Abstract
INTRODUCTION
The Boston criteria v2.0 for cerebral amyloid angiopathy (CAA) incorporated non‐hemorrhagic imaging markers. Their prevalence and significance in patients with cognitive impairment remain uncertain.
METHODS
We studied 622 memory clinic patients with available magnetic resonance imaging (MRI) and cerebrospinal fluid (CSF) biomarkers. Two raters assessed non‐hemorrhagic markers, and we explored their association with clinical characteristics through multivariate analyses.
RESULTS
Most patients had mild cognitive impairment; median age was 71 years and 50% were female. Using the v2.0 criteria, possible or probable CAA increased from 75 to 383 patients. Sixty‐eight percent of the sample had non‐hemorrhagic CAA markers, which were independently associated with age (odds ratio [OR] = 1.04, 95% confidence interval [CI] = 1.01–1.07), female sex (OR = 1.68, 95% CI = 1.11–2.54), and hemorrhagic CAA markers (OR = 2.11, 95% CI = 1.02–4.35).
DISCUSSION
Two‐thirds of patients from a memory clinic cohort had non‐hemorrhagic CAA markers, increasing the number of patients meeting the v2.0 CAA criteria. Longitudinal approaches should explore the implications of these markers, particularly the hemorrhagic risk in this population.
Highlights
The updated Boston criteria for cerebral amyloid angiopathy (CAA) now include non‐hemorrhagic markers.
The prevalence of non‐hemorrhagic CAA markers in memory clinic patients is unknown.
Two‐thirds of patients in our memory clinic presented non‐hemorrhagic CAA markers.
The presence of these markers was associated with age, female sex, and hemorrhagic CAA markers.
The hemorrhagic risk of patients presenting these type of markers remains unclear.
Keywords: Alzheimer's disease, cerebral amyloid angiopathy, cognitive impairment, CSF, memory clinic, microbleeds, MRI, multispot, non‐hemorrhagic markers, perivascular spaces, superficial siderosis, white matter hyperintensities
1. BACKGROUND
Cerebral amyloid angiopathy (CAA) is a cerebral small‐vessel disease caused by a progressive accumulation of amyloid beta (Aβ) in the wall of cortical and leptomeningeal arteries, arterioles, veins, venules, and capillaries. 1 One of its major manifestations is the occurrence of cerebral hemorrhages, which are thought to be associated with fibrinoid necrosis, microaneurysm formation, and vessel wall remodeling. 2 However, non‐hemorrhagic lesions such as microinfarcts and white matter lesions also develop in patients with CAA as a result of smooth muscle cell loss, reduced vascular reactivity, and obliterative intimal changes. 3 Another non‐hemorrhagic marker of CAA that may be involved in the disease pathophysiology is perivascular space dilation, which occurs in regions with significant Aβ vascular deposition. 4 There is evidence that decreased white matter volume, decreased global cortical thickness, and cortical microinfarcts are additional non‐hemorrhagic markers that are associated with cognitive impairment in these patients. 5 , 6 , 7
The updated Boston criteria v2.0 for CAA 8 reflect this growing body of knowledge and now include non‐hemorrhagic imaging markers, namely, the multispot pattern of white matter hyperintensities (WMH) and a severe burden of enlarged perivascular spaces (EPVS) in the centrum semiovale (CSO). Another change in the criteria, the possibility of diagnosing possible sporadic CAA in patients with cognitive impairment based on the isolated presence of a non‐hemorrhagic marker on magnetic resonance imaging (MRI), has raised some concerns. 9 The Boston criteria v2.0 were derived from a patient population in which one‐third of the patients had cognitive impairment. However, presentation with cognitive impairment was underrepresented in the validation cohort, thereby limiting the generalizability to memory clinic patients with the full range of neurodegenerative causes of cognitive impairment. 8 The fact that subcortical WMH of presumed vascular origin increase with age and are hallmarks of vascular risk factor–related arteriosclerotic small‐vessel disease, which may lead to vascular cognitive impairment, 10 and that the burden of EPVS also increases with age and is increased in the CSO of patients with Alzheimer's disease (AD), 11 confounds the interpretation of non‐hemorrhagic imaging markers of CAA in patients presenting with cognitive impairment in memory clinics.
Our aim was to investigate the frequency of non‐hemorrhagic imaging markers of CAA in a cohort of patients from a memory clinic and to analyze possible associations with clinical, imaging, and cerebrospinal fluid (CSF) markers of neurodegeneration.
2. METHODS
We performed a retrospective analysis based on the Aachen Memory Database, a prospective registry of consecutive patients presenting to the memory clinic of the University Hospital RWTH Aachen Department of Neurology, for diagnostic evaluation of suspected cognitive impairment. The study period was between July 2009 and May 2019. Patients were selected according to the following criteria: (1) available MRI images for visual rating (including fluid‐attenuated inversion recovery [FLAIR], T2‐weighted, and blood‐sensitive sequences); (2) available CSF neurodegeneration markers (Aβ42, Aβ40, Aβ42/Aβ40 ratio, total tau [t‐tau] and phosphorylated tau [p‐tau]). To reduce the bias associated with the interpretation of hemorrhagic and non‐hemorrhagic imaging markers of CAA, we excluded patients with the following conditions: primary or secondary central nervous system (CNS) vasculitis; inflammatory CNS disease (e.g., multiple sclerosis); intracranial tumors causing cognitive impairment; prion diseases; genetic causes of hereditary small vessel disease (e.g., cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy [CADASIL]); mitochondrial diseases; previous treatment with extracorporeal membrane oxygenation; and hematologic and coagulation diseases associated with increased bleeding risk. The study was approved by the local ethics committee (EK 018‐19). The study was conducted in accordance with the recommendations of the Strengthening the Reporting of Observation Studies in Epidemiology (STROBE). 12
RESEARCH IN CONTEXT
Systematic review: The authors reviewed the literature using traditional (e.g., PubMed) sources. The prevalence and significance of the newly proposed non‐hemorrhagic imaging markers for cerebral amyloid angiopathy (CAA) are not yet as widely studied in patients presenting with cognitive impairment. For the multispot pattern, no publications were found.
Interpretation: Our results are consistent with the hypothesis that non‐hemorrhagic imaging markers for CAA are very common in patients in a memory clinic setting, leading to more patients fulfilling the updated CAA criteria (Boston criteria v2.0). They also add evidence to the association between CAA markers, aging, and sex differences in patients with cognitive impairment.
