Eric Westman

Background: Anti‐amyloid therapies lecanemab and donanemab are the first potentially disease‐modifying treatments for Alzheimer's disease (AD) approved in Europe. We estimated real‐world eligibility for these treatments in a large multicenter memory clinic patient cohort in Sweden (MemClin). Methods: MemClin included patients with a first diagnostic visit between 2016 and 2022 at seven specialist memory clinics in Stockholm, Sweden. To estimate eligibility for anti‐amyloid therapy, we applied key inclusion/exclusion criteria from the lecanemab and donanemab appropriate use recommendations. Eligible patients were required to have (1) a clinical diagnosis of AD‐type dementia or mild cognitive impairment (MCI), (2) confirmed Aβ pathology (abnormal CSF Aβ42) and (3) no apparent medical exclusion criteria. Eligibility was calculated for different scenarios to provide both upper‐bound estimates (main analysis) and more conservative, implementation‐ready estimates of near‐term treatable patients (sensitivity analysis). Results: Cohort consisted of 2126 patients (mean age 77.1 ± 6.6 years; 61.8% MCI, 24.2% dementia, 13.7% subjective cognitive impairment). In the main analysis, 11.5% of the whole cohort was estimated to be eligible for either treatment (lecanemab 10.6%, donanemab 11.5%). Most eligible patients had early‐stage disease (61.9% MCI vs. 38.1% dementia) and an A + T– biomarker profile, that is, normal biomarker of tau (57.9%). Eligibility was markedly lower in sensitivity analyses (e.g., 2.7% when all patients without MRI were assumed ineligible). Conclusions: Even upper‐bound estimates indicate that a minority of real‐world memory clinic patients may currently qualify for anti‐amyloid therapy. These findings provide timely evidence to inform the ongoing debate on the implementation of anti‐amyloid therapy in public healthcare.

Population-based studies on the relation between polygenic longevity scores (PGLSs) and specific brain regions on magnetic resonance imaging (MRI) are few. We aimed to investigate this relationship in a population of 70-year-olds from Gothenburg, Sweden ( N = 675). We constructed two PGLSs (one with and one without the APOE locus) using longevity GWAS summary statistics. MRI outcomes included total grey matter, white matter (WM), WM hyperintensities (WMHs), subcortical volumes, small vessel disease (SVD) score, cortical thicknesses, and fractional anisotropy (FA) fibre tracts. We adjusted for sex, age, principal components, and sex interactions, and corrected for multiple comparisons. We found no associations in the total sample. In males, both PGLSs were associated with smaller WMHs volume and lower SVD burden, and a thicker cortex of pars triangularis, pars opercularis, and insula. Our findings suggest a possible link between genetic propensity for longevity and lower cerebrovascular burden and certain cortical grey matter areas in males.

Background: The cortical thinning signature of early-onset Alzheimer’s disease (EOAD) is well established in dementia, but its emergence during earlier disease stages remains unclear. We investigated when it becomes detectable and whether it differs according to fluid amyloid and phosphorylated-tau (A/T1) biomarker profiles. Methods: We studied 292 individuals p -tau181. Participants were classified as A−T1− subjective cognitive decline (SCD; reference group), A+T1−, or A+T1+. Cortical thickness and hippocampal volume were measured using FreeSurfer. EOAD signature regions were defined by comparing biomarker-confirmed AD dementia with the reference group and subsequently examined across A/T1 profiles in SCD and mild cognitive impairment (MCI). Brain–cognition associations were assessed. Results: The cohort comprised 107 reference and 185 amyloid-positive participants (76 A+T1− and 109 A+T1+). EOAD dementia was characterized by marked cortical thinning in posterior parietal and lateral temporal regions. In MCI, cortical thinning was focal in A+T1−, but more extensive in A+T1+ who also showed hippocampal atrophy. At the SCD stage, focal thinning of the EOAD signature, precuneus, and banks of the superior temporal sulcus was observed in A+T1+, whereas no structural alterations were detected in A+T1−. Cortical thinning was associated with episodic and visuospatial memory performance in MCI, while no brain–cognition associations were observed in SCD. Conclusion: EOAD-related cortical thinning is detectable before dementia and varies across A/T1 profiles, with combined amyloid and p -tau181 positivity associated with earlier and more extensive structural involvement.

Background and ObjectivesInterpretation of brain atrophy in multiple sclerosis (MS) relies on group-level research and lacks individualized reference standards. Normative modelling can enable patie...

Background: To optimise management of patients with covert brain infarction (CBI), personalised estimates of stroke and dementia risks are needed. We determined age- and sex-specific stroke and dementia risks in individuals with CBI and identified clinical and imaging prognostic indicators. Methods: We included 2287 participants with MRI-defined CBI from eight community-based cohorts across Europe and the USA, who were age/sex-matched within each cohort to 6861 participants without CBI (1:3). We used Cox models to determine relative risks of stroke and dementia, and subdistribution hazard models to estimate 10-year absolute risks and prognostic indicators among participants with CBI. Findings: Of 9148 participants (mean age 74 years, 53% women), 716 (7·8%) were diagnosed with a stroke and 1311 (14·3%) with dementia during 10-year follow-up. Individuals with CBI had a 2·01-fold (95%CI: 1·66−2·42) increased risk of stroke and a 1·34-fold (1·11−1·62) increased risk of dementia compared to those without CBI. Stroke and dementia risks relative to no CBI were higher with multiple CBI than with single CBI (hazard ratio stroke: 2·51 [1·94–3·26] vs 1·79 [1·49–2·15]; dementia: 1·74 [1·39–2·17] vs 1·22 [1·04–1·44]), but did not differ between lacunes, cortical infarcts or cerebellar infarcts. In individuals with CBI, smoking, hypertension, diabetes, atrial fibrillation, education, APOE-ε4 genotype, and multiple CBI predisposed to stroke and/or dementia. With CBI, ten-year absolute stroke risk was 10·9% (95%CI: 7·4%–15·1%) and 10-year absolute dementia risk was 18·9% (95%CI: 10·5%–29·4%). Absolute risks increased with age from 5·4% (50-59 years) to 16·2% (aged >80) for stroke and from 2·5% to 31·2% for dementia. Interpretation: Community-dwelling individuals with CBI are at increased risk of stroke and dementia, particularly in the presence of multiple infarcts. Demographic, clinical, and imaging characteristics may guide risk factor management, interventions, and trial inclusion.

