Animal models of cerebral palsy.

Abstract

Animal models of a human disease are laboratory animals that show some or all of the pathological features that are observed in the actual disease. They can be studied extensively to provide insights into pathophysiological processes. They can also reveal clues for diagnosis. And, most importantly, they can serve for early testing of potential treatment approaches. Interpretation and generalization of the findings must be cautious because of limitations relating to differences between the studied animals and humans, between the model and the actual disease, and the risk of experimental bias. The latter may relate to many factors, including species, (sub)strain, sample size, randomization, blinding, analysis, and reporting. However, animal models have unquestionably contributed to many invaluable advances in medicine, and it is hard to imagine how the treatment of disease in humankind would be if the generated knowledge were not available. Animal models have been developed to study and test hypotheses regarding several aspects of cerebral palsy (CP), though as a heterogeneous condition it is probably not best approached as a disease. Most studies using such models have concentrated on brain response to early injury, with a special focus on situations considered to be similar to factors associated with CP. With regard to (human) preterm birth, for example, various types of insult have been imposed on animals, none of them entirely satisfactory, taking account of species-specific maturational organization. Thanks to these experiments, we have learned a lot about maturation, blood supply, metabolism, inflammation, and plasticity. Yet, differences in the timing of the occurrence of key brain maturation events can make comparison of regional vulnerability, regenerative potential, and functional consequences difficult. Attempts to translate findings from models of perinatal arterial ischaemic stroke and white matter injury to CP have proved challenging.1 Even long-held hypotheses such as the role of hypoxia in brain injury associated with preterm birth are still debated,2, 3 in spite of a significant body of research conducted in various rat, rabbit, sheep, and baboon models. This preoccupation is hampering the development of effective strategies for primary and secondary prevention. Furthermore, many studies of suggested CP models have not analyzed the motor phenotype, and many of those that did failed to document a phenotype evocative of CP. Thus, it might be misleading to refer to these as models of CP. There is evidence that many of the processes addressed in this research are also being implicated in other neurodevelopmental disorders, such as intellectual disability, attention-deficit/hyperactivity disorder, and autism spectrum disorder.4 But very few studies are pursued beyond the perinatal period or use appropriate methods to assess features that would provide more direct insights into those disorders, besides the potential for early neuroprotection. Indeed, changes in muscle tone or reflexes after early brain injury have been sparsely documented in a handful of animal models (non-human primates, pigs, sheep, and more practically, rabbits).5 Non-motor features are also highly relevant to CP and they should, therefore, be studied in CP models. In such models, outcomes of experimental therapies could be assessed, generating arguments for reflection on the management of individuals with CP. For this purpose, more clarity is required on the type of CP that is being modelled, based on the timing and nature of the insult, lesion, and phenotype. In recent years, disease modelling has also benefited from dramatic advances in stem cell research, biomaterials, and machine learning. These could provide potentially useful complementary models to contribute to progress in treatment approaches in CP. Basic research, rigour in the nature of that research, and rigour in translation of pre-clinical findings are indispensable to human CP.