Abstract
Insomnia Insomniais a very frequent symptom, usually due to non organic brain diseases. In some organic brain disorders, however, sleep impairment occurs through a series of mechanisms: structures responsible for need for sleep are lesionned ; the biological clock doesn’t give the start for sleep; sleep networks responsible for inhibition of waking structures are not efficient; mechanisms carrying on sleep or responsible for waking stages are damaged.
In each case, examples of those brain disorders leading to insomnia (tumors, strokes, traumas, neurodegenerative disorders) are reviewed, focusing on the neuropathological description of structures involved in sleep network. When possible, clinicopathological correlates are suggested.
Introduction
Whatever the cause, insomnia is usually related to a stress-induced hyperarousal state thought to be mainly mediated by the limbic system and forebrain areas. This hyperarousal state leads to an over-stimulation of the hypothalamo-hypophyseal axis, inducing chronic insomnia. (Vgontzas A, 2001; Buckley R et al, 2008). No organic brain lesion is seen, but functional neuroimaging shows an increased brain metabolism associated with inability to get asleep. In contrast, it has been suggested that decreased activity in the prefrontal cortex may be related to sleep deprivation effect (Nofzinger EA et al, 2004).
In a few occurrences, however, insomnia is associated with identified brain lesions. This may occur when the hypothalamic suprachiasmatic nucleus doesn’t give the start for sleep, when waking structures are not inhibited by sleeping structures, when thalamocortical bundles that carry on sleep are lesionned, or even when some waking structures are impaired. In addition, it may be due to a dysfunction of structures responsible for need for sleep.
Some strategically located brain tumours, strokes, and consequences of brain traumas are characterized by insomnia. Several neurodegenerative diseases such as Steele-Richardson, Alzheimer, Parkinson or Huntington diseases may also lead to insomnia. A series of different disorders, Fatal insomnia, rare cases of Creutzfeldt-Jakob disease, Morvan’s chorea, limbic encephalitis and delirium tremens, often associated with severe insomnia, have been grouped together under the name "agrypnia excitata"(Montagna P and Lugaresi E, 2002).
Most of the time, on the other hand, when recognised, the lesions in these disorders are widespread, making definite correlations difficult. In addition, combined factors (organic, psychic, cognitive or therapeutic) are often associated. In some cases, however, the lesions involve various regions, some of which are responsible for sleep organisation, which makes clinico-pathological correlations possible. In addition, in a few selected cases, organic brain lesions are localized enough to throw light on the mechanism of associated sleep disorders.
The Biological Clock
Hypothalamic suprachiasmatic nucleus is supposed to start sleep. Cells of this "pacemaker" contain a series of clock genes with rhythmic expression. It is influenced by light stimulation from the retina through the retinohypothalamic tract (glutamatergic), and indirectly from the geniculohypothalamic tract (GABAergic) and projects widely (among other regions) towards the pineal gland which secretes melatonin. Circadian synthesis of melatonin in the pineal gland is part of the rhythmic controls set up by the suprachiasmatic nucleus. The neuronal inputs originate in the retina which project fibers to the suprachiamatic nucleus; then the signal passes through the paraventricular nucleus, follows the medial forebrain bundle, and ends in the intermediolateral cell column of the upper thoracic spinal cord, from where it projects to the superior cervical ganglion, which sympathetic neurons innervate the pineal gland (M0ller M and Baeres FM, 2002). The signal to the pineal gland is mediated by norepinephrine, which is inhibited by light. Melatonin is implicated in numerous physiological processes, including circadian rhythms, stress and reproduction, many of which are mediated by the hypothalamus and pituitary gland. Melatonin receptors MT1 immunoreactivity is seen in the suprachiasmatic nucleus, paraventricular nucleus, periventricular nucleus, supraoptic nucleus, sexually dimorphic nucleus, the nucleus of the diagonal band of Broca and the nucleus basalis of Meynert, infundibular, tuberomamillary nuclei and the mamillary body, thalamic ventromedial, dorsomedial nuclei and paraventricular nuclei (Wu YH et al, 2006).
The suprachiasmatic nucleus, optic tracts, retinohypothalamic pineal gland and MT1 receptors are crucial to induce sleep. Lesions of this circuit may induce insomnia characterized by delayed sleep onset, eventually leading to day-night reversal :
a) Tumors of the anterior hypothalamus occur more frequently in children. The anterior hypothalamic syndrome consists of insomnia and loss of thirst regulatory mechanisms (Carmel PW 1980). In occasional larger lesions which interrupt the output from the supraoptic and paraventricular nuclei, diabetes insipidus is noticed. Tumors or cyst of the pineal gland may also lead to delayed sleep onset (Etzioni A et al, 1996; Taraszweska A et al, 2008). A 17-year-old boy, presented with severe headaches, associated with insomnia. Preoperative MR imaging suggested tumour of the pineal gland. Histopathological and immunohistochemical examination of the specimens from this surgical case revealed a characteristic pattern of cystic structures within the pineal gland, surrounded by layers of a dense fibrillar glial tissue and pineal parenchyma, consistent with non-neoplastic glial cysts (Taraszweska A et al, 2008).
b) Delayed sleep phase syndrome is also frequently associated to brain injury. It probably results from alterations either of the suprachiasmatic nucleus, or of its connecting fibers (Quinto C et al, 2000), which could imply basal contusion or diffuse axonal injury (Wang JY et al, 2008 ; Kumar R et al, 2009)
c) In the "normal old"aged persons, who pathologically may correspond to pre-clinical stages I and II of Braak’s classification of abnormal tau-associated lesions of Alzheimer disease (Braak H and Braak E, 1991), sleep initiation can be much delayed. In Alzheimer disease (AD), sleep disorders consist mainly in hypersomnia at the onset of the disease. Then, insomnia and inversion of circadian rhythm take place, the patient wandering during the night and sleeping during the day.
Melatonin, which inhibits neurons mediated by receptor MT1, may regulate neuronal sensitivity to stimuli that set biological clock in phase with environmental pressures (Wu YH et al, 2006). Melatonin has been reported to be low in old persons, but much variations were found: some have high levels, others have low levels and a few have even no evidence of melatonin secretion, whatever the differences in self reported quality of sleep (Baskett et al, 2001). It may be emphasized, however, that these data have been found in aged persons in the absence of neuropathological data. Melatonin levels in the cerebrospinal fluid decrease during the progression of AD associated tau neuropathology, tightly linked to neuronal loss (as assessed by the Braak’s stages), and this is already observed in cognitively intact subjects with the earliest AD neuropathology (Braak stages I-II), the so-called preclinical AD (Wu YH et al, 2003). It has been shown that dysfunction of the sympathetic regulation of pineal melatonin synthesis by the suprachiasmatic nucleus is responsible for melatonin changes during the early AD stages (Wu YH, 2005).
In neuropathologicaly assessed Alzheimer’s disease, two main lesions (already found in some aging "normal" brain) are seen: essentially extracellular deposits of Abeta (diffuse and focal deposits, including the core of senile plaques) and mostly intraneuronal inclusions of hyperphosphorylated tau, including neurofibrillary tangles, neuropil threads and the corona of senile plaques. The latter are highly linked to neuronal loss (Duyckaerts C, 2008 ; Dixon DW, 2008). The regions implicated in circadian rhythms (hypothalamus and pineal gland) have been thoroughly studied in AD by Dick Swaab’s group (2004) using early postmortem samples from the Amsterdam tissue bank. In 68 pineal glands, some clock genes (hBmal1, hCry1and hPer1) were found rhythmically expressed in aged control cases (stage 0 of Braak), and there were some correlations between hPerl expression and adrenergic beta1 (hbeta1-ADR) receptors. These data were linked to the hour of death. In AD (late stages V-VI of Braak), and even at pre-clinical stages I-II, ("old aged"persons), rhythmic expression of clock genes was no longer found and the correlations disappeared. Surprisingly, hCryl mRNA was increased in AD. The authors suggest that brain control upon pineal gland is lost, in reference to similar results obtained in the pineal gland of de-afferented rats (Wu YH et al, 2006). These data may be match up to the reduction of the hypothalamic neurons that express melatonin MT1 receptors, vasopressin and VIP (vasoactive intestinal peptide) at stages V-VI of Braak, whereas in aged controls (Braak’s stage 0) and at early stages I-II, the reduction of neurons concerns only those expressing melatonin receptors.