Future directions: Our results substantiate the need for further replication in other cohorts to guarantee and explore possible long‐term implications of non‐hemorrhagic CAA markers in patients with cognitive impairment, namely, (1) hemorrhagic risk and (2) eligibility and safety profile of anti‐amyloid beta (Aß) immunotherapies.
2.1. Clinical, laboratorial, and neuropsychological variables
We collected data on demographics, symptom duration, comorbidities, history of stroke, and family history of dementia in first‐degree relatives, clinical dementia severity at presentation (Clinical Dementia Rating [CDR]), comprehensive neuropsychological assessment, CSF neurodegeneration markers, and clinical diagnosis according to national guidelines. CSF neurodegeneration markers were measured at the University Medical Center Göttingen Neurochemical Laboratory using commercially available assays that have been validated in clinical populations. Patients were classified according to the National Institute on Aging and Alzheimer's Association (NIA‐AA) Research Framework AT(N) system into “normal AD biomarkers,” “Alzheimer's continuum” and “non‐AD pathological change.” 13 The presence of possible or probable CAA was defined according to the Boston criteria v1.5 14 and v2.0. 8 We calculated the 10‐year atherosclerotic cardiovascular disease (ASCVD) 15 risk for individual patients as a marker of global atherosclerotic burden, using the “PooledCohort” package in R version 4.2.2 (R Project for Statistical Computing, Vienna, Austria). Comprehensive neuropsychological assessment included the Mini‐Mental State Examination (MMSE), the Montreal Cognitive Assessment (MoCA), and the Consortium to Establish a Registry for Alzheimer's Disease neuropsychological assessment battery (CERAD‐NAB). We calculated composite scores for a priori defined cognitive domains by averaging individual z‐scores on the CERAD‐NAB based on published normative data, 16 adjusted for age, education, and sex. Cognitive domains included attention (Trail Making Test A [TMT Part A]), executive function (TMT Part B, phonemic fluency), language (naming, semantic fluency), memory (verbal learning, verbal recall, verbal savings and verbal recognition, nonverbal recall), and visuospatial processing (figure copy).
2.2. Magnetic resonance imaging features
The available MRI scans were performed as part of the clinical routine. 1.5‐Tesla MRI was performed in 66% of patients and a 3‐Tesla MRI was performed in 34% of patients. The local standard MRI protocol (Table S1) includes coronal T1‐weighted spin‐echo images, sagittal T2‐weighted turbo spin‐echo images, axial FLAIR images, and axial T2*‐weighted gradient‐echo images. Susceptibility‐weighted imaging (SWI) was performed in only three patients. All MRI features were evaluated independently by two raters (one senior neuroradiologist and one senior neurologist) and included mesial temporal atrophy score (MTA), 17 global cortical atrophy score (GCA), 18 parietal atrophy score (Koedam score), 19 EPVS score for basal ganglia and for CSO, 20 Microbleed Anatomical Rating Scale (MARS), 21 cortical superficial siderosis (cSS), 22 and multispot pattern of WMH. The multispot pattern of WMH of presumed vascular origin was defined as >10 small circular or ovoid non‐confluent hyperintense lesions in the subcortical white matter of both cerebral hemispheres using FLAIR sequences. 8 Severe burden of EPVS was defined as the presence of >20 EPVS. 8 To reduce bias, assessment of the non‐hemorrhagic imaging markers of CAA was performed before the rating of microbleeds and cSS and evaluators were blinded to patients' characteristics.
2.3. Statistical analyses
The study population was grouped according to the presence or absence of non‐hemorrhagic imaging markers of CAA (severe burden of CSO‐EPVSs and/or multispot pattern). Baseline demographic, clinical imaging, and laboratory parameters were compared between groups using Pearson chi‐square, Fischer's exact, Mann–Whitney, and Jonckheere–Terpstra tests. We also present group comparisons between included patients and excluded patients without CSF neurodegeneration markers. In addition, we performed exploratory analyses by separately comparing baseline characteristics of patients with no non‐hemorrhagic markers with patients with multispot pattern and with severe burden of CSO‐EPVS, separately. To understand the impact of reclassification of CAA diagnosis, we compared baseline characteristics of possible and probable CAA defined by the Boston criteria v1.5 with reclassified possible and probable CAA defined by the Boston criteria v2.0, respectively. To identify independent contributors to the presence of non‐hemorrhagic imaging markers of CAA, we constructed a multivariable logistic regression model using the presence of multispot pattern and/or severe burden of CSO‐EPVS as the dependent variable. Age at diagnosis, sex, clinical disease severity at presentation (CDR), atherosclerotic disease burden (10‐year ASCVD risk), amyloid pathology (Aβ42 in CSF), and the presence of possible or probable CAA (according to the Boston criteria v1.5) were included in the model as independent variables. The inclusion of possible or probable CAA according to the Boston criteria v1.5 in this model equates, in this context, to the presence of hemorrhagic markers of CAA. To avoid collinearity between the presence of amyloid pathology (CSF Aβ42) and the other covariates, we constructed a model that only included age, sex, and CSF Aβ42. Odds ratios (ORs) and corresponding 95% confidence intervals (95% CIs) are presented. Three sensitivity analyses were performed: (1) the regression model was performed using Aβ42/40 ratio instead of Aβ42, because the ratio has been suggested to be more specific for detecting amyloid pathology 23 ; (2) the regression model was performed using only data from patients who underwent 3‐Tesla MRI; and (3) the regression model was performed separately for prediction each of the non‐hemorrhagic markers.
Inter‐rater reliability (IRR) analyses of the non‐hemorrhagic imaging markers of CAA visual ratings were performed using intraclass correlation (ICC) for ordinal scales with a two‐way mixed‐effects model and Cohen's weighted kappa. The results showed good IRR values for both the CSO‐EPVS (ICC = 0.76, 95% CI = 0.72–0.80) and the multispot pattern (k = 0.72, 95% CI = 0.65–0.78).
No method for imputing missing data was used. The statistical significance threshold was set at an α value of 0.05. Statistical analyses and visualizations were performed with SPSS version 28.0.1.0 (IBM SPSS Statistics for Windows, IBM Corp, Armonk NY).
