2_01 Food chemicals Toxicity Text - ILSI Globalilsisea-region.org/wp-content/uploads/sites/21/2016/08/2... · Web view2016/08/02  · , The NOAEL depends on the adverse effect endpoint

Chapter 2: Food chemicals – their toxicity

Content: Understanding the steps from intake of food chemicals, their metabolism, the observed toxic effects, assessing the severity of effects in animals and humans.

1. Absorption, distribution, metabolism, and excretion (ADME)2. Toxicokinetics 3. Toxicodynamics (Effects)4. Tests5. Dose-response observations6. Threshold vs non-threshold mode of actions7. Examples of chemicals and toxicity endpoints

1. Absorption, distribution, metabolism, and excretion (ADME)

In order to assess the risk of a food chemical on humans it is of central importance to understand some basic processes within the human organism: the absorption, distribution, metabolism and the excretion of food chemicals.

There are different ways how a chemical compound can enter the human body: through ingestion of food or water, inhalation via the lungs, epidermal through the skin, or through injection (Greim, Snyder 2008). For the chemical risk assessment of food chemicals only the adsorption of orally administered substances is of interest.Upon oral intake of food or drinks etc a chemical compound can be absorbed from the oral cavity, the stomach, the small intestine and the large intestine (Brimer 2011). The absorption can either be passive by diffusion etc. or active as facilitated by a carrier under the use of energy in the form of ATP. In the following we will look at two important factors for the absorption of food chemicals: the chemical speciation of a substance and the nutritional state of the individual (EU_Hansen_ ADME). The former is decisive for the bioavailability of a compound, thus the fraction of the administered dose, which is systematically available (Greim, Snyder 2008). An example concerning bioavailability: humans take up almost 100% of methylmercury whereas the uptake of metallic mercury is limited. In order to understand the concept of bioavailability one must understand some basic processes of human physiology concerning the ingestion of food. After ingestion the food will first reach the stomach, where the food enters an acid environment and can thus be rapidly hydrolysed if the food chemical is not stable in acid (Greim, Snyder 2008). This will result in a reduced uptake of the genuine compound. After the stomach the chemical compounds reach the intestine where they will be taken up via the intestinal walls and are transported to the liver, i.e. before entering the systemic blood circulation the blood, from the gastrointestinal organs, first flows via the portal vein to the liver. Any metabolism happing in the liver as a result of this is called the “first-pass effect”(first-pass metabolism or presystemic metabolism) and usually results in a partial removal of the administered/absorbed dosage. Because the chemical compound thus is transformed before reaching the systemic blood circulation, the bioavailability is reduced (Greim, Snyder 2008). Some substances also undergo a first biotransformation in the intestinal walls of the gastrointestinal tract (Greim, Snyder 2008). The nutritional state of an individual is especially of interest with regard to divalent metal ions like calcium, zinc and iron. Iron deficiency thus increases absorption of cadmium, lead, and aluminum. Figure 1 shows an overview of the different pathways for adsorption, distribution, metabolism and excretion of food chemicals.

1

Absorption

Bioavailability

First-pass effect

Nutritional state

Figure 1: Different pathways for adsorption, distribution, metabolism and elimination of food chemicals in the human body. (Greim, Snyder 2008), page 20.

After the absorption of the food chemicals into the blood the distribution within the body starts. Generally it can be said: In order to harm or damage an organ the food chemical needs to reach the respective organ. The main transport medium in the human body is the blood cells and the plasma. The plasma contains a lot of proteins, to which chemical compounds can bind . The degree of protein binding has been investigated thoroughly for many drugs and can vary considerably. Alternatively a chemical compound can also bind to the erythrocytes (red blood cells). An example: When the toxic heavy metal lead is introduced to the body, 99% of the lead is bound to the erythrocytes, while only 1% is found in the plasma, where most of it is bound to albumin (Brimer 2011). For the further distribution in the body the extent of protein binding is crucial because only the unbound fraction can diffuse through the capillary wall into cells and produce systemic effects (Brimer 2011). For the passive diffusion of the unbound fraction of food chemicals through membranes it is furthermore of paramount importance whether a chemical compound is hydrophilic or lipophilic. Hydrophilic compounds are highly water-soluble and will in general distribute relatively slowly into the cells because they are transported rather slowly through the phospholipid bilayer, which forms the biological membrane of tissues and organs. Lipophilic substances, which are highly lipid-soluble chemical compounds, may cross cell membranes more rapidly (Brimer 2011). Alternatively compounds can be filtered through a membrane through pores in the double lipid layer due to hydrostatic or osmotic pressure (Brimer 2011).

Beside the ability of the compound to cross a cell membrane, the blood flow to the organ in question (the perfusion) is decisive for the rate at which unbound compounds will enter the target organ (Brimer 2011). Thus, when trying to assess the distributin rate of a substance the perfusion of organs or tissues needs to be accounted for. However the perfusion of tissues and organs can vary considerably. There are well-perfused tissues like the liver, muscle and the lungs, while the adipose tissue exhibits a more restrictive blood flow (Greim, Snyder 2008). Yet the adipose tissue can store compounds effectively by dissolving them in the fatty matrix of the tissue leading to accumulation of chemical compounds. The accumulation of chemical substances can also take place in the bone marrow, or the peripheral or central nervous system (PNS and CNS). The accumulation in the fat tissue or the bone marrow is, per se, no

2

Distribution

Protein binding

Crossing of cell membrane

Perfusion

Accumulation

problem as long as the stored substances are not mobilized. However if fat is mobilized due to starvation, the concentration of the substance in the plasma can rise drastically, which can lead to acute intoxication (Brimer 2011). Examples of substances that accumulate in the fatty tissue are chlorinated pesticides (e.g. DDT), pollutants (e.g. PCBs) or metal-organic compounds like methylmercury (Brimer 2011).The accumulation of substances in the peripheral or the CNS on the other hand is always an issue (Brimer 2011). In order to quantify the distribution of a chemical compound radioactive tracers or microscopy can be used, thus lead accumulated in the kidneys may be seen as protein-lead bodies in the cells. As an example of the tissue dependent distribution of a compound Figure 2 shows the concentration of thiomebumal in the plasma, the muscles, the brain and in lipids after intravenous injection as a function of time (Brimer 2011).

Figure 2: Example for the concentration of thiomebumal in the plasma, the muscle, the brain and in lipids after intravenous injection as a function of time. (Brimer 2011) page 37.

However there are some important physiological barriers of toxicological importance that need to be mentioned: the blood-brain barrier, the blood-testis barrier and the placental barrier (Brimer 2011).

After the absorption and distribution of a substance in the human body, it is of great importance to understand the metabolism (biotransformation) of a substance. After a substance has entered the human body and is distributed it is usually metabolised by means of enzyme-catalysed chemical reactions, which may alter the structure and reactivity of the compound (Brimer 2011). Generally metabolism may occur in all cells of the body. However the main detoxification pathway in humans (and most animals) is via the liver, which is the major metabolising organ. Other important sites for metabolism are the kidneys, the lungs and the skin. Metabolism reactions can be divided in two types: phase I and phase II reactions (Brimer 2011). Phase I reactions are generally degradation reactions, where a substance is oxidised, reduced or hydrolysed. Phase II reactions are termed conjugation reactions. This kind of reaction comprises the formation of a conjugate that is biosynthesized from the toxicant or from one of its metabolites. Additionally an endogenous metabolite is involved to form the conjugate. It is possible but not necessary that a phase I reactions precedes a phase II reactions. However it is also possible that a substance undergoes only a phase I or phase II reaction before being excreted. The polarity of the genuine compound determines among others which pathway

3

Physiological barriers

Metabolism

Metabolism reactions

prevails. The rate and the type of metabolism also depend on the species, age, gender, environment, and the ingested food (Brimer 2011).

