General Overview of the Safety Evaluation of Chemicals

PAUL ILLING

1 Introduction

Society demands that chemicals be 'safe'. This requirement is exercised through legislation concerned with the safety of chemicals, usually introduced post hoc, in response either to disasters or to perceived inadequacies in the previous legislation. Public perceptions of risk drive priorities and legislative agendas. Generally legislatures are concerned with setting frameworks and setting up bodies to develop detailed requirements, evaluate the evidence and enforce the legislation.

When human health is the issue, the aim is to prevent ill health. In order to do this it is necessary to have means of predicting ill health. Ethically, it is difficult to justify predictive testing in humans unless a substance is intended for administration to humans. Thus tests are normally conducted in laboratory animals as surrogates for humans. Even when testing in humans is possible, preliminary testing in animals is usually required. The bodies concerned with evaluating safety therefore set minimum testing requirements for those who wish to place on the market a chemical substance or to market it for a specified use. These requirements include:

• when and what tests should be conducted (testing strategies);

• the protocols setting out how the tests should be conducted (test methods); and

• the audit procedures to ensure that tests have been properly conducted and reported ('Good Laboratory Practice'; GLP).

Furthermore, the bodies may either evaluate the results of the tests or set guidance as to how the results should be interpreted.

In the context of safety testing, chemicals includes natural products as well as synthetic chemicals. While the general public may consider 'natural' as good (less harmful) and 'synthetic' as bad (more harmful), there is no scientific reason to differentiate between 'natural' and 'synthetic' chemicals. Ricin (from castor beans), the toxins produced by Cl. botulinus and a number of mycotoxins are examples of natural materials of great potency as toxic agents. Generally, therefore, legislation does not differentiate between 'natural' and 'synthetic' chemicals.

One of the main tools used to predict what the potential ill health effects of chemicals might be is animal testing. A second societal demand, sometimes expressed with considerable force by groups within society who are against conducting work with animals, is associated with the welfare of animals used in experimental procedures. The aim is to prevent, or, at least, to minimize experiments on, usually, vertebrates. Generally, society has been less tolerant of animal testing for substances intended for use as cosmetics or cosmetic ingredients than for substances as a whole.

This clash between the demand for safety and the limitations on human experimentation is the reason why testing is undertaken in animals. The requirements concerning animal welfare dictate that, wherever possible, alternatives to animal testing should be sought.

2 Legislation and Regulatory Requirements

Predictive testing is best instigated when a licence or some other form of permission is required before a substance can be used. It may also be required in law prior to marketing even when there is no formal assessment of the data by a regulator. Thus, predictive testing is required for 'new substances' about to be placed on the market. It is also required prior to use for substances intended for specific uses, such as drugs, veterinary medicines, cosmetics/personal care products, food additives, plant-protection products and biocides. In these cases the person or corporate body desiring to market the substance carries out the required testing. Predictive testing almost always includes experiments in animals.

Legislation may also cover the presence of chemicals in air, water or soil. This legislation generally deals with evaluating existing situations rather than new ones and post hoc information from studies in humans should be available. Under these circumstances, predictive testing may be required for endpoints not easily addressed through studies in humans. However, in these cases it is usually Government that sponsors and finances this work.

Originally legislation concerning the safety of chemicals was based on the individual national government. Nowadays it is mostly based on regional (e.g. European Union [EU]) or international (Organization for Economic Co-operation and Development [OECD]; World Trade Organization [WTO] or United Nations [UN]) groupings.

Two areas of legislation are particularly pertinent to the current debate on alternatives to animal testing as the testing requirements associated with them are currently being revised. The first is that concerned with the placing on the market of chemicals generally. Currently this is governed by EU Council Directive 67/548/EEC. A Council Directive, implemented in the UK as the 'Notification of New Substances Regulations' deals with 'new substances', i.e. those placed on the market after 1981. Under the seventh amendment, a dossier, containing, inter alia, information on testing in animals, is required prior to placing the substance on the market. Testing is tiered, some testing being conducted before the substance is placed on the market and further testing being required as the amounts placed on the market increase. Currently testing strategies are given in Directive 92/32/EEC and in a technical guidance document and test protocols are given in Annex V of Directive 67/548/EEC. The test protocols are those set up by the OECD. Tests have to be conducted to these protocols and audited in accordance with the principles of good laboratory practice. Animal welfare considerations, set out in Directive 86/609/EC, should be taken into account when deciding on the need for a specific test.

Existing substances are still classified and labelled on the basis of existing information.

