Months to several years before amyotrophic lateral sclerosis (ALS) symptoms arise, levels of certain blood proteins may dramatically shift. Researchers analyzed data from the long-running, National Institutes of Health (NIH)-funded Pre-symptomatic Familial ALS (Pre-fALS) study, to identify a lineup of key proteins that may predict the emergence of clinically manifest ALS. By anticipating the arrival of symptoms, investigators could intervene with preventative therapies before the irreversible motor neuron damage that is characteristic of ALS sets in.

If someone carrying an ALS-associated genetic variant had asked me in the past when they would become symptomatic, I would have struggled to provide a reasonable estimate. These biomarkers give us a far better idea of the timing, allowing us to estimate the time to symptom onset with an average error of about 18 months. That's something we can work with."

Michael Benatar, M.D., Ph.D., senior author, professor of neurology and public health sciences, University of Miami

For nearly 20 years, the Pre-fALS study, led by Benatar and Joanne Wuu, Sc.M., research associate professor of neurology and public health sciences at the University of Miami, has collected data and biological samples from people who are at significantly elevated genetic risk for ALS but have not yet progressed, or phenoconverted, to the disease. While this cohort is unique, permitting the examination of presymptomatic ALS, recent studies suggest that findings from Pre-fALS are likely relevant to the broader population.

In 2017, an analysis of ten Pre-fALS participants who had developed symptoms showed that neurofilament light chain (NfL), a structural protein in neurons, spiked in their blood in the months preceding ALS phenoconversion.

As more study participants have begun showing symptoms or signs of disease, new opportunities to search for other pre-symptomatic ALS biomarkers have emerged.

Now, the investigators applied a high-throughput protein, or proteomic, analysis method called Olink to plasma samples from 137 study participants, 33 of whom had demonstrated clinical manifestations of ALS or frontotemporal dementia. Having collected data on the levels of more than 5,000 proteins, the team identified 92 whose levels differed in people before they eventually showed symptoms.

Using machine-learning techniques, the authors tested how various combinations of proteins could predict future risk of phenoconversion. They settled on a select panel of 19, including NfL, that maximized predictive accuracy across time horizons ranging from six months to five years. The researchers demonstrated that, with data on these 19 proteins, their predictive models could estimate a patient's onset within less than two years of the actual time that they showed signs of disease.

They also produced similar results using data from the UK Biobank, which, despite some limitations, is more representative of the general population than the genetically predisposed cohort of Pre-fALS.

"With preventative gene-targeting treatments now becoming available, there is a particularly urgent need for reliable biofluid-based signatures that indicate near-term onset in individuals that carry ALS risk genes," said Amy Bany Adams, Ph.D., acting director of NIH's National Institute of Neurological Disorders and Stroke (NINDS).

Tofersen, a drug approved for symptomatic ALS, is currently being evaluated as a preventative therapeutic in pre-symptmoatic ALS through ATLAS, a clinical trial designed by Benatar in partnership with the company Biogen. ATLAS will test whether starting treatment shortly before symptoms appear could avert or delay the onset of ALS.

"This is all possible because of the members of the carrier community who believe in our mission of preventing ALS and have supported and participated in our research. It has been one of my life's greatest privileges to give something back," Benatar said.

This research was supported by NIH through NINDS grants R01NS105479 and U54NS092091.

Source:

Journal reference: