Dépression après accident vasculaire cérébral et localisation de la lésion

Abstract

Post-stroke depression constitutes a challenging field of research in clinical neurology because it has both a high prevalence (20-60%) and a negative impact on the functional outcome. The possibility of correlating depressive behaviours and symptoms with neural circuits and regions not only allows further insight into brain functioning but also permits to identify patients with an increased risk and suitable for treatment. The aetiopathology of post-stroke depression remains not yet completely understood and the neurocognitive approach for linking symptoms to regional brain dysfunction is of limited value in such a research. The complexity of behavioural and cognitive symptoms and the marked inter-patient variation in clinical expression is a limit for functional neuroimaging. The consequence is that the neural systems processing mood regulation and their neurotransmitters are only partially known. Until now, there are no biological markers and experimental models suitable for post-stroke depression. Univocal clinical diagnostic criteria (DSM-IV) are actually available to diagnose post-stroke depression, nevertheless the possibility of subgroups of syndromes (major and minor, predominantly with vegetative signs or with associated anxiety) makes problematic not only to carry out clinical investigations on large population samples, but also to maximise treatment of individual patients. For those reasons a vast amount of scientific literature has failed to define clear anatomical-clinical correlations for post-stroke depression. Methodological limitations such as selection bias (e.g. the exclusion of aphasic patients from most prospective cohort studies) or differences in clinical and radiological tools for investigation raise significant doubts about the results of previous studies on the role of a left-right and an anterior-posterior cerebral gradient for post-stroke depression. Rigorous meta-analysis studies suggest that the lesion location appears to contribute only to a small extent to the risk of developing post-stroke depression. Nevertheless, some clinical data are noteworthy for localisation purposes. The association of post-stroke depression with a specific cognitive syndrome (psychomotor slowing, executive and verbal learning dysfunction, amotivational states), and the presence of equivalent mood changes in other neurologic diseases with a documented regional pattern of lesion (CADASIL, vascular dementia, multiple sclerosis, Hungtington's disease, Parkinson's disease, HIV encephalopathy) suggest the role of the prefontal regions, the frontal-subcortical networks (including striatum, caudate nucleus and the thalamus) and the lymbic and paralymbic regions (temporal lobe and amygdala). Clinical similarities of post-stroke depression with endogenous depression and their response to serotonin reuptake inhibitors indicate a serotonin imbalance as the common metabolic substrate. For future research on post-stroke depression and lesion localisation a major reworking of criteria for case definition is necessary, placing great emphasis on the profile of cognitive, behavioural, somatic or vegetative changes. It is also important to determine the impact of specific functional deficits due to the stroke itself (i.e. aphasia, hemiplegia, visual disturbances) that account, at least in part, for the subsequent mood disorder.