Parkinson disease (PD) is a common neurodegenerative disease that affects 6 to 10 million people worldwide, incurred an estimated $14.4 billion dollars in medical costs alone in the United States in 2010, and is projected to become a greater burden as the population ages.¹ The cardinal signs of PD are bradykinesia, tremor, rigidity, and postural instability. The degenerative process leads to the destruction of dopaminergic neurons in the substantia nigra pars compacta (SNc) that project to the striatum. Neuronal loss and the presence of Lewy bodies in the SN and locus coeruleus (LC) provide a histopathologic diagnosis of PD, though the implications of Lewy bodies outside the SN or LC are uncertain.²

On imaging, narrowing of the SNc on T2-weighted images may be seen, but the finding is highly variable, and the main role of MRI is to exclude other etiologies (ie, neoplasm, vascular, etc) or support the diagnosis of atypical forms of parkinsonism, including multisystem atrophy (MSA) and progressive supranuclear palsy (PSP). The “hummingbird sign” in PSP or the “hot-cross bun sign,” T2 lateral putaminal hyperintensity, and brain stem atrophy in MSA are useful radiologic findings suggesting atypical parkinsonism. Distinguishing idiopathic PD from atypical parkinsonism is important, as treatments and prognoses differ. In light of the fact that idiopathic PD is by far the most common cause of parkinsonism, however, the current role of diagnostic imaging is limited.

While the classic features of PD revolve around debilitating motor symptoms, a substantial percentage of patients with PD have cognitive deficits, with up to 75% of those with PD developing dementia after 8 years,³ possibly rising to 83% at 20 years.⁴ The mechanisms of cognitive decline are unknown, though there is an association between the presence of intraneuronal Lewy bodies and cognitive impairment. A previous study showed that cognitive decline correlated with regional Lewy body density predominantly in the anterior cingulate and entorhinal cortex.⁵ Compared with healthy controls, patients with PD with mild cognitive impairment have lower D2 dopamine receptor levels in the associative (ie, cognitive) striatum as well as the insular cortex, suggesting coupling between striatal and limbic systems.⁶ The hippocampus has also been implicated as a mediator of cognitive dysfunction in PD, based on preclinical and in vivo volumetric and functional work in PD.⁷ These studies suggest that neuropsychological features of PD may be understood as circuit pathologies, involving limbic, executive function, or memory pathways, and consistent with parallel cortico-striato-pallido-thalamic-cortico circuit models of the disease.⁸

High-frequency deep brain stimulation (DBS) is used increasingly to treat medically refractory tremors or treatment-related complications such as dyskinesias in PD. The subthalamic nucleus (STN), located between the red nucleus medially and the SN laterally; internal pallidum; and thalamus are potential targets, associated with different degrees of efficacy and side effects. Efficacy rate as measured by percent of patients with off-period reduced or abolished tremor has been reported to be 54% for STN and 33% for pallidal DBS. Efficacy for dyskinesia is approximately 75% for STN and 65% for pallidal DBS.⁹ Although DBS was first described in the 1980s,¹⁰ the technique is still relatively novel and mechanisms of action are not well understood. Functional and structural imaging, in conjunction with neural modeling, electrophysiologic recordings, and better clinical characterization will improve our ability to predict efficacy and adverse side effects associated with DBS.¹¹

The current AJNR News Digest focuses on advanced imaging to optimize anatomic delineation of midbrain structures that are etiologically involved in PD,¹² distinguish idiopathic PD from atypical parkinsonism,^13,14 investigate pathway disruptions as possible mechanisms of cognitive impairment,^15,16 and optimize coregistration methods to improve DBS targeting.¹⁷

Delineating midbrain anatomy could lead to studies that will improve our understanding of both motor and nonmotor aspects in PD, but anatomic detail is limited at clinical field strengths. Gore et al investigated midbrain dopaminergic subcortical structures in healthy controls at 7T. Qualitative assessment, contrast-to-noise ratio (CNR) and volumes of the ventral tegmental area, SN, and red nucleus were studied. They compared a mixed T2- and T2*-weighted gradient-echo (GRASE) to a T2*-weighted fast field echo (FFE) sequence. Both GRASE and FFE clearly delineated the midbrain anatomy. Compared with FFE, GRASE better distinguished densely versus poorly vascularized subregions of the red nucleus. As expected, CNR was higher at 7T compared with values previously reported at 4T. Although the distinction between SN pars compacta (SNc) and pars reticulata (SNr) at 7T was judged to be good, the authors note that reliable boundaries between SNc and SNr have yet to be established. Future studies applying these high-field metrics in patients with PD are needed to determine the utility of 7T MRI.

