NAMs Beyond Pharma: Advancing Food Safety With Alternatives to Animal Testing

When a new snack food or ingredient appears in your local grocery store, your first thought is probably "looks tasty," not "is that safe to eat?"

That is because new foods and food ingredients are required to undergo extensive food safety testing and toxicology assessments to prove that they are safe for consumption before they ever reach our shelves.

Conventional food safety testing has primarily relied on animal testing to identify any hazards associated with long-term consumption of new food products. However, these animal tests have notable limitations, beyond even the general rising opposition to animal testing in society; these tests can be time-consuming, incur high costs, and may not have translational relevance to human health.

As science and technology have advanced, new methods to replace animal testing have risen in prominence within the drug and pharmaceutical sectors. Known collectively as New Approach Methodologies (NAMs), this broad assortment of techniques offers alternative ways to assess the safety and efficacy of drugs, biologics, and other products being investigated for human use. In silico techniques using computer modeling and artificial intelligence, as well as advances in organoids and organ-on-a-chip technologies, are all examples of NAMs that can reduce or replace the need for animal tests.

Taking a leaf from this book, food safety testing is also increasingly making use of NAMs for the same reasons—to reduce the use of animals and improve the relevance of testing to human health. To learn more about the use of NAMs in food safety testing, Technology Networks spoke with Dr. Alie de Boer, an associate professor in nutrition and food information at Maastricht University and founder of the Food Claims Research Centre, a hub of expertise on functional foods and food law.

How are animals used in food safety testing?

Food testing requirements vary by geographical region. Across the United States (US) and Europe, animal testing remains a common practice in food safety assessment.

"Animal testing is mostly used in the safety assessment of new foods or new food ingredients, before these substances are brought to market," de Boer told Technology Networks.

As de Boer explained, in the European Union (EU), new foods can only be introduced to the market once they have received proper authorization, which requires the submission of a scientific dossier detailing the safety of that new food or ingredient.

"The submitted scientific dossier contains data and results from studies conducted by the applicant, in which, for example, the kinetics of a substance are assessed, toxicological aspects of a product have been studied, or its allergenic potential [assessed]," de Boer said. "Different methods are used for these studies, including in vitro, in silico, and in vivo studies—the latter mostly using experimental animals and sometimes human data."

"For genetically modified (GM) products, EU law requires that this scientific dossier contains a so-called 90-day toxicity study—one of these in vivo animal studies," she continued. "In the repeated-dose 90-day oral toxicity study, mostly rodents are exposed to a substance daily, for a period of 90 days. The study aims to characterize potential hazards for human health through such sub-chronic, daily exposure for at least 90 days to different levels of a test substance."

Episode 4: Acceleration of Process Development & Characterization With Hybrid Modeling & Transfer Learning

Discover how hybrid modeling and transfer learning are transforming therapeutic antibody process development. Learn how machine learning enables knowledge reuse to improve process understanding, reduce experimental effort and accelerate development.
View Webinar

Advertisement

For novel foods—defined by EU law as any food not commonly consumed in the EU before 15 May, 1997—these toxicity studies are not a legal requirement. However, many novel food dossiers still contain data generated from animal experiments to prove other aspects of safety, de Boer noted. Additionally, while this approval process differs in the US, toxicological tests on rodents are still a recommended (though not legally required) part of testing for food additive petitions.

"So in practice, especially for assessing subchronic toxicity of foods and thereby for establishing safe levels of intake, animal testing is still used for food safety assessments," she said.

NAMs: An alternative to animal testing

In recent years, regulatory agencies such as the US Food and Drug Administration and the European Commission have begun to encourage moves away from animal testing for the safety evaluation of foods and additives. NAMs form a key part of these approaches.

"NAMs are not generally developed specifically for 'food safety' assessments, but for toxicology and safety assessment more broadly," de Boer clarified. "Food safety 'borrows' from this large toolbox that is ever-increasing. So almost any in vitro or in silico method with a plausible link to a relevant toxicological endpoint or mechanism can, in principle, serve a purpose in food safety testing."

While limited in number, there are examples of scientific dossiers that have been submitted and accepted without the need for in vivo animal toxicity data. One notable example of this is the case of egg membrane hydrolysate, a novel food ingredient derived from chicken eggshells, which was approved as a novel food largely based on a history of safe use and an in vitro micronucleus genotoxicity study. While an oral toxicity study in rats was also submitted, the European Food Safety Authority (ESFA) "concluded [it] did not add anything to the scientific risk assessment," stated a 2020 paper authored by de Boer and colleagues.

