On Monday 8th and Tuesday 9th of May we had a meeting with the head of the IARC-monographs Kurt Straif in Lyon to discuss our project research and how it could impact the practices of IARC. IARC is short for the International Agency for Research on Cancer.
In their monographs, a committee of the world leading experts will meet after months of preparation, and then for 10 days (and often nights), assess the evidence pertinent to whether a certain agent (for instance, coffee, red meat or radio frequencies) causes cancer. Basically, each agent is categorised into one of five different groups ranging from “causes cancer” to “probably not causing cancer”. One of the key practices of IARC is to consider evidence of mechanism alongside epidemiological studies in humans. This is necessary not only because there is virtually no randomised controlled trial available in cancer research, but also because exposing humans to agents that are under suspicion of causing cancer is clearly unethical.
Now, observational studies are subject to confounding, chance and bias. To rule out these possibilities, IARC devotes a subgroup of experts to assess the evidence of mechanisms. Furthermore, another subgroup reviews the evidence pertinent to animals. If an agent causes cancer in animals, this is already an indicator that it causes cancer in humans. The more similar the mechanisms in humans and in the particular animal species considered are, the more relevant the results in animals are to humans (here is also a role for evidence of mechanisms to play). Finally, the three subgroup results (mechanistic, animals and epidemiology in humans) are put together to arrive at an overall evaluation of the carcinogenicity of the agent. The preamble to the monographs gives detailed guideline how all of this should be done.
In this connection, some of our project members co-authored the “Evaluating Evidence of Mechanisms in Medicine: a handbook for practitioners” (see Jon’s blog for more information). This handbook aims to improve our understanding of how evidence of mechanisms is best considered in practice. And thus, we think that some of the insights developed in this document could supplement the preamble of the IARC-monographs. We discussed our handbook with Kurt and compared the methods proposed there with those advocated by IARC.
All in all, I think those two methods are very similar in highlighting the importance of evidence of mechanisms for hazard identification. However, while we are in broad agreement, there were also important differences between our two approaches. Some concerned the correct picture of causality and some concerned the most appropriate categories for classifying agents according to their cancer risk. During our discussions, however, I think these differences became smaller. We also agreed that there is need for more detailed guidance on how to best evaluate evidence of mechanisms.
Thanks to Kurt for his very insightful comments and for hosting us. We learned much in these two days and took a lot away to think about. We are looking forward to our next meeting, which we all agreed will not be too far in the future.
Words by Christian Wallman.