Les études de neuroimagerie structurale montrent, dans le trouble bipolaire, la présence d’un certain nombre d’anomalies cérébrales. Les anomalies les plus souvent retrouvées sont des hyperintensités de la substance blanche dans les régions périventriculaires et sous-corticales profondes, un élargissement modéré des ventricules cérébraux touchant surtout le ventricule latéral droit, des altérations de la substance grise et de la substance blanche dans les régions frontale, cingulaire et temporale, ainsi que des changements de volume des structures sous-corticales, comprenant en particulier l’amygdale, le thalamus et les ganglions de la base. Il existe une grande hétérogénéité des changements morphométriques observés dans ces travaux, qui peut s’expliquer en particulier par la variabilité des méthodes utilisées et des caractéristiques cliniques des patients étudiés. Néanmoins, l’ensemble des résultats de ces études, associés à ceux obtenus grâce aux approches lésionnelles, tend à montrer une distribution non symétrique des anomalies cérébrales, et qui touche en particulier les régions frontales et sous-corticales impliquées dans le traitement des informations émotionnelles. Confirmant cette latéralisation, les anomalies d’activation cérébrale observées en neuroimagerie fonctionnelle permettent en outre de préciser le rôle particulier de certains circuits fronto-limbiques dans l’expression symptomatique du trouble bipolaire.

The number of structural neuroimaging studies of bipolar disorder have increased during recent years, expanding the literature on the nature of cerebral abnormalities underlying this disorder. The purpose of this paper is to provide a selective review on the main issues concerning this literature. Consistent findings are higher rate of periventricular and deep subcortical white matter hyperintensites seen on MRI. Although there is strong evidence for links between hyper-intense lesions and age or cardio-vascular risk factors, some authors have observed the presence of these abnormalities early in the course of the illness. There are also frequent reports on ventricular enlargement, which has been described as mild and predominant in the right lateral ventricle. Total cerebral volume appears to be preserved. Whereas changes in total grey matter volume are uncertain, evidence suggests that reduced white matter volume reflects genetic factors predisposing to the disorder. Recent studies have reported volume changes in several cortical areas including the subgenual cingular, frontal and temporal cortices. Additionally, a number of reports described morphometric abnormalities in various subcortical structures, such as amygdala, basal ganglia and thalamus. Part of the variability in the morphometric abnormalities might be attributable to differences in clinical status and demographic characteristics of patient groups. Despite some inconsistencies across the studies, it emerges that abnormalities are asymmetrically distributed throughout the two cerebral hemispheres. When increase in volume is reported, it is preferentially localised in the left cerebral hemi-sphere and more specifically in prefrontal and temporal cortices and in amygdala. By contrast, when structural abnormalities concern the right cerebral hemisphere, they are identified as deficits. These latter results are in direct line with those of studies of mania following brain injuries, which report that these secondary mania result mainly from right cerebral lesions. It is also important to notice that most of the abnormalities concern both the cortical and subcortical level, ie frontal, striatal, thalamic and limbic regions. These abnormalities highlight the role in the pathophysiology of bipolar disorder of the loops involved in emotional information processing. The particular role of fronto-limbic loops in the phenomenology of bipolar disorder have been emphasised by recent data from functional neuroimaging studies.