Cerebral amyloid angiopathy

INTRODUCTION — Cerebral amyloid angiopathy (CAA) is characterized by amyloid beta-peptide deposits within small- to medium-sized blood vessels of the brain and leptomeninges. CAA is an important cause of lobar intracerebral hemorrhage in older adults [1,2]. In addition to intracerebral hemorrhage, CAA may present with transient neurological symptoms, an inflammatory leukoencephalopathy, as a contributor to cognitive impairment, or with incidental microbleeds or hemosiderosis on magnetic resonance imaging.

The clinical features, diagnosis, and management of CAA is discussed here. Alzheimer disease, which is also characterized by abnormal amyloid beta-peptide deposits in the brain, is discussed separately. (See "Epidemiology, pathology, and pathogenesis of Alzheimer disease" and "Clinical features and diagnosis of Alzheimer disease".)

Other causes of intracerebral hemorrhage are discussed in detail separately. (See "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis" and "Spontaneous intracerebral hemorrhage: Acute treatment and prognosis" and "Superficial siderosis".)

EPIDEMIOLOGY — The incidence of CAA is strongly age dependent. By autopsy, CAA can be identified by the replacement of at least some cerebral blood vessel walls with amyloid beta-peptide. In one series of 784 autopsy cases, the prevalence of CAA ranged from 2.3 percent for patients between the ages of 65 and 74, 8.0 percent for those age 75 to 84, and 12.1 percent in patients over the age of 85 [3]. In an autopsy study of 1079 patients whose mean age at death was 89.7 years, the prevalence of moderate-to-severe CAA was 36 percent [4]. CAA-related symptoms are uncommon at ages younger than 60 to 65 and are rare in individuals less than 60 years [5,6].

The prevalence of CAA among older patients with dementia is higher than those without dementia. In a systematic review of population-based studies, nearly 60 percent of patients with dementia showed CAA pathology compared with less than 40 percent among those without dementia [7]. Among patients with Alzheimer disease, more than 80 percent had pathologic evidence of CAA [8].

PATHOPHYSIOLOGY

Pathogenesis — The pathology of CAA involves deposition of amyloid beta-peptide within the cerebral vasculature. Vascular amyloid deposits in CAA are biochemically similar to the material comprising senile plaques in Alzheimer disease [1]. The primary constituent of each is amyloid beta-peptide, a 39 to 43 amino acid fragment of the amyloid precursor protein (APP).

●Development of CAA – The factors other than aging that initiate and promote amyloid beta-peptide deposition leading to CAA are not well understood. Genetic factors can cause CAA in an autosomal dominant manner and can increase the risk for sporadic CAA (see 'Genetic susceptibilities' below). Rarely, deposition of amyloid beta-peptide may occur, and clinical manifestations of CAA may develop decades after a neurosurgical procedure and/or exposure to cadaveric central nervous system tissue [9-11].

●Development of hemorrhage – Vascular rupture and bleeding in CAA appear to be a multistep process involving the deposition of amyloid beta-peptide in the vascular wall and subsequent vascular changes such as concentric splitting of the vascular wall. The relationship between CAA and hypertension is debated. Although many patients with CAA-related hemorrhage are normotensive [12-14], elevated blood pressure nonetheless appears to contribute to the risk of hemorrhage recurrence [15].

Despite shared pathologic features, the pathophysiology of CAA and Alzheimer disease appear distinct [16]. There is no clinical overlap between CAA and the non-central nervous system systemic amyloidoses, such as primary (amyloid AL) and secondary (amyloid AA) amyloidosis.

Genetic susceptibilities — CAA is sporadic in most patients who lack an identified genetic cause. A small minority of patients have a genetic susceptibility to CAA; such patients have a more severe course and an earlier age of onset [17,18].

●Amyloid precursor protein variant – Variant forms of the gene that encodes the APP are responsible for some cases of early-onset CAA that are inherited in an autosomal dominant pattern. While most of these variants are also associated with at least some of the neuropathologic features of Alzheimer disease, at least two APP variants (Glu693Gln and Leu705Val) have been reported to cause autosomal-dominant CAA with minimal parenchymal amyloid plaques or neurofibrillary tangles [19-21]. The Dutch-type Glu693Gln APP pathologic variant is associated with cerebral amyloid deposition with a more aggressive course than patients with sporadic CAA [18]. The causative amino acid substitutions in these hereditary forms of CAA may increase the toxic effects of amyloid beta-peptide on the vessel wall [22,23], decrease the peptide's susceptibility to proteolysis [24], or impair its clearance from the central nervous system [25].

●Apolipoprotein E – Patients carrying the apolipoprotein E (APOE) epsilon 2 (e2) or epsilon 4 (e4) alleles appear to be at greater risk for CAA-related hemorrhage than those with only the common APOE epsilon 3 (e3) allele [26-29]. One systematic review found that there is good evidence for a dose-dependent association between APOE e4 and sporadic CAA [30]. Among patients with severe head injury, for example, the APOE e4 allele is associated with an increased risk of deposition of amyloid beta-peptide [31]. This allele also promotes deposition of amyloid beta-peptide in Alzheimer disease [32].

APOE e2 or e4 alleles are present in approximately two-thirds of patients with CAA compared with only about one-quarter of older adult controls without evidence of CAA. These alleles are associated with an increased likelihood of having a CAA-related hemorrhage [17,26], earlier age of disease onset (mean age of first hemorrhage 75 versus 82 years in patients who are not carriers of APOE e2 or e4), and a greater risk of hemorrhage recurrence (two-year cumulative recurrence rate of 28 percent in carriers of e2 or e4 versus 10 percent for the APOE e3/e3 genotype) [33].

The APOE e2 and e4 alleles act via separate mechanisms. APOE e4 increases amyloid beta-peptide deposition [32]. APOE e2 causes amyloid-laden vessels to undergo changes such as concentric wall splitting and necrosis that predispose to rupture [26,34,35]. Patients with CAA who have both APOE e2 and e4 alleles appear to have a particularly early onset of disease and a high risk of early recurrence [26,33]. Carriers of the APOE e2 allele also have larger intracerebral hemorrhage (ICH) volumes, increased mortality, and worse functional outcomes compared with noncarriers, while these associations are not seen for carriers of the APOE e4 allele [36].

