Regulatory context

Because foodstuffs themselves are difficult to analyse and have a variable composition, testing packaging materials for migration almost invariably uses model foods, known as food simulants , which are designed to mimic the main classes of foodstuffs. There are some minor differences in detail between EU and US-FDA requirements but the essential elements are listed in Table 9.3. 

Acidic foods pH 4.5 Dilute (3%) acetic acid solution in water 

Fatty foods Olive oil, heptane, 95% ethanol, Miglyol 

In the European Union (EU), the legal foundations governing materials and articles intended to come into contact with food are contained in the Framework Directive 89/109/EEC, which states in Article 2  

Materials and articles must be manufactured in compliance with good manufacturing practice so that, under their normal or foreseeable conditions of use, they do not transfer their constituents into foodstuffs in quantities which could  

---

Regarding the last but not the least principle, it is recognized that it is not always possible to provide a mechanistic interpretation of a given QSAR model but the absence of this information does not preclude the use of the given model in the regulatory context. 

The BCS has been developed primarily for regulatory applications, although its use has been extended beyond this area (as discussed in more detail below). The aim of the BCS in a regulatory context is to provide a basis for replacing certain bioequivalence studies by equally or more accurate in vitro dissolution tests. This could reduce costs and time in the development process as well as reducing unnecessary drug exposure in healthy volunteers, which is normally the study population in bioequivalence studies. 

When epidemiological evidence is limited, and insufficient to establish causation, it may remain important in the hazard identification step if it is supported by reliable animal data. Within the regulatory context, convincing animal evidence of toxicity, even in the absence of strong epidemiological evidence, or indeed any epidemiological evidence, will still be used for hazard characterization. 

The threshold for the human population is estimated by dividing the NOAEL (or, alternatively, the BMD), by an uncertainty factor (UF), the size of which depends upon the nature and quality of the toxicity data and the characteristics of the human population. (The estimated human threshold dose has several different names, depending upon the regulatory context, see later.)... 

It is possible to conduct animal studies in an infinite number of ways. Although individually designed studies are often scientifically sound, and in many cases serve a particular purpose very well, they pose problems in a regulatory context. Free movement of chemicals between countries is based on the mutual acceptance of the risk evaluation made by each country and this, in turn, relies on the mutual acceptance of the data generated when testing the chemicals. Experience has shown this acceptance to be extremely difticult, if chemicals have been tested by different methods. 

The approach to exposure assessment is not as internationally harmonized as hazard assessment. A synopsis of current activities regarding exposure assessment for industrial chemicals in a number of OECD Member countries has been published (OECD 2006). The executive summary of this document states that while there is a significant level of sharing of approaches used for hazard characterization for risk assessment, this is not the case for exposure characterization. Although broad consistency in the overall approaches used by different countries in conducting exposure assessment exists, there is variation in policy-related factors, including the regulatory context for assessment and the way that information is applied, as well as in the types of approaches and tools used. 

Empirical Bayes methodology and other kinds of shrinkage estimation may be considered in situations where there is some, perhaps limited information for a situation of specific interest, but also a desire to give some weight to data from situations less representative. The term shrinkage expresses the idea that an estimate from the situation of specific interest is shrunk toward some prior estimate such as an estimate from less strictly representative situations. As yet the methods have seen little or no use for pesticide ecological risk assessment in regulatory contexts. 

There is some USEPA precedent for use of statistical meta-analysis in a regulatory context, including the recent meta-analysis of organophosphate-related acetylcholinesterase inhibition data and meta-analysis of epidemiological studies on effects of 2nd hand tobacco smoke exposure. Warren-Hicks and Moore (1998) provide some discussion of the potential applicability of meta-analysis to ecological risk assessments. 

