Substantia nigra neuromelanin magnetic resonance imaging in patients with different subtypes of Parkinson disease

Parkinson disease (PD) is a common neurodegenerative disorder, its pathological changes mainly focus on the dopaminergic neurons containing neuromelanin (NM) lost in the subatantia nigra pars compacta (SNc), and both noradrenergic and dopaminergic neurons containing NM lost in the locus coeruleus (Hirsch et al. 1988; Gibb 1992; Kastner et al.1992; Zarow et al. 2003). NM is a special dark polymer pigment, it mainly exist in the SNc and other catecholaminergic neurons. It has shown that NM is an important factor in promoting dopaminergic neurons degeneration in Parkinson disease (Hirsch et al. 1989; Kastner et al. 1992; Zecca et al. 2002). The mechanism may be that NM derived from dying neurons can activate microglia to release cytoactive factors, inducing the accumulation of the major histocompatibility complex class I (MHC-I), which can bind antigen and presenting on the membrane of the NM contained neurons, and then CD8 + T cells recognize and target these neurons, inducing neurodegeneration (Cerbrian et al. 2014). Pathological studies have shown that the degeneration of NM neurons occurs earlier than clinical symptoms appeared in PD patients, and it was exponentially lost with a 45% loss in the first decade of onset, and the number of NM in substantia nigra of PD patients after death is less than that 50% of the healthy control group (Fearnley and Lees 1991; Zecca et al. 2002). Therefore, NM may participate in the occurrence and development of PD through contribute to neuronal death, and become one of the pathogenic factors of PD. Sasaki et al. invented high-resolution spin-echo T1-weighted neuromelanin magnetic resonance imaging (NM-MRI), which revolutionized the visualization of neurons in the substantia nigra (Sasaki et al. 2006). Using high-resolution spin-echo T1-weighted NM magnetic resonance imaging (NM-MRI) can make the neurons with NM in the brain show high signal area to reflect the signal intensity and distribution of NM (Sulzer D et al. 2018). The loss of high signal area can be detected by NM-MRI in the early stage of Parkinson's disease. Therefore, we can clarify the progress of PD by noninvasive monitoring the attenuation degree of NM signal, and make NM become an in vivo biomarker reflecting the severity of PD (Cassidy et al. 2019).

Parkinson disease is a highly heterogeneous. Due to the lack of a unified classification standard, PD patients are often classified by clinical features into tremor dominant (TD) subtype and postural instability gait difficulty (PIGD) subtype (Stebbins et al. 2013). The clinical manifestations of the two subtypes were significantly different. Compared with patients with TD subtype, patients with PIGD subtype showed more severe symptoms, faster disease progression, higher depression tendency and more severe cognitive impairment (Konno et al. 2018; Dissanayaka et al. 2011). At present, the difference of neural mechanism between the two subtypes is not clear. The difference of NM concentration between different subtypes of PD detected by NM-MRI is helpful to deepen the understanding of the pathogenesis of these two subtypes and improve the efficacy of clinical intervention. Therefore, it is of great clinical significance for the treatment and prognosis of PD patients to strengthen the research on the neural mechanism of PD subtypes.

The purpose of this study was to explore the relationship between the concentration of NM and the progression of PD, and to explore the difference of NM signal intensity between the two subtypes of PD, so as to verify the feasibility of NM as a biomarker for identifying PD progression and subtypes.

In the present study, we confirmed the correlation between NM and the progression of PD, and found that there was a significant difference in the degree of NM signal attenuation between the subtypes of PD patients: (1) CNR values were significantly decreased in PD patients in medial and lateral SN compared to controls, and decreased linearly with PD progression. (2) In PIGD patients, CNR mean values were significantly decreased compared to those in TD patients. H-Y stage and UPDRSIII scores were also significantly higher in the PIGD patients than those in the TD group.

PD is a progressive neurodegenerative disease whose characteristic pathological changes include progressive loss of NM-containing dopaminergic neurons and formation of Lewy bodies in the SN of the midbrain (Ehringer and Hornykiewicz 1960). The results of pathological anatomy showed that the loss rate of dopaminergic neurons in the SN of PD patients is proportional to the number of NM-containing cells therein, and that the loss rate of each NM-containing cell is positively associated with the amount of its melanin pigmentation. Therefore, there is a close clinicopathological correlation between NM and PD (Hirsch et al. 1988; 1989; Kastner et al. 1992; Zucca et al. 2018). In this study, NM-MRI results revealed that CNR values of NM were significantly lower than that of the healthy control group, which were consistent with previous research results (Reimão et al. 2015a, 2015b; Ohtsuka et al. 2013; Isaias et al. 2016). Meanwhile, we further analyzed the correlation between NM and the course of PD, and concluded that with the progression of PD, the clinical symptoms of patients were gradually aggravated, and the CNR values of medial and lateral SN decreased, especially in the lateral part. This result was consistent with the previous pathological research results. Previous pathological results showed that during the first 10 years of Parkinson's disease, the loss of pigmented neurons was exponential, with the largest loss in the ventrolateral layer (91%), followed by the ventrolateral layer (71%) and the dorsal lateral layer (56%) (Schwarz et al. 2017; Fearnley and Lees 1991). Most previous results of NM-MRI research also found that the decrease of NM-MRI signal in PD subjects was more inclined to the lateral, posterior and ventral voxels of SN, and NM is negatively associated with PD severity indicators, such as disease course, Hoehn–Yahr (H–Y) stage, and Unified Parkinson's Disease Rating Scale (UPDRS) motor score, but there are still a few studies showing no significant correlation (Cassidy et al. 2019; Schwarz et al. 2017; Reimão et al. 2015; Kashihara et al. 2011; Fabbri et al. 2017). This difference may be due to the application of different techniques in experimental data processing, which indicates that we still need to continue to explore the relationship between NM concentration and PD progression.

