I. GENERAL OVERVIEW
On September 19-20, 2010, FDA held a meeting of the Veterinary Medicine Advisory Committee (VMAC) to consider issues regarding the safety and effectiveness of the new animal drug that is the subject of a new animal drug application (NADA) concerning AquAdvantage Salmon produced by AquaBounty Technologies, Inc. (ABT).1 AquAdvantage Salmon are intended to grow faster than farm-raised Atlantic salmon that have not been genetically engineered.2
FDA simultaneously made available to the VMAC and the public an extensive “Briefing Packet” that contained a summary of the data and information on which the agency had drawn its preliminary conclusions. This Briefing Packet, copies of the presentations made at the VMAC meeting, and a copy of the VMAC chairman’s report can be found at VMAC Meeting Materials. Interested persons were invited to present data, information, or views, orally or in writing, on issues pending before the VMAC.
FDA received 322,031 written comments on the issues the VMAC considered for the AquAdvantage Salmon. Of these, 321,993 comments were form letters or general statements about GE animals. Thirty eight (38) comments were what FDA considers to be substantive because they provided specific analysis, recommendations, or opinions. The comments were submitted by a broad range of stakeholders, including consumers, academics, trade and professional organizations, consumer and environmental groups, and other developers of GE animals.
In addition, as part of the public comments on the draft Environmental Assessment (EA) and Preliminary Finding of No Significant Impact (FONSI) (Docket # 2011-N-0899 closed April 26, 2013)3, the agency received comments that, although substantive, were not relevant to the draft EA/preliminary FONSI, and in almost all cases, were similar to, and suitable for combining with, previously submitted comments to the VMAC. We have included those comments in the topic-based comment summaries, and addressed them immediately thereafter.
Additional information regarding the NADA concerning AquAdvantage Salmon, including the final EA and FONSI and Freedom of Information (FOI) Summary, can be found at AquAdvantage Salmon - Response to Public Comments on the Environmental Assessment; FONSI; and FOI Summary.
II. SUMMARY OF COMMENTS AND FDA’S RESPONSES
A. Food and Food Safety
i. Food Labeling
Comment Summary: Several comments addressed the issue of labeling food from AquAdvantage Salmon. Two comments requested that the Agency require food from AquAdvantage Salmon bear special labeling indicating it was developed using biotechnology, while others opined that special labeling should not be required in the absence of a nutritional difference in the food.
Response: FDA held a public hearing on September 21, 2010 on labeling of food derived from AquAdvantage Salmon. The public submitted comments concerning labeling of food derived from AquAdvantage Salmon to the hearing docket. FDA is releasing separately a document that responds to these public comments. See Comments and Agency's Responses on the Public Hearing on the Labeling of Food Made from Aquadvantage Salmon. Please refer to that document for responses to comments concerning food labeling.
ii. Food Safety Assessment
Comment Summary: Several comments addressed the methods FDA used to assess the safety of food from AquAdvantage Salmon. Some comments disagreed with various aspects of FDA’s approach to the safety assessment of food from AquAdvantage Salmon or requested additional studies, analyses, or the development of a new system consisting of a public committee of food safety experts to decide the components of the food safety assessment. Others found FDA’s process for food safety assessment of AquAdvantage Salmon to be rigorous and appropriate.
Response: FDA’s food safety analysis for food derived from AquAdvantage Salmon meets applicable requirements of the Federal Food, Drug, and Cosmetic Act (FD&C Act) and FDA’s implementing regulations, and is science-based and thorough. Accordingly, we are not requiring additional studies to determine that food from AquAdvantage Salmon is safe. Under the FD&C Act and FDA’s implementing regulations, a new animal drug cannot be approved for use in food animals unless FDA determines that the drug in or on the food, and any substance formed in or on food resulting from the use of the drug, is safe for consumption. 21 U.S.C. § 360b(d); 21 CFR 514.1(b)(7). The standard that is used to determine the safety of food under the new animal drug provisions is the same as FDA’s standard for approving a food additive: reasonable certainty of no harm. See 21 U.S.C. § 360b(d)(2).
As discussed in Guidance For Industry 187: Regulation of Genetically Engineered Animals Containing Heritable Recombinant DNA Constructs (GF1 187), the food safety assessment for an NADA related to a GE animal focuses on the issue of whether food or feed derived from a GE animal is safe for humans or animals consuming edible products from the GE animal. The risks for food and feed safety can be divided into two overall categories. The first addresses whether there is any direct toxicity, including allergenicity, via food or feed consumption of the expression product of the article. The second category addresses potential indirect toxicity associated with both the article and its expression product perturbing the physiology of the animal. Accordingly, as part of the safety determination for food from AquAdvantage Salmon, we evaluated the health of the AquAdvantage Salmon, potential toxicity to the consumer, and potential allergenicity of the expression product of the rDNA construct, Chinook salmon growth hormone, along with changes in the levels of hormones that could potentially be affected by the expression of growth hormone. We also carefully reviewed the potential for unintended effects arising from the insertion of the rDNA construct, including possible effects on the composition and allergenicity of AquAdvantage relative to appropriate comparators, including farm-raised Atlantic salmon. As we explained in GFI 187, our approach to the safety assessment of food from GE animals, including AquAdvantage Salmon, is consistent with that recommended by the Codex Alimentarius Commission, the international food safety standard setting body, in its Guideline for the Conduct of the Food Safety Assessment of Foods from Recombinant-DNA Animals (CAC/GL 68-2008) (Codex rDNA Animal Guideline).
Comment Summary: One comment stated that in the food safety studies “[s]almon were screened visually for general health status and traits relevant to commercial marketability,” but that not all of those fish were tested and there were no data kept on the rejected fish.
Response: In a food safety study involving food animals, the general objective is to study such animals to determine whether they are suitable for consumption as food. In general, there are usually many more animals at the beginning of such studies than are selected as a sample for testing. The reason there are more animals in these safety studies is to assure that there are sufficient animals to meet the study acceptance criteria at critical time points in the study. This is done to account for normal losses due to disease, accident, or other factors that may result in losses from the population. Based on pre-established criteria, animals for inclusion in the study may then be selected from the larger sampling pool in a non-biased manner. Animals that are not selected are simply not used in the study, either because they fail to meet the acceptance criteria or because they are in excess of the number of animals required.
In ABT’s safety study on composition, salmon were screened to determine which ones had traits, such as visual appeal, that would make them commercially marketable as food. As a result, in the salmon food safety studies, those fish that were judged not likely to be used as food were not included in the study. This is consistent with general practices for animal food safety studies and FDA’s approach for eliminating from safety studies animals that are unlikely to be used as food. We also note that the safety study on composition was not the only data and information evaluated in our review. As explained in detail in the FOI summary, we evaluated data and information submitted by ABT, considered relevant peer-reviewed publications, and conducted our own analysis of certain data and information. Based on our review of all relevant data and information, we concluded that food from AAS is as safe as food from non-GE salmon and that there is a reasonable certainty of no harm of consumption of food from AAS. FOI Summary at Section IX.
iii. rDNA Construct Safety
Comment Summary: A few comments addressed the safety of the rDNA construct when consumed in food, and one comment raised the possibility of the transfer of DNA from the rDNA construct in AquAdvantage Salmon into the genome of gut bacteria.
Response: The genetic construct that is present in AquAdvantage Salmon is composed of deoxyribonucleic acid (DNA) which is present in all living things. DNA is a normal component of food derived from animals, plants, and microorganisms, regardless of the food’s production method. Because DNA, regardless of its sequence, is present in the cells of every living organism, including every plant and animal used for food by humans or animals, and has been safely consumed in food, FDA has presumed it to be generally recognized as safe (GRAS) (FDA, Statement of Policy: Foods Derived from New Plant Varieties 57 FR 22984, 1992). Therefore, the DNA that makes up the construct in AquAdvantage Salmon poses no more food consumption risk than DNA present in other foods. That is, the DNA present in food from AquAdvantage Salmon is no different from, and is as safe as, the DNA that we eat in other foods every day. In addition, transfer of DNA from the rDNA construct in AquAdvantage Salmon into the genome of gut bacteria would be no more or less likely to occur and would present no additional risks than the transfer of DNA from non-GE salmon into the genomes of gut bacteria.