Data from a population-based, case-control study suggest that the risk of lobar ICH associated with APOE alleles may be modified by variation elsewhere in the APOE gene [37]. Notably, overall APOE haplotype (combination of multiple alleles on one chromosome) was also independently associated with lobar ICH, suggesting the presence of regulatory variants that could influence the effect of APOE e2 or e4 on the risk of lobar ICH [38].

●CR1 gene variant – A case-control genetic association study that included a prospective followup of 178 ICH survivors found that a variant within the CR1 gene (rs6656401) influences the risk and recurrence of CAA-related ICH [39]. In the same report, this genetic variant was also associated with the severity of vascular amyloid deposition on pathologic examination of 544 autopsy studies from two community-based clinical-pathological studies of aging.

ACUTE INTRACEREBRAL HEMORRHAGE — − The most common clinical manifestation of CAA is acute lobar intracerebral hemorrhage (ICH) (image 1) [40]. The term "lobar" refers to location in the cortex and subcortical white matter of a hemispheric lobe of the brain; this contrasts with the deep locations, such as putamen, thalamus, and pons, which are characteristic of hypertensive hemorrhage. The lobar location of the hemorrhages reflects the underlying distribution of the vascular amyloid deposits, which favor cortical vessels and largely spare white matter, deep gray matter, and the brainstem. Involvement of the blood vessels in the cerebellum and leptomeninges can also give rise to the less common clinical presentations of cerebellar or subarachnoid/subdural hemorrhage.

Clinical features — The clinical presentation of CAA-related hemorrhage varies with the lesion size and brain region impacted. Lobar hemorrhages can cause hemiparesis from involvement of pyramidal motor neurons and tracts. A large lobar hemorrhage may cause depressed consciousness from direct involvement or secondary mass effect on reticular activating system networks. In comparison, smaller lobar or cerebellar hemorrhages may cause more limited focal deficits related to the underlying brain structure impacted. The clinical presentation of intracerebral hemorrhage is described in more detail separately. (See "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis", section on 'Clinical presentation'.)

In rare cases, small cortical hemorrhages may irritate meningeal nociceptors to cause isolated headache. They may also infrequently be asymptomatic, found on imaging pursued for other indications [41,42]. (See 'Microbleeds' below.)

Imaging features — Imaging findings in acute hemorrhage can vary in size and location, but specific imaging patterns and associated findings are common in CAA-related hemorrhage.

●Lobar regions – CAA-related lobar hemorrhages most commonly arise in posterior lobar brain regions. Analysis of the spatial distribution of 321 intracerebral hemorrhages in 59 patients with CAA revealed that hemorrhages were significantly more likely to occur in the temporal and occipital than the frontal and parietal lobes (ratio of actual to expected hemorrhages 1.37, 1.43, 0.58, and 1.0, respectively) [43]. The explanation for the posterior brain clustering of CAA hemorrhages is undetermined, but it may be related to as-yet unknown characteristics of posterior circulation vessels that influence amyloid beta-peptide elimination or to increased vulnerability of these brain regions to minor trauma [43,44]. CAA-related lobar hemorrhages often extend beyond the brain tissue into the subarachnoid and subdural spaces and, less frequently, may rupture into ventricles [45,46]. Extension of the hemorrhage into the subarachnoid space and the presence of elongated "finger-like" projections appear to be characteristic features of CAA-related lobar hemorrhages that may assist in diagnosis [47].

●Cerebellum – The cerebellum may contain variable amounts of vascular amyloid in individuals with CAA. The cerebellum is also a site of CAA-related hemorrhage, with predilection for the cerebellar cortex and vermis rather than the cerebellar nuclei and deep white matter [48-50].

●Convexity subarachnoid hemorrhage – Subarachnoid hemorrhage can also occur within the convexity of the cerebral hemispheres (cSAH) in patients with CAA due to amyloid deposition at the cortical surface. cSAH may occur along with acute ICH, located either adjacent to or remote from the ICH. In addition, patients may present with isolated cSAH and seizures or other focal symptoms due to dysfunction of underlying cortex. (See 'Transient focal neurologic episodes' below.)

●Associated chronic hemorrhagic findings – Gradient-echo or susceptibility-weighted sequence brain magnetic resonance imaging (MRI) obtained in patients with an acute lobar hemorrhage may also show chronic cerebral microbleeds (CMBs) and/or cortical superficial siderosis (cSS). CMBs are typically asymptomatic lesions and are found in the juxtacortical and cortical lobar regions with a predilection for temporal and occipital lobes. cSS represents cSAH in the chronic phase. The presence of cSS likely reflects severe CAA in the leptomeningeal vessels [51]. (See 'Incidental chronic imaging features' below.)

Differential diagnosis — Differentiation of nontraumatic lobar ICH related to CAA from other causes depends upon the clinical and radiographic appearance. Other causes include:

●Lobar extension of a hypertensive hemorrhage (see "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis")

●Hemorrhagic transformation of an ischemic stroke (see "Neuroimaging of acute stroke")

●Hemorrhagic venous infarction from cerebral venous thrombosis (see "Cerebral venous thrombosis: Etiology, clinical features, and diagnosis")

●Hemorrhage of arteriovenous malformation (AVM) (see "Vascular malformations of the central nervous system")

●Hemorrhagic tumor (see "Overview of the clinical features and diagnosis of brain tumors in adults")

Clinical features that favor diagnoses other than CAA include younger age (many hemorrhages attributed to AVM occur before age 35 to 40 [52]), prodromal symptoms (progressive headache may suggest cerebral venous thrombosis), or patient-level risk factors (eg, active metastatic cancer may suggest hemorrhagic tumor, high thromboembolic risk in a patient with atrial fibrillation not treated with anticoagulation may suggest hemorrhagic transformation of ischemic infarct).

Imaging features may also help identify diagnoses other than CAA in patients with lobar hemorrhage (table 1). Brain MRI can help identify evidence of acute ischemia or enhancement associated with tumors. Corresponding computed tomographic (CT) angiography or MR angiographic studies can help identify the presence of an associated arterial or venous occlusion or AVM. This differential diagnosis and the evaluation to determine etiology is discussed separately. (See "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis", section on 'Subsequent imaging'.)

Diagnostic approach — The presence of CAA should be suspected clinically in patients age 50 or older with or without a clinical manifestation of CAA who have acute or chronic hemorrhagic findings or characteristic white matter features on brain MRI in the absence of an obvious alternative cause [53].

Definite CAA is only diagnosed postmortem. A full pathologic examination of the brain showing amyloid deposition with vasculopathy, evidence of brain hemorrhage, and absence of other diagnostic lesions confirms CAA [54]. During life, the diagnosis of probable CAA can be made with clinical evaluation and MRI of the brain. Examination of a sample of brain tissue obtained by brain biopsy further supports the diagnosis but is infrequently performed.

●For all patients, we use hemorrhagic imaging features on T2*-weighted MRI sequences and additional white matter markers to support the diagnosis of probable CAA. (See 'Imaging-based diagnosis' below.)

●For select patients in whom imaging features are equivocal or the cause of a clinical manifestation is uncertain, we use additional adjunctive testing. (See 'Adjunctive diagnostic testing for some patients' below.)

Imaging-based diagnosis — All patients with suspected CAA should undergo brain MRI, including T2*-susceptibility-weighted or gradient-echo sequences, which accentuate the signal dropout caused by iron-containing deposits left by old hemorrhages (image 2) [55]. Chronic lobar hemorrhages, cSAH, cSS, and CMBs, appear dark on such sequences. Inclusion of white matter features, including enlarged perivascular spaces and multispot pattern subcortical hyperintensities on T2-weighted sequences in the MRI criteria for probable CAA, increased the diagnostic sensitivity in version 2.0 of the Boston criteria [56]. (See 'The Boston criteria for CAA' below.)

To support the diagnosis of probable CAA with MRI, we look for the presence of either of the following findings not attributable to another cause [56] (see 'Differential diagnosis' above):

●Two or more hemorrhagic lesions (ICH, CMB, cSAH, or cSS foci in any combination) in the lobar brain regions, entirely sparing regions typical of hypertensive hemorrhage (basal ganglia, thalamus, or pons) (image 3 and image 4 and image 5 and image 6) [53,57,58]

●One lobar hemorrhagic lesion and one white matter lesion, defined as either severe (ie, >20 per hemisphere) dilated perivascular spaces in the centrum semiovale or multiple (ie, >10) ovoid-shaped white matter hyperintensities on T2-weighted imaging in bilateral subcortical regions

A single hemorrhagic lesion (eg, lobar hemorrhage, cSAH, cSS focus, or lobar CMB) or isolated white matter features create less diagnostic certainty but can be suggestive of the diagnosis (classified as "possible CAA").

The Boston criteria for CAA — The Boston criteria provide a framework for various levels of diagnostic certainty in patients with intracerebral hemorrhage with or without pathologic tissue analysis (table 2).

Initial versions of the criteria included clinical, radiologic, and pathologic data to support the diagnosis of CAA [54,57]. The Boston criteria version 2.0 for sporadic CAA update the modified Boston criteria to include expanded imaging findings (hemorrhagic and white matter features) with validation data across multiple time epochs, medical centers, and both hemorrhagic and nonhemorrhagic clinical presentations (table 2). (See 'Transient focal neurologic episodes' below and 'Cerebral amyloid angiopathy-related inflammation' below and 'Cognitive impairment' below and 'Incidental chronic imaging features' below.)

In a retrospective, multicenter review of 341 patients who presented with clinical features consistent with CAA and had both MRI and brain tissue available for analysis, the inclusion of additional imaging findings to the Boston criteria version 2.0 produced a higher diagnostic accuracy for probable CAA than the previous modified Boston criteria (84.8 versus 79.8 percent) [56]. The specificity was 95 percent for both sets of criteria.

Adjunctive diagnostic testing for some patients — While not commonly performed for evaluation and diagnosis of most patients with typical symptoms and imaging findings, additional testing to support the diagnosis of probable CAA may be performed in circumstances when the above criteria are not met.

●Follow up brain MRI – Follow-up imaging, typically three to six months after the acute event, may show resolution of the acute hemorrhage and help exclude underlying alternative causes. Interval development of subclinical neuroimaging findings, namely strictly lobar CMB or cSS, may provide further support for the diagnosis of CAA. (See 'Microbleeds' below and 'Cortical superficial siderosis' below.)

●Brain biopsy – Brain biopsies are done rarely for the diagnosis of CAA. However, brain tissue may be obtained during surgical evacuation of select acute lobar hemorrhages. (See "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis", section on 'Subsequent imaging'.)

Evacuated hematoma specimens and accompanying leptomeningeal or parenchymal tissue from older adult patients should routinely be examined with Congo red stain or beta-amyloid immunostain for CAA. Based upon data from a postmortem model, nearly all blocks of tissue from brains with CAA-related hemorrhage demonstrate some degree of CAA [3], often with evidence of advanced disease such as complete amyloid replacement of the smooth muscle layer or the appearance of vessel breakdown [59-61]. Signs of advanced disease are rare in single-tissue specimens from asymptomatic older adult brains; thus, their presence points toward a degree of CAA severe enough to cause hemorrhage.

●Cerebrospinal fluid analysis – Up to a 50 percent reduction in levels of cerebrospinal fluid (CSF) amyloid beta 42 and beta 40 protein have been found in patients with CAA [62]. In combination with the finding of mildly increased total tau levels, CSF analysis distinguished patients with CAA from normal controls with a high degree of accuracy. Similar results were obtained in an independent study [63]. We obtain CSF when the imaging diagnosis is uncertain and when confirming the diagnosis may affect clinical decisions, such as whether to give or withhold antithrombotic treatment.

●Positron emission tomography – Positron emission tomography (PET) using 11C-Pittsburgh compound B (PIB), a ligand that binds to beta-amyloid, demonstrates increased uptake in patients with CAA-related hemorrhage compared with normal controls [64,65]. Another PET study of the amyloid ligand florbetapir showed elevated retention in patients with CAA but not hypertension-related intracerebral hemorrhages [66]. Compared with patients with Alzheimer disease, median binding of PIB is lower in CAA and may differ in its distribution. Current and future hemorrhagic lesions in patients with CAA appear to occur preferentially in local regions of concentrated amyloid detected by PIB [67,68]. PET imaging is typically used in research settings.

●Genetic testing – There is no clearly defined clinical role for genetic testing in sporadic CAA. In particular, apolipoprotein E (APOE) genotype is neither sensitive nor specific for the diagnosis of CAA, as the epsilon 2 (e2) and epsilon 4 (e4) alleles are present in only a subset of patients [17,26].

