4
Factors Considered in Screening, Setting Priorities, and Safety Evaluation

In any scientific evaluation there are different categories of data that are useful and that could be termed as key factors to consider. It is helpful to collect and sort relevant information according to these categories. This chapter describes the different types of scientific evidence and other information, herein termed “factors,” thought to be most useful when screening, setting priorities, and conducting a critical safety evaluation of a dietary supplement ingredient. Different factors contribute to each step of the framework process to a different degree, and different sources of information are necessary to examine and evaluate the factors in the various steps of the process.

DESCRIPTION AND USE OF KEY FACTORS

This chapter includes a description of each factor, limitations when considering the different types of information grouped under the factor, and suggestions for how each factor is used in each step of the process. In addition, the different sources of information for each factor are outlined in Table 4–1. These sources of information may change over time and new sources may be added. It is likely that, with use, the systematic approach described in this report will eventually evolve into an increasingly efficient and effective system as experience and an accumulating database inform and organize the process.

Of the key factors described below (human data, animal data, biological activity of related substances, and in vitro data), the primary factor that contributes to decision making at all steps of the framework is the evidence of harm in humans. This is provided by data collected in observational studies and clinical trials, spontaneously reported adverse events, and other sources of information about the consequences of use in humans. Whether or not the ingredient is new, and thus safety cannot be ascertained as readily, is also considered in the screening/flagging step. This is classified as the “new ingredient status” question.

In the descriptions below, each factor is defined, a rationale for its use provided, and limitations in the use of these types of data are described. A general description of how each factor can be used at each step of the process is then outlined, including a description of the appropriate information sources to consider at each step in the process—ranging from easily obtainable information for the screening/flagging step to an increasingly comprehensive information collection for the priority-setting and critical safety evaluation steps. Finally,



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 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements 4 Factors Considered in Screening, Setting Priorities, and Safety Evaluation In any scientific evaluation there are different categories of data that are useful and that could be termed as key factors to consider. It is helpful to collect and sort relevant information according to these categories. This chapter describes the different types of scientific evidence and other information, herein termed “factors,” thought to be most useful when screening, setting priorities, and conducting a critical safety evaluation of a dietary supplement ingredient. Different factors contribute to each step of the framework process to a different degree, and different sources of information are necessary to examine and evaluate the factors in the various steps of the process. DESCRIPTION AND USE OF KEY FACTORS This chapter includes a description of each factor, limitations when considering the different types of information grouped under the factor, and suggestions for how each factor is used in each step of the process. In addition, the different sources of information for each factor are outlined in Table 4–1. These sources of information may change over time and new sources may be added. It is likely that, with use, the systematic approach described in this report will eventually evolve into an increasingly efficient and effective system as experience and an accumulating database inform and organize the process. Of the key factors described below (human data, animal data, biological activity of related substances, and in vitro data), the primary factor that contributes to decision making at all steps of the framework is the evidence of harm in humans. This is provided by data collected in observational studies and clinical trials, spontaneously reported adverse events, and other sources of information about the consequences of use in humans. Whether or not the ingredient is new, and thus safety cannot be ascertained as readily, is also considered in the screening/flagging step. This is classified as the “new ingredient status” question. In the descriptions below, each factor is defined, a rationale for its use provided, and limitations in the use of these types of data are described. A general description of how each factor can be used at each step of the process is then outlined, including a description of the appropriate information sources to consider at each step in the process—ranging from easily obtainable information for the screening/flagging step to an increasingly comprehensive information collection for the priority-setting and critical safety evaluation steps. Finally,

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements TABLE 4–1 Sources of Information for Key Factors and Modifiersa Key Factors and Modifiers Screening/Flagging Priority Setting Monograph/Critical Evaluation Key factors Human data Serious adverse events: MedWatch Poison Control Center Cursory search in scientific and medical literature including IBIDS, Medline, Toxline Letters to the Center for Food Safety and Applied Nutrition Sources listed at left Secondary reviews Sources listed at left All available sources, including: Published case reports—available through MedLine or other literature Unpublished safety information requested from published clinical studies Unpublished safety information requested from manufacturers Prepublication safety information requested from clinical trials Discovery materials from tort litigation Animal data Consider under “Other Concerns” Literature searches (e.g., IBIDS, MedLine, Toxline, Embase) Database searches (e.g., NAPRALERT, Poisindex, Naurac) Secondary reviewsb Sources listed at left Data voluntarily provided by industry Data provided by animal poison control centersc Biological activity of structurally related or taxonomically related substances Consider under “Other Concerns” Poisonous plants (Kingsbury, 1964) NAPRALERT Sources listed at left Data voluntarily provided by industry In vitro data Consider under “Other Concerns” Literature searches (e.g., IBIDS, MedLine, Toxline, Embase) Database searches (e.g., NAPRALERT, Poisindex, Naurac) Secondary reviews Sources listed at left Data voluntarily provided by industry

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements Modifiers Prevalence of use Industry estimates of production and sales (e.g., U.S. Consumer [2000]) Surveys describing supplement use Large-scale, cross-sectional data collection (e.g., the National Health and Nutrition Examination Survey, the Centers for Disease Control and Prevention, the U.S. Department of Agriculture, the Continuing Survey of Food Intakes by Individuals, the Food and Drug Administration) Sources listed at left Not applicable Vulnerable groups Same sources as key factors (human data, animal data, structure/chemotaxonomy, in vitro) Same sources as key factors Same sources as key factors New ingredient status 75-day advance notifications Not applicable Not applicable aThese information sources are likely to change over time. b Secondary reviews include Commission E monographs, Agency for Healthcare Research and Quality evidence-based reports, American Herbal Products Association monographs, Natural Medicine Comprehensive Database, World Health Organization monographs, Dietary Reference Intakes by the Institute of Medicine, and National Toxicology Program reviews. c The Animal Poison Control Center of the American Society for the Prevention of Cruelty to Animals (ASPCA) will provide database information on its cases if requested by the Food and Drug Administration (Personal communication, S. Hanson, ASPCA Animal Poison Control Center, May 17, 2002).

