4
Journal of the Ncurologwal S( wtz(e.s, I i3 (1992) 4 1~ ~' 1992 Elsevier Science Publishers B.V. All rights reserved 0()22-510X/t;2/'$1)5.()0
JNS 03883
Review article
What do we really know about amyotrophic lateral sclerosis? Michael Swash and Martin S. Schwartz Department of Clinical Neuroscience, The Royal London Hospital, London, UK, and Department of Neurology, Atkinson Morley's Hospital, London, UK
(Received 13 January, 1992) (Revised, received 19 May, 1992) (Accepted 30 June, 1992) Key words: Amyotrophic lateral sclerosis; Pathogenesis
Summary The cause of amyotrophic lateral sclerosis is unknown. In this review clinical and scientific data that are pertinent to understanding this disease are reviewed. There are currently several major controversies concerning the possible role of immunological factors, genetic factors, environmental toxins, and viral infection in pathogenesis. These concepts must be considered in relation to what is known about the disease in all its aspects, including epidemiologieal data, information on the classical and molecular pathology of the disease, and on associated involvement of other systems, e.g., the spinocerebellar pathways and frontal dementia. Only when all this information is assimilated can full understanding of the disease and, hopefully, a logical approach to treatment and prevention, be achieved.
Contents Presentation and prognosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Risk factors for ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Is ALS becoming more common? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atypical forms of ALS: problems in nosology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dementia and ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Is ALS a pure motor disorder? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Poliomyelitis and ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anterior horn celt loss and declining strength; electrophysiological correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The immunological theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The neurotoxic hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genetic aspects of ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clues from pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intraneuronal cytoplasmic inclusion bodies in ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Animal models of ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A m y o t r o p h i c lateral sclerosis (ALS) was described by C h a r c o t in 1869. Since then, o t h e r a n t e r i o r h o r n cell diseases have b e e n recognised, some of which, such as
Correspondence to: Dr. M. Swash, The Royal London Hospital, London E1 1BB, UK. Tel.: (71) 377 7472; Fax: (71) 377 7008.
5 5 6 6 7 7 7 8 9 10 10 I1 1! 12 12
the spinal m u s c u l a r atrophies, are familial. In addition, s y n d r o m e s involving a n t e r i o r h o r n cells b u t associated with p r e s u m e d n e u r o t o x i n ingestion, such as the A L S s y n d r o m e f o u n d o n the W e s t e r n Pacific island of G u a m , have b e e n recognised. A l t h o u g h the majority of cases of A L S are readily d i a g n o s e d , o t h e r related syndromes may closely r e s e m b l e ALS. I n the U K a n d r e l a t e d countries, the t e r m m o t o r n e u r o n disease ( M N D ) is
often used as a generic term to include the four major clinical syndromes: amyotrophic lateral sch'rosis, in which therc is both upper and lower motor neuron inw~lvemcnt; progressire muscular atrophy, in which the clinical features imply progressive lower motor neuron disease limited to thc spinal cord: primao' lateral sch'rests, in which there is progressive involvement of upper nlotor neurons without signs of lower motor neuron diseasc; and progressire hulbar palsy, in which there is predominant involvement of motor systems in lhc brain stem. Familial ALS, a feature of about 5-10¢~ of patients with otherwise typical ALS, must be differcntiated from the spinal muscular atrophies (SMA), disordcrs with an autosomal reccssive pattern of inheritancc. The clinical features of familial ALS typically resemble those of the more common sporadic form of disease, but controversy exists concerning the classification of certain rare SMA syndromes associated with additional clinical features, e.g. deafness, and of atypical familial ALS patients, with long survival, in whom corticospinal inwHvement may bc minimal. Rowland (ICJ91) noted that the WFN classification of motor neuron disorders includcs 28 autosomal dominant syndromes, 37 autosomal recessive syndromes and 7 X-linked syndromes. In addition, 16 different syndromcs of ALS with dementia are recognised in this classification, including dementia with otherwise typical ALS. This apparent heterogeneity is difficult to reconcile with at single undcrlying pathological process. lndced, pathological studies of thcse clinically defined disorders arc incomplete.
Presentation and prognosis ALS is a progressive disorder, of insidious onset, with a fatal outcomc. Thc presenting complaint is usually weakness, often asymmetrical and focal, involving a few adjacent muscles. However, even at this early stage there is electrophysiological evidence of widespread involvement (Stalberg et al. 1975). Weakness of a leg is a slightly more common presentation than weakness of a hand (Jokelainen 1977). Signs of tipper motor neuron involvement are almost invariable at the time of presentation (Eisen et al. 1992). About a third of cases present with bulbar symptoms, whether predominantly upper or lower motor neuron in type. Fasciculations are a characteristic feature, nearly always first noted by the clinician, mostly in the arm muscles. 'Fhcrc is an excess of women presenting with bulbar ALS after the age of 59 years (Li et al. 1990). Survival is shorter when there is bulbar inw)lvement at presentation. Jokelainen (1977) found that patients with predominant bulbar features survived for a mean period of 2.2 years and that those with mainly spinal signs survived longer. The median survival for all ADS
cases is 3 - 4 years (Caroscio et al. 1984; Jokelainen et al. 1989), but in patients with progrcssivc nluscular atrophy the median survival was 1() years (Mortara ct al. 1984). Generally, the earlier the age of onset lhe longer the survival (Kondo 1978). Mulder and Howard (1976) and Mortara et al. (1984) found that 2(Vf of their patients were alive 5 years after onset, and I(Vi 10 ycars after onset, Spontaneous remission in ALS is controversial. T u c k c r et al. (1991) reported four paticnts with an
ALS-like syndrome in whom complete recovcry occurred, without treatment, 5-12 months after onset. Nonc of these patients had fcaturcs of nlotor ncuropathy or toxic exposure, and the CSF protein was normal in the three patients in whom lumbar puncture was performed. Three additional cases of AI,S with recovery have been reported (Mulder and Howard 1076: Engcl et at. 1969: Rowland 1980). The patient dcscribed by Engcl et al. (1909) begal/ to improve 20 y c a r s i/lter onset, a n d recovcred to 85~i of hernial strength. In addition, recovery from an Al~S-like syndrome has been rcportcd in patients with renal carcinoma (Buchanan and Malamud 1973: Evans el al. 1990), and bronchial clircinonla (Mitchcll anti Okzak 1979}. Given the abscncc of a diagnostic test for thc disease it is not possible to decide whether thcsc paticnts had typical sporadic AI,S. It is difficuh to envisage that anterior horn cells lost in tile progressive phase of the disease could rcgenerate, and it seems likely, thercfore, that these unusual cases r c p r c s e n t another syndrome, perhaps a fornl of motor n e u r o p a thy. In longer surviving patients Patten c t a l . (1979) found that the muscles showcd typc 1 fibre grouping and, conversely, a poor prognosis was noted in patients whose muscle biopsies showed clusters of small atrophic fibres. Secondary myopathic fcatures develop in the muscle biopsies of long surviw~rs, and tile blood CK may be elevatcd, but these arc featnrcs that tire associated with long survival, rather than prognostic indicators. Although symptomatic treatment, such as gastrostomy fecding and ventilatory support, mercascs the duration of survival, no treatment has yet beell found effective in preventing muscle wasting and progrcssion in thc classical form of thc disease.
Risk ['actors for ALS Epidcmiological studies offer an atlractivc option for formulating hypotheses in a disease of unknown causation. However, although thcrc havc been many epidemiological studies of ALS, none has provided a clear, testable hypothesis regarding pathogenesis. Ideas dcrived from observations of relatively small populations, or from recognition of apparent "clusters" of
cases have led to larger studies, but these, usually based on death certification or hospital activity analyses have failed to demonstrate any consistent variation in prevalence or incidence rates for typical sporadic ALS in developed countries, and have similarly failed to link ALS with any other disease. In addition, no relationship with life-style, social class, occupation, dietary habits or with previous illness has been established, despite suggestive early reports (Armon et al. 1991; Kurtzke 1991). However, the disease is more common in men than in women, and it may be more frequent in white than in black people (Lilienfeld et al. 1989; Kurtzke 1991). Environmental factors have been considered in some detail, but early reports of an association with heavy metal exposure, e.g. to selenium, lead and mercury, or to chemicals and leather products have not been confirmed (Li et al. 1985; Roelofs-Iversen et al. 1984; Kurtzke 1991). Heavy labouring occupations, particularly in farmers, have been associated with an increased incidence of the disease (Holloway and Mitchell 1986). Head injury is not a risk factor for ALS (Williams et al. 1991). An association with athleticism, suspected because of a few well-publicised instances of ALS occurring in retired sportsmen, has proven difficult to verify (Roelofs-lversen 1984). Despite the considerable research investment, information pertinent to aetiology has not emerged from epidemiological studies.
Is ALS becoming more common? Kurtzke (1982) noted a 42% increase in deaths certified due to ALS in the USA in the period 19491976, but attributed this to better ascertainment of cases rather than to a rising incidence of the disease itself. Recent studies have confirmed this trend in France, the United Kingdom, the USA, Italy and Scandinavia, in all age groups, but more in men than in women. The annual incidence has risen from about 0.6/100,000/year to about 1.3/100,000/year (Durleman and Alperovitch 1989; Jokelainen 1987; Yoshida et al. 1986; Lilienfeld et al. 1989). The uniformity of these data suggests that the reported increase in incidence is real, and not due to better case ascertainment.
