Article Text

Download PDFPDF

Cause of death and clinical grading criteria in a cohort of amyotrophic lateral sclerosis cases undergoing autopsy from the Scottish Motor Neurone Disease Register
  1. K M Kurian1,
  2. R B Forbes2,
  3. S Colville1,
  4. R J Swingler1
  1. 1
    Department of Neurology, Ninewells Hospital and Medical School, Dundee, UK
  2. 2
    Department of Neurology, Craigavon Area Hospital, Portadown County Armagh, UK
  1. Dr K M Kurian, Neuropathology, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2QQ, UK; kathreena.kurian{at}


Background: The Scottish Motor Neurone Disease Register is a population based register of amyotrophic lateral sclerosis/motor neurone disease (ALS/MND) in Scotland, with high case ascertainment levels.

Objective: To investigate the cause of death by autopsy and assess grading criteria in a cohort of cases of ALS from the Scottish MND Register.

Methods: The records of 44 patients undergoing autopsy were reviewed to determine the cause of death, clinical assessment (El Escorial and modified World Federation of Neurology criteria) during life and neuropathological autopsy findings.

Results: In a cohort of 44 cases undergoing autopsy between 1989 and 1998, the cause of death could be directly or indirectly (bronchopneumonia, aspiration/pneumonia and respiratory failure) attributed to MND in 32/44 (73%) cases. The clinical diagnosis of MND was confirmed at autopsy in 44/44 (100%) cases, 3/44 (7%) cases showed coexistent neurodegenerative disease and 5/44 (11%) were familial MND cases.

Conclusions: Within our cohort, MND contributes to death in the majority of cases and there is excellent clinicopathological correlation, irrespective of the clinical grading criteria used. However, the autopsy rate is low (4%) and further larger studies are required to identify heterogeneity within the disease.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Motor neurone disease (MND) was first clearly delineated by Charcot and Joffrey (1869). They described a condition in which there is primary degeneration of the upper and lower motor neurones, manifested by weakness and wasting of affected muscles with evidence of corticospinal tract involvement.1 Both the clinical and pathological diagnosis of this disease can be challenging and the criteria for both have changed in recent years.

Previous studies have found an overall misdiagnosis rate varying between 2% and almost 10%2 3 but these studies did not include autopsy findings. MND mimics include multifocal motor neuropathy, motor neuropathy, inclusion body myositis, cervical spondylitic myelopathy, Kennedy disease, spinal muscular atrophy, Pancoast tumour, postpolio syndrome, hereditary spastic paraparesis, multiple sclerosis and hyper or hypothyroidism.3 4

In the literature, one previous pathological study has reported that the categories of clinically definite and probable amyotrophic lateral sclerosis (ALS) according to the El Escorial criteria correlate with pathological confirmation of disease with ubiquitin positive neuronal cytoplasmic inclusions.5

We examined the cause of death at autopsy in 44 cases undergoing autopsy from the Scottish MND register6 between 1989 and 1998 and compared the neuropathological records with the clinical grading of disease.


Forty-four neuropathological and general post mortem records of cases that underwent autopsy were available for study with appropriate consent from the Scottish MND Register between 1989 and 1998. During this period, 1226 clinical cases of ALS/MND had been registered. The 44 autopsy cases were anonymised and full ethics approval was obtained. The reports were derived from the four main neuropathology centres within Scotland with the kind permission of the local pathologists and hospitals. The local clinicians had assessed these cases clinically. Two of the authors (RF and RS) classified the diagnosis after examining the medical records. Between 1989 and 1993, patients were classified according to the modified World Federation of Neurology criteria.7 Between 1994 and 1998, the El Escorial criteria were used.8 Information regarding age, sex, cause of death at main post mortem, neuropathological diagnosis and other findings were recorded by the local neuropathologists and the reports reviewed by KMK.


The results of the comparison between clinical and autopsy data with respect to MND are presented in table 1.

Table 1 Comparison between clinical and autopsy data with respect to motor neurone disease

Clinically there were 23 men and 21 women (table 1). Mean age at death was 65 years (range 22–92) and was the same for men (mean 65 years, range 51–82) and women (mean 65 years, range 22–92).

The causes of death recorded at the post mortem examination are displayed in table 2. The primary causes of death registered at the post mortem examination were: bronchopneumonia (n = 18), aspiration/pneumonia (n = 5), MND (n = 5), respiratory failure (n = 4), pulmonary oedema/congestive cardiac failure (n = 4), cardiac tamponade/acute myocardial infarction (n = 2), malignancy (n = 2), sepsis (n = 1), chronic obstructive pulmonary disease (n = 1) and epistaxis (n = 1). Other findings recorded at post mortem included the presence of familial MND (n = 5), frontotemporal dementia with MND (n = 1), Alzheimer-type pathology (n = 2), Lewy body pathology (n = 1), subependymoma (n = 1), Hashimoto’s thyroiditis (n = 1), chronic hepatitis (n = 1) and inflammation of the trigeminal nerve (n = 1).

