Autosomal dominant cerebellar ataxia type I
This is a heterogeneous category characterized by progressive cerebellar ataxia with additional features including supranuclear ophthalmoplegia, slow eye movements, optic atrophy, dementia, extrapyramidal features, dysphagia, pyramidal signs, amyotrophy, and peripheral neuropathy in various combinations (Harding 1984). Pathologically, most patients have olivopontocerebellar atrophy. In contrast to the olivopontocerebellar atrophy form of multiple system atrophy (Chapter 31), glial and neuronal cytoplasmic inclusions are usually absent (Section 32.3.8). Age of onset is typically between 30 and 49 with a median of 40 years (Klockgether et al. 1998) but a minority of cases develop in childhood or old age. The degree of dementia and any visual impairment due to optic atrophy are usually mild. Parkinsonism or dystonia are the usual extrapyramidal manifestations but chorea is occasionally seen. Tendon reflexes may be increased, normal, or absent depending on the degree of pyramidal or peripheral nerve dysfunction. Anterior horn cell involvement is usually manifest as lingual, facial, or limb fasciculation. Bladder dysfunction is common. Progression is relentless and treatment is entirely palliative. Most patients are wheelchair dependent after 17 years and survive up to 25 years after disease onset (Klockgether et al. 1998). Several subtypes within ADCA type I have been defined genotypically. Details of these mutations are shown in Table 31.10.
- The SCA1 mutation is present in 13–35 per cent of ADCA type I cases (Giunti et al. 1998; Klockgether et al. 1998) but the proportion varies considerably among different series (Hammans 1996). As with other trinucleotide repeat expansions, increasing repeat length is associated with earlier age of onset, increased disease severity, and reduced survival (Banfi and Zoghbi 1994; Goldfarb et al. 1996) although one series was unable to demonstrate a relationship between repeat length and rate of disease progression (Klockgether et al. 1998). SCA1 families also demonstrate anticipation with increased severity and earlier onset in succeeding generations, especially with paternal transmission of the SCA1 gene (Genis et al. 1995); the molecular basis for this is the tendency for the SCA1 mutation, like other unstable trinucleotide repeats, to expand during meiosis especially in spermatogenesis (Banfi and Zoghbi 1994).
Clinically, SCA1 patients more commonly have optic atrophy, dysphagia, dysarthria, spasticity, and increased tendon reflexes compared with SCA2 or SCA3 cases (Burk et al. 1996). Hyporeflexia and ophthalmoplegia are less common. MR scanning reveals cerebellar and brainstem atrophy. The mechanism by which the SCA1 mutation causes selective neurodegeneration is unknown. In normal alleles, the CAG repeat sequence is usually interrupted, while pathological expansions are continuous and incorporate a polyglutamine sequence into the translated ataxin 1 protein (as seen with the Huntington's disease mutation and the Huntington protein). The function of ataxin 1 is unclear and the regional selectivity of the neuropathology in SCA1 patients difficult to explain; this does not appear to be due to somatic mosaicism of CAG repeat length within the brain (Lopes-Cendes et al. 1996).
- The SCA2 mutation has been reported in 21–40 per cent of ADCA type I cases in two large recent series (Giunti et al. 1998; Klockgether et al. 1998).
The effect of repeat length on age of onset (Fig. 31.9) and severity is similar to SCA1 and large expansions are associated with paternal transmission (Giunti et al. 1998); as with SCA1, pathological expansions are continuous while normal alleles often have interruptions within the CAG repeat sequence. As expected, anticipation is observed in SCA2 kindreds. Neurological deterioration in SCA2 patients is faster in females and with increasing CAG repeat lengths (Klockgether et al. 1998). Clinically, the SCA2 mutation is consistently more frequently associated with ophthalmoplegia and areflexia (Durr et al. 1995; Filla et al. 1995; Burk et al. 1996; Giunti et al. 1998). Dementia and fasciculations have been reported more often than in SCA1 or SCA3 by some authors. The rate of progression is probably similar to other forms of ADCA type I. MR scanning reveals cerebellar and brainstem atrophy, the latter tending to be more severe than in SCA1 or 3 (Burk et al. 1996).