Differentiating idiopathic PD from atypical parkinsonian syndromes such as MSA and PSP can be difficult. Qualitatively, basal ganglia hypointensity and linear putaminal hyperintensity on T2-weighted images suggest the diagnosis of MSA, and further separation into MSA Parkinson (MSA-P) and cerebellar (MSA-C) subtypes is a goal for diagnosis and treatment. Focke et al extend this by quantifying R2, R2*, DTI, and magnetization transfer parameters in patients with PD, MSA-P, MSA-C, PSP, and in healthy controls.¹³ Pallidal and putaminal iron content, as measured with R2*, differentiated MSA-P from PD and healthy controls with high predictive power. There was a small but significant increase in mean diffusivity in the SN of PSP compared with the other groups. Overall, however, R2, DTI, and MT measures were relatively insensitive to group differences. Strengths of this study included its quantitative approach, prospective design, and rigorous statistical approach. However, due to small samples sizes (9 PSP, 10 MSA-P, 12 idiopathic Parkinson syndrome, 13 controls) these interesting results will have to be replicated in larger studies.

As an alternative to studying PD disease-related DTI at the group level, Haller et al used machine learning to classify patients at the individual level. Support vector machine (SVM) analysis of imaging data fundamentally differs from the more common group-level voxel-wise analyses. Voxel-wise analyses are mass-univariate tests that separately analyze each voxel across 2 or more groups. SVM is a multivariate analysis of patterns across voxels, resulting in each individual having one pattern, with the goal of identifying patterns that allow discrimination between individuals. Haller et al retrospectively investigated 17 patients with PD and

23 with atypical parkinsonism. Importantly, classifier studies are most accurate when groups reflect real-world prevalence rather than selected "classic cases." Thus, consecutive and unselected patients over a 5-year period were included. All patients had 30-direction DTI at 3T. SVM analysis of fractional anisotropy (FA) correctly classified PD versus atypical PD with 97.5 ± 7.5% accuracy. SVM analysis on DTI data is a potentially clinically applicable method that provides accurate detection of PD. Though useful at an individual level, the disadvantage of classifier studies is a limited ability to interpret results from a neuropathologic perspective.

Kamagata et al investigated limbic pathway integrity using DTI. They compared patients with PD without and with dementia (PDD) and healthy controls, and found that FA was significantly decreased in the anterior cingulum in PD and PDD compared with controls. In addition, FA was lower in the posterior cingulum in PDD vs. controls. FA correlated with Mini-Mental State Examination scores, suggesting that DTI measures may be a marker of cognitive performance. The FA differences could reflect one of several pathologic changes including Lewy body deposition, neuronal loss, or demyelination.

The thalamus plays a role in the pathophysiology and treatment of PD. Thalamic dysfunction appears to result from direct pathology as well as downstream effects of nigrostriatal degeneration. Vaillancourt et al investigated thalamic integrity in early, unmedicated PD. DTI measures were studied in 6 thalamic nuclei: anterior nucleus (AN), ventral anterior (VA), dorsal medial (DM), ventral posterior medial (VPM), ventral posterior lateral (VPL), and pulvinar (PU). Compared with controls, PD had lower FA in thalamic nuclei involved in cognitive and affective processing (AN, VA, DM), but there was no difference in FA within nuclei involved in sensory processing (VPL, VPM, PU). Vaillancourt et al show that it is insufficient to consider the thalamus as a single entity. Individual nuclei have specific connectivity and function, and may be differentially affected in PD.

Motor outcomes following DBS surgery depend on electrode location, but the optimal anatomic targets and placement within those targets are not fully understood. For most DBS surgeries, targets are defined using atlas-based coordinates. Differences in brain size and shape are accounted for by coregistering the patient’s MRI to a standard atlas. Kupsch et al investigated the accuracy of 3 coregistration algorithms: a conventional global algorithm and 2 optimized subcortical coregistration masking algorithms. The spatial dispersion of 16 anatomic landmarks relative to an atlas was compared across the 3 registration methods in 71 patients who had DBS placement. Compared with conventional registration used in most DBS surgeries, optimized registration with subcortical masks was more accurate, suggesting there is substantial room for refining registration methods to improve outcomes and further our understanding of the mechanism of action.

