Abstract:The opening plenary session on Moral & Ethical Issues in MRI Research at the 2023 annual meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) highlighted several contemporary issues in preclinical research for the community to consider. Translation is bidirectional; laboratory observations improve human health; and clinical studies are refined based on the human experience.1 Unfortunately, most preclinical research cannot be replicated in studies involving human subjects, possibly due to limited internal (i.e., study design) and external (i.e., replicability across sites) validation.2 Here, we discuss the methodological and ethical challenges of preclinical models in imaging research and suggest best practices. Most preclinical studies are performed in rodents (mice and rats), nonhuman primates (NHPs), and domesticated large animals such as pigs. Animal models are needed to translate findings to humans, and human studies can help refine animal models. Humans are evolutionarily closer to NHPs than rodents, and although the study of diseases and new treatments can be more easily translated from NHPs to humans, their use in research is highly regulated. Rodent models are the most prevalent as they are easy to manage, allowing the study of cause-and-effect relationships with larger sample sizes, fewer individual variations, and shorter time frames. Nevertheless, these models often have lower translational value, mainly attributed to the vast differences from humans, among other confounds. Current approaches that aim to improve their translational value include the development of imaging biomarkers that can be reproduced between species and in humans to study disease and cross-species convergence and divergence,3 and humanized mouse models4 where mice are engrafted with human cells or tissue to study infections, cancer (patient-derived xenograft [PDX] models), among others. This has also led to the creation of co-clinical trials, where a humanized preclinical model informs a simultaneous human clinical trial.5 Finally, in genetically modified mouse models, human genes can be introduced to model systems to produce and study human disease.6 Although these efforts have increased the translational value of animal models, it remains challenging to select the optimal preclinical model.7 New efforts to increase external validity by openly sharing data between groups have facilitated reproducibility studies and will likely support translation to humans.5, 8 Initiatives such as OpenNeuro (https://openneuro.org), Zenodo (https://zenodo.org), and Open Science Framework (https://osf.io) are a few examples. To maximize resources and the impact of preclinical research, collaboration is essential. An underlying tenet is to ensure our moral and ethical approaches are shared by our collaborators. Different countries have varying legislation and cultural norms on general animal welfare,9 and these may also influence the approach to animal welfare in preclinical research. It is debatable whether the “Western concept” should be imposed on other cultures.10 To date, there is no analog to the Declaration of Helsinki for animal work, and the most common internationally used agreements are the Animal Welfare Act, the American Veterinary Medical Association Guidelines for Euthanasia, the Council for International Organizations of Medical Sciences International Guiding Principles for Biomedical Research Involving Animals, and the U.S. Public Health Service Policy on Humane Care and Use of Laboratory Animals. There remains a need to harmonize these rules across the world. To ensure the highest ethical treatment of preclinical models in research, we suggest best practices including: (1) a single international agreement for animal welfare across teams; (2) following the three Rs principle (replace, reduce, and refine)11; and (3) using the ARRIVE (Animal Research: Reporting of in Vivo Experiments) guidelines12 in reporting. Importantly, discussions with colleagues, without a moral high ground, about concerns and possible solutions is crucial. Involving research staff in on-site visits to colleagues' labs is also a beneficial goal, perhaps by leveraging the ISMRM research exchange program. In conclusion, although challenging, the use of preclinical models is important to understanding and treating human disease and developing cross-species MRI technology and biomarkers. The translational value of preclinical research can be improved by using new models such as transgenic and humanized mouse models. To ensure our colleagues sharing our ethical concerns, a push toward unified guidelines from the onset while maintaining respect for distinct cultural and institutional norms are key goals. Andrew D. Scott is funded by British Heart Foundation Grant RG/19/1/34160.

Ethical Considerations of Preclinical Models in Imaging Research