Simulating the outcome of amyloid treatments in Alzheimer's disease from imaging and clinical data

Abstract

In this study, we investigate SimulAD, a novel quantitative instrument for the development of intervention strategies for disease-modifying drugs in Alzheimer's disease. SimulAD is based on the modeling of the spatio-temporal dynamics governing the joint evolution of imaging and clinical biomarkers along the history of the disease, and allows the simulation of the effect of intervention time and drug dosage on the biomarkers' progression. When applied to multi-modal imaging and clinical data from the Alzheimer's Disease Neuroimaging Initiative the method enables to generate hypothetical scenarios of amyloid lowering interventions. The results quantify the crucial role of intervention time, and provide a theoretical justification for testing amyloid modifying drugs in the pre-clinical stage. Our experimental simulations are compatible with the outcomes observed in past clinical trials, and suggest that anti-amyloid treatments should be administered at least 7 years earlier than what is currently being done in order to obtain statistically powered improvement of clinical endpoints.

A multidimensional ODE-based model of Alzheimer’s disease progression

“…Alternatively, we can change the population characteristics to identify the subgroup where the target outcome changes the fastest. In 22 , the authors use their model to simulate the effect of amyloid treatments on the disease course. We do not recommend using our model for this type of causal prediction because we did not do any causal analysis.…”

“…However, other functions can be used to describe the velocity. For example, Gaussian Processes were used in 23 , and a second-order polynomial was used in 22 .…”

“…But it is also an extension of the differential equation models because velocity is the key modelling feature. In particular, our model is conceptually very similar to 22 , the most notable differences being the mathematical implementation and the focus on the causal interpretation. Specifically, our model is more straightforward because we model the progression speed directly on the observed biomarkers without the need to define latent variables; it is more general because we do not prescribe a specific functional form for the velocities; our paper focuses on the prediction of progression, while in 22 , the authors interpret their model from a causal perspective.…”

A multidimensional ODE-based model of Alzheimer's disease progression

“…Alternatively, we can change the population characteristics to identify the subgroup where the target outcome is changing the fastest. In 22 , the authors use their model for simulating the effect of amyloid treatments on the disease course. We do not recommend using our model for this type of causal prediction, because we did not do any causal analysis.…”

“…But it is also an extension of the differential equation models because velocity is the key modelling feature. In particular, our model is conceptually very similar to 22 , the most notable differences being the mathematical implementation and the focus on the causal interpretation. Specifically, our model is more simple because we model the progression speed directly on the observed biomarkers, without the need to define latent variables; it is more general because we don't prescribe a specific functional form for the velocities; the focus of our paper is on prediction of progression, while in 22 the authors interpret their model from a causal perspective.…”