The Canadian Food Inspection Agency's Online Consultation on Plant Molecular Farming: Summary of Feedback

(April 26 - May 23, 2006)

PDF (183 kb)

1. Table of Contents

• 2 EXECUTIVE SUMMARY
• 3 DEMOGRAPHICS OF PARTICIPANTS
  • 3.1 BREAKDOWN OF PARTICIPANTS BY SECTOR
  • 3.2 OFFICIAL LANGUAGE OF RECEIVED SUBMISSIONS
• 4 SUMMARIES OF RESPONSES
  • 4.1 GOVERNMENT OVERSIGHT
    • 4.1.1 What form should the government regulatory oversight of commercial-scale plant molecular farming (PMF) take place?
  • 4.2 CATEGORIES OF RISK
    • 4.2.1 PNTs intended for PMF will not all carry the same level of associated risk to the environment or to human and animal health. To ensure that regulation of these PNTs is commensurate with their individual associated risk, the Plant Biosafety Office (PBO) is considering grouping these plant products into categories of risk.Do you agree/disagree with this approach?
    • 4.2.2 In determining which category of risk a particular PNT intended for PMF will belong to, what scientific criteria should be used (e.g., ease of segregation and confinement, biology of the novel compound's production platform, toxicity or allergenicity of the novel compound, etc.)?
  • 4.3 ACCEPTABILITY OF FOOD AND FEED CROPS AS PMF PRODUCTION PLATFORMS
    • 4.3.1 Which elements/characteristics would enhance the acceptability of food and feed crops as PMF production platforms?
    • 4.3.2 Should the acceptance of food and feed crop species as PMF platforms be based on the risk of the expressed novel gene product to the environment and to human and animal health?
  • 4.4 RISK MANAGEMENT
    • 4.4.1 The PBO is considering using the Canadian Seed Growers Association (CSGA) guidelines as a basis for developing the relevant reproductive isolation distances for commercial PMF fields. What other information should the PBO be incorporating or considering when establishing these required isolation distances?
    • 4.4.2 On what scientific basis can limits, if any, be applied to the size of commercial PMF fields?
    • 4.4.3 The cultivation of PNTs intended for PMF may affect existing crop rotations used by farmers/producers. Are there particular crop species that would have large impacts on crop rotations in Canada if they are used as a platform for PMF? Would the impact be dependant on whether the PNT is expressing a pharmaceutical compound compared to an industrial compound?
    • 4.4.4 It is expected that those applying for authorization of commercial production of PNTs intended for PMF will be required to submit a contingency plan in the event an unintended release occurs. What components should be required to be included in such a plan?
  • 4.5 SEGREGATION AND HANDLING OF PMF CROPS
    • 4.5.1 Under what circumstances can the current segregation and grain handling systems deal with PMF commodities (e.g., the production of a vaccine in a grain crop within a closed-loop production system)?
    • 4.5.2 Under what circumstances could low risk PNTs intended for PMF be placed into the bulk handling system (e.g., a grain crop producing a novel compound already present in the food or feed systems)?
  • 4.6 OTHER
    • 4.6.1 Is there anything else of a scientific nature pertaining to PMF production in Canada that you would like to comment on?

2. Executive Summary

From April 26 - May 23, 2006, the Canadian Food Inspection Agency (CFIA) hosted an online consultation on commercial plant molecular farming (PMF). This online consultation was a follow-up to a previous face-to-face multi-stakeholder consultation (March 2005) and was designed to solicit further feedback, from a broader audience, on the approach taken by the CFIA in developing guidelines for regulating the environmental release of plants with novel traits (PNTs) intended for PMF.

Most participants agreed that strict regulatory oversight by the Government of Canada (GoC) was the only acceptable form of oversight for this activity. Although licensing of PMF producers gathered some support, several participants advocated for a mixed GoC / third party inspection service system, where regulatory oversight would transition over time from the GoC to an accredited third party inspection service.

Many participants conceded that the stringency of the regulatory oversight of these PNTs should be commensurate with their individual risk, but debated whether the use of risk categories would be the most appropriate mechanism to achieve this. Most agreed that the acceptance of categorization would be largely dependant on:

• the criteria used in assigning a PNT to a particular risk category
• the impartial mechanisms implemented to ensure that all PNTs would be assigned to the appropriate risk category.

Participant responses regarding the use of food or feed crops as PMF production platforms were polarized; many were strongly opposed while others felt that the use of redundant confinement measures would increase their acceptability. However, most participants agreed that the acceptability of using a particular food or feed crop species as a production platform should consider the potential impact(s) the subsequent PNT intended for PMF would have not only on food and feed safety, but environmental safety as well. Adventitious presence (AP) and unintentional releases of PNTs intended for PMF were their main concerns.

With regards to confinement measures, some participants were strongly opposed to using the Canadian Seed Growers Association (CSGA) isolation distance guidelines as a starting point for developing appropriate reproductive isolation measures for commercial PMF fields as they were designed for maintaining varietal purity and not for the purposes of confinement. Particular concern was raised with respect to the potential for PMF gene bridging into seed stocks and fields of non-PMF counterparts. Despite their concerns, many participants offered their input into adapting the CSGA guidelines for their use in commercial PMF.

Various suggestions were given for components or elements which should be included in an acceptable contingency plan; these included strategies for public notification of intended commercial planting, long-term monitoring of contaminated areas, and a validated detection method for the specific PMF trait. It was felt that all plans should reflect the potential and known hazards and risks of the PNT intended for PMF in question.

3. Demographics of Participants

3.1 Breakdown of Participants by Sector

Image - This image is a bar chart that shows the breakdown by sector of the online consultation's participants.

Figure 1: Breakdown by sector of online consultation's participants. The sector "Other" includes the general public, citizens' groups, students, private companies, as well as members from the information technology sector, the horticulture sector, and the education sector. Please note: that some participants identified themselves with more than one sector. Please note: Numbers represent actual number of participants selecting the option in question. Some participants self-identified with more than one sector.

