Archived – What we heard report - Consultations on contaminant standards for aflatoxins, deoxynivalenol, fumonisins, ergot alkaloids and salmonella in livestock Feeds

CFIA response

With respect to these key respondent messages, the CFIA has reflected further on the suitability and timing of including all of the identified contaminants and the respective proposed maximum levels into the modernized regulatory framework via a document for incorporation by reference. These considerations include:

• with regards to considerations related to a contaminant, the
  • scientific soundness of the proposed standard
  • international alignment of the proposed standard
  • availability, access and suitability of analytical methodologies that would enable regulated parties to monitor for the presence of the contaminant in ingredients and mixed feeds
  • availability of surveillance data to benchmark current compliance rates
  • equity of enforcement for products in domestic, import and export trade
• with regards to considerations related to the risk management of a contaminant, the
  • adequacy of current compliance rates
  • achievability of compliance by regulated parties along the supply chain
  • availability of reasonable risk management measures to respond to situations where the contaminant is present in the supply chain

Having given careful consideration to these criteria, the CFIA plans to move forward with including the maximum levels set out in this summary report for aflatoxins, deoxynivalenol, and fumonisins into the modernized regulatory framework by listing them in a document to be incorporated by reference. In addition, the proposed outcome-based standard for salmonella will be included in the proposed regulatory text.

Maximum limits for ergot alkaloids will remain in guidance at this time.

A more detailed discussion on contaminant-specific feedback and concerns and the CFIA's response follows below.

CFIA response

The CFIA proposed maximum limits in feed ingredients, as well as in the total diet. In most cases, the maximum for the total diet is lower than in the ingredient. This difference between the two levels accounts for how ingredients are used in total diets. While the CFIA has a "no-blending" policy and does not allow mixing a contaminated feed to bring it into compliance, we do recognize that individual ingredients are not the only thing the animal eats so in most cases the tolerance for contamination of ingredients is higher than for mixed feed.

The proposed maximum levels are in line with those from other international jurisdictions such as the European Union and the United States. As well, current data from CFIA's multi-year (1990-2015) mycotoxin survey suggests that naturally occurring levels of individual mycotoxins are typically much lower than the proposed limits.

Based on the feedback received from stakeholders and our re-evaluation of the available data, the CFIA has reflected on the suitability of including maximum limits for fumonisins and deoxynivalenol for single ingredients in the IbR document. Instead, these maximums will be moved to guidance and industry is encouraged to use these levels in evaluating the safety and suitability of ingredients for use in livestock feed. The maximum for aflatoxin in single ingredients will remain as this is a current regulatory requirement.

The following factors influenced the CFIA's decision consider mycotoxin standards at the single ingredient level:

• the proposed Canadian maximum limits for ingredients are consistent with the standards set by other jurisdictions worldwide, including European Union and the United States
• provides a standard to industry on what is considered acceptable and safe for distribution
• prevents the use of poor quality feeds by encouraging the use of best practices in contaminant prevention and control
• helps industry decide to direct the ingredient to be used for a specific livestock species depending on the maximum level in total diet

In addition, setting a maximum limit for each ingredient reinforces the safety of feed ingredients by taking into account potential mycotoxin synergistic effects. For example, DON contamination in feeds has been found to co-exist with other fusario-toxins including nivalenol and zearalenone. In a worldwide survey, it was also observed that 38% of the naturally contaminated feeds had co-contamination of two or more mycotoxins. The presence of multiple fusarium mycotoxins could act synergistically and cause adverse health impact on animals.

The CFIA encourages regulated industry to continue the use of best practices for contaminant prevention and control. The CFIA is also open to working with industry regarding best practices (example, analytical testing and segregation of contaminated grains) to prevent contamination in livestock feed. The CFIA is committed to exploring opportunities for developing guidance materials for industry in this regard.

The CFIA and Health Canada have been working on new policies to facilitate the marketing of various products which previously would have been classified as veterinary drugs based on their claims. Mycotoxin detoxification agent (MDA) is one category where the CFIA will be offering increased flexibilities by allowing for appropriate mycotoxin control claims when the mode of action occurs in the feed matrix. MDAs are not permitted to be used for mycotoxins that exceed the maximum level in feed. More recently, Appendix E-1 to the Drug-Feed Classification document was published to clarify this policy.