Comment Summary: Two comments raised food safety concerns regarding issues sometimes associated with farm-raised salmon including the use of antibiotics in aquaculture and the accumulation of toxic substances in farm-raised salmon.
Response: FDA requires that all seafood that it regulates, including all farm-raised Atlantic salmon, whether non-GE or AquAdvantage Salmon, meet all existing, applicable food safety standards, including tolerance levels for residues of animal drugs and levels of toxic substances.
Comment Summary: Some of the food safety comments focused on the composition of food from AquAdvantage Salmon. A subset of these cited specific differences in composition between food from AquAdvantage Salmon and Atlantic salmon (farm-raised salmon or wild and farm-raised salmon), and one comment disagreed with a specific aspect of FDA’s compositional analysis. Others stated that food from AquAdvantage Salmon is nutritionally equivalent to non-GE Atlantic salmon.
Response: FDA performed a detailed compositional analysis of ABT’s GE salmon containing the AquAdvantage rDNA construct (ABT salmon), including AquAdvantage Salmon in order to determine whether consumers eating food from AquAdvantage Salmon would be receiving the nutrients they expect from farm-raised Atlantic salmon, and whether there were any differences from farm-raised Atlantic salmon that would pose any nutritional or health risks. Because there are differences between the composition of wild-caught and farm-raised Atlantic salmon due to the differences in their diets, the appropriate comparator for composition for ABT salmon is farm-raised rather than wild-caught Atlantic salmon.
Of the proximate, vitamin, mineral and amino acid analytes in the study examined for compositional analysis, only three analytes were present at levels in TX salmon (i.e., GE salmon) that were statistically significantly different from levels in non-GE control Atlantic salmon: vitamin B6 (when diploid and triploid salmon were considered together for GE and non-GE control Atlantic salmon), folic acid (when diploid salmon were compared in GE and non GE-control Atlantic salmon) and niacin (when triploid salmon were compared in GE and non-GE control Atlantic salmon). Based on all previous criteria including statistical analysis, FDA concluded that the levels of all proximate, vitamin, mineral and amino acid analytes in TX salmon except vitamin B6 are similar to levels in one or more appropriate groups of control salmon, and therefore the statistical differences are not biologically relevant.[FOI summary at Section IX C 2 a].
For the statistically significant increase in the level of vitamin B6 in diploid ABT salmon (the diploid broodstock that are not to be the commercialized food product), FDA performed an additional analysis. Following that analysis, which took the form of a “margin of exposure” assessment common in risk and safety assessment, in which we compared intake of vitamin B6 from diploid ABT salmon to the daily intake of Vitamin B6 that would be experienced at the most “conservative” (i.e., health protective) levels, we concluded that the difference was not meaningful in terms of safety or nutrition because even if the highest observed level of vitamin B6 shown in diploid ABT salmon were consumed, the assessment indicated that such level would still be well within the upper bound of the recommended daily intake for vitamin B6 [FOI Summary at Section IX C 2 a I (b) (iii)]. Furthermore, diploid ABT salmon are not intended to be the commercialized food product. Nevertheless, even if such fish were consumed, the observed level of B6 in such fish would not present a safety or nutritional concern.
In conclusion, no biologically relevant differences were observed in the general (e.g., proximates, including total protein and total fat) or detailed (e.g., specific amino acids, fatty acid ratios of fatty acids, including omega-3 and omega-6 fatty acids) composition of food from AquAdvantage Salmon and other non-GE farm-raised Atlantic salmon.
Comment Summary: Several comments addressed the hormone analysis performed on AquAdvantage Salmon. Of these, some criticized the hormone study in a scientific publication reviewed by FDA. Some comments noted elevated IGF1 levels in GE salmon compared with comparator salmon, with some questioning FDA’s analysis of IGF1, citing irregularities in FDA’s analysis. A few comments expressed concern regarding association of IGF1 and cancer. Another few comments addressed the quantity of hormone data and the interpretation of data below the limit of detection.
Response: In FDA’s evaluation of the safety of food from AquAdvantage Salmon, we considered all relevant data and information, including data from scientific publications, in a weight-of-evidence approach. See VMAC Briefing Packet at 2.
Insulin-like growth factor 1 (IGF1) is a necessary endogenous hormone in the body whose synthesis is stimulated by growth hormone. IGF1 plays a critical role in growth in children, and is also responsible for maintaining the normal growth and replacement of cells in adults. Humans produce significant levels of IGF1 themselves, and IGF1 is a normal component of food derived from animals.
Some have considered IGF1 a potential hazard for human consumption following increased growth hormone levels in food producing animals, and have claimed that IGF1 is linked to increased levels of cancers. FDA has previously conducted a thorough investigation of the potential associations, and has found the there is no demonstrable causative association between levels of IGF1 and increased human cancer risk, or the ability to transform non-cancer cells to cancer cells.
IGF1 levels in AquAdvantage Salmon (the fish that are intended for marketing as food in the U.S.) were so low that they were below the limit the assay could measure.
Two comments characterized differences in the number of Atlantic salmon with IGF1 levels above the limit of quantitation (LOQ) of the assay in Tables 15 and 16 of the Briefing Packet (Tables 26 and 27 of the FOI Summary) as “manipulation of” or “unexplained alterations to” IGF1 data by FDA. Table 15 (FOI Summary Table 26) is a summary of data above the LOQ for hormone analysis in salmon muscle/skin for all Atlantic salmon tested : (GE = ABT salmon; SControl = farm-raised non-GE Atlantic salmon raised under near identical conditions to ABT salmon; and Fcontrol = non-GE farm-raised Atlantic salmon from commercial Atlantic salmon farms). For IGF1, 6 ABT fish had values above the LOQ and 11 non-GE, farm-raised Atlantic salmon raised under near identical conditions had values above the LOQ. In the paragraph immediately preceding Table 16 (FOI Summary 27), we explained that there initially appeared to be an increase in IGF1 in the ABT salmon and further evaluation showed that this potential difference was most apparent in mature diploid non-GE farm-raised Atlantic salmon and mature diploid ABT salmon, a subset of ABT salmon that are not the subject of the application (AquAdvantage Salmon are by definition triploid). Data from these specific subsets were further evaluated in Table 16; FOI Summary Table 27. The “apparent irregularities” cited by submitters to the docket were the additional analysis that FDA performed on this particular subset, and not what the submitters characterized as “manipulation of” or “unexplained alterations to” data. FOI Summary at Section IX C.
IGF1 levels appeared to be elevated in mature diploid ABT salmon compared with mature diploid non-GE farm-raised Atlantic salmon (data presented in Table 16 (FOI Summary Table 27)), but the difference was not statistically significant. In order to determine whether the apparent elevation in IGF1 in these diploid broodstock fish was biologically relevant, even though these fish are the breeding animals and not the AquAdvantage Salmon that are the subject of this approval, we used the margin of exposure approach. Such an approach is a common risk assessment tool used by expert bodies such as the Food and Agriculture Organisation/World Health Organisation Joint Expert Committee on Food Additives. This analysis determined that the potential increase in IGF1 levels in the diploid broodstock fish was not meaningful in terms of safety even when using the most conservative, health-protective assumptions. In other words, the amount of additional IGF1 that could be consumed using the most health-protective assumptions is so small compared to the amount experienced as the result of daily exposure from an individual’s own level of IGF1 in human plasma as to be not biologically meaningful. FOI Summary at Section IX C 1 iii.
Du et al. 1992 presented data from GE Atlantic salmon containing the opAFP-GHc2 construct and provided insight into potential food safety concerns that were addressed in subsequent studies in FDA’s evaluation of the safety of food from AquAdvantage Salmon. They reported that mean plasma growth hormone concentrations did not differ statistically between the ABT salmon and either non-GE siblings or control samples.
