Listen Up: Emerging Vocal Biomarker Could Aid ALS Drug Development

Phone analyzing sound_Taylor Tieden

Pictured: A smartphone analyzing sound/Taylor Tieden for BioSpace

It’s been a disappointing spring for the amyotrophic lateral sclerosis community, with the Phase II failure of Sanofi and Denali Therapeutics’ candidate and Amylyx’s decision to pull Relyvrio from the market. But drugmakers are hopeful that the development and validation of novel biomarkers for the disease, including neurofilament, genetic markers and, more recently, speech, can help improve patient care and expedite the development of effective treatments.

Objective Voice Biomarkers

Amyotrophic lateral sclerosis (ALS) is a fatal, neurodegenerative disease that causes muscles to weaken and waste and the brain to lose control over movements such as walking and breathing. Most patients with ALS die from respiratory failure within five years of symptom onset.

In most afflicted individuals, speech is affected, beginning either at the onset of the disease or as it progresses. Speech is included on the “gold standard” Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised (ALSFRS-R). The 10-item inventory measures the progression of disability, with each item receiving a 0-4 score, ranked by the patient or their caregiver.

This highly subjective scale bothered Indu Navar, a tech professional, after her husband was diagnosed with ALS. “When we start relying on us to measure our own disease, it’s just a disaster,” she said. “What you can’t measure, you can’t fix.”

Navar was driven to start a nonprofit dubbed EverythingALS to develop tools to measure each aspect of the ALSFRS-R objectively. While the organization’s first focus was speech, it has also collected data on respiratory and gross motor biomarkers, and is targeting fine motor skills next.

Through a citizen-driven initiative, EverythingALS completed and will soon publish the Austen Speech Study, which the organization used to develop a digital biomarker that measures changes in a patient’s speech too subtle for the human ear to pick up. The study recruited a diverse population of 7,000 patients and caregivers and had them collect data using a phone app that records both facial expression and speech through a series of speech tasks performed up to once a week.

The method draws in a “huge longitudinal data set,” Lyle Ostrow, a neurologist and researcher at Temple University specializing in ALS, told BioSpace.

Ostrow is particularly excited about the promise of greater care equity that digital biomarkers bring. “They allow people to participate in research wherever they live because it can be done remotely. People who normally perhaps would not be able to participate in a trial can also allow us to get data much more frequently,” he said.

The study has collected a massive amount of data in just two and a half years. Navar explained that the data are shared not only with EverythingALS but also with the participating patients as well as openly for use by other entities working on this technology and pharmaceutical companies working in the ALS space. Data are de-identified to protect patient privacy.

EverythingALS is now in talks with a few pharma companies to include the voice biomarker in their clinical trials, Navar said, adding that the team is currently analyzing speech data from a drug trial that was already completed.

FDA Backing

The FDA’s Center for Drug Evaluation and Research (CDER) has a program dedicated to biomarker qualification to develop novel biomarkers for better efficiency and innovation in drug development. ALS desperately needs clinical endpoints that are precise, said Stacy Lindborg, co-CEO at BrainStorm Cell Therapeutics, which is developing the cell therapy NurOwn. “That will speed up our goals of bringing new therapies forward” for patients, she told BioSpace.

As an example, both Lindborg and Ostrow pointed to the approval of Biogen and Ionis’ Qalsody last year, in which the agency accepted neurofilament light chain, a marker of neurodegeneration, as a clinically validated endpoint.

“The FDA was perfectly willing to pay attention to rigorous, well-done science. So that’s sort of become the goal—rigorous science that proves these outcomes are meaningful,” Ostrow said of the approval. Navar said EverythingALS is “very closely aligned with the FDA” and working to make their objectively tracked biomarkers, starting with vocal but eventually covering every aspect of motor skills, meaningful clinical endpoints.

Potential Impact on Drug Development

Research shows less than 12% of candidate medicines that enter Phase I trials will be approved by the FDA, and Lindborg said she expects the percentage of ALS drugs to be even lower.

That’s because ALS is a particularly complicated disease, Lindborg said. About 90% of cases are sporadic, with no familial history, and its causes are not well understood. Despite decades of research, there are very few treatment options available to these patients.

After the first drug was approved for the disease in 1995, the next new treatment specifically for ALS, Mitsubishi Tanabe Pharma’s Radicava, didn’t come until 22 years later in 2017. Five years after that, Relyvrio was approved, followed by Qalsody in 2023 for patients specifically with the SOD1 gene mutation.

Earlier this month, Amylyx voluntarily pulled Relyvrio from U.S. and Canadian markets after a Phase III study showed the drug did not improve ALSFRS-R scores after 48 weeks compared to placebo.

“We have to de-risk” ALS drug development, Ostrow said, both for companies and investors. In addition to providing an objective measure for patients to monitor their disease progression, the goal in developing these novel biomarkers is to provide an assessment tool that can be used as a clinical endpoint. “We’re sick of trials failing and not knowing why they failed or . . . what to do next,” Ostrow said.

He pointed out the advantage of digital biomarkers for identifying the correct patient populations for potential therapies. “We can cluster people into fast progressors or slow progressors,” he said, adding that a drug will often work in a subpopulation of patients. Biomarkers can help identify the right subset to target, hopefully improving clinical trial outcomes.

“No two ALS patients are the same, but there are clusters,” Ostrow said. “If you can identify a cluster and know how they behave, that makes it easier to tell if your drugs are making a difference.”

BrainStorm has encountered turbulence in its bid to bring NurOwn to the market. In September 2023, an advisory committee concluded that the treatment was not effective in terms of survival. Yet a subgroup of patients with earlier-stage disease had a wider margin of benefit. The company will now target this earlier-stage population in a Phase IIIb trial.

“There’s no question that as these [biomarkers] emerge . . . we will want to include them,” Lindborg said.

A more objective view of a patient’s disease can help researchers to see potential treatment benefits that may otherwise be missed, she added.  “Companies are not likely to make a funding decision . . . based on speech [alone], but it’s a fundamental part of the disease.”

Kate Goodwin is a freelance life science writer based in Des Moines, Iowa. She can be reached at and on LinkedIn.  

Back to news