3. The toxicological evaluation of compounds on the agenda.(EVALUATION OF CERTAIN FOOD CONTAMINANTS)

The Committee considered three food contaminants for the first time and re-evaluated three others. Information on the safety evaluations is summarized in Annex 2.

3.1 Acrylamide

Explanation

Acrylamide (C[H.sub.2]=[CHCONH.sub.2] CAS Registry Number 79-06-1) is an important industrial chemical that has been used since the mid 1950s as a chemical intermediate in the production of polyacrylamides, which are used as flocculants for clarifying drinking-water and in other industrial applications. It is well established that acrylamide is neurotoxic in humans, as revealed by the consequences of occupational and accidental exposures. In addition, experimental studies in animals have shown that acrylamide has reproductive toxicity, and is genotoxic and carcinogenic.

Studies conducted in Sweden in 2002 showed that high concentrations of acrylamide are formed during the frying or baking of a variety of foods. Owing to concerns about the possible public health risks associated with dietary exposure to acrylamide, a consultation was held by FAO/WHO in June 2002 (7). On the basis of the recommendations arising from this consultation, numerous studies of metabolism, bioavailability, toxicokinetics, DNA adduct formation, and mutagenicity in vitro and in vivo have been performed. Concurrently, a major worldwide effort has produced extensive survey data that can be used to estimate the extent and levels of contamination in food and to estimate national intakes.

At its present meeting, the Committee responded to a request from the Codex Committee on Food Additives and Contaminants (CCFAC) at its Thirty-sixth Session (8) to:

* comment on the extent to which acrylamide is bioavailable in food and on the safety implications;

* consider the threshold-based end-points of concern, such as neurotoxicity and reproductive toxicity, and eventually derive a tolerable dietary intake;

* evaluate the degree of uncertainty related to the assessments made;

* provide estimates of dietary intake for various population groups, including susceptible groups such as young children and regional populations, and to identify and quantify as far as possible the major sources of dietary intake;

* provide estimates and MOEs, safety, and intake for various end-points of concern (non-cancer and cancer). These estimates should contain comparisons between the levels of exposure shown to produce effects in animal studies and demonstrated no-effect levels versus estimates of dietary intake for humans;

* provide quantitative estimates of risk for various end-points, including cancer, for varying degrees of dietary exposure to acrylamide; and

* provide comments on the toxicological significance of the main metabolite, glycidamide, and whether this may be more genotoxic than the parent compound.

Acrylamide has not been evaluated previously by the Committee.

Absorption, distribution, metabolism, and excretion

In animals, acrylamide administered orally is rapidly and extensively absorbed from the gastrointestinal tract and is widely distributed to the tissues, as well as the fetus. It has also been found in human milk. Acrylamide is metabolized to a chemically reactive epoxide, glycidamide, in a reaction catalysed by cytochrome P450 2E1 (CYP2E1). An alternate pathway for the metabolism of acrylamide is conjugation with glutathione. Acrylamide and its metabolites are rapidly eliminated in the urine, primarily as mercapturic acid conjugates of acrylamide and glycidamide. The absolute bioavailablity of acrylamide (i.e. the fraction entering the circulation as parent compound) is in the range of 23% to 48% in rodents for a dose of 0.1 mg/ kg of body weight administered in the diet over a period of 30min. The relative internal exposure to glycidamide is much higher after dietary administration than after intravenous administration, owing to extensive first-pass metabolism of acrylamide to glycidamide.

Glycidamide is much more reactive than acrylamide with DNA, and several purine base adducts have been identified in vitro. Studies in knockout and wild-type mice have shown that CYP2El-mediated oxidation is the predominant pathway leading to the formation of glycidamide--DNA adducts. In rodents given acrylamide, glycidamide--DNA adducts are formed at comparable levels in all tissues examined and accumulate to apparent steady-state levels after regimens involving repeated dosing. DNA adducts have been found in the liver, lung, testis, leukocytes, and kidney of mice, and in the liver, thyroid, testis, mammary gland, bone marrow, leukocytes, and brain of rats treated with either acrylamide or glycidamide. The formation of DNA adducts in mice shows a monotonic dependence on the dose of acrylamide administered, from measurable levels of adduct at background exposure, to evidence for saturation of levels of adduct at higher doses. Kinetic studies of adduct loss from DNA in vitro and in vivo showed that spontaneous depurination, as opposed to active repair, is operative.

Both acrylamide and glycidamide also bind covalently to amino acids in haemoglobin, and adducts with the N-terminal valine residue have been widely used to estimate internal exposures in biomonitoring studies in humans. Preliminary studies measuring concentrations of acrylamide--haemoglobin and glycidamide--haemoglobin adducts in rodents and humans with background exposure to acrylamide in the diet suggested that there may be species differences in the relative formation of glycidamide, (relative formation, mice > rats > humans). However, the long half-life of haemoglobin means that the measured levels of adduct reflect a time-weighted average over the lifetime of the erythrocyte. Thus the same total exposure over an extended period of time or over a short period of time could produce similar levels of adducts. This has limited the usefulness of these biomarkers for dose--response modelling under circumstances where there is variability in the magnitude and frequency of exposure.

Toxicological data

Single oral doses produced acute toxic effects only at doses of >100 mg/kg of body weight, and reported median lethal doses ([LD.sub.50]s) are generally >150 mg/kg of body weight.