Acute management — Acute CAA-related hemorrhage is treated like other acute nontraumatic intracerebral hemorrhages. Of note, surgical biopsy or hematoma resection appears to carry little or no additional risk in CAA compared with other types of ICH and can be performed when indicated [69]. (See "Spontaneous intracerebral hemorrhage: Pathogenesis, clinical features, and diagnosis" and "Evaluation and management of elevated intracranial pressure in adults".)

Prevention of recurrent hemorrhage — Survivors of lobar hemorrhage and patients with other clinical manifestations of CAA are at risk for future hemorrhagic complications (see 'Prognosis' below). This risk should be factored into decision-making when assessing the risks and benefits of other medications for patients with CAA.

Avoiding anticoagulants and antiplatelet agents — Because of the risk of spontaneous lobar hemorrhage in patients with CAA, we weigh the risks and benefits of using of anticoagulant and antiplatelet agents at an individual-level, in agreement with guidelines from the American Heart Association [70].

The major determinants of ICH risk in CAA patients are:

●History of prior ICH

●Presence of particular CAA-associated imaging features such as disseminated cSS (see 'Incidental chronic imaging features' below)

●Class of agent used (highest ICH risk with warfarin, then the direct oral anticoagulants [DOACs; also called non-vitamin K antagonist oral anticoagulants, or NOACs], then antiplatelet agents, and lowest with no agent)

●Duration of treatment

Warfarin in particular increases both the frequency (approximately 7- to 10-fold) and severity (approximately 60 percent mortality) of cerebral hemorrhage [71-73].

The benefit of anticoagulants or antiplatelets is determined by the strength of the clinical indication and the availability of alternative treatments. In select patients with compelling indications for anticoagulation due to high risk of thromboembolism and absence of alternatives, anticoagulation may be used after discussion with the patient regarding risks and benefits. We reserve anticoagulation in CAA patients for those at high-risk for thromboembolic complications related to specific indications. These may include:

●Cancer-related thrombophilia with high risk of or prior venous thromboembolism (see "Risk and prevention of venous thromboembolism in adults with cancer")

●Hypercoagulable (acquired and inherited) conditions (see "Evaluating adult patients with established venous thromboembolism for acquired and inherited risk factors")

●Mechanical prosthetic heart valve replacement (see "Antithrombotic therapy for mechanical heart valves")

●Other temporary high-risk indications for anticoagulation (see "Atrial fibrillation: Left atrial appendage occlusion", section on 'Postprocedure management' and "Overview of the treatment of proximal and distal lower extremity deep vein thrombosis (DVT)" and "Prevention of venous thromboembolism in adults undergoing hip fracture repair or hip or knee replacement")

Because ICH risk is a function of antithrombotic treatment duration as well as type, we recommend in these situations the shortest and least intense treatment course that is still compatible with effective thrombosis prevention.

In other patients, we pursue alternatives to anticoagulation, based on indication.

●Atherosclerotic disease – Aspirin also increases the risk of hemorrhage but to a lesser extent than anticoagulants. We reserve aspirin for selected patients with CAA who have clear indications for antiplatelet therapy. In one prospective cohort of patients with primary lobar ICH, aspirin was associated with an increased risk of ICH recurrence (hazard ratio 3.95, 95% CI 1.6-8.3) when controlling for other hemorrhage risk factors [74].

●Atrial fibrillation – The management of atrial fibrillation in patients with CAA-related ICH is uncertain. Left atrial appendage closure is a reasonable treatment option for individuals with CAA who are at high risk for atrial fibrillation-related cardioembolic stroke [75]. (See "Atrial fibrillation: Left atrial appendage occlusion" and "Atrial fibrillation in adults: Selection of candidates for anticoagulation".)

Of note, some retrospective studies have reported good outcomes for patients with atrial fibrillation who were restarted on anticoagulation with warfarin after recovery from an anticoagulant-related intracerebral hemorrhage [76]. Conclusions from these retrospective analyses are limited by the likelihood of indication bias regarding which patients were or were not re-anticoagulated, but they offer a rationale for randomized trials to address this question [77].

The DOACs appear at least as effective as warfarin for prevention of ischemic strokes in patients with atrial fibrillation and confer lower risks for ICH. While they have not been studied for this indication, some experts use these agents (dabigatran, apixaban, edoxaban, rivaroxaban) for patients with atrial fibrillation and CAA who are at high risk for both ischemic and hemorrhagic stroke.

Other potential alternatives to anticoagulation in high-risk patients with atrial fibrillation are discussed separately.

Some nonsteroidal anti-inflammatory medications have weak antithrombotic properties. We prefer the nonacetylated salicylates (eg, magnesium salicylate) drugs over other nonsteroidal anti-inflammatory drugs as they do not appear to affect platelet function.

Thrombolytic therapy — We do not routinely offer thrombolytic therapy for indications such as acute ischemic stroke, myocardial infarction, or pulmonary embolism in patients with a history of CAA-related ICH. Intravenous thrombolysis for ischemic stroke is contraindicated in patients with a history of ICH [78]. Endovascular mechanical thrombectomy may be an option for patients with a history of lobar ICH. Because patients with CAA are at risk of hemorrhagic complications from thrombolytic therapy, the risks and benefits of acute therapies should be discussed with patients or their proxies whenever possible.

A large trial evaluating the use of tissue-type plasminogen activator for acute myocardial infarction identified severe CAA at postmortem examination in two of five patients with intracerebral hemorrhagic complications [79]. Additionally, in one analysis of acute ischemic stroke, patients reported an association between number of cortical microbleeds and an elevated risk of hemorrhagic complications after systemic thrombolysis [80]. However, the specific treatment benefits of thrombolytic therapy in such patients were not assessed.

Blood pressure control — Although the vascular pathology in CAA does not appear to be driven by hypertension, control of blood pressure within normal limits is nonetheless advisable. We use intensive blood pressure goals, as tolerated (table 3). (See "Antihypertensive therapy for secondary stroke prevention", section on 'Patients with intracerebral hemorrhage'.)

Support for lowering of blood pressure in patients diagnosed with CAA came from a secondary analysis of data from the Perindopril Protection Against Recurrent Stroke Study (PROGRESS) trial [14]. At a median follow up of 3.9 years, those assigned to active treatment (perindopril plus indapamide) had a 77 percent reduction (3 versus 13 events) of probable CAA-related ICH. These results are consistent with observational data showing reduced risk of ICH recurrence among ICH survivors with lower ambulatory blood pressures [15].

Managing statin use — We do not withhold statin agents for most CAA patients when otherwise indicated. While a number of studies have found an inverse relationship between total and low-density lipoprotein cholesterol and the risk of ICH [81,82], treatment with statins does not appear to increase the risk of primary ICH or to negatively impact prognosis according to a number of studies and meta-analyses [83-86]. (See "Spontaneous intracerebral hemorrhage: Secondary prevention and long-term prognosis", section on 'Management of statins'.)

Prognosis — Lobar hemorrhage in general is associated with features that are both favorable (their superficial location and tendency to spare the ventricles) and unfavorable to outcome (older age and somewhat larger hematoma size) [87-89].

●Mortality – Overall mortality in acute lobar hemorrhage due to CAA is in the range of 10 to 30 percent [88,89], with the best prognosis for patients with smaller hematomas (<50 mL) and higher level of consciousness on admission (Glasgow coma scale ≥8).

●Recurrence – CAA carries a substantially higher risk of hemorrhage recurrence than for hypertensive hemorrhage. In two case series of ICH survivors, recurrence rates were 21 percent at 2 years and 24 percent at a median 2.6 years, respectively [33,90]. Another report of 104 survivors of primary lobar ICH found that patients with history of ICH prior to the index event were at approximately six times the risk as those without a previous history. The risk for recurrent ICH was even higher in those who had more than one prior ICH [74].

Recurrent hemorrhages have a tendency to arise in areas of prior hemorrhage [43]. Areas of high amyloid deposition on amyloid imaging appear to predict sites for future hemorrhage [67].

Recurrent ICH is more likely with increasing number of microbleeds, disseminated siderosis, and with posterior white matter hypodensity on CT scan (see 'Incidental chronic imaging features' below). In a study of 94 consecutive survivors of primary lobar hemorrhage, the three-year cumulative risk of recurrent hemorrhage for patients with one, two, three to five, and six or greater CMBs on the baseline gradient-echo MRI was 14, 17, 38, and 51 percent, respectively (hazard ratio 1.7, 95% CI 1.2-2.4 for each increase in category) [91]. In a study of 118 patients diagnosed with CAA (104 with a symptomatic lobar ICH), the cumulative risk of new ICH at four years was 74 percent for the 27 individuals with disseminated siderosis (defined as involving more than three cortical sulci) versus only 25 percent for the 77 without siderosis at baseline [92]. Similarly, the presence of cSAH in patients with CAA is associated with an increased risk of recurrent ICH [93].

Carriers of the e2 or e4 APOE alleles are also at increased risk compared with the more common APOE e3/e3 genotype [33]. (See 'Genetic susceptibilities' above.)

●Incident dementia – The association of subsequent dementia in patients with lobar ICH may reflect an elevated rate of ICH recurrence and/or the relationship of CAA with cognitive impairment. In a study of 255 patients with lobar and nonlobar ICH followed for five years, patients with lobar ICH had higher rates of dementia (36 versus 21 percent) and higher rates of disability (60 versus 31 percent) compared with those with nonlobar ICH [94]. This finding is consistent with a prior study of 218 ICH survivors with one-year incidence of new-onset dementia of 23.4 percent for lobar ICH versus 9.2 percent for nonlobar ICH [95]. In another cohort of 97 patients with lobar ICH, the rate of dementia after median 2.5-year follow-up was 26 percent [96]. (See 'Cognitive impairment' below.)

INCIDENTAL CHRONIC IMAGING FEATURES — Patients with CAA are frequently asymptomatic unless they develop a specific clinical episode. Evidence suggestive of CAA may be identified in asymptomatic patients who undergo brain magnetic resonance imaging (MRI) for unrelated symptoms (eg, chronic headaches).

Microbleeds — Chronic evidence of tiny asymptomatic bleeding can be detected on brain MRI and may reflect CAA. These microbleeds (or microhemorrhages) appear as 2 to 10 mm focal areas of hemosiderin deposition on gradient echo or other T2*-weighted sequences (image 7). In population-based studies, cerebral microbleeds are detected in 5 to 23 percent of older individuals [97-100].

Microbleeds can arise from small-vessel disease (CAA or hypertensive vasculopathy) and may represent a milder and somewhat distinct manifestation from overt (macro-)hemorrhage. Analysis of CAA patients in a neuroimaging study who underwent autopsy showed those with a large number of microbleeds had thicker-walled vessels due to beta-amyloid deposition compared with patients with a lower proportion or absence of microbleeds [101]. Microbleeds are also more prevalent among those using antiplatelet agents than nonusers [102,103].

●Evaluation and differential diagnosis – Focal areas of hemosiderin deposition on MRI should be reviewed with some care since lesions and structures other than hemorrhage can produce signal dropout. These include mineralization (particularly of the basal ganglia), flow-void from a cortical vessel, or adjacent air in the nasal sinuses. When establishing the presence of at least two hemorrhages, lesions that do not clearly represent independent hemorrhagic foci (for example, small blood deposits close to a larger hematoma) should not be counted.

Cerebral microbleeds are not specific for CAA as they can be seen in multiple conditions. Such conditions may include:

•Hypertension [104,105]

•Cerebral cavernous malformations [106]

•Coagulopathy [107,108]

•Thrombocytopenia [109,110]

•Anticoagulant medications [111]

•Central nervous system vasculitis [112]

•Infective endocarditis [113]

•End-stage kidney failure [114]

While microbleeds in general are not specific to CAA, cortical microbleeds (CMBs; (ie, those restricted to the cerebral cortex or superficial cerebellar regions [cerebellar cortex and vermis]), suggest CAA [104,105]. In contrast, microbleeds that involve the basal ganglia, thalamus, or pons are believed to result from hypertensive microangiopathy. The association between lobar microbleeds and APOE e4 in the Rotterdam and other studies support the hypothesis that these more superficial lesions often originate from CAA [29,99,100,115,116].

●Management and prognosis – For patients with acute ischemic stroke and a known history of CAA with CMBs only (ie, without a history of lobar hemorrhage or cortical superficial siderosis), we do not withhold systemic intravenous thrombolytic therapy if they otherwise meet eligibility criteria (table 4), in agreement with guideline statements from the American Heart Association [78].

Over the long-term, the risk of incident hemorrhage in patients with multiple CMBs is nonetheless substantial [117]. Thus, other measures for primary prevention are similar as for patients with an ICH presentation. (See 'Prevention of recurrent hemorrhage' above.)

CMBs are associated with an elevated risk of death. One study found that an incidental finding of multiple CMBs in an older adult was associated with a sevenfold risk of stroke-related death compared to individuals without CMBs [118].

Cortical superficial siderosis — Features of chronic hemorrhage in the cortical sulci, typically referred to as cortical superficial siderosis (cSS), can represent a focus of remote bleeding related to CAA (image 6) and is the chronic form of acute convexity subarachnoid hemorrhage [54,119,120]. cSS is typical at the convexities of the cerebral hemispheres in patients with CAA but may also be found along the cerebellar folia [50]. cSS is frequently found in CAA patients with microbleeds and appears to signal higher risk for future ICH, particularly disseminated cSS. (See 'Microbleeds' above and 'Prognosis' above.)

cSS typically is an asymptomatic imaging finding but is also detected in many CAA patients who present with transient neurologic symptoms [120,121]. (See 'Transient focal neurologic episodes' below.)

●Evaluation and differential diagnosis – cSS is common in patients with CAA (40 to 60 percent) but is unusual in patients with ICH of other cause (0 to 4 percent) [54,121]. However, cSS and superficial siderosis involving the cerebellum or brainstem can also arise from several other unrelated causes. These are discussed separately. (See "Superficial siderosis", section on 'Etiology'.)

Patients with isolated hemorrhagic imaging findings such as a single CMB or equivocal focus of cSS may develop progressive subclinical findings to further support the diagnosis of CAA. As an example, in one cohort of 118 patients with probable CAA who underwent follow-up imaging at a mean interval of 2.2 years, progression of cSS was found in 28 percent and new CMBs were found in 18 percent [122].

●Management and prognosis – For patients with acute ischemic stroke and a known history of CAA with cSS, we do not routinely offer systemic thrombolytic therapy. cSS represents an imaging marker of prior ICH, a contraindication for intravenous thrombolysis (table 4). Endovascular mechanical thrombectomy may be an option for such patients when treatment benefits are felt to outweigh the hemorrhagic risks.

Superficial siderosis is associated with poor functional outcome in CAA [123]. Disseminated SS in CAA appears to be an independent predictor of recurrent ICH [92,124], and radiologic progression of SS on serial imaging has been associated with an increased risk of subsequent ICH [122].

Nonhemorrhagic imaging findings — Nonhemorrhagic findings on brain MRI may also be found in patients with CAA [56].

●Acute ischemic micro-infarcts – Asymptomatic punctate hyperintense lesions on diffusion-weighted image (DWI) sequencing are occasionally found in patients with CAA [125]. In one retrospective review of brain MRIs from 78 patients with probable CAA, subacute infarcts were identified on DWI in 15 percent [126].

●Brain atrophy – Cerebral atrophy of white matter is associated with CAA. In one study, white matter volume was lower in patients with CAA than age-matched patients with Alzheimer disease and healthy controls [127]. In CAA patients, atrophy was most pronounced in occipital regions and was more severe in those with higher cortical microbleeds.

●White matter hyperintensities – Chronic white matter hyperintensities on T2-weighted sequences of brain MRI frequently represent small-vessel disease related to multiple conditions. Imaging patterns suggestive of CAA include the "multispot" pattern of multiple (ie, >10), typically bilateral lesions with a round or ovoid appearance [128]. In more advanced cases, white matter hyperintensities may be confluent. Such prominent findings may be seen preferentially in the subcortical parietal and occipital lobes of patients with advanced CAA [129].

●Centrum semiovale perivascular spaces – Dilated perivascular spaces are identified typically on brain MRI as hyperintense lesions on T2-weighted sequences. They have been associated with vascular disease and aging, but the presence of multiple (>20) dilated perivascular spaces that predominate in the subcortical centrum semiovale regions has been associated with CAA with and without ICH [56,124,130-132].

TRANSIENT FOCAL NEUROLOGIC EPISODES — A less common clinical manifestation of CAA is transient focal neurologic episodes (TFNE) [133-135]. These have also been called "amyloid spells."

Pathogenesis and clinical features — Patients report recurrent, brief, and often stereotyped spells of weakness, numbness, paresthesias, or other cortical symptoms that can spread smoothly over contiguous body parts over several minutes. In one cohort, transient neurologic symptoms occurred in 14 percent of patients with CAA, and positive symptoms (positive visual aura, limb jerking) were as common as negative symptoms (vision loss, limb weakness, dysphasia) [136].

These episodes may reflect abnormal activity (ie, cortical spreading depression) of the surrounding cortex typically in response to the small hemorrhages [137]. In one study, CAA patients with cortical superficial siderosis (cSS) or convexity subarachnoid hemorrhage (cSAH) were more likely to have transient neurologic symptoms compared with CAA patients without these symptoms (50 versus 19 percent) [136].

Evaluation and diagnosis — The description of TFNEs is similar to other transient neurologic attacks such as transient ischemic attacks (TIAs), seizures, and migraine auras.

Clinical features more suggestive of TFNE over other diagnoses include the smooth spread of the symptoms over minutes and the stereotypic recurrence of symptoms over time. Symptoms of TFNE tend to localize to the site of a cSAH, prior lobar hemorrhage, or focus of cSS [138]. (See 'Incidental chronic imaging features' above.)

Additionally, diagnostic evaluation at the time of acute symptoms can support the diagnosis of TFNE and exclude alternative causes. This may include:

●Brain MRI with gradient echo or other T2*-weighted sequences to identify cSAH, cSS, or cortical microbleeds (CMBs) in the region of cortex corresponding to TFNE symptoms

●Vascular imaging to exclude hemodynamically significant stenosis in the relevant vascular supply

●Electroencephalogram showing no epileptiform activity or seizure in the brain region corresponding to TFNE symptoms

Additional testing may be performed to exclude TIA or ischemic stroke if indicated by clinical features or patient risk profile. (See "Differential diagnosis of transient ischemic attack and acute stroke" and "Initial evaluation and management of transient ischemic attack and minor ischemic stroke".)

Features that support an alternative diagnosis include:

●TIAs may present in a patient with vascular risk factors as sudden loss of focal neurologic function attributable to a specific cerebrovascular arterial territory.

●Seizures may occur in a patient with a history of epilepsy or seizures as focal or diffuse symptoms. They may be associated with temporary postictal weakness or other loss of function.