Atypical forms of ALS: problems in nosology The majority of patients present with the core syndrome of the disease. However, there are a number of recognised variant forms of the disease. These are often of limited severity, and of restricted anatomical and geographical distribution. There is striking heterogeneity among these variant forms of ALS. In one
form, monomelic motor neuron disease, the distal parts of the upper limbs are affected asymmetrically, without spasticity or bulbar involvement, and with no signs of corticospinal involvement of arms or legs. Most reports of this disorder have come from Asia and from the Indian subcontinent. The disorder begins in the second or third decade, mainly affects men, and ceases to progress some 2-3 years after the onset. Although the legs are not clinically affected there is EMG evidence of reinnervation in lower limb muscles (Singh et at. 1980; Chan et al. 1991). in view of the absence of upper motor neuron features, and the relatively benign clinical course, this disorder has also been classified as a spinal muscular atrophy, rather than ALS. in a second atypical variant, the wasted leg, syndrome (Chopra et al. 1984), wasting confined to one leg is the main presenting complaint. In most patients with this syndrome there is diffuse wasting of the affected limb, but major disability is uncommon and corticospinal features, or sensory disturbances, are not present. This disorder also is probably a form of SMA rather than ALS. A related heterogenous syndrome has been reported from Madras, India (Valmikinathan et al. 1973). The age of onset ranged from less than 10 years to 30 years, 90% were male, 15% had associated sensorineural deafness, thus resembling the Vialetto-van Laere syndrome (Summers et al. 1987), and 40% had bulbar involvement. Some had increased reflexes and extensor plantar responses. The distribution of weakness and wasting was variable. These atypical syndromes of neurogenic wasting probably represent several different disorders, including SMA syndromes and juvenile or early-onset ALS. Juvenile ALS is a term used to describe the typical sporadic ALS syndrome beginning in childhood or adolescence. This is a rare syndrome in developed Western countries; only 4 - 6 % of all cases of ALS begin under the age of 40 years (Mortara et al. 1984; Li et al. 1990). Familial ALS may have an adult or juvenile onset. Familial cases comprise 5 - 1 0 % of cases in large series, and the majority of these show an autosomal dominant pattern of inheritance. The mean age of onset is about 10 years less than in typical sporadic ALS. The clinical features and course are similar to those of sporadic cases, but the sex ratio is equal (Emery and Holloway 1982; Williams et al. 1988; Li et al. 1988). Juvenile-onset familial ALS is characterised by prominent spasticity and prolonged survival, sometimes for as long as 10 years after diagnosis. Recessively-inherited cases have an early onset, usually before the age of 5 years, and show bulbar features in as many as 80% of cases (Emery and Holloway 1982; Ben Hamida et al. 1988). By contrast, Emery and Holloway (1982) found that in dominantly inherited cases the mean age of onset was ll years, and butbar involvement was uncommon.
l ) e m e n t i a and ALS
Dementia is l'arc ill sporadic ALS, occurring in only -,c; of the 255 cases reviewed by Jokchtinen (lt)77). lhtdson (19NI) studied a number of such cases and cndcawnu-cd to separate them from Creutzfeldt-Jakob disease, (iuamanian ALS, and post-encephalitic syndromes, t'aticnts with cortical Lewy bodies show parkinsonian features alld cognitive impairment, and this may sometimes be accompanied by clinical features consistcrii with ALS. Mitsuyama (1984) found degeneration of the substantia nigra iri such cases. Ncary c t a l . (199()) noted thal the ncuropsychiatric IealtllCS of the dementia in these eases were consistent with a fioiltal lobe disordcr, consisting of disinhibition, inattention, restlessness (i.e., impairment of personality arid conduct), diminished spcccll output and poor riaming abilities without apraxia or agnosia. Formal men> err tests rcx,calcd impairment of recent recall. Pathological studies showed frontal atrophy with neuronal loss in the frontal cortex, subcortical gliosis and spongitornl change. In the spinal cord changes consistent with ADS were evident. In addition, despite the absence of clinical fcatuleS of extrapyramidal disease, there was loss of cells and gliosis in the substantia nigra pars conlpacla, btl! Lcwy bodies were not seen. No senile plaques or ncurofibrillary tangles were found in the cortex in these cases, although the coexistence of Alzhcmicr changes. Pick's disease, or Lcwy bodies with A I £ has bccn noted in some cases of familial AI_S and, ralCly, in the sporadic syndronle (see Rowland lggl for re,,icw). Attempts to transrnit AI,S by inoculation of CNS material into priniatcs have been carried ()tit in both the sporadic s}ndromc and in Guamanian ALS (Gibbs and (;aidusck 197{'4). These inoculated primates were followed for Up tO 5 years, but no neurological illness developed cven in those animals inoculated intracercbrally with brain honlogeriates or with visceral material ['l'CIlll patients with ALS associated with denlcntia (Salazar cl al. 1983). Nonetheless, it is tempting to spccuhttc l]lal ALS with spongiform change in the frontal cortex rcprescrits a prion disease (see Allen cl al. 1971), aim one case of the amyotrophic form of ('reutzfeldt-,lakob disease has been shown to trarismit by intracercbral inoculation of brain homogenatc to the squirrel monkcy (('onnollv et al. 1988).
Is ALS a pure motor disorder?
The clinical data suggest that early in the natural history of the disease, sporadic ALS is charactcriscd by predominant involvement of bulbar and spimll motor neurons, with a similar distribution of cortico-spinal tract in'>olx,ct+lenl. However, tipper motor neuron ab-
normalities arc rarely the major presenting featu;'c, except in primaw hiteral sclerosis (Yc.urlgcr ct al. lt)SS). As the disease progresses subclinical involvement of other neuronal systems occurs, inwlriably inchiding spinocercbelhir pathways and nuclei, less consistently involving frontal cortex and substantia nigra. Some patients with ALS complain of non-specific sensory phenorncna, especially tingling paracsthcsia. Muldcr ct al. (1983) found abnormal detection thresholds for cutaneous sensation in lgef of patients tested. Shcfncr ct al. (1991) showed that the minimum sensory conduction velocity was slowed in lhe sural nerve m 9 of 18 patients with ALS, and Bchnia and Kelly (lqgl) found that the sensory nerve action potential amplitude was reduced in a small proportion of their patients. These observations imply that both the sporadic and the familial forms of ALS consist of a multisystcm disorder in which the brunt of the palhological process involves the Motor system, rather lhail ,2 ptiie motor system degeneration.
Poliomyelitis and ALS
Since poliomyelitis causes loss of anterior horn cells, and of motor cells in thc bulbar nuclei, it is not surprising that it I]as bccn suggested that AIrS rcprcsents a late con3plication of poliovirus infection. This hypothesis was stated most clearly by Zilkha (It;h2) who described 11 cases of ALS associated with antecedent poliomyelitis infection, representing Ilia/- of all cases of ALS diagnosed at the National Ilospital, Ouccn Square, during a 30-year period to that date. The relatively small ntlnlbcf of patients in whom this diagnosis was made at this time suggests that selection wets an inadvertent factor in these data. All these patients survived more than 2 years, and other atypical features were noted. Poskanzer ci al. ( lt)('lt,1) reported thal 5 of lt)h patients with chissic ALS had a history of paralytic poliomyelitis. However, subsequent st.rvcvs of several hundred patients with previous poliomyelitis have failed to conl:irm a relationship with the later development of ALS (Howard ct al. 19SS). Armon ct al. (199()) reported a patient in ,ahem ALS developed 30 years after acute poliomyelitis, but noted that this association had been reporlcd in only two patients in the USA in the 30-ycar period 196()-109(L whereas 7 new cases would bc expected annually in the 30ql,()0() polio survivors in the USA, simply on the basis of an atmual incidcnce of ALS of 2.4,/10(k000/ycar in the Mayo Clinic series (Yoshida ct al. 1986). They concluded that antecedent paralytic poliotnyelitis might have a protective effect against the deve[opment of ALS, rather than the causative role predicted by others (scc Martyn c t a l . 1088). The post poliomyelitis nluscuhu- atrophy svndrcunc is
recognised as a distinct clinical entity, in which progressive weakness and atrophy develop, especially in muscles weakened by the acute disease, on average 33 years after the acute infection. This syndrome is due to loss of reinnervative capacity in large motor units, causing new or possibly continuing instability in previously damaged motor units. There is no evidence of new motor neuronal loss other than that associated with ageing (Dalakas et al. 1986), although Sharief et al. (1991) have described increased intrathecal synthesis of IgM antibodies to poliovirus in 58% of 36 patients with post-polio syndrome. This response was not found in patients with ALS or other neurological disorders. This observation raises important questions concerning the ability of poliovirus to survive in a latent state in the CNS in patients previously infected with poliovirus, or of CNS tissue to continue to express antibodies to poliovirus, perhaps to some other stimulus. It is conceivable that an immunologically distinct virus, directed against both anterior horn cells and the upper motor neuron could cause ALS, but no such virus has yet been isolated.
