A lesson from anatomy: Focal brain lesions causing REM sleep behavior disorder

The brain modulates muscle tone during sleep. Although the precise neuronal network that induces muscle atonia during REM sleep is not fully understood, it has been shown that the pontine tegmentum and the medial medulla are critical areas for its generation [1], [2], [3]. In animals, experimental lesions in the dorsolateral pontine tegmentum eliminate muscle atonia and induce what looks like dream-enacting behaviors during unequivocal REM sleep. The site and extent of the lesions in the pons determine the severity of the behaviors that the animal releases, ranging from prominent jerks and limb twitches to locomotion and attack behaviors [2]. Experimental lesions confined to the ventromedial medulla in animals also produce REM sleep without atonia associated with abnormal motor behaviors [3].

In humans, lack of muscle atonia and excessive phasic muscular activity during REM sleep are the cardinal neurophysiologic abnormalities of REM sleep behavior disorder (RBD), a parasomnia clinically characterized by dream-enacting behaviors and unpleasant dreams. This condition is very common in several neurodegenerative diseases such as multiple system atrophy, dementia with Lewy bodies and Parkinson’s disease [4]. In these diseases, neuronal loss is found in the lower brainstem and limbic system, two anatomic regions that modulate most of the components of REM sleep. It can be speculated that focal brain lesions damaging the key structures that modulate REM sleep can cause RBD. In fact, small ischemic lesions [5], [6], [7], [8], [9], hemorrhages from vascular malformations [10], [11], tumors [6], [12], demyelinating plaques [13], [14], [15] and inflammatory lesions [16], [17], [18] have been found in a few patients with RBD or status dissociatus (the most extreme form of RBD where wakefulness and sleep stages are not distinguishable). These reports have important implications for understanding the underlying mechanisms of RBD. In this issue of Sleep Medicine, Xi and Luning report a 68-year-old male with RBD of subacute onset in whom brain MRI disclosed a small lacunar infarct in the right pontine tegmentum [19].

One crucial question raised here and in other cases previously published [5], [6], [7], [8] is whether a lesion found in a neuroimaging study is the direct cause of RBD or if it is simply an incidental finding. We here propose the following criteria to determine when a focal brain structural lesion (e.g., vascular, inflammatory, tumoral, etc.) is the direct cause of RBD: (1) RBD onset should be temporally associated with the appearance of the brain lesion; (2) RBD onset should be coincident to the onset of other symptoms caused by the lesion if they do appear (e.g., oculomotor abnormalities, hypersomnia, limbic syndrome, etc.); (3) the lesion should be located in a brain area known to regulate REM sleep (e.g., mesopontine tegmentum, ventromedial medulla, amygdala, hypothalamus, etc.); (4) disappearance of the lesion when possible (e.g., by surgery in tumors [12] or by immunotherapy in multiple sclerosis and autoimmune mediated limbic encephalitis [16]) is associated with remission or improvement of the RBD-related nocturnal symptoms and polysomnographic abnormalities; and (5) RBD is not better explained by another current disorder (e.g., Parkinson’s disease), medication use or withdrawal.

Infratentorial and supratentorial small lacunar infarcts and nonspecific subcortical white matter lesions, similar to those reported in some patients with RBD [5], [6], [7], [8], [9], are frequent in asymptomatic people with advanced age, particularly in those with vascular risk factors. In one study, three RBD patients had small lacunar infarcts in the pons or in the medulla but all three had an unrelated neurodegenerative disorder known to be associated with RBD [8]. Similarly, in one patient from our series with isolated RBD of six years duration and normal neurologic examination, brain MRI showed mass effect on the left pons and medulla caused by vertebral artery dolicoectasy (Fig. 1). One would think that this finding was the cause of RBD, but the observations that the patient had hyposmia and the fact(s) that neuropsychological tests showed marked visuospatial and memory impairment (and that FP-CIT demonstrated decreased striatal dopamine transporter binding) suggest that RBD in this case could represent an early marker of an underlying evolving neurodegenerative disease. Moreover, in one of our RBD patients, brain CT demonstrated a giant thalamic-parieto-occipital arteriovenous malformation in the left hemisphere (Fig. 2). We could only conclude that this lesion was not causing RBD after seven years of clinical follow-up when we noticed that the patient gradually developed the classical symptoms of dementia with Lewy bodies (cognitive impairment, fluctuation of symptoms, delusion, hallucination(s) and akinetic-rigid syndrome responsive to levodopa).