●Migraines typically occur in younger patients with a prior history of migraines. Neurologic symptoms typifying the aura may wax over minutes prior to the onset of a headache.

The differential diagnosis and evaluation of patients with transient neurologic episodes is discussed in greater detail separately. (See "Differential diagnosis of transient ischemic attack and acute stroke", section on 'Transient neurologic events' and "Differential diagnosis of transient ischemic attack and acute stroke", section on 'Distinguishing transient attacks'.)

Management — Avoiding misdiagnosis is critical since the administration of antithrombotics for TFNE symptoms instead presumed to be a TIA can increase the risk of hemorrhagic stroke from CAA [138]. In one series, transient neurologic symptoms in a patient with CAA appeared to predict a high early risk of symptomatic intracerebral hemorrhage, which occurred in 50 percent of patients over a median follow-up period of 14 months [136].

TFNE are typically brief and self-limited but may recur. Patients should be reassured that TFNE that recur typically resolve within weeks [137]. Medications used for seizures or migraine such as topiramate or levetiracetam have been used for those with recurrent or bothersome attacks [137,139].

Other general measures, applicable to patients with hemorrhagic features of CAA, are discussed separately. (See 'Prevention of recurrent hemorrhage' above.)

CEREBRAL AMYLOID ANGIOPATHY-RELATED INFLAMMATION — Cerebral amyloid angiopathy-related inflammation (CAA-ri) appears to represent a distinct manifestation of CAA characterized by an inflammatory response to amyloid deposition in the brain with subacute and often progressive neurologic symptoms [140,141].

CAA-ri may be a milder form of a condition called Abeta-related angiitis. CAA-ri is characterized by perivascular inflammation, while Abeta-related angiitis is a true vasculitis with inflammation throughout the vessel wall [142-144]. These two inflammatory syndromes share similar clinical presentations, imaging properties, and response to treatment, except that Abeta-related angiitis typically requires more aggressive immunosuppressive treatment. (See 'Management' below.)

Clinical and diagnostic features — The clinical presentation of a CAA-ri syndrome is that of acute or subacute cognitive decline rather than hemorrhage [145,146]. Patients or family members may report memory impairment, personality changes, confusion, or alterations in level of consciousness. Seizures, new and persistent headaches, and focal and progressive neurologic signs are common.

Patients with CAA-ri are younger than those with other manifestations of CAA. In one review of case reports of 72 patients with CAA-ri, the mean age of diagnosis was 63 years [146]. Cognitive symptoms were reported in 76 percent, focal neurologic symptoms in 46 percent, headaches in 41 percent, and seizures in 31 percent.

The differential diagnosis includes primary, viral, and autoimmune encephalitides as well as cerebral neoplasm and other causes of rapidly progressive dementia [147-149]. (See "Primary angiitis of the central nervous system in adults", section on 'Diagnostic approach'.)

Diagnostic studies in this setting typically include brain magnetic resonance imaging (MRI) with contrast, serologic, and cerebrospinal fluid (CSF) analysis. Brain biopsy may be performed both to make a positive diagnosis of CAA-related inflammation as well to exclude other conditions. Findings that support the diagnosis of CAA-ri include:

●Brain MRI typically shows a (potentially reversible) leukoencephalopathy consisting of patchy or confluent white matter hyperintensities on T2-weighted sequences, characteristically in subcortical white matter, often asymmetric; multiple microbleeds are seen on gradient echo sequences (image 8). Gadolinium enhancement may be found in approximately one-third of cases [150].

●Angiography or magnetic resonance angiography does not show evidence of large- or medium-vessel vasculitis.

●Neuropathology shows perivascular inflammation with multinucleated giant cells that is associated with amyloid-laden vessels.

●Erythrocyte sedimentation rate and C-reactive protein are normal [146].

●CSF analysis may be normal but often shows a pleocytosis and/or mildly elevated protein. Anti-amyloid autoantibodies have been detected in the CSF during the acute phase of inflammation and return to control levels during remission [151-153]. CSF assay for anti-amyloid antibodies is not yet commercially available but may eventually emerge as a diagnostic test. Clinical and radiographic improvement may occur with immunosuppressive treatment [140,141,146,154]. (See 'Management' below.)

The diagnosis of CAA-ri is made in symptomatic patients with diagnostic imaging evidence of inflammation and hemorrhagic features of CAA who have undergone evaluation to exclude other sources (table 5). A validation study of proposed criteria for the diagnosis of probable CAA-related inflammation showed high sensitivity (82 percent) and specificity (97 percent) [150].

Management — We typically treat patients with suspected CAA-related inflammation with a course of high-dose glucocorticoids followed by a gradual oral taper. Our treatment protocol is usually methylprednisolone 1000 mg per day for three to five days, followed by an oral glucocorticoid taper over approximately 6 to 12 weeks. A repeat brain MRI scan approximately four to six weeks after treatment onset use clinical symptoms and changes in the subcortical white matter hyperintensities as evidence for treatment response. Although glucocorticoid therapy is most often used in this setting [153,155-157], other immunosuppressive treatments, such as cyclophosphamide, methotrexate, and mycophenolate mofetil, have also been used with a favorable response in isolated cases [146,158,159].

While data are limited, available evidence from observational studies supports treating inflammatory forms of CAA with immunosuppressive therapy [141,155-160]. In one series of 48 patients with CAA-related inflammation, those who were treated with immunosuppression (mostly glucocorticoids) were more likely to improve clinically (94 versus 50 percent) and radiographically (86 versus 29 percent) [160]. Treated patients were also less likely to have recurrent symptoms during a median of 2.7 years of follow-up (26 versus 71 percent).

The response to therapy may be seen within the first few months and may be more durable when glucocorticoids are tapered slowly. In an observational study of 113 patients with CAA-related inflammation of whom 88 percent received immunosuppressive therapy, clinical recovery was reported by three months in 70 percent and resolution of inflammatory features on neuroimaging occurred in 45 percent [155]. By 12 months, the rates of clinical and radiologic recovery were 84 and 77 percent, respectively. Symptomatic recurrence was likelier in patients treated with high-dose pulse of intravenous glucocorticoids alone than those treated with pulse intravenous glucocorticoids followed by a gradual oral glucocorticoid taper (hazard ratio 4.68, 95% CI 1.57–13.93).

Additional measures to mitigate the associated risks of future bleeding associated with CAA should be based on the risks associated with hemorrhagic features. (See 'Prevention of recurrent hemorrhage' above.)