For the sponsor’s hormone assessment, samples were analyzed from a total of 73 salmon, including 10, farm-raised Atlantic salmon control fish from a commercial farm different from the sponsor’s; 33 non-GE farm-raised Atlantic salmon that came from the sponsor’s farm; and 30 ABT salmon. All 73 salmon were tested for all listed hormones, but only the ones with measurable levels of the hormone in muscle/skin tissue are listed in the corresponding table in the Briefing Packet and FOI Summary. From the sponsor’s study, mean levels of estradiol, testosterone, 11-ketotestosterone, T3, and T4 were not different in the ABT salmon than levels in comparator non-GE salmon (Table 22). The apparent difference in IGF1 in mature diploid ABT salmon compared to sponsor control non-GE salmon was small, and FDA determined it to be of no biological relevance because the margin of exposure to this endogenous component of food would be well within levels of exposure from other dietary sources of IGF1, and poses no additional risk. See FOI Summary, Section IX C 1a ii.
All of the 73 samples tested for growth hormone fell below the limit of quantitation, including all non-GE farm-raised Atlantic salmon, and all ABT salmon. That these values were below the limit of quantitation indicates that they are too low to be measured accurately; concentrations of hormones below the limit of quantitation of the analytical method are not of public health concern. Our public health concerns would arise at the opposite extreme, that is, if the levels were very high, rather than very low.
In summary, mean levels of estradiol, testosterone, 17-ketotestosterone, T3, and T4 were not different in the ABT salmon compared with comparator non-GE salmon. The apparent difference in IGF1 in mature diploid ABT salmon compared to sponsor control non-GE salmon is relatively small. Even if the expression of IGF1 were present at the highest levels measured, and even if expected high consumers of salmon ate nothing but ABT salmon containing this likely upper bound level of IGF1, the margin of exposure to this endogenous component of food would be well within levels of exposure from other dietary sources of salmon, and poses no additional risk. We further note that the highest level of IGF1 that was used for this analysis was from the diploid ABT salmon that are not the subject of the approval.
Comment Summary: Several comments addressed concerns regarding the allergenicity of food from AquAdvantage Salmon. Some expressed concern about allergic reactions that could be caused by food from AquAdvantage Salmon. Others cited flaws in the studies performed, or requested additional studies or more research on the allergenic potential of foods from GE organisms in general, and the AquAdvantage Salmon in particular. Other comments expressed concerns about diploid ABT salmon in the triploid ABT salmon population (AquAdvantage Salmon). One requested that the Agency consult with an outside expert.
Response: Fish are one of the eight major allergenic foods or food groups.4 People who are allergic to salmon will likely avoid all salmon, including food from AquAdvantage Salmon. The results of the immunoassays that we reviewed show that that AquAdvantage Salmon are similar in allergenic potency to farm-raised Atlantic salmon, meaning that if someone is allergic to, farm-raised Atlantic salmon, she/he will likely also be allergic to AquAdvantage Salmon as well. FOI Summary at Section VIII C 2 c.
Some of the commenters cited limitations in aspects of the allergenicity studies. In our review, available in the Briefing Packet, and in the FOI Summary, we noted these limitations and interpreted the results accordingly. As part of FDA’s review of this and any other new animal drug application, we always carefully and deliberately evaluate all studies, taking into account any study limitations, and interpret those studies within the constraints imposed by their limitations. Further, it is important to note that in the evaluation of the overall allergenicity of food from AquAdvantage Salmon, we were not determining whether AquAdvantage Salmon are allergenic, but rather, whether they differ in allergenicity from non-GE, farm-raised Atlantic salmon.
In assessing the allergenicity of food from AquAdvantage Salmon, we used an approach that is consistent with that recommended Codex Alimentarius Commission in Guideline for the Conduct of Food Safety Assessment of Foods Derived from Recombinant-DNA Animals (Codex rDNA Animal Guideline) (CAC/GL 68-2003). Triploid ABT salmon pose no additional allergenic risk than comparator Atlantic salmon. Because AAS are an all female hemizygous subset of triploid ABT salmon (that include male triploid ABT salmon), the conclusions for the triploid ABT salmon also apply to AAS. FOI Summary at Section IX C 2 b iii.
With respect to the concern regarding the potential presence of diploid ABT salmon in the population of what should be triploid ABT AquAdvantage Salmon, we note that under the conditions established in the approved application, ABT must follow manufacturing specifications to ensure to the greatest extent possible that AquAdvantage Salmon are an all triploid population. Further, we identified no novel food safety concerns that would be encountered were diploid ABT salmon consumed, as people who are allergic to salmon, will likely avoid all salmon, regardless of whether it is genetically engineered. FOI Summary, Appendix A.
Regarding the request that FDA consult an outside expert, FDA independently consulted with an outside food allergy expert on assessing the allergenicity of food from GE organisms at the National Institutes of Health (NIH). The outcome of that discussion can be found at Appendix A of the FOI Summary, and supports the agency’s approach.
As for the request for additional research, FDA’s current safety assessment framework for foods from GE organisms is rigorous, science-based, protective of public health, and consistent with the current international standard (the Codex rDNA Animal Guideline cite above). As indicated in FDA’s consultation with an outside expert on food allergenicity (see FOI Summary, Appendix A) there is a great deal that is still unknown about food allergies in general. However, with respect to AAS, there is sufficient data and information for FDA to conclude that AAS are no more allergenic than their non-GE comparators.
viii. Long-Term Safety
Comment Summary: A few comments expressed concern regarding the long term safety of consumption of food from AquAdvantage Salmon.
Response: The Agency’s approach to assessing the safety of food from AquAdvantage Salmon included an assessment of animal health, identification and characterization of new or altered hazards, and an assessment of potential unintended effects that may impact the safety or nutritional value of the food. This approach is consistent with that recommended by the Codex Alimentarius Commission in Guideline for the Conduct of Food Safety Assessment of Foods from Recombinant-DNA Animals (CAC/GL 68-2008), which addresses factors that might relate to both short-and long-term consumption. Using this approach, FDA has determined that no hazards from consumption of food from AquAdvantage Salmon were identified that differed from those posed from consumption of farm-raised Atlantic salmon, or that required the evaluation of long term consumption of food from AquAdvantage Salmon. FOI Summary at Section VIII D.
i. General Comments on Phenotype and Phenotypic Assessment
Comment Summary: Several comments were critical of aspects of individual studies, datasets, or information used to evaluate the phenotype (physical characteristics) of AquAdvantage Salmon that the commenters believed weakened the study.
Response: In FDA’s evaluation of the phenotype of AquAdvantage Salmon, we considered all relevant data and information, including data from scientific publications, in a weight-of-evidence approach. This approach, described in our Briefing Packet, describes that certain levels of deference are given to data and observations depending on their direct relevance to the question at hand, as well as the quality of the study. We draw on a variety of types of data and information from a number of sources. These include the following, listed in rank order (from highest to lowest) of importance in the overall weight-of-evidence evaluation: (1) controlled studies conducted on the specific animals being considered for approval; (2) non-controlled studies on these same animals and historical records and data for these animals; and (3) studies reported in the scientific literature investigating these or similar animals. Each source, in turn, is given appropriate deference with respect to its relevance to the specific question under consideration.
What is important is that regardless of the source or order of deference given to a given dataset, all of the data and information are evaluated in the context of basic scientific principles and external validity. This means that studies, data, and information with flaws or weaknesses are given less weight than studies, data, or information without such flaws or weaknesses. Nonetheless, we consider all of the data and information available that is relevant to a particular question.
Comment Summary: One comment stated that the phenotypic characterization data (and the nutritional and food safety assessment data) all come from GE Salmon raised in the PEl facility, not at the facility in Panama, and that the different facility could affect the GE salmon's phenotype.
Response: Significant differences in husbandry conditions such as feed composition and feeding rate, biomass density, water temperature, and oxygen level can affect the growth and phenotype of any fish. It is possible that rearing at a different facility could affect a fish’s phenotype if the husbandry conditions differed significantly between the facilities; however, we do not believe this is the case for facilities located in PEI and Panama. Although the rearing conditions at these two facilities are not identical, they are not so different from each other or conditions commonly used in farm-raised Atlantic salmon aquaculture to elicit concern regarding the key parameters evaluated. Instead, feed composition is one factor that can have a large impact on growth and phenotype, and the feed that an animal eats can have the largest influence on the composition of the food from that animal. Commercially available standard fish feeds were used in both the PEI and Panama locations; these may vary as a function of the supplier and type of diet (e.g., in general, there has been a reduction in the amount of fish meal and fish protein, and an increasing amount of plant protein over time in most aquaculture commercial diets), but both are representative of feeds commonly used in salmon aquaculture. Thus, we do not believe that the location of the facility has a significant impact on our assessments.
