Marina Savelieva (1), Luyuan Qi (2), Helene Karcher (3), Gorana Capkun-Niggli (1), Angelika Caputo (1), Vladimir Bezlyak (1), Valery Risson (1)
(1) Novartis Pharma AG, Basel, Switzerland; (2) Certara France, Paris, France, (3) PAREXEL International AG, Basel, Switzerland
Introduction: Alzheimer’s disease (AD) is a degenerative brain disease and the most common form of dementia. Current research is focusing on diagnosing the AD prior to clinical symptoms appearance [1]. The early stages of disease are defined clinically by the level of cognitive decline alone since functional decline is not apparent [2]. Recent studies suggested that progress of cognitive impairment is followed by subsequent decline in functional abilities [3]; however, the causal and temporal relationship between the two was uncharted. The irreversible nature of the disease makes the use of the neuropsychological tests that are sensitive to the first signs of cognitive decline essential, together with the development of adequate treatments for the early AD stages. On the other hand, demonstrating the benefits of pre-clinical AD treatments on functional impairment is inherently difficult [4]. To address this gap, we developed a disease progression model describing the dynamic relationship between cognitive and functional declines. The model also provides a promising tool for evidencing the value of future pre-clinical treatments in AD.
Objectives:
- Build a disease progression model for and to determine the temporal relationship between the cognitive and the functional declines in patients with Alzheimer’s disease using a longitudinal mixed-effects model
- Evaluate the impact of the APOE4 genotype (carrier vs. non-carrier) on the dynamics of AD progression (i.e. cognition and function)
Methods: Longitudinal data of 659 patients diagnosed with AD dementia were sourced from the Alzheimer’s disease neuroimaging initiative (ADNI) database and modeled in two steps using mixed-effects Emax models (R, version 3.4.3). A cognitive subscale, delayed word recall, of the Alzheimer disease assessment scale and a functional assessment questionnaire were selected as endpoints to characterize cognitive and functional declines, respectively. To evaluate the extent of the causality between cognition and function, individual parameter estimates derived from the cognitive decline model were used to partially drive and explain the functional decline. Furthermore, the impact of the APOE4 genotype status on the dynamics of AD progression as characterized by cognition and function was evaluated.
Results: Mixed-effects Emax models adequately quantified the individual and population disease trajectories as well as the relationship between cognitive and functional decline. While APOE4 carriers had a higher initial cognitive impairment than non-carriers, decline in function was similar between the two populations. The population-average time when patients reached their half of maximum cognitive decline proceeded the population-average time when they reached half of maximum functional decline.
Conclusions: The mixed-effects Emax models provide a tool to characterize the population- and individual-level AD progression and its dependence on patient-specific characteristics, e.g., APOE4 status. The analysis is also a first case study of a methodology to bridge dynamically the two outcomes sensitive in different stages of (pre-)AD spectrum, thus provided a promising tool for evidencing the value of future pre-clinical treatments of AD.
References:
[1] Chong MS and Sahadevan S. Lancet Neurol. 2005;4:576–9.
[2] Albert MS, et al. Alzheimers Dement. 2011;7:270–9.
[3] Burton RL, et al. Arch Clin Neuropsychol. 2017:1–13.
[4] Liu-Seifert H, et al. J Alzheimers Dis. 2015;47:205–14.
Reference: PAGE 28 (2019) Abstr 9080 [www.page-meeting.org/?abstract=9080]
Poster: Drug/Disease Modelling - CNS