d) In Parkinson disease (PD), characterised by a synucleinopathy (Lewy bodies and Lewy neuritis) involving progressively some selected areas of the peripheral and the central nervous system (Hauw JJ et al, 2008 ; Braak H and Del Tredici K, 2008), as expanded later on, insomnia is the most common of the sleep disorders: 67% of patients have problem initiating sleep and 88% have sleep maintenance impairment (Factor SA et al, 1990). The number of neurons expressing melanoconcentrating hormone (MCH) is reduced from 12% at stage I to 74% at stage V, at the same time as their size increases significantly. Hypothalamic neuronal loss may play a more significant role in this disease than previously thought (Braak H and Del Tredici K, 2008).
e) Huntington disease (HD) is characterized by dementia, chorea and a dominant mode of inheritance. It is caused by an abnormal expansion of CAG repeats in the gene encoding Huntingtin. Sleep disturbances have been described in about 20% of patients with late-onset HD. Insomnia with impaired initiation and maintenance of sleep is a common complaint particularly in moderate to severe cases. With progression of the disease, sleep fragmentation increases (Myers RH et al, 1985). Sleep records used to show increased spindle density (Emser W et al, 1988). More recent sleep records showed low sleep efficiency, increased stage 1, delayed and shortened REM sleep and increased periodic leg movements (Arnulf I et al, 2008). On cerebral imagery, global cerebral atrophy did not significantly correlate with sleep parameters while atrophy of caudate nuclei was associated with reduced slow wave sleep and increased periods of awakeness (Wiegand M et al, 1991).
Hypothalamic atrophy occurs at early stages, associated with loss of orexin (hypocretin) and somatostatin-containing cell populations. Endocrine changes including increased cortisol levels, reduced testosterone levels and increased prevalence of diabetes are also found in HD patients. In HD mice, alterations in the hypothalamic-pituitary-adrenal axis occurs, as well as pancreatic beta-cell and adipocyte dysfunction (Petersen A and Bjorkqvist M, 2006).
The Structures Inhibiting Waking Systems
Sleep is set on only when structures responsible for waking state are inhibited: the ventrolateral preoptic area, which is in charge of sleep, inhibits the lateral hypothalamus (orexin/hypocretin system), Meynert nucleus, posterior hypothalamus and periaqueductal grey matter (Mignot E et al, 2002; McGinty D and Szymusiak R, 2003; Swaab D, 2004; Saper C et al, 2005; Sakurai T, 2007).
Stress, which stimulates CRF and orexin/hypocretin system is the usual cause of lack of inhibition of the waking structures. Lesions of the ventrolateral preoptic area or its connecting fibers to the waking structures may also impair the setting of sleep and lead to initiation of sleep insomnia:
a) Tumors of the anterior hypothalamus that impair biological clock, may also impair ventrolateral preoptic areas, when they are large enough.
b) In Parkinson’s disease, ventrolateral preoptic areas is involved at an early stage (Braak H and Del Tredici K, 2008). That could account for the severe insomnia occurring in this disease.
c) Whipple disease is a rare infectious disease that typically infects the bowel. It is caused by a bacteria named Tropheryma whippelii. It can affect any system of the body. At times, brain is the only organ affected. Symptoms include diarrhea, intestinal bleeding, abdominal pain, loss of appetite, weight loss, fatigue, weakness, arthritis and neurological signs and symptoms. Patients could suffer from hypersomnia or insomnia. Hypothalamic manifestations occur in 31% of Whipple encephalopathies, including polydipsia, hyperphagia, change in libido and insomnia (Papadopoulou M et al, 2005). Cerebral lesions of Whipple disease are disseminated throughout hemispheric and brain stem grey matter much more than in the white matter. They are more prominent in the hypothalamo-hypophyseal region and the upper brain stem especially in the periaqueductal grey matter (Powers J and Rawe SI, 1979).
