A-maize-ing Chardon LL fiasco
By Lim Li Ching, TWN/ISIS
In March 2000, the UK government proposed adding Chardon LL, a variety of genetically modified T25 maize developed by Aventis (formerly AgrEvo), to the National List. Chardon LL is a fodder maize for cattle, genetically modified to be resistant to Aventis’ own herbicide, glufosinate ammonium (Liberty).
GM varieties may only be added to the National List after fulfilling EC legislative requirements. T25 maize had already received marketing consent in 1998 under Council Directive 90/220/EEC on the deliberate release into the environment of GMOs, which allows commercial cultivation and import for processing into animal feed and non-food uses, and novel food authorisation for use as a novel food or food ingredient, under Council regulation 258/97. Since a variety must be included in the National List or the EC Common Catalogue before it can be sold commercially, National Listing is the final hurdle in the regulatory process. If approved, Chardon LL would be the first GM variety on the National List.
There were widespread public objections to the government’s proposal, and a public hearing was convened from October to November 2000, to consider these. During the hearing, it became apparent that Aventis’ application to support commercialisation of Chardon LL was based on highly suspect, flawed science.
Dr Steve Kestin and Dr Toby Knowles from the University of Bristol, presented evidence that nutrition tests on broiler chickens, fed GM and non-GM maize, submitted by Aventis to support their application, was woefully “inadequate” [1, 2, 3]. They asserted that the research “…is not of a standard that would be acceptable for publication in a scientific journal”.
The research methods were strongly criticised. There were “far too few” replicates (only 4), limiting its statistical power. Yet, using such a small number of replicates would be a convenient way of showing no effect, which is precisely what Aventis claimed. And the absence of a positive control was a “major flaw in the design”, which would not enable determination of whether the lack of effects noted was really because there is no difference between the diets, or because the experiment had not worked due to other factors. Similarly, the study’s statistical analysis was found sorely wanting.
Concerns were also raised about “suspicious” higher death rates among chickens that ate GM maize – twice as many chickens died, compared with those fed on conventional maize. Whilst the results were not statistically significant, the trend for higher mortality in GM-fed chicken should have alerted to either flaws in the study or a real direct effect of the GM diet. The 8% mortality rate recorded was double the usual rate amongst the UK broiler industry.
Dr Kestin and Dr Knowles have reaffirmed their earlier conclusions, on BBC Radio 4’s Farming Today programme on 27 April 2002 [4, 5]. Dr Kestin said ‘the chicken study’ was “not really good enough to base a student project on, let alone a marketing consent for a GM product. But it does surprise me that we have got so far down the line of licensing a GM crop apparently based on very weak and thin science”.
Having gone through a battery of regulatory and approval scrutiny, it is shocking that questions are still being raised as to the adequacy of the chicken study. That this particular piece of poor, incomplete science was submitted to support Aventis’ application, and that government accepted it, is astonishing. What’s more, it is symptomatic of the other studies Aventis submitted.
The Advisory Committee on Releases to the Environment (ACRE), earlier responsible for UK approval for marketing of T25 maize under EC rules, conducted its own hearing on the criticisms of T25 GM maize risk assessment, on 20 February 2002. The criticisms highlight that the scientific evidence forming the basis of T25 maize approval is still hotly disputed [6, 7].
Dr Mae-Wan Ho and Dr Ricarda Steinbrecher were concerned about mistaken assumptions by Aventis and the EU Scientific Committee on Plants that the cauliflower mosaic virus (CaMV) 35S promoter driving the expression of the pat gene for glufosinate tolerance is inactive in bacteria, and so would not matter even if taken up by bacteria [6, 8]. In fact, the CaMV 35S promoter is active across the entire living world, including in bacteria and human cells. Recombination hotspots associated with CaMV 35S could enhance horizontal gene transfer and recombination. If transferred to human or animal cells, it could activate cancer-associated genes and dormant viruses that are in all genomes.
Worst of all, as Dr Ho argued, Aventis had not provided evidence that Chardon LL is genetically stable [9]. Structural instability, the tendency for transgenic DNA to come loose, to rearrange or become lost in part or in whole in successive generations, could change the transgenic line in unpredictable ways in terms of health and environmental risks. In effect, blanket approval was granted for what may be an unstable, non-uniform variety that will likely undergo changes affecting safety in subsequent generations [10]. Instability increases the chance of horizontal gene transfer and recombination. New disease-causing viruses and bacteria could be generated, and the ampicillin resistance gene in the insert could spread to pathogens. Horizontal gene transfer of the pat gene to soil micro-organisms would have far-reaching impacts on soil ecology. Transgenic DNA may insert into the human genome, with possible adverse side-effects. Unfortunately, the question of horizontal gene transfer was not addressed by the risk assessments on Chardon LL [8].
