ROCKVILLE, Md. -- While a genetically engineered salmon is almost certainly safe to eat, the government should pursue a more rigorous analysis of the fish's possible health effects and environmental impact, members of a federal advisory committee said yesterday.
The independent panel made up largely of veterinary scientists was convened by the Food and Drug Administration (FDA) to discuss the agency's upcoming decision on whether to allow the commercial cultivation of a fast-growing salmon engineered by the biotech firm AquaBounty Technologies. The panel's deliberations, which included no formal vote, are not binding on the agency, but could prompt an internal review of FDA's protocols.
The modified Atlantic salmon would be the first bioengineered fish approved for human consumption; fluorescent aquarium fish carrying a jellyfish gene have been marketed in the United States for nearly a decade. The FDA finalized its oversight of biotech animals last year, and its current deliberation will likely carry heavy precedent in the future.
Because of this precedent, the agency should be more rigorous in its evaluations, said Dicky Dee Griffin, a professor of cattle production at the University of Nebraska, Lincoln.
"It is extremely important how this precedent gets set," he said. "And it's not an economic issue. It may be, but it can't be. Economics is the shovel with which we dig the grave at the very end of these [deliberations]."
The FDA regulates genes spliced into animals somewhat awkwardly as always-on drugs, since Congress has never passed a bill creating new regulations for biotech crops or animals. If the agency moves forward with its approval of the salmon, it would still need to prepare an environmental assessment, subject to public comment. Once approved, the company says, the salmon could be on supermarket shelves in several years. It remains uncertain if the salmon would be labeled when sold, an issue the FDA is discussing in public meetings today.
AquaBounty's salmon, which are effectively sterile, are designed for growth at inland fish farms. Such facilities have proved to be an unprofitable, if environmentally benign, technique in the United States, where much salmon comes from ocean-based fish farms. It is widely agreed that these farms exact, through excess feeding and escaped fish, a heavy toll on surrounding waters and wildlife.
Safety studies conducted by AquaBounty and made public by the FDA earlier this month often relied on small samples, ranging from 60 fish to only a half-dozen or so. Often, the company's research was conducted upon biotech salmon that were fertile, along with the sterile varieties AquaBounty will sell, causing confusing results. And the methods used to cull the salmon prior to study were not described, prompting concern from the panel.
"Given its study design, sample sizes and mixture of fish ... I would have to characterize this body of work as potentially [promising] but preliminary work that would need to be validated and confirmed in other studies," said Jodi Ann Lapidus, a professor of biostatistics at Oregon Health and Science University.
"Because of the issues related to culling and the data that seems to be floating around that's not in our materials leaves a cloud. It's not partly sunny, necessarily," said Jim McKean, a professor of swine veterinary science at Iowa State University. "There are questions that have not been answered by the data."
While the study design may not have been perfect, the sheer biology of AquaBounty's salmon implies that it will be safe to eat, said Kevin Wells, a professor of animal science at the University of Missouri. The biotech salmon carry a growth hormone gene from chinook salmon that is kept continuously active, even in cold conditions, by a DNA switch from the eel-like ocean pout.
"Everything here is something we eat," Wells said.
Rethinking biotech assessments
Lapidus advocated that FDA completely reconsider how it evaluates biotech animals.
Rather than search for how the salmon differs from unaltered varieties, it should mandate studies that seek to prove the animals are equivalent. Such work would define, in consultation with human nutrition experts, a continuum of biologically relevant ingredients found in typical Atlantic salmon. AquaBounty's salmon would then have to be found to sit within these boundaries.
The intensive study of the salmon also underscores uncertainty in some scientific realms. Like most fish, for example, AquaBounty's salmon can cause allergic responses in one subset of the population. When considering the possibility of additional allergenicity, however, there is no sound science describing what allergen levels would have to be present to cause allergies in previously unallergic consumers.
The engineered salmon would, in the application FDA is currently considering, be approved only for growth at the company's closed spawning facility on Prince Edward Island, Canada, and at a small inland fish farm in the highlands of Panama. However, if successful, the company will seek to sell its eggs to additional farms, each of which would trigger a separate FDA review.
Such a cumulative collection of applications, each separately considered, could see the fish grown at a large scale without ever triggering a broader environmental review. While each farm may have redundant containment systems, future agency assessments would not be available to the public, making it necessary to study the broader effects now, said Anne Kapuscinski, a professor of sustainability science at Dartmouth College.
"The assessment does not adequately address the major questions that should be asked about genetic and ecological risks," she said.
Kapuscinski, who did not serve on the panel, urged the FDA to look beyond the firm's current containment efforts, which she agreed were strong. The agency must complete an environmental impact statement that will fully flesh out the ramifications, however highly unlikely, that fertile biotech salmon could escape to the wild, Kapuscinski said. The company and agency's current assessments stop short of such evaluation, citing redundant containment strategies.
While lauding AquaBounty's containment strategy and saying he would eat the salmon without alarm, Gary Thorgaard, a marine biologist at Washington State University, echoed Kapuscinski's call for a full environmental assessment, to be conducted with other federal agencies.
The accelerated growth provided by the chinook gene would not necessarily cause the fish to enjoy a selective advantage over wild salmon, scientists said. (The fish, while fast-growing, does not grow larger than its domesticated peers.) The engineered salmon's increased food needs could make it more vulnerable to predators. In fact, the salmon could even be preferable to typical domesticated salmon, which often cause havoc on wild salmon populations.
"I'm worried about the transgene getting out much less than [the escape of] domesticated fish," Wells said.