Sequences of cognitive decline in typical Alzheimer's disease and posterior cortical atrophy estimated using a novel event‐based model of disease progression

Abstract

Introduction This work aims to characterize the sequence in which cognitive deficits appear in two dementia syndromes. Methods Event‐based modeling estimated fine‐grained sequences of cognitive decline in clinically‐diagnosed posterior cortical atrophy (PCA) ( ) and typical Alzheimer's disease (tAD) ( ) at the UCL Dementia Research Centre. Our neuropsychological battery assessed memory, vision, arithmetic, and general cognition. We adapted the event‐based model to handle highly non‐Gaussian data such as cognitive test scores where ceiling/floor effects are common. Results Experiments revealed differences and similarities in the fine‐grained ordering of cognitive decline in PCA (vision first) and tAD (memory first). Simulation experiments reveal that our new model equals or exceeds performance of the classic event‐based model, especially for highly non‐Gaussian data. Discussion Our model recovered realistic, phenotypical progression signatures that may be applied in dementia clinical trials for enrichment, and as a data‐driven composite cognitive end‐point.

Quantitative detection and staging of presymptomatic cognitive decline in familial Alzheimer’s disease: a retrospective cohort analysis

“…From cross-sectional data, we estimated the probable sequence of cognitive decline in presymptomatic FAD using an event-based model [25,26]. The event-based model converts input data into severity scores (probability of abnormality) to estimate an order in which a set of measures become abnormal, and also estimate uncertainty in that ordering.…”

“…We first removed healthy linear trends (if present, in noncarriers) of test score with years of formal education, age and sex. Event severity (probability of cognitive abnormality) in each test score is determined here using Kernel Density Estimation mixture modelling [25]. Using a mixture model replaces disease labels (mutation carrier/noncarrier) with pre-/post-event labels to allow for different cognitive profiles (and disease severity) across mutation carriers.…”

Targeted screening for Alzheimer’s disease clinical trials using data-driven disease progression models

“…This is achieved directly from the data distributions in diseased and healthy groups and without a priori -defined disease stages or biomarker cutpoints. The EBM, in its various versions, has been applied to a variety of diseases since 2011, e.g., (19, 8, 22, 23, 24, 25). For a detailed intuitive description of the EBM, we refer the reader to (22).…”

“…For example, potentially targeting the “wrong” pathology at the wrong time — typically amyloid protein pathogens are the target but if a treatment is given to symptomatic individuals, it may be too late to halt or reverse any damage done. Notwithstanding this, enrolling the right people at the right time (disease stage) into a clinical trial remains a considerable challenge because of undetected heterogeneity in phenotype/presentation (8) and/or ensuring the underlying pathology is present (9), which can be a general problem because clinical trials often cannot adapt their designs to accommodate research discoveries made after they have begun. This can result in enrolment of non-responders into a clinical trial that wash out treatment effect in any subgroup of responders.…”

Tau-first subtype of Alzheimer's disease consistently identified across in vivo and post mortem studies

“…|¬ ! ( for each biomarker using the kernel density estimation based mixture modeling approach described by Firth et al 49 We optimized the number of subtypes ( ) in an iterative manner using ten-fold crossvalidation. For each fold we ran SuStaIn on the training data and evaluated the average out-ofsample log likelihood of the held-out testing data, so that we had ten measures for each model.…”