Over the last several years, a group of scientists have put forward the so-called “Key Characteristics of Carcinogens” (KCCs) approach for collecting, organizing and evaluating mechanistic study results to classify the carcinogenic potential of chemicals. The KCCs approach was developed by a Working Group of the International Agency for Research on Cancer (IARC) and since has been used in recent IARC Monographs. The bottom line (explained in detail below) is that IARC’s approach to cancer classification is different than US EPA’s, the KCCs have been proven not to work using EPA’s approach, and TSCA has scientific standards.
Whether or not a chemical is ultimately classified as a carcinogen, as well as the language used to describe carcinogenic potential, depend to an appreciable extent on the guidelines (or ‘instructions’) that are followed. While IARC’s guidelines (stated in the Preamble to its Monographs) and US EPA’s 2005 Guidelines for Carcinogen Risk Assessment are different, both have used, or propose to use, KCC mechanistic data to inform their determinations of the potential of a chemical to cause cancer.
In order to determine how well the KCCs actually predict cancer classifications made by EPA, an extensive analysis was carried out using a set of explicit and objective scientific methods. This was done by comparing data from effects of chemicals in cells (bioactivity data) in EPA’s ToxCast assays that mapped to KCCs to EPA’s previously derived cancer classifications for the same chemicals. The results? Assays measuring bioactivity of chemicals according to the so-called Key Characteristics of Carcinogens were no better than chance alone in predicting cancer classifications. What does this mean? It means that when put up against the rigorous standards for causality in EPA’s 2005 Guidelines, KCCs performed no better than a coin toss.
TSCA’s Section 26 specifies EPA must use ‘scientific information, technical procedures, measures, methods, protocols, methodologies, or models, employed in a manner consistent with the best available science’. Does EPA think that predictive performance no better than a coin flip meets these standards? Even if the use of KCCs were limited to a way to collect, organize, and evaluate mechanistic data to reduce the use of vertebrate animals, TSCA Section 4 specifies EPA must facilitate the use of ‘scientifically valid test methods and strategies that reduce or replace the use of vertebrate animals while providing information of equivalent or better scientific quality and relevance’. Does EPA think that KCCs proven not to work in EPA’s own cancer classification framework provide information of ‘equivalent or better scientific quality’?
Based on EPA’s actions, the answer to the above questions appears to be ‘Yes’. Proof that KCCs don’t work hasn’t stopped KCCs from gaining a firm foothold at the EPA. Recently, the EPA Integrated Risk Information System (IRS) Program’s draft ORD Staff Handbook for Developing IRIS Assessments went so far as to assert, without any evidence to back this claim, that “Instead of identifying and organizing the mechanistic evidence according to predefined MOAs [modes of action], a more objective approach is to categorize the literature from a broad search for chemical-specific mechanistic information according to commonly recognized properties of carcinogens.” The KCCs are also being used at the State level. California’s Office of Environmental Health Hazard Assessment (OEHHA) has begun to insert the KCCs approach into Proposition 65.
EPA and OEHHA have continued to pursue this flawed KCC approach, even after additional studies by Bus (2017) and Goodman and Lynch (2017) have raised concern with using the KCCs as a tool for assessing cancer hazards. Smith et al (2021) recently affirmed that the KCCs are too broad and non-specific for evaluating the potential cancer hazard of chemicals. The warnings from scientists are getting ever-louder as more studies support the seminal analysis showing that KCCs simply don’t work. Putting science first and applying a more balanced data driven framework are key. Box-checking, while convenient isn’t science. It’s bureaucracy and TSCA’s scientific standards are higher than that. The facts are the facts – EPA and IARC need to implement a more rigorous mode-of-action pathway-based framework to organize, evaluate, and integrate mechanistic evidence with animal toxicity, epidemiological investigations, and knowledge of exposure and dosimetry to evaluate potential carcinogenic hazards and risks to humans.
