Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
The recent French study that suggested a link between genetically modified (GM) maize and cancer has attracted considerable debate and criticism on scientific grounds (1). Francois Houllier in his comment in Nature (2) asked for more rigour in research of this kind but he did not include animal welfare.
In the experiment, rats were fed a lifetime diet of GM maize that had been sprayed with weed killer. Rats were allowed to develop cancerous tumours up to 25% body weight (photographs in the paper show rats with tumours of 70mm in length) (3). The authors set humane endpoints far in excess of established recommendations which did not seem at all necessary to achieve the aims of the study. 50% of males and 70% of females during the study were found dead or were killed at the ?humane? endpoint. The National Research Institute states that multiple tumours should not exceed the maximum burden of a single tumour of 25mm diameter for rats (4), the US Institutional Animal Care and Use Committee says that tumours should be no larger than 40mm in adult rats (5) and the US National Academies National Research Council says that tumour burden should not usually exceed 5% of the animal's normal body weight (6). It is important that humane endpoints are set but they should not be set any higher than is scientifically necessary. Broader international consensus on what some common humane endpoints should be will help in future.
Regulatory studies have established methodologies that ensure that the results will be accepted within the international community. Academic studies that fall below these standards are likely, quite rightly, to be treated with scepticism, as has happened here. Local and national ethical review bodies need to pay more attention to studies intending to improve existing knowledge on the hazards of substances, including GM food. They should ensure that such studies do not fall below the rigour and ethics expected not only from the regulatory community but the animal research community at large. We are taking up our concerns with the local review board at Caen University that granted the authorisation for this particular study and the journal that published it.
Dr. Katy Taylor British Union for the Abolition of Vivisection, London, UK. Katy.taylor@buav.org
1. GM study rebutted, Nature 492, 6 December 2012, p 12. 2. Houllier, F. Bring more rigour to GM research. Nature 491, 15 November 2012 p 327. 3. Seralini, G-E et al. Food Chem. Toxicol. 50, 4221-4231 (2012). 4. Workman P. et al. Br. J. Cancer 102, 1555-1577 (2010). 5. The US Institutional Animal Care and Use Committee (IACUC) 2009 IUCAC/LARC Standard Procedures Guidelines For Tumor Production In Rats And Mice. Available at: http://www.iacuc.ucsf.edu/P... 6. US National Academies National Research Council report on Recognition and Alleviation of Pain in Laboratory Animals. National Academies Press 2009.
Vincent DEMOULIN
	Katy Taylor point is well taken and one may wonder if one of the inadequacies of the Seralini et al. study, that is to use in a life-long study a strain of rats prone to spontaneous tumours, was not deliberate in order to produce spectacular pictures. However it is not only superfluous animal sufferings that may result from the repetition of this kind of experiment but a waste of research funds. 	Indeed, if one cannot but agree with the spirit of François Houiller comment, one can also understand that the head of a big and excellent research organisation takes up an opportunity to obtain more funds. However, in those days of tightening budgets, is not the study of the toxicity of classical agricultural OGM the superfluous chase of a red herring? 	From the first critiques published on-line in the Huffington Post on 21 sept. by J. Fr. Narbonne and H. Bleiberg, to the detailed examinations by experts consulted by six European governments, available as Annex to the final report of the European Food Safety Authority (EFSA Journal 2012; 10(11): 2986) it is evident the Séralini et al. paper is not scientifically valid. Does this mean the study should be started again with an adequate protocol? The only positive result of this non-event is that the French agency ANSES did commission the translation of the paper in Japanese by Sakamoto et al. (?A 104-week feeding study of genetically modified soybeans in f344 rats? Shokuhin Iseigaku Zasahi 49: 272-282, 2008). Obviously because of the language problem, the importance of this study was not fully appreciated in the review by Snell et al. (Food and Chem. Toxicol. 50: 1134-1138, 2012), which failed to note this was a CP4 EPSPS strain. Snell et al. concluded GM plants are safe but they could have been more affirmative. The Sakamoto et al. study is the well designed (50 rats by group) equivalent of the one by Séralini et al., with the expression of the same enzyme CP4 EPSPS, but on soybeans It does not show a significant effect. Of course opponents to GMO will say ?yes, but this is soybean, we want to know with corn? but when a good two years study with maize NK603 will show its innocuity, they will endlessly ask for one with every other strain. 	That the good Sakamoto et al. research was locally published is indicative of the difficulty to publish negative results in high-impact journals. Few biologists will expect significant results from a toxicity study of Round-up ready crops. Is it worthwhile to encourage them to spend time and money on such studies? I doubt it. The fight between GM-crops haters and rat lovers will be interesting to follow, but for once sound scientists should stand with the rat lovers. 	The Séralini et al. paper has been reviewed by something like 60 experts in Europe. A conservative estimation of the average time spent would be 4 hours (I know of a statistician who spent 24 hours making simulations to demonstrate the non significance of the results) and of the salary cost, 80 euros by hour. This means, beside spreading unjustified concern in the public, Séralini et al. have already cost some 20 000 euros to the collectivity, is this is not enough?