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements overarching guiding principles are presented to explain the scientific basis for the suggested use of each factor throughout the framework. These guiding principles will again be emphasized in the critical safety evaluation process described in Chapter 6. Key Factor: Data from Humans and Clinical Evidence of Harm To identify possible concerns regarding safety, it is essential to examine data and information on undesired effects that may have occurred in humans. These undesired effects are referred to as adverse events, a term that does not imply that a particular substance caused the event, but simply indicates that the untoward effects observed were associated with its use and might be related to the ingested substance (ICH, 1996). Information about the occurrence of adverse events in humans is obviously the data most relevant to the supplement ingredient in other humans. Information about adverse event occurrence in humans has its own limitations, however, and must therefore be interpreted carefully. This section describes different sources of information about data on adverse effects of dietary supplement ingredients in humans and considerations for using the different types of available information. When human data suggest risk, these are the most relevant data for consideration of human safety. However, it is anticipated that applicable human data from experimental studies, observational studies, spontaneous reports, and historical use sources often will not be available. As discussed below, the limitations in using available human data often lead to its value only as a signal generator, but even weak data may be useful in this capacity. Information about untoward effects associated with the use of supplement ingredients may come from experimental studies designed to examine the efficacy or safety of a substance, epidemiological studies, case reports or series, spontaneous adverse event reports to the Food and Drug Administration (FDA) or to poison control centers, or anecdotal reports in the history of the substance’s use. Each of these sources provides a different type of potentially valuable information. Clinical Trials It is helpful to first consider the ideal source of data and then consider limitations of other sources of data. If available, the “gold standard” for determining the safety of an ingested substance is considered by many to be the randomized, controlled clinical trial (RCT) that is designed to assess safety as well as efficacy. The ideal RCT would enroll a sufficient number of subjects who are systematically monitored for a sufficient amount of time to detect a wide array of adverse effects or physiological changes that might warrant concern. It is the usual practice in an RCT to query subjects for possible adverse events at defined intervals and to record and evaluate these events as “definitely,” “probably,” “possibly,” or “not” related to the ingested substance (ICH, 1995). The use of randomization and control groups enables scientists to determine the likelihood that adverse effects are actually due to the substance rather than to confounding factors. Limits to the generalizability of the study include the statistical power of the study to detect adverse events, differences between the study and target populations, and differences between how a substance is administered during the RCT and its actual use by the general population. Most RCTs are designed to assess beneficial effects. Thus, in general, efficacy results are more reliably reported than safety data (Ioannidis and Lau, 2001). Although those conducting efficacy trials are expected to observe and report adverse reactions, the extent and detail of this

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements reporting is highly variable (Ioannidis and Lau, 2001). In some cases, however, investigators may be able to supply unpublished data useful in the safety evaluation, even if the published results do not contain all the available information about adverse events (Ioannidis et al., 2002). While investigators may be able to provide unpublished additional data, characteristics of the study design itself may limit usefulness in predicting safety because even large studies may lack sufficient statistical power to detect adverse events of low incidence. Adverse events generally occur at rates much lower than desired effects (FDA, 1995). Clinical trials generally are designed to detect one primary endpoint, thus secondary events, such as adverse effects, will typically be inadequately reported (Ioannidis and Lau, 2001). A major cause of an incomplete safety evaluation is that the unexpected adverse events may not be noticed by the subject or detected by the investigator if they fall outside the investigator protocol. For these reasons, a study to test the effects of a supplement ingredient on mood, for example, may not detect potentially dangerous cardiovascular effects if heart function is not monitored. Even if investigators are alert for adverse effects, the limited number of subjects, the limited duration, and the unrepresentative nature of populations studied limit the sensitivity in detecting adverse events that would occur infrequently, after extended exposure, or in subpopulations. For example, events that occur at the rate of 1 in 1,000 would require a study with at least 3,000 subjects at risk to have a 95 percent chance of being detected (Lewis, 1981). Although RCTs can be limited in their sensitivity, they do provide valuable information when adverse events are detected. Information from clinical studies is strengthened by the following information (Counsell, 1997; ICH, 1995; Moher et al., 2001): demographic information on the study population; inclusion and exclusion criteria to determine if the results are generalizable; description of the condition or disease and comorbidities of the study population; description of the intervention (supplement ingredient [composition], dose, and duration of exposure); list of prior and concomitant ingested substances, including dietary supplements and drugs; and description of the adverse event including temporal relationship to ingestion of supplement ingredient (response to discontinuation or rechallenge). Observational Epidemiological Research As discussed above, a limitation inherent to many RCTs is that size and duration limit sensitivity to detect adverse events (FDA, 1995). Latent or delayed effects that occur long after exposure may not be detected. Information about these latent and infrequent effects often comes from observational or epidemiological studies that retrospectively or prospectively examine the effects of ingested substances on large populations. Like RCTs, the value of observational studies also depends on the endpoints examined. For example, if a study evaluates the incidence of cancer, death, or liver damage but does not evaluate anemia, the study is unlikely to detect interference with iron absorption. For the endpoints examined, cohort studies using registries and other sources of information about large populations are a valuable source of safety information about ingested substances such as pharmaceutical drugs. These types of studies would likely be informative about the safety of particular dietary supplement ingredients, but unfortunately there are few studies of

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements this type conducted on them. The type of information needed to conduct such studies is rarely available for supplement ingredients because their use is not systematically tracked in a manner similar to use of prescription drugs. Non-Study Information: Spontaneously Reported Adverse Events Adverse events that are spontaneously reported to FDA, poison control centers, or as case reports or case series in the medical literature are also important sources of information. Generally, these voluntary spontaneous reports are made by consumers, physicians, or pharmacists who notice an untoward effect following ingestion of a substance. It is assumed that adverse effects most likely to be attributed to the ingested substance are unusual, persistent, or severe, and occur shortly following ingestion. Thus, effects that are not noticeable enough to garner attention are unlikely to be associated with the ingestion of the ingredient and thus unlikely to be reported to FDA or another data-collecting entity. Given these limitations, it is not surprising that the total rate of spontaneous adverse-event reporting is very low (OIG, 2001), and that even fewer reports are made to FDA (Chyka and McCommon, 2000). It has been estimated that adverse events spontaneously reported to FDA account for only 1 percent of serious drug reactions that occur outside the parameters of clinical studies (Scott et al., 1987). It is unknown if spontaneous adverse event reporting may be even less frequent with dietary supplements, because it is unknown if consumers are less likely to associate dietary supplements with untoward effects than to associate drugs with untoward effects.Unlike drugs, supplement manufacturers and distributors are not required to share with FDA the adverse event reports they receive (CFSAN, 2001a; OIG, 2001). Nonetheless, MedWatch and other sources of reported serious adverse effects will often be the first line of evidence that indicates a substance might warrant a higher priority review. Even when reports are inadequately documented and causation difficult to assess, the reports should serve as sentinel events that alert regulators and the medical community to potential adverse effects of a product. In summary, when they exist, spontaneous reports of adverse events and published case reports are useful for generating hypotheses about relationships between supplement ingredients and untoward effects. However, due to the nature of adverse event reporting, especially for dietary supplement ingredients, a lack of reports does not imply that a dietary supplement ingredient is safe. Similarly, the existence of adverse event reports does not, without extensive critical evaluation of the reports, establish a causal relationship between the adverse event and the ingredient. Non-Study Information: Historical Use Experience from generations of use in humans is often referred to as evidence of safety for modern day supplements that bear resemblance to substances used historically. Some botanicals, for example, have had a long history of medicinal use in many cultures. Historical use is of less importance when relevant clinical, epidemiological, or animal toxicity data exist. For many supplements, however, the amount of scientific and experimental data that exists ranges from scant to nil. Recognizing that a full range of data is unlikely to be available for many dietary supplement ingredients, historical use may be taken into account as a limited surrogate measure for toxicity in the absence of relevant scientific and experimental data. In doing so, it is important to consider the relevance of the traditional use to the current use. Historical information is only useful if the product in question is not so far removed from