Anterior horn cell loss and declining strength; electrophysiological correlations Wohlfart (1957) noted that in ALS muscular atrophy developed only when more than 30% of anterior horn cells were lost. This retention of muscular function despite loss of anterior horn cells is a function of the relative effectiveness of collateral axonal sprouting and reinnervation of denervated muscle cells by surviving motor units. This process results in an increased fibre density in single fibre E M G recordings (Stalberg et al. 1975). There is no direct correlation between strength and fibre density in affected muscles in ALS, since the relationship between these two variables differs in different stages of the disease (Swash and Schwartz 1984). As motor units are lost in a muscle there is a gradual increase in fibre density, implying effective functional reinnervation. Later, the fibre density decreases, reflecting a less effective capacity for axonal sprouting in the few surviving anterior horn ceils. McComas (1977) showed that reinnervation was effective even when only half the motor units remained functional. When only 5% of motor units remain, reinnervation virtually ceases (Hansen and Ballantyne 1978). The complex relation of motor unit loss with weakness is shown by the clinical finding that strength declines in a linear fashion during the course of the disease. Before clinical involvement is evident the muscle is normal, and at the advanced stage of the disease a low level of strength may persist for a long period (Munsat et al. 1988). Motor unit loss, measured by functional motor unit counting using the fractional incrementing stimulation
technique of McComas et al. (1971), showed that the functioning motor unit population was halved in each 6-month period in the first year, and that it diminished more slowly later in the course of the disease (Dantes and McComas 1991), thus representing an approximately logarithmic function. In some instances, there was marked variation in the numbers of functional motor units in a muscle at different times in the course of the disease, implying that periods of reversible dysfunction occurred, an observation that could explain transient clinical improvement following T R H therapy (Modarres-Sadeghi et al. 1988). At any one time in the course of the disease the functional state of motor units in a muscle will not be uniform. Some units will be dysfunctional, others will be sustaining an increased axonal load, and others may be normal (Dantes and McComas 1991). In the latter study, the rate of progression was faster in arm muscles than in leg muscles. Other comparisons showed no significant differences. There was no relation between the pattern of progression and age, region of onset or asymmetry at onset (Munsat et al. 1988). Because of the compensatory effect of axonal sprouting in sustaining the number of functional motor units in affected muscles, the exponential loss of motor units found etectrophysiologically is consistent with a linear loss of strength. The relation of corticospinal tract degeneration to the lower motor neuron syndrome associated with anterior horn cell dysfunction, has not been adequately addressed. Abnormalities in central motor conduction have been demonstrated using magnetic stimulation of the motor cortex (Ingram and Swash 1987; Berardelli et al. 1987; Eisen et al. 1990). No responses were observed in limb muscles in 12 patients with pseudobulbar symptoms, and in 25 of 28 additional patients the latency from cortical stimulation was increased in at least one of the three upper limb muscles studied (Eisen et al. 1990). Ingrain and Swash (1987) noted increased latencies to lower limb muscles, and correlated these findings with the presence of extensor plantar responses. Similar abnormalities in central motor conduction have been noted in patients with multiple sclerosis, and in other disorders involving the corticospinal tracts (Gandevia 1990). Dengler et al. (1990) noted that there are three reasons for weakness in ALS. These are: (1) corticospinal degeneration, (2) loss of motor neurons, and (3) reduced force production in surviving motor units. In more advanced stages of the disease they noted that the decline in force seemed to particularly affect high threshold motor units. The pattern of degeneration of anterior horn cells in the cord appears random at any given cord segmental level, and there is no predilection for certain motor neuron cell columns (Swash et al. 1986). Nonetheless the brachial segments, particularly C6, are especially likely to be involved (Tsukagoshi et al. 1979; Swash et
al. 1986). A characteristic feature of the disease, however, is rclative sparing of certain muscle groups, for example, the external ocular muscles, cricopharyngeus and pelvic floor striated sphincter muscles (Mannen el al. 1977). This relative resistance to the degenerativc process is a major early feature of the disease but, in the later stages, especially if lifc is prolonged by ventilatory support, these musclc groups are clearly inw)lved (Hayashi and Kaio It)St)). In this cnd-stage state there is marked degcncration of the spinocerebellar pathways, of the small fibre sensory system, brain stem reticular formation and cercbellar and vestibular systems, features that arc not so marked in the earlier stages of the disease (Swash ct al. 1988). The pelvic floor sphinctcrs atrc also relatively uninvolved in acute poliomyelitis (Koshima c t a l . 1980), and in muscular dystrophics and spinal muscular attrophy. The multisystern involvcnlent found in the late stages of ALS clearly indicates that the discasc is not at disorder restricted to the nlotor system (see above)
The immunological t h e o n In general, ALS is not associated with autoimmunc disease. Nonethcless, a number of autoimmunc clucs have bccn ascertained in relation to disorders that resemble ALS. Howcvcr, almost all the patients reported in whom an AkS-likc syndrome was associated with an autoinmmnc disorder have had a restricted lower motor neuron syndrome. The possiblc relation of ALS to cancer is a hmgstanding clinical problem that exemplifies the immunological hypothesis (Brain and Norris 1965). The strongest reported association concerns ALS and lymphoma, Younger ct al. (1991) reviewed the 30 known cases of ALS associated with lymphoma, two of which wcrc confirmcd at autopsy. All thc patients who improved werc lrcated with immunosuppressivc drugs. Three of seven investigated patients in this group of 30 had a paraproleinaemia. There is no rccognised association of ALS with malignancies other than lymphoma. and cvcn this apparent association hats not been subjcctcd to statistical analysis. The relation of paraproteinaemia to ALS is also controversial. Rowland el al. (1981) reported a patient with paraprotcinacmia associated with peripheral ncuropathy and fasciculation, a clinical syndrome that sonlcwJlal resembled motor neuron disease. Subsequently ,Shy el al. (1986) and Younger c t a l . (1990) found paraproteinaemia in 10G+ of patients with motor neuron syndromes, and in 3(V7 of such patients when thc CSF protein was increased. This incidence of paraprotcinacmia was greater than that found in a popul:.ilion of similar age. EMG data were not given in these
reports, but patients with conduction block were meluded in the series of Shy et al. (1986). As ill paraprotein-associatcd neuropathies, the circulating monoclonal lgM paraprotcin can bind to peripheral nerve, especially at the nodes of Ranvicr, a phenomenon that probably accounts lk~r the conduction block found at EMG (Santoro e t a [ . 1990). (kmduction block is not a feature of classical ALS. Anti-GM~ antibodies arc found in the serum of 10-21tC~ of patients with lower motor neuron syndromes resembling ALS. Thc GM~ gangliosidcs ,|re acidophilic glycolipids that arc a constituent of neuronal cell membranes, inchiding ,lcrvc terminals and nodes of Ranvier (Pestronk 1991 ). The reported incidence of circulating anti-GM~ antibodies in patients with ALS has varied greatly from series 1o series, suggesting differences in laboratory t c c h n i q u c (Pcstronk et al. 1989). In typical ALS the rclevancc of anti-GM~ antibodies is unccrtain, since this finding may rcprcsent an epiphcnomenon, and lhe association has been most clearly recognised in lower motor neuron syndromcs (Pcstronk ct al. 1989). tto~vcvcr, antiGM I antibodies are found ill titres less tilnll [; 35(I ill a varictv of ncurological disorders, e.g., othcr autoimmulle diseases alnd a:,lmm( ~J )pathics, ' and inflanmlah)rv polyneuropatthics, and titrcs at this level :,lrc not clinically significant (Pcstronk 1991). l~amb and Patten (1991) flmnd thai only one of 16 pclticnls with classical ALS had anti-GM I antibcMies in a iitrc gl-caicr than 1: 15(i. Pestronk c t a l . (19911) dcfincd three lower motor neuron syndronles associated with high tiircs ~l ;lilliGM~ antibodies in the serum: (l) A multifoca[ motor ncuropattl), with weakness of distal ;.irln muscles, principally affecting men younger than 45 years. Ncrvc conduction studies show multifocal conduction block. (2) A nlultifocaI motor neuropathy, with simihu clinical fcatulcS, but vvithout conduclit)n block. (3) An asymmetrical lowcr motor neuron syndrome, affecting proximal muscles more lhan distal, mostly found in tncn older than 45 vcars. There is no conduction block. In practice, the specificity of these syndromes is difficult to define. It has bcen suggested that the triad of asymmetrical lower motor neuron wcakncss, motor conduction block and high anti-GM~ antibody litrcs represents a treatable syndrome. ('yclophosphan]idc therapy for 3 to 6 months nlay result ill improvenlcnt, correlating with reduction in sel'unl anti-tiM I antibody titrcs to 311r+~ of previous lcvcls. In considering the clinical significance of anti-GM I antibodies associated with these neurological syndromes, it should bc noted that there is no simple relation bctwccn the occurrence of peripheral neuropathy and nlonoclonal gamnlopaithy, witl~ or without anti-glycoprotein aniiboclies, and
10 that many patients with monoclonal gammopathy do not develop neuropathies.