Table 2 Causes of death recorded at the post mortem examination

The pathological criteria used in the reports reflected practice over the study period (1989–1998). In 24 cases the criteria used for MND diagnosis included significant microscopic loss of motor neurones and astrogliosis in the spinal cord, brainstem or motor cortex, in addition to atrophy of the anterior nerve roots with or without spinal cord atrophy alone. Ubiquitin positive inclusions were reported in 20 cases only. Inclusions were not subclassified into skein-like, Lewy-like or Bunina bodies. Immunohistochemistry for TDP-63, p62 a tyrosine phosphoprotein, was not performed and the presence of argyrophilic grains was not investigated in these cases.


The commonest proximate cause of death recorded at autopsy was bronchopneumonia in 18/44 (41%) cases. If aspiration/pneumonia (n = 5), MND (n = 5) and respiratory failure (n = 4) are added to this figure, it is likely that underlying MND contributed to death in 32/44 (73%) cases. Moreover, the clinical diagnosis was confirmed pathologically in all 44/44 (100%) cases. Within our cohort, the clinical diagnosis of MND, when care has been taken to exclude mimic syndromes, is very specific for the pathological process of MND when using the modified World Federation of Neurology and El Escorial criteria. These findings are interesting and may suggest that either ALS is a homogenous disease or that perhaps, given the clinical heterogeneity, that all patients reach the same pathological end stage. Although our sample size is limited (4% of the total number of 1226 clinical registered cases during the period), the demographic characteristics of our autopsy series matches those of the register as a whole.9

For the first 5 year cohort (1989–93), the modified World Federation of Neurology criteria were used, and for the second 5 year period (1994–98), the El Escorial criteria were used. There does not appear to be any difference in the specificity of the two sets of criteria but the series is small. The high level of clinical correlation for clinically diagnosed MND is different to the imperfect clinicopathological correlation observed in other neurodegenerations such as dementia or Parkinson’s disease. In Parkinson’s disease, the “error” rate for a clinical diagnosis may be between 10% and 20%.10 11 In dementia, the clinicopathological correlation is even less.3 It is possible that the high level of clinicopathological correlation in MND is an effect of the striking clinical picture, and the propensity for the disease to predominantly affect the voluntary motor system to the exclusion of others. In Parkinson’s disease, cognitive, autonomic or oculomotor features may escape clinical detection and obscure a correct pathological diagnosis of Lewy body dementia, multiple system atrophy or progressive supranuclear palsy.10 12 13

The coexistence of Alzheimer’s disease or Alzheimer-type pathology has been previously described in approximately 30% of cases of ALS with dementia and in the same study significant Alzheimer’s disease pathology was identified in 2/8 (13%) cases without frank dementia.14 The observation of Lewy body pathology in 1/44 (2%) case in our cohort is interesting, as we were not aware of any movement disorder in this patient. Coexistence of Lewy body pathology and MND has been described previously in the literature.15 16 The incidence of familial MND in 5/44 (11%) cases within the cohort may reflect selection bias in terms of the wish of relatives to be involved in research and the awareness of a genetic basis to the disease.4

Pathological assessment of MND has changed considerably since the study period but this would not have altered the final diagnosis of MND.17 18 Only 20/44 (45%) cases had documented the presence of ubiquitin positive cytoplasmic inclusions. Current best practice would involve subclassification of inclusions and immunohistochemistry for TDP43 and p62 as a minimum.19 20 Full assessment of non-motor systems post dated the practice during the study period21 although in one case an associated frontotemporal dementia with ubiquitin inclusions was identified.

In summary, in a cohort of cases undergoing autopsy that were available for study within the Scottish MND register, in over 32/44 (73%) cases the cause of death could be attributed to the underlying MND. Clinicopathological correlation within the small cohort was excellent, irrespective of the clinical grading criteria used. However, larger autopsy studies are required to confirm the high diagnostic specificity of the El Escorial and the modified World Federation of Neurology criteria and to further investigate coexistent pathology and heterogeneity in the disease.


We would like to thank the deceased and their families who consented to research at a time of grief; the local neuropathologists at Aberdeen Royal Infirmary, Glasgow Southern General Hospital and the Western General Hospital, Edinburgh; and the Scottish MND Association.



  • Competing interests: None.

  • Ethics approval: Full ethics approval was obtained.