Neuropathology and cognitive impairment in Alzheimer disease: a complex but coherent relationship

Abstract

Amyloid plaques and neurofibrillary tangles (NFTs) are the pathological hallmarks of Alzheimer disease (AD). There is controversy regarding the use of current diagnostic criteria for AD and whether amyloid plaques and NFTs contribute to cognitive impairment. Because AD is specific to humans, rigorous and comprehensiveclinicopathologic studies are necessary to test and refine hypotheses of AD diagnosis and pathogenesis. Neither the clinical nor the pathological aspects of AD evolve in a linear manner, but thepredictable sequence of AD pathology allows for stage-based correlations with cognitive deterioration. We discuss subsets of patients with clinical dementia who lack amyloid plaques and NFTs and, conversely, whether individuals without antemortem cognitive impairment can harbor severe AD-type pathological findings at autopsy. There are many medical, technical, and anatomical challenges to clinicopathologic studies in AD. For example, at least two thirds of persons older than 80 years have non-AD brain diseases that can effect on cognitive function. We argue that existing data strongly support the hypothesis that both amyloid plaques and NFTs contribute to cognitive impairment.

FIGURE 1

Histopathologic hallmarks of Alzheimer disease demonstrated with Bielschowsky silver impregnation. (A) Neuritic plaques are extracellular fibrillary amyloid deposits, surrounded by swollen, degenerating, argyrophilic neurites. (B) Neurofibrillary tangles are composed of intracellular, insoluble, and protease-resistant fibrillary polymers of tau protein. In both panels, there are wispy argyrophilic neuropil threads. Scale bars = 25 µm.

FIGURE 2

(A) An idealized clinicopathological correlation may not be applicable to Alzheimer disease. (B) A conventional metric for cognition, the Mini-Mental State Examination (MMSE) and the Braak staging method provide a relatively strong correlation between pathological severity and antemortem cognitive decline. Data depicted are from the University of Kentucky Alzheimer's Disease Center (63). Each patient in this series lacked concomitant pathological processes, for example, advanced cerebrovascular disease, and died within a year of their last MMSE score. Numbers of patients from each group: Braak Stage 0, n = 8; I, n = 15; II, n = 18; III, n = 13; IV, n = 12; V, n = 6; and VI, n = 25. Error bars are SD.

FIGURE 3

Alzheimer disease (AD) progresses in a nonlinear manner that renders clinicopathological correlation a challenge. The disease duration is long, there is a preclinical phase, in which pathological findings are present, but there are no clinical manifestations. Further, each patient has a unique constellation of clinical and pathological features. Nonetheless, there is a relatively predictable progression of both clinical and pathological indices. This schematic illustrates that the association between mesial temporal neurofibrillary tangles (NFTs) and neocortical neuritic plaques with cognitive decline may be weak, whereas the association between neocortical NFTs and cognitive decline is strong. Early mesial temporal NFTs may play a role in memory dysfunction in early AD.

FIGURE 4

(A) The epidemiology of Alzheimer disease (AD) and the duration of the disease process predict that many persons with appreciable AD pathology die before the onset of dementia. Assuming a 10-year preclinical stage, approximately as many individuals would have significant AD pathology without the clinical disease as would manifest the disease during life. The epidemiological data represent an average of 6 large studies, as summarized previously (16). (B) Empirical data confirm the prediction that many individuals who die of other causes show appreciable AD-type pathology. Empirical data are from 3,928 cases in the indicated age ranges (25). Note that Braak Stages III/IV correspond to National Institute on Aging–Reagan Institute “intermediate likelihood” for the diagnosis of AD (46).

FIGURE 5

Variations in clinical and pathological indices in Braak Stage VI cases from the University of Kentucky Alzheimer’s Disease Center database (63). Cases are arranged left-to-right by increasing Mini-Mental State Examination (MMSE) scores. All patients had undergone MMSE testing within a year before death. Patients with higher MMSE scores tended to have fewer neocortical neurofibrillary tangles (NFTs). One nondemented patient (MMSE score, 29) had a relatively low number of neocortical NFTs compared with the other more cognitively impaired patients.

FIGURE 6

Data from the University of Kentucky Alzheimer’s Disease Center database demonstrate that neocortical neurofibrillary tangle (NFT) counts predict dementia in patients without severe concomitant brain pathologies (n = 168). Neocortical counts comprised total counts from occipital (Brodmann area 17 and 18), inferior parietal lobule, superior and midtemporal gyri, and middle frontal gyrus (63). Neocortical amyloid plaque counts are the combined amount of neuritic and diffuse amyloid plaques in the same sections. Above a certain neocortical count (blue arrow), all patients were demented (blue diamonds = patients with the Mini-Mental State Examination [MMSE] <20). By contrast, patients with mild or no dementia (MMSE, >20; red circles) tend to have lower neocortical NFTs, and all patients below a threshold of neocortical NFT counts (shown by red arrow) have MMSE greater than 20. There is only a weak correlation between the amyloid plaques (green triangles) and the NFTs, but cases with high neocortical NFTs generally have high amyloid plaques.

FIGURE 7

Neurofibrillary tangles (NFTs) in ascending reticular activating system (ARAS) including the acetylcholinergic nucleus basalis of Meynert (NbM) do not map directly onto most clinicopathologic studies. (A) Many NFTs are stained using an antibody against phosphorylated tau protein in the NbM of a midstage Alzheimer disease (AD). Scale bar = 50 µM. (B) Other ARAS nuclei, including the tuberomammillary nucleus (TM), the locus caeruleus (LC), ventral tegmental area (VTA), and the brainstem raphe areas also show many NFTs and neuronal loss in AD patient. Neurotransmitters that are known to augment memory, arousal, attention, and mood are indicated alongside the ARAS cell groups that produce them. The ARAS neurons are the only source of those neurotransmitters in cerebral cortex, and thus pathological alterations in ARAS may contribute to clinical symptoms.