There are different ways how the metabolised compounds can be excreted: via faeces, urine, air, sweat or as constituents in hair, nails or dead skin cells (Brimer 2011). In order to be released via urine the substance has to be (or be metabolised to) a water soluble structure and pass through the kidneys. The functional units of the kidneys, the so called nephrons, in total filter about a quarter of the cardiac blood output (about 180 litre) per day) for an average adult person) and substances that are leaving the blood are excreted with the urine if not reabsorbed in the tubuli (happens e.g. for small proteins, phosphorous etc. that are essential for the body). Examples of compounds that are secreted by the kidneys are Penicillin G, atropine, and quinine (Brimer 2011). The kidneys primarily excrete small and water-soluble molecules, while the biliary excretion favours compounds with very high polarity, compounds bound to plasma proteins, and compounds with high molecular weight. The bile is produced by the liver and is then released to the small intestine. Some compounds may be reabsorbed from the small intestine to the liver and again secreted into the bile (enterohepatic circulation). If a compound does not take part in such an enterohepatic circulation it is readily excreted via the faeces. Another excretion pathway is by air via the lungs. Pulmonary excretion is mainly used for volatile metabolites of certain selenium compounds, for example of dimethylselenium (Brimer 2011); still another elimination pathway is via the mammary glands. Particularly slowly metabolized, highly lipophilic compounds are eliminated by means of the mammary glands, as are to a certain extend weak bases such as alkaloids (e.g. caffeine). The latter is due to the slightly lower pH of mother’s milk (7.0-7.4) compared to that of the organism (7.4) (Greim, Snyder 2008). The pH of cow’s milk is around 6.6. Many medicines (e.g. antibiotics) and everyday drugs (e.g. nicotine, ethanol) as well as heavy metals are to a certain extent eliminated with the milk. Thus special caution should be taken in order to spare breast-fed babies form toxic influences (Greim, Snyder 2008). For all pathways a compound may remain unchanged and leave the body by 100% while others are excreted as a mixture of the pristine compound and metabolites (Brimer 2011).

2. Toxicokinetics

What is now still lacking is more detailed knowledge concerning i) the concentration-time relationship of the chemical compound and its metabolites in different target and non-target organs, ii) the mechanisms of toxicity on cellular, subcellular and molecular levels, and iii) a description of the concentration-effect correlation for a given toxicant and its effects (Brimer 2011). The first point is covered in toxicokinetics where the uptake, the distribution, metabolism and elimination of a food chemical is of interest. The second and the third point belong to the toxicodynamics and will be discussed in the following chapter. In Figure 3 below the differentiation between toxicokinetics and toxicodynamics is illustrated.

Figure 3: Conceptual pathway of toxicokinetics and toxicodynamics (Wikipedia1).

1http://en.wikipedia.org/wiki/ File:Diagram_showing_the_conceptual_pathway_of_toxicokinetics_and_toxicodynamics.png#mediaviewer/

4

Excretion

Urine

Faeces

Lungs

Mammary glands

Toxicokinetics

Because we only discuss food chemicals we will focus on the orally administered compounds. Yet it is important to mention that the concentration-time relationship of a chemical compound within the body strongly depends on the rout of administration (Brimer 2011). Figure 4 shows an example of how the time-blood concentration curve may look for a given compound for different kinds of administration-

Figure 4: “Plasma concentration as a function of time after intravenous injection (IV), intramuscular injection (IM), subcutaneous injection (SC), or dermal (D, percutaneous) or oral (PO, per os) administration of a hypothetical compound with an adsorption fraction of 100%.” (Brimer 2011) page 54.

For the orally administered compounds it is important to account for the “first-pass effect”. As discussed before the primary compound may undergo biotransformation in the intestinal wall or the liver before they reach the general blood circulation. This is due to the fact that the blood from the gastrointestinal organs first reaches the liver via the portal vein before entering the systemic blood circulation (see Figure 5). After reaching the general circulation fractions of the chemical compound still will a re-enter the liver. (Fig. 5).

File:Diagram_showing_the_conceptual_pathway_of_toxicokinetics_and_toxicodynamics.png5

Rout of administration

Enterohepatic circulation

Figure 5: “Routes of exposure and systematic distribution of a compound within the organism.” (Greim, Snyder 2008) page 9.

In toxicokinetics we try to assess the concentration-time relationship of a compound in different organs/tissues including the blood.. For convenience The concentration is usually measured in the blood/bloood plasma and is then plotted over time in an x-y-graph. The shape of the curve for this concentration-time relationship strongly depends on both the chemical structure and the physico-chemical properties of the compound (Gupta 2007; Brimer 2011). Normally for the adsorption and the elimination of a compound first-order reactions (kinetics) are assumed. A first-order chemical reaction depends on the concentration of one reactant and the rate is proportional to the amount of this reactant. The concentration of the reactant over time usually varies between different tissues and organs (Brimer 2011).

