Researchers have identified specific blood proteins that change months to years before ALS symptoms appear, utilizing advanced proteomic analysis and machine learning to improve early detection.
New findings could help clinicians predict amyotrophic lateral sclerosis (ALS) symptom onset, according to a National Institutes of Health (NIH) news release.
A group of researchers investigated data from the NIH-funded Pre-symptomatic Familial ALS (Pre-fALS) study and discovered a change that happens to levels of certain blood proteins months to years before symptoms start. Specifically, they built on the 2017 finding that neurofilament light chain (NfL) levels in the blood spike months before ALS phenoconversion.
Using Olink, “a high-throughput protein, or proteomic, analysis method” on 137 participants’ plasma samples, they uncovered 92 proteins that shift before ALS symptom onset. The researchers utilized machine-learning techniques to assess how different combinations of the proteins could reveal clues about symptoms. They found 19 proteins “that maximized predictive accuracy across time horizons ranging from six months to five years.” With this data, the models “could estimate a patient’s onset within less than two years of the actual time that they showed signs of disease.”
Additionally, similar outcomes were observed using data from the UK Biobank. These findings could lead to earlier intervention and treatment for patients.