In a well-studied case of Whipple disease, progressive and almost complete sleep loss was the initial and predominant symptom. Polysomnographic recording over 24 h, and during several consecutive nights, showed a complete abolition of the sleep-wake cycle with nocturnal sleep duration of less than 15 min. Hypothalamic dysfunction occurred, with flattening of circadian rhythmicity of cortisol, TSH, GH and melatonin, and hypocretin was reduced in CSF. However, FDG-PET scan revealed hypermetabolic zones in cortical and subcortical areas, including the brainstem. In this case, hypothalamic dysfunction was probably not the only cause of insomnia (Vodeholzer U et al, 2002).
The Structures Carrying on Sleep
Slow Wave Sleep
After the two previous steps have been achieved, thalamus switches to sleep (Akire MT et al, 2000). Slow sleep is the product of synchronized rhythmic activity of the thalamic reticular nuclei projecting to the cortex (thalamocortical activity) and hyperpolarizing cortical layer V. Sleep spindle and delta waves are thought to be due to these thalamocortical networks (Steriade M et Timofeev I, 2003). The main neurotransmittor is GABA. Deactivation of the orbital, dorsolateral prefrontal and inferior parietal cortices occurs during slow wave sleep and REM sleep, as shown by PET study (Braun AR, 1997).
Various brain lesions may involve the thalamocortical network and induce insomnia: stroke, trauma, normal aging and Alzheimer disease, and prion infection, a) Insomnia is associated with acute stroke, whatever the location, and sleep duration is reduced proportionally to the size of the stroke (Muller C et al, 2002). When stroke involves the middle cerebral artery territory, stages 2 and 3 and spindle density decrease during the few days following stroke. Following a thalamic infarct involving postero-lateral, ventral postero-lateral, dorso-medial and centromedial nuclei (a mixture of reticular/intralaminar, limbic and specific sensory nuclei) a 48hour insomnia has been reported (Garrel S et al, 1966). Surprisingly, thalamic infarcts are seldom reported as a cause of insomnia, since thalamus plays an important role in generating sleep spindles. As thalamocortical activity is also involved in waking state, patients are frequently neither asleep nor totally awake. Actually, thalamic infarcts are more responsible for hypersomnia or altered states of consciousness than for insomnia (Castaigne P et al; 1981; Schmahmann JD, 2003). The clinical presentations of thalamic stroke is not specific to individual nuclei because even small, focal ischemic lesions are seldom confined within nuclear boundaries. For instance, there is nothing such as an "infarct of the reticular/intralaminar nuclei", that would be the best candidate to produce insomnia (De Girolami U et al, 2009).
b) Insomnia is a frequent consequence of brain injury. It is attributed to shearing mechanism that induces diffuse axonal injury, as already mentionned. Most of the time it is associated to depressive mood and is usually reversible (Fichtenberg NL et al, 2002). Very few cases of localised brain injury leading to persistent insomnia are available. A case of total insomnia during 6 days was reported after right thalamotomy in a Parkinsonian patient who had benefited from a left sided alcoholic thalamotomy in the past (Bricolo A,1967). The right lesion involved ventral lateral, ventral intermedial and anterior part of the ventral intermedial nuclei, medial center and the subthalamic area, the dento-rubro-thalamic bundle and the H1 Forel region. The left sided lesion involved the medioventral part of the lateral ventral nucleus, ventral intermedial nucleus and anterior part of the ventral intermedial, median center and dorsal subthalamic bundles.