2.4. Data availability
Anonymized data may be shared upon reasonable request to the corresponding author, considering data sharing restrictions imposed by the local ethics committee approval and applicable privacy and data protection laws.
3. RESULTS
Among 1778 patients included in the Aachen Memory Database, the main reasons for exclusion from the current study were: unavailable CSF neurodegeneration markers (n = 869), unavailable MRI scans (n = 176), and unavailable blood‐sensitive MRI sequences (n = 84) (Figure 1). The final study population consisted of 622 patients with a median age at diagnosis of 71 years (interquartile range [IQR] = 63–77), 311 were female (50%), median years of education was 11 (IQR = 10–13), and median duration of symptoms was 29 months (IQR = 18–41). Most patients presented with mild cognitive impairment (n = 347, 55.8%), followed by dementia (n = 247, 39.7%) and subjective cognitive impairment (n = 28, 4.5%). The most frequent underlying diagnosis was AD (n = 249, 40.0%), followed by vascular cognitive impairment (n = 104, 16.7%) and affective disorder (n = 59, 9.5%) (a comprehensive list of diagnoses is provided in Table 1). Excluded patients due to unavailability of CSF neurodegeneration markers (list of diagnosis in Table S2) were younger (median age at diagnosis 67 years, IQR = 56–75, p < 0.001), had a higher proportion of subjective cognitive impairment (n = 223, 25.5%, p < 0.001), and a lower proportion of dementia (n = 183, 21.1%, p < 0.001) than included patients. There were no differences in sex distribution (50% vs 46.7%, p = 0.245).
FIGURE 1.
Patient flow chart. CAA, cerebral amyloid angiopathy; CADASIL, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; CNS, central nervous system; MRI, magnetic resonance imaging.
TABLE 1.
List of diagnoses of the final study population.
| Clinical diagnoses | n (%) |
|---|---|
| Alzheimer's disease | 249 (40.0) |
| Vascular cognitive impairment | 104 (16.7) |
| Affective disorder | 59 (9.5) |
| Mixed Alzheimer's disease and vascular cognitive impairment | 51 (8.2) |
| Frontotemporal dementia | 20 (3.2) |
| Parkinson's disease | 14 (2.3) |
| Lewy body disease | 11 (1.8) |
| Primary progressive aphasia | 11 (1.8) |
| Normal pressure hydrocephalus | 11 (1.8) |
| Progressive supranuclear palsy | 5 (0.8) |
| Unclear etiology | 76 (12.2) |
| Other diagnoses | 11 (1.7) |
In the study population, a multispot pattern of WMH was found in 238 patients (38.2%) and severe burden of CSO‐EPVS was found in 329 patients (52.9%). Multispot pattern and/or a severe burden of CSO‐EPVS was present in a total of 425 patients (68.3%).
3.1. Baseline clinical and imaging characteristics
Baseline characteristics of the study population according to the presence of non‐hemorrhagic imaging markers of CAA are shown in Table 2. Patients with non‐hemorrhagic imaging markers of CAA were older (72.8 vs 68.2 years), more often female (53.4% vs 42.6%), more often hypertensive (64.7% vs 52.8%), and had a higher 10‐year ASCVD risk (23.2% vs 14.1%, p < 0.001). Although patients with non‐hemorrhagic imaging markers of CAA tended to have more severe clinical disease according to the CDR, this difference was not statistically significant (p = 0.057). Similarly, no statistically significant difference was found between the two groups on global measures of cognitive function (MMSE, MoCA) or in any of the cognitive composites (attention, executive function, language, memory, visuospatial function). Patients with non‐hemorrhagic imaging markers of CAA had significantly higher MTA scores (p = 0.002), indicating greater severity of mesial temporal atrophy. No differences were found in global cortical atrophy or parietal atrophy measures. All hemorrhagic imaging markers, including strictly lobar microbleeds (p = 0.022), deep microbleeds (p = 0.016), and cSS (p = 0.016), were significantly more frequent in the group of patients with non‐hemorrhagic imaging markers of CAA. Probable or possible CAA, as defined by the Boston criteria v1.5, was significantly more frequent in the group of patients with the multispot pattern and/or severe burden of CSO‐EPVS (14.8% vs 6.1%). We compared separately the group of patients with no non‐hemorrhagic markers of CAA with the group of patients with multispot pattern and with the group of patients with severe burden of CSO‐EPVS (Table S3). The only discrepancies found between the comparisons with no non‐hemorrhagic markers of CAA were that, although patients with multispot pattern were more frequently female, more frequently presented atrial fibrillation, and had a better performance in executive functions in comparison to patients with no non‐hemorrhagic markers of CAA, these associations were not found for patients with severe CSO‐EPVS.
TABLE 2.
Baseline characteristics of the study population according to the presence of non‐hemorrhagic imaging markers of cerebral amyloid angiopathy (multispot pattern and/or severe burden of enlarged perivascular spaces in the centrum semiovale).