3. Effects (Toxicodynamics)

After looking at the the fate of a given chemical compoun in the human body we will now look at the effects of this chemical in the body, the toxicodynamics. Effects of chemicals on the human body include changes in the morphology, physiology, growth, development, reproduction or life span of an organism (EU_Sharma_Intro_risk_analysis), and are differentiated in non-adverse and adverse effects. Many fluctuations in enzyme levels or other biochemical parameters, a decrease in body weight or gain due to palatability of feed, some discolorations of organs or tissue, or other effects with no statistical significance will be regarded as non-adverse.. Examples for adverse effects are cancer, damage on organs, damage on DNA (genotoxicity), damage on the central nervous system (CNS) (neurotoxicity) or damage on the reproductive system. However not all damage on organs, DNA or the CNS is irreversible (Greim, Snyder 2008). There is still the possibility that the damage is repaired or that the affected cells die without giving rise to any further disturbance (apoptosis).

In the case of primary DNA damage the cell may still repair this. If the cell is not repaired and undergoes division mutation occurs. At this point the mutated cell can still be repaired or die (apoptosis). If neither occurs there are different consequences of a DNA damage, which may either concern somatic cells or germ cells. Any cell that forms the body of an organism is a somatic cell with the expectation of germ cells, which give rise to gametes and thus the offspring of an organism that reproduces sexually. Damage in somatic cells can lead to cancer,

6

Concentration-time relationship

First-order reaction

Adverse and non-adverse effects

Genotoxicity

Somatic cells

Germ cells

Genotoxic and non-genotoxic carcinogens

Example genotoxic: PAH

premature ageing, cardio-vascular diseases or damage of the immune system. Cancer is initiated if tumour suppressor genes are affected. Damaged DNA of germ cells on the other hand can lead to heritable diseases, malformation or decreased fertility. Genotoxic effects can occur through direct and indirect pathways. For direct genotoxicity the administered chemical or its metabolites are DNA reactive and causes immediate damage. Chemicals with this mode of action are also called genotoxic carcinogens. For indirect genotoxicity first a chronic inflammation occurs which releases reactive species, which then damage the DNA. Compounds with an indirect pathway are called non genotoxic carcinogens and first need to overcome a threshold before adverse effects occur. In order to discriminate between genotoxic and non-genotoxic carcinogens genotoxicity tests can be used. An example of a substance that is genotoxic is the polyaromatic hydrocarbon (PAH) benzopyrene, which can be produced when cooking meat at high temperatures like barbequing. After entering the body the benzopyrene is transformed by enzymes (Cytochrom P450) to metabolites that react with the guanine base of the DNA and are thus carcinogenic.

Neurotoxicity concerns the adverse change in the structure or a functional alteration of the nervous system originating from exposure to chemicals, biological and physical agents. Neurotoxic damage can be reversible or irreversible and strongly depends on the dose and the duration of the exposure to a chemical, on the genetic pre disposition, the liver function and the developmental stage of the targeted organism and other factors. (Quelle: Berlin Vortrag, Nikolopoulou). The signal transporting cells of the nervous system are called neurons and are the functional units of the nervous system. In order to understand the possible toxic effects on the nervous system the fundamental structure needs be understood. Neurons are highly branched cells that consist of cell bodies and appendices; i.e. several dendrites and one axon.. Dendrites are responsible for receiving the communication from other neurons and are highly branched (Greim, Snyder 2008). Thus, each neuron receives signals from other neurons and transmits the information further through the long axon. For this communication the action potential needs to be transferred along the axon.Therefore neurons exhibit a high density of ion channels along the axon and at the synapses, which are specialized junctions through which cells of the nervous system signal to one another and to non-neuronal cells such as muscle cells. In the synapses the signal is forwarded by means of a chemical transmitter, called a neurotransmitter, which are released in the synaptic cleft and diffuse to the postsynaptic site where they bind to postsynaptic receptors. Upon binding to the postsynaptic receptors a new action potential is induced by means of a biochemical response and the information can be forwarded (Greim, Snyder 2008). Parts of the axon are surrounded by myelin layers that have high electric impedance. The myelin shields are disconnected at the Ranvier nodes, which occur every 1.5 mm along the axon (Greim, Snyder 2008). The myelin layers around the axons allow for a fast signal transmission because the electrical signals jump from one Ranvier node to another (Greim, Snyder 2008).

An impairment of the CNS, consisting of the brain and the spinal cord, is troubling to the fetal brain still under development but also to because neurons of the adult brain which are terminally differentiated cells that cannot undergo proliferative responses to repair damage. Beside the CNS there is the peripheral nervous system (PNS), which consists of sensory neurons that transmit information from peripheral receptors to the CNS and motor neurons that transfer information from the CNS to the muscles and glands. In the PNS damage can be reduced if there are surviving neurons, which expand their territory by axonal branching and can thus overtake the territory of dead neuronal cells.

