Premium
Is there still a rationale for BACE‐inhibition in Alzheimer’s disease? A quantitative systems pharmacology analysis
Author(s) -
Geerts Hugo,
Spiros Athan
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.038596
Subject(s) - cognition , neuroscience , pharmacodynamics , pharmacology , genotype , medicine , psychology , bioinformatics , oncology , biology , pharmacokinetics , genetics , gene
Background So far all clinical trials with BACE inhibitors have failed to generate a cognitive benefit and many have demonstrated a reversible cognitive worsening. Lack of understanding of the biological reason for this clinical effect hampers the study of which patient subpopulations might still benefit from lower levels of target exposure. Method We present computer‐based Quantitative Systems Pharmacology (QSP) simulations of cognitive ADAS‐Cog readouts to identify relevant PD‐PD interactions between comedications, common genotype variants, disease state and amyloid modulating agents, in particular BACE inhibition. This platform is based on a computer model of biophysically realistic neuronal cortical microcircuit with 35 CNS targets and calibrated for ADAS‐Cog. Long and short amyloid peptides affect glutamate and nAChR neurotransmission differentially that reproduces three clinical datasets. APOE, COMTVal158Met and 5‐HTTLPR rs23351 genotype are introduced using human imaging studies. Result The 5‐HTTLPR rs23351 L/L genotype defines a set of responders that are more resilient to amyloid increase with a 3‐4 point smaller worsening on ADAS‐Cog. Differential interaction with tau pathology can further reduce the effect with up to a 0.5 point. When simulating 2 year BACE‐I treatment in MCI, the platform predicts a worsening of up to 3 points compared to baseline on ADAS‐Cog for subjects with 5‐HTTLPR s/s genotype and slow amyloid progression while placebo deteriorates only 2 points. Conversely, BACE‐I is anticipated to improve cognition by about 0.4 point compared to baseline in subjects with 5‐HTTLPR L/L genotype and fast amyloid progression with placebo deteriorating 0.1 point. Lower levels of target engagement slightly improve the effect of BACE‐I but only in subjects with already elevated amyloid baseline levels. This is likely because of the beneficial effect of low doses of short amyloid forms that are also reduced by BACE‐I. Conclusion QSP is a hypothesis‐generating engine based on mechanistic, integrative and quantitative modeling of pharmacological and biological data and is particular informative at the individual patient level. The platform explains the reversible worsening of cognition with high exposure levels of BACE‐I in prodromal AD. Further simulations suggest that BACE inhibition can have modest benefits at lower levels of target engagement and in well‐defined patient populations.