3.2 Official Language of Received Responses

Image - This image is a bar chart that shows the official language used by participants in each sector.

Figure 2. Official languages of participants by sector (Academia = A; Government Research = B; Canadian Provincial/Territorial Governments = C; Producers/Grain Handlers = D; Non-Governmental Organizations = E; Industry Associations = F; International Organizations = G; Other = H). Please note: Numbers represent actual number of participants selecting the option in question.

4. Summaries of Responses

4.1 Government Oversight

4.1.1 What form should the government regulatory oversight of commercial-scale plant molecular farming (PMF) take place?

• Strict oversight by the Government of Canada (GoC), with complete inspection programs for 100% of production field sites, regardless of novel compound being produced
• GoC licensing of PMF producers, where producers are audited by the GoC against their submitted quality assurance (QA) systems
• Accreditation of third parties by the GoC to deliver inspection programs on its behalf, where third parties are audited by the GoC
• Other

Image - This image is a graph representing the number of respondents in each sector selecting the proposed approach for regulating the commercial environmental release of Plant wit novel traits intended for plant molecular farming.

Figure 3: Graph representing the number of respondents in each sector selecting the proposed approach for regulating the commercial environmental release of PNTs intended for PMF.

Strict GoC oversight

Many participants agreed that only strict oversight by the Government of Canada (GoC) of commercial PMF was acceptable and in accordance with the GoC's use of the precautionary principle for regulating products of biotechnology. In addition, this option would be more likely to assure the credibility and public trust in the regulatory oversight of commercial PMF in Canada.

Given the possibility that agricultural crops could be used as production platforms for novel compounds, some participants felt that this more stringent approach to oversight should be used until there is sufficient scientific research/data that has been published in peer-reviewed journals stating that such strict oversight may not be required in all cases (i.e., until the potential risks and impacts of commercial PMF are better defined). At this time, the required level of oversight could be assessed by the CFIA on a case-by-case basis. This approach would provide a more solid basis for justifying the approach taken and level of oversight imposed in regulating the commercial environmental release of PNTs intended for PMF.

This stringent regulatory oversight was also thought to provide developers greater flexibility in choosing the most appropriate plant species for their production platform, as the commercial environmental release of all plants with novel traits (PNTs) intended for PMF would be subject to the same level of oversight and scrutiny. In addition, participants felt that such strict oversight would help alleviate concerns from Canada's trading partners with respect to the potential adventitious presence (AP) of PMF crops in Canadian non-PMF commodities.

Licensing of PMF producers

Some participants felt that strict GoC oversight would lead to over regulating the environmental release of many PNTs intended for PMF as their regulation would often not be commensurate with their risk.

It was noted that many of the biotechnology companies already develop good manufacturing practices (GMP) which would allow for the assessment and auditing of standard operating procedures (SOPs) on an on-going basis. These SOPs would then be modified accordingly if compliance issues were identified during audits. Some participants believed that this option would be more cost-effective and provide adequate regulatory oversight. However, given that 100% segregation is not possible, some felt that, for PNTs intended for PMF with a food or feed crop as a production platform, this form of regulatory oversight would only be acceptable if it is was determined that low levels of AP in food and feed would pose no human and animal health safety concerns. If such concerns were raised in the risk assessments of these PNTs, then participants believed the only acceptable option was strict GoC oversight.

Other participants were very uncomfortable with the GoC delegating the regulatory oversight of commercial PMF to the industry itself as they feared that this approach would not allow sufficient awareness on the GoC's part of the industry's activities and would increase the likelihood of non-compliance going unnoticed.

Third party inspections services

This option was preferred by some as they felt that regulation by third party inspection services would provide for a more impartial regulatory oversight (i.e., inspection services would have no ties or interests with either the GoC, the PMF industry, or licensed producers). These participants stressed that stringent accreditation systems and oversight programs to oversee the accredited third party inspection services would need to be developed to ensure that all accredited third parties possess the capacity and technical expertise to carry out the inspection programs delegated to them by the GoC. Some participants suggested provincial governments may possess the capacity and the expertise to act as third party inspection services.

Other participants expressed some concerns with this approach to regulation. Specifically, this approach would require rigorous management by the GoC, an endeavour that would most likely not provide any relief in terms of pressure on the GoC resources. For example, where a non-compliance by a company/producer is not identified by an accredited third party, the GoC would be required to not only deal with the company/producer, but the third party inspection service that failed to identify the compliance issue.

Another concern was the possibility of an accredited third party inspection services having a vested interest in the commercial PMF operations for which they would be responsible for overseeing.

Other

Some participants felt that commercial PMF should be overseen by both a third party inspection service and the GoC to ensure both the safety of this activity as well as the PMF industry's compliance with its legal responsibilities. It was also suggested that the GoC might want to consider starting off with either strict GoC oversight or third party inspection services and transitioning, once experience has been gained by the industry, to licensing those PMF producers having displayed exemplary compliance records.

Additional comments:

There was agreement among several participants with regards to the GoC approach that the level of regulatory oversight of a PNT intended for PMF should be commensurate with its real and potential risks to the environment as well as to the safety of human and animal health (a risk that would be determined on a case-by-case basis). However, given the concerns over potential inadvertent and negative health impacts that could result from exposure to these plant products, it was recommended that the precautionary approach be taken when developing, implementing, and carrying out risk assessments and inspection programs.

Although not against the use of plants in molecular farming, certain participants felt that this type of agriculture would only be acceptable if it took place in containment facilities, under the strict GoC regulatory oversight.

4.2 Categories of Risk

4.2.1 PNTs intended for PMF will not all carry the same level of associated risk to the environment or to human and animal health. To ensure that regulation of these PNTs is commensurate with their individual associated risk, the Plant Biosafety Office (PBO) is considering grouping these plant products into categories of risk. The proposed categories are the following:

• Low to medium risk (e.g., plants producing compounds already present in the food system, such as a food processing enzyme)
• Medium to high risk (e.g., plants producing vaccines in a non-food, non-feed platform)
• High risk (e.g., a therapeutic compound being produced in a food or feed crop)
• Unacceptable risk (e.g., a highly toxic compound produced in a food crop)

Do you ________ with this approach?