The frequency of the SCA3 mutation within ADCA type I is also uncertain, with estimates of 17–40 per cent in different series (Hammans 1996; Silveira et al. 1996; Giunti et al. 1998; Klockgether et al. 1998). CAG repeat size and age of onset are inversely related as with SCA1 and 2 and anticipation is also observed in SCA3 families (Giunti et al. 1995; Burk et al. 1996; Higgins et al. 1996) but a clear effect of parental sex on repeat size has not been demonstrated so far. As with SCA2, females and those with larger repeat sizes tend to deteriorate more quickly (Klockgether et al. 1998). Clinically, ADCA type I patients with a SCA3 mutation do not differ significantly from those with the SCA1 gene and MRI appearances are also similar. It is now clear that the SCA3 mutation is also the genetic basis of Machado–Joseph disease (MJD) (Matilla et al. 1995; Higgins et al. 1996). This condition, which is especially common among those of Portuguese or Azorean descent, was for many years regarded as a distinct clinical entity although Harding (1984) included MJD within ADCA type I. Progressive cerebellar ataxia was the salient feature of MJD but a high frequency of parkinsonism, dystonia, eyelid retraction, and bulbar fasciculation was considered characteristic (Rosenberg 1992) although all of these features are seen in SCA1, SCA2, and SCA3. Accordingly, MJD does not, after all, appear to be clinically or genetically distinct from SCA3. Although the cerebellar syndrome is the most common presentation of the SCA3 mutation, there is considerable variation between and within families as noted in earlier descriptions of MJD. Some patients have pure cerebellar ataxia without prominent additional features (Ishikawa et al. 1996) or ‘non-cerebellar’ presentations including levodopa responsive parkinsonism with peripheral neuropathy (Tuite et al. 1995), dystonia, and spastic paraplegia (Sakai and Kawakami 1996). It is therefore likely that if the whole SCA3 phenotype is considered, rather than just the usual cerebellar syndrome, the frequency of extrapyramidal features is greater with SCA3 than SCA1 or 2. The reasons for the phenotypic variability of SCA3 or the mechanism by which the polyglutamine sequence in ataxin 3 causes neurodegeneration are unclear, but as with SCA1, this cannot be accounted for simply by CAG repeat length variability (Lopes-Cendes et al. 1996). Generalized dystonia was the main clinical feature of a patient homozygous for the SCA3 mutation, suggesting that gene dosage also influences the phenotype (Lang et al. 1994).
SCA8 is a CTG expansion mutation in the 3′ untranslated region of a gene on chromosome 13q21. It is a rare cause of ADCA type I; affected patients have had dysarthria, spasticity, and sensory loss in addition to slowly progressive cerebellar ataxia (Koob et al. 1999). However, expanded SCA8 repeats have been described in some normal individuals and in conditions other than cerebellar ataxia. Accordingly, the existence of this gene as a definite ataxia locus is uncertain (Vincent, et al. 2000).
SCA12 is also a rare cause of the ADCA I phenotype; onset ranges from 8 to 55 years but is typically in the 4th decade of life. Tremor of the head and upper limbs is prominent, with mild cerebellar signs, hyperreflexia, parkinsonism and dementia or psychiatric features. Neuroimaging shows cerebral and cerebellar atrophy. In some cases, subtle neurological signs are present in infancy (O'Hearn et al. 2001).
SCA13 has been described in a single family with ataxia, learning disability, motor developmental delay and pyramidal features. MR scanning shows cerebellar and pontine atrophy. A gene locus has been mapped to chromosome 19q (Herman-Bert et al. 2000). Although currently classified among the dominant later onset ataxias (Subramony and Filla 2001), this condition is of congenital onset and should probably be included in the congenital ataxias.
SCA15 and SCA16 are also described only in single families; at this stage, detailed clinical features are unclear and their ultimate classification is uncertain (Subramony and Filla 2001).
Some ADCA type I families do not have any of the currently known SCA mutations and although the exact proportion is unclear it is probably small. Different reports have quoted various frequencies of SCA1–3 in ADCA I, probably reflecting differences in ascertainment which has usually not been systematic. Nevertheless, two large recent ADCA type I series demonstrated SCA1, 2, or 3 mutations in 74 and 90 per cent of families (Giunti et al. 1998; Klockgether et al. 1998).
Autosomal dominant cerebellar ataxia type II
This form of ADCA is characterized clinically by the presence of visual loss due to a pigmentary retinopathy (Harding 1984). Pathologically there is olivopontocerebellar atrophy with additional pregeniculate visual pathway involvement but these findings are not unique. ADCA II has recently been mapped to chromosome 3 and the pathogenic mutation (SCA7) identified as another unstable CAG trinucleotide repeat expansion (Benomar et al. 1995; Lindblad et al. 1996) as previously predicted clinically (Enevoldson et al. 1994). The SCA7 CAG repeat is more unstable than any other trinucleotide repeat expansion, especially with paternal transmission; increased repeat size, as in other neurodegenerative CAG repeat disorders, is associated with earlier onset, greater severity, and the phenomenon of anticipation. The age of onset in patients with SCA7 is wider than in other forms of ADCA and it is with this condition that the approximate guideline of autosomal recessive and autosomal dominant cerebellar ataxia usually developing either side of about 25 years of age is least dependable. Many patients are first affected in their 30s or later but children may be affected with a more severe form of the disorder from infancy onwards. Occasionally patients develop symptoms in their 60s and there are some asymptomatic gene carriers indicating a gene penetrance of 95 per cent (Enevoldson et al. 1994). Patients with later onset show less rapid progression. The usual presenting features are ataxia or visual failure but both are usually present within a few years. Pyramidal signs and ophthalmoplegia are commonly associated while chorea, dementia, amyotrophy, and sensory loss are less common. With this combination of features it is not surprising that misdiagnoses of multiple sclerosis have been reported. The visual failure is due to retinal degeneration with loss of central visual field and eventually severe visual loss in most cases. However, patients are often unaware of the insidious progression of visual loss until quite late in the disease and although ophthalmoscopic examination reveals a pigmentary retinopathy in advanced cases (Fig. 31.11), the fundoscopic appearances may be normal for many years leading to diagnostic difficulty. Careful ophthalmological examination, and especially electroretinogram (ERG) recordings may be needed to reveal the characteristic retinal involvement (Enevoldson et al. 1994). In contrast, visual evoked potentials are less sensitive and may be misleadingly normal. Early onset infantile cases are much more severe and rapidly fatal with wasting, weakness, fasciculations, dementia, and blindness; retinal pigmentary changes are usually prominent. ADCA II is probably heterogeneous; at least one family with the phenotype does not have the SCA7 mutation (Giunti et al. 1999).