The Registration, Evaluation, Authorization and restriction of CHemicals (REACH) proposals currently before the EU Council and Parliament, intend to bring the regulation of both new and existing substances into a common scheme, and to require that specific information is available under specified conditions. For existing substances this implies those placing on the market the substance or preparations containing it may need to set up consortia to conduct testing to fill data gaps. The REACH proposals also have a testing strategy contained within them, and the test protocols are those set out in the OECD Guidelines. Although currently the classification is that for Directive 67/548/EEC, it is likely that the UNECE Globally Harmonized System (GHS) will be substituted before implementation of the proposals. It is claimed that the REACH proposals should encourage the use of alternatives to animal testing.

The second is the area of 'cosmetics' – more properly personal care products. These are governed by Council Directive 76/768/EEC. There are lists of acceptable ingredients and of banned or restricted ingredients. The regulator evaluates dossiers on certain ingredients, information on other ingredients and products are evaluated by the manufacturer/formulator placing the product on the market. The testing requirements have been set out in guidance from the Scientific Committee on Cosmetic Products and Non-Food Products (SCCNFP) intended for Consumers, the latest version being the fifth revision. The tests protocols are based on those set out in Annex V of Directive 67/48/EEC, and are supplemented with SCCNFP opinions concerning studies on human volunteers and mutagenicity/genotoxicity testing. The seventh amendment to this Directive requires that testing of cosmetic products and ingredients in live animals to meet the requirements of the Cosmetics Directive should cease. It gives the European Commission powers to set deadlines after which testing of cosmetics ingredients in animals will not be permitted if the intention is to meet the requirements of the Cosmetics Directive. The Directive sets maxima of a 6 year time limit for short-term testing and a 10 year time limit for repeated dose toxicity, reproductive toxicity and toxicokinetics, with provision that shorter deadlines can be introduced if suitable test methods become available earlier.

One difference between the two sets of Directives is the attitude to animal experimentation. Perhaps society has taken the view that the use of cosmetics and the choice of preparation used is, compared to the generality of the exposure of chemicals, relatively voluntary, and that animal welfare considerations should be given greater prominence in deciding on the need for animal testing. A second difference concerns human exposure, and hence the ability to carry out studies in humans. Studies in humans have to comply with ethical guidelines laid out in the World Medical Association Declaration of Helsinki, as amended and clarified. Generally, studies in humans on chemicals are not encouraged, and this is reflected in the regulatory requirements for testing of chemicals placed on the market. However, cosmetics (personal care products) are a group of chemicals intended for application to human skin. Thus it is normally possible, with care, to undertake certain tolerance studies for short-term end points in humans, and this is reflected in the regulatory guidance for their testing.

3 The Regulatory Paradigm

Safety evaluation is essentially risk evaluation, with the intent that the risk of ill-health occurring should be acceptably low (i.e. 'safe'). The concepts of risk assessment and risk management go back to two key documents published in 1983. The first is the US 'Red Book' and the second is a Royal Society Study Group report. Some definitions are required if the regulatory paradigm is to be understood. A recent set of definitions is that of the OECD and IPCS.

3.1 Hazard and Risk

Probably the most important concept in risk analysis is the distinction between hazard and risk.

Hazard: 'The inherent property of an agent or situation having the potential to cause adverse effects when an organism, system or (sub)population is exposed to that agent.

Risk: The probability of an adverse effect in an organism, system or (sub)population caused under specified circumstances to an agent'.

A good short statement relating these definitions is that 'risk is the possibility of suffering harm from a hazard'. In the case of human health, the hazard is the interaction of a toxic agent with a receptor (the human), and the risk is the likelihood of sufficient interaction occurring such that ill-health results. Because the human is a fixed element, the properties of the interaction are taken as the hazardous properties of the agent, for our purposes the chemical substance.

3.2 Risk Assessment and Risk Management

The OECD/IPCS definitions for risk assessment and risk management are given below.

'Risk assessment. A process intended to calculate or estimate the risks to a given target organism, system or (sub) population, including the identification of attending uncertainties, following exposure to a particular agent, taking into account the inherent characteristics of the agent of concern as well as the characteristics of the specific target system.

'The risk assessment process includes four steps: hazard identification, hazard characterization (related term: dose-response assessment), exposure assessment and risk characterization. It is the first component in a risk analysis process'.

Risk management. '[A] Decision-making process involving considerations of political, social, economic and technical factors with relevant risk assessment information relating to hazard so as to develop, analyse, and compare regulatory and non-regulatory options and to select and implement [an] appropriate regulatory response to that hazard.

Risk management comprises three elements: risk evaluation; emission and exposure control; and risk monitoring'.

Safety evaluation of chemicals involves all the stages of risk assessment and the stage of risk evaluation. These are:

Hazard identification: 'The identification of the type and nature of adverse effects that an agent has the inherent capacity to cause an organism, system or (sub)population'.

Hazard characterization: 'The qualitative, and wherever possible, quantitative description of the inherent properties of an agent or situation having the potential to cause adverse effects. This should, where possible, include a dose-response assessment and its attendant uncertainties'.

Exposure assessment. 'Evaluation of the exposure of an organism, system or (sub) population to an agent (and its derivatives)'.

Risk characterization: 'The qualitative and, whenever possible, quantitative determination, including attendant uncertainties, of the probability of an occurrence of known and potential adverse effects of an agent to a given organism, system or (sub)population under defined exposure conditions'.

Risk evaluation: 'Establishment of a qualitative or quantitative relationship between risks and benefits of exposure to an agent, involving the complex process of determining the significance of the identified hazards and estimated risks to the system concerned or affected by the exposure, as well as the significance of the benefits bought about by the agent' [my italics].

In the case of human health the 'given organism' is human. As risk evaluation involves assessment of the 'significance of the identified hazards and estimated risks' to 'those concerned with or affected by the exposure' it involves more than a technical assessment of the toxicological and exposure data. 'Significance to those concerned with or affected by the exposure' implies that the way in which those concerned with or affected by the exposure perceive the risks, and hence the sociological and psychological factors affecting how people perceive risk become important. Sociological and psychological opinions concerning risk have to be taken into account when evaluating risks.

3.3 The Risk Evaluation Framework

The Royal Society study group first put forward a risk evaluation framework. This framework (the 'tolerability of risk' concept) has been restated and slightly developed in 'Reducing risks, protecting people'. It was originally developed to handle engineering risk, but it is equally applicable to risks to human health. Illing and the Health and Safety Executive (HSE) have discussed the application of this framework to the evaluation of health risks arising from exposure to chemicals.

Criteria for reaching decisions can be classified according to three 'pure' criteria. These are:

• An equity based criterion, which starts from the premise that all individuals have unconditional rights to certain levels of protection. This leads to standards, applicable to all, held to be usually acceptable in normal life. In practice this leads to fixing a limit to represent the maximum level of risk above which no individual can be exposed. If the risk characterization indicates that the risk is above this limit the risk is held to be unacceptable, whatever the benefits.

• A utility based criterion, which applies to the comparison between incremental benefits of measures to prevent the risk of injury or detriment [for health effects, ill health], and the cost of the measures. The utility based criterion compares the relevant benefits (e.g. statistical lives saved, life-years extended, reduced ill-health and better quality of life) obtained by adoption of a particular risk prevention measure with the net cost of introducing it, and requires that a balance be struck between the two. This balance can be deliberately skewed towards benefits by ensuring gross disproportion between costs and benefits.

• A technology based criterion, which essentially reflects the idea that a satisfactory level of risk prevention is attained when 'state of the art' control measures (technological, managerial, organizational) are employed to control risks, whatever the circumstances.

These criteria underlie the regulatory process first outlined by the Royal Society. The scheme is based on:

• an upper limit of risk which should not be exceeded for any individual ['unacceptable'];

• further control, so far as is reasonably practicable, making allowances if possible for aversions to the higher levels of risk or detriment ['tolerable']; and

• a cut-off in the deployment of resources below some level of exposure or detriment judged to be trivial ['broadly acceptable']. The scheme is outlined in Figure 1.

This approach to risk evaluation can be applied to health effects. For many health effects, the risk evaluation is concerned only with determining what constitutes a 'broadly acceptable' risk, and hence with the equity criterion. This is the case if any exceedance of an equity criterion for 'safe' (the 'broadly acceptable' level of risk), such as a residue level in a foodstuff, results in its immediate withdrawal from the market. However, the current approach to human medicines clearly makes use of a utility criterion, in that the risks and benefits associated with the use of a drug in patients in general are evaluated by a licensing body. In a specific patient they are evaluated by the person prescribing the drug. In other fields it can also be seen that utility/technology based criteria are employed. 'As low as is reasonably practicable' and 'best available technology' incorporate such criteria. In addition, the use of reducing limits for air quality indicates that there has been exceedance of the 'broadly acceptable' risk, and that time is needed to restore the exposures to those considered to be the maximum 'broadly acceptable' risk.