NAMs: a paradigm change in biomedical research and toxicology

Historically, animal models have been a central pillar of biomedical, pharmacological and toxicological research. While they have enabled significant advances, their limitations are now well documented, particularly in terms of their transferability to humans. These findings have led to the emergence of New Approach Methodologies (NAM), which bring together a set of tools and approaches based on human biology, modelling and data integration.

In line with European Directive 2010/63/EU, which encourages replacement, NAMs are gradually establishing themselves as the solutions of choice for meeting the needs of biomedical research and risk assessment. More than just alternatives, NAMs represent a paradigm shift that leverages advances in cell engineering, microfluidics, and computer modelling to better represent the complexity of the human body.

This article aims to present the diversity of these new models, their applications, their contributions and their limitations, in order to fuel societal debate and inform public decision-making.

What are New Approach Methodologies?

The acronym NAM is a generic term that encompasses a wide range of methods, mainly non-animal, used to explore human biology and generate data on the biological and toxic effects of substances. 

In simple terms, NAMs seek to answer questions such as: Is a substance effective? Is it toxic? At what dose? Through what biological mechanisms?, using scientific strategies that differ from those used in animal models. NAMs are not strictly defined in most regulations. Instead, they are referred to using terms such as “alternative or substitute” or “non-animal” methods. Nevertheless, the term NAM is gradually becoming established to encompass these innovative approaches in scientific, regulatory and industrial circles.

The main categories of NAMs

In vitro methods: Cell/tissue culture models

In vitro methods are a long-standing and central pillar of NAMs. They involve culturing cells or tissues, often of human origin, in the laboratory in order to observe their responses to substances or physiological or pathological stimuli. 

Models are simple when they comprise a single cell type. In recent years, more complex systems have been developed with different types of cells cultured in 3D: organoids, reconstructed tissues, tissue sections. Tumoroids, for example, are structures formed from cells taken from patients’ tumours.

Intestine organoid. Image credits: Prisca Liberali/FMI

The most advanced models, such as organs-on-chip (OOC), use fluid circulation in microchannels (known as microfluidics) to precisely control the microenvironment and thus reproduce the interactions between the different cell types that make up an organ, as well as with the dynamic environment in which they live. These microphysiological systems (MPS) involve modulating mechanical and environmental stresses to more accurately reproduce certain aspects of tissue architecture, cellular diversity and interactions specific to human organs.

Organ-on-chip. Image credits: NETRI

In silico approaches: mathematical modelling, bioinformatics and AI

In silico methods are based on computer models that predict the behaviour of substances in biological systems and their interactions with certain targets such as nuclear or membrane receptors, proteins that modify the activity and/or function of cells in the body. They use existing databases, mathematical models and artificial intelligence tools.

Among the most common examples in toxicology:

• Read-across approaches, which assume that two chemically similar substances will have similar effects [1];
• SAR (structure – activity relationship) models capable of linking the chemical structure of a molecule to its biological or toxic effects, in some cases quantitatively (QSAR) [2];
• PBPK (pharmacokinetics based on physiology) models, which estimate the absorption, distribution, metabolism and elimination of a substance in the human body.

These methods and others (such as 3D modelling) are particularly useful for the initial screening of large libraries of substances, to limit or even replace the usual experimental testing. They also provide, in record time, information on the behaviour of these substances in the body and the targets they are able to reach.

A rapidly growing concept originating in industry, the digital twin in healthcare is a dynamic model of a patient, organ or physiological process, built from medical, biological, environmental and behavioural data. It can be used to simulate the progression of a disease or the potential effect of a treatment with a view to providing more personalised medicine.

In chemico methods: chemistry as a biological indicator

In chemico approaches focus on the key chemical reactions involved in toxicity mechanisms, particularly the interaction of substances with proteins (which cause allergic reactions) or with DNA (which causes harmful genetic mutations). They are widely used to measure the effect of substances on skin proteins, which can cause skin sensitisation. Today, the combination of in vitro and cell-based tests has replaced most animal testing for skin sensitisation, although some immunological tests are still performed in mice. In the field of drug development, in vitro tests are used to assess potential side effects or possible interactions with other drugs.

Integrated approaches: combining evidence

One of the major developments in NAMs is the shift from isolated tests to integrated assessment frameworks, combining different data sources to support regulatory decisions. Key concepts include: Adverse Outcome Pathways (AOPs), which describe the chain of events linking an initial molecular interaction to an observable adverse effect; Integrated Approaches to Testing and Assessment (IATA), which rely on combining in silico, in chemico and in vitro data; and Next Generation Risk Assessment (NGRA) approaches, which integrate hazard, exposure and mechanism of action data to assess real risks to humans.
In addition, there are “omic” analyses (transcriptomics, proteomics, lipidomics, metabolomics) to measure molecular and biochemical responses in cells, tissues and biological fluids. In toxicology, these approaches are used to characterise and quantify the effects of exposure to chemicals or drugs, contributing to a more mechanistic and predictive assessment.

In which areas are NAMs already being used?

Safety assessment of substances

NAMs are incorporated into regulatory assessment in several sectors. Since 2013, animal testing has been prohibited in the EU for cosmetic products and their ingredients when validated alternative methods exist, such as reconstructed human skin models.

In chemical regulations, particularly REACH (Registration, Evaluation, Authorisation and Restriction of Chemicals), companies must avoid animal testing. NAMs (in vitro, in chemico and in silico) are used via weight-of-evidence approaches; animal testing should only be used as a last resort.

In the food sector, risk assessments (additives, contaminants, contact materials, etc.) are still largely based on animal data, but European authorities are aiming to increase the proportion of data derived from NAMs. Similarly, regulations on biocides and plant protection products (pesticides) promote the reduction of animal testing.

