32 Specific myoclonic syndromes

In this chapter we will deal with three specific myoclonic syndromes. They are separated out for they present particular problems of recognition, differential diagnosis, and investigation for the clinician. They are asterixis, progressive myoclonus ataxia, and progressive myoclonic epilepsy (the Ramsay Hunt syndrome).

Adams and Foley coined the term asterixis for the lapses of posture, or ‘flapping tremor’ of the outstretched arms, which they first described in hepatic encephalopathy (‘liver flap’) in 1949. They found that the bilateral jerky movements of the outstretched hands were caused by intermittent pauses in ongoing electromyographic (EMG) activity, rather than by excessive muscle contraction as had been supposed previously. They also noted that the postural lapses frequently were repetitive and almost rhythmical. They and others (Conn 1960, Lance and Adams 1963, Leavitt and Tyler 1964, Tyler and Leavitt 1965) soon recognized that asterixis occurred in a wide variety of conditions (Table 32.1). In particular, it is associated with the metabolic disturbance produced by organ failure (liver, kidneys, lungs, and heart), septicaemia, electrolyte disturbances (hyponatraemia, hypokalaemia, hypercalcaemia), and delirium tremens. In addition, a variety of drug intoxications can produce asterixis. In all these situations, asterixis invariably is accompanied by tremor and clouding of consciousness. Thus, the clinical picture is of the confused sick patient who shakes and flaps. The alterations in consciousness, tremor, and asterixis are manifestations of a toxic or metabolic encephalopathy, and the electroencephalogram (EEG) usually shows slow rhythms reflecting widespread central nervous system (CNS) dysfunction.

Asterixis is recognized clinically as a brief, involuntary lapse of posture with the arms outstretched and the wrists dorsiflexed (Fig. 32.1). The hand suddenly falls and then jerks back to its previous position. It is this jerk that resembles myoclonus (Ugawa et al. 1990). However, the preceding involuntary postural lapse is associated with a 50- to 200-ms silence in the ongoing EEG, followed by an EMG burst which restores the hand to its previous position (Fig. 32.2). This contrasts with other types of myoclonus in which the jerk is produced by an active burst of EMG activity (positive myoclonus). Young and Shahani (1976) introduced the term negative myoclonus to describe asterixis, so as to emphasize the EMG silent periods responsible for the abnormal movements, even though the latter looked myoclonic (see also Chapter 30 for discussion on negative myoclonus).

Such brief lapses in posture, associated with brief cessation of tonic muscle contraction, can be seen in any skeletal muscle engaged in maintaining a posture, for example in the neck, fingers, or wrists (flexors or extensors depending upon whether the hand is held palm down or up), hip flexors if the leg is elevated from the bed, and dorsiflexors of the feet. When marked, asterixis is easily recognized and can be separated from the associated tremor. The latter is a rhythmic oscillation on which are superimposed the intermittent, arrhythmic, postural lapses of asterixis. However, in the absence of such gross lapses it may be difficult to discern the asterixis against the background of tremor. In these circumstances surface EMG recordings may reveal brief, irregular, silent periods associated with small lapses of posture, which constitute the tremor (Artieda et al. 1992). The frequency of asterixis in toxic or metabolic encephalopathies may be as often as 5–6/second or as rare as several times a minute. Young and Shahani (1986) stressed that its presence can be determined by asking a patient to simply dorsiflex the index finger for 30 seconds while watching carefully.

Young and Shahani (1986) drew attention to the similarity of pathological asterixis in toxic or metabolic encephalopathies to the sudden body jerks that occur in normal people who are simply drowsy. The sudden fall of the head followed by a jerk back to the upright position occurs in all of us during lectures!

The negative myoclonus of asterixis may coexist with conventional positive myoclonus (Fig. 32.3). This was emphasized in the original description of post-anoxic action myoclonus by Lance and Adams (1963). In this condition, jerks on movement are produced not only by bursts of excessive myoclonic EMG activity but also by silent periods of asterixis in the ongoing EMG. It is often impossible to tell clinically whether individual jerks in such patients, or in others with action myoclonus due to different causes, are due to positive or negative myoclonus; only EMG recordings can make the differentiation. A dramatic example of the contribution of negative myoclonus in such cases is seen when they attempt to stand. Frequently, they bounce up and down on their legs as a result of repetitive postural lapses in antigravity muscles and subsequent corrective movement.

Apart from the many metabolic and toxic causes of myoclonus, drugs are frequently responsible. Anticonvulsants are often incriminated (see Table 32.1), including phenytoin, carbamazepine (Rittmannsberger et al. 1991, Rittmannsberger and Leblhuber 1992), valproate (Aguglia et al. 1995), and newer agents such as gabapentin (Jacob et al. 2000, Babiy et al. 2005). Young and Shahani (1973) drew attention to ‘phenytoin flap’. This may be a sign of anticonvulsant toxicity, with other evidence of encephalopathy (Murphy and Goldstein 1974). The bilateral asterixis disappears as anticonvulsant concentrations are reduced to normal levels. However, on occasions, anticonvulsant blood levels may be within the therapeutic range. There are even examples of unilateral asterixis induced by phenytoin (Young and Shahani 1973), when the condition may be mistaken for focal motor epilepsy. One patient described by these authors had had an acute paralysis of the left leg due to a stroke, followed by what were thought to be motor seizures in the left arm. Anticonvulsant treatment led to a worsening of the jerky arm movements which were then recognized to be those of asterixis.

This report by Young and Shahani (1973) highlighted the fact that while most cases of asterixis are associated with a generalized encephalopathy and are bilateral, occasionally there occur examples of unilateral asterixis associated with focal brain lesions. Leavitt and Tyler (1964) also mentioned patients with unilateral thalamic or parietal disorders. However, in these cases there was severe sensory loss which might have disturbed postural control. Young and Shahani (1973) emphasized that unilateral asterixis could follow a stroke without weakness or loss of sensation. Since then, many examples of unilateral asterixis have been described with focal brain lesions. Young and Shahani (1986) collected 42 examples of unilateral asterixis from the literature, due to the various lesions such as those shown in Table 32.1. In seven of their patients, the asterixis was only apparent while the patient was receiving phenytoin. They noted that asterixis may be seen as an accompaniment of cerebral vasospasm following a subarachnoid haemorrhage (in the absence of a demonstrable infarct), or prior to a cerebral infarct caused by carotid occlusions. Thus, asterixis may be a prodromal symptom for an ischaemic hemiparesis. However, once the hemiparesis has developed, it does not prevent the appearance of asterixis (Goldblatt and Griggs 1979). Asterixis does, however,

disappear in the presence of a complete hemiplegia. Focal brain lesions may cause bilateral asterixis (see Table 32.1).

Young and Shahani (1986) collected from the literature 32 single lesions that had produced asterixis. Localization was by CT scan in all but one case, who had autopsy-confirmed medial frontal infarction. Fifteen of these patients had lesions in the thalamus, nine of whom had undergone a ventrolateral thalamotomy. Such post-stereotactic asterixis appeared acutely and disappeared over several days or weeks, perhaps reflecting damage to the adjacent internal capsule. In total, 15 of the responsible lesions were in the thalamus, seven in the parietal lobe, six in the internal capsule, three in the medial frontal cortex, and one was a bilateral lesion in the rostral midbrain tegmentum (the latter had bilateral asterixis).

Rio et al. (1995) performed a prospective study in a general hospital in Barcelona, Spain, over 5 years and found 45 cases in which asterixis occurred in association with a focal cerebral lesion. The average age was 61.8 years (range 35–86 years) and in 82% the disorder was unilateral, while in the remainder it involved both sides. A vascular pathology was demonstrated in 96%, with 4% each resulting from subarachnoid or subdural haemorrhage and the remainder being split more or less equally between intracerebral bleeding and cerebral ischaemia. The non-vascular 4% consisted of a patient with primary cerebral lymphoma and another with a small tuberculous lesion. In those with unilateral asterixis the thalamus was the most frequently affected structure (45%), but in other patients the caudate, lenticular nucleus, frontal lobe, occipital lobe, parietal lobe, midbrain, and pons were the sites of the pathology. Bilateral asterixis resulted from subarachnoid haemorrhage, subdural haematoma, or pathologies involving the brainstem.

Tatu et al. (1996) reported 12 cases with unilateral asterixis caused by focal brain lesions and noted that in all cases the upper limb was involved and that in only two was the lower limb similarly affected. Vascular lesions were present in nine, abscesses in two, and a meningioma in one. Multiple lesions were found in five cases. As with other series, the thalamus was most frequently affected (seven cases), but the lenticular nucleus, frontal lobe, precentral regions, internal capsule, and cerebellum were involved in some patients.

Lee and Marsden (1994) analyzed reports of 62 cases of movement disorder associated with a focal lesion in the thalamus and/or subthalamic region, 18 of whom had asterixis. In 13 of these 18 cases the damage was confined to the thalamus, eight resulting from thalamotomy and five from stroke. The lesion was contralateral to the asterixis in all cases.

Although stroke appears to be the commonest focal pathology causing asterixis, this movement disorder is a relatively rare complication. In a study of acute and delayed hyperkinetic movement disorders in 2500 consecutive stroke patients, asterixis was only noted in two (Ghika-Schmid et al. 1997). In spite of this the same authors have claimed it can be found in 15% of patients with pure parietal strokes and have postulated that in this situation it might reflect loss of multiple sensory feedback to motor programmes (Ghika et al. 1998). Others have suggested that when asterixis occurs in association with stroke involving the primary motor cortex, it may be due to impairment of motor-command signals controlling postural tone (Nighoghossian et al. 1995).

