Following is a commentary on the findings of Prof GE Seralini and colleagues that NK603 GM maize and Roundup produced severe toxic effects in rats. The author is John J Moore, former professor of biology at the University of Zambia specialising in ecology and biostatistics, former professor of botany at UC Dublin, and a Jesuit priest.

EXCERPT: "I find it amazing that a body [European Food Safety Authority] set up to ensure the safety of food offered on the market within the European Union should shut its eyes to the data reported in Seralini's paper. Despite the academic flaws in its presentation, the data is clear and frightening, a red light warning us of a very real possibility of danger."
---
---
A red warning light: Reflections on Seralini
20 December 2012
John J. Moore, S.J., D.Sc.

Two months have passed since Prof G.-E. Seralini and his colleagues published a scientific paper which indicated clearly that the European Food Safety Authority (EFSA) was ill-advised to authorize in 2009 the cultivation of the GMO maize variety NK603 in the countries of the EU. By repeating the feeding experiment on which the judgment was largely based (Hammond et al., 2004) but extending it for the normal life span of the experimental rats used – i.e. two years – they showed that, starting from about the second year of the life of the rats, there was a massive outbreak of different kinds of cancer in those rats fed with GMO maize. These cancers were similar to the cancers that had already been noted by previous scientists in control rats towards the end of their normal life span of 2 years. However, in Seralini's study they seemed to appear at an earlier stage and in greater numbers.

A Bombshell?

What has been the effect of this "bombshell" as some of the French media called it?  As usually happens when there is question of GMO foods, there have been two diametrically opposed reactions. 

Let us first look at those who, like myself, see the results of this study as a red warning light to all involved with GM food crops.  In Africa an urgent appeal was sent to the African Union (AU) requesting that the next AU Summit, to be held in January 2013, should discuss imposing a ban on the cultivation, import and export of genetically modified (GM) crops in Africa.    This appeal came from several African organisations representing small-scale farmers, religious groups, social movements, organic producers, consumers, business people and ordinary citizens. The statement was supported by a substantive document detailing the failure of GM technology to deliver any of its promised benefits since its global introduction some 16 years ago. In Kenya the government has already banned all GMO food imports. The reason given is because of the inadequate research done on the safety of GMO foods. 

In India a six-member panel of technical experts set up by the Supreme Court of India has unanimously recommended a ten-year moratorium on field trials of all genetically modified food crops.

However, the continuing sniping against Seralini's paper has not ceased.  Apart from the frivolous outpourings of uninformed 'bloggers' there are some contributions from those with some training and expertise in science which we should look at.  They continue to focus on the flaws that were already pointed out in the media within hours of the release of the paper – too few controls, cancer-prone experimental rats, puzzling statistics, etc.  A new approach is to appeal to the editor of the journal "Food and Chemical Toxicology" to withdraw the Seralini paper. A "holier than thou" attitude is assumed – "Seralini's paper is an insult to our integrity as scientists".  They also demand to see the original data and lab notes, something that the GM industry itself has repeatedly refused to do.

The most striking criticism of the Seralini study is the final report of the European Food Safety Authority (EFSA). It criticises the Seralini work very severely, claiming that it is "inadequately designed, analysed and reported. It finds that the paper is of insufficient scientific quality for safety assessments".  Moreover, it refuses to re-examine its own earlier judgement in 2009 legitimising the cultivation of Roundup-tolerant Maize NK603. It states that the Authority is still "of the opinion that maize NK603 is as safe as conventional maize" and passed it for import and planting.

One reviewer, (Dr. Brian John) declares that "vitriol drips from every page" of this final report. Perhaps this is an exaggeration, but it is not far from the truth. I find it amazing that a body set up to ensure the safety of food offered on the market within the European Union should shut its eyes to the data reported in Seralini's paper. Despite the academic flaws in its presentation, the data is clear and frightening, a red light warning us of a very real possibility of danger.

A Scientist's Reaction

It is obvious that GMO controversies are not based merely on arguments between scientists, nor are they just a row between fanatical "greens" and those conservative souls who wish to maintain the status quo.  Big and powerful commercial interests are obviously very active in this area. For me as a scientist who spent most of my life training students to be scrupulously honest and accurate reporters of what they observed, it is the scientific data that interests me not the commercial interests. What is the significance of these horribly deformed rats?

I have found the assessment of Seralini's work by the French High Council for Biotechnology[1] (HCB) extremely helpful.  It is written in a clear detached way and also performs the required statistical tests that were so obviously missing in Seralini's paper.  I cannot quarrel with their conclusion that the paper "does not present conclusive results regarding possible toxicity of maize NK603".  Four French government ministries had asked the HCB to determine precisely this - whether the paper might be considered "conclusive regarding the possible health risk of food derived from GM plants containing event NK603".

I agree that the paper itself cannot be considered "conclusive", i.e. proving "beyond all reasonable scientific doubt", but it does present very important and disturbing data.  There was something going on in that laboratory at Caen that I consider to be a warning red light. 

