Why do you want some one else's report. You have read the publication of 2012 and you agree with the short falls in that report. Why don't you try to work and seen how far these can be corrected?

Thank you for answering, Jai. Unfortunately, for some reason I cannot see your answer under my question, so I copy it here: 

"Why do you want some one else's report. You have read the publication of 2012 and you agree with the short falls in that report. Why don't you try to work and seen how far these can be corrected?"

Well, Jai, I do not work on that field myself. I have neither competence nor resources to conduct that research. If I did work on toxicology and plant biotechnology, I probably would try to get such a research funded. Which kind of is the point: it seems obvious to me that a long term animal testing of round up would be an interesting research, so maybe somebody has conducted a more methodologically sound research after Séralini et al., but I cannot find a one.

I did not have the opportunity to read the paper you are referring to. However, I want to add this:

Dpeartment of Agriculture and Food Western Australia has been conducting effect of low and high rate of glyphosate on resistance eveolution in rigid ryegrass and weed species shift since 1999. While we have not examined the effect of glyphosate on soil health (level of toxicity), we are now undertaking another project in collabration with University of Western Australian microbiologists to examine the effect of  continuous  glyphostae application on soil micro-organisms. If you have any great idea about the the methodology to get the best results, please feel free to share.

Thanks for answer, Abul. The research I am referring to was an experiment on rats and concerns toxicity to animals. I am not sure if that has many points in common with the research you are conducting.

As far as animal studies intended to inform about human risk, I've not seen much literature on glyphosate.  There is plenty of literature on aspects of environmental toxicity (just do a search for glyphosate toxicity in pubmed).  I'd say the reason is that it is not considered a serious risk for humans (see the pesticide fact sheet at http://npic.orst.edu/factsheets/glyphotech.html and you can also look up its IRIS review summary at the US EPA - http://www.epa.gov/iris ).  Additionally, the estimates of typical human exposures indicates levels hundreds of times less than the oral reference dose (EPA).

Funds for those sorts of very expensive, long term animal studies tend to get allocated towards compounds with significantly less information available, or for which there already is a much clearer basis for concern over risk.  It would be a tough sell to try and raise the funds for say, a 2 year mouse and/or rat exposure study, given the backlog of potentially far more serious chemicals, many of which have not been screened at all.

P.S. note that the paper you mentioned was also formally retracted (http://www.sciencedirect.com/science/article/pii/S0278691512005637 ).

Thank you Michael. This was informative. If I understood correctly you suggest that there is no such research, because it would be expensive and other, less-known chemicals are priority to such expensive studies, as there is low risk that this one would be toxic for humans? I had started to expect something to that effect.

About the interest of such research, I realize that, following the original paper and discussion around it, I only speak about toxicity in my question and should have probably precised the suspicion that roundup, rather than glyphosate itself, was a carcinogenic (see earlier research by Peluso et al. in Environmental and Molecular Mutagenesis 31:55-59 (1998)).

I am aware that Séralini's paper was retracted (the discussion around the retraction is also interesting, at least for a philosopher of sciences like myself, but that's another issue). Had the research been better, I would probably not be asking this question here. It is precisely because it raised an issue that seems interesting - namely that the researches detecting its carcinogenic potential were all made according to the standard of three months period, but it is possible that the effect would only appear after three months - but failed to solve it, that I am interested in potential new long term research with greater number of rats of more suiting strain, in each group, more complete statistical analysis, and so on.

That this issue is a valid one was conceded in the advice not of the French national Academies:

"[H]owever, it remains true that we have now undoubtedly reached an opportune moment to raise and examine the question of experimental protocols to be implemented to detect carcinogenic potential effects of food-stuffs. For instance, is the period of 3 months (the most commonly used time horizon) sufficient or not? This question is relevant probably for both pesticide and herbicide uses. But the problem is not as simple given that the time-scale, in particular as far as life expectancy is concerned, in rats or human beings. But it is certainly not this paper that should lead on to this sort of reflexion since it does not contain any proof as such. It would be especially dangerous even to evoke the possible need to conduct long-term experiments on the basis of this paper, insofar as it would give the impression that the results presented by G.E.Séralini et al. have sufficient validity to justify that the public at large suffers from a fear of GMOs, etc., knowing full well the damage that such fears would cause both in France and round the world."

http://www.academie-sciences.fr/activite/rapport/avis1012_gb.pdf

While they stay firm that the study conducted by Séralini et al. contains no proof that  3 months period was insufficient, they do reckon that studies trying to find potential carcinogenics might require longer time span than 3 months. Now, they do not say that precisely a long term research of roundup should be conducted, but is it not implied that it could be? If not, why not? That they do not suggest such research, may also be because they want to avoid giving Séralini et al. study any merit, even that of raising the issue, as they say in the end of the quoted passage.