c) In old persons, sleep is interrupted by frequent awakenings. There is age-dependent changes in sleep EEG, with a shift of power from the anterior towards the central regions (Landolt HP and Borbely AA, 2001). In Alzheimer’s disease, sleep records show reduced sleep efficiency, increased amounts of stage I and increased numbers of awakenings. The typical lesions of tauopathy in clinical AD involve limbic, then hemispheric cortex with associative areas being predominantly injured. A number of nuclei that give projections to the cortex (involving mainly cholinergic, serotoninergic and histaminergic systems), are also injured (Duyckaerts C and Dickson DW, 2003), although initial tauopathy occurs in the peri-rhinal and entorhinal areas in presymptomatic patients. As a matter of fact, the development of abnormally phosphorylated tau occurs typically as described by Braak (I and II: preclinical stages, thus including "normal aged people"; III and IV: limbic stages: Ammon horn involvement ; V and VI : isocortex involvement) (Braak H and Braak E, 1991). The ventro-medial amygdala (Tsuchiya K and Kosaka K, 1990), and the basal nucleus of Meynert are involved precociously (Sassin I et al, 2000). On the contrary, the striatum and the thalamus are usually spared, with the exception of thalamic limbic nuclei which are also affected early (Braak H and Braak E, 1991).
d) Prion diseases, or transmissible spongiform encephalopathies (Creutzfeldt-Jakob diseases, fatal insomnias, Gerstmann-Strausler-Scheinker disease…) are due to a misfolded isoform enriched in beta-sheet of a normal protein (PrP). We will call PrPres this abnormally conformed protein for it is partly resistant to proteinases. In addition, it is transmissible to some animal species, including human. In PrP knockout mice devoid of prion protein, there is an alteration in both circadian activity rhythms and patterns. This could imply that prion protein is strongly involved in sleep regulation (Tobler I et al, 1996). However, in humans affected by Prion disorders, sleep disturbances are related more to the localisation of lesions than to prion infection by itself. For instance, in fatal insomnias, insomnia is constant; in Creutzfeldt-Jakob diseases, sleep disorders are a variable feature and in Gerstman-Straussler-Scheinker disease, sleep is usually normal. These different phenotypes are likely related to different conformations of PrPres. The distinct localization of neuronal loss or systems that could explain the different phenotypes may be difficult to assess, however, unless specific markers (such as enzymes of neurotransmitter metabolism) allow numerating definite neuronal classes. The classical triad of lesions in prion diseases is not always easily recognisable: in contrast to spongiform change and gliosis involving both astrocytes and, at a lesser degree, microglial cells, neuronal loss is often difficult to detect in the early lesions. However, it is conspicuous in the severe lesions where it induces gross atrophy. Early, severe, and selective neuronal loss affects a subset of parvalbumin positive GABAergic inhibitory interneurons both in human and experimental prion diseases, except in FFI (Guentchev M et al., 1997; Budka H et al, 2003). Electron microscopy does not disclose any known infectious agents, but tubulovesicular structures, particles of unknown origin and chemical composition which seem to occur only in prion diseases (Liberski PP et al, 2008). These lesions are associated with various types of PrPres deposits seen by immunohistochemistry (Privat N et al, 2008) or immunobloting. Western blotting provides most sensitive, but less precise data.
Fatal insomnia (FI) is a rare prion disease which affects as well men as women, at an age ranging from 36 to 62 years (Lugaresi E et al, 1986 ; Manetto V et al, 1992 ; Krasnianski et al, 2008). Most FI have a genetic autosomic dominant aetiology (Montagna P et al, 1998), usually linked to a mutation at codon 178 (D178N) of the gene of prion protein (PRNP). Interestingly enough, in the very same family, fatal familial insomnia (FFI) phenotype usually occurs only when there is a methionine (M) on the mutated allele at the polymorphism methionine-valine (MV) of codon 129 of the PRNP. When a valine (V) is present on this allele, a genetic disease with Creutzfeldt-Jakob phenotype occurs. Variants of this scheme have been reported, however, including typical Creutzfeldt-Jakob disease, FFI, and what was thought to be an autosomal dominant cerebellar ataxia in the same family (McLean CA et al, 1997) and different involvement of thalami and frontal lobes on PET scan in half-brothers (Johnson MD et al, 1998), suggesting that other genetic or environmental factors may play a role in the disease process. Mutations at codon 200 (E200K-129M), classified usually as familial Creutzfeldt-Jakob disease (Parchi P et al., 2003) have also lead to FFI phenotypes in cases with homozygoty VV at codon 129 (Taratuto AL et al, 2002). In sporadic fatal insomnia (sFI), also called the MM2-T subtype of sporadic Creutzfeldt-Jakob, PRNP shows no mutation. MM homozygosity at codon 129 polymorphism is also found (Scaravilli F et al, 2000).
