Protocole - Results - Discussion

SECTION ONE : PROTOCOL

Why not use a varied "reference" food on top of your controls, as did Monsanto?

In science we study one variable at a time. We can thus only seriously compare to a genetically similar control shown to be substantially equivalent to precisely study the GMO effect.

You are being criticized for not having mentioned the composition of the feed, how do you explain this?

We have scrupulously followed GPL (Good Laboratory Practice) standards which impose a balanced diet, as well as reinforced controls (with the analysis of mycotoxins, pesticide residues in the feed, as well the detection of pathogens in food, air and water). All feed formulations were made following a standard recipe that ensures a balanced nutrition for laboratory animals.

What do you respond to the criticisms regarding your choice of rat breed?

We used the very rat breed Monsanto use (Sprague-Dawley) in their own toxicological studies on GMOs. This particular rat strain is mostly used in toxicology. The National Toxicology Program in the USA has even recently decided that this particular rat strain was the best model for toxicological studies and it would therefore be their top choice for any future study. What would be the interest of selecting a strain that doesn’t develop tumors to look for carcinogenic effects? Females in this strain develop mammary tumors (whereas males tend to develop kidney pathologies) towards the end of their lives, but what is really important is the difference between the treatments and the controls in terms of tumor occurrence and incidence. We noticed how early and often mammary tumors occurred in the treated females in comparison to the controls. Our study is not about carcinogenesis. We could never have expected such a strong incidence of hormone-dependent tumors, since Monsanto never measured hormone levels. What we have done is to conduct a toxicological life-long experiment, that studies all pathologies. This decision follows more than ever a logical scientific method. We have tested the repeatability of the measured effects in the Monsanto experiment, following OECD guidelines, with exactly the same parameters, and we even went further to try and find out whether the first signs of toxicity confirmed pathologic profiles in the long term. 

What about the number of rats per group?

10 rats per group: it is what was accepted by Monsanto and up to now, it is as well the maximum number of rats per group in GMO assessment and more generally for chemical products toxicology following OECD guidelines. To be perfectly clear, we always compared 10 treated rats to 10 controls.

It is quite true that it results in a quite a low statistical power, that prevents from finding significant statistical effects, and we actually pointed this out in an earlier study (de Vendômois, IJBS 2009). We have therefore decided to be factual and to deliver raw data on the mortality and number of tumors. However we have conducted the highest performance statistics for biochemical data analyses of blood and urine. The method is called OPLS-DA (Orthogonal Partial Least Squares Discriminant Analysis), it is adapted to the analysis of a very large number of variables, as in the case of genomics and proteomics. Such statistics make it possible to determine discriminant variables between each treated group and the control group, and they corroborate what we have clinically observed, in anatomopathology, and optical and electronic microscopy.

Why not have used a standard statistical method?

These methods have not been judged satisfactory by expert agencies to demonstrate toxicity for groups of 10 rats. However, the maximum differences for deaths or tumors with controls (600 days, 2-5 times more) speak for themselves.
In addition, there is an underestimation of the tumorigenic effects at the end of two years compared to controls according to these curves data. This underestimation is due to the fact that the controls are living longer and developing pathologies, including tumors, towards the end of life.


What degree of confidence is there in the significant differences found by the statistical method OPLS-DA, there is no p-values?

This is one of the most modern methods to treat a large number of variables, such as in genomics, indeed the significance does not pass through the p-value reserved for other tests.


Why have you shown biochemical analysis at month 15?

We could not put everything in a first article. This is the last time point when there are the most live rats, which demonstrates significance.


Were the rats treated during the experiment by other molecules? Sanitizers? Antibiotics?

No. Not with GLP (Good Laboratory Practice) in general, otherwise they are excluded from the experiment.


Why were you using two different formulations of Roundup (with crops and contaminated water)?

The formulations contain about 500 g / L of glyphosate. They have different names in different countries.


