USA – The U.S. Food and Drug Administration (FDA) has released the Expanded Decision Tree (EDT), a structure-driven tool designed to support screening and toxicity evaluation of chemicals in food.
By examining a compound’s molecular structure and estimated toxicity, the EDT offers a consistent method for assessing chemical safety where direct testing data may be limited or unavailable.
The tool is part of the FDA’s New Approach Methods (NAMs) initiative, which seeks to incorporate technology, large datasets, and innovative risk models into regulatory processes.
The EDT builds on the original Cramer Decision Tree, a decades-old framework used to sort chemicals into broad toxicity classes using structural features.
Unlike its predecessor, the EDT applies a more detailed set of structure-based questions, enabling more refined classification into multiple toxicity categories. This allows regulators and scientists to assess chemical risk with greater precision.
Peer review and scientific oversight
In March 2024, the FDA submitted the EDT for independent peer review, in line with the Information Quality Act.
The review included scientific experts from academia, public health agencies, and international food safety bodies. The tool was revised based on that feedback and is now being shared publicly for technical use and comment.
Use in pre- and post-market evaluations
According to the FDA, the EDT will assist in both pre-market evaluations of new food substances and post-market reassessments of existing ones. It is expected to help the agency prioritize chemicals for further review, especially where historical approvals may lack current toxicological context.
The tool can be used to guide decisions about whether more data or testing is needed. It may also inform how the industry prepares safety documentation for new food contact materials or substances generally recognized as safe (GRAS).
Automation in progress
At present, the EDT is used manually and requires specialized expertise in organic chemistry and metabolic pathways.
To broaden access and improve efficiency, the FDA is developing an automated software version. Once complete, the automation is expected to reduce the time needed to classify a chemical from hours to minutes, while maintaining consistency and reproducibility.
Foundation in international toxicology data
The EDT database is built on toxicological, metabolic, and chemical data from authoritative global sources.
These include the Joint FAO/WHO Expert Committee on Food Additives, the U.S. Environmental Protection Agency (EPA), the European Chemicals Agency (ECHA), and the European Food Safety Authority (EFSA), among others.
The data span food ingredients, packaging materials, contaminants, and chemicals found in related sectors such as pharmaceuticals and cosmetics. Chemicals from the Munro database served as a starting point, with additional compounds and study data incorporated over time.
Threshold and classification tiers
The EDT estimates a chemical’s toxic potential by predicting its Threshold of Toxicological Concern (TTC), expressed in micrograms per kilogram of body weight per day (µg/kg bw/d). It also classifies chemicals into more categories than the original Cramer model, allowing more precise identification of substances that may warrant closer review.
However, the EDT’s scope does not currently include nanomaterials, non-hydrolyzable polymers, proteins, or certain metals such as arsenic, mercury, cadmium, and lead.
Tool to supplement, not replace, regulatory oversight
The FDA has emphasized that the EDT is not a substitute for pre-market approval processes required under the Federal Food, Drug, and Cosmetic Act. Instead, it is intended to serve as one of several tools used to inform regulatory decisions and identify potential risks in the food chemical landscape.
Stakeholders in the food industry may also choose to use the EDT to assess safety risks of novel ingredients or impurities before submitting them for regulatory review.
The release of the EDT aligns with the FDA’s stated goals of improving post-market chemical oversight and reducing unnecessary animal testing. By utilizing existing data and predictive modeling, the agency hopes to better prioritize substances of potential concern and modernize its risk assessment practices.
In the coming months, the FDA plans to release an informational video on how to use the EDT and will organize listening sessions with scientists, technical experts, and industry representatives. These sessions will inform further updates to the tool and offer a platform for feedback.
The agency has invited the scientific community and other interested parties to explore the tool and contribute insights that may enhance its future development.