A Framework for Assessing and Prioritising Mycotoxin Risks
Background
Mycotoxins are naturally-occurring toxic metabolites produced by fungi (moulds) and can be formed during the growth and storage of food and feed, particularly raw foods. It is estimated that about a quarter of the world’s food crops (WHO/FAO Food risk analysis guidelines ), including many basic foods, are affected by mycotoxin-producing fungi. Many mycotoxins are stable and can be carried over into processed foods, which presents human health risks depending on toxicity, contamination level and the amount of contaminated food that is consumed. Mycotoxins are often associated with particular food types, but, because they occur naturally, it is impossible to impose a total ban. However, although EU and UK regulators recommend that sampling is risk-based, none provide clear guidance on the actual number of checks that should be performed. Guidance will be given here on what aspects food and feed businesses need to include in a robust risk-based sampling scheme for mycotoxins.
Food legislation in the EU demands food businesses to be responsible for the safety of the food and feed they sell, and not to place on the market unsafe (including mycotoxin-contaminated) feed or food. These businesses are also required to identify and review the risks associated with mycotoxins, and, where practicable, put in place processes and controls to reduce these risks. Since a zero-tolerance for mycotoxins in foods and feeds is impractical, legislation has been set in terms of maximum limits (maximum residue limits, MRLs) for specific mycotoxins in certain food and feed (Worldwide regulations for mycotoxins in food and feed 2003). These levels are set for mycotoxins with the greatest health concern and are based on scientific advice. The aim of maximum limits is to minimise human exposure and the risks of both acute and long-term adverse health effects and to support international trade.
Food regulators monitor food safety via a number of control and response procedures based on random checks, including the Rapid Alert System for Food and Feed (RASFF) in the EU (RASFF website ). Additionally, food and feed operators need to demonstrate that their raw materials and finished products are safe, comply with appropriate maximum limits and meet customer specification. However, it can be difficult to prioritise mycotoxins for analysis and set appropriate sampling rates, especially for smaller business that lack food chemical risk assessment professionals.
This document aims to provide small and medium size food and feed businesses with an overview of food safety risk analysis and its relation to mycotoxins. It then outlines published guidelines and other sources of information that can be used in a practical way within a risk assessment to rank the risks associated with mycotoxins and derive sampling schedules to verify food safety. The aim of this document is to highlight important features to consider in developing a risk-based sampling strategy, not to prescribe a sampling rate, since this is best met on an individual business basis in relation to the acceptance of risk and budget.
Food safety risk analysis
It is crucial to document mycotoxin risk analysis (or other chemical and microbiological contaminants) since this forms the foundation for prioritisation. The Food and Agriculture Organisation of the United Nations (FAO) and the World Health Organisation (WHO) have developed benchmarks for chemical and microbiological food safety risk assessment for national authorities ((WHO/FAO Food risk analysis guidelines ). It describes the three stages of risk analysis: risk assessment, risk management and risk communication. Risk assessment involves hazard identification (i.e. type of mycotoxin), hazard characterization (including toxicity and dose-response), exposure assessment (including occurrence in food, levels and amount of food consumed) and risk characterisation to assess food safety risks. Using this information risk management strategies can be put in place and their performance evaluated. The final stage of risk analysis - risk communication - needs careful consideration as to its aims, which should cover the information to put across and the intended audience.
Many of these aspects are covered in more detail in other factsheets on this website, which should be consulted as part of a comprehensive risk analysis.
In terms of mycotoxin risk analysis, most compounds are already recognised as food safety issues and there are several published risk assessments;
EFSA Risk Assessment for ochratoxin A
EFSA risk assesssment for deoxynivalenol
EFSA risk assessment for zeralenone
EFSA risk assessment for fumonisins
These documents are comprehensive and often highly technical. However, they can be can be used as source of information for assessing mycotoxin risks and for identifying data of particular relevance to individual food or feed business.
The next stage is to use this information in conjunction with data on mycotoxin occurrence and content, including risk management policies and procedures, to develop a risk-based sampling programme to prioritise mycotoxins/food combinations and set sampling rates for verification of food safety.
The following sections outline a framework and the aspects to consider when developing a risk-based sampling plan.
