Cerebral Amyloid Angiopathy, App-Related

A number sign (#) is used with this entry because cerebral amyloid angiopathy (CAA) can be caused by mutation in the gene encoding the amyloid precursor protein (APP; 104760). Mutations in the APP gene can also cause autosomal dominant Alzheimer disease-1 (AD1; 104300), which shows overlapping clinical and neuropathologic features.

Other forms of CAA include the Icelandic type (105150), caused by mutation in the CST3 gene (604312), and the so-called British (176500) and Danish (117300) types, both caused by mutation in the ITM2B gene (603904).

Description

Cerebral amyloid angiopathy, or cerebroarterial amyloidosis, refers to a pathologic process in which amyloid protein progressively deposits in cerebral blood vessel walls with subsequent degenerative vascular changes that usually result in spontaneous cerebral hemorrhage, ischemic lesions, and progressive dementia. APP-related CAA is the most common form of CAA (Revesz et al. (2003, 2009)).

Clinical Features

Wattendorff et al. (1982) reported a Dutch family in which 11 members had cerebral and cerebellar hemorrhage and infarction at ages ranging from 44 to 58 years. Affected family members comprised 5 sibships spanning 2 generations. The principal clinical characteristic was recurring cerebral hemorrhages, sometimes preceded by migrainous headaches or mental changes. Of the 11 who presented with acute stroke, all who survived eventually developed dementia. In addition, there were 5 family members who developed dementia, with or without accompanying stroke. Neuropathology showed hyaline thickening of the walls of cortical arterioles in 6 autopsied cases and 1 biopsy specimen. The arteries of the arachnoid showed marked tortuosity, concentric proliferation, and focal hyalinization. Amyloid was demonstrated in the hyalinized vessels but was not found outside the nervous system. The kindred of Wattendorff et al. (1982) was from Scheveningen, the patients were descendants of a couple who married in 1871. Luyendijk and Bots (1986) wrote: 'As the hereditary disease is well-known to the co-members of the respective families they usually inform the doctors on the probable diagnosis themselves, when such a patient is admitted into the hospital. Besides which they usually add all kinds of genealogical information.' The disorder was referred to as HCHWAD for 'hereditary cerebral hemorrhage with amyloidosis, Dutch type.'

In studies of the Dutch form of hereditary cerebral hemorrhage with amyloidosis, van Duinen et al. (1987) demonstrated that the vascular amyloid deposits were related to the beta-protein associated with Alzheimer disease and Down syndrome (190685). The findings indicated that the 'Dutch type' is genetically distinct from the 'Icelandic type' of cerebroarterial amyloidosis (105150), which is due to a defect in cystatin C (CST3; 604312).

Luyendijk et al. (1988) described 136 patients with hereditary cerebral hemorrhage, all belonging to families originally resident in Katwijk, Netherlands. No genealogic connection had been established between the Dutch and Icelandic pedigrees. Moreover, the findings in all of the Dutch cases were identical and differed from the findings in the Icelandic cases in age at onset and involvement of cystatin C. Among 78 males and 58 females with HCHWAD, Luyendijk et al. (1988) found that the sex ratio for the proven cases was nearly equal (29 males and 26 females). There were numerous examples of father-to-son transmission.

Roosen et al. (1985) and Smith et al. (1985) provided case reports of patients with intracerebral hemorrhage or transient ischemic attacks, respectively, resulting from cerebral amyloid angiopathy.

Cosgrove et al. (1985) reviewed 24 cases of autopsy-proven cerebral amyloid angiopathy; death was caused by intracranial hemorrhage in 16. None had systemic amyloidosis.

Haan et al. (1990) found that all 16 patients they examined with the Dutch type of hereditary cerebral hemorrhage with amyloidosis had psychiatric abnormalities; dementia was present in 12. Three patients tested twice at an interval of some years exhibited progressive intellectual deterioration and memory disturbance; in 2 of them there was no evidence of intercurrent strokes.

Fernandez-Madrid et al. (1991) reported a 63-year-old woman of Dutch extraction living in the United States with HCHWA confirmed by molecular analysis. The patient was normotensive and was well until age 47 years, when she began to have attacks approximately every 2 weeks.

