In 1906, Alois Alzheimer described a 51-year-old lady with dementia, and senile plaques and neurofibrillary tangles at autopsy. Originally Alzheimer's disease was viewed as a rare pre-senile dementia. However, clinicopathological studies in the 1960s demonstrated an overall relationship between dementia and the presence of senile plaques in both young and elderly demented patients (Blessed et al. 1968), and the view emerged that Alzheimer's disease and senile dementia of the Alzheimer type was a single disease.
It is now recognized that Alzheimer's disease is the major cause of dementia. Epidemiological studies indicate a doubling of the prevalence of dementia with each decade above 65 years to a prevalence of nearly 50% in those aged 85 and above (Evans et al. 1989). Approximately 70% of cases of dementia will be due to Alzheimer's disease, either alone or in combination with vascular disease.
Up to 40% have a family history of an affected first-degree relative (Farrer et al. 1990); rarely there is a clear autosomal dominant history. Pathogenic mutations in three different genes have been identified in this group, namely presenilin 1 and 2 and amyloid precursor protein (APP) genes, which account for up to 50% of the autosomal dominant familial Alzheimer's disease cases described (Cruts et al. 1998). Inheritance of an E4 allele of the apolipoprotein E gene is associated with an increased risk of developing Alzheimer's disease (Corder et al. 1993).
Alzheimer's disease is a disorder of middle and late life. Early onset cases are described in the fourth and fifth decade but these are rare and almost exclusively familial. The clinical features of familial Alzheimer's disease associated with mutations in the APP and presenilin genes are broadly similar to sporadic disease, apart from the age at onset. However, cases with mutation at APP 692 have more amyloid angiopathy with cerebral haemorrhages and, thus, share similarities to hereditary cerebral haemorrhage with amyloidosis of the Dutch type due to mutations at APP 693.
The classical presentation of Alzheimer's disease is with memory impairment and, in some patients, there may be a relatively prolonged course, with isolated memory deficits until late into the disease.
Dyspraxia is generally a late feature (Della et al. 1987), although ideomotor dyspraxia is often found if specifically sought. Visuospatial and visuoperceptual deficits are also prominent and, in some patients, may be the presenting feature. Patients are quite often unaware of their cognitive deficits, often being brought to the attention of doctors by their relatives. The denial or anosognosia is not related to severity (Feher et al. 1991).
The general neurological examination is relatively normal in Alzheimer's disease at presentation, although motor abnormalities of extrapyramidal type emerge in up to 60 per cent in some series (Sulkava 1982). Increased muscle tone or Gegenhalten is the main feature and appears unrelated to the nigrostriatal deficit seen in Parkinson's disease (Tyrrell et al. 1990a), although many patients with additional bradykinesia are found to have dementia with Lewy bodies (Perry et al. 1996). Primitive reflexes such as the instinctive grasp reaction, rooting, and sucking occur late. Generalized seizures occur in 10–20 per cent over the total course of the disease, and myoclonus is seen in approximately 10 per cent of younger cases, and in some it may be a prominent feature (Faden and Townsend 1976).
Structural neuroimaging characteristically shows medial temporal lobe atrophy which can be seen on CT scan (Jobst et al. 1992) or, more specifically, hippocampal atrophy on MRI (Kennedy 1998) (Fig. 1). Functional imaging with SPECT or PET reveals a posterior biparietal bitemporal pattern of hypometabolism. EEG shows slowing and loss of alpha rhythm relatively early in the disease. Many CSF markers have been sought, and increased tau and decreased Aβ1–42 is claimed to be of value (Hulstaert et al. 1999) but is not yet in routine clinical practice.
Fig. 1. Coronal T1-weighted MRI, showing progressive hippocampal atrophy in a patient with Alzheimer's disease.