The neurotoxic hypothesis Several geographic foci of ALS are recognised. In Guam, the Kii peninsula of Japan, and in Western New Guinea there is a high incidence of ALS. These ALS syndromes are associated with other neurodegenerative syndromes, e.g., the A L S / p a r k i n s o n i s m / d e mentia complex, but ALS occurs in these areas without features of these other neurodegenerative disorders (Hudson 1981). The causation of these syndromes is controversial. The suggestion that the Guam ALS syndrome resulted from the ingestion of a dietary excitotoxin, BMAA, found in cycad nuts, which were used as a form of flour (Spencer et al. 1987), has been disputed on the grounds that insufficient of the active substance is present in cooked food to act as a neurotoxin (Duncan 1991) and because, with changing dietary habits, the disease in these endemic areas is not decreasing in incidence. There may thus be genetic factors as well as environmental factors. In addition, oral BMAA is not neurotoxic to mice (Perry et al. 1989). Guam ALS shows neurofibrillary tangles and granulovacuolar degeneration in neurons in subcortical areas of the brain, features that differ from the typical sporadic or familial ALS syndrome in Western countries. Dietary calcium depletion with coincidental aluminium intoxication has been employed as a model of the diet of the Chamorro people of Guam, in experiments with non-human primates (Garruto et al. 1989). Motor neuron degeneration, with accumulation of phosphorylated neurofilaments, and with basophilic inclusions, were found in these experimental studies (Garruto et al. 1989), but the relevance of these findings to sporadic ALS is uncertain. Affected motor neurons contained excess aluminium. Garruto et al. (1989) suggested that this resulted in disruption of intraneuronal cytoplasmic membranes, and thus in the neuronal cytopathology. Lathyrism, an acute or insidious spastic paraparesis, developing in malnourished people, sometimes associated with sensory symptoms at the onset, is a syndrome that occurs when wheat flour is unavailable and flour made from the chickling pea, Lathyrus sativus, is substituted in the diet. This flour contains an excitotoxin, B O A A (/3-n-oxalylamino-L-alanine), that is capable of inducing corticospinal tract degeneration in non-human primates (Spencer et al. 1987). This experimental disorder mimics the human disease. A cytotoxic effect on motor neurons, however, is not established in human lathyrism (Streifler et al. 1977). Since the characteristic combination of upper and lower motor neuron
involvement does not occur in lathyrism, this disorder is not a neurotoxic model for sporadic ALS. Plaitakis et al. (1988) found abnormalities of glutamate metabolism, with increased blood levels of glutamate and reduced glutamate content in brain and, particularly, in spinal cord in sporadic ALS. They suggested that over-excitation of glutamate receptorbearing cells in the CNS resulted in degeneration and death of these cells, and thus in the clinical features of ALS. Increased taurine levels have been reported in CNS tissue, perhaps secondary to excitotoxic neuronal damage (Plaitakis et al. 1988). Rothstein et al. (1990) confirmed Plaitakis' (1991)) findings, but Perry et al. (1990) found that CNS tissue levels of these amino acids in ALS did not differ from those of controls. Young (1990)concluded that the excitotoxic hypothesis of the causation of ALS was not proven. A similar controversy surrounds the observation that plasma cysteine levels are raised in ALS and Parkinson's disease (Steventon et al. 1988), a finding that Perry et al. (1991) could not confirm.
Genetic aspects of ALS The familial occurrence of ALS suggests the possibility of a genetic factor not only in these cases but also in the sporadic disease. In late onset disorders family studies are likely to be incomplete and, inevitably, there has been much speculation concerning the genetic hypothesis of ALS. In spinal muscular atrophy (SMA) the genetic locus has been mapped on chromosome 5ql 1-2-13.3 (SMA5q). This locus has been found both in acute, infantile onset SMA (Werdnig-Hoffmann disease) and in the juvenile onset (KugetbergWelander disease) variants. Further. no evidence of allelic heterogeneity between the acute and chronic forms of the disease has been detected (Munsat et al. 1990), although there may be alletic heterogeneity within the chronic late onset form itself. In ALS the search for a genetic marker has proven more difficult. but Siddique et al. (1991) have demonstrated linkage with a gene located on the long arm of chromosome 21 (21q 22.1-22.2). Linkage of the disease with this locus in informative families is currently being explored. There appears to be genetic heterogeneity, within the locus. Although this locus on chromosome 21q is close to that associated with the locus for the amyloid precursor protein gene that is important in one form of familial Alzheimer's disease, the two loci are far apart (Goate et al. 1991). Genetic heterogeneity in ALS is consistent with the differing phenotypes observed in clinical practice, and might result from the interactive expression of several dominant genes within the locus. as in any oligogenic disorder. This hypothesis will be important in formulating concepts of the pathogenesis
11 of sporadic A I £ , and pcrhaps offer the possibility of deducing the naturc of the abnormal protein product o f thc defective gone(s). T h c locus f o r X-linked, adult-onset, bulbo-spinal SMA (Kenncdv syndromc) has been identified on the long arm of the X c h r o m o s o m e in linkage with th0 :indrogcn receptor gonc. The latl0r is especially i n t e r csting in that gynaocomastia and reduced fertility are features of this syndrome (La Spada ot al. 1991).
Clues front pathology The principal pathological features of ALS arc well rccogniscd ( t t o l m c s 1909: Brownell ct al. 1970). Firstly, there is loss of antcrior h o r n cells, and of somatic motor neurons ii1 lhc lower cranial nerve nucl0i. The ocular motor nuclei, and thc $ 2 - 3 0 n u f nucleus that inncrvatcs the striatcd pelvic floor sphincter muscles arc rolati\cly spared ( M a n n c n o t al. 1977) but u n d e r g o degeneration hltcr in the course of the disoasc (Kihira c t a l . 19q I). ( } a n l n l a m o t o r ncurons are also inwHvcd, since there is h~ss o f thc m o t o r innorvation of intrafusal musclc fibres in muscle spindlcs (Swash and Fox 1974). The widespread pattcrn of loss of neurons in the antcrior horn stlggosls that intcrncurons are probably also affectcd. These features imply that thcrc is a degrce of selective vulnerability within the m o t o n e u renal population in the spinal cord and brain stcm (Swash ct al. 1986). Sccondly. there is loss of Bctz cells in the ccrcbral cortex, and degeneration of the corticospinal tracts ( t t o l m c s 19(19). ('orticospinal tract dogcncration, like loss of a n t e r i o r h o r n cells, is often strikingly asymmetrical, a fcaturc that parallels the clinical findings (Swash ct oil. 19SN). No relationship has bccn ostablishcd between the development of the u p p e r and lower m o t o r neuron fcalurcs. Thirdly, other neural systems arc inwHvcd in the disease. For cxamplc, there is d c g c n c r a tion of the spinoccrobcllar pathways in the cord and brain stcn~ (Swash et al. It,iNN), with loss of neurons in ('larkc's column (Avcrback and C r o c k e r 1982) and, in some cases, in t h a l a m u s , corpus callosum and superior
colliculi. This pathological cvidcncc of widespread involvement of neural systems in the central nerwms systcm, beyond the motor system, has bocn confirmed by a magnetic rcsonancc imaging study of ALS in which high intensity Icsions wcrc found widcly distribut0d in the white matter of the brain and brain stem (Salcs-Luis ot al. 199(I). In addition, abnormalities have becn reported in surat nervc biopsies, consisting of axonal atrophy, romyclination and a shift in thc diameter spcctrnm toward smaller diameter fibres (Dyck ci al. 1975: Heads c t a l . 19ql). At autopsy there is often slight pallor in the medial parts of the gracile columns, a featurc that is said to be more promincnt in familial cases.
lntraneuronal cytoplasmic inclusion bodies in ALS A n u m b c r of diffcrcnt types of intrancuronal inchisions (Martin ct al. 1990) have bccn recorded in anterior h o r n cclls in A L S ( T a b l e 1). 13unina (1962) doscribcd intracytoplasmic acidophilic inclusions, 2 3 # m in diameter, that a p p e a r as electron-dense, anlorphous material, often with a vesicular appc;,lrance, in thc clcctron microscopc (Chou 1979: Hirano ct al. 1984). Bunina bodies are found in thc pcrikarya, in the axes, and in the soma of motor ncurons, t | i r a n o and ]wata (1979) found B u n i n a bodies in 1(t o f 20 cases of AI,S. Howcvcr, Bunina bodies are not specific to ADS ( ( ' h o u 1979). Axonal spheroids are c o m m o n in the discasc ((Tarpcnter 1968; Hirano ct al. 1984). Spheroids contain neurofihtmontous material, consisting of filaments Ill nln in d i a m c t c r ,
arrangcd
individually
o r in s m a l l
bundles, often associ:,ited with other cell organcllcs. Those neurofilamcntous conglomerates are associated with incrcascd conccntrations of aluminium (Kihira ct al. 19911. P h o s p h o w l a t c d cpitopcs arc prcscnt in increased concentration in some degenerating anterior horn coils ( M a n c t t o ct al. 1088), but this appears to be a nonispccific finding (Leigh ct al. 1988). Ubiquitin, a highly conscrvcd hcal shock protcin of molccuhir weight 8565 kDa is associated with soxoral diffcrcnt types of intermediate l'ilamcnl inclusion hod-
TAFII.I{ I IN('I.USI()N l'l()l)lES IN AI,,%(FROM MARTIN ET AI,. 1991)) Inclusion
Appcalance
('omposition
{Jhiquifinalud hotlies l]tillina hodics llclsophilic hodi~:s
aggregai.csor skein-like [ilalllerlls I lfl/xrn diamclcr: 7 |ilamcnh)us at EM cosinl~philic, round or ¢hmgatud, 1-4 #m diameter, variabh_' LM appear:.lllCC hcisophilic, globoid or asteroid, 3 4/xm cliamcler. Irregular chains of rough
ubiquithl core ill)l kiloWil nol klll)'~VI1
[ lhano
bodies
Spheroids
[:,R. glycogen and vesicles as EM eosin~)philic, hyaline, rod-shaped. 4 6 /xm diameter. Paracryslalline arrays olh lit nm filaments at EM cosimiphilic, hyaline, argyrophilic, rounded, I(I ll}0#m diamcier. Ill nm filaments a! EM
actm, acfinin, tau. xincnlm, [ I'Ol]OlllyOSill n e u r o f i l a l l l c n l
Journal of the Ncurologwal S( wtz(e.s, I i3 (1992) 4 1~ ~' 1992 Elsevier Science Publishers B.V. All rights reserved 0()22-510X/t;2/'$1)5.()0
JNS 03883
Review article
What do we really know about amyotrophic lateral sclerosis? Michael Swash and Martin S. Schwartz Department of Clinical Neuroscience, The Royal London Hospital, London, UK, and Department of Neurology, Atkinson Morley's Hospital, London, UK
(Received 13 January, 1992) (Revised, received 19 May, 1992) (Accepted 30 June, 1992) Key words: Amyotrophic lateral sclerosis; Pathogenesis
Summary The cause of amyotrophic lateral sclerosis is unknown. In this review clinical and scientific data that are pertinent to understanding this disease are reviewed. There are currently several major controversies concerning the possible role of immunological factors, genetic factors, environmental toxins, and viral infection in pathogenesis. These concepts must be considered in relation to what is known about the disease in all its aspects, including epidemiologieal data, information on the classical and molecular pathology of the disease, and on associated involvement of other systems, e.g., the spinocerebellar pathways and frontal dementia. Only when all this information is assimilated can full understanding of the disease and, hopefully, a logical approach to treatment and prevention, be achieved.