The exact mechanism of asterixis is uncertain. The anatomical data surveyed above suggest that lesions in the thalamus, parietal lobe, internal capsule, and frontal cortex may be especially liable to produce the phenomenon. There is some evidence that it might arise as a result of discharge in the motor cortex. Marsden et al. (1982) noted that electrical stimulation of the human motor cortex produced not only positive EMG bursts to cause twitches of activated muscles, but occasionally silent periods in voluntary contracting muscles, associated with postural lapses. Transcranial magnetic stimulation of the motor cortex can have a similar effect [see Hallett (1995) for a review]. Electrical stimulation of the human internal capsule has also been reported (Pagni et al. 1964) to produce not only short-latency EMG activation but also 50 to 200 ms periods of EMG silence during postural voluntary contraction. This suggests that there might exist inhibitory corticomotor neuron pathways which could be responsible for asterixis.

Early attempts to identify a cortical correlate prior to the EMG silences responsible for asterixis were unsuccessful. However, Ugawa et al. (1989 and 1990) were able to demonstrate a cortical event prior to typical EMG silences in three out of 11 patients with metabolic or toxic encephalopathy, using back-averaging of the EEG. These three patients all had a brief increase in EEG activity immediately preceding the silent period, whereas this was not the case in the remaining eight. The authors considered that the sharp wave arose from the motor cortex. Artieda et al. (1992) also demonstrated a large biphasic wave from parietal and central electrodes preceding the EMG silence of asterixis, which in 70% of records was preceded by a small amplitude short burst of EMG activity. The topography of this positive EEG wave was similar to that of the P22 wave of the somatosensory evoked potential elicited by stimulating the contralateral median nerve. The P22 was abnormally enlarged and situated over the sensorimotor cortex. Magnetic stimulation of the cortex produced a normal motor response, but ongoing asterixis made it impossible to determine whether magnetic stimulation of the cortex produced silent periods. The authors concluded that either the abnormal EEG discharges, which originated in the region of the sensorimotor cortex and preceded the asterixis, recruited inhibitory corticospinal projections or the descending volley activated spinal inhibitory interneurons. Another possibility entertained, however, was that cortical motor neuron discharges producing the small EMG bursts were followed by a period of inhibition at the cortical or spinal level.

This evidence suggests that asterixis is generated by motor cortical discharge in the same way that positive myoclonus often is. This again emphasizes the close co-existence of both positive and negative myoclonus.

In contrast to the primary generalized epilepsies, the secondary generalized epilepsies (that is, generalized epilepsies associated with a static or progressive encephalopathy) remain difficult to classify and to treat (Berkovic et al. 1986[b]). Secondary generalized epilepsies account for approximately 9% of all patients with epilepsy seen in special clinics (Gastaut et al. 1975). Patients with progressive myoclonus epilepsy form a small but clinically discrete subgroup amongst those with secondary generalized epilepsy.

Progressive myoclonus epilepsy is characterized by severe myoclonus (spontaneous, action, and stimulus-sensitive induced), severe generalized tonic-clonic seizures, and progressive neurological decline, particularly dementia and ataxia (Berkovic et al. 1986[b]) (see Table 32.2). Progressive myoclonic ataxia (sometimes known as the Ramsay Hunt syndrome) is distinguished from

progressive myoclonus epilepsy by seizures being mild or absent and dementia being slight or non-existent (Table 32.3) (Fig. 32.4). Progressive myoclonic ataxia is dealt with later.

Watson and Denny-Brown (1953), from their own experience and from a review of the literature, concluded that progressive myoclonus epilepsy was not a disease in its own right but a symptom complex that could be produced by a variety of diffuse neuronal insults to the CNS.

The clinical characteristics of the myoclonus in all the causes of the progressive myoclonus epilepsy syndrome are very similar. Myoclonus on action is typical, but many cases may have spontaneous myoclonus and some have stimulus-sensitive myoclonus, in response to somaesthetic stimuli, bright light, or noise (Table 32.4). The myoclonic jerks are not usually accompanied by loss of consciousness.

The myoclonic jerks must be distinguished from non-epileptic myoclonus (physiological or due to spinal cord lesions, brainstem, or startle syndromes) and other movement disorders associated with basal ganglia disease.

The next step is to distinguish progressive myoclonus epilepsy from other causes of generalized epileptic myoclonus (see Chapter 30) which are far more common. At the onset of the diseases causing progressive myoclonus epilepsy, the diagnostic triad of myoclonic attacks, tonic-clonic seizures, and progressive neurological decline may be incomplete. Accordingly, many patients with progressive myoclonus epilepsy are initially regarded as having more benign forms of primary generalized epileptic myoclonus, particularly juvenile myoclonic epilepsy of Janz (see Chapter 30). The beginning of progressive myoclonus epilepsy may even mimic more typical absence seizures due to primary generalized epilepsy (Andermann 1967). The diagnosis of progressive myoclonus epilepsy, however, becomes inescapable as increasingly severe seizures and progressive neurological signs emerge. Nevertheless, in those with primary generalized epilepsy, there may be concern if aggressive treatment with anticonvulsants causes ataxia, cognitive impairment, and worsening fits. This problem is particularly evident in those with the Lennox-Gastaut syndrome (see Chapter 30), where multiple forms of seizures including myoclonic seizure occur against a basically fixed neurological deficit but one that can appear progressive in the face of frequent seizures and/or anticonvulsant toxicity.

The differential diagnosis of progressive myoclonus epilepsy is accounted for by five uncommon conditions (Table 32.5) and a group of even rarer causes (Table 32.6) (see also Chapter 41 for a discussion of these). Amongst the uncommon conditions, Lafora body disease and neuronal ceroid lipofuscinosis generally produce severe neurological decline with dementia or regression and will be discussed in detail here. Unverricht-Lundborg disease, sialidosis, and mitochondrial encephalomyopathy tend to produce less dementia and present more with the picture of progressive myoclonic ataxia; they will be discussed below.

Lafora described inclusion bodies in a case of progressive myoclonic epilepsy in 1911 (Lafora 1911, Lafora and Glueck 1911). These are abundant in the brain, skeletal muscle, liver, and other tissues. The inclusions consist largely of polyglucosans. The disorder begins with seizures and these are followed by rapid progressive cognitive decline. The illness is inherited as an autosomal recessive disorder and can be caused by mutations in the laforin (EPM2A at 6q24) or the malin (NHLRC1; EPM2B at 6p22) gene; however, at least three genes underlie Lafora body disease (Chan et al. 2003) [also see Ganesh et al. (2006) for review]. The term EPM2 is as opposed to the more common EPM1 or Unverricht-Lundborg disease.

The laforin gene, mapped to 6q24, encodes a putative protein laforin phosphatase (tyrosine phosphatase), and DNA sequence variations and microdeletions have been reported (Serratosa et al. 1995, 1999, Minassian et al. 1998). A mouse model suggests that laforin is a glycogen phosphatase and deficiency leads to elevated phosphorylation of glycogen (Tagliabracci et al. 2007). The second gene, NHLRC1, encodes an E3 ubiquitin ligase, malin, which interacts with the laforin protein in vivo, as shown by Gentry et al. (2005) in HEK293T cells. Laforin is polyubiquitinated in a malin-dependent manner which leads to laforin degradation. The authors concluded that malin regulates laforin protein concentrations and that mutations in the NHLRC1 gene result in loss of the E3 ligase activity of malin. Ganesh et al. (2006) provide a detailed review of the molecular mechanisms.

Generally, the disease commences in adolescence, usually between the ages of 11 and 18 years, occasionally between 5 and 20 years (Van Heycop ten Ham and de Jager 1963, Janeway et al. 1967), but rarely younger or older than this (Kaufman et al. 1993, Footitt et al. 1997).

Seitelberger et al. (1964) collected 33 autopsy-proven cases. In the classic form, the first symptom usually is a generalized seizure, followed by absences and drop attacks (Rapin 1986). Sometimes behavioural change or school failure may be the initial presentation (Van Heycop ten Ham and de Jager 1963). Characteristically, the seizures are prefaced by a visual aura of flashing or coloured lights and such typical focal occipital seizures occur in about a half of cases (Roger et al. 1983, Tinuper et al. 1983). Myoclonus appears some months after the onset, initially occasional and of relatively small amplitude, but becoming increasingly severe and frequent, sometimes building up into a generalized seizure (Rapin 1986). As the disease progresses, the myoclonus becomes almost constant and grossly incapacitating.

Regression and rapidly progressive dementia develop usually within months but may be delayed for up to 1–2 years from the onset of the seizures. Behavioural disturbances, agitation and delusions, and executive dysfunction suggesting involvement of frontal areas occur. Vision deteriorates markedly despite normal fundi. Motor signs are relatively inconspicuous until late in the disease. Eventually, over a period of years, the patient becomes totally disabled, mute, and dies after a variable period of vegetative existence (Rapin 1986).

There are also atypical cases. Among those with EPM2A mutations, childhood-onset is characterized by dyslexia and learning disorder, which is then followed by epilepsy and neurologic deterioration, and this is associated mainly with mutations in exon 1, of the lafora gene as based on an analysis by Ganesh et al. (2002) of 22 patients from 14 families. Rarely the illness may begin in early adult life, with a milder more protracted course (Kraus-Ruppert et al. 1970, Diebold 1972).

Death occurs within 2–10 years after onset, on average at the age of about 20 years (between 16 and 24 years). There is suggestion that EPM2A-associated Lafora body disease may have a more severe clinical course compared to the EPM2B variant. (Gomez-Abad et al. 2005, Singh et al. 2006). A more prolonged course may be seen in adult-onset cases (Footitt et al. 1997).

The EEG may show focal or multifocal posterior epileptiform discharges, in addition to generalized discharges and progressive slowing of background rhythms (Tassinari et al. 1978, Ponsford et al. 1993, Striano et al. 2008) (Fig. 32.5). There are frequent runs of 2.5–3.5 Hz slow waves and sporadic spike and sharp waves. The spikes may not appear temporarily related to the myoclonic jerks (Roger et al. 1965, Janeway et al. 1967, King 1975). Patients are photosensitive, with high flash rates producing multiple spike and spike-wave complexes, associated with myoclonus, and sometimes precipitating seizures. The induced occipital waves seem to rapidly spread forwards and activate the motor corticies, followed by rapid conduction to the limbs, probably via corticospinal pathways (Rubboli et al. 1999). Sound, touch, and stretch rarely provoke myoclonus, but it does occur on voluntary action.