|
No non‐hemorrhagic imaging markers of CAA (n = 197) |
Presence of non‐hemorrhagic imaging markers of CAA (n = 425) |
p | |
|---|---|---|---|
| Age at diagnosis (years) | 68.2 (57.4–76.1) | 72.8 (65.8–77.6) | <0.001 |
| Female sex | 84 (42.6) | 227 (53.4) | 0.012 |
| Education (years) a | 11 (10‐14) | 11 (10.13) | 0.271 |
| Symptom duration (months) | 31.4 (21.4–45.7) | 28.3 (16.3–40.2) | 0.120 |
| Dementia in first‐degree relative b | 51 (34.0) | 104 (33.3) | 0.887 |
| Vascular risk factors or diseases | |||
| Arterial hypertension | 104 (52.8) | 275 (64.7) | 0.005 |
| Diabetes | 39 (19.8) | 66 (15.5) | 0.186 |
| Dyslipidemia | 80 (40.6) | 191 (44.9) | 0.311 |
| Current or past smoking | 53 (26.9) | 117 (27.5) | 0.871 |
| Previous ischemic stroke | 20 (10.2) | 55 (12.9) | 0.320 |
| Previous spontaneous ICH/SAH | 3 (1.5) | 8 (1.9) | 0.752 |
| Coronary heart disease | 20 (10.2) | 39 (9.2) | 0.699 |
| Atrial fibrillation | 6 (3.0) | 25 (5.9) | 0.130 |
| High‐density lipoprotein (mg/dL) | 54 (45–66) | 56 (47–73) | 0.029 |
| ASCVD 10‐year risk (%) c | 14.1 (6.3–28.1) | 23.2 (13.1–33.5) | <0.001 |
| ASCVD 10‐year risk c | <0.001 | ||
| Low (<5%) | 42 (23.6) | 30 (8.4) | |
| Borderline (5%–7.4%) | 11 (6.2) | 22 (6.1) | |
| Intermediate (7.5%–19.9%) | 51 (28.7) | 100 (27.9) | |
| High (>20%) | 74 (41.6) | 207 (57.7) | |
| Clinical Dementia Rating Scale d | 0.057 | ||
| Subjective cognitive impairment | 15 (7.7) | 13 (3.1) | |
| Mild cognitive impairment | 115 (59.3) | 241 (57.4) | |
| Mild dementia | 54 (27.8) | 133 (31.7) | |
| Moderate dementia | 10 (5.2) | 30 (7.1) | |
| Severe dementia | 0 | 3 (0.7) | |
| Mini‐Mental State Examination | 25 (22–28) | 26 (22–28) | 0.611 |
| Montreal Cognitive Assessment | 20 (15–24) | 20 (16–23) | 0.793 |
| Cognitive composite scores | |||
| Attention | −1.10 (−2.15 to −0.35) | −1.25 (−2.10 to −0.50) | 0.195 |
| Executive function | −1.15 (−1.90 to −0.38) | −0.97 (−1.70 to −0.25) | 0.071 |
| Language | −0.85 (−1.88 to −0.14) | −0.85 (−1.65 to −0.18) | 0.737 |
| Memory | −1.47 (−2.34 to −0.75) | −1.63 (−2.60 to −0.68) | 0.649 |
| Visuospatial | −1.10 (−2.20 to −0.70) | −0.80 (−2.0 to 0.70) | 0.451 |
| Magnetic resonance imaging | |||
| Mesial temporal atrophy (right) | 1 (0–2) | 2 (1–2) | 0.002 |
| Mesial temporal atrophy (left) | 1 (0–2) | 2 (1–2) | 0.002 |
| Global cortical atrophy | 1 (0–1) | 1 (1–1) | 0.190 |
| Parietal atrophy (Koedam) | 0 (0–1) | 0 (0–1) | 0.210 |
| Strictly lobar microbleeds (≥1) | 13 (6.6) | 54 (12.7) | 0.022 |
| Deep microbleeds (≥1) | 10 (5.1) | 47 (11.1) | 0.016 |
| Superficial cortical siderosis | 1 (0.5) | 17 (4.0) | 0.016 |
| Cerebrospinal fluid | |||
| Aβ42 (pg/mL) | 677 (470–961) | 598 (431–882) | 0.011 |
| Aβ40 (pg/mL) e | 10190 (7473–13602) | 10550 (7701–13517) | 0.647 |
| Aβ42/40 ratio e | 0.71 (0.45–0.97) | 0.58 (0.40–0.83) | 0.008 |
| Total tau protein (pg/mL) | 294 (182–501) | 348 (211–576) | 0.042 |
| Phosphorylated tau (pg/mL) | 51 (38–81) | 61 (41–93) | 0.006 |
| AT(N) classification | 0.107 | ||
| Normal AD biomarkers | 98 (49.7) | 173 (40.7) | |
| Alzheimer's continuum | 64 (32.5) | 163 (38.4) | |
| Non‐AD pathological change | 35 (17.8) | 89 (20.9) | |
| CAA (Boston criteria v1.5) | 0.008 | ||
| Possible CAA | 8 (4.1) | 43 (10.1) | |
| Probable CAA | 4 (2.0) | 20 (4.7) | |
| CAA (Boston criteria v2.0) | <0.001 | ||
| Possible CAA | 7 (3.6) | 307 (72.2) | |
| Probable CAA | 5 (2.5) | 64 (15.1) |
Abbreviations: AD, Alzheimer´s disease; ASCVD, atherosclerotic cardiovascular disease; CAA, cerebral amyloid angiopathy; ICH, intracerebral hemorrhage; SAH, subarachnoidal hemorrhage.
Available for 584 patients.
Available for 462 patients.
Available for 537 patients.
Available for 614 patients.
Available for 548 patients.
3.2. CSF neurodegeneration markers and AT(N) classification
The distribution of CSF neurodegeneration markers in the groups of patients with and without multispot pattern and/or a severe CSO‐EPVS burden is shown in Figure 2. Patients with non‐hemorrhagic imaging markers of CAA had lower Aβ42 (median 598 vs 677 pg/mL, p = 0.011), lower Aβ42/40 ratio (median 0.58 vs 0.71, p = 0.008), higher t‐tau (median 348 vs 294 pg/mL, 0.042), and higher p‐tau (median 61 vs 51 pg/mL, p = 0.006) in CSF. Despite significant differences in concentrations and ratios of CSF neurodegeneration markers between the two groups, there were no statistically significant differences with respect to AT(N) classification (Table 2). These differences in the distribution of CSF neurodegeneration markers remained unchanged when comparisons were done separately for each of the non‐hemorrhagic markers (Figure S1).
FIGURE 2.
Differences between patients with and without non‐hemorrhagic imaging markers of cerebral amyloid angiopathy (CAA) in cerebrospinal fluid (CSF) biomarkers. Patients with non‐hemorrhagic imaging markers of CAA show significantly lower levels of Aβ42 and Aβ42/40 ratio, as well as higher concentrations of t‐tau and p‐tau. In multivariate analyses, none of the cerebrospinal fluid (CSF) biomarkers were independently associated with an increased likelihood of non‐hemorrhagic imaging markers of CAA in patients presenting with cognitive impairment. Aβ, amyloid β; p‐tau, phosphorylated tau; t‐tau, total tau.
3.3. Reclassification of CAA diagnosis
Possible CAA was found in 51 patients (8.2%) according to the v1.5 criteria, and in 314 patients (50.5%) according to the v2.0 criteria. Probable CAA was found in 24 patients (3.9%) according to the v1.5 criteria, and in 69 patients (11.1%) according to the v2.0 criteria. Using the v2.0 criteria for patient reclassification, 308 patients were reclassified as possible CAA and 6 patients maintained the possible CAA diagnosis, whereas 45 patients were reclassified as probable CAA and 24 maintained the probable CAA diagnosis.
