IT is extraordinarily confronting to realise, after devoting much of your career to a worthy endeavour, that the primary system for achieving your goals is flawed.
From fresh analyses and thinking on animal models in fundamental biomedical research, a bracing and uncomfortable message has emerged for scientists.
This current trepidation on animal models is not a fad or fringe discussion, but has found forums in well established and highly regarded scientific and medical journals, for example, Science and The BMJ, as well as recently in mainstream media.
Discussion, controversy and debate on the experimental use of animals can be traced to the beginning of the scientific revolution. Over the course of the 19th century, with the increasing modernisation of scientific inquiry and its links to informing medical knowledge, tensions arose that we still recognise today, in terms of human health advancement via animal experiments versus unjustified animal cruelty.
The wider public had a role, with British physiologists of the 19th century, for example, sensitive to public concerns about animal cruelty during experimental procedures, informed by reaction to gruesome experiments conducted by some European physiologists of the time.
From this tradition we gained the “3Rs” — replacement, reduction, refinement —to promote and guide humane experimentation, while ultimately aspiring to absolute animal replacement.
The 3Rs are central tenets of the Australian code for the care and use of animals for scientific purposes, providing practical guidelines on best humane practices for using animals experimentally. They are clearly a significant advance to satisfy both the conventional wisdom that animal experimentation is essential for health progress, as well as the animal welfare obligations broader society demands.
What is different now is that the science itself is under fire, stemming from the re-evaluation of evidence from the burgeoning biomedical literature, the growing concerns about the “valley of death” when developing fundamental discoveries towards health interventions, the costs of research and so on. There is often mention of a “90% failure rate” for animal to human translation.
Disquiet has appeared through the recognition that, in spite of the massive discovery science and preclinical research effort, the expected acceleration in delivery of new therapies has not eventuated — and has in fact declined — while costs to bring a new medicine to market have increased.
Comments and analysis of reasons contributing to this unexpected trend were presented by a 2004 US Food and Drug Administration report, Innovation or stagnation: challenge and opportunity on the critical path to new medical products. This report covered many contributing issues, including animal testing and experimentation.
Others have offered candid appraisals of the multiple problems facing the extrapolation of animal results to humans for medical intervention (Can animal models of disease reliably inform human studies?), citing the 90% attrition rate in the context of eventual use in humans.
The gravity of this contention was further amplified because the scientific literature surveyed to reach this conclusion was only from highly cited research papers published in prestige journals. And to add discomfort, concerns about experimental practices have been revealed.
Terms like “publication bias” or “optimism bias” and “validity” appear through these discussions to capture concerns. Therefore, is the problem to do with extrapolating from animals to humans per se, or is it the way animals are used, or a combination of both?
This is not to say that all animal models are inadequate, as suggested by the 10% success rate. Murine studies of fundamental immunology have provided insights (eg, major histocompatibility complex-restricted antigen recognition, dendritic cell function) that have contributed valuable knowledge to human health.
Yet, even with such success, there are also the failures that lead to calls to consider how we might shift focus to human immunology.
Having been confronted by this situation, it seems a natural instinct to dismiss or ignore the evidence on animals as proxies for fundamental investigations into human disease.
Another approach is to engage with this evidence and take a proactive position, which, for example, might be to determine which animal studies contribute to the successful 10%, while also considering how we may develop alternatives to animal experiments that we know will not add value.
Much progress has been achieved across the 3Rs without it being recognised as such (eg, stem cells, 3D cell culture), with adherence to refinement measures for experimental animals now commonplace.
With our natural curiosity and expertise as scientists, a new horizon of biomedical innovation beckons.
Brett A Lidbury is associate professor with the Genomics and Predictive Medicine group at the John Curtin School of Medical Research, the Australian National University, Canberra, and a scientific advisor to Humane Research Australia.