The diagnosis is established by the finding of characteristic inclusions, which stain positively with the periodic acid-Schiff reaction. These inclusions are present in many tissues, including the brain (liver, skeletal and cardiac muscle, and skin. Their distribution in the brain (de Ajuriaguerra et al. 1954, Seitelberger et al. 1964) shows greatest density in the substantia nigra, dentate nucleus, superior olive, pontine reticular nuclei, thalamic nuclei, lateral geniculate body, globus pallidus, and sensorimotor cortex. However, most areas of the brain are affected to some degree. Neuronal loss is not a striking feature of the illness, which produces relatively little brain atrophy.

Skin biopsy of course is less invasive as a diagnostic procedure (Carpenter and Karpati 1981). The characteristic inclusions are found in eccrine sweat-gland duct cells. They consist largely of polyglucosans, with a variable small component of phosphate and sulphate groups (Robitaille et al. 1980). DNA analysis provides definitive evidence of the underlying defect.

Proton magnetic resonance spectroscopy has shown reduction of N-acetyl aspartate in the grey matter, mostly in frontal regions, and in the white matter, frontally and parietally (Pichiecchio et al. 2008).

There is no curative treatment for this rapidly progressive disease. Initially, generalized seizures may respond to anticonvulsant drugs. However, as the fits become more severe they become less responsive, and the myoclonus is very difficult to treat. Some benefit may be obtained from the use of clonazepam and sodium valproate.

(See also Chapter 41).

This group of illnesses, inherited as autosomal recessive disorders and associated with the names of Batten and ‘familial amaurotic idiocy’, is characterized by retinal degeneration, dementia, seizures, myoclonus, and a variety of motor deficits. They have in common the storage in lysosomes of neurons and other tissues of autofluorescent lipopigments resembling ceroid and lipofuscin. The storage material contains excessive amounts of long-chain polyisoprenyl alcohols (dolichols). At least four sub-types were defined on the basis of age of onset, clinical signs, and histological appearance of the intracellular accumulations (Zeman et al. 1970), including infantile, late infantile, juvenile, and adult forms (Table 32.7), although it was recognized there was clinical overlap. Genetic studies have largely supported this division and at least ten forms and genetic loci have been described (Table 32.8); some of these will be discussed in the following sections.

The storage material can be divided into two types, sphingolipid activator proteins (SAP) and mitochondrial ATP synthase subunit c. The former results in a granular osmiophilic appearance, is found in the earliest onset cases, and is due to a mutation of the CLN1 gene, which controls production of the lysosomal enzyme palmitoyl-protein thioesterase (PPT). Occasionally mutations of this gene with similar ultrastructural inclusions can cause late infantile and juvenile onset of the disease. Storage of subunit c, which can arise from defects of at least six other genes, is responsible for the inclusions in other types of neuronal ceroid lipofuscinosis. Storage of this material can lead to other histological appearances or profiles, including curvilinear, fingerprint, and rectilinear.

The gene that causes the classical late infantile form (CLN2) encodes for a lysosomal enzyme tripeptidyl peptidase 1 (TPP1). Other late infantile forms result from mutation of other genes (see Table 32.8).

The classical juvenile form is caused by mutation of the CLN3 gene, which is thought to encode a membrane protein, possibly lysosomal. Mutations of PPT or TPP1, which are characteristic of childhood or late infantile onset disease, can also occur in a milder form and cause a variance of juvenile onset neuronal ceroid lipofuscinosis. The gene responsible for Kufs’ disease has been designated CLN4. The basic defect in the neuronal ceroid lipofuscinoses appears to be one of lysosomal dysfunction (Mole et al. 1999), resulting in impairment of degradation of post-transitionally modified proteins (Dawson and Cho 2000). While some of these disorders may result from specific genetically encoded defects in the degrading peptidases (Vines and Warburton 1999), others might be caused by factors such as an alteration in lysosomal pH which might alter the protease activity (Pearce et al. 1999). Abnormalities of mitochondrial energy production may also be a common feature of a number of these disorders (Das et al. 1999). As a group, neuronal ceroid lipofuscinoses are the commonest neurodegenerative disease of children, adolescents, and young adults. In Europe and the United States of America, the disorder has been estimated to affect 1:12,500–100,000 people (Uvebrant and Hagberg 1997, Mole 1999), with late infantile and juvenile cases generally being more common. Distribution in Japan seems similar (Oishi et al. 1999).

The infantile form, which can result from a great number of different mutations of the gene coding for palmitoyl-protein thioesterase-1 (PPT) (Salonen 2000), will not be discussed in detail here since it does not present as a progressive myoclonus epilepsy syndrome (Berkovic et al. 1986[b]). It presents between 8 and 18 months with progressive loss of motor milestones, hypotonia, myoclonus, ataxia, dementia, and visual failure. Within 3–4 years the children are in a vegetative state, in which they may have stimulus-sensitive myoclonus (Santavuori et al. 1973).

In about 80% of late infantile onset neuronal ceroid lipofuscinosis the defective gene is CLN2 and the activity of lysosomal TPP1 is abnormally reduced (Sleat et al. 1999). Onset, which is between the ages of 2.5 and 4 years, is usually characterized by seizures (Nardocci et al. 1995). These include tonic-clonic seizures, atonic attacks, and atypical absences. Within a few months, stimulus-sensitive myoclonus becomes prominent; this becomes progressively more severe and frequent, and occurs spontaneously and in response to photic stimulation. As the severity of the myoclonus increases, it becomes increasingly disabling and continuous, often building up to culminate in generalized seizures. Intellectual deterioration is an early sign, followed by loss of vision, with a granular mahogany appearance of the macula, which is not a cherry-red spot, and optic atrophy. The children become ataxic, spastic, and survive for a few years in a vegetative decorticate state. Age at death is usually 6–10 years (Rapin 1986).

The EEG is disorganized early in the disease with slowing of background rhythms, spikes, sharp-waves, and polyphasic spikes (Fig. 32.6). Myoclonic jerks are sometimes but not always preceded by a spike or spike-wave discharge. The characteristic feature is the response to slow photic stimulation; flicker slower than 3 Hz triggers high-amplitude polyphasic spikes usually in posterior leads; these have the appearance of a giant visual evoked response, and are often associated with repetitive myoclonic jerks. The electroretinogram is abnormal and becomes increasingly small in size until it disappears. The visual evoked response, in contrast, is of high amplitude, as are sensory evoked potentials (Pampiglione and Harden 1973, Westmoreland et al. 1979). Central conduction time is frequently prolonged (Schmitt et al. 1994).

MRI abnormalities in the brain appear at an early stage with cerebellar shrinkage being followed by cerebral atrophy. High signal develops in the periventricular white matter on T2-weighted images, which correlates with demyelination and gliosis (Autti et al. 1992, Petersen et al. 1996, Tyynela et al. 1997, Seitz et al. 1998). The thalamus may show hypointensity on T2-weighted images in some variants. Magnetic resonance spectroscopy is abnormal with reduced N-acetylaspartate and elevated lactate in grey and white matter, along with increased creatinine and choline-containing compounds in white matter (Brockmann et al. 1996, Autti et al. 1997). SPECT Tc-99m-Hexamethylpropyleneamine Oxime (HMPAO) brain scan shows reduced cerebral blood flow (Liewendahl et al. 1997), while PET with [18F]fluoro-D-glucose reveals widespread hypometabolism which sometimes seems particularly severe in the thalamus (Philippart et al. 1997).

Variants of infantile ceroid lipofuscinoses have been described and include CLN5 (Finnish variant), CLN6 (Costa Rica variant) and CLN7 (Turkish variant) of these, CLN5 is characterized by mixed combinations of ‘granular,’ ‘curvilinear’, and ‘fingerprint’ profiles. The clinical course includes progressive dementia, seizures, and progressive visual failure (Savukoski et al. 1998, Mole et al. 2005).

CLN9 refers to the juvenile-onset variant described by Schulz et al. (2004) in two Serbian sisters and two German brothers. The Serbian sisters developed declining vision, progressive ataxia, and seizures by age 4 years and immobility and mustism by age 10. The German brothers showed a similar course with onset at age 4 years with declining vision and seizures. Cognitive decline (age 6 years), ataxia and rigidity (age 9), and mutism (age 12) also occurred. Following pneumonia the younger brother died at age 15 years. The older brother, who later developed hallucinations and dysphagia, died at age 19 years. EEG was slow in all patients with polyspike wave discharges. Neurons were ballooned with autofluorescent fine granular material. A brain biopsy showed neurons with granular and curvilinear bodies. Fibroblasts from all patients were small and rounded with prominent nucleoli, attached poorly, and were highly apoptotic.

In Caucasians 80–85% of cases result from a 1.02 kb genomic deletion in the CLN3 gene (Haskell et al. 2000), although this percentage may be less in other racial groups (Oishi et al. 1999). This mutation results in truncation of the protein by a premature stop codon. While the majority of cases of typical juvenile neuronal ceroid lipofuscinosis are homozygous for this deletion, in a small number of families the abnormality is only found on one chromosome. Similarly, a minority of patients with the atypical variant of the disease, based on clinical and pathological features, are heterozygous for this deletion (Wisnieski et al. 1998). Missense and nonsense mutations of the CLN3 gene have also been described.

The juvenile type begins between the ages of 4 and 10 and commences with visual failure (juvenile amaurotic idiocy) (Lou and Kristensen 1973, Sorensen and Parnas 1979). The retinal findings are of macular degeneration with pigmentary changes (Fig. 32.7). Visual loss progresses to blindness over the next few years. Intellectual decline is often present from onset and seizures usually appear within 2–4 years (Rapin 1986, Nardocci et al. 1995). Most of the seizures are generalized, and myoclonus appears relatively late in the course of the illness. Many of the children attend a school for the blind until their teens, when increasing motor disability becomes disabling. They develop a stooped posture and accelerated stuttering speech which are pathognomonic of the disease (Rapin 1986). Marked extrapyramidal signs appear with an expressionless face, increasingly severe rigidity, bradykinesia, apraxia, and inability to walk. Seizures become more frequent, as does motor disability and dementia. Death usually occurs in the late teens or early 20s.