Contents Presentation and prognosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Risk factors for ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Is ALS becoming more common? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atypical forms of ALS: problems in nosology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dementia and ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Is ALS a pure motor disorder? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Poliomyelitis and ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Anterior horn celt loss and declining strength; electrophysiological correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The immunological theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The neurotoxic hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Genetic aspects of ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clues from pathology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Intraneuronal cytoplasmic inclusion bodies in ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Animal models of ALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A m y o t r o p h i c lateral sclerosis (ALS) was described by C h a r c o t in 1869. Since then, o t h e r a n t e r i o r h o r n cell diseases have b e e n recognised, some of which, such as
Correspondence to: Dr. M. Swash, The Royal London Hospital, London E1 1BB, UK. Tel.: (71) 377 7472; Fax: (71) 377 7008.
5 5 6 6 7 7 7 8 9 10 10 I1 1! 12 12
the spinal m u s c u l a r atrophies, are familial. In addition, s y n d r o m e s involving a n t e r i o r h o r n cells b u t associated with p r e s u m e d n e u r o t o x i n ingestion, such as the A L S s y n d r o m e f o u n d o n the W e s t e r n Pacific island of G u a m , have b e e n recognised. A l t h o u g h the majority of cases of A L S are readily d i a g n o s e d , o t h e r related syndromes may closely r e s e m b l e ALS. I n the U K a n d r e l a t e d countries, the t e r m m o t o r n e u r o n disease ( M N D ) is
often used as a generic term to include the four major clinical syndromes: amyotrophic lateral sch'rosis, in which therc is both upper and lower motor neuron inw~lvemcnt; progressire muscular atrophy, in which the clinical features imply progressive lower motor neuron disease limited to thc spinal cord: primao' lateral sch'rests, in which there is progressive involvement of upper nlotor neurons without signs of lower motor neuron diseasc; and progressire hulbar palsy, in which there is predominant involvement of motor systems in lhc brain stem. Familial ALS, a feature of about 5-10¢~ of patients with otherwise typical ALS, must be differcntiated from the spinal muscular atrophies (SMA), disordcrs with an autosomal reccssive pattern of inheritancc. The clinical features of familial ALS typically resemble those of the more common sporadic form of disease, but controversy exists concerning the classification of certain rare SMA syndromes associated with additional clinical features, e.g. deafness, and of atypical familial ALS patients, with long survival, in whom corticospinal inwHvement may bc minimal. Rowland (ICJ91) noted that the WFN classification of motor neuron disorders includcs 28 autosomal dominant syndromes, 37 autosomal recessive syndromes and 7 X-linked syndromes. In addition, 16 different syndromcs of ALS with dementia are recognised in this classification, including dementia with otherwise typical ALS. This apparent heterogeneity is difficult to reconcile with at single undcrlying pathological process. lndced, pathological studies of thcse clinically defined disorders arc incomplete.
Presentation and prognosis ALS is a progressive disorder, of insidious onset, with a fatal outcomc. Thc presenting complaint is usually weakness, often asymmetrical and focal, involving a few adjacent muscles. However, even at this early stage there is electrophysiological evidence of widespread involvement (Stalberg et al. 1975). Weakness of a leg is a slightly more common presentation than weakness of a hand (Jokelainen 1977). Signs of tipper motor neuron involvement are almost invariable at the time of presentation (Eisen et al. 1992). About a third of cases present with bulbar symptoms, whether predominantly upper or lower motor neuron in type. Fasciculations are a characteristic feature, nearly always first noted by the clinician, mostly in the arm muscles. 'Fhcrc is an excess of women presenting with bulbar ALS after the age of 59 years (Li et al. 1990). Survival is shorter when there is bulbar inw)lvement at presentation. Jokelainen (1977) found that patients with predominant bulbar features survived for a mean period of 2.2 years and that those with mainly spinal signs survived longer. The median survival for all ADS
cases is 3 - 4 years (Caroscio et al. 1984; Jokelainen et al. 1989), but in patients with progrcssivc nluscular atrophy the median survival was 1() years (Mortara ct al. 1984). Generally, the earlier the age of onset lhe longer the survival (Kondo 1978). Mulder and Howard (1976) and Mortara et al. (1984) found that 2(Vf of their patients were alive 5 years after onset, and I(Vi 10 ycars after onset, Spontaneous remission in ALS is controversial. T u c k c r et al. (1991) reported four paticnts with an
ALS-like syndrome in whom complete recovcry occurred, without treatment, 5-12 months after onset. Nonc of these patients had fcaturcs of nlotor ncuropathy or toxic exposure, and the CSF protein was normal in the three patients in whom lumbar puncture was performed. Three additional cases of AI,S with recovery have been reported (Mulder and Howard 1076: Engcl et at. 1969: Rowland 1980). The patient dcscribed by Engcl et al. (1909) begal/ to improve 20 y c a r s i/lter onset, a n d recovcred to 85~i of hernial strength. In addition, recovery from an Al~S-like syndrome has been rcportcd in patients with renal carcinoma (Buchanan and Malamud 1973: Evans el al. 1990), and bronchial clircinonla (Mitchcll anti Okzak 1979}. Given the abscncc of a diagnostic test for thc disease it is not possible to decide whether thcsc paticnts had typical sporadic AI,S. It is difficuh to envisage that anterior horn cells lost in tile progressive phase of the disease could rcgenerate, and it seems likely, thercfore, that these unusual cases r c p r c s e n t another syndrome, perhaps a fornl of motor n e u r o p a thy. In longer surviving patients Patten c t a l . (1979) found that the muscles showcd typc 1 fibre grouping and, conversely, a poor prognosis was noted in patients whose muscle biopsies showed clusters of small atrophic fibres. Secondary myopathic fcatures develop in the muscle biopsies of long surviw~rs, and tile blood CK may be elevatcd, but these arc featnrcs that tire associated with long survival, rather than prognostic indicators. Although symptomatic treatment, such as gastrostomy fecding and ventilatory support, mercascs the duration of survival, no treatment has yet beell found effective in preventing muscle wasting and progrcssion in thc classical form of thc disease.
Risk ['actors for ALS Epidcmiological studies offer an atlractivc option for formulating hypotheses in a disease of unknown causation. However, although thcrc havc been many epidemiological studies of ALS, none has provided a clear, testable hypothesis regarding pathogenesis. Ideas dcrived from observations of relatively small populations, or from recognition of apparent "clusters" of
cases have led to larger studies, but these, usually based on death certification or hospital activity analyses have failed to demonstrate any consistent variation in prevalence or incidence rates for typical sporadic ALS in developed countries, and have similarly failed to link ALS with any other disease. In addition, no relationship with life-style, social class, occupation, dietary habits or with previous illness has been established, despite suggestive early reports (Armon et al. 1991; Kurtzke 1991). However, the disease is more common in men than in women, and it may be more frequent in white than in black people (Lilienfeld et al. 1989; Kurtzke 1991). Environmental factors have been considered in some detail, but early reports of an association with heavy metal exposure, e.g. to selenium, lead and mercury, or to chemicals and leather products have not been confirmed (Li et al. 1985; Roelofs-Iversen et al. 1984; Kurtzke 1991). Heavy labouring occupations, particularly in farmers, have been associated with an increased incidence of the disease (Holloway and Mitchell 1986). Head injury is not a risk factor for ALS (Williams et al. 1991). An association with athleticism, suspected because of a few well-publicised instances of ALS occurring in retired sportsmen, has proven difficult to verify (Roelofs-lversen 1984). Despite the considerable research investment, information pertinent to aetiology has not emerged from epidemiological studies.
Is ALS becoming more common? Kurtzke (1982) noted a 42% increase in deaths certified due to ALS in the USA in the period 19491976, but attributed this to better ascertainment of cases rather than to a rising incidence of the disease itself. Recent studies have confirmed this trend in France, the United Kingdom, the USA, Italy and Scandinavia, in all age groups, but more in men than in women. The annual incidence has risen from about 0.6/100,000/year to about 1.3/100,000/year (Durleman and Alperovitch 1989; Jokelainen 1987; Yoshida et al. 1986; Lilienfeld et al. 1989). The uniformity of these data suggests that the reported increase in incidence is real, and not due to better case ascertainment.
