Introduction
Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease clinically characterized by a progressive weakness and signs of upper (UMN) and lower motor neuron (LMN) impairment. According to the degree of UMN and LMN involvement, three phenotypes can be distinguished: classical ALS (cALS), primary lateral sclerosis (PLS) and progressive muscular atrophy (PMA) [
1].
The hallmark of ALS is the presence of TDP-43 aggregates in degenerating UMN and LMN [
1]. However, these aggregates can be found far beyond motor neurons in many ALS patients, including
substantia nigra (SN) [
2‐
4], probably accounting for the extra-motor features seen in ALS patients [
1].
By means of transcranial sonography (TCS), a larger area of hyperechogenicity in SN (SNh) has been described in patients with several neurodegenerative diseases, including Parkinson’s disease (PD), atypical parkinsonisms, and ALS [
5‐
14]. In them, the SNh has been proposed to be a marker of vulnerability for neuronal degeneration, probably caused by a disturbance in iron metabolism [
10,
12,
15,
16].
Two different parts can be distinguished in the SN in terms of anatomy and function: the pars reticulata (SNpr) and the pars compacta (SNpc). The latter contains dopaminergic neurons, some of which are packed in clusters called nigrosomes [
17‐
19]. In PD patients, postmortem studies have shown that dopaminergic neuronal loss typically affects nigrosomes, especially the largest one, called nigrosome 1 [
20]. Nigrosome 1 is located along the rostro-caudal axis of the SNpc, in its dorsal part, and is visible as hyper-intense ovoid sub-structure within the dorsal hypo-intense midbrain region, medial to the cerebral peduncles in high-resolution susceptibility weight (SW) images of 3T and 7T brain magnetic resonance imaging (MRI) [
20,
21].
Interestingly, the loss of nigrosome 1 has been demonstrated in 7T and 3T brain MRI of PD patients with postmortem confirmation and has been suggested to be a consequence of neuromelanin loss, an increase in iron content and/or a change in iron oxidation state, associated to the degeneration of dopaminergic neurons of the SNpc [
20]. Moreover, the absence of nigrosome 1 differentiates with high-specificity and -sensitivity healthy controls from PD patients [
22], where it has been associated with the disease severity and clinical laterality of symptoms [
23].
Despite clinical, imaging and neuropathological data suggesting the impairment of SNpc, at least in a subset of ALS patients [
2‐
4], no previous study has assessed nigrosome 1 status in ALS patients.
Therefore, the aims of this study were: (1) to evaluate the nigrosome 1 status in patients with ALS; (2) to evaluate the association between nigrosome 1 status and SNh; (3) to address the contribution of demographical, clinical and genetic factors to the absence or presence of nigrosome 1 in patients with ALS and finally; (4) to evaluate the prognostic value of the absence of nigrosome 1 in the disease.
Discussion
The SNpc is a small midbrain region comprising neuro-melanin-containing dopaminergic neurons (some of them packed in nigrosomes), which becomes frequently affected in neurodegenerative diseases with motor involvement, such as parkinsonisms and motor neuron diseases [
2,
4,
10]. Up till now, several imaging biomarkers have been described in the SN in these diseases. The SNh, measured by TCS, is thought to reflect iron deposits and has been shown to be a marker of SN vulnerability rather than neuronal loss [
10,
16]. Accordingly, the SNh is not related with motor symptoms, but with genetic and constitutional factors that increase the risk of those neurodegenerative diseases [
10,
16]. Unlike the SNh, the absence of nigrosome 1, detected in high SWI of brain MRI, is thought to represent a neuro-melanin/iron imbalance due to neuronal loss, and has been associated with motor symptoms in the contralateral brain hemisphere in PD patients [
20,
23]. Despite the profuse literature about both biomarkers, to the best of our knowledge, no previous study has compared the absence of nigrosome 1 with the SNh in any neurodegenerative disease. Here, we demonstrate that there is not an association between the SNh and the absence of nigrosome 1, reinforcing the idea that they measure different pathophysiological processes. In short, the SNh probably measures an increase in iron content in the SNpc acting as a marker of neuronal vulnerability, while the nigrosome 1 absence is rather a marker of neuro-melanin loss in the SNpc. Considering this, it is not surprising that, in some diseases specifically causing SNpc degeneration (such as PD), the SNh is as frequently found as the nigrosome 1 absence [
22]. Conversely, in other diseases, such as ALS, the SNh is much more frequently found than the absence nigrosome 1 on MRI or on the neuropathological assessment.