Brain computed tomography (CT) is an accessible and commonly utilised technique for assessing brain structure. In cases of idiopathic normal pressure hydrocephalus (iNPH), the presence of ventriculomegaly is often neuroradiologically evaluated by visual rating and manually measuring each image. Previously, we have developed a deep-learning-model that utilises transfer learning from magnetic resonance imaging (MRI) for CT-based intracranial tissue segmentation. Accordingly, herein we aimed to enhance the segmentation of ventricular cerebrospinal fluid (VCSF) in brain CT scans and assess the performance of automated brain CT volumetrics in iNPH patient diagnostics. This retrospective study employed a two-stage approach in developing the model. Initially, a 2D U-Net model was trained to predict VCSF segmentations from CT scans, using paired MR-VCSF labels from healthy controls. This model was subsequently refined by incorporating manually segmented lateral CT-VCSF labels from iNPH patients, building on the features learned from the initial U-Net model. The training dataset included 734 CT datasets from healthy controls paired with T1-weighted MRI scans from the Gothenburg H70 Birth Cohort Studies and 62 CT scans from iNPH patients at Uppsala University Hospital. To validate the model's performance across diverse patient populations, external clinical images including scans of 11 iNPH patients from the Universitätsmedizin Rostock, Germany, and 30 iNPH patients from the University of Alabama at Birmingham, United States were used. Further, we obtained three CT-based volumetric measures (CTVMs) related to iNPH. Our analyses demonstrated strong volumetric correlations (ρ=0.91, p<0.001) between automatically and manually derived CT-VCSF measurements in iNPH patients. Based on the ventricular volume, the CTVMs exhibited high accuracy in differentiating iNPH patients from controls in external clinical datasets with an AUC of 0.97 (95% CI: 0.94-1.00) and in the Uppsala University Hospital datasets with an AUC of 0.99 (95% CI: 0.98-1.00). CTVMs derived through deep learning show potential for assessing and quantifying morphological features in hydrocephalus. Critically, these measures performed comparably to gold-standard neuroradiology assessments in iNPH patients and healthy controls, even in the presence of intraventricular shunt catheters. Accordingly, such an approach may serve to improve the radiological evaluations of the diagnostic work-up and treatment response monitoring in patients with hydrocephalus. Since CT is much more widely available than MRI, our results have considerable clinical impact.

Background: Brain atrophy subtypes are increasingly recognized in Alzheimer’s disease (AD) dementia. However, their relevance across the real-world memory clinic spectrum, from subjective cognitive impairment (SCI) and mild cognitive impairment (MCI) to AD and non-AD dementias, remains unclear. This cross-sectional study aimed to identify MRI-based atrophy subtypes in a relatively young memory clinic and examine associations with demographic, cerebrospinal fluid (CSF) biomarkers, and cerebrovascular burden to inform precision medicine approaches. Methods: We included all consecutive patients (SCI to dementia), evaluated at the Karolinska University-Hospital Memory Clinic (Stockholm, Sweden) between 2018 and 2023 with available clinical and 3T MRI data. Subtypes were defined using FreeSurfer-derived volumetric measures and a validated algorithm combining categorical classification (typical, limbic predominant, cortical predominant, minimal atrophy) with continuous indices of typicality (cortical predominant–limbic predominant) and severity (minimal atrophy–typical). Demographics, cognitive profiles, APOE ε4 status, CSF biomarkers (Aβ42, Aβ42/40, phosphorylated [p]-tau181, total tau, neurofilament light chain [NFL]), and cerebrovascular burden were compared across subtypes. Analyses were replicated in Aβ-positive individuals and those eligible for anti-Aβ therapy. Results: Among 809 patients (median age 60.0 years [interquartile-range 56.0–63.0], 56.1% female), 38.2% had SCI, 44.4% MCI, and 17.4% dementia. CSF biomarkers were available in 596 (73.7%). Limbic predominant and typical subtypes had more males (59.3% and 50.0%, respectively; group-wise p < 0.001), higher APOE ε4 frequency (47.7% and 41.0%, p = 0.02), greater cerebrovascular burden, and poorer memory. These subtypes were more often Aβ positive (46.1% and 46.5%, p = 0.01). A cortical predominant pattern was frequent in females (66.0%, p < 0.001), while minimal atrophy was associated with milder cognitive impairment (49.0% SCI, 45.5% MCI) and higher depressive symptoms. In Aβ-positive patients (n = 231), typical and limbic subtypes had higher p-tau181 (median: 83.0 and 84.5 pg/mL, respectively; p < 0.001), NFL (1120.0 and 1125.0 pg/mL, p < 0.001), and lower Aβ42/40 ratios (0.051 and 0.049, p = 0.02). Findings remained consistent across continuous atrophy measures and in the 14.9% (n = 89) eligible for anti-Aβ therapy. Discussion: MRI-based atrophy subtypes exhibit distinct clinical and biomarker profiles, consistently observed in Aβ-positive and anti-Aβ-therapy-eligible patients. These findings support their diagnostic utility in memory clinics and relevance for biologically targeted AD trials.