Just a heads-up – this is a pretty in-depth investigation post. We touch on portions of 9 different EPA documents, discuss study results, look at documents within documents… it is a fun-filled carnival of glyphosate, Monsanto and EPA chaos. As an alternative, feel free to check out GMO Truth Episode #7, if you’d prefer to watch and listen rather than read.
Either way, TRUTH JUST AHEAD, so if you’re interested in learning the facts on what happened between Monsanto and the EPA regarding glyphosate, at a time that was key to the launch of Roundup Ready GMOs, then please throw caution to the wind and read on. Come on, I said it’s a carnival right? Who doesn’t love a carnival!??
So in our last investigation, we documented 5 potential problems with glyphosate, a path we were sent down courtesy of the International Agency for Research on Cancer (IARC) of the World Health Organization, who declared the substance “probably carcinogenic to humans” back in March 2015.
That kicked off a whole new research path for us here at Walk a Mile, and we went all the way back to 1974, when Roundup first hit the market. But nothing in that investigation really delved into this whole “probable carcinogen” hypothesis that the World Health Organization put out there, which is what we’re covering today.
Blind Renal Tubular Adenoma-Having Mice
It seems a little much to think that the EPA would be duped into letting a probable carcinogen slip out in such mass quantity again – that’s a big part of why they were created in the first place. Unfortunately, there is reason for concern. Back in March of 1985, an EPA committee classified glyphosate as a Group C oncogen, “based on an increased evidence of renal tubular adenomas in male mice.” So 11 years after stores first started selling Roundup, its active ingredient suddenly came under fire by the EPA. That’s where we begin.
Before we dive in here though, it’s worth noting that the EPA’s final classification on glyphosate was given in 1991. There were only a handful of studies used to evaluate the oncogenicity (tumor forming) and carcinogenicity of glyphosate. And those studies, at least in their entirety, are still hidden from public view, despite being over 30 years old at this point. There were 3 key studies, two in rats, and one in mice. We’re focusing on the mouse study here and will cover the rat studies in our next investigation. Also, this whole oncogen vs. carcinogen ambiguity will be cleared up in the next post, thanks to an EPA resource now assisting us.
The mouse study took place in July of 1983, and in a February 1984 review of that study, the EPA determined that Glyphosate was oncogenic, as it produced a rare tumor in a dose-related manner. A risk assessment was then required by the EPA’s Toxicology Branch, standard protocol. They saw rare tumors show up in a study, and the tumors looked dose-related, so yes, they called for a risk assessment to further evaluate. After all, glyphosate was already in use for a decade at this point.
Fast forward an entire year later, however, and there was no real movement, other than Monsanto trying to contest the EPA’s findings. In an EPA document from February of 1985, Monsanto claimed the rare tumors were unrelated to treatment. To support their argument, they provided historical data on mice from Bio/Dynamics, the lab they paid to do the mouse study in the first place, plus data from two other laboratories as well. In their plea to the EPA, they also suggested that the results may have included some false positives – some tumors that weren’t really tumors.
Let the Battle Begin…
As you might guess, that didn’t sit too well with the EPA. In fact, for close to 3 pages, an EPA statistician goes to great lengths in evaluating Monsanto’s argument, and concludes the following:
Under such circumstances a prudent person would reject the Monsanto assumption that Glyphosate dosing has no effect on kidney tumor production. Another way of saying this is that if Glyphosate were truly unrelated to kidney tumor production we would expect to see 4 or more tumors in less than 1 out of 100 experiments of the type sponsored by Monsanto. Thus, glyphosate is suspect.
In response to the false positives assertion, he gives multiple reasons for rejecting the idea, but states this specifically in his final summary:
False positive results are less likely to occur with rare tumors. And the tumors in question are rare. Viewpoint is a key issue. Our viewpoint is one of protecting the public health when we see suspicious data. It is not our job to protect registrants from false positives. We sympathize with the registrants problem; but they will have to demonstrate that this positive result is false. Finally, we mention that none of the tumors occurred in the control or low dose groups. Instead there was one at 5000 ppm and 3 at the 30000 ppm dose level. This together with the previous comments make it likely that there is a dose-tumor relationship for Glyphosate.