The EEG is slow and disorganized with sharp waves, spikes, and spike-wave complexes. Sensory evoked potentials may be enlarged and central conduction time slowed, as in the late infantile-onset variety (Schmitt et al. 1994). Similar changes can be seen in somatosensory evoked magnetic fields over the primary sensorimotor cortex, along with changes in peak latencies (Lauronen et al. 1997). The electroretinogram is progressively lost, and photic driving produces neither behavioural nor EEG responses (Green 1971).

MRI brain scan appearances are generally similar to those in the late infantile-onset form but appear later and are milder. Under 11 years scans often appear normal on visual inspection. Cerebral and cerebellar atrophy is mainly apparent after 14 years of age. On T2-weighted images periventricular white matter changes may be apparent. Diminished thalamic and basal ganglia signal intensity may be seen earlier in the course of the disease, especially in the former structure (Autti et al. 1996, Jarvela et al. 1997, Aberg et al. 2000). Blood flow studies using Tc-99m-HMPAO SPECT show cerebral hypoperfusion, particularly in the temporal lobes (Launes et al. 1996, Liewendahl et al. 1997), while glucose metabolism on PET may reveal a pattern of reduction which spreads gradually from occipital cortex anteriorly, sparing the subcortical areas and brainstem (Philippart et al. 1994 and 1997). PET scanning using [18F] fluorodopa reveals reduced uptake in the putamen and caudate nucleus, which has a modest relation with the extrapyramidal features (Ruottinen et al. 1997). SPECT studies show a decrease in striatal dopamine transporter in the neostriatum, which also correlates with the parkinsonian signs (Aberg et al. 2000).

CLN8 refers to ‘northern epilepsy’, also known as progressive epilepsy with mental retardation (EPMR). It is caused by a Finnish founder mutation and presents between 5 and 10 years of age with frequent tonic-clonic seizures followed by progressive mental retardation. Visual loss is not a prominent feature of northern epilepsy, there is no myoclonus, and the clinical progression is slower.

This is the least common type, named after Hugo Friedrich Kufs, born 1871 (Fig. 32.8A). The various adult forms of neuronal ceroid lipofuscinosis are variable in their expression, but they usually do not have retinal degeneration. Berkovic et al. (1988) reviewed 118 cases published as Kufs’ disease but accepted only 50. Two phenotypes were evident: (1) progressive myoclonus epilepsy (29 cases), and (2) dementia with motor disturbances (21 cases). The progressive myoclonus epilepsy type presented with seizures between the ages of 11 and 50 years, behavioural change, or both. Progressive ataxia and dysarthria were common in the course of increasing epilepsy, myoclonus, and dementia. Visual impairment and pyramidal and extrapyramidal signs were absent or only seen terminally. Death occurred within about 9 years (range 3–47). The dementia type was not dominated by seizures but by prominent cognitive impairment, cerebellar ataxia, facial dyskinesias, and dystonia, with myoclonus occasionally. Onset was between 11 and 47 years and death occurred after some 11 years (range 7–25). Others have confirmed these two clinical pictures (Nardocci et al. 1995), although in occasional cases onset may be delayed until the seventh decade (Donnet et al. 1992, Gambardella et al. 1998).

Evoked potentials may by enlarged (Donnet et al. 1992). MRI brain scan may show diffuse cerebral and cerebellar atrophy and some patients have also had white matter changes, lesions in their cortical grey matter, and diminished signal intensity in the putamen (Augustine et al. 1993, Gille et al. 1995, Nardocci et al. 1995, Cottier et al. 1996).

The electrophysiological findings may point to a diagnosis of neuronal ceroid lipofuscinosis, particularly in the late infantile and juvenile forms. The EEG shows marked photosensitivity in both types and the electroretinogram degrades, with an absent B-wave in both forms. However, the visual evoked response is enlarged in the late infantile type but diminished in the juvenile type (Pampiglione and Harden 1977). Enlargement of the somatosensory evoked responses has been mentioned above.

Urinary sediment dolichol estimation may be valuable and is elevated in about 90% of late infantile cases and in 85% of patients with the juvenile type (Wolfe et al. 1983, 1986). Elevated levels may also be found in Kufs’ disease, but the sensitivity and specificity of the test in adults is not established. Unfortunately, false positive results are seen in 15% of healthy controls (Wolfe et al. 1986) and elevated levels may be found in some other adult degenerative dementias (Mandell et al. 1984).

The MRI changes in the brain are relatively late but may point to the underlying correct diagnosis. Studies of blood flow, glucose metabolism, and dopaminergic function in the brain are largely experimental.

The definitive diagnosis is established by the demonstration of characteristic inclusions on electromicroscopy of biopsy tissue. These inclusions are found reliably in neurons of the brain, appendix, rectal mucosa, and in eccrine secretory cells in skin; they are usually but not always found in skeletal muscle (Carpenter et al. 1977, Lake 1984). The storage material reacts with lipid stains and autofluoresces in ultraviolet light. On electron microscopy the inclusions are of various types, which are relatively specific for the classical varieties of the different clinical forms (Fig. 32.8B and C and Table 32.8). Curvilinear profiles are characteristic of the late infantile type, and they also occur in association with fingerprint profiles in late onset cases. Fingerprint profiles are typical of the juvenile and adult forms and may be the predominant finding in some late infantile cases. Granular osmiophilic deposits are invariably found in the infantile type but are also seen rarely in juvenile and adult cases. Although diagnosis in most reported cases of Kufs’ disease has been made on histology of the brain, characteristic deposits can also be seen in peripheral tissues (Gelot et al. 1998).

Genetic analysis is clearly of major importance in establishing the diagnosis in many cases and has a particular role to play in asymptomatic at-risk individuals and in antenatal diagnosis.

Treatment is largely symptomatic and includes the use of anticonvulsants to try to control seizures. Bone marrow transplantation has been attempted without convincing evidence of success (Lake et al. 1997).

(See also Chapter 41)

Gaucher's disease is caused by deficiency of the lysosomal enzyme glucocerebrosidase (GBA), the gene for which is located on chromosome 1q21. It has a highly homologous pseudogene situated 16 kb downstream. A large number of mutations of this gene can be responsible for the development of the disease. Gaucher's disease is among the most frequently inherited disorders of Ashkenazi Jews.

Classically, according to the presence of neurological symptoms and the dynamics of the neurological picture, Gaucher's disease is divided into three types: type 1 – the non-neuronopathic form; type 2 – the acute neuronopathic form; and type 3 – the subacute neuronopathic form. Of these, type 1 is typically not associated with development of neurological signs; however, there are reports discussing an association between type 1 Gaucher's disease and parkinsonism (not further discussed here; see Chapter 5 under ‘Hereditary’).

Similarly, there is a classification by age of onset. Infantile neuronopathic Gaucher's disease leads to death before the age of about 2 years and is not associated with myoclonus. Classical juvenile Gaucher's disease does not affect the brain. However, in the juvenile neuronopathic variant, many cases have severe myoclonus (spontaneous action and photosensitive), associated with a supranuclear gaze palsy involving saccadic horizontal eye movements but sparing pursuit and vertical movements (King 1975, Tripp et al. 1977, Nishimura and Barranger 1980, Winkelman et al. 1983). Dementia may or may not be evident. Most cases have other types of seizures and splenomegaly. The disease is compatible with survival to mid-adult life, although the myoclonus becomes so severe as to considerably disable patients and render them anarthric.

Genetic analysis in 16 patients with Gaucher's-associated myoclonic epilepsy, nine of whom were diagnosed by age 4 years with severe visceral involvement and myoclonus and seven had a more chronic course, has revealed different genotypes (Park et al. 2003). Yet there were several shared alleles, including V394L (seen on two alleles), G377S (seen on three alleles), and L444P, N188S, and recombinant alleles (each found on four alleles). V394L, G377S, and N188S are mutations that have also been associated with non-neuronopathic Gaucher's disease. The spectrum of genotypes differed significantly from other patients with type 3 Gaucher's disease, where genotypes L444P/L444P and R463C/null allele predominated. The authors concluded that both the lack of a shared genotype and the variability in clinical presentations suggest that other modifiers must contribute to this rare phenotype. Further studies in this direction were undertaken by Montford et al. (2004), but more work needs to be done.

Other investigations may reveal pancytopenia, elevated serum acid phosphatase levels, and a low leukocyte beta-glucocerebrosidase activity. Brain autopsy samples from two patients (Park et al. 2003) demonstrated elevated levels of glucosylsphingosine, a toxic glycolipid, which could contribute to the development of myoclonus.

Despite improvement in hepatosplenomegaly with intravenous glucocerebrosidase, improvement of neurological manifestations is negligible. Decreasing blocking synthesis of glucocerebroside offers possibilities (Platt and Butters 1998), but although such therapy also helps the non-neuronopathic form of Gaucher's disease (Cox et al. 2000) the effect on the nervous system is uncertain. Bone marrow transplantation, however, may be successful (Krivit et al. 1999).

(See also Chapter 41.)

Classical Tay-Sachs disease is due to hexosaminidase A deficiency and Sandhoff's disease to both hexosaminidase A and B deficiency, inherited as autosomal recessive traits. Infantile GM2 gangliosidosis is characterized by a typical startle reaction to sound, rather than the myoclonus under consideration here (see Chapter 31, Startle Syndromes). This exaggerated startle is recognized from birth. After about the age of 18 months, the increased sound-induced startle is often followed by focal or generalized clonic movements, and occasionally by generalized tonic-clonic seizures. There is progressive dementia, spasticity, and visual loss associated with a macular cherry-red spot. Death occurs before the age of 3 years.

