Biopsy Pathology of Neurodegenerative Disorders in Adults

Definition and Synonyms

Alzheimer disease (AD) is the prototype neurodegenerative disease associated with adult-onset dementia. Initially thought to have only “presenile” (i.e., before old age) onset, AD was later expanded to include patients with “senile” onset of dementia (synonym: senile dementia of Alzheimer type) and identical neuropathology.

Brief Historical Overview

Alois Alzheimer presented a clinicopathologic case report of presenile dementia in 1906, in which he described severe gross atrophy and the microscopic presence of neuritic plaques (NPs) and neurofibrillary tangles (NFTs) (demonstrated by Bielschowsky silver stain) in the cerebral cortex of a 55-year-old woman who suffered from a progressive dementing illness for at least 5 years. Prior to Alzheimer's article, NPs had been reported by Blocq and Marinesco in 1892 in aging brains. Nonetheless, the association of NFT with dementia was first made by Alzheimer, for which Kraepelin later proposed the eponym for the disease.

Incidence and Demographics

AD is the most common (60% to 80% of total) neurodegenerative disorder in adults, with either senile or presenile (younger than 65 years of age) symptomatic onset. Both incidence and prevalence of AD increase exponentially with age, with prevalence rates in the United States estimated to reach 9.7% and 37.4% among elderly individuals older than 70 and 89 years of age, respectively. The strongest risk factors are age and family history, followed by low education level, cardiovascular disease, and prior head injury. Women have a slightly higher incidence of AD than men in advanced age.

Clinical Manifestations and Localization

AD is recognized in three clinical stages in a continuum: (1) preclinical, (2) mild cognitive impairment (MCI), and (3) dementia. Only a fraction of patients with MCI (especially those with amnestic MCI) develop clinical AD in 5 to 10 years. Clinical onset of sporadic AD usually occurs in the seventh decade of life, but patients with familial AD may present as early as in their 20s. Symptoms of AD vary, but often reflect global dysfunction and do not localize to specific regions of brain. Early progressive loss of short-term memory with preservation of social skills is common. CSF levels of total and phosphorylated tau are elevated, while that of Aβ ₄₂ is reduced. Over a period of 6 to 10 years, patients gradually lose remote memory, calculation, speech, ability to perform complex tasks, and, eventually, the basic skills of daily living, ultimately becoming totally dependent on others for their care. Direct cause of death is usually related to malnutrition, dehydration, and infection.

Radiologic Features and Gross Pathology

Brain imaging studies (structural or functional) of AD patients show varying degrees of atrophy, affecting cerebral cortex globally (occipital lobes and paracentral lobules are usually less affected), hippocampus, and amygdala, with corresponding white matter atrophy and ventricular dilatation (hydrocephalus ex vacuo) ( Fig. 27.1 ). Subcortical deep gray matter (basal ganglia and thalamic complexes) usually appears less affected, except in advanced disease. Cortical atrophy is usually most striking in mesotemporal areas, with variable involvement of parietal and frontal lobes, but this overlaps with that seen in nondemented age-matched controls. Atrophy of the hippocampus and amygdala quantified by MRI volumetry has been reported as a predictor of subsequent dementia in elderly patients who are followed prior to the appearance of cognitive impairment. Nuclear imaging modalities, such as positron emission tomography (PET) using ¹⁸ F-FDG, show hypometabolism in the temporoparietal region of AD. Amyloid PET using various tracers (e.g., Pittsburgh Compound B or PiB) shows increased Aβ deposition in brain parenchyma, corresponding to early and precipitous reduction of Aβ ₄₂ level in CSF of these individuals, and holds promise for detecting preclinical cases of AD. The radiographic abnormalities so detected are not disease specific, and gross examination of a biopsy specimen is not instructive.

