|
Project Topic
|
Background: Alzheimer’s disease and related dementias are becoming an epidemic and there is an urgent need
to develop effective therapies to prevent or delay onset. However, clinical trials to date have failed to find an
effective drug even though there is some evidence of adequate target engagement in many studies, but without a
corresponding clinical benefit. One reason for these failures may be that multiple co-pathologies, including
neurodegenerative causes together with cerebrovascular disease, underlie the more common sporadic forms of
dementia. Thus drugs used in trials targeting a single neuropathological entity, such as beta-amyloid, do not cover
the full spectrum of underlying pathology.
Overall goal: To leverage large, multimodal datasets from patients representing the full spectrum of
neurodegenerative dementia as well as those at risk and apply well-informed data-driven analytic approaches to
identify neurodegenerative dementia spectra and biotypes that can discern between pure from mixed forms of
dementia during both the presymptomatic and symptomatic stages, i.e., that represent the true complexity of the
underlying pathologies, which does not rely on original clinical diagnosis.
Study Design: We will combine multivariate, heterogeneous, and multimodal genomic, neuroimaging,
cognitive/behavioural, biomarker and demographic datasets from four Canadian dementia cohorts (Ontario
Neurodegenerative Disease Research Initiative [ONDRI], Brain Eye Amyloid Memory [BEAM] study,
Comprehensive Assessment of Neurodegeneration and Dementia (COMPASS-ND) study and the Sunnybrook
Dementia Study [SDS]), a Czech dementia cohort (Pre-clinical genotype-phenotype predictors of Alzheimer's
disease and other dementias, a JPND cohort acquired at CEITEC, MU [APGEM_MU]) and a Czech prodromal
Lewy body diseases cohort (proLBD), and an Italian dementia cohort (Ca‘ Granda Cohort [CGC]), including the full
spectrum of pathologies associated with dementia (Alzheimer’s disease, Parkinson’s-Lewy body disease,
frontotemporal dementia/ amyotrophic lateral sclerosis, cerebrovascular disease), and a longitudinal aging cohort
(Rotterdam Study) to achieve this goal. Specifically, we will evaluate how genomic, age, education, sex, and
cardiovascular risk factors impact neuroimaging, biomarker, and cognitive/behavioural signatures in a diseaseagnostic fashion, i.e. blinded to diagnosis, and use this information to locate an individual subject along dementia
spectra. Dementia biotypes will also be defined using a variety of data-driven analytic approaches developed by the
expertise of our neuroinformatics team. These dementia spectra will also be applied to the aging cohorts to see how
they predict development of dementia/cognitive impairment and which specific biotype, i.e., presymptomatic
disease signatures. Validation will take place using an autopsy subset of the data.
Significance: Understanding shared mechanisms leading to this underlying pathological complexity in late onset
sporadic dementia represent a critical knowledge gap and may inform future clinical trial design and more
appropriate patient selection a priori (i.e., applied precision medicine approaches).
|