In FI, disease onset is characterized by subtle personality changes, indifference and unability to express emotions and feelings, hallucinations and depression (87%). Cognitive dysfunction (87%) is characterized by lack of attention and hypovigilance. Myoclonus (70%), ataxia, dysarthria (61%), dysphagia (13%), pyramidal (43%) and extra-pyramidal (35%) signs occur. Profuse vegetative signs including perspiration and salivation, tachycardia, hypertension, fever, impotence, are present at the onset or a few weeks later (83%) (Krasnianski A et al, 2008). Hypovigilance is often prominent. Peudohypersomnia could also happen with profound alteration in the sleep-wake cycle (Dauvilliers Y et al, 2004). However, the hallmark of the disease is sleep loss with inability to produce the physiological patterns of slow sleep and REM sleep (96%) as well as hormonal and vegetative circadian fluctuations.
Sleep records in 5 cases showed REM sleep reduction (100%), reduced efficiency of sleep (80%), deep sleep reduction (80%), periodic limb movements (60%), central apneas (40%) (Krasnianski A et al, 2008). Hypnograms of an italian case fluctuated between stage 1 and REM sleep (Provini F et al, 2008).
CT and MRI are normal except for slight atrophy of the cerebrum and the cerebellum in the most advanced cases. A mild hypometabolism initiating in the thalamus, then extending to the mesial frontal areas and the whole hemispheres and basal ganglia, could be observed as soon as 13 months before the disease onset in a longitudinal 18FDT-PET scan study performed in carriers of the D178N mutation of PRNP (Cortelli P et al, 2006). Magnetic resonance spectroscopy combined with the measurement of apparent diffusion coefficient of water in different brain areas were performed 4 days before death in a 55-year-old man with familial fatal insomnia linked to D178N mutation. They showed an increase of apparent diffusion coefficient of water, and a metabolic pattern indicating gliosis in the thalamus but not in the other brain regions studied (Hai’k S et al, 2008).
In FI, the neuropathological lesions include mainly gliosis and neuronal loss, due to apoptosis (Daurandeu A et al., 1998) which is difficult to detect, as mentioned, in the early lesions. It is associated with gross atrophy in the severe lesions only. Spongiform change is mild or lacking, except in cases with very long duration (Hirose K et al, 2006). Some gliosis of the hypothalamus and periaqueductal gray matter is a common finding. In contrast, there is a severe bilateral and symmetrical thalamic degeneration with marked gliosis, neuronal loss, and usually lack of spongiform change. Gross atrophy is seen in the most affected areas, i. e. the anteroventral, mediodorsal and pulvinar nuclei. Other thalamic nuclei are usually less damaged, but intralaminar nuclei can be severely injured, especially in long duration cases. A morphometric investigation showed, indeed, that most thalamic nuclei had severely degenerated in two patients with FFI. Associative and motor nuclei lost 90% neurons, and limbic-paralimbic, intralaminar and reticular nuclei lost 60%. Although PrP immunohistochemistry shows various demonstrable deposits of PrP, biochemical studies usually detect PrPres at highest levels in the thalamus, and at lower levels in the brain stem and limbic structures (Gambetti P et al, 2003 ; Sazaki K et al, 2005 ; Ironside JW and Head MW, 2008). Serotoninergic neurons are selectively decreased in the raphe nuclei (Wanschitz J et al, 2000).
It may be mentioned that in mutations at codon 200 (E200K-129M), which may also lead to FFI phenotype, the usual lesions are alike those of sporadic Creutzfeldt-Jakob of the MM1 type (Parchi P et al, 1999; Parchi P et al, 2003), affecting the cerebral cortex, striatum, medial thalamus, and cerebellum. There is severe gliosis and moderate to severe neuronal loss in the thalamic nuclei and the inferior olivary nuclei, and spongiform change is particularly evident in the medio-dorsal nucleus of the thalamus, with rare and coarse spongiform degeneration in the isocortex and the putamen in a case with severe insomnia. In a E200K-129M case of CJD with thalamic involvement, insomnia occurred as the initial symptom, and was severe and intractable. Polysomnographic studies showed an absence of deep sleep and of REM sleep. On neuropathological examination, there was major involvement of the thalamus and of the inferior olivary nucleus, as could be seen in D178N mutation. Spongiform changes were mild in the neocortex and not patent in the cerebellum (Taratuto AL et al, 2002). ).
