Seeking Huntington disease biomarkers by multimodal, cross-sectional basal ganglia imaging

Abstract

Neurodegeneration of the striatum in Huntington disease (HD) is characterized by loss of medium-spiny neurons, huntingtin nuclear inclusions, reactive gliosis, and iron accumulation. Neuroimaging allows in vivo detection of the macro- and micro-structural changes that occur from presymptomatic stages of the disease (preHD). The aim of our study was to evaluate the reliability of multimodal imaging as an in vivo biomarker of vulnerability and development of the disease and to characterize macro- and micro-structural changes in subcortical nuclei in HD. Macrostructure (T1-weighted images), microstructure (diffusion tensor imaging), and iron content (R 2* relaxometry) of subcortical nuclei and medial temporal lobe structures were evaluated by a 3 T scanner in 17 preHD carriers, 12 early-stage patients and 29 matched controls. We observed a volume reduction and microstructural changes in the basal ganglia (caudate, putamen, and globus pallidus) and iron accumulation in the globus pallidus in both preHD and symptomatic subjects; all these features were significantly more pronounced in patients, in whom degeneration extended to the other subcortical nuclei (i.e., thalamus and accumbens). Mean diffusivity (MD) was the most powerful predictor in models explaining more than 50% of the variability in HD development in the caudate, putamen, and thalamus. These findings suggest that the measurement of MD may further enhance the well-known predictive value of striatal volume to assess disease progression as it is highly sensitive to tissue microimpairment. Multimodal imaging may detect brain changes even in preHD stages.

Figure 1

Histogram showing volume (mm³), R ${}_{\text{2}}^{\text{*}}$, MD, and FA mean values from the seven structures considered. *Significant differences between preHD and control preHD; as well as between HD and control HD. ^†Significant differences between preHD and HD subjects. (A) Volume mean values, (B) R ${}_{\text{2}}^{\text{*}}$ mean values, (C) MD mean values, and (D) FA mean values.