Vincent DEMOULIN, Hon. Prof., University of Liege, Dept. of Life Sciences, B.22, B-4000 LIEGE, BELGIUM V.Demoulin@ulg.ac.be
Peter Cary
On behalf of Gilles-Eric Séralini:
Health effects of pesticides are overlooked in GMO risk assessments G.E. Séralini, R. Mesnage, N. Defarge. University of Caen, Institute of Biology, CRIIGEN and Risk Pole, MRSH-CNRS, EA 2608, Esplanade de la Paix, Caen Cedex 14032, France An international debate arose after we published the first long-term toxicological assessment of the whole Roundup pesticide formulation at environmentally relevant levels (from 0.1 ppb) and a Roundup-tolerant genetically modified (GM) maize (NK603)1. Our results challenge claims that these products are safe. The concern expressed by Francois Houllier (Nature 491)2 that our results could damage public opinion about GM crops gives precedence to economic interests over public health. Nevertheless, we agree that more rigour in GM research is needed. The usual way of commercializing agricultural GMOs and formulated pesticides with short-term or no in vivo testing, as well as keeping raw data hidden for reasons of commercial confidentiality, is not scientifically rigorous. The first focus should be the toxicity of pesticide residues. Current regulatory assessment overlooks the fact that all GM crops are genetically modified to contain pesticide residues (mostly Roundup residues and Bt toxins) together with adjuvants. These pesticide residues are new elements in our diet, either in type or amount and are often not tested in vivo. We showed that Roundup (sprayed in 80% of GMOs) contains new toxics compounds, more toxic than glyphosate which is surprisingly tested alone for chronic effects 3. In order to take the debate forward, more data are needed. First, our experiment should be repeated. The outcome of the debate relies on the access by the scientific community to the raw data that allowed the commercialization of Roundup and NK603. Without this step, the debate cannot be on scientific issues. Our raw data1 were already given to a notary, expecting comparisons. We have answered to all other criticisms4.
1	Séralini, G. E. et al. Food Chem Toxicol 50, 4221-4231 (2012). 2	Houllier, F. Nature 491, 327 (2012). 3	Mesnage, R. et al.Toxicology, doi:S0300-483X(12)00345-9 (2012). 4	Séralini, G. E. et al. Food Chem Toxicol 53, 461-468 (2013).
Katy Taylor
The recent French study that suggested a link between genetically modified (GM) maize and cancer has attracted considerable debate and criticism on scientific grounds (1). Francois Houllier in his comment in Nature (2) asked for more rigour in research of this kind but he did not include animal welfare.
In the experiment, rats were fed a lifetime diet of GM maize that had been sprayed with weed killer. Rats were allowed to develop cancerous tumours up to 25% body weight (photographs in the paper show rats with tumours of 70mm in length) (3). The authors set humane endpoints far in excess of established recommendations which did not seem at all necessary to achieve the aims of the study. 50% of males and 70% of females during the study were found dead or were killed at the ?humane? endpoint. The National Research Institute states that multiple tumours should not exceed the maximum burden of a single tumour of 25mm diameter for rats (4), the US Institutional Animal Care and Use Committee says that tumours should be no larger than 40mm in adult rats (5) and the US National Academies National Research Council says that tumour burden should not usually exceed 5% of the animal's normal body weight (6). It is important that humane endpoints are set but they should not be set any higher than is scientifically necessary. Broader international consensus on what some common humane endpoints should be will help in future.
Regulatory studies have established methodologies that ensure that the results will be accepted within the international community. Academic studies that fall below these standards are likely, quite rightly, to be treated with scepticism, as has happened here. Local and national ethical review bodies need to pay more attention to studies intending to improve existing knowledge on the hazards of substances, including GM food. They should ensure that such studies do not fall below the rigour and ethics expected not only from the regulatory community but the animal research community at large. We are taking up our concerns with the local review board at Caen University that granted the authorisation for this particular study and the journal that published it.
Dr. Katy Taylor British Union for the Abolition of Vivisection, London, UK.
Katy.taylor@buav.org
1. GM study rebutted, Nature 492, 6 December 2012, p 12.
2. Houllier, F. Bring more rigour to GM research. Nature 491, 15 November 2012 p 327.
3. Seralini, G-E et al. Food Chem. Toxicol. 50, 4221-4231 (2012).
4. Workman P. et al. Br. J. Cancer 102, 1555-1577 (2010).
5. The US Institutional Animal Care and Use Committee (IACUC) 2009 IUCAC/LARC Standard Procedures Guidelines For Tumor Production In Rats And Mice. Available at: http://www.iacuc.ucsf.edu/P...
6. US National Academies National Research Council report on Recognition and Alleviation of Pain in Laboratory Animals. National Academies Press 2009.