Late infantile and juvenile forms of GM2 gangliosidosis may show myoclonus (spontaneous or stimulus-sensitive) or seizures as prominent early symptoms, but dementia, ataxia, spasticity, or dystonia then develop. In the late infantile and juvenile forms, cherry-red spots are typically absent. The diagnosis is made on the basis of low hexosaminidase activity in the serum, leukocytes, and cultured fibroblasts (Brett et al. 1973). Genetic testing provides a more definitive analysis of the underlying defect.

This illness may have features of progressive myoclonus epilepsy, with dementia and ataxia, along with chorea, dystonia, parkinsonism, and a neuropathy (Vuia 1976, Dorfman et al. 1978, Scheithauer et al. 1978). The diagnosis is established by demonstrating the presence of axonal spheroids, which may be evident in peripheral nerves or brain, and by electron microscopy of autonomic terminals around eccrine secretory cells. This form of neuroaxonal dystrophy shows similarities to pantothenate kinase-associated neurodegeneration (PKAN) (Hallervorden-Spatz disease) (see Chapter 11), and may not only show iron deposition in globus pallidus and substantia nigra, but also diffuse Lewy bodies in neurites, which are immunolabelled by anti-alpha-synuclein (Hayashi et al. 1992, Sugiyama et al. 1993, Wakabayashi et al. 2000).

While the infantile variant of neuroaxonal dystrophy has been found to be due to mutations in the PLA2G6 gene (Morgan et al. 2006), it is not certain whether the progressive myoclonus epilepsy phentyope with juvenile onset is due to the same or another gene.

Another very rare condition is atypical inclusion body disease (Dastur et al. 1966, Berkovic et al. 1986[a], Yerby et al. 1986) which may present with progressive myoclonus epilepsy. The inclusions differ from those of Lafora body disease.

The characteristic features of progressive myoclonic ataxia resemble those of progressive myoclonus epilepsy in some respects (spontaneous, action, and stimulus-induced myoclonus, with progressive ataxia) but differ in others (epilepsy usually is mild or easily controlled or may not exist at all; cognitive impairment also is mild and tends to be non-progressive) (Table 32.3). The term progressive myoclonic ataxia was introduced (Marseille Consensus Group 1990) in an attempt to overcome the confusion surrounding the Ramsay Hunt syndrome. The latter has caused vigorous debate (Andermann et al. 1989, Marsden and Obeso 1989).

In 1921, Ramsay Hunt described a group of six patients, four sporadic cases and a pair of twin boys, under the title ‘dyssynergia cerebellaris myoclonica – primary atrophy of the dentate system.’ The twins had a neurological syndrome resembling Friedreich's ataxia, with onset in early childhood; myoclonus and epilepsy subsequently developed in the third decade of life. One of these patients died at the age of 36 years; autopsy showed degeneration of the spinocerebellar tracts and posterior columns, and atrophy of the dentate nuclei and superior cerebellar peduncles. The four sporadic cases presented with myoclonus and epilepsy before the age of 20 years, and progressive cerebellar ataxia developed later. Ramsay Hunt (1921) drew attention to the combination of myoclonus and progressive cerebellar ataxia, which he believed arose as a result of degeneration of the dentate system. He did not claim that this was a specific disease.

Since then, many patients have been described under the term ‘Ramsay Hunt syndrome’, but much of the relevant literature is confusing. There are problems in defining the limits of the syndrome and in separating it from other conditions with similar clinical features. First, the presence of action myoclonus often makes it difficult to assess cerebellar function. Second, patients with long-standing idiopathic epilepsy and myoclonic seizures may develop ataxia after many years, perhaps related to anticonvulsant therapy and toxicity. Third, the Ramsay Hunt syndrome may be confused with that of progressive myoclonus epilepsy, although severe progressive dementia and severe seizures distinguish the latter. Fourth, many patients have been described as having Ramsay Hunt disease in the literature. Most of these were inadequately investigated to exclude other causes of the syndrome, or had Unverricht-Lundborg disease. Fifth, the distinction between myoclonic seizures and non-epileptic myoclonus is difficult, both clinically and electrophysiologically. A myoclonic fit is due to a pathological paroxysmal electrical discharge in the cerebral cortex, usually accompanied by loss of consciousness. However, spontaneous and action myoclonus in many of the cases of progressive myoclonic ataxia (and progressive myoclonus epilepsy) can also be shown electrophysiologically to be due to an abnormal discharge of cortical motor neurons (see Chapter 29, Cortical Myoclonus).

Accordingly, Berkovic et al. (1986[b]) came to the conclusion that ‘the term “Ramsay Hunt syndrome” has caused considerable confusion in the literature. It has no agreed definition, it does not represent a specific disease, and its use should now be abandoned’. They included patients with conditions under the rubric of the Ramsay Hunt syndrome as examples of progressive myoclonus epilepsy. However, Marsden and Obeso (1989) argued that many of those with progressive myoclonus and ataxia have only mild generalized epilepsy, and some have no tonic-clonic seizures at all. This makes it difficult to include such patients within the syndrome of progressive myoclonus epilepsy, particularly as dementia also is not a prominent feature of such cases. Hence, the new term ‘progressive myoclonic ataxia’ was born (Marseille Consensus Group 1990) to accommodate these patients.

The causes of progressive myoclonic ataxia (Table 32.9) overlap with those responsible for progressive myoclonus epilepsy. However, the majority of such patients are found to have Unverricht-Lundborg disease, a mitochondrial encephalomyopathy, sialidosis, or a spinocerebellar degeneration. These will be discussed in more detail here. Other rarer causes of progressive myoclonic ataxia will also be mentioned. Some idea of the frequency of the various causes for the syndrome of progressive myoclonic ataxia was obtained by Marsden et al. (1990) in a series of 30 patients (Table 32.10). This series did not include patients with proven sialidosis. In nine cases it was impossible to arrive at a firm pathological diagnosis.

Unverricht (1891) reported a family from Estonia of two boys and eight girls, half of whom were affected by myoclonic jerks and epilepsy with onset between the ages of 9 and 15 years. Subsequently, the same author (1895) described a family of five children, three of whom were affected by jerks and epilepsy beginning between the ages of 1–3 and 15 years. None of these patients was severely demented. Lundborg (1903), in an important review of the literature, produced a classification of myoclonus in which he divided cases into three groups: (1) symptomatic myoclonus; (2) essential myoclonus; and (3) familial myoclonic epilepsy. The latter he divided into two groups: (a) a non-progressive form, following the report of Rabot (1899) who described a family with jerks and myoclonus in whom the disorder was clearly not progressive (this may be an example of hereditary essential myoclonus); (b) a slowly progressive form as described by Unverricht. Lundborg himself reported 18 cases of myoclonus from 10 families in a peasant community in Sweden. Their illness was characterized by onset in late childhood or adolescence, no other neurological deficits besides myoclonus and epilepsy, long survival, and no progressive dementia. Siblings only were affected and consanguinity was evident, suggesting autosomal recessive inheritance.

Koskiniemi (1986), based upon her experience in Finland (Fig. 32.9), summarized Unverricht-Lundborg disease as being an illness inherited as an autosomal recessive trait, characterized by stimulus-sensitive myoclonic jerks, onset around the age of 6–15 years, generalized tonic-clonic seizures, a characteristic EEG, and a progressive course.

The illness is common in Scandinavia and related countries, which led Eldridge et al. (1983) to term the condition ‘Baltic myoclonus’. The incidence of the disease is estimated to exceed 1 in 20,000 of the population in Finland (Koskiniemi 1986). However, the disease is also recognized in countries world-wide, including North America (Eldridge et al. 1983, Berkovic et al. 1986[b]), the United Kingdom (Marsden et al. 1990), the rest of Europe and North Africa (Roger et al. 1968, Roger 1985, Tassinari et al. 1989, Gouider et al. 1998), and Japan (Marseille Consensus Group 1990). So-called Mediterranean myoclonus, which is relatively frequent

in the western Mediterranean region (Genton et al. 1990), is also phenotypically and genotypically identical (Lehesjoki et al. 1994).

Only a few cases of Unverricht-Lundborg disease have come to autopsy. Reported abnormalities have included a diffuse loss of Purkinje cells as the only consistent finding in the brain (Haltia et al. 1969, Koskiniemi et al. 1974[a]). It has been suggested, however, that this might be secondary to treatment with phenytoin (Eldridge et al. 1983). Membrane-bound vacuoles have been reported in eccrine cells in sweat glands (Cochius et al. 1994), but the significance of this is uncertain.

Unverricht-Lundborg disease, which has also been designated EPM1, results from a genetic abnormality located at 21q23.3. The gene encodes for cystatin B, an inhibitor of cysteine protease. The most common abnormality is an expansion mutation of a dodecamer repeat located in a non-coding region upstream of the transciption start site of the cystatin B gene and this accounts for over 90% of patients with this diease who have been tested world-wide (Lafreniere et al. 1997, Lalioti et al. 1997, Virtaneva et al. 1997). Virtually all patients have been shown to have at least one of the disease chromosomes and most are homozygous for this abnormality (Pennacchio et al. 1996, Lafreniere et al. 1997, Lalioti et al. 1997, Virtaneva et al. 1997). Normal alleles contain only two or three copies of this repeat, but affected individuals may have between 30 and 75 copies (Lalioti et al. 1998). Preliminary analysis has not shown any correlation between the size of the expansion and age of onset or clinical severity (Lalioti et al. 1998). In addition, several ‘minor’ mutations affecting one or two nucleotides in the gene have been reported (Lehesjoki and Koskiniemi 1999), but, as mentioned above, these account for less than 10% of reported cases. Abnormalities of the gene result in dramatically reduced levels of cytostatin B mRNA. Cytostatin B is thought to protect against intracellular degradation by proteases that leak from lysosomes and its reduced level may be the direct cause of the condition.

