Several factors, both of a scientific and socio-economic nature, may influence the establishment of mycotoxin limits and regulations. These include:
availability of toxicological data;
availability of data on the occurrence of mycotoxins in various commodities;
knowledge of the distribution of mycotoxin concentrations within a lot;
availability of analytical methods;
legislation in countries with which trade contacts exist; and
need for sufficient food supply.
The first two factors provide the necessary information for hazard assessment and exposure assessment respectively, the main ingredients for risk assessment. Risk assessment is the scientific evaluation of the probability of occurrence of known or potential adverse health effects resulting from human exposure to food-borne hazards; it is the primary scientific basis for the establishment of regulations.
Regulations are primarily made on the basis of known toxic effects. For the mycotoxins currently considered most significant - aflatoxins, ochratoxin A, patulin, fumonisins, zearalenone and some trichothecenes including deoxynivalenol - the Joint Expert Committee on Food Additives (JECFA), a scientific advisory body of FAO and WHO, has recently evaluated their hazards. JECFA provides a mechanism for assessing the toxicity of food additives, veterinary drug residues and contaminants. Safety evaluation of contaminants incorporates various steps in a formal health risk assessment approach.
The qualitative indication that a contaminant can cause adverse effects on health (hazard identification) is usually included in the information presented to JECFA for evaluation. Similarly, qualitative and quantitative evaluation of the nature of the adverse effects (hazard characterization) is embodied in the data sets that are presented. The evaluation of toxicological data carried out by JECFA normally results in the estimation of a Provisional Tolerable Weekly Intake (PTWI) or a Provisional Tolerable Daily Intake (PTDI).
The use of the term "provisional" expresses the tentative nature of the evaluation in view of the paucity of reliable data on the consequences of human exposure at levels approaching those with which JECFA is concerned. In principle, the evaluation is based on the determination of a No-Observed-Adverse-Effect-Level (NOAEL) in toxicological studies and the application of an uncertainty factor. The uncertainty factor means that the lowest NOAEL in animal studies is divided by 100, 10 for extrapolation from animals to humans and 10 for variation between individuals, to arrive at a tolerable intake level. In cases where the data are inadequate, JECFA uses a higher safety factor.
This hazard assessment approach does not apply for toxins where carcinogenicity is the basis for concern as is, for example, the case with the aflatoxins. Assuming that a no-effect concentration limit cannot be established for genotoxic compounds, any small dose will have a proportionally small probability of inducing an effect. Imposing the absence of any amount of genotoxic mycotoxins would then be appropriate, if these toxins were not natural contaminants that can never completely be eliminated without outlawing the contaminated food or feed. In these cases, JECFA does not allocate a PTWI or PTDI. Instead it recommends that the level of the contaminant in food should be reduced so as to be As Low As Reasonably Achievable (ALARA). The ALARA level, which may be viewed as the irreducible level for a contaminant, is defined as the concentration of a substance that cannot be eliminated from a food without involving the discard of that food altogether or without severely compromising the availability of major food supplies. This covers the case of the JECFA evaluations of the aflatoxins made in 1987 and 1997. On some occasions in the 1990s, JECFA also evaluated the risk of other mycotoxins: ochratoxin A, patulin and zearalenone.
In February 2001, a special JECFA session was completely devoted to mycotoxins (FAO, 2001; WHO, 2002b). The mycotoxins evaluated or re-evaluated at this 56th JECFA meeting included fumonisins B1, B2 and B3, ochratoxin A, deoxynivalenol, T-2 and HT-2 toxins, and aflatoxin M1. The report addressed several concerns about each mycotoxin including explanation of the mycotoxin, absorption through excretion, toxicological studies and final evaluation. Along with the mycotoxin evaluations, the committee put forth general considerations on analytical methods, sampling, associated intake issues and control.
The evaluation of aflatoxin M1 is the more interesting as JECFA responded to a request by the Codex Committee on Food Additives and Contaminants (CCFAC, see also Section 3.5.5.) at its 32nd session (CAC, 2000) to "examine exposure to aflatoxin M1 and to conduct a quantitative risk assessment" to compare the application of two standards for contamination of milk (0.05 mg/kg and 0.5 mg/kg), limits that are currently applied in the European Union (EU) and the United States respectively. The calculations showed that, with worst case assumptions, the projected risks for liver cancer attributable to use of the proposed maximum levels of aflatoxin M1 of 0.05 mg/kg milk and 0.5 mg/kg milk are very small, and that there is no significant health benefit when a 0.5 mg/kg limit would be reduced to 0.05 mg/kg.
In the further development of tolerable daily intake (TDI) levels for mycotoxins in food for national or international (Codex Alimentarius) purposes, factors other than hazard assessment play a role. These will be discussed below.
In addition to information about toxicity, exposure assessment is another main ingredient of the risk assessment. Reliable data on the occurrence of mycotoxins in various commodities and data on food intake are needed to prepare exposure assessment. The quantitative evaluation of the likely intake of mycotoxins is quite difficult. At its 56th Meeting, JECFA stressed the importance of the use of validated analytical methods and the application of analytical quality assurance (see also Section 2.4 on methods of analysis) to ensure that the results of surveys provide a reliable assessment of intake (WHO, 2002b).