Where did you grow the maize and treated the rats?

We used maize from Canada because the culture of this GMO is permitted in Canada, unlike in France. Rats were under experiment in France and analysis was performed in different laboratories, some wish to remain anonymous, in France and Italy.

Proponents of this type of study are generally industrial companies and the fact that a NGO and a university developed this study together means that it is independent and unique.


Is the maize used as a control exactly the same as the GMO?

A genetic identity is not possible given the method of seed production but it is the closest genetically and phenotypically.


Were the different maize grown at the same time? Were the climatic conditions the same?

Absolutely yes, and geographically very close, avoiding cross contamination though.


How many times have the maize been treated? At what time?

Once during the culture by the Roundup, when treated. Residues of glyphosate and AMPA in GMOs are recognized and regulated, even in the tissues of animals consuming them. Waivers for overruns are regularly granted, unfortunately. This was not the case for us.


Was there glyphosate in the controls' water?

Not within the limits of detection. We changed the water weekly and contaminated it with the precise doses indicated for treatments.

Why did you use an average and not a median for the aging limit?

It is the general custom; there is little talk of the median life of the French for example. We have not drawn from it any statistical calculations but a graphic mark.

SECTION TWO : THE RESULTS


What is the magnitude of the difference in mortality of the controls compared to the historical norm?

Each experiment having its own conditions, the historical norm is too large to be a relevant comparator. The controls are in the average normal life, and our differences are compared to the controls of the experiment.

How do you explain the absence of manifestation of biochemical disturbances in the males?

On the opposite, there are many. However, all the results are not presented in the study, it was impossible because of their number. There is always a time difference between biochemical disturbances, the firsts to appear, and pathological lesions that we observe in both sexes. In the males the pathological lesions were earlier and larger than in the females and these lesions are the most noticed.

The same differences can be found in all treatments, how do you know that your controls are not the abnormal ones? Or that it is not due to luck?

Our controls correspond to the values observed in the species. Pathological findings have logical explanations for all treatments, they are consistent and numerous enough to not be related to chance; extensive statistics at the biochemical level are consistent and show it. Our in vitro studies are consistent.

How can you be sure that such a little depletion in ferulic and caffeic acids also explains a wide range of pathologies?

There are for us understandable indicators of the changes in the metabolism of the maize that could have happened, the very interesting logical tracks from the scientific literature and our work, they in no way exclude the action of other metabolites toxic due to the GMO and that is why we are asking for funding for analysis in proteomics, transcriptomics, in order to know the events' mechanistic key details.

Do you have any interesting results for the doses of Roundup in the tissues, microbiology and transgene dosing?

Yes we have results that must be completed that give us very interesting leads that will be published later. Several publications are planned after this preliminary work.

Why use the threshold of 17.5 * 17.5 mm in males and 20 * 20 mm in females to
count tumors?

Because it is the threshold size at which more than 95% of tumors are non-regressive.

What is the basis for determining the pathogenicity criteria used in Table 2? What classification do you use?

By differential elimination of the smallest gaps.

Have you measured glyphosate residues in the NK603 or the dry food?

Yes we checked its use and the presence of all pesticides. Values were below regulatory thresholds. The limits of quantification in the different matrices are different.

You indicate an effect on oxidative stress in rats due to your treatments, are the oxidative stress markers disrupted?

Yes for the cytochromes in the liver, and the GST for instance.

Have the maize been sprayed with other pesticides? Have you found other residues?

Yes, normally, they were not organic crops that we could test thereafter. There are no pesticides above the threshold of quantification in food.


SECTION THREE : DISCUSSION


Are the results that you find in this study corresponding to disturbances found in the subchronic tests reanalyzed previously in your publications, that Monsanto sub-interpreted?

Yes, there are signs of hepatorenal toxicity that were published previously after only 90 days of treatment, which are reported as pathologies firmly in our long-term experience.

Do you think that these pathologies may be transferable to humans?

