Abstract Body

Aims

Examine baseline and longitudinal brain pathology (amyloid and tau) and memory decline in cognitively-unimpaired carriers of autosomal dominant Alzheimer's disease.

Methods

Members of the Colombian kindred with the PSEN1 E280A ADAD were enrolled in our ongoing study (40 nondemented carriers and 40 age-matched non-carriers; age 36 +/- 6 years old). Participants underwent a comprehensive neuropsychological evaluation, and Positron Emission Tomography (PET) scanning to measure cortical amyloid and regional tau burden (i.e., inferior temporal tau and entorhinal tau). Mann-Whitney tests and Spearman correlations were conducted to examine group differences and relationships among cognitive performance and AD pathology.

Results

Carriers performed significantly worse on MMSE (p=.001) and CERAD Word List Learning (p=.003), and had greater cortical amyloid (p<.001) and entorhinal tau burden (p=.025), compared to non-carriers. Memory scores negatively correlated with markers of AD pathology (p<.001).

Conclusions

Our results suggest that verbal memory decline relates to markers of AD pathology in cognitively-unimpaired individuals with PSEN1 E280A ADAD. Future research is needed to examine whether memory measures such as the CERAD Word List Learning may be useful in detecting subtle cognitive changes in those at risk for AD, prior to clinical onset.

Aims

There are currently no established tools for monitoring disease progression in Posterior Cortical Atrophy (PCA), an atypical form of Alzheimer's disease. The characteristic visual dysfunction in PCA tends to be so severe at presentation that symptoms cannot be used to track further decline. Here, we explore the potential of Addenbrooke's Cognitive Examination (ACE) as an instrument for measuring PCA progression across cognitive subdomains.

Methods

Twenty participants with PCA and seventeen healthy adults completed repeated ACE assessments over 3-44 months. Trajectories of cognitive change on subdomains of the ACE were assessed using linear mixed effect models. Specifically, we investigated rate of decline in attention/orientation, memory, verbal fluency, and visuospatial skills.

Results

PCA patients had significantly lower scores on all cognitive domains at presentation compared to healthy controls. At baseline, PCA patients had an average score of 28% of visuospatial tasks, with a predicted decline to an average of 2% over the course of 30 months. Compared to controls, PCA patients showed a greater decrease in attention and verbal fluency scores over time. In contrast, longitudinal changes in memory scores did not differ between the groups.

Conclusions

As predicted, PCA patients were severely impaired on visuospatial tasks, impeding their use as a monitoring tool. In contrast, although scores on attention and verbal fluency also decreased to a greater extent than in healthy adults, they did not reach floor levels over a 30-month period. ACE measures of attention and verbal fluency may therefore hold promise as a marker of disease progression in PCA.

Aims

Medial temporal lobe atrophy (MTA), a hallmark of AD, is widely used in the clinical workup of cognitive impairment and is conventionally related to age in order to assess abnormality. Here, we study the effect of multiple pathologies, amyloid, tau, neurodegeneration as well as small vessel disease (SVD) on MTA.

Methods

Data from cognitively unimpaired (N=171) as well as MCI (n=135) and AD (n=98) people from the BioFINDER study included MTA score, hippocampal volume, CSF levels of amyloid, tau and NfL as a proxy for neurodegeneration, tau PET and for a subgroup amyloid PET as well as white matter lesions (WML) as a proxy for SVD. Markers of neuroinflammation, YKL-40, VEGF, sICAM and vICAM were studied in a separate cohort (n=738). Associations were tested using linear regression models.

Results

The effect of age on the hippocampal volume and on MTA decreased with increasing AD stage. There was no effect of amyloid (measured in CSF or with PET) or markers of neuroinflammation. The effect of tau was present only in a whole group analysis, but not in the diagnostic groups separately. The relatively large effect of WML was stronger for the MTA score than for the hippocampal volume.

Conclusions

Our results suggest that apart from age, the largest effect on MTA of any pathologies studied is that of WML.

Aims

Down syndrome(DS) is a genetically determined form of Alzheimer’s disease(AD). Objective: to describe the clinical progression along the ADcontinuum and the associated cognitive decline in a cohort of adults with DS.

Methods

Single-centre longitudinal study. We included adults with DS with at least one follow-up. Participants were clinically classified as: asymptomatic (aDS), prodromalAD (pDS) or ADdementia (dDS). For prognostic evaluation, aDS and pDS were also classified as “progressors” and “non-progressors” based on clincal criteria. We performed KaplanMeier regression analyses to study progression across diagnostic categories and linear mixed effect models to estimate the longitudinal cognitive decline. For cognitive measures we used CAMCOG-DS and CRT.

Results

From February'14 to May'20 we assessed 540 adults with DS (19.1%mild, 50.6%moderate, 21.9%severe, 8.3%profound intellectual disability), 69.7% were aDS, 10.9% pDS and 19.4% dDS. aDS were younger than pDS and dDS (mean age 38, 51, 53 respectively; p=0.001). aDS had a median time to progression to symptomatic AD of 5.4 years. When stratifying by baseline age, we observed a clear age-dependence risk for progression in the aDS group, but not in pDS. Progressors had higher annual cognitive decline in neuropsychological test than non-progressors, 0.25 points and 0.05 respectively (p<0.05). There were no differences in the longitudinal performance in neuropsychological testing when stratifying by intellectual disability (0.005 and -0.017 points/year for CAMCOG and CRT respectively, (p>0.05)).

