Sunday, October 03, 2010

GMO proponents should call for stronger safety testing of the AquAdvantage genetically modified salmon

The FDA recently held hearings to discuss the safety and labeling of a genetically modified salmon, called AquAdvantage, from a company called AquaBounty. Scientists have added a gene to the DNA of this fish, so that it has a longer period of rapid growth each year.

FDA will now take some time to consider approval for human consumption.  While milk is already commonly produced using a genetically modified growth hormone, the new salmon would be the first genetically engineered animal approved for direct human consumption.

As you know, genetically modified organisms (GMOs) are one of the most controversial topics in U.S. food policy.  Many critics argue that GMOs, along with many other modern production technologies, promote an environmentally unsustainable industrialized food system.  Many supporters argue that GMO technology is essential to feed the world's growing population. I don't like either of these across-the-board claims, myself, but that is a topic for another day.  Because of this larger controversy, if this salmon is the first such product approved by FDA, the consequences will be closely scrutinized and will influence the broader debate about agricultural technologies for years to come.

GMO supporters should reflect on what standard of food safety they want this new product to meet.  For example, if the new salmon turns out to have a slightly higher allergenicity than conventional salmon, leading to a slight increase in hospitalizations and deaths, it will not fly to say, later, "well, people with allergies probably shouldn't even be eating finfish like salmon anyway."  If the new salmon turns out to have slightly elevated average concentrations of a growth factor that has been associated with risk of cancer, it will not be adequate to say, later, "only a few of the additional cancers can be directly attributed to the GMO technology."  The politics of this debate are such that GMO proponents should want the first GMO animal food to be as safe as conventional food.

If they have any strategic vision at all, GMO proponents should call on FDA to require additional tests before approving the AquAdvantage salmon.  The alternative is to risk a damaging mistake. An excellent prescription for the additional testing is provided by testimony to FDA (.pdf) from Michael Hansen of the Consumers Union.  Hansen is widely considered a committed critic of GMO technology, so GMO proponents who read this far will now be tempted to quit and ignore the rest of the post.  That would be a mistake.

I think GMO proponents should read Hansen's testimony as both friendly and frightening.  The testimony is friendly, because it is not religiously anti-GMO, instead offering a long but finite list of complaints about the AquaBounty and FDA safety analyses.  The testimony is frightening, because it seems quite convincing that the safety evidence is incomplete.

If you read Hansen's testimony as a prescription for additional studies, you would recognize that FDA's decision would be delayed by perhaps a year or two, but not a decade.  The main studies that would be required are: (a) studies of veterinary effects from Panama, where the fish will be grown, to supplement the existing studies that are all from Canada, (b) better studies of allergenicity, which remedy some serious flaws identified by Hansen, (c) studies of changes in concentrations of growth hormone and a related growth factor, which remedy flaws identified by Hansen.

The great thing is that you can read Hansen's testimony (.pdf) and FDA's safety submission (.pdf) side by side for yourself.  You don't have to believe Hansen's judgment more broadly.
  • If Hansen claims that a study of allergenicity was done with too few fish, you can read pages 98-100 in the FDA submission to confirm that FDA recognizes he is correct.  
  • If Hansen says that a study of the presence of growth hormone lacked sufficient sensitivity to detect any growth hormone in either GMO fish or control fish -- even though it seems the whole point of the genetic modification is to increase growth -- you can read Table 15 in the FDA submission to verify that all growth hormone measurements seem to be indistinguishable from zero. 
  • If Hansen claims that FDA strangely appears to have made it look like there was no detectable difference in a growth factor (IGF-1), by adding to the sample fish for which the research methods could not detect the growth factor at all, you can read for yourself the footnote and sample sizes in FDA's Table 16.  I am not sure, but FDA may have done what Hansen claimed.
For the second of these three points, about the inability to detect any growth hormone, Hansen is quite damning.
This is not a scientifically valid statement. How can FDA conclude that there are no biologically relevant differences in growth hormone levels between GE and non-GE salmon when the study uses a methodology that cannot detect growth hormone in these fish? This would be like the police using a radar gun that cannot detect speeds below 120 mph and concluding that there is no “relevant difference” in the speed of cars versus bicycles.
I can think of a couple rebuttals GMO proponents might make to Hansen's testimony.  The proponents could say that these safety concerns are secondary, and that most of the safety evidence was favorable.  But, I disagree.  My impression after close reading is that Hansen picked on the same safety issues that FDA focused on, implicitly accepting FDA's and AquaBounty's definition of the leading concerns.  That was fundamentally generous of him -- and many GMO critics would have offered a longer list of concerns -- but it means that Hansen cannot be accused of cherry-picking safety concerns.

The proponents could say that these additional tests are just the beginning, and that there would be no satisfying Consumers Union and Hansen after any volume of testing.  I think that misses the point.  Proponents might not want to offer to go ten more rounds on this scientific debate, but they will look like they are fleeing the scientific evidence if they don't go at least one more round.

As for the science, there are three possibilities: (a) the new studies that answer Hansen's challenge may show no evidence of increased risk, (b) they may show small evidence of increased risk, or (c) they may show strong evidence of increased risk.  GMO proponents should want FDA approval only in case (a).  For any other result, they should wait for a different first genetically modified animal for human consumption.

No comments: