Expanding the 3R principles

Abstract

The implementation of the 3R in research using animals has already improved animal welfare and benefited science. Further progress could be achieved by expanding the 3R to include “Robustness” and “Reproducibility”.

The use of animals in research dates back to ancient Greece when physicians and natural philosophers performed experiments on live animals to understand anatomy and physiology. Throughout history, animal research played a vital role in many scientific and medical advances, and continues to aid our understanding of human diseases and the development of new therapies and cures.

Historically, the 3R have been advocated for social and ethical reasons, with less focus on how they can contribute to advances in research itself

An increased demand for animal models, together with a pervasive remonstrance against animal experimentation, caused an expansion of Laboratory Animal Science in the 1950s both to professionalize the field and to improve the humane treatment and care of animals in research. This was primarily guided by the principles of replacement, reduction and refinement (3R) defined by Russell and Burch in 1959. Historically, the 3R have been advocated for social and ethical reasons, with less focus on how they can contribute to advances in research itself. Nonetheless, adoption of the 3R has significantly improved medical research, but more improvements are required: specifically, transparency and reporting of animal experiments. The expansion of the 3R to the 5R, to include “Robustness” and “Reproducibility”, may provide a framework for ameliorating these issues.

Although most scientists agree that animal models are a central component of biomedical research, there have been critical comments lately regarding their reliability and contribution to medical progress 1. One major issue is the fact that 95% of drugs that enter clinical trials do not make it to the market, despite promising results in animal models 2. The contributions of animal models to science are further impaired by poor standards of statistics and reporting, which impair reliability and reproducibility 2. More recently, genomic data and experimental results revealed significant differences between humans and animal models, which further question the value of various animal models for research on human diseases 3. Notwithstanding that animal experimentation remains central to our understanding of human disease and the biological effects of many substances, they clearly have limitations, which can, at least in part, be ameliorated by the expansion of the 3R framework and other regulatory or reporting measures.

The impact of the 3R

There are concerns amongst scientists that a strict adherence to the 3R could compromise the value of animal studies 4, while other researchers do not see the 3R as achievable or realistic. There is an overall attitude that application of the 3R not only disturbs the research process in terms of time and cost involved owing to additional administration, but also compromises research data, mainly owing to a loss of statistical significance when reducing the number of animals in an experiment 5. However, a recent study indicates that these concerns may be overstated: after animal experiments were subjected to modulation by the 3R, the study demonstrated by internal validation that the quality of the data did not change and substantially increased cost effectiveness 4.

We therefore argue that the 3R approach is an ethically and scientifically valid framework to reform the use of animals in research and to improve scientific quality, validity and reproducibility. A strong 3R culture would provide a solid framework for implementing strategies, such as improved study design, method development and project coordination. In fact, implementation of the 3R has already yielded benefits for research in terms of quality, reliability and the development of new, alternative methods to animal experimentation.

Minimizing unnecessary stress during an experiment will reduce the variability of results along with the number of animals needed

One concern is that environmental refinement conflicts with the standardization of experiments as animals from a refined environment have a more diverse behaviour and show more variability in their response to experimental procedures. However, animals housed in refined conditions are physiologically and psychologically more stable, and are likely to generate more generalizable results 6. Moreover, compromising the well‐being of an animal can have downstream effects: these animals are often physiologically and immunologically abnormal, which might invalidate the conclusions made from experiments 6. Minimizing unnecessary stress during an experiment will reduce the variability of results along with the number of animals needed 6. The application of refinement thereby increases the validity of scientific research.

… even though many journals refer to the ARRIVE guidelines in their “guide to authors”, many authors do not adhere to them.

To achieve reduction in the number of animals in a research project often involves developing alternative methods: by way of example, microsampling has reduced the number of animals needed for toxicokinetic evaluation 4. Such reduction projects also help to identify and develop new biomarkers. Thus, developing and validating new sampling methods and improving analysis of biological samples fulfil the reduction requirement and lowers costs as less animals are needed to generate solid results.

Applying the replacement principle can also advance medical research through the development of new technologies and methods. For example, a recent study assessed efforts to replace animal models for placental diseases with human cell lines and human tissue samples in combination with human subject studies, and found this approach superior to using animals 3.