The expansion of the gene underlying most cases was the first case of instability of a repeat unit other than trinucleotides to be found in association with human disease, but, unlike the latter, intergenerational instability does not seem to be a prominent feature (Lehesjoki and Koskiniemi 1999).

The early development of these children is normal. The first symptom is either stimulus-sensitive myoclonic jerks or generalized tonic-clonic seizures, appearing between the ages of around 6 and 15 years (mean age of onset 10.8 years) (Koskiniemi et al. 1974[a], Koskiniemi 1986, Lehesjoki and Koskiniemi 1999). Initially, myoclonic jerks appear in the morning or when the patient is tired. They occur on movement or in response to external stimuli which may be visual, auditory, or somatosensory. Repetitive morning myoclonus is also typical, frequently building up and culminating in a clonic-tonic seizure (Koskiniemi et al. 1974[a], Norio and Koskiniemi 1979). The myoclonus gets progressively more frequent and severe. There is gradual deterioration of gait, speech, swallowing, and manual dexterity. Over the course of approximately 5 years, the myoclonic jerks incapacitate patients such that they need support in walking by the age of about 18 years and are eventually confined to bed. The mean age of death in earlier reports was about 24 years, approximately 14 years after the onset (Koskiniemi 1986). Advances in medical treatment have resulted in a better outlook. The frequency of myoclonus and generalized seizures may stabilize or even lessen with time (Lehesjoki and Koskiniemi 1999). Dystonia may also be present and oculomotor apraxia has been described (Chew et al. 2008).

On examination, dysarthria, and intention tremor are prominent features within 1–3 years of the onset of the disease. Pyramidal signs are rare, as is evident for a peripheral neuropathy and optic atrophy.

Emotional lability and depression are frequent. Patients’ intelligence is relatively unimpaired (Koskiniemi 1974, Koskiniemi et al. 1974[a]). The mean Wechsler Adult Intelligence Scale (WAIS) intelligence quotient at the onset of the disease is around 92; this diminishes about 10 points over 10 years. This cognitive impairment, which is mild, may to a large extent be due to the effect of myoclonus and immobility leading to little contact outside of the home and interruption of schooling, as well as the deleterious effects of anticonvulsant drugs (Eldridge et al. 1983). Rapid dementia is never seen (Koskiniemi 1986). Even in the final stages, patients are alert and are able to communicate, although they may have developed urinary and faecal incontinence.

Survival into adult life is usual and some patients have reached their sixth decade (Norio and Koskiniemi 1979, Leino et al. 1982). Chew et al. (2008), for example, studied the natural history of eight genetically proven cases and after a mean duration of disease of 30 years, four patients were walking with aids while another four were wheelchair bound. However, a more malignant course with death in adolescence has also been described. This may well be due, at least in part, to complications of the seizures, unrecognized anticonvulsant toxicity, and infections due to recumbency and aspiration (Eldridge et al. 1983).

The EEG is abnormal from the onset of the disease with decreased alpha activity and slow-waves; follow-up studies (Ferlazzo et al. 2007) suggest there is gradual reduction of changes over time, correlating with the good seizure outcome in this condition.

Generalized spike-wave paroxysms are seen early, and polyspike-wave paroxysms appear in 50% of patients (Koskiniemi et al.

1974[b] (Fig. 32.10). Photic sensitivity to flashes of light at around 4 Hz commonly leads to paroxysmal activity and may precipitate a generalized seizure. Intermittent photic stimulation at frequencies up to 12 Hz can evoke a reflex myoclonic jerk with a 1:1 relationship and back-averaging of EEG shows an occipital wave followed 10 ms later by an ipsilateral positive-negative transient in the central region. The subsequent myoclonic jerk is propagated to cranial and limb muscles in a rostro-caudal pattern with latencies suggesting transmission along the fast-conducting corticospinal motor pathways (Rubboli et al. 1999). The amplitude of visual and brainstem auditory evoked potentials shows little, if any, alteration, although latencies are significantly increased (Hari et al. 1983, Mervalla et al. 1986). By contrast ‘giant’ somatosensory evoked potentials are present (Shibasaki et al. 1985). It has been suggested that this results from hyperactivity in the thalamocortical system (Karhu et al. 1994). Cortical stimulation elicits not only an initial motor response in peripheral muscles but also a second one, the latency of the latter suggesting that it may be due to a long-latency or C response, involving afferent stimulation from the first motor response (Rubboli et al. 1990). Such a C reflex may also occur following peripheral nerve or nerve root stimulation, and a magnetic stimulation over the cerebral cortex timed to coincide with the arrival of this C reflex at the cortex shows marked facilitation, suggesting the cortex is rendered hyperexcitable with this somatosensory input (Reutens et al. 1993).

Brain imaging by CT and cerebral glucose metabolism measured by PET scan shows no abnormality (Lehesjoki and Koskiniemi 1999). However, there are abnormalities of dopaminergic function (dopamine D2-like receptor availability in both the striatum and the thalamus) as shown by 11C raclopride-PET. (Korja et al. 2007).

Routine examination of blood and urine is normal, as is the CSF. Unverricht (1895) reported an increased excretion of indican in the urine, which has also been commented upon by Koskiniemi and Palo (1978). Koskiniemi (1980) also reported a lower plasma concentration of tryptophan in such patients, and Leino et al. (1980) described reduced concentrations of 5-hydroxyindoleacetic acid and of vanillic mandelic acid in the CSF.

Clinical suspicion of Unverricht-Lundborg disease can be confirmed by genetic analysis. This also allows presymptomatic and prenatal diagnosis, as well as carrier detection.

There are no comparative trials of anticonvulsants in this disorder, but valproate is usually regarded as the drug of choice. It has been claimed that if started immediately after symptom onset it may delay or stop disease progression (Lehesjoki and Koskiniemi 1999). It also diminishes the excretion of indican in the urine (Koskiniemi and Palo 1978). Combination with clonazepam has been reported to be helpful (Goldgerg and Dorman 1976, Iivanainen and Himberg 1982). Zonismanide has been considered useful (Kyllerman and Ben-Menachem 1998). Piracetam also results in an improvement in myoclonus and reduction in functional disability (Koskiniemi et al. 1998). There is evidence to suggest that phenytoin may contribute to progressive disability (Eldridge et al. 1983) and should be avoided.

Interestingly, alcohol may decrease myoclonus and some patients may use it for social events (Genton and Guerrini et al. 1990). It has also been claimed that the antioxidant N-acetylcysteine, a sulfhydryl antioxidant, may decrease myoclonus (Hurd et al. 1996). Edwards et al. (2002) found variable responses to N-acetylcysteine. In one patient there was improvement in seizures but not in myoclonus or ataxia. Glutathione levels were low before treatment and increased during treatment. However, three other patients showed a variable response and some notable side effects occurred during treatment with N-acetylcysteine.

(See also Chapter 41.)

Diseases of muscle and the nervous system raise the suspicion of disturbance of mitochondrial function (Berenberg et al. 1977, DiMauro et al. 1985, Morgan-Hughes 1986, Petty et al. 1986, Harding 1991). Encephalopathy and myopathy may be intermingled in different proportions and result in three major syndromes (Table 32.11). A variety of biochemical abnormalities of mitochondrial function have been established (Table 32.12), although there is no strict correlation between a primary biochemical defect and the clinical phenotype. Inheritance of the defect via maternal mitochondrial DNA is usual (Harding 1991), but interaction between genes encoded by nuclear DNA and those encoded by mitochondrial DNA can result in a wide range of abnormalities through effects on oxidative phosphorylation (Fig. 32.11).

One phenotype is known as the ‘myoclonus epilepsy and ragged-red fibre syndrome’, or MERRF (Fukuhara et al. 1980, So et al. 1989). One of the first patients to be reported was that of Spiro et al. (1970) who described a 16-year-old boy with slow intellectual development. At the age of 8 years he developed decreased visual acuity and at the age of 11 years had a generalized motor seizure which was followed by progressive gait ataxia, dementia, and weakness. Examination showed mental retardation, generalized proximal weakness, bilateral foot drop, severe ataxia, and many small-amplitude random asymmetric myoclonic jerks. He was areflexic and had bilateral extensor plantar responses and decreased joint position sense in the legs. Muscle biopsy revealed ragged-red fibres (Fig. 32.12) and biochemical analysis showed an abnormality of oxidative phosphorylation with a reduction in the concentration of cytochrome b. In 1973 Tsairis et al. reported a family with an inherited disorder characterized by myoclonic epilepsy, ataxia, and diffuse weakness associated with a sensorineural deafness but

sparing of intelligence. Subsequent early reports of the combination of myoclonus, ataxia, and a mitochondrial myopathy include those of Fukuhara et al. (1980), Fitzsimons et al. (1981), Morgan-Hughes et al. (1982), Sasaki et al. (1983), Riggs et al. (1984), and  Rosing et al. (1985). MERRF is concentrated on here because of its propensity to cause myoclonus.

There are a number of defects of mitochondrial DNA which have been found to underlie MERRF (Fig. 32.13), for example mutations in the MTTK, MTTL1, MTTH, MTTS1, MTTS2, MTTF, and the POLG gene (Nakamura et al. 1995, von Goethem et al. 2003, Mancuso et al. 2004, Melonem et al. 2004).

An A to G transition at position 8344 in the tRNALys gene, MTTK, is the commonest, accounting for about 80–90% of patients (Yoneda et al. 1990, Berkovic et al. 1991, Seibel et al. 1991, Zeviani et al. 1991). A T to C transition at nucleotide 8356 of the same gene is the second most common defect (Silvestri et al. 1992,

Zeviani et al. 1993). It may also present as so-called May White syndrome (progressive myoclonus, ataxia, and deafness, accompanied by lipomas) (Ekbom 1975, Berkovic et al. 1986[a], Calabrese et al. 1994, Larsson et al. 1995).

There are a number of other abnormalities of mitochondrial DNA which can cause a similar phenotype (Ozawa et al. 1997, Blumenthal et al. 1998, Arenas et al. 1999), including A to G translation at position 3243 in the tRNALeu (UUR) gene (MTTL1), which, however, more commonly results in the syndrome of mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS) (Hammans 1995, Fabrizi et al. 1996). The converse also applies and MELAS can result from mutations which normally produce MERRF. There is thus significant overlap between these disorders (Campos et al. 1996, Serra et al. 1996). Other clinical pictures may also be associated with the mutations which usually cause MERRF, such as Leigh's syndrome (Silvestri et al. 1993, Rahman et al. 1996, Santorelli et al. 1998), progressive external ophthalmoplegia (Hammans et al. 1993, Naumann et al. 1997), multiple subcutaneous lipomata, which are especially found in a ‘collar’ distribution around the neck, trunk, and arms (Calabresi et al. 1994, Klopstock et al. 1997, Naumann et al. 1997, Munoz-Malaga et al. 2000), diabetes mellitus and hypertension (Austin et al. 1998). The genetic defect mentioned above, which most commonly underlies MELAS, can also cause most of these features (Hammans et al. 1995, Verma et al. 1996) so that there is merging clinical phenotype. It has been suggested that additional base changes may influence the neurological and other clinical features (Hammans et al. 1995) [for review see Finsterer (2007)].

Although a number of different mitochondrial gene mutations can underlie MERRF, they appear to have similar effects on cell functions with reduced rates of protein synthesis, decreased rate of ATP formation, derangement in mitochondrial calcium homeostasis, decrease in mitochondrial membrane potential, and multiple respiratory chain defects, with diminished activity of cytochrome c oxidase (COX) (Masucci et al. 1995, Antonicka et al. 1999, Brini et al. 1999, James et al. 1999). There is thus interruption of the generation of cellular energy by oxidative phosphorylation.

One outcome of this biochemical defect is a decrease in the number of cells that histochemically stain for COX, not only in muscle but also in other cells including neurons (Sparaco et al. 1995). Even in cells that may be COX-positive, a number of mitochondria may be COX-negative (Moslemi et al. 1998). Most patients are heteroplasmic and contain both wild-type and mutant DNA. Ragged-red fibres and those with reduced COX activity tend to have a higher portion of mutant DNA (Ozawa et al. 1997, Mita et al. 1998, Tiranti et al. 1999). The amount of mutant DNA in neurons, however, does not bear a direct relationship to the amount of histological degeneration they undergo (Zhou et al. 1997).

Hopkins and Rosing (1986) reviewed a total of some 19 patients who had a combination of both mitochondrial myopathy and frequent myoclonus. Berkovic et al. (1989) and So et al. (1990) reported the clinical pathological and electrophysiological findings in 13 patients with MERRF (six from one family). Symptoms typically began in the second decade, but onset as late as the age of 42 years had been reported (Sasaki et al. 1983). Myoclonus and ataxia were constant features, tonic-clonic seizures and dementia were usual, and short stature, hearing loss, optic atrophy, neuropathy, hypoventilation, and migraine were sometimes present. Muscle weakness was variable.

Truong et al. (1990) reviewed the spectrum of movement disorders in 85 consecutive patients with proven mitochondrial disease and ragged-red fibres on muscle biopsy; these included 66 patients originally reported by Petty et al. (1986) and 19 further cases. Twenty-nine patients had significant CNS involvement as well as a myopathy. Nine of these patients (11%) had a movement disorder: myoclonus in five, myoclonus and chorea in one, chorea in one, and dystonia in two. Eight were ataxic and four had seizures; four had a retinopathy; five were deaf; four were of short stature; and six were demented. Only one patient conformed to the MERRF phenotype; five with myoclonus had a progressive external ophthalmoplegia; one had the Kearns-Sayre phenotype; and one had strokes.

In a study of seven index cases and 11 symptomatic relatives who had the A to G mutation of tRNALys, Hammans et al. (1993) noted that there was a wide spectrum of clinical signs which virtually encompassed all of those seen in patients with mitochondrial cytopathy, with the exception of retinopathy (Table 32.13). The core features were myoclonus, ataxia, and seizures. They found the myoclonus was usually stimulus sensitive and the seizures were tonic-clonic, absence or atonic in type, and that there was often photosensitivity. Unlike most other causes of progressive myoclonic epilepsy or ataxia, there was a wide range of age of onset and severity, from commencement in childhood and death in early adult life, to symptoms starting in the sixth decade and being limited to mild epilepsy. Mild to moderate proximal limb weakness, deafness, cognitive impairment, axonal peripheral neuropathy, optic atrophy, and hemicranial headaches were seen commonly, while progressive external ophthalmoplegia, ptosis, lactic acidosis, stroke-like episodes, infertility, cataracts, and cervical lipomas were seen less frequently.

Hammans et al. (1995) reported the clinical features of the A to G tRNALeu (UUR) in 22 unrelated patients and 14 of their relatives (Table 32.14). MELAS was the commonest clinical phenotype, although less than half of the index cases showed this. Myoclonus was uncommon and only seen in 8%. Much greater phenotypic heterogeneity is seen in this mutation than with the A to G tRNALys 8344 mutation (Hammans et al. 1991 and 1995).

In a study of 245 patients who had either of these two mitochondrial mutations carried out by Chinnery et al. (1997), there were significantly more stroke-like episodes, progressive external ophthalmoplegia, diabetes, and pigmentary retinopathy associated with A to G tRNALeu (UUR) 3243, while myopathy, myoclonus, ataxia, peripheral neuropathy, optic atrophy, and lipomas were significantly more frequent with A to G tRNALys 8344 (Fig. 32.14).

Inheritance in these disorders is generally by means of maternally transmitted mitochondrial DNA (DiMauro et al. 1985, Rosing et al. 1985, Morgan-Hughes 1986, Harding 1991). The risk of transmission appears to relate to the level of mutated mitochondrial DNA in the mother and it has been postulated that if this exceeds 35–40% for the A to G 8344 translation, then all children will be affected (Larsson et al. 1995). This effect has been found to be stronger for the A to G 3243 MELAS mutation (Chinnery et al. 1998). The severity of the disease may vary considerably in different members of the same family and may be related to the ‘dose’ of the defective gene (Hammans et al. 1993, Arpa et al. 1997). Some asymptomatic older relatives may have only sensorineural deafness associated with ragged-red fibres (Rosing et al. 1985). The myoclonus occurs spontaneously, on movement, or in response to somatosensory stimuli. It can be multifocal or generalized and massive. Photosensitivity may be present.

The prognosis varies considerably. Mildly affected patients may not deteriorate (Rosing et al. 1985), but a proportion develops a severe disabling illness with marked myoclonus and ataxia, often with dementia. Lifespan may be shortened and occasionally overwhelming lactic acidosis can result in death at a young age (Sanger and Jain 1996).

The EEG may show slowing with spike and wave complexes and photosensitivity (Ohtsuka et al. 1993). In general, somatosensory evoked potentials are of normal size (Truong et al. 1990, Ohtsuka et al. 1993, Reutens et al. 1993). Some authors, however, have found them to be enlarged (Thompson et al. 1994). Prolonged latencies in visual and brainstem auditory evoked potentials have been reported (Ohtsuka et al. 1993). A long latency response to peripheral nerve stimulation and an exaggerated facilitatory effect of peripheral stimulation on motor evoked potentials using transcranial magnetic stimulation may be seen, similar to that described above in the section on Unverricht-Lundborg disease (Reutens et al. 1993).

Autopsy of patients with MERRF has revealed marked neuronal degeneration and gliosis which has particularly involved the dentate nucleus, with loss of cerebellar Purkinje and granule cells, and the inferior olivary nucleus. Extensive degeneration of superior cerebellar peduncles, posterior columns, spinocerebellar tracts, and pyramidal pathways has been found. There has also been damage to the subthalamic nucleus, neostriatum, and globus pallidus (Fukuhara et al. 1980, Nakano et al. 1982, Sasaki et al. 1983, Fukuhara 1985, Takeda et al. 1988, Berkovic et al. 1989, McKelvie et al. 1991, Sparaco et al. 1995, Zhou et al. 1997) [for a review of neuropatholigical findings in MERRF in other mitochondrial disorders see Filosto et al. (2007)].

Any patient with progressive myoclonic ataxia in whom the cause is not evident should be investigated for mitochondrial disease. A raised serum and CSF lactate (and pyruvate) are suggestive. EMG may show evidence of a myopathy and nerve conduction studies may reveal a peripheral neuropathy. The EEG may show slow background rhythms, with spike or spike-wave complexes. CT or MRI brain scans may show atrophy, white matter changes, and calcification in the basal ganglia, brainstem, cerebellum, and white matter. The CSF is normal or has a slightly raised protein.

Muscle biopsy may reveal the presence of ragged-red fibres in specimens stained with Gomori's trichrome reaction (Fig. 32.12). Electron microscopy also may show abnormal muscle mitochondria. Specific biochemical defects of mitochondrial energy metabolism may be identified by appropriate investigation. Unfortunately, ragged-red fibres are not invariably found in affected members of well-studied families (Rosing et al. 1985, Jaksck et al. 1998), nor are they present in all sporadic cases with biochemically proven mitochondrial dysfunction (DiMauro et al. 1985).

Abnormalities of glucose and oxygen metabolism can be studied using PET or SPECT (Frackowiack et al. 1988, Watanabe et al. 1998). Magnetic-resonance spectroscopy may also be abnormal (Arnold et al. 1985). Definitive diagnosis, however, depends on studies of mitochondrial DNA. Because of heteroplasmia, however, blood tests may be negative and other tissue may need to be examined. Many cases in whom a diagnosis of mitochondrial cytopathy is suspected on good clinical and laboratory grounds are not as yet able to have this confirmed by genetic testing.