The Harlan company, which supplied the rats for the Seralini experiments, was able to supply data on the survival rate Sprague-Dawley laboratory rats at the end of 2 years.  Statisticians from the French HCB, basing themselves on these data, were able to show that none the mortality data reported by Seralini et al. for their different treatments were outside the range of 95% probability.    In their own words - "we cannot explain the differences observed in the survival curves of different experimental and control groups by invoking the effect of diet". 
However, the survival rates at the end of two years do not really provide us with any useful data about the "mid-life" period of the rats.  Seralini's diagrams seem to suggest that this "mid-life" period might be the crucial time when the GMO maize was influencing the development of tumours. 

Fortunately a paper published by McMartin and his colleagues in 19922 may enable us to make a more precise statistical evaluation of Seralini's data. McMartin reported on data that had accumulated in the Ciba-Geigy laboratory between 1984 and 1991. His main interest was to report on the various cancer-like disorders that developed in 1,170 control Sprague-Dawley rats which had been used as controls in 9 separate studies. As well as the survival rates at the end of two years, the paper also gave the figures for the average time (in days) when precisely 75% of the rats survived (or 25% had died) during these 2-year long experiments. 

If we make the assumption that these 1,170 rats could be considered as a 'quasi-control group' for Seralini's data, then we should be able to calculate the probability of the premature deaths of Seralini's rats fed on GMO maize.

McMartin reports that 75% of his control rats were alive, on average, after 573 days (for males) and 582 days for females.  Since he also gives the standard deviations and range of these averages, we can calculate the probability of rats dying prematurely 'by mere chance'. My calculations tell me that the probability of male rats dying naturally before they are 484 days old is less than 0.1%. For females the corresponding 0.1% point is less than 523 days. Three out of the six GM treatments of the male rats fall below this level; 2 out of six for the females. This is the basis of my red light warning. 

Many objections can be made against my using McMartin's data as a 'quasi-control group'. The data is old, it is from USA, I am using 70% survival data for the Seralini data, I am not dealing with a continuous variable, etc., etc.

A still larger data-set exists in Hannover (Germany) at the Fraunhofer Institute of Toxicology and Experimental Medicine with similar data for 17,380 laboratory rats used as controls in different feeding experiments over the years. I am convinced that if somebody has the leisure to examine this database, strong and valid evidence will be found to prove that the Seralini data is a genuine warning red light.

To sum up: Seralini's paper has done two things to promote the advance of our scientific knowledge in this whole area of GMOs. 

(1) It has shown conclusively that the 90-days feeding tests previously used to validate the release of GMOs for commercial or agricultural use were of little use. The tests must extend over the full life expectancy of the rats, preferably 2 years, if the danger signals are to be picked up.

(2) The onset of tumors in rats fed on GMO Roundup-resistant maize NK603 is advanced from the last weeks of a rat's life to the early second year of a rat's natural lifespan.

One way of bringing out the importance of these data may be to put them in human terms by equating the life span of a rat to our human one. Two years (700 days) for rats can be taken as the equivalent of 70 years for a human.  The surge of new tumours begins at around 400 days in the rat experiment. This would correspond to the early forties for a human.

Those responsible for the 'safety' of our staple foodstuffs thus have a real responsibility to demand that a well designed 2 year feeding experiment must be carried out whenever the Biotech Companies are seeking permission for the release of GM varieties of our staple foods.

Ireland's GM Potatoes

The case of Ireland's GMO potato field trials illustrates this very well.  When the Irish agricultural research institute (Teagasc) applied for clearance for field trials on GM Potatoes modified for resistance to Potato Blight, 83 objections were submitted to the authorising body (EPA). Many people wished to maintain the image of Ireland as a GM-free zone since this image was an important element in overseas sales techniques for Irish food products. There were also scientific objections from the Anti-GM lobby. However the Teagasc submission merely follows closely (even slavishly) the format set out in the legislation which was enacted in 2003.

In the light of these new data presented by Seralini, the guidelines drawn up a decade ago need revising, not just in Ireland, but worldwide, especially for human food crops. If the GM crop for which clearance is sought is a major component of human diet, the first question must always be: "have these genetically modified varieties been satisfactorily shown to be without any unwanted harmful effects for human populations who eat them in large quantities?" In a word, "are they safe"? Normally the answer should be based on a feeding experiment on laboratory rats lasting for 2 years.

The vociferous attacks on Seralini in the media have thus not ended the GMO controversies – on the contrary, they have opened up a new area of research that is urgently needed. The lights have not yet turned to green!

Notes

1. HIGH COUNCIL FOR BIOTECHNOLOGY. Opinion on the paper by Seralini et al. (Food and Chemical Toxicology, 2012).
2.  McMartin, D. et al. Neoplasms and Related Proliferative Lesions in Control Sprague-Dawley Rats from Carcinogenicity Studies. Historical Data and Diagnostic Considerations.  Toxicol Pathol 1992. 20: 212.