Image - This image is a graph representing the number of respondents in each sector selecting the degree at which they agree with the proposed categories of risk.

Figure 4: Graph representing the number of respondents in each sector selecting the degree at which they agree with the proposed categories of risk.

A number of participants agreed with the general approach given both the broad range of novel compounds that could be produced via PMF as well as the wide spectrum of potential risks that could be associated with these PNTs. They believed that a system based on categories of risk would ensure that the level of scrutiny would be appropriate to the risk posed by the product. In addition, these participants felt that this would encourage the industry to develop low risk production platforms as well as applications. Some participants also believed that such a system could prove successful so long as the risk categories were agreed upon by all parties involved (i.e., government, industry, producers, as well as other relevant stakeholders).

However, numerous participants disagreed with how the categories were defined. In particular, it was noted that it could be difficult to categorize certain products given how the categories are currently defined as the risk categories seem to only take into account health safety risks as oppose to environmental risks. For example, a compound being produced in a non-food or non-feed production platform may be considered as low risk in terms of health safety, but could still pose a significant risk to the environment. It was suggested that a more differentiated approach be considered with respect to defining the proposed risk categories, one that would taken into account both potential environmental and health safety risks when assigning a particular PNT to a risk category. In addition, some participants suggested that this system would be clearer if categories were defined as "low risk" and “medium risk” as opposed to "low to medium risk" and "medium to high risk". It was also suggested that the “unacceptable risk” category should include all food or feed crops expressing compounds which are physiologically active (e.g. therapeutic or pharmaceutical compounds).

Certain participants indicated that it was difficult to agree or disagree with the CFIA's outlined approach to regulating the environmental release of PNTs intended for PMF given that no information on anticipated risk management strategies, confinement requirements, and/or restrictions was provided for each risk category. These participants would like to see some analysis on the impact of accidental commingling on food or feed companies and the ensuing consumer confidence in the food supply as well as the regulatory process. For example, what level of product recall would be available for each risk category and what would be the ensuing costs for affected companies?

Others stressed that specific “types” of PMF applications (i.e., production platform + novel compound) should not automatically be assigned to a particular category, but rather assigned to a category only once its risk has been assessed. As a result, they suggested that the CFIA clarify in their guidelines that these risk categories are intended to be used as pre-defined levels of regulatory oversight and risk management requirements which would be assigned to PNTs intended for PMF on a case-by-case basis.

Some participants felt that these risk categories were too permissive and that all modified plants intended for PMF should be under strict regulatory oversight. These respondents felt that only two categories were appropriate: “high risk” and “unacceptable risk”. Another subset of respondents suggested that the only two risk categories should be “low to medium risk” and “unacceptable risk,” while another subset felt that PMF production should only be carried out in containment facilities. Respondents also noted:

• if human toxicity was the basis for the various risk categories, then the risk of using any food crop for PMF should be considered unacceptable
• the dose at which it is safe to consume the PMF compound is another important criteria that needs to be considered when assigning a PNT intended for PMF to a particular risk category .

Some participants disagreed with this approach to regulation entirely as they felt that the potential risks of these plant varieties were being over stated. In particular, they felt that PNTs intended for PMF with non-food and non-feed production platforms should always be considered low risk and that risk should only be assessed as being higher when the production platform is a food or feed crop species and that there is a risk of a breach in confinement in its commercial production. They also believed that risk should not only be a function of the plant product itself, but also of its production location and timing. For example, planting PMF crops later in the season may be a potential option in mitigating the risk of cross pollination as pollen release in the PMF field would occur after that of its domestic and wild relatives.

Some concerns were raised with regards to the mechanism that will be established to assign a PNT intended for PMF to a particular risk category. Such concerns included:

• Who will be responsible for establishing the parameters or criteria used in determining the risk level of a particular PNT intended for PMF?
• Who will be responsible for determining the risk level of a PNT intended for PMF?
• How can we ensure that a PNT intended for PMF will always be assigned to the appropriate risk level, both in the short and long term?
• What criteria will be used to assess whether or not a PNT intended for PMF will pose negative long-term effects on human and animal health and on the environment?
• How will we ensure that cross-pollination of PMF with both wild and domestic relatives would not result in hybrids with negative synergistic impacts on the environment?

4.2.2 In determining which category of risk a particular PNT intended for PMF will belong to, what scientific criteria should be used (e.g., ease of segregation and confinement, biology of the novel compound's production platform, toxicity or allergenicity of the novel compound, etc.)?

Participants recommended that the following criteria be used in assigning individual PNTs intended for PMF to a risk category:

• biology of the production platform (e.g. outcrossing potential or gene flow to both domestic and wild relatives, annual versus perennial, etc.)
• agronomic practices used in the cultivation of the production platform's species
• the food and/or feed uses of the production platform (if any)
• potential negative impacts on soil and non-target organisms (NTOs) (e.g., soil micro-organism, insects, flora wildlife, field workers, etc.)
• potential routes that humans, livestock animals, and other NTOs could be exposed to the novel compound (e.g., dermal, respiratory, etc.)
• potential synergistic effects of novel compound (i.e. potential of novel compound to interact with the platform to produce a negative biological response in not only the plant, but in humans or animals as well)
• potential for novel compound to give a selective advantage to weeds in the event of gene flow (rare, but potentially significant)
• intended production system (contained, open field, use of transplants or direct seedling, etc.)
• potential effects on post-harvest management, such as volunteer management, seed segregation, involuntary seed dispersal, etc.
• intended geographic location of production area (i.e., whether open-field cultivation will take place within the usual Canadian production range for that plant species or in a new Canadian production range)
• intended size of commercial fields (10-acre field versus a 1000-acre field)
• method of expression of novel compound (e.g. stable transformation versus viral transient transformation)
• potential or degree of biological activity of novel compound in humans and livestock
• health impacts of novel compound at the potential doses that humans and livestock could be exposed at both low and high levels of food contamination
• quantity of tissue expressing the novel compound that would need to be ingested to contain a dose toxic to humans and animals
• the degree of ease or difficulty in traceability of novel compound in imports, exports, and processed food and feed products
• ability (or not) to be able to distinguish a PMF crop from its non-PMF counterparts (e.g. distinctive colour or other morphological characteristics)
• degree of ease in segregating the PMF crop from its non-PMF counterparts (e.g. ease of preventing cross pollination, of cleaning up spills, of distinguishing PMF variety from non-PMF varieties, potential for PNT to enter the food and feed chains, etc.)
• intended use of the novel compound (e.g. food processing enzyme that is already present in the food chain versus a therapeutic drug)
• presence or absence of biological tools for the confinement of a PMF crop in open-field commercial cultivation (e.g., male sterility, genetic use restriction technologies (GURTs), chloroplast transgenic systems, etc.)
• overall objective of cultivation/production (e.g. seed multiplication versus commercial production of novel compound)
• constitutive expression versus inducible, tissue-specific, or growth-stage specific expression of novel compound