In comparison to patients with possible CAA (v1.5), patients reclassified as possible CAA (v.2.0) were younger (median age 69 vs 74 years, p < 0.001), less frequently had previous ischemic stroke (8.4 vs 19.6%, p = 0.014), less frequently had previous spontaneous intracerebral hemorrhage or subarachnoidal hemorrhage (0 vs 3.9%, p < 0.001), had a lower ASCVD 10‐year risk (21.8 vs 28.4%, p = 0.003), had higher global cognition scores on MMSE (26 vs 23, p = 0.027) and on MoCA (20 vs 18, p = 0.005), less frequently had strictly lobar microbleeds (0 vs 82.4%, p < 0.001), less frequently had superficial cSS (0.3 vs 15.7%, p < 0.001), and a different AT(N) classification distribution (p = 0.020) (Table S4).
Baseline characteristics of patients with probable CAA (v1.5) and of patients reclassified as probable CAA (v2.0) were similar, except that the latter presented higher parietal atrophy scores (p = 0.040) and less frequently presented strictly lobar microbleeds (80 vs 100%, p = 0.019) (Table S5).
In comparison to patients without CAA, patients reclassified as possible or probable CAA (v2.0) presented lower Aβ42/40 ratio (median 0.60 vs 0.70, p = 0.015), and higher p‐tau (median 61 vs 52 pg/mL, p = 0.010). In comparison to patients without CAA, patients reclassified as possible CAA (v2.0) presented lower Aβ42/40 ratio (median 0.60 vs 0.70, p = 0.018), and higher p‐tau (median 61 vs 52 pg/mL, p = 0.011). In comparison to patients without CAA, patients reclassified as probable CAA (v2.0) presented lower Aβ42 (median 541 vs 671 pg/mL, p = 0.026), lower Aβ42/40 ratio (median 0.48 vs 0.70, p = 0.003), higher p‐tau (median 72 vs 52 pg/mL, p = 0.027), and higher t‐tau (median 380 vs 299 pg/mL, p = 0.005). , ,
3.4. Multivariable logistic regression models
Variables independently associated with the presence of non‐hemorrhagic CAA imaging markers in the multivariable regression model were age at diagnosis (OR per 1‐year increase = 1.04, 95% CI = 1.01–1.07, p = 0.015), female sex (OR = 1.68, 95% CI = 1.11–2.54, p = 0.013), and the presence of possible or probable CAA according to the v1.5 criteria (OR = 2.11, 95% CI = 1.02–4.35, p = 0.043) (Table 3). Replacing Aβ42 with Aβ42/40 ratio in the multivariable regression model did not significantly change the results (Table S6). Aβ42 in CSF was also not independently associated with the presence of non‐hemorrhagic CAA imaging markers when including only age and sex as covariates (Table S7). When only patients with 3‐Tesla MRI were selected (n = 210), no variable was independently associated with the presence of multispot pattern and/or a severe burden of CSO‐EPVS (Table S8). When the multivariable logistic regression analysis was performed separately for each of the non‐hemorrhagic markers, the results did not change significantly, except for age not being an independent predictor of CSO‐EPVS (Tables S9–10).
TABLE 3.
Multivariable logistic regression model using the presence of non‐hemorrhagic imaging markers of cerebral amyloid angiopathy (multispot pattern and/or severe burden of enlarged perivascular spaces in the centrum semiovale) as the dependent variable.
| Odds ratio (95% confidence interval) | p | |
|---|---|---|
| Age (per 1‐year increase) | 1.04 (1.01–1.07) | 0.015 |
| Female sex | 1.68 (1.11–2.54) | 0.013 |
| Clinical Dementia Rating (per 1‐point increase) | 1.19 (0.76–1.85) | 0.443 |
| ASCVD 10‐year risk (per‐1% increase) | 1.01 (0.99–1.03) | 0.479 |
| Aβ42 in CSF (per 1‐pg/mL increase) | 1.00 (0.99–1.00) | 0.651 |
| Possible or probable CAA a | 2.11 (1.02–4.35) | 0.043 |
Abbreviations: ASCVD, atherosclerotic cardiovascular disease; CAA, cerebral amyloid angiopathy; CSF, cerebrospinal fluid.
According to the Boston criteria v1.5.
4. DISCUSSION
This study shows a high frequency of non‐hemorrhagic imaging markers of CAA in patients presenting to a memory clinic, with a multispot pattern of WMH and/or a severe CSO‐EPVS burden identified in two‐thirds of patients. Consequently, the number of patients fulfilling the diagnostic criteria for possible or probable CAA also increased significantly by using the Boston criteria v2.0.
Our results are consistent with previous reports which, although using different definitions and methodological approaches, described a severe burden of CSO‐EPVS in memory clinics and AD patients ranging from 20% to 58%. 11 , 24 , 25 , 26 It is also comparable to observational studies, such as the Framingham Heart Study, which found a severe CSO‐EPVS burden in over 60% of participants over 60 years of age. 27 Similarly, the prevalence of a multispot pattern of WMH found in our cohort (38%) is comparable to that found in the derivation and validation cohorts of the Boston criteria v2.0 (33% and 35%, respectively). 8 It is generally accepted that WMH are common in patients presenting with cognitive decline, namely, in patients with AD and mild cognitive impairment, and they appear to contribute to the degree of cognitive impairment. 28 , 29 For example, in the Amsterdam Dementia Cohort, 29% of patients with AD had WMH classified as Fazekas grade 2 or 3, which were associated with older age and vascular risk factors. 30 However, none of the previous studies in memory clinic settings specifically investigated the presence of the multispot pattern. Because the multispot pattern cannot be inferred from scales such as the Fazekas Scale or the Age‐Related White Matter Changes Scale (ARWMC), its prevalence in other memory clinic cohorts remains unknown.
Evidence from clinicopathological studies shows that CAA is a common (co)pathology in cognitive impairment and dementia. A recent meta‐analysis of the prevalence of CAA suggested that in patients with AD, the prevalence of CAA based on pathology is twice as high as the prevalence based on the presence of strictly lobar cerebral microbleeds, used as an imaging marker of CAA, and in the general population, this difference is even three times higher. 31 Thus, the high prevalence of CAA diagnoses in our cohort is plausible, given the likely overlap with AD, which seems to be supported by the finding that reclassified patients had a CSF profile characterized mainly by lower Aβ42/40 ratio levels and higher p‐tau levels, suggesting amyloid pathology.