Although new cells cannot be developed in the CNS the brain is protected by other adaptive mechanisms that provide the brain with considerable capacity for structural and functional modification. Dysfunctions in selected areas of the brain can be compensated for by other areas. Yet there are specialized areas of the brain where no compensation is possible.Overall it can be said that neurons are very active cells that exhibit a very high metabolic demand. Some neurons overcome very long distances with their axons and present a very effective system for

7

Neurotoxicity

Neurons

Dendrites andaxons

Synapses

Neurotransmitter

Axons

Myelin layers

Ranvier node

Impairment of CNS

Impairment of PNS

Damage of neurons

Example: lead

distributing metabolites between the cell body, the dendrites, and the axons. Neurons are thus especially susceptible to chemicals that destroy the myelin layers or the cytoskeleton of the cells, or interfere with the energy supply (Greim, Snyder 2008).

A common toxicant for nervous tissue is lead, which is known to be a toxic agent for the cognitive development of children and the mental abilities of adults by environmental, occupational or food exposure (Greim, Snyder 2008). Well known neurotoxic effects of lead are abnormal myelin formation, altered neurotransmitter release and receptor density, the disruption of the blood-brain barrier and lowered IQ.

8

Figure 6: Schematic diagram of neurons. From (Greim, Snyder 2008), page 251.

9

4. Experimental data acquisition

There are different approaches in order to measure the adverse effects of a specific chemical on a target organism. It is possible to conduct studies as experimental animal assays (in vivo) or as cell culture assays (in vitro) or to analyse epidemiological data from groups of humans; exposed e.g. through their work. The results of the experiments/studies should help to identify and risk assess the adverse effects that a chemical compound can initiate in a target organism. When talking about food constituents, the overall goal of experiments is to simulate lifelong exposure of humans to a specific chemical.

The first international organization that developed internationally agreed guidelines for the testing of chemicals with regard to their toxicity was the Organisation for Economic Co-operation and Development (OECD) in 19812 (Brimer 2011). Meanwhile guidelines were published by other organisations amongst others the US FDA. Yet most international and national organizations stipulate that their tests are performed according to the OECD guidelines for the testing of chemicals (Brimer 2011).

In the following we will focus on in vivo experimental studies, because they are frequently used when assessing the risk of food chemicals for humans. Results from suchexperimental animal studies need to be extrapolated to other species (e.g. man) or to a longer time span than the exposure time in the experiment (Gupta 2007). This is done by division with uncertainty factors.The use of animal testing for the risk assessment of food chemicals, drugs or cosmetics is increasingly criticized, however. In line with tis, the EU, to take an example, supports the development of alternative (typically in vitro methods) that can reduce, replace or refine (RRR) the use of animal experiments (Brimer 2011).

The experimental design must be adapted to the specific hypothesis that should be tested. At this point two experimental designs for in vivo experiments will be presented. For all experiments it is important to consider the rout of administration because the distribution and the effect of a toxicant may vary depending on the type of administration (see Figure 4).

Figure 7:” Schematic rendering of an experimental design for evaluating the kinetics of an administered toxicant” (Gupta 2007), page 15.

The first experimental setting here described should render results that help to understand the link between the exposure and the internal dose and can be used to better understand the

2 OECD (revised 1993) Guidelines for the Testing of Chemicals, Sections 1–5. http://www.oecd.org/env/ehs/testing/oecdguidelinesforthetestingofchemicals.htm

10

Different methods

OECD guidelines

In vivo

RRR

Experimental design

Toxicokintetics experiment

toxicokinetics of a substance (see Figure 7). Data rfom this kind of experiments is used to create a mass balance between the administered quantity and the quantity recovered. For this all the animals are exposed to a specific dose of a possible toxic substance. At different time points after the administration a certain number of experimental animals will be euthanized and tissue is collected. Like this a dynamic profile can be created of how the body handles the administered chemical (Gupta 2007). The time points when the animals are sacrificed are guided by the anticipated kinetic profile of the compound and its metabolites (Gupta 2007). A possible output of such an experiment can be seen in Figure 8, which shows an example for the blood concentration plotted against time. The graph is purely hypothetical.

Figure 8: An example for an output of an experiment to understand the toxicokinetics. Blood concentration-time curve (―) as the sum of input (- - -) and output (…) in linear coordinates. (Brimer 2011), page 55.

In order to make dose-response observations another experimental setting is needed (Gupta 2007). Also here the administration route should be chosen according to the most likely exposure conditions to be encountered; i.e. in our case exposure through food or beverages. In this example we will look at a 2-year assay that is typically used to evaluate carcinogenicity in rats or in mice. However, the same study design can be used to evaluate other endpoints and to conduct shorter-term studies (Gupta 2007).Common study durations include the evaluation of acute effects, and further study durations such as 28 days, 90 days, and maybe finally chronic studies, which last at least one year or two years for carcinogenicity in rats or mice. It is critical that multiple exposure levels are used (see Figure 9). The choice of the exposure levels is very important, because they are used for the identification of threshold values for regulatory documents and laws. In this experimental design multiple sacrifice times are used for all exposure levels (Gupta 2007). This setting can provide valuable insight into the progression of disease processes. As the name of this kind of experiments already states the output are dose-response relationships that are plotted in an x-y-graph, which are discussed in the following chapter.

11

Dose-response observations

Figure 9: "Schematic rendering of an experimental design for evaluating exposure (dose)- response relationships for a toxicant. (Gupta 2007), page 15.

Beside in vivo experiments it is common to conduct in vitro experiments, which for example can be used to test for acute general cytotoxicity and for genotoxicity (Brimer 2011). In vitro experiments are biological experiments that are conducted outside the intact live organism using cultures of isolated organs (also termed ex vivo), primary cells,cells in continuous culture or subcellular fractions (Brimer 2011). The development of in vitro methods was enhanced by the implementation of the Animal Welfare Guidelines in 1986. At this time the EU institutions started to support the development of alternatives to in vivo methods by supporting the RRR principle (see beginning of this chapter). Common effects (end-points) that are tested with in vitro methods concern cell morphology, cell viability, cell metabolism, cell membrane integrity, cell proliferation, cell adhesion but also genotoxicity (Brimer 2011). It is also possible to conduct some skin irritation tests with in vitro methods and thus replacing in vivo methods where acute dermal irritation/corrosion is tested. An example for an in vitro test for skin irritation is the non-biological test called SKINTEX TM, which is commercially available. In this test the chemical compound is administered to a membrane of collagen, keratin and colorant, which is in contact with a proteinous reagent. If the compound is irritating it may alter the membrane and provoke a colorant liberation, and/or pass through the membrane r reacting with the reagent inducing different levels of precipitation, according to the sample irritating capacity (Bason et al. 1992). Anr example of an in vitro experiment for the test of genotoxicity is the Ames test, which is used to test whether a chemical substance induces mutation. For this bacteria, usually salmonella typhimurium, are deprived of their ability to produce a specific amino acid, namely the amino acid histidine. The salmonelle typhimurium strain is then administered onto an agar plate where only very little histidine is present in the growth medium, just sufficient to allow the bacteria to grow for an initial time and have the opportunity to mutate. Some of the plates do not contain any of the chemical compound to be investigated (control plates), while two or three groups are made containing different concentrations of the compound. In case the compound is mutagenic, it will among others back-mutate the bacteria to the histidine producing “wild-type”, which within the next 48-72 hours will form visible colonies. The mutagenicity of a substance is proportional to the number of colonies observed. It is also possible to use Escherichia coli for such a test yet Salmonelle typhimurium is more suitable because it has a defect in the DNA repairmen system, thus mutations cannot be eliminated. Additionally the permeability of the membrane of Salmonelle typhimurium is higher so that mutagenic substances are not detained by it.

The third possibility to assess the toxicity of chemicals is through observations in humans/human studies(Integrated Risk Information System (IRIS) 1993). For this different methods can be used: case reports, studies in volunteers (normally only low dose studies of the kinetics), occupational experiences, and more general epidemiological studies (Nielsen et al.