In sporadic fatal insomnia (sFI), also called the MM2-T subtype of sporadic Creutzfeldt-Jakob, there is a marked atrophy of the thalamus and inferior olive, with mild lesions in other areas (Scaravilli F et al, 2000, Piao YS et al, 2005).
In summary, thalamus is always involved in FI and thalamic lesions appear responsible for disappearance of deep stages of sleep, delta waves and spindles. Degeneration of mediodorsal, anteroventral and pulvinar nuclei of the thalamus is prominent. The disorganization of most thalamic circuits is a condition necessary for the sleep-wake rhythm being affected (Macchi G et al, 1997). This could be confirmed when comparing with the mild hypometabolism initiating in the thalamus observed as soon as 13 months before the disease onset when sleep was normal (Cortelli P et al, 2006). Mediodorsal nucleus has important connections both with reticular nucleus of the thalamus and cerebral cortex. Its lesion, or the frequent lesions of the intralaminar nuclei, are thus likely responsible for disappearance of spindles and delta sleep. By contrast, REM sleep often (but not always) persists because the tegmentum pontis sublaterodorsal nucleus and precoereuleus area, ventrolateral periaqueductal grey matter and tegmentum latero-pontis are not always severely affected. Hypovigilance and sleepiness could depend on damage of other thalamic structures, i.e. reticular nuclei for instance.
In sporadic Creutzfeldt-Jakob disease (sCJD), PrPres seems to occur in the brain for no obvious reason ("natural stochastic conversion") and the disease manifests as a rapidly evolving dementia. A number of clinical criteria for suspicion of sCJD have been proposed (Brandel JP et al., 2000; Zerr I, 2008); every one includes myoclonus, cerebellar signs, pyramidal or/and extrapyramidal signs, and some akinetic mutism or lower motor neuron involvement. None of them, however, mention insomnia or other sleep alteration. Typical EEG (pseudoperiodic sharp waves) and detection of 14-3-3 protein in the CSF are also considered in some criteria.
Even if insomnia is rare, sleep disturbances are frequent, however, and night-time behaviour disorders have been noted in 37% of 30 cases of sCJD (Zerr I, 2008). Sleep disorders can sometimes be similar to those found in FFI (Landolt HP et al, 2006). These authors followed seven consecutive patients with definite sCJD who underwent a systematic assessment of sleep-wake disturbances, including clinical history, video-polysomnography, and actigraphy. Sleep-wake symptoms were observed in all patients, and were prominent in four of them. All patients had severe sleep EEG abnormalities with loss of sleep spindles, vertex sharp waves, K-complexes, and slow sleep, very low sleep efficiency, and virtual absence of REM sleep. The EEGs were similar to that of wakening state, at times interrupted by muscular atonia without the other characteristics of REM sleep.
Neuropathological examination (Budka H et al., 2003; Landolt HP et al., 2006; Ironside JW and Head MW, 2008) show that, macroscopically, there may be either no gross alteration or a diffuse cerebral and cerebellar atrophy. The lesions typical of sCJD are patent spongiform change, gliosis, neuronal loss, and deposits of PrP. In about 10% of cases, these deposits aggregate into amyloid plaques of the "kuru" type. Several regional involvements could occur, in relation to the type of mutation or, in sporadic cases, to the M/V distribution at polymorphic codon 129 of the PrP gene in combination with the migration pattern of PrPres on Western blotting:1 or 2 following the classification of Parchi P et al. (1999). Six main molecular subtypes can thus been recognized (Parchi P et al., 1999; Zerr I, 2008). Whatever either the prominent types of lesions (spongiform change, gliosis, marked neuronal loss, plaques, type of immunodeposit) or their localisation (cortical, basal ganglia, cerebellum, brain stem nuclei), no insomnia was expressly described (Zerr I, 2008), but insomnia is not a systematically noticed symptom.