COGNITIVE IMPAIRMENT — Advanced CAA is associated with cognitive impairment. The majority of patients diagnosed with CAA appear to have cognitive impairment in at least one domain on neuropsychological testing [16]. An autopsy series found that moderate to severe CAA (present in one-third of the study population) was associated with faster rates of decline in global cognition, perceptual speed, episodic memory, and semantic memory, independently of age, sex, education, Alzheimer disease pathology, and other potential covariates [161].

Pathogenesis — The pathogenesis of cognitive impairment in CAA is well understood and may be multifactorial. Researchers have investigated a relationship between Alzheimer disease and CAA as both are associated with abnormal cerebral deposition of amyloid beta-peptide. Additionally, CAA has been linked to vascular dementia because of overlapping epidemiology and risk factors.

Relationship to Alzheimer disease — CAA is common in conjunction with Alzheimer disease, appearing in moderate to severe form in 30 of 117 (26 percent) Alzheimer disease brains in an autopsy series; CAA with hemorrhage occurred in six (5.1 percent) [5,162]. Another autopsy study found that patients with both CAA and Alzheimer disease had more severe cognitive impairment than patients with Alzheimer disease alone [163]. Similarly, an MRI study in Alzheimer disease patients found that the presence of multiple microbleeds was associated with worse cognitive performance [164]. However, only about 25 percent of CAA patients appear to have clinical histories of dementia prior to their first hemorrhage [87]. (See "Clinical features and diagnosis of Alzheimer disease".)

Relationship to vascular dementia — Cerebrovascular disease may contribute to cognitive impairment in patients with CAA. Studies in population- and hospital-based subjects have correlated the number and presence of microbleeds with cognitive impairment and dementia, raising the possibility that these lesions are contributors to neurologic dysfunction, as well as markers of small-vessel disease [165,166]. In one study of patients who underwent MRI because of transient ischemic attack or stroke, a finding of lobar microhemorrhage but not deep microhemorrhage was associated with cognitive impairment [167].

Additionally, clinically silent acute or subacute cerebral infarcts on diffusion-weighted imaging have been detected in 15 to 23 percent of patients with CAA [126,168], and cerebral microinfarcts on T1 and fluid-attenuated inversion recovery (FLAIR) imaging have been found in 35 to 39 percent [169,170]. These data are consistent with autopsy and imaging studies showing an association between CAA severity and volume of white matter hyperintensity and/or microinfarct burden [133,171-175].

Vascular dementia is discussed in more detail separately. (See "Etiology, clinical manifestations, and diagnosis of vascular dementia".)

Evaluation and management — The evaluation and management of patients with cognitive impairment in the setting of CAA does not differ from other settings. Supportive care is the mainstay for patients with cognitive impairment related to CAA, like that of patients with other forms of cognitive impairment and dementia. This is discussed in greater detail separately. (See "Management of the patient with dementia".)

Other general measures, applicable to all patients with CAA, are discussed separately. (See 'Prevention of recurrent hemorrhage' above.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Stroke in adults".)

SUMMARY AND RECOMMENDATIONS

●Definition – Cerebral amyloid angiopathy (CAA) is characterized by amyloid beta-peptide deposits within small- to medium-sized blood vessels of the brain and leptomeninges. CAA is an important cause of lobar intracerebral hemorrhage in older adults. (See 'Introduction' above.)

●Diagnosis – The diagnosis of probable CAA can be made with clinical evaluation and magnetic resonance imaging (MRI) of the brain (table 2). Follow-up MRI examinations may be helpful if the initial study is equivocal. (See 'Diagnostic approach' above.)

●Acute intracerebral hemorrhage – The most common clinical manifestation of CAA is acute lobar intracerebral hemorrhage (ICH) (image 1). CAA-related lobar hemorrhages most commonly arise in posterior lobar brain regions. Because of their superficial location, CAA-related hemorrhages often extend beyond the brain tissue into the subarachnoid and subdural spaces and less frequently rupture into ventricles. (See 'Acute intracerebral hemorrhage' above.)

Because of the risk of spontaneous lobar hemorrhage in patients with CAA, we weigh the risks and benefits of using of anticoagulant and antiplatelet agents at an individual level. We reserve anticoagulation in CAA patients for those at high risk for thromboembolic complications related to specific indications. (See 'Avoiding anticoagulants and antiplatelet agents' above.)

●Incidental imaging features – Chronic evidence of asymptomatic bleeding and other imaging findings can be detected on brain MRI in CAA patients with or without lobar hemorrhage. (See 'Incidental chronic imaging features' above.)

•Microbleeds (or microhemorrhages) appear as 2 to 10 mm focal areas of hemosiderin deposition on gradient echo or other T2*-weighted sequences (image 7) and have an anatomic distribution similar to that of intracerebral hemorrhage, with a predilection for the cerebral cortex.

•Cortical superficial siderosis (cSS) can represent a focus of remote bleeding related to CAA (image 6) and is thought to be the chronic form of acute convexity subarachnoid hemorrhage.

•Nonhemorrhagic imaging findings such as ischemic microinfarcts, cerebral atrophy, white matter hyperintensities, and multiple dilated perivascular spaces may also be found in patients with CAA.

●Transient focal neurologic episodes – Patients with CAA may present with transient focal neurologic episodes (TFNE) described as recurrent, brief, and often stereotyped spells of weakness, numbness, paresthesias, or other cortical symptoms. Symptoms spread smoothly over contiguous body parts over several minutes and tend to localize to the site of a prior lobar hemorrhage or focus of cSS. (See 'Transient focal neurologic episodes' above.)

●Cerebral amyloid angiopathy-related inflammation – Cerebral amyloid angiopathy-related inflammation (CAA-ri) is a distinct manifestation of CAA. Patients present with acute or subacute cognitive decline, seizures, new persistent headaches, or other progressive neurologic signs (table 5); imaging demonstrates a patchy or confluent immediately subcortical leukoencephalopathy along with lobar microhemorrhages (image 8).

For patients with suspected CAA-ri, we suggest treatment with a course of immunosuppressive therapy, most often with glucocorticoids (Grade 2C). (See 'Cerebral amyloid angiopathy-related inflammation' above.)

●Cognitive impairment – Advanced CAA is associated with cognitive impairment, which may reflect comorbid Alzheimer disease, vascular dementia, or both. (See 'Cognitive impairment' above.)