Predicting the efficacy of donepezil intervention in Alzheimer's disease patients using regional homogeneity in the inferior orbitofrontal cortex.

ABSTRACT

BACKGROUND: Although donepezil is a commonly used drug for treating Alzheimer's disease (AD), the mechanisms by which it affects patients' functional brain activity, and thus modulates clinical symptoms, remain unclear. METHODS: In the present study, we used resting-state functional magnetic resonance imaging (MRI) and regional homogeneity (ReHo) to investigate the effects of donepezil on local brain activity in AD patients. Resting-state functional MRI data were collected from 32 subjects: 16 healthy controls and 16 AD patients. All 16 AD patients underwent 6 months of donepezil treatment and received two MRI scans (pre- and post-intervention). Analysis of covariance and post hoc analyses were used to compare ReHo differences among the healthy controls, pre-intervention AD patients, and post-intervention AD patients. Pearson correlation analysis was used to examine relationships between ReHo values in differential brain regions and clinical symptoms. RESULTS: Compared with healthy controls, post-intervention AD patients had reduced ReHo in the orbital part of the inferior frontal gyrus, and pre-intervention AD patients had reduced ReHo in the orbital part of the right inferior frontal gyrus. Pattern recognition models revealed that pre-intervention ReHo values in abnormal brain regions of AD patients were 76% accurate for predicting the efficacy of donepezil on cognitive function and 65% accurate for predicting its efficacy on depressive symptoms. CONCLUSIONS: These findings deepen our understanding of the brain mechanisms underlying the clinical efficacy of donepezil in AD patients, and provide a novel way to predict its clinical efficacy in such patients.