It must be stressed upon that in Landolt HP et al (2006) patients with definite sCJD who underwent a systematic assessment of sleep-wake disturbances, and had severe sleep EEG abnormalities with loss of sleep spindles, very low sleep efficiency, and virtual absence of REM sleep, autopsy revealed prominent changes in cortical areas, and only mild changes in the thalamus.
A severe involvement of the thalamus can be observed in some cases of CJD (Nagashima T et al, 1999, Yamashita M et al, 2001). The MV2 subtype, where the thalamus is predominantly affected, is easily distinguished from FI, always MM at codon 129 polymorphism.
Studies concerning hypothalamus have mainly concerned pituitary infectivity. Hypothalamus per se is seldom mentioned.
In summary: In CJD, insomnia is frequent, but is not a constant symptom. It appears to be associated with involvement of the thalamus in some cases, of the cortex in others.
Variant CJD (Ironside JW and Head MW, 2008), iatrogenic CJD (Brown P et al, 2001) and Gerstman-Straussler-Scheinker disease (GSS) have quite different clinical and pathological characteristics, but in none of them insomnia is a salient feature even if variant CJD is characterised by severe thalamic lesions. A few exceptions are provided, however, for example by P102L GGS, where two sisters showed protracted awakenings, reduced sleep efficiency, brief daytime naps. Slow wave sleep and REM sleep were preserved (Provini F et al, 2008), or in a case of variant CJD (Limousin N et al, 2008).
e) In Progressive supranuclear palsy (PSP), insomnia is a frequent and important feature (Petit D et al., 2004; Sixel-Doring F et al., 2008; Hauw JJ et al., 2008). Polysomnographic studies show short total sleep time, lower sleep efficiency, drastic reduction in sleep spindles, atonic slow-wave sleep. Standard EEG shows slowing in the frontal lobes (Montplaisir J et al., 1997). Although described as normal in the seminal descriptions, the isocortex is commonly affected by lesions of glial cells and neurons predominating in the motor and temporal cortex (Hauw JJ and Agid Y, 2003; Dickson DW, 2008). Some degree of neuronal loss is seen in the striatum and thalamus, especially ventral anterior and lateral thalamic nuclei. Thalamic intra-laminar caudal nuclei are often involved (Henderson J et al, 2000).Other lesions will be described later on.
f) In Huntington disease (HD), there is usually a cortical atrophy in the frontal lobes, and significant neuronal loss has been found by morphometric studies, notably in layer VI of HD brains at stages 3 or 4 of the disease. These cells project principally to the thalamus, the claustrum and other regions of cerebral cortex. Layer V neurons have also be found in decreased densities (Hedreen JC et al, 1991). Using anti huntingtin and antipolyglutamin immunohistochemistry, neuronal intranuclear ubiquitinated inclusions can be detected long before the onset of symptoms. An increase of CD15 (carbohydrate epitope 3-fucosyl-N-acetyl-lactosamine)-positive astrocytes has been described in the lateral part of the nucleus basalis of Meynert (Morres SA et al, 1992). The most striking neuronal loss and astrocytic gliosis in the brains of HD occurs in the striatum, predominant in the caudate nucleus. However only 1-4% of striatal neurons in all grades of HD have huntingtin nuclear inclusions (Gutekunst CA et al, 1999). Medium spiny projection neurons containing GABA colocalized with substance P and especially with enkephalin degenerate first. The most severe lesions of the thalamus are seen in the centrum medianum. In seven cases, the ventrolateral thalamus was studied by quantitative cytometry. A selective 50% atrophy of microneurons (internuncial cells) was found while the macro neurons did not show significant atrophy (Dom R et al, 1976). There is also a loss of neurons of the substantia nigra and the subthalamicus nucleus.
In summary: In HD, insomnia seems to be related to the involvement of the thalamus and even more to the involvement of the layers V and VI of the cortex. The role of the caudate nuclei in the occurrence of sleep disorders in HD is unknown.