The most relevant consequences of CAA reclassification introduced by the Boston criteria v2.0 was the 6‐fold increase in the number of possible CAA and associated demographic and clinical profile. Reclassified patients with possible CAA were younger, had higher global cognition scores, and less frequently presented hemorrhagic imaging markers. This suggests that the v2.0 criteria may detect possible CAA at an earlier stage of disease than the v1.5 criteria, aligning with the pathophysiological CAA disease timeline proposed by Koemans et al. 32 Despite this, reclassified patients still exhibited lower Aβ42/40 ratios and higher p‐tau levels, supporting the presence of CNS amyloid pathology.
The high prevalence of non‐hemorrhagic imaging markers of CAA in memory clinic patients raises the question of whether these imaging markers are specific for CAA 9 or whether they may also be a manifestation of aging, other neurodegenerative processes, and/or arteriosclerotic cerebral small vessel disease. Our data demonstrate that although the frequency of non‐hemorrhagic CAA markers is high in this patient population, they are indeed independently associated with hemorrhagic CAA markers, which are the hallmark criteria in the previous Boston criteria. Thus, this association supports the link between non‐hemorrhagic markers and vascular Aβ deposition. It is hypothesized that this type of non‐hemorrhagic cerebral lesions occur in an advanced stage of CAA, relatively late in the disease's progression, but before the onset of intracerebral hemorrhage, which is the most widely acknowledged clinical manifestation of CAA. 32 Therefore, future steps should clarify the clinical significance and long‐term implications of non‐hemorrhagic markers of CAA in this specific population, namely, their intracranial hemorrhagic risk. 33
Another long‐term implication relates to the safety profile of anti‐Aβ monoclonal antibody (Aβ‐mAbs) treatments in patients with AD. A substantial number of patients with AD who would otherwise be eligible for Aβ‐mAbs treatment may also have CAA, given the overlap between the two pathologies. This is particularly critical because pre‐existing CAA, defined by the presence of hemorrhagic markers such as lobar microbleeds or cSS, is recognized as a major risk factor for the development of amyloid‐related imaging abnormalities (ARIA) 34 , 35 in patients receiving Aβ‐mAbs. ARIA can manifest as parenchymal edema or sulcal effusion (ARIA‐E) or as hemorrhage or hemosiderin deposition (ARIA‐H), and may cause clinical symptoms, neurological deficits, and, rarely, death. Because non‐hemorrhagic CAA markers are thought to reflect an already advanced stage of CAA, 32 it is plausible that they may also increase the risk of ARIA. However, there are currently no data on non‐hemorrhagic CAA markers in patients who have developed Aβ‐mAbs‐related ARIA to support this hypothesis.
We found that increasing age was an independent and robust predictor of the presence of non‐hemorrhagic CAA markers. This finding is not unexpected, as age is one of the most important risk factors for both neurodegenerative and vascular cerebral pathologies. 36 In a recently published pooled analysis of individual participant data from 10 population‐based studies, increasing age was found to be one of the most important determinants of increased burden of EPVS in several brain regions, namely, in the CSO. 37 However, as age was not an independent predictor of CSO‐EPVS in our study, the independent association between age and non‐hemorrhagic CAA markers may be driven primarily by the presence of WMH. This association between age and increasing volume of WMH has also been demonstrated in several population‐based studies of healthy older adults. 38 , 39
We also found an independent association between female sex and the presence of non‐hemorrhagic CAA markers. Although in some population‐based studies, 37 cohorts of patients with CAA, 40 and memory clinic cohorts, 41 no association between sex and CSO‐EPVS burden was found, other studies have found an increased CSO‐EPVS burden in women with CAA 42 and a higher age‐related rate of increased EPVS in women. 43 Very little was found in the literature on the question of sex differences and the presence of the multispot pattern. However, there is emerging evidence, consistent with our observation, that women with and without cognitive impairment have a greater WMH burden, 44 , 45 which may reduce their resilience to other types of neuropathology. 46
In our study, we found that an ASCVD 10‐year risk was not linked independently with non‐hemorrhagic CAA markers. This challenges the assumption that WMH and increased EPVS burden, common in cognitive impairment cohorts, have a primary vascular origin. Recent studies on various cohorts, including those with autosomal dominant AD and late‐onset AD, showed that traditional cardiovascular risk scores did not predict volume or longitudinal increases of WMH. 47 These findings suggest that WMH in cognitive impairment cohorts may be more associated with neurodegenerative processes related to amyloid deposition rather than traditional vascular risk factors. Similarly, our study found no independent association between Aβ42 CSF levels and non‐hemorrhagic CAA markers, possibly due to similarities in Aβ42 levels between CAA and AD without CAA markers, 48 and the enrichment of our study population with AD. In addition, this lack of association may also be related to the high prevalence of patients with isolated non‐hemorrhagic CAA markers, which appears to represent an earlier stage of the disease and may be associated with higher Aβ42 levels compared to CAA patients with hemorrhagic markers. 32
Given the study's monocentric focus, results need cross‐validation in further cohorts. Further research in heterogeneous samples can elucidate the role of various clinical and imaging features. The lack of significant results in the sensitivity analysis using only 3‐Tesla MRI data likely stems from reduced statistical power. Despite limitations in assessing non‐hemorrhagic CAA markers and the absence of validated automated approaches, efforts were made to enhance visual rating quality using standardized methods and independent raters, with resulting good inter‐rater reliability. Retrospective use of clinical data and patient selection based on available CSF introduces potential heterogeneity and bias. Although clinicopathological validation of v2.0 criteria was beyond the scope of this study, the absence of neuropathological data weakens the evidence level of our findings.
The most striking finding to emerge from this study is that two‐thirds of patients in our memory clinic cohort had non‐hemorrhagic markers of CAA, resulting in a notable increase in the number of patients meeting the v2.0 criteria for possible or probable CAA. Overall, these data suggest that the specificity of the Boston criteria v2.0 may be lower in patients with cognitive decline than in those with hemorrhagic presentations. This is particularly important given that the Boston criteria are designed to predict the neuropathological presence of CAA and not to exclude other neurodegenerative disorders, such as AD, or to suggest whether or how CAA contributes to cognitive symptoms. Therefore, additional pathological validation studies, as well as longitudinal clinical data are needed to further clarify how reliably the updated criteria for CAA can be translated into clinical routine in memory clinics, and what are their implications in patient management.