For those of you keeping score at home, that’s EPA 2, Monsanto 0, but it’s very early in the game here…
A month later, on March 4th 1985, an extensive EPA memo surfaced, with a ton of information, including two other disturbing studies, but then a bunch of other studies quickly referenced that were “supportive of non-oncogenic potential of Glyphosate.” It provided little help and contained mixed data, but the conclusion, however, was as follows:
It was the consensus of the group that the renal tubular adenomas were related to compound administration, since their frequency was NOT consistent with the historical controls and there is a trend indicating dose dependency.
Then under “Other Considerations”, they also state this:
The panel believes that additional sectioning of new blocks of male kidneys might help in the interpretation of the study results.
And at the very end of the document, here we go:
In accordance with EPA proposed guidelines, the panel has classified Glyphosate as a Category C oncogen.
Essentially, the mouse study raised enough of a red flag within the EPA to warrant the C classification, in 1985, although that specifically said oncogen, not carcinogen, which we will clarify in the next GMO Investigation post. What would really be nice, however, would be to see the actual study in question, which again, we can’t, because it’s still on lockdown after 30+ years — even though glyphosate came of patent 15 years ago!
Something Doesn’t Add Up
So a month later, April 1985, another EPA memo surfaced, which was strangely difficult to find online, and it contains the most details available for the mouse study. It does reveal kind of a mess in some of the other data. The control groups were all over the place for some of the lesions reported, which you can see on page 4. In fact, one of the male control groups showed 15/49 mice affected, while the high dose group showed the least affected of the 4 groups, with only 7 of the 50 mice affected. Meanwhile, in the female control group of the same category, there was a clear and steady progression from control (0) all the way up to the high dose (9).
And another weird thing… there are places in here where the numbers don’t match. I don’t know if it’s just a timing issue, perhaps related to mortality, where maybe one of the mice died so the group count was reduced by the time a specific condition was accounted for… I don’t know, but it makes you wonder if someone wasn’t trying to manipulate data here. Not accusing anyone, just saying something doesn’t seem quite right.
Bottom line though, the rare tumor in 3 mice from the high dose group meant something to the EPA, and here they also note that there were decreased body weights in male and female mice in the high dose as well, which they considered compound-related. So to them, there was some significance that needed addressing.
It’s Not a Tumor
Fast forward 8 months later to our 5th EPA document, and here’s where things really start to get interesting. Remember, the EPA suggested that additional sectioning of new blocks of the male kidneys might help in interpreting the study results more accurately. What they found was a small, localized change in one kidney of the control group. Monsanto’s quoted response was this:
The pathology staff at Bio/dynamics and I (Dr. McConnell) reviewed the lesion and concur that it may be representative of a developing tumor.
So the EPA’s Dr. Louis Kasza did his own microscopic examination, and this was his conclusion:
With regard to the questionable male control kidney, it is my opinion that the presence of a tumor can not definitely be established. My interpretation is similar to the conclusion of Bio/dynamics’ pathology staff and Dr. McConnell, that the lesion ‘may be’ a proliferative change having the potential to lead to the development of a frank tumor. But as the tissue can be seen under the microscope as a small well-demarcated focal cell aggregate morphologically different from the healthy looking surrounding kidney tissue, this morphological alteration does not represent a pathophysiologically significant change.
Is it a tumor, or isn’t it? That document sort of kicks off the war, because this is a huge question. If there is indeed the exact same tumor in the control group, that damages the statistical significance of the 3 tumors found in the high dose group – something Monsanto would obviously want to see.
Just over a week later, on December 12th, the EPA announced the results of all the additional evaluations on re-cut kidney sections, and they found no additional tumors present. But the re-evaluation also found 4 more mice who had chronic interstitial nephritis, a kidney disease that eventually leads to kidney failure. There were 3 times the number of mice affected in the high dose group as there were in the control group, so at the end of this document, it’s noted that the issue was compound-related. At this point, everything’s kind of backfiring on Monsanto, and they really want to see the control group tumor affirmed.
