| Comment Record|
Dr. Dirk Maier ||
2003-02-06 23:56:13 |
Purdue University |
| Comments for FDA General |
1. General Comments
I am responding to FDA’s request for public comments on FDA’s Draft Guidance for Industry Drugs, Biologics, and Medical Devices Derived from Bioengineered Plants for Use in Humans and Animals (published in the Federal Register, September 12, 2002, Volume 67, Number 177, pp. 57828-57829).
Earlier this year I outlined my concerns about pharmaceutical traits in grains and oilseeds in a fact sheet (http://www.agcom.purdue.edu/AgCom/Pubs/GQ/GQ_47/gqtf47.html) based on over 20 years of first hand experience with food and feed grain production, handling and processing. Since then I have continued to study and learn more about this issue, discussed it with many colleagues, grain producers, handlers and processors, reflected on my experience as a member of the last two EPA StarLink FIFRA SAPs, reviewed various proposed containment protocols including ISO-like process-based quality management systems (including BIO's May 17 reference document), and studied the FDA draft guidance for industry on drugs, biologics and medical devices from bioengineered plants for use in humans and animals.
It is my professional opinion that the current FDA Draft Guidance is inherently flawed because it is based on the premise that large scale containment systems can be implemented that involve open pollinating plants grown on unsecured farm fields while assuring compliance with a zero tolerance for non-food/feed approved products that would be considered adulterants if they were detected in our food and feed crops, and their processed products. Given today's concerns over homeland security, an effective containment system must not only prevent unintentional compliance failures but also intentional sabotage that is targeted to cause physical and/or economic harm to the United States. Growing regulated products in unsecured farm fields exposes our food and feed supply to what I would consider an unacceptably high exposure risk to acts of terrorism.
My specific concerns with the FDA Draft Guidance as currently proposed are:
(1) In Section II.A., the FDA states that the developer “should give careful consideration to choosing the plant species that will be used as the source of the desired regulated product. … Concerns to be addressed include: …the measures to ensure confinement; and, if it is a food crop species engineered to produce non-food material, the measures to ensure that non-food (or non-feed) material will not get into food or feed. The presence of any such material in food or feed could render such products adulterated under the FD&C Act (21 U.S.C. 342).” In my opinion, food and feed crops should not be used to produce non-food/non-feed materials unless reasonably achievable, risk-based tolerance limits are established for the global crop marketing system below which trace amounts of these materials are safe to consume. The U.S. food crop production system is based on allowable limits for a host of sanitary and food safety attributes (including mycotoxins, foreign matter, pesticide residues, etc.) which allow the market place to function efficiently. The U.S. food and feed crop production system was never designed and is not capable to comply with a zero tolerance for traits that if detected would be considered adulterants. The recent market disruption due to the StarLink corn contamination of our domestic food and export supplies exemplifies how easily an adulterant can be introduced, how quickly it can spread and how long it can linger. The fact remains that if regulated traits are allowed in food and feed crops then it will not be a matter of whether contamination will occur but rather when it will occur, where it will occur, what will be its extent, and who will pay for the damages and clean-up.
(2) In Section C.1., the FDA “strongly recommends that you have tests available that can detect the presence of the target gene and the protein product in the raw agricultural commodity”. This should be a requirement (must) not a recommendation (should). Unless required, there is no economic incentive for technology providers to release test kit or analysis tool. The burden of proof and costs of testing will fall on those trying to assure buyers that regulated traits are not present in their commodity and specialty cops. Test kits must be available at the time of the first APHIS-licensed field experiments.
(3) In Section III.A., the FDA states that “confinement measures […] may be needed to control the spread of the bioengineered pharmaceutical plants and to keep them from entering the food or feed supply.” There is no doubt that confinement will be needed and should be required as part of any permitting process whether for experimental or commercial uses.