Atypical forms of ALS: problems in nosology The majority of patients present with the core syndrome of the disease. However, there are a number of recognised variant forms of the disease. These are often of limited severity, and of restricted anatomical and geographical distribution. There is striking heterogeneity among these variant forms of ALS. In one
form, monomelic motor neuron disease, the distal parts of the upper limbs are affected asymmetrically, without spasticity or bulbar involvement, and with no signs of corticospinal involvement of arms or legs. Most reports of this disorder have come from Asia and from the Indian subcontinent. The disorder begins in the second or third decade, mainly affects men, and ceases to progress some 2-3 years after the onset. Although the legs are not clinically affected there is EMG evidence of reinnervation in lower limb muscles (Singh et at. 1980; Chan et al. 1991). in view of the absence of upper motor neuron features, and the relatively benign clinical course, this disorder has also been classified as a spinal muscular atrophy, rather than ALS. in a second atypical variant, the wasted leg, syndrome (Chopra et al. 1984), wasting confined to one leg is the main presenting complaint. In most patients with this syndrome there is diffuse wasting of the affected limb, but major disability is uncommon and corticospinal features, or sensory disturbances, are not present. This disorder also is probably a form of SMA rather than ALS. A related heterogenous syndrome has been reported from Madras, India (Valmikinathan et al. 1973). The age of onset ranged from less than 10 years to 30 years, 90% were male, 15% had associated sensorineural deafness, thus resembling the Vialetto-van Laere syndrome (Summers et al. 1987), and 40% had bulbar involvement. Some had increased reflexes and extensor plantar responses. The distribution of weakness and wasting was variable. These atypical syndromes of neurogenic wasting probably represent several different disorders, including SMA syndromes and juvenile or early-onset ALS. Juvenile ALS is a term used to describe the typical sporadic ALS syndrome beginning in childhood or adolescence. This is a rare syndrome in developed Western countries; only 4 - 6 % of all cases of ALS begin under the age of 40 years (Mortara et al. 1984; Li et al. 1990). Familial ALS may have an adult or juvenile onset. Familial cases comprise 5 - 1 0 % of cases in large series, and the majority of these show an autosomal dominant pattern of inheritance. The mean age of onset is about 10 years less than in typical sporadic ALS. The clinical features and course are similar to those of sporadic cases, but the sex ratio is equal (Emery and Holloway 1982; Williams et al. 1988; Li et al. 1988). Juvenile-onset familial ALS is characterised by prominent spasticity and prolonged survival, sometimes for as long as 10 years after diagnosis. Recessively-inherited cases have an early onset, usually before the age of 5 years, and show bulbar features in as many as 80% of cases (Emery and Holloway 1982; Ben Hamida et al. 1988). By contrast, Emery and Holloway (1982) found that in dominantly inherited cases the mean age of onset was ll years, and butbar involvement was uncommon.
l ) e m e n t i a and ALS
Dementia is l'arc ill sporadic ALS, occurring in only -,c; of the 255 cases reviewed by Jokchtinen (lt)77). lhtdson (19NI) studied a number of such cases and cndcawnu-cd to separate them from Creutzfeldt-Jakob disease, (iuamanian ALS, and post-encephalitic syndromes, t'aticnts with cortical Lewy bodies show parkinsonian features alld cognitive impairment, and this may sometimes be accompanied by clinical features consistcrii with ALS. Mitsuyama (1984) found degeneration of the substantia nigra iri such cases. Ncary c t a l . (199()) noted thal the ncuropsychiatric IealtllCS of the dementia in these eases were consistent with a fioiltal lobe disordcr, consisting of disinhibition, inattention, restlessness (i.e., impairment of personality arid conduct), diminished spcccll output and poor riaming abilities without apraxia or agnosia. Formal men> err tests rcx,calcd impairment of recent recall. Pathological studies showed frontal atrophy with neuronal loss in the frontal cortex, subcortical gliosis and spongitornl change. In the spinal cord changes consistent with ADS were evident. In addition, despite the absence of clinical fcatuleS of extrapyramidal disease, there was loss of cells and gliosis in the substantia nigra pars conlpacla, btl! Lcwy bodies were not seen. No senile plaques or ncurofibrillary tangles were found in the cortex in these cases, although the coexistence of Alzhcmicr changes. Pick's disease, or Lcwy bodies with A I £ has bccn noted in some cases of familial AI_S and, ralCly, in the sporadic syndronle (see Rowland lggl for re,,icw). Attempts to transrnit AI,S by inoculation of CNS material into priniatcs have been carried ()tit in both the sporadic s}ndromc and in Guamanian ALS (Gibbs and (;aidusck 197{'4). These inoculated primates were followed for Up tO 5 years, but no neurological illness developed cven in those animals inoculated intracercbrally with brain honlogeriates or with visceral material ['l'CIlll patients with ALS associated with denlcntia (Salazar cl al. 1983). Nonetheless, it is tempting to spccuhttc l]lal ALS with spongiform change in the frontal cortex rcprescrits a prion disease (see Allen cl al. 1971), aim one case of the amyotrophic form of ('reutzfeldt-,lakob disease has been shown to trarismit by intracercbral inoculation of brain homogenatc to the squirrel monkcy (('onnollv et al. 1988).
Is ALS a pure motor disorder?
The clinical data suggest that early in the natural history of the disease, sporadic ALS is charactcriscd by predominant involvement of bulbar and spimll motor neurons, with a similar distribution of cortico-spinal tract in'>olx,ct+lenl. However, tipper motor neuron ab-
normalities arc rarely the major presenting featu;'c, except in primaw hiteral sclerosis (Yc.urlgcr ct al. lt)SS). As the disease progresses subclinical involvement of other neuronal systems occurs, inwlriably inchiding spinocercbelhir pathways and nuclei, less consistently involving frontal cortex and substantia nigra. Some patients with ALS complain of non-specific sensory phenorncna, especially tingling paracsthcsia. Muldcr ct al. (1983) found abnormal detection thresholds for cutaneous sensation in lgef of patients tested. Shcfncr ct al. (1991) showed that the minimum sensory conduction velocity was slowed in lhe sural nerve m 9 of 18 patients with ALS, and Bchnia and Kelly (lqgl) found that the sensory nerve action potential amplitude was reduced in a small proportion of their patients. These observations imply that both the sporadic and the familial forms of ALS consist of a multisystcm disorder in which the brunt of the palhological process involves the Motor system, rather lhail ,2 ptiie motor system degeneration.
Poliomyelitis and ALS
Since poliomyelitis causes loss of anterior horn cells, and of motor cells in thc bulbar nuclei, it is not surprising that it I]as bccn suggested that AIrS rcprcsents a late con3plication of poliovirus infection. This hypothesis was stated most clearly by Zilkha (It;h2) who described 11 cases of ALS associated with antecedent poliomyelitis infection, representing Ilia/- of all cases of ALS diagnosed at the National Ilospital, Ouccn Square, during a 30-year period to that date. The relatively small ntlnlbcf of patients in whom this diagnosis was made at this time suggests that selection wets an inadvertent factor in these data. All these patients survived more than 2 years, and other atypical features were noted. Poskanzer ci al. ( lt)('lt,1) reported thal 5 of lt)h patients with chissic ALS had a history of paralytic poliomyelitis. However, subsequent st.rvcvs of several hundred patients with previous poliomyelitis have failed to conl:irm a relationship with the later development of ALS (Howard ct al. 19SS). Armon ct al. (199()) reported a patient in ,ahem ALS developed 30 years after acute poliomyelitis, but noted that this association had been reporlcd in only two patients in the USA in the 30-ycar period 196()-109(L whereas 7 new cases would bc expected annually in the 30ql,()0() polio survivors in the USA, simply on the basis of an atmual incidcnce of ALS of 2.4,/10(k000/ycar in the Mayo Clinic series (Yoshida ct al. 1986). They concluded that antecedent paralytic poliotnyelitis might have a protective effect against the deve[opment of ALS, rather than the causative role predicted by others (scc Martyn c t a l . 1088). The post poliomyelitis nluscuhu- atrophy svndrcunc is
recognised as a distinct clinical entity, in which progressive weakness and atrophy develop, especially in muscles weakened by the acute disease, on average 33 years after the acute infection. This syndrome is due to loss of reinnervative capacity in large motor units, causing new or possibly continuing instability in previously damaged motor units. There is no evidence of new motor neuronal loss other than that associated with ageing (Dalakas et al. 1986), although Sharief et al. (1991) have described increased intrathecal synthesis of IgM antibodies to poliovirus in 58% of 36 patients with post-polio syndrome. This response was not found in patients with ALS or other neurological disorders. This observation raises important questions concerning the ability of poliovirus to survive in a latent state in the CNS in patients previously infected with poliovirus, or of CNS tissue to continue to express antibodies to poliovirus, perhaps to some other stimulus. It is conceivable that an immunologically distinct virus, directed against both anterior horn cells and the upper motor neuron could cause ALS, but no such virus has yet been isolated.