Regardless however, the obvious answer here was to just re-do the study (which would unfortunately mean another 2+ years in limbo). But as you’ll see, Monsanto adamantly did not want to go that route.
Moving to Group D
Two months later, in February of 1986, the EPA asked for assistance from the FIFRA SAP (Federal Insecticide, Fungicide, and Rodenticide Act Scientific Advisory Panel).
On page 5, there is a long paragraph that might just make your head explode. First, they state that the control group tumor should be counted, and then they say that, when the statistical analysis is age-adjusted, “no oncogenic effect of Glyphosate is demonstrated using concurrent controls.” They then propose that Glyphosate be moved into Group D, meaning not classified, that there’s not enough evidence either way at this point, and that there should be a data call-in for further studies in rats and/or mice to clarify unresolved questions.
So this is the 2nd pivotal moment, where we now have a panel stating that glyhposate should indeed not be kept in Group C. They also mention historical control data, and state that the historical control data could actually give reason for concern, but that they feel using the concurrent control data is more prudent. Remember that, because we’ll come back to it in a moment…
This all sounds fair enough when you think about it, I mean 3 out of 50 mice had tumors. That’s still only 6%. And their suggestion for doing more studies and/or repeating the mouse study, well, when we’re talking about something potentially causing cancer, it’s kind of important to follow through there. Now Monsanto’s heard the call for that again, from a second group, that more data is needed.
Instead, however, Monsanto then asked several doctors to respond directly to them regarding the mouse study (which if you want to talk corruption potential, this little detour certainly holds as much as anything), and they then submitted those responses to the EPA in a document from March of 1986. Four doctors refuted the EPA’s take on the study, and then an opinion issued by the PWG (Pathology Working Group) also echoed the same sentiment. And the EPA’s response was simple, “This information will be submitted to the TB Ad Hoc Committee for evaluation to determine if concurrence is possible.”
Slower Than GMO Molasses in January
Again, conflicting opinions, and we’re talking about concerns over cancer potential, so let’s re-do the study. Or – let’s come up with a couple of better studies. Let’s do something. Instead… nothing. For nearly 3 years… nothing. The final significant document I could dig up was from June of 1989. In the “Background” section of that document, it states:
On November 10, 1988, a meeting was held between EPA staff and representatives of Monsanto to discuss the Agency’s requirement that the mouse oncogenicity study with glyphosate be repeated.
We can safely assume that request was floating out there the entire time, because it was mentioned in multiple documents years earlier. But still, no compliance from Monsanto. Then, in this late 1988 meeting, they again bring up historical control data. Apparently there were new people working in the EPA at this point, because they don’t seem to remember that this was all discussed already, back in February of 1985 (see document #2 for the details).
Here’s the EPA’s summary:
Monsanto stated that there were historical control data demonstrating that the incidence of mouse kidney neoplasms ranged from 0 to 7.1 percent. This incidence exceeded the incidence of 6 percent from the high-dose group in the glyphosate study. Monsanto indicated that a repeat mouse oncogenicity study was not required. EPA stated that the historical control data should be submitted in order to re-evaluate the Agency’s position on the repeat study.
Before I respond to that, let’s look at the actual letter Monsanto wrote to the EPA in October of 1988. Here’s a quote:
Monsanto’s position regarding this requirement has been, and remains to be, that there is no scientific justification for repeating the mouse oncogenicity study.
Based upon the above discussion, Monsanto contends that there is no relevant scientific or regulatory justification for repeating the glyphosate mouse oncogenicity study.
Respect My Authority!
Wow, can you say non-compliance? Does anyone else find it disturbing that Monsanto could have a product on the market for nearly 15 years at this point, and for 5 years it was flagged as a potential cancer concern by the EPA, yet there was little to no action taken? I mean Monsanto can just avoid re-testing for 4 or 5 years? And then they think they can just throw the same disallowed data back at the EPA from years prior, and expect a free pass? Why wouldn’t there have been a mandate, and I’m guessing the EPA just doesn’t have this kind of power, but I’d like to see a mandate that says hey, you can get the study re-done, or get better studies done, OR, we can take glyphosate off the market as potentially unsafe until you do! Meanwhile, Monsanto is working hard on the Roundup Ready GMO crops that will take glyphosate use to record-breaking heights.
