Platon Chitzos, exports manager, said the US is among the most problematic areas due to its diverse climate.
“The US has frequently high problems with corn produced in south-eastern states and wheat in mid-western states also has DON problems on a yearly basis,” he told FoodQualityNews.
“In the EU you can have very high levels of aflatoxin contamination on the Mediterranean, almost no aflatoxin contamination in central and northern Europe but you can have ZON contamination in northern Europe and a very low percentage in the Mediterranean.
“I wouldn’t say it is an issue of developing and developed countries, I would say it is an issue of what you are producing, corn is the number one product susceptible to aflatoxin, and under which climatic conditions.”
Standards needed for trade
ProGnosis Biotech has lateral flow tests for milk as well as ELISA assays for mycotoxins in food and feed and aflatoxin M1 in milk.
Chitzos said trade, mostly in raw materials but also in milk, has brought a greater need for setting standards.
“We were informed that in Ireland and the UK for example due to the climatic change there has been a small increase in the demand for testing. In Ireland especially because of the exports to China,” he said.
“If countries did not export they would not test, only maybe in crisis situations. But exports do bring the demand for such kinds of paper and when you start testing you will see you have some smaller or bigger problems and then you put it in your routine. The exporting business brings more requirements and specifications from the importers side.”
Different country, different regulation
Legislation is not uniform worldwide and few countries outside the EU have sophisticated regulations.
The US does not allow sale and export of food with aflatoxin levels exceeding 20 parts per billion (ppb) and European Union standards are 2 ppb.
“Most multinational companies due to their transactions with other parts of the world, do test,” said Chitzos.
“But there are countries, most of them are in the developing world, which do not test as much although there is a legislation. Local services are not obliged to take big numbers of samples, frequently.
“Mycotoxins are not yet in many parts of the world considered to be something as important as Salmonella or Listeria because it does not have immediate impact, it is a gradual, low intoxication.”
There are differences on which mycotoxin to test for depending on the matrix.
Spices are tested for aflatoxin and ochratoxin, nuts for aflatoxins and raisins for ochratoxins.
In cereals it is deoxynivalenol (DON) and zearalenone (ZON) and T2-H2 mostly in wheat, barley and cereals from a colder climate. Fusarium is prevalent almost only in corn associated with very humid climatic conditions, so are prevalent in Latin America and North America.
Representability of a sample
Sampling can be a major headache, said Chitzos.
“In a private dock if you want to have a substantial picture of what your cargo is then you cannot make just one sample,” he said.
“If we take the scale of a truck, a truck can transport 30 tonnes of corn, these 30 tonnes will produce a first sample of less than 50kg, this 50kg must be reduced by sub-sampling and sequence sub-sampling to a final sample of 5-50g.
“Therefore it is almost impossible to say a ship containing hundreds of tonnes is ok with just one sample. It won’t be a representative one and that is something we have talked about again and again – representability of a sample.
“In reality it is very difficult to clear one cargo, one ship, one truck with just one sample. Just to give you an example, four seeds with a high concentration can produce a 5ppb result for Aflatoxin B1. The probability of testing these four seeds together or getting a good representative idea of your truck is very difficult.
“The best way to test a sample, which is not a fluid like milk or wine, would be to ground up the samples. You test per truck, you get the result that you get I would say you do not discard anything, and then you try to test within your lines of processing.”
Chitzos said if samples and sub-samples are not efficient it will make no difference whether ELISA, HPLC or another technique is used.
“What we recommend is according to the size of each company they should all develop a policy of testing raw materials and testing final products, that is the optimum. If they are a smaller company they should do rapid tests, if they are a bigger company they should use ELISA.
“In bigger companies you do not have the time to check with rapid tests as you would have to check one sample at a time and whatever comes above the internal limits should go to an external or your internal lab for chromatography.”
HPLC, ELISA or lateral flow?
Chitzos said chromatography, ELISA and lateral flow are used for mycotoxin detection.
“Chromatography has a much broader range of mycotoxin tests, masked mycotoxins are much easier to trace with HPLC. The problem is due to the fact they test one sample at a time and each sample needs a specific separation and treatment and sometimes by specialised technicians,” he said.
“That renders HPLC as a very difficult tool for mass testing. It is gradually being substituted by LC-MS, in terms of accuracy, cost and time. In non-crisis times, external labs using HPLC or LC-MS are the prevalent method. Many industries will reduce the sample tested, they test periodically and do more risk management.
“In milk it is easier to use rapid tests as they are more in the spotlight than other industries. When a crisis comes in ZON, DON or aflatoxins, ELISA comes back into the game to test multiple samples and reduce the cost per sample and keep good time.”
ELISA, due to the aflatoxin crisis in 2013, which started in the Balkans and spread in the EU, has improved a lot, said Chitzos.
“I mean it has been simplified, all conjugates are ready to use nowadays, the duration of the test can be reduced,” he said.
“Our ELISA kits besides milk take up to 20 minutes; accuracy and precision has been improved; we have excellent results in different kinds of matrixes in proficiency tests for FAPAS and BIPEA in France.
“Lateral flow is still quite far away – what somebody always notices and asks is what your coefficient variant (CV) would be, how precise and how accurate your test is.
“But in the case of lateral flow nobody used to ask, so there was a common knowledge that when it had 20% CV it is a good lateral flow. 20% is too much, our ELISA kits have a CV of around 6-8%, our lateral flow for M1 has a CV of less than 9%.
“Lateral flow has not moved forward in terms of accuracy but it has moved forward in terms of easiness to use and shorter time.”
Tackling the problem
Education of different players in the supply chain is vital, said Chitzos.
“Educating consumers of what to seek and how to evaluate a company which is testing or not testing, they will be posing questions so educating the consumer public is number one,” he said.
“Number 1.1 would be educating industry and the farmers, the industries is easier because they won’t be burdened with big costs but you have to find the correct people that will be able to absorb this kind of information.”
Chitzos said the most difficult part, especially in developing countries where production is fragmented, are the farmers.
“The farmers have to be able to report simple things, for example, many farmers in Greece confuse organic with mycotoxin free,” he said.
“They believe if you grow something without fertiliser, without the insecticide or fungicide that you will be clean of any kind of problem. So you have to educate on how to store, storage is a very common reason why things go bad with raw materials.
“Secondly, testing raw materials and the final product in every case and in the case of animal foods testing animal feed, raw materials and final product and milk.
“In the EU we have a less pre-emptive policy we try and test the final product, in the US they have a more pre-emptive policy in theory, which is testing every truck that is received must be tested. This is correct in theory but as we discussed before it is very difficult to have a true picture, you should try and test the final product.”