ansci.cornell.edu/plants/index.html. Accessed July 8, 2003. Dourson ML, Stara JF. 1983. Regulatory history and experimental support of uncertainty (safety) factors. Regul Toxicol Pharmacol 3:224–238. Dourson ML, Felter SP, Robinson D. 1996. Evolution of science-based uncertainty factors in noncancer risk assessment. Regul Toxicol Pharmacol 24:108–120. Everist SL. 1981. Poisonous Plants of Australia. London: Angus & Robertson Publishers. Garland T, Barr AC, eds. 1998. Toxic Plants and Other Natural Toxicants. Wallingford, UK: CAB International. Hayes W. 2001. Principles and Methods of Toxicology. 4th ed. Philadelphia: Taylor and Francis. Huff J. 1999. Long-term chemical carcinogenesis bioassays predict human cancer hazards. Issues, controversies, and uncertainties. Ann NY Acad Sci 895:56–79. IARC (International Agency for Research on Cancer). 1999. Cyclamates. IARC Monogr Eval Carcinog Risks Hum 73:195–222. ICH/FDA (International Conference on Harmonisation/Food and Drug Administration). 2001. Guidance for Industry: S7A Safety Pharmacology Studies for Human Pharmaceuticals. Rockville, MD: FDA. James LF, Keeler RF, Cheeke PR, Bailey EM Jr, Hegarty MP, eds. 1992. Poisonous Plants. Ames, IA: Iowa State University Press. Keeler RF, Tu AT, eds. 1983. Handbook of Natural Toxins. Volume 1. Plant and Fungal Toxins. New York: Marcel Dekker. Keeler RF, Tu AT. 1991. Handbook of Natural Toxins, Volume 6. Toxicology of Plant and Fungal Compounds. New York: Marcel Dekker. Keeler RF, Van Kampen KR, James LF, eds. 1978. Effects of Poisonous Plants on Livestock. New York: Academic Press. Kellerman TS, Coetzer JAW, Naudé TW. 1988. Plant Poisonings and Mycotoxicoses of Livestock in Southern Africa. Cape Town, South Africa: Oxford University Press.

OCR for page 156
Dietary Supplements: A Framework for Evaluating Safety Kingsbury JM. 1964. Poisonous Plants of the United States and Canada. Englewood Cliffs, NJ: Prentice-Hall. Klaassen CD, ed. 2001. Casarett and Doull’s Toxicology: The Basic Science of Poisons. New York: McGraw-Hill. Mishra HN, Das C. 2003. A review on biological control and metabolism of Aflatoxin. Crit Rev Food Sci Nutr 43:245–264. NRC (National Research Council). 1994. Science and Judgment in Risk Assessment. Washington, DC: National Academy Press. NRC. 1996. Understanding Risk: Informing Decisions in a Democratic Society. Washington, DC: National Academy Press. NRC. 2001. Evaluating Chemical and Other Agent Exposures for Reproductive and Developmental Toxicity. Washington, DC: National Academy Press. NTP (National Toxicology Program). 2002. Report on Carcinogens. 10th ed. Research Triangle Park, NC: NTP. OFAS (Office of Food Additive Safety). 2001. Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food. Redbook II-Draft. Washington, DC: OFAS, Food and Drug Administration. OFAS. 2003. Redbook 2000. Toxicological Principles for the Safety of Food Ingredients. Online. CFSAN Food and Drug Administration. Available at http://www.cfsan.fda.gov/~redbook/red-toca.html. Accessed June 19, 2003. Rodricks JV, Gaylor DW, Turnbull D. 2001. Quantitative extrapolations in toxicology. In: Hayes AW, ed. Principles and Methods of Toxicology. 4th ed. Philadelphia: Taylor and Francis. Seawright AE, Hegarty MP, Keeler RF, James LF, eds. 1985. Plant Toxicology. Brisbane: Queensland Poisonous Plants Committee. Tomatis L. 2001. The identification of human carcinogens and primary prevention of cancer. Mutat Res 462:407–421. USDA/ARS (United States Department of Agriculture/Agricultural Research Service). 2003. Poisonous Plant Research Laboratory (PPRL). Online. Available at http://www.pprl.usu.edu. Accessed July 8, 2003. WHO (World Health Organization). 1999. Principles for the Assessment of Risk to Human Health from Exposure to Chemicals. Geneva: WHO.