CNS injury research; reviewing the last decade: methodological errors and a proposal for a new strategy.

During the last decades the field of Traumatic Brain Injury (TBI) has been characterized by a paucity of new treatments. This is in contrast to the amount of pre-clinical experimental work and the number of clinical trials done. This paper aims to contribute to the ongoing debate on the reasons that have led to this phenomenon. A reasonable suggestion could be the presence of methodological limitations when comparing and integrating experimental results. The first methodological drawback, which is shortly discussed, is the insistence (during the last decades) on the concept of "similarity to the human pathology" as the main criterion to evaluate results, and the constant effort to create a "super model" that would fully replicate human TBI cases. The second methodological limitation examined is the lack of a common way to present and analyze data. It is proposed that the basic neuro-histo-pathology of each injury model should serve as the ground on which hypotheses should be built, as it could constitute the common basis for comparisons between different experimental settings. In this context, 95 papers reporting experimental results from various models of animal CNS injury were reviewed in order to examine the extent to which results were presented and analyzed using a common basis. No such common basis was observed; moreover, the review revealed a remarkable lack of histopathological examination of the animals, especially when biochemical and/or behavioral endpoints were assessed. It is argued that this practice deprives data of an objective common basis. Conclusively, a new theoretical way of organizing experimental work in the field of TBI is briefly presented.