Participants were quick to note that most criteria used in assigning a PNT intended for PMF to a risk category would be dependant on the species of the production platform in question as well as the commercial production system. For example, the risk of a PNT intended for PMF of being invasive of natural or managed ecosystems could be considered as minimal or even negligible for a PNT whose species propagates itself by seed if the PNT in question is harvested prior to flowering or seed set. However, participants also stressed that unless there is impartial regulatory oversight of commercial PMF, the use of these criteria will be ineffective in the management of risk.

Some participants did not feel that the biological activity and stability of the novel compound in the PNT should be used in assignment it to a risk category. Even if the novel compound is expressed as an inactive precursor, it was felt that activation could accidentally occur during production, extraction, or shipping, and maybe even via spontaneous mutation. With this in mind, it was felt that the novel compound should always be assumed to be in its biologically active form. Furthermore, the toxicity and allergenicity of possible degradation products of a novel compound should also be used as a criterion in risk assignment.

Some felt that using the presence or absence of the novel compound in the food and feed chains as a criterion for assignment was inappropriate. These respondents felt that the AP of a PMF crop expressing a novel compound that is already present in the food or feed chain could result in a potentially toxic level of the novel compound for humans and/or livestock. In addition, higher levels of a particular novel compound (whether in its native or modified state) than what is normally seen for a particular plant species could also change the PNTs interactions with its surrounding environment(s), for both in the short and long term (e.g. interactions with symbionts, pathogens, insects, soil organisms, etc.). It was not felt that companies/applicants would be able to collect safety data that would cover all these possibilities.

Given that risk = hazard x exposure, it was felt that the hazard of the PNT intended for PMF would be determined or quantified by the potential toxicity and allergenicity of the novel compound, and that the required information to make this determination is currently available. However, it would be more challenging to quantify the “exposure” variable at this time as it was not felt that there are currently proper scientific approaches to do so. Exposure would include variables such as the management of seeds (transport, seeding, harvesting, storage, processing, etc.) as well as potential for gene flow (vegetative reproduction versus insect or wind-pollinated, versus self-pollination). It was recommended that the CFIA fill the knowledge gaps that still exist in the fundamentals of reproductive biology, such as pollination biology. It was believed that these fundamentals would enable better prediction the behaviour of commercial fields of PNTs intended for PMF.

Some participants felt that an international multi-disciplinary group of scientific researchers, similar to that set up for climate change, should be established. This group's mandate would be to review scientific research as well as practical experiences with respect to potential short- and long-term impacts on health and environmental safety of genetically modified organisms and report back with their findings and recommendations.

Concern was raised that risk assignment seemed to be largely dependant on the novel compound's potential toxicity and allergenicity. It was stressed that available risk management strategies should also be considered when assigning a PNT to a risk category. The example provided was a therapeutic compound being produced in a plant which has been assessed as not posing a measurable hazard to humans, but its potential presence in the food supply could be extremely upsetting for both the consuming public and Canada's trading partners. The commercial production of the PNT expressing this therapeutic compound would require “extra” risk management measures that could go well beyond the PNT's species ability to be segregated and confined.

4.3 Acceptability of Food and Feed Crops as PMF Production Platforms

4.3.1 Which elements/characteristics would enhance the acceptability of food and feed crops as PMF production platforms?

• Ease of confinement and segregation (A)
• Use/inclusion of biological containment measures (e.g., the use of chloroplast transgenic systems or genetic use restriction technologies (GURTs) (B)
• Availability of effective risk management tools (e.g., inducible promoters, tissue-specific or growth phase-specific expression, etc. (C)
• Always acceptable to use food and feed species as PMF production platforms (D)
• Never acceptable to use food and feed species as PMF production platforms (E)
• Other (F)

Image - This image is a graph representing the number of participants in each sector selecting the element(s) or characteristic(s) they feel would enhance the acceptability of using a food or feed crop as a production platform for Plant molecular farming.

Figure 5: Graph representing the number of participants in each sector selecting the element(s) or characteristic(s) they feel would enhance the acceptability of using a food or feed crop as a production platform for PMF.

Some participants felt that if the Canadian regulatory system is truly science-based and case-by-case, it cannot make a generalization regarding whether it is always or never acceptable to use food or feed crops as production platforms for PMF. They highlighted that a main consideration when developers choose a food or feed species as their production platform is the overall risk associated with the PNT they intend to develop. In some cases, the risk to environmental safety and human and animal health could be no greater than the PNT's non-PMF counterparts (e.g., a PNT developed to express a modified starch or oil composition for industrial purposes, but is equivalent in terms of food and feed quality as its non-PMF counterparts). In other cases, where the PNT's food or feed quality would be compromised, an assessment of appropriate risk mitigation tools or strategies would be necessary to minimize any risks to the environment and to the food and feed chains.