(4) In Section III.A., the FDA further states that “for most initial experiments and commercial uses of these plants, a USDA/APHIS/BRS permit will be needed.” If a crop does not have full food and feed approval, federal oversight must always be a requirement. Thus, it is unclear what products would be exempted from permitting. Also, the permitting process must also include enough federal auditors to be in the field and confirm that protocols are followed.
(5) In Section III.C.1., the FDA states that “you should implement procedures to ensure that such a plant line is used only for its intended purpose as a source material for a regulated product”. This should be a requirement (must) not a recommendation (should) (see item 1 for reasons).
(6) In Section III.C.1., the FDA further states that “when a plant species […] for food or feed is bioengineered to produce a regulated product, you should consider the use of strategies that allow the bioengineered pharmaceutical plant line to be readily distinguished from its food or feed counterpart.” Although it appears desirable it will likely be totally ineffective. For example, if pharma corn was genetically altered to be of a different or distinct color, the inadvertent mixing with food corn or the intentional insertion of random kernel amounts a few inches below the grain surface of a storage structure could occur. One would have to run potentially huge amounts of grain through a color sorting machine to find them, which are machines that have less than a 100% accuracy. There are unlikely enough color sorters available in the world -- and none in the U.S. corn processing industry -- to handle such a vast undertaking. This would also be extremely impractical and cost prohibitive given that one typical 40,000 bu farm bin holds over 3 billion kernels of corn.
(7) In Section III.C.1., the FDA states that “for such plants that outcross, you may want to consider growing them in regions of the country where little or none of its food/feed counterparts are grown”. Corn is our major food and feed crop valued at $19 billion per year. Any adventitious presence of regulated traits due to cross-pollination, inadvertent or intentional commingling would have a huge economic impact on U.S. corn producers, handlers, processors and exporters, as well as on the overall U.S. economy. The closer geographically crops with regulated traits are grown to “commodity” crops, the higher the probability that contamination will occur. Thus, corn with regulated traits should not be allowed to be grown in any of the major U.S. corn regions unless reasonably achievable, risk-based tolerance limits are established (see item 1 for reasons).
(8) In Section III.C.1., the FDA states that “measures should be in place to ensure that there is no inadvertent mixing of the bioengineered pharmaceutical plant with plant material intended for food or feed (including inadvertent mixing with seeds for food or feed crops).” Although inadvertent and unintentional commingling is a significant concern, the intentional mixing is of even greater concern. Let me outline a scenario, which I would have dismissed as possible but purely speculative if it had not been for Sept 11, the anthrax scare and the current War on Terrorism. The primary vulnerability that is easily overlooked is that crop production fields are typically not secure from outside intruders. Sufficient public information exists to identify the producers and companies who have APHIS permits and to pinpoint the areas where they may be growing these regulated bioengineered crops. For example, corn fields are easily identified because the current APHIS isolation requirements are so stringent that a regulated pharma corn test plot stands out in a larger field either because it is in the middle of a soybean field or it is temporally offset in a corn field and remains standing long after the commercial corn has been harvested. Corn physiology indicates that once kernels form on the ear, protein production is initiated and continues until physiological maturity, which under typical Corn Belt conditions may last from mid-July until mid-September (about 60 days). Once maturity has been reached, ears are left on the plant for field dry down, which may typically take another 2-4 weeks (15-30 days). Thus, the regulated pharma trait is available for the taking from an unsecured farm field for 75 - 90 days. A terrorist or any malicious intruder (including a competitor) could easily enter a field undetected and steal a number of ears. Having walked through many corn fields and removed ears from plants to assess grain quality problems before harvest, I can assure you that it would not take more than 5 minutes for one person to snap off 50 ears, toss them into a bag, and run off with them. There are up to 25,920 five-minute periods during which such a theft could thus take place in a regulated but unsecured pharma corn field. Just imagine how many ears a small group of intruders could collect!? What could a terrorist do with those ears? Each ear has typically about 500-700 kernels, which weigh about 200 grams total; 50 ears would yield about 10 kg of regulated product (which is significantly more than the amount of stalk material that was suspected to be contained in the 500,000 bu soybean tank that