Histopathology

The Consortium to Establish a Registry for Alzheimer Disease (CERAD) first met in 1986 with the goal of standardizing its diagnostic criteria. This led to the development of the CERAD criteria that rely mainly on a semiquantitative assessment of NP densities correlated with the patient's age and clinical presence or absence of dementia. Importantly, these early diagnostic criteria did not include NFT density in various brain regions as part of the diagnostic algorithm. In 1997 the NIA-AA issued a set of consensus postmortem diagnostic criteria for AD, incorporating both NP density and NFT stage in the clinical context. These criteria were updated in 2012 based on analysis of 562 cases registered at the National Alzheimer's Coordinating Center. The term AD pathologic changes refers only to neuropathology found at autopsy, which should include a careful assessment of other potentially contributing abnormalities (e.g., vascular injury, Lewy body pathology, hippocampal sclerosis, cerebral amyloid angiopathy [CAA], and transactive response DNA-binding protein 43 kDa [TDP-43] pathology). Three parameters were used to calculate a composite ABC score of AD pathologic changes: (1) the distribution phase of Aβ deposition, referred to as the Thal phase; (2) Braak neurofibrillary stage; and (3) CERAD NP score. Each ABC score is assigned a level (no, low, intermediate, high) of severity of AD pathologic changes, and it is recommended that intermediate and high levels of AD pathologic changes are adequate to explain the patient's dementia. Some nondemented elderly individuals have been found to show AD pathologic changes focally. On the practical level, advanced AD is readily recognized at autopsy with widespread NFTs in nontemporal isocortex (Braak stages V and VI) and deep/brainstem nuclei as well as many NPs in the background.

The four microscopic features of AD include (1) neuronal loss, (2) gliosis, (3) deposition of β-amyloid peptides (Aβ, predominantly Aβ ₄₂ , derived from the transmembranous APP) in neuropil as diffuse and neuritic plaques (NPs) ( Fig. 27.2 ), and (4) formation of fibrillary polymers of ubiquitinated and hyperphosphorylated microtubule-associated protein tau (encoded by MAPT ) as NFTs and neuropil threads in neurons and their processes, respectively ( Fig. 27.3 ). Of these changes, gliosis is readily appreciated on routine histologic and IHC preparations, but is not disease specific. Severe loss of neurons is not difficult to recognize by those familiar with neuroanatomy, but mild neuronal loss may be difficult to detect since neuronal density on histologic sections does not correlate well with total neuronal count. Neurofibrillary change (NFTs and neuropil threads) and NPs are more specific for AD as its diagnostic hallmarks.

Based on the absence or presence of associated thick wavy dystrophic neurites filled with phosphorylated tau, Aβ plaques are classified as diffuse or neuritic types. Over time diffuse plaques mature into neuritic plaques, with the latter being more specific for AD. Diffuse plaques are not visible on H & E stain with or without LFB, and their visualization is greatly facilitated by silver (see Fig. 27.2A ) or amyloid stains. The latter include histochemical stains (e.g., thioflavin S) and IHC stains for CNS-specific Aβ peptides or its precursor protein APP ( Fig. 27.2B ). NPs appear as spherical eosinophilic structures in neuropil, measuring up to 200 µm in diameter, with ( Fig. 27.2C ) or without ( Fig. 27.2D ) a central amyloid core and variable numbers of associated dystrophic neurites ( Fig. 27.2D ), reactive astrocytes, and microglia. Rarely NPs may be associated with so many microglia and even lymphocytes (so-called “inflamed” NPs) ( Fig. 27.2E ) that they mimic microglial nodules and small epithelioid granulomas. An occasional NP core may show prominent radial spikes, but most are roughly spherical and otherwise amorphous. Some plaques consist of amyloid cores without surrounding dystrophic neurites or cellular reaction ( Fig. 27.2F ); these so-called “burned-out” plaques may mimic the amyloid plaques seen in prion diseases.

On the other hand, NFTs appear as perinuclear, curvilinear, flame-shaped or globular, pale basophilic profiles in neuronal cell bodies on H & E stain with or without LFB (see Fig. 27.3A ). These structures and neuropil threads are much easier to detect on silver stains ( Fig. 27.3B ) or on the thioflavin S fluorescent stain. The density of neocortical NFTs in AD is usually high (several per high-power field [HPF]), and they are characteristically concentrated in layers II (external granular layer), III (external pyramidal layer), and V (internal pyramidal layer). The shape and size of NFTs conform to the neuronal cell bodies in which they reside. Therefore, they appear flame shaped in pyramidal neurons and globose (resembling a ball of yarn) in large polygonal neurons. They fill the cytoplasm of the cell body, often partially or completely encircling the nucleus, and sometimes extending into proximal apical dendrites. An alternative for NFT detection is IHC using antibodies against phosphorylated tau ( Fig. 27.3C ) and ubiquitin ( Fig. 27.3D ), although the latter is considerably less specific since it stains many different types of neurodegenerative inclusions.