CFIA response

Mycotoxins are generally not distributed evenly throughout a feed or grain. This uneven distribution results in "hot spots" where large amounts of mycotoxins are present in small areas. Extensive sampling would be required to determine the starting contamination level in order to safely blend a contaminated feed or grain. Given this information, the CFIA does not permit grains or feeds contaminated with mycotoxins, including aflatoxins, to be blended with uncontaminated product in order to bring the aflatoxin levels to within the maximum levels. This "no-blending" policy is also found in other jurisdictions. In addition, mycotoxins continue to be produced by the mould under favourable storage conditions. This can result in the actual contamination level continuing to increase in the time since the sample was taken for analytical testing. Blending in these situations would not guarantee a feed that would meet the established standards and thus be safe for animal or human health.

Aflatoxins in particular have the potential to negatively impact the health and productivity of a variety of livestock at very low levels. In addition, aflatoxins consumed by livestock from feed are readily transferred to milk and other animal products, which could represent a food safety risk. Sources of aflatoxin contaminated products are usually imported into the country and typically originate from single ingredient feeds where the intended livestock species is unknown.

To safeguard all livestock species, and the resulting food, the CFIA intends to proceed with a maximum level for aflatoxins of 20 ppb in single ingredient feeds and 20 ppb in the total diet (Table 1 of Appendix I).

The proposed Canadian maximum limits for ingredients are consistent with most of the standards set by other jurisdictions worldwide, including the United States and the European Union. International standards are provided below for comparison:

From: CPG Sec. 683.100 Action Levels for Aflatoxins in Animal Feeds

Source: Directive 2002/32/ec of the european parliament and of the council; of 7 May 2002 on undesirable substances in animal feed: Legislation on mycotoxins (PDF 84 kb)

CFIA response

The CFIA reviewed scientific studies identifying the levels at which adverse animal effects were observed and then focussed on whether these levels represented a food safety, or environmental concern. The negative impact of DON in livestock is usually first observed as a reduction in animal performance followed by adverse health effects. As a result, it was reasonable to consider animal performance as well as animal health in the scope of the review.

Suggestion to increase maximum levels

The CFIA agrees with the respondents that there is not sufficient scientific evidence to support a 1 ppm DON maximum in total diet for lactating dairy animals. Thus, the CFIA will remove the maximum level for lactating dairy animals from the DON list of standards. Feeds for lactating dairy animals will need to meet the "Cattle – dairy" levels.

Stakeholders suggested increasing the maximum level of DON in total diets for various livestock species (see Table 2 above) and provided EFSA's review, 2017 as supporting evidence. In the process of investigating and proposing the standards for DON, the CFIA reviewed the studies referenced in the EFSA (2017) document noted in the proposal, as well as other animal feeding studies. Additionally, risk assessments carried out by international scientific bodies, such as EFSA (2004 and 2013), FDA (2011) and Food Safety Commission of Japan (FSCJ, 2010) were also considered in the CFIA's review. Based on these reviews, the CFIA has determined that there is not sufficient scientific evidence to warrant increasing the maximum level solely based on the stakeholders' recommendations.

Considering these data gaps, CFIA took a conservative approach to establish the maximum level for DON in feed and align with international organizations including European Commission and US FDA. CFIA's maximum levels follow the same approach as the FDA has published in their guidance, with the exception of distiller grains. Given the increased risk of contamination of distiller grains and their frequent use in livestock diets, CFIA is setting the same maximum level as for other single ingredient feeds.

Suggestion for fish

Per Table 2a above, stakeholders suggested a maximum level of DON be set for total diets for certain fish species. CFIA acknowledges that the scientific studies on adverse health effects of DON in fish are limited. However, current scientific data shows a growing occurrence of mycotoxins, including DON in fish feed due to increased replacement of marine ingredients with those from plants sources. Additionally, the data from recent scientific studies demonstrate adverse health effects of DON in fish, particularly trout. Given the supporting evidence, the CFIA is adding a maximum level of DON for Salmonids (as per Table 2b). Feeds for other types of fish would need to meet the "Other animals" levels.

Single ingredient feeds

The CFIA proposed the maximum level in ingredients (single ingredient feeds) by considering the industry's most common inclusion rates of ingredients in the total diet of a particular livestock species, as well as the sensitivity of the livestock species to a specific mycotoxin. Based on the feedback received, the CFIA will move these proposed ingredient maximum levels for the various livestock species to guidance.