Very generally, yes, but not all. In fact, any signs of toxicity in rats must be taken into account for the prohibition of a product. For 50 years the studies are performed in rat or human cells for products that are not tested in humans (where they test only drugs, not GMOs, nor pesticides nor chemicals). And for drugs, tests on rats or 2-3 mammals precede any clinical trial. If they show serious effects, humans are not treated then. Hormonal disturbances are certainly relevant for women to contribute to breast tumors and hepatorenal effects were found in vitro on human cells.

Why do you quote Zhang et al. 2012 as a reference? This reference concerns on no account GMOs.

One hypothesis is that new micro RNA produced by GMOs may interfere with the metabolism. We could not leave it untold, but we have other explanatory hypotheses.

How do you explain that the reported effects are not found in the human populations? Nobody has ever noticed an increase in breast cancer in the populations exposed to Roundup?

There is an explosion in the number of breast tumors that are not explained by epidemiological studies. We remind you that GMOs not being labeled, the GMO consumption in the U.S. is not listed, nor for the use of Roundup all around the world. The impact of Roundup and its GMOs add to the multitude of xenobiotics present in the environment and in the blood of populations exposed.

It is recommended to experiment on 50 rats for a statutory study on carcinogenesis. What value to bring to your results on 10 rats?

We studied 200 rats, 10 rats/group. Statutory biochemical studies are recommended by the OECD on 10 rats per group minimum.
No statutory study which allowed the authorization of GMOs had more than 10 rats measured per group.
We therefore made the most robust tests in the world, especially as we were examining the long term.
We could not anticipate the results of the tumors, but we observed and recorded them in this study, what was normal, it is not the study of specific carcinogenesis effects with 50 rats/group that would not have allowed to observe the hepatorenal effects and others.

Are the concentrations of ferulic acid found correspond to those indicated in the experiment of Monsanto?

Monsanto has unfortunately not measured the concentrations of hepato-renal and mammo-protective acids directly in the diet, but only once ferulic acid in maize and GMO control.

You say that 76% of the parameters are perturbed in the kidney, in what does that show toxicity? I do not understand this method.

We record all parameters disturbed compared to controls and compare them to the number of parameters related to renal activity over the set of all parameters. We have 48% of renal parameters amongst all measured parameters, yet 76% of the disturbed ones are markers of renal activity! Any doctor would panic for a patient in this case. Distribution cannot be due to chance. (This figure was 42% in one of our previous publications, of disturbed parameters, for 24.9% measured in the male kidney and for a quarterly consumption on average of 19 GMO, testing regulations). The kidney is 1.5 times more affected than other organs.

Roundup increases lifespan in males? Isn't this the best indicator of the safety of this herbicide? This increase is even dose-dependent. Strangely you do not talk about it, why?

We leave you this misinterpretation! Treated males were sicker than controls in all cases, even if in one case out of 6 treatments (3 males + 3 females) there was no extra mortality at any dose before average life expectancy. These rats lose weight however (discussed in another publication) and this can give them some resistance.

Have you compared your results with those of the Japanese study of Sakamoto or another? Contrary to what you say you're not the first to study the safety of one GMO for 2 years.

Yes. None has been as comprehensive as ours, and none is on the maize NK 603 beyond 3 months.

Why did you choose Sprague Dawley rats?

This is the most common model in these studies, the best known.

You've done the study in a GLP environment, is the study GLP? Doesn't the fact that it is not undermine it?
 

There was no standard protocol for this type of very long study with GMO, we establish it while improving it. This is a world first. Thus there could not be any pre-established standard for this type of test. In addition, this research is a protocol where we added analysis along, biochemical and microscopic to understand what was happening. Now it may serve as an example to establish standards for GLP for GMO, much more serious than what the health agencies do today who do work neither scientifically nor honestly. However, we did that in a GLP lab environment, of course.

Produced by the team of Prof. Séralini.