Other disputed science relates to Aventis’ food and feed safety assessment of T25 maize, which it submitted as evidence of “no concern” for its safety. This despite statistically significant differences found between GM and non-GM varieties, both in grain and silage [3, 11, 12]. Differences were dismissed without adequate explanation and the varieties were considered “not materially different”. Furthermore, its assertions of compositional equivalence, based on chemical tests, say nothing about biological risks, or that novel genes could cause changes that affect toxicity. The compositional analysis was judged “not optimally designed” by Dr Vyvyan Howard, with too few specimens to allow suitably powerful statistical analysis.
Aventis had also examined how quickly the pat gene was broken down by gastric juices of swine, chicken and cattle. (If a protein takes a long time to break down this indicates it may be toxic or allergenic). The experiment, conducted in a laboratory simulation of digestion, was deemed unrealistic by Dr Howard because the acidity of the simulation was much higher than in a real animal gut [3, 12]. This gave the impression that the pat gene could break down more quickly than in reality. Further, the pat gene used was extracted from GM oilseed rape, irrelevant to safety tests on GM maize.
When marketing consent was first granted, the chicken study was one of two animal studies relating to T25 maize submitted; the other was on rats. Chickens and rats are monogastrics and have completely different digestive systems from ruminants. What relevance do these experiments have to assessing the safety of feeding GM maize to cows?
The rat study aimed to evaluate toxicity effects, using refined extract from GM oilseed rape. However, the relatively short-term exposure of 14 days cannot realistically model the toxicity that might occur from lifelong exposure to GM maize-derived forage at > 50% of the diet [12]. Rats are also able to detoxify toxic substances at much higher rates than larger animals. And feeding rats purified protein (not even extracted from GM maize) is immaterial to the entire GM plant being consumed by cattle. In fact, Dr Howard countered, by feeding purified proteins, the experiment is designed not to detect pleiotropic impacts, which would only be evident when using the whole plant.
In any case, Aventis was strongly criticised by Dr Bob Orskov for not carrying out microbial and host animal metabolism tests [13]. Although forage is fed directly to animals, microbes inhabiting the stomach are actually the ones being fed, as they degrade the digestible parts. And whilst chemical analysis might be used to characterise new feedstuff, it is inappropriate for GM feedstuff, as it would not pick up poisonous plant secondary compounds that may be subsequently produced.
The original risk assessment for T25 maize would be unacceptable today, as it did not include assessment of indirect environmental impacts particularly under actual conditions of use, a monitoring plan, or consider its safety as animal feed [3, 14]. Since EU approval was granted, the Deliberate Release Directive has undergone major revisions, which substantially increases the rigour of risk assessments required. Moreover, any application must now include a monitoring plan to detect unexpected impacts. Since T25 maize had been approved without meeting these requirements, it can’t be said that all measures have been taken to avoid adverse effects.
ACRE is considering the disputed evidence and will advise the government on the validity of the original decision to grant marketing consent, and review its advice on T25 GM maize being grown in the UK crop trials. In the meantime, ACRE’s view is that there are no urgent safety concerns that need to be immediately notified to the government [15].
Lord Alan Gray, chair of ACRE, has however recently admitted that the chicken study should have been reanalysed, and that safety tests are not sufficient to accurately assess risks [5]. He also admitted that at the time of original approval, ACRE only saw a summary of the chicken study. He however defended the original decision, saying that expert advice indicated there was “nothing in any of the data… which made them believe there was a risk to the animals, humans and the environment” from feeding T25 maize. Yet, this is only valid because the science and research were so poor that no useful information could be derived.
The truth is T25 maize was approved based on bad science, and Aventis got away with submitting poor data. Surely it is not enough to rely on data supplied by the very same corporations who apply to market their GM products, for which there is no independent verification of accuracy. There is urgent need for sound, independent science to be the basis of risk assessments and decisions on GMOs.
Likewise, the regulatory process on GMOs needs to be vastly improved. Peter Ainsworth, Department for Environment, Food and Rural Affairs (DEFRA) shadow secretary of state, called for an overhaul of the whole approvals process for GM crops [4, 5]. “This GM maize should never have been given commercial approval in the first place. The fact that poor and incomplete science didn’t prevent government advisors from giving it the go-ahead is a major cause for concern and highlights the lack of rigour applied to GM food and crops”.
The precautionary principle, as affirmed in the Cartagena Protocol on Biosafety, should be the basis of regulation of all GMOs, including Chardon LL. The Biosafety Protocol is the first international law to regulate genetic engineering, and recognises GMOs as inherently different, carrying special hazards and risks. As such, in the absence of scientific certainty, a party should err on the side of caution, in view of the potential serious threats.