12

In vitro

Skin irritation test

Ames test

Human studies

Case reports

Studies in volunteer

Occupational experiences

Epidemiological studies

2008). Human studies (data) obviate the extrapolation from animals to humans and are thus given first priority whenever available. Case reports describe particular observed effects in individuals or groups that have been exposed, usually accidentally or in suicidal attempts, to a substance (Nielsen et al. 2008). Therefore case reports are mainly used to assess acute toxic effects and clinical symptomsOccupational experiences refer to the monitoring of workers in their working environment with regard to the compliance of occupational exposure to the limits required by national laws (Nielsen et al. 2008). In epidemiological studies the distribution and determinants of health-related states and events in human population is scrutinized (Greim, Snyder 2008). It is a non-experimental approach that presumes that a certain group of individuals happened to have been exposed to a specific compound. Commonly measured endpoints are mortality, morbidity, medical visits or hospital admissions, and clinical signs and symptoms (Nielsen et al. 2008). Epidemiological studies can be used to assess long-term effects from repeated exposure for a long time or can help to reveal effects from short-term exposure.

5. Dose-response observations

In order to assess the effects of a toxic substance, dose-response relationships are plotted. The dose response curve describes the change in effect on an organism caused by different levels of exposure (dosage) to a chemical compound. The dose is plotted on the x-axis and the response or the effect on the y-axis. As the toxic effect of a substance is the function of the dose and time, the dose-response curve is referring to a certain exposure duration, and thus independent of time. The response(s) (effect(s)) is defined in each study and can be physiological or biochemical responses, or number of fatalities (mortality). The effect can furthermore be quantified at different levels (molecular, cellular, tissue or organ etc.) and may thus be expressed in different entities (Brimer 2011). A commonly used measurement for the effect is percentage of responses in the target organism. A formerly frequently used measurement is the percentage of fatalities, whereas the dose where 50% of the experimental animals died after specified test duration is termed as the LD50 (Brimer 2011). Today a number of alternative assays have been developed to reduce the number of animals and their sufferings. Figure 10 shows a hypothetical dose-response curve for two different toxicants with the dosage on the x-axis and the response in percent on the y-axis.

Figure 10: Hypothetical dose-response curves for two different toxicants, A and B. Extracted from Brimer (2011), page 66.

6. Threshold vs non-threshold mode of actions13

Dose-response relationship

Normally we differentiate between two modes of action that can lead to adverse effects: the direct and the indirect pathways. The direct mode of action has no threshold and can be displayed as a linear relationship between the dose of a food chemical and the adverse effects (see figure 11, left graph). This relationship describes the mode of action of direct genotoxicity, where a small number of molecular events can evoke changes in a single cell that can lead to uncontrolled cellular proliferation – cancer (Brimer 2011).

In the indirect mode of action the organic homeostatic, compensating and adaptive mechanisms need to be overcome before a toxic effect occurs. Thus there is a threshold that needs to be reached before adverse effects take place (Brimer 2011). These kinds of effects are also referred to as “systemic toxicity”. They describe effects other than gene mutation related cancer o, which on the other hand are often treated as non-threshold processes as described above (Integrated Risk Information System (IRIS) 1993).

7. Examples of chemicals and toxicity endpoints for risk assessment

Until now we have looked at several basic toxicological principles and assay methods. In this last chapter an important output of the risk assessment is discussed. For the utilisation of the bulk of results obtained from different assays and observations (e.g. case reports etc.) for regulatory purpose,s it is important to generate key figures. There are different measures for each toxicity endpoint that can be used in this regard.

The most prominent through time one is probably the NOAEL (No Observed Adverse Effect Level.), The NOAEL depends on the adverse effect endpoint of the experiment. For example: If the endpoint of an experiment is skin irritation and the subject under study exhibits skin irritation at the dosage of 50 ppm but none at the next lower level of 30 ppm, then the NOAEL is 30 ppm. Normally the NOAEL is determined for different endpoints (e.g. skin irritation, genotoxicity, etc.) for the same chemical compound. In the end the lowest NOAEL of all the experiments is chosen (Brimer 2011). The NOAEL can only be determined for chemical compounds that follow a threshold mode of action (United States Environmental Protection Agency (EPA) 2012).

As an example to illustrate the above discussed issues we can take a look at the results of a study conducted by Littlefield et al. (1980) where over 24’000 mice and 81 different treatment groups were used to determine the shape of the dose-response relationship for two different adverse effects of the exposure to 2-acetylaminofluorine (2-AAF) (Casarett et al. 2001). The mice were exposed to 2-AAF to one of seven doses between 30 to 150 ppm (in the food) plus one control group with 0 ppm. Figure 12 shows the dose-relationship between 2-AAF exposure and liver and bladder cancer at 24 months of exposure. Although for both types of tumours the incidents increase with increasing dose the shape of the curves are dramatically different. For liver tumours no evident threshold can be determined, whereas for bladder tumours an apparent threshold is evident. For the bladder tumours the NOAEL was set at 75 ppm.