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements BOX 4–1 Questions to be Answered When Considering Relevance of Information About Historical Use Is the supplement ingredient one that was commonly used within the context of a traditional medical system? If the supplement ingredient is a botanical, is the part of the plant marketed the same as the part that was traditionally used? Is the preparation a crude preparation, extract, or concentrate; a selected fraction; an isolated compound; or a mixture of these? How similar is the current preparation to that used traditionally? Are current intake levels or recommended intake levels clearly different from traditional use? Is the modern duration of use consistent with historical use? Is the modern indication consistent with historical use? If there are traditional cautions in the use of the supplement ingredient, are these cautions typically heeded? Are there other reasons to expect a different toxicity profile for the modern formulation than for the traditional preparations? the original plant use as to constitute a distinct entity. For example, a whole root extract that was traditionally used for three days to treat a cold is not comparable to a fraction of a leaf extract promoted for long-term use to treat cancer. The discussion in this section focuses on questions to help assess how to consider the relevance of information about traditional use. These questions are listed in Box 4-1 and are explained in more detail below. If the supplement ingredient is a botanical, is the part of the plant marketed the same as the part that was traditionally used? Safety comparisons for botanicals can only be made when the same plant part used in traditional preparations is used in the modern preparation. Seeds, roots, leaves, and other parts may have distinct safety profiles due to different composition. For this reason, this report defines the specific plant part as the ingredient under question. Indication of safe use of one plant part should not be used as prima facie evidence that other plant parts might also be used safely. Is the preparation a crude preparation, extract, or concentrate; a selected fraction; an isolated compound; or a mixture of these? The method of preparation can have an impact on an ingredient’s safety. This is most clearly illustrated in botanicals with traditional medicinal uses. Traditionally, most orally ingested medicinal herbs were administered as crude aqueous extractions of plant parts that were soaked, steeped, or boiled in water. Today’s supplements are often sold in a different form—as encapsulated dried herbs, fluid extracts, solid extracts (such as capsules or tablets), or foodstuffs containing herbal extracts. While these modem formulations are not equivalent to traditional preparations and may not have exactly the same effect as teas and infusions, they could have safety profiles similar to traditional preparations. A botanical with a history of benign use in infusions may or may not manifest new toxic effects when concentrated, lyophilized, or encapsulated. Teas (infusions) are typically extracts prepared from dried plant materials, while lyophilized plants are made from whole fresh materials. The chemical composition and concentrations could be sufficiently different between the two forms to result in different safety profiles. Even if dried and lyophilized materials were identical in all respects, an infusion of a botanical should not be thought of as necessarily comparable to whole

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements dried botanicals because the chemistry of the extract (tea) may be different. Differences in safety profiles could also be expected for alcoholic extracts of plants with known toxic components. Alcohol draws out different compounds, so alcohol extracts may contain a higher concentration of toxic compounds than aqueous extracts. An example is wormwood (Artemisia absinthium), which in an aqueous extract contains little thujone (a neurotoxin) (Tegtmeier and Harnischfeger, 1994), but may contain substantial amounts of thujone in alcohol extracts. Additionally, isolated compounds may be dissimilar to traditional plant extracts in safety, as could extracts to which isolated compounds (e.g., yohimbine, ephedrine, hypericin) have been added. Are current intake levels or recommended intake levels clearly different from traditional use? A frequently quoted axiom of toxicology from Paracelcus is that “dose makes the poison.” Unfortunately, differences in traditional and modern formulations render dose comparisons difficult or even impossible. In the rare cases where active compounds or groups of compounds are known and have been quantified (e.g., kavalactones in kava [Piper methysticum], ephedrine alkaloids in Ephedra sinica), doses can be compared. In most cases, however, dosing comparisons are so imprecise that it should probably only be attempted in cases where the modern formulation is clearly providing doses that are orders of magnitude higher than traditional doses. For example, consumption of a culinary herb in small amounts in cooked food may have different effects than medicinal consumption of large amounts of the same herb, rendering a safety extrapolation from culinary to supplemental use inappropriate. Is the modern duration of use consistent with historical use? Is the modern indication consistent with historical use? The duration of use is another component of dosage that should be considered. Acute, short-term, and long-term intakes all have different safety implications. A lack of adverse events reported for an herb traditionally used only for a few days has little relevance to safety of the same herb chronically ingested. When considering how the current duration of use compares to traditional duration of use, it may be helpful to also consider whether the modern day indication is consistent with traditional indications. The modern uses of some botanicals, especially for nonmedical indications such as memory enhancement and ergogenics, for example, might lead consumers to use supplements chronically that were never used chronically in traditional medicine. If there are traditional cautions in the use of the supplement ingredient, are these cautions typically heeded? Some dietary supplement ingredients, such as some botanicals, were traditionally prescribed by practitioners knowledgeable about contraindications to their use. It is scientifically appropriate to take contraindications in traditional use into account when considering the safety of the ingredient. If, for example, an ingredient traditionally contraindicated for pregnant women is currently being marketed to pregnant women or frequently consumed by pregnant women due to its expected effects, then FDA should be more concerned about the safety of this ingredient. In summary, it is clear from these questions that historical use, even widespread historical use, is no guarantor of long-term safety. Historical use information is very useful when it describes a relationship between untoward effects and an ingested substance. It is less useful in predicting harmful effects, especially those effects that do not occur immediately following exposure. However, in the absence of scientific or experimental data, historical use may provide indirect evidence for lack of serious acute harmful effects, and it may be useful to compare with current cautions and exposures (intake levels and duration). Because little other data may be

