IC‐P2‐073: Feasibility and test‐retest reliability of fMRI in an Alzheimer's disease clinical trial

With promising disease-modifying therapies for Alzheimer's Disease (AD) entering clinical trials, there remains a critical need for biomarkers that can rapidly detect a “signal of efficacy”. Functional MRI (fMRI) has the potential to detect acute pharmacological effects, but has not yet been widely used in AD clinical trials. To examine the feasibility of implementing an fMRI memory paradigm in a blinded clinical trial and to assess fMRI test-retest reliability in subjects with mild to moderate AD. Twenty-four subjects with mild to moderate Probable AD (mean MMSE 21.6, age 71.6) participated in a randomized, double-blinded, placebo-controlled pharmacological fMRI study. Using a face-name encoding fMRI paradigm, whole-brain activity for novel versus repeated (NvR) face-name pairs were assessed at baseline (week0) for all subjects (n=24). To assess test-retest reliability, NvR activations in the placebo group (n=12) were compared at week0 (T1), week6 (T2) and week12 (T3) by whole-brain analyses and in hippocampal ROIs using measures of extent and magnitude of activations. All 24 subjects completed the study, and 23/24 had good quality fMRI data at all time-points. Whole-brain analyses of baseline data for all 24 subjects showed significant (p<0.001, 5-voxel extent) N>R activations in bilateral hippocampi (R.HC p=0.014; L.HC p=0.007,) R.inf frontal cortex (p=3.72e-04) and cingulate cortex (p=0.016). Small volume corrections used six SPM-MarsBaR ROIs (bilateral hippocampus, inferior frontal, and mid-cingulum cortices). The placebo group (n=12) demonstrated stability of N>R whole-brain activation maps over the 3 fMRI sessions (T1, T2, T3) and on clinical/behavioral measures. Hippocampal ROIs demonstrated similar extent and magnitude of activations across the 3 placebo sessions (RM-ANOVAs, age & education covaried, all p's>0.25). Intraclass Correlation Coefficients (ICC) ranged from 0.43–0.72 for right, and 0.64–0.75 for left hippocampal ROIs. Cognitive fMRI paradigms can be successfully implemented into AD clinical trials. Whole-brain and hippocampal fMRI encoding-related activations remained stable in subjects with mild to moderate AD across three fMRIs in 12 weeks, and showed satisfactory test-retest reliability. These results lend further support to the promise of fMRI methods as a potential biomarker to detect acute and subacute drug effects in “Proof of Concept” trials in AD. Atri et al. Feasibility and Test-retest Reliability of fMRI in an Alzheimer's Disease Clinical Trial