14

effect

dose

effect

dose

threshold

Figure 11: Dose-effect relationship without (left) and with a threshold (right).

Non-threshold mode of action

Threshold mode of action

NOAEL

Figure 12: Dose-response relationship for carcinogens adapted from (Littlefield et al. 1980), extracted from (Casarett et al. 2001),page 22.

If the NOAEL is not available the Lowest Observed Adverse Effect Level (LOAEL) can be used. Both methods have the advantage that they do not depend on a mathematical model, are easy to understand and that it can be applied to all data. On the other hand they only provide knowledge at the dose level of the experimental study and are thus strongly dependent on the study design.

Figure 13: Illustration of the NOAEL, the LOAEL and the benchmark dose (BMD) approach. Extracted from Brimer (2010) page 261.

Alternatively the benchmark dose (BMD) can be assessed, based on the statistically best-fitting dose-response curve derived from the experimental results. . The BMD is the dose that produces a predetermined change (usually 5-10%) in the rate of an adverse response compared with the background level. The mathematical model applied is fitting the experimental

15

LOAEL

Benchmark dose

data within the observable range and refers to the central estimate of the dose that is expected to yield the benchmark response (BMR). The statistical method applied in the BMD approach thus uses the information in the complete dataset instead of making pair wise comparison using subsets of the data (Brimer 2011). The BMD is thus less dependent on the experimental setting than the NOAEL and is thus conceptually superior.

The determined NOAEL or BMD(L) can be used in the deviation (an estimate) of the acceptable daily intake (ADI) for humans of a substance. For this, the selected NOAEL (BMDL) is divided by an uncertainty factor (range 10 -100). The lowest NOAEL (BMDL) of the different experiments within a study should be chosen. For example: In a study the NOAEL was determined for the effects on the liver and the germ cells,for the testis weight reduction, and for reduced F2 generation. The NOAEL for the liver experiment was the lowest, thus the calculation of the ADI should be based upon this NOAEL (Brimer 2011). The ADI represents the maximum dose that is believed a person can ingest throughout his/her life without any adverse effects.

16

ADI

Publication bibliography

Bason, M. M.; Harvell, J.; Realica, B.; Gordon, V.; Maibach, H. I. (1992): Comparison of in vitro and human in vivo dermal irritancy data for four primary irritants. In Toxicol In Vitro 6 (5), pp. 383–387.

Brimer, Leon (2011): Chemical food safety. Nosworthy Way Wallingford Oxfordshire UK ;, Cambridge MA: CABI (Modular texts).

Casarett, Louis J.; Doull, John; Klaassen, Curtis D. (2001): Casarett and Doull's toxicology. The basic science of poisons. 6th ed. New York: McGraw-Hill Medical Pub. Division.

Greim, Helmut; Snyder, Robert (2008): Toxicology and risk assessment. A comprehensive introduction / edited by Helmut Greim and Robert L. Snyder. Chichester: John Wiley.

Gupta, Ramesh C. (2007): Veterinary toxicology. Basic and clinical principles / Ramesh C. Gupta, editor. Amsterdam ; London: Academic.

Integrated Risk Information System (IRIS) (1993): Reference Dose (RfD): Description and Use in Health Risk Assessments. Background Document 1A. Environmental protection agengy of the United States. Available online at http://www.epa.gov/iris/rfd.htm, updated on 31/10/2014, checked on 13/01/2015.

Littlefield, N. A.; Farmer, J. H.; Gaylor, D. W.; Sheldon, W. G. (1980): Effects of dose and time in a long-term, low-dose carcinogenic study. In J Environ Pathol Toxicol 3 (3 Spec No), pp. 17–34.

Nielsen, Elsa; Østergaard, Grete; Larsen, John Christian (2008): Toxicological risk assessment of chemicals. A practical guide. New York: Informa Healthcare.

United States Environmental Protection Agency (EPA) (2012): EPA Risk Assessment - Human Health Risk. Step 2 - Dose-Response ASsessment. United States Environmental Protection Agency (EPA. Available online at http://www.epa.gov/risk_assessment/dose-response.htm, updated on 31/07/2012, checked on 23/01/2015.

17