This is illustrated by the additional data submitted by ABT since the Briefing Packet was released, which include data and information on growth performance, mortality, culling/planned reductions to optimize stocking density, occurrence of irregularities, and rank scores on irregularities for the 2010, 2011, and 2012 year classes (also referred to as the 2011-2013, 2012, and 2013-2015 production cycles) from both the PEI and Panama facilities. All parameters were within norms for other farm-raised Atlantic salmon (see FOI Summary, Phenotypic Characterization).
Comment Summary: Some comments opined that AquAdvantage Salmon were significantly different from farm-raised non-GE Atlantic salmon, and because they are triploid and all female, differ from such salmon by more than just one added gene.
Response: Although AquAdvantage Salmon contain the opAFP-GHc2 rDNA construct, and therefore differ from farm-raised non-GE Atlantic salmon at that locus (a locus is a specific site in the genome of an organism) and thereby grow more quickly than their non-GE counterparts, they are not otherwise significantly different from farm-raised Atlantic salmon. The comparators that were used throughout the studies submitted by ABT are all St. John’s strain Atlantic salmon, which is the strain of the progenitors of AquAdvantage Salmon. In many cases, the comparators are non-GE half-siblings of the AquAdvantage Salmon, that is, they were bred from the same “father” that provided milt (sperm) to produce the hemizygous AquAdvantage Salmon (milt from wild-type male fish are used to fertilize eggs from homozygous5 female GE salmon to yield the hemizygous6 animals that will undergo triploidization).
Farm-raised fish often are treated with the same or very similar methods used by ABT to produce either triploid or single sex populations of fish (or both). FDA does not regulate this process of reproduction used in aquaculture. Fish that are triploid or single sex are currently found in the marketplace.
Finally, FDA assessed the extent to which AquAdvantage Salmon could be considered Atlantic salmon by employing the methodology used by the agency in determining fish identity in FDA’s Regulatory Fish Encyclopedia (RFE). The RFE is a compilation of data in several formats that assists with the accurate identification of fish species, developed by FDA's RFE Team composed of scientists at the Seafood Products Research Center (SPRC, Seattle District), and the Center for Food Safety and Applied Nutrition (CFSAN) to help federal, state, and local officials and purchasers of seafood identify species substitution and economic deception in the marketplace.
In addition to high-resolution photographs for each of a number of commercially relevant fish species for sale in the U.S. market, the RFE includes chemical taxonomic information consisting of species-characteristic biochemical patterns which may be compared quantitatively to patterns obtained by an appropriate laboratory analysis of the fish species specimen in question. Currently the RFE includes tissue protein patterns determined by Isoelectric Focusing (IEF) electrophoresis and cytochrome c oxidase 1 (CO1) mitochondrial DNA sequence patterns determined by DNA Barcoding.
FDA/CVM conducted IEF analysis of AquAdvantage Salmon and non-genetically engineered farm-raised Atlantic salmon. The finding of identical IEF banding patterns confirmed that ABT’s salmon, and thus, AquAdvantage Salmon met the FDA’s standard for Atlantic salmon under the criteria developed by FDA for IEF for species identification.
Using DNA barcoding, FDA/CFSAN compared CO1 mitochondrial gene sequences from AquAdvantage Salmon with FDA reference standard fish sequences. The AquAdvantage Salmon matched both FDA reference standards for Atlantic salmon under the criteria developed by FDA for DNA barcoding for species identification.
AquAdvantage Salmon thus meets FDA’s standard for Atlantic salmon under both criteria established for the RFE. See FOI Summary, Appendix B.
ii. Study Design
Comment Summary: Some comments questioned the comparators chosen for the controlled study assessing the health of AquAdvantage Salmon, i.e., the study assessing the safety of the rDNA construct to animal.
Response: Selection of control animals or “comparators” in controlled studies is a very important part of any study design and interpretation of the results of the studies. In general, when conducting safety studies such as these in livestock, investigators attempt to match comparators and test animals as closely as possible so variations due to their age, size, gender, etc. do not “confound” (make difficult to interpret) the results because of the multiple variables at play. The goal of the controlled studies used to evaluate the phenotype of AquAdvantage Salmon was to determine if the rDNA construct was harmful to the health of the animals. The difficulty in finding the appropriate comparator for the AquAdvantage Salmon was that if the AquAdvantage Salmon were matched with farm-raised Atlantic salmon that were similar in size, the ages of the fish would be very different; if the fish were matched on age, their sizes and possibly developmental stages would be different. Because of these complications, the sponsor consulted with FDA on how to prepare the protocol for the study. After deliberation, FDA recommended that the sponsor use multiple comparators: age matched non-GE fish that, as expected, were smaller than the AquAdvantage Salmon at the end of the study, and size-matched non-GE fish that were older than the AquAdvantage Salmon, but were the same size as the AquAdvantage Salmon at the end of the study because they had more time to grow.
Although there were other options available for use as comparators, they were less desirable. For example, using other lines of farm-raised Atlantic salmon would introduce different genetic background. Use of farm-raised Atlantic salmon grown under different conditions (e.g., ocean net pens) would introduce additional variables (water quality, composition, temperature, parasite load, etc.) not related to the rDNA construct. These comparators were also excluded from consideration, as net pen growth is not being considered for this application. Thus, FDA concluded that the comparator fish used in the animal safety study were the appropriate, and probably the best comparators that could be used for the question being asked (i.e., did the rDNA construct affect the health of the target animal).
Comment Summary: Some comments stated that FDA only evaluated fish “that would enter commerce” although other fish also contained the construct. Related to this point, some comments opined that ABT’s culling practices skewed the findings on animal safety by removing any fish with abnormalities prior to conducting an analysis, and that the culling practices at the hatchery influenced the outcome of the study. Some comments also alleged that “the most severely deformed and unhealthy animals were excluded from the study.”
Response: ABT’s application is for approval of the inserted form of the rDNA construct (the EO-1α locus) in triploid Atlantic salmon (the production fish). Before approving such an application, FDA must find the rDNA construct to be safe and effective for the animal, among other things. Although FDA conducted the most stringent analysis on the triploid fish intended to serve as food, all of the data that the sponsor submitted, including data from the diploid breeding fish, were evaluated. Using our weight-of-evidence paradigm, the data from the diploid breeding fish helped to inform the agency’s decision-making process, although they received lower deference because they were derived from diploid breeding fish and not AquAdvantage Salmon. These analyses are also found in the Briefing Packet and the FOI Summary.
During the initial phase of CVM’s review, we stated that, as with all data sets, there are some uncertainties. See VMAC Briefing Packet at 21.7 At the time FDA performed its preliminary review, which was released in the briefing packet for the VMAC meeting, the primary area of uncertainty was in determining the actual rate of adverse outcomes in grow-out facilities, as the process of selecting animals for the initial sponsor study, which has been referred to as “culling,” may have influenced the apparent rate of abnormalities observed. Because of concerns that the culling procedures for the initial study may not have reflected typical aquaculture procedures and may have obscured adverse outcomes, the agency requested and received from the sponsor additional information regarding culling practices, the health of the ABT fish populations at the grow-out facilities, and the potential role that culling could have had in masking adverse outcomes. These data and information are found in the FOI Summary, Section VII B 2, and Tables 8-13 and include information on morbidity and mortality from more than 150,000 ABT salmon and approximately 9,000 non-GE Atlantic salmon from both the PEI and Panama facilities. These new data did not reveal any new abnormalities or altered rates of abnormalities beyond those identified in the initial study, and did not indicate any bias in the initial study’s estimation of rates of morphologic abnormalities, mortality, or morbidity.
Rates of morphological abnormalities, mortality and morbidity in these data establish the expected rate of such events for use in determining adverse events that must be reported as part of the post-approval reporting requirements (see Approval Letter, Appendix 1). FDA typically continues to monitor products that it regulates in order to detect any unanticipated effects that develop after an approval is issued.