Autosomal dominant cerebellar ataxia type III
This form of ADCA typically has a later onset, after 50 years of age, and slow progression; there may or may not be mild additional pyramidal signs (Harding 1984). Such cases are not common and ADCA III is heterogeneous (Table 31.11). Pathologically, there is cerebellar cortical atrophy which can lead to severe cerebellar atrophy but only very slow clinical deterioration (Frontali et al. 1992). One form of mild late onset ataxia has been described in the descendants of President Lincoln (Lincoln ataxia) and a gene locus designated SCA5 mapped to chromosome 11 (Ranum et al. 1994). The SCA3 mutation may also produce a pure cerebellar phenotype occasionally (see above). Other cases are associated with a small CAG repeat expansion within the alpha 1A calcium channel gene on chromosome 19 (Zhuchenko et al. 1997) which has been designated SCA6 (Table 31.10). Clinically these patients also have late onset, slowly progressive ataxia which is usually not associated with additional signs; cerebellar atrophy is marked on magnetic resonance scans but the brainstem is preserved (Stevanin et al. 1997). Two families with an ADCA III phenotype and epilepsy have a gene designated SCA10 which has been mapped to chromosome 22 (Zu et al. 1999). Two other ADCA III families have now been linked to the SCA11 locus on chromosome 15 (Worth et al. 1999); the nature of this mutation is currently unknown. A single family with mild, late onset cerebellar ataxia and cerebellar atrophy on neuroimaging studies has been described with linkage to a gene (SCA14) on chromosome 19q13.4-ter (Yamashita et al. 2000). The same family contains patients with younger onset ataxia associated with axial myoclonus.
Autosomal dominant cerebellar ataxia with sensory neuropathy
This is a rare form of autosomal dominant cerebellar ataxia in which there is a prominent sensory peripheral neuropathy; it is likely to be the same disorder previously referred to as Biemond's ataxia (Harding 1984). Pathologically there is degeneration of cerebellar Purkinje cells, dorsal columns, dorsal root ganglion cells, and peripheral nerves (Nachmanoff et al. 1997). The sensory symptoms, affecting the face and limbs, and rapid progression of the disorder are unlike ADCA I, II, or III. In one family, linkage to a locus on chromosome 16, designated SCA4, has been reported (Flanigan et al. 1996). 31.8.6 Dentatorubropallidoluysian atrophy (DRPLA)
This rare autosomal dominant condition was until recently described mainly in Japan (Ilzuka et al. 1984) but also occurs in Europe. It has been reported elsewhere as the ‘Haw River syndrome’. DRPLA is caused by a CAG repeat expansion mutation of the atrophin 1 gene on chromosome 12 (Koide et al. 1994). Pathologically there is degeneration in the dentate nucleus and external segment of the pallidum (Gpe) along with the projections of these areas to the red and subthalamic nuclei (Warner et al. 1994). However, the neuropathological findings are variable and may not suggest the diagnosis. Most patients develop symptoms in adult life although earlier onset in childhood may occur. The phenotype is highly variable both within and between families; some patients have progressive cerebellar ataxia with prominent chorea and dystonia while others have dementia with chorea (pseudo-Huntingtonian type) or a progressive myoclonic epilepsy (Ilzuka et al. 1984).
Other patients have various combinations of these subtypes and a psychotic presentation has also been described. As expected for a trinucleotide expansion disorder, age of onset is earlier with increased repeat length (Koide et al. 1994) and DRPLA families show anticipation and more severe manifestations following paternal transmission (Warner et al. 1995). MR scans show cerebellar and brainstem atrophy as well as multiple white matter hyperintensities on T2 sequences (Koide et al. 1997). DRPLA needs to be considered in the differential diagnosis of Huntington's disease (especially if DNA testing is negative), ADCA type I cases with prominent dystonic or choreiform manifestations and progressive myoclonic epilepsy. However, a DRPLA mutation is usually absent in patients without a family history (Warner et al. 1995).
As with SCA mutations, the regional selectivity of the neuropathology in DRPLA is not due to differences in CAG repeat length or gene expression in different brain regions (Nishiyama et al. 1997). There is possibly a defect of mitochondrial respiration in muscle in DRPLA based on results of in vivo magnetic resonance spectroscopy studies (Lodi et al. 2000).