Iglesias et al. (2000) reported a patient of Spanish descent who presented at age 62 years with intracerebral hemorrhage in a background of progressive mental deterioration. Neuroimaging revealed fine tram-line bilateral occipital calcifications, extensive leukoencephalopathy, and bilateral external carotid artery dysplasia. Skin biopsy with ultrastructural study revealed novel changes in the basal lamina of capillaries, with multilayered appearance and round-shaped microcalcifications. Of 19 next of kin who survived beyond 60 years of age, 6 had brain disorders; 4 of the 6 presented at least 3 components of the syndrome. The proband's mother had died at age 83 with profound dementia; one sister, who was diagnosed with dementia with occipital calcifications and leukoencephalopathy at age 67, died 2 years later from intracranial hemorrhage; a brother had an occipital hemorrhage at age 58, at which time occipital calcifications and leukoencephalopathy were discovered; and another brother died after a minor stroke at age 70 with dementia, occipital calcifications, and external carotid artery dysplasia. Iglesias et al. (2000) suggested that this represented a novel familial cerebrovascular entity with widespread microvascular calcifications and presumably autosomal dominant inheritance. They suggested the acronym FOCHS-LADD, for 'familial occipital calcifications, hemorrhagic strokes, leukoencephalopathy, arterial dysplasia, and dementia.' Iglesias et al. (2000) emphasized that subjects had no seizures, facial angioma, or intracranial vascular malformation, and that arterial hypertension was neither constant nor severe.

Grabowski et al. (2001) reported a 3-generation Iowa family with autosomal dominant dementia beginning in the sixth or seventh decade of life. The proband and an affected brother had progressive aphasic dementia, leukoencephalopathy, and occipital calcifications. Neither had intracerebral hemorrhage. Neuropathologic examination of the proband revealed severe cerebral amyloid angiopathy, widespread neurofibrillary tangles, and extensive distribution of beta-amyloid-40 in plaques.

Using T2 gradient-echo MRI, van den Boom et al. (2005) identified microbleeds (less than 5 mm in diameter) in 18 (69%) of 27 patients with HCHWAD confirmed by genetic analysis. Three of the patients with microbleeds were asymptomatic. The microbleeds occurred at the gray-white matter junction in the cerebral hemispheres or in the cerebellum; none were found in the basal ganglia, thalamus, or brainstem. Mutation carriers with hypertension had more microbleeds in the cerebellum than those without hypertension, but there was no association between number of microbleeds and hypertension. Twenty-two (81%) mutation carriers had white matter hyperintensities, and 16 (62%) had intracranial hemorrhages. All hemorrhages were supratentorial and spared the thalamus and basal ganglia. The number of microbleeds correlated with increasing age, possibly reflecting disease progression.

Obici et al. (2005) reported an Italian family with autosomal dominant cerebral amyloid angiopathy. Clinically, the patients had multiple intracerebral hemorrhages, but only 1 affected family member had cognitive impairment. Neuropathologic analysis of 2 patients showed severe selective cerebroarterial amyloidosis in leptomeningeal and cortical vessel walls with secondary microvascular degeneration and 'vessel-within-vessel' changes. There were no parenchymal amyloid plaques or neurofibrillary tangles.

Bugiani et al. (2010) reported 4 unrelated Italian families with autosomal dominant hereditary cerebral hemorrhage with amyloidosis caused by a heterozygous mutation in the APP gene (E693K; 104760.0014). Affected individuals presented with recurrent headache and multiple hemorrhagic strokes between age 44 and 63, followed by epilepsy and cognitive decline in most of them. Several affected individuals became comatose or bedridden, and some died as a result of cerebral hemorrhage. Neuroimaging demonstrated small to large hematomas, subarachnoid bleeding, scars with hemosiderin deposits, multi-infarct encephalopathy, and leukoaraiosis. Multiple brain regions were involved, including both gray and white matter. Postmortem examination of 1 patient showed many small vessels with thickened and/or split walls due to a hyaline congophilic material that was immunoreactive for beta-amyloid-40. Most of the abnormal vessels were in the leptomeninges, in the cerebral and cerebellar cortex, and in the white matter close to the cortex. Beta-amyloid-40 was also detectable in cortical capillaries, and beta-amyloid-42 was found in neuropil of the gray structures. Neurofibrillary tangles and neuritic plaques were not present. The progression of the clinical phenotype correlated with that pathologic findings.