Comment Summary: A few comments asserted that the controlled animal health studies were too small or lacked statistical power.
Response: As described above, the controlled animal health studies are one part of the weight-of-evidence evaluation, and were granted greatest deference in our analysis. As designed, these studies were intended to examine a large number of variables in minute detail for each fish. The intent in doing so was to determine if subtle, but reproducible, abnormalities would occur and pose risks to animal health or food safety. Because of the relatively homogenous genetic background of these fish, abnormalities would be expected to be detected in a small number of fish at a higher frequency than in outbred populations, which are often the subject of safety studies for traditional drugs. In addition, multiple, different approaches were used to identify potential anomalies (see the discussion above of the weight-of-evidence evaluation). No abnormalities affecting animal safety or food safety were identified. The studies fulfilled the function they were designed to accomplish, especially within the context of the weight-of-evidence approach used. Further, the Agency reviewed additional data and information from the sponsor on over 150, 000 ABT Atlantic salmon, including AAS, and approximately 9,000 farm-raised comparators (see immediately previous response). The data demonstrate the initial culling did not mask adverse health outcomes for the parameters evaluated. See FOI Summary at Section VI Tables 8-13.
Comment Summary: A few comments asserted that because the studies were conducted by ABT (the sponsor) those studies are biased.
Response: Sponsors of both human and animal drug applications for FDA approval commonly conduct the safety and effectiveness studies that support those applications. Under section 512(b)(1) of the FD&C Act, the new animal drug sponsor is required to submit to FDA full reports of tests showing whether or not the new animal drug is safe and effective for use, among other things. In addition, FDA’s implementing regulations provide that applicants must submit all information pertinent to an evaluation of the safety and effectiveness of the new animal drug, including reports in the scientific literature, both favorable and unfavorable. 21 CFR 514.1(b)(8)(iv). FDA’s implementing regulations further provide that applicants must “explain any omission of reports from any investigator to whom the investigational new animal drug has been made available,” and “[t]he unexplained omission of any reports of investigations . . . or the unexplained omission of any pertinent reports of investigations or clinical experience received or otherwise obtained by the applicant . . .that would bias an evaluation of the safety of the new animal drug or its effectiveness in use, constitutes grounds for the refusal or withdrawal of the approval of an application.” 21 CFR 514.1(b)(8)(vii). The Agency reviews the data and information submitted to decide if the drug is safe and effective and if the approval requirements are otherwise met. Although ABT conducted the studies, as most animal and human drug sponsors do, FDA Center for Veterinary Medicine (CVM) experts concurred with study designs prior to the initiation of several studies, reviewed the data and information generated in those studies, and on multiple occasions inspected or performed extensive site visits to the facilities used to house and grow the fish. We found no evidence of bias.
To understand the drug approval process better, interested parties may find helpful “From an Idea to the Marketplace: The Journey of an Animal Drug through the Approval Process,” which is posted on the CVM website.
Comment Summary: Some comments stated that there were deviations from study protocols.
Response: Many studies have minor deviations from the established protocols; these are common and generally have no impact on study outcomes. Under FDA’s Good Laboratory Practice (GLP) regulations for non-clinical studies (21 CFR Part 58) and good clinical practices guidance for clinical studies (Guidance for Industry 85), all changes in or revisions of an approved protocol and the reasons for the change are documented. 21 CFR 58.120(b), GFI 85 at 9. Where ABT deviated from study protocols, they explained these deviations to FDA’s satisfaction.
Comment Summary: One comment asserted that cloned lines of GE animals should be treated differently, and should have separate INAD files.
Response: None of the AquAdvantage Salmon are clones. They are derived from a single micro-injected fertilized egg that has given rise to a lineage that has been bred for several generations.
iii. Health of the Fish
Comment Summary: A few comments asserted that AquAdvantage Salmon are unhealthy and that abnormalities were found in them, either as general statements or citing particular concerns.
Response: AquAdvantage Salmon are as healthy as farm-raised non-GE Atlantic salmon grown under similar fresh water aquaculture conditions. All populations of animals (or any other organisms) will exhibit a range of health conditions, regardless of whether they are genetically engineered or not. These conditions can range from very minor deviations from “normality” to some conditions that are lethal to the fish. The summary reports released for the VMAC included descriptions (in some cases quite technical) of ABT’s observations of the health status and phenotypes of their GE salmon. FDA performed a very careful and rigorous evaluation of every observation (also summarized in the Briefing Packet). As expected, these detailed analyses discovered some abnormalities, and grouped them into severity classes. After much further evaluation, FDA scientists concluded that the abnormalities do not differ appreciably from those in non-GE farm-raised Atlantic salmon raised under similar conditions, and have no impact on the health of the AquAdvantage Salmon. The agency therefore concluded that AquAdvantage Salmon are as healthy as non-GE farm-raised Atlantic salmon grown under similar fresh water aquaculture conditions. See FOI Summary at Sections VI B 2 a, C.
It is important to note that the abnormalities that were observed are common in farm-raised Atlantic salmon raised under typical aquaculture conditions. As explained above in the “Study Design” section, the sponsor used farm-raised Atlantic salmon raised under similar aquaculture conditions as comparators for AquAdvantage Salmon. Comparison of AquAdvantage Salmon with wild Atlantic salmon would not be appropriate, as the conditions under which the fish live would introduce too many variables to make the study scientifically valid. Further, “abnormal” animals, including wild Atlantic salmon, do not usually survive for very long in the wild. They are less fit and are either eaten by other fish or die as the result of not being able to feed properly. Because access to food and predation are not issues for fish raised in confined tanks, it is entirely possible that an apparent higher rate of abnormalities may be observed in both GE and non-GE fish because the more supportive environment for these fish means that more fish could survive and grow to market weight with abnormalities than could such fish in the wild.
Comment Summary: A few comments included assertions that because some GE animals are unhealthy, AquAdvantage Salmon also must be unhealthy.
Response: These sorts of generalizations are not scientifically valid. In reviewing the data relating to AquAdvantage Salmon, FDA had to determine, among other things, if the rDNA construct in the salmon is safe for that particular line of fish, i.e., the AquAdvantage Salmon. FDA is making a decision only for this one line of fish raised under the particular set of conditions specified in the approved application. The agency is not extending that decision to any other conditions, and not to any other fish, and certainly not making overall decisions about all GE animals. As we state in GFI 187, FDA makes determinations regarding the evaluation of new animal drug applications related to GE animals on a case-by-case basis.
In this instance, we used a weight-of-evidence evaluation methodology to make our decisions. As described beginning on page 2 of the Briefing Packet and in the FOI Summary, the weight-of-evidence evaluation allows us to make use of as much information as we can gather, including information from other lines of GE fish. It is very important to note, however, that the weight-of-evidence approach assigns different degrees of “deference” to studies. That means that we pay closest attention to (i.e., give most deference to) the highest quality, most directly relevant studies and observations -- data and information specific to AquAdvantage Salmon’s particular lineage of GE animals containing its specific rDNA construct at its specific genomic location. But we still considered other more general information during the review process. For example, FDA’s evaluation of AquAdvantage Salmon considered numerous studies and reports from the scientific literature that evaluate GE animals. Nevertheless, when considered as a whole, the evidence we considered under the weight-of-evidence evaluation shows that AquAdvantage Salmon are healthy.
Comment Summary: A few comments claimed that AquAdvantage Salmon are more susceptible to disease and asserted that AquAdvantage Salmon will require use of “more antibiotics” than non-GE salmon. These comments contended that Atlantic salmon farming in general requires extensive use of antibiotics and further opined that because AquAdvantage Salmon are genetically engineered, they will be sicker and require even more use of antibiotics.
Response: No scientific data support the assertions that AquAdvantage Salmon are any more susceptible to disease or require more antibiotics than non-GE farm-raised Atlantic salmon under the aquaculture conditions being used. As described previously in this document, AquAdvantage Salmon are healthy and are no more susceptible to disease than non-GE farm-raised Atlantic salmon. ABT has instituted a rigorous testing program for fish diseases and the program has not detected disease since 2009. Absent the presence of bacterial disease, ABT should not have had a need to use antibiotics at its breeding facility.
