The International Agency for Research on Cancer (IARC) Monographs Program evaluates the potential for substances and behaviors, or “agents,” to cause cancer. Each assessment is performed by a separate Working Group comprised of various experts declared “free” from conflicts of interest. While IARC has established general principles for the evaluation of evidence and provides participants with monograph-specific instructions that are not released publicly, these groups are largely free to take their own approach to the monograph evaluation process. Essentially, there are no standard methodologies or detailed procedures dictating how Working Groups evaluate scientific evidence.
A new paper, “Improving the International Agency for Research on Cancer’s consideration of mechanistic evidence,”* highlights why this lack of standard and clear guidance is significant and offers suggestions for improvement. Writing in Toxicology and Applied Pharmacology, Julie Goodman, Ph.D., and Heather Lynch, M.P.H., specifically examine IARC’s recently developed framework for evaluating mechanistic evidence, or how carcinogens cause cancer.
In the following Q&A, Dr. Goodman and Ms. Lynch explain the paper’s approach and findings.
Q: Historically, IARC monographs have focused on identifying cancer hazards regardless of the underlying mechanism. What is mechanistic evidence and why is it important to consider mechanistic evidence as part of IARC’s assessments of cancer hazards?
A: Some studies evaluate whether agents are associated with or cause cancer. Mechanistic studies focus on how chemicals act or might act; that is, the way that chemicals might cause cancer. This is important because if a chemical causes an effect that has been shown to lead to cancer in people, then this provides evidence that the chemical itself could cause cancer in people. In contrast, a chemical may cause cancer in animals, but the mechanism by which it causes cancer does not operate in people, so it is not likely to be a human carcinogen. This is why it is important to study mechanistic evidence.
Q: Though IARC has created this basic framework, identifying 10 key characteristics common to cancer cells, your paper points out that some of these 10 characteristics are also shared by non-carcinogenic agents. Could you explain what these 10 key characteristics are and why the current application of this framework is concerning?
A: IARC defines these as characteristics shared by many carcinogenic substances, and they correspond to mechanisms by which chemicals may influence or directly initiate cancer. The 10 characteristics are that an agent 1) is electrophilic or can be metabolically activated, 2) can directly damage DNA, 3) can alter DNA repair or cause a high frequency of mutations (genomic instability), 4) can alter gene expression, 5) can induce oxidative stress, 6) can induce chronic inflammation, 7) is immunosuppressive, 8) can modulate receptor-mediated effects (i.e., can activate or inactivate receptors or affect the molecules that bind to them), 9) can cause cells to have an infinite lifespan (immortalization), and 10) can alter cell proliferation, cell death, or nutrient supply.
The IARC framework has no specific guidance regarding how the presence of a characteristic should be interpreted in the context of other evidence. Considering the presence of one or more characteristics as evidence of carcinogenic potential, without assessing whether there is evidence that the agent causes cancer in animal experiments or in people, could lead to a misinterpretation of mechanistic evidence, and possibly to labeling an agent as carcinogenic when it is not.
Q: How does IARC’s approach differ from how other public health agencies and regulatory bodies weigh and interpret mechanistic evidence?
A: One critical difference is that IARC takes a strength-of-evidence approach to its hazard classifications, while many other regulatory agents use weight-of-evidence (WOE) approaches. In a WOE approach, all evidence is considered concurrently, and several well-conducted studies with consistent results are required to formulate a causal classification. In contrast, strength-of-evidence approaches, such as IARC’s, may base a classification primarily on a single study or a few studies, without fully considering study quality or the human relevance of the studies.
Q: What guidance could IARC offer Working Group participants to ensure each monograph is developed taking a similar approach to the interpretation of mechanistic evidence? How should mechanistic evidence be integrated into the IARC monograph process?
A: IARC should provide clear, explicit guidance regarding how to consider the entire body of mechanistic evidence, including study strengths and limitations, and how they impact the interpretation of results. IARC could do this by adapting an existing framework that addresses study quality and human relevance. For example, IARC could follow the Mode of Action/Human Relevance (MOA/HR) framework, which provides a systematic, clear method for conducting a thorough analysis of mechanistic evidence within a larger WOE assessment to determine whether any observed mechanisms could occur in people.
Q: Why is clear and transparent guidance on how Working Groups should consider mechanistic evidence important? Could you give an example of a substance that an IARC monograph classified as a possible, probable, or known human carcinogen based on inappropriately used mechanistic evidence?
A: Without clear and transparent guidance, each Working Group may approach and interpret mechanistic information differently, leading to inconsistent IARC evaluations.
In 2016, IARC released a monograph that included the organophosphate pesticide diazinon. The Working Group determined that while the animal and human data were limited, the strong evidence that diazinon can operate through two key characteristics of carcinogens provided support for the classification that diazinon is probably carcinogenic to humans. However, the Working Group did not fully consider factors that affect the interpretation of mechanistic evidence, such as study quality and whether the study outcomes were relevant to humans. Also, the monograph did not discuss how the lack of positive findings in certain mechanistic studies impacted the interpretation of positive findings in other studies, nor how the available mechanistic evidence was considered in light of other types of studies.
Q: IARC’s cancer determinations have an impact on U.S. policy, such as California’s chemical labeling law known as Proposition 65, and are often widely covered by the media around the globe. Based on your knowledge of IARC’s Monographs Program, what advice would you give non-scientists who are trying to determine the relevance of IARC’s cancer classification?
A: IARC ranks agents based on their potential hazard, but it’s important to put this in perspective. One must look at how an agent is used, and what potential exposures could be, to assess the real-world risk.
IARC does not evaluate the likelihood of risk associated with any agent or the potential of any individual or group of people to actually develop cancer from exposure. Even agents with high hazards may not pose risks to people.
You can read the paper, “Improving the International Agency for Research on Cancer’s consideration of mechanistic evidence,” online here.
*The authors are employed by Gradient, a private environmental consulting company. The authors have sole responsibility for the writing and contents of this paper. This paper represents the professional opinions of the authors and not necessarily those of the American Chemistry Council, which provided funding for this paper.