Pathogenesis

Torack (1975) reported the postmortem studies of 3 patients with congophilic angiopathy who had a surgical procedure and died from a subsequent massive hemorrhagic episode. Two of the patients had clinical evidence of a dementing syndrome. Ultrastructural studies confirmed the amyloid nature of the congophilic material in the 2 biopsied cases. The deposition of amyloid in these cases was believed to be a primary event and related to a generalized systemic disorder.

Cerebral amyloid angiopathy is frequently found in demented and nondemented elderly persons. Natte et al. (2001) investigated the relationship between the amount of cerebral amyloid angiopathy and the presence of dementia in 19 patients with hereditary cerebral hemorrhage with amyloidosis of the Dutch type. They found that the amount of cerebral amyloid angiopathy, as quantified by computerized morphometry, was strongly associated with the presence of dementia independent of neurofibrillary pathology, plaque density, or age. A semiquantitative score, based on the number of amyloid beta-laden severely stenotic vessels, completely separated demented from nondemented patients. The results suggested that extensive (more than 15 amyloid beta-laden severely stenotic vessels in 5 frontal cortical sections) CAA alone is sufficient to cause dementia in hereditary cerebral hemorrhage with amyloidosis of the Dutch type.

Revesz et al. (2003) reviewed the pathology and genetics of APP-related CAA and discussed the different neuropathologic consequences of different APP mutations. Those that result in increased beta-amyloid-40 tend to result in increased deposition of amyloid in the vessel walls, whereas those that result in increased beta-amyloid-42 tend to result in parenchymal deposition of amyloid and the formation of amyloid plaques. These latter changes are common in classic Alzheimer disease.

Attems et al. (2004) presented evidence that deposition of beta-amyloid in cerebral capillary amyloidosis is distinct from the deposition of beta-amyloid in cerebral arterial amyloidosis. Neuropathologic examination of 100 postmortem brains from elderly patients showed AD pathology in 8 (22.2%) of 36 nondemented individuals and in 42 (65.6%) of 64 individuals with a clinical diagnosis of dementia. The results indicated that CAA is characterized by beta-amyloid-40 and -42 deposits in leptomeningeal and cortical arterial vessels, with heavier deposition of beta-amyloid-40. Involvement of the capillaries was rare. In contrast, capillary CAA was characterized by globular deposition of beta-amyloid-42 in the glial limitans of cortical capillaries and in pericapillary compartments, often in conjunction with parenchymal beta-amyloid-42 deposits. Attems et al. (2004) postulated that the deposition in capillaries was a result of drainage of beta-amyloid-42 from senile and neuritic plaques, i.e., from perivascular drainage along basement membranes in the central nervous system.

In a study of 17 patients with cerebral amyloid angiopathy, 72 with AD, and 58 controls, Verbeek et al. (2009) found that patients with CAA had strongly decreased CSF levels of beta-amyloid-40 and beta-amyloid-42 compared to both AD patients and controls. CSF tau levels were significantly higher in the AD patients compared with the CAA group, although the concentrations among CAA patients were also higher than in controls. The combination of beta-amyloid-42 and total tau discriminated CAA from controls, with an area under the receiver operator curve (ROC) of 0.98. The data were consistent with CAA neuropathologic evidence indicating that beta-amyloid-40 and -42 are selectively trapped in the cerebral vasculature from interstitial fluid drainage pathways that usually transport these amyloid proteins to the CSF.

Mapping

By linkage analysis, Van Broeckhoven et al. (1990) determined that the APP gene on chromosome 21 was a candidate gene for the Dutch form of cerebroarterial amyloidosis.

Molecular Genetics

In 2 patients from presumably unrelated Dutch families with hereditary cerebral hemorrhage with amyloidosis, Levy et al. (1990) identified a glu693-to-gln mutation in the APP gene (E693Q; 104760.0001). The authors noted that amyloid precursor proteins in the Dutch and Icelandic forms of cerebroarterial amyloidosis are both protease inhibitors and both have been found to have a substitution in their genes that give rise to a substitution of glutamine (see 604312.0001).

Prelli et al. (1990) demonstrated that both the normal and the variant alleles were expressed in vascular amyloid in this disorder.

