An analysis of the correlation of lesion size, localization and behavioral effects in 283 published studies of cortical and subcortical lesions in old-world monkeys

Abstract

The present article evaluates the quality and magnitude of the effects of lesion size and location and their interaction, on the behavioral performance of old world monkeys by a quantitative comparison of 283 published studies. The results indicate that lesion size alone is a poor predictor of the behavioral performance of monkeys, as opposed to Lashley's work in rats. Lesion location is a reliable predictor of the behavioral performance for brain regions thought to be primarily involved in a specific behavior; however, similar behavioral effects, although less reliable, can be observed for many different lesion loci, suggesting a specialized and a holistic brain functioning to be working at the same time. Some lesion loci are, in sharp contrast to current hypotheses about functional localization in the brain, not associated with impairments, but with significant improvements of a specific behavior. For such lesion loci the correlation of lesion size and behavioral performance may yield significant positive relationships (that is, increasing behavioral improvement with increasing lesion size); these relationships are contrasted by the significant negative relationships obtained for lesions of brain regions thought to be primarily involved in a given behavior. Thus, the lesion size may be a good predictor of the behavioral performance, depending on the lesion location and on the behavior under measurement. The behaviors analysed in this study were discrimination or delayed reaction or delayed matching-to-sample. The former two behaviors involve habit-like learning and are thought to be mediated by corticostriate functional pathways in the brain and the latter behavior implies the learning of single events, being thought to be mediated by corticolimbic functional pathways in the brain. Improved performances were observed for habit-like behaviors after lesions of brain regions (lateral frontal, premotor/motor, parietal, inferotemporal cortex, amygdala and fornix) being not primarily involved in a given behavior but possibly being able to inhibit the corticostriate pathways. Interestingly, lesions of subareas of the neostriatum were found to cause impairments in habitlike behaviors presumably being processed via these subareas (e.g. head of the caudate nucleus and delayed reaction), but to cause significant improvements in other behaviors (e.g. head of the caudate nucleus and visual discrimination). Thus, it may be concluded that diverse systems of functionally interconnected brain regions may maintain reciprocal inhibitions, with the result that a lesion within one system not only leads to a loss of one behavior, but in addition leads to a modification, may be a facilitation, of another behavior. In addition to a neural reorganization by disinhibition, more fundamental reorganizational processes may take place in the presence of very large lesions. Impairments of a certain behavior are sometimes missing after massive lesions involving much more tissue than that of the brain regions being thought to mediate the respective behavior (e.g. large bilateral temporo-occipital removals and visual discrimination). Obviously, the likeliness for plastic changes increases with the extent of the lesion and thus the necessity of the brain to fulfill plastic changes. When the amount of explained behavioral variability is assessed by a multiple regression model, the estimates ( r 2) of explained variance are maximally 8% for the lesion size and 31% for the lesion location (depending on the type of behavior measured), thus roughly comparing to results of similar studies using brain-damaged humans. The conclusion out of this may be that the lesions and behaviors used in the current experimental designs are not yet specific enough to allow a more powerful explanation of behavioral variability. However, a similarly qualifying conclusion may be that the concepts of functional localization must be complemented by nonlocalizational ones, emphasizing a holistic, that is very widespread functional organization of the brain and emphasizing organic and environmental factors, leading to interindividually different functional organizations of the brain and thus interindividually different reactions to brain damage.