Transcranial Magnetic Stimulation based Network Integrity Assessment in Alzheimer’s disease

Abstract

AbstractBackgroundAlzheimer’s disease (AD) has a relatively well‐characterized pathology of amyloid‐β and tau accumulation. Yet, disease progression from preclinical to clinical stage remains largely unpredictable at the individual level. Growing evidence shows disrupted synaptic signaling in functionally connected brain networks early in the preclinical stage, and suggests impaired network integrity, assessed with functional connectivity measures, as potential markers for disease symptomatology in AD.MethodHere propose a novel TMS‐induced causal connectivity measure reflecting integrity of multi‐synaptic signaling within the stimulated brain networks. In a preliminary sample of 9 cognitively normal older adults (Healthy Controls: HCs) and 11 patients with amnestic mild cognitive impairment or mild dementia (early AD), we applied 150 single pulses of TMS to inferior parietal lobule (IPL), a distinct brain region within the default mode network (DMN) showing decreased metabolism in early AD and recorded TMS evoked neural responses with simultaneous electroencephalography (TMS‐EEG). We also characterized spontaneous brain activity with resting‐state EEG (rsEEG) measurements. We performed source space spectral power analyses and seed‐based power envelope correlations (PEC) to estimate cortical distribution of relative alpha power and spontaneous IPL connectivity as rsEEG measurements. Network integrity was defined as the ratio of significantly activated vertices to the total number of vertices within the DMN following TMS of IPL.ResultAlthough relative alpha power (8‐12 Hz) was slightly lower and seed based connectivity was slightly higher within the IPL node in AD participants, no network specific differences was observed across the cortex for both rsEEG measures. On the other hand, DMN integrity was substantially lower in AD participants with an average ratio of 0.53 (0.79 in HC) indicating that almost half of the vertices within the DMN were not activated by TMS.ConclusionOverall, these preliminary analyses suggest that perturbation‐based (TMS‐EEG) measures may reveal neurophysiological signatures of brain network dysfunction that are not observable in spontaneous brain oscillations, and thus may characterize defining features of network failure in preclinical and mild AD.