International standards are provided below for comparison:

Reference
European Commission, 2006/576/EC. Commission Recommendation of 17 August 2006 on the presence of deoxynivalenol, zearalenone, ochratoxin A, T-2 and HT-2 and fumonisins in products intended for animal feeding. Official Journal of the European Union.
https://ec.europa.eu/jrc/en/eurl/mycotoxins/legislation

Reference
FDA, 2011. FDA Mycotoxin Regulatory Guidance: A Guide for Grain Elevators, Feed Manufacturers, Grain Processors and Exporters. National Grain and Feed Associations.

CFIA response

Stakeholders suggested increasing the maximum level for fumonisins in total diets for horses, rabbits, poultry (layers) and breeding/lactating ruminants. While the supporting evidence (EFSA's 2005 review) discusses certain studies with NOAELs (no observable adverse effects levels) higher than CFIA's proposed limits, these studies alone are insufficient to warrant a limit change. The studies referenced in the review were not robust enough to justify CFIA increasing the proposed levels. The studies had issues such as, limited testing in target species, lack of appropriate endpoints, study duration, and dose range. CFIA has already taken into consideration most papers from EFSA's review and has also identified a few other cases where LOAELs (lowest-observed-adverse-effect level) were detected at limits below the stakeholder's proposed limits. Given the data gaps mentioned above, a more conservative and harmonized approach was taken when establishing the limits in order to align with international organizations like the US FDA. Therefore, the CFIA will not be changing the current proposed limits for fumonisins.

With respect to maximum levels for ingredients, the CFIA's proposed levels are based on considering the industry's most common inclusion rates of ingredients in the total diet of a particular livestock species, as well as the sensitivity of the livestock species to a specific mycotoxin. Based on the feedback received, the CFIA will move these proposed ingredient maximum levels for the various livestock species to guidance.

International standards provided below for comparison:

European Union

Source: The Commission of the European Communities. 2006. Commission recommendation on the presence of deoxynivalenol, zearalenone, ochratoxin A, T-2 and HT-2 and fumonisins in products intended for animal feeding. Official Journal of the European Union.

Source: FDA. 2011. FDA Mycotoxin Regulatory Guidance: A Guide for Grain Elevators, Feed Manufacturers, Grain Processors and Exporters. National Grain and Feed Associations. Retrieved on January 21st.

CFIA response

As described in the proposal and mentioned by external stakeholders, a number of assumptions were made (worst-case scenario/conservative approach/weight of evidence approach) when setting the maximum limits of EAs in feed ingredients and livestock feeds. Some of the assumptions included that processing had no impact on the toxicity of the EAs present, there are no interactions along EAs or between EAs and other toxins or other nutrients in the diet, etc. It has been mentioned by CFIA and stakeholders that EA toxicity is quite complex. Due to this complexity, CFIA took a more conservative approach to establish maximum limits for ergot alkaloids. It is known that the effects of EAs observed in livestock species not only depends on the dose of the EAs, pattern of EAs and duration of exposure, but also on the health of the animals, age, gender, physiological state, production stage, etc. Unfortunately, there is no data or limited data to set standards based on gender differences or production classes (for example, breeding).

Based the feedback received from stakeholders and our re-evaluation of the available data, the CFIA has reflected on the suitability and timing of including maximum levels for EAs in the regulatory framework at this time. Given the limited feeding studies for each livestock species/production class and the limited availability of analytical methods to determine the presence of EAs in feeds, the proposed EA limits in feed ingredients and livestock feeds will remain in guidance and will not be included in the maximum contaminants IbR document at this time.

There is limited data from published literature with regards to the transfers of EAs from feed into foods of animal origin. The limited published data suggest that EAs do not transfer readily and it is unlikely foods of animal origin will be a main source of human exposure to EAs.

The current tolerance levels for EAs are listed in RG-8 Regulatory Guidance: Contaminants in Feed. There are scientific studies and information available that EAs have an impact on animal safety. The studies reviewed in developing the proposed maximum limits for EAs clearly indicate that the current tolerance levels are not protective of animal health. The current tolerance levels for EAs are too high as adverse effects are observed at limits that are much lower.