However, participants felt that no food or feed crop should be chosen as a production platform for PMF if it is a crop species that is deemed critical to Canadian agriculture (e.g. corn, wheat, soybeans) unless the developer provides the GoC with a compelling justification as to why the species was chosen as the production platform. In cases where a valid justification is provided, the GoC should consider only allowing these production platforms in containment facilities.

Other factors suggested that would enhance the acceptability of food and feed crops as PMF production platforms included:

• the presence of a visually distinctive characteristic, such as colour, to facilitate the early detection of a failure in the segregation system
• the inclusion of redundant measures of biological containment
• the use of multiple risk management tools for commercial cultivation (geographic restrictions, temporal and spatial isolation techniques, use of dedicated planting/harvest machinery and grain elevators, storage and transport bins, on-going monitoring, etc.)

Some participants believe that the use of particular biological containment tools, such as GURTs, actually increases the perceived and real risks of using a particular food or feed crop species due to the potential of GURTs to be introduced into regular (non-PMF) crops.

Several participants preferred that food and feed crop species not be used as production platforms for PMF, but felt it could be acceptable for low risk compounds or where the production platform is a plant species which can be easily segregated (e.g., potatoes). Other felts that a limited use of food and feed crops could be acceptable, but only in containment settings as they could not conceive of any food/feed PMF crop which would pose a low enough risk to be cultivated outside of containment facilities. One participant expressed concern over the loss of effectiveness of biological containment tools (and hence confinement) resulting from spontaneous mutations occurring in the PNTs.

Some participants felt that none of the above-mentioned characteristics would substantially reduce the risk associated in using food or feed crops as production platforms. Others also stated that the effectiveness of many biological containment tools, including GURTs, has yet to be proven and that as long as there is risk associated with using food or feed crops, however minimal, their open-field cultivation is not acceptable. These participants expected the cultivation of such crops to only take place in containment facilities, where the entire production, from planting to extraction, purification and packaging, should take place under one roof.

4.3.2 Should the acceptance of food and feed crop species as PMF platforms be based on the risk of the expressed novel gene product to the environment and to human and animal health?

Image - This image is a graph representing the number of respondents in each sector agreeing or not regarding whether the acceptability of a food or feed crop species as a production platform for plant molecular farming should be primarily based on the novel compound's risk to human, animal, and environmental safety (as opposed to other criteria, such as ease of confinement or segregation).

Figure 6: Graph representing the number of respondents in each sector agreeing or not regarding whether the acceptability of a food or feed crop species as a production platform for PMF should be primarily based on the novel compound's risk to human, animal, and environmental safety (as opposed to other criteria, such as ease of confinement or segregation).

Some participants stated that the use of a food or feed crop species as a PMF production platform should not always be automatically considered as being unacceptable; their acceptability as a production platform for a particular novel biomolecule should be assessed on a case-by-case basis, an approach consistent with Canada's regulatory framework for new substances (which includes the regulation of products of biotechnology).

Some participants agreed that the acceptance of food and feed crop species as PMF platforms should be primarily based on the novel compound's risk to human and animal health safety as well as environmental safety. However, they added that, where potential risks to safety are identified, the benefit of the novel compound as well as the benefits of producing the novel compound in a food and/or feed crops species should also be considered in assessing acceptability. In addition, some participants suggested that other factors also be considered when assessing the acceptability of a food or feed crop species as a production platform for a particular novel biomolecule; such factors included:

• impact of PNT on biodiversity
• availability of effective risk mitigation tools for the plant species in question
• reproductive biology of plant species and potential gene flow to domestic and wild relatives
• ease of segregation of PNT from non-PMF commodities/counterparts
• potential for PMF gene products to interact with native gene products
• plant tissues/parts where novel product is expressed (i.e., part/tissues used for food and feed versus those not use in food or feed)

Other participants argued that acceptability of food or feed crops as PMF production platforms is also dependant on whether the regulatory oversight is considered as effective and efficient during:

• the authorization/licensing/registration process
• the growing and processing of the crops
• a response to an event of contamination of the food supply with a PMF product.

Furthermore, some also reasoned that part of the acceptance will be the ability of technology providers, developers, and growers to demonstrate that they can meet their responsibility in ensuring that the safety and integrity of the food and feed supply will not be compromised by PMF crops.

A number of participants stressed that short and long term risks related to the use of any plant species as a production platform should be assessed using research done by independent groups and not that of industry.

Some participants felt that if the definition of risks was based only on potential allergenic or toxic effects of the novel biomolecule, then food and feed crops should not be permitted as production platforms for PMF. Others felt that consequences from the potential adventitious mixing of PMF crops with commodities destined for the food and feed chains are much too substantive to allow any food or feed crop to be used as a production platform other than for expressing nutraceuticals or other to be used as functionally enhanced crops. Some also stated that the potential risk of losing existing markets and public trust in this technology is far too substantive to allow the use of food and feed species as production platforms.

4.4 Risk Management

4.4.1 The PBO is considering using the Canadian Seed Growers Association (CSGA) guidelines as a basis for developing the relevant reproductive isolation distances for commercial PMF fields. What other information should the PBO be incorporating or considering when establishing these required isolation distances?

Several participants cautiously agreed with this approach but emphasized that the CSGA guidelines should only be used as a starting point in developing adequate reproductive isolation distances for commercial PMF as the CSGA guidelines were designed for maintaining varietal purity and not for preventing material escape from a particular field. These participants were particularly concerned about PMF crops considered to be high risk (in terms of outcrossing potential and/or the novel compound expressed), for which they expected stricter reproductive isolation mechanisms.

A few participants noted that the isolation distance imposed will influence the environment's exposure to a PNT. Given that risk = hazard x exposure, it was felt that the same overall risk could be maintained by imposing varying isolation distances. As a result, the appropriate isolations distance for a given PNT could be more or less than what is currently prescribed in the CSGA guidelines for a particular plant species. For example, smaller isolation distances may be appropriate for PNTs intended for PMF that include proven means of genetic confinements whereas isolation distances would likely be much greater for production platforms which are insect-pollinated.