During the third quarter of 2009, a disease outbreak later determined to be infectious salmon anemia (ISA) occurred at the PEI facility. Prior to this, the PEI facility had been considered “disease free” for many years based on periodic inspections and testing by Canadian authorities. (For a more detailed discussion of this disease outbreak, see EA Section 188.8.131.52.) Once the presence of the infectious salmon anemia virus (ISAV) was confirmed, ABT notified Canada’s Department of Fisheries and Oceans and the Canadian Food Safety Inspection Service. Although the source of the infection was never definitively identified, such outbreaks are not rare in aquaculture operations, including non-GE fish, in Canada and around the world.
In response to such a disease outbreak, the only way to restore a healthy fish population is to extirpate the infection by culling all diseased and co-cultured fish. ABT did so by implementing standard Atlantic salmon mitigation strategies appropriate for this disease in its facility and implementing an ISAV detection and monitoring program. This means that all fish displaying any characteristic of poor health or high viral load were culled from the facility. As part of its risk mitigation strategy, ABT constructed quarantine areas to house and isolate the early rearing and grow out areas, and installed ultraviolet lights and other equipment to disinfect both incoming well and recirculating water.
All year classes of fish produced since the 2009 ISA outbreak have tested negative for ISAV when assayed using the most sensitive assays. Since November 2009, there has been no detectable evidence of ISA disease in the PEI facility. As of late November 2014, no notifiable diseases or disease agents for finfish per Canadian or international (World Organisation for Animal Health (OIE)) requirements have been detected in fish or eggs from either the early rearing or grow out areas of the PEI facility since the ISA outbreak that occurred in 2009. Negative results have been found in all subsequent inspections of each area. For further information on the ISA outbreak and present disease status of the PEI facility, again, refer to Section 5.4.2 of the EA. From this history at the ABT facilities, as well as other studies conducted by ABT (see Briefing Packet and FOI Summary), AquAdvantage Salmon do not appear to be any more sensitive to pathogens than non-GE farm-raised Atlantic salmon, also housed at the ABT facilities.
iv. Data Considered and Disclosed
Comment Summary: A few comments expressed concerns about the variability in the breadth and depth of the historical data.
Response: In its submissions, ABT included historical data and information from its hatchery records, which capture the hatchery practices as the sponsor was developing and propagating successive generations of AquAdvantage Salmon and refining practices in its facility. Consistent with our weight-of-evidence approach, we evaluated all the material in the hatchery record submissions. This information included larger numbers of fish than would be practical to include in a controlled prospective study, and supports the conclusion that, overall, the health of AquAdvantage Salmon is comparable to comparators at the facility and comparable to other farm-raised non-GE Atlantic salmon. Subsequent to the release of the Briefing Packet, ABT submitted additional data that provided the Agency a more complete overview of ABT’s hatchery practices and the fish that result from those conditions of use (see FOI Summary, Phenotypic Characterization).
Comment Summary: Some comments asked that FDA require submission of samples of the fish throughout their lifespan.
Response: FDA evaluated AquAdvantage Salmon over seven generations, and many samples came from different generations. No marked differences have been observed over these generations. As established in the approved application, the sponsor is required to institute a rigorous durability plan that requires the reporting of data to FDA on the health, genotype, and phenotype of the AquAdvantage Salmon over the entire course of its commercial lifespan. In addition, there are post-market requirements for reporting of adverse events at different life stages of the animals. We therefore believe that we have adequately addressed the issue of ongoing evaluation, and do not believe it is necessary to require submission of additional samples from throughout the salmon’s lifespan.
Comment Summary: A few comments asked that the Agency disclose varying amounts of additional data submitted by the sponsor, claiming that the data release was incomplete.
Response: As explained above in section II(B)(ii), although generally FDA may not disclose to the public confidential commercial information contained in an NADA, in the case of AquAdvantage Salmon, ABT agreed to allow FDA to release a summary of all submitted data prior to the September 20, 2010 VMAC meeting. FDA has thus already made more extensive data available than it ordinarily could release for an NADA approval.
C. Molecular Biology
Comment Summary: Two comments expressed concern that the rDNA construct insertion in AquAdvantage Salmon is inadequately characterized.
Response: FDA considered multiple factors in its weight-of-evidence evaluation of the rDNA construct insertion site in AquAdvantage Salmon. These include careful characterization of the rDNA construct prior to insertion into the salmon. Following insertion, the rDNA construct has been characterized repeatedly in the genomes of every generation of the salmon produced from the founder of the AquAdvantage Salmon lineage. The data show that the inserted form of the rDNA construct (the EO-1α locus) is stably inherited over at least seven generations. The rDNA construct insertion perfectly interrupts a known sequence (termed a 35 base repeat) at a single position without adding or removing any nucleotides from the repeat. Studies also show that the EO-1α locus is the only insertion site of the rDNA construct in AquAdvantage Salmon.
The rapid growth of the AquAdvantage Salmon is strongly correlated with the presence of the EO-1α locus and follows a standard (Mendelian) inheritance pattern. Other than their intended rapid growth, AquAdvantage Salmon are phenotypically, or physically, quite similar to non-GE comparators. This supports the conclusion that no important genomic sequences were altered or deleted when the rDNA construct was inserted at the EO-1α locus. All of these factors taken together demonstrate that the rDNA construct does not interfere with an important region of the chromosome and does confer the rapid growth phenotype (physical characteristic), as can be seen from the weight of evidence assessment conducted in the phenotypic characterization and the adequate and well-controlled study described in the claim validation portions of the FOI Summary.
Comment Summary: Two comments expressed concern that the sponsor has failed to prove that the rDNA construct that has been characterized is the sequence responsible for the growth phenotype of AquAdvantage Salmon.
Response: There is abundant evidence that the rDNA construct at the EO-1α locus (the insertion site) in AquAdvantage Salmon is responsible for the rapid growth phenotype observed for this lineage of GE Atlantic salmon (see Briefing Packet and FOI Summary). The rDNA construct at the EO-1α locus contains a growth hormone gene of Chinook salmon, driven by a constitutive promoter (meaning that the promoter is not only active seasonally, but rather, constantly directs the transcription of the growth hormone gene) and is present in a stable form in the fast-growing fish for at least seven generations with the associated fast-growing phenotype. The rDNA construct at the EO-1α locus have been detected by multiple techniques in AquAdvantage Salmon, and no other rDNA construct insertions (or fragments of the rDNA construct) have been detected in the developed AquAdvantage Salmon lineage.
D. Claim Validation
Comment Summary: Some comments stated that the data and information demonstrating that AquAdvantage Salmon reach their growth endpoint (i.e., at least 100 g within 2,700 0C-days) were flawed because of the choice of comparator. Others opined that the agency should not approve AquAdvantage Salmon because it would be possible to breed Atlantic salmon to grow as quickly without the use of genetic engineering.
Response: The comparators used in the study were derived from the same crosses (breeding events) that produced the test subjects, and thus provide the same genetic background, except for the introduced gene.
With respect to whether the agency should approve AquAdvantage Salmon because it might be possible to generate similar growth rates by conventional breeding, we note that FDA regulates GE animals under the new animal drug provisions of the FD&C Act, which generally require that a sponsor must demonstrate the safety and effectiveness of a new animal drug for its intended use in order for an NADA to be approved (21 U.S.C. § 360b). Further, under these provisions, FDA must approve an NADA unless there is a lack of evidence of safety or effectiveness, inadequate manufacturing methods, or other applicable requirements have not been satisfied, in which case the agency must deny approval (21 U.S.C. §§ 360b(c) and (d)). The Agency cannot deny an approval solely because the intended effect of the new animal drug under review (in this case, the rDNA construct) can be produced by other means (e.g., conventional selective breeding, administration of exogenous growth hormone) (see 21 U.S.C. § 360b(d)).