In the past, action limits, maximum limits, etc. have been established for ergot bodies (ergot sclerotia) or "ergot alkaloids" in which it was unclear what alkaloids were being referred. It has been stated by CFIA and stakeholders that some ergot alkaloids are more toxic to livestock species compared to other alkaloids. It is also known that the "-inine" forms can be described as biologically inactive, but interconversion does occur between the "-ine" forms and the "-inine" forms during various alkaline conditions, long periods of storage, presence of light, etc. EAs are unstable and can epimerize. Therefore, it is important to characterize the toxicity and safety of both the "-ine" and "-inine" forms and be able to detect both the "-ine" and "-inine" forms of EAs.

The maximum limits being proposed focus on the main EAs from Claviceps purpurea known to infect a number of plant species and in particular rye, wheat, triticale, barley, and maybe oats. The maximum limits are based on the limits of the dominant 12 EAs (or 6 EA pairs) from C. purpurea including Ergometrine and Ergometrinine, Ergosine and Ergosinine, Ergotamine and Ergotaminine, Ergocornine and Ergocorninine, Ergocryptine and Ergocryptinine, Ergocristine and Ergocristinine. Reference standards for these particular EAs are commercially available. Based on information in the Guo et al. article, there are standards available for 25 EAs.

One of the concerns raised by stakeholders was with respect to the availability of analytical methods for determining EAs. Currently, there is a lack of analytical methods commercially available to detect EAs in feed matrices. There are a few published studies that focus on the development of analytical methods to determine certain EAs in different matrices (Guo et al., 2016, Krska et al., 2008, Tittlemier et al., 2015, Mulder et al., 2015). In addition, a couple of review papers have been published with regards to the analysis of EAs in certain matrices (Krska and Crews, 2008 and Crews, 2015).These can be used as a starting point in the development of analytical methods to determine the concentrations of EAs in livestock feeds. In addition, there has been some research that has focussed on appropriate extraction and sample preparation techniques and methods to analyze for EAs in different commodities.

CFIA is well aware that research is being conducted with regards to the safety of EAs fed to livestock species. Once this research has been completed, there is an opportunity for stakeholders to provide this information to CFIA to evaluate. This information can be used to update the guidance document or to allow the limits for EAs to be added to the IbR document.

The CFIA will update the levels for EAs that are found in RG-8. In addition, the CFIA will provide additional guidance on managing and controlling EAs to help mitigate some of the risks that EAs pose. The CFIA will revisit EAs in the future and reassess them based on our criteria to determine when it is suitable to move the EAs into the contaminants maximum limit document that is incorporated by reference.

CFIA response

Animal feed controls play an important role in protecting the health of animals and the safety of food that we eat. With the proposed regulatory approach, the CFIA is establishing an outcome based standard that protects human and animal health. Salmonella is by far the most significant bacterial pathogen that can be present in livestock feeds.

After considering all the stakeholders' input the CFIA will establish:

• an outcome-based standard for Salmonella that will be added directly to the Feeds Regulations, indicating that feeds shall not contain Salmonella that is likely to be deleterious to livestock or present a risk to human health

The standard will be supported by regulatory guidance. The guidance will provide background information regarding salmonella and the risk it poses to animal and human health, preventative control procedures to control the risk posed by salmonella, risk management options that should be considered by a facility when they identify salmonella, the CFIA's inspection approach, and the risk management options, implemented through compliance and enforcement activities, that will be considered by the CFIA when they identify salmonella (including consideration of the serotype). This guidance will be based on peer-reviewed scientific evidence and international standards where possible. The guidance will be posted for consultation with all interested stakeholders.

In addition, Health Canada's Veterinary Drugs Directorate and the Animal Feed Division have been working on new policies to facilitate the marketing of various products which previously would have been classified as having veterinary drug claims. Salmonella control agents are one category where there will be increased flexibilities to allow for appropriate salmonella control agents whose purpose is to act in preventing or controlling contamination in feed.

CFIA response

The CFIA appreciates the importance of trade and has regular discussions with our international regulatory counterpoints (European Commission, EFSA and US FDA) with regards to standards for contaminants. At this time there are no accepted international standards for biological contaminants in feed. Given that the regulatory systems and authorities in countries differ, it is not always possible to completely align.

