A near kin to the GMO controversies that ITIF deconstructed earlier this year in a comprehensive guide for policymakers is the drumbeat of often specious criticism directed at some of the other essential inputs for innovative, commercial-scale agriculture. The latest case in point was an announcement last week from the International Agency for Research on Cancer (IARC) that one of the world’s most widely used weed killers, this time 2,4-D, is “possibly carcinogenic.” Following on the heels of a similar, highly suspect and widely criticized claim regarding glyphosate a few months ago, one would be justified in asking the question: Does the evidence really indicate 2,4-D, widely used for seven decades, has been causing cancer in humans, and requires we change the way we handle it? In fact, no, it doesn’t. Let’s take a closer look.
The announcement from IARC comes in the form of a brief report of barely 1,000 words, of which one-third (about one page of text) is devoted to 2,4-D. The details of the evidence considered and the reasoning applied are largely relegated to the monograph that will follow, after the indefensible headlines are in the rear view mirror. This is not a reputable approach to science, nor to policy.
There are several words of interest among the few shared with the world in the brief IARC announcement. It notes that, “In humans, 2,4-D is eliminated largely unchanged in the urine.” This is reassuring for a number of reasons. Obviously, it implies a dramatic reduction of exposure compared to compounds that are, for example, lipid soluble, which are therefore absorbed and sequestered in the body’s energy reserves, potentially hanging around for a long time, extending exposure. Also, being excreted “unchanged” means it is not metabolized, which means there are no potentially toxic breakdown products to wreak their own variety of havoc. Elimination via urine is one of the best possible outcomes.
IARC also states that, “The Working Group conducted a meta-analysis of 11 studies that showed no association of non-Hodgkin lymphoma with ever-exposure to 2,4-D…. The consensus of the Working Group was that there is inadequate evidence in humans for the carcinogenicity of 2,4-D, although a substantial minority considered that the evidence was limited.” [Emphasis added.]
Wait, what? The consensus of the group is that there is inadequate evidence in humans for the carcinogenicity of 2,4-D? But that’s not what the headline says!? What on earth is going on here?
The next paragraph goes on to reveal, “The Working Group concluded that there was limited evidence in experimental animals for the carcinogenicity of 2,4-D due to methodological concerns regarding the positive studies, although a substantial minority judged the evidence to be sufficient.” [Emphasis added.]
So there appear to be no data indicating 2,4-D is carcinogenic—zero, nada, zip, none. “A substantial minority considered that the evidence was limited,” according to the IARC announcement, but still agreed there is no evidence in humans for carcinogenicity. And while “there was limited evidence in experimental animals for the carcinogenicity of 2,4-D,” that evidence is itself suspect, due to “methodological concerns,” which is to say, problems with the design or execution of the tests.
So how does this translate into “possibly carcinogenic?” One could with justice infer from the report that a minority among the panelists were intent on reporting a positive finding, no matter what. There are zero data indicating human carcinogenicity and no clear or unambiguous suggestions of animal carcinogenicity, except in cases that involve methods of exposure (subcutaneous injection) that are simply irrelevant to the way humans are exposed to 2,4-D when using it to control weeds. And yet all that leads to a finding of “possibly carcinogenic?” The conclusion does not appear to follow, at all, from the premises.
Part of what’s going on here is admitted by IARC in its press release, which says, “The IARC Monographs Programme evaluates cancer hazards but not the risks associated with exposure.” Risk equals hazard multiplied by exposure, so in the absence of exposure there is no risk. If there is exposure, the type and extent may make a difference, though they would appear of no relevance if there is no mechanism of carcinogenesis, nor any reports, ever, anywhere, supporting a link in humans with cancer.
It is worth noting that no regulatory body in the world has found 2,4-D to be carcinogenic. The medical literature is conspicuously bereft of reports of cancers caused by 2,4-D exposure, despite its enormously wide use for the past 70 years. Regulatory agencies, unlike IARC, typically conduct risk assessments to examine the relationships between the presence and extent of exposure, and negative outcomes (“dose/response”). All of this is required to introduce any real-world relevance to what would otherwise be an ivory tower exercise of no practical import or value.
An industry task force notes that the “IARC cancer classification [is] at odds with health and safety regulators worldwide.” They go on to say that, “The studies that IARC considered have previously been reviewed by health and safety regulators around the world. Based on published, ongoing and continually updated scientific studies, health and safety authorities in more than 90 countries—including US Environmental Protection Agency, Health Canada, the European Food Safety Authority and the World Health Organization—continue to find that 2,4-D meets modern safety standards. In 2014, the US EPA stated: ‘…[B]ased on weight of evidence consideration of the available data, 2,4-D would be classified as ‘Not Likely to be Carcinogenic to Humans’.”
The most difficult issues with which regulatory agencies around the world must deal are how to assess risk accurately (exposure to a hazard) and how to manage risks appropriately. To stop at merely identifying a potential hazard, as IARC does, is not helpful to regulators, and an IARC spokesperson admits as much. The question, then, is what purpose is served by provoking concerns with classifications unsupported by data, and failing to provide regulators with something they might actually find to be helpful?