Structural MRI studies

The most consistent change in the HD brain is a significant progressive volumetric loss of the striatum[4,11–20]. A reduction of 50%-54% in mean putamen volume and 28%-29% in mean caudate volume has been reported in patients with mild to moderate HD[11,12]. Striatal atrophy has been also documented in early HD patients with Total Functional Capacity (TFC) scores between I-II[14,15] and in premanifest HD gene carriers who were even 15-20 years before predicted disease onset[4,13,16–20]. The amount of volume loss in the striatum correlates with the age of onset, the disease duration and the CAG repeat length[14,15,21]. While motor impairment correlates with increased putamen atrophy, Mini-Mental Status Examination scores (MMSE) and cognitive assessments are inversely correlated with the amount of caudate volume loss[11,12].

Cortical volume loss has been also reported in HD patients[17–20,22,23]. Cortical thinning occurs early during the course of the disease and seems to be topographically selective proceeding from posterior to anterior cortical regions as the disease progresses[22,23]. Individual variability in regional cortical thinning may also have a role in explaining phenotypic variability. For example, HD patients with more prominent bradykinesia showed significant cortical volume loss in frontal regions including the pre-motor and supplementary motor areas compared to HD patients with chorea[23]. Additionally, regional cortical atrophy correlates with clinical measures such as TFC, Unified HD rating scale (UHDRS) and cognitive tests enhancing the role of this measurement as potential biomarker for assessing neuroprotective therapies[23]. Widespread white matter (WM) atrophy has been identified in HD patients and has been associated with longer CAG length and decline in cognitive and motor performance[24]. Changes in WM volume are detectable up to 12-15 years before the predicted onset and correlate with cognitive functions underlining the role of structural connectivity degeneration in the pathogenesis of HD[25]. Diffusion tensor imaging (DTI) studies have also reported WM tract abnormalities in premanifest HD gene carriers and alterations in diffusion indices were correlated with cognitive performance[26–28]. Dumas and coworkers[28] have found abnormal WM connections of the sensori-motor cortex, which correlated with the 5-year probability for symptomatic conversion.

TRACK-HD is a multicentre longitudinal study, which focused in identifying sensitive and reliable biomarkers in premanifest HD gene carriers and early HD patients[17–20]. Four groups were enrolled in TRACK-HD: 120 premanifest HD gene carriers which were subdivided in pre-HD A and pre-HD B according to the proximity to predicted disease onset (pre-HD A > 10.8 years; pre-HD B < 10.8 years), and 123 early HD patients subdivided in two groups according to the TFC scores (HD stage I, HD stage II). At 12 months follow-up significantly increased total brain volume atrophy rates were reported in both premanifest HD gene carriers and early HD patients. Caudate and putamen volume was reported reduced by 1.4% to 4.5% compared with baseline in premanifest and early HD group. Atrophy of WM was also increased in all groups[18]. Over 24 mo, greater increases in caudate and putamen atrophy were observed in all four subgroups. Higher rates of whole brain and grey matter (GM) loss were reported in pre-HD B, HD-I and HD-II; whereas in the pre-HD A GM atrophy was confined to the striatum. Interestingly, WM atrophy around the striatum and within the corpus callosum and posterior WM tract was observed even in the earliest premanifest stage[19]. At 36 mo, early HD patients showed further significant increases in whole brain, caudate, putamen and GM atrophy and these measures were strongly associated to TFC decline. Although in pre-HD A group increased rates of whole brain, striatal and WM atrophy were observed, these were not accompanied by progressive worsening of motor and cognitive performance. On the contrary, pre-HD B showed higher rates of brain structural loss compared to pre-HD A group and these were associated with significant decline in several motor and cognitive tests. Furthermore, striatal and GM volume measures were sensitive predictors of subsequent clinical diagnosis of HD in the pre-HD B group[20]. Taken together, these findings suggest that MRI measures are able to track pathology in premanifest and manifest HD gene carriers and could be useful for the designing of future clinical trials.

Dopaminergic system

Altered dopamine signalling may play a key role in the pathogenesis of HD[49,50]. In particular, striatal MSNs expressing dopamine receptors are primarily affected in HD, whereas presynaptic dopaminergic nerve terminals are relatively spared[51]. PET studies in premanifest and manifest HD gene carriers have shown severe involvement of the postsynaptic dopaminergic system, whereas the dopaminergic nerve terminals seem to be less affected[52–55]. An 18F-fluorodopa case-study did not demonstrate diminished striatal dopamine synthesis capacity suggesting an intact nigrostriatal pathway[52]. However, Ginovart and coworker[56], using PET with 11C-b-CIT, have found a 50% decrease in striatal dopamine transporter (DAT) binding. In line with this finding, nigrostriatal density of the type-2 vesicular monoamine transporter (VMAT2) was found reduced in HD patients[57]. It still remains unclear whether degeneration of nigrostriatal dopaminergic neurons or presynaptic terminal dysfunction takes place in HD.