Anterior horn cell loss and declining strength; electrophysiological correlations Wohlfart (1957) noted that in ALS muscular atrophy developed only when more than 30% of anterior horn cells were lost. This retention of muscular function despite loss of anterior horn cells is a function of the relative effectiveness of collateral axonal sprouting and reinnervation of denervated muscle cells by surviving motor units. This process results in an increased fibre density in single fibre E M G recordings (Stalberg et al. 1975). There is no direct correlation between strength and fibre density in affected muscles in ALS, since the relationship between these two variables differs in different stages of the disease (Swash and Schwartz 1984). As motor units are lost in a muscle there is a gradual increase in fibre density, implying effective functional reinnervation. Later, the fibre density decreases, reflecting a less effective capacity for axonal sprouting in the few surviving anterior horn ceils. McComas (1977) showed that reinnervation was effective even when only half the motor units remained functional. When only 5% of motor units remain, reinnervation virtually ceases (Hansen and Ballantyne 1978). The complex relation of motor unit loss with weakness is shown by the clinical finding that strength declines in a linear fashion during the course of the disease. Before clinical involvement is evident the muscle is normal, and at the advanced stage of the disease a low level of strength may persist for a long period (Munsat et al. 1988). Motor unit loss, measured by functional motor unit counting using the fractional incrementing stimulation
technique of McComas et al. (1971), showed that the functioning motor unit population was halved in each 6-month period in the first year, and that it diminished more slowly later in the course of the disease (Dantes and McComas 1991), thus representing an approximately logarithmic function. In some instances, there was marked variation in the numbers of functional motor units in a muscle at different times in the course of the disease, implying that periods of reversible dysfunction occurred, an observation that could explain transient clinical improvement following T R H therapy (Modarres-Sadeghi et al. 1988). At any one time in the course of the disease the functional state of motor units in a muscle will not be uniform. Some units will be dysfunctional, others will be sustaining an increased axonal load, and others may be normal (Dantes and McComas 1991). In the latter study, the rate of progression was faster in arm muscles than in leg muscles. Other comparisons showed no significant differences. There was no relation between the pattern of progression and age, region of onset or asymmetry at onset (Munsat et al. 1988). Because of the compensatory effect of axonal sprouting in sustaining the number of functional motor units in affected muscles, the exponential loss of motor units found etectrophysiologically is consistent with a linear loss of strength. The relation of corticospinal tract degeneration to the lower motor neuron syndrome associated with anterior horn cell dysfunction, has not been adequately addressed. Abnormalities in central motor conduction have been demonstrated using magnetic stimulation of the motor cortex (Ingram and Swash 1987; Berardelli et al. 1987; Eisen et al. 1990). No responses were observed in limb muscles in 12 patients with pseudobulbar symptoms, and in 25 of 28 additional patients the latency from cortical stimulation was increased in at least one of the three upper limb muscles studied (Eisen et al. 1990). Ingrain and Swash (1987) noted increased latencies to lower limb muscles, and correlated these findings with the presence of extensor plantar responses. Similar abnormalities in central motor conduction have been noted in patients with multiple sclerosis, and in other disorders involving the corticospinal tracts (Gandevia 1990). Dengler et al. (1990) noted that there are three reasons for weakness in ALS. These are: (1) corticospinal degeneration, (2) loss of motor neurons, and (3) reduced force production in surviving motor units. In more advanced stages of the disease they noted that the decline in force seemed to particularly affect high threshold motor units. The pattern of degeneration of anterior horn cells in the cord appears random at any given cord segmental level, and there is no predilection for certain motor neuron cell columns (Swash et al. 1986). Nonetheless the brachial segments, particularly C6, are especially likely to be involved (Tsukagoshi et al. 1979; Swash et
al. 1986). A characteristic feature of the disease, however, is rclative sparing of certain muscle groups, for example, the external ocular muscles, cricopharyngeus and pelvic floor striated sphincter muscles (Mannen el al. 1977). This relative resistance to the degenerativc process is a major early feature of the disease but, in the later stages, especially if lifc is prolonged by ventilatory support, these musclc groups are clearly inw)lved (Hayashi and Kaio It)St)). In this cnd-stage state there is marked degcncration of the spinocerebellar pathways, of the small fibre sensory system, brain stem reticular formation and cercbellar and vestibular systems, features that arc not so marked in the earlier stages of the disease (Swash ct al. 1988). The pelvic floor sphinctcrs atrc also relatively uninvolved in acute poliomyelitis (Koshima c t a l . 1980), and in muscular dystrophics and spinal muscular attrophy. The multisystern involvcnlent found in the late stages of ALS clearly indicates that the discasc is not at disorder restricted to the nlotor system (see above)
The immunological t h e o n In general, ALS is not associated with autoimmunc disease. Nonethcless, a number of autoimmunc clucs have bccn ascertained in relation to disorders that resemble ALS. Howcvcr, almost all the patients reported in whom an AkS-likc syndrome was associated with an autoinmmnc disorder have had a restricted lower motor neuron syndrome. The possiblc relation of ALS to cancer is a hmgstanding clinical problem that exemplifies the immunological hypothesis (Brain and Norris 1965). The strongest reported association concerns ALS and lymphoma, Younger ct al. (1991) reviewed the 30 known cases of ALS associated with lymphoma, two of which wcrc confirmcd at autopsy. All thc patients who improved werc lrcated with immunosuppressivc drugs. Three of seven investigated patients in this group of 30 had a paraproleinaemia. There is no rccognised association of ALS with malignancies other than lymphoma. and cvcn this apparent association hats not been subjcctcd to statistical analysis. The relation of paraproteinaemia to ALS is also controversial. Rowland el al. (1981) reported a patient with paraprotcinacmia associated with peripheral ncuropathy and fasciculation, a clinical syndrome that sonlcwJlal resembled motor neuron disease. Subsequently ,Shy el al. (1986) and Younger c t a l . (1990) found paraproteinaemia in 10G+ of patients with motor neuron syndromes, and in 3(V7 of such patients when thc CSF protein was increased. This incidence of paraprotcinacmia was greater than that found in a popul:.ilion of similar age. EMG data were not given in these
reports, but patients with conduction block were meluded in the series of Shy et al. (1986). As ill paraprotein-associatcd neuropathies, the circulating monoclonal lgM paraprotcin can bind to peripheral nerve, especially at the nodes of Ranvicr, a phenomenon that probably accounts lk~r the conduction block found at EMG (Santoro e t a [ . 1990). (kmduction block is not a feature of classical ALS. Anti-GM~ antibodies arc found in the serum of 10-21tC~ of patients with lower motor neuron syndromes resembling ALS. Thc GM~ gangliosidcs ,|re acidophilic glycolipids that arc a constituent of neuronal cell membranes, inchiding ,lcrvc terminals and nodes of Ranvier (Pestronk 1991 ). The reported incidence of circulating anti-GM~ antibodies in patients with ALS has varied greatly from series 1o series, suggesting differences in laboratory t c c h n i q u c (Pcstronk et al. 1989). In typical ALS the rclevancc of anti-GM~ antibodies is unccrtain, since this finding may rcprcsent an epiphcnomenon, and lhe association has been most clearly recognised in lower motor neuron syndromcs (Pcstronk ct al. 1989). tto~vcvcr, antiGM I antibodies are found ill titres less tilnll [; 35(I ill a varictv of ncurological disorders, e.g., othcr autoimmulle diseases alnd a:,lmm( ~J )pathics, ' and inflanmlah)rv polyneuropatthics, and titrcs at this level :,lrc not clinically significant (Pcstronk 1991). l~amb and Patten (1991) flmnd thai only one of 16 pclticnls with classical ALS had anti-GM I antibcMies in a iitrc gl-caicr than 1: 15(i. Pestronk c t a l . (19911) dcfincd three lower motor neuron syndronles associated with high tiircs ~l ;lilliGM~ antibodies in the serum: (l) A multifoca[ motor ncuropattl), with weakness of distal ;.irln muscles, principally affecting men younger than 45 years. Ncrvc conduction studies show multifocal conduction block. (2) A nlultifocaI motor neuropathy, with simihu clinical fcatulcS, but vvithout conduclit)n block. (3) An asymmetrical lowcr motor neuron syndrome, affecting proximal muscles more lhan distal, mostly found in tncn older than 45 vcars. There is no conduction block. In practice, the specificity of these syndromes is difficult to define. It has bcen suggested that the triad of asymmetrical lower motor neuron wcakncss, motor conduction block and high anti-GM~ antibody litrcs represents a treatable syndrome. ('yclophosphan]idc therapy for 3 to 6 months nlay result ill improvenlcnt, correlating with reduction in sel'unl anti-tiM I antibody titrcs to 311r+~ of previous lcvcls. In considering the clinical significance of anti-GM I antibodies associated with these neurological syndromes, it should bc noted that there is no simple relation bctwccn the occurrence of peripheral neuropathy and nlonoclonal gamnlopaithy, witl~ or without anti-glycoprotein aniiboclies, and
10 that many patients with monoclonal gammopathy do not develop neuropathies.