It was also recommended that the CFIA ascertain the scientific criteria that were used in developing the isolation distances listed in the CSGA guidelines. These criteria could then be supplemented with additional criteria based on data from gene flow studies when determining the acceptable isolations distances for commercial fields of PNTs intended for PMF. In addition, data regarding how often the impurity thresholds are exceeded despite the use of the CSGA guidelines should also be taken into account when evaluating the potential for a breach in confinement.

Several participants strongly opposed the use of the CSGA guidelines as a basis for developing appropriate isolation distances for commercial PMF as they were designed to help growers meet the established standards for varietal purity needed by seed lots to be certified and were not designed to prevent/minimize potential safety concerns. In particular, some participants stressed that the CSGA guidelines allowed for a percentage of contamination, whereas the isolation distances for commercial PMF should be striving for achieving as close as possible to 0% contamination.

Some felt that given the possible risks associated with PNTs intended for PMF, the isolation distances prescribed for these plants should taken into account criteria such as:

• the amount of food crops cultivated and the potential sexually compatible wild species in the intended growing region of the PNT intended for PMF
• the current available segregation and handling mechanisms/systems
• the possible implications for neighbouring fields/farmers if there is pollen spread
• the relative risk of the expressed trait of the PNT intended for PMF
• the expression pattern of the novel traits (e.g. constitutive expression versus tissue-specific expression, presence/absence of gene or gene products in pollen, etc.)
• the ability to predict the behaviour of insect pollinators (e.g. distances and patterns travelled) and non-target organisms (e.g. wild life carrying material off site)
• the possibility of inclement weather and its adverse effects on genetic confinement.

Other suggestions included:

• requiring literature reviews on gene flow studies for each new plant variety used as a novel production platform for PMF
• requiring PNTs intended for PMF to be infertile or male-sterile
• only permitting open-field commercial cultivation of PNTs intended for PMF in geographic areas which are outside the traditional production regions for that crop .
• requiring all growers intending to cultivate a PNT intended for PMF to notify neighbouring land owners and get their written consent for each field in question.
• limiting PMF in Canada to containment facilities until there is more information or in-depth studies/analysis regarding the open-field commercial PMF in other countries
• developing guidelines for crops not currently covered under the CSGA guidelines
• commissioning modelling studies on the various potential routes for environmental contamination and contamination of the food and feed supply
• forming an independent group of academics (with no vested interest in the industry) tasked with developing a mechanism for determining the appropriate isolation distance for individual commercial PMF fields

Some participants expressed some concern regarding quality control for pharmaceutical drugs produced in plants; namely, that the safety and quality of these compounds would be reduced if the open-field cultivation of such plants was to occur. Note: The CFIA is responsible for regulating the environmental release of PNTs, including those producing pharmaceutical drugs. However, any new drug manufacturing process, whether for an existing drug or a new one, is required to be assessed by Health Canada prior to being implemented to ensure that established quality and consistency standards are being met by the new manufacturing process. For pharmaceutical drugs being produced in a plant, the manufacturing process would include the cultivation of the plant producing it. Therefore, a company would be required to obtain approval not only from the CFIA, but Health Canada as well prior to commercially cultivating a plant producing a pharmaceutical drug. Unless the company's manufacturing process meets Health Canada's rigorous standards, such a manufacturing process would not be approved.

Some participants felt that, in the absence of biological containment tools, no isolation distances could ever be great enough to protect wild species and other crops. They noted that there are some biological containment tools available for many plant species. One example given was chloroplast transformation in which the novel gene would not be present in the pollen, hence preventing the potential for gene bridging. Another possible option would be expressing the novel compound from tissues other than seeds. These tissues could be collected from the field prior to flowering or pollen shed.

Another concern raised was whether the GoC will have the required resources to be able to effectively regulate this agricultural activity. For example, what control points would be set up for the GoC to be able to detect and respond quickly in the event that a breach in confinement occurs? How will the GoC minimize the possibility of errors occurring?

Other participants raised the fact that commercial PMF has the potential to bring economic diversification for the production platforms; however, this diversification could be inhibited if overly stringent isolation distances are mandatory. It was suggested that isolation distances should be enforced between all varieties of PMF plants of a particular species and all food and feed plants of the same species and that isolation distances between different PMF crop varieties of that species should not be imposed.

4.4.2 On what scientific basis can limits, if any, be applied to the size of commercial PMF fields?

Participants suggested the following criteria as scientific basis for limiting the size of commercial PMF fields:

• ease of physical segregation of a particular PNT intended for PMF from its non-PMF counterparts and other crops intended for use as food or livestock feed (e.g., whether it is distinguishable from its non-PMF counterparts, its potential weediness and/or seed dormancy, the ease of PMF volunteer management, etc.)
• potential for escape of genetic material through gene flow to sexually compatible non-PMF crops and wild plant species (e.g., cross-pollination versus vegetative reproduction)
• potential for exposure of NTOs (including humans) to the novel compound ( i.e. in what tissues is the novel compound being expressed?)
• the intended geographic area for the PNT's cultivation (i.e., Is the area located in the Canadian production range for non-PMF varieties of the same plant species?)
• effectiveness of biological containment tools included in PNT intended for PMF
• field orientation (for wind-pollinated crops)

Given that the effects of field size on insect movements (and resulting levels of cross-pollination) are not yet fully understood, it was recommended that commercial PMF fields should be restricted to areas where commercial production of sexually compatible non-PMF plants is limited, thus minimizing the potential exposure of non-PMF crops. In addition, since the level of gene flow from pollinators bringing in genetic material from outside a field is a function of its size, smaller non-PMF fields would be more vulnerable to cross-pollination than larger ones (Cresswell & Osborne 2004, J. Applied Ecology 41: 539-546). Consequently, it was suggested that PMF crops should not be planted beside non-PMF fields (of a sexually compatible species) which are below a certain size.