References

1. Kowal SL, Dall TM, Chakrabarti R, et al. The current and projected economic burden of Parkinson's disease in the United States. Mov Disord 2013;28:311–18, 10.1002/mds.25292
2. Gelb DJ, Oliver E, and Gilman S. Diagnostic criteria for Parkinson disease. Neurol 1999;56:33–39, 10.1001/archneur.56.1.33
3. Aarsland D, Andersen K, Larsen JP, et al. Prevalence and characteristics of dementia in Parkinson disease: an 8-year prospective study. Neurol 2003;60:387–92, 10.1001/archneur.60.3.387
4. Hely MA, Reid WGJ, Adena MA, et al. The Sydney multicenter study of Parkinson's disease: The inevitability of dementia at 20 years. Mov Disord 2008;23:837–44, 10.1002/mds.21956
5. Kövari E, Gold G, Herrmann FR, et al. Lewy body densities in the entorhinal and anterior cingulate cortex predict cognitive deficits in Parkinson’s disease. Acta Neuropathol (Berl) 2003;106:83–88, 10.1007/s00401-003-0705-2
6. Christopher L, Marras C, Duff-Canning S, et al. Combined insular and striatal dopamine dysfunction are associated with executive deficits in Parkinson’s disease with mild cognitive impairment. Brain 2014:137:565–75, 10.1093/brain/awt337
7. Calabresi P, Castrioto A, Di Filippo M, et al. New experimental and clinical links between the hippocampus and the dopaminergic system in Parkinson’s disease. Lancet Neurol 2013;12:811–21, 10.1016/S1474-4422(13)70118-2
8. Alexander GE, DeLong MR, Strick PL. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annu Rev Neurosci 1986;9:357–81, 10.1146/annurev.ne.09.030186.002041
9. Volkmann J. Deep brain stimulation for the treatment of Parkinson’s Disease. J Clin Neurophysiol 2004;21:6–17
10. Benabid AL, Pollack P, Hommel M, et al. Treatment of Parkinson tremor by chronic stimulation of the ventral intermediate nucleus of the thalamus. Rev Neurol Paris 1989;145:32–23
11. McIntyre CC, Savasta M, Walter BL, et al. How does deep brain stimulation work? Present understanding and future questions. J Clin Neurophysiol 2004;21:40–50
12. Eapen M, Zald DH, Gatenby JC, et al. Using high-resolution MR imaging at 7T to evaluate the anatomy of the midbrain dopaminergic system. AJNR Am J Neuroradiol 2011;32:688–94, 10.3174/ajnr.A2355
13. Focke NK, Helms G, Pantel PM, et al. Differentiation of typical and atypical Parkinson syndromes by quantitative MR imaging. AJNR Am J Neuroradiol 2011;32:2087–92, 10.3174/ajnr.A2865
14. Haller S, Badoud S, Nguyen D, et al. Individual detection of patients with Parkinson disease using support vector machine analysis of diffusion tensor imaging data: initial results. AJNR Am J Neuroradiol 2012;33:2123–28, 10.3174/ajnr.A3126
15. Kamagata K, Motoi Y, Abe O, et al. White matter alteration of the cingulum in Parkinson disease with and without dementia: evaluation by diffusion tensor tract–specific analysis. AJNR Am J Neuroradiol 2012;33:890–95, 10.3174/ajnr.A2860
16. Planetta PJ, Schulze ET, Geary EK, et al. Thalamic projection fiber integrity in de novo Parkinson disease. AJNR Am J Neuroradiol 2013;34:74–79, 10.3174/ajnr.A3178
17. Schönecker T, Kupsch A, Kühn AA, et al. Automated optimization of subcortical cerebral MR imaging−atlas coregistration for improved postoperative electrode localization in deep brain stimulation. AJNR Am J Neuroradiol 2009;30:1914–21, 10.3174/ajnr.A1741

 

Image modified from: Focke NK, Helms G, Pantel PM, et al. Differentiation of Typical and Atypical Parkinson Syndromes by Quantitative MR Imaging.