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements available for many ingredients, it is important to judge the relevance of traditional use information to current use conditions. Causation in the Consideration of Adverse Events For adverse events from any type of study or nonstudy source, ascertaining whether or not a causal relationship exists between the adverse events and the ingestion of the ingredient is likely to be a challenging aspect of considering human data. At the screening/flagging step of the process, ingredients should be flagged without a burdensome evaluation of actual causation. At the priority-setting step to a limited degree, and in the critical evaluation step to a much greater degree, the evidence should be evaluated for causation. Generally accepted causation criteria for assessing the relationship between adverse events and drugs are outlined by Sackett and colleagues (1991). These criteria, listed below, should generally be applicable to other ingested substances as well, including dietary supplement ingredients. The adverse effect is well accepted as an adverse reaction. There is no good alternative candidate (unexplained exacerbation or recurrence of underlying illness). The timing is as expected for an adverse reaction to this compound. The blood level or other biomarker provides unequivocal evidence of overdose. The adverse effect improves suitably if the individual is not rechallenged with the compound. The adverse effect unequivocally recurs or is exacerbated on rechallenge. These causation criteria and the quality and documentation of the data will be helpful in weighting the information collected about adverse events reported, but it is not necessary for causation to be clearly demonstrated or refuted during any step of the process. The different types of human data discussed in the sections above are considered in all steps of the framework, but the degree to which the data are evaluated varies significantly with each step. The differences are summarized below and discussed further in Chapters 5 and 6. Use of Human Data in the Screening/Flagging and Priority-Setting Steps There are several primary differences in how human data are considered at each step. The first difference is that in the screening/flagging step, the occurrence of serious adverse events, rather than all adverse events, is used to flag ingredients that should be considered in the priority-setting step. This distinction is made because including nonserious adverse events in the screening could serve to dilute the efforts with untoward consequences that are nuisances or inconveniences (flatulence or halitosis, for example), but do not cause morbidity or mortality. Serious adverse events are defined in the Guideline for Good Clinical Practice issued by the International Committee on Harmonization (ICH, 1996) and endorsed by FDA (FDA, 2002). A serious adverse event is an untoward effect that is a death, life-threatening event, initial or prolonged hospitalization, disability, congenital anomaly, birth defect, or other important medical event (ICH, 1996). In contrast to the screening/flagging step, serious and nonserious adverse events are considered in the priority-setting and critical safety evaluation steps. The second difference in how human data are considered at the screening/flagging and priority-setting steps is the degree to which causation is considered. As described above, the

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements term adverse event encompasses all untoward effects that may be associated with the supplement ingredient, even if the ingredient has not been demonstrated to actually cause the effect. At the screening/flagging step, the relationship of the adverse event and the ingredient (i.e., the causality) is considered to some degree, but it is considered less at this step than in the priority-setting and critical safety evaluation steps when more information and resources are available to examine causality. That is, ingredients are flagged in the screening/flagging step if a causative relationship between the ingredient and the adverse event cannot easily be ruled out. The level of information gathering at each step also varies. For the screening/flagging step, readily or easily obtainable sources of information about serious adverse events should be explored. Data sources could include MedWatch and the Poison Control Center database that collect adverse event reports from consumers, pharmacists, physicians, and hospitals. A cursory search of the medical literature should also be conducted to determine if there are any reports of serious adverse events associated with supplement ingredients.10 At the priority-setting step, FDA should invest additional effort into weighing the strength of the evidence that a relationship exists between the supplement ingredient and the adverse event. Although the consideration of causation at this stage is not comprehensive, the reviewer makes some judgments about the strength of the evidence. Historical use information may also be used in the screening/flagging and priority-setting steps. The screening/flagging step is focused on responding to indicators of concern, rather than on considering information that may suggest an ingredient is safe. To the degree that historical use information provides insight into possible concerns or subpopulations that may be harmed by the ingredient, it would be considered in these two steps of the process. As stated above, historical use information can be considered as a surrogate indicator that acute serious toxic effects are unlikely when other, more relevant safety information is not available. During the priority-setting step, it therefore may be appropriate for the reviewer to consider relevant historical use information to determine if it provides insight into areas of concern; such consideration may influence the priority score (see Chapter 5 for the proposed scoring system). Use of Human Data in the Critical Safety Evaluation Step The purpose of the critical safety evaluation step is to consider all of the available and relevant information about possible effects of the ingredient on humans, and to consider these data in the context of other types of data collected (e.g., animal or in vitro). Available information should be collected, which includes published case reports available through the National Library of Medicine databases and other sources. Also, clinical investigators may have adverse event information that was not published. This information should be solicited, as well as adverse event information from federal agency-sponsored studies in progress and materials discovered by plaintiff lawyers in tort litigation. Importantly, the manufacturers and distributors of ingredients that reach the critical safety evaluation stage should also be asked to provide data voluntarily on adverse events reported to them or other relevant evidence they have regarding safety evaluation.11 10   For example, the National Library of Medicine’s PubMed, available at http://www.ncbi.nlm.nih.gov/entrez/query.fcgi; TOXNET, available at http://toxnet.nlm.nih.gov/; and EMBASE, available at http://www.embase.com/. 11   Manufacturers and distributors are always welcome to submit adverse event reports and other safety information. They are specifically requested to submit data after the priority-setting step because at this step additional effort can be expended to request information about specific ingredients in the Federal Register and/or via letters to individual manufacturers and distributors, if they are known.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements In the critical safety evaluation step, experts will weigh the evidence and consider the likelihood that a substance poses a risk to human health. At this point, consideration of human adverse event data in relation to the other factors, such as animal, in vitro, and biological activity of related substances, will provide an overall picture of what is known about the ingredient’s safety and may provide plausible biological explanations for reported human adverse events for which causation is not clear. While it is hoped that eventually all dietary supplement ingredients will reach the critical safety evaluation step, given the large number of dietary supplement ingredients, it is unlikely that those ingredients for which significant safety concerns have not been raised will reach this step in the near future. Thus, ingredients that have been safely used historically are unlikely to be reviewed unless other information leads to questions about their safety. Information about the historical use of an ingredient will be the most useful during the critical safety evaluation step, when it can be considered along with the in-depth analysis of potential for harm derived from other information. In this step, historical use information will be considered to the degree that the historical use is similar to current use. The historical use information should not be considered as more important than the scientific evidence, but it may be appropriate to take information about the history of use into account if it is relevant to understanding the likelihood of the potential harm being considered and it is relevant to current use conditions. For example, it would be important to determine whether the potential harm being considered would be expected to have been detected during years of previous use. In such cases, historical use information may mitigate concerns to some degree. Guiding Principle for Human Data A credible report of a serious adverse event in humans that is associated with use of a dietary supplement ingredient raises concern about the ingredient’s safety and requires further information gathering and evaluation. A final judgment about the safety of the supplement ingredient, however, will require consideration of the totality of the evidence. Historical use should not be used as prima facie evidence that the ingredient does not cause harm. It may be appropriate, however, to give considerable weight to a lack of adverse events in large, high-quality, randomized clinical trials or retrospective or prospective cohort studies that are adequately powered and designed to detect adverse effects. Key Factor: Animal Data Animal testing provides invaluable evidence about the potential for ingested substances to cause harm in humans. Thus, studies on animals are regularly employed as an important step in attempting to predict untoward effects of these substances on humans (see, for example, Redbook 2000 [CFSAN, 200 1b] or guidance documents for new drugs [CDER, 2002). They are powerful because controlled studies can be conducted to predict effects that might not be detectable with customary use by humans until they lead to harmful results. Animal studies serve as important signal generators and, in some cases, may stand alone as indicators of unreasonable risk.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements Key Factor: In Vitro Data In vitro studies are defined here as studies not conducted in humans or other whole animals. A wide range of in vitro experimental systems are used to gain insight into the risk of adverse clinical effects of compounds. These systems include isolated cells, microorganisms, subcellular components, and isolated organs. In vitro assays often focus on measuring effects on cells and subcellular targets such as enzymes, receptors, and DNA. The primary advantage of conducting in vitro studies is that their reductionist approach allows insight into a compound’s mechanisms of action that might be more difficult to obtain in a whole animal study, making in vitro studies useful screening tools. They are also generally more rapid and less expensive, leading to a greater amount of this type of data being available in the literature. As mentioned above, it is the reductionist approach of in vitro studies that makes them powerful and inexpensive assays useful for learning about effects and mechanisms of actions of compounds. The reductionist approach of in vitro assays, however, requires that reviewers of these studies carefully consider their limitations and caveats. It is very important, for example, to consider whether the compound applied to the in vitro system is similar in identity and concentration to the compound that reaches the target (e.g., tissue, receptor, subcellular component) in the human. After a substance is ingested, the metabolic fate of the compound and the amount of the biologically active compound that actually reaches the target site is dependent on a multitude of processes including absorption, distribution within the body, metabolism by liver and intestinal enzymes, and rate of excretion. Knowledge of an ingredient’s pharmacokinetics and in vivo metabolism will allow the most appropriate interpretation of the relevancy of the dose used in the in vitro tests. Botanical extracts provide an example of how important it is to consider bioavailability of the ingested substances. When applied to cells in vitro, these extracts often contain polyphenolic compounds (e.g., tannins and related compounds) that may reversibly or irreversibly bind to subcellular components such as enzymes, signal transduction factors, and receptors where they cause effects. In a human, however, these compounds can bind to the food bolus or be metabolized by gastrointestinal enzymes, becoming unavailable for absorption, and therefore not exert the same effects on receptors and enzymes (Bravo, 1998; Yang et al., 2001). Another example of problematic interpretation can be in hepatocyte cultures that do not always support expression of metabolizing enzymes, causing some data to be misleading. In contrast, some cell cultures are established specifically to evaluate metabolism of substances and can provide useful information. All cell types do not respond similarly to a single substance, even when the cells originate from the same organ; one cell type may exclude or excrete a compound whereas another cell will not, and another may behave differently due to its unique biochemical pathways. In the drug development world, results from some in vitro assays are considered predictive enough of toxicological problems that the assays are used to screen compounds in development and influence decisions about further development. For example, assays have been developed to identify compounds that may contribute to the development of torsades depointes cardiac arrhythmia by slowing cardiac repolarization. Drugs that may potentially contribute to this condition can be identified by in vitro experiments conducted with isolated organs and measurement of potassium channel activity in isolated cells (Liu et al., 1998; Wang et al., 1998; Zabel and Franz, 2000). It is also possible to use in vitro assays to anticipate which dietary supplement ingredients may contribute to supplement-drug interactions by studying the effects of a dietary supplement