But do you think that for some reason the precise issue of carcinogenic potential of roundup is not an interesting subject for a long term research?

In the USA, a normal NTP cancer study is a two year exposure regimen using both male and female rats and mice.  Typical quoted costs are $2M-$4M per compound tested, and total time of study, from planning to final analyzed results is 4-5 years per compound.  Funding usually allows for only about a dozen or so such studies each year.  As such, of the tens of thousands of industrial and agricultural chemicals produced and in use around the world, only a very relatively few have ever been characterized for cancer causing affects.

Roundup is one of those compounds where it seems the overall weight of evidence, and the estimated exposure levels, are such that no, I do not think it is on anyone's priority list for a major toxicity study, at least not one aimed at human risk.  Money, people and time are always highly limited resources when it comes to toxicity testing.  Detailed mode of action studies are particularly time consuming and very expensive, often running into the many millions of dollars per chemical studied.  As such, both government agencies and private corporations have no choice but to prioritize and pick and choose what they study, when, and to what level of detail.

I don't think the paper you mention even really plays into the discussion.  It was highly criticized right from publication and has now been retracted.  I'd think most toxicologists would generally dismiss it as not having anything to contribute to the discussion of roundup or glyphosate toxicity.  That is, I don't think anyone will be basing any funded study decisions on that publication at all - it will be dismissed as simple a case of bad science, and hence of no relevant value in decision making since its results are simply ambiguous (at best).

Nowadays, there is intense interest in developing low cost, very high throughput assays for toxicity testing.  I do not know for certain if roundup was included in the EPA's toxcast assays, but I imagine it was as many commonly used agricultural chemicals were screened.  However, high throughput assays are still very much in a developmental phase.  These assays are also primarily designed for rapid assessment of human exposure hazard, and not intended for interpretation of detailed mode of action. When it comes to carcinogenicity specifically, it is not clear how well such high throughput assays will be in assessing cancer risk (as opposed, say, to affects like acute hepatotoxicity).

So my assessment would be that no, roundup is likely not on anyone's priority list for a multimillion dollar, long term human or mammalian cancer study.  That's not to say it may not be of interest to many, but under the circumstances, it would not be a priority compound for such a major research investment, at least not by industry nor government agencies.

Edit: I don't know why, but a long answer I wrote got cut in the middle and became non-sensical. Another try:

Thank you Michael. I take your word that we will not see that new long term study on roundup, then. It is too bad that not all interesting researches can be conducted. I hope that those low cost very high throughput researches develop well.

One thing I want to comment, though, before I let this one go:

"I don't think the paper you mention even really plays into the discussion. It was highly criticized right from publication and has now been retracted. I'd think most toxicologists would generally dismiss it as not having anything to contribute to the discussion of roundup or glyphosate toxicity. That is, I don't think anyone will be basing any funded study decisions on that publication at all - it will be dismissed as simple a case of bad science, and hence of no relevant value in decision making since its results are simply ambiguous (at best)."

I thought I made it clear already in the question that the study was flawed and I do not think that the new 2 years research should be conducted because of the results of this study. On the contrary, it is because its methodological flaws, most importantly the low number of rats in each group, and the strain of rats, it was unable to produce any reliable results. However, it is not because of these results that I think the long term research would be interesting, nor do I mention the Séralini paper for its results, but because it appeared that all the earlier studies with roundup had the lack that they only lasted 3 months. As a result a potentially significant lack in field of research has become clear, but the study itself was unable to rectify this situation in any manner.  I would have thought that the immediate reaction would have been to conduct this new, better, long term research.