Table IV lists the results of risk calculations provided in the preliminary proposal for the substances that were proposed as potential carcinogens in the regulatory context at that time (44). 1,1-Dichloroethylene was later converted to a listing of equivocal evidence of carcinogenicity. The table includes calculations made by the USEPA CAG and the NAS Safe Drinking Water Committee. These calculations attempt to project concentrations of each chemical in drinking water that, if consumed for a lifetime (70 years) at the rate of 2 L of water per day would contribute an excess lifetime cancer risk of up to 1 in 100,000 and up to 1 in 1,000,000. The quality of evidence of carcinogenicity ranging from sufficient in humans to limited in animals is also included for each chemical. Provisional ADI values calculated from chronic toxicity data only are included for the sake of comparison.

The Need for Estimation Methods 1-1.1 Regulatory Context 1-1.2 Chemical Design 1-2 Overview of this Work 1-2.1 Chapter Contents 1-2.2 Properties of Pure Substances 1-2.3 Partitioning Properties 1-2.4 Reactivity or Persistence 1-2.5 Specific Classes of Substances 1-2.6 Benchmark Chemicals References... 

The fifth principle, i.e., the mechanistic interpretation, is an added value. The OECD principle means that The absence of a mechanistic interpretation for a model does not mean that a model is not potentially useful in the regulatory context. The intent of Principle 5 is not to reject models that have no apparent mechanistic basis [18]. We notice, as previously discussed, that there are two different perspectives to build up a model, which predicts a value, or to develop a model that is useful to study a phenomenon. These two targets are independent. Different techniques can be applied, and we can imagine a model that is only predictive, but without studying the reasons why certain descriptors are used, or vice versa, we can imagine a model that is only dedicated to understand why certain phenomena occur, for instance, if we want to study why the chlorinated methanes have a given toxicity. In this case, all toxicity data are available, but we may want to investigate the role of the chlorine atom in the phenomenon. 

An array of extrapolation types for matrix and media extrapolations has been given, with specihc approaches being dependent on compound, medium and matrix, and species and ecosystem properties. Guidance to provide a systematic orientation in the array of methods is useful when considering their practical use in a regulatory context. Using a tiered approach (see Chapter 1), the tiers for media and matrix extrapolation listed in Table 2.11 are recommended. 

A mechanistic interpretation, if possible Models without mechanistic interpretation can be used in a regulatory context. However, consideration of the mechanistic association between descriptors used in the model and the predicted endpoint should be sought to improve the regulatory applicability of the (Q)SAR. Guides to (Q)SARs currently used in OECD member states has been published (OECD 2005a, 2005b). 

Mesman M, Posthuma L. 2003. Ecotoxicity of toxicant mixtures in soils recommendations for application in the Dutch regulatory context, as derived from a scientific review on approaches, models and data. No. 711701035. Bilthoven (The Netherlands) National Institute of Public Health and the Environment (RIVM), 70 p. 

The book is based on contributions from thirty-five scientists, regulators, and policy makers from eleven countries with individual expertise across disciplines such as risk assessment, environmental, health, economic, and social sciences. These scientists summarize current knowledge on aquatic and terrestrial environmental quality standards, placing these standards in a wider socioeconomic and regulatory context. The book explains how to derive environmental standards that are defensible from a scientific and socioeconomic perspective. Using multidisciplinary techniques applicable to water, sediments, and soils, the text demonstrates how to select the best form and derivation method relative to individual environmental standards. 

Which effect assessment method should be applied in a particular situation depends on the nature of the mixture problem at hand. Because the diversity in assessment methods is large, it is important to clearly describe the problem. For example, derivation of a safe level for a proposed industrial mixture emission requires a different approach than the prioritization of a number of sites contaminated with mixtures. The former problem requires the assessment of realistic risks, for example, by the application of a suite of fate, exposure, and effect models, whereas the application of a simple consistent method suffices to address the latter problem, for example, a toxic unit approach. A successful and efficient assessment procedure thus starts with an unambiguous definition of the mixture problem at hand. The problem definition consists of the assessment motive, the regulatory context, the aim of the assessment, and a structured or stepwise approach to realize the aim. Elaboration of the problem definition is an iterative process (Figure 5.1) that strongly depends on factors such as resources, methods, data availability, desired level of accuracy, and results of previous studies. 

---