StrokeCoMorb: Preclinical systematic review and meta-analysis of the effects of age and comorbidities on ischaemic stroke

CAMARADES (Collaborative Approach to Meta-Analysis and Review of Animal Data from Experimental Studies) Berlin applies and supports systematic review and meta-analysis of preclinical studies to investigate challenges in the translation of biomedical research. These meta-research tools provide a summary of research findings and allow evaluation of both the range and quality of evidence addressing specific research questions. The results can help inform decision-making and the development of improvement strategies for preclinical research and translation.

In this series of systematic reviews and meta-analyses, we investigate the effects of age and comorbidities on ischaemic stroke outcome and treatment efficacy in preclinical studies. In stroke research, the results from animal models often fail to translate to clinical benefit in humans. One of the reasons for this may be limited external validity of preclinical models. External validity is the extent to which the results of a given study are generalisable when applied to other study conditions. Our ability to generalise results beyond the laboratory may be affected by a number of factors, including how well we model relevant patient characteristics. Stroke affects women and men, who are usually elderly, and often suffer from comorbidities such as hypertension and diabetes. In contrast, the most commonly used models of stroke are young healthy male rodents. In vitro stroke experiments that recreate ischaemia-like conditions are also used to investigate stroke pathophysiology and treatments. These experiments often inform animal and human studies; however, their external validity has also not been systematically studied to date.

In this project we systematically identify, evaluate, and synthesise in vivo and in vitro stroke studies modelling advanced age and comorbidities. An important feature of this project is the use of new automation tools including machine learning and text mining to improve the accuracy and speed of the review, and to contribute to the development and validation of these tools.