The relative importance of NPs and NFTs in AD pathogenesis has been the subject of intense debate. For practical interpretation of brain biopsies, all cases of AD would be expected to contain many NPs, neocortical NFTs, and neuropil threads (Braak and Braak stages V and VI; see following section), as well as abundant amyloid plaques (both diffuse and neuritic types). When global in distribution and relatively high in density, these changes are readily detected on frontal lobe biopsies from AD patients. It is important to note that NFTs may be found in the relative absence of NPs in the elderly, as confirmed by numerous studies. In 2014, Crary et al. proposed the term primary age-related tauopathy (PART) to describe the pathologic continuum ranging from focally (medial temporal) distributed NFTs in cognitively normal-aged individuals to widespread cortical NFTs (without accompanying NPs) in rare elderly individuals with dementia known as “tangle-only dementia.” The relation between AD and PART is still unclear.

One common coexisting neuropathology seen with AD is cerebral amyloid angiopathy (CAA) caused by deposition of Aβ (predominantly Aβ ₄₀ ) in the walls of cerebral blood vessels. Most examples are sporadic in nature; however, rare CAA cases are hereditary (e.g., hereditary cerebral hemorrhage with amyloidosis of the Dutch type or HCHWA-D) and caused by mutations in the APP gene or, even more rarely, in the PRNP gene encoding for human prion protein (PrP). Small and midsize blood vessels (arterioles and venules) in the leptomeninges and small arterioles and capillaries in cerebral cortex are preferentially affected, whereas vessels in the white matter are mostly spared. On routine H & E stain, affected small vessels often assume perfectly round cross-sectional configurations with eosinophilic, thickened, and rigid walls containing fibrillary or amorphous deposits ( Fig. 27.4A and B ). Affected cortical capillaries may become serpiginous profiles heavily impregnated by amyloid without residual endothelia or lumens (called “dyshoric” or “dysphoric” angiopathy; Fig. 27.4C ). Occasionally, abnormal vessels affected by CAA are actively inflamed (so-called amyloid-beta–related angiitis or ABRA) with associated foreign body giant cells, macrophages, neutrophils, luminal stenosis, and extravasation ( Fig. 27.4D and E ). Perivascular reactive gliosis, siderophages, and microinfarcts can be detected as well. Optically empty vacuoles or irregular spaces (presumably representing spaces left by degenerating vascular smooth muscle cells) surrounded by amyloid are a frequent finding in the walls of affected larger meningeal vessels ( Fig. 27.4F ). Congo red stain shows the typical apple green birefringence under polarized light, although this stain can be fickle and difficult to interpret. More reliable confirmatory stains for CAA are thioflavin S and IHC stains for APP or Aβ. Patients with CAA have an increased risk of hemorrhage, ranging from small cortical petechiae to massive (and recurrent) lobar hemorrhages. The former may be asymptomatic or be associated with symptoms of transient ischemic attacks (TIA), but the latter may present as a life-threatening event or neurosurgical emergency. Evacuated intracerebral hematomas often include scant leptomeninges and brain parenchyma that may show CAA. It should be emphasized, however, that CAA does not always coexist with AD.

Histologic Variants and Grading

The neuropathology seen in familial AD does not differ in form from sporadic disease, but is generally more severe, especially in regard to the deposition of Aβ. Therefore, suggestion for familial disease is based mainly on clinical grounds (young age of onset and positive family history). In 1991, Braak and Braak reported the topographic distributions of neurofibrillary change (NFTs and neuropil threads) in brains of 83 elderly individuals (demented and nondemented) and outlined their progressive appearance in six stages that correlated with clinical cognitive decline. Stages I (transentorhinal and entorhinal cortex) and II (CA1 and CA2 sectors of hippocampus) are transentorhinal stages; stages III (hippocampus and amygdala) and IV (thalamus, hypothalamus, and basal forebrain) are limbic stages; and stages V (association cortex) and VI (primary sensory cortex such as occipital) are neocortical stages. Subsequently, a good correlation between the Braak and Braak stage of neurofibrillary changes (as opposed to NP density) found at autopsy and the degree of clinical cognitive impairment has been confirmed by other researchers.

Differential Diagnosis

The most important differential diagnosis of AD is age-related changes (and PART) not associated with dementia. Both NPs and NFTs are found in brains of older individuals without neurodegenerative disorders. It is essential to be familiar with the neuropathology of aging in comparison with AD in terms of quantity and topographic distribution of these hallmarks. A limited distribution of NFTs and neuropil threads can be found in nondemented elderly individuals in mesotemporal cortex (entorhinal, transentorhinal, and periamygdaloid), hippocampus, amygdala, paralimbic cortex (cingulate and parahippocampal gyri, temporal pole, and anterior insula), substantia innominata, and locus ceruleus. A low density of NPs (mostly of the diffuse type) can be noted in these same nonsymptomatic aging individuals within allocortex and isocortex. These changes develop slowly after the age of 60 years and are considered by many experts to represent “normal” (as opposed to “pathologic”) aging. For this reason, a diagnosis of definite AD should not be made based solely on a limited histologic examination of hippocampal or other mesotemporal structures. In AD patients, these hallmarks develop ubiquitously in nontemporal cortex and in high densities (hence the analogy to “accelerated” aging) at the time of clinical diagnosis. In particular, the association cortex of frontal and parietal lobes are invariably affected. NFTs are often also found to a limited degree in the deep gray matter (neostriatum and thalamus) and substantia nigra of AD patients—a phenomenon almost never seen in normal aging. The pathologic distinction between normal aging and AD may not be clear cut. Multiple studies have found widespread AD pathologic changes in up to 20% of nondemented octogenarians, putting these individuals in the preclinical stage of AD.

Two relatively common neurodegenerative disorders of non-Alzheimer type (i.e., DLB and FTLD) may mimic AD clinically. The neuropathologic features of these two entities are detailed in later sections. A less common clinical mimicker of AD is vascular dementia caused by multiple old infarcts or diffuse white matter ischemia. Having AD does not preclude the possibility of other coexisting neurodegenerative disorders, the most commonly encountered being DLB. Also, AD is the most common, but not the only neurodegenerative disease with formation of NFTs. In fact, NTFs are found in more than 20 different neurodegenerative disorders with different densities, topographic distributions, and biochemical characteristics. Several of these so-called “tauopathies” show neocortical NFTs and should be carefully differentiated from AD in interpreting brain biopsy, including PSP, Pick disease, PART, and frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17).

Ancillary Diagnostic Studies

A reduced level of Aβ ₄₂ and elevated levels of total tau and phosphorylated tau in CSF are recognized biomarkers for AD. Amyloid PET and functional MRI studies are very helpful. When familial AD is suggested by a positive family history or young patient age, DNA sequencing for mutations in the three known causative genes after appropriate genetic counseling should be considered (see following section).

Genetics

Up to 10% of AD patients have familial disease. Many familial AD patients have presenile (before age 65, sometimes as early as in their third or fourth decades) onset of symptoms. Currently, mutations in three genes (presenilin-1 [PSEN1] on chromosome 14, presenilin-2 [PSEN2] on chromosome 1, and APP on chromosome 21) are known to cause autosomal dominant familial AD via dysregulated APP metabolism and deposition of Aβ ₄₂ in brain tissue. Together they account for 20% to 50% of familial AD cases, and clearly other unidentified genes are involved. DNA extracted from blood of individuals thought to have AD can be tested for mutations in these known genes, obviating the need for brain biopsy. In addition, ApoE4 is an ApoE allele on chromosome 19 known to be associated with an increased risk for late-onset, sporadic AD and with increased Aβ deposition in neuropil (Aβ ₄₂ ) and cerebral vascular walls (Aβ ₄₀ ). The tau gene (MAPT) on chromosome 17 encodes six tau isoforms via alternative splicing. Biochemical and immunohistochemical analyses of tissue from distinct tauopathies have revealed accumulation of different sets (4-repeat or 3-repeat) of tau isoforms. Interestingly, mutations in MAPT have been found in some (not all) families with FTDP-17, but not in familial AD.