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements BOX 4–4 Useful In Vitro Assays apoptosis induction ATP synthesis inhibition cell cycle effects cell proliferation effects cell transformation effects cholinergic effects cholinesterase inhibition or induction cytolytic effects cytotoxic effects detoxification enzyme inhibition or induction DNA damage Epstein-Barr virus activation histaminergic effects hormone receptor binding studies immunosuppressant effects mitochondrial respiration inhibition mitogenic effects parasympatholytic and parasympathomimetic effects pharmacokinetic alterations phototoxicity effects prooxidation effects sympatholytic/sympathomimetic effects ingredient on cytochrome P450 enzymes, which are important in the liver metabolism of drugs and supplements (Budzinski et al., 2000; Obach, 2000; Piscitelli et al., 2000). In vitro assays that assess enzymes, receptors, tissues, or other biological endpoints that might provide useful information in the context of other data are listed in Box 4-4. As additional in vitro assays are developed and validated, they will be useful in identifying which dietary supplement ingredient may potentially be associated with risk. Because of the difficulties that often exist in their interpretation, it is often appropriate to use in vitro data as hypothesis generators or potential indicators of harmful health effects rather than as stand-alone demonstrated indicators that in themselves suggest possible risk. However, some in vitro assays, when carefully conducted and interpreted, provide valuable information beyond simply reinforcing observations from other systems or generating hypotheses. When the relationship between the results of an in vitro assay and actual clinical or animal outcomes has been demonstrated, thus validating the predictive value of the assay, then the in vitro assay warrants careful attention. Use of In Vitro Data in the Screening/Flagging Step In vitro data are not a separate, independent factor of the initial screening/flagging step. It is expected that in vitro data that raise concerns about the safety of a supplement ingredient will come to FDA’s attention through secondary reviews or other outlets for expression of public concern. Use of In Vitro Data in the Priority Setting Process Information about in vitro effects can be obtained from many of the same sources used to locate animal data. These sources include literature databases and secondary and tertiary reviews. As with the other factors, at this step it is important to consider what the in vitro data suggest about the risk and seriousness of possible harm.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements Use of In Vitro Data in the Critical Evaluation Step The same information sources used for the priority-setting step should be searched in greater depth for the critical safety evaluation. In addition, industry and other public communities should specifically be queried for any in vitro data on the safety of the ingredient being evaluated. As discussed above, some in vitro effects in themselves raise substantial concerns about potential for harm. These effects, as well as other in vitro assays with less clinical validation and independent predictive value, become very important in assessing biological plausibility of observations made or predicted by other systems, such as animal, human, or structural association. While it is not necessary to determine a rational mechanism of harm to determine that an ingredient is potentially unsafe, it is valuable to identify possible mechanisms that explain the totality of the data. In vitro studies can be very useful and irreplaceable in this regard. If in vitro data are a central focus of the critical safety evaluation, expert opinion in the extrapolation of in vitro data will be sought. Such expert opinion will help ensure that the data are appropriately reviewed in the context of all information available regarding the ingredient. Guiding Principle for In Vitro Data In vitro studies can serve as signals of potential harmful effects in humans, but not as independent indicators of risk unless an ingredient causes an effect that has been associated with harmful effects in animals or humans and there is evidence that the ingredient or its metabolites are present at physiological sites where they could cause harm. Alone, in vitro data should serve only as hypotheses generators and as indicators of possible mechanisms of harm when the totality of the data from the different factors is considered. MODIFYING FACTORS In addition to the key factors that explicitly contribute to the screening/flagging and priority-setting steps, prevalence of use and vulnerability of subpopulations are considered as modifying factors. These two factors must be considered when evaluating the different types of scientific evidence. Both are considered to some extent in the screening/flagging step and when setting priorities for evaluation. Prevalence of use in the general population, however, is not a factor considered during the critical safety evaluation. Prevalence of Use in the Population The number of individuals who could be at risk of harm due to overall use of the dietary ingredient in the United States (or “prevalence of use”) can be estimated from various types of data that provide estimates of the relative popularity of different ingredients. Across the wide variety of dietary supplement ingredients that are currently available, there is a wide range of usage patterns in the population. Some ingredients are used only rarely or by a small fraction of the population, while others are used by a considerable fraction. Dietary supplements that are readily available or are consumed or marketed for common concerns and conditions are more likely to result in a high level of usage, while those that are less widely available, used only