Our results also showed that there was a significant difference in signal attenuation in substantia nigra between pigd and TD groups. The signal attenuation of SN in the PIGD group was more serious than that in the TD group, which may be attributed to the difference of the two subtypes in pathophysiological mechanism. Autopsy results showed that compared with pigd patients, TD patients had less loss of pigment cells in SN, and less Lewy small volume aggregation in cortex (Selikhova et al. 2009). Therefore, it is speculated that the incidence of tremor is not closely associated with nigrostriatal pathway disorder, but instead it may result from the cerebellar–thalamic–cortical pathway. Therefore, the pigmented neurons containing NM in the substantia nigra of TD group had less degeneration and necrosis (Helmich et al. 2012). Our results showed that the signal attenuation of lateral SN in PIGD group was more serious than that in TD group (AUC 79%, sensitivity 76.5%, specificity 78.6%). There was no significant difference in CNR between the two groups. However, a previous small sample imaging study found that the signal attenuation difference between the two subgroups was more significant in the medial part of SN (Yuanyuan Xiang et al. 2017), while another NM-MRI study on de novo Parkinson disease showed no significant difference between the motor subgroups (Wang et al. 2018).

As the sample size of this study is relatively small, and the classification of PD motor subtypes may change with the progress of the disease (Eisinger et al. 2017), we need to follow-up the subjects for a longer time, and conduct a large sample longitudinal study to further clarify the difference of NM signal attenuation in substantia nigra between the two subgroups. Previous studies have shown that tremor is a feature of benign PD; compared with patients with bradykinesia and rigidity dominant PD, tremor dominant PD patients enter H–Y 4 more slowly (Rajput et al. 2009), have lower dementia and depression tendency (Vakil and Herishanu-Naaman 1998; Lewis et al. 2005; Williams-Gray et al. 2007; Aarsland et al. 2003; Williams-Gray et al. 2007; Van Rooden et al. 2011), Our results showed that the H–Y stage and updrsiii score of pigd group were significantly higher than those of TD group, but MMSE and hamd-24 scores had no significant difference, which could be attributed to the small number of Parkinson's disease patients registered in this study.

As a non-traumatic resting state brain imaging technology, NM-MRI does not require subjects to perform tedious tasks or receive invasive stimulation, provides favorable operability, and can be used as a routine clinical application to assist PD diagnosis and disease follow-up. Recent studies have shown that the volume and CNR of PD patients’ SN are positively associated with the dopamine transporter (DAT) density of its corresponding corpus striatum (Isaias et al. 2016). Therefore, NM-MRI can be used to quantitatively study the pathology of SN in PD patients, especially considering its close association with the dopaminergic innervation of the human striatum in PD.

This study has some limitations, including its relatively small sample size and cross-sectional study design. Considering the instability of the subtypes of some PD patients, it is necessary for the study to replicate in a larger sample. Another small sample study with 14 PD patients detected longitudinal changes in their SN values using NM-MRI. In that study, follow-up of the PD patients found them to have a significantly smaller total area of SN and a lower CNR value than the initial NM-MRI values (Matsuura et al. 2016). Therefore, it is feasible to consider NM-MRI as an indicator of PD progression. Furthermore, the motion artifacts during the acquisition of the NM-MRI sequence may have been generated by some TD patients and may have affected the sensitivity and specificity of the results. To confirm our results, additional prospective study with a larger sample and longitudinal exploration of the correlation between NM and PD severity is still needed.

NM concentration in PD patients decreases gradually during disease progressionm with CMR values correlating linearly with the course of PD. In addition, PD patients in different subtypes demonstrate different rules of change for NM in the SN, NM may therefore be a reliable biomarker for PD severity and subtype identification. As such, we suggest that NM-MRI can be used for the identification of specific brain function changes during PD, thereby shedding new light on its pathological mechanism, diagnosis, and treatment.