For this application, the evidence clearly demonstrated that compared to diploid near-siblings (the comparator farm-raised Atlantic salmon were derived from the same parental line as the AquAdvantage Salmon), a greater proportion of AquAdvantage Salmon grow to at least 100 g within 2,700 degree-days of first feed than their closely related, diploid near-siblings (the comparator fish) under normal commercial aquaculture conditions. The issue is not, therefore, whether any other Atlantic salmon, bred in any other way, or raised under any other conditions can reach this milestone at the same time. An analogous situation could be to consider a conventional animal drug that is intended to impart carcass leanness. The appropriate comparison in such an effectiveness trial would be to a population of commercial animals that were not treated with the drug, but otherwise raised under identical conditions, and not to a population of commercial animals that had been selected for carcass leanness, and raised under a different feeding regimen.
Comment Summary: Some comments indicated that selective breeding of non-genetically engineered Atlantic salmon may reach or exceed the growth rate of AquAdvantage Salmon, and thus the agency should not approve the application.
Response: As explained in the previous response, FDA regulates GE animals under the new animal drug provisions of the FD&C Act. Under these provisions, FDA must approve an NADA unless there are specific statutory grounds to deny approval, in which case FDA must refuse to approve the NADA. 21 U.S.C. 360b(c), 360b(d). Among the grounds to deny approval are a lack of evidence of safety or effectiveness, or inadequate manufacturing methods. FDA may not deny approval solely because other means can cause the same outcomes as the regulated article under review. Moreover, even if there were another approved new animal drug intended to cause salmon to grow faster, FDA could approve multiple drugs for that same indication.
E. Approval Process
i. Adequacy and Appropriateness of Using the NADA Provisions to Regulate GE Animals
Comment Summary: A number of comments expressed support for the NADA regulatory process, calling it the most effective way to demonstrate the safety and effectiveness of the rDNA construct in GE animals. Some comments disagreed, stating that the NADA provisions of the FD&C Act were not designed to cover GE animals or an rDNA construct in GE animals.
Response: The definition of a drug in section 201(g) of the FD&C Act includes “articles intended for use in the diagnosis, cure, mitigation, treatment, or prevention of disease in man or other animals” and “articles (other than food) intended to affect the structure or any function of the body of man or other animals.” Because the rDNA construct in a GE animal, such as AquAdvantage Salmon, is intended to affect the structure or function of the body of the GE animal, it meets this definition. FDA therefore regulates rDNA constructs in GE animals under the new animal drug provisions of the FD&C Act. Although these provisions may not have been written specifically with GE animals in mind, they require a thorough evaluation of the safety of the regulated article with respect to the animal and, where applicable, to humans who might eat food derived from the animal, as well as a determination of effectiveness. These provisions are therefore well-suited to assuring the safety and effectiveness of rDNA constructs in GE animals.
Comment Summary: A few comments asserted that FDA does not have expertise relevant to fish and aquaculture.
Response: CVM has many years of experience regulating drugs used in aquaculture including review of data relating to food safety and environmental impacts of aquaculture drugs. CVM is one of the founding members of the U.S. Government’s interagency Joint Subcommittee on Aquaculture (JSA), formed in 1990. CVM’s Aquaculture Team in the Division of Therapeutic Drugs for Food Animals, and another group of experts devoted to aquaculture research at the Center’s Office of Research have been engaged in product review and regulatory research related to aquaculture for decades. Additional expertise, particularly with respect to salmonids and other fish, is found in the CVM Environmental Safety Team, which is responsible for NEPA-related assessments at the Center. In addition, the Agency conferred with experts from the National Marine Fisheries Service (NMFS), which is part of the National Oceanic and Atmospheric Administration, and included a NMFS expert as part of the team of Agency subject matter experts that conducted an extensive site visit of the ABT Panama grow-out facility in 2009.
Comment Summary: A few comments stated that the NADA process does not take into consideration animal health and welfare.
Response: Under the FD&C Act and FDA’s implementing regulations, in order to approve a new animal drug, FDA must find the drug safe and effective for the animal (21 U.S.C. § 360b(d)). Therefore, as part of the NADA approval process, FDA will not approve an NADA unless it meets the animal safety standards set forth in the FD&C Act and FDA’s implementing regulations. Under this review, we take into account physical health, and to the extent that it can be measured in a species, behavioral health. To the extent that animal welfare encompasses animal health the NADA approval process does take it into account.
Comment Summary: Some comments argued that FDA should take public opposition to AquAdvantage Salmon into account in making its decision on the NADA related to AquAdvantage Salmon.
Response: As explained above, FDA regulates GE animals under the new animal drug provisions of the FD&C Act. Under these provisions, FDA must approve an NADA unless there is a lack of evidence of safety or effectiveness, inadequate manufacturing methods, or other applicable requirements have not been satisfied, in which case FDA must refuse to approve the NADA (21 U.S.C. §§ 360b(c) and(d)). The statutory grounds to deny approval of an NADA do not include public opposition to the approval. Accordingly, FDA may not deny approval of the NADA related to AquAdvantage Salmon based upon public opposition to approval. FDA has carefully considered the public comments that it has received, and has considered whether such comments raise any grounds to deny approval under the FD&C Act. With respect to comments that contend that consumers will not buy food from AquAdvantage Salmon, market considerations also are not a permissible basis for FDA to deny an NADA approval.
Comment Summary: Some comments suggested that non-governmental third-party organizations such as the American Fisheries Society should review the NADA.
Response: Congress granted FDA authority to review and approve NADAs under the FD&C Act. 21 U.S.C. 360b. While FDA may not delegate this authority to a third-party organization, FDA may seek input from the public and from outside experts subject to certain limitations. These limitations include restrictions related to the Federal Advisory Committee Act, which governs how an agency may obtain advice or recommendations from a group of outside experts, and restrictions on the release of confidential information. While adhering to these restrictions and complying with applicable requirements, FDA did seek comments both from the public and from an advisory committee. Any organization was welcome to review the NADA information released in advance of the VMAC meeting and to provide input through the related public comment process.
Comment Summary: A comment asked that FDA regulate each animal line derived from a separate transformation event as a new animal drug because the insertion site for the rDNA construct varies from one rDNA insertion to another.
Response: FDA agrees with this comment and, as described in Guidance For Industry 187: Regulation of Genetically Engineered Animals Containing Heritable Recombinant DNA Constructs, “[b]ecause the site at which an rDNA construct is located can affect both the health of the animal and the level and control of expression of the construct (i.e., its effectiveness), in general, each animal lineage derived from a separate transformation event (or series of transformation events) is considered to contain a separate new animal drug subject to a separate new animal drug approval.” As discussed in the Guidance for Industry 187, each new animal drug approval covers all animals containing the same rDNA construct (the regulated article or new animal drug) derived from the same transformation event. All of the fish evaluated in our review of AquAdvantage Salmon were derived from a single transformation event as described in detail in the FOI Summary.
Comment Summary: A comment asserted that the products of GE animals should not be approved for use in food unless they are also regulated under the food additive provisions of the FD&C Act.
Response: Under the definition of a “food additive” in the FD&C Act, a food additive does not include a new animal drug. 21 U.S.C. 201(s). Therefore, a substance cannot be regulated as both a new animal drug and a food additive. Further, in order to approve a new animal drug for use in food animals, FDA must find that food or feed derived from the animal is safe for humans or animals consuming it. 21 U.S.C. 360b(d); see also 21 CFR 514.1(b)(8). In determining whether a new animal drug is safe for humans to consume, FDA must consider a number of factors, including the probable consumption of the drug and of any substance formed in or on the food because of the use of the drug, the cumulative effect on man or animal of the drug, and other safety factors. These factors are the same factors that the agency is required to consider when determining whether a proposed food additive is safe. See 21 U.S.C. §§ 348(c)(5) and 360b(d)(2). Further, the same standard that FDA uses to determine whether a food additive is safe for consumption applies when the Agency determines the food safety of a new animal drug intended for use in food-producing animals. Specifically, FDA determines whether there is a reasonable certainty in the minds of competent scientists that the substance is not harmful under the intended conditions of use (i.e., there is a reasonable certainty of no harm when the drug is used as intended). See 21 CFR 570.3(i) and 170.3(i). Accordingly, there is no substantive difference in the food safety standard that FDA would use for a food additive approval.
ii. Transparency of the Regulatory Process
Comment Summary: The largest portion of the comments focused on the degree of transparency during the approval process, claiming that the FDA disclosed only a fraction of the total data that the sponsor submitted. Some of the comments criticized the existing NADA process and the Agency as a whole for not requiring the release of the complete sets of raw data and studies submitted as part of the NADA.