Several participants felt that restrictions on field sizes should not be based on scientific criteria but rather on segregation and agronomic practices criteria, as all commercial fields will vary in their characteristics (e.g., physical landscape of field, biology of crop being cultivated, potential toxicity and allergenicity of novel compound to NTOs, types of NTOs present, geographic location, etc.). In particular it was felt that restrictions on size should be based on the ability of the producer to meet all the terms and conditions of confinement imposed on the commercial field in question.

For example, it is expected that the larger the field, the greater the pollen cloud produced. Consequently, larger fields might require greater isolation distances to ensure the segregation of the PNT intended for PMF from related domestic and wild sexually compatible species. If an imposed isolation distance is so great that proper monitoring and maintenance of isolation distance is not possible, it was felt that field size should be restricted to a size where the related isolation distance can be properly maintained. Another example would be a case where the field is so big that the producer/grower is unable to harvest the field all at once making the logistics of confinement more difficult.

Restrictions on size could also be imposed based on:

• landscape (streams/rivers, trees, etc.)
• whether or not field is located within the Canadian production range of non-PMF varieties of the same plant species
• whether fences are present to help prevent wildlife from moving material off site (although this would not help prevent material being moved off site by birds, wind, or insects)

In addition, it was also recommended that statistical epidemiological data and the outcome of risk assessments be considered when assessing whether field size restrictions should apply.

Other comments

Some participants felt that, initially, the size of all commercial field sizes should be restricted until there is sufficient published, peer-reviewed field data (under confinement conditions) available to determine, on a case-by-case basis, whether limits should be imposed on field sizes. Others stated that they would prefer to see large fields of PMF crops concentrated in one area rather than a series of smaller fields spread out, as this would reduce the potential for gene bridging.

Others felt that the field size or size of containment facility used should not be an issue so long as all PMF plant material, including roots, is accounted for. However, if plant residues of PMF crops could impact soil biology or ground water, these participants wanted appropriate safe guards to be put in place.

Some felt that commercial PMF should only be carried out in containment facilities (i.e., no open-field commercial cultivation of these plants should occur) as gene flow with sexually compatible wild species and domestic crops cannot be 100% prevented or avoided.

4.4.3 The cultivation of PNTs intended for PMF may affect existing crop rotations used by farmers/producers. Are there particular crop species that would have large impacts on crop rotations in Canada if they are used as a platform for PMF? Would the impact be dependant on whether the PNT is expressing a pharmaceutical compound compared to an industrial compound?

Several participants stated that using corn, canola, as well as cereals (e.g. wheat, barley, and oats) as production platforms for PMF could have larger impacts on crop rotations. In particular, difficulty in volunteer management, harvest losses, long seed dormancy, propensity to weediness, as well as the potential for gene bridging for highly out-crossing species (in particular Brassica spp.) were all cited as reasons for their greater impact. Others felt that any crops that are persistent in the environment, such as forage crops, would have a greater impact on existing crop rotations.

It was stressed that crop management practices were of great importance in answering this particular question. For example, it would be simpler to destroy any seed having fallen to the ground during harvest if these seeds are left on the surface of the soil instead of being ploughed into the soil. Concern was also raised with respect to stacking herbicide traits along with PMF traits as herbicides would play a significant role in volunteer management of PNTs intended for PMF.

Some participants felt that the impact would also be dependant on the actual trait expressed (irrespective of whether it is pharmaceutical or industrial) and whether there is a potential for the novel compound to leach into soils. Other factors that could impact existing crop rotation include the agricultural sectors that exist in the intended growing region of the PNT intended for PMF as well as availability and application of control measures for PMF volunteers. In addition, the ability to be able to easily visually identify PMF volunteers in subsequent crops could also have an impact on existing crop rotations. For example, to be able to easily identify a corn volunteer, subsequent crops may have to be ones with shorter plant heights, such as alfalfa or peas. In other words, these participants believed the actual impact of a PNT intended for PMF on a crop rotation could only be determined on a case-by-case basis.

Some participants stated that existing crop rotations would be impacted irrespective of the species of the PMF production platform, as a crop rotation which includes a PNT intended for PMF would likely be subject to the same guidelines for crop rotations as confined research field trials.

Some participants strongly opposed the cultivation of PNTs intended for PMF in rotation with crops intended for food or livestock feed; in particular, they felt that, even with aggressive monitoring programs, this would contribute to a higher adventitious presence of PMF crops in the food and feed chains.

4.4.4 It is expected that those applying for authorization of commercial production of PNTs intended for PMF will be required to submit a contingency plan in the event an unintended release occurs. What components should be required to be included in such a plan?

Participants suggested that the following components be required to be included in a contingency plan for the open-field commercial production of a PNT intended for PMF:

• a reporting mechanism for both PMF and non-PMF producers to be able to notify of any uncovered problems or unintended releases (including names and contact information for key contacts)
• a clear reporting process and quick turnaround time to alert the appropriate regulatory authorities (federal, provincial, and/or municipal) as well as a detailed plan on how the company will work collaboratively with the regulatory authorities
• a strategy for public notification of the unintended release
• a traceability/tracking system to link shipments of bulk or processed products back to their original source, which will include standard operating procedures for record keeping as well as tracking and recovery of material (this system will help determine the extent of spread and level of contamination and isolate affected material from food and feed supply chain)
• a notification strategy to producers/exporters potentially affected by an unintended release (which would be determined with information gathered with the tracking system)
• a validated detection method specific to the PNT intended for PMF and sensitive enough to detect it at low levels (to determine whether the PNT is present in suspected contaminated shipments/fields and for on-going monitoring after removal and clean-up is completed to ensure that it is no longer present); such method(s) must be useable under a range of conditions (e.g., in the field, grain silo, production facility) and not just in a specialized laboratory.
• a recall plan for removing all contaminated material (i.e., from fields, in transit, and in the market place, if applicable); such a plan should include a lock down procedure that would be triggered along the entire chain of custody
• a disposal plan for the destruction of all contaminated fields, harvested plant material and processed products
• a long-term monitoring plan of contaminated areas once clean up is done
• protocols for both internal and CFIA audits of contingency plan once it is put into action

Several participants indicated that contingency plans should include a clause stating the company's commitment to being financially responsible for all costs related to environmental/market place clean up as well as to compensate all affected farmers/producers of non-PMF commodities. Others felt that a legal requirement should be set up to prevent a company from declaring bankruptcy prior to compensating farmers and completing any required clean-up.