Thus, in postmortem studies, the SNpc becomes involved in about 20–30% of patients [
2,
4,
33], a similar proportion to the nigrosome 1 loss found in our study. Different neuropathological patterns of involvement of the SNpc can be distinguished among ALS cases. On the one hand, the SNpc involvement is a feature of a subset of cALS patients being in an advanced neuropathological stage (2–4) [
2,
4]. On the other hand, the SNpc involvement is a hallmark of ALS with pallidonigroluysian degeneration [
2,
4,
33]. Similarly, different clinical patterns of patients combining motor neuron disease and parkinsonism have been described (e.g. ALS-parkinsonism, PLS-parkinsonism, ALS-FTD-parkinsonism…) [
34]. Moreover, parkinsonian features and a dopaminergic deficit have also been described in the striatum of a variable proportion of cALS patients [
30,
35‐
37]. Intriguingly, no association between extrapyramidal features and dopaminergic striatal deficit in ALS patients has been found [
30,
35,
36], suggesting that other circuits are responsible of these symptoms. Accordingly, the SNh was not associated with clinical features in ALS patients, but with genetic and constitutional factors [
16]. However, SNh is not a marker of neuronal loss and, up till now, no previous study has assessed nigrosome 1 in ALS patients.
In this study, we show for the first time that nigrosome 1 is absent, on at least one side of the
SNpc, in 30% of ALS patients with different phenotypes, compared to about 95% of PD patients and 10% of healthy controls [
22]. Interestingly, the neuropathologic involvement of the SNpc has been found in a similar percentage of patients in postmortem studies [
2,
4,
33].
Unlike in PD patients [
23], in most ALS patients, the nigrosome 1 was absent in both sides and, in those with an asymmetric involvement, no relationship with the clinical side of onset was found. Moreover, no associations between nigrosome 1 absence and age, disease duration or disability were found. Furthermore, although the frequency of FTD and
C9ORF72 mutations was proportionally higher in patients with absent nigrosome 1, no independent association could be found in the multivariable model (perhaps due to the small sample size). However, male sex and the UMN score (a proxy of the degree of UMN impairment) were independently associated with the nigrosome 1 absence. Moreover, the nigrosome 1 absence was most frequent in PLS patients and least frequent in PMA patients. These findings are not surprising, considering previous literature. Firstly, male sex is a well-known risk factor for ALS, Parkinson’s disease and larger area of hyperechogenicity of
substantia nigra [
16]. Second, both parkinsonian features and dopaminergic dysfunction have been more frequently described in male and UMN-predominant ALS patients [
30,
35]. Third, extrapyramidal features are much more frequent in PLS than in ALS patients and can respond to levodopa [
38]. All these suggest a relationship between male sex, pyramidal and extrapyramidal involvement. The distinction between extrapyramidal and pyramidal symptoms and signs is complex and it has been hypothesized that parkinsonian traits in UMN-predominant ALS patients could actually be attributed to spasticity [
30]. However, our data showing an association between UMN signs and the absence of nigrosome 1, rather suggest a direct relationship between the degree of pyramidal and extrapyramidal involvement in ALS patients. This fits well with the proposed model of corticofugal axonal spread of the disease [
39] and the existence of a direct cortico-nigral pathway [
40]. Thus, the absence of nigrosome 1 in ALS patients could be a marker of the cerebral neuropathological expansion of the disease to the second neuropathological stage, as proposed by Brettschneider et al. [
2]. In this respect, the lack of association between the nigrosome absence and disability or disease duration is not surprising, since the neuropathological extension in ALS is not related with none of these variables [
2,
3,
33].
Finally, we found that the nigrosome 1 absence is an independent prognostic factor, when adjusted for other well-known risk factors. As commented above, the absence of nigrosome 1 could be a marker of neuropathological multisystem degeneration, which ultimately associates to poor prognosis [
3], as found in ALS patients with concomitant FTD or other extramotor symptoms [
31,
41].
Strengths and limitations
Our study represents the first study assessing the status of nigrosome 1 and SNh in a large cohort of thoroughly characterized ALS patients, which allowed the use of multivariable analysis.
The main limitation is that parkinsonian features were not assessed systematically and consequently have not been analyzed. However, interpreting parkinsonian features in ALS patients may be challenging, since they can be masked by UMN and LMN signs [
30]. Moreover, previous studies have not found relationship between extrapyramidal traits and SN dysfunction [
30,
35,
36]. Another limitation is that the assessment of the nigrosome 1 is observer-dependent. To minimize this limitation, the nigrosome 1 was assessed by two experienced readers and discrepancies between the two were resolved by consensus. The use of 123 I-FP-CIT scintigraphy could have help to validate our results of nigrosome 1 absence, but was not available in this study. Although we studied a large cohort of ALS patients, some characteristics (FTD, C9ORF72 expansion) are infrequent, which could have limited the power to detect an association between nigrosome 1 and those variables. Finally, healthy controls or ALS mimics were not included to assess its role as a diagnostic biomarker. However, given its frequency in ALS patients, it does not appear to be useful in the diagnostic work up of ALS patients.