The expansion of the 3R

Overall, the 3R have in fact already generated benefits for both animal care and research. A strong 3R culture can therefore provide a framework to further improve scientific reliability, validity and reproducibility. However, there is no global standard for reporting on whether and how any of these principles have been applied 4. Registering animal studies, or at least studies that use higher vertebrates, could improve global standards of reporting and transparency—and therefore validity, reproducibility and reliability of animal experiments—akin to registries of clinical trials. A review in 2009 calculated that 85% of basic and clinical research was wasted because of inadequate or inappropriate design, unpublished negative results and poor reporting 7. Some of these problems could be ameliorated by expanding the 3R into 5R to include “Rigour” and “Reproducibility”. Animal studies that adhere to scientific rigour—robust and unbiased experimental design and methods and full transparency in reporting methods, experiments and results—would therefore lead to more reproducibility and transparency in research.

In 2010, the UK National Centre for the Replacement Refinement & Reduction of Animals in Research published the Animal Research Reporting of in vivo experiments (ARRIVE) guidelines, which have now been endorsed by more than 300 major journals. Although the guidelines were developed after consultation with the scientific community, they have been criticized since for not being comprehensive enough and for not having the same standards as for clinical trials.

Moreover, even though many journals refer to the ARRIVE guidelines in their “guide to authors”, many authors do not adhere to them. A recent study assessed publications that involve major surgery on laboratory animals whether these completely describe the procedures and adherence to animal welfare norms, found that whether the journal asked authors to adhere to ARRIVE publishing guidelines or not had no effect on the quality of the methods section 8. It concluded that many publications cannot be trusted to present full details, at least in this case, on the use of anaesthetics and analgesics. Moreover, animal welfare regulations in general do not include any guidance on publishing relevant data, leaving it to journals and authors to decide which details on animal experiments and housing are published in the Materials and Methods section.

Given the unsatisfactory outcome so far in reporting on animal experiments, it could be more efficient to place the onus on funders and research institutions to encourage or require their investigators to adhere to reporting guidelines. The US National Institutes of Health (NIH), for example, now requires that grant applications must address the implementation of the 3R (https://grants.nih.gov/reproducibility/index.htm) to enhance reproducibility. Making this a requirement for funding would be much more efficient in assuring that scientists adhere to rigour and reporting guidelines than a reference to journal guidelines at the publication stage after all experiments are done. More transparency in grant applications is also necessary for reviewers to properly assess the proposed studies.

More detailed descriptions of experimental procedures, animal welfare measures, housing conditions and so on will help to improve refinement and the welfare of animals, and enhance reproducibility

Full transparency and accurate and detailed descriptions of materials and methods are essential for scientists to evaluate results and replicate findings by someone else. As progress in science depends on previous observations and experiments and the dissemination of correct and reliable knowledge, more transparency along with making negative results publicly available will therefore improve research itself. More detailed descriptions of experimental procedures, animal welfare measures, housing conditions and so on will help to improve refinement and the welfare of animals, and enhance reproducibility. Reduction can be further achieved by publishing both positive and negative results so as to avoid repeating studies that have yielded solid and reproducible data. Lastly, transparency creates an incentive for researchers to apply scientific robustness, and to better explain the use of animals in lieu of possible replacements.

Conclusion

Since their introduction as a guiding principle for animal experimentation, the 3R have served as a framework for improving scientific validity and reproducibility in biomedical research. The ARRIVE guidelines provide an additional framework to improve the still incomplete reporting of experimental details. However, adherence to these systematic guidelines has been lacking, and many publications based on animal models either fail to describe the experimental procedures and conditions in sufficient detail or are of poor scientific quality. To improve the quality, reproducibility and transparency of scientific publications that describe animal experiments, and to further enhance biomedical research, the 3R should be expanded to the 5R to include “Rigour” and “Reproducibility”. The 5R, together with the ARRIVE guidelines, must also be more strictly implemented and regulated by funding bodies, research institutions and publishing bodies.

Conflict of interest

The authors declare that they have no conflict of interest.