Carcinogen or not a carcinogen? A tale of two WHO Agencies, and the importance of evaluating study quality and human relevance
How is it possible that two World Health Organization (WHO) agencies could evaluate the same chemical’s potential to cause cancer and come to seemingly opposite conclusions?
Dr. David Eastmond explored this question in a presentation at the Summer Toxicology Forum meeting comparing the approaches taken by the International Agency for Research on Cancer (IARC) and the Joint Meeting on Pesticide Residues (JMPR), in their evaluations of the herbicide glyphosate. You can view the slides from his presentation here.
IARC: Glyphosate is “Probably Carcinogenic to Humans”
The objective of IARC, the cancer agency of the WHO, is to promote international collaboration in cancer research. The IARC Monographs on the Evaluation of Carcinogenic Risks to Humans identify environmental factors that have the potential to cause cancer in humans. Since 1971, more than 900 agents have been evaluated.
In March 2015, IARC published a final evaluation and rationale relating to the carcinogenicity of glyphosate. IARC classified glyphosate as probably carcinogenic to humans (Group 2A). This was based on IARC’s finding of limited evidence of carcinogenicity in humans for non-Hodgkin lymphoma. In addition, IARC noted that there is also convincing evidence that glyphosate can cause cancer in laboratory animals, and that glyphosate caused DNA and chromosomal damage in human cells.
JMPR: Glyphosate is Unlikely to Pose a Carcinogenic Risk to Humans Exposed via Diet
JMPR is an expert ad hoc body administered jointly by the Food and Agriculture Organization of the United Nations and WHO for the purpose of harmonizing the requirement and the risk assessment on the pesticide residues. JMPR provides advice on the acceptable levels of pesticide residues in food moving in international trade. In May 2016, JMPR re-evaluated the risk posed to human health of consuming residues of glyphosate in food. JMPR determined that glyphosate is unlikely to pose a carcinogenic risk to humans exposed via the diet.
Study Quality and Human Relevance
While the findings by IARC and JMPR may seem contradictory, they actually are not. As stated by JMPR: “IARC reviews published studies to identify potential cancer hazards. It does not estimate the level of ‘risk’ to the population associated with exposure to the hazard. In contrast, JMPR reviews both published and unpublished studies to assess the level of health risk to consumers associated with dietary exposure to pesticide residues in food.”
The JMPR finding of an unlikely risk results from the fact that JMPR gave greater weight to the studies most relevant to humans and to oral exposure. JMPR gave less weight to studies on species far removed from humans, like plants, earthworms, fish and caiman, and routes other than oral exposures. This review led to the conclusion that glyphosate is unlikely to be genotoxic at anticipated dietary exposures.
Dr. Eastmond was on the JMPR panel and at the Toxicology Forum meeting and provided insights into why IARC and JMPR reached these differing conclusions, including:
- While IARC focuses only on the published studies, the JMPR evaluation was able to include many guideline studies, which are not published. JMPR evaluated just over 246 studies, compared to IARC’s evaluation of ~113.
- JMPR gives more weight to higher quality studies, in vivo studies over in vitro studies, and more weight to mammals than other species. JMPR also gave higher weight to relevant exposure routes, in this case the oral route. While these seem like minor points, Dr. Eastmond’s presentation shows the importance of digging into the study details and looking at more than results. The JMPR group looked at the methods and the quality of the studies. They articulated their weighting criteria in advance of the evaluation and benchmarked each of the studies to these clear criteria.
- When it came to genotoxicity, JMPR gave more weight to serious endpoints like mutations and chromosome alterations than they did to less serious or transient endpoints (e.g., single strand breaks).
- While IARC did not have original reports for some animal studies and often worked with re-evaluations of the data, JMPR looked at the data in original reports.
The differences in some cases were striking. While IARC found 80 percent of the human biomonitoring studies to be positive, JPRM considered 0 percent to be positive with most being evaluated as equivocal. For studies that looked at bacteria, where IARC found 67 percent to be positive, JMPR found only 13 percent to be positive. The findings were similarly dissimilar for other types of studies. When looking at studies that IARC said were highly significant, JMPR found inconsistencies and concerns with the studies, leading the results to be deemed only equivocal.
Dr. Eastmond acknowledged that the JMPR review required considerable expertise and scientific judgment. For the glyphosate review, JMPR was able to conclude that the overall weight of evidence, for oral exposure, did not support genotoxicity in mammals, despite IARC’s finding of strong evidence of genotoxicity.
The differences in outcomes were largely due to the information available for evaluation, the rigor and weights applied to the different types of studies, and the criteria used to incorporate the important issue of human relevance. The JMPR report is not yet final so it is difficult to evaluate the full extent of the differences. However, it is easy to see that the strength of the evidence matters: having clear criteria and giving more weight to data that are most relevant to exposure routes in humans, is going to create differing results.
This example with two agencies with two different reviews reaching different conclusions about the same chemical is not just confusing for chemical specialists and toxicologists like Dr. Eastmond. Confusing results – especially when it comes to whether or not a substance causes cancer in people – creates confusion for the public.
IARC and JMPR aren’t the only agencies to look at the same chemical and publish seemingly contradictory results. Varying standards in evaluating quality of evidence have led government bodies such as the National Toxicology Program, the U.S. Environmental Protection Agency’s (EPA) IRIS program, EPA’s pesticides program, the U.S. Food and Drug Administration, and California’s Proposition 65 agency to come to different conclusions regarding a chemical’s health effects.
Examples like these emphasize why the new requirements in the Lautenberg Chemical Safety Act (LCSA) calling for weight of the scientific evidence evaluations will truly help to increase the quality of chemical assessments done by EPA – and give consumers greater confidence about the safety of chemicals in their products.
