Basic research: Standardizing animal atherosclerosis studies to improve reproducibility

Abstract

leading cause of death worldwide, and tremen­ dous research activities have improved our understanding of the underlying pathology. Whereas human studies often remain obser­ vational, research on laboratory animals in principle allow functional, interventional, and possibly even longitudinal observations. The current mechanistic understanding of athero­ sclerosis has led to the emergence of numerous preclinical developments. However, the liter­ ature is full of examples of poor reproducibil­ ity of animal studies between laboratories and, consequently, limited translational success. Against this background, a scientific statement in Circulation Research provides guidelines for the selection of suitable animal models, experi­ mental design, and standardized assessment of the atherosclerotic phenotype1. The first chapter of the statement gives an overview of available animal models of atherosclerosis with specific emphasis on differences between species and the model­ ling of comorbidities. Although the authors provide an extensive overview, they focus on models that mimic early human atherosclero­ sis, or that imitate some features of advanced plaques. However, generating lesions with features of unstable plaques or even of plaque rupture and/or erosion is of great relevance to translational studies. Indeed, the fact that The third chapter provides guidelines for the standardized assessment of lesion sizes and composition (FIG. 1). Traditionally in animal atherosclerosis studies, lesion bur­ den is quantified as a measure of lesion size, but the statement makes an important point that the lesion composition is more relevant. In addition, objective analysis of the various parameters collected from the lesions, rather than use of selected parameters, might be a superior approach to understanding the overall vulnerability status of a lesion3. Overall, the statement makes an impor­ tant contribution to standardizing mouse experiments of atherosclerosis and might, therefore, help to reduce variability between laboratories and improve reproducibility. However, one should not forget that the major source of error is humans involved in animal experi ments ranging from the animal carers to the principal investigator. For example, the inclusion and display of control groups is an increasing cause of concern. Complex animal studies using (possibly inducible) conditional mouse models require the inclu­ sion of several control groups. Mostly, how­ ever, only one control group is displayed in these studies. Similarly, depletion studies using diph theria toxin or antibodies, or intervention studies involving targeted deliv­ ery or complex vehicle composition require meticu lous inclusion of possibly several control groups. In addition, mechanistic conclusions drawn from a finding might not necessarily be accurate, especially when not all para meters are taken into consideration. A classic observation is the striking reduction rupture events are rarely observed in ani­ mal models has hindered the development of novel models of plaque destabilization. Regardless, multiple pathological processes that are the basis of atherothrombosis in humans (fibrous­cap thinning, necrotic­core expansion) are also observed in animal models. Various approaches to generate such lesions based on surgical or genetic manipu­ lation have been reported2, compared with one another3, and summarized previously4. Considering the importance of understand­ ing mechanisms of plaque destabilization to translate this knowledge into therapeutic strategies, increased standardization of the use of such animal models is required. The second chapter highlights important aspects of experimental design and standard­ ized reporting. Important details discussed in the statement include the genetic background of mice, possible sex­specific differences, dietary considerations, as well as consider­ ations on the appropriate sizes of groups. Studies — even in high­impact journals — frequently use underpowered group sizes5. When inadequate group sizes are used, the initial observation might not be reproduci­ ble6. Moreover, not only the group size, but also the independent repetition of an animal study is important — a statement rarely made in animal­based studies. B A S I C R E S E A R C H

DOI: 10.1038/nrcardio.2017.132

Cite this paper

@article{Soehnlein2017BasicRS, title={Basic research: Standardizing animal atherosclerosis studies to improve reproducibility}, author={Oliver Soehnlein and Carlos Silvestre-Roig}, journal={Nature Reviews Cardiology}, year={2017}, volume={14}, pages={574-575} }