Response: Confidential commercial information is protected by law. FDA, therefore, may not release such information when included in an application unless the sponsor has disclosed the information or has authorized its disclosure. In anticipation of the VMAC meeting, with the sponsor’s consent, the Agency released to the public an unprecedented amount of information. To increase transparency, the Agency has made available to the public the entire summary of all data submitted by the sponsor (the Briefing Packet). In addition, the sponsor and outside collaborators have published a substantial amount of these data in peer-reviewed scientific journals. As part of the FOI Summary, the agency has also made available a summary of the data on which it has made its approval decision.
Comment Summary: A number of comments stated that the NADA process does not allow for significant public participation, noting that it does not require public input prior to application approval or the convening of scientific advisory committees.
Response: Considering the public’s high level of interest in GE animal technology, the Agency has taken unprecedented steps to ensure the transparency of the review process for the ABT NADA. FDA held a VMAC meeting on the ABT application on September 19-20, 2010; we accepted public comment on AquAdvantage Salmon at the VMAC meeting and to a public docket; we made available for public comment the draft EA and proposed FONSI; and we are now responding to the public comments we have received. See Response to Comments on the EA. None of these steps is required for new animal drug approvals. Nonetheless, FDA has committed to providing heightened transparency for this GE animal-related application and, as demonstrated by our actions on this GE salmon-related application, we are carrying through on that commitment.
iii. VMAC Organization
Comment Summary: A number of the comments addressed overall organization of the VMAC meeting held on September 19-20, 2010. Some of these comments noted the limited time they were given to review the released data and materials before the VMAC meeting. Other comments stated that the charge to the VMAC was made public on too short notice, which precluded meaningful public participation. In addition, several comments noted the absence of a public comment period and suggested that a 90 day comment period would be adequate for full public participation.
Response: The VMAC meeting was convened and organized by the Agency in accordance with applicable requirements under the Federal Advisory Committee Act (FACA), Federal Advisory Committee Management regulations (41 CFR Parts 101-6 and 102-3), FDA’s regulation for Advisory Committee meetings (21 CFR Part 14), and Agency procedures for Advisory Committees Meetings outlined in Guidance for Industry Advisory Committee Meetings – Preparation and Public Availability of Information Given to Advisory Committee Members. As required by 21 CFR 14.20, FDA published in the Federal Register a notice of the VMAC meeting, including the nature of the subjects to be discussed at the meeting, at least 15 days in advance of the meeting. The VMAC meeting was held on September 19-20, 2010, and FDA published the notice of the meeting in the Federal Register and on its website on August 26, 2010 (75 FR 52605). The notice also included website addresses to two background documents and additional information regarding CVM’s regulatory oversight of GE animals. Further, the notice specified that the Agency intended to make meeting materials available to the public on the Agency’s website approximately 16 days before the meeting. The Agency also posted the notice on its website. As required by FACA, materials made available to the VMAC were also made available to the public, with the exception of materials that were exempted from public disclosure under the Freedom of Information Act (FOIA), including confidential commercial information. Although FDA’s Guidance on Advisory Committee Meetings provides that FDA intends to post a publicly available version of all briefing materials for an advisory committee meeting at least two days prior to the meeting, FDA made data and information available to the public for the VMAC meeting on AquAdvantage Salmon 24 days in advance. We believe that this additional time increased transparency and public participation and was sufficient for the public to make the VMAC and the Agency aware of any additional information that should be taken into consideration prior to acting on the application. Additionally, on December 26, 2012, FDA published a draft EA and preliminary FONSI with a 60 day public comment period and then extended the comment period to 120 days ending April 26, 2013.
Comment Summary: One comment stated that the composition of the VMAC was not adequate and could not properly address major issues associated with AquAdvantage Salmon. Others opined that members of the VMAC were biased either in favor of, or against the subject of the application or that they had conflicts of interest.
Response: The experts on the VMAC at the time of the AquAdvantage Salmon meeting were selected to serve on a standing committee that would provide recommendations and opinions to the FDA Commissioner on various scientific and technical issues. In addition to those on the standing committee, CVM could invite additional experts to provide their scientific opinion in a particular area. To provide a rigorous and robust discussion for this application, FDA invited leading experts in salmon biology, fisheries, aquaculture, as well as molecular biology and production of GE animals to participate. A consumer group advocate with expertise in GE food-related issues was also a member of the VMAC. All of the members, including the additional experts, were subject to prior screening for conflicts of interest. See 18 U.S.C. § 208; 21 U.S.C. § 379d-1(c)(2); Guidance for the Public, FDA Advisory Committee Members, and FDA Staff on Procedures for Determining Conflict of Interest and Eligibility for Participation in FDA Advisory Committees, http://www.fda.gov/downloads/regulatoryinformation/guidances/ucm125646.pdf. This screening determined that committee members had no such conflicts. Accordingly, the Agency believes that the VMAC was fairly balanced in its membership in terms of the points of view represented and the functions to be performed. 5 U.S.C., App 2; 41 CFR 102-3.30.
iv. Post-Approval Issues
Comment Summary: One comment expressed a concern that ABT may sell AquAdvantage Salmon eggs to other companies for grow-out and import to other countries; another comment stated that it was not clear whether FDA imposed any restrictions on ABT’s sale of the AquAdvantage Salmon eggs.
Response: The conditions established in the approved application, AquAdvantage Salmon are to be bred and raised only under the conditions specified in the approved application: breeding in a specific FDA-inspected facility in Canada and grown-out in a specific facility in Panama where FDA and NMFS conducted an extensive site visit. This approval does not extend to any other locations, facilities, or conditions. It does not allow raising AquAdvantage Salmon in the United States. Other sovereign nations may determine whether they wish to approve or import AquAdvantage Salmon for themselves. If the sponsor wishes to grow AquAdvantage Salmon in the United States, before doing so it would be required to submit a supplemental NADA, which would require FDA evaluation and approval, as well as meeting other federal, state, and local requirements.
Comment Summary: Some comments asked enforcement-related questions such as about FDA’s authority to address escaped fish, what FDA would do if the sponsor is not in compliance with its durability plan, and what FDA would do if the construct changes enough that it no longer constitutes the same drug.
Response: ABT has multiple, redundant systems in place to prevent escape of fish. In the extremely unlikely event that the entire system failed and fish did escape, FDA would consider whether to take appropriate enforcement action under applicable provisions of the FD&C Act and its implementing regulations. Provisions accounting for escaped or missing fish are described in the record-keeping and reporting requirements specified in Appendix A of the Approval Letter. Similarly, if the sponsor was not in compliance with a condition established in the approved application, including the durability plan, this likely would violate the FD&C Act and implementing regulations and, in that case, FDA could take enforcement action. FDA will monitor AquAdvantage Salmon post-approval for, among other issues, the stability of the rDNA construct. If there is a change to the rDNA construct such that it no longer constitutes the same drug, then FDA could take enforcement action and/or require submission of a new application.
1 The NADA is for approval of the α-form of the opAFP-GHc2 recombinant DNA (rDNA) construct at the α-locus in the triploid, all female EO-1α line of Atlantic salmon under the conditions of use specified in the application. For ease of reference, this document refers to the application as being for approval of the AquAdvantage Salmon.
2 AquAdvantage Salmon are the triploid, hemizygous, all-female Atlantic salmon from the E0-1α lineage genetically engineered (GE) Atlantic salmon subject to this application. They are a subset of ABT salmon, which are defined as any GE Atlantic salmon from the E0-1α lineage, irrespective of ploidy, zygosity, or gender (i.e., the set of salmon that includes diploid GE salmon that may be used as broodstock, as well as AquAdvantage Salmon or other triploid GE salmon).
4 The eight major food allergens are peanuts, tree nuts, milk, eggs, wheat, soybeans, fin fish, and crustacean shellfish. 21 U.S.C. § 321(qq).
5 A homozygous individual has inherited the same alleles for a particular gene from both parents.
6 A hemizygous individual has only one copy (or allele) of a given gene.