In addition to all this, let’s have the EPA supervise the study, but Monsanto pay for it. That’s how it should ethically work in the first place, but it doesn’t. Instead, we see Monsanto refuse to supply the required safety studies and then go back once again to this historical data mess.
And yes, we’re talking about some of the same historical control data that two different documents mention as not supporting Monsanto’s position whatsoever, with the last one going so far as to say it actually raised concern about the study they were trying to refute!
But since they were so adamant about this control data exonerating them from any further studies… and basically anointing glyphosate as this super-safe chemical, let’s take a look at this historical data, which is on p.4 of that June 1989 document.
Monsanto Puts a Spin on Historical Data
This is historical data from the same lab, and for the same type of mice used in the study in question. There are 815 control mice in this data, and out of that 815, 3 had tumors. The study they are trying to refute, showed 3 tumors out of 50 mice. So why would you even submit this kind of data?
Apparently they were trying for the smoke and mirrors approach. Instead of looking at the totals, Monsanto completely cherry picks the data (which they LOVE to accuse others of doing by the way, as they just did to the World Health Organization), so forgoing the actual 3 out of 815 number, they suggest to the EPA that the occurrence in these historical controls ranges from 0 to 3.3 percent. Well, in 5 out of the 7 studies with control data, the percentage was a big fat ZERO. Zero control mice developed the tumor in 5 of the 7 studies!
Update: July 15, 2017 — We originally referenced 14 control studies here, but the EPA and Monsanto were only looking at 7, and it’s not explained why in any document. As discussed both above and below this update, we now only reference the 7 studies and the 815 mice.
However, if you look at page 4 on the PDF we’ve linked to HERE, the data actually does show 14 studies, but one of them was clearly irrelevant because kidneys in those studies were never examined.
That still leaves 13 studies, with a total of 1,371 mice, and only 3 total mice exhibiting the rare tumor, or 2/10ths of one percent. We were unable to find any rationale for why they narrowed the control mice data down to this lesser number, which obviously worked in Monsanto’s favor.
In one study, 2 out of 60 mice developed tumors, so there’s your 3.3 percent, one study out of 7. In another study, there was 1 out of 54 mice, and that’s 1.9%. But when you add up all the controls, 3 out of the 815 total, that’s less than 4 tenths of one percent. In other words, congratulations Monsanto, you just re-submitted data that again, actually supports the study you’re trying so hard to refute.
But their big one was on page 6, from Hazelton Labs, and I think this might have been new data, although I’m not sure because the EPA documents don’t tell us. Regardless, if you look at this amazingly un-detailed data set that is supposed to be the end-all-be-all for the mouse controversy, you will simply shake your head. As you might expect, here’s how the EPA responded:
Apparently, this historical control data, which range from 0 to 7.1 percent, are the historical control data cited by Monsanto… Dr. Ward (of Monsanto) indicated that Hazelton was unable to provide any additional details (dates of study, supplier, pathologists, etc.) about these particular historical controls. Therefore, in light of this telephone communication, TB concludes that these particular historical controls from F1 male mice cannot be used to substantiate the Monsanto position.
Here’s the final conclusion of this document:
TB concludes that a repeat of the mouse oncogenicity study is not required at this time. After the results of the new 2-year rat chronic toxicity and oncogenicity study are reviewed, TB will consider whether the repeat of the mouse oncogenicity study is required.
Now in defense of Monsanto, it looks like they started that rat study in late 1988, so it was probably already underway when these letters were flying back and forth. That doesn’t excuse the years of push-back, however, and what essentially looks like bullying of the EPA. But regardless, it meant that the 1990 rat study, using the same rats vilified for long-term use in the famous Seralini study by the way, would likely be the final evaluation piece on this whole “glyphosate – is it a carcinogen or isn’t it” question.
We will dive into that into our next investigation post, but for today, this was all about a mouse study that, time will tell, was either completely irrelevant (meaning that those 3 kidney tumors really weren’t dose-related), or it will go down in history as the first ignored warning sign of serious glyphosate problems to come…