The histopathological hallmarks of the disease are neurofibrillary tangles and senile plaques. Neurofibrillary tangles are intraneuronal and found predominantly in the allocortex and temporoparietal neocortex (Fig. 2). There is a predilection for pyramidal cells to be involved, particularly in layers 3 and 5 of the neocortex, the CA1 layer of the hippocampus, subiculum, and layers 2 and 5 of the entorhinal cortex. Braak and Braak (1991) quantified regional tangle formation in normal elderly and mild to severe Alzheimer's disease cases, and suggested a staging system with progression of the disease from entorhinal cortex and hippocampus to neocortex. Neurofibrillary tangles consist of paired helical filaments which can be seen under electron microscopy (Kidd 1963), with a filament diameter of 10 nm wound in a double helix with a periodicity of 160 nm. Neurofibrillary tangles are also found within the dystrophic neurites of senile plaques. The major component of the paired helical filament has been shown to be the microtubule-associated protein, tau, which is abnormally phosphorylated (Lee et al. 1991). All six tau isoforms are deposited in Alzheimer's disease (Goedert et al. 1992), which distinguishes the tau pathology from that found in progressive supranuclear palsy (PSP), Pick's disease, and the other tauopathies.
Fig. 2. (a) Senile plaques (arrows) and neurofibrillary tangles (arrowheads) in the hippocampus in Alzheimer's disease (modified Bielschowsky silver impregnation; 300). (b) Positive immunostaining of (i) plaques (250) and (ii) blood vessels (250), with an antibody to A4 protein. Avidinbiotin complex method. (Antibody to A4 kindly provided by Professor B. H. Anderton, Department of Neuroscience, Institute of Psychiatry, London; the micrographs were kindly provided by Professor P. L. Lantos, Department of Neuropathology, Institute of Psychiatry, London; the photographic assistance of Mr A Brady, Department of Neuropathology is acknowledged.)
Senile plaques are also found predominantly in neocortical association areas and consist of glial processes, abnormal nerve endings or dystrophic neurites, and a central core of β-amyloid; they vary between 25 and 200 µm in diameter. β-Amyloid is also deposited in cerebral blood vessels. The β-amyloid protein has been isolated and shown to be derived from a much larger transmembrane molecule, the amyloid precursor protein (Kang et al. 1987). Diffuse plaques are not associated with dystrophic neurites and are believed to precede the classical mature plaque.
The central amyloid core consists predominantly of the Aβ1–42 species (see below), which may act as a nidus for subsequent fibrillary amyloid deposition (Iwatsubo et al. 1995).
Neuronal cell loss is also maximal in the hippocampus and association areas of the neocortex. Cell loss also occurs in subcortical nuclei, which include the amygdala and the origins of the subcortical projection systems, the nucleus basalis of Meynert, the nucleus raphe, and the locus ceruleus. The damage to the nucleus basalis and septal nuclei results in the cholinergic deficit, a consistent but not specific feature of Alzheimer's disease. Replacement of the cholinergic deficit is the basis of current symptomatic treatment. There is an overall association between cell loss and the histological features of plaques and tangles, and it is assumed that cells that contain neurofibrillary tangles are degenerating. Other histological changes found predominantly in the hippocampus include eosinophilic Hirano bodies and vacuolar changes in cytoplasm referred to as granulovacuolar degeneration. The severity of the histological changes of senile plaques and neurofibrillary tangles shows an overall association with severity of dementia. They are not, however, specific to Alzheimer's disease and neurofibrillary tangles in the hippocampus are found commonly in normal old age, as are limited numbers of senile plaques throughout the cortex. This has led to quantitative criteria for diagnosis (Khachaturian 1985) which have largely been superseded by the CERAD (Consortium to Establish a Registry of Alzheimer's Disease) criteria, based on numbers of neuritic plaques irrespective of neurofibrillary tangles (Mirra et al. 1991).