Sialidases cleave sialyl linkages of oligosaccharides and glycoproteins, a reaction that is involved with a number of fundamental biological processes, including antigenic expression and recognition of cell surface receptors (Saito and Yu 1995, Reuter and Gabius 1996). They have been classified into three types on the basis of their subcellular distribution and substrate specificity, namely cytosolic, plasma membrane, and lysosomal. It is the lysosomal form that is defective in human disease. The gene for this is at 6p21.3 (Pshezhetsky et al. 1997). Lysosomal sialidase is a glycoprotein which is only active when it is part of a high molecular weight multienzyme complex that also contains β-glactosidase and cathepsin A (Potier et al. 1990).

The sialidoses (which are included in the mucolipidoses – see also Chapter 41) consist of a group of lysosomal storage disorders associated with alpha-N-acetylneuraminidase deficiency and, in some phenotypes, with an additional deficiency of beta-galactosidase (Lowden and O’Brien 1979, Warner and O’Brien 1983). There is tissue accumulation and urinary excretion of sialylated oligosaccharides and glycolipids. A number of different missense and frameshift mutations have been described, which result in rapid degradation of the enzyme and reduction in its activity (Lukong et al. 2000).

Sialidoses are characterized by the onset in childhood or adolescence of myoclonus in association with a characteristic macular cherry-red spot (Fig. 32.15). They are inherited as autosomal recessive conditions. Two main varieties are recognized: sialidosis types I and II are both caused by mutation in the gene encoding neuraminidase (NEU1; 6p21).

This usually begins in adolescence but onset may range from 8 to 38 years (Tana et al. 1995). The first symptoms are myoclonus and gradual visual failure, although the characteristic macular cherry-red spot may be evident some years earlier (Fig. 32.15). Myoclonus is of two kinds (Engel et al. 1977): (1) massive myoclonus, which occurs spontaneously, in response to movement or somatosensory stimulation, and is associated with small vertex positive spikes; (2) facial myoclonus that is irregular, asymmetric, and stimulus insensitive, and which does not have an EEG correlate (Rapin 1986). Other features include tonic-clonic seizures, progressive ataxia with variable nystagmus, mild pyramidal signs, and occasionally a neuropathy with ‘burning feet’ or evidence of anterior horn cell degeneration (Rapin et al. 1978, Thomas et al. 1979: Steinman et al. 1980). Characteristically, dementia is absent, although performance may deteriorate because of visual failure and myoclonus. The prognosis is relatively good, with survival into adult life. At autopsy, fine cytoplasmic vaculation is seen in the cerebral cortex, basal ganglia, and thalamus, along with intracytoplasmic storage of lipofuscin-like pigment. Although the CNS bears the brunt of the pathology, there may be microscopic involvement of the liver and kidney (Allegranza et al. 1998).

This group has a somewhat variable phenotype (Warner and O’Brien 1983) and has been termed ‘Hurler-type’ because of dysmorphism, hepatosplenomegaly, and mental retardation. Features include a coarse face, dysostosis multiplex, angiokeratoma, and hearing loss. One type most common in Japan develops onset of myoclonus, tonic-clonic seizures, and cherry-red spots in the second or third decade of life (Tsuji et al. 1982, Matsuo et al. 1983, Sakuraba et al. 1983).

Patients with sialidosis type II have a deficiency of neuraminidase, but most also have a partial defect of beta-galactosidase activity, which has led to the designation of this syndrome as galacto-sialidosis (D’Azzo et al. 1982, Sakuraba et al. 1983, Warner and O’Brien 1983, Verheijen et al. 1985).

The EEG may show slow background or fast rhythms. The massive myoclonus is associated with trains of vertex positive spikes, each of which precedes a muscle jerk with short latency (Fig. 32.16) (Franceschetti et al. 1980). The electroretinogram is normal, but visual evoked responses are reduced. Auditory evoked responses are normal, but somatosensory evoked responses are enlarged (Engel et al. 1977, Sakuraba et al. 1983, Tana et al. 1995).

MRI may show atrophy of the cerebral hemispheres, corpus callosum, cerebellum, and pons (Palmeri et al. 2000), focal areas of high signal intensity in the cerebral white matter (Tana et al. 1995), or be normal (Nishiyama et al. 1997). SPECT has shown decreased cerebral blood flow and PET has revealed diminished glucose metabolism in the occipital cortex (Nishiyama et al. 1997).

The diagnosis of sialidosis is confirmed by demonstrating grossly elevated levels of sialyl-oligosaccharides in urine, with a deficiency of cryolabile alpha-N-acetylneuraminidase in leukocyte or cultured fibroblasts, with or without an associated beta-galactosidase deficiency. The presence of an abnormality in beta-galactosidase, however, does not influence the nature of the excreted material

(Van Pelt et al. 1991). The 24 hour excretion of sialic acids and their derivatives is age dependent and increases throughout life (Fang-Kircher 1997). Enzyme analysis of cultured amniotic fluid cells permits prenatal diagnosis (Sasagasako et al. 1993). Identification of mutations in the DNA gene allows definitive diagnosis (Lukong et al. 2000).

(See also Chapter 21)

The relationship between myoclonus and hereditary ataxias is confusing. Boschi (1913) described a case of the association of myoclonus with ataxia (which he considered to be a form of Unverricht-Lundborg disease) but one brother had an illness resembling Friedreich's ataxia and another had a pure cerebellar degeneration. Ramsay Hunt (1921) also described two siblings with myoclonus and a progressive ataxia reminiscent of Friedreich's disease. Lance (1986) reviewed many other isolated reports of the association between myoclonus and cerebellar degenerative disease. Marsden et al. (1990) drew attention to the frequency of this association in adult-onset cerebellar ataxias.

The issues that cloud this discussion are:

However, despite these reservations, there was sufficient information to indicate that some examples of familial or sporadic cerebellar degenerations were associated with prominent myoclonus, which could not be explained by mitochondrial disease or Unverricht-Lundborg disease (Marsden et al. 1990). The advent of DNA analysis, however, clarified this situation. Myoclonus can be seen in most genetic types of spinocerebellar ataxia, including those due to SCA1, SCA3, SCA6, SCA14, SCA17, and SCA19, but is most common in SCA2 (Cancel et al. 1997, Schols et al. 1997[a],[b], Watanabe et al. 1998). This is often of the action type. Other features that differentiate the SCA2 phenotype are the more frequent occurrence of slowed ocular movements, postural and action tremor, and hyporeflexia. The size of the expanded CAG trinucleotide repeat is directly related to the age of onset and severity (Schols et al. 1997[a]). MRI scan shows pontine and cerebellar atrophy (Schols et al. 1997[b]). This feature is also seen in SCA1 and in this disorder degeneration of the dentate nucleus, brainstem motor nuclei, and spinocerebellar tract is more marked than in SCA2, whereas the converse applies to the degeneration of the substantia nigra (Sasaki 1983).

Facial action myoclonus, which is seen as rhythmic muscle twitching of facial muscles during voluntary movement, has been reported in patients with olivopontocerebellar atrophy and has been associated with neurophysiological abnormality of brainstem reflexes (Valls-Sole et al. 1994).

In addition, there is another mixed cerebellar and basal ganglia degeneration, one of whose manifestations involves severe myoclonus. This is hereditary dentatorubral-pallidoluysian atrophy (DRPLA).

(See also Chapter 21)

The first description of DRPLA is credited to Titeca and Van Bogaert (1946), but Smith et al. (1958) clearly defined the entity and Neumann (1959) subsequently described two additional cases. Naito and Oyanagi (1982) reported this condition from Japan in five different families. The manifestations were myoclonus, tonic-clonic seizures, cerebellar ataxia, choreoathetosis, and dementia. Some patients presented in childhood with epilepsy and dementia, whereas ataxia was the first symptom when the onset was in adult life. Degenerative changes, without inclusion bodies, were described in the globus pallidus, subthalamic nucleus, and the fasiculus lenticularis, as well as in the dentatorubral pathway. Occasional cases also have involvement of the posterior columns and spinocerebellar tracts (Pfeiffer and McComb 1990).

Subsequently, Iizuka et al. (1984) described further cases from Japan and delineated various subtypes including:

It is now known that this disorder is caused by an unstable expansion of a CAG repeat in a gene called atrophin 1 located on chromosome 12p. This is very unstable and particularly likely to increase in size during spermatogenesis. The gene is much more frequent in Japanese than Caucasians. Larger gene size results in earlier onset and the phenotype of progressive myoclonus epilepsy. Most people with this clinical subtype have expansions between 62 and 79 repeats, compared with 54 and 67 repeats in those with other subtypes (Ikeuchi et al. 1995[a], [b], Kanazawa 1999). This disorder is discussed more fully in Chapter 21.

Occasional infants have been reported in whom there has been an encephalopathy associated with myoclonic jerks, generalized tonic-clonic seizures, and mental deterioration, in which the clinical picture has been dramatically improved by administration of biotin. In some cases serum biotinase activity has been reduced (Colamaria et al. 1989, Heron et al. 1993) and in others no primary underlying metabolic disturbance has been identified (Low et al. 1986). Rarely, a somewhat similar picture has been reported later in life. Bressman et al. (1986) described a single case of a 30-year-old woman who developed in early adult life progressive unsteadiness, myoclonic jerks, hearing loss, and ataxia followed by severe seizures. She then developed a right hemiparesis followed by a left hemianopia. Despite the resemblance to a mitochondrial encephalomyopathy, plasma lactate and muscle biopsy were normal. Because of the resemblance to biotin-responsive multiple carboxylase deficiencies in infants and young children, biotin was measured in the plasma and found to be undetectable. Pyruvate carboxylase activity in fresh leukocytes was also undetectable. However, all four biotin-dependent carboxylase enzyme activities were normal in skin fibroblasts. The organic acids 3-hydroxyisovaleric and 3-hydroxypropionic were elevated in urine. She was treated with oral biotin (3–20 mg/day) and after 48 hours there was a marked improvement. Within 4 days, the myoclonus, ataxia, hemianopia, and hemiparesis had all resolved. However, despite increasing the dose of biotin, neurological deterioration occurred. Biotin withdrawal led to further deterioration.