Numerous studies in a number of animal species have shown that the nervous system is a principal target site for the toxic effects of acrylamide. Sufficient repeated exposure to acrylamide causes degenerative peripheral nerve changes that result from an accumulation of damage at the sites of toxicity (Table 1). For example, the same degree of neurotoxicity was observed in rats given acrylamide at a dose of 50 mg/kg of body weight per day by intraperitoneal administration for 11 days and in rats given drinking-water containing acrylamide at a dose of 21 mg/kg of body weight per day for 40 days. Continued dosing with acrylamide has been shown to induce degeneration of nerve terminals in brain areas (i.e. cerebral cortex, thalamus, and hippocampus) critical for learning, memory and other cognitive functions, and these lesions may precede the morphological changes in nerves. In rats given drinking-water containing acrylamide for 90 days, the NOEL was 0.2 mg/kg of body weight per day for morphological changes in nerves, detected by electron microscopy, and no exposure-related non-neoplastic lesions were found in other tissues at doses of <5 mg/kg of body weight per day.

In studies of reproductive toxicity, male rodents given acrylamide showed reduced fertility, dominant lethal effects, and adverse effects on sperm count and morphology at oral doses of [greater than or equal to] 7 mg/kg of body weight per day. In female rodents, no adverse effects on fertility or reproduction were observed, apart from slight reductions in the body weight of rat offspring at oral doses of 2.5 mg/kg of body weight per day (the lowest-observed-effect level, LOEL) and above. In studies of developmental toxicity, acrylamide was fetotoxic in mice only at a maternally toxic oral dose of 45 mg/kg of body weight per day, and was not teratogenic in mice or rats. In a study of developmental neurotoxicity, in which rats were given acrylamide orally from day 6 of gestation until day 10 of lactation, the NOEL for developmental neurotoxicity was 10 mg/kg of body weight per day. The overall NOEL for reproductive and developmental effects was 2 mg/kg of body weight per day.

Genotoxicity

Although acrylamide was not mutagenic in the Ames assay in Salmonella, glycidamide clearly was. Acrylamide was both clastogenic and mutagenic in mammalian cells in vitro and in vivo. In addition, studies of dominant lethality have shown that acrylamide is a germ cell mutagen in male rodents. The mutational spectra produced by acrylamide and glycidamide in transgenic mouse cells are consistent with the formation of promutagenic purine DNA adducts in vivo.

The metabolism of acrylamide to glycidamide appears to be a prerequisite for the genotoxicity caused by acrylamide in vitro and in experimental animals. Studies using knockout and wild-type mice showed that CYP2E1-mediated oxidation is the predominant pathway leading to the formation of DNA adducts. Estimates of internal exposures to glycidamide, based on measurements of haemoglobin adducts after administration of either acrylamide or glycidamide, indicated that glycidamide was the active clastogen responsible for induction of micronuclei in mice. Studies in wild-type and CYP2E1 knockout mice have also shown that glycidamide is the active metabolite of acrylamide responsible for germ cell mutations and dominant lethality in spermatids of male mice. Glycidamide is the presumed active mutagen because dosing with glycidamide produced increases in the frequency of mutation at the Hprt and cII loci in Big Blue transgenic mice that were comparable to or greater than those resulting from dosing with acrylamide.

Carcinogenicity

Acrylamide, administered in drinking-water, has been tested for carcinogenicity in two experiments in Fischer 344 rats. There were increases in the incidence of tumours at a variety of sites (see Tables 2 and 3). Information about the total number of tumour-bearing animals was not available for either study.

Acrylamide was evaluated by the International Agency for Research on Cancer (IARC) in 1994 and classified as "probably carcinogenic to humans (IARC Group 2A)" (16) on the basis of a positive result in a bioassay for cancer (Table 2), supported by evidence that acrylamide is efficiently biotransformed to a chemically reactive genotoxic metabolite, glycidamide, in both rodents and humans. The endocrine-responsive nature of several tumour sites from the two long-term bioassays with acrylamide in F344 rats has elicited speculation about neuroendocrine-mediated mechanisms. However, no published studies have linked hormonal changes with the carcinogenicity of acrylamide in any tissue, nor is there any indication of hormonal effects in studies of reproductive toxicity. Moreover, the wide body of evidence supporting a genotoxic mechanism is not incompatible with hormonal dysregulation by acrylamide, because it is clear that other factors beyond DNA damage are probably required for the observed target tissue specificity of tumourigenesis caused by acrylamide.

Observations in humans

Epidemiological studies in humans exposed in industry or accidentally suggest that the nervous system is a principal target site for toxicity caused by acrylamide in humans.

In workers exposed occupationally to acrylamide, exposure was not associated with an increase in overall mortality caused by cancer, nor with any statistically significant dose-related increase in risk of cancer at any organ site, except for a statistically significant doubling of risk for pancreatic cancer in workers with the highest cumulative exposure. These studies, however, were based on small numbers of cases, measurements of dietary intake of acrylamide were not made and potential confounders, such as tobacco smoking, were not considered.

The only information available that included dietary intake of acrylamide came from case-control studies originally designed to assess the potential risk of cancer associated with dietary factors other than acrylamide. The available results from epidemiological studies that estimated oral exposure to acrylamide were not suitable for use in risk assessment for acrylamide.

The formation of acrylamide adducts to haemoglobin has been used as a biomarker of exposure in humans. Although levels of acrylamide adducts were often found to be higher among exposed workers and smokers, and there was a positive correlation with the amount of tobacco product smoked, some uncertainties remained that precluded use of this measure as a marker of dietary intake of acrylamide at the present time. Because analytical methods may vary between laboratories, there is a need for improved and validated analytical methodology. At the time of the present meeting, it was not possible to link biomarkers of exposure to acrylamide in humans with measurements …