How did the US EPA and IARC reach diametrically opposed conclusions on the genotoxicity of glyphosate-based herbicides?

The US EPA considers glyphosate as “not likely to be carcinogenic to humans.” The International Agency for Research on Cancer (IARC) has classified glyphosate as “probably carcinogenic to humans (Group 2A).” EPA asserts that there is no convincing evidence that “glyphosate induces mutations in vivo via the oral route.” IARC concludes there is “strong evidence” that exposure to glyphosate is genotoxic through at least two mechanisms known to be associated with human carcinogens (DNA damage, oxidative stress). Why and how did EPA and IARC reach such different conclusions? A total of 52 genotoxicity assays done by registrants were cited by the EPA in its 2016 evaluation of technical glyphosate, and another 52 assays appeared in the public literature. Of these, one regulatory assay (2%) and 35 published assays (67%) reported positive evidence of a genotoxic response. In the case of formulated, glyphosate-based herbicides (GBHs), 43 regulatory assays were cited by EPA, plus 65 assays published in peer-reviewed journals. Of these, none of the regulatory, and 49 published assays (75%) reported evidence of a genotoxic response following exposure to a GBH. IARC considered a total of 118 genotoxicity assays in six core tables on glyphosate technical, GBHs, and aminomethylphosphonic acid (AMPA), glyphosate’s primary metabolite. EPA’s analysis encompassed 51 of these 118 assays (43%). In addition, IARC analyzed another 81 assays exploring other possible genotoxic mechanisms (mostly related to sex hormones and oxidative stress), of which 62 (77%) reported positive results. IARC placed considerable weight on three positive GBH studies in exposed human populations, whereas EPA placed little or no weight on them. EPA and IARC reached diametrically opposed conclusions on glyphosate genotoxicity for three primary reasons: (1) in the core tables compiled by EPA and IARC, the EPA relied mostly on registrant-commissioned, unpublished regulatory studies, 99% of which were negative, while IARC relied mostly on peer-reviewed studies of which 70% were positive (83 of 118); (2) EPA’s evaluation was largely based on data from studies on technical glyphosate, whereas IARC’s review placed heavy weight on the results of formulated GBH and AMPA assays; (3) EPA’s evaluation was focused on typical, general population dietary exposures assuming legal, food-crop uses, and did not take into account, nor address generally higher occupational exposures and risks. IARC’s assessment encompassed data from typical dietary, occupational, and elevated exposure scenarios. More research is needed on real-world exposures to the chemicals within formulated GBHs and the biological fate and consequences of such exposures.