In most of the JECFA reviews of mycotoxins, the analytical data on the levels of contamination were often inadequate from developed countries and non-existent for developing countries. Because most mycotoxin contamination is heterogeneous, sampling is another important consideration in the development of information on the levels of contamination (Page, 2003) (see also Section 2.3 on sampling procedures).
In the EU, efforts to assess exposure are undertaken within Scientific Cooperation on Questions relating to Food (SCOOP) projects, funded by the European Commission. The SCOOP projects are targeted to make the best estimates of intake of several mycotoxins by EU inhabitants. In the 1990s, these activities resulted in a report on the exposure assessment of aflatoxins (European Commission, 1997). SCOOP reports were later published for several other mycotoxins including: ochratoxin A (Miraglia and Brera, 2002); patulin (Majerus and Kapp, 2002); and several Fusarium toxins, trichothecenes, fumonisins and zearalenone (Gareis et al., 2003). The SCOOP data have been used by the European Food Safety Authority (EFSA) for its evaluation and advisory work on the risks to public health arising from dietary exposure to certain mycotoxins.
The distribution of the concentration of mycotoxins in products is an important factor to be considered in establishing regulatory sampling criteria. The distribution can be very heterogeneous, as is the case with aflatoxins in peanuts. The number of contaminated peanut kernels in a lot is usually very low, but the contamination level within a kernel can be very high. If insufficient care is taken for representative sampling, the mycotoxin concentration in an inspected lot may therefore easily be wrongly estimated. Also, consumption of peanuts could lead to an accidental high single dose of aflatoxins, rather than a chronic intake at a relatively low level.
A similar situation could occur with pistachio nuts and figs. The risk to both consumer and producer must be considered when establishing sampling criteria for products in which mycotoxins are heterogeneously distributed. The design of sampling procedures has been an international concern for several years (FAO, 1993; CAC, 2000). Working groups and discussions are being organized by FAO and the Codex Alimentarius Commission in an attempt to find a harmonized international approach.
Examples of official sampling plans for mycotoxins are those for aflatoxins in peanuts and corn carried out in the United States (Food and Drug Administration, 2002) and for peanuts in the EU (European Commission, 2002b). In the United States, the USDA requires three 22 kg laboratory samples to average less than 15 mg total aflatoxins/kg for acceptance. In the EU, one 30 kg laboratory sample is required to test less than 15 mg total aflatoxins/kg for raw peanuts destined for further processing, and three 10 kg laboratory samples to all test less than 4 mg total aflatoxins/kg (and 2 mg aflatoxin B1/kg) for finished peanuts sold for direct human consumption.
Although the approaches are different, the United States peanut industry, in cooperation with USDA, has recently developed an Origin Certification Program (OCP) with several key EU countries that import United States peanuts into Europe. These key markets, in a memorandum of understanding, have agreed to recognize the sampling and testing of United States peanuts for aflatoxin before being exported to these markets (Trucksess et al., 2003). Documents showing positive lot identification and aflatoxin test results can be used to certify that the peanuts meet EU aflatoxin regulations.
In the OCP, the United States exporter uses a first 22 kg sample test result for screening lots. A second USDA 22 kg sample is tested according to EU protocol for lot certification. The OCP will reduce lots rejected at the port of entry, reduce the disruption in supply for the importer, reduce economic losses for the exporter and the importer, and maintain EU standards for consumer safety. The OCP is an example of an agreement between two countries that is mutually beneficial to both while maintaining high standards for consumer safety (Adams and Whitaker, 2004).
Food legislation calls for methods of control. Reliable analytical methods will have to be available to make enforcement of the regulations possible. Tolerance levels that do not have a reasonable expectation of being met are both wasteful in the resources that they utilize, and may well condemn products that are perfectly fit for consumption (Smith et al., 1994). In addition to reliability, simplicity is desired, as it will influence the amount of data that will be generated and the practicality of the ultimate measures taken. The reliability of analysis data can be improved through use of methods that fulfil certain performance criteria (as can be demonstrated in interlaboratory studies).
The Association of Official Analytical Chemists (AOAC International) and the European Standardization Committee (CEN), the European equivalent of ISO, have a number of standardized methods of analysis for mycotoxins that have been validated in formal interlaboratory method validation studies, and this number is gradually growing. The latest edition of Official Methods of Analysis of AOAC International (Horwitz, 2000) contains approximately 40 validated methods for mycotoxin determination, and a recent review has been published about the validation of methods of analysis for mycotoxins (Gilbert and Anklam, 2002).
CEN has produced a document that provides specific criteria for various mycotoxin methods that can be used for official purposes (Comité Européen de Normalisation, 1999). This document presents information concerning method performance, which can be expected from experienced analytical laboratories. The CEN criteria are currently reflected as method performance requirements in official EU legislation on aflatoxins, ochratoxin A and patulin (European Commission, 1998; 2002a; 2003a). They are expected to appear as well in future EU legislation for other mycotoxins in food and feed.