Graffagnino et al. (1994) failed to find the amyloid mutation in any of 48 consecutive patients with sporadic intracerebral hemorrhage admitted to Duke University Hospital. No pathologic examinations were made to determine if any of these patients had amyloid deposition.

In 4 affected members of an Italian family with cerebral amyloid angiopathy, Obici et al. (2005) identified a mutation in the APP gene (104760.0019).

In 2 brothers from an extensive Iowa kindred with progressive dementia and cerebroarterial amyloidosis, Grabowski et al. (2001) identified a heterozygous mutation in the APP gene (N694D; 104760.0016). Greenberg et al. (2003) identified the N694D mutation in 2 affected members of the Spanish family reported by Iglesias et al. (2000).

Rovelet-Lecrux et al. (2006) identified duplication of the APP gene (104760.0020) in affected members of a family with early-onset Alzheimer disease and prominent cerebral amyloid angiopathy.

Modifier Genes

The majority of beta-amyloid CAA is sporadic, affecting elderly individuals who may or may not have accompanying Alzheimer disease pathology. The incidence of both diseases steadily increases with age, with the incidence of CAA reaching 50% in those older than 70 years (Revesz et al., 2009).

Greenberg et al. (1995) found that the presence of apolipoprotein E4 (107741) significantly increased the odds ratio for moderate or severe cerebral amyloid angiopathy, even after controlling for the presence of Alzheimer disease. Yamada et al. (1996) reported a lack of association between the E4 allele and cerebral amyloid angiopathy in elderly Japanese patients.

Nicoll et al. (1996, 1997) did not find an association between the E4 allele and CAA-related hemorrhage. However, they did find a high frequency of the E2 allele in patients with CAA-related hemorrhage, regardless of the presence of AD. The authors suggested that patients with the E2 allele may be protected from parenchymal AD but may be susceptible to the rupture of amyloid-laden vessels.

In a postmortem study, Greenberg et al. (1998) found an association between apolipoprotein E2 and vasculopathy in cerebral amyloid angiopathy. Of 75 brains with complete amyloid replacement of vessel walls, only 23 had accompanying signs of hemorrhage in cracks of the vessel wall. The frequency of apolipoprotein E2 was significantly higher in the group with vasculopathy. The authors suggested that apolipoprotein E2 and E4 might promote hemorrhage through separate mechanisms: E4 by enhancing amyloid deposition and E2 by promoting rupture.

Hemorrhages related to amyloid angiopathy generally occur in the cortical and cortico-subcortical (lobar) brain regions where vascular amyloid deposits are most frequent, and occur less commonly in the cerebellum. Most patients recover from an initial lobar hemorrhage. Recurrent lobar hemorrhages are relatively common, however, and may cause greater morbidity and mortality than first hemorrhages. O'Donnell et al. (2000) identified a specific apolipoprotein E genotype as a risk factor for early recurrence: carriers of the E2 (107741.0001) or E4 (107741.0016) allele had an increased risk for early recurrence compared to individuals with the E3/E3 (107741.0015) genotype.

Animal Model

Herzig et al. (2004) found that transgenic mice expressing the human E693Q APP mutation developed extensive cerebral amyloid angiopathy in the leptomeningeal and cortical vessels, cerebral hemorrhages, and neuroinflammation with astrogliosis similar to that found in HCHWAD. Human APP mRNA was detected in neurons and neuronal processes, but not in vessel walls. There was smooth muscle degeneration and irregular thickening of the basement membrane in some vessels, whereas the endothelial cell layer appeared to be intact, and there was no parenchymal amyloid deposition. The ratio of amyloid-beta(40) to amyloid-beta(42) was approximately 4-fold higher than in wildtype mice or human Alzheimer disease.

Herzig et al. (2006) extended their earlier studies by developing several murine models of APP-related CAA and APP-related parenchymal amyloid deposition. The findings indicated that APP-related CAA is sufficient to induce cerebral hemorrhage and neuroinflammation; the origin of vascular amyloid is mainly neuronal; APP-related CAA results largely from impaired clearance; a high ratio of beta-40 to beta-42 favors vascular over parenchymal amyloidosis; and genetic factors such as ApoE (107741) can modulate the occurrence of hemorrhages.