Conclusions

This longitudinal study confirms the ultra-high risk to develop symptomatic AD in DS. The rate of progression in aDS confirms that a preventive trial would have high statistical power in this population.

Aims

Neuroimaging biomarkers of Alzheimer’s disease are systematically associated, suggesting that a single neuroimaging biomarker could provide pathology-related information reflected by other biomarkers. Here, we aimed at predicting global amyloid status (AB+/-) from regional glucose metabolism in cognitively normal subjects (CN) and patients with mild cognitive impairment (MCI) using various machine learning models.

Methods

We used FDG- and ¹⁸F-Florbetapir-PET (amyloid) scans of 348 CN (35% AB+) and 510 age-matched MCI individuals (58% AB+) from ADNI. Using SPM12, FDG-PET scans were spatially normalized and intensity-standardized (reference: pons) and mean SUVRs were extracted for 140 brain regions using the AAL3 atlas. Individual AB+/- was determined by a cut-off of 1.1 global SUVR based on pre-processed amyloid scans available at ADNI.

We compared models predicting AB+/- from regional FDG-PET signal, employing different classifiers (random forest (RF), support vector machine, k-nearest neighbors, and neural network) and different subsets of brain regions using 10-fold cross-validation. Finally, glucose metabolism in regions selected by the best-performing model was summarized to clusters (frontal, parietal, temporal, occipital, limbic) and compared between AB+ and AB- subjects per group.

Results

In the CN group, FDG-PET signal in 11 predominantly parietal regions dissociated AB+ from AB- with an accuracy of 73% (classifier: RF). In MCI, AB+/- was predicted with 67% accuracy (classifier: RF), based on 17 mainly parietal and limbic regions.

Conclusions

Amyloid status can be predicted from regional FDG signal within groups of non-demented individuals using machine learning. If confirmed in future studies, FDG-PET may serve as a gatekeeper test for subsequent Amyloid-PET.

Aims

Disruption of the endosomal pathway in Down syndrome (DS) results from increased levels of the betaCTF. Our objective was to determine whether targeting betaCTFs with antibodies can prevent the betaCTF-induced endosomal pathway dysfunction in DS.

Methods

Anti-betaCTF monoclonal antibodies were either added to cultured DS fibroblasts or infused into the hippocampi of a DS mouse model (Ts2) and diploid controls. RNA sequencing (RNAseq) was performed comparing gene expression in control and anti-betaCTF-antibody-treated DS fibroblasts. Immunolabeling of early endosomes and retromers and morphometric analysis were performed of cultured fibroblasts and of neurons within the cingulate cortex of the mice. Cell-surface binding and uptake of an anti-cation-dependent mannose 6-phosphate receptor (CD-MPR) antibody in cultured cells was used to assess flux through the endosome-retromer pathway.

Results

KEGG analysis of the RNAseq data identified endocytosis and neurodegenerative disease pathway groups (Alzheimer, Parkinson, Huntington diseases) as being altered by anti-betaCTF-antibody treatment. Multiple endocytic pathway rab-GTPases and retromer genes were differentially expressed. Immunolabeling with early endosome markers revealed an anti-betaCTF-antibody-induced reduction in the endosome area and increase in the retromer area per fibroblast and per neuron in both diploid and DS models. Increased internalization of the CD-MPR from the plasma membrane occurred with anti-betaCTF-antibody in DS fibroblasts, consistent with an increase in retromer-to-Golgi recycling of the CD-MPR.

Conclusions

Anti-betaCTF antibodies rescue early-endosome morphological changes in DS by activating the early-endosome-to-retromer recycling pathway, suggesting anti-betaCTF antibodies as novel therapeutics for neurodegenerative disorders with endosomal dysfunction.

Abstract Body

Mayo Clinic Florida (MCF) Ataxia Center of Excellence was established in 2017. The Center’s aims are to deliver an excellent care and to provide the support for clinical and basic research studies on genetic forms of spinocerebellar ataxias (SCA).

On research side, MCF Ataxia Center of Excellence collects biomarker specimens including urine, blood, skin, and cerebrospinal fluid (CSF). As of writing this abstract (September of 2020), MCF Ataxia Center of Excellence personnel either directly or through the collaborative efforts were able to collect biospecimens from 130 SCA patients and controls. The collection contains specimens from patients with SCA1, SCA2, SCA3, SCA5-SCA8, SCA12, and others. Our sample collection includes 139 blood specimens, 91 urine samples, 78 CSF specimens, and 124 skin fibroblast cultures. Since earlier this year, a longitudinal component has been implemented to track disease progression by re-enrolling participants every two years. As of today, ten patients have returned to partake in the longitudinal arm of the study.

This unique collection allowed us to develop an immunoassay to measure polyQ ATXN3 proteins in biological fluids that can be used as a pharmacodynamic marker specific for SCA3 patients. This discovery will be presented during the meeting.

MCF Ataxia Center of Excellence is rapidly growing and will attract even more patients with SCA who will contribute to further discoveries in ataxia field. A discovery of a pharmacodynamics marker for SCA3 is of paramount importance for future planning of medication trials. Perhaps, it will allow the monitoring of the response to the experimental therap