A pilot safety and tolerability study of scanning ultrasound as a neuromodulation therapy in Alzheimer's disease.

Clearing amyloid-β pathology in Alzheimer's disease (AD) has been considered a prerequisite for restoring cognitive functions. Intriguingly, by application of a modality of scanning ultrasound (SUS) to mice that does not remove amyloid-β, we previously achieved significant cognitive improvements. This prompted us to explore SUS as a non-invasive brain stimulation strategy in an open-label safety trial in AD. We conducted a human pilot study in 12 participants with AD with the primary objective of determining feasibility, safety and tolerability. Exploratory secondary end-points were cognitive and behavioural measures, resting-state EEG and functional MRI. A portable device termed UltraTheraPilot was built under medical device standard guidelines, integrating a Brainsight image-guided neuronavigation system. A single-element 286-kHz transducer was programmed to deliver non-derated ultrasound doses of 2.6, 1.95 or 1.3 MPa. With four treatment sessions spaced fortnightly, four participants received 30 sonications per session (precuneus, ∼30 cm3 brain tissue) and the remaining 8 received 100 sonications per session (bilateral precuneus and temporo-parietal association cortex, ∼100 cm3). Safety monitoring, EEG, MRI, cognitive and neuropsychiatric evaluations were performed. The treatment was fast, safe and well-tolerated at the 1.95 MPa dose. MRI showed no changes, whereas changes were observed in aperiodic EEG content. Cognitive performance did not change but statistically significant improvements in behavioural and psychological symptoms were found using the Neuropsychiatric Inventory test. In conclusion, this SUS safety trial met its primary and secondary end-points in biomarker-confirmed mild-to-moderate AD. It informs our future work in an upcoming efficacy trial in an AD population.