CONFLICT OF INTEREST STATEMENT
The authors report no conflict of interest with regard to the current work. Author disclosures are available in Supporting Information.
CONSENT STATEMENT
The need for informed consent was waived by the ethics committee, given the retrospective nature of the data.
Supporting information
Supporting information
Supporting information
ACKNOWLEDGMENTS
International Training Grant (T2201) from the Alzheimer Forschung Initaitive (AFI), START grant T2201 from the Faculty of Medicine RWTH Aachen University, grant START118/20, HabilitationStipend2023 to Ana Sofia Costa.
Open access funding enabled and organized by Projekt DEAL.
Costa AS, Albrecht M, Reich A, et al. Non‐hemorrhagic imaging markers of cerebral amyloid angiopathy in memory clinic patients. Alzheimer's Dement. 2024;20:4792–4802. 10.1002/alz.13920
REFERENCES
- 1. Thal DR, Ghebremedhin E, Rüb U, Yamaguchi H, Del Tredici K, Braak H. Two types of sporadic cerebral amyloid angiopathy. J Neuropathol Exp Neurol. 2002;61:282‐293. [DOI] [PubMed] [Google Scholar]
- 2. Vinters HV, Vonsattel JP. Neuropathologic features and grading of Alzheimer‐related and sporadic CAA. Cereb Amyloid Angiopathy Alzheimes Dis Relat Disord. 2000:137‐155. [Google Scholar]
- 3. van Veluw SJ, Scherlek AA, Freeze WM, et al. Different microvascular alterations underlie microbleeds and microinfarcts. Ann Neurol. 2019;86:279‐292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Perosa V, Oltmer J, Munting LP, et al. Perivascular space dilation is associated with vascular amyloid‐β accumulation in the overlying cortex. Acta Neuropathol. 2022;143:331‐348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Fotiadis P, Reijmer YD, Van Veluw SJ, et al. White matter atrophy in cerebral amyloid angiopathy. Neurology. 2020;95:e554‐e562. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Subotic A, McCreary CR, Saad F, et al. Cortical thickness and its association with clinical cognitive and neuroimaging markers in cerebral amyloid angiopathy. J Alzheimers Dis. 2021;81:1663‐1671. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Xiong L, Van Veluw SJ, Bounemia N, et al. Cerebral cortical microinfarcts on magnetic resonance imaging and their association with cognition in cerebral amyloid angiopathy. Stroke. 2018;49:2330‐2336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Charidimou A, Boulouis G, Frosch MP, et al. The Boston criteria version 2.0 for cerebral amyloid angiopathy: a multicentre, retrospective, MRI–neuropathology diagnostic accuracy study. Lancet Neurol. 2022;21:714‐725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Biessels GJ, Costa AS. Cerebral amyloid angiopathy‐how to translate updated diagnostic criteria for this multifaceted disorder to clinical practice? JAMA Neurol. 2023;80:225‐226. [DOI] [PubMed] [Google Scholar]
- 10. Duering M, Biessels GJ, Brodtmann A, et al. Neuroimaging standards for research into small vessel disease—advances since 2013. Lancet Neurol. 2023;22:602‐618. [DOI] [PubMed] [Google Scholar]
- 11. Banerjee G, Kim HJ, Fox Z, et al. MRI‐visible perivascular space location is associated with Alzheimer's disease independently of amyloid burden. Brain. 2017;140:1107‐1116. [DOI] [PubMed] [Google Scholar]
- 12. von Elm E, Altman DG, Egger M, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet Lond Engl. 2007;370:1453‐1457. [DOI] [PubMed] [Google Scholar]
- 13. Jack CR, Bennett DA, Blennow K, et al. NIA‐AA Research Framework: toward a biological definition of Alzheimer's disease. Alzheimers Dement. 2018;14:535‐562. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Linn J, Halpin A, Demaerel P, et al. Prevalence of superficial siderosis in patients with cerebral amyloid angiopathy. Neurology. 2010;74:1346‐1350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Goff DC, Lloyd‐Jones DM, Bennett G, et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129:S49‐S73. [DOI] [PubMed] [Google Scholar]
- 16. Schmid NS, Ehrensperger MM, Berres M, Beck IR, Monsch AU. The extension of the German CERAD Neuropsychological Assessment Battery with tests assessing subcortical, executive and frontal functions improves accuracy in dementia diagnosis. Dement Geriatr Cogn Disord Extra. 2014;4:322‐334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Scheltens P, Leys D, Barkhof F, et al. Atrophy of medial temporal lobes on MRI in “probable” Alzheimer's disease and normal ageing: diagnostic value and neuropsychological correlates. J Neurol Neurosurg Psychiatry. 1992;55:967‐972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Pasquier F, Leys D, Weerts JG, Mounier‐Vehier F, Barkhof F, Scheltens P. Inter‐ and intraobserver reproducibility of cerebral atrophy assessment on MRI scans with hemispheric infarcts. Eur Neurol. 1996;36:268‐272. [DOI] [PubMed] [Google Scholar]
- 19. Koedam ELGE, Lehmann M, van der Flier WM, et al. Visual assessment of posterior atrophy development of an MRI rating scale. Eur Radiol. 2011;21:2618‐2625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Doubal FN, MacLullich AMJ, Ferguson KJ, Dennis MS, Wardlaw JM. Enlarged perivascular spaces on MRI are a feature of cerebral small vessel disease. Stroke. 2010;41:450‐454. [DOI] [PubMed] [Google Scholar]
- 21. Gregoire SM, Chaudhary UJ, Brown MM, et al. The Microbleed Anatomical Rating Scale (MARS): reliability of a tool to map brain microbleeds. Neurology. 2009;73:1759‐1766. [DOI] [PubMed] [Google Scholar]
- 22. Charidimou A. Cortical superficial siderosis presumed due to cerebral amyloid angiopathy: minimum standards for rating and reporting. Am J Neuroradiol. 2016;37:E43‐E44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Hansson O, Lehmann S, Otto M, Zetterberg H, Lewczuk P. Advantages and disadvantages of the use of the CSF Amyloid β (Aβ) 42/40 ratio in the diagnosis of Alzheimer's disease. Alzheimers Res Ther. 2019;11:1‐15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Choe YM, Baek H, Choi HJ, et al. Association between enlarged perivascular spaces and cognition in a memory clinic population. Neurology. 2022;99:e1414‐e1421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Gertje EC, van Westen D, Panizo C, Mattsson‐Carlgren N, Hansson O. Association of enlarged perivascular spaces and measures of small vessel and Alzheimer disease. Neurology. 2021;96:e193‐e202. [DOI] [PubMed] [Google Scholar]