I take your word on that there are other chemicals on which conducting a long and expensive studies is more important. What makes roundup of particular interest is, however, that it and the associated GM crops have been in the centre of a heated debate in EU law-making, and maybe that well conducted 2 year research on rats would decide that part of the debate, at least. No doubt there are heated debates about other chemicals as well, but legalisation of roundup and the associated seeds would probably have a massive impact in European agriculture, in many ways.

It is too bad that Séralini et al. ended up wasting the funding they got for that research. If you are right, it may have been a once in a life time chance.

The decisions to conduct those expensive, long term cancer risk studies are made based on the acute toxicity studies which rarely go beyond a 13 week time point (for both economic reasons, as well as the need to evaluate acute toxicity in as timely a manner as possible to provide a human risk assessment).  In the absence of any real short term data to indicate a basis for concern of long term cancer or other risk, a compound would rarely receive attention for funding for such studies as the list for those is already long and the available funding limited.

From my admittedly limited knowledge of roundup/glyphosate specifically, it does seem that there really is almost no such short term data to justify an expensive long term large scale cancer risk study.  So again, the weight of evidence from short term studies appears not to warrant the expense of a long term animal study.

So I do not think it is a lack of some specific result (nor the lack of specific results from any given study) from the short term studies that is at issue.  It is the actual accumulated direct results of many of those same studies, which in turn, generally indicate little to no evidence of a significant risk.

Keep in mind that any chemical can be toxic under the right set of exposure concentration, or repeated or continuous exposures over a long enough time.

Whether or not someone will fund the kind of study you suggest should be done I cannot say.  I simply don't think it would be done by industry or government funding based on overall available evidence.  Within the EU, the currently changing regulatory landscape is also driving to move away from animal based testing as quickly as possible.  That means right now, a lot of research dollars are going into developing new (non-animal) based technologies for toxicity testing.  Economics and budgets being what they are, that in turn further strains already limited research budgets and so limits the scope of projects that can be funded.   Whether some other (non-industry or non-government) funding source would be interested I do not know.

"Keep in mind that any chemical can be toxic under the right set of exposure concentration, or repeated or continuous exposures over a long enough time."

But precisely for this reason, even if short term studies indicate no concern (and even the one long term one does not prove that there would be one) is it not very well possible that a long term exposure could reveal a cancer risk that does not show in even a very high short term exposure? At least this is the expressed worry, and I don't think it is going away without a long term study that could test it. I know long term studies are more expensive than short term ones, but many short term studies have been conducted on this topic. As it is manifest that this is a substance that we are exposed to repetedly in small doses through our whole lifetime, intuitively it would make more sense to conduct a couple of long term studies rather than many short term ones, right?

In my experience, working in chemical toxicology research, no that is not at all how the decision making process goes.  Short term studies are done in part as they are less expensive (although by no means inexpensive as even they can also easily run into millions of dollars), and they give some useful indication of risk quickly.  However, in the event that little risk is apparent from a short term study, then only very rarely will long term studies be done.

The reason is simple economics.  There are tens of thousands of un- or under-studied compounds already in production and in use.  There are thousands of new ones being cataloged by the chemical industry every year.  There is never enough money, let alone time, to fully characterize all of those compounds as you suggest.  Everyone from industry to government has to make very hard choices of the very, very few compounds that get the sort of study you are suggesting for roundup.  Even then, the backlogs for such studies, on compounds that people would like to see analyzed, is huge, and grows larger every year.

I understand what you are saying, and sure, it would be great to do intensive, long term exposure, detailed mode of action studies on almost every compound.  But logistically it simply could not be done, and there would never be enough money for them all, even if research budgets increased many-fold.

The emphasis thus tends to be on a tiered approach.  Short term assays, like many of the high throughput, low cost per compound in development or under assessment/validation would be the first cut.  That would be used to eliminate further testing of compounds that show no evidence of compelling concern for human risk.  Some form of intermediate throughput and cost assays could then be used to further differentiate those that remain.  And from that, one would pick those exceptional compounds where very high risk is potentially evident and study those in detail.

But that means that many of the compounds considered low or no risk from the initial screening will likely never receive further study - there just is neither the time, manpower, and most importantly, the funds, to do so.

My comment that anything and everything can be toxic, given the right dose or exposure doesn't mean everything needs be studied in great detail.  Modern risk assessment also takes into account exposure.  So while a compound like roundup does indeed show some moderate toxicity at very high doses, it has been noted that those doses are many, many times higher than the average human exposure (as measured in a worst case scenario of agricultural works actually handling the compound according to prescribed procedures).  Thus the perceived human risk from roundup is considered minimal, making it a non-priority compound for further expensive testing.