AI-Enabled Standardization for Scalable Cell & Gene Therapy Trial Matching

Discover how a fine-tuned GPT-4o agentic AI platform extracts oncology data from unstructured EHRs, standardizes it to mCODE 3.0, and automates patient pre-screening and trial matching to accelerate cell and gene therapy clinical trial enrollment.
View Webinar

Advertisement

"In our paper from 2020, we gave different examples [of animal-free strategies] that were shown to work well for specific food safety assessments. Bacterial reverse mutation and in vitro micronucleus assays for genotoxicity are already accepted as Tier 1 for additives, flavorings, and novel foods. Advanced 3D or organotypic models such as intestinal organoids and gut-on-a-chip systems have been shown to be useful in food safety testing also, as well as simulation models, e.g., for the gastro-intestinal tract," de Boer said.

"Next to these in vitro studies, in silico methods are also increasingly used in food safety analyses," de Boer continued. "Quantitative Structure-Activity Relationship (QSAR) and Read Across (RA) models have been used by the EFSA in grouping food improvement agents; the threshold of toxicological concern (TTC) is an approach increasingly used for low-exposure substances; and various digestion/bioaccessibility simulation models are used for micronutrient assessments."

The benefits of NAMs for food testing

While some of the interest in NAMs is due to regulators issuing directives and guidance to refine, reduce, and replace animal studies, there are other reasons to pursue the implementation of such next-generation risk assessment tools.

"Next to having the potential of reducing or avoiding animal suffering, NAMs offer the option to generate more human-relevant data," de Boer emphasized. "Animal studies generate data that needs to be extrapolated to other species, and it may therefore fail to predict some adverse effects in humans."

This is just as true for food safety assessment as it is for drug discovery and related fields; animal-based methods cannot provide the same human-relevant mechanistic understandings of toxicological events at the molecular level. In vitro tests using organoids or other materials are a better model for human exposure scenarios, and can better capture the mechanisms and pathways relevant to human biology.

"An additional benefit of NAMs is that they may offer efficient approaches: high-throughput assays and in silico tools can screen lots of substances faster and cheaper than sequential animal studies; and approaches such as RA and grouping allow for better use of existing evidence, thereby reducing duplication," de Boer added.

The future outlook on NAMs in food safety analysis

Predict Drug-Induced Liver Injury With Greater Confidence

This poster introduces programmed hiPSC-derived hepatocytes that combine the scalability and batch consistency researchers need to address liver functionality and clinically relevant toxicity responses.
View Poster

Advertisement

While NAMs hold great promise as a replacement for animal testing, moving away from animal-based toxicology assessments will not be as simple as switching out one animal experiment for one non-animal experiment.

"To start with the scientific barriers: currently, very limited NAMs can deal with complex, chronic, systemic effects that would be relevant to understand the effect of longer food exposure. Similarly, testing whole foods in NAMs is more challenging than single chemicals," de Boer said.

Scientists will also need to tackle the question of how to derive a point of departure—the exposure level on a toxicity dose-response curve at which harmful effects are first observed—from purely mechanistic data. This is required in order to set acceptable daily intake limits for new ingredients and additives. Additionally, while NAMs offer a significant opportunity for integrating AI tools into food safety testing, the "black box" nature of some AI tools may pose challenges for regulators. "On a more general note, NAMs are [being] developed at a high speed, but limited NAMs have been validated and well-established for use," de Boer added.

In addition to scientific barriers, there are also important cultural barriers to overcome. Despite their translational limitations, animal testing is still generally thought of as the "gold standard" in food safety testing. "The GM 90-day feeding study is the clearest example of that; even though different EU-funded projects found no added value of the method, the requirement persists," de Boer said. Still, regulators are used to the data generated by such trials; for NAMs to be more commonly accepted in regulatory assessments, risk assessors will need to gain more experience with NAMs data and better understand how it compares to the insights captured in animal trials.

"When it comes to regulatory barriers, we see that guidance seems to lag behind the science. Tiered testing schemes still require animal confirmation past Tier 1—a positive in vitro genotoxicity result still needs in vivo confirmation," de Boer added. "Acceptance of NAMs is case-by-case rather than systematic, and there is no agreed way to validate a NAM’s ‘level of protection’ without benchmarking against animal or human data."

Still, de Boer is hopeful that NAMs will one day have a transformative effect on food safety assessments, just as they are beginning to transform drug discovery, precision medicine, cancer research, proteomics, and more.

"I'm optimistic about the future of NAMs: the scientific groundwork, in method development and validation, is advancing steadily. Turning that promise into practice, though, will depend on continued commitment from regulators, risk assessors, and food business operators using these methods working together, rather than scientific progress alone," de Boer concluded.