Some participants highlighted the fact that the response to an unintended release would be largely dependant on where and when a system failure occurred in a production operation. They believed that a response resulting from the accidental commingling of harvested PMF crops with commodities intended for the food and feed system could be quite different from the response resulting from an identified leak in a transport container. In addition, some felt that contingency plans should include responses to an unintended release due to sabotage, theft, or other emergencies. Therefore, these participants felt that contingency plans should be designed not only for each production stage but also take into account the different courses of action that would be taken depending on the nature of the failure. In addition, participants stated that companies should be required to implement a system of prevention, not simply contingency plans alone.

Other participants felt that the components that are currently required as part of a contingency plan for confined research field trials were adequate while others suggested that required components of a contingency plan for individual PNTs intended for PMF should be relative to their associated potential and real risk(s).

Some participants expressed concern over the feasibility of detecting, in a timely manner, an unintended release and want to know what type of monitoring measures the GoC would require from companies to ensure that any unintended release was identified as early as possible so as to minimize any contamination of both fields, as well as raw and processed products. Others expressed scepticism that a complete clean-up of an unintended release could ever be achieved, insisting that there could always be a very low level of AP in the environment as well as in the food and feed supply chains.

Some participants voiced their disagreement regarding delegating the responsibility of developing such contingency plans to the industry. They believed these plans should be developed by impartial independent working groups. Note: The submission of such a contingency plan is expected to be part of an application for the approval of the commercial cultivation of a PNT intended for PMF. No such approval would be granted until the GoC, among other information requirements, would be satisfied that the submitted contingency plan meets all the requirements outline in relevant technical guidelines (guidelines that would be developed in consultation with all stakeholders).

4.5 Segregation and Handling of PMF Crops

4.5.1 Under what circumstances can the current segregation and grain handling systems deal with PMF commodities (e.g., the production of a vaccine in a grain crop within a closed-loop production system)?

Some participants felt that the current handling and segregation systems used within closed-loop production systems would be sufficient to safely handle many PMF commodities. For example, some current crops, such as Ontario soybeans, have well-developed segregation systems which could be used for most PNTs intended for PMF. Others felt that it would only be appropriate in cases where the PNT could be easily distinguished from other commodities (e.g. being a different colour).

It was suggested that the CFIA include a requirement for producers/companies to develop and submit standard operating procedures (SOPs) that would provide for the secure and segregated handling of PMF commodities. It was felt that these SOPs should be required to address:

• the reproductive isolation of commercial PMF fields (which could include not only temporal or spatial isolation, but also protocols for minimizing the inadvertent transport of material off site though worker clothing and footwear, wild life, etc.)
• the use of dedicated harvest and transport equipment (including proper cleaning procedures)
• the use of dedicated storage and handling facilities
• the initial and on-going training of personnel
• the methods for disposal of residual plant matter as well as cullage or off-grade material
• the record-keeping system (including how they will keep thorough records on crop pedigree, seed sales, acreages planted, quantities marketed and quantities of cullage or off-grade material, as well as the disposal of cullage or off-grade material)

A suggested requirement was for producers/growers to specify a time period for the handling of PMF material, where an inspector could be called on-site to monitor the handling and clean-up afterwards. Some participants felt that the experience with current identity preservation programs or programs implemented for the production of organic crops could provide some guidance in developing SOPs for commercial PMF. It was stressed, however, that the stringency of such closed-loop systems should be dependent on the relative level of risk associated with the particular PNT.

Several participants strongly felt that current grain handling and segregation systems are not adequate to safely deal with PMF commodities and should be handled in separate systems altogether; StarLink Corn and Prodigene corn were cited as examples substantiating that current systems are not equipped to handle this type of plant material. It was recommended that the CFIA only allow commercial PMF where the entire production system is either in-house or under contract to producers.

4.5.2 Under what circumstances could low risk PNTs intended for PMF be placed into the bulk handling system (e.g., a grain crop producing a novel compound already present in the food or feed systems)?

Some participants felt that if a low risk PNT intended for PMF can be assessed as:

• posing minimal or no risk to human and animal health when consumed at low levels (i.e., low level adventitious presence of the PMF crop in non-PMF commodities), or
• equivalent to its counterparts grown for traditional purposes in terms of environmental safety as well as human and animal health safety,

then it would be acceptable for this PMF commodity to be handled safely using the bulk grain handling system, as long as it's produced using an identity preservation (IP) system.

However, several participants felt that no PNT intended for PMF should be permitted to be handled in the bulk grain handling system regardless of the associated risk. In particular, it was felt that PNTs expressing a novel compound already present in the food or feed supply would be at a greater risk of being present in non-PMF commodities and their presence would be harder to detect in food or feed commodities. These participants stressed once again that even if a PNT is expressing a compound already present in the food supply, the potential dose that would be consumed from this commodity's low level AP in food or feed could be at a level high enough to be toxic to human and animal health. In addition, a particular compound could be approved to be present in specific foods, but could end up being present in foods for which it is unapproved.

A few participants felt that this was an issue that would be addressed by the market itself. For example, if there is a particular buyer requiring assurance that the product he or she is buying is free of a particular trait, then it would be up to the supplier to comply with the buyer's demands. Others felt that whether or not it is acceptable for PMF crops to be handled in the bulk grain handling system would be dependant largely on the agreed upon tolerance level(s) for the AP of PMF crops in non-PMF commodities. If the tolerance level is not achievable using the bulk grain handling system, then PMF producers would not use this system to handle their crops.

4.6 Other

4.6.1 Is there anything else of a scientific nature pertaining to PMF production in Canada that you would like to comment on?

Some provincial governments have indicated that they wish to be more involved in the CFIA's stakeholder consultation process with regards developing technical guidelines for the commercial environmental release of PNTs intended for PMF.