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements rarely, or used by few consumers would be expected to result in a low level of exposure on a population basis. Estimating prevalence of use allows a qualitative consideration of population exposure and therefore how much of the population may be at risk if an ingredient is harmful—a factor that is important in optimizing the impact of a rigorous safety evaluation on public health. That is, from a public health perspective, it is more logical to first allocate resources to evaluation of potentially harmful ingredients that have the potential to harm many people before evaluating those ingredients that may only affect a small fraction of the population (assuming other information is equivalent). Relative consumption and prevalence of use of various dietary supplement ingredients in the general population can be estimated from two types of data. One type is industry estimates of production or sales. Dietary supplement industry publications such as the Nutrition Business Journal provide such data. Additionally, manufacturers and distributors collect production data, unit sales data, and total sales information in dollars as a normal component of business operations. The industry may be willing to make this information available. Neither unit sales nor total sales data are ideal, but both can serve as proxy indicators useful in developing a qualitative understanding about prevalence of use. Unit sales data, especially, allow a rough comparison of relative potential use among different ingredients, serving as a surrogate marker of actual use of each ingredient—information that is usually not readily available. One limitation with the use of sales figures, however, is that these numbers are often collected and collated by methods that make cross-category comparisons difficult. This is important to keep in mind when estimating relative ingredient use. The second type of data about prevalence of use is that collected in surveys about supplement use. Increasingly, national surveys that have traditionally collected information from a large number of persons regarding health issues and conventional food consumption information are also collecting valuable information about specific supplement use. An example is the expanded monitoring efforts of the National Health and Nutrition Examination Survey. Although it may be several years before this information is available, the expanded data collection will provide more detailed and useful information than is currently available. Such surveys can provide data on prevalence of use in the general population as well as in specific population groups. Most surveys also ask subjects about patterns of use and other information that is helpful in estimating the effect of potential adverse effects on a population. Although some publications based on such surveys group supplement ingredients into broad categories (e.g., vitamins, botanicals) for the purpose of data analysis, other sources of information are likely to list specific ingredients or products. Even when data on specific ingredients are not included in the published articles, such data might have been collected and might be available from the investigators upon request. One deficiency of older survey data sets is that frequency of use information was rarely collected (i.e., differentiating supplement use for only short intermittent periods versus chronic use). This type of information is important because it augments the evaluation of total sales figures. Some ingredients that are widely sold may be used less frequently than others. Whether a substance is used for short periods of time or chronically is particularly important in evaluation of safety, because a product or ingredient that is used intermittently will usually pose a smaller or different type of risk than one used chronically. Another limitation of available information is in the collection and interpretation of data on combination products or ingredients that are typically used in combination products (in addition