I apologize for my very poor English and to get out of your interesting discussion.

I believe that the discussion is very interesting because it summarizes a conflict that goes beyond the topic glyphosate.

On one hand, the "precautionary principle" (which is not in what Paul says but that is present in many discussions "culturally" or "ideologically" oriented), which if strictly applied, would not allow that any chemical-technological innovation; Michael makes us see, from another point of view a realistic picture of how the toxicological research are funded.

As Paul said, the particular interest on glyphosate is given by its use in GM plants with all the consequences he mentioned; I would only add that with the reputation of harmless glyphosate has become a long time the most widely used herbicide in GM and conventional crops with very very casual ways in the use; the second point is the “time factor” that binds different things with great evocative power; GMO, the use of meat and bone meal in cattle and Creutzfeldt Jakob's disease and so on. Currently when a substance or procedure appears economically advantageous quickly spreads throughout the world and are not those niche uses that, in the past, formed a kind of experiment that revealed the possible hazards of the substance (an example would be the asbestos; although its use had not been very circumscribed when it exit from the workplace with the use of asbestos-cement building).

I started to follow the story of glyphosate many years ago when I began to study the visceral lesions of tuberous sclerosis (a human genetic syndrome). I went almost immediately to come across in the shikimic acid pathway for the biosynthesis of aromatic amino acids (phenylalanine, tyrosine, and tryptophan). Glyphosate kills plants by inhbiting the shikimate pathway; it targets EPSP synthase that is changed into GM crops and makes them resistant to the herbicide; but this pathway is not found in animals and for this reason I put a little aside the thing.

After a few years it became clear that the TSC2 gene (mutated and ineffective in tuberous sclerosis) was a potent inhibitor of mTOR (mammalian target of rapamycin). Rapamycin is a macrolide produced by the bacteria Streptomyces hygroscopicus using the shikimic acid pathway.

The effect of rapamycin, as many chemicals, is very different to different dosages. At high doses, has an antibiotic effect, while acting in small doses (I'm not talking about therapy) inhibiting mTOR. mTOR pathway is a sort of a central hub node involved in pathophysiological most varied situations. mTOR is the subject of several studies and the use of rapamycin and its analogues in many metabolic and degenerative diseases and cancer is increasing day by day.

Glyphosate definitely change the bacterial flora of the rhizosphere. It would be interesting to know what happens to the Streptomyces hygroscopicus in the rhizosphere or intestines. On the assumption that there has been a sort of evolutionary adaptation to the assumption of small doses of rapamycin and this can be reduced by the use of glyphosate.

This assumption tells us that we can not choose to study a chemical with many years toxicity tests referring to its acute effects (which seems to me a very unreliable way in a general sense).

The above also implies that the toxicological studies should be targeted on many seemingly unrelated characteristics but which indicate the alteration of many molecular pathways that may occur in pathological phenotypes in a long time. From these points of view, as suggested by Abul Hashem makes sense, even as preliminary data for possible studies in animals.

Maurizio, you touch on a number of things that were addressed in a US National Academy of Sciences report  - "Toxicity Testing in the 21st Century: A Vision and a Strategy" (executive summary is available at http://dels.nas.edu/resources/static-assets/materials-based-on-reports/reports-in-brief/Toxicity_Testing_final.pdf ).

There are big hurdles to over come though (what are pathways of toxicity, exactly?, how do we design appropriate in vitro assays to adequately mimic in vivo response?, how do we translate short term dose response into long term risk assessment?, etc).

The days of long term, large scale animal testing are numbered though - the politicians have made that clear.  So however we address these sorts of questions in the future is going to involve some very different approaches and technologies.

One way, relatively simple and inexpensive, to see if what I said above may have some attachment point might be to check with immunohistochemical method (also on tissues and / or organs, paraffin-embedded) if there are signs of hyperactivation of the mTOR pathway. As a first step could also fit animals grown in polluted environment from glyphosate (with particular reference to fish gills, liver and kidney), at a later time should pass in animals chronically exposed to pure glyphosate to exclude non-specific activation of mTOR by other toxic substances and / or additives present in the commercial product of the herbicide (probably for the activation of the dioxin receptor / ARNT ___ HIF1___ etc..). I continue to talk in terms of conventional toxicology, but maybe you can do everything in the lab.