was quarantined by APHIS in Nebraska in November 2002). The ears are easily hand-shelled, and small amounts of kernels could be tossed randomly into various food and feed corn bins on farms and elevators, as well as into rail cars and river barges throughout a fairly wide geographical region of the Midwestern Corn Belt. This could take place within a few hours or days and long before anyone would detect missing ears from corn plants in the test plot. As a matter of fact, a terrorist would not even have to go through the trouble of actually tossing kernels into these generally unsecured storage structures and transport vessels but would only need to claim to have done so because there is no way to detect and differentiate these corn kernels from others. If the FDA and APHIS reacted to such a breach in containment in the same manner as it did in November 2002 to the apparent breach in compliance when voluntary pharma corn plant material was suspected in commodity soybeans, they would have to quarantine potentially all of the corn in the U.S. Corn Belt, and depending on the time between theft and threat, initiate a massive recall of all corn-containing food and feed products as well as of all in-transit export corn shipments. On the other hand, if the government declared an emergency exemption to quarantining because it considers pharma corn in such potentially minute quantities to be safe enough for humans and animals to consume, it would still be an economic disaster. Our export customers would respond as they did with the StarLink corn incident and refuse to accept shipments (millions of dollars in lost sales, depressed corn prices throughout the Midwestern Corn Belt), and our food manufacturers would still have to consider the liability associated with false consumer claims of getting sick from possibly contaminated products. This is a no-win situation that would result in huge liability, testing and cleanup costs in order to prove that no trace amounts above detectable limits are in the food and feed supply system. The above scenario becomes an even greater concern if these pharma crops were commercialized on more acres and a wider geographic area using the proposed FDA/APHIS approach of allowing production in unsecured farm fields with recommended rather than required procedures and a lack of federal oversight.
(9) In Section III.C.3., the FDA states that “you … must have control over the growing process from planting through harvesting and over the disposition of remaining crops and/or crop residue and, if required, over the subsequent use of the field if for growth of food or feed or as a pasture during subsequent seasons.” It is unclear what defines control. For example, production in open and unsecured fields allows for little control during the growing process as strong winds could carry pollen or plant material for miles, as animals could scatter protein-carrying seed from bioengineered plants well beyond field borders, and as human intruders could steal the regulated product an sabotage our crop marketing and processing system (see Item 8 above).
(10) In Section III.C.3., the FDA further states that “all persons involved in field growth of the product should be adequately trained to perform the duties for which they are responsible.” This should also be a requirement (must) not a recommendation (should). I would strongly urge that all persons involved in producing a regulated product must be federally licensed in a similar manner as applicators of restricted use pesticides must be state licensed in the U.S. Licensing would assure consistent and continuous professional training and a database of such professionals. Training could be offered by third parties including state ag departments and land grant universities.
(11) In Section III.C.3., the FDA “recommends that you consider the use of perimeter fencing to help exclude wildlife and escaped livestock.” Not only should this be a requirement, but the perimeter fencing should consist of high security fencing to avoid the willful intrusion of bioterrorists and avoid the scenario described under Item 8).
(12) In Section III.C.4., the FDA does not address the bulk handling and storage of the harvested material in any detail. It should be pointed out that typical conveying and storage equipment for handling agricultural grains and oilseeds does not contain every kernel and does not clean out easily. Thus, industrial standard, fully enclosed stainless steel equipment should be required for conveying. Storage structures and transport vessels should be sealed. Biosecurity concerns also need to be addressed while the bioengineered product is stored at the production site even temporarily.
(13) In Section III.C.6., it is unclear how the FDA will determine how and under what conditions waste material might be allowed to enter the human or animal food chain. It appears that even if exclusion of regulated traits were successfully accomplished during the production period in the field, contamination could occur if waste material was allowed under any circumstances to enter the food and feed processing stream. It would seem best to categorically exclude this option because contamination of food and feed products could occur if inadvertent trace amounts were present in such waste material.
