The 3Rs (Replacement, Reduction and Refinement) originating from Russell and Burch (1959) are the guiding principles in preclinical animal research. They had major influence on the legislation on animal testing in many countries and have been incorporated into the US Guide for the Care and Use of Laboratory Animals (Chapter 1) and the European Directive 2010/63 EU (Article 4), the latter of which forms the basis of national legislation on animal care and use in EU member states:
Directive 2010/63 EU, Article 4: Principle of replacement, reduction and refinement
1. Member States shall ensure that, wherever possible, a scientifically satisfactory method or testing strategy, not entailing the use of live animals, shall be used instead of a procedure.
2. Member States shall ensure that the number of animals used in projects is reduced to a minimum without compromising the objectives of the project.
3. Member States shall ensure refinement of breeding, accommodation and care, and of methods used in procedures, eliminating or reducing to the minimum any possible pain, suffering, distress or lasting harm to the animals.
There is a close interplay between Animal Welfare and Data Quality in that the latter has impact on the principles of the 3Rs.
How Data Quality Affects Animal Welfare and the 3Rs
Refinement:Fully validated, robust methods, full disclosure of all data and the integrity of data forms the basis for robust data which in turn allow for educated post-approval reviews and decisions by the Institutional Animal Care and Use Committees (IACUCs) / Animal Welfare Bodies (AWBs) and competent authorities on refinement opportunities. The sharing of experimental protocols and data (in the spirit of transparency and full disclosure, e.g. on a data sharing platform) will allow further exchange of experiences and refinement of assays across scientists from different institutions and research fields.
Moreover, transparency enables assurance of compliance with IACUC-approved protocols. While not leading to refinement per se, it mitigates the risk that animals experience more suffering than allowed for.
Reduction: More robust statistics with appropriate power ensure reduction of animal numbers to the minimum required to still allow generation of meaningful data. The use of validated preclinical model systems that are fit for purpose, use of authenticated research tools and better experimental design with appropriate controls/baseline, pre-specified in- and exclusion criteria, start- and endpoints, and randomization and blinding leads to more conclusive animal studies, which abolishes the need for unnecessary repetition of experiments. Full reporting of all experiments (valid and failed) with results accurately reflecting the raw data, possibly on a platform to share raw data and metadata, and prevention of bias serves the same purpose. This, as well as appropriate sample size calculations, use of historical controls (where appropriate) and re-use of data may further reduce the need for additional animal experiments.
Also for authorization of animal experiments it is important to provide sufficient detail to enable IACUCs / AWBs and competent authorities in the evaluation of the scientific rigor of a proposal. It will help to decide whether subjecting animals to an additional test will indeed result in additional knowledge.
This is not to say that there is no need to repeat animal experiments: replication or reproduction of experimental results represents a major principle of the scientific method. However, full reporting of methodological detail (research resources and experimental design) from the original study will facilitate understanding of what was actually done in that study, will help to judge the quality of the animal experiment and will increase the likelihood for successful replication.
Further, it must be realized that, in the short term, the need to validate experiments and to conduct more appropriately powered studies (many animal experiments are underpowered and hence inconclusive) will lead to an increase in animal numbers, but that this will be outweighed by the generation of more reliable and conclusive animal studies, which in the long-term will reduce the risk that animals are wasted.
Replacement:In the extreme, this greater transparency could lead to evidence-basedconclusions that an animal procedure offers no advance over an alternative method, or even that some assays are of so limited validity that, until the issues inherent to the assays have been addressed, they should be replaced by non-animal approaches. Important, these conclusions must be data-driven and must not be based on poorly substantiated claims or impressions.
How Animal Welfare Affects Data Quality
Clearly, Animal Welfare will also impact on Data Quality: it will assure that the animals that are used in a study will be of high quality, are defined by genotype, strain, gender, age and source. As such, high Animal Welfare standards form the basis for the authentication of the most critical part of an animal experiment, that is the animal. The monitoring of environmental factors that could impact Animal Welfare and adherence to acceptable standards, ranging from breeding conditions, transport, holding and handling to health monitoring and specific experimental procedures, allows control over relevant variables that influence experimental outcomes and ensures the full disclosure of those variables. Implementation of guidelines such as the Directive 2010/63 into national legislation further facilitates harmonization of experimental conditions across laboratories from different countries. Thus, high Animal Welfare standards should be considered a prerequisite for the generation of validated, robust methods and reproducible data in animal experiments.
Also, the adherence to the principles of the 3Rs provides a constant challenge to the researcher and thereby promotes innovation and efficiencies as scientists improve their models and assays in the search for alternative or refined methods and reduce the number of animals to the minimum required to generate meaningful data. IACUCs/AWBs can support the scientists in these efforts.
However, care has to be taken to strive for the right balance. In particular, it seems important to correctly interpret the meaning of reduction in the context of animal experiments. ‘Reduction’ does not merely mean to decrease the number of animals used indefinitely, but to use an optimal number of animals. Numbers should be the minimum required to obtain meaningful results, not too high, but also not too low! Instead, care should be taken to use the right number of animals, based on formal statistical criteria, e.g., power calculations, as an insufficient number of animals would result in underpowered studies with unreliable data, with the outcome that those animals would be wasted, which would not be in the spirit of the 3Rs.
Thus, there are many areas of overlap between Animal Welfare and Data Quality, and while there may be some, usually short-lived, ‘negative’ influence of one over the other (e.g., the need to add animals for validation studies), it seems obvious that in the long-term these interactions are beneficial both for Data Quality and Animal Welfare.
BioResearch Quality & Compliance, Janssen R&D