NAMs are also transforming the evaluation of medical devices. Agencies are encouraging their use (particularly in the United States and Japan). Despite these advances, obstacles remain (see next chapter).

NAMs in biomedical research

In biomedical research, NAMs, in vitro and in silico, offer original perspectives not only for better understanding diseases and discovering new targets (biomarkers), but also for developing and testing new therapeutic strategies. Today, approximately 90% of drug candidates validated in preclinical animal tests fail in clinical trials [1,2]; contributing to a long and costly development process. NAMs, which are more predictive, faster and less costly, have the potential to improve the R&D cycle: understanding mechanisms, screening and optimising therapeutic candidates, evaluating efficacy and side effects, etc.

Another perspective is to be able to offer treatments tailored to each patient or group of patients, with a view to “stratified” or “personalised” medicine. The use of organoids and patient data combined with AI is being considered, for example, for conditions such as cancer, diabetes and neurodegenerative diseases, but also for so-called vulnerable populations (pregnant women, infants, children, etc.).

In regenerative medicine, combining advanced human models (organoids, organs-on-chips) with 3D bioprinting makes it possible to study tissue repair and regeneration mechanisms with increased biological relevance.

Limitations, challenges and conditions for success

Current scientific limitations

No model, animal or otherwise, can fully represent the complexity of a human organism. Certain issues, such as long-term systemic effects, interactions between organs, and the impact of complex exposure pathways (e.g., transgenerational alterations), remain difficult to study, even with NAMs. Some tissues, stages of development, or pathological conditions do not yet have sufficiently mature or scientifically validated models. Even though the predictability of animal models for humans does not exceed 65% and the reproducibility of trials is regularly questioned, much remains to be done to move away from animal models, particularly in the decision-making process [3].

Validation and standardisation issues

In order to be widely adopted, NAMs must demonstrate their reliability, reproducibility and biological relevance for their intended use. Scientific validation establishes this credibility, in particular through inter-laboratory comparisons and clear interpretation criteria. Where possible, standardising protocols facilitates their adoption and regulatory use. For example, in accordance with the ISO 10993 standard, authorities favour in vitro (cytotoxicity, irritation, sensitisation) and chemical characterisation coupled with in silico modelling (QSAR). The standardisation of new methods, which can be a lengthy and demanding process, is essential to build confidence among authorities, industry and the public. Harmonisation and standardisation initiatives led by the Organisation for Economic Co-operation and Development (OECD) and the International Cooperation on Alternative Test Methods (ICATM) aim to address this challenge.

Regulatory approval

NAMs can be used in innovative approaches to hazard and risk assessment. Numerous initiatives and research projects are promoting greater regulatory acceptance. While some regulations are pioneering, such as those for cosmetics, and some countries encourage their use (e.g. the FDA (Food & Drug Administration), USA) and the PMDA (Pharmaceuticals and Medical Devices Agency), Japan), the regulatory framework for NAMs is still evolving. Drug Administration, USA) and the PMDA (Pharmaceuticals and Medical Devices Agency, Japan), the acceptability of these approaches remains limited at European and global levels. Chronic and environmental effects are still largely assessed on the basis of animal studies. At the same time, mutual recognition of data from NAMs still depends on the validation and standardisation of protocols. Significant work therefore needs to be done to compare existing data in order to help regulatory bodies move towards better integration of NAMs. The International Council for Harmonisation has begun updating the appendices to its guidelines -ICH S7A and S7B- (particularly for cardiac and hepatic toxicology).

Interoperability, quality and data sharing

Interoperability, quality, and data sharing are key to the rise of NAMs. The diversity of formats, models, and platforms complicates the comparison and reuse of results. However, solutions are gradually emerging, combining common standards, harmonised metadata and FAIR principles (Findable, Accessible, Interoperable, Reusable). The development of secure infrastructures and controlled sharing spaces will improve data quality and access while protecting industrial (particularly sensitive data) and regulatory issues.

Need for new skills

NAMs can be a fundamentally different source of data from animal testing, posing a significant challenge to the current system. To ensure that we are able to understand and assimilate them now and in the future, a major effort must be made in terms of training for students, researchers and experts. Training programmes must therefore reflect the complexity and interdisciplinary requirements of NAMs.

Socio-technical barriers

The implementation of a valid and accepted method is not only a matter for science, but also for politics, law and society. Out of habit or ignorance, certain influential communities of experts tend to denigrate the usefulness of NAMs and hinder their development. The transition to these new methods requires deep reflection on the foundations of modern research. It requires collective support, bringing together researchers, competent authorities, industry and civil society.

Conclusion & Opening

The diversity of uses for NAMs in academic and industrial settings is a strength, although they are not a perfect solution for all research needs. It is therefore essential not to overpromise.

However, NAMs are establishing themselves as a major scientific and economic lever for transforming biomedical research, toxicology and risk assessment on an international scale. They therefore offer public decision-makers an opportunity to base public health policies on data that is more reliable and transferable to humans, as well as industrial policies based on cutting-edge technologies.

Their use is set to grow rapidly and the robustness of the models offered will increase, with a view to better understanding and responding to contemporary challenges in health, competitiveness and sovereignty.

However, accelerating their adoption while ensuring scientific rigour and collective confidence requires a structured approach: training and uniting stakeholders who are still too fragmented, organising shared governance of the transition, and targeted, coordinated investments that are commensurate with the challenges. The speed of progress depends on the resources allocated to validation, research, stakeholder support and infrastructure development.

Without claiming to be exhaustive on a subject as rich and promising as NAMs, this article outlines its scope, main areas of research and numerous fields of application. Other publications will follow, providing more concrete examples of how these methodologies are implemented, particularly in the area of risk assessment.

Pro Anima Scientific Advisory Board

References