Neuromelanin-sensitive imaging in patients with idiopathic rapid eye movement sleep behaviour disorder

Abstract

This scientific commentary refers to ‘The coeruleus/subcoeruleus complex in idiopathic rapid eye movement sleep behaviour disorder’, by Ehrminger et al. (doi: 10.1093/brain/aww006 )

This scientific commentary refers to ‘The coeruleus/subcoeruleus complex in idiopathic rapid eye movement sleep behaviour disorder’, by Ehrminger et al. (doi: 10.1093/brain/aww006 )

Rapid eye movement sleep behaviour disorder (RBD) is an intriguing parasomnia that has been recognized in humans for ∼30 years. When RBD is not associated with medication or other obvious features of a co-existing neurological disorder, the term ‘idiopathic’ RBD is often used. Idiopathic RBD continues to gain attention as it relates to neurodegenerative disease—particularly the α-synucleinopathies of Lewy body disease (manifested by the clinical phenotypes of Parkinson’s disease, dementia with Lewy bodies, or pure autonomic failure) and multiple system atrophy ( Boeve, 2010 a ). Retrospective and prospective analyses have shown that RBD often precedes the onset of these classic motor and/or dementia syndromes by many years.

There are numerous questions pertaining to RBD, several of which relate to the RBD-neurodegenerative disease association and challenges in predicting if, how and when phenoconversion to a more overt, defined neurodegenerative syndrome will occur ( Boeve, 2010 b ). Furthermore, while animal studies have suggested that dysfunction of the dorsal pontine sublaterodorsal nucleus and/or medullary magnocellular reticular nucleus, and related nuclei and networks, are key to human RBD pathophysiology ( Fig. 1 A) ( Boeve et al. , 2007 ), more data are needed to further elucidate their involvement.

Figure 1.

Schematic depicting key aspects of normal REM sleep and REM sleep without atonia (the electrophysiological substrate for REM sleep behaviour disorder) associated with neurodegenerative disease in humans. Normal REM sleep ( A ). The glutaminergic sublaterodorsal nucleus is in the subcoeruleus area (just caudal to the locus coeruleus), and is part of the REM-on region. The sublaterodorsal nucleus (or homologous nucleus in humans) projects to spinal interneurons (‘direct route’, denoted by the line from the sublaterodorsal nucleus to spinal interneurons). The ‘indirect route’, denoted by the line from the sublaterodorsal nucleus to the magnocellular reticular formation to the anterior horn cells of the spinal cord, may also contribute to muscle atonia. Either or both routes represent the final common pathway(s) that cause active inhibition of the cranial motor nuclei (not shown) and anterior horn cells of the spinal cord, resulting in suppression of skeletal muscle activity in REM sleep. REM sleep without atonia ( B ). According to the schema by Braak et al. (2003) , the locus coeruleus, sublaterodorsal nucleus and magnocellular reticular formation degenerate early in the course of Lewy body disease underlying evolving Parkinson’s disease (stage 2), and prior to significant changes in the substantia nigra (stage 3). A similar topography of degeneration may explain REM sleep behaviour disorder associated with multiple system atrophy. The loss of active inhibition of the sublaterodorsal nucleus ± magnocellular reticular formation on the anterior horn cells of the spinal cord combines with influences by the locomotor generators (not shown) to cause REM sleep without atonia and the syndrome of REM sleep behaviour disorder. Excitatory projections represented by encircled ‘plus’ sign, inhibitory projections represented by encircled ‘minus’ sign, with the size of these symbols representing the relative effect of each projection on the synapsing nuclei. Nuclei are represented by circles or ovals, with solid coloured circles and ovals reflecting those with normal populations of neurons, and speckled circles and ovals reflecting those with significantly reduced populations of neurons. The relative tonic influences of each projection are represented by line thickness, with thicker lines depicting stronger influences, thinner lines depicting weaker influences, and dashed and dotted lines depicting weak influences due to damage to neurons in the respective nuclei. AHC = anterior horn cell; EMG = electromyographic; LC = locus coeruleus; MCRF = magnocellular reticular formation; SLD = sublaterodorsal nucleus; SN = substantia nigra.

In this issue of Brain , Ehrminger et al. (2016) provide data and insights pertinent to several of these questions. They confirm the observations of many other investigators (reviewed in Boeve, 2010 a ) by showing that compared to age- and gender-matched controls ( n = 21), patients with idiopathic RBD ( n = 21) are more likely to have symptoms or signs of constipation, olfactory deficits, orthostatic hypotension, and subtle motor impairment. These clinical findings also relate to unique neuroimaging findings. Neuromelanin-sensitive MRI revealed reduced signal intensity in the locus coeruleus/subcoeruleus complex of those with idiopathic RBD compared to controls. Furthermore, this decrease in neuromelanin signal intensity was proportional to the loss of muscular atonia during REM sleep (also known as REM sleep without atonia on polysomnography), which Ehrminger et al. interpret as suggesting a direct and causal link. Specifically, the authors identify a negative correlation between radiographic neuromelanin signal intensity and polysomnographic REM sleep without atonia. Given current understanding of REM atonia control, this negative correlation is in the expected direction, as one would anticipate that neuronal loss or dysfunction in the dorsal pontine sublaterodorsal nucleus (and/or related ‘REM-on’ structures responsible for REM atonia control) would cause both decreased neuromelanin signal intensity on MRI and increased amounts of polysomnographic REM sleep without atonia through both direct and indirect mechanisms of REM atonia control exerted by the sublaterodorsal nucleus ( Fig. 1 B). Possible concerns regarding this finding of a negative overall association between radiographic neuromelanin signal and REM sleep without atonia are that no significant correlation was found in either the RBD or control subgroups (possibly suggesting a spurious correlation due to type 1 error), as well as the use of a non-validated method of polysomnographic REM sleep without atonia scoring. Further studies using well-validated polysomnographic REM sleep without atonia methods ( Frauscher et al. , 2012 ) will be necessary to determine whether the degrees of neuromelanin signal loss and REM sleep without atonia are truly correlated.

The neuromelanin-sensitive imaging findings—along with the results in RBD patients using other imaging modalities ( Boeve, 2010 a )—underscore the importance of differentiating an association from a direct aetiological link when considering human RBD. First, there are many examples of RBD-associated structural lesions in the posterior fossa, most of which implicate network(s) in the dorsal pons and/or medulla as causally related ( McCarter et al. , 2015 ), and the localization of pontomedullary circuit(s) involved in human RBD appears clear. One could interpret the association between RBD and the α-synucleinopathies that manifest clinically with Parkinsonism, as well as the reduced striato-nigral uptake on ¹²³ I-FP-CIT SPECT and substantia nigra hyperechogenicity on transcranial sonography in those with idiopathic RBD, as indicating that substantia nigra dysfunction (and hence dopamine deficiency) is a direct cause of human RBD pathophysiology. However, the structural lesion data, pathological data, and other recent imaging data ( Boeve, 2010 a ; Mayer et al. , 2015 )—showing the absence of nigral dysfunction in some cases—strongly suggest that human RBD may be associated with, but is not directly linked to, substantia nigra (or dopamine) dysfunction. This is not to suggest that ¹²³ I-FP-CIT SPECT and transcranial sonography are not useful biomarkers; on the contrary, both appear to be predictive for future phenoconversion of RBD to a more defined neurodegenerative syndrome ( Iranzo et al. , 2010 ). Similarly, a reduction in signal intensity on neuromelanin-sensitive imaging in the area of the coeruleus/subcoeruleus complex must also be interpreted with caution. First, as shown in the images presented by Ehrminger et al. , the coeruleus/subcoeruleus complex is small, and depending on the slice thickness and volume averaging, the signal intensity may be variable. This may lead to inconsistencies between quantitative analysis and visual evaluation, and also increase the variability in longitudinal analysis of the signal change. Also, the presence of neuromelanin is an indirect marker of neurodegeneration. Importantly, the subcoeruleus (which is the region where the sublaterodorsal nucleus or its human homologue resides) is just caudal to the locus coeruleus, but the signal intensity on neuromelanin-sensitive imaging may be driven primarily or exclusively by the locus coeruleus (see Fig. 10 in Boeve et al. , 2007 ). This nucleus degenerates early in the evolution of Parkinson’s disease ( Braak et al. , 2003 ), and may also do so in dementia with Lewy bodies (although there is little direct evidence for this yet). The changes in the coeruleus/subcoeruleus complex in those with RBD might therefore be an association but not a causation, similar to the nigral story. Finally, the locus coeruleus degenerates early in the course of Alzheimer’s disease too ( Pamphlett and Kum Jew, 2015 ), and atrophy in this region is a feature of both Alzheimer’s disease and dementia with Lewy bodies. While RBD is only rarely associated with pathologically-proven Alzheimer’s disease ( Boeve, 2010 a ), the specificity of signal reduction on neuromelanin-sensitive imaging for the α-synucleinopathies is not clear. One of the more compelling arguments for a causal link in the analysis by Ehrminger et al. may be the correlation between the signal intensity and the percentage of REM sleep without atonia on polysomnography. The authors themselves note many of the strengths and limitations described above in discussing their data.

Some obvious next steps for these and other investigators to extend this line of work will be to perform longitudinal clinical imaging (including neuromelanin-sensitive imaging) and polysomnography studies using well-validated REM sleep without atonia measures ( Frauscher et al. , 2012 ) in idiopathic RBD patients, and correlate these findings with pathology. It is particularly important to determine whether these quantitative imaging biomarkers are predictive of phenoconversion to disorders such as Parkinson’s disease, dementia with Lewy bodies or multiple system atrophy. Using neuromelanin-sensitive imaging in those with Alzheimer’s disease dementia and progressive supranuclear palsy, with and without co-existing RBD, will also provide insights on the specificity of this imaging technique.

We commend these investigators for demonstrating that another imaging technique is worthy of further study in characterizing patients with idiopathic RBD, along with several key clinical measures. Future studies will determine whether neuromelanin-sensitive imaging should be included as one of the key biomarkers in disease-modifying trials involving subjects with idiopathic RBD.

Funding

B.F.B., E.K.S.L., and K.K. are funded by the National Institutes of Health.

Glossary

¹²³ I-FP-CIT SPECT: ¹²³ I-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl) nortropane (FP-CIT) single photon emission computed tomography (SPECT) demonstrates dopamine transporter integrity in the basal ganglia. Reduced striatonigral uptake of the ¹²³ I-FP-CIT tracer on SPECT reflects dopamine deficiency. An abnormal ¹²³ I-FP-CIT SPECT is typically present in patients with disorders associated with dopamine deficiency such as Lewy body disease.

Hyperechogenicity/transcranial sonography: Transcranial sonography focused on the brainstem can demonstrate midbrain hyperechogenicity, a finding thought to reflect increased iron content in the substantia nigra. Hyperechogenicityin the midbrain on transcranial sonography is often present in patients with Lewy body disease.