Factors to consider in a risk-based sampling scheme for setting sampling
Legislation
Legislation and emerging science should be used to guide food and feed businesses to the risks associated with mycotoxins (Worldwide regulations for mycotoxins in food and feed 2003). EU regulation sets maximum limits for mycotoxins in certain foods (Regulations 1881/2006) and recommends maximum limits in animal feed (Recommendation 2006/576/EC). It is clear from the regulations that not all raw materials and finished products need to be analysed for every mycotoxin, which helps to focus sampling programmes.
However, businesses need to consider carefully the composition of the foods they produce. In the case of single component foods such as intake raw materials including fruits and cereals, it is straightforward to identify the mycotoxins for which maximum limits have been set. Manufacturers of multicomponent processed foods, such as fruit pies and cakes, will need to take account of dilution and concentration steps that occur during production for the various ingredients for which there are maximum limits.
For example, maximum limits are set for aflatoxin M1 in dairy foods but not for patulin in dairy products. Therefore, an apple would not need to be analysed for aflatoxin M1 (or milk for patulin). However, products that contain dairy and apple raw materials that are used to make a composite food would need to be analysed for aflatoxin M1 and patulin, respectively. This is probably best done on intake raw materials but finished products might need to be analysed for both patulin and aflatoxin M1 to ensure compliance with legislation.
The table below is a non-exhaustive list of foods and mycotoxins controlled by EU and UK food contaminants legislation to guide businesses to important food/mycotoxin combinations.
Mycotoxin(s) |
|
Apple juice, solid apple products & apple juice |
Patulin |
Cereals and cereal products |
Aflatoxins |
Coffee |
Ochratoxin A |
Dairy products |
Aflatoxin M1 |
Dried
fruit |
Aflatoxins |
Nuts |
Aflatoxins |
Spices |
Aflatoxins |
Wines and grape juice |
Ochratoxin A |
Having identified the food/feed-mycotoxin combinations, the next stages in the process of setting a sampling rate involve assessment and scoring of technical and scientific aspects to prioritise the risks and then to derive a sampling rates. The overall aim is to rank the relative risk of food or feed exceeding a maximum limit and to use this to establish sampling rates that cover the range of food-mycotoxin combinations in a proportionate manner.
Factors to consider during technical assessment of risks
Rationale |
|
Legislation – including interaction with food & feed type |
It is an offence to place on the market certain foods if they contain contaminants of any kind specified in particular regulations or which may be injurious to health. Food & feed businesses are responsible for ensuring that food and feed satisfy the requirements of general food law (Regulation 178/2002) and regulations for maximum limits for specific contaminants (Regulation 1881/2006). In the UK, similar compliance demands are placed on food businesses by the Food Safety Act 1990 and The Contaminants in Food (England) Regulations 2007. It is also an offence to mix foods & feed that do not comply with the maximum limits (i.e. not appropriate to dilute them) or to detoxify by chemical treatment. |
Likelihood of exceeding a maximum limit (i.e. safety margin) |
A safety margin, which is derived from the reciprocal of the value from dividing the contamination level in food by the maximum limit may be used to rank the individual risks. Larger numbers (i.e. >1) indicate higher safety margin. Similarly, the safety margin can be determined using toxicology metrics such as the level of mycotoxin known to cause an adverse effect in man divided by the estimated daily intake from food, such as derived from nutritional surveys. |
Prevalence & survey data |
Mycotoxin levels in food can be obtained from data produced by food regulators (including the FSA and EU) and/or from in-house surveillance/due diligence data. This data should be combined so as to assess the underlying level of a product exceeding a mycotoxin maximum limit. |
Toxicity |
Can be based on toxic metrics but to a large extent this is taken into account within the MRL. Can be useful for new and emerging mycotoxins consider genotoxicity. |
Consumption pattern among consumers |
Whether the food is a staple the general intake. |
Number of people likely to be affected |
Whether the product is a local, national or international brand. |
Food safety controls that might have prevented the fault |
Food safety procedures such as HACCP, storage transport conditions, processing reduction/concentration. |
Likelihood of a consumer recognizing the fault |
Easy for the consumer to recognize and does it change during storage. |
Many of these are included in the key obligations for food and feed businesses (Regulation 178/2002).
Relation between risk score and sampling rate
The output from the different sections in the risk assessment document can be classified using a qualitative scoring system based on traffic lights (i.e. like the FSA nutritional systems), high, medium and low. A quantitative system could also be developed. Whichever system is used, it is critical that the logic used to derive the scoring system is sensible and transparent as this will serve as a foundation for setting priorities and sampling rates. These scores are then combined to give an overall risk score. Products with higher scores present the greater risks and should be sampled more often than products with low scores.
However, although EU and UK regulators recommend that sampling is risk-based, none provide clear guidance on the actual number of checks that should be performed.
This leads to a number of scenarios in the food industry. For example, businesses may specify that mycotoxin analysis must be performed a certain number of times per year, regardless of the volume of value of material.
Basing the sampling rate on the risk score helps to move away from this type of prescriptive system. It can also help food businesses target their food safety resources more effectively in areas where it is most needed (including those with a high risk score or where the score has changed owing to new information).
For example, food business may specify that that every product must be sampled at least once in a given period (i.e. a minimum sampling rate per week/month/year). The testing rate could then be scaled in relation to the distribution of risk scores taking account of the resources (budget and analytical capacity).
A key point to note is that risk-based sampling schemes for mycotoxins are most efficient and effective in relation to food safety when sampling occurs as early in the supply chain as possible. This is because mycotoxins are normally formed during plant growth or during storage and many are not generally reduced to any significant extent during food manufacturing. This approach and rational may mean that more raw material analysis needs to be undertaken but it also helps to reduce the number of samples required of finished products (especially retail samples).
Sampling and analysis
The performance of sampling methods and analytical methods for mycotoxins are two critical aspects that are often overlooked. The best risk assessment and risk-based sampling schemes are almost useless if inappropriate sampling and analysis methods are employed because they may either fail to detect (sample) mycotoxins or provide very poor accuracy in measuring contamination level. This is because mycotoxins are heterogeneously distributed in samples (particularly in raw cereals and nuts). Taking just one spot-sample from a bulk might overestimate or underestimate the underlying average level (which is the value important for risk assessment) because the sample came from a localised hot-spot. Good sampling schemes involve taking several small sub-samples from a bulk or at regular time intervals in a continuous process and then combining, mixing and comminuting the sub-samples to give an aggregate sample. The aggregate sample is used for mycotoxin analysis.
The EU has regulations (Regulation 401/2006) to help food operators develop sampling methods and choose appropriate analytical determination methods. The regulation is primarily for official control (i.e. clearance of bulk deliveries by port authorities) but has useful information for use by food businesses for raw materials and retail products. It also gives performance criteria for analytical methods for mycotoxin analysis in terms of recovery and method repeatability and reproducibility relative standard deviation that need to be met.
There are two approaches for determination of mycotoxins. One involves sophisticated chromatography instrumental methods in laboratories, which allows the largest range of mycotoxins to be determined with the highest sensitivity. However, the drawbacks include long turnaround times and some of the tests can be expensive. Biosensor and lateral-flow devices are rapid and portable technologies that have been developed for a limited number of mycotoxins. The performance of the various commercial kits varies, but they are inexpensive and can be used in the field or food store to give near instant results. However, some are not as sensitive or specific as chromatography methods. Biosensors are useful for rapid high throughput screening (including raw material intake testing) to detect batches with high levels of contamination, which can then be checked using chromatography methods.
Risk assessments - documentation, communication and review
Risk assessments, including rationale and supporting evidence for their basis, must be documented and reviewed. This will be useful to food regulators to support a due diligence defense following a food incident (including retrospective identification of foods that exceed mycotoxin maximum limits). The risk assessment documentation and sample results can be used with FSA guidelines for how food and feed business should respond to a food safety incident (FSA incidents response).
Analytical is also useful to food business to verify that food safety measures (including HACCP) are working properly and for demonstrating to suppliers food safety standards and for demonstrating performance.
As with other food safety management systems, the information should be reviewed in the light of appropriate emerging science and/or annually. For example the FSA and the WHO have published best-practice guidelines for good agricultural practice (GAP, GAP for mycotoxins) and storage practices to minimize mycotoxin contamination (FSA guidance on Fusarium). Review of these sources of information, emerging science and actual data will help food businesses to re-evaluate the risk scores and the overall risk-based sampling rate. For example, introduction of a new type of processing method or use of a novel cereal variety that destroy or prevent the development of mycotoxins could substantially reduce the risk scores, which then alters the sampling rate and need for testing. This information is documented and the sampling rate for the next period is established.