The public’s right to object to the safety of GM crops is also crucial given that bad science and bad decisions prevail. Yet, even this right of public objection may soon be removed. A leaked government memorandum proposes an “urgent” change in the law to stop the public challenging the safety of GM crops at public hearings [16]. Such an amendment will make it difficult for the public to object to commercialisation of GM crops, as the hearings are the only point in the process where they have a legal right to raise safety and other concerns.
DEFRA proposes restricting representations at future hearings only to matters relevant to the criteria for National Listing –“distinctiveness, uniformity and stability and value for cultivation and use”. In effect, the changes mean that consideration of GM safety would revert solely to the government’s advisory committees – which have previously approved GMOs and whose scientific bases of decisions are now being questioned.
The memo urges the government to “proceed rapidly” to amend the law, immediately after the reconvened Chardon LL hearing ends in June 2002, to come into effect before Ministers decide on the listing of Sheridan (another Aventis T25 GM maize) around October 2002.
The Chardon LL hearing, suspended after it became apparent that French tests for distinctiveness, uniformity and stability did not conform to regulatory requirements, resumed on 29 April 2002. Aventis, who at the start of the hearing announced it would not present evidence or a witness, (which meant that it could not be cross-examined), introduced new evidence in its submission of 29-30 May. Objections were raised to this procedural anomaly, and a compromise eventually reached: the evidence was allowed, but objectors would have the right to present new evidence in reply. After considering evidence from the hearing, and advice from ACRE and the Advisory Committee on Novel Foods and Processes (ACNFP), Ministers will decide on Chardon LL’s National Listing fate.
1. ‘Scientists slam GM research: Key industry study unfit for publication’, Friends of the Earth press release, 3 November 2000, http://www.foeeurope.org/press/scientists_slam.htm
2. Kestin, S. and T. Knowles (2000) An analysis of “the Chicken Study”, joint proof of evidence on behalf of Friends of the Earth.
3. Sound science? The evidence against Aventis’ GM maize, Friends of the Earth, February 2001.
4. GM safety tests ‘flawed’, BBC News, 27 April 2002, http://news.bbc.co.uk/hi/english/sci/tech/newsid_1954000/1954408.stm
5. ‘Concern over safety of GM crops’, transcript from Radio 4’s Farming Today programme, 27 April 2002.
6. Ho, M.W (2002) ‘GM maize approved on bad science in the UK’, ISIS Report, 26 February 2002, http://www.i-sis.org.uk/GMmaize.php
7. Verbatim transcript of the Advisory Committee on Releases to the Environment (ACRE) T25 maize hearing, February 20th 2002.
8. Steinbrecher, R. (2000) Risks posed by horizontal gene transfer with regard to Chardon LL maize, proof of evidence on behalf of Friends of the Earth.
9. Ho, M.W. (2002)‘The best kept secret of GM Crops’, ISIS witness statement to ACRE, 14 February 2002, http://www.i-sis.org.uk/secretGMcrops.php
10. Ho, M.W. (2000) ‘The proposed decision to add Chardon LL (Aventis-T25 Maize) to the National List’, ISIS written representation, 20 April 2000, http://www.i-sis.org.uk/appeal.php
11. Kearney, C. and E. Diamond (2001) Bad science, bad decisions: the evidence against Aventis’ GM maize, Friends of the Earth.
12. Howard, C.V. (2000) Analysis of key documents relevant to the safety of Chardon LL for animal feed purposes, proof of evidence on behalf of Friends of the Earth.
13. Orskov, B. (2000) Statement on required testing of GM forage for feeding to animals,relevant to the proposed decision to add GM maize variety Chardon LL to the National List, proof of evidence on behalf of Friends of the Earth.
14. Mayer, S. (2000) Changes to the regulation of the environmental safety of releasing genetically modified organisms into the environment since December 1998, proof of evidence on behalf of Friends of the Earth.
15. ‘Proposed spring GM crop trial sites published’, DEFRA News Release, 14 March 2002, http://www.defra.gov.uk/news/2002/020314a.htm
16. ‘Outrage at Labour gag on protests’, by Rob Edwards, The Sunday Herald, 19 May 2002, www.sundayherald.com/24815; ‘GM crop protesters to be silenced’, by Geoffrey Lean, The Independent, 19 May 2002, http://news.independent.co.uk/uk/environment/story.jsp?story=296623 and ‘Leaked report shows government scrambling to change GM law’, Friends of the Earth press release, 20 May 2002, www.foe.co.uk/pubsinfo/infoteam/pressrel/2002/20020520103930.html Chardon LL Hearing transcripts available at: www.defra.gov.uk/planth/pvs/chardon/transc.htm