The CFIA took into consideration the maximum levels for contaminants in other jurisdictions when setting the biological contaminant maximum levels and has aligned with the jurisdictions to the extent possible. Differences between countries could be impacted by factors such as country-specific weather patterns which would impact the presence or absence of a particular biological contaminant or country-specific animal consumption data which would influence the total diet levels used. How standards are set, for example via guidance versus directly in the regulations, are impacted by the authorities and regulatory approaches that specific countries take. Setting maximum levels using incorporation by reference provides greater clarity to stakeholders with respect to the standards they need to meet, while allowing flexibility to update and change the values as new scientific information becomes available.

CFIA response

Given the additional feedback received from stakeholders, and greater clarity on the approach that the food program is taking, it is proposed that for feed the CFIA follow the approach for the export of food products as set out in the Safe Food for Canadians Regulations.

Feeds intended for export would be required to be manufactured:

• by a CFIA licence holder
• in accordance with a preventive control plan
• in accordance with documentation demonstrating that the importing country's requirements are being met

If the importing country does not specify its own requirements, Canadian safety standards would apply.

CFIA response

The identification of hazards in feed ingredients and the sampling and testing frequency of a feed product should be part of a facility's preventative control plan (PCP). It is the responsibility of the feed manufacturer or supplier to ensure that the feeds they import, export, manufacture, or sell, are safe and meet the maximum levels of chemical and biological contaminants for feed.

Certificates of analysis from accredited labs, labs that are able to demonstrate their method validation or results generated using validated testing procedures or kits, may be incorporated into a PCP. Accreditation bodies such as the Standards Council of Canada can provide insight into the proficiency of labs and the use of validated standard methods such as those developed or endorsed by bodies like AOAC International are recommended. Neither the frequency nor method of sampling and testing are prescribed by the regulations, but could be part of a validated PCP.

Please note that part of CFIA's National Feed Inspection Program consists of sample monitoring programs for biological and chemical contaminants. These sampling programs are used to monitor for background levels of various contaminants in feed and feed ingredients and verify compliance with contaminant standards. The CFIA will provide guidance on best practices for sampling and testing.

CFIA response

Based on the comments received, it appears there are inconsistencies on how "action levels", "guidelines", "maximum limits", and "regulatory maximum limits" are interpreted or used by various stakeholders and between different regulatory agencies. In addition, there seem to be inconsistencies in the interpretation of non-compliance and on the actions to be taken when contaminants are found in feeds above action levels or guidelines versus those prescribed in regulations as maximum limits.

Currently, product control actions are taken when feeds are found to be out of compliance with standards such as:

• ingredient definitions (example, Schedule IV and V of the Feed Regulations)
• contaminant action levels (example, RG-8)
• regulatory maximums (example, from Section 19 of the Feeds Regulations)

The risk management goal of product control actions is to bring the feeds into compliance and prevent the entrance or continued entrance of contaminated feeds into the feed chain.

Standards, such as those found in regulations, incorporated by reference documents, or in guidance, should be integrated into industry PCPs and addressed by industry accordingly. If standards are not met, the CFIA would take regulatory action in cases where the industry is not demonstrating that feeds they are responsible for are meeting standards.

The CFIA will better define the terms "action limit" or "maximum level," so all regulated parties and interested stakeholders have a clear and concise understanding of the use of these terms in the context of the CFIA and what they mean (that is, regulatory actions taken if the action limits or maximum limits are exceeded).

The Agency's proposed approach to incorporate a document containing the contaminant maximum levels by reference in the modernized regulatory framework represents an opportunity to take a consistent approach to the setting of such safety standards while providing more flexibility to review and amend the standards on a more timely basis.

In keeping with the CFIA's Incorporation by Reference Policy, the CFIA will provide a plan regarding the frequency and process for the review and revision of the incorporated document and the maximum limits.

Note, in the Tables below

"Proposed Maximum Level" refers to a level set out in the August 2017 CFIA Proposal - Contaminant Standards for Aflatoxins, Deoxynivalenol, Fumonisins, Ergot Alkaloids and Salmonella in Livestock Feeds

"Final Maximum Level for IbR" refers to the level the CFIA proposes to include in the modernized regulations following consideration of stakeholder feedback received in response to the August 2017 Proposal