In addition to the use of analytical methods of demonstrated reliability, the application of analytical quality assurance (AQA) procedures is recommended including the use of (certified) reference materials, especially when a high degree of comparison and accuracy is required. Further developments in AQA and the use of reference materials are likely to emerge in the future for the control of mycotoxins in foods. Several (certified) reference materials for mycotoxins have been developed in projects funded by the European Commissions Standards, Measurements and Testing (SMT) Programme (previously known as the Bureau Communautaire de Référence [BCR]), or are currently being redeveloped (Josephs et al., 2004). In Table 1 (in Annex) an overview is given of the BCR (certified) mycotoxin reference materials that have been developed since the 1980s. The mycotoxin reference materials are now worldwide available through the European Commissions Joint Research Centre/Institute for Reference Materials and Measurements (JRC/IRMM)[1].
Certified reference materials are relatively expensive because of the enormous amount of time and money invested in their development, and current supplies are limited. Therefore, laboratories are advised to develop their own reference materials for routine use, the toxin content of which should be established on the basis of the certified materials.
Besides the application of (certified) reference materials, regular participation in interlaboratory comparisons such as proficiency testing schemes is becoming increasingly important as part of AQA measures that a laboratory must undertake to demonstrate acceptable performance. A number of proficiency testing schemes for mycotoxins exist at the international level including: i) those organized by the Food Analysis Performance Assessment Scheme (FAPAS®) operated by the Central Science Laboratory in the United Kingdom of Great Britain and Northern Ireland (Richard et al., 2003); and ii) those organized by the American Oil Chemists Society (AOCS) based in the United States (AOCS, 2003).
Good analytical methodology and AQA are prerequisites for adequate food law enforcement. Also important, especially in free trade areas, is the way in which enforcement bodies handle an issue as measurement uncertainty. Within the EU and the European Free Trade Area, approaches are not yet harmonized between countries, which may lead to different action levels, e.g. for aflatoxins. Therefore, the Food Law Enforcement Practitioners (FLEP) Working Party on "Mycotoxins" has recommended a uniform approach (Jeuring, 2004). It is of interest to note that the new EU Commission Directive on patulin in foodstuffs, which came into force in November 2003, gives some guidance about how to deal with measurement uncertainty (European Commission, 2003a). So far this is rather unique but the issue of measurement uncertainty is expected to become part of more regulatory documents in the near future.
Preferably, regulations should be brought into harmony with those in force in other countries with which trade contacts exist. In fact, this approach has been applied in the regions of Australia and New Zealand, the EU and MERCOSUR (Mercado Comun del Sur), where harmonized regulations for some mycotoxins now exist. Strict regulative actions may lead importing countries to ban or limit the importing of commodities such as certain food grains, which can cause difficulties for exporting countries in finding or maintaining markets for their products. For example, the stringent regulations for aflatoxin Bl in animal feedstuffs in the EU (Commission of the European Communities, 1991) led European animal feed manufacturers to switch from groundnut meal to other protein sources to include in feeds; this had an impact on the export of groundnut meal of some developing countries (Bhat, 1999).
The distortion of the market caused by regulations in importing countries may lead to export of the less contaminated foods and feeds leaving those inferior foods and feeds for the local market. Some countries apply different limits for aflatoxins in certain products depending on the destination.
The World Bank has published a study on impact of the adoption of international food safety standards, and the harmonization of standards, on global food trade patterns (Wilson and Otsuki, 2001). Several scenarios led to estimates of the effects of aflatoxin regulatory standards in 15 importing countries (including four developing countries) on exports from 31 countries (including 21 developing countries). In one of the scenarios, the authors examined trade flows when all countries would adopt an international standard for aflatoxin B1 in food at 9 mg/kg (equivalent to the Codex guidelines of 15 mg/kg for total aflatoxins) in contrast to all importing countries remaining at the (generally lower) limits of 1998. This would lead to an increase of the cereal and nut trade among these countries by US$6.1 billion (or 51 percent).
The regulatory philosophy should not jeopardize the availability of some basic commodities at reasonable prices. Especially in the developing countries, where food supplies are already limited, drastic legal measures may lead to lack of food and to excessive prices. At the time of writing, for instance, the dramatic food security situation in parts of Africa leads to measures that prioritize food sufficiency above food safety. Mycotoxins are an important problem as evidenced by occasional outbreaks of human mycotoxicoses and the role of aflatoxins in liver cancer in West Africa and fumonisins in oesophageal cancer in South Africa (Shephard, 2004).
Weighing the various factors at the interface of science, food security and regulations is not a trivial activity, and common sense is a major factor for reaching a decision. Public health officials are confronted with a complex problem: mycotoxins, and particularly the carcinogenic mycotoxins, should be excluded from food as much as possible. Since the substances are present in foods as natural contaminants, however, human exposure cannot be completely prevented, and exposure of the population to some level of mycotoxins has to be tolerated. Despite the dilemmas, mycotoxin regulations have been established during the past decades in many countries, and newer regulations are still being drafted.
[1] See http://www.irmm.jrc.be |