Investigations of postsynaptic dopaminergic systems, specifically the role of D1 and D2 receptors, which are highly expressed in MSNs, have shown reduced receptor densities and activity in the striatum of HD patients even at the early stage of the disease. The radioligand 11C-SCH23390 is a selective antagonist of D1 receptors while 11C-raclopride is a selective reversible antagonist of D2 receptors. Striatal D1-dopamine receptor density was found reduced by 75% in five HD patients with mild to moderate disease compared to a group of healthy controls[58]. Additionally, one premanifest HD gene carrier showed D1 binding in the lower range of the control subjects[58]. Turjanski and colleagues[55] have studied 10 non-neuroleptic treated patients with HD with either the choreic or the akinetic-rigid predominant phenotypes of the disease. They found severe parallel reduction of striatal D1 and D2 receptor binding with greater loss of mean striatal D1 and D2 binding in the akinetic-rigid patients than those choreic patients without rigidity[55]. However, there were no significant correlations between D1 and D2 striatal receptor binding and the duration of symptoms. Mean 11C-SCH23390 and 11C-raclopride binding was found to be reduced by 40% in the striatum of five patients with HD[56]. The degree of the decrease in D1 and D2 binding in the striatum was significantly associated with the duration of symptoms indicating that these two receptors may be reliable quantitative markers for monitoring disease progression[56]. Moreover, a reduction in D1 receptor binding was found also in the temporal cortex suggesting that dopaminergic abnormalities occur in cortical areas and may play a role in the development of cognitive dysfunction observed in HD[56]. Specifically, striatal D1 and D2 receptor density showed strong relationships with performance in several tasks assessing executive function, visuospatial ability, episodic memory, verbal fluency, perceptual speed and reasoning in a group of five HD patients[59]. Thus, cortico-striatal and/or thalamo-cortical circuity may be associated with cognitive impairment in HD[59]. A correlation between striatal D1 and D2 receptors binding, but mainly D2, and cognitive performance was found also in 17 premanifest HD gene carriers, in whom both striatal dopamine receptor levels and cognitive performance were lower in the subjects closer to the predicted disease onset[60]. Using 11C-raclopride PET and statistical parametric mapping, Pavese and coworkers[61] have found a reduction in D2 receptor density in cortical regions of symptomatic HD patients, which were also evident in frontal and/or temporal regions in 55% of premanifest HD gene carriers[62], suggesting that changes in cortical D2 receptor availability might be an early event in HD pathophysiology. Van Oostrom and colleagues[63] have also reported a reduction in striatal D2 receptor availability in 50% of premanifest HD gene carriers and these reductions correlated with increases in cumulative disease load as measured by disease burden (CAG index).

Clinically manifested HD patients have been shown to have constant loss of D2 receptor availability at around 5% per year in striatal and extrastriatal regions including frontal and temporal cortex, though no correlation between changes in UHDRS motor scores and reductions in striatal binding were observed[61]. Longitudinal 11C-raclopride PET studies in premanifest HD gene carriers have reported rates of decline from 4%[64] up to 6.3%[65]. Andrews and coworkers[64] investigated striatal dopamine D1 and D2 receptor binding over a follow-up period of 40 mo in nine premanifest HD gene carriers and four symptomatic HD patients. They reported a mean annual loss of D1 and D2 binding of 2% and 4% respectively in the group of premanifest HD gene carriers and a mean annual loss of D1 binding of 5% and D2 binding of 3% in symptomatic HD patients[64]. Additionally, UHDRS motor scores and TFC correlated with PET measures of striatal dopamine receptor in both groups. Interestingly, premanifest HDgene carriers who demonstrated active progression had an increased mean annual loss of D1 and D2 receptor binding (5% and 6.5% respectively). Thus, the authors conclude that PET measures of striatal D1 and D2 dopamine binding may be used to identify asymptomatic HD gene carriers who are actively progressive[64]. A reduction in the striatal dopamine D2 binding, in particular in the putamen, correlates weakly with the increasing probability of symptomatic conversion within 5 years, as calculated by an age and CAG repeat based model[51]. Although, putaminal D2 binding correlated with predicted time to disease onset, the rate of change of D2 receptor changes were not increased around the onset of HD symptoms[51]. A cross-sectional study by Antonini and colleagues[66] indicated that striatal degeneration in HD patients might proceed in a non-linear fashion. They found a correlation between CAG repeat length and the estimated percentage loss of striatal D2 binding after age correction in premanifest HD gene carriers and symptomatic HD patients. While CAG repeat length influenced the rate of disease progression, the slopes of the correlation for asymptomatic mutation carriers and patients were significantly different, implying that the rate of disease progression is faster during the earlier asymptomatic stages of the disease[66]. These data suggest that striatal D2 measures are more sensitive in premanifest HD than later in the disease.

While the loss of striatal dopamine D2 receptors is well known, few studies have addressed the extrastriatal D2 receptor distribution in patients with HD. Statistical parametric mapping of 11C-raclopride binding in patients with HD suggest a loss of cortical dopamine D2 receptors in symptomatic HD patients[61,62]. A significant reduction in postsynaptic dopamine D2 receptor binding was also found in the hypothalamus of nine premanifest HD patients and in 10 asymptomatic HD gene carriers[67]. These findings suggest that hypothalamic dysfunction occurs early during the course of the disease and may be responsible for the development of commonly reported nonmotor symptoms in HD including progressive weight loss, alterations in sexual behaviour and disturbances in the wake-sleep cycle[67].

Using PET with 11C-FLB457, a radioligand with high affinity for dopamine D2 receptor, Esmaeilzadeh and coworkers[68] have investigated density of dopamine D2 receptors in extrastriatal brain regions in patients with mild to moderate HD. They found that unlike from striatum, D2 receptors seem to be relatively spared in the brain extrastriatal regions in HD patients suggesting that D2 receptor binding in brain regions outside the striatum may not be a reliable biomarker in HD[68].

Moreover, PET with D1 and D2 receptor radioligands has been used to assess the efficacy of restorative therapy. In 1998, a multicentre open label pilot study was designed to evaluate the safety and efficacy of bilateral fetal striatal transplantation in HD[69]. Five HD patients were transplanted and followed up clinically and with PET over a 3-10 year postoperative period[70,71]. No significant differences were found over time between patients, grafted and non-grafted on the UHDRS and striatal D1 and D2 binding suggesting that there was no obvious surviving striatal graft tissue[70,71].