View Large Image Download to PowerPoint

Fig. 1. Axial (A) and sagittal (B) brain MRI of a 63-year-old man with RBD showing vertebral artery dolicoectasy producing mass effect on the left and ventral pons and medulla. The patient had hyposmia, (decreased striatal dopamine transporter binding demonstrated by FP-CIT scan) and mild cognitive impairment suggesting that the vascular malformation was not the direct casue of RBD.

View Large Image Download to PowerPoint

Fig. 2. Axial brain CT showing a giant thalamic-parieto-occipital arteriovenous malformation in the left hemisphere of a 71-year-old man with RBD. After seven years of follow-up the patient developed dementia with Lewy bodies and no changes in RBD symptoms.

In the patient presented by Xi and Luning, certain aspects of the patient merit additional comments. The authors recommend performing brain MRI in subjects diagnosed with idiopathic RBD in order to exclude brainstem lesions. We think that a structural lesion in the brain should be ruled out by neuroimaging studies only when RBD onset is acute or subacute, additional new neurologic signs and symptoms are detected, and RBD appears in a young non-narcoleptic subject. The lesion described by these authors was confined to the dorsolateral pontine tegmentum involving the subcoeruleus region which has been implicated in the pathogenesis of RBD [1]. Although in most of the published cases RBD has been linked to a structural lesion confined to the pontine tegmentum [5], [6], [7], [8], [9], [10], [13], [14], [15], [18], some acute lesions in the pontine tegmentum may not cause RBD but other sleep abnormalities including peduncular hallucinosis [20] (Fig. 3) or marked reduction of REM sleep with physiological muscle atonia [21]. On the other hand, RBD may also result from a direct dysfunction of anatomic connections of the subcoeruleus nucleus (e.g., with limbic system, thalamus, hypothalamus, medial medulla, etc.). RBD has been reported to occur in neurological disorders involving amygdala [16] or anterior thalamus [22] with no apparent primary brainstem damage, as well as with lesions in amygdala and mesopontine tegmentum [17]. Narcolepsy, a condition associated with loss of hypocretinergic neurons in the hypothalamus, is also linked to RBD [23]. We recently reported an 81-year-old patient with RBD plus Wallenberg syndrome resulting from an acute hemorrhage from a cavernous hemangioma in the left medulla, probably damaging the magnocellularis nucleus and descending projections from the pons (Fig. 4) [11]. The lesion in the patient described by Xi and Luning spared the substantia nigra, suggesting that dopaminergic deficiency is not necessary to cause RBD. The lesion was located in the right side of the pons indicating that an unilateral pontine lesion by itself may be sufficient to cause RBD. The patient’s clinical RBD picture included not only abnormal motor behavior during sleep but also frightening dream recall. This suggests the occurrence of functional dysregulation of those supratentorial structures that modulate intense emotions and are anatomically connected with the pons such as the amygdala. A satisfactory response of clonazepam to both idiopathic and secondary RBD suggests that different forms of RBD share similar pathophysiologic mechanisms, perhaps involving GABAergic pathways.

View Large Image Download to PowerPoint

Fig. 3. Axial (A) and sagittal (B) brain MRI of a 56-year-old man showing an acute hemorrhage in the dorsomedial pontine tegmentum. The patient developed acute horizontal oculomotor palsy, (facial palsy) and peduncular hallucinosis. (video-polysomnography ruled out the presence of RBD).

View Large Image Download to PowerPoint

Fig. 4. Axial (A) and coronal (B) brain MRI of a 81-year-old man in whom RBD plus Wallenberg syndrome coincided temporarily with an acute hemorrhage from a cavernous hemangioma in the left medulla. Two years later the Wallenberg syndrome resolved but RBD symptoms persisted. Clonazepam ameliorated dramatically the RBD symptoms.