107 Etiology in Lesion-Symptom Mapping: Tumor vs. Stroke INTRODUCTION Topological organization of brain function has been widely studied in patients with brain lesions.1 It remains the most powerful method to infer causality between brain structures and behavior.2 The past decades, lesion analyses methods have evolved3, and lesion-symptom mapping (LSM) is currently performed in large samples and is a frequently utilized method in behavioral neurology. LSM has most frequently been applied in stroke patients, which raises questions about generalizability. The behavioral consequences of lesions from different etiologies, such as brain tumors or stroke, may vary as a result of how they affect the brain.1,4 Firstly, the lesion distribution differs between etiologies, with certain brain areas more likely to be damaged than others. The possible locations for a stroke are confined by the architecture of the vascular system, with a stroke most often occurring in subcortical areas in the territory of the middle cerebral artery.5–7 Primary brain tumors, on the other hand, frequently involve both subcortical and cortical structures in the frontotemporoinsular areas.8,9 This nonrandom lesion distribution, which is inherent to each pathology, affects the spatial accuracy of LSM analyses since it limits the possibility to analyze rarely affected brain areas.10,11 First, areas that are rarely damaged are often excluded beforehand from the analyses. Second, for the areas that are included in the analysis, statistical power will vary across the brain and power problems arise when areas are very rarely (or very often) damaged.12 Moreover, if neighboring or downstream areas are always damaged together, the unique involvement of each voxel to a given behavioral deficit cannot be distinguished. Furthermore, disease-specific characteristics determine the behavioral consequences. An ischemic stroke is defined by an acute event causing immediate cell death, whereas the rate of brain tumor growth is much slower (months to years, dependent on grade), and neural activity can continue to persist after infiltration by tumor cells.13–15 This raises the possibility that the brain recruits mechanisms for neural plasticity in different ways when affected by different pathologies, which in turn may lead to different functional outcomes.16,17 The use of etiologies other than stroke in lesion-deficit mapping is a highly discussed and controversial topic.18–21 Recently, there has been a rise in lesion studies involving brain tumor patients. The classically theorized neural correlates of behavior, for example the involvement of the posterior perisylvian areas and Broca’s area in language, have mostly been corroborated in this population.22–26 In addition, a recent 5
RkJQdWJsaXNoZXIy MTk4NDMw