Treatment and Prognosis

Currently, no treatment effectively improves cognitive function in individuals with advanced AD, but certain approved pharmaceutical agents (e.g., cholinesterase inhibitors and memantine) have been shown to slow memory impairment and disease progression, especially at its early stages. Research suggests that individuals with MCI may be in the earliest or preclinical stage of AD with Braak and Braak stages III and IV neurofibrillary change. Since 1999, immunization (active or passive) against Aβ ₄₂ has been shown to clear amyloid deposits in brain tissue in animal models and to prevent/reverse functional memory deficits of APP transgenic mice. However, human clinical trials of Aβ immunotherapy have so far yielded mixed results, sometimes complicated with encephalitis. These and other developments of AD treatment have been reviewed recently by Agadjanyan, Rafii, and coworkers.

Definition and Synonyms

DLB is characterized clinically by dementia associated with hallucinations, fluctuation in consciousness, parkinsonism, rapid eye movement (REM) sleep disorders, and sensitivity to neuroleptics, and is pathologically characterized by the presence of readily found Lewy bodies in cerebral cortex (both isocortex and allocortex), basal forebrain, brainstem, and spinal cord. Starting in the 1980s, the spectrum of Lewy body neuropathology has been delineated while various terms were coined for the dementing illness, including diffuse Lewy body disease, cortical Lewy body disease, Lewy body dementia, senile dementia of Lewy body type, and the Lewy body variant of Alzheimer disease. In 1996, after an international workshop, the Consortium on Dementia with Lewy Bodies recommended the name in current use and published a set of consensus guidelines for clinical and pathologic diagnosis of DLB, which were later updated and validated.

Idiopathic Parkinson disease (iPD), on the other hand, has been known since ancient times, with the synonyms of shaking palsy and paralysis agitans. Parkinsonism is a set of clinical symptoms (triad of resting tremors, truncal rigidity, and bradykinesia) that result from various conditions, including but not limited to several neurodegenerative diseases, exposure to toxins, chronic use of some neuroleptics, and head trauma. The neuropathology seen in iPD is degeneration of substantia nigra with formation of Lewy bodies in residual nigral neurons and in locus ceruleus. Mild to moderate dementia appears at late stages of iPD in up to 40% of patients. The DLB Consortium utilizes the term Parkinson disease with dementia instead of DLB when the dementia appears 1 year or longer after the initial appearance of parkinsonism.

Brief Historical Overview

In 1861, the French neurologist Jean-Martin Charcot named the disease after James Parkinson of London, England, who wrote the first comprehensive clinical account of the illness in 1817 to bring medical attention to this common disease. The deficiency of the neurotransmitter dopamine in the affected basal ganglia and the associated degeneration of the substantia nigra were discovered much later in the 1950s and 1960s. Since then, the effects of levodopa (1,3-dihydroxy- l -phenylalanine, or l -dopa) on increasing central dopamine levels and alleviating parkinsonism have been firmly established. Interestingly, Lewy bodies were first observed in the basal nucleus of Meynert and dorsal motor nucleus of the vagus, as reported in 1912 by Frederich H. Lewy. They were only later reported in the substantia nigra as “Corps de Lewy” by Tretiakoff in 1919. The neuropathology of DLB (abundant Lewy bodies in cerebral cortical neurons) was first described in 1961 by Okazaki et al. The similarity between cortical Lewy bodies and classic brainstem Lewy bodies found in iPD was immediately apparent. Many cases of DLB are misdiagnosed clinically as AD. In fact, DLB was considered a variant of AD due to the abundance of coexisting NPs often seen at postmortem examination. With better characterization of clinical manifestations at early stages of disease, neuroimaging findings, and cases showing pure DLB on pathology, many neuropathologists now believe that although DLB and AD may coexist in the same patient, they are two distinct clinicopathologic entities.

Incidence and Demographics

DLB is regarded by many as the second most common neurodegenerative cause of progressive dementia in the elderly, accounting for 10% to 25% of all cases confirmed by autopsy. Patients typically present in their 60s, with an average disease duration of 13 years and no clear-cut gender preference. A comprehensive review of the literature showed that the incidence of DLB ranges from 50 to 160 per 100,000 persons per year, and its prevalence ranges as high as 6350 per 100,000 persons; both of which increase with age.