Conference Agenda

This panel brings together transdisciplinary contributions from natural and social scientific scholars and practitioners contributing to accelerating the transition towards animal-free innovation. Attention for non-humans is increasing in sustainability debates, including through discussions on rights of nature. There is, however, less attention for animal interests (animal health, welfare and rights), in other words, seeing animals as individuals instead of representatives of species. Especially the issue of animal testing is largely absent from scholarly and practitioner debates on sustainable development: developments in the transition to animal-free innovation and other sustainability transitions and transformations are largely disconnected. This panel zooms in on this issue of animal-free innovation, which is highly relevant to many sustainability debates, since animal testing is done for, among others, food and agriculture, health, and (environmental) safety. With this, it represents an integral part of environmental justice debates.

Animal testing continues despite political and societal support for alternatives. Animal testing is not only problematic from an ethical perspective, but also from a scientific point of view. There is increasing evidence for the variability in animal data and the poor predictability of animal studies for humans. Yet, almost 200 million animals are estimated to be used annually in research around the world. In 2018, in the EU alone, almost 10.6 million animals were used for research and testing, the majority being mice. Most animals were used for basic research (46%), a further 28% for translational and applied research, and 23% were registered to satisfy legislative requirements, followed by 5% routine production. The development of animal-free New Approach Methods (NAMs) is widespread and encompasses e.g. in vitro, ex vivo, in chemico and/or in silico alternatives. The panel discusses key issues for replacing animal testing by NAMs, including the use of artificial intelligence, communication, and transformative governance.

Safety assessment of chemicals and pharmaceuticals is traditionally performed using animal studies which is considered to be ‘the gold standard’ by both industry and regulators. In the last decades a 3R mission has been formulated aimed at replacing, reducing and refining animal studies. Concerns around animal welfare and the ambition to strive for increased relevance of test results for health and sustainability has been recognized as a powerful motor to realize this mission. However, while the call to adopt 3R principles is growing and innovations in new approach methodologies (NAMs) increasingly allow for animal-free prediction of toxicity, this has not yet materialized in a sustained reduction of animal studies used for toxicity assessments.

It has been argued that to implement NAMs, changing the safety assessment system through multi-actor collaborations is needed. Facilitating systemic change also requires a thorough understanding of all relevant system elements at play. Previous studies in the sustainability transitions field provide a systemic framework that not only focusses on technological fixes, but also takes into account socio-institutional aspects that need to be changed and fulfilled in order for a systemic change to occur. The mission-oriented innovation system (MIS) framework in particular discusses systemic change necessary to achieve a specific mission when coordination between different actors with an important role for state-led actors is required, as is the case for the 3R mission. To what extent this framework can be used and applied to the 3R mission is however yet unknown. We thus research the following question in this paper:

How can we apply insights from the mission-oriented innovation systems literature to better understand the process of developing and implementing new approach methodologies and their contribution to the transition towards animal-free safety assessment?

To develop a systemic framework, we have combined an existing linear implementation framework specifically for NAMs and developed by the Dutch National Institute for Public Health and the Environment (RIVM), with the MIS framework that adds a focus on socio-institutional aspects and system dynamics. We are currently validating the integrated framework’s applicability in various domains (e.g., pharmaceuticals, pesticides) through interviews and workshop sessions with professionals in a research consortium that focuses on the development of NAMs.

Finding existing alternatives for animal experiments is currently challenging because of the lack of comprehensive structured databases and the lack of balanced keyword-based search strategies to mine unstructured textual databases. In this paper we describe 3Ranker which is a fast, keyword-independent, and unbiased algorithm for finding alternative methods to animal experiments in biomedical research. The 3Ranker algorithm was created using a machine learning approach, by a combination of initial building of a RandomForest model on a data set of 35 million abstracts constructed with weak supervision, followed by iterative model improvement with expert curated data. We found a satisfactory trade-off between sensitivity and specificity (Area Under the Curve (AUC) values ranging from 0.85-0.95). In a practical application of these models, these were able to specifically identify articles that describe potential alternatives for animal studies from thousands of articles returned by generic PubMed queries on dermatitis and Parkinson's disease. Application of the model on time series data show the early use and acceptance of these models in the area of cosmetics and skin research. The 3Ranker algorithm is freely available at www.open3r.org, a web based application in which PubMed query results are filtered for papers that describe alternatives to animal experiments. It is the aim to develop this further for other fields besides cosmetics and skin research, and to implement the broad use by researchers, policy makers, funders and ethical review boards, in order to avoid animal studies in research to the maximum extent.

New approaches avoiding and minimizing animal testing have to be developed and implemented in the triangle between science, politics, and the public. The technological advances of recent years are in principle disrupting technologies. The pace of their development, e.g., bioengineered microphysiological systems or artificial intelligence (AI), is overwhelming for many and present a challenge for timely implementation especially in the safety sciences, which are particularly conservative. The dissemination of knowledge to all stakeholder groups is therefore critical. Scientists are usually good at teaching, sometimes even to lay audiences. It more often gets difficult when the audience is journalists, activists, or policy-makers – hesitancy to make mistakes as well as the desire to disproportionally promote their case comes into play. At the same time, the means of communication have diversified and accelerated especially with the multitude of social media channels and web-based outlets. Real-time reactions, sharing of data, tools and results, increasing invitation for personal opinion, demand for transparency, political correctness but also increasingly tribal communication (https://paulsutton.co/2010/03/05/the-tribal-characteristics-of-social-networks/), and loss of trust in experts are challenges to researchers in general. The field of alternatives to animal testing is from the start more political and important to lay audiences and cannot avoid these trends. They offer the opportunity to form community and create support for research and its implementation. Especially for academics, this requires moving out of the ivory tower of education and entering the engagement of stakeholders, such as community outreach, citizen scientists, and other forms of engagement.

This paper represents the first output of the (Research project name removed to annonymise review process) consortium. The paper presents a state-of-the-art of the science and policy practice relevant for the transition to animal-free safety assessment, as a foundation for the (Research project name removed to annonymise review process).

The aim of the research project is to contribute to the acceleration of the transition to animal-free safety assessment for chemicals and pharmaceuticals in the EU, including the Netherlands, and the USA. To achieve this aim, it applies a transformative governance approach. Transformative governance is focused on the underlying causes of societal problems, and incorporates three levels of transitions. The project operationalizes these niche, regime and landscape levels as follows. At the niche level, it develops the transdisciplinary knowledge needed to demonstrate the usability and applicability of Next Generation Risk Assessment (NGRA) for chemicals and pharmaceuticals. At the regime level, it facilitates the (regulatory) acceptance of NGRA for chemicals and pharmaceuticals, and draws broader lessons for the transition to animal-free safety assessment. The project will develop an enhanced understanding of how to govern the acceleration of this transition by experimenting with transformative governance as a consortium. At the landscape level, the project analyzes the societal underlying causes of the lack of progress, with a focus on the values, convictions and interests of different societal groups and the political economies of pharmaceuticals, chemicals, and animal testing. The project will build experience with governing the acceleration of transitions, which is relevant for animal-free safety assessment and other sustainability transitions.