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements to formulation and sales as a single-ingredient product). Obtaining reasonably accurate estimates of usage of ingredients that are largely consumed as components of combination products may be more difficult but may be available from manufacturers of raw ingredients or if registration procedures for dietary supplement ingredients, similar to that currently required for drugs, is implemented at some point in the future. Prevalence of Use as a Modifying Factor in the Screening/Flagging Step Information about prevalence of use is not considered as a key factor, but it is used as a modifying factor for human data. Prevalence of use is considered in the screening of human data in that it may mitigate or exacerbate concern. That is, if an ingredient is widely used but few adverse events are spontaneously reported, it is less likely to be flagged than a rarely used ingredient with a similar number of spontaneously reported adverse events. Prevalence of Use as a Modifying Factor in the Priority-Setting Step The prevalence of use is considered in the priority-setting step only in establishing relative rank within a Priority Group. Within each Priority Group (explained in Chapter 5), items that are widely used are moved to the top of that Priority Group. Prevalence of Use as a Modifying Factor in the Critical Evaluation Step Prevalence of use is not considered in the critical safety evaluation step. Guiding Principle for Prevalence of Use Data Ingredients that are widely used by the general population should be given higher priority for critical safety evaluation than less widely used ingredients with similar degrees of safety concerns. This is consistent with the public health goal of producing the most impact from limited resources. Use by Vulnerable Subpopulations When considering the safety of supplement ingredients or other substances, it is important to consider that some individuals may be particularly vulnerable to adverse effects from certain supplement ingredients. Vulnerable subpopulations can be defined as groups of individuals who are more likely to experience an adverse event related to the use of a particular dietary supplement ingredient, or individuals in whom such events are more likely to be serious in comparison with the general population. Characteristics that contribute to such vulnerability may be physiological, disease-related, or due to other aspects, such as therapeutic interventions that are commonly utilized by the subgroup. An example of a physiological characteristic that results in an individual’s increased susceptibility compared to the general population is the change in the capacity for metabolism of various dietary supplement ingredients across the lifespan. Changes in metabolism may lead to variable concentrations of active compounds at sites of action and result in different responses.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements For example, elderly individuals are a potential vulnerable subpopulation for some ingredients in that aging is associated with changes in the ability to digest, metabolize, or excrete some ingested substances (Munro, 1989; Rosenberg et al., 1989). Supplement ingredients that are normally cleared by, or altered by, the kidney or liver may potentially pose a greater risk to this subgroup than to a younger population. This factor should be considered for supplements specifically directed toward an older population. Children also metabolize some chemical substances differently than do adults, which for certain supplement ingredients might make children more susceptible to adverse effects and should be taken into consideration for any supplements marketed toward children. Likewise, differences in metabolism between children and adults may make children more resistant to adverse effects of certain substances (Guzelian et al., 1992). Infants have limited hepatic function that may make them particularly susceptible to certain hepatotoxic substances. Other age-related changes may involve receptors or kinetic parameters such as the volume of distribution. Physiological changes that occur during pregnancy may also influence susceptibility to adverse effects associated with particular supplement ingredients. In addition to life stages that may alter responses to ingested substances, the presence of disease may also result in enhanced susceptibility to adverse effects from particular ingredients. For example, hepatitis or renal disease can significantly alter xenobiotic clearance, allowing compounds that are normally cleared rapidly to accumulate to toxic levels. People who are prescribed critical medications to be used on a chronic basis may be at greater risk of harm from drug interactions with various supplement ingredients. For example, people living with HIV/AIDS or other chronic diseases may be taking drug combinations that may interact with supplement ingredients, such as St. John’s Wort, that alter cytochrome P450 activity (Ernst, 1999; Piscitelli et al., 2000). Interactions between drugs and dietary supplements may be of particular concern when both are recommended for the same pathology and are thus potentially taken at the same time. For example, vitamin E supplements, which are often recommended to patients with atherosclerotic vascular disease, may have an interaction with statin drugs (Brown et al., 2001). Disease or pre-existing conditions, such as hypertension, cardiac arrhythmias, or other early stages of cardiovascular disease, can also be expected to exacerbate susceptibility to products that specifically affect the organ exhibiting the disease or condition. Another example is the potential for supplements that affect insulin and glucose regulation to affect persons with diabetes. Thus, factors such as age, disease, pre-existing conditions, ethnicity, gender, or history of specific xenobiotic exposure such as pesticides can alter supplement exposure by altering pharmacodynamics and clearance of a drug. Alternatively, these factors may alter the dose-response, causing certain individuals to be more sensitive to a specific supplement than the majority of the population. The paragraphs above describe several general reasons for particular susceptibilities. In addition, special concerns are warranted for supplement ingredients that may have teratogenic effects. Fetuses may be harmed if exposed to dangerous substances in utero as may infants if exposed to substances released into human milk. A well-known example is the teratogenicity of high doses of vitamin A and related retinoids in the periconceptual period (Eckoff and Nau, 1990; Lammer et al., 1985; Rothman et al., 1995). Animal studies or chemical characteristics may provide clues that fetuses or infants are particularly susceptible to other supplement ingredients as well.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements In summary, certain segments of the population may be particularly susceptible to the effects of some supplement ingredients for a variety of reasons. When reviewing data, it is important to ask if ingredients are more likely to cause harmful effects to particular subgroups of the population. In the proposed framework, “vulnerable groups” are described as a modifying factor, in that whether identifiable subpopulations are particularly susceptible to harm should always be taken into consideration when assigning screening and priority-setting scores (see Chapter 5 for more details). Vulnerable Subpopulation Information as a Modifying Factor in the Screening/Flagging Step In examining the human data used to screen/flag dietary supplement ingredients, the reviewer should consider if the dietary supplement ingredient is being used by subpopulations that are particularly susceptible to serious adverse effects. Vulnerable Subpopulation Information as a Modifying Factor in the Priority-Setting and Critical Safety Evaluation Steps Particular susceptibility of identifiable subgroups of the population is taken into account when considering the scientific evidence (human, animal, structure/chemotaxonomy, and in vitro evidence) in this step, as noted in the description of scoring provided in Chapter 5. Guiding Principle for Vulnerable Subpopulation Data When data indicate that an identifiable Subpopulation may be especially sensitive to adverse effects from a certain supplement ingredient, then this higher level of concern should be taken into account when scoring the ingredient. New Ingredient Status A new dietary supplement ingredient is one that was introduced to the U.S. market after October 1994, when the Dietary Supplement and Health Education Act was implemented. In this proposed framework, the notification of intent to market a new dietary supplement ingredient by a manufacturer automatically moves the ingredient through the screening/flagging step and on to the priority-setting step. Noting new ingredient status allows the ingredient to be considered during the priority-setting process in relation to other new ingredients and other ingredients flagged in the first step of the process. If the new ingredient has been used in other countries there may be a considerable amount of clinical, animal, and in vitro data; data on serious adverse events; or data on usage patterns. If this type of information is not available, it is expected that considering the biological activity of related substances will be helpful in setting priorities in Step Two of the process. If the new dietary supplement ingredient is marketed after the 75-day notification period, more data may become available, and the priority level of the ingredient may change. The rationale behind channeling all new ingredients directly into the priority-setting process is that the limited information on the other factors, especially human data, is no indication of