Sorry for abandoning my question, I have been busy with other things. Also thank you for your contributions Maurizio and Michael. I would like to tackle one point that perhaps was not clear to Michael. I can also tell that I have stumbled into a response to my question elsewhere: studies I asked for are probably about to begin soon, or have already recently begun. But I return to that later [I moved it to its own post].

Michael, my point was that  if we are talking about a chemical to which nobody basically is exposed in large amounts at one instance, but everybody everywhere where it is legalized is exposed in moderate amounts throughout their whole lives, then we should primarily fund fewer long term studies rather than dozens of short term ones. Your counter argument was that "no, that is not at all how the decision making process goes." That is irrelevant, my point was how it should go.

If we have only limited resources, which is true, we should use it to make studies that explore the real risks related to the chemical. One short term study is less expensive than one long term study. Maybe that is still true for two or three studies. But numerous 90 days studies have been cited to tell that roundup is not toxic. Sure it would have been possible to fund fewer of them, and direct some resources to, say, two 1-year-long studies, to see if something appears in long term.

If the symptoms take long time to develop, then even a high dose does not suffice, if we only follow a short time. A fortiori if a long, moderate exposure is required for the effects to be produced. As Lynn Goldman said about these 90-days feeding studies, "they don't tell you is whether there's a toxin that works very slowly. If it's a toxin that kills you that's one thing, but what if it causes neurological damage, something that is more like Parkinson's?"

In the case of roundup, it is obvious that any risks that there are for humans, are related to exposure to it in long time. People who have born in USA during last decade will have been exposed to Roundup constantly during their whole life. Because the Roundup Ready crops swiftly take over the whole market, replacing other crops, the political decision is so important that it is easily worth an extra attention relative to a chemical that would only play a minor role or that we could easily just stop using.

Toxicity is, of course, only one aspect of it and I do not expect everything to be solved by a long term study of carcinogenesis and other possible toxicity. And maybe a long-term feeding study is not the right way to go, but in some form a long term low- or moderate exposure study should be conducted. However, if, when deciding what research is funded, the actual use of the chemical is not taken into account, as you suggest, and we thus fund a lot of essentially similar of studies that bear little relevance to the actual use of the chemical and do not tackle some clear, potential problems with its use, and as a result we cannot conduct the ones that could decisively tackle those problems, then that is money wasted. Such decision making process does not appear as optimal. Far from it.

Now, back to my promise that I actually can partially answer my original question myself now. It seems I'm not the only one who considers that a better long-term study would be worthwhile conducting, as I happened to find out that both EU and France launched appeals for long-term research precisely 2012, in aftermath of the Séralini case. See the attached link.

They are funding precisely life long experiments on animals, in order to follow what effects GMO's, here particularly mentioned the Roundup*, could have in life-long exposure:

"Ces propositions pourront porter sur la surveillance d’animaux d’élevages dont la durée de vie commerciale devra être suffisamment longue pour pouvoir juger des effets sanitaires à long terme de la consommation d’aliments contenant des PGM."

*Commission is funding a 2-year study on toxicity of NK603: "Au niveau européen, la Commission (DG Santé et Consommateurs) a décidé de financer une étude de toxicité à long terme sur le maïs NK603, sur une durée de deux ans.

So, a couple of long term studies is about to be conducted, and one of them on Roundup Ready corn.

http://www.developpement-durable.gouv.fr/IMG/pdf/APR__Risk_OGM_rel_pbch_pbj_rs2.pdf

Hi! I am not sure whether you heard of 2 EU projects called GRACE and G-TwYST. Many questions you discussed are addressed there (http://www.grace-fp7.eu/content/eu-research-project-grace-publishes-first-study-findings as well as http://www.g-twyst.eu/). Best regards!

Thank you Roland, I had not heard of this, but I guess G-Twyst and the French GMO90+ have to do with the projects mentioned in the previous post. Thank you, and thanks to everybody else, too, who has contributed to this question. Concerning my original question, at least, I think it has been quite well answered. Now just to wait for the results from these researches.