(14) In Section D. 2., the FDA “recommends that you follow current Good Agricultural Practices.” This should be a requirement (must) not a recommendation (should).
(15) In Section D.3., the FDA “recommends the use of dedicated equipment… [and] … that equipment-cleaning procedures be developed and that cleaning agents used on harvesting equipment be described”. This should be a requirement (must) not a recommendation (should).
(16) In Section D.4., the FDA states that “if the harvested source material is to be stored prior to further processing, the storage conditions (e.g., temperature, humidity, volume, density, storage time, etc.) should be fully described in your application”. An important consideration with grain crops is their harvest moisture content. Most of the U.S. corn crop has to be dried from typical harvest moisture contents of 20% or higher to recommended safe storage moisture contents of 14.5% or below. Drying of corn involves the movement of large volumes of heated air and results in the exhausting of particulate matter. Grain dust typically contains traces of protein and thus will contain traces of regulated traits. Therefore, emission control from regulated crops dryers needs to be addressed.
It seems to me that given the fact that plant-made pharmaceuticals will always be regulated products, the pharmaceutical industry would only adopt this technology in the future if it were able to impose total control and security over the entire regulated product chain in order to manage risk, reduce liability, and prevent intellectual property theft. The only way to achieve such control and security will be to grow these pharma crops inside fenced or walled compounds, and if gene flow remains a liability issue, to add a roof on top of the walls. If that is in fact the likely scenario for future commercial production of regulated PMPPs, why risk experimental trials in unsecured fields where both gene flow and terrorism will make our food and feed supply so vulnerable? The final FDA Guidance to Industry should thus be consistent with the strictest commercial scenario that will most likely be required.
In light of what the likely commercial future of PMPPs will be as far as the pharma industry is concerned, I am beginning to worry more about the possibility of non-regulated industrial traits (PMIPs) that may be commercialized without food/feed safe approval. My understanding is that several of the initial products would likely test to be food and feed safe if those tests were required. If that is indeed the case, I would like to recommend that the FDA and APHIS require food and feed safe approval of these industrial traits as part of the approval process and before commercialization. Given the potential for the commercialization of many different future industrial traits, the more of them that are commercialized with food and feed safe approval, the more farmers will have an opportunity to grow them using reasonable identity preservation (IP) protocols and thus be able to participate in these value-added opportunities. Otherwise too many farmers will likely be shut out because they cannot afford to follow overly stringent containment protocols or their local buyers may threaten not to accept their commercial grain because of contamination concerns. Even though I remain philosophically opposed to using our food and feed crops for the production of any regulated traits that are not food and feed safe, I can see myself compromising on this issue if -- and that is a big if –those industrial traits that will not pass the food and feed safe test were regulated (similar to PMPPs), that reasonably achievable and risk-based thresholds were established for their inadvertent presence in commercial crops, and that these regulated pharma and industrial traits were only allowed to be grown on secured fields (preferably with walls and roofs) outside the major food and feed crop production areas.
One of the lessons from Sept 11 is that terrorism succeeded by turning our own technology against us with relative ease. Going down the path of allowing non-food/feed uses of our food and feed crops in unsecured farm fields would potentially allow terrorists to turn this tremendous biotechnology tool into a disastrous and disruptive economic weapon against the United States. Having not seen this security concern addressed with sufficient seriousness in the proposed government guidelines and private industry containment system descriptions, I would like to encourage the USDA-APHIS and FDA to consider the reality of this potential security risk carefully. I am open and available for any follow up dialogue, and welcome your comments on my outlined concerns.
Dirk E. Maier, Ph.D., P.E.
Professor and Extension Agricultural Engineer
Post-Harvest Education & Research Center
Agricultural & Biological Engineering Building
225 S. University Street
West Lafayette IN 47907-2064
* Quality Grain is Job #1! *