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements safety because the ingredient has not been marketed in the United States. Thus, the direct channeling of all new products to the priority-setting step assures that scientific data and theoretical prediction of harm are considered by the framework to some degree. Use of New Ingredient Status in the Screening/Flagging Step Consideration of new ingredient status in the screening/flagging step does not require that FDA actively collect information. The 75-day notification, which comes to FDA directly from the manufacturer, indicates that the item should be flagged and moved forward to the priority-setting step, simply by virtue of being a new ingredient about to enter the market. Use of New Ingredient Status in the Priority-Setting and Critical Safety Evaluation Steps New ingredient status is not considered in the priority-setting step or in the critical safety evaluation of an ingredient. The new ingredient status simply moves it into the priority-setting process, where the evaluation of evidence of the other key factors occurs. SUMMARY In summary, several different key factors and modifying factors should be taken into consideration when evaluating the safety of dietary supplement ingredients. The four primary key factors (data from humans and clinical evidence of harm, animal data, bioactivity of structurally related and taxonomically related substances, and in vitro data) contribute to different extents in the two first steps of the proposed process, screening/flagging and priority-setting, as compared to their contribution to the third step, the critical safety evaluation. The same holds true for the two modifying factors (prevalence of use in the population and use by vulnerable subpopulations). In Chapter 5, the processes for screening/flagging and priority-setting are described, with more suggestions about how different types of data are appropriately weighted when setting priorities. A systematic approach to this weighting process is described. After ingredients are categorized and prioritized, in-depth safety evaluations should be conducted for the ingredients with the greatest safety concerns (and which thus have the highest priority scores). Chapter 6 outlines a system for conducting these reviews and also revisits the guiding principles underlying the consideration of different types of data that were first outlined in this chapter. REFERENCES AIAI (Artificial Intelligence Applications Institute). 2002. TOPKAT—The Open Practical Knowledge Acquisition Toolkit. Online. University of Edinburgh. Available at http://www.aiai.ed.ac.uk/-jkk/topkat.html. Accessed June 19, 2002. Bravo L. 1998. Polyphenols: Chemistry, dietary sources, metabolism, and nutritional significance. Nutr Rev 56:317–333. Brown BG, Zhao XQ, Chait A, Fisher LD, Cheung MC, Morse JS, Dowdy AA, Marino EK, Bolson EL, Alaupovic P, Frohlich J, Albers JJ. 2001. Simvastatin and niacin, antioxidant vitamins, or the combination for the prevention of coronary disease. N Engl J Med 345: 1583–1592.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements Budzinski JW, Foster BC, Vandenhoek S, Arnason JT. 2000. An in vitro of human cytochrome P450 3A4 inhibition by selected commercial herbal extracts and tinctures. Phytomedicine 7:273–282. 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. CFSAN (Center for Food Safety and Applied Nutrition). 1993. Toxicological Principles for the Safety Assessment of Direct Food Additives and Color Additives Used in Food. Draft Redbook II. Washington, DC: CFSAN, Food and Drug Administration. CFSAN. 2001a. Overview of Dietary Supplements. Online. Food and Drug Administration. Available at http://www.cfsan.fda.gov/-dms/ds-oview.html. Accessed February 22, 2002. CFSAN. 2001b. Toxicological Principles for the Safety of Food Ingredients. Redbook 2000. Online. Food and Drug Administration. Available at http://www.cfsan.fda.gov/-redbook/red-toca.html. Accessed June 19, 2002. Chyka PA, McCommon SW. 2000. Reporting of adverse drug reactions by poison control centers in the US. Drug Safety 23:87–93. Counsell C. 1997. Formulating questions and locating primary studies for inclusion in systematic reviews. Ann Intern Med 127:380–387. Eckhoff C, Nau H. 1990. Vitamin A supplementation increases levels of retinoic acid compounds in human plasma: Possible implications for teratogenesis. Arch Toxicol 64:502–503. Ernst E. 1999. Second thoughts about safety of St John’s wort. Lancet 354:2014–2016. FDA (Food and Drug Administration). 1995. Clinical Therapeutics and the Recognition of Drug-Induced Disease. A MedWatch Continuing Education Article. Online. Available at http://www.fda.gov/medwatch/articles/dig/rcontent.htm. Accessed February 25, 2002. FDA. 2002. Good Clinical Practice in FDA-Regulated Clinical Trials. Online. Available at http://www.fda.gov/oc/gcp/guidance.html. Accessed February 25, 2002. Guzelian PS, Henry CJ, Olin SS, eds. 1992. Similarities and Differences Between Children and Adults: Implications for Risk Assessment. Washington, DC: ILSI Press. ICH (International Conference on Harmonisation). 1995. Guidance for Industry. E3. Structure and Content of Clinical Study Reports. Rockville, MD: Drug Information Branch, CDER, FDA. ICH. 1996. Guidance for Industry. E6. Good Clinical Practice: Consolidated Guidance. Rockville, MD: CDER, FDA. ICH. 2001. Guidance for Industry. S7A. Safety Pharmacology Studies for Human Pharmaceuticals. Rockville, MD: Division of Drug Information, CDER, FDA. ILAR (Institute of Laboratory Animal Research). 1996. Guide for the Care and Use of Laboratory Animals. Washington, DC: National Academy Press, Ioannidis JPA, Lau J. 2001. Reporting of the safety of medications in randomized trials in neglected: An evaluation of 7 medical areas. J Am Med Assoc 285:437–443. Ioannidis JPA, Chew P, Lau J. 2002. Standardized retrieval of side effects data for meta-analysis of safety outcomes: A feasibility study in acute sinusitis. J Clin Epidemiol 55:619–626. Kingsbury JM. 1964. Poisonous Plants of the United States and Canada. Englewood Cliffs, NJ: Prentice Hall.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements Lammer EJ, Chen DT, Hoar RM, Agnish ND, Benke PJ, Braun JT, Curry CJ, Fernhoff PM, Grix AW Jr, Lott IT, Richard JM, Sun SC. 1985. Retinoic acid embryopathy. N Engl J Med 313:837–841. Lewis JA. 1981. Post-marketing surveillance: How many patients? Trends Pharmacol 2:93–94. Liu XK, Katchman A, Ebert SN, Woosley RL. 1998. The antiestrogen tamoxifen blocks the delayed rectifier potassium current, IKr, in rabbit ventricular myocytes. J Pharmacol Exp Ther 287:877–883. Moher D, Schulz KF, Altman DG. 2001. The CONSORT statement: Revised recommendations for improving the quality of reports of parallel group randomized trials. BMC Med Res Methodol 1:2–8. Munro HN. 1989. The challenges of research into nutrition and aging. Introduction to a multifaceted problem. In: Munro HN, Danford DE, eds. Nutrition, Aging, and the Elderly. New York: Plenum. Pp. 1–21. Obach RS. 2000. Inhibition of human cytochrome P450 enzymes by constituents of St. John’s Wort, an herbal preparation used in the treatment of depression. J Pharmacol Exp Ther 294:88–95. OIG (Office of Inspector General). 2001. Adverse Event Reporting for Dietary Supplements. An Inadequate Safety Valve. OEI-01–00–00180. Boston, MA: OIG, Food and Drug Administration. Piscitelli SC, Burstein AH, Chaitt D, Alfaro RM, Falloon J. 2000. Indinavir concentrations and St John’s wort. Lancet 355:547–548. Prevention Magazine. 2001. Consumer Use of Dietary Supplements. Emmaus, PA: Rodale. Rosenberg IH, Russell RM, Bowman BB. 1989. Aging and the digestive system. In: Munro HN, Danford DE, eds. Nutrition, Aging, and the Elderly. New York: Plenum. Pp. 43–60. Rothman KJ, Moore LL, Singer MR, Nguygen UDT, Mannino S, Milunsky B. 1995. Teratogenicity of high vitamin A intake. N Engl J Med 333:1369–1373. Sackett DL, Haynes BR, Tugwell P. 1991. Clinical Epidemiology: A Basic Science for Clinical Medicine. Boston: Little, Brown. Pp. 283–302. Scott HD, Rosenbaum SE, Waters WJ, Colt AM, Andrews LG, Juergens JP, Faich GA. 1987. Rhode Island physicians’ recognition and reporting of adverse drug reactions. Rhode Island Med J 70:311–316. Tegtmeier M, Harnischfeger G. 1994. Methods for the reduction of thujone content in pharmaceutical preparations of Artemisia, Salvia and Thuja. Eur J Pharm Biopharm 40:337–340. U.S. Consumer Supplement Use Summary 1999. 2000. Online. Nutrition Business Journal. Available at http://www.nutritionbusiness.com/sub/consume.xls. Accessed March 1, 2002. Wang WX, Ebert SN, Liu XK, Chen YW, Drici MD, Woosley RL. 1998. “Conventional” antihistamines slow cardiac repolarization in isolated perfused (Langendorff) feline hearts. J Cardiovasc Pharmacol 32:123–128. Yang CS, Landau JM, Huang MT, Newmark HL. 2001. Inhibition of carcinogenesis by dietary polyphenolic compounds. Annu Rev Nutr 21:381–406. Zabel M, Franz MR. 2000. The electrophysiological basis of QT dispersion: Global or local repolarization? Circulation 101:E235–E236.

OCR for page 47
For Comment: Proposed Framework for Evaluating the Safety of Dietary Supplements This page intentionally left blank.