Lithium treatment impacts nucleus accumbens shape in bipolar disorder

The effects of lithium treatment duration on deep grey matter structures in bipolar disorder are not well known. In this cross-sectional neuroimaging case-control study, we tested the hypothesis that shape characteristics of deep grey matter structures in bipolar disorder are associated with the duration of lithium treatment and with clinical phenomenology. In a setting of neuropsychiatry outpatient clinic, we included 74 patients with bipolar disorder (BD) and 74 matched healthy control subjects (HC). Both groups underwent a Magnetic Resonance Imaging acquisition and an exhaustive assessment of clinical and psychiatrics dimensions. Shape measures of seven deep grey matter structures (hippocampus, amygdala, caudate, nucleus accumbens, putamen, globus pallidus and thalamus) were obtained from T1 weighted images in both groups, using FSL FIRST segmentation tool. The segmented structures were then analysed vertex-by-vertex with FSL Randomise tool. First, we investigated the presence of significant associations between the duration of lithium treatment and shape measures in BD sample. Then, for structures that resulted significantly associated with the duration of lithium treatment, comparisons between BD and HC were performed either considering the BD group as a whole or dividing it in three groups based on the duration of treatment (lithium drug-naïve, short and long treated). Any deformation uncovered by group comparisons was subsequently associated with depressive and hypomanic/manic symptoms. The relationship between structures shape and the duration of lithium treatment in BD sample was significant for bilateral nucleus accumbens. Specifically, significant bilateral extroflection effects, related to longer duration of lithium treatment, were found bilaterally over the surface shape of core accumbens nuclei (r2R-Accu-Core = 0.12, p = 0.016, r2L-Accu-Core = 0.1, p = 0.031). Moreover, introflection effect related to longer duration of treatment resulted over the shell of right accumbens (r2R-Accu-Shell = 0.17, p = 0.002). Nucleus accumbens shape did not differ between BD and HC considering BD group as a whole. By contrast, categorizing BD in subgroups as a function of the duration of lithium treatment revealed significant inward deformation on the core of left accumbens nucleus and outward deformation on the shell of the right accumbens nucleus in lithium-naive patients, compared to both patients with long duration of lithium treatment (pL-Accu-Core = 0.016, pR-Accu-Shell = 0.005) and HC (pL-Accu-Core = 0.002; pR-Accu-Shell = 0.005). Moreover, compared to HC, inward deformation on the core of the left accumbens surface was found for patients with short duration of treatment (pLAccu-Core = 0.027). Finally, measures of surface deformation on the core of left accumbens observed in the group comparison showed significant positive correlations with depressive symptoms severity, as assessed by the Hamilton Depression Rating Scale (total score: r2L-AccuCore = 0.07, p = 0.02, somatic score: r2L-Accu-Core = 0.1, p = 0.005) and Beck Hopelessness Scale (r2LAccu-Core = 0.05, p = 0.03). Findings demonstrate that lithium untreated BD patients are characterised by localized shape abnormalities in the nucleus accumbens. Lithium treatment could act modulating these morphometric features as part of its mechanism of action in mood stabilizing.