The neurotoxic hypothesis Several geographic foci of ALS are recognised. In Guam, the Kii peninsula of Japan, and in Western New Guinea there is a high incidence of ALS. These ALS syndromes are associated with other neurodegenerative syndromes, e.g., the A L S / p a r k i n s o n i s m / d e mentia complex, but ALS occurs in these areas without features of these other neurodegenerative disorders (Hudson 1981). The causation of these syndromes is controversial. The suggestion that the Guam ALS syndrome resulted from the ingestion of a dietary excitotoxin, BMAA, found in cycad nuts, which were used as a form of flour (Spencer et al. 1987), has been disputed on the grounds that insufficient of the active substance is present in cooked food to act as a neurotoxin (Duncan 1991) and because, with changing dietary habits, the disease in these endemic areas is not decreasing in incidence. There may thus be genetic factors as well as environmental factors. In addition, oral BMAA is not neurotoxic to mice (Perry et al. 1989). Guam ALS shows neurofibrillary tangles and granulovacuolar degeneration in neurons in subcortical areas of the brain, features that differ from the typical sporadic or familial ALS syndrome in Western countries. Dietary calcium depletion with coincidental aluminium intoxication has been employed as a model of the diet of the Chamorro people of Guam, in experiments with non-human primates (Garruto et al. 1989). Motor neuron degeneration, with accumulation of phosphorylated neurofilaments, and with basophilic inclusions, were found in these experimental studies (Garruto et al. 1989), but the relevance of these findings to sporadic ALS is uncertain. Affected motor neurons contained excess aluminium. Garruto et al. (1989) suggested that this resulted in disruption of intraneuronal cytoplasmic membranes, and thus in the neuronal cytopathology. Lathyrism, an acute or insidious spastic paraparesis, developing in malnourished people, sometimes associated with sensory symptoms at the onset, is a syndrome that occurs when wheat flour is unavailable and flour made from the chickling pea, Lathyrus sativus, is substituted in the diet. This flour contains an excitotoxin, B O A A (/3-n-oxalylamino-L-alanine), that is capable of inducing corticospinal tract degeneration in non-human primates (Spencer et al. 1987). This experimental disorder mimics the human disease. A cytotoxic effect on motor neurons, however, is not established in human lathyrism (Streifler et al. 1977). Since the characteristic combination of upper and lower motor neuron
involvement does not occur in lathyrism, this disorder is not a neurotoxic model for sporadic ALS. Plaitakis et al. (1988) found abnormalities of glutamate metabolism, with increased blood levels of glutamate and reduced glutamate content in brain and, particularly, in spinal cord in sporadic ALS. They suggested that over-excitation of glutamate receptorbearing cells in the CNS resulted in degeneration and death of these cells, and thus in the clinical features of ALS. Increased taurine levels have been reported in CNS tissue, perhaps secondary to excitotoxic neuronal damage (Plaitakis et al. 1988). Rothstein et al. (1990) confirmed Plaitakis' (1991)) findings, but Perry et al. (1990) found that CNS tissue levels of these amino acids in ALS did not differ from those of controls. Young (1990)concluded that the excitotoxic hypothesis of the causation of ALS was not proven. A similar controversy surrounds the observation that plasma cysteine levels are raised in ALS and Parkinson's disease (Steventon et al. 1988), a finding that Perry et al. (1991) could not confirm.
Genetic aspects of ALS The familial occurrence of ALS suggests the possibility of a genetic factor not only in these cases but also in the sporadic disease. In late onset disorders family studies are likely to be incomplete and, inevitably, there has been much speculation concerning the genetic hypothesis of ALS. In spinal muscular atrophy (SMA) the genetic locus has been mapped on chromosome 5ql 1-2-13.3 (SMA5q). This locus has been found both in acute, infantile onset SMA (Werdnig-Hoffmann disease) and in the juvenile onset (KugetbergWelander disease) variants. Further. no evidence of allelic heterogeneity between the acute and chronic forms of the disease has been detected (Munsat et al. 1990), although there may be alletic heterogeneity within the chronic late onset form itself. In ALS the search for a genetic marker has proven more difficult. but Siddique et al. (1991) have demonstrated linkage with a gene located on the long arm of chromosome 21 (21q 22.1-22.2). Linkage of the disease with this locus in informative families is currently being explored. There appears to be genetic heterogeneity, within the locus. Although this locus on chromosome 21q is close to that associated with the locus for the amyloid precursor protein gene that is important in one form of familial Alzheimer's disease, the two loci are far apart (Goate et al. 1991). Genetic heterogeneity in ALS is consistent with the differing phenotypes observed in clinical practice, and might result from the interactive expression of several dominant genes within the locus. as in any oligogenic disorder. This hypothesis will be important in formulating concepts of the pathogenesis
11 of sporadic A I £ , and pcrhaps offer the possibility of deducing the naturc of the abnormal protein product o f thc defective gone(s). T h c locus f o r X-linked, adult-onset, bulbo-spinal SMA (Kenncdv syndromc) has been identified on the long arm of the X c h r o m o s o m e in linkage with th0 :indrogcn receptor gonc. The latl0r is especially i n t e r csting in that gynaocomastia and reduced fertility are features of this syndrome (La Spada ot al. 1991).
Clues front pathology The principal pathological features of ALS arc well rccogniscd ( t t o l m c s 1909: Brownell ct al. 1970). Firstly, there is loss of antcrior h o r n cells, and of somatic motor neurons ii1 lhc lower cranial nerve nucl0i. The ocular motor nuclei, and thc $ 2 - 3 0 n u f nucleus that inncrvatcs the striatcd pelvic floor sphincter muscles arc rolati\cly spared ( M a n n c n o t al. 1977) but u n d e r g o degeneration hltcr in the course of the disoasc (Kihira c t a l . 19q I). ( } a n l n l a m o t o r ncurons are also inwHvcd, since there is h~ss o f thc m o t o r innorvation of intrafusal musclc fibres in muscle spindlcs (Swash and Fox 1974). The widespread pattcrn of loss of neurons in the antcrior horn stlggosls that intcrncurons are probably also affectcd. These features imply that thcrc is a degrce of selective vulnerability within the m o t o n e u renal population in the spinal cord and brain stcm (Swash ct al. 1986). Sccondly. there is loss of Bctz cells in the ccrcbral cortex, and degeneration of the corticospinal tracts ( t t o l m c s 19(19). ('orticospinal tract dogcncration, like loss of a n t e r i o r h o r n cells, is often strikingly asymmetrical, a fcaturc that parallels the clinical findings (Swash ct oil. 19SN). No relationship has bccn ostablishcd between the development of the u p p e r and lower m o t o r neuron fcalurcs. Thirdly, other neural systems arc inwHvcd in the disease. For cxamplc, there is d c g c n c r a tion of the spinoccrobcllar pathways in the cord and brain stcn~ (Swash et al. It,iNN), with loss of neurons in ('larkc's column (Avcrback and C r o c k e r 1982) and, in some cases, in t h a l a m u s , corpus callosum and superior
colliculi. This pathological cvidcncc of widespread involvement of neural systems in the central nerwms systcm, beyond the motor system, has bocn confirmed by a magnetic rcsonancc imaging study of ALS in which high intensity Icsions wcrc found widcly distribut0d in the white matter of the brain and brain stem (Salcs-Luis ot al. 199(I). In addition, abnormalities have becn reported in surat nervc biopsies, consisting of axonal atrophy, romyclination and a shift in thc diameter spcctrnm toward smaller diameter fibres (Dyck ci al. 1975: Heads c t a l . 19ql). At autopsy there is often slight pallor in the medial parts of the gracile columns, a featurc that is said to be more promincnt in familial cases.
lntraneuronal cytoplasmic inclusion bodies in ALS A n u m b c r of diffcrcnt types of intrancuronal inchisions (Martin ct al. 1990) have bccn recorded in anterior h o r n cclls in A L S ( T a b l e 1). 13unina (1962) doscribcd intracytoplasmic acidophilic inclusions, 2 3 # m in diameter, that a p p e a r as electron-dense, anlorphous material, often with a vesicular appc;,lrance, in thc clcctron microscopc (Chou 1979: Hirano ct al. 1984). Bunina bodies are found in thc pcrikarya, in the axes, and in the soma of motor ncurons, t | i r a n o and ]wata (1979) found B u n i n a bodies in 1(t o f 20 cases of AI,S. Howcvcr, Bunina bodies are not specific to ADS ( ( ' h o u 1979). Axonal spheroids are c o m m o n in the discasc ((Tarpcnter 1968; Hirano ct al. 1984). Spheroids contain neurofihtmontous material, consisting of filaments Ill nln in d i a m c t c r ,
arrangcd
individually
o r in s m a l l
bundles, often associ:,ited with other cell organcllcs. Those neurofilamcntous conglomerates are associated with incrcascd conccntrations of aluminium (Kihira ct al. 19911. P h o s p h o w l a t c d cpitopcs arc prcscnt in increased concentration in some degenerating anterior horn coils ( M a n c t t o ct al. 1088), but this appears to be a nonispccific finding (Leigh ct al. 1988). Ubiquitin, a highly conscrvcd hcal shock protcin of molccuhir weight 8565 kDa is associated with soxoral diffcrcnt types of intermediate l'ilamcnl inclusion hod-
TAFII.I{ I IN('I.USI()N l'l()l)lES IN AI,,%(FROM MARTIN ET AI,. 1991)) Inclusion
Appcalance
('omposition
{Jhiquifinalud hotlies l]tillina hodics llclsophilic hodi~:s
aggregai.csor skein-like [ilalllerlls I lfl/xrn diamclcr: 7 |ilamcnh)us at EM cosinl~philic, round or ¢hmgatud, 1-4 #m diameter, variabh_' LM appear:.lllCC hcisophilic, globoid or asteroid, 3 4/xm cliamcler. Irregular chains of rough
ubiquithl core ill)l kiloWil nol klll)'~VI1
[ lhano
bodies
Spheroids
[:,R. glycogen and vesicles as EM eosin~)philic, hyaline, rod-shaped. 4 6 /xm diameter. Paracryslalline arrays olh lit nm filaments at EM cosimiphilic, hyaline, argyrophilic, rounded, I(I ll}0#m diamcier. Ill nm filaments a! EM
actm, acfinin, tau. xincnlm, [ I'Ol]OlllyOSill n e u r o f i l a l l l c n l