Vincent DEMOULIN
	Katy Taylor point is well taken and one may wonder if one of the inadequacies of the Seralini et al. study, that is to use in a life-long study a strain of rats prone to spontaneous tumours, was not deliberate in order to produce spectacular pictures. However it is not only superfluous animal sufferings that may result from the repetition of this kind of experiment but a waste of research funds.
	Indeed, if one cannot but agree with the spirit of François Houiller comment, one can also understand that the head of a big and excellent research organisation takes up an opportunity to obtain more funds. However, in those days of tightening budgets, is not the study of the toxicity of classical agricultural OGM the superfluous chase of a red herring?
	From the first critiques published on-line in the Huffington Post on 21 sept. by J. Fr. Narbonne and H. Bleiberg, to the detailed examinations by experts consulted by six European governments, available as Annex to the final report of the European Food Safety Authority (EFSA Journal 2012; 10(11): 2986) it is evident the Séralini et al. paper is not scientifically valid. Does this mean the study should be started again with an adequate protocol? The only positive result of this non-event is that the French agency ANSES did commission the translation of the paper in Japanese by Sakamoto et al. (?A 104-week feeding study of genetically modified soybeans in f344 rats? Shokuhin Iseigaku Zasahi 49: 272-282, 2008). Obviously because of the language problem, the importance of this study was not fully appreciated in the review by Snell et al. (Food and Chem. Toxicol. 50: 1134-1138, 2012), which failed to note this was a CP4 EPSPS strain. Snell et al. concluded GM plants are safe but they could have been more affirmative. The Sakamoto et al. study is the well designed (50 rats by group) equivalent of the one by Séralini et al., with the expression of the same enzyme CP4 EPSPS, but on soybeans It does not show a significant effect. Of course opponents to GMO will say ?yes, but this is soybean, we want to know with corn? but when a good two years study with maize NK603 will show its innocuity, they will endlessly ask for one with every other strain.
	That the good Sakamoto et al. research was locally published is indicative of the difficulty to publish negative results in high-impact journals. Few biologists will expect significant results from a toxicity study of Round-up ready crops. Is it worthwhile to encourage them to spend time and money on such studies? I doubt it. The fight between GM-crops haters and rat lovers will be interesting to follow, but for once sound scientists should stand with the rat lovers.
	The Séralini et al. paper has been reviewed by something like 60 experts in Europe. A conservative estimation of the average time spent would be 4 hours (I know of a statistician who spent 24 hours making simulations to demonstrate the non significance of the results) and of the salary cost, 80 euros by hour. This means, beside spreading unjustified concern in the public, Séralini et al. have already cost some 20 000 euros to the collectivity, is this is not enough?
Vincent DEMOULIN, Hon. Prof., University of Liege, Dept. of Life Sciences, B.22, B-4000 LIEGE, BELGIUM
V.Demoulin@ulg.ac.be
Peter Cary
On behalf of Gilles-Eric Séralini:
Health effects of pesticides are overlooked in GMO risk assessments
G.E. Séralini, R. Mesnage, N. Defarge.
University of Caen, Institute of Biology, CRIIGEN and Risk Pole, MRSH-CNRS, EA 2608, Esplanade de la Paix, Caen Cedex 14032, France
An international debate arose after we published the first long-term toxicological assessment of the whole Roundup pesticide formulation at environmentally relevant levels (from 0.1 ppb) and a Roundup-tolerant genetically modified (GM) maize (NK603)1. Our results challenge claims that these products are safe. The concern expressed by Francois Houllier (Nature 491)2 that our results could damage public opinion about GM crops gives precedence to economic interests over public health.
Nevertheless, we agree that more rigour in GM research is needed. The usual way of commercializing agricultural GMOs and formulated pesticides with short-term or no in vivo testing, as well as keeping raw data hidden for reasons of commercial confidentiality, is not scientifically rigorous. The first focus should be the toxicity of pesticide residues. Current regulatory assessment overlooks the fact that all GM crops are genetically modified to contain pesticide residues (mostly Roundup residues and Bt toxins) together with adjuvants. These pesticide residues are new elements in our diet, either in type or amount and are often not tested in vivo. We showed that Roundup (sprayed in 80% of GMOs) contains new toxics compounds, more toxic than glyphosate which is surprisingly tested alone for chronic effects 3.
In order to take the debate forward, more data are needed. First, our experiment should be repeated. The outcome of the debate relies on the access by the scientific community to the raw data that allowed the commercialization of Roundup and NK603. Without this step, the debate cannot be on scientific issues. Our raw data1 were already given to a notary, expecting comparisons. We have answered to all other criticisms4.
1	Séralini, G. E. et al. Food Chem Toxicol 50, 4221-4231 (2012).
2	Houllier, F. Nature 491, 327 (2012).
3	Mesnage, R. et al.Toxicology, doi:S0300-483X(12)00345-9 (2012).
4	Séralini, G. E. et al. Food Chem Toxicol 53, 461-468 (2013).