PROSA-a multicenter prospective observational study to develop low-burden digital speech biomarkers in ALS and FTD

Area of Science:

• Neuroscience
• Digital Health
• Biomarker Discovery

Background:

• Amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) require novel biomarkers for disease monitoring and treatment assessment.
• Digital biomarkers, particularly speech biomarkers, offer non-invasive, low-burden, and cost-effective methods for remote patient monitoring.
• Speech analysis can objectively capture cognitive, motor, and respiratory symptoms relevant to neurodegenerative diseases.

Purpose of the Study:

• To develop and validate frequent, prognostic digital speech biomarkers for ALS and FTD.
• To create a unified, easy-to-use speech assessment battery serving as a proxy for cognitive, respiratory, and motor functions.
• To leverage artificial intelligence for advanced speech analysis to identify speech-based phenotypes.

Main Methods:

• A 12-month, multicenter observational study including 75 ALS patients, 75 FTD patients, and 50 healthy controls.
• Collection of comprehensive speech data via a remote, automated telephone protocol at four time points.
• Integration of speech data with extensive clinical phenotyping from established longitudinal cohorts (DANCER, DESCRIBE-ALS, DESCRIBE-FTD).

Main Results:

• Longitudinal speech data analyzed with AI to develop speech-based phenotypes for ALS and FTD.
• Identification of speech patterns correlating with cognitive, motor, and respiratory symptoms.
• Validation of speech biomarkers as reliable proxies for disease status.

Conclusions:

• Speech-based phenotypes can inform the development of diagnostic and prognostic models for predicting clinical changes in ALS and FTD.
• These findings have significant implications for stratifying patients in clinical trials and designing innovative trial methodologies.
• Digital speech biomarkers represent a promising tool for advancing research and patient care in neurodegenerative diseases.