The cause of Alzheimer's disease is still not established, but considerable advances have been made which identify a central role for amyloid deposition. This is best understood in patients with autosomal dominant familial Alzheimer's disease. The rare mutations in the APP gene either increase the total amount of β-amyloid produced or alter the processing to favour amyloid deposition. The APP 670–671 mutation increases the total amount of β-amyloid peptide produced from the precursor protein, and is similar in this regard to Down syndrome, where there is an increased amount of amyloid produced due to a gene dosage effect arising from the trisomy 21. The amyloid precursor protein is normally cleaved at putative β- and γ-secretase sites to release the 40-amino-acid Aβ peptide, whose physiological function is unclear. It can be measured as a soluble peptide in both plasma and CSF but will form fibrillary aggregates of β-amyloid in senile plaques. Mutations at APP 717, close to the putative γ-secretase site, result in a subtle increase in the proportion of molecules extended at the C terminus of Aβ1–42 (Fig. 3). Aβ1–42 has a greater propensity to fibril formation and can act as a nidus for subsequent Aβ1–40 deposition. The familial Alzheimer's disease cases associated with mutations in the presenilin 1 and 2 genes (Cruts and Van Broeckhoven 1998a) also result in altered metabolism of APP and a relative increase in Aβ1–42; there has even been speculation that presenilin 1 may be the γ-secretase. ApoE4, also known to be an important genetic risk factor, is associated with enhanced amyloid deposition and may stabilize fibril formation.
Fig. 3. (a)Metabolism of APP. (b) Metabolism of APP with extension at the C terminus.
These lines of evidence all support a central role for amyloid deposition and the ‘amyloid cascade’ hypothesis (Hardy and Higgins 1992). This model predicts that the neuronal loss and neurofibrillary tangle formation are secondary events, but as yet, it is not known why amyloid is neurotoxic. Recent evidence for this central role is provided by transgenic mouse models arising from overexpression of a human mutated APP gene, which result in senile plaque formation which can be prevented by immunization with Aβ1–42 (Schenk et al. 1999). The secondary role for neurofibrillary tangles in this pathogenic cascade does not diminish their potential importance as a final common pathway. Moreover, in some of the hereditary tauopathies presenting with frontotemporal degeneration, mutations in the tau gene are the key pathogenic event.
Advice, support and a sensible prognosis for the carer should be available, and information can be obtained from national Alzheimer's Disease Societies and Alzheimer's Disease International. Treatment of systemic disease such as infections and heart failure are important as these may cause further deterioration in cognition, and patients with dementia are very sensitive to the cognitive side-effects of drugs. Agitation and delusional symptoms may need to be treated, but neuroleptic medication should be kept to an absolute minimum (McShane et al. 1997). Depression is a common early feature of Alzheimer's disease and a trial of a selective serotonin re-uptake inhibitor (SSRI) is indicated if there is a clinical suspicion of depressive symptomology. Tricyclics should be avoided because of anticholinergic side-effects. Patients who hold a driving licence should inform the national driving authority at the time of diagnosis, according to national guidelines, but the decision of when to stop driving will depend upon their cognitive function (Waldemar et al. 2000).
There is no treatment to affect the progression of Alzheimer's disease. A trial of selegiline and vitamin E, both alone and in combination, suggested a possible delay in progression (Sano et al. 1997). Although this trial has not been replicated and there were methodological problems, many patients are prescribed vitamin E in doses of up to 2000 units daily. Inhibition of cholinesterase activity in the brain will enhance levels of acetylcholine and thus ameliorate the cholinergic deficit arising from degeneration of the ascending projections from basal forebrain to hippocampus and the neocortex. Tacrine (Cognexreg;) has been largely superseded by donepezil (Ariceptreg;) and rivastigmine (Exelonreg;). The symptomatic effects are modest, but improvement in cognitive measures and clinical global impression is seen in 20–25 per cent of patients over and above a placebo response (therapeutic gain) (Rogers et al. 1998; Rosler et al. 1999). There is no evidence of any effect on disease progression.