- 26. Nagaraja N, DeKosky S, Duara R, et al. Imaging features of small vessel disease in cerebral amyloid angiopathy among patients with Alzheimer's disease. NeuroImage Clin. 2023;38:103437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Lara FR, Scruton AL, Pinheiro A, et al. Aging, prevalence and risk factors of MRI‐visible enlarged perivascular spaces. Aging. 2022;14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. van den Berg E, Geerlings MI, Biessels GJ, Nederkoorn PJ, Kloppenborg RP. White matter hyperintensities and cognition in mild cognitive impairment and alzheimer's disease: a domain‐specific meta‐analysis. J Alzheimers Dis. 2018;63:515‐527. [DOI] [PubMed] [Google Scholar]
- 29. Kučikienė D, Costa AS, Banning LCP, et al. The role of vascular risk factors in biomarker‐based AT(N) groups: a German‐Dutch memory clinic study. J Alzheimers Dis. 2022;87:185‐195. [DOI] [PubMed] [Google Scholar]
- 30. Benedictus MR, Goos JDC, Binnewijzend MAA, et al. Specific risk factors for microbleeds and white matter hyperintensities in Alzheimer's disease. Neurobiol Aging. 2013;34:2488‐2494. [DOI] [PubMed] [Google Scholar]
- 31. Jäkel L, De Kort AM, Klijn CJM, Schreuder FHBM, Verbeek MM. Prevalence of cerebral amyloid angiopathy: a systematic review and meta‐analysis. Alzheimers Dement. 2022;18(1):10‐28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Koemans EA, Chhatwal JP, Van Veluw SJ, et al. Progression of cerebral amyloid angiopathy: a pathophysiological framework. Lancet Neurol. 2023;22:632‐642. [DOI] [PubMed] [Google Scholar]
- 33. Fandler‐Höfler S, Gattringer T, Enzinger C, Werring DJ. Comparison of Boston criteria v2.0/v1.5 for cerebral amyloid angiopathy to predict recurrent intracerebral hemorrhage. Stroke. 2023;54(7):1901‐1905. [DOI] [PubMed] [Google Scholar]
- 34. Hampel H, Elhage A, Cho M, Apostolova LG, Nicoll JAR, Atri A. Amyloid‐related imaging abnormalities (ARIA): radiological, biological and clinical characteristics. Brain J Neurol. 2023;146(11):4414‐4424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Smith EE, Greenberg SM, Black SE. The impending era of beta‐amyloid therapy: clinical and research considerations for treating vascular contributions to neurodegeneration. Cereb Circ—Cogn Behav. Forthcoming. [Google Scholar]
- 36. Stephan BCM, Gaughan DM, Edland S, Gudnason V, Launer LJ, White LR. Mid‐ and later‐life risk factors for predicting neuropathological brain changes associated with Alzheimer's and vascular dementia: the Honolulu Asia Aging Study and the Age, Gene/Environment Susceptibility‐Reykjavik Study. Alzheimers Dement. 2023;19:1705‐1713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Evans TE, Knol MJ, Schwingenschuh P, et al. Determinants of perivascular spaces in the general population: a pooled cohort analysis of individual participant data. Neurology. 2023;100:e107‐e122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Vernooij MW, Ikram MA, Tanghe HL, et al. Incidental findings on brain MRI in the general population. N Engl J Med. 2007;357:1821‐1828. [DOI] [PubMed] [Google Scholar]
- 39. Lampe L, Kharabian‐Masouleh S, Kynast J, et al. Lesion location matters: the relationships between white matter hyperintensities on cognition in the healthy elderly. J Cereb Blood Flow Metab Off. 2019;39:36‐43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Koemans EA, Castello JP, Rasing I, et al. Sex differences in onset and progression of cerebral amyloid angiopathy. Stroke. 2023;54:306‐314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Shams S, Martola J, Charidimou A, et al. Topography and determinants of magnetic resonance imaging (MRI)‐visible perivascular spaces in a large memory clinic cohort. J Am Heart Assoc. 2017;6:e006279. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Pinho J, Almeida FC, Araújo JM, et al. Sex‐specific patterns of cerebral atrophy and enlarged perivascular spaces in patients with cerebral amyloid angiopathy and dementia. Am J Neuroradiol. 2023;44(7):792‐798. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Lynch KM, Sepehrband F, Toga AW, Choupan J. Brain perivascular space imaging across the human lifespan. NeuroImage. 2023;271:120009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Alqarni A, Jiang J, Crawford JD, et al. Sex differences in risk factors for white matter hyperintensities in non‐demented older individuals. Neurobiol Aging. 2021;98:197‐204. [DOI] [PubMed] [Google Scholar]
- 45. Exalto LG, Boomsma JMF, Babapour Mofrad R, et al. Sex differences in memory clinic patients with possible vascular cognitive impairment. Alzheimers Dement. 2020;12(1):e12090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Morrison C, Dadar M, Collins DL, Alzheimer's Disease Neuroimaging Initiative . Sex differences in risk factors, burden, and outcomes of cerebrovascular disease in Alzheimer's disease populations. Alzheimers Dement. 2024;20(1):34‐46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Shirzadi Z, Schultz SA, Yau W‐YW, et al. Etiology of white matter hyperintensities in autosomal dominant and sporadic Alzheimer disease. JAMA Neurol. 2023;80(12):1353‐1363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. van den Berg E, Nilsson J, Kersten I, et al. Cerebrospinal fluid panel of synaptic proteins in cerebral amyloid angiopathy and Alzheimer's disease. J Alzheimers Dis. 2023;92:467‐475. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting information
Supporting information
Data Availability Statement
Anonymized data may be shared upon reasonable request to the corresponding author, considering data sharing restrictions imposed by the local ethics committee approval and applicable privacy and data protection laws.